Compare commits

...

1102 Commits

Author SHA1 Message Date
Jeremy Huddleston
a96360a951 apple: Implement applegl_unbind_context
glXMakeCurrent(dpy, None, NULL) would not correctly unbind the context
causing subsequent GLX requests to fail in peculiar ways

http://xquartz.macosforge.org/trac/ticket/514

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 5c44c1348e)
2011-10-24 16:26:46 -07:00
Jeremy Huddleston
c8f2665f02 apple: Silence some debug spew
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 098ecfad83)
2011-10-24 15:54:31 -07:00
Dave Airlie
1e279de4b0 r600g: bump domain selection up one layer.
this is taken from a patch from Mathias Froehlich, just going to
stage it in a few pieces.

Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit ecc051d65b, see
https://bugs.freedesktop.org/show_bug.cgi?id=40979)
2011-10-03 17:00:35 +02:00
Jeremy Huddleston
7c3cf50d99 Fix PPC detection on darwin
Fixes regression introduced by 7004582c18

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit e737a99a6f)
2011-07-31 09:49:02 -07:00
Paul Berry
7b8e7f4450 glsl: Rewrote _mesa_glsl_process_extension to use table-driven logic.
Instead of using a chain of manually maintained if/else blocks to
handle "#extension" directives, we now consult a table that specifies,
for each extension, the circumstances under which it is available, and
what flags in _mesa_glsl_parse_state need to be set in order to
activate it.

This makes it easier to add new GLSL extensions in the future, and
fixes the following bugs:

- Previously, _mesa_glsl_process_extension would sometimes set the
  "_enable" and "_warn" flags for an extension before checking whether
  the extension was supported by the driver; as a result, specifying
  "enable" behavior for an unsupported extension would sometimes cause
  front-end support for that extension to be switched on in spite of
  the fact that back-end support was not available, leading to strange
  failures, such as those in
  https://bugs.freedesktop.org/show_bug.cgi?id=38015.

- "#extension all: warn" and "#extension all: disable" had no effect.

Notes:

- All extensions are currently marked as unavailable in geometry
  shaders.  This should not have any adverse effects since geometry
  shaders aren't supported yet.  When we return to working on geometry
  shader support, we'll need to update the table for those extensions
  that are available in geometry shaders.

- Previous to this commit, if a shader mentioned
  ARB_shader_texture_lod, extension ARB_texture_rectangle would be
  automatically turned on in order to ensure that the types
  sampler2DRect and sampler2DRectShadow would be defined.  This was
  unnecessary, because (a) ARB_shader_texture_lod works perfectly well
  without those types provided that the builtin functions that
  reference them are not called, and (b) ARB_texture_rectangle is
  enabled by default in non-ES contexts anyway.  I eliminated this
  unnecessary behavior in order to make the behavior of all extensions
  consistent.

Some changes were made in glsl_parser_extras.h during cherry pick to
7.10 because 7.11 and master support many extensions that 7.10 does
not.  The unsupported extensions were removed.

NOTE: This is a candidate for the 7.10 and 7.11 branches.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
2011-07-20 13:36:46 -07:00
Paul Berry
db3d21d676 glsl: Changed extension enable bits to bools.
These were previously 1-bit-wide bitfields.  Changing them to bools
has a negligible performance impact, and allows them to be accessed by
offset as well as by direct structure access.

Some changes were made in glsl_parser_extras.h during cherry pick to
7.10 because 7.11 and master support many extensions that 7.10 does
not.  The unsupported extensions were removed.

NOTE: This is a candidate for the 7.10 and 7.11 branches.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 9c4445de6e)
2011-07-20 13:36:46 -07:00
Paul Berry
0a838d1d32 glsl: Ensure that sampler declarations are always uniform or "in" parameters.
This brings us into compliance with page 17 (page 22 of the PDF) of
the GLSL 1.20 spec:

    "[Sampler types] can only be declared as function parameters or
    uniform variables (see Section 4.3.5 "Uniform"). ... [Samplers]
    cannot be used as out or inout function parameters."

The spec isn't explicit about whether this rule applies to
structs/arrays containing shaders, but the intent seems to be to
ensure that it can always be determined at compile time which sampler
is being used in each texture lookup.  So to avoid creating a
loophole, the rule needs to apply to structs/arrays containing shaders
as well.

Fixes piglit tests spec/glsl-1.10/compiler/samplers/*.frag, and fixes
bug 38987.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=38987
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit f07221056e)
2011-07-20 13:36:46 -07:00
Paul Berry
94a1fe1fd8 glsl: Move type_contains_sampler() into glsl_type for later reuse.
The new location, as a member function of glsl_type, is more
consistent with queries like is_sampler(), is_boolean(), is_float(),
etc.  Placing the function inside glsl_type also makes it available to
any code that uses glsl_types.
(cherry picked from commit ddc1c96390)
2011-07-20 13:36:46 -07:00
Ian Romanick
f2ef6036e8 linker: Only over-ride built-ins when a prototype has been seen
The GLSL spec says:

    "If a built-in function is redeclared in a shader (i.e., a
    prototype is visible) before a call to it, then the linker will
    only attempt to resolve that call within the set of shaders that
    are linked with it."

This patch enforces this behavior.  When a function call is processed
a flag is set in the ir_call to indicate whether the previously seen
prototype is the built-in or not.  At link time a call will only bind
to an instance of a function that matches the "want built-in" setting
in the ir_call.

This has the odd side effect that first call to abs() in the shader
below will call the built-in and the second will not:

float foo(float x) { return abs(x); }
float abs(float x) { return -x; }
float bar(float x) { return abs(x); }

This seems insane, but it matches what the spec says.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=31744
(cherry picked from commit 66f4ac988d)
2011-07-20 13:36:45 -07:00
Henri Verbeet
f81058140f glx: Avoid calling __glXInitialize() in driReleaseDrawables().
This fixes a regression introduced by commit
a26121f375 (fd.o bug #39219).

Since the __glXInitialize() call should be unnecessary anyway, this is
probably a nicer fix for the original problem too.

NOTE: This is a candidate for the 7.10 and 7.11 branches.

Signed-off-by: Henri Verbeet <hverbeet@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Tested-by: padfoot@exemail.com.au
(cherry picked from commit 0f20e2e18f)
2011-07-19 23:34:55 +02:00
Marek Olšák
3871c55a18 swrast: fix depth/stencil blits when there's no colorbuffer
NOTE: This is a candidate for the 7.10 and 7.11 branches.
(cherry picked from commit d1214cca08)
2011-07-11 22:59:48 +02:00
Marek Olšák
7a75fcd657 mesa: return early if mask is cleared to zero in BlitFramebuffer
From ARB_framebuffer_object:
    If a buffer is specified in <mask> and does not exist in both the
    read and draw framebuffers, the corresponding bit is silently
    ignored.
(cherry picked from commit 83478e5d59)
2011-07-11 22:59:38 +02:00
Marek Olšák
e041956cb2 st/mesa: use the first non-VOID channel in st_format_datatype
Otherwise PIPE_FORMAT_X8B8G8R8_UNORM and friends would fail.

NOTE: This is a candidate for the 7.10 and 7.11 branches.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 292148dc4b)
2011-07-08 13:24:25 +02:00
Ian Romanick
c286f7870f linker: Assign locations for fragment shader output
Fixes an assertion failure in the piglib out-01.frag
ARB_explicit_attrib_location test.  The locations set via the layout
qualifier in fragment shader were not being applied to the shader
outputs.  As a result all of these variables still had a location of
-1 set.

This may need some more work for pre-3.0 contexts.  The problem is
dealing with generic outputs that lack a layout qualifier.  There is
no way for the application to specify a location
(glBindFragDataLocation is not supported) or query the location
assigned by the linker (glGetFragDataLocation is not supported).

NOTE: This is a candidate for the 7.10 and 7.11 branches.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=38624
Reviewed-by: Eric Anholt <eric@anholt.net>
Cc: Kenneth Graunke <kenneth@whitecape.org>
Cc: Vinson Lee <vlee@vmware.com>
(cherry picked from commit d32d4f780f)
2011-07-07 14:06:01 -07:00
Paul Berry
4e2a5d013d glsl: permit explicit locations on fragment shader outputs, not inputs
From the OpenGL docs for GL_ARB_explicit_attrib_location:

    This extension provides a method to pre-assign attribute locations to
    named vertex shader inputs and color numbers to named fragment shader
    outputs.

This was accidentally implemented for fragment shader inputs.  This
patch fixes it to apply to fragment shader outputs.

Fixes piglit tests
spec/ARB_explicit_attrib_location/1.{10,20}/compiler/layout-{01,03,06,07,08,09,10}.frag

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>

NOTE: This is a candidate for the 7.10 and 7.11 branches.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=38624
(cherry picked from commit b078aad8ab)
2011-07-07 14:06:00 -07:00
Ian Romanick
cd73c06eed glsl: Don't choke when printing an anonymous function parameter
This is based on commit 174cef7fee, but
the code is heavily changed.  The original commit modifies
ir_print_visitor::unique_name, but there is no such method in 7.10.
Instead, this code just modifies ir_print_visitor::visit(ir_variable
*ir) to "do the right thing" when ir_variable::name is NULL.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=38584
2011-07-06 17:24:54 -07:00
Ian Romanick
7424f9c1fe ir_to_mesa: Allocate temporary instructions on the visitor's ralloc context
And don't delete them.  Let ralloc clean them up.  Deleting the
temporary IR leaves dangling references in the prog_instruction.  That
results in a bad dereference when printing the IR with MESA_GLSL=dump.

NOTE: This is a candidate for the 7.10 and 7.11 branches.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=38584
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit dbda466fc0)
2011-07-06 17:17:34 -07:00
Ian Romanick
cb6dd6c399 glsl: Track initial mask in constant propagation live set
The set of values initially available (before any kills) must be
tracked with each constant in the set.  Otherwise the wrong component
can be selected after earlier components have been killed.

NOTE: This is a candidate for the 7.10 and 7.11 branches.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=37383
Reviewed-by: Eric Anholt <eric@anholt.net>
Cc: Kenneth Graunke <kenneth@whitecape.org>
Cc: Matthias Bentrup <matthias.bentrup@googlemail.com>
(cherry picked from commit 0eb9797958)
2011-07-06 17:17:25 -07:00
Ian Romanick
85b965b462 linker: Reject shaders that use too many varyings
Previously it was up to the driver or later code generator to reject
these shaders.  It turns out that nobody did this.

This will need changes to support geometry shaders.

NOTE: This is a candidate for the stable branches.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=37743
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit de77324d8f)
2011-06-27 16:39:43 -07:00
Marek Olšák
459012b148 r600g: bump shader input limits
(cherry picked from commit 1e5cef96d1)
2011-06-27 16:39:43 -07:00
Ben Skeggs
bd40e7cebd nouveau: fix includes for latest libdrm
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
(cherry picked from commit 5c102dd94f)
2011-06-26 18:09:23 +02:00
Marek Olšák
9370ddde85 r300g: drop support for ARGB, ABGR, XRGB, XBGR render targets
Blending and maybe even alpha-test don't work with those formats.

Only supporting RGBA, BGRA, RGBX, BGRX.

NOTE: This is a candidate for the 7.10 and 7.11 branches.
(cherry picked from commit bc517d64da)

Conflicts:

	src/gallium/drivers/r300/r300_texture.c
2011-06-25 20:01:23 +02:00
Marek Olšák
26ab29cf79 mesa: fix a memory leak in _mesa_unpack_depth_span
NOTE: This is a candidate for the 7.10 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 12c105b5de)
2011-06-24 23:18:54 +02:00
Marek Olšák
a312052add mesa: fix texstore of DEPTH24_STENCIL8 if srcFormat is STENCIL_INDEX
NOTE: This is a candidate for the 7.10 branch.
(cherry picked from commit e41a91cea7)
2011-06-24 23:17:34 +02:00
Brian Paul
6128739b75 st/wgl: return height, not width for WGL_PBUFFER_HEIGHT_ARB
Fixes https://bugs.freedesktop.org/show_bug.cgi?id=38599
(cherry picked from commit 8a5a28b731)
2011-06-23 06:56:46 -06:00
Alex Deucher
bc46f0c969 r600c: add missing bank tiling case for evergreen
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
2011-06-22 16:51:58 -04:00
Marek Olšák
1ad06c7a25 r300g: fix handling PREP_* options
This should fix rendering >65532 vertices using draw_arrays on r300-r400.

NOTE: This is a candidate for the 7.10 branch.
(cherry picked from commit 7df7eaf845)

Conflicts:

	src/gallium/drivers/r300/r300_render.c
2011-06-18 22:59:28 +02:00
Alex Deucher
bdc518e341 r600c: add tiling support for evergreen+
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
2011-06-14 18:47:11 -04:00
Jeremy Huddleston
338e8e5f14 apple: Dead code removal
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit e903cc17bb)
(cherry picked from commit 5078cb6858)
2011-06-13 23:23:09 -07:00
Jeremy Huddleston
5255e844af apple: applegl_destroy_context: Pass along the correct display
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit c6cf82fb55)
2011-06-13 21:41:26 -07:00
Brian Paul
45f369c74d docs: fix 'release release' typos 2011-06-13 18:29:44 -06:00
Ian Romanick
39ad7dc7b1 docs: Add 7.10.3 md5sums 2011-06-13 16:53:24 -07:00
Ian Romanick
2d0fd07037 docs: Add change log to 7.10.3 release notes 2011-06-13 16:22:29 -07:00
Ian Romanick
edbaa9e856 docs: Add list of bugs fixed in 7.10.3 release 2011-06-13 16:22:29 -07:00
Ian Romanick
a40eed0a10 mesa: Regenerate parser files from previous two commits 2011-06-13 16:22:17 -07:00
Eric Anholt
2b7ee98588 mesa: Add support for OPTION ATI_draw_buffers to ARB_fp.
Tested by piglit ati_draw_buffers-arbfp.
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit fb6e39737a)
2011-06-13 15:36:02 -07:00
Eric Anholt
ff07695e19 mesa: Add support for the ARB_fragment_program part of ARB_draw_buffers.
Fixes fbo-drawbuffers-arbfp.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=34321
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 28cec9e832)
2011-06-13 15:35:55 -07:00
José Fonseca
fc23cc06af wgl: Don't hold on to user supplied HDC.
Certain applications (e.g., Bernina My Label, and the Windows
implementation of Processing language) destroy the device context used when
creating the frame-buffer, causing presents to fail because we were still
referring to the old device context internally.

This change ensures we always use the same HDC passed to the ICD
entry-points when available, or our own HDC when not available (necessary
only when flushing on single buffered visuals).
2011-06-12 09:16:00 +01:00
José Fonseca
195e15559e st/wgl: Remove buggy assertion.
The assertion is wrong, now that state tracker can cope with a window with
zero width or height.
2011-06-12 09:15:58 +01:00
José Fonseca
22bf0ec991 st/wgl: Allow to create pbuffers bigger than the desktop.
We use a hidden window for pbuffer contexts, but Windows limits window
sizes to the desktop size by default. This means that creating a big
pbuffer on a small resolution single monitor would truncate the pbuffer
size to the desktop.

This change overrides the windows maximum size, allow to create windows
arbitrarily large.
2011-06-12 09:15:52 +01:00
José Fonseca
3f3d199121 st/wgl: Cope with zero width/height windows.
While ensuring the framebuffer area is never zero.
2011-06-12 09:15:49 +01:00
José Fonseca
ac28843c1a st/wgl: Prevent spurious framebuffer sizes when the window is minimized.
When the window is minimized GetClientRect will return zeros.

Instead of creating a 1x1 framebuffer, simply preserve the current window
size, until the window is restored or maximized again.
2011-06-12 09:15:45 +01:00
José Fonseca
c78ac563f1 st/wgl: Fix debug output format specifiers of stw_framebuffer_get_size(). 2011-06-12 09:15:42 +01:00
José Fonseca
b6c601a5e8 st/wgl: Adjust the pbuffer invisible window size.
Thanks to Brian Paul for diagnosing the issue.
2011-06-12 09:15:35 +01:00
José Fonseca
27f6db0f38 gallivm: Fix for dynamically linked LLVM 2.8 library.
This prevents the error

    prog: for the -disable-mmx option: may only occur zero or one times!

when creating a new context after XCloseDisplay with DRI drivers linked
with a shared LLVM 2.8 library.
2011-06-12 09:13:24 +01:00
José Fonseca
c1da25aede gallivm: Tell LLVM to not assume a 16-byte aligned stack on x86.
Fixes fdo 36738.
2011-06-12 09:13:24 +01:00
Jeremy Huddleston
05d9a4ab1c osmesa: Fix missing symbols when GLX_INDIRECT_RENDERING is defined.
When GLX_INDIRECT_RENDERING is defined, some symbols are used in
libglapi.a but are not defined.  Define them through the help of
glapitemp.h.

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
(cherry picked from commit 0e8d045bf8)
2011-06-11 22:16:43 -07:00
Marek Olšák
b0625f8486 mesa: return after invalidating renderbuffer
(cherry picked from commit a674ef7814)
2011-06-12 05:58:28 +02:00
Marek Olšák
b10abac70e mesa: invalidate framebuffer if internal format of renderbuffer is changed
RenderTexture doesn't have to be called in invalidate_rb, I guess.
(cherry picked from commit df818d572e)
2011-06-12 05:57:26 +02:00
Marek Olšák
7d30582c91 mesa: fix up assertion in _mesa_source_buffer_exists
This was probably missed when implementing luminance and luminance alpha
render targets.

_mesa_get_format_bits checks for both GL_*_BITS and GL_TEXTURE_*_SIZE.

This fixes:
main/framebuffer.c:892: _mesa_source_buffer_exists: Assertion `....' failed.
(cherry picked from commit c0110d5450)
2011-06-12 05:56:18 +02:00
Dave Airlie
08c47e4851 st/mesa: fix compressed mipmap generation.
If the underlying transfer had a stride wider for hw alignment reasons,
the mipmap generation would generate badly strided images.

this fixes a few problems I found while testing r600g with s3tc

Signed-off-by: Dave Airlie <airlied@redhat.com>

(cherry picked from commit fdb4373a20 by Marek)
This fixes the DXT1 tests from fbo-generatemipmap-formats on some drivers.
2011-06-12 05:45:07 +02:00
Ian Romanick
39865f5c46 mesa: Ignore blits to/from missing buffers
The EXT_framebuffer_object spec (and later specs) say:

     "If a buffer is specified in <mask> and does not exist in both
     the read and draw framebuffers, the corresponding bit is silently
     ignored."

Check for color, depth, and stencil that the source and destination
FBOs have the specified buffers.  If the buffer is missing, remove the
bit from the blit request mask and continue.

Fixes the crash in piglit test 'fbo-missing-attachment-blit from', and
fixes 'fbo-missing-attachment-blit es2 from'.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=37739
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Eric Anholt <eric@anholt.net>

NOTE: This is a candidate for the stable branches.
(cherry picked from commit bb4758669c by Marek)
2011-06-12 05:29:18 +02:00
Ian Romanick
6586475f37 mesa: Don't try to clear a NULL renderbuffer
In an ES2 context (or if GL_ARB_ES2_compatibility) is supported, the
framebuffer can be complete with some attachments be missing.  In this
case the _ColorDrawBuffers pointer will be NULL.

Fixes the crash in piglit test fbo-missing-attachment-clear.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=37739
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Eric Anholt <eric@anholt.net>

NOTE: This is a candidate for the stable branches.
(cherry picked from commit 7f9c17aaa8 by Marek)
2011-06-12 05:29:14 +02:00
Marek Olšák
b9f2eefd1d st/mesa: remove asserts in st_texture_image_copy
This is for 7.10 only. The fix in master looks too complex to be
cherry-picked.

The assertions fail when generating mipmaps for NPOT textures.

This fixes:
- fbo-generatemipmap-formats

Reviewed-by: Brian Paul <brianp@vmware.com>
2011-06-11 22:34:01 +02:00
Brian Paul
48a3b03409 mesa: bump version to 7.10.3 2011-06-10 15:25:38 -06:00
Brian Paul
29164dcac8 docs: 7.10.3 release notes skeleton file, links 2011-06-10 15:25:38 -06:00
Brian Paul
f2901d3f98 mesa: add include/c99/*.h files to tarballs
See https://bugs.freedesktop.org/show_bug.cgi?id=36238

NOTE: This is a candidate for the 7.10 branch.
2011-06-10 15:25:38 -06:00
Brian Paul
3b89e1c0e6 st/mesa: fix software accum buffer format bug 2011-06-10 15:25:38 -06:00
Brian Paul
8a78e6cf80 vbo: remove node->count > 0 test in vbo_save_playback_vertex_list()
See piglit dlist-fdo31590.c test and
http://bugs.freedesktop.org/show_bug.cgi?id=31590

In this case we had node->prim_count=1 but node->count==0 because the
display list started with glBegin() but had no vertices.  The call to
glEvalCoord1f() triggered the DO_FALLBACK() path.  When replaying the
display list, the old condition basically no-op'd the call to
vbo_save_playback_vertex_list call().  That led to the invalid operation
error being raised in glEnd().

NOTE: This is a candidate for the 7.10 branch.
(cherry picked from commit 62811057f4)
2011-06-10 13:09:41 -06:00
Brian Paul
788dda53cf vbo: check array indexes to prevent negative indexing
See the piglit dlist-fdo31590.c test

NOTE: This is a candidate for the 7.10 branch.
(cherry picked from commit f1cdce95f6)
2011-06-10 13:09:41 -06:00
Brian Paul
cc5d54e797 draw: fix edge flag handling in clipper (for unfilled tris/quads/polygons)
Previously, we were errantly drawing some interior edges of clipped
polygons and quads.  Also, we were introducing extra edges where
polygons intersected the view frustum clip planes.

The main problem was that we were ignoring the edgeflags encoded in
the primitive header's 'flags' field which are set during polygon/quad
->tri decomposition.  We need to observe those during clipping.  Since
we can't modify the existing vert's edgeflag fields, we need to store
them in a parallel array.

Edge flags also need to be handled differently for view frustum planes
vs. user-defined clip planes.  In the former case we don't want to draw
new clip edges but in the later case we do.  This matches NVIDIA's
behaviour and it just looks right.

Finally, note that the LLVM draw code does not properly set vertex
edge flags.  It's OK on the regular software path though.
(cherry picked from commit f6572017b9)
2011-06-10 13:09:40 -06:00
Brian Paul
59b147c6a3 st/mesa: fix incorrect texture level/face/slice accesses
If we use FBOs to access mipmap levels with glRead/Draw/CopyPixels()
we need to be sure to access the correct mipmap level/face/slice.
Before, we were just passing zero in quite a few places.

This fixes the new piglit fbo-mipmap-copypix test.

NOTE: This is a candidate for the 7.10 branch.
(cherry picked from commit bf14ab417c)

Conflicts:

	src/mesa/state_tracker/st_cb_drawpixels.c
2011-06-10 13:07:03 -06:00
Brian Paul
5ec931eb7d mesa: check that flex/bison are installed
Fixes https://bugs.freedesktop.org/show_bug.cgi?id=36651

NOTE: This is a candidate for the 7.10 branch.
(cherry picked from commit de1df26b5c)
2011-06-10 13:07:03 -06:00
Brian Paul
9c7e1b054d mesa: add some missing GLAPIENTRY keywords
NOTE: this is a candidate for the 7.10 branch.
(cherry picked from commit 3559440873)
2011-06-10 13:07:03 -06:00
Marek Olšák
3d5e6c4c1e r300g: clear can be killed by render condition
Fixes piglit:
- NV_conditional_render/clear
(cherry picked from commit 76056510bc)

Conflicts:

	src/gallium/drivers/r300/r300_blit.c
2011-06-10 19:45:45 +02:00
Marek Olšák
bbebbdd3e7 r300g: fix occlusion queries when depth test is disabled or zbuffer is missing
From now on, depth test is always enabled in hardware.

If depth test is disabled in Gallium, the hardware Z function is set to ALWAYS.

If there is no zbuffer set, the colorbuffer0 memory is set as a zbuffer
to silence the CS checker.

This fixes piglit:
- occlusion-query-discard
- NV_conditional_render/bitmap
- NV_conditional_render/drawpixels
- NV_conditional_render/vertex_array
(cherry picked from commit f76787b3ea)

Conflicts:

	src/gallium/drivers/r300/r300_state.c

Squashed with cherry-picked b1246cf13b.
2011-06-10 19:45:45 +02:00
Marek Olšák
e5408efb20 r300g: fix texturing with non-3D textures and wrap R mode set to sample border
If the wrap R (3rd) mode is set to CLAMP or CLAMP_TO_BORDER and the texture
isn't 3D, r300 always samples the border color regardless of texture
coordinates.

I HATE THIS HARDWARE.

NOTE: This is a candidate for the 7.10 branch.
(cherry picked from commit da8b4c0798)

Conflicts:

	src/gallium/drivers/r300/r300_state_derived.c
2011-06-09 04:23:14 +02:00
Marek Olšák
8f8d7d0803 r300g: fix draw_vbo splitting on r3xx-r4xx
NOTE: This is a candidate for the 7.10 branch.
(cherry picked from commit 578d4539ba)

Conflicts:

	src/gallium/drivers/r300/r300_render.c
2011-06-09 01:38:24 +02:00
Jeremy Huddleston
addc396d18 darwin: Fix VG_LIB_GLOB to also match the unversioned symlink
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 6ea70600b9)
2011-06-07 13:28:19 -04:00
Jeremy Huddleston
8ea26afd8b darwin: Don't link against libGL when building libOSMesa
Everything should be resolved through glapi.

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit cebdffaa2a)
2011-06-07 13:28:14 -04:00
Jeremy Huddleston
ca9ab0a6a1 darwin: Set VG_LIB_{NAME,GLOB} to fix make install
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 22c320aa2c)
2011-06-06 19:57:00 -04:00
Jeremy Huddleston
1a79cde8fa apple: Package applegl source into MesaLib tarball
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit e5d241ddb2)
2011-06-06 12:48:54 -04:00
Jeremy Huddleston
4d934efa19 darwin: Define GALLIUM_DRIVERS_DIRS in darwin config
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit d563009cd1)
2011-06-06 12:00:16 -04:00
Jeremy Huddleston
aba30d8fbe apple: Fix build failures in applegl_glx.c
See https://bugs.freedesktop.org/show_bug.cgi?id=29162

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>

This commit squashes three cherry-picks:
(cherry picked from commit 5d35343d12)
(cherry picked from commit 7c5f37c032)
(cherry picked from commit 2ee5272e16)
2011-06-06 12:00:16 -04:00
Jeremy Huddleston
4e18ad9d71 apple: Build darwin using applegl rather than indirect
This reverts portions of 6849916170 that caused
the darwin config to fail to build due to missing implementations in that
commit.

See https://bugs.freedesktop.org/show_bug.cgi?id=29162

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 1885cf27c9)
2011-06-06 12:00:16 -04:00
Jeremy Huddleston
12537b0baf glx: Dead code removal
Remove a redundant ifndef GLX_USE_APPLEGL

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 3843bbcb4c)
2011-06-06 12:00:16 -04:00
Jeremy Huddleston
891ce8aaa8 apple: ifdef out come glapi-foo on darwin
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 8593bb32ea)
2011-06-06 12:00:15 -04:00
Jeremy Huddleston
0f11d05e81 apple: Change from XExtDisplayInfo to struct glx_display
Fixes regression introduced by: ab434f6b76 and
                                c356f5867f

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 7cdf969527)
2011-06-06 11:51:46 -04:00
Jeremy Huddleston
1605525111 apple: Rename GLXcontext
Fixes regression introduced by: c356f5867f

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 279e471750)
2011-06-06 11:51:37 -04:00
Jeremy Huddleston
d92931467d apple: Rename _gl_context_modes_find_visual to glx_config_find_visual
Fixes regression introduced by: 6ddf66e923

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit b7f0ed8444)
2011-06-06 11:51:28 -04:00
Jeremy Huddleston
ee004cc681 apple: Re-add driContext and do_destroy
Fixes regression introduced by: c491e585e4

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 488d6c739f)
2011-06-06 11:51:22 -04:00
Jeremy Huddleston
60291c0798 apple: Rename GLXcontext
Fixes regression introduced by: c356f5867f

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 22613d1670)
2011-06-06 11:51:12 -04:00
Jeremy Huddleston
56930eccb2 apple: Rename __GLcontextModes to struct glx_config
Fixes regression introduced by: 6ddf66e923

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit bb621cb61e)
2011-06-06 11:51:06 -04:00
Jeremy Huddleston
6ad7721500 apple: Rename glcontextmodes.[ch] to glxconfig.[ch]
Fixes regression introduced by: 65d98e2577

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 4c9bab78a1)
2011-06-06 11:50:58 -04:00
Jeremy Huddleston
5e1199ff12 apple: Update GL specs
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 8e89d0bea7)
2011-06-06 11:50:46 -04:00
Brian Paul
a10cba3a7f mesa: fix void pointer arithmetic warnings
And fix a couple logic errors in the put_*_generic() functions.
(cherry picked from commit 7ca38f5d97)
2011-06-01 08:33:09 -06:00
Marek Olšák
f8b4ca7e47 mesa: queries of non-existent FBO attachments should return INVALID_OPERATION
OpenGL 4.0 Compatibility, page 449:

If the value of FRAMEBUFFER_ATTACHMENT_OBJECT_TYPE is NONE, no
framebuffer is bound to target. In this case querying pname FRAMEBUFFER_-
ATTACHMENT_OBJECT_NAME will return zero, and all other queries will generate
an INVALID_OPERATION error.

Reviewed-by: Chad Versace <chad@chad-versace.us>
(cherry picked from commit b9e9df78a0)
2011-06-01 16:10:56 +02:00
Marek Olšák
9e1f40c182 mesa: UseShaderProgramEXT and Uniform* shouldn't be allowed inside Begin/End
I couldn't find this being required by the spec.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit d69dc2e203)
2011-06-01 16:10:49 +02:00
Eric Anholt
4d08ca20ba Revert "intel: Add spans code for the ARB_texture_rg support."
This reverts what remains of commit
28bab24e16.  It was garbage, trying to
use a MESA_FORMAT enum as a preprocessor token, and I don't know how I
thought it was even tested.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit bc57df0356)
2011-05-31 17:45:42 -07:00
Eric Anholt
fd9a093086 intel: Use mesa core's R8, RG88, R16, RG1616 RB accessors.
Fixes:
ARB_texture_rg/fbo-alphatest-formats

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 73f0700481)
2011-05-31 17:45:42 -07:00
Eric Anholt
c7e0c21d42 swrast: Don't try to adjust_colors for <8bpc when handling R16, RG1616.
The GL_RED and GL_RG were tricking this code into executing, but it's
totally unprepared for a 16-bit channel and just rescaled the values
down to 0.  We don't have anything with <8bit channels alongside >8bit
channels, so disabling it should be safe.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 2b624634dd)
2011-05-31 17:45:42 -07:00
Eric Anholt
dc8599cdf1 mesa: Add renderbuffer accessors for R8/RG88/R16/RG1616.
This will replace the current (broken by trying to use an enum in the
preprocessor) spantmp2.h support I wrote for the intel driver.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit f0471d904c)
2011-05-31 17:45:42 -07:00
Eric Anholt
25134f877b mesa: Use _mesa_get_format_bytes to refactor out the RB get_row_*
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit a52803e108)
2011-05-31 17:45:42 -07:00
Eric Anholt
27fb522faa mesa: Use _mesa_get_format_bytes to refactor out the RB get_pointer_*
This is a squash of the following two commits:

    mesa: Use _mesa_get_format_bytes to refactor out the RB get_pointer_*

    Reviewed-by: Brian Paul <brianp@vmware.com>
    (cherry picked from commit 6ab9889a27)

    mesa: Fix return type of  _mesa_get_format_bytes() (#37351)

    Despite that negative values aren't sensible here, making this unsigned
    is dangerous.  Consider get_pointer_generic, which computes a value of
    the form:

        void *base + (int x * int stride + int y) * unsigned bpp

    The usual arithmetic conversions will coerce the (x*stride + y)
    subexpression to unsigned.  Since stride can be negative, this is
    disastrous.

    Fixes at least the following piglit tests on Ironlake:

        fbo/fbo-blit-d24s8
        spec/ARB_depth_texture/fbo-clear-formats
        spec/EXT_packed_depth_stencil/fbo-clear-formats

    NOTE: This is a candidate for the 7.10 branch.

    Reviewed-by: Chad Versace <chad.versace@intel.com>
    Signed-off-by: Adam Jackson <ajax@redhat.com>
    (cherry picked from commit e8b1c6d6f5)
2011-05-31 17:45:41 -07:00
Eric Anholt
3436f5d82f intel: Use Mesa core's renderbuffer accessors for depth.
Since we're using GTT mappings now (no manual detiling), there's
really nothing special to accessing these buffers, other than needing
the new RowStride field of gl_renderbuffer to accomodate padding.

Reduces the driver size by 2.7kb, and improves glean depthStencil
performance 3-10x (!)

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 0778fdb002)
2011-05-31 17:45:41 -07:00
Eric Anholt
77be3db45c mesa: Add a function to set up the default renderbuffer accessors.
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 9f164823c7)
2011-05-31 17:45:41 -07:00
Eric Anholt
1d58f70e56 mesa: Add a gl_renderbuffer.RowStride field like textures have.
This will allow some drivers to reuse the core renderbuffer.c get/put
row functions in place of using the spantmp.h macros.  Note that
unlike textures, we use a signed integer here to allow for handling
FBO orientation.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 155200c154)
2011-05-31 17:45:41 -07:00
Eric Anholt
bb8336f82c swrast: Don't assert against glReadPixels of GL_RED and GL_RG.
Everything appears to already be in place for this.  Fixes aborts in:
ARB_texture_rg/fbo-alphatest-formats-float
ARB_texture_rg/fbo-blending-formats-float.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 641dd899bd)
2011-05-31 17:45:41 -07:00
Eric Anholt
d264ed0b48 intel: Use _mesa_base_tex_format for FBO texture attachments.
The _mesa_base_fbo_format variant doesn't handle some texture
internalformats, such as "3".

Fixes:
fbo-blending-formats.
fbo-alphatest-formats
EXT_texture_sRGB/fbo-alphatest-formats

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit cd3568c329)
2011-05-31 17:45:41 -07:00
Eric Anholt
e78908d152 glsl: Perform type checking on "^^" operands.
We were letting any old operand through, which generally resulted in
assertion failures later.

Fixes array-logical-xor.vert.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chad Versace <chad.versace@intel.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 756c262756)
2011-05-31 17:45:41 -07:00
Eric Anholt
1d779672fa glsl: When we've emitted a semantic error for ==, return a bool constant.
This prevents later errors (including an assertion failure) from
cascading the failure.

Fixes invalid-equality-04.vert.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=33303
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chad Versace <chad.versace@intel.com>
(cherry picked from commit 175829f1a8)
2011-05-31 17:45:41 -07:00
Eric Anholt
0ab66b7d0b glsl: Semantically check the RHS of `||' even when short-circuiting.
We just do the AST-to-HIR processing, and only push the instructions
if needed in the constant false case.

Fixes glslparsertest/glsl2/logic-02.frag

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chad Versace <chad.versace@intel.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 9e04b190b5)
2011-05-31 17:45:41 -07:00
Eric Anholt
b47825e626 glsl: Semantically check the RHS of `&&' even when short-circuiting.
We just do the AST-to-HIR processing, and only push the instructions
if needed in the constant true case.

Fixes glslparsertest/glsl2/logic-01.frag

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chad Versace <chad.versace@intel.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 7ec0c97896)
2011-05-31 17:45:41 -07:00
Eric Anholt
7609af4228 glsl: Avoid cascading errors when looking for a scalar boolean and failing.
By always using a boolean, we should generally avoid further
complaints.  The failure case I see is logic_not, where the user might
understandably make the mistake of using `!' on a boolean vector (like
a piglit case did recently!), and then get a further complaint that
the new boolean type doesn't match the bvec it gets assigned to.

Fixes invalid-logic-not-06.vert (assertion failure when the bad type
ends up in an expression and ir_constant_expression gets angry).

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=33314
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 01822706ec)
2011-05-31 17:45:41 -07:00
Kenneth Graunke
ef96b6148b i965: Never enable the GS on Gen6.
Prior to Gen6, we use the GS for breaking down quads, quad-strips,
and line loops.  On Gen6, earlier stages already take care of this,
so we never need the GS.

Since this code is likely completely untested, remove it for now.
We can write new code when enabling real geometry shaders.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 35b3f597bd)
2011-05-31 17:45:41 -07:00
Kenneth Graunke
4d9becfb1f i965: Rename various gen6 #defines to match the documentation.
This should make it easier to cross-reference the code and hardware
documentation, as well as clear up any confusion on whether constants
like CMD_3D_WM_STATE mean WM_STATE (pre-gen6) or 3DSTATE_WM (gen6+).

This does not rename any pre-gen6 defines.
(cherry picked from commit e31defc825)
2011-05-31 17:45:40 -07:00
Hans de Goede
157f4a9d28 texstore: fix regression stricter check for memcpy path for unorm88 and unorm1616
According to https://bugs.freedesktop.org/show_bug.cgi?id=34280
commit 5d1387b2da causes the font corruption
problems people have been seeing under various apps and gnome-shell on r200
cards.

This commit changed (loosened) the check for using the memcpy path in the
former al88 / al1616 texstore functions, which are now also used to
store rg texures. This patch restores the old strict check in case of
al textures. I've no idea why this fixes things, since I don't know the
code in question at all. But after seeing the bisect in bfdo34280 point
to this commit, I gave this fix a try and it fixes the font issues seen on
r200 cards.

[airlied:
r200 has no native working A8, so it does an internal storage format of AL88
however srcFormat == internalFormat == ALPHA when we get to this point,
so it copies, but it wants to store into an AL88 not ALPHA so fails,
I'll also push a piglit test for this on r200].

Many thanks to Nicolas Kaiser who did all the hard work of tracking this down!

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit e338a1b0ce)
2011-05-31 17:45:40 -07:00
Ian Romanick
5f8729859e i965: Remove hint_gs_always and resulting dead code
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit a7fa203f0d)
2011-05-31 17:45:40 -07:00
Henri Verbeet
db17691e28 glx: Only remove the glx_display from the list after it's destroyed.
Signed-off-by: Henri Verbeet <hverbeet@gmail.com>
(cherry picked from commit a26121f375)
2011-05-31 17:45:40 -07:00
Henri Verbeet
43ad25baa7 glx: Destroy dri2Hash on DRI2 display destruction.
Signed-off-by: Henri Verbeet <hverbeet@gmail.com>
(cherry picked from commit a75de67c51)
2011-05-31 17:45:40 -07:00
Henri Verbeet
9e5da5894d mesa: Also update the color draw buffer if it's explicitly set to GL_NONE.
NOTE: This is a candidate for the 7.10 branch.

Signed-off-by: Henri Verbeet <hverbeet@gmail.com>
(cherry picked from commit 158d42c8b0)
2011-05-31 17:45:40 -07:00
Marek Olšák
10bf7aeeeb mesa: forbid UseProgram to be called inside Begin/End
The spec doesn't state it should be an error, but. We have this piglit test
useprogram-inside-begin that passes with this commit. No idea what's correct.

NOTE: This is a candidate for the 7.10 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 29ceeeba20)
2011-06-01 01:44:41 +02:00
Marek Olšák
e6c031c440 st/mesa: conditional rendering should not kill texture decompression via blit
NOTE: This is a candidate for the 7.10 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 8d45bbc422)
2011-06-01 01:43:28 +02:00
Marek Olšák
6850899f44 st/mesa: CopyTex(Sub)Image should not be killed by conditional rendering
NOTE: This is a candidate for the 7.10 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 91e56c8897)
2011-06-01 01:42:43 +02:00
Marek Olšák
c6b6688f88 st/mesa: BlitFramebuffer should not be killed by conditional rendering
NOTE: This is a candidate for the 7.10 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit ccfeb90b75)
2011-06-01 01:42:34 +02:00
Marek Olšák
831c7b1768 swrast: BlitFramebuffer should not be killed by conditional rendering
NOTE: This is a candidate for the 7.10 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit c0277d9539)
2011-06-01 01:42:25 +02:00
Marek Olšák
06b04a2a64 st/mesa: GenerateMipmap should not be killed by conditional rendering
NOTE: This is a candidate for the 7.10 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 1c95c3ef9c)

Conflicts:

	src/mesa/state_tracker/st_context.h
2011-06-01 01:41:54 +02:00
Marek Olšák
23d9baa5a1 st/mesa: fix changing internal format via RenderbufferStorage
The problem is: The second time the function is called with a new
internal format, strb->format is usually not PIPE_FORMAT_NONE.

RenderbufferStorage(... GL_RGBA8 ...);
RenderbufferStorage(... GL_RGBA16 ...); // had no effect on the format

Broken with: fd6f2d6e57
Test: piglit/fbo-storage-completeness

NOTE: This is a candidate for the 7.10 branch.
(if fd6f2d6e57 is cherry-picked as well)

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 49d4e803d3)

Conflicts:

	src/mesa/state_tracker/st_cb_fbo.c
2011-06-01 01:39:36 +02:00
pepp
1b8df41f75 st/mesa: assign renderbuffer's format field when allocating storage
See http://bugs.freedesktop.org/show_bug.cgi?id=36173

NOTE: This is a candidate for the 7.10 branch.

Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit fd6f2d6e57)
2011-06-01 01:27:57 +02:00
Marek Olšák
224ed851ed tgsi/ureg: bump the limit of immediates
Lowered indirect addressing can create lots of immediates.

Fixes piglit/glsl-fs-uniform-array-7 on r300g.

NOTE: This is a candidate for the 7.10 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit ff038170ff)
2011-06-01 01:26:31 +02:00
Kenneth Graunke
b531e75cc5 glsl: Regenerate autogenerated file builtin_function.cpp.
For changes way back in ab58b21634.

Should fix fd.o bug #35603.
2011-05-31 12:02:53 -07:00
Brian Paul
eaadbacb5c mesa: s/height/depth/ in texsubimage()
Fixes http://bugs.freedesktop.org/show_bug.cgi?id=37648

(cherry picked from commit 4609e80288)
2011-05-26 19:26:36 -06:00
Kenneth Graunke
9286d0ddd3 glsl: Fix memory error when creating the supported version string.
Passing ralloc_vasprintf_append a 0-byte allocation doesn't work.  If
passed a non-NULL argument, ralloc calls strlen to find the end of the
string.  Since there's no terminating '\0', it runs off the end.

Fixes a crash introduced in 14880a510a.

(cherry-picked from commit a7d350790b)
2011-05-26 10:45:19 -07:00
Kenneth Graunke
c66ffcf663 intel: Support glCopyTexImage() from ARGB8888 to XRGB8888.
Nexuiz was hitting a software fallback.

(cherry-picked from commit d1fc920f61)
2011-05-19 11:00:10 -07:00
José Fonseca
4a14f76c69 draw: Fix draw_variant_output::format's type.
(cherry picked from commit b79b05e17e)
2011-05-18 13:43:35 -06:00
Brian Paul
068926aaea glsl: add cast to silence signed/unsigned comparison warning 2011-05-18 13:42:37 -06:00
Brian Paul
a5c0969087 glsl: add static qualifier to silence warning 2011-05-18 13:42:13 -06:00
Matt Turner
b6bca28113 r300/compiler: align memory allocations to 8-bytes
Eliminates unaligned accesses on strict architectures. Spotted by Jay
Estabrook.

Signed-off-by: Matt Turner <mattst88@gmail.com>

NOTE: This is a candidate for the 7.10 branch.
(cherry picked from commit 86852236a3)
2011-05-13 01:23:08 +02:00
José Fonseca
348779a1fc mesa: Fix GetVertexAttrib* inside display lists.
GetVertexAttrib*{,ARB} is no longer aliased to the NV calls.

This fixes tracing yofrankie with apitrace, given it requires accurate
results from GetVertexAttribiv*.

NOTE: This is a candidate for the stable branches.
2011-05-12 14:12:26 +01:00
Tom Stellard
abf9217ea4 r300/compiler: Limit instructions to 3 source selects
Some presubtract conversions were generating more than 3 source
selects.

https://bugs.freedesktop.org/show_bug.cgi?id=36527

(cherry picked from commit 4612554dce)

Conflicts:

	src/mesa/drivers/dri/r300/compiler/radeon_compiler_util.c
2011-05-11 18:46:02 -07:00
Alex Buell
0309089e5a configure: bump LIBDRM_REQUIRED to 2.4.24
Signed-off-by: Brian Paul <brianp@vmware.com>
2011-05-06 12:01:35 -06:00
Kostas Georgiou
a8032483ec r600c/g: Add pci id for FirePro 2270
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
2011-05-06 13:10:38 -04:00
Alex Deucher
8963295b1b r600g: add new pci ids
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
2011-05-03 16:20:24 -04:00
Alex Deucher
3371397b1a r600c: add new pci ids
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
2011-05-03 16:18:58 -04:00
José Fonseca
b8c6362389 mesa: GL_PROVOKING_VERTEX_EXT is a GLenum, not GLboolean. 2011-04-30 22:46:16 +01:00
Ian Romanick
73f4273b86 glsl: Regenerate compiler and glcpp files from cherry picks 2011-04-21 17:26:39 -07:00
Carl Worth
6d35d0bda6 glcpp: Fix attempts to expand recursive macros infinitely (bug #32835).
The 095-recursive-define test case was triggering infinite recursion
with the following test case:

	#define A(a, b) B(a, b)
	#define C A(0, C)
	C

Here's what was happening:

  1. "C" was pushed onto the active list to expand the C node

  2. While expanding the "0" argument, the active list would be
     emptied by the code at the end of _glcpp_parser_expand_token_list

  3. When expanding the "C" argument, the active list was now empty,
     so lather, rinse, repeat.

We fix this by adjusting the final popping at the end of
_glcpp_parser_expand_token_list to never pop more nodes then this
particular invocation had pushed itself. This is as simple as saving
the original state of the active list, and then interrupting the
popping when we reach this same state.

With this fix, all of the glcpp-test tests now pass.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=32835
Signed-off-by: Carl Worth <cworth@cworth.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-and-tested-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 9dacbe2226)
2011-04-21 11:35:15 -07:00
Carl Worth
29c2e1f3f7 glcpp: Simplify calling convention of parser's active_list functions
These were all written as generic list functions, (accepting and returning
a list to act upon). But they were only ever used with parser->active as
the list. By simply accepting the parser itself, these functions can update
parser->active and now return nothing at all. This makes the code a bit
more compact.

And hopefully the code is no less readable since the functions are also
now renamed to have "_parser_active" in the name for better correlation
with nearby tests of the parser->active field.
(cherry picked from commit 02d293c08e)
2011-04-21 11:35:06 -07:00
Kenneth Graunke
dca5ddf471 i965: Allocate the whole URB to the VS and fix calculations for Gen6.
Since we never enable the GS on Sandybridge, there's no need to allocate
it any URB space.

Furthermore, the previous calculation was incorrect: it neglected to
multiply by nr_vs_entries, instead comparing whether twice the size of
a single VS URB entry was bigger than the entire URB space.  It also
neglected to take into account that vs_size is in units of 128 byte
blocks, while urb_size is in bytes.

Despite the above problems, the calculations resulted in an acceptable
programming of the URB in most cases, at least on GT2.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>

(cherry picked from commit 42a8057000)
2011-04-20 15:28:45 -07:00
Ian Romanick
ff77a69ae3 intel: Fix ROUND_DOWN_TO macro
Previously the macro would (ALIGN(value - alignment - 1, alignment)).
At the very least, this was missing parenthesis around "alignment -
1".  As a result, if value was already aligned, it would be reduced by
alignment.  Condisder:

     x = ROUND_DOWN_TO(256, 128);

This becomes:

    x = ALIGN(256 - 128 - 1, 128);

Or:

    x = ALIGN(127, 128);

Which becomes:

    x = 128;

This macro is currently only used in brw_state_batch
(brw_state_batch.c).  It looks like the original version of this macro
would just use too much space in the batch buffer.  It's possible, but
not at all clear to me from the code, that the original behavior is
actually desired.

In any case, this patch does not cause any piglit regressions on my
Ironlake system.

I also think that ALIGN_FLOOR would be a better name for this macro,
but ROUND_DOWN_TO matches rounddown in the Linux kernel.

Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Keith Whitwell <keithw@vmware.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 7e809f0b8d)
2011-04-20 15:28:45 -07:00
Eric Anholt
510fb3269e i965: Fix the VS thread limits for GT1, and clarify the WM limits on both.
(cherry picked from commit 904b8ba1bb)
2011-04-20 15:28:45 -07:00
Kenneth Graunke
62fad6cb30 intel: Add IS_GT2 macro for recognizing Sandybridge GT2 systems.
Also, refactor IS_GEN6 to use the IS_GT1 and IS_GT2 macros.

(cherry picked from commit ee8d182426)
2011-04-20 15:28:45 -07:00
Kenneth Graunke
0f02b4253d i965: Resolve implied moves in brw_dp_READ_4_vs_relative.
Fixes piglit test glsl-vs-arrays-3 on Sandybridge, as well as garbage
rendering in 3DMarkMobileES 2.0's Taiji demo and GLBenchmark 2.0's
Egypt and PRO demos.

NOTE: This a candidate for stable release branches.  It depends on
commit 9a21bc6401.
(cherry picked from commit 9d60a7ce08)
2011-04-20 15:28:45 -07:00
Kenneth Graunke
773ea1a234 i965: Refactor Sandybridge implied move handling.
This is actually a squash of the following two commits.  The first
caused a regression, and the second fixes it.  The refactor of the
first is needed for another patch that fixes an SNB bug.

    i965: Refactor Sandybridge implied move handling.

    This was open-coded in three different places, and more are necessary.
    Extract this into a function so it can be reused.

    Unfortunately, not all variations were the same: in particular, one set
    compression control and checked that the source register was not
    ARF_NULL.  This seemed like a good idea, so all cases now do so.
    (cherry picked from commit 9a21bc6401)

    i965: Fix null register use in Sandybridge implied move resolution.

    Fixes regressions caused by commit 9a21bc6401, namely GPU hangs when
    running gnome-shell or compiz (Mesa bugs #35820 and #35853).

    I incorrectly refactored the case that dealt with ARF_NULL; even in that
    case, the source register needs to be changed to the MRF.

    NOTE: This is a candidate for the 7.10 branch (if 9a21bc6401 is
    cherry-picked, take this one too).
    (cherry picked from commit a019dd0d6e)
2011-04-20 15:28:45 -07:00
Zou Nan hai
41d5dd4a6e i965: Align interleaved URB write length to 2
The BSpec says that interleave URB writes must be aligned, so this
patch fulfills that requirement.

This is half of patch 6c32477 from master.

Signed-off-by: Zou Nan hai <nanhai.zou@intel.com>
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
2011-04-20 15:28:45 -07:00
Brian Paul
2c83c28608 docs: replace llvmpipe/README with docs/llvmpipe.html 2011-04-20 09:57:48 +01:00
Brian Paul
d4a82b3146 llvmpipe: document issue with LLVM 2.8 and earlier with AVX 2011-04-20 09:57:07 +01:00
José Fonseca
8d539264ca llvmpipe: Update readme. 2011-04-20 09:56:39 +01:00
Alan Hourihane
c13a7748de Check for out of memory when creating fence 2011-04-20 09:51:48 +01:00
Vinson Lee
f814e28e6d gallivm: Disable MMX-disabling code on llvm-2.9.
The disable-mmx option was removed in llvm-2.9svn by revisions 122188
and 122189.

Fixes FDO bug 32564.
2011-04-20 09:50:55 +01:00
Brian Paul
7e325d5e62 Makefile: add missing Scons files 2011-04-19 09:11:29 -06:00
Brian Paul
4373f6b9ad scons: remove dangling reference to state_trackers/python/SConscript 2011-04-19 09:06:54 -06:00
Tom Stellard
1f8c5611a2 r300/compiler: Fix dataflow analysis bug with ELSE blocks
Writes within ELSE blocks were being ignored which prevented us from
discovering all possible writers for some register values.

Fixes piglit glsl-fs-raytrace-bug27060

(cherry picked from commit ffc1d166d2)
2011-04-18 21:54:58 -07:00
Tom Stellard
bf9a469cb6 r300/compiler: Fix incorrect presubtract conversion
ADD instructions with constant swizzles can't be converted to
presubtract operations.

(cherry picked from commit 0fa81d6d05)
2011-04-16 14:40:20 -07:00
Ian Romanick
f890661a9a ir_to_mesa: Handle shadow compare w/projection and LOD bias correctly
The code would previously handle the projection, then swizzle the
shadow comparitor into place.  However, when the projection is done
"by hand," as in the TXB case, the unprojected shadow comparitor would
over-write the projected shadow comparitor.

Shadow comparison with projection and LOD is an extremely rare case in
real application code, so it shouldn't matter that we don't handle
that case with the greatest efficiency.

NOTE: This is a candidate for the stable branches.

Reviewed-by: Brian Paul <brianp@vmware.com>
References: https://bugs.freedesktop.org/show_bug.cgi?id=32395
(cherry picked from commit 9996a86085)
2011-04-11 19:06:48 -07:00
Ian Romanick
b31425aae9 glsl: Fix off-by-one error setting max_array_access for non-constant indexing
NOTE: This is a candidate for the stable branches.
(cherry picked from commit 0d9d036004)
2011-04-11 19:06:48 -07:00
Kenneth Graunke
6c7a5d52ee i965/fs: Switch W and 1/W in Sandybridge interpolation setup.
Various documentation mentions that "W" is handed to the WM stage,
but further digging seems to indicate that they really mean 1/W.

The code here is still unclear, but changing this fixes piglit
test "fragcoord_w" on Sandybridge as well as a Khronos ES2 conformance
test.  I also tested 3DMarkMobile ES2.0's taiji and hoverjet demos, as
well as Nexuiz, just to be safe.

(cherry-picked from commit 5d7fefb9af)
2011-04-11 13:21:53 -07:00
Brian Paul
3c405079fd docs: add link to 7.10.2 release notes 2011-04-11 13:13:58 -06:00
Ian Romanick
1fb1012bf1 docs: Add 7.10.2 md5sums 2011-04-06 13:44:29 -07:00
Ian Romanick
b0866f6cfd docs: update news.html with 7.10.2 release 2011-04-06 13:41:43 -07:00
Ian Romanick
c6a68814b4 docs: Add change log to 7.10.2 release notes 2011-04-06 13:32:09 -07:00
Ian Romanick
3831ba6dd1 mesa: Remove nonexistant files from _FILES lists
This allows 'make -j1 tarballs' to work.
2011-04-06 13:29:59 -07:00
Ian Romanick
812e11f4b4 mesa: set version string to 7.10.2 2011-04-06 10:01:02 -07:00
Ian Romanick
0c69a2fda5 docs: Initial bits of 7.10.2 release notes 2011-04-06 09:59:32 -07:00
Tom Stellard
6e08ceb77d r300/compiler: Don't try to convert RGB to Alpha in full instructions
(cherry picked from commit cd2857fae1)
2011-04-06 00:41:29 -07:00
Ian Romanick
50dccfdbef Revert "i965: bump VS thread number to 60 on SNB"
Increasing the number of VS threads beyond 1 causes some regressions
in vertex shader tests on Sugar Bay GT1 systems.

This reverts commit c21a44463a.

References: https://bugs.freedesktop.org/show_bug.cgi?id=35730
2011-04-05 16:14:39 -07:00
Ian Romanick
22035e3d84 glcpp: Refresh autogenerated lexer files
These are the results of commit 9ebb904 and 7cf7c966.
2011-04-05 16:09:47 -07:00
Tom Stellard
62b75d889c r300/compiler: Fix vertex shader MAD instructions with constant swizzles
(cherry picked from commit d8361400b7)
2011-04-05 09:26:26 -07:00
Brian Paul
7c9d66c60f glsl: silence warning in printf() with a cast
(cherry picked from commit 0eab3a8a97)
2011-04-05 09:12:20 -06:00
Fabian Bieler
1068b7f9ed st/mesa: Apply LOD from texture object
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 5d5db24a26)
2011-04-05 09:10:49 -06:00
Brian Paul
e17ac39d38 st/mesa: Apply LOD bias from correct texture unit
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit d863bd7d7b)
2011-04-05 09:10:26 -06:00
Brian Paul
33129c040b mesa: Guard against null pointer deref in fbo validation
This matches the behaviour below when numSamples is compared.

At least with the gallium state tracker this can actually occur if st_render_texture fails.

Signed-off-by: Brian Paul <brianp@vmware.com>

(cherry picked from commit c7339d42c6)
2011-04-05 07:54:04 -06:00
Brian Paul
ed5c9ae016 docs: update prerequisites, remove old demo info
(cherry picked from commit 32a11e5324)
2011-04-04 11:31:30 -06:00
Brian Paul
06422ce0d2 docs: update info about Mesa packaging/contents
(cherry picked from commit 48f696c793)
2011-04-04 11:31:16 -06:00
Marek Olšák
cd2cf02139 r300/compiler: apply the texture swizzle to shadow pass and fail values too
Piglit tests:
- glsl-fs-shadow2d-01
- glsl-fs-shadow2d-02
- glsl-fs-shadow2d-03
- fs-shadow2d-red-01
- fs-shadow2d-red-02
- fs-shadow2d-red-03

NOTE: This is a candidate for the stable branches.
(cherry picked from commit 0d96ae8fc7)
2011-04-04 19:14:29 +02:00
Marek Olšák
f3a21be95e r300/compiler: propagate SaturateMode down to the result of shadow comparison
NOTE: This is a candidate for the stable branches.
(cherry picked from commit 2679760834)
2011-04-04 19:14:20 +02:00
Alex Deucher
5a3f1aee64 r600g: add some additional ontario pci ids
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
2011-04-04 11:27:25 -04:00
Alex Deucher
0372db6fa8 r600c: add new ontario pci ids
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
2011-04-04 11:27:09 -04:00
Marek Olšák
868ca30235 r300g: tell the GLSL compiler to lower the continue opcode
NOTE: This is a candidate for the stable branches.
(cherry picked from commit f814dd7a81)
2011-04-03 19:37:43 +02:00
Kenneth Graunke
b39603e008 glsl: Accept precision qualifiers on sampler types, but only in ES.
GLSL 1.30 states clearly that only float and int are allowed, while the
GLSL ES specification's issues section states that sampler types may
take precision qualifiers.

Fixes compilation failures in 3DMarkMobileES 2.0 and GLBenchmark 2.0.

(cherry-picked from commit 8752824f27)
2011-04-02 20:11:10 -07:00
Kenneth Graunke
73f7453318 i965/fs: Fix linear gl_Color interpolation on pre-gen6 hardware.
Civilization 4's shaders make heavy use of gl_Color and don't use
perspective interpolation.  This resulted in rivers, units, trees, and
so on being rendered almost entirely white.  This is a regression
compared to the old fragment shader backend.

Found by inspection (comparing the old and new FS backend code).

References: https://bugs.freedesktop.org/show_bug.cgi?id=32949

(cherry-picked from commit 0c8beb0ab5)
2011-04-02 20:10:40 -07:00
Tom Stellard
a947d9be61 prog_optimize: Fix reallocating registers for shaders with loops
Registers that are used inside of loops need to be considered live
starting with the first instruction of the outermost loop.

https://bugs.freedesktop.org/show_bug.cgi?id=34370

(cherry picked from commit 18dcbd358f)

Reviewed-by: Eric Anholt <eric@anholt.net>
2011-03-31 21:24:32 -07:00
Jerome Glisse
ed4aa47d42 r600g: move user fence into base radeon structure
This avoid any issue when context is free and we still try to
access fence through radeon structure.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
(cherry picked from commit 63b9790a55)
2011-03-30 11:06:52 +02:00
Michel Dänzer
5bbaf1992c Use proper source row stride when getting depth/stencil texels.
(cherry picked from commit b082e04619)
2011-03-30 10:54:51 +02:00
Benjamin Franzke
71b06c63ce st/dri: Fix surfaceless gl using contexts with previous bound surfaces
ctx->dPriv might be != NULL then draw which is NULL is accessed:

struct dri_drawable *draw = dri_drawable(driDrawPriv);
[..]
if (ctx->dPriv != driDrawPriv) {
      ctx->dPriv = driDrawPriv;
      draw->texture_stamp = driDrawPriv->lastStamp - 1;
}

Cherry-picked from 0acb31be17

Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2011-03-30 02:26:41 +02:00
Henri Verbeet
7fcaf9ae80 st/mesa: Validate state before doing blits.
Specifically, this ensures things like the front buffer actually exist. This
fixes piglt fbo/fbo-sys-blit and fd.o bug 35483.

Signed-off-by: Henri Verbeet <hverbeet@gmail.com>
(cherry picked from commit 5d7c27f5ec)
2011-03-30 00:53:07 +02:00
Marek Olšák
d464f5908f r300g: fix alpha-test with no colorbuffer
Piglit:
- fbo-alphatest-nocolor

NOTE: This is a candidate for the stable branches.
(cherry picked from commit 226ae9d6c8)
2011-03-24 23:46:28 +01:00
Eric Anholt
78fa94f091 i965: Fix alpha testing when there is no color buffer in the FBO.
We were alpha testing against an unwritten value, resulting in garbage.
(part of) Bug #35073.
(cherry picked from commit a99447314c)
2011-03-24 13:42:18 -07:00
Eric Anholt
1ae33556dd i965: Apply a workaround for the Ironlake "vertex flashing".
This is an awful hack and will hurt performance on Ironlake, but we're
at a loss as to what's going wrong otherwise.  This is the only common
variable we've found that avoids the problem on 4 applications
(CelShading, gnome-shell, Pill Popper, and my GLSL demo), while other
variables we've tried appear to only be confounding.  Neither the
specifications nor the hardware team have been able to provide any
enlightenment, despite much searching.

https://bugs.freedesktop.org/show_bug.cgi?id=29172
Tested by:	Chris Lord <chris@linux.intel.com> (Pill Popper)
Tested by:	Ryan Lortie <desrt@desrt.ca> (gnome-shell)
(cherry picked from commit 1a57717bbe)
2011-03-24 13:40:43 -07:00
Zou Nan hai
c21a44463a i965: bump VS thread number to 60 on SNB
Signed-off-by: Zou Nan hai <nanhai.zou@intel.com>
(cherry picked from commit 6c324777a6)
2011-03-24 13:40:25 -07:00
Ian Romanick
386921cf45 glsl: Only allow unsized array assignment in an initializer
It should have been a tip when the spec says "However, implicitly
sized arrays cannot be assigned to. Note, this is a rare case that
*initializers and assignments appear to have different semantics*."
(empahsis mine)

Fixes bugzilla #34367.

NOTE: This is a candidate for stable release branches.
(cherry picked from commit 85caea29c1)
2011-03-23 18:59:31 -07:00
Ian Romanick
310d85c492 glsl: Use insert_before for lists instead of open coding it
(cherry picked from commit bdb6a6ef83)
2011-03-23 18:59:22 -07:00
Ian Romanick
55d86204f3 linker: Add imported functions to the linked IR
Fixes piglit test glsl-function-chain16 and bugzilla #34203.

NOTE: This is a candidate for stable release branches.
(cherry picked from commit 60f898a90e)
2011-03-23 18:12:55 -07:00
Ian Romanick
da8c178c8b glsl: Add several function / call related validations
The signature list in a function must contain only ir_function_signature nodes.

The target of an ir_call must be an ir_function_signature.

These were added while trying to debug Mesa bugzilla #34203.
(cherry picked from commit 8bbfbb14ee)
2011-03-23 18:12:24 -07:00
Ian Romanick
9dec904ef3 glsl: Function signatures cannot have NULL return type
The return type can be void, and this is the case where a `_ret_val'
variable should not be declared.
(cherry picked from commit 2df56b002d)
2011-03-23 18:12:16 -07:00
Ian Romanick
856a661d2f glsl: Process redeclarations before initializers
If an array redeclaration includes an initializer, the initializer
would previously be dropped on the floor.  Instead, directly apply the
initializer to the correct ir_variable instance and append the
generated instructions.

Fixes bugzilla #34374 and piglit tests glsl-{vs,fs}-array-redeclaration.

NOTE: This is a candidate for stable release branches.  0292ffb8 and
8e6cb9fe are also necessary.
(cherry picked from commit 09a4ba0fc3)
2011-03-23 16:55:08 -07:00
Ian Romanick
f0231a44b9 glsl: Refactor AST-to-HIR code handling variable redeclarations
Some significant edits were made to this patch during cherry picking.
There some fairly major conflicts due to GLSL 1.30 features and
extensions added in master that do not exist in the 7.10 branch.

(cherry picked from commit 8e6cb9fe51)
2011-03-23 16:55:07 -07:00
Ian Romanick
a28cef5e6e glsl: Refactor AST-to-HIR code handling variable initializers
(cherry picked from commit 0292ffb85c)
2011-03-23 16:55:07 -07:00
José Fonseca
1efb7428ed mesa: More glGet* fixes.
glGet(GL_NORMAL_ARRAY) giving potentially wrong results.

Most of glGet(GL_XXX_ARRAY_BUFFER_BINDING) giving totally bogus results.
2011-03-23 17:18:54 +00:00
Chad Versace
b8a077cee0 i965: Fix tex_swizzle when depth mode is GL_RED
Change swizzle from (x000) to (x001).

Signed-off-by: Chad Versace <chad.versace@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
2011-03-22 19:55:05 -07:00
Kenneth Graunke
9ebb904bbd glcpp: Remove trailing contexts from #if rules.
These are now unnecessary.

(cherry picked from commit 09e1bebc25)
2011-03-22 19:55:05 -07:00
Kenneth Graunke
7cf7c966f8 glcpp: Rework lexer to use a SKIP state rather than REJECT.
Previously, the rule deleted by this commit was matched every single
time (being the longest match).  If not skipping, it used REJECT to
continue on to the actual correct rule.

The flex manual advises against using REJECT where possible, as it is
one of the most expensive lexer features.  So using it on every match
seems undesirable. Perhaps more importantly, it made it necessary for
the #if directive rules to contain a look-ahead pattern to make them
as long as the (now deleted) "skip the whole line" rule.

This patch introduces an exclusive start state, SKIP, to avoid REJECTs.
Each time the lexer is called, the code at the top of the rules section
will run, implicitly switching the state to the correct one.

Fixes piglit tests 16384-consecutive-chars.frag and
16385-consecutive-chars.frag.

(cherry picked from commit f20656e944)
2011-03-22 19:44:37 -07:00
José Fonseca
7628e489e4 mesa: Fix typo glGet*v(GL_TEXTURE_COORD_ARRAY_*). 2011-03-22 23:00:13 +00:00
Tom Stellard
d525a1b468 r300/compiler: Use a 4-bit writemask in pair instructions
We now use a 4-bit writemask for all instruction types, which makes it
easier to write generic helper functions to manipulte writemasks.

(cherry picked from commit 9d2ef284bb)
2011-03-18 19:52:14 -07:00
Dave Airlie
d59da64817 r600: don't close fd on failed load
This fd gets passed in from outside, closing it causes the X.org server
to crap out when the driver doesn't identify the chipset.

Signed-off-by: Dave Airlie <airlied@redhat.com>
2011-03-14 07:11:16 +10:00
Marek Olšák
ea26cc8696 st/mesa: fix crash when using both user and vbo buffers with the same stride
If two buffers had the same stride where one buffer is a user one and
the other is a vbo, it was considered to be one interleaved buffer,
resulting in incorrect rendering and crashes.

This patch makes sure that the interleaved buffer is either user or vbo,
not both.
(cherry picked from commit 695cdee678)
2011-03-12 22:10:44 +01:00
Marek Olšák
954787cee1 r300/compiler: remove unused variables
(cherry picked from commit ff8baec5bc)
2011-03-12 22:04:17 +01:00
Marek Olšák
75fd54e7bd r300/compiler: fix equal and notequal shadow compare functions
(cherry picked from commit 4dfcf3c4fe)
2011-03-12 22:03:55 +01:00
Marek Olšák
8453dce232 r300/compiler: saturate Z before the shadow comparison
This fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=31159

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit eb1acd1613)
2011-03-12 22:03:41 +01:00
Marek Olšák
e48b0b9032 r300/compiler: do not set TEX_IGNORE_UNCOVERED on r500
The docs say it can be set for direct texture lookups, but even that
causes problems.

This fixes the wireframe bug:
https://bugs.freedesktop.org/show_bug.cgi?id=32688

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 6da4866ffd)
2011-03-12 21:51:28 +01:00
Marek Olšák
3a1b2bb372 r300/compiler: TEX instructions don't support negation on source arguments
This fixes piglit:
- glsl-fs-texture2d-dependent-4

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 1e97b4dd10)
2011-03-12 21:51:17 +01:00
Marek Olšák
38f0e9b651 r300/compiler: Abs doesn't cancel Negate (in the conversion to native swizzles)
NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 589d835dfd)
2011-03-12 21:51:05 +01:00
Marek Olšák
aaf7c86b30 r300/compiler: fix translating the src negate bits in pair_translate
(1, -_, ...) was converted to (-1, ...) because of the negation
in the second component.
Masking out the unused bits fixes this.

Piglit:
- glsl-fs-texture2d-branching

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit d96305e4fc)
2011-03-12 21:50:53 +01:00
Marek Olšák
c3fcd8af4f r300/compiler: fix the saturate modifier when applied to TEX instructions
This bug can only be triggered if the source texture is either signed or float.
(cherry picked from commit e4707604ab)
2011-03-12 21:50:36 +01:00
Marek Olšák
b5d293e4fd st/mesa: fail to alloc a renderbuffer if st_choose_renderbuffer_format fails
This fixes:
  state_tracker/st_format.c:401:st_pipe_format_to_mesa_format:
  Assertion `0' failed.
(cherry picked from commit fb5d9e1199)

Conflicts:

	src/mesa/state_tracker/st_cb_fbo.c
2011-03-12 21:46:01 +01:00
Marek Olšák
3a02a2bf25 st/mesa: fix crash when DrawBuffer->_ColorDrawBuffers[0] is NULL
This fixes the game Tiny and Big.
(cherry picked from commit 7942e6a5ae)
2011-03-12 21:40:47 +01:00
Brian Paul
3ed9054cc5 docs: add, fix release notes links 2011-03-08 09:16:30 -07:00
Brian Paul
9a5de0895e docs: fill in 7.10.1 release data 2011-03-08 09:16:14 -07:00
Brian Paul
2aebe261aa docs: update news.html with 7.10.1 and 7.9.2 releases 2011-03-08 09:15:56 -07:00
Brian Paul
d7c64d7c36 docs: pull 7.9.2 release notes into 7.10 branch 2011-03-08 09:14:33 -07:00
Ian Romanick
68cdea9fb2 docs: Add 7.10.1 md5sums 2011-03-02 14:14:33 -08:00
Ian Romanick
565caabf40 docs: Add change log to 7.10.1 release notes 2011-03-02 13:54:20 -08:00
Ian Romanick
e4fefc3c32 mesa: set version string to 7.10.1 (final) 2011-03-02 13:49:17 -08:00
Ian Romanick
b0a7492aeb intel: Remove driver date and related bits from renderer string
Not only did this contain lies, it contained lies that wouldn't be
useful even if true.
2011-03-01 13:35:39 -08:00
Ian Romanick
8aabb1bc99 docs: Clean up bug fixes list
All the unnumbered bugs are first.  These are followed by numbered
bugs sorted by bug number.
2011-03-01 13:19:58 -08:00
Ian Romanick
a67a0a0589 docs: Update 7.10.1 with (hopefully) the last of the cherry picks 2011-03-01 13:03:47 -08:00
Cyril Brulebois
a6263f2738 Point to bugs.freedesktop.org rather than bugzilla.freedesktop.org
Suggested by a freedesktop.org admin.

Signed-off-by: Cyril Brulebois <kibi@debian.org>
(cherry picked from commit d252db7af1)
2011-03-01 12:57:29 -08:00
Brian Paul
7e158e85bd docs: updated environment variable list
(cherry picked from commit 1bf9954bb4)

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
2011-03-01 12:55:44 -08:00
Sam Hocevar
2fc2553261 docs: fix glsl_compiler name
(cherry picked from commit fde4943688)
2011-03-01 12:54:25 -08:00
Sam Hocevar
355601812a docs: add glsl info
(cherry picked from commit 3e8fb54fb8)
2011-03-01 12:53:56 -08:00
Chad Versace
8a27f9845b tnl: Add support for datatype GL_FIXED in vertex arrays
Before populating the vertex buffer attribute pointer (VB->AttribPtr[]),
convert vertex data in GL_FIXED format to GL_FLOAT.

Fixes bug: http://bugs.freedesktop.org/show_bug.cgi?id=34047

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit a231ac23f4)
2011-03-01 11:53:58 -08:00
Ian Romanick
e160c815c2 i915: Force lowering of all types of indirect array accesses in the FS
NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 4c1dc1c4d7)
2011-03-01 10:47:01 -08:00
Ian Romanick
a9d3cce8f3 i915: Calculate partial result to temp register first
Previously the SNE and SEQ instructions would calculate the partial
result to the destination register.  This would cause problems if the
destination register was also one of the source registers.

Fixes piglit tests glsl-fs-any, glsl-fs-struct-equal,
glsl-fs-struct-notequal, glsl-fs-vec4-operator-equal,
glsl-fs-vec4-operator-notequal.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 53b8b68843)
2011-03-01 10:46:52 -08:00
Ian Romanick
a62d3534e1 i915: Only mark a register as available if all components are written
Previously a register would be marked as available if any component
was written.  This caused shaders such as this:

  0: TEX TEMP[0].xyz, INPUT[14].xyyy, texture[0], 2D;
  1: MUL TEMP[1], UNIFORM[0], TEMP[0].xxxx;
  2: MAD TEMP[2], UNIFORM[1], TEMP[0].yyyy, TEMP[1];
  3: MAD TEMP[1], UNIFORM[2], TEMP[0].zzzz, TEMP[2];
  4: ADD TEMP[0].xyz, TEMP[1].xyzx, UNIFORM[3].xyzx;
  5: TEX TEMP[1].w, INPUT[14].xyyy, texture[0], 2D;
  6: MOV TEMP[0].w, TEMP[1].wwww;
  7: MOV OUTPUT[2], TEMP[0];
  8: END

to produce incorrect code such as this:

  BEGIN
  DCL S[0]
  DCL T_TEX0
  R[0] = MOV T_TEX0.xyyy
  U[0] = TEXLD S[0],R[0]
  R[0].xyz = MOV U[0]
  R[1] = MUL CONST[0], R[0].xxxx
  R[2] = MAD CONST[1], R[0].yyyy, R[1]
  R[1] = MAD CONST[2], R[0].zzzz, R[2]
  R[0].xyz = ADD R[1].xyzx, CONST[3].xyzx
  R[0] = MOV T_TEX0.xyyy
  U[0] = TEXLD S[0],R[0]
  R[1].w = MOV U[0]
  R[0].w = MOV R[1].wwww
  oC = MOV R[0]
  END

Note that T_TEX0 is copied to R[0], but the xyz components of R[0] are
still expected to hold a calculated value.

Fixes piglit tests draw-elements-vs-inputs, fp-kill, and
glsl-fs-color-matrix.  It also fixes Meego bugzilla #13005.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit a04582739e)
2011-03-01 10:46:41 -08:00
Kenneth Graunke
f9f01e40c7 Revert "i965/fs: Correctly set up gl_FragCoord.w on Sandybridge."
This reverts commit 2171197559, as it
caused a regression on Ironlake (bug #34646).
2011-03-01 01:09:52 -08:00
Kenneth Graunke
c9ded4d418 glsl: Use reralloc instead of plain realloc.
Plugs a memory leak when compiling shaders with user defined structures.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit eb639349e2)
2011-02-28 16:40:58 -08:00
Kenneth Graunke
022f1110dd i965: Increase Sandybridge point size clamp in the clip state.
255.875 matches the hardware documentation.  Presumably this was a typo.

NOTE: This is a candidate for the 7.10 branch, along with
      commit 2bfc23fb86.

Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit e6e5c1f46d)
2011-02-28 16:40:58 -08:00
Kenneth Graunke
bb90087eda i965/fs: Refactor control flow stack handling.
We can't safely use fixed size arrays since Gen6+ supports unlimited
nesting of control flow.

NOTE: This is a candidate for the 7.10 branch.

Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit df2aef0e19)
2011-02-28 16:40:58 -08:00
Kenneth Graunke
e0f6193024 i965: Increase Sandybridge point size clamp.
255.875 matches the hardware documentation.  Presumably this was a typo.

Found by inspection.  Not known to fix any issues.

Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 2bfc23fb86)
2011-02-28 16:40:57 -08:00
Kenneth Graunke
2171197559 i965/fs: Correctly set up gl_FragCoord.w on Sandybridge.
pixel_w is the final result; wpos_w is used on gen4 to compute it.

NOTE: This is a candidate for the 7.10 branch.

Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 4a3b28113c)
2011-02-28 16:40:57 -08:00
Kenneth Graunke
ec4822a316 i965/fs: Avoid register coalescing away gen6 MATH workarounds.
The code that generates MATH instructions attempts to work around
the hardware ignoring source modifiers (abs and negate) by emitting
moves into temporaries.  Unfortunately, this pass coalesced those
registers, restoring the original problem.  Avoid doing that.

Fixes several OpenGL ES2 conformance failures on Sandybridge.

NOTE: This is a candidate for the 7.10 branch.

Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 2c2686b912)
2011-02-28 16:40:57 -08:00
Eric Anholt
f13e45d45d i965/fs: Add a helper function for detecting math opcodes.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>

(cherry picked from commit 382c2d99da)
2011-02-28 16:40:57 -08:00
Kenneth Graunke
4fef0bc115 i965: Fix shaders that write to gl_PointSize on Sandybridge.
gl_PointSize (VERT_RESULT_PSIZ) doesn't take up a message register,
as it's part of the header.  Without this fix, writing to gl_PointSize
would cause the SF to read and use the wrong attributes, leading to all
kinds of random looking failure.

Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 3e91070ea8)
2011-02-28 16:32:15 -08:00
Kenneth Graunke
8bf3a4f05e i965/fs: Apply source modifier workarounds to POW as well.
Single-operand math already had these workarounds, but POW (the only two
operand function) did not.  It needs them too - otherwise we can hit
assertion failures in brw_eu_emit.c when code is actually generated.

NOTE: This is a candidate for the 7.10 branch.

Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 72cd7e87d3)
2011-02-28 16:32:04 -08:00
Ian Romanick
aa180f2786 docs: update 7.10.1 release notes with Ian's recent cherry picks 2011-02-28 15:58:05 -08:00
Ian Romanick
52a274a4c0 linker: Fix off-by-one error implicit array sizing
Arrays are zero based.  If the highest element accessed is 6, the
array needs to have 7 elements.

Fixes piglit test glsl-fs-implicit-array-size-03 and bugzilla #34198.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 25b36e8ff8)
2011-02-28 15:16:56 -08:00
nobled
d7d55ab841 glx: Put null check before use
'dpy' was being checked for null *after* it was already used once.

Also add a null check for psc, and drop gc's redundant initialization.
(cherry picked from commit b5dc40710d)
2011-02-28 15:16:56 -08:00
Ian Romanick
bba05bc699 glsl: Regenerate compiler and glcpp files from cherry picks 2011-02-28 15:16:55 -08:00
Ian Romanick
bcdb23ef8a glsl: Finish out the reduce/reduce error fixes
Track variables, functions, and types during parsing.  Use this
information in the lexer to return the currect "type" for identifiers.

Change the handling of structure constructors.  They will now show up
in the AST as constructors (instead of plain function calls).

Fixes piglit tests constructor-18.vert, constructor-19.vert, and
constructor-20.vert.  Also fixes bugzilla #29926.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 8842158944)
2011-02-28 14:58:19 -08:00
Keith Packard
5db7ee0fde glsl: Eliminate reduce/reduce conflicts in glsl grammar
This requires lexical disambiguation between variable and type
identifiers (as most C compilers do).

Signed-off-by: Keith Packard <keithp@keithp.com>

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit f4b812e1a6)
2011-02-28 14:58:11 -08:00
Chad Versace
614eff1fc6 glsl: Reinstate constant-folding for division by zero
Fixes regression: https://bugs.freedesktop.org/show_bug.cgi?id=34160

Commit e7c1f058d1 disabled constant-folding
when division-by-zero occured. This was a mistake, because the spec does
allow division by zero. (From section 5.9 of the GLSL 1.20 spec: Dividing
by zero does not cause an exception but does result in an unspecified
value.)

For floating-point division, the original pre-e7c1f05 behavior is
reinstated.

For integer division, constant-fold 1/0 to 0.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>

(cherry picked from commit 62c8c77333)
2011-02-28 14:52:09 -08:00
Chad Versace
89576ea75b glsl: Set operators '%' and '%=' to be reserved when GLSL < 1.30
From section 5.9 of the GLSL 1.20 spec:
   The operator modulus (%) is reserved for future use.

From section 5.8 of the GLSL 1.20 spec:
   The assignments modulus into (%=), left shift by (<<=), right shift by
   (>>=), inclusive or into ( |=), and exclusive or into ( ^=). These
   operators are reserved for future use.

The GLSL ES 1.00 spec and GLSL 1.10 spec have similiar language.

Fixes bug:
https://bugs.freedesktop.org//show_bug.cgi?id=33916

Fixes Piglit tests:
spec/glsl-1.00/compiler/arithmetic-operators/modulus-00.frag
spec/glsl-1.00/compiler/assignment-operators/modulus-assign-00.frag
spec/glsl-1.10/compiler/arithmetic-operators/modulus-00.frag
spec/glsl-1.10/compiler/assignment-operators/modulus-assign-00.frag
spec/glsl-1.20/compiler/arithmetic-operators/modulus-00.frag
spec/glsl-1.20/compiler/assignment-operators/modulus-assign-00.frag
(cherry picked from commit 82f994f386)
2011-02-28 14:48:02 -08:00
Chad Versace
0ca5a1593d glcpp: Raise error when modulus is zero
For example, this now raises an error:
   #define XXX 1 / 0

Fixes bug: https://bugs.freedesktop.org//show_bug.cgi?id=33507
Fixes Piglit test: spec/glsl-1.10/preprocessor/modulus-by-zero.vert

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit fd1252ab67)
2011-02-28 14:47:37 -08:00
Ian Romanick
ec162447a5 mesa: Initial size for secondary color array is 3
See table 6.7 on page 347 of the OpenGL 3.0 specification.
(cherry picked from commit 09e15ac76a)
2011-02-28 14:47:01 -08:00
Eric Anholt
8b91cf406a i965: Fix a bug in i965 compute-to-MRF.
Fixes piglit glsl-fs-texture2d-branching.  I couldn't come up with a
testcase that didn't involve dead code, but it's still worthwhile to
fix I think.
(cherry picked from commit 8ce425f3e3)
2011-02-28 14:46:02 -08:00
Kenneth Graunke
ab58b21634 glsl: Fix use of uninitialized values in _mesa_glsl_parse_state ctor.
This has probably existed since e5e34ab18e or so.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit dfdb9fda82)
2011-02-28 14:33:56 -08:00
Ian Romanick
d92cc2f45f glsl: Reject shader versions not supported by the implementation
Previously we'd happily compile GLSL 1.30 shaders on any driver.  We'd
also happily compile GLSL 1.10 and 1.20 shaders in an ES2 context.
This has been a long standing FINISHME in the compiler.

NOTE: This is a candidate for the 7.9 and 7.10 branches
(cherry picked from commit 14880a510a)
2011-02-28 14:33:33 -08:00
Ian Romanick
a527411469 glsl: Ensure that all GLSL versions are supported in the stand-alone compiler
NOTE: This is a candidate for the 7.9 and 7.10 branches
(cherry picked from commit e5e34ab18e)
2011-02-28 14:33:19 -08:00
Ian Romanick
f9e01af6c3 intel: Fix typeos from 3d028024 and 790ff232
...and remove egg from face.
(cherry picked from commit 2fb0aebd4a)
2011-02-28 13:55:36 -08:00
Ian Romanick
0a92301c80 i915: Set correct values for range/precision of fragment shader types
(cherry picked from commit 790ff232e2)
2011-02-28 13:55:36 -08:00
Ian Romanick
3274681f80 i965: Set correct values for range/precision of fragment shader types
(cherry picked from commit 3d028024e5)
2011-02-28 13:55:36 -08:00
Ian Romanick
96917f1f0c mesa: Connect glGetShaderPrecisionFormat into the dispatch table
(cherry picked from commit dde3270c19)
2011-02-28 13:55:36 -08:00
Brian Paul
1d35ffc541 mesa: implement glGetShaderPrecisionFormat()
Drivers should override the default range/precision info as needed.
No drivers do this yet.
(cherry picked from commit 3ee60a3558)
2011-02-28 13:55:35 -08:00
Chia-I Wu
15e6d05650 mesa: Add glDepthRangef and glClearDepthf to APIspec.xml.
Core mesa has gained support for GL_ARB_ES2_compatibility.  Make GLES
generated dispatch table use them.
(cherry picked from commit a4a5a9a5ce)
2011-02-28 13:55:35 -08:00
Eric Anholt
d3c1fb7775 mesa: Add getter for GL_SHADER_COMPILER with ARB_ES2_compatibility.
Fixes piglit arb_es2_compatibility-shadercompiler
(cherry picked from commit 4620de7eea)
2011-02-28 13:55:35 -08:00
Eric Anholt
6428ca32c0 mesa: Add getters for ARB_ES2_compatibility MAX_*_VECTORS.
Fixes piglit arb_es2_compatibility-maxvectors.
(cherry picked from commit 8395f206a8)
2011-02-28 13:55:35 -08:00
Eric Anholt
281d3fe3c0 mesa: Add support for glDepthRangef and glClearDepthf.
These are ARB_ES2_compatibility float variants of the core double
entrypoints.  Fixes arb_es2_compatibility-depthrangef.
(cherry picked from commit e12c4faf7e)
2011-02-28 13:55:35 -08:00
Eric Anholt
88f24e2598 mesa: Add actual support for glReleaseShaderCompiler from ES2.
Fixes no-op dispatch warning in piglit
arb_es2_compatibility-releaseshadercompiler.c.
(cherry picked from commit 7b987578a9)
2011-02-28 13:55:35 -08:00
Eric Anholt
7992b59087 mesa: Add extension enable bit for GL_ARB_ES2_compatibility.
(cherry picked from commit 9c6954fc9d)
2011-02-28 13:55:34 -08:00
Ian Romanick
ac06d610fb glapi: Regenerate for GL_ARB_ES2_compatibility.
This is not a cherry pick, but it matches 841ad6bf.
2011-02-28 13:55:34 -08:00
Eric Anholt
9d1b17059d glapi: Add entrypoints and enums for GL_ARB_ES2_compatibility.
(cherry picked from commit 8560cb939b)
2011-02-28 13:24:08 -08:00
Chad Versace
525c5458f5 i915: Disable extension OES_standard_derivatives
OES_standard_derivatives must be manually disabled for i915 because Mesa
enables it by default.
(cherry picked from commit 7b9dc40b0d)
2011-02-25 17:24:34 -08:00
Chad Versace
3a4ab56f32 mesa: Change OES_standard_derivatives to be stand-alone extension
Add a bit in struct gl_extensions for OES_standard_derivatives, and enable
the bit by default. Advertise the extension only if the bit is enabled.

Previously, OES_standard_derivatives was advertised in GLES2 contexts
if ARB_framebuffer_object was enabled.
(cherry picked from commit 78838b2d1b)
2011-02-25 17:24:34 -08:00
Vinson Lee
654ee9f282 mesa: Move loop variable declarations outside for loop in extensions.c.
Fixes MSVC build.
(cherry picked from commit 31b1051663)
2011-02-25 17:24:34 -08:00
Vinson Lee
f012e8832b mesa: Move declaration before code in extensions.c.
Fixes SCons build.
(cherry picked from commit 356e2e962f)
2011-02-25 17:24:34 -08:00
Chad Versace
1328fbdefb mesa: Change OES_point_sprite to depend on ARB_point_sprite
The extension string in GLES1 contexts always advertised
GL_OES_point_sprite. Now advertisement depends on ARB_point_sprite being
enabled.

Reviewed-by: Ian Romanick <idr@freedesktop.org>
(cherry picked from commit a7b5664c05)
2011-02-25 17:24:34 -08:00
Chad Versace
d9d1b8dab0 mesa: Change dependencies of some OES extension strings
Change all OES extension strings that depend on ARB_framebuffer_object to
instead depend on EXT_framebuffer_object.

Reviewed-by: Ian Romanick <idr@freedesktop.org>
(cherry picked from commit 039150169e)
2011-02-25 17:24:34 -08:00
Chad Versace
21e44e947a mesa: Add/remove extensions in extension string
Add GL_OES_stencil8 to ES2.

Remove the following:
   GL_OES_compressed_paletted_texture : ES1
   GL_OES_depth32                     : ES1, ES2
   GL_OES_stencil1                    : ES1, ES2
   GL_OES_stencil4                    : ES1, ES2
Mesa advertised these extensions, but did not actually support them.

Reviewed-by: Ian Romanick <idr@freedesktop.org>
(cherry picked from commit 19418e921a)
2011-02-25 17:24:34 -08:00
Chad Versace
2d1b154f73 mesa: Refactor handling of extension strings
Place GL, GLES1, and GLES2 extensions in a unified extension table. This
allows one to enable, disable, and query the status of GLES1 and GLES2
extensions by name.

When tested on Intel Ironlake, this patch did not alter the extension
string [as given by glGetString(GL_EXTENSIONS)] for any API.

Reviewed-by: Ian Romanick <idr@freedesktop.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 9b260c377f)
2011-02-25 17:24:34 -08:00
Dave Airlie
76366ee06b radeon: avoid segfault on 3D textures.
This is a candidate for 7.9 and 7.10
2011-02-22 15:27:02 -05:00
Dave Airlie
d3d2be2de1 radeon: calculate complete texture state inside TFP function
(really not sure why I'm doing this).

This is a candidate for 7.9 and 7.10 branches.
2011-02-22 15:26:44 -05:00
Dave Airlie
8402caf752 radeon/r200: fix fbo-clearmipmap + gen-teximage
sw clears were being used and not getting the correct offsets in the span
code.

also not emitting correct offsets for CB draws to texture levels.

(I've no idea why I'm playing with r100).

This is a candidate for 7.9 and 7.10
2011-02-22 15:26:24 -05:00
Paulo Zanoni
f0fa040d94 dri_util: fail driCreateNewScreen if InitScreen is NULL
Without this, X doesn't start with UMS on r300g.

NOTE: This is a candidate for the 7.9 and 7.10 branches.

Signed-off-by: Paulo Zanoni <pzanoni@mandriva.com>
Signed-off-by: Brian Paul <brianp@vmware.com>
2011-02-22 15:20:20 -05:00
Fredrik Höglund
b847da213e st/mesa: fix a regression from cae2bb76
stObj->pt is null when a TFP texture is passed to st_finalize_texture,
and with the changes introduced in the above commit this resulted in a
new texture being created and the existing image being copied into it.

NOTE: This is a candidate for the 7.10 branch.

Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
2011-02-22 15:19:54 -05:00
Marek Olšák
e7d1b5489e st/dri: Track drawable context bindings
Needs to track this ourself since because we get into a race condition with
the dri_util.c code on make current when rendering to the front buffer.

This is what happens:
Old context is rendering to the front buffer.

App calls MakeCurrent with a new context. dri_util.c sets
drawable->driContextPriv to the new context and then calls the driver make
current. st/dri make current flushes the old context, which calls back into
st/dri via the flush frontbuffer hook. st/dri calls dri loader flush
frontbuffer, which calls invalidate buffer on the drawable into st/dri.

This is where things gets wrong. st/dri grabs the context from the dri
drawable (which now points to the new context) and calls invalidate
framebuffer to the new context which has not yet set the new drawable as its
framebuffers since we have not called make current yet, it asserts.
(cherry picked from commit 94ccc31ba4)

Conflicts:

	src/gallium/state_trackers/dri/common/dri_context.c
2011-02-20 17:03:43 +01:00
Brian Paul
6c7adb0822 docs: add link to 7.10.1 release notes 2011-02-21 18:05:56 -07:00
Brian Paul
917c44aa52 docs: update 7.9.2 release notes with Brian's cherry-picks 2011-02-21 18:05:39 -07:00
Brian Paul
49a190bb0e st/mesa: need to translate clear color according to surface's base format
When clearing a GL_LUMINANCE_ALPHA buffer, for example, we need to convert
the clear color (R,G,B,A) to (R,R,R,A).  We were doing this for texture border
colors but not renderbuffers.  Move the translation function to st_format.c
and share it.

This fixes the piglit fbo-clear-formats test.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit e2d108ec82)
2011-02-21 17:48:47 -07:00
Brian Paul
71eee987d9 st/mesa: fix the default case in st_format_datatype()
Part of the fix for piglit fbo-clear-formats

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit c966c6980c)
2011-02-21 17:38:59 -07:00
Brian Paul
0ed0670fa9 st/mesa: set renderbuffer _BaseFormat in a few places
NOTE: This is a candidate for the 7.9 and 7.10 branches
(cherry picked from commit 633c9fcf78)
2011-02-21 17:38:35 -07:00
Brian Paul
f3f0e30e8e st/mesa: fix incorrect glCopyPixels position on fallback path
If we hit the pipe_get/put_tile() path for setting up the glCopyPixels
texture we were passing the wrong x/y position to pipe_get_tile().
The x/y position was already accounted for in the pipe_get_transfer()
call so we were effectively reading from 2*readX, 2*readY.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit d1becefb05)

Conflicts:

	src/mesa/state_tracker/st_cb_drawpixels.c
2011-02-21 17:37:07 -07:00
Brian Paul
a835f586c6 cso: fix loop bound in cso_set_vertex_samplers()
Before we were looping to nr_samplers, which is the number of fragment
samplers, not vertex samplers.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit d087cfaabf)
2011-02-21 17:31:44 -07:00
Julien Cristau
d91ec5322a glx: fix length of GLXGetFBConfigsSGIX
The extra length is the size of the request *minus* the size of the
VendorPrivate header, not the addition.

NOTE: This is a candidate for the 7.9 and 7.10 branches

Signed-off-by: Julien Cristau <jcristau@debian.org>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit cbe9fc12a6)
2011-02-21 17:30:48 -07:00
Julien Cristau
739d099d43 glx: fix GLXChangeDrawableAttributesSGIX request
xGLXChangeDrawableAttributesSGIXReq follows the GLXVendorPrivate header
with a drawable, number of attributes, and list of (type, value)
attribute pairs.  Don't forget to put the number of attributes in there.
I don't think this can ever have worked.

NOTE: This is a candidate for the 7.9 and 7.10 branches

Signed-off-by: Julien Cristau <jcristau@debian.org>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit e27913f805)
2011-02-21 17:30:29 -07:00
Dimitry Andric
dd34903790 glapi: add @GOTPCREL relocation type
Fixes http://bugs.freedesktop.org/show_bug.cgi?id=33440
This replaces commit 731ec60da3

NOTE: This is a candidate for the 7.9 and 7.10 branches

Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit cfb9aae3ec)
2011-02-21 17:30:04 -07:00
Brian Paul
d15da60f3b softpipe: fix off-by-one error in setup_fragcoord_coeff()
If we invert Y, need to subtract one from the surface height.

Fixes https://bugs.freedesktop.org/show_bug.cgi?id=26795
for softpipe.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 40ac24e631)
2011-02-21 17:29:15 -07:00
Brian Paul
63733afc48 st/mesa: fix incorrect fragcoord.x translation
emit_adjusted_wpos() needs separate x,y translation values.  If we
invert Y, we don't want to effect X.

Part of the fix for http://bugs.freedesktop.org/show_bug.cgi?id=26795

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit bb56631f0c)
2011-02-21 17:28:57 -07:00
Dimitry Andric
2fa6aef594 glapi: adding @ char before type specifier in glapi_x86.S
Fixes http://bugs.freedesktop.org/show_bug.cgi?id=33433
NOTE: This is a candidate for the 7.9 and 7.10 branches.

Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 37bffe8d12)
2011-02-21 17:28:32 -07:00
Julien Cristau
24797bd375 glx: fix request lengths
We were sending too long requests for GLXChangeDrawableAttributes,
GLXGetDrawableAttributes, GLXDestroyPixmap and GLXDestroyWindow.

NOTE: This is a candidate for the 7.9 and 7.10 branches

Signed-off-by: Julien Cristau <jcristau@debian.org>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 4324d6fdfb)
2011-02-21 17:26:15 -07:00
Dimitry Andric
0ec3ec8086 mesa: s/movzxw/movzwl/ in read_rgba_span_x86.S
Fixes http://bugs.freedesktop.org/show_bug.cgi?id=33386
NOTE: This is a candidate for the 7.9 and 7.10 branches

Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 811ee32a9e)
2011-02-21 17:25:13 -07:00
Dimitry Andric
5a4be4455e mesa: s/movzx/movzbl/
Fixes http://bugs.freedesktop.org/show_bug.cgi?id=33388
NOTE: This is a candidate for the 7.9 and 7.10 branches.

Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 3fda80246f)
2011-02-21 17:24:02 -07:00
Brian Paul
cdcad7bb31 llvmpipe: make sure binning is active when we begin/end a query
This fixes a potential failure when a begin/end_query is the first
thing to happen after flushing the scene.

NOTE: This is a candidate for the 7.10 and 7.9 branches.
(cherry picked from commit 42dbc2530b)
2011-02-21 17:23:09 -07:00
Brian Paul
a23311e5c7 mesa: check for dummy renderbuffer in _mesa_FramebufferRenderbufferEXT()
Fixes a failed assertion when a renderbuffer ID that was gen'd but not
previously bound was passed to glFramebufferRenderbuffer().  Generate
the same error that NVIDIA does.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit dd973cd9e8)
2011-02-21 17:22:03 -07:00
Brian Paul
89fb9a94bb mesa: don't assert in GetIntegerIndexed, etc
We were getting an assertion upon invalid pname.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 67722ae403)
2011-02-21 17:21:30 -07:00
Brian Paul
7739b6b54c mesa: fix num_draw_buffers==0 in fixed-function fragment program generation
This fixes a problem when glDrawBuffers(GL_NONE).  The fragment program
was writing to color output[0] but OutputsWritten was 0.  That led to a
failed assertion in the Mesa->TGSI translation code.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 2fa6012f6a)
2011-02-21 17:21:02 -07:00
Brian Paul
9d54f6bf83 mesa: fix a few format table mistakes, assertions
The BaseFormat field was incorrect for a few R and RG formats.
Fix a couple assertions too.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 0073f50cd4)
2011-02-21 17:19:56 -07:00
Brian Paul
1a6154e022 gallivm: fix copy&paste error from previous commit
Fixes piglit regression, http://bugs.freedesktop.org/show_bug.cgi?id=32452

NOTE: This is a candidate for the 7.10 branch
(cherry picked from commit 3ecf47af12)
2011-02-21 17:18:41 -07:00
Brian Paul
5b844eff0a gallivm: work around LLVM 2.6 bug when calling C functions
Create a constant int pointer to the C function, then cast it to the
function's type.  This avoids using trampoline code which seem to be
inadvertantly freed by LLVM in some situations (which leads to segfaults).
The root issue and work-around were found by José.

NOTE: This is a candidate for the 7.10 branch
(cherry picked from commit ee16e97ed1)
2011-02-21 17:18:41 -07:00
Ian Romanick
fef0bf65a7 docs: Update 7.10.1 release notes
Add recent cherry picks
2011-02-21 13:35:59 -08:00
Ian Romanick
5ad00ef7a3 Use C-style system headers in C++ code to avoid issues with std:: namespace
Based on commit 497baf4e4a from master.
2011-02-21 13:19:34 -08:00
Ian Romanick
60675572f7 mesa: Fix error checks in GetVertexAttrib functions
Querying index zero is not an error in OpenGL ES 2.0.

Querying an index larger than the value returned by
GL_MAX_VERTEX_ATTRIBS is an error in all APIs.

Fixes bugzilla #32375.
(cherry picked from commit 5c3f1cdbbe)
2011-02-21 13:19:33 -08:00
Ian Romanick
6b7b2af43a linker: Generate link errors when ES shaders are missing stages
ES requires that a vertex shader and a fragment shader be present.

Fixes bugzilla #32214.
(cherry picked from commit ce9171f9d8)
2011-02-21 13:19:33 -08:00
Ian Romanick
9a9bd548b0 mesa: glGetUniform only returns a single element of an array
Also return it as the correct type.  Previously the whole array would
be returned and each element would be expanded to a vec4.

Fixes piglit test getuniform-01 and bugzilla #29823.
(cherry picked from commit 20d278a7ff)
2011-02-21 13:19:33 -08:00
Marek Olšák
d5a1325f81 mesa: fix texture3D mipmap generation for UNSIGNED_BYTE_3_3_2 2011-02-16 20:49:52 +01:00
Christoph Bumiller
0555e04aaa nv50,nvc0: do not forget to apply sign mode to saved TGSI inputs
fixes 34179.

Reported-by: Sense Hofstede
Requested-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
2011-02-16 10:33:00 +10:00
Tom Stellard
cc1636b6db r300/compiler: Don't erase sources when converting RGB->Alpha
https://bugs.freedesktop.org/show_bug.cgi?id=34030

(cherry picked from commit 9106b98766)
2011-02-11 20:16:25 -08:00
Bryce Harrington
995edd4c0a r300g: Null pointer check for buffer deref in gallium winsys
radeon_drm_bufmgr_create_buffer_from_handle() can return NULL buffers
sometimes (seen when alt-tabbing in compiz).  Avoid dereferencing the
buffer pointer in this case.

Ref.: https://bugs.launchpad.net/ubuntu/+source/mesa/+bug/691653
Also: https://bugzilla.redhat.com/show_bug.cgi?id=660143

Signed-off-by: Bryce Harrington <bryce@canonical.com>
2011-02-11 02:33:36 +01:00
Vinson Lee
546aade286 ralloc: Add missing va_end following va_copy.
(cherry picked from commit cde443e0b9)
2011-02-07 15:02:53 -08:00
Kenneth Graunke
77e3c82ad1 Remove the talloc sources from the Mesa repository.
(cherry picked from commit 1568b19e3b)
2011-02-07 15:02:39 -08:00
Kenneth Graunke
5c1e361b8e Remove talloc from the SCons build system.
(cherry picked from commit 8aac5d123c)

Conflicts:
	src/SConscript
	src/gallium/targets/egl-static/SConscript
	src/glsl/SConscript
	src/mesa/SConscript
2011-02-07 15:02:39 -08:00
Kenneth Graunke
72f90dc3ee Remove talloc from the make and automake build systems.
(cherry picked from commit d1d8120545)

Conflicts:
	src/glsl/Makefile
	src/mesa/Makefile
	src/mesa/drivers/osmesa/Makefile
2011-02-07 15:02:39 -08:00
Kenneth Graunke
6fb448c314 ralloc: a new MIT-licensed recursive memory allocator.
(cherry picked from commit 42fd9c2ebb)
2011-02-07 15:02:39 -08:00
Kenneth Graunke
2e226777a4 Convert everything from the talloc API to the ralloc API.
(cherry-picked from d3073f58c1 and
 cfd8d45ccd then squashed)

Conflicts:
	src/glsl/builtin_function.cpp
	src/glsl/glcpp/glcpp-parse.c
	src/glsl/ir_reader.cpp
	src/glsl/loop_analysis.cpp
	src/glsl/main.cpp
	src/mesa/drivers/dri/i965/brw_fs_schedule_instructions.cpp
	src/mesa/program/ir_to_mesa.cpp
	src/mesa/program/register_allocate.c
2011-02-07 15:02:39 -08:00
Kenneth Graunke
4e5b184a61 ralloc: Add a fake implementation of ralloc based on talloc.
(cherry picked from commit dc55254f5b)
2011-02-07 15:02:39 -08:00
Kenneth Graunke
511e5e30a5 glcpp: Remove use of talloc reference counting.
We almost always want to simply steal; we only need to copy when copying
a token list (in which case we're already cloning stuff anyway).

(cherry picked from commit 6ecee54a9a)
2011-02-07 15:02:39 -08:00
Kenneth Graunke
281e3bee4b glsl, i965: Remove unnecessary talloc includes.
These are already picked up by ir.h or glsl_types.h.

(cherry picked from commit e256e4743c)
2011-02-07 15:02:39 -08:00
Kenneth Graunke
ea4df94d61 glsl: Don't bother unsetting a destructor that was never set.
This was totally copied and pasted from glsl_symbol_table.

(cherry picked from commit 21031b4e88)
2011-02-07 15:02:39 -08:00
Tom Stellard
92a619b43f r300/compiler: Disable register rename pass on r500
The scheduler and the register allocator are not good enough yet to deal
with the effects of the register rename pass.  This was causing a 50%
performance drop in Lightsmark.  The pass can be re-enabled once the
scheduler and the register allocator are more mature.  r300 and r400
still need this pass, because it prevents a lot of shaders from using
too many texture indirections.

(cherry picked from commit 68b701f5de)
2011-02-05 23:30:08 -08:00
Ian Romanick
b54faf45dc docs: Update 7.10.1 release notes
Add recent cherry picks for precision qualifers, linker bugs, and other issues.
2011-02-04 16:51:17 -08:00
Ian Romanick
3370f9b606 linker: Propagate max_array_access while linking functions
Update the max_array_access of a global as functions that use that
global are pulled into the linked shader.

Fixes piglit test glsl-fs-implicit-array-size-01 and bugzilla #33219.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 0f4b2a0a23)
2011-02-04 15:14:01 -08:00
Ian Romanick
bdcdcb5c18 linker: Set sizes for non-global arrays as well
Previously only global arrays with implicit sizes would be patched.
This causes all arrays that are actually accessed to be sized.

Fixes piglit test glsl-fs-implicit-array-size-02.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit c87e9ef4d2)
2011-02-04 15:14:01 -08:00
Ian Romanick
b51b8db450 glsl: Don't assert when the value returned by a function has no rvalue
The rvalue of the returned value can be NULL if the shader says
'return foo();' and foo() is a function that returns void.

Existing GLSL specs do *NOT* say that this is an error.  The type of
the return value is void.  If the return type of the function is also
void, then this should compile without error.  I expect that future
versions of the GLSL spec will fix this (wink, wink, nudge, nudge).

Fixes piglit test glsl-1.10/compiler/expressions/return-01.vert and
bugzilla #33308.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 2db46fe5f0)
2011-02-04 15:14:01 -08:00
Kenneth Graunke
32786b8a33 glsl: Expose a public glsl_type::void_type const pointer.
This is analogous to glsl_type::int_type and all the others.
(cherry picked from commit 5c229e5fbd)
2011-02-04 15:14:01 -08:00
Chad Versace
7bb3fe50c2 glsl: Mark 'in' variables at global scope as read-only
Fixes Piglit tests:
spec/glsl-1.30/compiler/storage-qualifiers/static-write-centroid-in-01.frag
spec/glsl-1.30/compiler/storage-qualifiers/static-write-in-01.frag
spec/glsl-1.30/compiler/storage-qualifiers/static-write-in-02.frag
(cherry picked from commit 01a584d093)
2011-01-31 16:19:50 -08:00
Chad Versace
c4b626018a glsl: Fix segfault due to missing printf argument
Fixes the following Piglit tests:
glslparsertest/shaders/array2.frag
glslparsertest/shaders/dataType6.frag

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 46f7105df4)
2011-01-31 16:19:50 -08:00
Chad Versace
a9eec42b0e glsl: Fix semantic checks on precision qualifiers
The check for
   Precision qualifiers only apply to floating point and integer types.
was incomplete. It rejected only type 'bool' and structures.
(cherry picked from commit 45e8e6c6b1)
2011-01-31 16:19:50 -08:00
Ian Romanick
0ee9a6698d glsl: Refresh autogenerated lexer and parser files.
For the previous few commits.
2011-01-31 16:19:50 -08:00
Chad Versace
10f6e286d5 glsl: Remove redundant semantic check in parser
The removed semantic check also exists in ast_type_specifier::hir(), which
is a more natural location for it.

The check verified that precision statements are applied only to types
float and int.
(cherry picked from commit a9bf8c12ee)
2011-01-31 16:08:10 -08:00
Chad Versace
b2b1b1f596 glsl: Add support for default precision statements
* Add new field ast_type_specifier::is_precision_statement.
* Add semantic checks in ast_type_specifier::hir().
* Alter parser rules accordingly.
(cherry picked from commit 08a286c9cc)
2011-01-31 16:07:33 -08:00
Chad Versace
757a49cafd glsl: Add semantic checks for precision qualifiers
* Check that precision qualifiers only appear in language versions 1.00,
  1.30, and later.
* Check that precision qualifiers do not apply to bools and structs.

Fixes the following Piglit tests:
* spec/glsl-1.30/precision-qualifiers/precision-bool-01.frag
* spec/glsl-1.30/precision-qualifiers/precision-struct-01.frag
* spec/glsl-1.30/precision-qualifiers/precision-struct-02.frag

(cherry picked from commit 889e1a5b6c)

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
2011-01-31 16:06:42 -08:00
Chad Versace
9016ab24b8 glsl: Change default value of ast_type_specifier::precision
Change default value to ast_precision_none, which denotes the absence of
a precision of a qualifier.

Previously, the default value was ast_precision_high. This made it
impossible to detect if a precision qualifier was present or not.
(cherry picked from commit aaa31bf8f4)
2011-01-31 16:03:32 -08:00
Chad Versace
c597334ef2 glsl: Fix parser rule for type_specifier
Do not assign a value to ast_type_specifier::precision when no precision
qualifier is present.
(cherry picked from commit 33279cd2d3)
2011-01-31 16:03:06 -08:00
Tom Fogal
84b857ef73 Regenerate gl_mangle.h. 2011-01-27 15:09:12 -07:00
Ian Romanick
a80384d7f6 docs: Update 7.10.1 release notes
Replace "Fix an error in uniform arrays in row calculating" with the
actual bugzilla that was fixed.

Put the entry for bug #30156 in the correct order.
2011-01-26 10:16:36 -08:00
Ian Romanick
092b6f2ca8 glsl: Emit errors or warnings when 'layout' is used with 'attribute' or 'varying'
The specs that add 'layout' require the use of 'in' or 'out'.
However, a number of implementations, including Mesa, shipped several
of these extensions allowing the use of 'varying' and 'attribute'.
For these extensions only a warning is emitted.

This differs from the behavior of Mesa 7.10.  Mesa 7.10 would only
accept 'attribute' with 'layout(location)'.  This behavior was clearly
wrong.  Rather than carrying the broken behavior forward, we're just
doing the correct thing.

This is related to (piglit) bugzilla #31804.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 4bcff0c190)
2011-01-25 16:17:51 -08:00
Ian Romanick
c8e504eec5 doc: Update 7.10.1 release notes 2011-01-25 15:52:56 -08:00
Jian Zhao
acc7369285 mesa: fix an error in uniform arrays in row calculating.
Fix the error in uniform row calculating, it may alloc one line
more which may cause out of range on memory usage, sometimes program
aborted when free the memory.

NOTE: This is a candidate for 7.9 and 7.10 branches.

Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 2a7380e9c3)
2011-01-25 15:52:56 -08:00
Eric Anholt
2d22c508c7 docs: Add a relnote for the Civ IV on i965. 2011-01-18 12:06:54 -08:00
Eric Anholt
aba5e843dd i965: Fix dead pointers to fp->Parameters->ParameterValues[] after realloc.
Fixes texrect-many regression with ff_fragment_shader -- as we added
refs to the subsequent texcoord scaling paramters, the array got
realloced to a new address while our params[] still pointed at the old
location.
(cherry picked from commit e4be665bbd)
2011-01-18 11:34:54 -08:00
Eric Anholt
4e0d6cf7ba intel: Make renderbuffer tiling choice match texture tiling choice.
There really shouldn't be any difference between the two for us.
Fixes a bug where Z16 renderbuffers would be untiled on gen6, likely
leading to hangs.
(cherry picked from commit 29c4f95cbc)
2011-01-18 11:34:54 -08:00
Eric Anholt
2afa3f47af i965: Avoid double-negation of immediate values in the VS.
In general, we have to negate in immediate values we pass in because
the src1 negate field in the register description is in the bits3 slot
that the 32-bit value is loaded into, so it's ignored by the hardware.
However, the src0 negate field is in bits1, so after we'd negated the
immediate value loaded in, it would also get negated through the
register description.  This broke this VP instruction in the position
calculation in civ4:

MAD TEMP[1], TEMP[1], CONST[256].zzzz, CONST[256].-y-y-y-y;

Bug #30156
(cherry picked from commit 1d1ad6306d)
2011-01-18 11:34:54 -08:00
Eric Anholt
219b8b672d i965/fs: Do flat shading when appropriate.
We were trying to interpolate, which would end up doing unnecessary
math, and doing so on undefined values.   Fixes glsl-fs-flat-color.
(cherry picked from commit c3f000b392)
2011-01-18 11:34:54 -08:00
Eric Anholt
b90223f4cf i965/vs: When MOVing to produce ABS, strip negate of the operand.
We were returning the negative absolute value, instead of the absolute
value.  Fixes glsl-vs-abs-neg.
(cherry picked from commit 9351ef7a44)
2011-01-18 11:34:54 -08:00
Eric Anholt
e15ad414d0 i965/fs: When producing ir_unop_abs of an operand, strip negate.
We were returning the negative absolute value, instead of the absolute
value.  Fixes glsl-fs-abs-neg.
(cherry picked from commit ab56e3be9a)
2011-01-18 11:34:54 -08:00
Eric Anholt
a702858139 glsl: Fix the lowering of variable array indexing to not lose write_masks.
Fixes glsl-complex-subscript on 965.
(cherry picked from commit c00bc13564)
2011-01-18 11:34:53 -08:00
Ian Romanick
8f3eef1206 mesa: bump version to 7.10.1-devel 2011-01-17 16:28:02 -08:00
Ian Romanick
0a0b0c8f7e docs: Initial bits of 7.10.1 release notes 2011-01-17 16:27:58 -08:00
Ian Romanick
05ff61dc6b glsl: Allow 'in' and 'out' when 'layout' is also available
All of the extensions that add the 'layout' keyword also enable (and
required) the use of 'in' and 'out' with shader globals.

This is related to (piglit) bugzilla #31804.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 82c4b4f88a)
2011-01-17 16:24:45 -08:00
Ian Romanick
aff4170849 glsl: Track variable usage, use that to enforce semantics
In particular, variables cannot be redeclared invariant after being
used.

Fixes piglit test invariant-05.vert and bugzilla #29164.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit bd33055ef4)
2011-01-17 16:24:45 -08:00
Ian Romanick
d81615ee19 glsl: Disallow 'in' and 'out' on globals in GLSL 1.20
Fixes piglit tests glsl-1.20/compiler/qualifiers/in-01.vert and
glsl-1.20/compiler/qualifiers/out-01.vert and bugzilla #32910.

NOTE: This is a candidate for the 7.9 and 7.10 branches.  This patch
also depends on the previous two commits.
(cherry picked from commit 469ea695bb)
2011-01-17 16:24:45 -08:00
Ian Romanick
4ee68e2d47 glsl & glcpp: Refresh autogenerated lexer and parser files.
For the previous few commits.
2011-01-17 16:24:44 -08:00
Ian Romanick
27ef465276 glsl: Add version_string containing properly formatted GLSL version
(cherry picked from commit eebdfdfbcf)
2011-01-17 14:50:01 -08:00
Ian Romanick
50d40edb8c glcpp: Generate an error for division by zero
When GCC encounters a division by zero in a preprocessor directive, it
generates an error.  Since the GLSL spec says that the GLSL
preprocessor behaves like the C preprocessor, we should generate that
same error.

It's worth noting that I cannot find any text in the C99 spec that
says this should be an error.  The only text that I can find is line 5
on page 82 (section 6.5.5 Multiplicative Opertors), which says,

    "The result of the / operator is the quotient from the division of
    the first operand by the second; the result of the % operator is
    the remainder. In both operations, if the value of the second
    operand is zero, the behavior is undefined."

Fixes 093-divide-by-zero.c test and bugzilla #32831.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 9ca5300b6e)
2011-01-17 14:49:53 -08:00
Chad Versace
c2b721bc0e glcpp: Fix segfault when validating macro redefinitions
In _token_list_equal_ignoring_space(token_list_t*, token_list_t*), add
a guard that prevents dereferncing a null token list.

This fixes test src/glsl/glcpp/tests/092-redefine-macro-error-2.c and
Bugzilla #32695.
(cherry picked from commit 4fff52f1c9)
2011-01-17 14:49:48 -08:00
Chad Versace
45be27d09b glsl: At link-time, check that globals have matching centroid qualifiers
Fixes bug 31923: http://bugs.freedesktop.org/show_bug.cgi?id=31923
(cherry picked from commit 61428dd2ab)
2011-01-17 14:49:39 -08:00
Ian Romanick
6fded6d29d glsl: Support the 'invariant(all)' pragma
Previously the 'STDGL invariant(all)' pragma added in GLSL 1.20 was
simply ignored by the compiler.  This adds support for setting all
variable invariant.

In GLSL 1.10 and GLSL ES 1.00 the pragma is ignored, per the specs,
but a warning is generated.

Fixes piglit test glsl-invariant-pragma and bugzilla #31925.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 86b4398cd1)
2011-01-17 14:49:30 -08:00
Brian Paul
ea96167f2f docs: added news item for 7.9.1 and 7.10 release 2011-01-14 17:35:27 -07:00
Brian Paul
b8062cdf4b docs: add links to 7.9.1 and 7.10 release notes 2011-01-14 17:35:08 -07:00
Brian Paul
0e809808b1 draw: Fix an off-by-one bug in a vsplit assertion.
When use_spoken is true, istart (the first vertex of this segment) is
replaced by i0 (the spoken vertex of the fan).  There are still icount
vertices.

Thanks to Brian Paul for spotting this.
(cherry picked from commit abbb1c8f08)
2011-01-13 11:35:34 -07:00
Alberto Milone
b28a90c9a7 r600c: add evergreen ARL support.
Signed-off-by: Alberto Milone <alberto.milone@canonical.com>
2011-01-11 14:50:10 -05:00
Marek Olšák
38c3d8a828 docs: fix messed up names with special characters in relnotes-7.9.1
(cherry picked from commit 67aeab0b77)
2011-01-08 03:07:25 +01:00
Marek Olšák
36009724fd docs: fix messed up names with special characters in relnotes-7.10 2011-01-08 03:06:04 +01:00
Ian Romanick
50a82a8601 docs: Add 7.10 md5sums 2011-01-07 14:20:27 -08:00
Ian Romanick
7a3f869a47 mesa: set version string to 7.10 2011-01-07 14:09:03 -08:00
Ian Romanick
c18447bf97 docs: Update 7.10 release notes 2011-01-07 14:07:51 -08:00
Ian Romanick
5768445eaf docs: Import 7.9.1 release notes from 7.9 branch
(cherry picked from commit 46a360b26a)
2011-01-07 13:41:15 -08:00
Alex Deucher
fbd98eae6a r600c: fix up SQ setup in blit code for Ontario/NI 2011-01-07 03:12:28 -05:00
Marek Olšák
14950c50e1 r300/compiler: disable the rename_regs pass for loops
This workaround fixes rendering of kwin thumbnails.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 8543902bfb)
2011-01-07 07:10:49 +01:00
Alex Deucher
ca8e49f303 r600g: support up to 64 shader constants
From the r600 ISA:
Each ALU clause can lock up to four sets of constants
into the constant cache.  Each set (one cache line) is
16 128-bit constants. These are split into two groups.
Each group can be from a different constant buffer
(out of 16 buffers). Each group of two constants consists
of either [Line] and [Line+1] or [line + loop_ctr]
and [line + loop_ctr +1].

For supporting more than 64 constants, we need to
break the code into multiple ALU clauses based
on what sets of constants are needed in that clause.

Note: This is a candidate for the 7.10 branch.

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
2011-01-06 19:51:14 -05:00
Alex Deucher
f2f89f7b18 r600c: add support for NI asics 2011-01-06 18:41:23 -05:00
Alex Deucher
3285d1dc57 r600g: add support for NI (northern islands) asics
This adds support for barts, turks, and caicos.
2011-01-06 18:17:18 -05:00
Dave Airlie
9ba827100a r600g: hack around property unknown issues.
should fix https://bugs.freedesktop.org/show_bug.cgi?id=32619

Need to add proper support for properties later.

Signed-off-by: Dave Airlie <airlied@redhat.com>
2011-01-06 15:54:47 -05:00
Alex Deucher
002ce07abe r600g: remove useless switch statements
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
2011-01-06 15:54:47 -05:00
Dave Airlie
949c24862a r600g: fix evergreen segfaults.
evergreen was crashing running even gears here.

This is a 7.10 candidate if its broken the same.

Signed-off-by: Dave Airlie <airlied@redhat.com>
2011-01-06 15:54:47 -05:00
Jerome Glisse
34c58f6d46 r600g: avoid segfault
Candidates 7.10

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2011-01-06 15:54:47 -05:00
Jerome Glisse
ece71d605b r600g: properly unset vertex buffer
Fix bug http://bugs.freedesktop.org/show_bug.cgi?id=32455

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2011-01-06 15:54:47 -05:00
Jerome Glisse
0fc205152c r600g: need to reference upload buffer as the might still live accross flush
Can't get away from referencing upload buffer as after flush a vertex buffer
using the upload buffer might still be active. Likely need to simplify the
pipe_refence a bit so we don't waste so much cpu time in it.

candidates for 7.10 branch

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2011-01-06 15:54:47 -05:00
Jerome Glisse
4434614844 r600g: fix segfault when translating vertex buffer
Note the support for non float vertex draw likely regressed need to
find what we want to do there.

candidates for 7.10 branches

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2011-01-06 15:54:47 -05:00
Jerome Glisse
e7b12f2a0e r600g: fix bo size when creating bo from handle
Spoted by Alex Diomin

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2011-01-06 15:54:47 -05:00
Ian Romanick
b11623a5df glsl: Allow less restrictive uses of sampler array indexing in GLSL <= 1.20
GLSL 1.10 and 1.20 allow any sort of sampler array indexing.
Restrictions were added in GLSL 1.30.  Commit f0f2ec4d added support
for the 1.30 restrictions, but it broke some valid 1.10/1.20 shaders.
This changes the error to a warning in GLSL 1.10, GLSL 1.20, and GLSL
ES 1.00.

There are some spurious whitespace changes in this commit.  I changed
the layout (and wording) of the error message so that all three cases
would be similar.  The 1.10/1.20 and 1.30 text is the same.  The only
difference is that one is an error, and the other is a warning.  The
GLSL ES 1.00 wording is similar but not quite the same.

Fixes piglit test
spec/glsl-1.10/compiler/constant-expressions/sampler-array-index-02.frag
and bugzilla #32374.
2011-01-06 10:07:50 -08:00
Zou Nan hai
bbf7cc1f2a i965: skip too small size mipmap
this fixes doom3 crash.
2011-01-06 11:42:38 +08:00
Xiang, Haihao
4e8f123f14 i965: use BLT to clear buffer if possible on Sandybridge
This fixes https://bugs.freedesktop.org/show_bug.cgi?id=32713
(cherry picked from commit 266d8eed69)
2011-01-05 14:12:59 +08:00
Chia-I Wu
f5c1346932 autoconf: Fix --with-driver=xlib --enable-openvg.
st/egl should be enabled with --enable-openvg even the driver is xlib or
osmesa.  Also, GLX_DIRECT_RENDERING should not be defined because libdrm
is not checked.
(cherry picked from commit ada9c78c29)
2011-01-05 11:37:17 +08:00
Chia-I Wu
ebcb7f27e5 docs: Add an example for EGL_DRIVERS_PATH.
EGL_DRIVERS_PATH can be set to test EGL without installation.
(cherry picked from commit cba7786954)
2011-01-05 11:37:12 +08:00
Eric Anholt
52586ceb2b intel: When validating an FBO's combined depth/stencil, use the given FBO.
We were looking at the current draw buffer instead to see whether the
depth/stencil combination matched.  So you'd get told your framebuffer
was complete, until you bound it and went to draw and we decided that
it was incomplete.
(cherry picked from commit b7b2791c6b)
2011-01-04 13:01:42 -08:00
Eric Anholt
438fc337d4 intel: Fix segfaults from trying to use _ColorDrawBuffers in FBO validation.
The _ColorDrawBuffers is a piece of computed state that gets for the
current draw/read buffers at _mesa_update_state time.  However, this
function actually gets used for non-current draw/read buffers when
checking if an FBO is complete from the driver's perspective.  So,
instead of trying to just look at the attachment points that are
currently referenced by glDrawBuffers, look at all attachment points
to see if they're driver-supported formats.  This appears to actually
be more in line with the intent of the spec, too.

Fixes a segfault in my upcoming fbo-clear-formats piglit test, and
hopefully bug #30278
(cherry picked from commit 0ea49380e2)
2011-01-04 13:01:42 -08:00
Eric Anholt
7b6c5804f0 intel: Add a couple of helper functions to reduce rb code duplication.
(cherry picked from commit e339b669a1)
2011-01-04 13:01:42 -08:00
Eric Anholt
29bcf0a940 intel: Add spans code for the ARB_texture_rg support.
This starts spantmp2.h down the path of using MESA_FORMAT_* for
specifying the format instead of the crazy GL format/type combo.
(cherry picked from commit 28bab24e16)
2011-01-04 13:01:42 -08:00
Eric Anholt
fa61cb3609 intel: Use tri clears when we don't know how to blit clear the format.
Bug #32207.  Fixes ARB_texture_rg/fbo-clear-formats (see my
fbo-clear-formats piglit branch currently)
(cherry picked from commit 30fef21aa3)
2011-01-04 13:01:42 -08:00
Eric Anholt
db4e1c44b2 intel: Handle forced swrast clears before other clear bits.
Fixes a potential segfault on a non-native depthbuffer, and possible
accidental swrast fallback on extra color buffers.
(cherry picked from commit 94ed481131)
2011-01-04 13:01:41 -08:00
Eric Anholt
d4ae5f3411 intel: Only do frame throttling at glFlush time when using frontbuffer.
This is the hack for input interactivity of frontbuffer rendering
(like we do for backbuffer at intelDRI2Flush()) by waiting for the n-2
frame to complete before starting a new one.  However, for an
application doing multiple contexts or regular rebinding of a single
context, this would end up lockstepping the CPU to the GPU because
every unbind was considered the end of a frame.

Improves WOW performance on my Ironlake by 48.8% (+/- 2.3%, n=5)
(cherry picked from commit b01b73c482)
2011-01-04 13:01:41 -08:00
Zhenyu Wang
1feecbdb00 i965: Fix provoking vertex select in clip state for sandybridge
Triangle fan provoking vertex for first convention should be
'vertex 1' in sandybridge clip state.

Partly fix glean/clipFlat case
(cherry picked from commit 9977297ad9)
2011-01-04 13:01:41 -08:00
Zhenyu Wang
8847205976 i965: Use last vertex convention for quad provoking vertex on sandybridge
Until we know how hw converts quads to polygon in beginning of
3D pipeline, for now unconditionally use last vertex convention.

Fix glean/clipFlat case.
(cherry picked from commit bea6539abf)
2011-01-04 13:01:41 -08:00
Eric Anholt
8604d91ae4 i965: Do lowering of array indexing of a vector in the FS.
Fixes a regression in ember since switching to the native FS backend,
and the new piglit tests glsl-fs-vec4-indexing-{2,3} for catching this.
(cherry picked from commit df4d83dca4)
2011-01-04 13:01:41 -08:00
Eric Anholt
bba89b3793 i965: Fix regression in FS comparisons on original gen4 due to gen6 changes.
Fixes 26 piglit cases on my GM965.
(cherry picked from commit 54df8e48bc)
2011-01-04 13:01:41 -08:00
Eric Anholt
f4f3274ba3 i965: Factor out the ir comparision to BRW_CONDITIONAL_* code.
(cherry picked from commit 74dffb39c3)
2011-01-04 13:01:41 -08:00
Eric Anholt
7b1200901a i965: Improve the hacks for ARB_fp scalar^scalar POW on gen6.
This is still awful, but my ability to care about reworking the old
backend so we can just get a temporary value into a POW is awfully low
since the new backend does this all sensibly.

Fixes:
fp1-LIT test 1
fp1-LIT test 3 (case x < 0)
fp1-POW test (exponentiation)
fp-lit-mask
(cherry picked from commit d88aa6fe3e)
2011-01-04 13:01:40 -08:00
Tom Stellard
9dfa27c924 r300/compiler: Fix black terrain in Civ4
rc_inst_can_use_presub() wasn't checking for too many RGB sources in
Alpha instructions or too many Alpha sources in RGB instructions.

(cherry picked from commit e96e86d07b)
2011-01-04 11:37:08 -08:00
Kenneth Graunke
b71bff0100 i965: Internally enable GL_NV_blend_square on ES2.
Hopefully should fix bug #32520.
(cherry picked from commit 6bb1e4541e)
2011-01-04 09:47:03 -08:00
Kenneth Graunke
8cfce0c643 i965: Flatten if-statements beyond depth 16 on pre-gen6.
Gen4 and Gen5 hardware can have a maximum supported nesting depth of 16.
Previously, shaders with control flow nested 17 levels deep would
cause a driver assertion or segmentation fault.

Gen6 (Sandybridge) hardware no longer has this restriction.

Fixes fd.o bug #31967.
(cherry picked from commit 634a7dce9c)
2011-01-04 09:46:55 -08:00
Kenneth Graunke
9d3573c905 glsl: Support if-flattening beyond a given maximum nesting depth.
This adds a new optional max_depth parameter (defaulting to 0) to
lower_if_to_cond_assign, and makes the pass only flatten if-statements
nested deeper than that.

By default, all if-statements will be flattened, just like before.

This patch also renames do_if_to_cond_assign to lower_if_to_cond_assign,
to match the new naming conventions.
(cherry picked from commit 9ac6a9b2fa)
2011-01-04 09:46:17 -08:00
Marek Olšák
8d2c910e66 mesa: fix texel store functions for some float formats
These are copy-paste errors obviously.
(cherry picked from commit bf7b6f60ae)
2011-01-04 09:44:11 -08:00
Ian Romanick
f0c2420917 Refresh autogenerated file builtin_function.cpp.
See also a954dbeb.
2011-01-04 09:44:11 -08:00
Kenneth Graunke
83b39afc46 glsl/builtins: Compute the correct value for smoothstep(vec, vec, vec).
These mistakenly computed 't' instead of t * t * (3.0 - 2.0 * t).

Also, properly vectorize the smoothstep(float, float, vec) variants.

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit d7423a6531)
2011-01-04 09:39:01 -08:00
Brian Paul
adb49457c6 st/mesa: fix renderbuffer pointer check in st_Clear()
Fixes http://bugs.freedesktop.org/show_bug.cgi?id=30694

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit efbd33aff9)
2011-01-04 08:00:33 -07:00
Kenneth Graunke
9da0e20e46 Remove OES_compressed_paletted_texture from the ES2 extension list.
We don't support it.
(cherry picked from commit d0f8eea9a0)
2011-01-03 14:26:35 -08:00
Brian Paul
07342c84a9 glsl: new glsl_strtod() wrapper to fix decimal point interpretation
We always want to use '.' as the decimal point.

See http://bugs.freedesktop.org/show_bug.cgi?id=24531

NOTE: this is a candidate for the 7.10 branch.
(cherry picked from commit bb10e081c8)
2011-01-03 14:26:17 -08:00
Ian Romanick
3501fd8594 ir_to_mesa: Don't generate swizzles for record derefs of non-scalar/vectors
This is the same as what the array dereference handler does.

Fixes piglit test glsl-link-struct-array (bugzilla #31648).

NOTE: This is a candidate for the 7.9 and 7.10 branches.
(cherry picked from commit 2d577ee730)
2011-01-03 14:26:08 -08:00
Ian Romanick
4febfee3b7 linker: Allow built-in arrays to have different sizes between shader stages
Fixes pitlit test glsl-link-varying-TexCoord (bugzilla #31650).
(cherry picked from commit cb2b547a47)
2011-01-03 14:25:54 -08:00
Ian Romanick
d3fa3c60f2 glsl: Inherrit type of declared variable from initializer after processing assignment
do_assignment may apply implicit conversions to coerce the base type
of initializer to the base type of the variable being declared.  Fixes
piglit test glsl-implicit-conversion-02 (bugzilla #32287).  This
probably also fixes bugzilla #32273.

NOTE: This is a candidate for the 7.9 branch and the 7.10 branch.
(cherry picked from commit d7f27e2e76)
2011-01-03 14:25:34 -08:00
Henri Verbeet
4ad4c700bf st/mesa: Handle wrapped depth buffers in st_copy_texsubimage().
(cherry picked from commit 59051ad443)
2010-12-31 07:50:56 +01:00
Fredrik Höglund
aa196d047c r600g: fix pow(0, 0) evaluating to NaN
We have to use the non-IEEE compliant version of MUL here, since
log2(0) is -inf, and 0 * -inf is NaN in IEEE arithmetic.

candidates for 7.10 branch
2010-12-29 11:04:32 -05:00
Alex Deucher
747279c21c r600g: fix rendering with a vertex attrib having a zero stride
The hardware supports zero stride just fine.  This is a port
of 2af8a19831 from r300g.

NOTE: This is a candidate for both the 7.9 and 7.10 branches.

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
2010-12-29 10:59:53 -05:00
richard
0092793735 r600c : inline vertex format is not updated in an app, switch to use vfetch constants. For the 7.9 and 7.10 branches as well. 2010-12-29 10:59:00 -05:00
Zhenyu Wang
96685a662f i965: Fix occlusion query on sandybridge
Clear target query buffer fixed occlusion query on sandybridge.

https://bugs.freedesktop.org/show_bug.cgi?id=32167
(cherry picked from commit 689aca7822)
2010-12-29 09:41:52 +08:00
Marek Olšák
7e3c1f221a r300g: mark vertex arrays as dirty after a buffer_offset change
We shouldn't hit this bug in theory.

NOTE: This is a candidate for the 7.10 branch.
(cherry picked from commit d9b84017e0)
2010-12-28 19:41:46 +01:00
Marek Olšák
1e58915062 r300g/swtcl: re-enable LLVM
Based on a patch from Drill <drill87@gmail.com>.

NOTE: This is a candidate for the 7.10 branch.
(cherry picked from commit 88550083b3)

Conflicts:

	src/gallium/drivers/r300/r300_context.c
2010-12-28 19:41:00 +01:00
Xiang, Haihao
1ca240ec2d i965: don't spawn GS thread for LINELOOP on Sandybridge
LINELOOP is converted to LINESTRIP at the beginning of the 3D pipeline.
This fixes https://bugs.freedesktop.org/show_bug.cgi?id=32596
(cherry picked from commit b832ae8a4a)
2010-12-28 09:08:24 +08:00
Eric Anholt
d7e5620d6e i965: Add support for gen6 reladdr VS constant loading.
(cherry picked from commit 3a3b1bd722)
2010-12-27 14:30:42 -08:00
Eric Anholt
0bb9a3215e i965: Add support for gen6 constant-index constant loading.
(cherry picked from commit 15566183a6)
2010-12-27 14:30:42 -08:00
Eric Anholt
962ef4cada i965: Set the alternative floating point mode on gen6 VS and WM.
This matches how we did the math instructions pre-gen6, though it
applies to non-math as well.

Fixes vp1-LIT test 2 (degenerate case: 0 ^ 0 -> 1)
(cherry picked from commit c52adfc2e1)
2010-12-27 14:30:42 -08:00
Chris Wilson
a8e34dd8c2 intel: Check for unsupported texture when finishing using as a render target
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=32541
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
(cherry picked from commit 8b9570e685)
2010-12-27 14:30:42 -08:00
Zhenyu Wang
4fd0d556fc i965: explicit tell header present for fb write on sandybridge
Determine header present for fb write by msg length is not right
for SIMD16 dispatch, and if there're more output attributes, header
present is not easy to tell from msg length. This explicitly adds
new param for fb write to say header present or not.

Fixes many cases' hang and failure in GL conformance test.
(cherry picked from commit 4374703a9b)
2010-12-27 14:30:42 -08:00
Eric Anholt
8a908819cc i965: Avoid using float type for raw moves, to work around SNB issue.
The SNB alt-mode math does the denorm and inf reduction even for a
"raw MOV" like we do for g0 message header setup, where we are moving
values that aren't actually floats.  Just use UD type, where raw MOVs
really are raw MOVs.

Fixes glxgears since c52adfc2e1, but no
piglit tests had regressed(!)
(cherry picked from commit 4fe78d3e12)
2010-12-27 14:30:42 -08:00
Eric Anholt
6af8eac21b intel: Support glCopyTexImage() from XRGB8888 to ARGB8888.
The only mismatch between the two is that we have to clear the
destination's alpha to 1.0.  Fixes WOW performance on my Ironlake,
from a few frames a second to almost playable.
(cherry picked from commit 290a1141bc)
2010-12-27 14:30:42 -08:00
Eric Anholt
0874c37195 intel: Try to sanely check that formats match for CopyTexImage.
Before, we were going off of a couple of known (hopeful) matches
between internalFormats and the cpp of the read buffer.  Instead, we
can now just look at the gl_format of the two to see if they match.
We should avoid bad blits that might have been possible before, but
also allow different internalFormats to work without having to
enumerate each one.
(cherry picked from commit ec03b316b4)
2010-12-27 14:30:42 -08:00
Eric Anholt
0874356dbf intel: Drop commented intel_flush from copy_teximage.
The blit that follows appears in the command stream so it's serialized
with previous rendering.  Any queued vertices in the tnl layer were
already flushed up in mesa/main/.
(cherry picked from commit e65c643792)
2010-12-27 14:30:42 -08:00
Eric Anholt
3232019b67 intel: Update renderbuffers before looking up CopyTexImage's read buffer.
Not fixing a particular bug, just noticed by code inspection.
(cherry picked from commit 99c7840b0c)
2010-12-27 14:30:41 -08:00
Zhenyu Wang
fdf27dfdf8 i965: Use MI_FLUSH_DW for blt ring flush on sandybridge
Old MI_FLUSH command is deprecated on sandybridge blt.
(cherry picked from commit 845d651cf6)
2010-12-27 14:30:41 -08:00
Eric Anholt
10757e86f2 i965: Add support for using the BLT ring on gen6.
(cherry picked from commit c27285610c)
2010-12-27 14:30:41 -08:00
Eric Anholt
d62a383d53 i965: Fix gl_FragCoord.z setup on gen6.
Fixes glsl-bug-22603.
(cherry picked from commit 036c817f77)
2010-12-27 14:30:41 -08:00
Eric Anholt
c043c5ae00 i956: Fix the old FP path fragment position setup on gen6.
Fixes fp-arb-fragment-coord-conventions-none
(cherry picked from commit 5fbd8da8df)
2010-12-27 14:30:41 -08:00
Eric Anholt
da4ecaef8a i965: Fix ARL to work on gen6.
RNDD isn't one of the instructions that can do conversion from
execution type to destination type.

Fixes glsl-vs-arrays-3.
(cherry picked from commit 7cec7bf56c)
2010-12-27 14:30:41 -08:00
Eric Anholt
9a6d7d7eb8 intel: Include stdbool so we can stop using GLboolean when we want to.
This requires shuffling the driconf XML macros around, since they use
true and false tokens expecting them to not get expanded to anything.
(cherry picked from commit df9f891544)
2010-12-27 14:30:41 -08:00
Xiang, Haihao
1988cba847 i965: use align1 access mode for instructions with execSize=1 in VS
All operands must be 16-bytes aligned in aligh16 mode. This fixes l_xxx.c
in oglconform.
(cherry picked from commit dc987adc9f)
2010-12-27 08:49:05 +08:00
Xiang, Haihao
8953ac2570 i965: fix register region description
This fixes
 brw_eu_emit.c:179: validate_reg: Assertion `width == 1' failed.
(cherry picked from commit 8249321604)
2010-12-27 08:48:54 +08:00
Xiang, Haihao
639f595fa0 i965: support for two-sided lighting on Sandybridge
VS places color attributes together so that SF unit can fetch the right
attribute according to object orientation. This fixes light issue in
mesa demo geartrain, projtex.
(cherry picked from commit e47eacdc53)
2010-12-27 08:48:30 +08:00
Brian Paul
604009fa77 mesa/meta: fix broken assertion, rename stack depth var
assert(current_save_state < MAX_META_OPS_DEPTH) did not compile.

Rename current_save_state to SaveStackDepth to be more consistent with
the style of the other fields.
(cherry picked from commit 2a4df8933e)
2010-12-27 08:47:43 +08:00
Xiang, Haihao
7e856fd043 meta: allow nested meta operations
_mesa_meta_CopyPixels results in nested meta operations on Sandybridge.
Previoulsy the second meta operation overrides all states saved by the
first meta function.
(cherry picked from commit d1196bbc19)
2010-12-27 08:47:05 +08:00
Chia-I Wu
fd2b11e433 st/egl: Fix eglChooseConfig when configs is NULL.
When configs is NULL, the app wants to know the number of matching
configs.
(cherry picked from commit 9f2062fb12)
2010-12-26 23:39:33 +08:00
Chia-I Wu
662afccabe docs/egl: Update egl.html.
Various updates and a new section about packaging.
(cherry picked from commit 65e8f81110)
2010-12-25 02:57:58 +08:00
Eric Anholt
7d0c7d52e4 i965: Correct the dp_read message descriptor setup on g4x.
It's mostly like gen4 message descriptor setup, except that the sizes
of type/control changed to be like gen5.  Fixes 21 piglit cases on
gm45, including the regressions in bug #32311 from increased VS
constant buffer usage.
(cherry picked from commit 5dc53444c8)
2010-12-23 15:36:48 -08:00
Chia-I Wu
b7c187df9a st/egl: Assorted fixes for dri2_display_get_configs.
Set window_bit only when the visual id is greater than zero.  Correct
visual types.  Skip slow configs as they are not relevant.  Finally, do
not return duplicated configs.
(cherry picked from commit 445cb9e53b)
2010-12-22 16:07:18 +08:00
Chia-I Wu
55fb7269f0 st/egl: Fix eglCopyBuffers.
Flush before presenting.
(cherry picked from commit a31e2e3312)
2010-12-22 14:27:04 +08:00
Chia-I Wu
0dc5b97ddd st/egl: Plug pbuffer leaks.
Unreference validated resources or remove unnecessary validations.
(cherry picked from commit 18bc427ade)
2010-12-22 14:27:01 +08:00
Marek Olšák
612e26e82c r300g: finally fix the texture corruption on r3xx-r4xx
Even though a bound texture stays bound when calling set_fragment_sampler_views,
it must be assigned a new cache region depending on the occupancy of other
texture units.

This fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=28800

Thanks to Álmos <aaalmosss@gmail.com> for finding the bug in the code.

NOTE: This is a candidate for both the 7.9 and 7.10 branches.
(cherry picked from commit daffaca53e)
2010-12-18 10:19:16 +01:00
Marek Olšák
ed9eed088e r300g: fix rendering with a vertex attrib having a zero stride
The hardware apparently does support a zero stride, so let's use it.

This fixes missing objects in ETQW, but might also fix a ton of other
similar-looking bugs.

NOTE: This is a candidate for both the 7.9 and 7.10 branches.
(cherry picked from commit 2af8a19831)
2010-12-16 17:29:46 +01:00
Marek Olšák
ccff6dcf0d r300/compiler: fix swizzle lowering with a presubtract source operand
If a source operand has a non-native swizzle (e.g. the KIL instruction
cannot have a swizzle other than .xyzw), the lowering pass uses one or more
MOV instructions to move the operand to an intermediate temporary with
native swizzles.

This commit fixes that the presubtract information was lost during
the lowering.

NOTE: This is a candidate for both the 7.9 and 7.10 branches.
(cherry picked from commit d0990db6bd)
2010-12-16 17:29:32 +01:00
Marek Olšák
c547e230bb r300/compiler: fix LIT in VS
This fixes broken rendering of trees in ETQW. The trees still disappear
for an unknown reason when they are close.

Broken since:
2ff9d4474b
r300/compiler: make lowering passes possibly use up to two less temps

NOTE: This is a candidate for the 7.10 branch.
(cherry picked from commit 9e1fbd3d6e)
2010-12-16 17:29:16 +01:00
Dave Airlie
32218e4cc8 r300g: fixup rs690 tiling stride alignment calculations.
The RS690 memory controller prefers things to be on a different
boundary than the discrete GPUs, we had an attempt to fix this,
but it still failed, this consolidates the stride calculation
into one place and removes the really special case check.

This fixes gnome-shell and 16 piglit tests on my rs690 system.

NOTE: This is a candidate for both the 7.9 and 7.10 branches.

Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit d19b5cbd31)
2010-12-16 11:51:43 +10:00
Brian Paul
7db3e66ba8 mesa, st/mesa: disable GL_ARB_geometry_shader4
The new GLSL compiler doesn't support geom shaders yet so disable the
GL_ARB_geometry_shader4 extension.  Undo this when geom shaders work again.

NOTE: This is a candidate for the 7.10 branch.

(cherry picked from commit bb7c2691d2)
2010-12-14 16:30:01 -07:00
Brian Paul
ad523d08b5 tnl: a better way to initialize the gl_program_machine memory
This improves commit ef3f7e61b3

NOTE: This is a candidate for the 7.9 and 7.10 branches.

(cherry picked from commit 6577f753b2)
2010-12-14 16:29:31 -07:00
Brian Paul
7d38797ac9 tnl: Initialize gl_program_machine memory in run_vp.
Fixes piglit valgrind glsl-array-bounds-04 failure (FDO bug 29946).

NOTE:
This is a candidate for the 7.10 branch.
This is a candidate for the 7.9 branch.

(cherry picked from commit ef3f7e61b3)
2010-12-14 16:29:18 -07:00
Brian Paul
d50e8b2276 draw/llvm: don't flush in vs_llvm_delete()
Fixes piglit glx-shader-sharing crash.

When shaders are shared by multiple contexts, the shader's draw context
pointer may point to a previously destroyed context.  Dereferencing the
context pointer will lead to a crash.

In this case, simply removing the flushing code avoids the crash (the
exec and sse shader paths don't flush here either).

There's a deeper issue here, however, that needs examination.  Shaders
should not keep pointers to contexts since contexts might get destroyed
at any time.

NOTE: This is a candidate for the 7.10 branch (after this has been
tested for a while).

(cherry picked from commit becc4bb90c)
2010-12-14 16:28:41 -07:00
Luca Barbieri
0e50c21e24 glsl: Unroll loops with conditional breaks anywhere (not just the end)
Currently we only unroll loops with conditional breaks at the end, which is
the form that lower_jumps generates.

However, if breaks are not lowered, they tend to appear at the beginning, so
add support for a conditional break anywhere.

Signed-off-by: Luca Barbieri <luca@luca-barbieri.com>
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
2010-12-09 16:42:05 -08:00
Kenneth Graunke
13c45c590b glsl: Consider the "else" branch when looking for loop breaks.
Found this bug by code inspection.  Based off the comments just before
this code, the intent is to find whether the break exists in the "then"
branch or the "else" branch.  However, the code actually looked at the
last instruction in the "then" branch twice.
2010-12-09 16:42:05 -08:00
Kenneth Graunke
528fa8ce32 glsl: Clean up code by adding a new is_break() function. 2010-12-09 16:42:05 -08:00
Eric Anholt
b13a2e640f glsl: Correct the marking of InputsRead/OutputsWritten on in/out matrices.
If you used a constant array index to access the matrix, we'd flag a
bunch of wrong inputs/outputs as being used because the index was
multiplied by matrix columns and the actual used index was left out.

Fixes glsl-mat-attribute.
2010-12-09 14:41:50 -08:00
Eric Anholt
3fb18d6775 intel: Set the swizzling for depth textures using the GL_RED depth mode.
Fixes depth-tex-modes-rg.
2010-12-09 14:41:50 -08:00
Eric Anholt
b4e8ec3a57 intel: Use plain R8 and RG8 for COMPRESSED_RED and COMPRESSED_RG.
Fixes texture-rg.
2010-12-09 14:41:50 -08:00
Vinson Lee
c3ca384e71 i965: Silence uninitialized variable warning.
Fixes this GCC warning.
brw_fs.cpp: In function 'brw_reg brw_reg_from_fs_reg(fs_reg*)':
brw_fs.cpp:3255: warning: 'brw_reg' may be used uninitialized in this function
2010-12-09 14:17:17 -08:00
Vinson Lee
af5f7b3260 r600g: Fix SCons build. 2010-12-09 14:03:58 -08:00
Jerome Glisse
121079bd67 r600g: indentation cleanup
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2010-12-09 16:16:22 -05:00
Jerome Glisse
7055068eea r600g: specialized upload manager
Allow important performance increase by doing hw specific implementation
of the upload manager helper. Drop the range flushing that is not hit with
this code (and wasn't with previous neither). Performance improvement are
mostly visible on slow CPU.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2010-12-09 16:07:05 -05:00
Jerome Glisse
15753cf54d r600g: avoid using pb* helper we are loosing previous cpu cycle with it
r600g is up to a point where all small CPU cycle matter and pb* turn
high on profile. It's mostly because pb try to be generic and thus
trigger unecessary check for r600g driver. To avoid having too much
abstraction & too much depth in the call embedded everythings into
r600_bo. Make code simpler & faster. The performance win highly depend
on the CPU & application considered being more important on slower CPU
and marginal/unoticeable on faster one.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2010-12-09 16:07:01 -05:00
Fabian Bieler
ef534f3838 glsl: fix lowering conditional returns in subroutines
this fix applies to the lower_sub_return 'branch' of the lower_jumps pass

Fixes piglit tests glsl-functions-5 and glsl-functions-6.
2010-12-09 11:28:06 -08:00
Eric Anholt
834cc8e501 i965: remove unused variable since brw_wm_glsl.c removal. 2010-12-09 11:11:04 -08:00
Eric Anholt
cfcc2ef587 i965: Set render_cache_read_write surface state bit on gen6 constant surfs.
This is said to be required in the spec, even when you aren't doing writes.
2010-12-09 11:11:04 -08:00
Eric Anholt
30f25a1019 i965: Set up the correct texture border color state struct for Ironlake.
This doesn't actually fix border color on Ironlake, but it appears to
be a requirement, and gen6 needs it too.
2010-12-09 10:51:00 -08:00
Eric Anholt
14a9153a32 i965: Clean up VS constant buffer location setup. 2010-12-09 10:51:00 -08:00
Eric Anholt
8fab1c0e2e i965: Fix VS constants regression pre-gen6.
Last minute change for gen6 with 0 used params dropped the multiply.
2010-12-09 10:50:59 -08:00
José Fonseca
cdd4f04f80 llvmpipe: Plug fence leaks. 2010-12-09 16:48:26 +00:00
Shuang He
9946f15d30 mesa: allow GLfixed arrays for OpenGL ES 2.0
Signed-off-by: Brian Paul <brianp@vmware.com>
2010-12-09 08:23:54 -07:00
Chia-I Wu
0c0eda393a mesa: Fix glTexCoordPointer with type GL_FIXED.
GL_FIXED is also a legal type for glTexCoordPointer.
2010-12-09 19:37:56 +08:00
Chia-I Wu
2d270ac090 mesa: Fix GL_FIXED arrays.
It is broken since 433e5e6def.
2010-12-09 19:25:05 +08:00
Eric Anholt
05e534e6c4 i965: Drop push-mode reladdr constant loading and always use constant_map.
This eases the gen6 implementation, which can only handle up to 32
registers of constants, while likely not penalizing real apps using
reladdr since all of those I've seen also end up hitting the pull
constant buffer.  On gen6, the constant map means that simple NV VPs
fit under the 32-reg limit and now succeed.  Fixes around 10 testcases.
2010-12-08 22:26:18 -08:00
Alex Deucher
fd543e1f95 radeon: bump mip tree levels to 15
I forgot to bump this when I bumped the tex levels.
2010-12-09 00:16:02 -05:00
Brian Paul
09a5e028a6 mesa: simplify target checking for TexImage functions 2010-12-08 21:38:35 -07:00
Brian Paul
7404fa3f07 mesa: revamp error checking for compressed texture images
Simplify some code, remove unneeded checks, etc.
2010-12-08 21:38:35 -07:00
Chad Versace
f0f2ec4d8a glsl: In ast_to_hir, check sampler array indexing
Raise error if a sampler array is indexed with a non-constant expression.

From section 4.1.7 of the GLSL 1.30 spec:
  "Samplers aggregated into arrays within a shader (using square
  brackets [ ]) can only be indexed with integral constant
  expressions [...]."
2010-12-08 18:53:53 -08:00
Eric Anholt
d547ab150a i965: Drop KIL_NV from the ff/ARB_fp path since it was only used for GLSL. 2010-12-08 11:14:52 -08:00
Eric Anholt
dba6fde08c i965: Use the new pixel mask location for gen6 ARB_fp KIL instructions.
Fixes:
fp-kil
fp-generic/kil-swizzle.
2010-12-08 11:14:35 -08:00
Eric Anholt
39eaacff14 i965: Set the render target index in gen6 fixed-function/ARB_fp path.
Fixes:
fbo-drawbuffers2-blend
fbo-drawbuffers2-colormask
2010-12-08 10:51:04 -08:00
Eric Anholt
4b4dc778b6 i965: Set up the per-render-target blend state on gen6.
This will let us get EXT_draw_buffers2 blending and colormasking working.
2010-12-08 10:50:57 -08:00
Eric Anholt
0d3b8a5cc2 i965: Set up the color masking for the first drawbuffer on gen6.
Fixes glean/maskedClear
2010-12-08 09:53:16 -08:00
Chia-I Wu
d2028ba339 mesa: Do not advertise GL_OES_texture_3D.
GL_OES_texture_3D has a GLSL counterpart.  Since it is not implemented,
GL_OES_texture_3D should not be advertised.
2010-12-08 22:35:40 +08:00
Chia-I Wu
98ee6739d9 vbo: Fix GLES2 glVertexAttrib.
Attribute 0 has no special meaning in GLES2.  Add VertexAttrib4f_nopos
for that purpose and make _es_VertexAttrib* call the new function.

Rename _vbo_* to _es_* to avoid confusion.  These functions are only
used by GLES, and now some of them (_es_VertexAttrib*) even behave
differently than vbo_VertexAttrib*.
2010-12-08 22:19:19 +08:00
Chia-I Wu
9f0d7dd259 vbo: Drop second ATTR macro.
There is no need to have a special version of ATTR for
!FEATURE_beginend, since 81ccb3e2ce.
2010-12-08 22:18:37 +08:00
Brian Paul
7da704ee72 configure: use llvm-config --cppflags instead of --cflags 2010-12-08 06:45:00 -07:00
Brian Paul
c64447f47d mesa: make _mesa_test_proxy_teximage() easier to read 2010-12-07 21:37:20 -07:00
Brian Paul
4ff70b7a8f mesa: consolidate glCompressedTexImage1/2/3D() functions 2010-12-07 21:37:20 -07:00
Brian Paul
1c23b860ce mesa: consolidate glCopyTexSubImage1/2/3D() functions 2010-12-07 21:37:20 -07:00
Brian Paul
11f386fb50 mesa: consolidate glCopyTexImage1/2D() code 2010-12-07 21:37:19 -07:00
Brian Paul
45124e043d mesa: consolidate the glTexSubImage1/2/3D() functions 2010-12-07 21:37:19 -07:00
Brian Paul
35f620d55c mesa: simplify proxy texture code in texture_error_check() 2010-12-07 21:37:19 -07:00
Marek Olšák
c4f9e55d61 r300/compiler: remove at least unused immediates if externals cannot be removed 2010-12-08 04:39:51 +01:00
Marek Olšák
2ff9d4474b r300/compiler: make lowering passes possibly use up to two less temps
CMP may now use two less temps, other non-native instructions may end up
using one less temp, except for SIN/COS/SCS, which I am leaving unchanged
for now.

This may reduce register pressure inside loops, because the register
allocator doesn't do a very good job there.
2010-12-08 04:39:51 +01:00
Marek Olšák
93f2df0760 r300/compiler: handle DPH and XPD in rc_compute_sources_for_writemask
This bug can only be triggered if you put deadcode before native rewrite.
2010-12-08 04:39:50 +01:00
Marek Olšák
95080fb50f r300/compiler: do not print pair/tex/presub program stats for vertex shaders 2010-12-08 04:39:50 +01:00
Marek Olšák
8fac29d49e r300/compiler: cleanup rc_run_compiler 2010-12-08 04:39:50 +01:00
Marek Olšák
2592a8f506 r300/compiler: add a function to query program stats (alu, tex, temps..) 2010-12-08 04:39:50 +01:00
Marek Olšák
431b4c0c84 r300/compiler: don't terminate regalloc if we surpass max temps limit
The same check is already in a later pass (translate_vertex_program).
2010-12-08 04:39:50 +01:00
Eric Anholt
2f07a744f1 i965: Don't try to store gen6 (float) blend constant color in bytes.
Fixes glean/blendFunc
2010-12-07 19:33:47 -08:00
Ian Romanick
002cd2c8d4 linker: Fix regressions caused by previous commit
That's what I get for not running piglit before pushing.

Don't try to patch types of unsized arrays when linking fails.

Don't try to patch types of unsized arrays that are shared between
shader stages.
2010-12-07 19:00:44 -08:00
Ian Romanick
6f53921c4b linker: Ensure that unsized arrays have a size after linking
Fixes piglit test case glsl-vec-array (bugzilla #31908).

NOTE: This bug does not affect 7.9, but I think this patch is a
candiate for the 7.9 branch anyway.
2010-12-07 18:32:16 -08:00
Eric Anholt
b2167a6c01 i965: Fix flipped value of the not-embedded-in-if on gen6.
Fixes:
glean/glsl1-! (not) operator (1, fail)
glean/glsl1-! (not) operator (1, pass)
2010-12-07 17:46:47 -08:00
Ian Romanick
b0fc5103cb glsl: Inherrit type of declared variable from initializer
Types of declared variables and their initializer must match excatly
except for unsized arrays.  Previously the type inherritance for
unsized arrays happened implicitly in the emitted assignment.
However, this assignment is never emitted for uniforms.  Now that type
is explicitly copied unconditionally.

Fixes piglit test array-compare-04.vert (bugzilla #32035) and
glsl-array-uniform-length (bugzilla #31985).

NOTE: This is a candidate for the 7.9 branch.
2010-12-07 16:36:44 -08:00
Eric Anholt
7ca7e9b626 i965: Work around gen6 ignoring source modifiers on math instructions.
With the change of extended math from having the arguments moved into
mrfs and handed off through message passing to being directly hooked
up to the EU, it looks like the piece for doing source modifiers
(negate and abs) was left out.

Fixes:
fog-modes
glean/fp1-ARB_fog_exp test
glean/fp1-ARB_fog_exp2 test
glean/fp1-Computed fog exp test
glean/fp1-Computed fog exp2 test
ext_fog_coord-modes
2010-12-07 15:11:27 -08:00
Eric Anholt
2d7dfb8446 i965: Add disabled debug code for dumping out the WM constant payload.
This can significantly ease thinking about the asm.
2010-12-07 15:11:27 -08:00
Ian Romanick
6848e27e14 i965: Correctly emit constants for aggregate types (array, matrix, struct)
Previously the code only handled scalars and vectors.  This new code
is modeled somewhat after similar code in ir_to_mesa.

Reviewed-by: Eric Anholt <eric@anholt.net>
2010-12-07 15:03:14 -08:00
Jerome Glisse
b7617346dc r600g: fix userspace fence against lastest kernel
R6XX GPU doesn't like to have two partial flush writting
back to memory in row without a prior flush of the pipeline.
Add PS_PARTIAL_FLUSH to flush all work between the CP and
the ES, GS, VS, PS shaders.

Thanks a lot to Alban Browaeys (prahal on irc) for investigating
this issue.

Signed-off-by: Alban Browaeys <prahal@yahoo.com>
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2010-12-07 17:54:56 -05:00
Jerome Glisse
69251fc4cd r600g: remove dead code
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2010-12-07 16:14:18 -05:00
Eric Anholt
fa0d5a2c5b i965: Always hand the absolute value to RSQ.
gen6 builtin RSQ apparently clamps negative values to 0 instead of
returning the RSQ of the absolute value like ARB_fragment_program
desires and pre-gen6 apparently does.

Fixes:
glean/fp1-RSQ test 2 (reciprocal square root of negative value)
glean/vp1-RSQ test 2 (reciprocal square root of negative value)
2010-12-07 13:13:27 -08:00
Ian Romanick
6d36be508f glsl: Ensure that equality comparisons don't return a NULL IR tree
This fixes bugzilla #32035 and piglit test case array-compare-01 and
array-compare-02.

NOTE: This is a candidate for the 7.9 branch.
2010-12-07 12:50:38 -08:00
Eric Anholt
72845d206e i965: Handle saturates on gen6 math instructions.
We get saturate as an argument to brw_math() instead of as compile
state, since that's how the pre-gen6 send instructions work.  Fixes
fp-ex2-sat.
2010-12-07 12:21:08 -08:00
Eric Anholt
ed492e9544 i965: Fix comment about gen6_wm_constants.
This is the push constant buffer, not the pull constants.
2010-12-07 12:21:08 -08:00
Kenneth Graunke
bd74101aeb Refresh autogenerated glcpp parser. 2010-12-07 10:52:59 -08:00
Kenneth Graunke
800eed6765 glcpp: Don't emit SPACE tokens in conditional_tokens production.
Fixes glslparsertest defined-01.vert.

Reported-by: José Fonseca <jfonseca@vmware.com>
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Acked-by: Carl Worth <cworth@cworth.org>
2010-12-07 10:52:36 -08:00
Marek Olšák
d8b861987d r300g: also revalidate the SWTCL vertex buffer after its reallocation 2010-12-07 19:26:28 +01:00
Zhenyu Wang
27609a8267 i965: upload WM state for _NEW_POLYGON on sandybridge
Be sure polygon stipple mode is updated. This fixes 'gamma' demo.
2010-12-07 17:05:02 +08:00
Vinson Lee
6ee415cc6d r200: Silence uninitialized variable warning.
Fixes this GCC warning.
r200_maos_arrays.c: In function 'r200EmitArrays':
r200_maos_arrays.c:113: warning: 'emitsize' may be used uninitialized in this function
2010-12-07 00:58:54 -08:00
Xiang, Haihao
5ff6ed2b97 i965: set minimum/maximum Point Width on Sandybridge
It is used for point width on vertex. This fixes mesa demo spriteblast and pointblast.
2010-12-07 16:43:24 +08:00
Vinson Lee
dda3ccba8c mesa: Clean up header file inclusion in viewport.h. 2010-12-07 00:37:48 -08:00
Vinson Lee
dbd3f72662 mesa: Clean up header file inclusion in varray.h. 2010-12-07 00:33:36 -08:00
Vinson Lee
2aa36f78dc mesa: Clean up header file inclusion in transformfeedback.h. 2010-12-07 00:28:57 -08:00
Vinson Lee
8cc3d411ed mesa: Clean up header file inclusion in texrender.h. 2010-12-07 00:19:06 -08:00
Marek Olšák
4953ba6a71 r300g: validate buffers only if any of bound buffers is changed
This prevents needless buffer validation (CS space checking).
2010-12-07 08:46:18 +01:00
Marek Olšák
78068a5fbf r300g: cache packet dwords of 3D_LOAD_VBPNTR in a command buffer if possible
It's not always possible to preprocess the content of 3D_LOAD_VBPNTR
in a command buffer, because the offset to all vertex buffers (which
the packet depends on) is derived from the "start" parameter of draw_arrays
and the "indexBias" parameter of draw_elements, but we can at least lazily
make a command buffer for the case when offset == 0, which should occur
most of the time.
2010-12-07 06:42:05 +01:00
Marek Olšák
857d107bfe u_blitter: use util_is_format_compatible in the assert 2010-12-07 06:22:38 +01:00
Brian Paul
d0b2b8da7d mesa: consolidate glTexImage1/2/3D() code
Something similar could be done for glCopyTex[Sub]Image() and the
compressed texture image functions as well.
2010-12-06 17:10:05 -07:00
Brian Paul
6dca66b620 mesa: set gl_texture_object::_Complete=FALSE in incomplete() 2010-12-06 17:10:05 -07:00
Brian Paul
ecb7cc3319 mesa: test for cube map completeness in glGenerateMipmap()
The texture is not cube complete if the base level images aren't of
the same size and format.

NOTE: This is a candidate for the 7.9 branch.
2010-12-06 17:10:05 -07:00
Kenneth Graunke
c17c790387 glsl: Properly add functions during lazy built-in prototype importing.
The original lazy built-in importing patch did not add the newly created
function to the symbol table, nor actually emit it into the IR stream.

Adding it to the symbol table is non-trivial since importing occurs when
generating some ir_call in a nested scope.  A new add_global_function
method, backed by new symbol_table code in the previous patch, handles
this.

Fixes bug #32030.
2010-12-06 13:43:22 -08:00
Kenneth Graunke
a8f52647b0 symbol_table: Add support for adding a symbol at top-level/global scope. 2010-12-06 13:43:22 -08:00
Kenneth Graunke
6fae1e4c4d glsl: Factor out code which emits a new function into the IR stream.
A future commit will use the newly created function in a second place.
2010-12-06 13:43:22 -08:00
Jakob Bornecrantz
d72cb9c94d st/mesa: Unbind all constant buffers 2010-12-06 22:10:49 +01:00
Jerome Glisse
e0d554ab78 r600g: remove useless flush map
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2010-12-06 15:50:50 -05:00
Jerome Glisse
afc56b1861 r600g: avoid useless shader rebuild at draw call
Avoid rebuilding constant shader state at each draw call,
factor out spi update that might change at each draw call.
Best would be to update spi only when revealent states
change (likely only flat shading & sprite point).

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2010-12-06 15:50:50 -05:00
Jerome Glisse
fa86fc564a r600g: build fetch shader from vertex elements
Vertex elements change are less frequent than draw call, those to
avoid rebuilding fetch shader to often build the fetch shader along
vertex elements. This also allow to move vertex buffer setup out
of draw path and make update to it less frequent.

Shader update can still be improved to only update SPI regs (based
on some rasterizer state like flat shading or point sprite ...).

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2010-12-06 15:50:50 -05:00
José Fonseca
a9fa0f3a2f mesa: Bump the number of bits in the register index.
More than 1023 temporaries were being used for a Cinebench shader before
doing temporary optimization, causing the index value to wrap around to
-1024.
2010-12-06 20:03:51 +00:00
Brian Paul
cae2bb76c1 st/mesa: fix mipmap generation bug
In st_finalize_texture() we were looking at the st_texture_object::
lastLevel field instead of the pipe_resource::last_level field to
determine which resource to store the mipmap in.

Then, in st_generate_mipmap() we need to call st_finalize_texture() to
make sure the destination resource is properly allocated.

These changes fix the broken piglit fbo-generatemipmap-formats test.
2010-12-06 11:01:21 -07:00
Brian Paul
c1095f0b0c mesa/llvm: use llvm-config --cppflags
Use --cppflags instead of --cflags so that we get the -I and -D flags
we want, but not compiler options like -O3.

A similar change should probably be made for autoconf.
2010-12-06 10:11:00 -07:00
Brian Paul
4fc62a074c gallium/util: minor formatting fixes 2010-12-06 09:46:45 -07:00
Brian Paul
b89a731ff2 mesa: add error margin to clip mask debug/check code
When X or Y or Z is close to W the outcome of the floating point clip
test comparision may be different between the C and x86 asm paths.
That's OK; don't report an error.

See fd.o bug 32093
2010-12-06 09:46:01 -07:00
Eric Anholt
4538ce915b i965: Remove INTEL_DEBUG=glsl_force now that there's no brw_wm_glsl.c 2010-12-06 00:14:23 -08:00
Eric Anholt
5ba517baa2 i965: Nuke brw_wm_glsl.c.
It was only used for gen6 fragment programs (not GLSL shaders) at this
point, and it was clearly unsuited to the task -- missing opcodes,
corrupted texturing, and assertion failures hit various applications
of all sorts.  It was easier to patch up the non-glsl for remaining
gen6 changes than to make brw_wm_glsl.c complete.

Bug #30530
2010-12-06 00:14:23 -08:00
Eric Anholt
245662f308 i965: Add support for the instruction compression bits on gen6.
Since the 8-wide first-quarter and 16-wide first-half have the same
bit encoding, we now need to track "do you want instruction
compression" in the compile state.
2010-12-06 00:14:23 -08:00
Eric Anholt
3f8bcb0d99 i965: Align gen6 push constant size to dispatch width.
The FS backend is fine with register level granularity.  But for the
brw_wm_emit.c backend, it expects pairs of regs to be used for the
constants, because the whole world is pairs of regs.  If an odd number
got used, we went looking for interpolation in the wrong place.
2010-12-06 00:14:23 -08:00
Eric Anholt
237aa33c67 i965: Make the sampler's implied move on gen6 be a raw move.
We were accidentally doing a float-to-uint conversion.
2010-12-06 00:14:23 -08:00
Eric Anholt
5340dd8cca i965: Fix up gen6 samplers for their usage by brw_wm_emit.c
We were trying to do the implied move even when we'd already manually
moved the real header in place.
2010-12-06 00:14:22 -08:00
Eric Anholt
ad29e79850 i965: Fix gen6 interpolation setup for 16-wide.
In the SF and brw_fs.cpp fixes to set up interpolation sanely on gen6,
the setup for 16-wide interpolation was left behind.  This brings
relative sanity to that path too.
2010-12-06 00:14:22 -08:00
Eric Anholt
ae0df25ab4 i965: Don't smash a group of coordinates doing gen6 16-wide sampler headers. 2010-12-06 00:14:22 -08:00
Eric Anholt
d1ead22d1b i965: Fix up 16-wide gen6 FB writes after various refactoring. 2010-12-06 00:14:22 -08:00
Eric Anholt
ad35528944 i965: Provide delta_xy reg to gen6 non-GLSL path PINTERP.
Fixes many assertion failures in that path.
2010-12-06 00:14:22 -08:00
Eric Anholt
16f8c82389 i965: Move payload reg setup to compile, not lookup time.
Payload reg setup on gen6 depends more on the dispatch width as well
as the uses_depth, computes_depth, and other flags.  That's something
we want to decide at compile time, not at cache lookup.  As a bonus,
the fragment shader program cache lookup should be cheaper now that
there's less to compute for the hash key.
2010-12-06 00:14:22 -08:00
Chia-I Wu
8f2a974cf2 mapi: Rewrite mapi_abi.py to get rid of preprocessor magic.
The preprocessor magic in mapi was nothing but obfuscation.  Rewrite
mapi_abi.py to generate real C code.

This commit removes the hack added in
43121f2086.
2010-12-06 15:40:37 +08:00
Chia-I Wu
5ae4b6693a egl: _eglFilterArray should not allocate.
Otherwise, when it is called from within a driver, the caller cannot
free the returned data (on Windows).
2010-12-06 15:40:37 +08:00
Zhenyu Wang
2b1469340b i965: Fix GS state uploading on Sandybridge
Need to check the required primitive type for GS on Sandybridge,
and when GS is disabled, the new state has to be issued too, instead
of only updating URB state with no GS entry, that caused hang on Sandybridge.

This fixes hang issue during conformance suite testing.
2010-12-06 15:20:48 +08:00
Xiang, Haihao
08be8d6450 i965: fix for flat shading on Sandybridge
use constant interpolation instead of linear interpolation for
attributes COL0,COL1 if GL_FLAT is used. This fixes mesa demo bounce.
2010-12-06 09:42:28 +08:00
Henri Verbeet
4409435614 r600g: Cleanup fetch shader resources in r600_pipe_shader_destroy(). 2010-12-05 18:44:44 +01:00
Henri Verbeet
308cfb80f5 r600g: Cleanup block bo references in r600_context_fini(). 2010-12-05 18:44:44 +01:00
Marek Olšák
c0c929cdac st/mesa: initialize key in st_vp_varient
This fixes endless vertex shader recompilations in find_translated_vp
if the shader contains an edge flag output.

NOTE: This is a candidate for the 7.9 branch.

Signed-off-by Brian Paul <brianp@vmware.com>
2010-12-05 17:38:19 +01:00
Xavier Chantry
ccacabe86c gallium/trace: check bind_vertex_sampler_states and set_vertex_sampler_views
Signed-off-by: Xavier Chantry <chantry.xavier@gmail.com>
Reviewed-by: Jakob Bornecrantz <wallbraker at gmail.com>
Signed-off-by: Patrice Mandin <patmandin@gmail.com>
2010-12-05 12:12:25 +01:00
Xavier Chantry
e3256ccb04 init ps->context with util_surfaces_get and do_get
Signed-off-by: Xavier Chantry <chantry.xavier@gmail.com>
Reviewed-by: Jakob Bornecrantz <wallbraker at gmail.com>
Signed-off-by: Patrice Mandin <patmandin@gmail.com>
2010-12-05 12:09:38 +01:00
Xavier Chantry
af5345d937 nvfx: fixes after array textures merge
Signed-off-by: Xavier Chantry <chantry.xavier@gmail.com>
Signed-off-by: Patrice Mandin <patmandin@gmail.com>
2010-12-05 12:08:00 +01:00
Marek Olšák
66d45567b4 r300g: optimize looping over atoms
This also removes DBG_STATS (the stats can be obtained with valgrind instead).
2010-12-05 05:52:25 +01:00
Marek Olšák
6947e52548 r300g: cleanup winsys 2010-12-05 05:47:10 +01:00
Dave Airlie
c1365606c5 r300g: try and use all of vertex constant space
Finished up by Marek Olšák.

We can set the constant space to use a different area per-call to the shader,
we can avoid flushing the PVS as often as we do by spreading out the constants
across the whole constant space.

Signed-off-by: Marek Olšák <maraeo@gmail.com>
2010-12-05 05:47:03 +01:00
Marek Olšák
1774273bde r300g: do not use the index parameter in set_constant_buffer
It appears to be a constant buffer index (in case there are more constant
buffers explicitly used by a shader), i.e. something that Gallium currently
does not use. We treated it incorrectly as the offset to a constant buffer.
2010-12-05 05:46:56 +01:00
Vinson Lee
e72651dc5d gallium/noop: Add prototype for noop_init_state_functions.
Silences this GCC warning.
noop_state.c:247: warning: no previous prototype for
'noop_init_state_functions'
2010-12-04 17:30:08 -08:00
Eric Anholt
afb80c5315 i965: Fix compile warning about missing opcodes. 2010-12-04 16:27:57 -08:00
Eric Anholt
8f23039f00 i965: Update gen6 SF state on fragment program change too.
SF state depends on what inputs there are to the fragment program, not
just the outputs of the VS.
2010-12-04 16:26:55 -08:00
Eric Anholt
65570d0482 i965: Update gen6 WM state on compiled program change, not just FP change. 2010-12-04 16:26:55 -08:00
Eric Anholt
4ac2f09e20 intel: Add an env var override to execute for a different GPU revision.
Sometimes I'm on the train and want to just read what's generated
under INTEL_DEBUG=vs,wm for some code on another generation.  Or, for
the next gen enablement we'll want to dump aub files before we have
the actual hardware.  This will let us do that.
2010-12-04 16:26:55 -08:00
Chia-I Wu
859106f196 st/vega: Fix pipe blend state for various blend modes.
rgb_src_factor and rgb_dst_factor should be PIPE_BLENDFACTOR_ONE for
VG_BLEND_SRC_IN and VG_BLEND_DST_IN respectively.  VG_BLEND_SRC_OVER can
be supported only when the fb has no alpha channel.  VG_BLEND_DST_OVER
and VG_BLEND_ADDITIVE have to be supported with a shader.

Note that Porter-Duff blending rules assume premultiplied alpha.
2010-12-04 23:46:38 +08:00
Chia-I Wu
0ee73edecc st/vega: Add blend shaders for all blend modes. 2010-12-04 23:41:35 +08:00
Chia-I Wu
5d24411140 st/vega: Fix VG_BLEND_MULTIPLY.
TEMP[1].w will be needed for OUT.w just below.  Use TEMP[0] to store the
intermediate value.
2010-12-04 23:41:30 +08:00
Vinson Lee
09fba30fde mesa: Clean up header file inclusion in texobj.h. 2010-12-04 01:29:50 -08:00
Vinson Lee
f657ac951d mesa: Clean up header file inclusion in texgetimage.h. 2010-12-04 01:20:28 -08:00
Vinson Lee
0ed245bd57 mesa: Clean up header file inclusion in texformat.h. 2010-12-04 01:11:33 -08:00
Vinson Lee
f8a9f5458b mesa: Clean up header file inclusion in texenvprogram.h. 2010-12-04 01:03:52 -08:00
Vinson Lee
f618f4549a mesa: Clean up header file inclusion in texcompress_s3tc.h. 2010-12-04 01:00:21 -08:00
Vinson Lee
8aa4cd0e50 st/vega: Silence uninitialized variable warning.
Fixes this GCC warning.
api_filters.c: In function 'execute_filter':
api_filters.c:184: warning: 'tex_wrap' may be used uninitialized in this function
2010-12-04 00:53:44 -08:00
Vinson Lee
3132591a4d mesa: Clean up header file inclusion in texcompress.h. 2010-12-04 00:52:14 -08:00
Chia-I Wu
e87a0cd260 st/vega: Blending should use premultiplied alpha.
Convert color values to and back from premultiplied form for blending.
Finally the rendering result of the blend demo looks much closer to that
of the reference implementation.
2010-12-04 15:44:40 +08:00
Chia-I Wu
e8ff3931f8 st/vega: Add support for per-channel alpha.
Drawing an image in VG_DRAW_IMAGE_STENCIL mode produces per-channel
alpha for use in blending.  Add a new shader stage to produce and save
it in TEMP[1].

For other modes that do not need per-channel alpha, the stage does

  MOV TEMP[1], TEMP[0].wwww
2010-12-04 13:20:38 +08:00
Chia-I Wu
a19eaaa6c1 st/vega: Move masking after blending.
Masking should happen after blending.  The shader is not entirely
correct, but leave it as is for now.
2010-12-04 13:20:38 +08:00
Chia-I Wu
3b4c888653 st/vega: Refactor blend shaders.
Add a helper function, blend_generic, that supports all blend modes and
per-channel alpha.  Make other blend generators a wrapper to it.

Both the old and new code expects premultiplied colors, yet the input is
non-premultiplied.  Per-channel alpha is also not used for stencil
image.  They still need to be fixed.
2010-12-04 13:20:32 +08:00
Chia-I Wu
a09baf1668 st/vega: Add some comments to pipeline shaders. 2010-12-04 13:19:29 +08:00
Brian Paul
6c33e820d5 st/mesa: new comment about updating state vars 2010-12-03 17:07:16 -07:00
Brian Paul
64244dfd39 mesa: update comments, remove dead code 2010-12-03 16:45:44 -07:00
Brian Paul
6f851e6642 mesa: remove unneeded cast 2010-12-03 16:40:36 -07:00
Brian Paul
503983b09e mesa: make glGet*(GL_NONE) generate GL_INVALID_ENUM
In find_value() check if we've hit the 0th/invalid entry before checking
if the pname matches.

Fixes http://bugs.freedesktop.org/show_bug.cgi?id=31987

NOTE: This is a candidate for the 7.9 branch.
2010-12-03 16:04:26 -07:00
Brian Paul
40ee69b4f3 swrast: restructure some glReadPixels() code 2010-12-03 15:26:31 -07:00
Brian Paul
5fc2548fae swrast: accept GL_RG in glReadPixels()
Fixes http://bugs.freedesktop.org/show_bug.cgi?id=32088
2010-12-03 15:26:31 -07:00
Kenneth Graunke
b7acf538af ir_print_visitor: Print out constant structure values.
In the form (constant type ((field1 value) (field2 value) ...))
2010-12-03 13:59:21 -08:00
Brian Paul
8d6a0dc7f3 swrast: fix indentation 2010-12-03 14:49:19 -07:00
Brian Paul
75746e3779 swrast: allow GL_RG format in glDrawPixels()
Restructure the switch statement to avoid having to add additional
color formats in the future.

Fixes http://bugs.freedesktop.org/show_bug.cgi?id=32086
2010-12-03 14:48:03 -07:00
Brian Paul
b87369e863 mesa: consolidate some compiler -D flags
-D__STDC_CONSTANT_MACROS and -D__STDC_LIMIT_MACROS are only needed for
LLVM build.
2010-12-03 13:52:59 -07:00
Marek Olšák
9f7f093090 r300g: one more r500_index_bias_supported leftover 2010-12-03 20:59:55 +01:00
Marek Olšák
536d527020 r300g: add capability bit index_bias_supported
.. instead of calling r500_index_bias_supported(..) every draw call.
2010-12-03 20:34:56 +01:00
Jerome Glisse
edda44e0dc r600g: more indentation fix + warning silencing + dead code removal
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2010-12-03 13:06:53 -05:00
Jerome Glisse
119f00659c r600g: indentation fix
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2010-12-03 12:56:51 -05:00
Jerome Glisse
0b841b0349 r600g: update polygon offset only when rasterizer or zbuffer change
Aim is to build as little state as possible in draw functions.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2010-12-03 12:45:19 -05:00
Brian Paul
dbf996f308 llvmpipe: fix broken stencil writemask
Fixes http://bugs.freedesktop.org/show_bug.cgi?id=32070
2010-12-03 09:44:30 -07:00
Fabian Bieler
cd431a12bf r600g: set address of pop instructions to next instruction 2010-12-03 11:35:44 -05:00
Jerome Glisse
833f3a488a r600g: dump raw shader output for debugging
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2010-12-03 11:35:36 -05:00
Brian Paul
6d13ec7dc0 mesa: return GL_FRAMEBUFFER_DEFAULT as FBO attachment type
If querying the default/window-system FBO's attachment type, return
GL_FRAMEBUFFER_DEFAULT (per the GL_ARB_framebuffer_object spec).

See http://bugs.freedesktop.org/show_bug.cgi?id=31947

NOTE: This is a candidate for the 7.9 branch.
2010-12-03 08:32:29 -07:00
Brian Paul
20cf1851d8 mesa: fix GL_FRAMEBUFFER_ATTACHMENT_OBJECT_NAME query
Return 0 instead of generating an error.

See http://bugs.freedesktop.org/show_bug.cgi?id=30993

Note that piglit fbo-getframebufferattachmentparameter-01 still does
not pass.  But Mesa behaves the same as the NVIDIA driver in this case.
Perhaps the test is incorrect.

NOTE: This is a candidate for the 7.9 branch.
2010-12-03 08:24:51 -07:00
Brian Paul
14746b1d4f gallivm: fix null builder pointers 2010-12-03 07:38:02 -07:00
Chia-I Wu
5be631ce83 st/vega: Add a missing break. 2010-12-03 14:55:29 +08:00
Chia-I Wu
a84a1e344f st/vega: Move vertex transformation to shader.
It was done in path-to-polygon conversion.  That meant that the
results were invalidated when the transformation was modified, and CPU
had to recreate the vertex buffer with new vertices.  It could be a
performance hit for apps that animate.
2010-12-03 14:23:04 +08:00
Chia-I Wu
29bea39fde st/vega: Set pipe_resource::array_size to 1. 2010-12-03 14:22:32 +08:00
Chia-I Wu
9028f24b8a st/egl: Set pipe_resource::array_size to 1. 2010-12-03 14:22:32 +08:00
Marek Olšák
a60a5b850b r300g: do not remove unused constants if we are not near the limit 2010-12-03 06:32:10 +01:00
Marek Olšák
b088b255ec r300g: fix pointer arithmetic with void* in transfer_inline_write 2010-12-03 06:08:50 +01:00
Marek Olšák
d531f9c2f5 mesa, st/mesa: fix gl_FragCoord with FBOs in Gallium
gl_FragCoord.y needs to be flipped upside down if a FBO is bound.

This fixes:
- piglit/fbo-fragcoord
- https://bugs.freedesktop.org/show_bug.cgi?id=29420

Here I add a new program state STATE_FB_WPOS_Y_TRANSFORM, which is set based
on whether a FBO is bound. The state contains a pair of transformations.
It can be either (XY=identity, ZW=transformY) if a FBO is bound,
or (XY=transformY, ZW=identity) otherwise, where identity = (1, 0),
transformY = (-1, height-1).

A classic driver (or st/mesa) may, based on some other state, choose whether
to use XY or ZW, thus negate the conditional "if (is a FBO bound) ...".
The reason for this is that a Gallium driver is allowed to only support WPOS
relative to either the lower left or the upper left corner, so we must flip
the Y axis accordingly again. (the "invert" parameter in emit_wpos_inversion)

NOTE: This is a candidate for the 7.9 branch.

Signed-off-by: Marek Olšák <maraeo@gmail.com>
Signed-off-by: Brian Paul <brianp@vmware.com>
2010-12-03 05:56:40 +01:00
Marek Olšák
6a46fce14f r300g: implement simple transfer_inline_write for buffers
r600g might need something like that as well. This speeds up constant buffer
upload a bit.
2010-12-03 05:44:47 +01:00
Marek Olšák
3ba8843307 r300g: use internal BO handle for add_buffer and write_reloc
Small perf improvement in ipers.

radeon_drm_get_cs_handle is exactly what this commit tries to avoid
in every write_reloc.
2010-12-03 04:40:22 +01:00
Brian Paul
6299f241e9 gallivm/llvmpipe: remove lp_build_context::builder
The field was redundant.  Use the gallivm->builder value instead.
2010-12-02 18:11:16 -07:00
Brian Paul
36b09b5ded mesa: replace more MAX_WIDTH stack allocations with heap allocations 2010-12-02 18:07:03 -07:00
Marek Olšák
0c9a63db09 r300g: fix build 2010-12-03 00:54:41 +01:00
nobled
39c436a1a2 r300g: Drop unnecessary cast 2010-12-03 00:50:58 +01:00
nobled
7ed65f462a r300g: Abort if draw_create() fails
The other drivers need to be updated to do this, too.
2010-12-03 00:50:58 +01:00
nobled
d5bf231806 r300g: Abort if atom allocations fail 2010-12-03 00:50:58 +01:00
Ben Skeggs
a25cea19f2 nv50: silence some unknown get_param warnings
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
2010-12-03 08:23:44 +10:00
Brian Paul
77ea102735 st/mesa: avoid large stack allocations in readpixels code 2010-12-02 14:29:07 -07:00
Brian Paul
aa28efe60d swrast: avoid large stack allocations in tex combine code 2010-12-02 14:29:07 -07:00
Brian Paul
25662f878e swrast: avoid large stack allocations in blend code 2010-12-02 14:29:07 -07:00
Brian Paul
2addcb7b50 mesa: replace large/MAX_WIDTH stack allocations with heap allocations 2010-12-02 14:29:01 -07:00
Brian Paul
896a08bde6 mesa: replace large/MAX_WIDTH stack allocations with heap allocations 2010-12-02 14:26:55 -07:00
Alex Deucher
fae7cb8ed8 r600g: bump texture/cb limits appropriately for evergreen 2010-12-02 16:11:08 -05:00
Alex Deucher
a3dc947057 r600c: bump texture limits to hw limits 2010-12-02 16:11:07 -05:00
Zack Rusin
e737b9294a gallium/util: add states relevant to geometry shaders 2010-12-02 15:50:15 -05:00
José Fonseca
64da9a1a04 mesa: Temporary hack to prevent stack overflow on windows
e.g. st_readpixels is trying to alloca() an huge ammount of memory from
the stack.
2010-12-02 20:30:59 +00:00
José Fonseca
63df5a464e wgl: Fix visual's buffer_mask configuration. 2010-12-02 20:21:39 +00:00
José Fonseca
43121f2086 mapi: Hack to avoid vgCreateFont being generated as vgCreateFontA.
Right fix is probably stop C-preprocessor abuse and stick 100% with
scripted code generation.
2010-12-02 19:39:55 +00:00
Eric Anholt
43491adc44 mesa: Add getters for ARB_copy_buffer's attachment points.
Fixes more complaints by oglconform.
2010-12-02 10:28:55 -08:00
Eric Anholt
7cba339375 mesa: Add getters for the rest of the supported draw buffers.
MAX_DRAW_BUFFERS is 8, so allow all 8 GL_DRAW_BUFFER# to be retrieved.
Fixes complaints by oglconform.
2010-12-02 10:28:52 -08:00
Eric Anholt
b381eff141 glsl: Fix flipped return of has_value() for array constants.
Fixes glsl-array-uniform.
2010-12-02 10:28:51 -08:00
José Fonseca
e3659329e0 WIN32_THREADS -> WIN32
Fixes nasty bug where some parts of the code didn't define WIN32_THREADS
and were using the integer mutex implementation, causing even confusion
to the debuggers.

And there is little interest of other thread implemenation on Win32
besides Win32 threads.
2010-12-02 17:35:03 +00:00
Brian Paul
af4f75c8fe softpipe: increase max texture size to 16K 2010-12-02 10:10:05 -07:00
Brian Paul
4b08f35487 mesa: raise max texture sizes to 16K
This allows 16K x 16K 2D textures, for example, but we don't want to
allow that for 3D textures.  The new gl_constants::MaxTextureMBytes
field is used to prevent allocating too large of texture image.
This allows a 16K x 32 x 32 3D texture, for example, but prevents 16K^3.
Drivers can override this limit.  The default is currently 1GB.

Apps should use the proxy texture mechanism to determine the actual
max texture size.
2010-12-02 10:09:03 -07:00
Marek Olšák
23390e2f5c r300/compiler: disable the swizzle lowering pass in vertex shaders
It was a no-op because all swizzles are native there.
2010-12-02 17:48:08 +01:00
José Fonseca
14e2dc9c66 wgl: Unreference the current framebuffer after the make_current call.
To prevent a dangling pointer dereference.
2010-12-02 16:28:36 +00:00
José Fonseca
63c05c96e7 util: Don't try to use imagehlp on mingw. 2010-12-02 15:14:58 +00:00
José Fonseca
50a52ba67e util: __builtin_frame_address() doesn't work on mingw. 2010-12-02 15:14:58 +00:00
José Fonseca
744ef8721b util: Plug leaks in util_destroy_gen_mipmap. 2010-12-02 15:14:58 +00:00
José Fonseca
e5ffa9aa47 wgl: Fix double free. Remove dead code. 2010-12-02 15:14:58 +00:00
Marek Olšák
f3021c688f r300g: fix up cubemap texture offset computation
Broken since 4c70014626.
2010-12-02 14:44:28 +01:00
José Fonseca
431c478df9 util: C++ safe. 2010-12-02 12:26:55 +00:00
José Fonseca
27ba2eb33b retrace: Some fixes. 2010-12-02 12:17:07 +00:00
Chia-I Wu
cb2791213a st/vega: polygon_array requires a deep free.
A polygon array is an array of pointers to polygons.  The polygons
should be freed with the array.
2010-12-02 17:54:23 +08:00
Chia-I Wu
b950d6fa5d st/vega: Destroy the pipe context with vg_context. 2010-12-02 17:27:38 +08:00
Chad Versace
7528f143df glsl: Fix linker bug in cross_validate_globals()
Cause linking to fail if a global has mismatching invariant qualifiers.

See https://bugs.freedesktop.org/show_bug.cgi?id=30261
2010-12-01 20:40:07 -08:00
Roland Scheidegger
4c70014626 gallium: support for array textures and related changes
resources have a array_size parameter now.
get_tex_surface and tex_surface_destroy have been renamed to create_surface
and surface_destroy and moved to context, similar to sampler views (and
create_surface now uses a template just like create_sampler_view). Surfaces
now really should only be used for rendering. In particular they shouldn't be
used as some kind of 2d abstraction for sharing a texture. offset/layout fields
don't make sense any longer and have been removed, width/height should go too.
surfaces and sampler views now specify a layer range (for texture resources),
layer is either array slice, depth slice or cube face.
pipe_subresource is gone array slices (or cube faces) are now treated the same
as depth slices in transfers etc. (that is, they use the z coord of the
respective functions).

Squashed commit of the following:

commit a45bd509014743d21a532194d7b658a1aeb00cb7
Merge: 1aeca28 32e1e59
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Thu Dec 2 04:32:06 2010 +0100

    Merge remote branch 'origin/master' into gallium-array-textures

    Conflicts:
    	src/gallium/drivers/i915/i915_resource_texture.c
    	src/gallium/drivers/i915/i915_state_emit.c
    	src/gallium/drivers/i915/i915_surface.c

commit 1aeca287a827f29206078fa1204715a477072c08
Merge: 912f042 6f7c8c3
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Thu Dec 2 00:37:11 2010 +0100

    Merge remote branch 'origin/master' into gallium-array-textures

    Conflicts:
    	src/gallium/state_trackers/vega/api_filters.c
    	src/gallium/state_trackers/vega/api_images.c
    	src/gallium/state_trackers/vega/mask.c
    	src/gallium/state_trackers/vega/paint.c
    	src/gallium/state_trackers/vega/renderer.c
    	src/gallium/state_trackers/vega/st_inlines.h
    	src/gallium/state_trackers/vega/vg_context.c
    	src/gallium/state_trackers/vega/vg_manager.c

commit 912f042e1d439de17b36be9a740358c876fcd144
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Wed Dec 1 03:01:55 2010 +0100

    gallium: even more compile fixes after merge

commit 6fc95a58866d2a291def333608ba9c10c3f07e82
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Wed Dec 1 00:22:26 2010 +0100

    gallium: some fixes after merge

commit a8d5ffaeb5397ffaa12fb422e4e7efdf0494c3e2
Merge: f7a202f 2da02e7
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Tue Nov 30 23:41:26 2010 +0100

    Merge remote branch 'origin/master' into gallium-array-textures

    Conflicts:
    	src/gallium/drivers/i915/i915_state_emit.c
    	src/gallium/state_trackers/vega/api_images.c
    	src/gallium/state_trackers/vega/vg_context.c

commit f7a202fde2aea2ec78ef58830f945a5e214e56ab
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Wed Nov 24 19:19:32 2010 +0100

    gallium: even more fixes/cleanups after merge

commit 6895a7f969ed7f9fa8ceb788810df8dbcf04c4c9
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Wed Nov 24 03:07:36 2010 +0100

    gallium: more compile fixes after merge

commit af0501a5103b9756bc4d79167bd81051ad6e8670
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Tue Nov 23 19:24:45 2010 +0100

    gallium: lots of compile fixes after merge

commit 0332003c2feb60f2a20e9a40368180c4ecd33e6b
Merge: 26c6346 b6b91fa
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Tue Nov 23 17:02:26 2010 +0100

    Merge remote branch 'origin/master' into gallium-array-textures

    Conflicts:
    	src/gallium/auxiliary/gallivm/lp_bld_sample.c
    	src/gallium/auxiliary/util/u_blit.c
    	src/gallium/auxiliary/util/u_blitter.c
    	src/gallium/auxiliary/util/u_inlines.h
    	src/gallium/auxiliary/util/u_surface.c
    	src/gallium/auxiliary/util/u_surfaces.c
    	src/gallium/docs/source/context.rst
    	src/gallium/drivers/llvmpipe/lp_rast.c
    	src/gallium/drivers/nv50/nv50_state_validate.c
    	src/gallium/drivers/nvfx/nv04_surface_2d.c
    	src/gallium/drivers/nvfx/nv04_surface_2d.h
    	src/gallium/drivers/nvfx/nvfx_buffer.c
    	src/gallium/drivers/nvfx/nvfx_miptree.c
    	src/gallium/drivers/nvfx/nvfx_resource.c
    	src/gallium/drivers/nvfx/nvfx_resource.h
    	src/gallium/drivers/nvfx/nvfx_state_fb.c
    	src/gallium/drivers/nvfx/nvfx_surface.c
    	src/gallium/drivers/nvfx/nvfx_transfer.c
    	src/gallium/drivers/r300/r300_state_derived.c
    	src/gallium/drivers/r300/r300_texture.c
    	src/gallium/drivers/r600/r600_blit.c
    	src/gallium/drivers/r600/r600_buffer.c
    	src/gallium/drivers/r600/r600_context.h
    	src/gallium/drivers/r600/r600_screen.c
    	src/gallium/drivers/r600/r600_screen.h
    	src/gallium/drivers/r600/r600_state.c
    	src/gallium/drivers/r600/r600_texture.c
    	src/gallium/include/pipe/p_defines.h
    	src/gallium/state_trackers/egl/common/egl_g3d_api.c
    	src/gallium/state_trackers/glx/xlib/xm_st.c
    	src/gallium/targets/libgl-gdi/gdi_softpipe_winsys.c
    	src/gallium/targets/libgl-gdi/libgl_gdi.c
    	src/gallium/tests/graw/tri.c
    	src/mesa/state_tracker/st_cb_blit.c
    	src/mesa/state_tracker/st_cb_readpixels.c

commit 26c6346b385929fba94775f33838d0cceaaf1127
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Mon Aug 2 19:37:21 2010 +0200

    fix more merge breakage

commit b30d87c6025eefe7f6979ffa8e369bbe755d5c1d
Merge: 9461bf3 1f1928d
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Mon Aug 2 19:15:38 2010 +0200

    Merge remote branch 'origin/master' into gallium-array-textures

    Conflicts:
    	src/gallium/drivers/llvmpipe/lp_rast.c
    	src/gallium/drivers/llvmpipe/lp_rast_priv.h
    	src/gallium/drivers/r300/r300_blit.c
    	src/gallium/drivers/r300/r300_screen_buffer.c
    	src/gallium/drivers/r300/r300_state_derived.c
    	src/gallium/drivers/r300/r300_texture.c
    	src/gallium/drivers/r300/r300_texture.h
    	src/gallium/drivers/r300/r300_transfer.c
    	src/gallium/drivers/r600/r600_screen.c
    	src/gallium/drivers/r600/r600_state.c
    	src/gallium/drivers/r600/r600_texture.c
    	src/gallium/drivers/r600/r600_texture.h
    	src/gallium/state_trackers/dri/common/dri1_helper.c
    	src/gallium/state_trackers/dri/sw/drisw.c
    	src/gallium/state_trackers/xorg/xorg_exa.c

commit 9461bf3cfb647d2301364ae29fc3084fff52862a
Merge: 17492d7 0eaccb3
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Thu Jul 15 20:13:45 2010 +0200

    Merge commit 'origin/master' into gallium-array-textures

    Conflicts:
    	src/gallium/auxiliary/util/u_blitter.c
    	src/gallium/drivers/llvmpipe/lp_rast.c
    	src/gallium/drivers/llvmpipe/lp_surface.c
    	src/gallium/drivers/r300/r300_render.c
    	src/gallium/drivers/r300/r300_state.c
    	src/gallium/drivers/r300/r300_texture.c
    	src/gallium/drivers/r300/r300_transfer.c
    	src/gallium/tests/trivial/quad-tex.c

commit 17492d705e7b7f607b71db045c3bf344cb6842b3
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Fri Jun 18 10:58:08 2010 +0100

    gallium: rename element_offset/width fields in views to first/last_element

    This is much more consistent with the other fields used there
    (first/last level, first/last layer).
    Actually thinking about removing the ugly union/structs again and
    rename first/last_layer to something even more generic which could also
    be used for buffers (like first/last_member) without inducing headaches.

commit 1b717a289299f942de834dcccafbab91361e20ab
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Thu Jun 17 14:46:09 2010 +0100

    gallium: remove PIPE_SURFACE_LAYOUT_LINEAR definition

    This was only used by the layout field of pipe_surface, but this
    driver internal stuff is gone so there's no need for this driver independent
    layout definition neither.

commit 10cb644b31b3ef47e6c7b55e514ad24bb891fac4
Merge: 5691db9 c85971d
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Thu Jun 17 12:20:41 2010 +0100

    Merge commit 'origin/master' into gallium-array-textures

    Conflicts:
    	src/gallium/docs/source/glossary.rst
    	src/gallium/tests/graw/fs-test.c
    	src/gallium/tests/graw/gs-test.c

commit 5691db960ca3d525ce7d6c32d9c7a28f5e907f3b
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Thu Jun 17 11:29:03 2010 +0100

    st/wgl: fix interface changes bugs

commit 2303ec32143d363b46e59e4b7c91b0ebd34a16b2
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Wed Jun 16 19:42:32 2010 +0100

    gallium: adapt code to interface changes...

commit dcae4f586f0d0885b72674a355e5d56d47afe77d
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Wed Jun 16 19:42:05 2010 +0100

    gallium: separate depth0 and array_size in the resource itself.

    These fields are still mutually exclusive (since no 3d array textures exist)
    but it ultimately seemed to error-prone to adapt all code accept the new
    meaning of depth0 (drivers stick that into hardware regs, calculate mipmap
    sizes etc.). And it isn't really cleaner anyway.
    So, array textures will have depth0 of 1, but instead use array_size,
    3D textures will continue to use depth0 (and have array_size of 1). Cube
    maps also will use array_size to indicate their 6 faces, but since all drivers
    should just be fine by inferring this themselves from the fact it's a cube map
    as they always used to nothing should break.

commit 621737a638d187d208712250fc19a91978fdea6b
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Wed Jun 16 17:47:38 2010 +0100

    gallium: adapt code to interface changes

    There are still usages of pipe_surface where pipe_resource should be used,
    which should eventually be fixed.

commit 2d17f5efe166b2c3d51957c76294165ab30b8ae2
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Wed Jun 16 17:46:14 2010 +0100

    gallium: more interface changes

    In particular to enable usage of buffers in views, and ability to use a
    different pipe_format in pipe_surface.
    Get rid of layout and offset parameter in pipe_surface - the former was
    not used in any (public) code anyway, and the latter should either be computed
    on-demand or driver can use subclass of pipe_surface.
    Also make create_surface() use a template to be more consistent with
    other functions.

commit 71f885ee16aa5cf2742c44bfaf0dc5b8734b9901
Merge: 3232d11 8ad410d
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Mon Jun 14 14:19:51 2010 +0100

    Merge commit 'origin/master' into gallium-array-textures

    Conflicts:
    	src/gallium/auxiliary/util/u_box.h
    	src/gallium/drivers/nv50/nv50_surface.c
    	src/gallium/drivers/nvfx/nvfx_surface.c
    	src/gallium/drivers/r300/r300_blit.c
    	src/gallium/drivers/r300/r300_texture.c
    	src/gallium/drivers/r300/r300_transfer.c
    	src/gallium/drivers/r600/r600_blit.c
    	src/gallium/drivers/r600/r600_screen.h
    	src/gallium/include/pipe/p_state.h

commit 3232d11fe3ebf7686286013c357b404714853984
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Mon Jun 14 11:40:04 2010 +0100

    mesa/st: adapt to interface changes

    still need to fix pipe_surface sharing
    (as that is now per-context).
    Also broken is depth0 handling - half the code assumes
    this is also used for array textures (and hence by extension
    of that cube maps would have depth 6), half the code does not...

commit f433b7f7f552720e5eade0b4078db94590ee85e1
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Mon Jun 14 11:35:52 2010 +0100

    gallium: fix a couple of bugs in interface chnage fixes

commit 818366b28ea18f514dc791646248ce6f08d9bbcf
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:42:11 2010 +0200

    targets: adapt to interface changes

    Yes even that needs adjustments...

commit 66c511ab1682c9918e0200902039247793acb41e
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:41:13 2010 +0200

    tests: adapt to interface changes

    Everything needs to be fixed :-(.

commit 6b494635d9dbdaa7605bc87b1ebf682b138c5808
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:39:50 2010 +0200

    st: adapt non-rendering state trackers to interface changes

    might not be quite right in all places, but they really don't want
    to use pipe_surface.

commit 00c4289a35d86e4fe85919ec32aa9f5ffe69d16d
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:38:48 2010 +0200

    winsys: adapt to interface changes

commit 39d858554dc9ed5dbc795626fec3ef9deae552a0
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:26:54 2010 +0200

    st/python: adapt to interface changes

    don't think that will work, sorry.

commit 6e9336bc49b32139cec4e683857d0958000e15e3
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:26:07 2010 +0200

    st/vega: adapt to interface changes

commit e07f2ae9aaf8842757d5d50865f76f8276245e11
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:25:56 2010 +0200

    st/xorg: adapt to interface changes

commit 05531c10a74a4358103e30d3b38a5eceb25c947f
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:24:53 2010 +0200

    nv50: adapt to interface changes

commit 97704f388d7042121c6d496ba8c003afa3ea2bf3
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:24:45 2010 +0200

    nvfx: adapt to interface changes

commit a8a9c93d703af6e8f5c12e1cea9ec665add1abe0
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:24:01 2010 +0200

    i965g: adapt to interface changes

commit 0dde209589872d20cc34ed0b237e3ed7ae0e2de3
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:22:38 2010 +0200

    i915g: adapt to interface changes

commit 5cac9beede69d12f5807ee1a247a4c864652799e
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:20:58 2010 +0200

    svga: adapt to interface changes

    resource_copy_region still looking fishy.
    Was not very suited to unified zslice/face approach...

commit 08b5a6af4b963a3e4c75fc336bf6c0772dce5150
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:20:01 2010 +0200

    rbug: adapt to interface changes

    Not sure if that won't need changes elsewhere?

commit c9fd24b1f586bcef2e0a6e76b68e40fca3408964
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:19:31 2010 +0200

    trace: adapt to interface changes

commit ed84e010afc5635a1a47390b32247a266f65b8d1
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:19:21 2010 +0200

    failover: adapt to interface changes

commit a1d4b4a293da933276908e3393435ec4b43cf201
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:19:12 2010 +0200

    identity: adapt to interface changes

commit a8dd73e2c56c7d95ffcf174408f38f4f35fd2f4c
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:18:55 2010 +0200

    softpipe: adapt to interface changes

commit a886085893e461e8473978e8206ec2312b7077ff
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:18:44 2010 +0200

    llvmpipe: adapt to interface changes

commit 70523f6d567d8b7cfda682157556370fd3c43460
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:18:14 2010 +0200

    r600g: adapt to interface changes

commit 3f4bc72bd80994865eb9f6b8dfd11e2b97060d19
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:18:05 2010 +0200

    r300g: adapt to interface changes

commit 5d353b55ee14db0ac0515b5a3cf9389430832c19
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:17:37 2010 +0200

    cell: adapt to interface changes

    not even compile tested

commit cf5d03601322c2dcb12d7a9c2f1745e2b2a35eb4
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:14:59 2010 +0200

    util: adapt to interface changes

    amazing how much code changes just due to some subtle interface changes?

commit dc98d713c6937c0e177fc2caf23020402cc7ea7b
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Sat Jun 12 02:12:40 2010 +0200

    gallium: more interface fail, docs

    this also changes flush_frontbuffer to use a pipe_resource instead of
    a pipe_surface - pipe_surface is not meant to be (or at least no longer)
    an abstraction for standalone 2d images which get passed around.
    (This has also implications for the non-rendering state-trackers.)

commit 08436d27ddd59857c22827c609b692aa0c407b7b
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Thu Jun 10 17:42:52 2010 +0200

    gallium: fix array texture interface changes bugs, docs

commit 4a4d927609b62b4d7fb9dffa35158afe282f277b
Author: Roland Scheidegger <sroland@vmware.com>
Date:   Thu Jun 3 22:02:44 2010 +0200

    gallium: interface changes for array textures and related cleanups

    This patch introduces array textures to gallium (note they are not immediately
    usable without the associated changes to the shader side).
    Also, this abandons pipe_subresource in favor of using level and layer
    parameters since the distinction between several faces (which was part of
    pipe_subresource for cube textures) and several z slices (which were not part
    of pipe_subresource but instead part of pipe_box where appropriate for 3d
    textures) is gone at the resource level.
    Textures, be it array, cube, or 3d, now use a "unified" set of parameters,
    there is no distinction between array members, cube faces, or 3d zslices.
    This is unlike d3d10, whose subresource index includes layer information for
    array textures, but which considers all z slices of a 3d texture to be part
    of the same subresource.
    In contrast to d3d10, OpenGL though reuses old 2d and 3d function entry points
    for 1d and 2d array textures, respectively, which also implies that for instance
    it is possible to specify all layers of a 2d array texture at once (note that
    this is not possible for cube maps, which use the 2d entry points, although
    it is possible for cube map arrays, which aren't supported yet in gallium).
    This should possibly make drivers a bit simpler, and also get rid of mutually
    exclusive parameters in some functions (as z and face were exclusive), one
    potential downside would be that 3d array textures could not easily be supported
    without reverting this, but those are nowhere to be seen.

    Also along with adjusting to new parameters, rename get_tex_surface /
    tex_surface_destroy to create_surface / surface_destroy and move them from
    screen to context, which reflects much better what those do (they are analogous
    to create_sampler_view / sampler_view_destroy).

    PIPE_CAP_ARRAY_TEXTURES is used to indicate if a driver supports all of this
    functionality (that is, both sampling from array texture as well as use a range
    of layers as a render target, with selecting the layer from the geometry shader).
2010-12-02 04:33:43 +01:00
Xiang, Haihao
32e1e59146 i965: add support for polygon mode on Sandybridge.
This fixes some mesa demos such as fslight/engine in wireframe mode.
2010-12-02 09:54:35 +08:00
Jakob Bornecrantz
de3ff5af49 i915g: Make sure that new vbo gets updated
Malloc likes to reuse old address as soon as possible this would cause the
new vbo buffer to get the same address as the old. So make sure we set it to
NULL when we allocate a new one. This fixes ipers which will fill up a couple
of VBO buffers per frame.

Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-12-02 01:34:14 +01:00
Jakob Bornecrantz
442e567aa0 i915g: Improve debug printing for textures
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-12-02 01:34:14 +01:00
Chris Wilson
f6476822a0 i915g: Fix closure of full batch buffers
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
[danvet: incorporate comments by Dr_Jakob]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-12-02 01:34:14 +01:00
Daniel Vetter
600454084b i915g: track TODO items
Just as a reminder for all things currently broken with i915g.

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-12-02 01:34:14 +01:00
Daniel Vetter
0246b2bf27 i915g: assert(depth_surface->offset == 0)
Shouldn't happen and not supported, anyway.

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-12-02 01:34:14 +01:00
Daniel Vetter
1a47a25d2c i915g: enable x-tiling for render targets
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-12-02 01:34:13 +01:00
Daniel Vetter
8af684e37e i915g: switch rendering to mipmapped textures to (x,y) offsets
Byte offsets simply don't work with tiled render targets when using
tiling bits. Luckily we can cox the hw into doing the right thing
with the DRAWING_RECT command by disabling the drawing rect offset
for the depth buffer.

Minor fixes by Jakob Bornecrantz.

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-12-02 01:34:13 +01:00
Daniel Vetter
9493fe85d1 i915g: enable X-tiling for textures
Tiling is rather fragile in general and results in pure blackness when
unlucky.  Hence add a new option to disable tiling.

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-12-02 01:34:13 +01:00
Daniel Vetter
32345610cc i915g: don't pot-align stride for tiled buffers
libdrm will do this for us, if it's required (i.e. if tiling is
possible).

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-12-02 01:34:13 +01:00
Daniel Vetter
6ae6e0c6fe i915g: postpone mipmap/face offset calculation
libdrm-intel can refuse to tile buffers for various reasons. For
potentially tiled buffers the stride is therefore only known after
the iws->buffer_create_tiled call. Unconditionally rounding up to whatever
tiling requires wastes space, so rework the code to not use tex->stride
in the layout code.

Luckily only the mimap/face offset calculation uses it which can easily
be solved by storing an (x, y) coordinate pair. Furthermore this will
be usefull later for properly supporting rendering into the different
levels of tiled mipmap textures.

v2: switch to nblocks(x|y): More in line with gallium and better
suited for rendering into mipmap textures.

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-12-02 01:34:13 +01:00
Daniel Vetter
f34fd58ec9 i915g: implement unfenced relocs for textures using tiling bits
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-12-02 01:34:13 +01:00
Daniel Vetter
a95e694eaf i915g: implement unfenced color&depth buffer using tiling bits
v2: Clarify tiling bit calculation as suggested by Chris Wilson.

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-12-02 01:34:13 +01:00
Daniel Vetter
2ff0879a63 i915g: return tiling in iws->buffer_from_handle
This is needed to properly implement tiling flags. And the gem
implemention fo buffer_from_handle already calls get_tiling, so
it's for free.

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-12-02 01:34:13 +01:00
Daniel Vetter
135b083461 i915g: prepare winsys/batchbuffer for execbuf2
Wire up a fenced parameter, switch all relocations to _FENCED

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-12-02 01:34:13 +01:00
Daniel Vetter
1c60840338 i915g: switch to tiled allocations, kill set_fence
This way relaxed fencing is handled by libdrm. And buffers _can't_
ever change their tiling.

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-12-02 01:34:13 +01:00
Daniel Vetter
4a666488c4 i915g: add winsys function to create tiled buffers
Different kernels have different restrictions for tiled buffers.
Hence use the libdrm abstraction to calculate the necessary
stride and height alignment requirements.

Not yet used.

v2: Incorporate review comments from Jakob Bornecrantz

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-12-02 01:34:12 +01:00
Daniel Vetter
c62f5c7e7b i915g: drop alignment parameter from iws->buffer_create
It's unnecessary. The kernel gem ignores it totally and we can't
run on the old userspace fake bo manager due to lack of dri2.

Also drop the redundant name string from the sw winsys as suggested
by Jakob Bornecrantz

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-12-02 01:34:12 +01:00
Eric Anholt
b4f585665c glsl: Mark the array access for whole-array comparisons.
By not doing so, the uniform contents of
glsl-uniform-non-uniform-array-compare.shader_test was getting thrown
out since nobody was recorded as dereferencing the array.
2010-12-01 16:14:34 -08:00
Eric Anholt
f361f8a8a9 i965: Add support for loops in the VS.
This follows the changes done for the FS alongside the EU emit code.
2010-12-01 16:14:34 -08:00
Eric Anholt
251d15d888 i965: Enable IF statements in the VS.
While the actual IF instructions were fixed by Zhenyu, we were still
flattening them to conditional moves.
2010-12-01 16:14:34 -08:00
Eric Anholt
843a6a308e i965: Add support for gen6 CONTINUE instruction emit.
At this point, piglit tests for fragment shader loops are working.
2010-12-01 16:14:34 -08:00
Eric Anholt
00e5a743e2 i965: Add support for gen6 BREAK ISA emit.
There are now two targets: the hop-to-end-of-block target, and the
target for where to resume execution for active channels.
2010-12-01 16:14:33 -08:00
Eric Anholt
4890e0f09c i965: Add support for gen6 DO/WHILE ISA emit.
There's no more DO since there's no more mask stack, and WHILE has
been shuffled like IF was.
2010-12-01 16:14:31 -08:00
Eric Anholt
a9f62881a3 i965: Dump the WHILE jump distance on gen6. 2010-12-01 15:22:59 -08:00
Marek Olšák
fcf6b353bf r300g: disable ARB_texture_swizzle if S3TC is enabled on r3xx-only
r3xx cannot swizzle compressed textures. r4xx+ is unaffected.

NOTE: This is a candidate for the 7.9 branch.
2010-12-01 22:54:05 +01:00
Marek Olšák
6478a4de14 r300g: fix texture swizzling with compressed textures on r400-r500
This fixes all S3TC piglit/texwrap tests.

NOTE: This is a candidate for the 7.9 branch.
2010-12-01 22:29:09 +01:00
Ian Romanick
c92550be64 i915: Correctly generate unconditional KIL instructions
Fixes piglit test glsl-fs-discard-03.

NOTE: This is a candidate for the 7.9 branch.
2010-12-01 12:01:13 -08:00
Ian Romanick
b6dbc06742 i915: Request that POW instructions be lowered 2010-12-01 12:01:13 -08:00
Ian Romanick
c4285be9a5 glsl: Lower ir_binop_pow to a sequence of EXP2 and LOG2 2010-12-01 12:01:13 -08:00
Ian Romanick
da61afa738 glsl: Use M_LOG2E constant instead of calling log2 2010-12-01 12:01:12 -08:00
Kenneth Graunke
d2d7a273c5 glsl: Add comments to lower_jumps (from the commit message).
This is essentially Luca's commit message, but placed at the top of the
file.
2010-12-01 11:52:43 -08:00
Kenneth Graunke
1802cb9baf glsl: Remove "discard" support from lower_jumps.
The new lower_discard and opt_discard_simplification passes should
handle all the necessary transformations, so lower_jumps doesn't need to
support it.

Also, lower_jumps incorrectly handled conditional discards - it would
unconditionally truncate all code after the discard.  Rather than fixing
the bug, simply remove the code.

NOTE: This is a candidate for the 7.9 branch.
2010-12-01 11:52:43 -08:00
Kenneth Graunke
940df10100 glsl: Add a lowering pass to move discards out of if-statements.
This should allow lower_if_to_cond_assign to work in the presence of
discards, fixing bug #31690 and likely #31983.

NOTE: This is a candidate for the 7.9 branch.
2010-12-01 11:52:43 -08:00
Kenneth Graunke
9a1d063c6d glsl: Add an optimization pass to simplify discards.
NOTE: This is a candidate for the 7.9 branch.
2010-12-01 11:52:43 -08:00
Marek Olšák
ead2ea89f4 ir_to_mesa: Add support for conditional discards.
NOTE: This is a candidate for the 7.9 branch.

Signed-off-by: Marek Olšák <maraeo@gmail.com>
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
2010-12-01 11:52:36 -08:00
Alex Deucher
2ca9256911 r600c: fix some opcodes on evergreen
There were a few places where we were using the wrong opcodes
on evergreen.  arl still needs to be fixed on evergreen; see
r600g for reference.

NOTE: This is a candidate for the 7.9 branch.

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
2010-12-01 13:26:02 -05:00
Marek Olšák
e6d798948e r300/compiler: implement and lower OPCODE_CLAMP
Needed for st/vega.
2010-12-01 18:29:50 +01:00
José Fonseca
6f7c8c3cbf vega: Remove extraneous ; 2010-12-01 12:31:21 +00:00
José Fonseca
792caebced scons: Move MSVS_VERSION option to common module. 2010-12-01 12:23:12 +00:00
José Fonseca
2aa3203660 svga: Silence debug printf. 2010-12-01 12:23:12 +00:00
Chia-I Wu
0dadc0b808 st/vega: Avoid unnecessary constant bufer upload.
Remember the last uploaded data and avoid re-uploading.
2010-12-01 18:36:27 +08:00
Chia-I Wu
d7a6901cac st/vega: Initialize pipe states with renderer.
Initialize vertex elements, rasterizer, stencil ref, and vertex shader
with renderer_create.  Remove RASTERIZER_DIRTY and VS_DIRTY flags.
2010-12-01 18:07:00 +08:00
Chia-I Wu
c91c386012 st/vega: Create drawing surface mask as needed.
As the blend texture, a drawing surface mask is used when masking is
enabled.  It should be created as needed.

s/alpha_mask/surface_mask/ to follow OpenVG 1.1 naming.
2010-12-01 18:03:06 +08:00
Chia-I Wu
04f342b417 st/vega: Delay blend texture creation until needed.
It is used for more advanced blending or mask update.  It might not be
ever needed for some applications.
2010-12-01 17:46:34 +08:00
Chia-I Wu
f8e0dd246b st/vega: Remove st_inlines.h.
Per b0427bedde.
2010-12-01 17:00:22 +08:00
Chia-I Wu
2bb788ccc6 st/vega: Simplify radial gradient.
Eight less instructions with comments.
2010-12-01 16:46:01 +08:00
Chia-I Wu
d7aa03b4fe st/vega: Fix degenerate paints.
Fix the case that the two points of a linear gradient coincide, or the
case that the radius of a radial gradient is equal to or less than 0.
2010-12-01 16:46:01 +08:00
Zhenyu Wang
c530fd3f8a i965: also using align1 mode for math2 on sandybridge
Like Eric's workaround patch of commit 490c23ee6b.
This forces to align1 mode for math2 too.
2010-12-01 15:04:18 +08:00
Chia-I Wu
06e7a55028 st/vega: Fix negated logic in image_draw.
A typo from last commit.
2010-12-01 11:39:33 +08:00
Chia-I Wu
b06de80843 st/vega: Fix paint coordinates transformations.
Depending on whether vgDrawPath(mode), vgDrawImage, or vgDrawGlyph[s] is
called, different paint-to-user and user-to-surface matrices should be
used to derive the sample points for the paint.

This fixes "paint" demo.
2010-12-01 11:31:00 +08:00
Chia-I Wu
ca8bc9c05b st/vega: Bump version to 1.1. 2010-12-01 11:23:53 +08:00
Chia-I Wu
e360f91f15 st/vega: Add color transformation support.
Per OpenVG 1.1.  A new shader stage is added.  It uses the first two
constants of the fragment shader for color transformation parameters.
2010-12-01 11:23:52 +08:00
Chia-I Wu
213e288e78 st/vega: More flexible shader selection.
Divide bits of VegaShaderType into 6 groups: paint, image, mask, fill,
premultiply, and bw.  Each group represents a stage.  At most one shader
from each group will be selected when constructing the final fragment
shader.
2010-12-01 11:23:52 +08:00
Chia-I Wu
30cab4b6cb st/vega: Revive mask layer support. 2010-12-01 11:23:52 +08:00
Chia-I Wu
5d64a06a63 st/vega: Add primitive text support.
Optional features such as auth-hinting are not implemented.  There is no
anti-aliasing, and no effort is done to keep the glyph origin integral.
So the text quality is poor.
2010-12-01 11:23:52 +08:00
Chia-I Wu
34f466d4e6 st/vega: Make image_draw take a matrix. 2010-12-01 11:23:52 +08:00
Chia-I Wu
165cb19abc st/vega: Make path_render and path_stroke take a matrix. 2010-12-01 11:23:51 +08:00
Chia-I Wu
d873f1f5b6 st/vega: Fix image sampler views for alpha-only formats.
For alpha-only VG formats, R = G = B = 1.0.
2010-12-01 11:23:51 +08:00
Chia-I Wu
56f02cedfa st/vega: Update to latest headers. 2010-12-01 11:23:51 +08:00
Chia-I Wu
20ce148c30 st/vega: Get rid of renderer_copy_texture. 2010-12-01 11:23:51 +08:00
Chia-I Wu
33ca973e7a st/vega: vg_copy_texture and vg_copy_surface should share code. 2010-12-01 11:23:51 +08:00
Chia-I Wu
4690cdfe07 st/vega: Clean up renderer fields and functions. 2010-12-01 11:23:50 +08:00
Chia-I Wu
ace4539e88 st/vega: Clean up vg_context fields and functions. 2010-12-01 11:23:50 +08:00
Chia-I Wu
635fe3e192 st/vega: vg_manager should care about only the color buffer.
Move depth/stencil buffer, blend texture view, and alpha mask view
creation to vg_context.c.
2010-12-01 11:23:50 +08:00
Chia-I Wu
ee0f1ab923 st/vega: Make shader_bind call into the renderer.
With this commit, the pipe states are entirely managed by the renderer.
The rest of the code interfaces with the renderer instead of
manipulating the states directly.
2010-12-01 11:23:50 +08:00
Chia-I Wu
b730f0fc52 st/vega: Move g3d states to renderer.
Let vg_context focus on OpenVG states and renderer focus on gallium
states.
2010-12-01 11:23:50 +08:00
Chia-I Wu
96c6637a13 st/vega: Use st_framebuffer for fb width/height.
This allows us to eventually make g3d states opaque.
2010-12-01 11:23:50 +08:00
Chia-I Wu
438359597c st/vega: Delay fb state update to vg_validate_state.
vg_manager_validate_framebuffer should mark the fb dirty and have
vg_validate_state call cso_set_framebuffer.  Rename VIEWPORT_DIRTY to
FRAMEBUFFER_DIRTY.
2010-12-01 11:23:49 +08:00
Chia-I Wu
3b71cb6ad6 st/vega: Add POLYGON_STENCIL and POLYGON_FILL renderer state.
The states are designated for polygon filling.  Polygon filling is a
two-pass process utilizing the stencil buffer.  polygon_fill and
polygon_array_fill functions are updated to make use of the state.
2010-12-01 11:23:49 +08:00
Chia-I Wu
b23f732075 st/vega: Use the renderer for vgMask.
vgMask renders to the alpha mask with special fragment shaders.  The
operation can be supported by switching the renderer to FILTER state.
2010-12-01 11:23:49 +08:00
Chia-I Wu
e5968a5355 st/vega: Add FILTER renderer state for image filtering.
The state is designated to perform image filtering.  execute_filter is
updated to make use of the state.
2010-12-01 11:23:49 +08:00
Chia-I Wu
6b241f532a st/vega: Add CLEAR renderer state for vgClear.
This state provides the ability to clear rectangles of the framebuffer
to the specified color, honoring scissoring.  vegaClear is updated to
make use of the state.
2010-12-01 11:23:49 +08:00
Chia-I Wu
54cb382ea5 st/vega: Add SCISSOR renderer state.
The state can be used to set rectangles of the depth buffer to 0.0f.
update_clip_state is changed to use the state for scissor update.
2010-12-01 11:23:49 +08:00
Chia-I Wu
e31a04ea3b st/vega: Add DRAWTEX renderer state.
This state provides glDrawTex-like function.  It can be used for
vgSetPixels.

Rather than modifying every user of the renderer, this commit instead
modifies renderer_copy_surface to use DRAWTEX or COPY state internally
depending on whether the destination is the framebuffer.
2010-12-01 11:23:48 +08:00
Chia-I Wu
59309337e4 st/vega: Overhaul renderer with renderer states.
Renderer states are high-level states to perform specific tasks.  The
renderer is initially in INIT state.  In that state, the renderer is
used for OpenVG pipeline.

This commit adds a new COPY state to the renderer.  The state is used
for copying between two pipe resources using textured drawing.  It can
be used for vgCopyImage, for example.

Rather than modifying every user of the renderer, this commit instead
modifies renderer_copy_texture to use the COPY state internally.
2010-12-01 11:23:48 +08:00
Chia-I Wu
709e57ae4f llvmpipe: Fix build errors on x86.
The errors were introduced by
efc82aef35.
2010-12-01 11:23:48 +08:00
Kristian Høgsberg
7db49853f0 docs: Fix MESA_drm_image typo 2010-11-30 21:14:50 -05:00
Brian Paul
efc82aef35 gallivm/llvmpipe: squash merge of the llvm-context branch
This branch defines a gallivm_state structure which contains the
LLVMBuilderRef, LLVMContextRef, etc.  All data structures built with
this object can be periodically freed during a "garbage collection"
operation.

The gallivm_state object has to be passed to most of the builder
functions where LLVMBuilderRef used to be used.

Conflicts:
	src/gallium/auxiliary/gallivm/lp_bld_tgsi_soa.c
	src/gallium/drivers/llvmpipe/lp_state_setup.c
2010-11-30 16:35:12 -07:00
Marek Olšák
1f1375d4d8 r300g: fix texture border color once again
I made the texwrap test be more thorough and realized that this driver code
had not been quite right. This commit fixes the border color for depth
textures, compressed textures, and 16-bits-per-channel textures
with up to 2 channels (R16, RG16).

NOTE: This is a candidate for the 7.9 branch.
2010-11-30 23:31:16 +01:00
Kenneth Graunke
2da02e75b1 glsl/linker: Free any IR discarded by optimization passes.
Previously, IR for a linked shader was allocated directly out of the
gl_shader object - meaning all of it lived as long as the shader.

Now, IR is allocated out of a temporary context, and any -live- IR is
reparented/stolen to (effectively) the gl_shader.  Any remaining IR can
be freed.

NOTE: This is a candidate for the 7.9 branch.
2010-11-30 13:48:28 -08:00
Kenneth Graunke
ff994eeff8 glsl: Remove anti-built-in hacks from the print visitor.
Now that we only import built-in signatures that are actually used,
printing them is reasonable.
2010-11-30 13:48:28 -08:00
Kenneth Graunke
f5692f452f glsl: Lazily import built-in function prototypes.
This makes a very simple 1.30 shader go from 196k of memory to 9k.

NOTE: This -may- be a candidate for the 7.9 branch, as the benefit is
substantial.  However, it's not a simple change, so it may be wiser to
wait for 7.10.
2010-11-30 13:48:28 -08:00
Kenneth Graunke
01a25bb64e glsl: Refactor out cloning of function prototypes.
This allows us to reuse some code and will be useful later.
2010-11-30 13:48:28 -08:00
Aras Pranckevicius
4ce084c707 glsl: fix matrix type check in ir_algebraic
Fixes glsl-mat-mul-1.
2010-11-30 13:32:00 -08:00
Eric Anholt
d56c97413e glsl: Quiet unreachable no-return-from-function warning. 2010-11-30 13:29:28 -08:00
Zack Rusin
b22d65e9fc scons: add alias for identity 2010-11-30 16:13:11 -05:00
Eric Anholt
ff79633d9f glsl: Fix structure and array comparisions.
We were trying to emit a single ir_expression to compare the whole
thing.  The backends (ir_to_mesa.cpp and brw_fs.cpp so far) expected
ir_binop_any_nequal or ir_binop_all_equal to apply to at most a vector
(with matrices broken down by the lowering pass).  Break them down to
a bunch of ORed or ANDed any_nequals/all_equals.

Fixes:
glsl-array-compare
glsl-array-compare-02
glsl-fs-struct-equal
glsl-fs-struct-notequal
Bug #31909
2010-11-30 11:42:42 -08:00
Eric Anholt
6b937465d4 glsl: Add a helper constructor for expressions that works out result type.
This doesn't cover all expressions or all operand types, but it will
complain if you overreach and it allows for much greater slack on the
programmer's part.
2010-11-30 11:23:24 -08:00
Keith Whitwell
68a4f63247 llvmpipe: shortcircuit some calls to set_scene_state 2010-11-30 12:01:29 +00:00
Keith Whitwell
d9169364d4 llvmpipe: remove misleading debug string 2010-11-30 12:01:29 +00:00
Keith Whitwell
2d31f048ce llvmpipe: raise dirty flag on transfers to bound constbuf
Need this to trigger the scene to update its shadow of the constant
state.
2010-11-30 12:01:29 +00:00
José Fonseca
31aeac5bf9 wgl: More complete WGL_ARB_pbuffer support. 2010-11-30 10:49:08 +00:00
José Fonseca
c4a43873c5 wgl: Stub WGL_ARB_pbuffer support.
See http://www.opengl.org/registry/specs/ARB/wgl_pbuffer.txt
2010-11-30 10:47:49 +00:00
José Fonseca
af6407a500 scons: Alias for svga 2010-11-30 10:47:26 +00:00
José Fonseca
7bbf675b88 svga: Use consistent hexadecimal representation on debug output. 2010-11-30 10:45:31 +00:00
Marek Olšák
80f24c1575 util: rename u_mempool -> u_slab 2010-11-30 10:12:26 +01:00
Zack Rusin
5572805423 gallivm: fix storing of the addr register
we store into the index specified by the register index, not an
indirect register.
2010-11-30 02:01:43 -05:00
Eric Anholt
001eee52d4 glsl: Make the symbol table's add_variable just use the variable's name. 2010-11-29 17:08:27 -08:00
Eric Anholt
e8f5ebf313 glsl: Make the symbol table's add_function just use the function's name. 2010-11-29 17:08:27 -08:00
Eric Anholt
2927b6c212 i965: Fix type of gl_FragData[] dereference for FB write.
Fixes glsl-fs-fragdata-1, and hopefully Eve Online where I noticed
this bug in the generated shader.  Bug #31952.
2010-11-29 17:08:26 -08:00
Adam Jackson
1ccef926be drivers/x11: unifdef XFree86Server
This code was for the old GLcore build of the software rasteriser.  The
X server switched to a DRI driver for software indirect GLX long ago.

Signed-off-by: Adam Jackson <ajax@redhat.com>
2010-11-29 17:37:54 -05:00
Marek Olšák
e5aa69f6a6 st/mesa: fix texture border color for RED and RG base formats
The spec says the border color should be consistent with the internal
format.

Signed-off-by: Brian Paul <brianp@vmware.com>
2010-11-29 17:26:40 +01:00
pontus lidman
b1097607db mesa: check for posix_memalign() errors
Signed-off-by: Brian Paul <brianp@vmware.com>
2010-11-29 09:20:41 -07:00
Dave Airlie
a7cb673aa1 r600g: it looks like r600 can handle dword offsets in the indices.
Tested with piglit + ut2004 still seems to render okay (and it
definitely does this)
2010-11-29 11:38:24 +10:00
Marek Olšák
5d4d8b6205 u_blitter: interpolate clear color using a GENERIC varying instead of COLOR
There are also some u_simple_shaders changes.

On r300, the TGSI_SEMANTIC_COLOR varying is a fixed-point number clamped
to the range [0,1] and limited to 12 bits of precision. Therefore we can't
use it for passing through a clear color in order to clear high precision
texture formats.

This also makes u_blitter use only one vertex shader instead of two.
2010-11-28 17:45:39 +01:00
Henri Verbeet
c6ea4c0e8a r600g: Fix the PIPE_FORMAT_A8_UNORM color swap for Evergreen as well. 2010-11-27 17:40:47 +01:00
Henri Verbeet
7a4599c6f5 r600g: Fix the PIPE_FORMAT_L8A8_UNORM color swaps. 2010-11-27 17:40:47 +01:00
Mathias Fröhlich
c3602ff5ed st/mesa: Set PIPE_TRANSFER_DISCARD for GL_MAP_INVALIDATE_RANGE/BUFFFER_BIT
Signed-off-by: Brian Paul <brianp@vmware.com>

Note: this is a candidate for the 7.9 branch.
2010-11-26 14:03:42 -07:00
Brian Paul
97ae4dad1c st/mesa: fix mapping of zero-sized buffer objects
Fixes http://bugs.freedesktop.org/show_bug.cgi?id=31934
2010-11-26 13:48:49 -07:00
Thomas Hellstrom
5822484510 xorg/vmwgfx: Don't clip video to viewport
Since the viewport is not updated on RandR12 mode switches anymore,
clipping to viewport may incorrectly clip away the video.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
2010-11-26 10:28:01 +01:00
Thomas Hellstrom
28ee7561f9 xorg/vmwgfx: Flush even if we don't autopaint the color key
This may help paint the colorkey before overlay updates in some
situations where the app paints the color key (mainly xine).

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
2010-11-26 10:27:54 +01:00
Marek Olšák
e6e6fcd3a6 r300/compiler: move util functions to radeon_compiler_util
The compiler seriously needs a cleanup as far as the arrangement of functions
is concerned. It's hard to know whether some function was implemented or not
because there are so many places to search in and it can be anywhere and
named anyhow.
2010-11-26 02:23:13 +01:00
Marek Olšák
ea2f56b490 r300/compiler: add a function for swizzling a mask 2010-11-26 02:23:13 +01:00
Marek Olšák
7c29446232 r300/compiler: remove duplicate function rc_mask_to_swz 2010-11-26 02:23:13 +01:00
Marek Olšák
ae3e58d973 r300/compiler: fix rc_rewrite_depth_out for it to work with any instruction
It looks like the function was originally written for ARB_fragment_program.

NOTE: This is a candidate for the 7.9 branch.
2010-11-26 02:22:59 +01:00
Marek Olšák
2c1de07ddf u_blitter: use PIPE_TRANSFER_DISCARD to prevent cpu/gpu stall
But the driver must be smart here and follow PIPE_TRANSFER_DISCARD,
as it should.
2010-11-25 20:25:53 +01:00
Xavier Chantry
c14a261eb9 nvfx: reset nvfx->hw_zeta
If nvfx_framebuffer prepare and validate were called successively with
fb->zsbuf not NULL and then NULL, nvfx->hw_zeta would contain garbage and
this would cause failures in nvfx_framebuffer_relocate/OUT_RELOC(hw_zeta).

This was triggered by piglit/texwrap 2D GL_DEPTH_COMPONENT24 and caused
first a 'write to user buffer!!' error in libdrm and then worse things.

Signed-off-by: Xavier Chantry <chantry.xavier@gmail.com>
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
2010-11-25 17:04:29 +01:00
Xavier Chantry
049065cdfa nvfx: fb->nr_cbufs <= 1 on nv30
7e1bf94631 changed
PIPE_CAP_MAX_RENDER_TARGETS to 1 on nv30.

Signed-off-by: Xavier Chantry <chantry.xavier@gmail.com>
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
2010-11-25 17:04:02 +01:00
Kenneth Graunke
1eb7a81f2e glsl: Add a virtual as_discard() method.
NOTE: This is candidate for the 7.9 branch.
2010-11-25 03:26:55 -08:00
Kenneth Graunke
a82592de92 glsl: Use do_common_optimization in the standalone compiler.
NOTE: This is a candidate for the 7.9 branch.
2010-11-25 01:21:05 -08:00
Kenneth Graunke
e8a24c65bc glsl: Don't inline function prototypes.
Currently, the standalone compiler tries to do function inlining before
linking shaders (including linking against the built-in functions).
This resulted in the built-in function _prototypes_ being inlined rather
than the actual function definition.

This is only known to fix a bug in the standalone compiler; most
programs should be unaffected.  Still, it seems like a good idea.

NOTE: This is a candidate for the 7.9 branch.
2010-11-25 01:19:53 -08:00
Vinson Lee
2cc79acc1a r300/compiler: Move declaration before code.
Fixes this GCC warning with linux-x86 build.
radeon_pair_regalloc.c: In function ‘compute_live_intervals’:
radeon_pair_regalloc.c:222: warning: ISO C90 forbids mixed declarations and code
2010-11-24 23:03:26 -08:00
Vinson Lee
995de71550 r300/compiler: Move declaration before code.
Fixes this GCC warning with linux-x86 build.
radeon_pair_regalloc.c: In function ‘compute_live_intervals’:
radeon_pair_regalloc.c:221: warning: ISO C90 forbids mixed declarations and code
2010-11-24 23:00:03 -08:00
Chia-I Wu
ba1128db45 st/vega: Fix a typo in EXTENDED_BLENDER_OVER_FUNC.
The typo was introduced by commit
231d5457b2.
2010-11-25 13:34:24 +08:00
Chia-I Wu
c9fb8c3bcf st/vega: No flipping in vg_prepare_blend_surface.
The blend sampler view is addressed with unnormalized coordinates in the
fragment shader.  It should have the same orientation as the surface
does.
2010-11-25 13:33:59 +08:00
Chia-I Wu
37ec090ac9 st/vega: Masks and surfaces should share orientation.
The alpha mask is addressed with unnormalized coordinates in the
fragment shader.  It should have the same orientation as the surface
does.

This fixes "mask" OpenVG demo.
2010-11-25 13:33:59 +08:00
Chia-I Wu
9ea4936a36 st/vega: Fix a crash with empty paths. 2010-11-25 13:32:03 +08:00
Chia-I Wu
3965051dff auxiliary: util_blit_pixels_tex should restore the viewport.
Viewport state should be saved/restored.
2010-11-25 13:32:03 +08:00
Dave Airlie
3b5a3fd8d1 r300g/r600g: bump cache manager timeouts to 1s
On lightsmark on my r500 this drop the bufmgr allocations of the sysprof.
2010-11-25 09:18:19 +10:00
Kenneth Graunke
1197393faa mesa: Fix glGet of ES2's GL_MAX_*_VECTORS properties.
Previously, the get table listed all three as having custom locations,
yet find_custom_value did not have cases to handle them.

MAX_VARYING_VECTORS does not need a custom location since MaxVaryings is
already stored as float[4] (or vec4).  MaxUniformComponents is stored as
the number of floats, however, so a custom implementation that divides
by 4 is necessary.

Fixes bugs.freedesktop.org #31495.
2010-11-24 14:09:32 -08:00
Peter Clifton
ee88727df8 meta: Mask Stencil.Clear against stencilMax in _mesa_meta_Clear
This fixes incorrect behaviour when the stencil clear value exceeds
the size of the stencil buffer, for example, when set with:

glClearStencil (~1); /* Set a bit pattern of 111...11111110 */
glClear (GL_STENCIL_BUFFER_BIT);

The clear value needs to be masked by the value 2^m - 1, where m is the
number of bits in the stencil buffer. Previously, we passed the value
masked with 0x7fffffff to _mesa_StencilFuncSeparate which then clamps,
NOT masks the value to the range 0 to 2^m - 1.

The result would be clearing the stencil buffer to 0xff, rather than 0xfe.

Signed-off-by: Peter Clifton <pcjc2@cam.ac.uk>
Signed-off-by: Brian Paul <brianp@vmware.com>
2010-11-24 12:14:54 -07:00
Brian Paul
03a4f97a68 x11: remove test_proxy_teximage() function
This was really just for testing purposes.
2010-11-24 12:11:23 -07:00
Brian Paul
74c324fdba mesa: added _mesa_format_image_size64() 2010-11-24 12:11:23 -07:00
Brian Paul
7bfbd88d2c mesa: add assertion and update comment in _mesa_format_image_size() 2010-11-24 12:11:23 -07:00
Kristian Høgsberg
a889f9ee5c i965: Don't write mrf assignment for pointsize output
https://bugs.freedesktop.org/show_bug.cgi?id=31894
2010-11-24 11:27:43 -05:00
Thomas Hellstrom
f20a219e9e gallium/targets/xorg-vmwgfx: Xv fixes
Make sure regions are properly updated and that the colorkey painting is
flushed before we update the HW overlay.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
2010-11-24 15:23:10 +01:00
Thomas Hellstrom
0b1c0460a0 st/xorg: Add a function to flush pending rendering and damage
This is needed to properly sync with host side rendering. For example,
make sure we flush colorkey painting before updating the overlay.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
2010-11-24 15:23:10 +01:00
Chia-I Wu
1f4c55128b egl_dri2: Fix one context, multiple surfaces.
When a context was made current to another surface, the old code did
this

  dri2_dpy->core->bindContext(cctx, ddraw, rdraw);
  dri2_dpy->core->unbindContext(old_cctx);

and there will be no current context due to the second line.

unbindContext should be called only when bindContext is not.  This fixes
a regression since d19afc57.  Thanks to Neil Roberts for noticing the
issue and creating a test case.
2010-11-24 14:06:30 +08:00
Ian Romanick
78a340fd48 i915: Disallow alpha, red, RG, and sRGB as render targets
Fixes bugzilla #31832

NOTE: This is a candidate for the 7.9 branch.
2010-11-23 18:42:04 -08:00
Brian Paul
903ead0b26 glsl: start restoring some geometry shader code 2010-11-23 17:23:42 -07:00
Brian Paul
6162773ea4 glsl: better handling of linker failures
Upon link error, exit translation loop, free program instructions.
Check for null pointers in calling code.
2010-11-23 17:18:48 -07:00
Brian Paul
2900e56f9d mesa: use gl_shader_type enum 2010-11-23 17:00:08 -07:00
Brian Paul
c628fd743e mesa: replace #defines with new gl_shader_type enum 2010-11-23 15:52:43 -07:00
Brian Paul
512f840702 mesa: _mesa_valid_register_index() to validate register indexes 2010-11-23 15:52:43 -07:00
Brian Paul
b8dacaf174 mesa: rename, make _mesa_register_file_name() non-static
Plus remove unused parameter.
2010-11-23 15:52:42 -07:00
Brian Paul
caf974c525 glsl: use gl_register_file in a few places 2010-11-23 15:52:42 -07:00
Brian Paul
50fd99d172 glsl: fix off by one in register index assertion 2010-11-23 15:52:42 -07:00
Alex Deucher
ed8b5fb24e gallium/egl: fix r300 vs r600 loading
Should fix:
https://bugs.freedesktop.org/show_bug.cgi?id=31841
2010-11-23 15:18:31 -05:00
Eric Anholt
df24450bac i965: Use the new embedded compare in SEL on gen6 for VS MIN and MAX opcodes.
Cuts the extra CMP instruction that used to precede SEL.
2010-11-23 09:23:30 -08:00
Eric Anholt
8a7cf99457 i965: Don't upload line smooth params unless we're line smoothing. 2010-11-23 09:23:30 -08:00
Eric Anholt
008fd3779b i965: Don't upload line stipple pattern unless we're stippling. 2010-11-23 09:23:30 -08:00
Eric Anholt
e29e3c32d9 i965: Don't upload polygon stipple unless required. 2010-11-23 09:23:30 -08:00
Eric Anholt
7720bfffa3 i965: Move gen4 blend constant color to the gen4 blending file. 2010-11-23 09:23:29 -08:00
Tilman Sauerbeck
3688301c59 r600g: Removed duplicated call to tgsi_split_literal_constant().
Signed-off-by: Tilman Sauerbeck <tilman@code-monkey.de>
2010-11-23 09:20:54 +01:00
Tom Stellard
4265c2f819 r300/compiler: Don't allow presubtract sources to be remapped twice
https://bugs.freedesktop.org/show_bug.cgi?id=31193

NOTE: This is a candidate for the 7.9 branch.
2010-11-23 00:02:03 -08:00
Mathias Fröhlich
07e0424a17 r600g: Only compare active vertex elements
Signed-off-by: Tilman Sauerbeck <tilman@code-monkey.de>
2010-11-23 08:39:43 +01:00
Vinson Lee
f44d96e1af mesa: Clean up header file inclusion in syncobj.h. 2010-11-22 21:51:49 -08:00
Vinson Lee
37195b7f70 llvmpipe: Remove unnecessary headers. 2010-11-22 21:39:14 -08:00
Xiang, Haihao
93102b4cd8 mesa: fix regression from b4bb668020
Pending commands to the previous context aren't flushed since commit b4bb668

Reported-by: Oleksiy Krivoshey <oleksiyk@gmail.com>
Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
2010-11-23 08:59:44 +08:00
Alex Deucher
cb7a36b651 r600c: fix VC flush on cedar and palm 2010-11-22 19:27:58 -05:00
Alex Deucher
0e4c5f63b9 r600g: add support for ontario APUs
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
2010-11-22 18:01:26 -05:00
Alex Deucher
072f2cbf29 r600c: add Ontario Fusion APU support
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
2010-11-22 18:01:25 -05:00
Mathias Fröhlich
8d1ad3b21c r300g: Avoid returning values in a static array, fixing a potential race
(Marek: added the initializion of "vec" in the default statement)

NOTE: This is a candidate for the 7.9 branch.

Signed-off-by: Marek Olšák <maraeo@gmail.com>
2010-11-22 23:56:41 +01:00
Alex Deucher
271b7b5914 r600g: fix some winsys functions to deal properly with evergreen
Are these functions actually used anywhere?
2010-11-22 17:39:54 -05:00
Alex Deucher
bf9c80976f r600g: fix additional EVENT_WRITE packet
Add explicit EVENT_TYPE field
2010-11-22 17:39:16 -05:00
Marek Olšák
e7c74a7dfa st/mesa: set MaxUniformComponents
Signed-off-by: Brian Paul <brianp@vmware.com>
2010-11-22 21:44:35 +01:00
Brian Paul
6a0255122a swrast: init alpha value to 1.0 in opt_sample_rgb_2d() 2010-11-22 09:04:13 -07:00
Marek Olšák
9aa089eac0 gallium: add PIPE_SHADER_CAP_SUBROUTINES
This fixes piglit/glsl-vs-main-return and glsl-fs-main-return for the drivers
which don't support RET (i915g, r300g, r600g, svga).

ir_to_mesa does not currently generate subroutines, but it's a matter of time
till it's added. It would then break all the drivers which don't implement
them, so this CAP makes sense.

Signed-off-by: Marek Olšák <maraeo@gmail.com>
2010-11-22 12:41:22 +01:00
Keith Whitwell
b2ddb93ff3 Merge branch 'lp-offset-twoside' 2010-11-22 10:36:01 +00:00
Dave Airlie
d5aadf0d80 r600g: pick correct color swap for A8 fbos.
This fixes fdo bug 31810.

Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-11-22 16:05:44 +10:00
Tom Stellard
1b6ed80972 r300/compiler: Add a more efficient version of rc_find_free_temporary() 2010-11-21 18:48:31 -08:00
Tom Stellard
8833f53e65 r300/compiler: Enable rename_reg pass for r500 cards
In addition, the rename_reg pass has been rewritten to use
rc_get_readers().
2010-11-21 18:48:31 -08:00
Tom Stellard
bbe49bc585 r300/compiler: Use presubtract operations as much as possible
Previously, presubtract operations where only being used by instructions
with less than three source source registers.
2010-11-21 18:48:31 -08:00
Tom Stellard
ddceededf8 r300/compiler: Convert RGB to alpha in the scheduler 2010-11-21 18:48:31 -08:00
Tom Stellard
681f56af80 r300/compiler: Track readers through branches in rc_get_readers() 2010-11-21 18:48:31 -08:00
Tom Stellard
255860113f r300/compiler: Handle BREAK and CONTINUE in rc_get_readers() 2010-11-21 18:48:31 -08:00
Tom Stellard
96f9580160 r300/compiler: Add rc_get_readers() 2010-11-21 18:48:31 -08:00
Tom Stellard
23f577dbd4 r300/compiler: Ignore alpha dest register when replicating the result
When the result of the alpha instruction is being replicated to the RGB
destination register, we do not need to use alpha's destination register.
This fixes an invalid "Too many hardware temporaries used" error in
the case where a transcendent operation writes to a temporary register
greater than max_temp_regs.

NOTE: This is a candidate for the 7.9 branch.
2010-11-21 18:48:31 -08:00
Tom Stellard
d668659003 r300/compiler: Use zero as the register index for unused sources
This fixes an invalid "Too many hardware temporaries used" error in the
case where a source reads from a temporary register with an index greater
than max_temp_regs and then the source is marked as unused before the
register allocation pass.

NOTE: This is a candidate for the 7.9 branch.
2010-11-21 18:48:31 -08:00
Tom Stellard
3e5f9789d6 r300/compiler: Fix instruction scheduling within IF blocks
Reads of registers that where not written to within the same block were
not being tracked.  So in a situations like this:
0: IF
1: ADD t0, t1, t2
2: MOV t2, t1

Instruction 2 didn't know that instruction 1 read from t2, so
in some cases instruction 2 was being scheduled before instruction 1.

NOTE: This is a candidate for the 7.9 branch.
2010-11-21 18:48:31 -08:00
Tom Stellard
e2301b45c2 r300/compiler: Fix register allocator's handling of loops
NOTE: This is a candidate for the 7.9 branch.
2010-11-21 18:48:31 -08:00
Tom Stellard
412803b5cd r300/compiler: Make sure presubtract sources use supported swizzles
NOTE: This is a candidate for the 7.9 branch.
2010-11-21 18:48:31 -08:00
Vinson Lee
9d08902457 r600: Remove unnecessary header. 2010-11-21 15:03:27 -08:00
Marek Olšák
2892c8bdbc docs: add GL 4.1 status 2010-11-21 23:03:58 +01:00
Marek Olšák
e40a58b7f8 st/mesa: enable ARB_explicit_attrib_location and EXT_separate_shader_objects
Gallium drivers pass all piglit tests for the two (there are 12 tests
for separate_shader_objects and 5 tests for explicit_attrib_location),
and I was told the extensions don't need any driver-specific code.

I made them dependent on PIPE_CAP_GLSL.

Signed-off-by: Brian Paul <brianp@vmware.com>
2010-11-21 19:33:45 +01:00
Brian Paul
5e3733fadf mesa: fix get_texture_dimensions() for texture array targets
Fixes http://bugs.freedesktop.org/show_bug.cgi?id=31779
2010-11-21 10:05:51 -07:00
Brian Paul
0ec0f1025d docs: update some GL 3.0 status 2010-11-21 09:40:28 -07:00
Brian Paul
5ed51e950f mesa: hook up GL 3.x entrypoints
Fix up some details in the xml files and regenerate dispatch files.
2010-11-21 09:20:44 -07:00
Brian Paul
81c347ef79 glapi: rename GL3.xml to GL3x.xml as it covers all GL 3.x versions 2010-11-21 09:20:43 -07:00
Brian Paul
197b1d7898 mesa: fix error msg typo 2010-11-21 09:20:43 -07:00
Daniel Vetter
c8fca58d9d i915g: kill idws->pool
The drm winsys only ever handles one gem memory manager. Rip out
the unnecessary complication.

Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-11-21 16:41:19 +01:00
Daniel Vetter
e182618853 i915g: kill buf->map_gtt
Not using the gtt is considered harmful for performance. And for
partial uploads there's always drm_intel_bo_subdata.

Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-11-21 16:41:19 +01:00
Daniel Vetter
d54d67499c i915g: kill RGBA/X formats
It's intel, so always little endian!

Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-11-21 16:41:19 +01:00
Daniel Vetter
8624fe7a49 i915g: add pineview pci ids
Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-11-21 16:41:19 +01:00
Daniel Vetter
aba728eb25 i915g: s/hw_tiled/tiling
More in line with other intel drivers.

Change to use enum by Jakob Bornecrantz.

Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-11-21 16:41:18 +01:00
Daniel Vetter
f77a2690b4 i915g: rip out ->sw_tiled
It looks like this was meant to facilitate unfenced access to textures/
color/renderbuffers. It's totally incomplete and fundamentally broken
on a few levels:
- broken: The kernel needs to about every tiled bo to fix up bit17
  swizzling on swap-in.
- unflexible: fenced/unfenced relocs from execbuffer2 do the same, much
  simpler.
- unneeded: with relaxed fencing tiled gem bos are as memory-efficient
  as this trick.

Hence kill it.

Reviewed-by: Jakob Bornecrantz <wallbraker@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Jakob Bornecrantz <wallbraker@gmail.com>
2010-11-21 16:41:18 +01:00
Joakim Sindholt
bf10055cff r300g: silence guard band cap errors
Somebody should find out what these are. It can be found on Windows
getting a D3DCAPS9 from IDirect3D9::GetCaps() and reading the
GuardBand* values.
2010-11-21 15:45:20 +01:00
Chia-I Wu
b8f6cb3809 st/vega: Fix vgReadPixels with a subrectangle.
Fix a crash when the subrectangle is not inside the fb.  Fix wrong
pipe transfer when sx > 0 or sy + height != fb->height.

This fixes "readpixels" demo.
2010-11-21 19:32:22 +08:00
Chia-I Wu
e8bbaff22e st/vega: Set wrap_r for mask and blend samplers.
These two samplers use non-normalized texture coordinates.  wrap_r
cannot be PIPE_TEX_WRAP_REPEAT (the default).

This fixes

  sp_tex_sample.c:1790:get_linear_unorm_wrap: Assertion `0' failed

assertion failure.
2010-11-21 19:32:10 +08:00
Chia-I Wu
daa265e53c st/vega: vegaLookupSingle should validate the state.
Fix "lookup" demo crash.
2010-11-21 19:26:33 +08:00
Chia-I Wu
f90524a01b tgsi: Add STENCIL to text parser.
Fix OpenVG "filter" demo

  Program received signal SIGSEGV, Segmentation fault.
  0xb7153dc9 in str_match_no_case (pcur=0xbfffe564, str=0x0) at
  tgsi/tgsi_text.c:86
  86         while (*str != '\0' && *str == uprcase( *cur )) {
2010-11-21 19:26:29 +08:00
Vinson Lee
8bea7776a3 mesa: Clean up header file inclusion in stencil.h. 2010-11-20 22:44:33 -08:00
Vinson Lee
9732a93f40 mesa: Clean up header file inclusion in shared.h. 2010-11-20 22:30:27 -08:00
Vinson Lee
20f041952c mesa: Clean up header file inclusion in shaderapi.h. 2010-11-20 22:17:28 -08:00
Vinson Lee
baa0471597 mesa: Clean up header file inclusion in scissor.h. 2010-11-20 22:01:30 -08:00
Vinson Lee
27e96655c7 mesa: Clean up header file inclusion in renderbuffer.h. 2010-11-20 21:32:07 -08:00
Vinson Lee
b6215d18b5 mesa: Clean up header file inclusion in readpix.h. 2010-11-20 21:23:35 -08:00
Vinson Lee
bab188d22f mesa: Clean up header file inclusion in rastpos.h. 2010-11-20 21:14:06 -08:00
Vinson Lee
9b66305b8d mesa: Clean up header file inclusion in polygon.h. 2010-11-20 21:06:09 -08:00
Vinson Lee
f5cbe04b69 intel: Remove unnecessary header. 2010-11-20 20:13:50 -08:00
Vinson Lee
84f5229119 r600: Remove unnecesary header. 2010-11-20 19:04:30 -08:00
Vinson Lee
b59f3dd8ca swrast: Remove unnecessary header. 2010-11-20 19:00:18 -08:00
Vinson Lee
1310806872 st/mesa: Remove unnecessary headers. 2010-11-20 18:48:09 -08:00
Chia-I Wu
bb045d339b scons: Define IN_DRI_DRIVER.
The define is required for DRI drivers.  It is not needed for
libgl-xlib, but the overhead it introduces should be minor.
2010-11-20 17:47:11 -08:00
Xavier Chantry
7e1bf94631 nvfx: only expose one rt on nv30
We do not know how to use more, GL_ARB_draw_buffers is not exposed on blob.
2010-11-20 23:29:12 +01:00
Owen W. Taylor
c63a86e1e5 r600g: Fix location for clip plane registers
The stride between the different clip plane registers was incorrect.

https://bugs.freedesktop.org/show_bug.cgi?id=31788

agd5f: fix evergreen as well.
2010-11-20 12:18:56 -05:00
Marek Olšák
ffb732d8bd r300g: fix rendering with no vertex elements
Fixes glsl-vs-point-size, although I meant to fix glsl-novertexdata.
Since swrast fails glsl-novertexdata too, I guess it's a core issue.
2010-11-20 16:20:48 +01:00
Eric Anholt
b6b91fa029 i965: Remove duplicate MRF writes in the FS backend.
This is quite common for multitexture sampling, and not only cuts down
on the second and later set of MOVs, but typically also allows
compute-to-MRF on the first set.

No statistically siginficant performance difference in nexuiz (n=3),
but it reduces instruction count in one of its shaders and seems like
a good idea.
2010-11-19 20:05:56 -08:00
Eric Anholt
47b1aac1cf i965: Improve compute-to-mrf.
We were skipping it if the instruction producing the value we were
going to compute-to-mrf used its result reg as a source reg.  This
meant that the typical "write interpolated color to fragment color" or
"texture from interpolated texcoord" shader didn't compute-to-MRF.
Just don't check for the interference cases until after we've checked
if this is the instruction we wanted to compute-to-MRF.

Improves nexuiz high-settings performance on my laptop 0.48% +- 0.08%
(n=3).
2010-11-19 19:54:11 -08:00
Eric Anholt
ac89a90401 ir_to_mesa: Detect and emit MOV_SATs for saturate constructs.
The goal here is to avoid regressing performance on ir_to_mesa drivers
for fixed function fragment shaders requiring saturates.
2010-11-19 19:09:32 -08:00
Eric Anholt
19631fab35 i965: Recognize saturates and turn them into a saturated mov.
On pre-gen6, this turns 4 instructions into 1.  We could still do
better by folding the saturate into the instruction generating the
value if nobody else uses it, but that should be a separate pass.
2010-11-19 19:09:31 -08:00
Eric Anholt
02939d643f glsl: Add a helper function for determining if an rvalue could be a saturate.
Hardware pretty commonly has saturate modifiers on instructions, and
this can be used in codegen to produce those, without everyone else
needing to understand clamping other than min and max.
2010-11-19 19:09:18 -08:00
Eric Anholt
602ae2441a i965: Fold constants into the second arg of BRW_SEL as well.
This hits a common case with min/max operations.
2010-11-19 17:42:07 -08:00
Eric Anholt
f9b420d3bd i965: Remove extra \n at the end of every instruction in INTEL_DEBUG=wm. 2010-11-19 17:42:07 -08:00
Eric Anholt
5944cda6ed i965: Just use memset() to clear most members in FS constructors.
This should make it a lot harder to forget to zero things.
2010-11-19 17:42:07 -08:00
Eric Anholt
61126278a3 i965: Fix compute_to_mrf to not move a MRF write up into another live range.
Fixes glsl-fs-copy-propagation-texcoords-1.
2010-11-19 17:42:06 -08:00
Eric Anholt
6b1d7dd781 mesa: Include C++ files in the makedepend of DRI drivers. 2010-11-19 17:42:06 -08:00
Vinson Lee
a172368ef1 glsl: Fix type of label 'default' in switch statement. 2010-11-19 17:28:22 -08:00
Vinson Lee
7aebe181f3 glsl: Add lower_vector.cpp to SConscript. 2010-11-19 17:23:07 -08:00
Ian Romanick
bb756bb0a6 glsl: Fix matrix constructors with vector parameters
When the semantics of write masks in assignments were changed, this
code was not correctly updated.

Fixes piglit test glsl-mat-from-vec-ctor-01.
2010-11-19 17:17:25 -08:00
Kenneth Graunke
63684a9ae7 glsl: Combine many instruction lowering passes into one.
This should save on the overhead of tree-walking and provide a
convenient place to add more instruction lowering in the future.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
2010-11-19 15:56:28 -08:00
Kenneth Graunke
b943fb94bf glsl: Simplify a type check by using type->is_integer(). 2010-11-19 15:06:58 -08:00
Ian Romanick
11d6f1c698 glsl: Add ir_quadop_vector expression
The vector operator collects 2, 3, or 4 scalar components into a
vector.  Doing this has several advantages.  First, it will make
ud-chain tracking for components of vectors much easier.  Second, a
later optimization pass could collect scalars into vectors to allow
generation of SWZ instructions (or similar as operands to other
instructions on R200 and i915).  It also enables an easy way to
generate IR for SWZ instructions in the ARB_vertex_program assembler.
2010-11-19 15:00:26 -08:00
Ian Romanick
13f57d42b6 glsl: Add unary ir_expression constructor 2010-11-19 15:00:25 -08:00
Ian Romanick
8e498050dc glsl: Add ir_rvalue::is_negative_one predicate 2010-11-19 15:00:25 -08:00
Ian Romanick
fc92e87b97 glsl: Eliminate assumptions about size of ir_expression::operands
This may grow in the near future.
2010-11-19 15:00:25 -08:00
Ian Romanick
f2616e56de glsl: Add ir_unop_sin_reduced and ir_unop_cos_reduced
The operate just like ir_unop_sin and ir_unop_cos except that they
expect their inputs to be limited to the range [-pi, pi].  Several
GPUs require this limited range for their sine and cosine
instructions, so having these as operations (along with a to-be-written
lowering pass) helps this architectures.

These new operations also matche the semantics of the
GL_ARB_fragment_program SCS instruction.  Having these as operations
helps in generating GLSL IR directly from assembly fragment programs.
2010-11-19 15:00:25 -08:00
Alex Deucher
04ffbe1ac6 r600g: use full range of VS resources for vertex samplers
Now that we have fetch shaders, the full range of VS resources
can be used for sampling.
2010-11-19 15:51:24 -05:00
Alex Deucher
4afd068385 r600g: use meaningful defines for chiprev
Makes the code much clearer.
2010-11-19 15:32:02 -05:00
Alex Deucher
52c66120d8 r600g: translate ARR instruction for evergreen
evergreen variant of:
9f7ec103e2
2010-11-19 15:20:59 -05:00
Jerome Glisse
f609b2ab03 r600g: add fetch shader capabilities
Use fetch shader instead of having fetch instruction in the vertex
shader. Allow to restrict shader update to a smaller part when
vertex buffer input layout changes.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2010-11-19 13:40:55 -05:00
Alex Deucher
3e76ed4e25 r600g: All EVENT_WRITE packets need the EVENT_INDEX field
6xx-evergreen
2010-11-19 13:35:53 -05:00
Viktor Novotný
af17d89966 dri/nouveau: Clean up magic numbers in get_rt_format
Signed-off-by: Viktor Novotný <noviktor@seznam.cz>
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
2010-11-19 19:04:50 +01:00
Jerome Glisse
fab804bdfe r600g: fix occlusion query on evergreen (avoid lockup)
Occlusion query on evergreen need the event index field to be
set otherwise we endup locking up the GPU.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2010-11-19 11:53:01 -05:00
Keith Whitwell
546c5ffa11 llvmpipe: twoside for specular color also 2010-11-19 16:17:36 +00:00
Keith Whitwell
081ce2680e llvmpipe: fix up twoside after recent changes
Fix my slot/attr confusion.
2010-11-19 16:16:30 +00:00
Hui Qi Tay
d4b5cf6c05 llvmpipe: fix such that offset/twoside function only does in-place modification 2010-11-19 15:12:19 +00:00
Ian Romanick
c05ccc1ebd ir_to_mesa: Generate smarter code for some conditional moves
Condiation moves with a condition of (a < 0), (a > 0), (a <= 0), or (a
>= 0) can be generated with "a" directly as an operand of the CMP
instruction.  This doesn't help much now, but it will help with
assembly shaders that use the CMP instruction.
2010-11-18 18:19:45 -08:00
Ian Romanick
ad87f2ddc7 glsl: Make is_zero and is_one virtual methods of ir_rvalue
This eliminates the need in some cames to validate that an rvalue is
an ir_constant before checking to see if it's 0 or 1.
2010-11-18 18:19:45 -08:00
Brian Paul
83e93b6008 mesa: pass gl_format to _mesa_init_teximage_fields()
This should prevent the field going unset in the future.  See bug
http://bugs.freedesktop.org/show_bug.cgi?id=31544 for background.

Also remove unneeded calls to clear_teximage_fields().

Finally, call _mesa_set_fetch_functions() from the
_mesa_init_teximage_fields() function so callers have one less
thing to worry about.
2010-11-18 16:15:38 -07:00
José Fonseca
3dcc3153b0 scons: Use inline wrap helpers more consistently. 2010-11-18 13:02:36 +00:00
Dave Airlie
185d862cd8 gallium/noop: report GL 2.1
this should at least make app use the same paths as they would for a real
driver.
2010-11-18 18:11:27 +10:00
Vinson Lee
855c66bde7 glsl: Fix 'control reaches end of non-void function' warning.
Fix this GCC warning.
ir.cpp: In static member function
'static unsigned int ir_expression::get_num_operands(ir_expression_operation)':
ir.cpp:199: warning: control reaches end of non-void function
2010-11-17 22:43:52 -08:00
Chia-I Wu
ca9f99d9c5 mesa: Clean up core.h.
Remove version.h and context.h from core.h.
2010-11-18 11:56:01 +08:00
Chia-I Wu
997a2547d1 st/glx: Replace MESA_VERSION_STRING by xmesa_get_name.
xmesa_get_name returns the name of the st_api, which is the same as
MESA_VERSION_STRING.
2010-11-18 11:56:01 +08:00
Chia-I Wu
db1689c236 st/wgl: Use st_context_iface::share for DrvShareLists. 2010-11-18 11:56:01 +08:00
Chia-I Wu
4f38dcd974 gallium: Add st_context_iface::share to st_api.
It will be used to implement wglShareLists.  Fill st_context_iface::copy
for glXCopyContext as well.
2010-11-18 11:56:00 +08:00
Chia-I Wu
28105471af gallium: Add st_api::name.
It is the name of the rendering API.  This field is informative.
2010-11-18 11:56:00 +08:00
Chia-I Wu
cc5c908d7d st/vega: Do not wait NULL fences. 2010-11-18 11:56:00 +08:00
Eric Anholt
50b4508319 i965: Eliminate dead code more aggressively.
If an instruction writes reg but nothing later uses it, then we don't
need to bother doing it.  Before, we were just killing code that was
never read after it was ever written.

This removes many interpolation instructions for attributes with only
a few comopnents used.  Improves nexuiz high-settings performance .46%
+/- .12% (n=3) on my Ironlake.
2010-11-18 11:16:14 +08:00
Brian Paul
48af60b465 mesa: upgrade to glext.h version 66
The type of the num/count parameter to glProgramParameters4[df]vNV()
changed so some API dispatch code needed updates too.
2010-11-17 20:04:45 -07:00
Alex Deucher
a23f25eba1 r600g: fix buffer alignment
This should fix the remaining buffer alignment issues in r600g.
2010-11-17 21:33:40 -05:00
Eric Anholt
da35388044 i965: Fail on loops on gen6 for now until we write the EU emit code for it. 2010-11-18 09:18:47 +08:00
Eric Anholt
3c8db58a17 i965: Add dumping of the sampler default color. 2010-11-18 09:18:47 +08:00
Eric Anholt
95addca019 i965: Add state dumping for sampler state. 2010-11-18 09:18:47 +08:00
Eric Anholt
03ff02d08b mesa: Don't spam the console in a debug build unless some spam is requested.
It's annoying to use test suites under a Mesa debug build because
pretty output is cluttered with stderr's continuous reports that
you're still using the debug driver.
2010-11-18 09:18:47 +08:00
Eric Anholt
d512cbf58f i965: Shut up spurious gcc warning about GLSL_TYPE enums. 2010-11-18 09:18:47 +08:00
Jakob Bornecrantz
e89e8a4347 gallium: Remove redundant sw and debug target helpers 2010-11-17 23:49:09 +00:00
Jakob Bornecrantz
b46340c740 graw: Use inline debug helper instead of non-inline version 2010-11-17 23:49:09 +00:00
Jakob Bornecrantz
c30656d8c2 libgl-xlib: Use inline debug helper instead of non-inline version 2010-11-17 23:49:08 +00:00
Chad Versace
7819435f2e glsl: Improve usage message for glsl_compiler
The new usage message lists possible command line options. (Newcomers to Mesa
currently have to trawl through the source to find the command line options,
and we should save them from that trouble.)

Example Output
--------------
usage: ./glsl_compiler [options] <file.vert | file.geom | file.frag>

Possible options are:
    --glsl-es
    --dump-ast
    --dump-hir
    --dump-lir
    --link
2010-11-17 15:44:41 -08:00
Kenneth Graunke
007f488150 glsl: Refactor get_num_operands.
This adds sentinel values to the ir_expression_operation enum type:
ir_last_unop, ir_last_binop, and ir_last_opcode.  They are set to the
previous one so they don't trigger "unhandled case in switch statement"
warnings, but should never be handled directly.

This allows us to remove the huge array of 1s and 2s in
ir_expression::get_num_operands().
2010-11-17 15:44:41 -08:00
Jerome Glisse
7ffd4e976f r600g: code cleanup (indent, trailing space, empty line ...)
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2010-11-17 17:22:08 -05:00
Kenneth Graunke
9935fe705d glsl: Remove the ir_binop_cross opcode. 2010-11-17 13:59:17 -08:00
Kenneth Graunke
af1cba2260 Refresh autogenerated file builtin_function.cpp. 2010-11-17 13:37:16 -08:00
Kenneth Graunke
671ccf593e glsl: Reimplement the "cross" built-in without ir_binop_cross.
We are not aware of any GPU that actually implements the cross product
as a single instruction.  Hence, there's no need for it to be an opcode.
Future commits will remove it entirely.
2010-11-17 13:20:30 -08:00
Kenneth Graunke
302fe4049c Regenerate glcpp parser. 2010-11-17 12:53:07 -08:00
Kenneth Graunke
d719bf8fb4 glsl: Unconditionally define GL_FRAGMENT_PRECISION_HIGH in ES2 shaders.
This is really supposed to be defined only if the driver supports highp
in the fragment shader - but all of our current ES2 implementations do.
So, just define it.  In the future, we'll need to add a flag to
gl_context and only define the macro if the flag is set.

"Fixes" freedesktop.org bug #31673.
2010-11-17 12:50:35 -08:00
Robert Hooker
778917069c egl_dri2: Add missing intel chip ids.
Signed-off-by: Robert Hooker <robert.hooker@canonical.com>
2010-11-17 12:10:53 -08:00
Chad Versace
df883eb157 glsl: Fix Doxygen tag \file in recently renamed files 2010-11-17 12:07:24 -08:00
Chad Versace
b4cdba687c glsl: Fix erroneous cast in ast_jump_statement::hir()
Return values were erroneously cast from (ir_rvalue*) to (ir_expression*).

NOTE: This is a candidate for the 7.9 branch.
2010-11-17 11:59:32 -08:00
Kenneth Graunke
e16c9d5d03 glsl: Fix constant expression handling for <, >, <=, >= on vectors.
ir_binop_less, ir_binop_greater, ir_binop_lequal, and ir_binop_gequal
are defined to work on vectors as well as scalars, as long as the two
operands have the same type.

This is evident from both ir_validate.cpp and our use of these opcodes
in the GLSL lessThan, greaterThan, lessThanEqual, greaterThanEqual
built-in functions.

Found by code inspection.  Not known to fix any bugs.  Presumably, our
tests for the built-in comparison functions must pass because C.E.
handling is done on the ir_call of "greaterThan" rather than the inlined
opcode.  The C.E. handling of the built-in function calls is correct.

NOTE: This is a candidate for the 7.9 branch.
2010-11-17 11:58:56 -08:00
Marek Olšák
fb7ae06f59 r300g: print FS inputs uninitialized due to hardware limits to stderr 2010-11-17 19:01:12 +01:00
Alex Deucher
75e52556a8 r600c/evergreen: texture align is group_bytes just like 6xx/7xx
Default group bytes to 512 on evergreen.  Don't query
tiling config yet for evergreen, the current info returned is not
adequate for evergreen (no way to get bank info).
2010-11-17 11:30:52 -05:00
Brian Paul
1d39df42c4 mesa: minor clean-ups in context code 2010-11-16 20:12:34 -07:00
Brian Paul
85288ad722 mesa: reorder texture_error_check() params
To better match other functions.
2010-11-16 20:12:34 -07:00
Brian Paul
3c59febf05 mesa: 80-column wrapping 2010-11-16 20:12:34 -07:00
Brian Paul
76114d23d1 mesa: whitespace cleanups 2010-11-16 20:12:34 -07:00
Brian Paul
78527154bd mesa: fix error messages and minor reindenting 2010-11-16 20:12:34 -07:00
Kenneth Graunke
bee901a1cf Refresh autogenerated glcpp parser. 2010-11-16 16:22:13 -08:00
Kenneth Graunke
3fb83038a0 glcpp: Define GL_FRAGMENT_PRECISION_HIGH if GLSL version >= 1.30.
Per section 4.5.4 of the GLSL 1.30 specification.
2010-11-16 16:22:12 -08:00
Henri Verbeet
6bbe637c13 r600g: Synchronize supported color formats between Evergreen and r600/r700. 2010-11-17 00:41:42 +01:00
Henri Verbeet
7d0f45563d r600g: Swizzle vertex data only once.
Vertex data swizzles are already done in the vertex shader. Doing them twice
breaks BGRA vertex arrays for example.
2010-11-17 00:41:42 +01:00
Marek Olšák
b6e2c32626 r300g: remove the hack with OPCODE_RET
RET was interpreted as END, which was wrong. Instead, if a shader contains RET
in the main function, it will fail to compile with an error message
from now on.

The hack is from early days.
2010-11-16 22:39:27 +01:00
Ian Romanick
2d2d6a80c1 glsl: Simplify generation of swizzle for vector constructors 2010-11-16 12:11:02 -08:00
Ian Romanick
38e55153af glsl: Refactor is_vec_{zero,one} to be methods of ir_constant
These predicates will be used in other places soon.
2010-11-16 12:11:02 -08:00
José Fonseca
4f84a3aa32 libgl-gdi: Allow to pick softpipe/llvmpipe on runtime. 2010-11-16 18:56:39 +00:00
Vinson Lee
063c6b8f74 mesa: Add definitions for inverse hyperbolic function on MSVC. 2010-11-15 22:00:32 -08:00
Vinson Lee
1935bdca44 glsl: Add ir_constant_expression.cpp to SConscript.
This was accidentally removed in commit 32aaf89823.

Fixes SCons builds.
2010-11-15 20:56:11 -08:00
Brian Paul
9b4b70e7e2 glsl: remove opt_constant_expression.cpp from SConscript
And alphabetize the opt_* files.
2010-11-15 18:59:45 -07:00
Brian Paul
c1928c7f10 mesa: add more work-arounds for acoshf(), asinhf(), atahf() 2010-11-15 18:50:58 -07:00
Brian Paul
88f482a839 glsl: fix assorted MSVC warnings 2010-11-15 18:48:43 -07:00
Brian Paul
20c2debce8 st/mesa: fix glDrawPixels(depth/stencil) bugs
When drawing GL_DEPTH_COMPONENT the usual fragment pipeline steps apply
so don't override the depth state.

When drawing GL_STENCIL_INDEX (or GL_DEPTH_STENCIL) the fragment pipeline
does not apply (only the stencil and Z writemasks apply) so disable writes
to the color buffers.

Fixes some regressions from commit ef8bb7ada9
2010-11-15 18:40:32 -07:00
Kenneth Graunke
32aaf89823 glsl: Rename various ir_* files to lower_* and opt_*.
This helps distinguish between lowering passes, optimization passes, and
other compiler code.
2010-11-15 16:34:20 -08:00
Kenneth Graunke
46b80d6469 glsl: Remove unused and out of date Makefile.am.
This was from when glsl2 lived in a separate repository and used
automake.
2010-11-15 14:47:15 -08:00
Kenneth Graunke
3108095198 glsl: Add constant expression handling for asinh, acosh, and atanh. 2010-11-15 14:08:58 -08:00
Kenneth Graunke
91181c7dd4 glsl: Refresh autogenerated file builtin_function.cpp. 2010-11-15 14:02:52 -08:00
Kenneth Graunke
db9b8c062f glsl: Implement the asinh, acosh, and atanh built-in functions. 2010-11-15 14:02:52 -08:00
Kenneth Graunke
096d36872f generate_builtins.py: Fix inconsistent use of tabs and spaces warning. 2010-11-15 14:02:52 -08:00
Kenneth Graunke
0d082c0e06 glsl: Refresh autogenerated lexer and parser files.
For the last three commits.
2010-11-15 13:33:58 -08:00
Kenneth Graunke
7279feeb19 glsl: Add support for the 'u' and 'U' unsigned integer suffixes. 2010-11-15 13:33:58 -08:00
Kenneth Graunke
2b6c1a0b7c glsl: Add new keywords and reserved words for GLSL 1.30. 2010-11-15 13:33:58 -08:00
Kenneth Graunke
285036fbb0 glsl: Rework reserved word/keyword handling in the lexer.
This consolidates the TOKEN_OR_IDENTIFIER and RESERVED_WORD macros into
a single KEYWORD macro.

The old TOKEN_OR_IDENTIFIER macros handled the case of a word going from
an identifier to a keyword; the RESERVED_WORD macro handled a word going
from a reserved word to a language keyword.  However, neither could
properly handle samplerBuffer (for example), which is an identifier in
1.10 and 1.20, a reserved word in 1.30, and a keyword in 1.40 and on.

Furthermore, the existing macros didn't properly handle reserved words
in GLSL ES 1.00.  The best they could do was return a token (rather than
an identifier), resulting in an obtuse parser error, rather than a
user-friendly "you used a reserved word" error message.
2010-11-15 13:33:57 -08:00
Kenneth Graunke
5dc74e9c77 glsl: Convert glsl_type::base_type from #define'd constants to an enum.
This is nice because printing type->base_type in GDB will now give you a
readable name instead of a number.
2010-11-15 13:33:57 -08:00
Kenneth Graunke
ee36f14fa5 glsl: Remove GLSL_TYPE_FUNCTION define.
Functions are not first class objects in GLSL, so there is never a value
of function type.  No code actually used this except for one function
which asserted it shouldn't occur.  One comment mentioned it, but was
incorrect.  So we may as well remove it entirely.
2010-11-15 13:33:57 -08:00
Henri Verbeet
62fe9c4efc r600g: Add PIPE_FORMAT_L8A8_UNORM for Evergreen as well. 2010-11-15 22:20:12 +01:00
Henri Verbeet
228d0d1153 r600: Evergreen has two extra frac_bits for the sampler LOD state.
Note: this is a candidate for the 7.9 branch.
2010-11-15 22:20:12 +01:00
Henri Verbeet
aa3113ae20 r600g: Evergreen has two extra frac_bits for the sampler LOD state.
The (piglit) mipmap_limits test shows the issue very clearly.
2010-11-15 22:20:12 +01:00
Henri Verbeet
da8c877733 r600g: Cleanup the fenced_bo list in r600_context_fini(). 2010-11-15 22:20:12 +01:00
Jerome Glisse
5da246944a gallium/noop: no operation gallium driver
This driver is a fake swdri driver that perform no operations
beside allocation gallium structure and buffer for upper layer
usage.

It's purpose is to help profiling core mesa/gallium without
having pipe driver overhead hidding hot spot of core code.

scons file are likely inadequate i am unfamiliar with this
build system.

To use it simply rename is to swrast_dri.so and properly set
LIBGL_DRIVERS_PATH env variable.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2010-11-15 14:56:40 -05:00
Hui Qi Tay
e3f106b5fe llvmpipe: clean up polygon offset function in lp setup code 2010-11-15 17:21:35 +00:00
Francisco Jerez
88850b3e4f dri/nouveau: Kill a bunch of ternary operators. 2010-11-15 17:42:08 +01:00
Francisco Jerez
aceb5b3277 dri/nouveau: Fix typo. 2010-11-15 17:42:08 +01:00
Viktor Novotný
8c94c7138e dri/nouveau: Remove nouveau_class.h, finishing switch to rules-ng-ng headers
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
2010-11-15 17:42:07 +01:00
Viktor Novotný
69f54d2a7e dri/nouveau nv20: Use rules-ng-ng headers
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
2010-11-15 17:42:07 +01:00
Viktor Novotný
f4efc256fd dri/nouveau: nv10: Use rules-ng-ng headers
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
2010-11-15 17:42:07 +01:00
Viktor Novotný
8983855012 dri/nouveau: nv04: Use rules-ng-ng headers
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
2010-11-15 17:42:06 +01:00
Viktor Novotný
dfc2bf818b dri/nouveau: Import headers from rules-ng-ng
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
2010-11-15 17:42:06 +01:00
Brian Paul
3e910faed5 evergreen: set gl_texture_image::TexFormat field in evergreenSetTexBuffer()
See https://bugs.freedesktop.org/show_bug.cgi?id=31544

Note: this is a candidate for the 7.9 branch.
2010-11-15 09:31:33 -07:00
Brian Paul
f2f1c61950 r300: set gl_texture_image::TexFormat field in r300SetTexBuffer2()
See https://bugs.freedesktop.org/show_bug.cgi?id=31544

Note: this is a candidate for the 7.9 branch
2010-11-15 09:31:27 -07:00
Brian Paul
401f1efd3a r200: set gl_texture_image::TexFormat field in r200SetTexBuffer2()
See https://bugs.freedesktop.org/show_bug.cgi?id=31544

Note: this is a candidate for the 7.9 branch.
2010-11-15 09:31:21 -07:00
Brian Paul
0ef3b298e6 r600: set gl_texture_image::TexFormat field in r600SetTexBuffer2()
See https://bugs.freedesktop.org/show_bug.cgi?id=31544

Note: this is a candidate for the 7.9 branch.
2010-11-15 09:18:47 -07:00
Brian Paul
86abc1f104 radeon: set gl_texture_image::TexFormat field in radeonSetTexBuffer2()
See https://bugs.freedesktop.org/show_bug.cgi?id=31544

Note: this is a candidate for the 7.9 branch
2010-11-15 09:18:40 -07:00
Julien Cristau
e86b4c9194 Makefile: don't include the same files twice in the tarball
src/mesa/drivers/dri/*/*/*.[chS] is a superset of
src/mesa/drivers/dri/*/server/*.[ch] and
src/mesa/drivers/dri/common/xmlpool/*.[ch].
include/GL/internal/glcore.h is already in MAIN_FILES, no need for it in
DRI_FILES too.  src/glx/Makefile was listed twice.

Signed-off-by: Julien Cristau <jcristau@debian.org>
Signed-off-by: Brian Paul <brianp@vmware.com>
2010-11-15 08:46:22 -07:00
Daniel Lichtenberger
ef0720758e radeon: fix potential segfault in renderbuffer update
Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=31617

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
2010-11-15 01:32:42 -05:00
Marek Olšák
9cf25b3d1c r300g: return shader caps from Draw for SWTCL vertex shaders 2010-11-14 23:04:18 +01:00
Marek Olšák
ed7cb289b3 r300g: clean up redundancy in draw functions 2010-11-14 19:28:04 +01:00
Eric Anholt
3b337f5cd9 i965: Fix gl_FragCoord inversion when drawing to an FBO.
This showed up as cairo-gl gradients being inverted on everyone but
Intel, where I'd apparently tweaked the transformation to work around
the bug.  Fixes piglit fbo-fragcoord.
2010-11-14 22:35:17 +08:00
Vinson Lee
d11db2a857 i965: Silence uninitialized variable warning.
Silences this GCC warning.
brw_fs.cpp: In member function 'void fs_visitor::split_virtual_grfs()':
brw_fs.cpp:2516: warning: unused variable 'reg'
2010-11-13 21:19:59 -08:00
Marek Olšák
7e2256688a r300g: fix texture border color for all texture formats
This fixes 8 texwrap format tests.
The code should handle arbitrary formats now and is cleaner.

NOTE: This is a candidate for the 7.9 branch.
2010-11-13 16:43:20 +01:00
Vinson Lee
3f6b1756f8 mesa: Clean up header file inclusion in points.h. 2010-11-13 01:16:12 -08:00
Brian Paul
56b4819932 mesa: consolidate assertions in teximage code 2010-11-12 07:21:29 -07:00
Marek Olšák
93edd15178 svga: fill out CAPs for indirect addressing
As per the ps_3_0 and vs_3_0 documentation.
The aL register in D3D9 is quite tricky to use, though.
2010-11-12 03:13:23 +01:00
Marek Olšák
5c7127c07c r600g: fill out CAPs for indirect addressing 2010-11-12 03:13:23 +01:00
Marek Olšák
d279902b40 r300g: fill out CAPs for indirect addressing
To match shader model 2.0 (it's impossible to fully implement ARL
with shader model 3.0 relative addressing).
2010-11-12 03:13:22 +01:00
Marek Olšák
abe2c0d3b0 nvfx: fill out CAPs for indirect addressing
To match shader model 2.0.
2010-11-12 03:13:22 +01:00
Marek Olšák
3c6309e2f7 nv50: fill out CAPs for indirect addressing 2010-11-12 03:13:22 +01:00
Marek Olšák
04bafb2b55 i965g: fill out CAPs for indirect addressing 2010-11-12 03:13:22 +01:00
Marek Olšák
5bf7d668ac i915g: fill out CAPs for indirect addressing 2010-11-12 03:13:22 +01:00
Marek Olšák
53b7ec91ca tgsi: fill out CAPs for indirect addressing 2010-11-12 03:13:22 +01:00
Marek Olšák
cbfdf262cc gallium: add CAPs for indirect addressing and lower it in st/mesa when needed
Required because ATI and NVIDIA DX9 GPUs do not support indirect addressing
of temps, inputs, outputs, and consts (FS-only) or the hw support is so
limited that we cannot use it.

This should make r300g and possibly nvfx more feature complete.

Signed-off-by: Marek Olšák <maraeo@gmail.com>
2010-11-12 03:13:22 +01:00
Brian Paul
d18df9e336 tdfx: s/Format/_BaseFormat/
Fixes http://bugs.freedesktop.org/show_bug.cgi?id=31560
2010-11-11 16:56:45 -07:00
Eric Anholt
6929cdd14b glsl: Free the loop state context when we free the loop state.
Since this was talloced off of NULL instead of the compile state, it
was a real leak over the course of the program.  Noticed with
valgrind --leak-check=full --show-reachable=yes.  We should really
change these passes to generally get the compile context as an argument
so simple mistakes like this stop mattering.
2010-11-11 15:12:37 -08:00
Brian Paul
78587ea012 mesa: fix glDeleteBuffers() regression
This fixes a regression (failed assertion) from commit
c552f273f5 which was hit if glDeleteBuffers()
was called on a buffer that was never bound.

NOTE: this is a candidate for the 7.9 branch.
2010-11-11 15:31:36 -07:00
Brian Paul
c552f273f5 mesa: make glIsBuffer() return false for never bound buffers
Use a dummy buffer object as we do for frame/renderbuffer objects.
Fixes http://bugs.freedesktop.org/show_bug.cgi?id=31514

Note: this is a candidate for the 7.9 branch.
2010-11-11 14:49:53 -07:00
Aras Pranckevicius
d67df5dd9d glsl: fix crash in loop analysis when some controls can't be determined
Fixes loop-07.frag.
2010-11-11 10:49:37 -08:00
Keith Whitwell
7fb16423cc r600g: enforce minimum stride on render target texture images
Fixes piglit/fbo_readpixels since staging upload changes.
2010-11-11 16:20:24 +00:00
Keith Whitwell
8a3c181e9c r600g: do not try to use staging resource for depth textures
Currently r600_resource_copy_region() will turn these copies into
transfers + memcpys, so to avoid recursion we must not turn those
transfers back into blits.
2010-11-11 15:43:31 +00:00
Brian Paul
b3b6476695 mesa: handle more pixel types in mipmap generation code
NOTE: This is a candidate for the 7.9 branch.
2010-11-11 08:33:42 -07:00
Brian Paul
79c65410c1 mesa: add missing formats in _mesa_format_to_type_and_comps()
NOTE: this is a candidate for the 7.9 branch
2010-11-11 08:31:21 -07:00
Brian Paul
c9126d66fa mesa: improve error message 2010-11-11 07:43:46 -07:00
Brian Paul
624661cae4 mesa: #include mfeatures.h in enums.h 2010-11-11 07:43:46 -07:00
Keith Whitwell
6baad55f15 r600g: guard experimental s3tc code with R600_ENABLE_S3TC 2010-11-11 14:30:09 +00:00
Lucas Stach
089056a5f3 nvfx: fill PIPE_CAP_PRIMITIVE_RESTART and PIPE_CAP_SHADER_STENCIL_EXPORT
Signed-off-by: Lucas Stach <dev@lynxeye.de>
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
2010-11-11 14:55:46 +01:00
Francisco Jerez
cdb38b5d3d dri/nouveau: Split hardware/software TNL instantiation more cleanly. 2010-11-11 14:50:50 +01:00
Vinson Lee
dc524adee2 mesa: Fix printf format warnings. 2010-11-10 17:30:59 -08:00
Ian Romanick
bcef51c3b8 mesa: Allow query of MAX_SAMPLES with EXT_framebuffer_multisample
Previously queries of MAX_SAMPLES were only allowed with
ARB_framebuffer_object, but EXT_framebuffer_multisample also enables
this query.  This seems to only effect the i915.  All other drivers
support both extensions or neither extension.

This patch is based on a patch that Kenneth sent along with the report.

NOTE: this is a candidate for the 7.9 branch.

Reported-by: Kenneth Waters <kwaters@chromium.org>
2010-11-10 16:00:03 -08:00
Jakob Bornecrantz
0faa7ada84 libgl-xlib: Use sw helper instead of roll your own 2010-11-10 23:40:18 +00:00
Jakob Bornecrantz
89deebb1af graw: Use inline sw helper instead of roll your own loader 2010-11-10 23:05:17 +00:00
Jakob Bornecrantz
d4c60575f8 galahad: Correct the name of the scons library 2010-11-10 23:05:17 +00:00
Jerome Glisse
8e0230a85c r600g: allow driver to work without submitting cmd to GPU
For driver performance analysis it usefull to be able to
disable as much as possible the GPU interaction so that
one can profile the userspace only.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
2010-11-10 16:49:50 -05:00
Robert Hooker
e8b2d36723 intel: Add a new B43 pci id.
Signed-off-by: Robert Hooker <robert.hooker@canonical.com>
2010-11-10 13:37:47 -08:00
Eric Anholt
9effc1adf1 i965: re-enable gen6 IF statements in the fragment shader.
IF statements were getting flattened while they were broken.  With
Zhenyu's last fix for ENDIF's type, everything appears to have lined
up to actually work.

This regresses two tests:
glsl1-! (not) operator (1, fail)
glsl1-! (not) operator (1, pass)

but fixes tests that couldn't work before because the IFs couldn't be
flattened:
glsl-fs-discard-01
occlusion-query-discard

(and, naturally, this should be a performance improvement for apps
that actually use IF statements to avoid executing a bunch of code).
2010-11-10 13:33:27 -08:00
Eric Anholt
490c23ee6b i965: Work around strangeness in swizzling/masking of gen6 math.
Sometimes we swizzled in a different channel it looked like, and
sometimes we swizzled in zero.  Or something.

Having looked at the output of another code generator for this chip,
this is approximately what they do, too: use align1 math on
temporaries, and then move the results into place.

Fixes:
glean/vp1-EX2 test
glean/vp1-EXP test
glean/vp1-LG2 test
glean/vp1-RCP test (reciprocal)
glean/vp1-RSQ test 1 (reciprocal square root)
shaders/glsl-cos
shaders/glsl-sin
shaders/glsl-vs-masked-cos
shaders/vpfp-generic/vp-exp-alias
2010-11-10 12:36:23 -08:00
Francisco Jerez
47c471f281 meta: Handle bitmaps with alpha test enabled.
Acked-by: Brian Paul <brianp@vmware.com>
2010-11-10 21:24:31 +01:00
Zack Rusin
f623d0c1c2 gallivm: implement indirect addressing over inputs
Instead of messing with the callers simply copy our inputs into a
alloca array at the beginning of the function and then use it.

Reviewed-by: José Fonseca <jfonseca@vmware.com>
2010-11-10 13:00:35 -05:00
Roland Scheidegger
aad65fa112 mesa: remove unneeded DD_POINT_SIZE and DD_LINE_WIDTH tricaps
DD_POINT_SIZE was broken for quite some time, and the only driver (r200) relying
on this no longer needs it.
Both DD_POINT_SIZE and DD_LINE_WIDTH have no users left outside of debugging
output, hence instead of fixing DD_POINT_SIZE setting just drop both of them -
there was a plan to remove tricaps flags entirely at some point.
2010-11-10 17:24:42 +01:00
Roland Scheidegger
c7192ab11f r200: fix r200 large points
DD_POINT_SIZE got never set for some time now (as it was set only in ifdefed
out code), which caused the r200 driver to use the point primitive mistakenly
in some cases which can only do size 1 instead of point sprite. Since the
logic to use point instead of point sprite prim is flaky at best anyway (can't
work correctly for per-vertex point size), just drop this and always emit point
sprites (except for AA points) - reasons why the driver tried to use points for
size 1.0 are unknown though it is possible they are faster or more conformant.
Note that we can't emit point sprites without point sprite cntl as that might
result in undefined point sizes, hence need drm version check (which was
unnecessary before as it should always have selected points). An
alternative would be to rely on the RE point size clamp controls which could
clamp the size to 1.0 min/max even if the SE point size is undefined, but currently
always use 0 for min clamp. (As a side note, this also means the driver does
not honor the gl spec which mandates points, but not point sprites, with zero size
to be rendered as size 1.)
This should fix recent reports of https://bugs.freedesktop.org/show_bug.cgi?id=702.
This is a candidate for the mesa 7.9 branch.
2010-11-10 17:24:41 +01:00
Chia-I Wu
aa139a14ba egl_dri2: Fix __DRI_DRI2 version 1 support.
Correctly set __DRI_API_OPENGL flag.
2010-11-10 23:57:50 +08:00
Marek Olšák
93c04749be r300g: turn magic numbers into names in the hyperz code 2010-11-10 15:14:25 +01:00
Marek Olšák
1d28936dea r300g: rename has_hyperz -> can_hyperz 2010-11-10 15:14:25 +01:00
Marek Olšák
88ddfc57e4 r300g: mention ATI in the renderer string 2010-11-10 15:14:25 +01:00
Keith Whitwell
9a04eca2f8 ws/r600: match bo_busy shared/fence logic in bo_wait
Fixes crash in piglit depthrange-clear.
2010-11-10 11:04:38 +00:00
Vinson Lee
5b8ed2f6d2 mesa: Clean up header file inclusion in pixelstore.h. 2010-11-10 00:49:53 -08:00
Vinson Lee
700add5707 mesa: Clean up header file inclusion in pixel.h. 2010-11-10 00:46:27 -08:00
Zhenyu Wang
65972f992f Revert "i965: VS use SPF mode on sandybridge for now"
This reverts commit 9c39a9fcb2.

Remove VS SPF mode, conditional instruction works for VS now.
2010-11-10 08:17:45 -05:00
Zhenyu Wang
9249af17b8 i965: fix dest type of 'endif' on sandybridge
That should also be immediate value for type W.
2010-11-10 08:17:29 -05:00
Eric Anholt
f289dcd849 i965: Add support for math on constants in gen6 brw_wm_glsl.c path.
Fixes 10 piglit cases that were assertion failing.
2010-11-09 20:20:00 -08:00
Ian Romanick
ad8cb131d8 ir_to_mesa: Refactor code for emitting DP instructions 2010-11-09 18:09:41 -08:00
Eric Anholt
f00929cbdd i965: Allow OPCODE_SWZ to put immediates in the first arg.
Fixes assertion failure with texture swizzling in the GLSL path when
it's triggered (such as gen6 FF or ARB_fp shadow comparisons).

Fixes:
texdepth
texSwizzle
fp1-DST test
fp1-LIT test 3
2010-11-09 17:18:52 -08:00
Kenneth Graunke
afb6fb9a92 glsl: Remove unnecessary "unused variable" warning suppression.
The "instructions" variable -is- used, so the cast to void can go away.
2010-11-09 15:55:40 -08:00
Peter Clifton
efb0417040 intel: Add assert check for blitting alignment.
Also fixup code comment to reflect that the GPU requires DWORD
alignment, but in this case does not actually pass the value "in
DWORDs" as I previously stated.
2010-11-09 14:35:28 -08:00
Eric Anholt
00391c7941 Revert "intel: Fix the client-side swapbuffers throttling."
This reverts commit 76360d6abc.  On
second thought, it turned out that sync objects also used the
wait_rendering API like this, and would need the same treatment, and
so wait_rendering itself is fixed in libdrm now.
2010-11-09 14:03:04 -08:00
Eric Anholt
76360d6abc intel: Fix the client-side swapbuffers throttling.
We were asking for a wait to GTT read (all GPU rendering to it
complete), instead of asking for all GPU reading from it to be
complete.  Prevents swapbuffers-based apps from running away with
rendering, and produces a better input experience.
2010-11-09 13:30:27 -08:00
Ian Romanick
956ae44dcf glsl: Fix incorrect gl_type of sampler2DArray and sampler1DArrayShadow
NOTE: this is a candidate for the 7.9 branch.
2010-11-09 13:05:07 -08:00
José Fonseca
10740acf46 gallivm: Allocate TEMP/OUT arrays only once. 2010-11-09 20:36:28 +00:00
Zack Rusin
528c3cd241 gallivm: implement indirect addressing of the output registers 2010-11-09 20:36:28 +00:00
Vinson Lee
520140a6c9 winsys/xlib: Add cygwin to SConscript.
Fixes SCons NameError exception on Cygwin.
2010-11-09 12:31:11 -08:00
Keith Whitwell
63c3e3a3dc r600: fix my pessimism about PIPE_TRANSFER_x flags
For some reason I though we needed the _DISCARD flag to avoid
readbacks, which isn't true at all.  Now write operations should
pipeline properly, gives a good speedup to demos/tunnel.
2010-11-09 20:12:46 +00:00
Keith Whitwell
9f7ec103e2 r600g: translate ARR instruction 2010-11-09 20:12:46 +00:00
Keith Whitwell
c2c55547dc r600g: attempt to turn on DXTn formats
Seems to sort-of work for non-mipmapped textures.  Better than just
black anyway.
2010-11-09 20:12:46 +00:00
Keith Whitwell
e3ea4aec03 r600g: avoid recursion with staged uploads
Don't use an intermediate for formats which don't support hardware
blits under u_blitter.c, as these will recursively attempt to create a
transfer.
2010-11-09 20:12:46 +00:00
Brian Paul
6e2e136428 mesa: no-op glBufferSubData() on size==0
Fixes http://bugs.freedesktop.org/show_bug.cgi?id=31439

NOTE: this is a candidate for the 7.9 branch
2010-11-09 12:24:51 -07:00
Brian Paul
61ea76c8da softpipe: can't no-op depth test stage when occlusion query is enabled
Fixes http://bugs.freedesktop.org/show_bug.cgi?id=31479
2010-11-09 11:44:34 -07:00
Chia-I Wu
5b6ec5a553 st/dri: Add support for surfaceless current contexts.
Tested with Wayland.
2010-11-10 02:01:04 +08:00
Chia-I Wu
3418f74a94 docs: Update egl docs. 2010-11-10 01:31:38 +08:00
Chia-I Wu
dbacbb8219 autoconf: Add --enable-gallium-egl.
This option comes handy when we want to build gallium DRI drivers but
not st/egl.
2010-11-10 00:57:49 +08:00
Vinson Lee
3e6a05b1aa mesa: Clean up header file inclusion in nvprogram.h. 2010-11-09 06:22:25 -08:00
Vinson Lee
0c123679fc mesa: Clean up header file inclusion in multisample.h. 2010-11-09 06:08:29 -08:00
Vinson Lee
c509bf91ec mesa: Clean up header file inclusion in matrix.h. 2010-11-09 06:00:01 -08:00
Vinson Lee
e09800432b mesa: Clean up header file inclusion in lines.h. 2010-11-09 05:47:17 -08:00
Vinson Lee
a20e440c65 mesa: Clean up header file inclusion in light.h. 2010-11-09 05:35:24 -08:00
Vinson Lee
934fc80b06 mesa: Add missing header and forward declarations in dd.h. 2010-11-09 05:13:48 -08:00
Vinson Lee
90394b2d96 mesa: Clean up header file inclusion in image.h. 2010-11-09 05:00:44 -08:00
Thomas Hellstrom
24c6c41bd0 gallium/targets: Trivial crosscompiling fix
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
2010-11-09 12:50:12 +01:00
Thomas Hellstrom
0d5b4b320c svga/drm: Optionally resolve calls to powf during link-time
When linked with certain builds of libstdc++, it appears like powf is resolved
by a symbol in that library. Other builds of libstdc++ doesn't contain that
symbol resulting in a linker / loader error. Optionally
resolve that symbol and replace it with calls to logf and expf.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
2010-11-09 12:31:25 +01:00
Thomas Hellstrom
8e630fad72 st/egl: Fix build for include files in nonstandard places
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
2010-11-09 12:31:24 +01:00
Thomas Hellstrom
6af2a7fe2c mesa: Add talloc includes for gles
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
2010-11-09 12:31:24 +01:00
Thomas Hellstrom
675aec8178 egl: Add an include for size_t
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
2010-11-09 12:31:24 +01:00
Zack Rusin
9d9df964c4 scons: build the xorg state trackers only when env includes drm 2010-11-09 10:41:59 +00:00
Vinson Lee
d79d942b2e mesa: Clean up header file inclusion in histogram.h. 2010-11-09 01:14:55 -08:00
Vinson Lee
5b3d6bd39e mesa: Clean up header file inclusion in hint.h. 2010-11-09 01:12:34 -08:00
Vinson Lee
63f1740a5d mesa: Clean up header file inclusion in framebuffer.h. 2010-11-09 01:04:22 -08:00
Vinson Lee
b35d3b33e7 mesa: Clean up header file inclusion in fog.h. 2010-11-09 00:58:46 -08:00
Vinson Lee
08354667a3 mesa: Clean up header file inclusion in ffvertex_prog.h. 2010-11-09 00:56:02 -08:00
Vinson Lee
6121730e74 mesa: Clean up header file inclusion in fbobject.h. 2010-11-09 00:52:49 -08:00
Chad Versace
b62c1c4595 glsl: Fix ir_expression::constant_expression_value()
When the type of the ir_expression is error_type, return NULL.
This fixes bug 31371.
2010-11-09 00:50:54 -08:00
Johann Rudloff
d7855ee332 radeon: Implement GL_OES_EGL_image
agd5f: add support to radeon/r200/r300 as well
2010-11-08 19:59:53 -05:00
Johann Rudloff
b42e562a11 radeon: Implement __DRI_IMAGE and EGL_MESA_image_drm 2010-11-08 19:59:53 -05:00
Alex Deucher
4990b771de egl_dri2: Add radeon chip ids 2010-11-08 19:59:53 -05:00
Johann Rudloff
f9b5201dbd radeon: Implement EGL_MESA_no_surface_extension 2010-11-08 19:59:53 -05:00
Kenneth Graunke
a457ca7844 ir_dead_functions: Actually free dead functions and signatures.
This makes linked shaders use around 36k less memory since the
built-in prototypes are now freed.
2010-11-08 16:22:15 -08:00
Hui Qi Tay
315f8daab1 llvmpipe: added llvm offset setup code 2010-11-04 12:57:30 +00:00
Hui Qi Tay
7f0dc5ea1b llvmpipe: Moved draw pipeline twoside function to llvm setup code 2010-11-01 14:14:55 +00:00
Keith Whitwell
a1ca5ac7c2 llvmpipe: turn off draw offset/twoside when we can handle it 2010-10-22 18:58:36 +01:00
1023 changed files with 66787 additions and 49854 deletions

View File

@@ -180,7 +180,7 @@ ultrix-gcc:
# Rules for making release tarballs
VERSION=7.10-devel
VERSION=7.10.3
DIRECTORY = Mesa-$(VERSION)
LIB_NAME = MesaLib-$(VERSION)
GLUT_NAME = MesaGLUT-$(VERSION)
@@ -197,6 +197,9 @@ MAIN_FILES = \
$(DIRECTORY)/configure.ac \
$(DIRECTORY)/acinclude.m4 \
$(DIRECTORY)/aclocal.m4 \
$(DIRECTORY)/SConstruct \
$(DIRECTORY)/common.py \
$(DIRECTORY)/scons/*py \
$(DIRECTORY)/bin/config.guess \
$(DIRECTORY)/bin/config.sub \
$(DIRECTORY)/bin/install-sh \
@@ -209,7 +212,6 @@ MAIN_FILES = \
$(DIRECTORY)/docs/README.* \
$(DIRECTORY)/docs/RELNOTES* \
$(DIRECTORY)/docs/*.spec \
$(DIRECTORY)/include/GL/internal/glcore.h \
$(DIRECTORY)/include/GL/gl.h \
$(DIRECTORY)/include/GL/glext.h \
$(DIRECTORY)/include/GL/gl_mangle.h \
@@ -224,6 +226,7 @@ MAIN_FILES = \
$(DIRECTORY)/include/GL/vms_x_fix.h \
$(DIRECTORY)/include/GL/wglext.h \
$(DIRECTORY)/include/GL/wmesa.h \
$(DIRECTORY)/include/c99/*.h \
$(DIRECTORY)/src/glsl/Makefile \
$(DIRECTORY)/src/glsl/Makefile.template \
$(DIRECTORY)/src/glsl/SConscript \
@@ -233,7 +236,9 @@ MAIN_FILES = \
$(DIRECTORY)/src/glsl/glcpp/*.[chly] \
$(DIRECTORY)/src/glsl/glcpp/README \
$(DIRECTORY)/src/Makefile \
$(DIRECTORY)/src/SConscript \
$(DIRECTORY)/src/mesa/Makefile* \
$(DIRECTORY)/src/mesa/SConscript \
$(DIRECTORY)/src/mesa/sources.mak \
$(DIRECTORY)/src/mesa/descrip.mms \
$(DIRECTORY)/src/mesa/gl.pc.in \
@@ -259,6 +264,7 @@ MAIN_FILES = \
$(DIRECTORY)/src/mesa/tnl_dd/*.[ch] \
$(DIRECTORY)/src/mesa/tnl_dd/imm/*.[ch] \
$(DIRECTORY)/src/mesa/tnl_dd/imm/NOTES.imm \
$(DIRECTORY)/src/mesa/vf/*.[ch] \
$(DIRECTORY)/src/mesa/drivers/Makefile \
$(DIRECTORY)/src/mesa/drivers/beos/*.cpp \
$(DIRECTORY)/src/mesa/drivers/beos/Makefile \
@@ -271,6 +277,9 @@ MAIN_FILES = \
$(DIRECTORY)/src/mesa/drivers/osmesa/descrip.mms \
$(DIRECTORY)/src/mesa/drivers/osmesa/osmesa.def \
$(DIRECTORY)/src/mesa/drivers/osmesa/*.[ch] \
$(DIRECTORY)/src/mesa/drivers/dri/r300/compiler/*.[ch] \
$(DIRECTORY)/src/mesa/drivers/dri/r300/compiler/Makefile \
$(DIRECTORY)/src/mesa/drivers/dri/r300/compiler/SConscript \
$(DIRECTORY)/src/mesa/drivers/windows/*/*.[ch] \
$(DIRECTORY)/src/mesa/drivers/windows/*/*.def \
$(DIRECTORY)/src/mesa/drivers/x11/Makefile \
@@ -307,9 +316,9 @@ MAPI_FILES = \
$(DIRECTORY)/src/mapi/mapi/sources.mak \
$(DIRECTORY)/src/mapi/mapi/*.[ch] \
$(DIRECTORY)/src/mapi/vgapi/Makefile \
$(DIRECTORY)/src/mapi/vgapi/SConscript \
$(DIRECTORY)/src/mapi/vgapi/vgapi.csv \
$(DIRECTORY)/src/mapi/vgapi/vg.pc.in \
$(DIRECTORY)/src/mapi/vgapi/*.h
$(DIRECTORY)/src/mapi/vgapi/vg.pc.in
EGL_FILES = \
$(DIRECTORY)/include/KHR/*.h \
@@ -320,6 +329,7 @@ EGL_FILES = \
$(DIRECTORY)/src/egl/*/*.[ch] \
$(DIRECTORY)/src/egl/*/*/Makefile \
$(DIRECTORY)/src/egl/*/*/*.[ch] \
$(DIRECTORY)/src/egl/main/SConscript \
$(DIRECTORY)/src/egl/main/*.pc.in \
$(DIRECTORY)/src/egl/main/*.def
@@ -344,26 +354,34 @@ GALLIUM_FILES = \
$(DIRECTORY)/src/gallium/*/*/*/*.[ch] \
$(DIRECTORY)/src/gallium/*/*/*/*.py
APPLE_DRI_FILES = \
$(DIRECTORY)/src/glx/apple/Makefile \
$(DIRECTORY)/src/glx/apple/*.[ch] \
$(DIRECTORY)/src/glx/apple/*.tcl \
$(DIRECTORY)/src/glx/apple/apple_exports.list \
$(DIRECTORY)/src/glx/apple/GL_aliases \
$(DIRECTORY)/src/glx/apple/GL_extensions \
$(DIRECTORY)/src/glx/apple/GL_noop \
$(DIRECTORY)/src/glx/apple/GL_promoted \
$(DIRECTORY)/src/glx/apple/specs/*.spec \
$(DIRECTORY)/src/glx/apple/specs/*.tm
DRI_FILES = \
$(DIRECTORY)/include/GL/internal/dri_interface.h \
$(DIRECTORY)/include/GL/internal/glcore.h \
$(DIRECTORY)/include/GL/internal/sarea.h \
$(DIRECTORY)/src/glx/Makefile \
$(DIRECTORY)/src/glx/Makefile \
$(DIRECTORY)/src/glx/*.[ch] \
$(APPLE_DRI_FILES) \
$(DIRECTORY)/src/mesa/drivers/dri/Makefile \
$(DIRECTORY)/src/mesa/drivers/dri/Makefile.template \
$(DIRECTORY)/src/mesa/drivers/dri/dri.pc.in \
$(DIRECTORY)/src/mesa/drivers/dri/common/xmlpool/*.[ch] \
$(DIRECTORY)/src/mesa/drivers/dri/common/xmlpool/*.po \
$(DIRECTORY)/src/mesa/drivers/dri/*/*.[chS] \
$(DIRECTORY)/src/mesa/drivers/dri/*/*.cpp \
$(DIRECTORY)/src/mesa/drivers/dri/*/*/*.[chS] \
$(DIRECTORY)/src/mesa/drivers/dri/*/Makefile \
$(DIRECTORY)/src/mesa/drivers/dri/*/*/Makefile \
$(DIRECTORY)/src/mesa/drivers/dri/*/Doxyfile \
$(DIRECTORY)/src/mesa/drivers/dri/*/server/*.[ch]
$(DIRECTORY)/src/mesa/drivers/dri/*/Doxyfile
SGI_GLU_FILES = \
$(DIRECTORY)/src/glu/Makefile \
@@ -399,6 +417,7 @@ GLUT_FILES = \
$(DIRECTORY)/include/GL/glut.h \
$(DIRECTORY)/include/GL/glutf90.h \
$(DIRECTORY)/src/glut/glx/Makefile* \
$(DIRECTORY)/src/glut/glx/SConscript \
$(DIRECTORY)/src/glut/glx/depend \
$(DIRECTORY)/src/glut/glx/glut.pc.in \
$(DIRECTORY)/src/glut/glx/*def \

View File

@@ -32,7 +32,6 @@ import common
opts = Variables('config.py')
common.AddOptions(opts)
opts.Add(EnumVariable('MSVS_VERSION', 'MS Visual C++ version', None, allowed_values=('7.1', '8.0', '9.0')))
env = Environment(
options = opts,

View File

@@ -91,3 +91,4 @@ def AddOptions(opts):
opts.Add(BoolOption('llvm', 'use LLVM', default_llvm))
opts.Add(BoolOption('debug', 'DEPRECATED: debug build', 'yes'))
opts.Add(BoolOption('profile', 'DEPRECATED: profile build', 'no'))
opts.Add(EnumOption('MSVS_VERSION', 'MS Visual C++ version', None, allowed_values=('7.1', '8.0', '9.0')))

View File

@@ -34,9 +34,6 @@ LLVM_LIBS = @LLVM_LIBS@
GLW_CFLAGS = @GLW_CFLAGS@
GLUT_CFLAGS = @GLUT_CFLAGS@
TALLOC_LIBS = @TALLOC_LIBS@
TALLOC_CFLAGS = @TALLOC_CFLAGS@
# dlopen
DLOPEN_LIBS = @DLOPEN_LIBS@
@@ -56,6 +53,10 @@ INSTALL = @INSTALL@
PYTHON2 = @PYTHON2@
PYTHON_FLAGS = -t -O -O
# Flex and Bison for GLSL compiler
FLEX = @FLEX@
BISON = @BISON@
# Library names (base name)
GL_LIB = GL
GLU_LIB = GLU

View File

@@ -31,21 +31,23 @@ CXXFLAGS = -ggdb3 -Os -Wall -fno-strict-aliasing \
-I$(INSTALL_DIR)/include -I$(X11_DIR)/include $(OPT_FLAGS) $(PIC_FLAGS) $(ARCH_FLAGS) $(ASM_FLAGS) $(DEFINES)
# Library names (actual file names)
GL_LIB_NAME = libGL.dylib
GLU_LIB_NAME = libGLU.dylib
GLUT_LIB_NAME = libglut.dylib
GLW_LIB_NAME = libGLw.dylib
OSMESA_LIB_NAME = libOSMesa.dylib
GL_LIB_NAME = lib$(GL_LIB).dylib
GLU_LIB_NAME = lib$(GLU_LIB).dylib
GLUT_LIB_NAME = lib$(GLUT_LIB).dylib
GLW_LIB_NAME = lib$(GLW_LIB).dylib
OSMESA_LIB_NAME = lib$(OSMESA_LIB).dylib
VG_LIB_NAME = lib$(VG_LIB).dylib
# globs used to install the lib and all symlinks
GL_LIB_GLOB = libGL.*dylib
GLU_LIB_GLOB = libGLU.*dylib
GLUT_LIB_GLOB = libglut.*dylib
GLW_LIB_GLOB = libGLw.*dylib
OSMESA_LIB_GLOB = libOSMesa.*dylib
GL_LIB_GLOB = lib$(GL_LIB).*dylib
GLU_LIB_GLOB = lib$(GLU_LIB).*dylib
GLUT_LIB_GLOB = lib$(GLUT_LIB).*dylib
GLW_LIB_GLOB = lib$(GLW_LIB).*dylib
OSMESA_LIB_GLOB = lib$(OSMESA_LIB).*dylib
VG_LIB_GLOB = lib$(VG_LIB).*dylib
GL_LIB_DEPS = -L$(INSTALL_DIR)/$(LIB_DIR) -L$(X11_DIR)/$(LIB_DIR) -lX11 -lXext -lm -lpthread
OSMESA_LIB_DEPS = -L$(TOP)/$(LIB_DIR) -l$(GL_LIB)
OSMESA_LIB_DEPS =
GLU_LIB_DEPS = -L$(TOP)/$(LIB_DIR) -l$(GL_LIB)
GLUT_LIB_DEPS = -L$(TOP)/$(LIB_DIR) -l$(GLU_LIB) -l$(GL_LIB) -L$(INSTALL_DIR)/$(LIB_DIR) -L$(X11_DIR)/$(LIB_DIR) -lX11 -lXmu -lXi -lXext
GLW_LIB_DEPS = -L$(TOP)/$(LIB_DIR) -l$(GL_LIB) -L$(INSTALL_DIR)/$(LIB_DIR) -L$(X11_DIR)/$(LIB_DIR) -lX11 -lXt
@@ -57,3 +59,5 @@ GLU_DIRS = sgi
DRIVER_DIRS = osmesa
#DRIVER_DIRS = dri
DRI_DIRS = swrast
GALLIUM_DRIVERS_DIRS = softpipe trace rbug noop identity galahad failover
#GALLIUM_DRIVERS_DIRS += llvmpipe

View File

@@ -10,7 +10,7 @@ CONFIG_NAME = default
# Version info
MESA_MAJOR=7
MESA_MINOR=10
MESA_TINY=0
MESA_TINY=3
MESA_VERSION = $(MESA_MAJOR).$(MESA_MINOR).$(MESA_TINY)
# external projects. This should be useless now that we use libdrm.
@@ -37,6 +37,8 @@ MKLIB_OPTIONS =
MKDEP = makedepend
MKDEP_OPTIONS = -fdepend
MAKE = make
FLEX = flex
BISON = bison
# Use MINSTALL for installing libraries, INSTALL for everything else
MINSTALL = $(SHELL) $(TOP)/bin/minstall
@@ -82,9 +84,6 @@ GLESv1_CM_LIB_GLOB = $(GLESv1_CM_LIB_NAME)*
GLESv2_LIB_GLOB = $(GLESv2_LIB_NAME)*
VG_LIB_GLOB = $(VG_LIB_NAME)*
TALLOC_LIBS = `pkg-config --libs talloc`
TALLOC_CFLAGS = `pkg-config --cflags talloc`
# Optional assembly language optimization files for libGL
MESA_ASM_SOURCES =
@@ -119,7 +118,7 @@ EGL_CLIENT_APIS = $(GL_LIB)
# Library dependencies
#EXTRA_LIB_PATH ?=
GL_LIB_DEPS = $(EXTRA_LIB_PATH) -lX11 -lXext -lm -lpthread $(TALLOC_LIBS)
GL_LIB_DEPS = $(EXTRA_LIB_PATH) -lX11 -lXext -lm -lpthread
EGL_LIB_DEPS = $(EXTRA_LIB_PATH) -ldl -lpthread
OSMESA_LIB_DEPS = $(EXTRA_LIB_PATH) -L$(TOP)/$(LIB_DIR) -l$(GL_LIB)
GLU_LIB_DEPS = $(EXTRA_LIB_PATH) -L$(TOP)/$(LIB_DIR) -l$(GL_LIB) -lm

View File

@@ -45,7 +45,7 @@ EXTRA_LIB_PATH=-L/usr/X11R6/lib
LIBDRM_CFLAGS = $(shell pkg-config --cflags libdrm)
LIBDRM_LIB = $(shell pkg-config --libs libdrm)
DRI_LIB_DEPS = $(EXTRA_LIB_PATH) -lm -lpthread -lexpat -ldl -ltalloc $(LIBDRM_LIB)
DRI_LIB_DEPS = $(EXTRA_LIB_PATH) -lm -lpthread -lexpat -ldl $(LIBDRM_LIB)
GL_LIB_DEPS = $(EXTRA_LIB_PATH) -lX11 -lXext -lXxf86vm -lXdamage -lXfixes \
-lm -lpthread -ldl $(LIBDRM_LIB)

View File

@@ -15,7 +15,7 @@ ARCH_FLAGS = -mmmx -msse -msse2 -mstackrealign
DEFINES += -DNDEBUG -DGALLIUM_LLVMPIPE -DHAVE_UDIS86
# override -std=c99
CFLAGS += -std=gnu99 -D__STDC_CONSTANT_MACROS
CFLAGS += -std=gnu99
LLVM_VERSION := $(shell llvm-config --version)
@@ -30,7 +30,7 @@ else
endif
ifeq ($(MESA_LLVM),1)
# LLVM_CFLAGS=`llvm-config --cflags`
LLVM_CFLAGS=`llvm-config --cppflags`
LLVM_CXXFLAGS=`llvm-config --cxxflags backend bitreader engine ipo interpreter instrumentation` -Wno-long-long
LLVM_LDFLAGS = $(shell llvm-config --ldflags backend bitreader engine ipo interpreter instrumentation)
LLVM_LIBS = $(shell llvm-config --libs backend bitwriter bitreader engine ipo interpreter instrumentation)
@@ -41,4 +41,4 @@ else
endif
LD = g++
GL_LIB_DEPS = $(LLVM_LDFLAGS) $(LLVM_LIBS) $(EXTRA_LIB_PATH) -lX11 -lXext -lm -lpthread -ltalloc -lstdc++ -ludis86
GL_LIB_DEPS = $(LLVM_LDFLAGS) $(LLVM_LIBS) $(EXTRA_LIB_PATH) -lX11 -lXext -lm -lpthread -lstdc++ -ludis86

View File

@@ -18,7 +18,7 @@ AC_CONFIG_AUX_DIR([bin])
AC_CANONICAL_HOST
dnl Versions for external dependencies
LIBDRM_REQUIRED=2.4.15
LIBDRM_REQUIRED=2.4.24
LIBDRM_RADEON_REQUIRED=2.4.17
DRI2PROTO_REQUIRED=2.1
GLPROTO_REQUIRED=1.4.11
@@ -38,6 +38,12 @@ if test "x$MKDEP" = "x"; then
AC_MSG_ERROR([makedepend is required to build Mesa])
fi
AC_PATH_PROG([FLEX], [flex])
test "x$FLEX" = "x" && AC_MSG_ERROR([flex is needed to build Mesa])
AC_PATH_PROG([BISON], [bison])
test "x$BISON" = "x" && AC_MSG_ERROR([bison is needed to build Mesa])
dnl Our fallback install-sh is a symlink to minstall. Use the existing
dnl configuration in that case.
AC_PROG_INSTALL
@@ -572,10 +578,6 @@ xno)
;;
esac
PKG_CHECK_MODULES([TALLOC], [talloc])
AC_SUBST([TALLOC_LIBS])
AC_SUBST([TALLOC_CFLAGS])
dnl
dnl Driver specific build directories
dnl
@@ -714,8 +716,8 @@ xlib)
GL_PC_LIB_PRIV="$GL_LIB_DEPS"
GL_PC_CFLAGS="$X11_INCLUDES"
fi
GL_LIB_DEPS="$GL_LIB_DEPS $SELINUX_LIBS -lm -lpthread $TALLOC_LIBS"
GL_PC_LIB_PRIV="$GL_PC_LIB_PRIV $SELINUX_LIBS -lm -lpthread $TALLOC_LIBS"
GL_LIB_DEPS="$GL_LIB_DEPS $SELINUX_LIBS -lm -lpthread"
GL_PC_LIB_PRIV="$GL_PC_LIB_PRIV $SELINUX_LIBS -lm -lpthread"
# if static, move the external libraries to the programs
# and empty the libraries for libGL
@@ -964,7 +966,7 @@ if test "$mesa_driver" = dri -o "$mesa_driver" = no; then
fi
# put all the necessary libs together
DRI_LIB_DEPS="$SELINUX_LIBS $LIBDRM_LIBS $EXPAT_LIB -lm -lpthread $DLOPEN_LIBS $TALLOC_LIBS"
DRI_LIB_DEPS="$SELINUX_LIBS $LIBDRM_LIBS $EXPAT_LIB -lm -lpthread $DLOPEN_LIBS"
fi
AC_SUBST([DRI_DIRS])
AC_SUBST([EXPAT_INCLUDES])
@@ -1040,12 +1042,12 @@ case "$DRIVER_DIRS" in
*osmesa*)
# only link libraries with osmesa if shared
if test "$enable_static" = no; then
OSMESA_LIB_DEPS="-lm -lpthread $SELINUX_LIBS $DLOPEN_LIBS $TALLOC_LIBS"
OSMESA_LIB_DEPS="-lm -lpthread $SELINUX_LIBS $DLOPEN_LIBS"
else
OSMESA_LIB_DEPS=""
fi
OSMESA_MESA_DEPS=""
OSMESA_PC_LIB_PRIV="-lm -lpthread $SELINUX_LIBS $DLOPEN_LIBS $TALLOC_LIBS"
OSMESA_PC_LIB_PRIV="-lm -lpthread $SELINUX_LIBS $DLOPEN_LIBS"
;;
esac
AC_SUBST([OSMESA_LIB_DEPS])
@@ -1340,6 +1342,27 @@ AC_SUBST([LLVM_VERSION])
dnl
dnl Gallium state trackers configuration
dnl
AC_ARG_ENABLE([gallium-egl],
[AS_HELP_STRING([--enable-gallium-egl],
[enable gallium EGL state tracker @<:@default=auto@:>@])],
[enable_gallium_egl="$enableval"],
[enable_gallium_egl=auto])
if test "x$enable_gallium_egl" = xauto; then
case "$mesa_driver" in
dri|no)
enable_gallium_egl=$enable_egl
;;
*)
enable_gallium_egl=$enable_openvg
;;
esac
fi
case "x$enable_egl$enable_gallium_egl" in
xnoyes)
AC_MSG_ERROR([cannot build Gallium EGL state tracker without EGL])
esac
AC_ARG_WITH([state-trackers],
[AS_HELP_STRING([--with-state-trackers@<:@=DIRS...@:>@],
[comma delimited state_trackers list, e.g.
@@ -1352,8 +1375,6 @@ no)
GALLIUM_STATE_TRACKERS_DIRS=""
;;
yes)
st_egl="no"
# look at what else is built
case "$mesa_driver" in
xlib)
@@ -1362,14 +1383,11 @@ yes)
dri)
GALLIUM_STATE_TRACKERS_DIRS="dri"
HAVE_ST_DRI="yes"
st_egl="yes"
# Have only tested st/xorg on 1.6.0 servers
PKG_CHECK_MODULES(XORG, [xorg-server >= 1.6.0 libdrm >= $LIBDRM_XORG_REQUIRED libkms >= $LIBKMS_XORG_REQUIRED],
HAVE_ST_XORG="yes"; GALLIUM_STATE_TRACKERS_DIRS="$GALLIUM_STATE_TRACKERS_DIRS xorg",
HAVE_ST_XORG="no")
;;
no)
st_egl="yes"
esac
if test "x$enable_egl" = xyes; then
@@ -1378,7 +1396,7 @@ yes)
st_egl="yes"
fi
if test "$st_egl" = yes; then
if test "$enable_gallium_egl" = yes; then
GALLIUM_STATE_TRACKERS_DIRS="$GALLIUM_STATE_TRACKERS_DIRS egl"
HAVE_ST_EGL="yes"
fi
@@ -1451,10 +1469,6 @@ AC_SUBST([EGL_CLIENT_APIS])
if test "x$HAVE_ST_EGL" = xyes; then
GALLIUM_TARGET_DIRS="$GALLIUM_TARGET_DIRS egl"
# define GLX_DIRECT_RENDERING even when the driver is not dri
if test "x$mesa_driver" != xdri -a "x$driglx_direct" = xyes; then
DEFINES="$DEFINES -DGLX_DIRECT_RENDERING"
fi
fi
if test "x$HAVE_ST_XORG" = xyes; then
@@ -1542,7 +1556,7 @@ AC_ARG_ENABLE([gallium-llvm],
if test "x$enable_gallium_llvm" = xyes; then
if test "x$LLVM_CONFIG" != xno; then
LLVM_VERSION=`$LLVM_CONFIG --version`
LLVM_CFLAGS=`$LLVM_CONFIG --cflags`
LLVM_CFLAGS=`$LLVM_CONFIG --cppflags`
LLVM_LIBS="`$LLVM_CONFIG --libs jit interpreter nativecodegen bitwriter` -lstdc++"
if test "x$HAS_UDIS86" != xno; then
@@ -1692,6 +1706,19 @@ if test "x$enable_gallium_swrast" = xyes || test "x$enable_gallium_swrast" = xau
fi
fi
dnl
dnl Gallium noop configuration
dnl
AC_ARG_ENABLE([gallium-noop],
[AS_HELP_STRING([--enable-gallium-noop],
[build gallium radeon @<:@default=disabled@:>@])],
[enable_gallium_noop="$enableval"],
[enable_gallium_noop=auto])
if test "x$enable_gallium_noop" = xyes; then
GALLIUM_DRIVERS_DIRS="$GALLIUM_DRIVERS_DIRS noop"
GALLIUM_TARGET_DIRS="$GALLIUM_TARGET_DIRS dri-noop"
fi
dnl prepend CORE_DIRS to SRC_DIRS
SRC_DIRS="$CORE_DIRS $SRC_DIRS"

View File

@@ -20,7 +20,7 @@ Float textures, renderbuffers some infrastructure done
Framebuffer objects (GL_EXT_framebuffer_object) DONE
Half-float some infrastructure done
Multisample blit DONE
Non-normalized Integer texture/framebuffer formats not started
Non-normalized Integer texture/framebuffer formats ~50% done
1D/2D Texture arrays core Mesa, swrast done
Packed depth/stencil formats DONE
Per-buffer blend and masks (GL_EXT_draw_buffers2) DONE
@@ -31,10 +31,11 @@ Transform feedback (GL_EXT_transform_feedback) ~50% done
glBindBufferRange, glBindBufferBase commands
Vertex array objects (GL_APPLE_vertex_array_object) DONE
sRGB framebuffer format (GL_EXT_framebuffer_sRGB) not started
glClearBuffer commands DONE, except for dispatch
glGetStringi command DONE, except for dispatch
glTexParameterI, glGetTexParameterI commands DONE, except for dispatch
glVertexAttribI commands not started
glClearBuffer commands DONE
glGetStringi command DONE
glTexParameterI, glGetTexParameterI commands DONE
glVertexAttribI commands DONE (but converts int
values to floats)
GL 3.1:
@@ -93,6 +94,18 @@ GL_ARB_texture_buffer_object_rgb32 not started
GL_ARB_texture_cube_map_array not started
GL_ARB_texture_gather not started
GL_ARB_transform_feedback2 not started
GL_ARB_transform_feedback3 not started
GL 4.1:
GLSL 4.1 not started
GL_ARB_ES2_compatibility not started
GL_ARB_get_program_binary not started
GL_ARB_separate_shader_objects some infrastructure done
GL_ARB_shader_precision not started
GL_ARB_vertex_attrib_64bit not started
GL_ARB_viewport_array not started

View File

@@ -83,7 +83,7 @@ Additions to the EGL 1.4 Specification:
EGLImageKHR eglCreateDRMImageMESA(EGLDisplay dpy,
const EGLint *attrib_list);
In the attribute list, pass EGL_WIDTH, EGL_EIGHT and format and
In the attribute list, pass EGL_WIDTH, EGL_HEIGHT and format and
use in the attrib list using EGL_DRM_BUFFER_FORMAT_MESA and
EGL_DRM_BUFFER_USE_MESA. The only format specified by this
extension is EGL_DRM_BUFFER_FORMAT_ARGB32_MESA, where each pixel

View File

@@ -62,6 +62,7 @@ a:visited {
<LI><A HREF="perf.html" target="MainFrame">Performance Tips</A>
<LI><A HREF="extensions.html" target="MainFrame">Mesa Extensions</A>
<LI><A HREF="mangling.html" target="MainFrame">Function Name Mangling</A>
<LI><A href="llvmpipe.html" target="MainFrame">Gallium llvmpipe driver</A>
</ul>
<b>Developer Topics</b>

View File

@@ -21,76 +21,52 @@ When a new release is coming, release candidates (betas) may be found
<p>
Mesa is distributed in several parts:
The Mesa package is named MesaLib-x.y.z.{tar.bz2, tar.gz, zip} where x.y.z
is the version. There are three types of compressed archives.
</p>
<ul>
<li><b>MesaLib-x.y.z</b> - the main Mesa library source code, drivers
and documentation.
</li>
<li><b>MesaDemos-x.y.z</b> - OpenGL demonstration and test programs.
Most of the programs require GLUT (either the
<a href="http://www.opengl.org/resources/libraries/glut"
target="_parent">original GLUT by Mark Kilgard</a> or
<a href="http://freeglut.sourceforge.net" target="_parent">freeglut</a> or
<a href="http://openglut.sourceforge.net" target="_parent">OpenGLUT</a>).
</li>
<li><b>MesaGLUT-x.y.z</b> - Mark Kilgard's GLUT, easily compiled and used
with Mesa. Plus, other implementation of GLUT for DOS, OS/2, BeOS, etc.
</li>
</ul>
<p>
If you're not interested in running the demos, you'll only need the first
package.
There's also the MesaGLUT-x.y.z.{tar.bz2, tar.gz, zip} packages which
contain Mark Kilgard's GLUT library.
This is optional.
Most Linux distributions include an implementation of GLUT (such as freeglut).
</p>
<p>
In the past, the Mesa demos collection was distributed as
MesaDemos-x.y.z.{tar.bz2, tar.gz, zip}.
Now, the
<a href="ftp://ftp.freedesktop.org/pub/mesa/demos/" target="_parent">
Mesa demos</a> are distributed separately.
</p>
<p>
If you're new to this and not sure what you're doing, grab all three packages.
</p>
<p>
The packages are available in .tar.gz, .tar.bz2 and .zip formats.
Other sites might offer additional package formats.
</p>
<H1>Unpacking</H1>
<p>
All the packages should be in the same directory prior to unpacking.
To unpack .tar.gz files:
</p>
<ul>
<li>To unpack .tar.gz files:
<pre>
tar zxf MesaLib-X.Y.tar.gz
tar zxf MesaDemos-X.Y.tar.gz
tar zxf MesaGLUT-X.Y.tar.gz
tar zxf MesaLib-x.y.z.tar.gz
</pre>
or
<pre>
gzcat MesaLib-X.Y.tar.gz | tar xf -
gzcat MesaDemos-X.Y.tar.gz | tar xf -
gzcat MesaGLUT-X.Y.tar.gz | tar xf -
gzcat MesaLib-x.y.z.tar.gz | tar xf -
</pre>
or
<pre>
gunzip MesaLib-X.Y.tar.gz ; tar xf MesaLib-X.Y.tar
gunzip MesaDemos-X.Y.tar.gz ; tar xf MesaDemos-X.Y.tar
gunzip MesaGLUT-X.Y.tar.gz ; tar xf MesaGLUT-X.Y.tar
gunzip MesaLib-x.y.z.tar.gz ; tar xf MesaLib-x.y.z.tar
</pre>
<li>To unpack .tar.bz2 files:
<p>
To unpack .tar.bz2 files:
</p>
<pre>
bunzip2 -c MesaLib-X.Y.tar.gz | tar xf -
bunzip2 -c MesaDemos-X.Y.tar.gz | tar xf -
bunzip2 -c MesaGLUT-X.Y.tar.gz | tar xf -
bunzip2 -c MesaLib-x.y.z.tar.gz | tar xf -
</pre>
<li>To unpack .zip files:
<p>
To unpack .zip files:
</p>
<pre>
unzip MesaLib-X.Y.zip
unzip MesaDemos-X.Y.zip
unzip MesaGLUT-X.Y.zip
unzip MesaLib-x.y.z.zip
</pre>
</ul>
<h1>Contents</h1>
@@ -106,22 +82,13 @@ bin/ - shell scripts for making shared libraries, etc
docs/ - documentation
src/ - source code for libraries
src/mesa - sources for the main Mesa library and device drivers
src/gallium - sources for Gallium and Gallium drivers
src/glu - libGLU source code
src/glx - sources for building libGL with full GLX and DRI support
src/glw - Xt/Motif/OpenGL widget code
</pre>
If you downloaded and unpacked the MesaDemos.X.Y package:
<pre>
progs/demos - original Mesa demos
progs/xdemos - GLX OpenGL/Mesa demos
progs/redbook - examples from the OpenGL Programming Guide
progs/samples - examples from SGI
progs/images/ - image files
</pre>
If you downloaded and unpacked the MesaGLUT.X.Y package:
If you downloaded and unpacked the MesaGLUT.x.y.z package:
<pre>
src/glut - GLUT library source code
</pre>

View File

@@ -19,26 +19,22 @@ API entry points and helper functions for use by the drivers. Drivers are
dynamically loaded by the main library and most of the EGL API calls are
directly dispatched to the drivers.</p>
<p>The driver in use decides the window system to support. For drivers that
support hardware rendering, there are usually multiple drivers supporting the
same window system. Each one of of them supports a certain range of graphics
cards.</p>
<p>The driver in use decides the window system to support.</p>
<h2>Build EGL</h2>
<ol>
<li>
<p>Run <code>configure</code> with the desired state trackers and enable
the Gallium driver for your hardware. For example</p>
<p>Run <code>configure</code> with the desired client APIs and enable
the driver for your hardware. For example</p>
<pre>
$ ./configure --enable-gles-overlay --enable-openvg --enable-gallium-intel
$ ./configure --enable-gles2 --enable-openvg --enable-gallium-nouveau
</pre>
<p>The main library and OpenGL is enabled by default. The first option enables
<a href="opengles.html">OpenGL ES 1.x and 2.x</a>. The second option enables
<a href="openvg.html">OpenVG</a>.
</p>
<p>The main library and OpenGL is enabled by default. The first option above
enables <a href="opengles.html">OpenGL ES 2.x</a>. The second option enables
<a href="openvg.html">OpenVG</a>.</p>
</li>
@@ -80,22 +76,26 @@ types such as <code>EGLNativeDisplayType</code> or
<p>The available platforms are <code>x11</code>, <code>drm</code>,
<code>fbdev</code>, and <code>gdi</code>. The <code>gdi</code> platform can
only be built with SCons.</p>
</li>
<li><code>--enable-gles-overlay</code>
<p>OpenGL is built by default. To build OpenGL ES, this option must be
explicitly given.</p>
only be built with SCons. Unless for special needs, the build system should
select the right platforms automatically.</p>
</li>
<li><code>--enable-gles1</code> and <code>--enable-gles2</code>
<p>Unlike <code>--enable-gles-overlay</code>, which builds one library for each
rendering API, these options enable OpenGL ES support in OpenGL. The result is
one big library that supports multiple APIs.</p>
<p>These options enable OpenGL ES support in OpenGL. The result is one big
internal library that supports multiple APIs.</p>
</li>
<li><code>--enable-gles-overlay</code>
<p>This option enables OpenGL ES as separate internal libraries. This is an
alternative approach to enable OpenGL ES.</p>
<p>This is only supported by <code>egl_gallium</code>. For systems using DRI
drivers, <code>--enable-gles1</code> and <code>--enable-gles2</code> are
suggested instead as all drivers will benefit.</p>
</li>
@@ -105,6 +105,12 @@ one big library that supports multiple APIs.</p>
</li>
<li><code>--enable-gallium-egl</code>
<p>Explicitly enable or disable <code>egl_gallium</code>.</p>
</li>
</ul>
<h2>Use EGL</h2>
@@ -128,6 +134,16 @@ colon-separated directories where the main library will look for drivers, in
addition to the default directory. This variable is ignored for setuid/setgid
binaries.</p>
<p>This variable is usually set to test an uninstalled build. For example, one
may set</p>
<pre>
$ export LD_LIBRARY_PATH=$mesa/lib
$ export EGL_DRIVERS_PATH=$mesa/lib/egl
</pre>
<p>to test a build without installation</p>
</li>
<li><code>EGL_DRIVER</code>
@@ -143,7 +159,12 @@ specific driver. This variable is ignored for setuid/setgid binaries.</p>
<p>This variable specifies the native platform. The valid values are the same
as those for <code>--with-egl-platforms</code>. When the variable is not set,
the main library uses the first platform listed in
<code>--with-egl-platforms</code> as the native platform</p>
<code>--with-egl-platforms</code> as the native platform.</p>
<p>Extensions like <code>EGL_MESA_drm_display</code> define new functions to
create displays for non-native platforms. These extensions are usually used by
applications that support non-native platforms. Setting this variable is
probably required only for some of the demos found in mesa/demo repository.</p>
</li>
@@ -166,12 +187,26 @@ variable to true forces the use of software rendering.</p>
<h2>EGL Drivers</h2>
<ul>
<li><code>egl_dri2</code>
<p>This driver supports both <code>x11</code> and <code>drm</code> platforms.
It functions as a DRI driver loader. For <code>x11</code> support, it talks to
the X server directly using (XCB-)DRI2 protocol.</p>
<p>This driver can share DRI drivers with <code>libGL</code>.</p>
</li>
<li><code>egl_gallium</code>
<p>This driver is based on Gallium3D. It supports all rendering APIs and
hardwares supported by Gallium3D. It is the only driver that supports OpenVG.
The supported platforms are X11, DRM, FBDEV, and GDI.</p>
<p>This driver comes with its own hardware drivers
(<code>pipe_&lt;hw&gt;</code>) and client API modules
(<code>st_&lt;api&gt;</code>).</p>
</li>
<li><code>egl_glx</code>
@@ -180,18 +215,24 @@ The supported platforms are X11, DRM, FBDEV, and GDI.</p>
the EGL API. It supports both direct and indirect rendering when the GLX does.
It is accelerated when the GLX is. As such, it cannot provide functions that
is not available in GLX or GLX extensions.</p>
</li>
<li><code>egl_dri2</code>
<p>This driver supports the X Window System as its window system. It functions
as a DRI2 driver loader. Unlike <code>egl_glx</code>, it has no dependency on
<code>libGL</code>. It talks to the X server directly using (XCB-)DRI2
protocol.</p>
</li>
</ul>
<h2>Packaging</h2>
<p>The ABI between the main library and its drivers are not stable. Nor is
there a plan to stabilize it at the moment. Of the EGL drivers,
<code>egl_gallium</code> has its own hardware drivers and client API modules.
They are considered internal to <code>egl_gallium</code> and there is also no
stable ABI between them. These should be kept in mind when packaging for
distribution.</p>
<p>Generally, <code>egl_dri2</code> is preferred over <code>egl_gallium</code>
when the system already has DRI drivers. As <code>egl_gallium</code> is loaded
before <code>egl_dri2</code> when both are available, <code>egl_gallium</code>
may either be disabled with <code>--disable-gallium-egl</code> or packaged
separately.</p>
<h2>Developers</h2>
<p>The sources of the main library and the classic drivers can be found at

View File

@@ -9,16 +9,38 @@
<H1>Environment Variables</H1>
<p>
Mesa supports the following environment variables:
Normally, no environment variables need to be set. Most of the environment
variables used by Mesa/Gallium are for debugging purposes, but they can
sometimes be useful for debugging end-user issues.
</p>
<H2>LibGL environment variables</H2>
<ul>
<li>LIBGL_DEBUG - If defined debug information will be printed to stderr.
If set to 'verbose' additional information will be printed.
<li>LIBGL_DRIVERS_PATH - colon-separated list of paths to search for DRI drivers
<li>LIBGL_ALWAYS_INDIRECT - forces an indirect rendering context/connection.
<li>LIBGL_ALWAYS_SOFTWARE - if set, always use software rendering
<li>LIBGL_NO_DRAWARRAYS - if set do not use DrawArrays GLX protocol (for debugging)
</ul>
<H2>Core Mesa environment variables</H2>
<ul>
<li>MESA_NO_ASM - if set, disables all assembly language optimizations
<li>MESA_NO_MMX - if set, disables Intel MMX optimizations
<li>MESA_NO_3DNOW - if set, disables AMD 3DNow! optimizations
<li>MESA_NO_SSE - if set, disables Intel SSE optimizations
<li>MESA_DEBUG - if set, error messages are printed to stderr.
If the value of MESA_DEBUG is "FP" floating point arithmetic errors will
generate exceptions.
<li>MESA_DEBUG - if set, error messages are printed to stderr. For example,
if the application generates a GL_INVALID_ENUM error, a corresponding error
message indicating where the error occured, and possibly why, will be
printed to stderr.<br>
If the value of MESA_DEBUG is 'FP' floating point arithmetic errors will
generate exceptions.
<li>MESA_NO_DITHER - if set, disables dithering, overriding glEnable(GL_DITHER)
<li>MESA_TEX_PROG - if set, implement conventional texture env modes with
fragment programs (intended for developers only)
@@ -28,11 +50,14 @@ Setting this variable automatically sets the MESA_TEX_PROG variable as well.
<li>MESA_EXTENSION_OVERRIDE - can be used to enable/disable extensions.
A value such as "GL_EXT_foo -GL_EXT_bar" will enable the GL_EXT_foo extension
and disable the GL_EXT_bar extension.
<li>MESA_GLSL - <a href="shading.html#envvars">shading language options</a>
<li>MESA_GLSL - <a href="shading.html#envvars">shading language compiler options</a>
</ul>
<H2>Mesa Xlib driver environment variables</H2>
<p>
The following are only applicable to the Xlib software driver.
The following are only applicable to the Mesa Xlib software driver.
See the <A HREF="xlibdriver.html">Xlib software driver page</A> for details.
</p>
<ul>
@@ -51,9 +76,8 @@ See the <A HREF="xlibdriver.html">Xlib software driver page</A> for details.
</ul>
<p>
These environment variables are for the Intel i945/i965 drivers:
</p>
<h2>i945/i965 driver environment variables (non-Gallium)</h2>
<ul>
<li>INTEL_STRICT_CONFORMANCE - if set to 1, enable sw fallbacks to improve
OpenGL conformance. If set to 2, always use software rendering.
@@ -62,17 +86,71 @@ These environment variables are for the Intel i945/i965 drivers:
</ul>
<p>
These environment variables are for the Radeon R300 driver:
</p>
<h2>Radeon R300 driver environment variables (non-Gallium)</h2>
<ul>
<li>R300_NO_TCL - if set, disable hardware-accelerated Transform/Clip/Lighting.
</ul>
<h2>EGL environment variables</h2>
<p>
Mesa EGL supports different sets of environment variables. See the
<a href="egl.html">Mesa EGL</a> page for the details.
</p>
<h2>Gallium environment variables</h2>
<ul>
<li>GALLIUM_PRINT_OPTIONS - if non-zero, print all the Gallium environment
variables which are used, and their current values.
<li>GALLIUM_NOSSE - if non-zero, do not use SSE runtime code generation for
shader execution
<li>GALLIUM_NOPPC - if non-zero, do not use PPC runtime code generation for
shader execution
<li>GALLIUM_DUMP_CPU - if non-zero, print information about the CPU on start-up
<li>TGSI_PRINT_SANITY - if set, do extra sanity checking on TGSI shaders and
print any errors to stderr.
<LI>DRAW_FSE - ???
<LI>DRAW_NO_FSE - ???
<li>DRAW_USE_LLVM - if set to zero, the draw module will not use LLVM to execute
shaders, vertex fetch, etc.
</ul>
<h3>Softpipe driver environment variables</h3>
<ul>
<li>SOFTPIPE_DUMP_FS - if set, the softpipe driver will print fragment shaders
to stderr
<li>SOFTPIPE_DUMP_GS - if set, the softpipe driver will print geometry shaders
to stderr
<li>SOFTPIPE_NO_RAST - if set, rasterization is no-op'd. For profiling purposes.
</ul>
<h3>LLVMpipe driver environment variables</h3>
<ul>
<li>LP_NO_RAST - if set LLVMpipe will no-op rasterization
<li>LP_DEBUG - a comma-separated list of debug options is acceptec. See the
source code for details.
<li>LP_PERF - a comma-separated list of options to selectively no-op various
parts of the driver. See the source code for details.
<li>LP_NUM_THREADS - an integer indicating how many threads to use for rendering.
Zero turns of threading completely. The default value is the number of CPU
cores present.
</ul>
<p>
Other Gallium drivers have their own environment variables. These may change
frequently so the source code should be consulted for details.
</p>
<br>
<br>
</BODY>
</HTML>

View File

@@ -12,16 +12,16 @@
<ol>
<li><a href="#unix-x11">Unix / X11</a>
<ul>
<li><a href="#prereq">Prerequisites for DRI and hardware acceleration</a>
<li><a href="#prereq-general">General prerequisites for building</a>
<li><a href="#prereq-dri">Prerequisites for DRI and hardware acceleration</a>
<li><a href="#autoconf">Building with autoconf</a>
<li><a href="#traditional">Building with traditional Makefiles</a>
<li><a href="#libs">The Libraries</a>
<li><a href="#demos">Running the demos
<li><a href="#install">Installing the header and library files
<li><a href="#pkg-config">Building OpenGL programs with pkg-config
</ul>
<li><a href="#windows">Windows</a>
<li><a href="#scons">SCons</a>
<li><a href="#scons">Building with SCons</a>
<li><a href="#other">Other</a>
</ol>
<br>
@@ -31,8 +31,22 @@
<H2>1. Unix/X11 Compilation and Installation</H1>
<a name="prereq">
<h3>1.1 Prerequisites for DRI and hardware acceleration</h3>
<a name="prereq-general">
<h3>1.1 General prerequisites for building</h3>
<ul>
<li>lex / yacc - for building the GLSL compiler.
On Linux systems, flex and bison are used.
Versions 2.5.35 and 2.4.1, respectively, (or later) should work.
</li>
<li>python - Python is needed for building the Gallium components.
Version 2.6.4 or later should work.
</li>
</ul>
<a name="prereq-dri">
<h3>1.2 Prerequisites for DRI and hardware acceleration</h3>
<p>
The following are required for DRI-based hardware acceleration with Mesa:
@@ -49,7 +63,7 @@ version 2.4.15 or later
<a name="autoconf">
<h3>1.2 Building with Autoconf</h3>
<h3>1.3 Building with Autoconf</h3>
<p>
Mesa may be <a href="autoconf.html">built using autoconf</a>.
@@ -59,7 +73,7 @@ If that fails the traditional Mesa build system is available.
<a name="traditional">
<h3>1.3 Building with traditional Makefiles</h3>
<h3>1.4 Building with traditional Makefiles</h3>
<p>
The traditional Mesa build system is based on a collection of pre-defined
@@ -126,7 +140,7 @@ Later, if you want to rebuild for a different configuration run
<a name="libs">
<h3>1.4 The libraries</h3>
<h3>1.5 The libraries</h3>
<p>
When compilation has finished, look in the top-level <code>lib/</code>
@@ -185,81 +199,11 @@ If you built the DRI hardware drivers, you'll also see the DRI drivers:
-rwxr-xr-x 1 brian users 10997120 Jul 21 12:13 unichrome_dri.so
</pre>
<a name="demos">
<h3>1.5 Running the demos</h3>
<p>
If you downloaded/unpacked the MesaDemos-x.y.z.tar.gz archive or
obtained Mesa from CVS, the <b>progs/</b> directory will contain a
bunch of demonstration programs.
If you built with Gallium support, look in lib/gallium/ for Gallium-based
versions of libGL and device drivers.
</p>
<p>
Before running a demo, you'll probably have to set two environment variables
to indicate where the libraries are located. For example:
<p>
<blockquote>
<b>cd lib/</b>
<br>
<b>export LD_LIBRARY_PATH=${PWD}</b>
<br>
<b>export LIBGL_DRIVERS_PATH=${PWD}</b> (if using DRI drivers)
</blockquote>
<p>
Next, change to the Mesa/demos/ directory:
</p>
<blockquote>
<b>cd ../progs/demos</b>
</blockquote>
<p>
Run a demo such as gears:
</p>
<blockquote>
<b>./gears</b>
</blockquote>
<p>
If this doesn't work, try the <b>Mesa/progs/xdemos/glxinfo</b> program
and see that it prints the expected Mesa version number.
</p>
<p>
If you're using Linux or a similar OS, verify that the demo program is
being linked with the proper library files:
</p>
<blockquote>
<b>ldd gears</b>
</blockquote>
<p>
You should see something like this:
</p>
<pre>
libglut.so.3 => /home/brian/Mesa/lib/libglut.so.3 (0x40013000)
libGLU.so.1 => /home/brian/Mesa/lib/libGLU.so.1 (0x40051000)
libGL.so.1 => /home/brian/Mesa/lib/libGL.so.1 (0x400e0000)
libc.so.6 => /lib/i686/libc.so.6 (0x42000000)
libm.so.6 => /lib/i686/libm.so.6 (0x403da000)
libX11.so.6 => /usr/X11R6/lib/libX11.so.6 (0x403fc000)
libXmu.so.6 => /usr/X11R6/lib/libXmu.so.6 (0x404da000)
libXt.so.6 => /usr/X11R6/lib/libXt.so.6 (0x404f1000)
libXi.so.6 => /usr/X11R6/lib/libXi.so.6 (0x40543000)
libstdc++.so.5 => /usr/lib/libstdc++.so.5 (0x4054b000)
libgcc_s.so.1 => /lib/libgcc_s.so.1 (0x405fd000)
libXext.so.6 => /usr/X11R6/lib/libXext.so.6 (0x40605000)
libpthread.so.0 => /lib/i686/libpthread.so.0 (0x40613000)
/lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x40000000)
libdl.so.2 => /lib/libdl.so.2 (0x40644000)
libSM.so.6 => /usr/X11R6/lib/libSM.so.6 (0x40647000)
libICE.so.6 => /usr/X11R6/lib/libICE.so.6 (0x40650000)
</pre>
<p>
Retrace your steps if this doesn't look right.
</p>
<a name="install">

View File

@@ -1,74 +1,116 @@
LLVMPIPE -- a fork of softpipe that employs LLVM for code generation.
<HTML>
<TITLE>llvmpipe</TITLE>
<link rel="stylesheet" type="text/css" href="mesa.css"></head>
<BODY>
<H1>Introduction</H1>
<p>
The Gallium llvmpipe driver is a software rasterizer that uses LLVM to
do runtime code generation.
Shaders, point/line/triangle rasterization and vertex processing are
implemented with LLVM IR which is translated to x86 or x86-64 machine
code.
Also, the driver is multithreaded to take advantage of multiple CPU cores
(up to 8 at this time).
It's the fastest software rasterizer for Mesa.
</p>
Requirements
============
<h1>Requirements</h1>
- A x86 or amd64 processor. 64bit mode is preferred.
<dl>
<dt>An x86 or amd64 processor. 64-bit mode is preferred.</dt>
<dd>
<p>
Support for sse2 is strongly encouraged. Support for ssse3, and sse4.1 will
yield the most efficient code. The less features the CPU has the more
likely is that you ran into underperforming, buggy, or incomplete code.
</p>
<p>
See /proc/cpuinfo to know what your CPU supports.
- LLVM 2.6 (or later)
</p>
</dd>
<dt>LLVM. Version 2.8 recommended. 2.6 or later required.</dt>
<dd>
<p>
<b>NOTE</b>: LLVM 2.8 and earlier will not work on systems that support the
Intel AVX extensions (e.g. Sandybridge). LLVM's code generator will
fail when trying to emit AVX instructions. This was fixed in LLVM 2.9.
</p>
<p>
For Linux, on a recent Debian based distribution do:
</p>
<pre>
aptitude install llvm-dev
</pre>
For a RPM-based distribution do:
</p>
<pre>
yum install llvm-devel
</pre>
<p>
For Windows download pre-built MSVC 9.0 or MinGW binaries from
http://people.freedesktop.org/~jrfonseca/llvm/ and set the LLVM environment
variable to the extracted path.
</p>
<p>
For MSVC there are two set of binaries: llvm-x.x-msvc32mt.7z and
llvm-x.x-msvc32mtd.7z .
</p>
<p>
You have to set the LLVM=/path/to/llvm-x.x-msvc32mtd env var when passing
debug=yes to scons, and LLVM=/path/to/llvm-x.x-msvc32mt when building with
debug=no. This is necessary as LLVM builds as static library so the chosen
MS CRT must match.
</p>
</dd>
The version of LLVM from SVN ("2.7svn") from mid-March 2010 is pretty
stable and has some features not in version 2.6.
<dt>scons (optional)</dt>
</dl>
- scons (optional)
- udis86, http://udis86.sourceforge.net/ (optional). My personal repository
supports more opcodes which haven't been merged upstream yet:
git clone git://anongit.freedesktop.org/~jrfonseca/udis86
cd udis86
./autogen.sh
./configure --with-pic
make
sudo make install
Building
========
<h1>Building</h1>
To build everything on Linux invoke scons as:
<pre>
scons build=debug libgl-xlib
</pre>
Alternatively, you can build it with GNU make, if you prefer, by invoking it as
<pre>
make linux-llvm
</pre>
but the rest of these instructions assume that scons is used.
For windows is everything the except except the winsys:
<pre>
scons build=debug libgl-gdi
</pre>
Using
=====
<h1>Using</h1>
On Linux, building will create a drop-in alternative for libGL.so into
<pre>
build/foo/gallium/targets/libgl-xlib/libGL.so
</pre>
or
<pre>
lib/gallium/libGL.so
</pre>
To use it set the LD_LIBRARY_PATH environment variable accordingly.
@@ -81,26 +123,21 @@ replacing the native ICD driver, but it's quite an advanced usage, so if you
need to ask, don't even try it.
Profiling
=========
<h1>Profiling</h1>
To profile llvmpipe you should pass the options
<pre>
scons build=profile <same-as-before>
</pre>
This will ensure that frame pointers are used both in C and JIT functions, and
that no tail call optimizations are done by gcc.
To better profile JIT code you'll need to build LLVM with oprofile integration.
source_dir=$PWD/llvm-2.6
build_dir=$source_dir/build/profile
install_dir=$source_dir-profile
mkdir -p "$build_dir"
cd "$build_dir" && \
$source_dir/configure \
<pre>
./configure \
--prefix=$install_dir \
--enable-optimized \
--disable-profiling \
@@ -111,43 +148,57 @@ To better profile JIT code you'll need to build LLVM with oprofile integration.
make -C "$build_dir" install
find "$install_dir/lib" -iname '*.a' -print0 | xargs -0 strip --strip-debug
</pre>
The you should define
<pre>
export LLVM=/path/to/llvm-2.6-profile
</pre>
and rebuild.
Unit testing
============
<h1>Unit testing</h1>
<p>
Building will also create several unit tests in
build/linux-???-debug/gallium/drivers/llvmpipe:
</p>
- lp_test_blend: blending
- lp_test_conv: SIMD vector conversion
- lp_test_format: pixel unpacking/packing
</ul>
<li> lp_test_blend: blending
<li> lp_test_conv: SIMD vector conversion
<li> lp_test_format: pixel unpacking/packing
</ul>
<p>
Some of this tests can output results and benchmarks to a tab-separated-file
for posterior analysis, e.g.:
</p>
<pre>
build/linux-x86_64-debug/gallium/drivers/llvmpipe/lp_test_blend -o blend.tsv
</pre>
Development Notes
=================
<h1>Development Notes</h1>
- When looking to this code by the first time start in lp_state_fs.c, and
<ul>
<li>
When looking to this code by the first time start in lp_state_fs.c, and
then skim through the lp_bld_* functions called in there, and the comments
at the top of the lp_bld_*.c functions.
- The driver-independent parts of the LLVM / Gallium code are found in
</li>
<li>
The driver-independent parts of the LLVM / Gallium code are found in
src/gallium/auxiliary/gallivm/. The filenames and function prefixes
need to be renamed from "lp_bld_" to something else though.
- We use LLVM-C bindings for now. They are not documented, but follow the C++
</li>
<li>
We use LLVM-C bindings for now. They are not documented, but follow the C++
interfaces very closely, and appear to be complete enough for code
generation. See
http://npcontemplation.blogspot.com/2008/06/secret-of-llvm-c-bindings.html
for a stand-alone example. See the llvm-c/Core.h file for reference.
</li>
</ul>

View File

@@ -11,6 +11,46 @@
<H1>News</H1>
<h2>June 13, 2011</h2>
<p>
<a href="relnotes-7.10.3.html">Mesa 7.10.3</a> is released. This is a bug
fix release.
</p>
<h2>April 6, 2011</h2>
<p>
<a href="relnotes-7.10.2.html">Mesa 7.10.2</a> is released. This is a bug
fix release.
</p>
<h2>March 2, 2011</h2>
<p>
<a href="relnotes-7.10.1.html">Mesa 7.10.1</a> is released. This is a bug
fix release.
</p>
<p>
Also, <a href="relnotes-7.9.2.html">Mesa 7.9.2</a> is released.
This is a bug fix release.
</p>
<h2>January 7, 2011</h2>
<p>
<a href="relnotes-7.10.html">Mesa 7.10</a> (final) is released. This is a new
development release.
</p>
<p>
Also, <a href="relnotes-7.9.1.html">Mesa 7.9.1</a> (final) is released.
This is a bug fix release.
</p>
<h2>October 4, 2010</h2>
<p>

380
docs/relnotes-7.10.1.html Normal file
View File

@@ -0,0 +1,380 @@
<HTML>
<head>
<TITLE>Mesa Release Notes</TITLE>
<link rel="stylesheet" type="text/css" href="mesa.css">
<meta http-equiv="content-type" content="text/html; charset=utf-8" />
</head>
<BODY>
<body bgcolor="#eeeeee">
<H1>Mesa 7.10.1 Release Notes / March 2, 2011</H1>
<p>
Mesa 7.10.1 is a bug fix release which fixes bugs found since the 7.10 release.
</p>
<p>
Mesa 7.10.1 implements the OpenGL 2.1 API, but the version reported by
glGetString(GL_VERSION) depends on the particular driver being used.
Some drivers don't support all the features required in OpenGL 2.1.
</p>
<p>
See the <a href="install.html">Compiling/Installing page</a> for prerequisites
for DRI hardware acceleration.
</p>
<h2>MD5 checksums</h2>
<pre>
4b4cee19f3bf16eb78bd4cc278ccf812 MesaLib-7.10.1.tar.gz
efe8da4d80c2a5d32a800770b8ce5dfa MesaLib-7.10.1.tar.bz2
0fd2b1a025934de3f8cecf9fb9b57f4c MesaLib-7.10.1.zip
42beb0f5188d544476c19496f725fa67 MesaGLUT-7.10.1.tar.gz
637bb8a20fdad89f7382b4ea83f896e3 MesaGLUT-7.10.1.tar.bz2
bdbf3ffb2606d6aa8afabb6c6243b91b MesaGLUT-7.10.1.zip
</pre>
<h2>New features</h2>
<p>None.</p>
<h2>Bug fixes</h2>
<p>This list is likely incomplete.</p>
<ul>
<li>Fix an off-by-one bug in a vsplit assertion.</li>
<li>Fix incorrect handling of <tt>layout</tt> qualifier
with <tt>in</tt>, <tt>out</tt>, <tt>attribute</tt>, and <tt>varying</tt>.</li>
<li>Fix an i965 shader bug where the negative absolute value was generated instead of the absolute value of a negation.</li>
<li>Fix numerous issues handling precision qualifiers in GLSL ES.</li>
<li>Fixed a few GLX protocol encoder bugs (Julien Cristau)</li>
<li>Assorted Gallium llvmpipe driver bug fixes</li>
<li>Assorted Mesa/Gallium state tracker bug fixes</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=26795">Bug 26795</a> - gl_FragCoord off by one in Gallium drivers.</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29164">Bug 29164</a> - [GLSL 1.20] invariant variable shouldn't be used before declaration</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29823">Bug 29823</a> - GetUniform[if]v busted</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29927">Bug 29927</a> - [glsl2] fail to compile shader with constructor for array of struct type</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30156">Bug 30156</a> - [i965] After updating to Mesa 7.9, Civilization IV starts to show garbage</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31923">Bug 31923</a> - [GLSL 1.20] allowing inconsistent centroid declaration between two vertex shaders</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31925">Bug 31925</a> - [GLSL 1.20] "#pragma STDGL invariant(all)" fail</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32214">Bug 32214</a> - [gles2]no link error happens when missing vertex shader or frag shader</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32375">Bug 32375</a> - [gl gles2] Not able to get the attribute by function glGetVertexAttribfv</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32541">Bug 32541</a> - Segmentation Fault while running an HDR (high dynamic range) rendering demo</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32569">Bug 32569</a> - [gles2] glGetShaderPrecisionFormat not implemented yet</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32695">Bug 32695</a> - [glsl] SIGSEGV glcpp/glcpp-parse.y:833</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32831">Bug 32831</a> - [glsl] division by zero crashes GLSL compiler</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32910">Bug 32910</a> - Keywords 'in' and 'out' not handled properly for GLSL 1.20 shaders</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33219">Bug 33219</a> -[GLSL bisected] implicit sized array triggers segfault in ir_to_mesa_visitor::copy_propagate</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33306">Bug 33306</a> - GLSL integer division by zero crashes GLSL compiler</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33308">Bug 33308</a> -[glsl] ast_to_hir.cpp:3016: virtual ir_rvalue* ast_jump_statement::hir(exec_list*, _mesa_glsl_parse_state*): Assertion `ret != __null' failed.</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33316">Bug 33316</a> - uniform array will be allocate one line more and initialize it when it was freed will abort</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33386">Bug 33386</a> - Dubious assembler in read_rgba_span_x86.S</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33388">Bug 33388</a> - Dubious assembler in xform4.S</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33433">Bug 33433</a> - Error in x86-64 API dispatch code.</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33507">Bug 33507</a> - [glsl] GLSL preprocessor modulus by zero crash</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33508">Bug 33508</a> - [glsl] GLSL compiler modulus by zero crash</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33916">Bug 33916</a> - Compiler accepts reserved operators % and %=</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=34030">Bug 34030</a> - [bisected] Starcraft 2: some effects are corrupted or too big</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=34047">Bug 34047</a> - Assert in _tnl_import_array() when using GLfixed vertex datatypes with GLESv2</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=34114">Bug 34114</a> - Sun Studio build fails due to standard library functions not being in global namespace</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=34179">Bug 34179</a> - Nouveau 3D driver: nv50_pc_emit.c:863 assertion error kills Compiz</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=34198">Bug 34198</a> - [GLSL] implicit sized array with index 0 used gets assertion</li>
<li><a href="https://bugs.launchpad.net/ubuntu/+source/mesa/+bug/691653">Ubuntu bug 691653</a> - compiz crashes when using alt-tab (the radeon driver kills it) </li>
<li><a href="https://bugs.meego.com/show_bug.cgi?id=13005">Meego bug 13005</a> - Graphics GLSL issue lead to camera preview fail on Pinetrail</li>
<!-- <li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=">Bug </a> - </li> -->
</ul>
<h2>Changes</h2>
<p>The full set of changes can be viewed by using the following GIT command:</p>
<pre>
git log mesa-7.10..mesa-7.10.1
</pre>
<p>Alberto Milone (1):
<ul>
<li>r600c: add evergreen ARL support.</li>
</ul></p>
<p>Brian Paul (21):
<ul>
<li>draw: Fix an off-by-one bug in a vsplit assertion.</li>
<li>docs: add links to 7.9.1 and 7.10 release notes</li>
<li>docs: added news item for 7.9.1 and 7.10 release</li>
<li>gallivm: work around LLVM 2.6 bug when calling C functions</li>
<li>gallivm: fix copy&amp;paste error from previous commit</li>
<li>mesa: fix a few format table mistakes, assertions</li>
<li>mesa: fix num_draw_buffers==0 in fixed-function fragment program generation</li>
<li>mesa: don't assert in GetIntegerIndexed, etc</li>
<li>mesa: check for dummy renderbuffer in _mesa_FramebufferRenderbufferEXT()</li>
<li>llvmpipe: make sure binning is active when we begin/end a query</li>
<li>st/mesa: fix incorrect fragcoord.x translation</li>
<li>softpipe: fix off-by-one error in setup_fragcoord_coeff()</li>
<li>cso: fix loop bound in cso_set_vertex_samplers()</li>
<li>st/mesa: fix incorrect glCopyPixels position on fallback path</li>
<li>st/mesa: set renderbuffer _BaseFormat in a few places</li>
<li>st/mesa: fix the default case in st_format_datatype()</li>
<li>st/mesa: need to translate clear color according to surface's base format</li>
<li>docs: update 7.9.2 release notes with Brian's cherry-picks</li>
<li>docs: add link to 7.10.1 release notes</li>
<li>mesa: implement glGetShaderPrecisionFormat()</li>
<li>docs: updated environment variable list</li>
</ul></p>
<p>Bryce Harrington (1):
<ul>
<li>r300g: Null pointer check for buffer deref in gallium winsys</li>
</ul></p>
<p>Chad Versace (20):
<ul>
<li>glsl: At link-time, check that globals have matching centroid qualifiers</li>
<li>glcpp: Fix segfault when validating macro redefinitions</li>
<li>glsl: Fix parser rule for type_specifier</li>
<li>glsl: Change default value of ast_type_specifier::precision</li>
<li>glsl: Add semantic checks for precision qualifiers</li>
<li>glsl: Add support for default precision statements</li>
<li>glsl: Remove redundant semantic check in parser</li>
<li>glsl: Fix semantic checks on precision qualifiers</li>
<li>glsl: Fix segfault due to missing printf argument</li>
<li>glsl: Mark 'in' variables at global scope as read-only</li>
<li>mesa: Refactor handling of extension strings</li>
<li>mesa: Add/remove extensions in extension string</li>
<li>mesa: Change dependencies of some OES extension strings</li>
<li>mesa: Change OES_point_sprite to depend on ARB_point_sprite</li>
<li>mesa: Change OES_standard_derivatives to be stand-alone extension</li>
<li>i915: Disable extension OES_standard_derivatives</li>
<li>glcpp: Raise error when modulus is zero</li>
<li>glsl: Set operators '%' and '%=' to be reserved when GLSL &lt 1.30</li>
<li>glsl: Reinstate constant-folding for division by zero</li>
<li>tnl: Add support for datatype GL_FIXED in vertex arrays</li>
</ul></p>
<p>Chia-I Wu (1):
<ul>
<li>mesa: Add glDepthRangef and glClearDepthf to APIspec.xml.</li>
</ul></p>
<p>Christoph Bumiller (1):
<ul>
<li>nv50,nvc0: do not forget to apply sign mode to saved TGSI inputs</li>
</ul></p>
<p>Cyril Brulebois (1):
<ul>
<li>Point to bugs.freedesktop.org rather than bugzilla.freedesktop.org</li>
</ul></p>
<p>Dave Airlie (3):
<ul>
<li>radeon/r200: fix fbo-clearmipmap + gen-teximage</li>
<li>radeon: calculate complete texture state inside TFP function</li>
<li>radeon: avoid segfault on 3D textures.</li>
</ul></p>
<p>Dimitry Andric (4):
<ul>
<li>mesa: s/movzx/movzbl/</li>
<li>mesa: s/movzxw/movzwl/ in read_rgba_span_x86.S</li>
<li>glapi: adding @ char before type specifier in glapi_x86.S</li>
<li>glapi: add @GOTPCREL relocation type</li>
</ul></p>
<p>Eric Anholt (16):
<ul>
<li>glsl: Fix the lowering of variable array indexing to not lose write_masks.</li>
<li>i965/fs: When producing ir_unop_abs of an operand, strip negate.</li>
<li>i965/vs: When MOVing to produce ABS, strip negate of the operand.</li>
<li>i965/fs: Do flat shading when appropriate.</li>
<li>i965: Avoid double-negation of immediate values in the VS.</li>
<li>intel: Make renderbuffer tiling choice match texture tiling choice.</li>
<li>i965: Fix dead pointers to fp-&gt;Parameters-&gt;ParameterValues[] after realloc.</li>
<li>docs: Add a relnote for the Civ IV on i965.</li>
<li>glapi: Add entrypoints and enums for GL_ARB_ES2_compatibility.</li>
<li>mesa: Add extension enable bit for GL_ARB_ES2_compatibility.</li>
<li>mesa: Add actual support for glReleaseShaderCompiler from ES2.</li>
<li>mesa: Add support for glDepthRangef and glClearDepthf.</li>
<li>mesa: Add getters for ARB_ES2_compatibility MAX_*_VECTORS.</li>
<li>mesa: Add getter for GL_SHADER_COMPILER with ARB_ES2_compatibility.</li>
<li>i965: Fix a bug in i965 compute-to-MRF.</li>
<li>i965/fs: Add a helper function for detecting math opcodes.</li>
</ul></p>
<p>Fredrik Höglund (1):
<ul>
<li>st/mesa: fix a regression from cae2bb76</li>
</ul></p>
<p>Ian Romanick (42):
<ul>
<li>docs: Add 7.10 md5sums</li>
<li>glsl: Support the 'invariant(all)' pragma</li>
<li>glcpp: Generate an error for division by zero</li>
<li>glsl: Add version_string containing properly formatted GLSL version</li>
<li>glsl &amp; glcpp: Refresh autogenerated lexer and parser files.</li>
<li>glsl: Disallow 'in' and 'out' on globals in GLSL 1.20</li>
<li>glsl: Track variable usage, use that to enforce semantics</li>
<li>glsl: Allow 'in' and 'out' when 'layout' is also available</li>
<li>docs: Initial bits of 7.10.1 release notes</li>
<li>mesa: bump version to 7.10.1-devel</li>
<li>doc: Update 7.10.1 release notes</li>
<li>glsl: Emit errors or warnings when 'layout' is used with 'attribute' or 'varying'</li>
<li>docs: Update 7.10.1 release notes</li>
<li>glsl: Refresh autogenerated lexer and parser files.</li>
<li>glsl: Don't assert when the value returned by a function has no rvalue</li>
<li>linker: Set sizes for non-global arrays as well</li>
<li>linker: Propagate max_array_access while linking functions</li>
<li>docs: Update 7.10.1 release notes</li>
<li>mesa: glGetUniform only returns a single element of an array</li>
<li>linker: Generate link errors when ES shaders are missing stages</li>
<li>mesa: Fix error checks in GetVertexAttrib functions</li>
<li>Use C-style system headers in C++ code to avoid issues with std:: namespace</li>
<li>docs: Update 7.10.1 release notes</li>
<li>glapi: Regenerate for GL_ARB_ES2_compatibility.</li>
<li>mesa: Connect glGetShaderPrecisionFormat into the dispatch table</li>
<li>i965: Set correct values for range/precision of fragment shader types</li>
<li>i915: Set correct values for range/precision of fragment shader types</li>
<li>intel: Fix typeos from 3d028024 and 790ff232</li>
<li>glsl: Ensure that all GLSL versions are supported in the stand-alone compiler</li>
<li>glsl: Reject shader versions not supported by the implementation</li>
<li>mesa: Initial size for secondary color array is 3</li>
<li>glsl: Finish out the reduce/reduce error fixes</li>
<li>glsl: Regenerate compiler and glcpp files from cherry picks</li>
<li>linker: Fix off-by-one error implicit array sizing</li>
<li>docs: update 7.10.1 release notes with Ian's recent cherry picks</li>
<li>i915: Only mark a register as available if all components are written</li>
<li>i915: Calculate partial result to temp register first</li>
<li>i915: Force lowering of all types of indirect array accesses in the FS</li>
<li>docs: Update 7.10.1 with (hopefully) the last of the cherry picks</li>
<li>docs: Clean up bug fixes list</li>
<li>intel: Remove driver date and related bits from renderer string</li>
<li>mesa: set version string to 7.10.1 (final)</li>
</ul></p>
<p>Jian Zhao (1):
<ul>
<li>mesa: fix an error in uniform arrays in row calculating.</li>
</ul></p>
<p>Julien Cristau (3):
<ul>
<li>glx: fix request lengths</li>
<li>glx: fix GLXChangeDrawableAttributesSGIX request</li>
<li>glx: fix length of GLXGetFBConfigsSGIX</li>
</ul></p>
<p>Keith Packard (1):
<ul>
<li>glsl: Eliminate reduce/reduce conflicts in glsl grammar</li>
</ul></p>
<p>Kenneth Graunke (20):
<ul>
<li>glsl: Expose a public glsl_type::void_type const pointer.</li>
<li>glsl: Don't bother unsetting a destructor that was never set.</li>
<li>glsl, i965: Remove unnecessary talloc includes.</li>
<li>glcpp: Remove use of talloc reference counting.</li>
<li>ralloc: Add a fake implementation of ralloc based on talloc.</li>
<li>Convert everything from the talloc API to the ralloc API.</li>
<li>ralloc: a new MIT-licensed recursive memory allocator.</li>
<li>Remove talloc from the make and automake build systems.</li>
<li>Remove talloc from the SCons build system.</li>
<li>Remove the talloc sources from the Mesa repository.</li>
<li>glsl: Fix use of uninitialized values in _mesa_glsl_parse_state ctor.</li>
<li>i965/fs: Apply source modifier workarounds to POW as well.</li>
<li>i965: Fix shaders that write to gl_PointSize on Sandybridge.</li>
<li>i965/fs: Avoid register coalescing away gen6 MATH workarounds.</li>
<li>i965/fs: Correctly set up gl_FragCoord.w on Sandybridge.</li>
<li>i965: Increase Sandybridge point size clamp.</li>
<li>i965/fs: Refactor control flow stack handling.</li>
<li>i965: Increase Sandybridge point size clamp in the clip state.</li>
<li>glsl: Use reralloc instead of plain realloc.</li>
<li>Revert "i965/fs: Correctly set up gl_FragCoord.w on Sandybridge."</li>
</ul></p>
<p>Marek Olšák (4):
<ul>
<li>docs: fix messed up names with special characters in relnotes-7.10</li>
<li>docs: fix messed up names with special characters in relnotes-7.9.1</li>
<li>mesa: fix texture3D mipmap generation for UNSIGNED_BYTE_3_3_2</li>
<li>st/dri: Track drawable context bindings</li>
</ul></p>
<p>Paulo Zanoni (1):
<ul>
<li>dri_util: fail driCreateNewScreen if InitScreen is NULL</li>
</ul></p>
<p>Sam Hocevar (2):
<ul>
<li>docs: add glsl info</li>
<li>docs: fix glsl_compiler name</li>
</ul></p>
<p>Tom Fogal (1):
<ul>
<li>Regenerate gl_mangle.h.</li>
</ul></p>
<p>Tom Stellard (2):
<ul>
<li>r300/compiler: Disable register rename pass on r500</li>
<li>r300/compiler: Don't erase sources when converting RGB-&gt;Alpha</li>
</ul></p>
<p>Vinson Lee (3):
<ul>
<li>ralloc: Add missing va_end following va_copy.</li>
<li>mesa: Move declaration before code in extensions.c.</li>
<li>mesa: Move loop variable declarations outside for loop in extensions.c.</li>
</ul></p>
<p>nobled (1):
<ul>
<li>glx: Put null check before use</li>
</ul></p>
</p>
</body>
</html>

206
docs/relnotes-7.10.2.html Normal file
View File

@@ -0,0 +1,206 @@
<HTML>
<head>
<TITLE>Mesa Release Notes</TITLE>
<link rel="stylesheet" type="text/css" href="mesa.css">
<meta http-equiv="content-type" content="text/html; charset=utf-8" />
</head>
<BODY>
<body bgcolor="#eeeeee">
<H1>Mesa 7.10.2 Release Notes / April 6, 2011</H1>
<p>
Mesa 7.10.2 is a bug fix release which fixes bugs found since the 7.10 release.
</p>
<p>
Mesa 7.10.2 implements the OpenGL 2.1 API, but the version reported by
glGetString(GL_VERSION) depends on the particular driver being used.
Some drivers don't support all the features required in OpenGL 2.1.
</p>
<p>
See the <a href="install.html">Compiling/Installing page</a> for prerequisites
for DRI hardware acceleration.
</p>
<h2>MD5 checksums</h2>
<pre>
2f9f444265534a2cfd9a99d1a8291089 MesaLib-7.10.2.tar.gz
f5de82852f1243f42cc004039e10b771 MesaLib-7.10.2.tar.bz2
47836e37bab6fcafe3ac90c9544ba0e9 MesaLib-7.10.2.zip
175120325828f313621cc5bc6c504803 MesaGLUT-7.10.2.tar.gz
8c71d273f5f8d6c5eda4ffc39e0fe03e MesaGLUT-7.10.2.tar.bz2
03036c8efe7b791a90fa0f2c41b43f43 MesaGLUT-7.10.2.zip
</pre>
<h2>New features</h2>
<p>None.</p>
<h2>Bug fixes</h2>
<p>This list is likely incomplete.</p>
<ul>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29172">Bug 29172</a> - Arrandale - Pill Popper Pops Pills</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31159">Bug 31159</a> - shadow problem in 0ad game</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32688">Bug 32688</a> - [RADEON:KMS:R300G] some games have a wireframe or outline visible</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32949">Bug 32949</a> - [glsl wine] Need for Speed renders incorrectly with GLSL enabled</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=34203">Bug 34203</a> - [GLSL] fail to call long chains across shaders</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=34376">Bug 34376</a> - [GLSL] allowing assignment to unsized array
<ul>
<li>The commit message incorrectly
lists <a href="https://bugs.freedesktop.org/show_bug.cgi?id=34367">bug
34367</a>.</li>
</ul>
</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=34370">Bug 34370</a> - [GLSL] "i&lt;5 &amp;&amp; i&lt;4" in for loop fails</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=34374">Bug 34374</a> - [GLSL] fail to redeclare an array using initializer</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=35073">Bug 35073</a> - [GM45] Alpha test is broken when rendering to FBO with no color attachment</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=35483">Bug 35483</a> - util_blit_pixels_writemask: crash in line 322 of src/gallium/auxiliary/util/u_blit.c</li>
<!-- <li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=">Bug </a> - </li> -->
</ul>
<h2>Changes</h2>
<p>The full set of changes can be viewed by using the following GIT command:</p>
<pre>
git log mesa-7.10.1..mesa-7.10.2
</pre>
<p><em>Note:</em> Reverted commits and the reverts are not included in this list.</p>
<p>Alex Deucher (2):
<ul>
<li>r600c: add new ontario pci ids</li>
<li>r600g: add some additional ontario pci ids</li>
</ul></p>
<p>Benjamin Franzke (1):
<ul>
<li>st/dri: Fix surfaceless gl using contexts with previous bound surfaces</li>
</ul></p>
<p>Brian Paul (9):
<ul>
<li>docs: pull 7.9.2 release notes into 7.10 branch</li>
<li>docs: update news.html with 7.10.1 and 7.9.2 releases</li>
<li>docs: fill in 7.10.1 release data</li>
<li>docs: add, fix release notes links</li>
<li>docs: update info about Mesa packaging/contents</li>
<li>docs: update prerequisites, remove old demo info</li>
<li>mesa: Guard against null pointer deref in fbo validation</li>
<li>st/mesa: Apply LOD bias from correct texture unit</li>
<li>glsl: silence warning in printf() with a cast</li>
</ul></p>
<p>Chad Versace (1):
<ul>
<li>i965: Fix tex_swizzle when depth mode is GL_RED</li>
</ul></p>
<p>Dave Airlie (1):
<ul>
<li>r600: don't close fd on failed load</li>
</ul></p>
<p>Eric Anholt (2):
<ul>
<li>i965: Apply a workaround for the Ironlake "vertex flashing".</li>
<li>i965: Fix alpha testing when there is no color buffer in the FBO.</li>
</ul></p>
<p>Fabian Bieler (1):
<ul>
<li>st/mesa: Apply LOD from texture object</li>
</ul></p>
<p>Henri Verbeet (1):
<ul>
<li>st/mesa: Validate state before doing blits.</li>
</ul></p>
<p>Ian Romanick (13):
<ul>
<li>docs: Add 7.10.1 md5sums</li>
<li>glsl: Refactor AST-to-HIR code handling variable initializers</li>
<li>glsl: Refactor AST-to-HIR code handling variable redeclarations</li>
<li>glsl: Process redeclarations before initializers</li>
<li>glsl: Function signatures cannot have NULL return type</li>
<li>glsl: Add several function / call related validations</li>
<li>linker: Add imported functions to the linked IR</li>
<li>glsl: Use insert_before for lists instead of open coding it</li>
<li>glsl: Only allow unsized array assignment in an initializer</li>
<li>glcpp: Refresh autogenerated lexer files</li>
<li>docs: Initial bits of 7.10.2 release notes</li>
<li>mesa: set version string to 7.10.2</li>
<li>mesa: Remove nonexistant files from _FILES lists</li>
</ul></p>
<p>Jerome Glisse (1):
<ul>
<li>r600g: move user fence into base radeon structure</li>
</ul></p>
<p>José Fonseca (2):
<ul>
<li>mesa: Fix typo glGet*v(GL_TEXTURE_COORD_ARRAY_*).</li>
<li>mesa: More glGet* fixes.</li>
</ul></p>
<p>Kenneth Graunke (4):
<ul>
<li>glcpp: Rework lexer to use a SKIP state rather than REJECT.</li>
<li>glcpp: Remove trailing contexts from #if rules.</li>
<li>i965/fs: Fix linear gl_Color interpolation on pre-gen6 hardware.</li>
<li>glsl: Accept precision qualifiers on sampler types, but only in ES.</li>
</ul></p>
<p>Marek Olšák (15):
<ul>
<li>st/mesa: fix crash when DrawBuffer-&gt;_ColorDrawBuffers[0] is NULL</li>
<li>st/mesa: fail to alloc a renderbuffer if st_choose_renderbuffer_format fails</li>
<li>r300/compiler: fix the saturate modifier when applied to TEX instructions</li>
<li>r300/compiler: fix translating the src negate bits in pair_translate</li>
<li>r300/compiler: Abs doesn't cancel Negate (in the conversion to native swizzles)</li>
<li>r300/compiler: TEX instructions don't support negation on source arguments</li>
<li>r300/compiler: do not set TEX_IGNORE_UNCOVERED on r500</li>
<li>r300/compiler: saturate Z before the shadow comparison</li>
<li>r300/compiler: fix equal and notequal shadow compare functions</li>
<li>r300/compiler: remove unused variables</li>
<li>st/mesa: fix crash when using both user and vbo buffers with the same stride</li>
<li>r300g: fix alpha-test with no colorbuffer</li>
<li>r300g: tell the GLSL compiler to lower the continue opcode</li>
<li>r300/compiler: propagate SaturateMode down to the result of shadow comparison</li>
<li>r300/compiler: apply the texture swizzle to shadow pass and fail values too</li>
</ul></p>
<p>Michel Dänzer (1):
<ul>
<li>Use proper source row stride when getting depth/stencil texels.</li>
</ul></p>
<p>Tom Stellard (4):
<ul>
<li>r300/compiler: Use a 4-bit writemask in pair instructions</li>
<li>prog_optimize: Fix reallocating registers for shaders with loops</li>
<li>r300/compiler: Fix vertex shader MAD instructions with constant swizzles</li>
<li>r300/compiler: Don't try to convert RGB to Alpha in full instructions</li>
</ul></p>
</body>
</html>

303
docs/relnotes-7.10.3.html Normal file
View File

@@ -0,0 +1,303 @@
<HTML>
<head>
<TITLE>Mesa Release Notes</TITLE>
<link rel="stylesheet" type="text/css" href="mesa.css">
<meta http-equiv="content-type" content="text/html; charset=utf-8" />
</head>
<BODY>
<body bgcolor="#eeeeee">
<H1>Mesa 7.10.3 Release Notes / June 13, 2011</H1>
<p>
Mesa 7.10.3 is a bug fix release which fixes bugs found since the 7.10.2 release.
</p>
<p>
Mesa 7.10.3 implements the OpenGL 2.1 API, but the version reported by
glGetString(GL_VERSION) depends on the particular driver being used.
Some drivers don't support all the features required in OpenGL 2.1.
</p>
<p>
See the <a href="install.html">Compiling/Installing page</a> for prerequisites
for DRI hardware acceleration.
</p>
<h2>MD5 checksums</h2>
<pre>
d77b02034c11d6c2a55c07f82367d780 MesaLib-7.10.3.tar.gz
8c38fe8266be8e1ed1d84076ba5a703b MesaLib-7.10.3.tar.bz2
614d063ecd170940d9ae7b355d365d59 MesaLib-7.10.3.zip
8768fd562ede7ed763d92b2d22232d7a MesaGLUT-7.10.3.tar.gz
1496415b89da9549f0f3b34d9622e2e2 MesaGLUT-7.10.3.tar.bz2
1f29d0e7398fd3bf9f36f5db02941198 MesaGLUT-7.10.3.zip
</pre>
<h2>New features</h2>
<p>None.</p>
<h2>Bug fixes</h2>
<p>This list is likely incomplete.</p>
<ul>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29162">Bug 29162</a> - mesa/darwin is severly broken</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31590">Bug 31590</a> - Black space between colors on mole hill example</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32395">Bug 32395</a> - [glsl] Incorrect code generation for shadow2DProj() with bias</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32564">Bug 32564</a> - [llvmpipe] prog: Unknown command line argument '-disable-mmx'. Try: 'prog -help' with llvm-2.9svn</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32835">Bug 32835</a> - [glsl] recursive #define results in infinite stack recursion</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33303">Bug 33303</a> - [glsl] ir_constant_expression.cpp:72: virtual ir_constant* ir_expression::constant_expression_value(): Assertion `op[0]->type->base_type == op[1]->type->base_type' failed.</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33314">Bug 33314</a> - [glsl] ir_constant_expression.cpp:122: virtual ir_constant* ir_expression::constant_expression_value(): Assertion `op[0]->type->base_type == GLSL_TYPE_BOOL' failed.</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33512">Bug 33512</a> - [SNB] case ogles2conform/GL/gl_FragCoord/gl_FragCoord_xy_frag.test and gl_FragCoord_w_frag.test fail</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=34280">Bug 34280</a> - r200 mesa-7.10 font distortion</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=34321">Bug 34321</a> - The ARB_fragment_program subset of ARB_draw_buffers not implemented</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=35603">Bug 35603</a> - GLSL compiler freezes compiling shaders</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=36173">Bug 36173</a> - struct renderbuffer's 'format' field never set when using FBO</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=36238">Bug 36238</a> - Mesa release files don't contain scons control files</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=36410">Bug 36410</a> - [SNB] Rendering errors in 3DMMES subtest taiji</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=36527">Bug 36527</a> - [wine] Wolfenstein: Failed to translate rgb instruction.</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=36651">Bug 36651</a> - mesa requires bison and flex to build but configure does not check for them</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=36738">Bug 36738</a> - Openarena crash with r300g, swrastg + llvm &gt; 2.8</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=37648">Bug 37648</a> - Logic error in mesa/main/teximage.c:texsubimage</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=37739">Bug 37739</a> - Color clear of FBO without color buffer crashes</li>
<!-- <li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=">Bug </a> - </li> -->
</ul>
<h2>Changes</h2>
<p>The full set of changes can be viewed by using the following GIT command:</p>
<pre>
git log mesa-7.10.2..mesa-7.10.3
</pre>
<p>Alan Hourihane (1):
<ul>
<li>Check for out of memory when creating fence</li>
</ul></p>
<p>Alex Buell (1):
<ul>
<li>configure: bump LIBDRM_REQUIRED to 2.4.24</li>
</ul></p>
<p>Alex Deucher (2):
<ul>
<li>r600c: add new pci ids</li>
<li>r600g: add new pci ids</li>
</ul></p>
<p>Brian Paul (19):
<ul>
<li>docs: add link to 7.10.2 release notes</li>
<li>scons: remove dangling reference to state_trackers/python/SConscript</li>
<li>Makefile: add missing Scons files</li>
<li>llvmpipe: document issue with LLVM 2.8 and earlier with AVX</li>
<li>docs: replace llvmpipe/README with docs/llvmpipe.html</li>
<li>glsl: add static qualifier to silence warning</li>
<li>glsl: add cast to silence signed/unsigned comparison warning</li>
<li>mesa: s/height/depth/ in texsubimage()</li>
<li>mesa: fix void pointer arithmetic warnings</li>
<li>mesa: add some missing GLAPIENTRY keywords</li>
<li>mesa: check that flex/bison are installed</li>
<li>st/mesa: fix incorrect texture level/face/slice accesses</li>
<li>draw: fix edge flag handling in clipper (for unfilled tris/quads/polygons)</li>
<li>vbo: check array indexes to prevent negative indexing</li>
<li>vbo: remove node-&gt;count &gt; 0 test in vbo_save_playback_vertex_list()</li>
<li>st/mesa: fix software accum buffer format bug</li>
<li>mesa: add include/c99/inttypes.h include/c99/stdbool.h include/c99/stdint.h files to tarballs</li>
<li>docs: 7.10.3 release notes skeleton file, links</li>
<li>mesa: bump version to 7.10.3</li>
</ul></p>
<p>Carl Worth (2):
<ul>
<li>glcpp: Simplify calling convention of parser's active_list functions</li>
<li>glcpp: Fix attempts to expand recursive macros infinitely (bug #32835).</li>
</ul></p>
<p>Dave Airlie (1):
<ul>
<li>st/mesa: fix compressed mipmap generation.</li>
</ul></p>
<p>Eric Anholt (19):
<ul>
<li>i965: Fix the VS thread limits for GT1, and clarify the WM limits on both.</li>
<li>glsl: Avoid cascading errors when looking for a scalar boolean and failing.</li>
<li>glsl: Semantically check the RHS of `&amp;&amp;' even when short-circuiting.</li>
<li>glsl: Semantically check the RHS of `||' even when short-circuiting.</li>
<li>glsl: When we've emitted a semantic error for ==, return a bool constant.</li>
<li>glsl: Perform type checking on "^^" operands.</li>
<li>intel: Use _mesa_base_tex_format for FBO texture attachments.</li>
<li>swrast: Don't assert against glReadPixels of GL_RED and GL_RG.</li>
<li>mesa: Add a gl_renderbuffer.RowStride field like textures have.</li>
<li>mesa: Add a function to set up the default renderbuffer accessors.</li>
<li>intel: Use Mesa core's renderbuffer accessors for depth.</li>
<li>mesa: Use _mesa_get_format_bytes to refactor out the RB get_pointer_*</li>
<li>mesa: Use _mesa_get_format_bytes to refactor out the RB get_row_*</li>
<li>mesa: Add renderbuffer accessors for R8/RG88/R16/RG1616.</li>
<li>swrast: Don't try to adjust_colors for &lt8bpc when handling R16, RG1616.</li>
<li>intel: Use mesa core's R8, RG88, R16, RG1616 RB accessors.</li>
<li>Revert "intel: Add spans code for the ARB_texture_rg support."</li>
<li>mesa: Add support for the ARB_fragment_program part of ARB_draw_buffers.</li>
<li>mesa: Add support for OPTION ATI_draw_buffers to ARB_fp.</li>
</ul></p>
<p>Hans de Goede (1):
<ul>
<li>texstore: fix regression stricter check for memcpy path for unorm88 and unorm1616</li>
</ul></p>
<p>Henri Verbeet (3):
<ul>
<li>mesa: Also update the color draw buffer if it's explicitly set to GL_NONE.</li>
<li>glx: Destroy dri2Hash on DRI2 display destruction.</li>
<li>glx: Only remove the glx_display from the list after it's destroyed.</li>
</ul></p>
<p>Ian Romanick (9):
<ul>
<li>docs: Add 7.10.2 md5sums</li>
<li>glsl: Fix off-by-one error setting max_array_access for non-constant indexing</li>
<li>ir_to_mesa: Handle shadow compare w/projection and LOD bias correctly</li>
<li>intel: Fix ROUND_DOWN_TO macro</li>
<li>glsl: Regenerate compiler and glcpp files from cherry picks</li>
<li>i965: Remove hint_gs_always and resulting dead code</li>
<li>mesa: Don't try to clear a NULL renderbuffer</li>
<li>mesa: Ignore blits to/from missing buffers</li>
<li>docs: Add list of bugs fixed in 7.10.3 release</li>
</ul></p>
<p>Jeremy Huddleston (18):
<ul>
<li>apple: Update GL specs</li>
<li>apple: Rename glcontextmodes.[ch] to glxconfig.[ch]</li>
<li>apple: Rename __GLcontextModes to struct glx_config</li>
<li>apple: Rename GLXcontext</li>
<li>apple: Re-add driContext and do_destroy</li>
<li>apple: Rename _gl_context_modes_find_visual to glx_config_find_visual</li>
<li>apple: Rename GLXcontext</li>
<li>apple: Change from XExtDisplayInfo to struct glx_display</li>
<li>apple: ifdef out come glapi-foo on darwin</li>
<li>glx: Dead code removal</li>
<li>apple: Build darwin using applegl rather than indirect</li>
<li>apple: Fix build failures in applegl_glx.c</li>
<li>darwin: Define GALLIUM_DRIVERS_DIRS in darwin config</li>
<li>apple: Package applegl source into MesaLib tarball</li>
<li>darwin: Set VG_LIB_{NAME,GLOB} to fix make install</li>
<li>darwin: Don't link against libGL when building libOSMesa</li>
<li>darwin: Fix VG_LIB_GLOB to also match the unversioned symlink</li>
<li>osmesa: Fix missing symbols when GLX_INDIRECT_RENDERING is defined.</li>
</ul></p>
<p>José Fonseca (13):
<ul>
<li>llvmpipe: Update readme.</li>
<li>mesa: GL_PROVOKING_VERTEX_EXT is a GLenum, not GLboolean.</li>
<li>mesa: Fix GetVertexAttrib* inside display lists.</li>
<li>draw: Fix draw_variant_output::format's type.</li>
<li>gallivm: Tell LLVM to not assume a 16-byte aligned stack on x86.</li>
<li>gallivm: Fix for dynamically linked LLVM 2.8 library.</li>
<li>st/wgl: Adjust the pbuffer invisible window size.</li>
<li>st/wgl: Fix debug output format specifiers of stw_framebuffer_get_size().</li>
<li>st/wgl: Prevent spurious framebuffer sizes when the window is minimized.</li>
<li>st/wgl: Cope with zero width/height windows.</li>
<li>st/wgl: Allow to create pbuffers bigger than the desktop.</li>
<li>st/wgl: Remove buggy assertion.</li>
<li>wgl: Don't hold on to user supplied HDC.</li>
</ul></p>
<p>Kenneth Graunke (10):
<ul>
<li>i965/fs: Switch W and 1/W in Sandybridge interpolation setup.</li>
<li>i965: Refactor Sandybridge implied move handling.</li>
<li>i965: Resolve implied moves in brw_dp_READ_4_vs_relative.</li>
<li>intel: Add IS_GT2 macro for recognizing Sandybridge GT2 systems.</li>
<li>i965: Allocate the whole URB to the VS and fix calculations for Gen6.</li>
<li>intel: Support glCopyTexImage() from ARGB8888 to XRGB8888.</li>
<li>glsl: Fix memory error when creating the supported version string.</li>
<li>glsl: Regenerate autogenerated file builtin_function.cpp.</li>
<li>i965: Rename various gen6 #defines to match the documentation.</li>
<li>i965: Never enable the GS on Gen6.</li>
</ul></p>
<p>Kostas Georgiou (1):
<ul>
<li>r600c/g: Add pci id for FirePro 2270</li>
</ul></p>
<p>Marek Olšák (18):
<ul>
<li>tgsi/ureg: bump the limit of immediates</li>
<li>st/mesa: fix changing internal format via RenderbufferStorage</li>
<li>st/mesa: GenerateMipmap should not be killed by conditional rendering</li>
<li>swrast: BlitFramebuffer should not be killed by conditional rendering</li>
<li>st/mesa: BlitFramebuffer should not be killed by conditional rendering</li>
<li>st/mesa: CopyTex(Sub)Image should not be killed by conditional rendering</li>
<li>st/mesa: conditional rendering should not kill texture decompression via blit</li>
<li>mesa: forbid UseProgram to be called inside Begin/End</li>
<li>mesa: UseShaderProgramEXT and Uniform* shouldn't be allowed inside Begin/End</li>
<li>mesa: queries of non-existent FBO attachments should return INVALID_OPERATION</li>
<li>r300g: fix draw_vbo splitting on r3xx-r4xx</li>
<li>r300g: fix texturing with non-3D textures and wrap R mode set to sample border</li>
<li>r300g: fix occlusion queries when depth test is disabled or zbuffer is missing</li>
<li>r300g: clear can be killed by render condition</li>
<li>st/mesa: remove asserts in st_texture_image_copy</li>
<li>mesa: fix up assertion in _mesa_source_buffer_exists</li>
<li>mesa: invalidate framebuffer if internal format of renderbuffer is changed</li>
<li>mesa: return after invalidating renderbuffer</li>
</ul></p>
<p>Matt Turner (1):
<ul>
<li>r300/compiler: align memory allocations to 8-bytes</li>
</ul></p>
<p>Tom Stellard (3):
<ul>
<li>r300/compiler: Fix incorrect presubtract conversion</li>
<li>r300/compiler: Fix dataflow analysis bug with ELSE blocks</li>
<li>r300/compiler: Limit instructions to 3 source selects</li>
</ul></p>
<p>Vinson Lee (1):
<ul>
<li>gallivm: Disable MMX-disabling code on llvm-2.9.</li>
</ul></p>
<p>Zou Nan hai (1):
<ul>
<li>i965: Align interleaved URB write length to 2</li>
</ul></p>
<p>pepp (1):
<ul>
<li>st/mesa: assign renderbuffer's format field when allocating storage</li>
</ul></p>
</body>
</html>

File diff suppressed because it is too large Load Diff

406
docs/relnotes-7.9.1.html Normal file
View File

@@ -0,0 +1,406 @@
<HTML>
<head>
<TITLE>Mesa Release Notes</TITLE>
<link rel="stylesheet" type="text/css" href="mesa.css">
<meta http-equiv="content-type" content="text/html; charset=utf-8" />
</head>
<BODY>
<body bgcolor="#eeeeee">
<H1>Mesa 7.9.1 Release Notes / January 7, 2011</H1>
<p>
Mesa 7.9.1 is a bug fix release which fixes bugs found since the 7.9 release.
</p>
<p>
Mesa 7.9.1 implements the OpenGL 2.1 API, but the version reported by
glGetString(GL_VERSION) depends on the particular driver being used.
Some drivers don't support all the features required in OpenGL 2.1.
</p>
<p>
See the <a href="install.html">Compiling/Installing page</a> for prerequisites
for DRI hardware acceleration.
</p>
<h2>MD5 checksums</h2>
<pre>
78422843ea875ad4eac35b9b8584032b MesaLib-7.9.1.tar.gz
07dc6cfb5928840b8b9df5bd1b3ae434 MesaLib-7.9.1.tar.bz2
c8eaea5b3c3d6dee784bd8c2db91c80f MesaLib-7.9.1.zip
ee9ecae4ca56fbb2d14dc15e3a0a7640 MesaGLUT-7.9.1.tar.gz
41fc477d524e7dc5c84da8ef22422bea MesaGLUT-7.9.1.tar.bz2
90b287229afdf19317aa989d19462e7a MesaGLUT-7.9.1.zip
</pre>
<h2>New features</h2>
<p>None.</p>
<h2>Bug fixes</h2>
<p>This list is likely incomplete.</p>
<ul>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=28800">Bug 28800</a> - [r300c, r300g] Texture corruption with World of Warcraft</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29420">Bug 29420</a> - Amnesia / HPL2 RendererFeatTest - not rendering correctly</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29946">Bug 29946</a> - [swrast] piglit valgrind glsl-array-bounds-04 fails</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30261">Bug 30261</a> - [GLSL 1.20] allowing inconsistent invariant declaration between two vertex shaders</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30632">Bug 30632</a> - [softpipe] state_tracker/st_manager.c:489: st_context_notify_invalid_framebuffer: Assertion `stfb &amp;&amp; stfb-&gt;iface == stfbi' failed.</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30694">Bug 30694</a> - wincopy will crash on Gallium drivers when going to front buffer</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30787">Bug 30787</a> - Invalid asm shader does not generate draw-time error when used with GLSL shader</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30993">Bug 30993</a> - getFramebufferAttachmentParameteriv wrongly generates error</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31101">Bug 31101</a> - [glsl2] abort() in ir_validate::visit_enter(ir_assignment *ir)</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31193">Bug 31193</a> - [regression] aa43176e break water reflections</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31194">Bug 31194</a> - The mesa meta save/restore code doesn't ref the current GLSL program</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31371">Bug 31371</a> - glslparsertest: ir.cpp:358: ir_constant::ir_constant(const glsl_type*, const ir_constant_data*): Assertion `(type->base_type &gt;= 0) &amp;&amp; (type->base_type &lt;= 3)' failed.</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31439">Bug 31439</a> - Crash in glBufferSubData() with size == 0</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31495">Bug 31495</a> - [i965 gles2c bisected] OpenGL ES 2.0 conformance GL2Tests_GetBIFD_input.run regressed</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31514">Bug 31514</a> - isBuffer returns true for unbound buffers</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31560">Bug 31560</a> - [tdfx] tdfx_tex.c:702: error: const struct gl_color_table has no member named Format</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31617">Bug 31617</a> - Radeon/Compiz: 'failed to attach dri2 front buffer', error case not handled</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31648">Bug 31648</a> - [GLSL] array-struct-array gets assertion: `(size &gt;= 1) && (size &lt;= 4)' failed.</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31650">Bug 31650</a> - [GLSL] varying gl_TexCoord fails to be re-declared to different size in the second shader</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31673">Bug 31673</a> - GL_FRAGMENT_PRECISION_HIGH preprocessor macro undefined in GLSL ES</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31690">Bug 31690</a> - i915 shader compiler fails to flatten if in Aquarium webgl demo.</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31832">Bug 31832</a> - [i915] Bad renderbuffer format: 21</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31841">Bug 31841</a> - [drm:radeon_cs_ioctl] *ERROR* Invalid command stream !</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31894">Bug 31894</a> - Writing to gl_PointSize with GLES2 corrupts other varyings</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31909">Bug 31909</a> - [i965] brw_fs.cpp:1461: void fs_visitor::emit_bool_to_cond_code(ir_rvalue*): Assertion `expr-&gt;operands[i]-&gt;type-&gt;is_scalar()' failed.</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31934">Bug 31934</a> - [gallium] Mapping empty buffer object causes SIGSEGV</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31983">Bug 31983</a> - [i915 gles2] "if (expression with builtin/varying variables) discard" breaks linkage</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31985">Bug 31985</a> - [GLSL 1.20] initialized uniform array considered as "unsized"</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31987">Bug 31987</a> - [gles2] if input a wrong pname(GL_NONE) to glGetBoolean, it will not case GL_INVALID_ENUM</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32035">Bug 32035</a> - [GLSL bisected] comparing unsized array gets segfault</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32070">Bug 32070</a> - llvmpipe renders stencil demo incorrectly</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32273">Bug 32273</a> - assertion fails when starting vdrift 2010 release with shaders enabled</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32287">Bug 32287</a> - [bisected GLSL] float-int failure</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32311">Bug 32311</a> - [965 bisected] Array look-ups broken on GM45</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32520">Bug 32520</a> - [gles2] glBlendFunc(GL_ZERO, GL_DST_COLOR) will result in GL_INVALID_ENUM</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32825">Bug 32825</a> - egl_glx driver completely broken in 7.9 branch [fix in master]</li>
</ul>
<h2>Changes</h2>
<p>The full set of changes can be viewed by using the following GIT command:</p>
<pre>
git log mesa-7.9..mesa-7.9.1
</pre>
<p>Alex Deucher (5):
<ul>
<li>r100: revalidate after radeon_update_renderbuffers</li>
<li>r600c: add missing radeon_prepare_render() call on evergreen</li>
<li>r600c: properly align mipmaps to group size</li>
<li>gallium/egl: fix r300 vs r600 loading</li>
<li>r600c: fix some opcodes on evergreen</li>
</ul></p>
<p>Aras Pranckevicius (2):
<ul>
<li>glsl: fix crash in loop analysis when some controls can't be determined</li>
<li>glsl: fix matrix type check in ir_algebraic</li>
</ul></p>
<p>Brian Paul (27):
<ul>
<li>swrast: fix choose_depth_texture_level() to respect mipmap filtering state</li>
<li>st/mesa: replace assertion w/ conditional in framebuffer invalidation</li>
<li>egl/i965: include inline_wrapper_sw_helper.h</li>
<li>mesa: Add missing else in do_row_3D</li>
<li>mesa: add missing formats in _mesa_format_to_type_and_comps()</li>
<li>mesa: handle more pixel types in mipmap generation code</li>
<li>mesa: make glIsBuffer() return false for never bound buffers</li>
<li>mesa: fix glDeleteBuffers() regression</li>
<li>swrast: init alpha value to 1.0 in opt_sample_rgb_2d()</li>
<li>meta: Mask Stencil.Clear against stencilMax in _mesa_meta_Clear</li>
<li>st/mesa: fix mapping of zero-sized buffer objects</li>
<li>mesa: check for posix_memalign() errors</li>
<li>llvmpipe: fix broken stencil writemask</li>
<li>mesa: fix GL_FRAMEBUFFER_ATTACHMENT_OBJECT_NAME query</li>
<li>mesa: return GL_FRAMEBUFFER_DEFAULT as FBO attachment type</li>
<li>mesa: make glGet*(GL_NONE) generate GL_INVALID_ENUM</li>
<li>mesa: test for cube map completeness in glGenerateMipmap()</li>
<li>tnl: Initialize gl_program_machine memory in run_vp.</li>
<li>tnl: a better way to initialize the gl_program_machine memory</li>
<li>mesa, st/mesa: disable GL_ARB_geometry_shader4</li>
<li>glsl: fix off by one in register index assertion</li>
<li>st/mesa: fix mipmap generation bug</li>
<li>glsl: new glsl_strtod() wrapper to fix decimal point interpretation</li>
<li>mesa: no-op glBufferSubData() on size==0</li>
<li>tdfx: s/Format/_BaseFormat/</li>
<li>st/mesa: fix renderbuffer pointer check in st_Clear()</li>
<li>mesa: Bump the number of bits in the register index.</li>
</ul></p>
<p>Chad Versace (5):
<ul>
<li>glsl: Fix lexer rule for ^=</li>
<li>glsl: Fix ast-to-hir for ARB_fragment_coord_conventions</li>
<li>glsl: Fix ir_expression::constant_expression_value()</li>
<li>glsl: Fix erroneous cast in ast_jump_statement::hir()</li>
<li>glsl: Fix linker bug in cross_validate_globals()</li>
</ul></p>
<p>Chia-I Wu (10):
<ul>
<li>targets/egl: Fix linking with libdrm.</li>
<li>st/vega: Fix version check in context creation.</li>
<li>st/egl: Do not finish a fence that is NULL.</li>
<li>egl: Fix a false negative check in _eglCheckMakeCurrent.</li>
<li>st/mesa: Unreference the sampler view in st_bind_surface.</li>
<li>egl_dri2: Fix __DRI_DRI2 version 1 support.</li>
<li>st/vega: Do not wait NULL fences.</li>
<li>mesa: Do not advertise GL_OES_texture_3D.</li>
<li>egl_glx: Fix borken driver.</li>
<li>egl: Check extensions.</li>
</ul></p>
<p>Daniel Lichtenberger (1):
<ul>
<li>radeon: fix potential segfault in renderbuffer update</li>
</ul></p>
<p>Daniel Vetter (1):
<ul>
<li>r200: revalidate after radeon_update_renderbuffers</li>
</ul></p>
<p>Dave Airlie (1):
<ul>
<li>r300g: fixup rs690 tiling stride alignment calculations.</li>
</ul></p>
<p>Eric Anholt (13):
<ul>
<li>intel: Allow CopyTexSubImage to InternalFormat 3/4 textures, like RGB/RGBA.</li>
<li>glsl: Free the loop state context when we free the loop state.</li>
<li>i965: Allow OPCODE_SWZ to put immediates in the first arg.</li>
<li>i965: Add support for rendering to SARGB8 FBOs.</li>
<li>glsl: Add a helper constructor for expressions that works out result type.</li>
<li>glsl: Fix structure and array comparisions.</li>
<li>glsl: Quiet unreachable no-return-from-function warning.</li>
<li>glsl: Mark the array access for whole-array comparisons.</li>
<li>glsl: Fix flipped return of has_value() for array constants.</li>
<li>mesa: Add getters for the rest of the supported draw buffers.</li>
<li>mesa: Add getters for ARB_copy_buffer's attachment points.</li>
<li>i965: Correct the dp_read message descriptor setup on g4x.</li>
<li>glsl: Correct the marking of InputsRead/OutputsWritten on in/out matrices.</li>
</ul></p>
<p>Fabian Bieler (1):
<ul>
<li>glsl: fix lowering conditional returns in subroutines</li>
</ul></p>
<p>Francisco Jerez (3):
<ul>
<li>meta: Don't leak alpha function/reference value changes.</li>
<li>meta: Fix incorrect rendering of the bitmap alpha component.</li>
<li>meta: Don't try to disable cube maps if the driver doesn't expose the extension.</li>
</ul></p>
<p>Henri Verbeet (2):
<ul>
<li>r600: Evergreen has two extra frac_bits for the sampler LOD state.</li>
<li>st/mesa: Handle wrapped depth buffers in st_copy_texsubimage().</li>
</ul></p>
<p>Ian Romanick (33):
<ul>
<li>Add 7.9 md5sums</li>
<li>docs: Import 7.8.x release notes from 7.8 branch.</li>
<li>docs: download.html does not need to be updated for each release</li>
<li>docs: Update mailing lines from sf.net to freedesktop.org</li>
<li>docs: added news item for 7.9 release</li>
<li>mesa: Validate assembly shaders when GLSL shaders are used</li>
<li>linker: Reject shaders that have unresolved function calls</li>
<li>mesa: Refactor validation of shader targets</li>
<li>glsl: Slightly change the semantic of _LinkedShaders</li>
<li>linker: Improve handling of unread/unwritten shader inputs/outputs</li>
<li>glsl: Commit lexer files changed by previous cherry picking</li>
<li>mesa: Make metaops use program refcounts instead of names.</li>
<li>glsl: Fix incorrect gl_type of sampler2DArray and sampler1DArrayShadow</li>
<li>mesa: Allow query of MAX_SAMPLES with EXT_framebuffer_multisample</li>
<li>glsl: better handling of linker failures</li>
<li>mesa: Fix glGet of ES2's GL_MAX_*_VECTORS properties.</li>
<li>i915: Disallow alpha, red, RG, and sRGB as render targets</li>
<li>glsl/linker: Free any IR discarded by optimization passes.</li>
<li>glsl: Add an optimization pass to simplify discards.</li>
<li>glsl: Add a lowering pass to move discards out of if-statements.</li>
<li>i915: Correctly generate unconditional KIL instructions</li>
<li>glsl: Add unary ir_expression constructor</li>
<li>glsl: Ensure that equality comparisons don't return a NULL IR tree</li>
<li>glcpp: Commit changes in generated files cause by previous commit</li>
<li>glsl: Inherrit type of declared variable from initializer</li>
<li>glsl: Inherrit type of declared variable from initializer after processing assignment</li>
<li>linker: Ensure that unsized arrays have a size after linking</li>
<li>linker: Fix regressions caused by previous commit</li>
<li>linker: Allow built-in arrays to have different sizes between shader stages</li>
<li>ir_to_mesa: Don't generate swizzles for record derefs of non-scalar/vectors</li>
<li>Refresh autogenerated file builtin_function.cpp.</li>
<li>docs: Initial set of release notes for 7.9.1</li>
<li>mesa: set version string to 7.9.1</li>
</ul></p>
<p>Julien Cristau (1):
<ul>
<li>Makefile: don't include the same files twice in the tarball</li>
</ul></p>
<p>Kenneth Graunke (19):
<ul>
<li>glcpp: Return NEWLINE token for newlines inside multi-line comments.</li>
<li>generate_builtins.py: Output large strings as arrays of characters.</li>
<li>glsl: Fix constant component count in vector constructor emitting.</li>
<li>ir_dead_functions: Actually free dead functions and signatures.</li>
<li>glcpp: Define GL_FRAGMENT_PRECISION_HIGH if GLSL version &gt;= 1.30.</li>
<li>glsl: Unconditionally define GL_FRAGMENT_PRECISION_HIGH in ES2 shaders.</li>
<li>glsl: Fix constant expression handling for &lt, &gt;, &lt=, &gt;= on vectors.</li>
<li>glsl: Use do_common_optimization in the standalone compiler.</li>
<li>glsl: Don't inline function prototypes.</li>
<li>glsl: Add a virtual as_discard() method.</li>
<li>glsl: Remove "discard" support from lower_jumps.</li>
<li>glsl: Refactor get_num_operands.</li>
<li>glcpp: Don't emit SPACE tokens in conditional_tokens production.</li>
<li>glsl: Clean up code by adding a new is_break() function.</li>
<li>glsl: Consider the "else" branch when looking for loop breaks.</li>
<li>Remove OES_compressed_paletted_texture from the ES2 extension list.</li>
<li>glsl/builtins: Compute the correct value for smoothstep(vec, vec, vec).</li>
<li>Fix build on systems where "python" is python 3.</li>
<li>i965: Internally enable GL_NV_blend_square on ES2.</li>
</ul></p>
<p>Kristian Høgsberg (1):
<ul>
<li>i965: Don't write mrf assignment for pointsize output</li>
</ul></p>
<p>Luca Barbieri (1):
<ul>
<li>glsl: Unroll loops with conditional breaks anywhere (not just the end)</li>
</ul></p>
<p>Marek Olšák (17):
<ul>
<li>r300g: fix microtiling for 16-bits-per-channel formats</li>
<li>r300g: fix texture border for 16-bits-per-channel formats</li>
<li>r300g: add a default channel ordering of texture border for unhandled formats</li>
<li>r300g: fix texture border color for all texture formats</li>
<li>r300g: fix rendering with no vertex elements</li>
<li>r300/compiler: fix rc_rewrite_depth_out for it to work with any instruction</li>
<li>r300g: fix texture border color once again</li>
<li>r300g: fix texture swizzling with compressed textures on r400-r500</li>
<li>r300g: disable ARB_texture_swizzle if S3TC is enabled on r3xx-only</li>
<li>mesa, st/mesa: fix gl_FragCoord with FBOs in Gallium</li>
<li>st/mesa: initialize key in st_vp_varient</li>
<li>r300/compiler: fix swizzle lowering with a presubtract source operand</li>
<li>r300g: fix rendering with a vertex attrib having a zero stride</li>
<li>ir_to_mesa: Add support for conditional discards.</li>
<li>r300g: finally fix the texture corruption on r3xx-r4xx</li>
<li>mesa: fix texel store functions for some float formats</li>
<li>r300/compiler: disable the rename_regs pass for loops</li>
</ul></p>
<p>Mario Kleiner (1):
<ul>
<li>mesa/r300classic: Fix dri2Invalidate/radeon_prepare_render for page flipping.</li>
</ul></p>
<p>Peter Clifton (1):
<ul>
<li>intel: Fix emit_linear_blit to use DWORD aligned width blits</li>
</ul></p>
<p>Robert Hooker (2):
<ul>
<li>intel: Add a new B43 pci id.</li>
<li>egl_dri2: Add missing intel chip ids.</li>
</ul></p>
<p>Roland Scheidegger (1):
<ul>
<li>r200: fix r200 large points</li>
</ul></p>
<p>Thomas Hellstrom (17):
<ul>
<li>st/xorg: Don't try to use option values before processing options</li>
<li>xorg/vmwgfx: Make vmwarectrl work also on 64-bit servers</li>
<li>st/xorg: Add a customizer option to get rid of annoying cursor update flicker</li>
<li>xorg/vmwgfx: Don't hide HW cursors when updating them</li>
<li>st/xorg: Don't try to remove invalid fbs</li>
<li>st/xorg: Fix typo</li>
<li>st/xorg, xorg/vmwgfx: Be a bit more frendly towards cross-compiling environments</li>
<li>st/xorg: Fix compilation errors for Xservers compiled without Composite</li>
<li>st/xorg: Don't use deprecated x*alloc / xfree functions</li>
<li>xorg/vmwgfx: Don't use deprecated x*alloc / xfree functions</li>
<li>st/xorg: Fix compilation for Xservers &gt;= 1.10</li>
<li>mesa: Make sure we have the talloc cflags when using the talloc headers</li>
<li>egl: Add an include for size_t</li>
<li>mesa: Add talloc includes for gles</li>
<li>st/egl: Fix build for include files in nonstandard places</li>
<li>svga/drm: Optionally resolve calls to powf during link-time</li>
<li>gallium/targets: Trivial crosscompiling fix</li>
</ul></p>
<p>Tom Stellard (7):
<ul>
<li>r300/compiler: Make sure presubtract sources use supported swizzles</li>
<li>r300/compiler: Fix register allocator's handling of loops</li>
<li>r300/compiler: Fix instruction scheduling within IF blocks</li>
<li>r300/compiler: Use zero as the register index for unused sources</li>
<li>r300/compiler: Ignore alpha dest register when replicating the result</li>
<li>r300/compiler: Use correct swizzles for all presubtract sources</li>
<li>r300/compiler: Don't allow presubtract sources to be remapped twice</li>
</ul></p>
<p>Vinson Lee (1):
<ul>
<li>glsl: Fix 'control reaches end of non-void function' warning.</li>
</ul></p>
<p>richard (1):
<ul>
<li>r600c : inline vertex format is not updated in an app, switch to use vfetch constants. For the 7.9 and 7.10 branches as well.</li>
</ul></p>
</body>
</html>

336
docs/relnotes-7.9.2.html Normal file
View File

@@ -0,0 +1,336 @@
<HTML>
<head>
<TITLE>Mesa Release Notes</TITLE>
<link rel="stylesheet" type="text/css" href="mesa.css">
<meta http-equiv="content-type" content="text/html; charset=utf-8" />
</head>
<BODY>
<body bgcolor="#eeeeee">
<H1>Mesa 7.9.2 Release Notes / March 2, 2011</H1>
<p>
Mesa 7.9.2 is a bug fix release which fixes bugs found since the 7.9.1 release.
</p>
<p>
Mesa 7.9.2 implements the OpenGL 2.1 API, but the version reported by
glGetString(GL_VERSION) depends on the particular driver being used.
Some drivers don't support all the features required in OpenGL 2.1.
</p>
<p>
See the <a href="install.html">Compiling/Installing page</a> for prerequisites
for DRI hardware acceleration.
</p>
<h2>MD5 checksums</h2>
<pre>
eb4ab8c1a03386def3ea34b1358e9cda MesaLib-7.9.2.tar.gz
8f6d1474912787ce13bd35f3bae9938a MesaLib-7.9.2.tar.bz2
427a81dd43ac97603768dc5c6af3df26 MesaLib-7.9.2.zip
aacb8f4db997e346db40c6066942140a MesaGLUT-7.9.2.tar.gz
18abe6cff4fad8ad4752c7b7ab548e5d MesaGLUT-7.9.2.tar.bz2
3189e5732d636c71baf3d8bc23ce7b11 MesaGLUT-7.9.2.zip
</pre>
<h2>New features</h2>
<p>None.</p>
<h2>Bug fixes</h2>
<p>This list is likely incomplete.</p>
<ul>
<li>Fix an off-by-one bug in a vsplit assertion.</li>
<li>Fix incorrect handling of <tt>layout</tt> qualifier
with <tt>in</tt>, <tt>out</tt>, <tt>attribute</tt>, and <tt>varying</tt>.</li>
<li>Fix an i965 GPU hang in GLSL shaders that contain an unconditional <tt>discard</tt> statement.</li>
<li>Fix an i965 shader bug where the negative absolute value was generated instead of the absolute value of a negation.</li>
<li>Fix numerous issues handling precision qualifiers in GLSL ES.</li>
<li>Fixed a few GLX protocol encoder bugs (Julien Cristau)</li>
<li>Assorted Gallium llvmpipe driver bug fixes</li>
<li>Assorted Mesa/Gallium state tracker bug fixes</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=26795">Bug 26795</a> - gl_FragCoord off by one in Gallium drivers.</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29164">Bug 29164</a> - [GLSL 1.20] invariant variable shouldn't be used before declaration</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29823">Bug 29823</a> - GetUniform[if]v busted</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29927">Bug 29927</a> - [glsl2] fail to compile shader with constructor for array of struct type</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=30156">Bug 30156</a> - [i965] After updating to Mesa 7.9, Civilization IV starts to show garbage</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31923">Bug 31923</a> - [GLSL 1.20] allowing inconsistent centroid declaration between two vertex shaders</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=31925">Bug 31925</a> - [GLSL 1.20] "#pragma STDGL invariant(all)" fail</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32214">Bug 32214</a> - [gles2]no link error happens when missing vertex shader or frag shader</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32375">Bug 32375</a> - [gl gles2] Not able to get the attribute by function glGetVertexAttribfv</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32541">Bug 32541</a> - Segmentation Fault while running an HDR (high dynamic range) rendering demo</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32569">Bug 32569</a> - [gles2] glGetShaderPrecisionFormat not implemented yet</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32695">Bug 32695</a> - [glsl] SIGSEGV glcpp/glcpp-parse.y:833</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32831">Bug 32831</a> - [glsl] division by zero crashes GLSL compiler</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=32910">Bug 32910</a> - Keywords 'in' and 'out' not handled properly for GLSL 1.20 shaders</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33219">Bug 33219</a> -[GLSL bisected] implicit sized array triggers segfault in ir_to_mesa_visitor::copy_propagate</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33306">Bug 33306</a> - GLSL integer division by zero crashes GLSL compiler</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33308">Bug 33308</a> -[glsl] ast_to_hir.cpp:3016: virtual ir_rvalue* ast_jump_statement::hir(exec_list*, _mesa_glsl_parse_state*): Assertion `ret != __null' failed.</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33316">Bug 33316</a> - uniform array will be allocate one line more and initialize it when it was freed will abort</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33386">Bug 33386</a> - Dubious assembler in read_rgba_span_x86.S</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33388">Bug 33388</a> - Dubious assembler in xform4.S</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33433">Bug 33433</a> - Error in x86-64 API dispatch code.</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33507">Bug 33507</a> - [glsl] GLSL preprocessor modulus by zero crash</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33508">Bug 33508</a> - [glsl] GLSL compiler modulus by zero crash</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=33916">Bug 33916</a> - Compiler accepts reserved operators % and %=</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=34047">Bug 34047</a> - Assert in _tnl_import_array() when using GLfixed vertex datatypes with GLESv2</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=34114">Bug 34114</a> - Sun Studio build fails due to standard library functions not being in global namespace</li>
<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=34198">Bug 34198</a> - [GLSL] implicit sized array with index 0 used gets assertion</li>
<li><a href="https://bugs.launchpad.net/ubuntu/+source/mesa/+bug/691653">Ubuntu bug 691653</a> - compiz crashes when using alt-tab (the radeon driver kills it) </li>
<li><a href="https://bugs.meego.com/show_bug.cgi?id=13005">Meego bug 13005</a> - Graphics GLSL issue lead to camera preview fail on Pinetrail</li>
<!-- <li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=">Bug </a> - </li> -->
</ul>
<h2>Changes</h2>
<p>The full set of changes can be viewed by using the following GIT command:</p>
<pre>
git log mesa-7.9.1..mesa-7.9.2
</pre>
<p>Alberto Milone (1):
<ul>
<li>r600c: add evergreen ARL support.</li>
</ul></p>
<p>Brian Paul (19):
<ul>
<li>draw: Fix an off-by-one bug in a vsplit assertion.</li>
<li>mesa: fix a few format table mistakes, assertions</li>
<li>mesa: fix num_draw_buffers==0 in fixed-function fragment program generation</li>
<li>mesa: don't assert in GetIntegerIndexed, etc</li>
<li>mesa: check for dummy renderbuffer in _mesa_FramebufferRenderbufferEXT()</li>
<li>llvmpipe: make sure binning is active when we begin/end a query</li>
<li>st/mesa: fix incorrect fragcoord.x translation</li>
<li>softpipe: fix off-by-one error in setup_fragcoord_coeff()</li>
<li>cso: fix loop bound in cso_set_vertex_samplers()</li>
<li>st/mesa: set renderbuffer _BaseFormat in a few places</li>
<li>st/mesa: fix the default case in st_format_datatype()</li>
<li>st/mesa: need to translate clear color according to surface's base format</li>
<li>docs: update 7.9.2 release notes with Brian's cherry-picks</li>
<li>docs: add links to 7.9.1 and 7.9.2 release notes</li>
<li>mesa: include compiler.h for ASSERT macro</li>
<li>glsl: add ir_shader case in switch stmt to silence warning</li>
<li>glsl2: fix signed/unsigned comparison warning</li>
<li>mesa: implement glGetShaderPrecisionFormat()</li>
<li>docs: updated environment variable list</li>
</ul></p>
<p>Bryce Harrington (1):
<ul>
<li>r300g: Null pointer check for buffer deref in gallium winsys</li>
</ul></p>
<p>Chad Versace (14):
<ul>
<li>glsl: At link-time, check that globals have matching centroid qualifiers</li>
<li>glcpp: Fix segfault when validating macro redefinitions</li>
<li>glsl: Fix parser rule for type_specifier</li>
<li>glsl: Change default value of ast_type_specifier::precision</li>
<li>glsl: Add semantic checks for precision qualifiers</li>
<li>glsl: Add support for default precision statements</li>
<li>glsl: Remove redundant semantic check in parser</li>
<li>glsl: Fix semantic checks on precision qualifiers</li>
<li>glsl: Fix segfault due to missing printf argument</li>
<li>glsl: Mark 'in' variables at global scope as read-only</li>
<li>glcpp: Raise error when modulus is zero</li>
<li>glsl: Set operators '%' and '%=' to be reserved when GLSL &lt 1.30</li>
<li>glsl: Reinstate constant-folding for division by zero</li>
<li>tnl: Add support for datatype GL_FIXED in vertex arrays</li>
</ul></p>
<p>Chia-I Wu (1):
<ul>
<li>mesa: Add glDepthRangef and glClearDepthf to APIspec.xml.</li>
</ul></p>
<p>Chris Wilson (1):
<ul>
<li>intel: Check for unsupported texture when finishing using as a render target</li>
</ul></p>
<p>Cyril Brulebois (1):
<ul>
<li>Point to bugs.freedesktop.org rather than bugzilla.freedesktop.org</li>
</ul></p>
<p>Dave Airlie (2):
<ul>
<li>radeon/r200: fix fbo-clearmipmap + gen-teximage</li>
<li>radeon: avoid segfault on 3D textures.</li>
</ul></p>
<p>Dimitry Andric (4):
<ul>
<li>mesa: s/movzx/movzbl/</li>
<li>mesa: s/movzxw/movzwl/ in read_rgba_span_x86.S</li>
<li>glapi: adding @ char before type specifier in glapi_x86.S</li>
<li>glapi: add @GOTPCREL relocation type</li>
</ul></p>
<p>Eric Anholt (11):
<ul>
<li>i965: Avoid double-negation of immediate values in the VS.</li>
<li>docs: Add a relnote for the Civ IV on i965.</li>
<li>i965/vs: When MOVing to produce ABS, strip negate of the operand.</li>
<li>glsl: Fix the lowering of variable array indexing to not lose write_masks.</li>
<li>intel: Make renderbuffer tiling choice match texture tiling choice.</li>
<li>glapi: Add entrypoints and enums for GL_ARB_ES2_compatibility.</li>
<li>mesa: Add extension enable bit for GL_ARB_ES2_compatibility.</li>
<li>mesa: Add actual support for glReleaseShaderCompiler from ES2.</li>
<li>mesa: Add support for glDepthRangef and glClearDepthf.</li>
<li>mesa: Add getters for ARB_ES2_compatibility MAX_*_VECTORS.</li>
<li>mesa: Add getter for GL_SHADER_COMPILER with ARB_ES2_compatibility.</li>
</ul></p>
<p>Ian Romanick (42):
<ul>
<li>docs: Add 7.9.1 md5sums</li>
<li>glsl: Support the 'invariant(all)' pragma</li>
<li>glcpp: Generate an error for division by zero</li>
<li>glsl: Add version_string containing properly formatted GLSL version</li>
<li>glsl &amp; glcpp: Refresh autogenerated lexer and parser files.</li>
<li>glsl: Disallow 'in' and 'out' on globals in GLSL 1.20</li>
<li>glsl: Track variable usage, use that to enforce semantics</li>
<li>glsl: Allow 'in' and 'out' when 'layout' is also available</li>
<li>docs: Initial set of release notes for 7.9.2</li>
<li>mesa: bump version to 7.9.2-devel</li>
<li>docs: Update 7.9.2 release notes</li>
<li>i965: Make OPCODE_KIL_NV do its work in a temp, not the null reg!</li>
<li>glsl: Refresh autogenerated lexer and parser files.</li>
<li>glsl: Don't assert when the value returned by a function has no rvalue</li>
<li>linker: Set sizes for non-global arrays as well</li>
<li>linker: Propagate max_array_access while linking functions</li>
<li>docs: Update 7.9.2 release notes</li>
<li>Use C-style system headers in C++ code to avoid issues with std:: namespace</li>
<li>mesa: glGetUniform only returns a single element of an array</li>
<li>linker: Generate link errors when ES shaders are missing stages</li>
<li>mesa: Fix error checks in GetVertexAttrib functions</li>
<li>docs: Update 7.9.2 release notes</li>
<li>mesa: Remove unsupported OES extensions</li>
<li>glapi: Regenerate for GL_ARB_ES2_compatibility.</li>
<li>mesa: Connect glGetShaderPrecisionFormat into the dispatch table</li>
<li>i965: Set correct values for range/precision of fragment shader types</li>
<li>i915: Set correct values for range/precision of fragment shader types</li>
<li>intel: Fix typeos from 3d028024 and 790ff232</li>
<li>glsl: Ensure that all GLSL versions are supported in the stand-alone compiler</li>
<li>glsl: Reject shader versions not supported by the implementation</li>
<li>mesa: Initial size for secondary color array is 3</li>
<li>glcpp: Regenerate files from recent cherry picks</li>
<li>glsl: Finish out the reduce/reduce error fixes</li>
<li>glsl: Regenerate compiler files from cherry picks</li>
<li>linker: Fix off-by-one error implicit array sizing</li>
<li>i915: Only mark a register as available if all components are written</li>
<li>i915: Calculate partial result to temp register first</li>
<li>i915: Force lowering of all types of indirect array accesses in the FS</li>
<li>docs: Update 7.9.2 release notes for recent cherry picks</li>
<li>docs: Clean up bug fixes list</li>
<li>intel: Remove driver date and related bits from renderer string</li>
<li>mesa: set version string to 7.9.2 (final)</li>
</ul></p>
<p>Jian Zhao (1):
<ul>
<li>mesa: fix an error in uniform arrays in row calculating.</li>
</ul></p>
<p>Julien Cristau (3):
<ul>
<li>glx: fix request lengths</li>
<li>glx: fix GLXChangeDrawableAttributesSGIX request</li>
<li>glx: fix length of GLXGetFBConfigsSGIX</li>
</ul></p>
<p>Keith Packard (1):
<ul>
<li>glsl: Eliminate reduce/reduce conflicts in glsl grammar</li>
</ul></p>
<p>Kenneth Graunke (12):
<ul>
<li>glsl: Expose a public glsl_type::void_type const pointer.</li>
<li>glsl: Don't bother unsetting a destructor that was never set.</li>
<li>glsl, i965: Remove unnecessary talloc includes.</li>
<li>glcpp: Remove use of talloc reference counting.</li>
<li>ralloc: Add a fake implementation of ralloc based on talloc.</li>
<li>Convert everything from the talloc API to the ralloc API.</li>
<li>ralloc: a new MIT-licensed recursive memory allocator.</li>
<li>Remove talloc from the make and automake build systems.</li>
<li>Remove talloc from the SCons build system.</li>
<li>Remove the talloc sources from the Mesa repository.</li>
<li>glsl: Fix use of uninitialized values in _mesa_glsl_parse_state ctor.</li>
<li>glsl: Use reralloc instead of plain realloc.</li>
</ul></p>
<p>Marek Olšák (3):
<ul>
<li>docs: fix messed up names with special characters in relnotes-7.9.1</li>
<li>mesa: fix texture3D mipmap generation for UNSIGNED_BYTE_3_3_2</li>
<li>st/dri: Track drawable context bindings</li>
</ul></p>
<p>Paulo Zanoni (1):
<ul>
<li>dri_util: fail driCreateNewScreen if InitScreen is NULL</li>
</ul></p>
<p>Sam Hocevar (2):
<ul>
<li>docs: add glsl info</li>
<li>docs: fix glsl_compiler name</li>
</ul></p>
<p>Vinson Lee (1):
<ul>
<li>ralloc: Add missing va_end following va_copy.</li>
</ul></p>
<p>nobled (1):
<ul>
<li>glx: Put null check before use</li>
</ul></p>
</p>
</body>
</html>

View File

@@ -13,6 +13,12 @@ The release notes summarize what's new or changed in each Mesa release.
</p>
<UL>
<LI><A HREF="relnotes-7.10.3.html">7.10.3 release notes</A>
<LI><A HREF="relnotes-7.10.2.html">7.10.2 release notes</A>
<LI><A HREF="relnotes-7.10.1.html">7.10.1 release notes</A>
<LI><A HREF="relnotes-7.10.html">7.10 release notes</A>
<LI><A HREF="relnotes-7.9.2.html">7.9.2 release notes</A>
<LI><A HREF="relnotes-7.9.1.html">7.9.1 release notes</A>
<LI><A HREF="relnotes-7.9.html">7.9 release notes</A>
<LI><A HREF="relnotes-7.8.3.html">7.8.3 release notes</A>
<LI><A HREF="relnotes-7.8.2.html">7.8.2 release notes</A>

View File

@@ -167,7 +167,7 @@ Here's an example of using the compiler to compile a vertex shader and
emit GL_ARB_vertex_program-style instructions:
</p>
<pre>
src/glsl/glslcompiler --dump-ast myshader.vert
src/glsl/glsl_compiler --dump-ast myshader.vert
</pre>
Options include

View File

@@ -23,6 +23,7 @@ each directory.
<ul>
<li><b>docs</b> - EGL documentation
<li><b>drivers</b> - EGL drivers
<li><b>glsl</b> - the GLSL compiler
<li><b>main</b> - main EGL library implementation. This is where all
the EGL API functions are implemented, like eglCreateContext().
</ul>

View File

@@ -31,6 +31,7 @@
#define glAccum MANGLE(Accum)
#define glActiveProgramEXT MANGLE(ActiveProgramEXT)
#define glActiveShaderProgram MANGLE(ActiveShaderProgram)
#define glActiveStencilFaceEXT MANGLE(ActiveStencilFaceEXT)
#define glActiveTextureARB MANGLE(ActiveTextureARB)
#define glActiveTexture MANGLE(ActiveTexture)
@@ -56,6 +57,7 @@
#define glBeginOcclusionQueryNV MANGLE(BeginOcclusionQueryNV)
#define glBeginPerfMonitorAMD MANGLE(BeginPerfMonitorAMD)
#define glBeginQueryARB MANGLE(BeginQueryARB)
#define glBeginQueryIndexed MANGLE(BeginQueryIndexed)
#define glBeginQuery MANGLE(BeginQuery)
#define glBeginTransformFeedbackEXT MANGLE(BeginTransformFeedbackEXT)
#define glBeginTransformFeedback MANGLE(BeginTransformFeedback)
@@ -75,22 +77,27 @@
#define glBindBufferRange MANGLE(BindBufferRange)
#define glBindBufferRangeNV MANGLE(BindBufferRangeNV)
#define glBindFragDataLocationEXT MANGLE(BindFragDataLocationEXT)
#define glBindFragDataLocationIndexed MANGLE(BindFragDataLocationIndexed)
#define glBindFragDataLocation MANGLE(BindFragDataLocation)
#define glBindFragmentShaderATI MANGLE(BindFragmentShaderATI)
#define glBindFramebufferEXT MANGLE(BindFramebufferEXT)
#define glBindFramebuffer MANGLE(BindFramebuffer)
#define glBindImageTextureEXT MANGLE(BindImageTextureEXT)
#define glBindLightParameterEXT MANGLE(BindLightParameterEXT)
#define glBindMaterialParameterEXT MANGLE(BindMaterialParameterEXT)
#define glBindMultiTextureEXT MANGLE(BindMultiTextureEXT)
#define glBindParameterEXT MANGLE(BindParameterEXT)
#define glBindProgramARB MANGLE(BindProgramARB)
#define glBindProgramNV MANGLE(BindProgramNV)
#define glBindProgramPipeline MANGLE(BindProgramPipeline)
#define glBindRenderbufferEXT MANGLE(BindRenderbufferEXT)
#define glBindRenderbuffer MANGLE(BindRenderbuffer)
#define glBindSampler MANGLE(BindSampler)
#define glBindTexGenParameterEXT MANGLE(BindTexGenParameterEXT)
#define glBindTextureEXT MANGLE(BindTextureEXT)
#define glBindTexture MANGLE(BindTexture)
#define glBindTextureUnitParameterEXT MANGLE(BindTextureUnitParameterEXT)
#define glBindTransformFeedback MANGLE(BindTransformFeedback)
#define glBindTransformFeedbackNV MANGLE(BindTransformFeedbackNV)
#define glBindVertexArrayAPPLE MANGLE(BindVertexArrayAPPLE)
#define glBindVertexArray MANGLE(BindVertexArray)
@@ -112,18 +119,22 @@
#define glBlendColorEXT MANGLE(BlendColorEXT)
#define glBlendColor MANGLE(BlendColor)
#define glBlendEquationEXT MANGLE(BlendEquationEXT)
#define glBlendEquationiARB MANGLE(BlendEquationiARB)
#define glBlendEquationi MANGLE(BlendEquationi)
#define glBlendEquationIndexedAMD MANGLE(BlendEquationIndexedAMD)
#define glBlendEquation MANGLE(BlendEquation)
#define glBlendEquationSeparateATI MANGLE(BlendEquationSeparateATI)
#define glBlendEquationSeparateEXT MANGLE(BlendEquationSeparateEXT)
#define glBlendEquationSeparateiARB MANGLE(BlendEquationSeparateiARB)
#define glBlendEquationSeparatei MANGLE(BlendEquationSeparatei)
#define glBlendEquationSeparateIndexedAMD MANGLE(BlendEquationSeparateIndexedAMD)
#define glBlendEquationSeparate MANGLE(BlendEquationSeparate)
#define glBlendFunciARB MANGLE(BlendFunciARB)
#define glBlendFunci MANGLE(BlendFunci)
#define glBlendFuncIndexedAMD MANGLE(BlendFuncIndexedAMD)
#define glBlendFunc MANGLE(BlendFunc)
#define glBlendFuncSeparateEXT MANGLE(BlendFuncSeparateEXT)
#define glBlendFuncSeparateiARB MANGLE(BlendFuncSeparateiARB)
#define glBlendFuncSeparatei MANGLE(BlendFuncSeparatei)
#define glBlendFuncSeparateIndexedAMD MANGLE(BlendFuncSeparateIndexedAMD)
#define glBlendFuncSeparateINGR MANGLE(BlendFuncSeparateINGR)
@@ -153,6 +164,7 @@
#define glClearColor MANGLE(ClearColor)
#define glClearDebugLogMESA MANGLE(ClearDebugLogMESA)
#define glClearDepthdNV MANGLE(ClearDepthdNV)
#define glClearDepthf MANGLE(ClearDepthf)
#define glClearDepth MANGLE(ClearDepth)
#define glClearIndex MANGLE(ClearIndex)
#define glClear MANGLE(Clear)
@@ -215,6 +227,10 @@
#define glColorMaskIndexedEXT MANGLE(ColorMaskIndexedEXT)
#define glColorMask MANGLE(ColorMask)
#define glColorMaterial MANGLE(ColorMaterial)
#define glColorP3ui MANGLE(ColorP3ui)
#define glColorP3uiv MANGLE(ColorP3uiv)
#define glColorP4ui MANGLE(ColorP4ui)
#define glColorP4uiv MANGLE(ColorP4uiv)
#define glColorPointerEXT MANGLE(ColorPointerEXT)
#define glColorPointerListIBM MANGLE(ColorPointerListIBM)
#define glColorPointer MANGLE(ColorPointer)
@@ -236,6 +252,7 @@
#define glCombinerParameterivNV MANGLE(CombinerParameterivNV)
#define glCombinerStageParameterfvNV MANGLE(CombinerStageParameterfvNV)
#define glCompileShaderARB MANGLE(CompileShaderARB)
#define glCompileShaderIncludeARB MANGLE(CompileShaderIncludeARB)
#define glCompileShader MANGLE(CompileShader)
#define glCompressedMultiTexImage1DEXT MANGLE(CompressedMultiTexImage1DEXT)
#define glCompressedMultiTexImage2DEXT MANGLE(CompressedMultiTexImage2DEXT)
@@ -310,10 +327,18 @@
#define glCreateShader MANGLE(CreateShader)
#define glCreateShaderObjectARB MANGLE(CreateShaderObjectARB)
#define glCreateShaderProgramEXT MANGLE(CreateShaderProgramEXT)
#define glCreateShaderProgramv MANGLE(CreateShaderProgramv)
#define glCreateSyncFromCLeventARB MANGLE(CreateSyncFromCLeventARB)
#define glCullFace MANGLE(CullFace)
#define glCullParameterdvEXT MANGLE(CullParameterdvEXT)
#define glCullParameterfvEXT MANGLE(CullParameterfvEXT)
#define glCurrentPaletteMatrixARB MANGLE(CurrentPaletteMatrixARB)
#define glDebugMessageCallbackAMD MANGLE(DebugMessageCallbackAMD)
#define glDebugMessageCallbackARB MANGLE(DebugMessageCallbackARB)
#define glDebugMessageControlARB MANGLE(DebugMessageControlARB)
#define glDebugMessageEnableAMD MANGLE(DebugMessageEnableAMD)
#define glDebugMessageInsertAMD MANGLE(DebugMessageInsertAMD)
#define glDebugMessageInsertARB MANGLE(DebugMessageInsertARB)
#define glDeformationMap3dSGIX MANGLE(DeformationMap3dSGIX)
#define glDeformationMap3fSGIX MANGLE(DeformationMap3fSGIX)
#define glDeformSGIX MANGLE(DeformSGIX)
@@ -326,20 +351,25 @@
#define glDeleteFramebuffersEXT MANGLE(DeleteFramebuffersEXT)
#define glDeleteFramebuffers MANGLE(DeleteFramebuffers)
#define glDeleteLists MANGLE(DeleteLists)
#define glDeleteNamedStringARB MANGLE(DeleteNamedStringARB)
#define glDeleteNamesAMD MANGLE(DeleteNamesAMD)
#define glDeleteObjectARB MANGLE(DeleteObjectARB)
#define glDeleteOcclusionQueriesNV MANGLE(DeleteOcclusionQueriesNV)
#define glDeletePerfMonitorsAMD MANGLE(DeletePerfMonitorsAMD)
#define glDeleteProgram MANGLE(DeleteProgram)
#define glDeleteProgramPipelines MANGLE(DeleteProgramPipelines)
#define glDeleteProgramsARB MANGLE(DeleteProgramsARB)
#define glDeleteProgramsNV MANGLE(DeleteProgramsNV)
#define glDeleteQueriesARB MANGLE(DeleteQueriesARB)
#define glDeleteQueries MANGLE(DeleteQueries)
#define glDeleteRenderbuffersEXT MANGLE(DeleteRenderbuffersEXT)
#define glDeleteRenderbuffers MANGLE(DeleteRenderbuffers)
#define glDeleteSamplers MANGLE(DeleteSamplers)
#define glDeleteShader MANGLE(DeleteShader)
#define glDeleteSync MANGLE(DeleteSync)
#define glDeleteTexturesEXT MANGLE(DeleteTexturesEXT)
#define glDeleteTextures MANGLE(DeleteTextures)
#define glDeleteTransformFeedbacks MANGLE(DeleteTransformFeedbacks)
#define glDeleteTransformFeedbacksNV MANGLE(DeleteTransformFeedbacksNV)
#define glDeleteVertexArraysAPPLE MANGLE(DeleteVertexArraysAPPLE)
#define glDeleteVertexArrays MANGLE(DeleteVertexArrays)
@@ -348,7 +378,10 @@
#define glDepthBoundsEXT MANGLE(DepthBoundsEXT)
#define glDepthFunc MANGLE(DepthFunc)
#define glDepthMask MANGLE(DepthMask)
#define glDepthRangeArrayv MANGLE(DepthRangeArrayv)
#define glDepthRangedNV MANGLE(DepthRangedNV)
#define glDepthRangef MANGLE(DepthRangef)
#define glDepthRangeIndexed MANGLE(DepthRangeIndexed)
#define glDepthRange MANGLE(DepthRange)
#define glDetachObjectARB MANGLE(DetachObjectARB)
#define glDetachShader MANGLE(DetachShader)
@@ -363,6 +396,7 @@
#define glDisableVertexAttribArrayARB MANGLE(DisableVertexAttribArrayARB)
#define glDisableVertexAttribArray MANGLE(DisableVertexAttribArray)
#define glDrawArraysEXT MANGLE(DrawArraysEXT)
#define glDrawArraysIndirect MANGLE(DrawArraysIndirect)
#define glDrawArraysInstancedARB MANGLE(DrawArraysInstancedARB)
#define glDrawArraysInstancedEXT MANGLE(DrawArraysInstancedEXT)
#define glDrawArraysInstanced MANGLE(DrawArraysInstanced)
@@ -374,6 +408,7 @@
#define glDrawElementArrayAPPLE MANGLE(DrawElementArrayAPPLE)
#define glDrawElementArrayATI MANGLE(DrawElementArrayATI)
#define glDrawElementsBaseVertex MANGLE(DrawElementsBaseVertex)
#define glDrawElementsIndirect MANGLE(DrawElementsIndirect)
#define glDrawElementsInstancedARB MANGLE(DrawElementsInstancedARB)
#define glDrawElementsInstancedBaseVertex MANGLE(DrawElementsInstancedBaseVertex)
#define glDrawElementsInstancedEXT MANGLE(DrawElementsInstancedEXT)
@@ -386,7 +421,9 @@
#define glDrawRangeElementsBaseVertex MANGLE(DrawRangeElementsBaseVertex)
#define glDrawRangeElementsEXT MANGLE(DrawRangeElementsEXT)
#define glDrawRangeElements MANGLE(DrawRangeElements)
#define glDrawTransformFeedback MANGLE(DrawTransformFeedback)
#define glDrawTransformFeedbackNV MANGLE(DrawTransformFeedbackNV)
#define glDrawTransformFeedbackStream MANGLE(DrawTransformFeedbackStream)
#define glEdgeFlagFormatNV MANGLE(EdgeFlagFormatNV)
#define glEdgeFlag MANGLE(EdgeFlag)
#define glEdgeFlagPointerEXT MANGLE(EdgeFlagPointerEXT)
@@ -414,6 +451,7 @@
#define glEndOcclusionQueryNV MANGLE(EndOcclusionQueryNV)
#define glEndPerfMonitorAMD MANGLE(EndPerfMonitorAMD)
#define glEndQueryARB MANGLE(EndQueryARB)
#define glEndQueryIndexed MANGLE(EndQueryIndexed)
#define glEndQuery MANGLE(EndQuery)
#define glEndTransformFeedbackEXT MANGLE(EndTransformFeedbackEXT)
#define glEndTransformFeedback MANGLE(EndTransformFeedback)
@@ -447,6 +485,7 @@
#define glFlush MANGLE(Flush)
#define glFlushMappedBufferRangeAPPLE MANGLE(FlushMappedBufferRangeAPPLE)
#define glFlushMappedBufferRange MANGLE(FlushMappedBufferRange)
#define glFlushMappedNamedBufferRangeEXT MANGLE(FlushMappedNamedBufferRangeEXT)
#define glFlushPixelDataRangeNV MANGLE(FlushPixelDataRangeNV)
#define glFlushRasterSGIX MANGLE(FlushRasterSGIX)
#define glFlushVertexArrayRangeAPPLE MANGLE(FlushVertexArrayRangeAPPLE)
@@ -498,7 +537,6 @@
#define glFramebufferTextureEXT MANGLE(FramebufferTextureEXT)
#define glFramebufferTextureFaceARB MANGLE(FramebufferTextureFaceARB)
#define glFramebufferTextureFaceEXT MANGLE(FramebufferTextureFaceEXT)
#define glFramebufferTextureFace MANGLE(FramebufferTextureFace)
#define glFramebufferTextureLayerARB MANGLE(FramebufferTextureLayerARB)
#define glFramebufferTextureLayerEXT MANGLE(FramebufferTextureLayerEXT)
#define glFramebufferTextureLayer MANGLE(FramebufferTextureLayer)
@@ -521,23 +559,30 @@
#define glGenFramebuffersEXT MANGLE(GenFramebuffersEXT)
#define glGenFramebuffers MANGLE(GenFramebuffers)
#define glGenLists MANGLE(GenLists)
#define glGenNamesAMD MANGLE(GenNamesAMD)
#define glGenOcclusionQueriesNV MANGLE(GenOcclusionQueriesNV)
#define glGenPerfMonitorsAMD MANGLE(GenPerfMonitorsAMD)
#define glGenProgramPipelines MANGLE(GenProgramPipelines)
#define glGenProgramsARB MANGLE(GenProgramsARB)
#define glGenProgramsNV MANGLE(GenProgramsNV)
#define glGenQueriesARB MANGLE(GenQueriesARB)
#define glGenQueries MANGLE(GenQueries)
#define glGenRenderbuffersEXT MANGLE(GenRenderbuffersEXT)
#define glGenRenderbuffers MANGLE(GenRenderbuffers)
#define glGenSamplers MANGLE(GenSamplers)
#define glGenSymbolsEXT MANGLE(GenSymbolsEXT)
#define glGenTexturesEXT MANGLE(GenTexturesEXT)
#define glGenTextures MANGLE(GenTextures)
#define glGenTransformFeedbacks MANGLE(GenTransformFeedbacks)
#define glGenTransformFeedbacksNV MANGLE(GenTransformFeedbacksNV)
#define glGenVertexArraysAPPLE MANGLE(GenVertexArraysAPPLE)
#define glGenVertexArrays MANGLE(GenVertexArrays)
#define glGenVertexShadersEXT MANGLE(GenVertexShadersEXT)
#define glGetActiveAttribARB MANGLE(GetActiveAttribARB)
#define glGetActiveAttrib MANGLE(GetActiveAttrib)
#define glGetActiveSubroutineName MANGLE(GetActiveSubroutineName)
#define glGetActiveSubroutineUniformiv MANGLE(GetActiveSubroutineUniformiv)
#define glGetActiveSubroutineUniformName MANGLE(GetActiveSubroutineUniformName)
#define glGetActiveUniformARB MANGLE(GetActiveUniformARB)
#define glGetActiveUniformBlockiv MANGLE(GetActiveUniformBlockiv)
#define glGetActiveUniformBlockName MANGLE(GetActiveUniformBlockName)
@@ -589,16 +634,21 @@
#define glGetConvolutionParameteriv MANGLE(GetConvolutionParameteriv)
#define glGetDebugLogLengthMESA MANGLE(GetDebugLogLengthMESA)
#define glGetDebugLogMESA MANGLE(GetDebugLogMESA)
#define glGetDebugMessageLogAMD MANGLE(GetDebugMessageLogAMD)
#define glGetDebugMessageLogARB MANGLE(GetDebugMessageLogARB)
#define glGetDetailTexFuncSGIS MANGLE(GetDetailTexFuncSGIS)
#define glGetDoubleIndexedvEXT MANGLE(GetDoubleIndexedvEXT)
#define glGetDoublei_v MANGLE(GetDoublei_v)
#define glGetDoublev MANGLE(GetDoublev)
#define glGetError MANGLE(GetError)
#define glGetFenceivNV MANGLE(GetFenceivNV)
#define glGetFinalCombinerInputParameterfvNV MANGLE(GetFinalCombinerInputParameterfvNV)
#define glGetFinalCombinerInputParameterivNV MANGLE(GetFinalCombinerInputParameterivNV)
#define glGetFloatIndexedvEXT MANGLE(GetFloatIndexedvEXT)
#define glGetFloati_v MANGLE(GetFloati_v)
#define glGetFloatv MANGLE(GetFloatv)
#define glGetFogFuncSGIS MANGLE(GetFogFuncSGIS)
#define glGetFragDataIndex MANGLE(GetFragDataIndex)
#define glGetFragDataLocationEXT MANGLE(GetFragDataLocationEXT)
#define glGetFragDataLocation MANGLE(GetFragDataLocation)
#define glGetFragmentLightfvSGIX MANGLE(GetFragmentLightfvSGIX)
@@ -608,6 +658,7 @@
#define glGetFramebufferAttachmentParameterivEXT MANGLE(GetFramebufferAttachmentParameterivEXT)
#define glGetFramebufferAttachmentParameteriv MANGLE(GetFramebufferAttachmentParameteriv)
#define glGetFramebufferParameterivEXT MANGLE(GetFramebufferParameterivEXT)
#define glGetGraphicsResetStatusARB MANGLE(GetGraphicsResetStatusARB)
#define glGetHandleARB MANGLE(GetHandleARB)
#define glGetHistogramEXT MANGLE(GetHistogramEXT)
#define glGetHistogram MANGLE(GetHistogram)
@@ -678,6 +729,26 @@
#define glGetNamedProgramLocalParameterIuivEXT MANGLE(GetNamedProgramLocalParameterIuivEXT)
#define glGetNamedProgramStringEXT MANGLE(GetNamedProgramStringEXT)
#define glGetNamedRenderbufferParameterivEXT MANGLE(GetNamedRenderbufferParameterivEXT)
#define glGetNamedStringARB MANGLE(GetNamedStringARB)
#define glGetNamedStringivARB MANGLE(GetNamedStringivARB)
#define glGetnColorTableARB MANGLE(GetnColorTableARB)
#define glGetnCompressedTexImageARB MANGLE(GetnCompressedTexImageARB)
#define glGetnConvolutionFilterARB MANGLE(GetnConvolutionFilterARB)
#define glGetnHistogramARB MANGLE(GetnHistogramARB)
#define glGetnMapdvARB MANGLE(GetnMapdvARB)
#define glGetnMapfvARB MANGLE(GetnMapfvARB)
#define glGetnMapivARB MANGLE(GetnMapivARB)
#define glGetnMinmaxARB MANGLE(GetnMinmaxARB)
#define glGetnPixelMapfvARB MANGLE(GetnPixelMapfvARB)
#define glGetnPixelMapuivARB MANGLE(GetnPixelMapuivARB)
#define glGetnPixelMapusvARB MANGLE(GetnPixelMapusvARB)
#define glGetnPolygonStippleARB MANGLE(GetnPolygonStippleARB)
#define glGetnSeparableFilterARB MANGLE(GetnSeparableFilterARB)
#define glGetnTexImageARB MANGLE(GetnTexImageARB)
#define glGetnUniformdvARB MANGLE(GetnUniformdvARB)
#define glGetnUniformfvARB MANGLE(GetnUniformfvARB)
#define glGetnUniformivARB MANGLE(GetnUniformivARB)
#define glGetnUniformuivARB MANGLE(GetnUniformuivARB)
#define glGetObjectBufferfvATI MANGLE(GetObjectBufferfvATI)
#define glGetObjectBufferivATI MANGLE(GetObjectBufferivATI)
#define glGetObjectParameterfvARB MANGLE(GetObjectParameterfvARB)
@@ -700,6 +771,7 @@
#define glGetPointervEXT MANGLE(GetPointervEXT)
#define glGetPointerv MANGLE(GetPointerv)
#define glGetPolygonStipple MANGLE(GetPolygonStipple)
#define glGetProgramBinary MANGLE(GetProgramBinary)
#define glGetProgramEnvParameterdvARB MANGLE(GetProgramEnvParameterdvARB)
#define glGetProgramEnvParameterfvARB MANGLE(GetProgramEnvParameterfvARB)
#define glGetProgramEnvParameterIivNV MANGLE(GetProgramEnvParameterIivNV)
@@ -716,28 +788,42 @@
#define glGetProgramNamedParameterfvNV MANGLE(GetProgramNamedParameterfvNV)
#define glGetProgramParameterdvNV MANGLE(GetProgramParameterdvNV)
#define glGetProgramParameterfvNV MANGLE(GetProgramParameterfvNV)
#define glGetProgramPipelineInfoLog MANGLE(GetProgramPipelineInfoLog)
#define glGetProgramPipelineiv MANGLE(GetProgramPipelineiv)
#define glGetProgramRegisterfvMESA MANGLE(GetProgramRegisterfvMESA)
#define glGetProgramStageiv MANGLE(GetProgramStageiv)
#define glGetProgramStringARB MANGLE(GetProgramStringARB)
#define glGetProgramStringNV MANGLE(GetProgramStringNV)
#define glGetProgramSubroutineParameteruivNV MANGLE(GetProgramSubroutineParameteruivNV)
#define glGetQueryIndexediv MANGLE(GetQueryIndexediv)
#define glGetQueryivARB MANGLE(GetQueryivARB)
#define glGetQueryiv MANGLE(GetQueryiv)
#define glGetQueryObjecti64vEXT MANGLE(GetQueryObjecti64vEXT)
#define glGetQueryObjecti64v MANGLE(GetQueryObjecti64v)
#define glGetQueryObjectivARB MANGLE(GetQueryObjectivARB)
#define glGetQueryObjectiv MANGLE(GetQueryObjectiv)
#define glGetQueryObjectui64vEXT MANGLE(GetQueryObjectui64vEXT)
#define glGetQueryObjectui64v MANGLE(GetQueryObjectui64v)
#define glGetQueryObjectuivARB MANGLE(GetQueryObjectuivARB)
#define glGetQueryObjectuiv MANGLE(GetQueryObjectuiv)
#define glGetRenderbufferParameterivEXT MANGLE(GetRenderbufferParameterivEXT)
#define glGetRenderbufferParameteriv MANGLE(GetRenderbufferParameteriv)
#define glGetSamplerParameterfv MANGLE(GetSamplerParameterfv)
#define glGetSamplerParameterIiv MANGLE(GetSamplerParameterIiv)
#define glGetSamplerParameterIuiv MANGLE(GetSamplerParameterIuiv)
#define glGetSamplerParameteriv MANGLE(GetSamplerParameteriv)
#define glGetSeparableFilterEXT MANGLE(GetSeparableFilterEXT)
#define glGetSeparableFilter MANGLE(GetSeparableFilter)
#define glGetShaderInfoLog MANGLE(GetShaderInfoLog)
#define glGetShaderiv MANGLE(GetShaderiv)
#define glGetShaderPrecisionFormat MANGLE(GetShaderPrecisionFormat)
#define glGetShaderSourceARB MANGLE(GetShaderSourceARB)
#define glGetShaderSource MANGLE(GetShaderSource)
#define glGetSharpenTexFuncSGIS MANGLE(GetSharpenTexFuncSGIS)
#define glGetStringi MANGLE(GetStringi)
#define glGetString MANGLE(GetString)
#define glGetSubroutineIndex MANGLE(GetSubroutineIndex)
#define glGetSubroutineUniformLocation MANGLE(GetSubroutineUniformLocation)
#define glGetSynciv MANGLE(GetSynciv)
#define glGetTexBumpParameterfvATI MANGLE(GetTexBumpParameterfvATI)
#define glGetTexBumpParameterivATI MANGLE(GetTexBumpParameterivATI)
@@ -770,14 +856,17 @@
#define glGetTransformFeedbackVaryingNV MANGLE(GetTransformFeedbackVaryingNV)
#define glGetUniformBlockIndex MANGLE(GetUniformBlockIndex)
#define glGetUniformBufferSizeEXT MANGLE(GetUniformBufferSizeEXT)
#define glGetUniformdv MANGLE(GetUniformdv)
#define glGetUniformfvARB MANGLE(GetUniformfvARB)
#define glGetUniformfv MANGLE(GetUniformfv)
#define glGetUniformi64vNV MANGLE(GetUniformi64vNV)
#define glGetUniformIndices MANGLE(GetUniformIndices)
#define glGetUniformivARB MANGLE(GetUniformivARB)
#define glGetUniformiv MANGLE(GetUniformiv)
#define glGetUniformLocationARB MANGLE(GetUniformLocationARB)
#define glGetUniformLocation MANGLE(GetUniformLocation)
#define glGetUniformOffsetEXT MANGLE(GetUniformOffsetEXT)
#define glGetUniformSubroutineuiv MANGLE(GetUniformSubroutineuiv)
#define glGetUniformui64vNV MANGLE(GetUniformui64vNV)
#define glGetUniformuivEXT MANGLE(GetUniformuivEXT)
#define glGetUniformuiv MANGLE(GetUniformuiv)
@@ -803,6 +892,10 @@
#define glGetVertexAttribivARB MANGLE(GetVertexAttribivARB)
#define glGetVertexAttribiv MANGLE(GetVertexAttribiv)
#define glGetVertexAttribivNV MANGLE(GetVertexAttribivNV)
#define glGetVertexAttribLdvEXT MANGLE(GetVertexAttribLdvEXT)
#define glGetVertexAttribLdv MANGLE(GetVertexAttribLdv)
#define glGetVertexAttribLi64vNV MANGLE(GetVertexAttribLi64vNV)
#define glGetVertexAttribLui64vNV MANGLE(GetVertexAttribLui64vNV)
#define glGetVertexAttribPointervARB MANGLE(GetVertexAttribPointervARB)
#define glGetVertexAttribPointerv MANGLE(GetVertexAttribPointerv)
#define glGetVertexAttribPointervNV MANGLE(GetVertexAttribPointervNV)
@@ -864,20 +957,25 @@
#define glIsFramebufferEXT MANGLE(IsFramebufferEXT)
#define glIsFramebuffer MANGLE(IsFramebuffer)
#define glIsList MANGLE(IsList)
#define glIsNameAMD MANGLE(IsNameAMD)
#define glIsNamedBufferResidentNV MANGLE(IsNamedBufferResidentNV)
#define glIsNamedStringARB MANGLE(IsNamedStringARB)
#define glIsObjectBufferATI MANGLE(IsObjectBufferATI)
#define glIsOcclusionQueryNV MANGLE(IsOcclusionQueryNV)
#define glIsProgramARB MANGLE(IsProgramARB)
#define glIsProgram MANGLE(IsProgram)
#define glIsProgramNV MANGLE(IsProgramNV)
#define glIsProgramPipeline MANGLE(IsProgramPipeline)
#define glIsQueryARB MANGLE(IsQueryARB)
#define glIsQuery MANGLE(IsQuery)
#define glIsRenderbufferEXT MANGLE(IsRenderbufferEXT)
#define glIsRenderbuffer MANGLE(IsRenderbuffer)
#define glIsSampler MANGLE(IsSampler)
#define glIsShader MANGLE(IsShader)
#define glIsSync MANGLE(IsSync)
#define glIsTextureEXT MANGLE(IsTextureEXT)
#define glIsTexture MANGLE(IsTexture)
#define glIsTransformFeedback MANGLE(IsTransformFeedback)
#define glIsTransformFeedbackNV MANGLE(IsTransformFeedbackNV)
#define glIsVariantEnabledEXT MANGLE(IsVariantEnabledEXT)
#define glIsVertexArrayAPPLE MANGLE(IsVertexArrayAPPLE)
@@ -915,6 +1013,8 @@
#define glLogicOp MANGLE(LogicOp)
#define glMakeBufferNonResidentNV MANGLE(MakeBufferNonResidentNV)
#define glMakeBufferResidentNV MANGLE(MakeBufferResidentNV)
#define glMakeNamedBufferNonResidentNV MANGLE(MakeNamedBufferNonResidentNV)
#define glMakeNamedBufferResidentNV MANGLE(MakeNamedBufferResidentNV)
#define glMap1d MANGLE(Map1d)
#define glMap1f MANGLE(Map1f)
#define glMap2d MANGLE(Map2d)
@@ -928,6 +1028,7 @@
#define glMapGrid2d MANGLE(MapGrid2d)
#define glMapGrid2f MANGLE(MapGrid2f)
#define glMapNamedBufferEXT MANGLE(MapNamedBufferEXT)
#define glMapNamedBufferRangeEXT MANGLE(MapNamedBufferRangeEXT)
#define glMapObjectBufferATI MANGLE(MapObjectBufferATI)
#define glMapParameterfvNV MANGLE(MapParameterfvNV)
#define glMapParameterivNV MANGLE(MapParameterivNV)
@@ -963,8 +1064,10 @@
#define glMatrixScalefEXT MANGLE(MatrixScalefEXT)
#define glMatrixTranslatedEXT MANGLE(MatrixTranslatedEXT)
#define glMatrixTranslatefEXT MANGLE(MatrixTranslatefEXT)
#define glMemoryBarrierEXT MANGLE(MemoryBarrierEXT)
#define glMinmaxEXT MANGLE(MinmaxEXT)
#define glMinmax MANGLE(Minmax)
#define glMinSampleShadingARB MANGLE(MinSampleShadingARB)
#define glMinSampleShading MANGLE(MinSampleShading)
#define glMultiDrawArraysEXT MANGLE(MultiDrawArraysEXT)
#define glMultiDrawArrays MANGLE(MultiDrawArrays)
@@ -1048,6 +1151,14 @@
#define glMultiTexCoord4s MANGLE(MultiTexCoord4s)
#define glMultiTexCoord4svARB MANGLE(MultiTexCoord4svARB)
#define glMultiTexCoord4sv MANGLE(MultiTexCoord4sv)
#define glMultiTexCoordP1ui MANGLE(MultiTexCoordP1ui)
#define glMultiTexCoordP1uiv MANGLE(MultiTexCoordP1uiv)
#define glMultiTexCoordP2ui MANGLE(MultiTexCoordP2ui)
#define glMultiTexCoordP2uiv MANGLE(MultiTexCoordP2uiv)
#define glMultiTexCoordP3ui MANGLE(MultiTexCoordP3ui)
#define glMultiTexCoordP3uiv MANGLE(MultiTexCoordP3uiv)
#define glMultiTexCoordP4ui MANGLE(MultiTexCoordP4ui)
#define glMultiTexCoordP4uiv MANGLE(MultiTexCoordP4uiv)
#define glMultiTexCoordPointerEXT MANGLE(MultiTexCoordPointerEXT)
#define glMultiTexEnvfEXT MANGLE(MultiTexEnvfEXT)
#define glMultiTexEnvfvEXT MANGLE(MultiTexEnvfvEXT)
@@ -1080,6 +1191,7 @@
#define glMultTransposeMatrixf MANGLE(MultTransposeMatrixf)
#define glNamedBufferDataEXT MANGLE(NamedBufferDataEXT)
#define glNamedBufferSubDataEXT MANGLE(NamedBufferSubDataEXT)
#define glNamedCopyBufferSubDataEXT MANGLE(NamedCopyBufferSubDataEXT)
#define glNamedFramebufferRenderbufferEXT MANGLE(NamedFramebufferRenderbufferEXT)
#define glNamedFramebufferTexture1DEXT MANGLE(NamedFramebufferTexture1DEXT)
#define glNamedFramebufferTexture2DEXT MANGLE(NamedFramebufferTexture2DEXT)
@@ -1087,8 +1199,6 @@
#define glNamedFramebufferTextureEXT MANGLE(NamedFramebufferTextureEXT)
#define glNamedFramebufferTextureFaceEXT MANGLE(NamedFramebufferTextureFaceEXT)
#define glNamedFramebufferTextureLayerEXT MANGLE(NamedFramebufferTextureLayerEXT)
#define glNamedMakeBufferNonResidentNV MANGLE(NamedMakeBufferNonResidentNV)
#define glNamedMakeBufferResidentNV MANGLE(NamedMakeBufferResidentNV)
#define glNamedProgramLocalParameter4dEXT MANGLE(NamedProgramLocalParameter4dEXT)
#define glNamedProgramLocalParameter4dvEXT MANGLE(NamedProgramLocalParameter4dvEXT)
#define glNamedProgramLocalParameter4fEXT MANGLE(NamedProgramLocalParameter4fEXT)
@@ -1104,6 +1214,7 @@
#define glNamedRenderbufferStorageEXT MANGLE(NamedRenderbufferStorageEXT)
#define glNamedRenderbufferStorageMultisampleCoverageEXT MANGLE(NamedRenderbufferStorageMultisampleCoverageEXT)
#define glNamedRenderbufferStorageMultisampleEXT MANGLE(NamedRenderbufferStorageMultisampleEXT)
#define glNamedStringARB MANGLE(NamedStringARB)
#define glNewList MANGLE(NewList)
#define glNewObjectBufferATI MANGLE(NewObjectBufferATI)
#define glNormal3b MANGLE(Normal3b)
@@ -1121,6 +1232,8 @@
#define glNormal3s MANGLE(Normal3s)
#define glNormal3sv MANGLE(Normal3sv)
#define glNormalFormatNV MANGLE(NormalFormatNV)
#define glNormalP3ui MANGLE(NormalP3ui)
#define glNormalP3uiv MANGLE(NormalP3uiv)
#define glNormalPointerEXT MANGLE(NormalPointerEXT)
#define glNormalPointerListIBM MANGLE(NormalPointerListIBM)
#define glNormalPointer MANGLE(NormalPointer)
@@ -1140,6 +1253,9 @@
#define glOrtho MANGLE(Ortho)
#define glPassTexCoordATI MANGLE(PassTexCoordATI)
#define glPassThrough MANGLE(PassThrough)
#define glPatchParameterfv MANGLE(PatchParameterfv)
#define glPatchParameteri MANGLE(PatchParameteri)
#define glPauseTransformFeedback MANGLE(PauseTransformFeedback)
#define glPauseTransformFeedbackNV MANGLE(PauseTransformFeedbackNV)
#define glPixelDataRangeNV MANGLE(PixelDataRangeNV)
#define glPixelMapfv MANGLE(PixelMapfv)
@@ -1191,6 +1307,7 @@
#define glPrimitiveRestartNV MANGLE(PrimitiveRestartNV)
#define glPrioritizeTexturesEXT MANGLE(PrioritizeTexturesEXT)
#define glPrioritizeTextures MANGLE(PrioritizeTextures)
#define glProgramBinary MANGLE(ProgramBinary)
#define glProgramBufferParametersfvNV MANGLE(ProgramBufferParametersfvNV)
#define glProgramBufferParametersIivNV MANGLE(ProgramBufferParametersIivNV)
#define glProgramBufferParametersIuivNV MANGLE(ProgramBufferParametersIuivNV)
@@ -1231,39 +1348,123 @@
#define glProgramParameters4dvNV MANGLE(ProgramParameters4dvNV)
#define glProgramParameters4fvNV MANGLE(ProgramParameters4fvNV)
#define glProgramStringARB MANGLE(ProgramStringARB)
#define glProgramSubroutineParametersuivNV MANGLE(ProgramSubroutineParametersuivNV)
#define glProgramUniform1dEXT MANGLE(ProgramUniform1dEXT)
#define glProgramUniform1d MANGLE(ProgramUniform1d)
#define glProgramUniform1dvEXT MANGLE(ProgramUniform1dvEXT)
#define glProgramUniform1dv MANGLE(ProgramUniform1dv)
#define glProgramUniform1fEXT MANGLE(ProgramUniform1fEXT)
#define glProgramUniform1f MANGLE(ProgramUniform1f)
#define glProgramUniform1fvEXT MANGLE(ProgramUniform1fvEXT)
#define glProgramUniform1fv MANGLE(ProgramUniform1fv)
#define glProgramUniform1i64NV MANGLE(ProgramUniform1i64NV)
#define glProgramUniform1i64vNV MANGLE(ProgramUniform1i64vNV)
#define glProgramUniform1iEXT MANGLE(ProgramUniform1iEXT)
#define glProgramUniform1i MANGLE(ProgramUniform1i)
#define glProgramUniform1ivEXT MANGLE(ProgramUniform1ivEXT)
#define glProgramUniform1iv MANGLE(ProgramUniform1iv)
#define glProgramUniform1ui64NV MANGLE(ProgramUniform1ui64NV)
#define glProgramUniform1ui64vNV MANGLE(ProgramUniform1ui64vNV)
#define glProgramUniform1uiEXT MANGLE(ProgramUniform1uiEXT)
#define glProgramUniform1ui MANGLE(ProgramUniform1ui)
#define glProgramUniform1uivEXT MANGLE(ProgramUniform1uivEXT)
#define glProgramUniform1uiv MANGLE(ProgramUniform1uiv)
#define glProgramUniform2dEXT MANGLE(ProgramUniform2dEXT)
#define glProgramUniform2d MANGLE(ProgramUniform2d)
#define glProgramUniform2dvEXT MANGLE(ProgramUniform2dvEXT)
#define glProgramUniform2dv MANGLE(ProgramUniform2dv)
#define glProgramUniform2fEXT MANGLE(ProgramUniform2fEXT)
#define glProgramUniform2f MANGLE(ProgramUniform2f)
#define glProgramUniform2fvEXT MANGLE(ProgramUniform2fvEXT)
#define glProgramUniform2fv MANGLE(ProgramUniform2fv)
#define glProgramUniform2i64NV MANGLE(ProgramUniform2i64NV)
#define glProgramUniform2i64vNV MANGLE(ProgramUniform2i64vNV)
#define glProgramUniform2iEXT MANGLE(ProgramUniform2iEXT)
#define glProgramUniform2i MANGLE(ProgramUniform2i)
#define glProgramUniform2ivEXT MANGLE(ProgramUniform2ivEXT)
#define glProgramUniform2iv MANGLE(ProgramUniform2iv)
#define glProgramUniform2ui64NV MANGLE(ProgramUniform2ui64NV)
#define glProgramUniform2ui64vNV MANGLE(ProgramUniform2ui64vNV)
#define glProgramUniform2uiEXT MANGLE(ProgramUniform2uiEXT)
#define glProgramUniform2ui MANGLE(ProgramUniform2ui)
#define glProgramUniform2uivEXT MANGLE(ProgramUniform2uivEXT)
#define glProgramUniform2uiv MANGLE(ProgramUniform2uiv)
#define glProgramUniform3dEXT MANGLE(ProgramUniform3dEXT)
#define glProgramUniform3d MANGLE(ProgramUniform3d)
#define glProgramUniform3dvEXT MANGLE(ProgramUniform3dvEXT)
#define glProgramUniform3dv MANGLE(ProgramUniform3dv)
#define glProgramUniform3fEXT MANGLE(ProgramUniform3fEXT)
#define glProgramUniform3f MANGLE(ProgramUniform3f)
#define glProgramUniform3fvEXT MANGLE(ProgramUniform3fvEXT)
#define glProgramUniform3fv MANGLE(ProgramUniform3fv)
#define glProgramUniform3i64NV MANGLE(ProgramUniform3i64NV)
#define glProgramUniform3i64vNV MANGLE(ProgramUniform3i64vNV)
#define glProgramUniform3iEXT MANGLE(ProgramUniform3iEXT)
#define glProgramUniform3i MANGLE(ProgramUniform3i)
#define glProgramUniform3ivEXT MANGLE(ProgramUniform3ivEXT)
#define glProgramUniform3iv MANGLE(ProgramUniform3iv)
#define glProgramUniform3ui64NV MANGLE(ProgramUniform3ui64NV)
#define glProgramUniform3ui64vNV MANGLE(ProgramUniform3ui64vNV)
#define glProgramUniform3uiEXT MANGLE(ProgramUniform3uiEXT)
#define glProgramUniform3ui MANGLE(ProgramUniform3ui)
#define glProgramUniform3uivEXT MANGLE(ProgramUniform3uivEXT)
#define glProgramUniform3uiv MANGLE(ProgramUniform3uiv)
#define glProgramUniform4dEXT MANGLE(ProgramUniform4dEXT)
#define glProgramUniform4d MANGLE(ProgramUniform4d)
#define glProgramUniform4dvEXT MANGLE(ProgramUniform4dvEXT)
#define glProgramUniform4dv MANGLE(ProgramUniform4dv)
#define glProgramUniform4fEXT MANGLE(ProgramUniform4fEXT)
#define glProgramUniform4f MANGLE(ProgramUniform4f)
#define glProgramUniform4fvEXT MANGLE(ProgramUniform4fvEXT)
#define glProgramUniform4fv MANGLE(ProgramUniform4fv)
#define glProgramUniform4i64NV MANGLE(ProgramUniform4i64NV)
#define glProgramUniform4i64vNV MANGLE(ProgramUniform4i64vNV)
#define glProgramUniform4iEXT MANGLE(ProgramUniform4iEXT)
#define glProgramUniform4i MANGLE(ProgramUniform4i)
#define glProgramUniform4ivEXT MANGLE(ProgramUniform4ivEXT)
#define glProgramUniform4iv MANGLE(ProgramUniform4iv)
#define glProgramUniform4ui64NV MANGLE(ProgramUniform4ui64NV)
#define glProgramUniform4ui64vNV MANGLE(ProgramUniform4ui64vNV)
#define glProgramUniform4uiEXT MANGLE(ProgramUniform4uiEXT)
#define glProgramUniform4ui MANGLE(ProgramUniform4ui)
#define glProgramUniform4uivEXT MANGLE(ProgramUniform4uivEXT)
#define glProgramUniform4uiv MANGLE(ProgramUniform4uiv)
#define glProgramUniformMatrix2dvEXT MANGLE(ProgramUniformMatrix2dvEXT)
#define glProgramUniformMatrix2dv MANGLE(ProgramUniformMatrix2dv)
#define glProgramUniformMatrix2fvEXT MANGLE(ProgramUniformMatrix2fvEXT)
#define glProgramUniformMatrix2fv MANGLE(ProgramUniformMatrix2fv)
#define glProgramUniformMatrix2x3dvEXT MANGLE(ProgramUniformMatrix2x3dvEXT)
#define glProgramUniformMatrix2x3dv MANGLE(ProgramUniformMatrix2x3dv)
#define glProgramUniformMatrix2x3fvEXT MANGLE(ProgramUniformMatrix2x3fvEXT)
#define glProgramUniformMatrix2x3fv MANGLE(ProgramUniformMatrix2x3fv)
#define glProgramUniformMatrix2x4dvEXT MANGLE(ProgramUniformMatrix2x4dvEXT)
#define glProgramUniformMatrix2x4dv MANGLE(ProgramUniformMatrix2x4dv)
#define glProgramUniformMatrix2x4fvEXT MANGLE(ProgramUniformMatrix2x4fvEXT)
#define glProgramUniformMatrix2x4fv MANGLE(ProgramUniformMatrix2x4fv)
#define glProgramUniformMatrix3dvEXT MANGLE(ProgramUniformMatrix3dvEXT)
#define glProgramUniformMatrix3dv MANGLE(ProgramUniformMatrix3dv)
#define glProgramUniformMatrix3fvEXT MANGLE(ProgramUniformMatrix3fvEXT)
#define glProgramUniformMatrix3fv MANGLE(ProgramUniformMatrix3fv)
#define glProgramUniformMatrix3x2dvEXT MANGLE(ProgramUniformMatrix3x2dvEXT)
#define glProgramUniformMatrix3x2dv MANGLE(ProgramUniformMatrix3x2dv)
#define glProgramUniformMatrix3x2fvEXT MANGLE(ProgramUniformMatrix3x2fvEXT)
#define glProgramUniformMatrix3x2fv MANGLE(ProgramUniformMatrix3x2fv)
#define glProgramUniformMatrix3x4dvEXT MANGLE(ProgramUniformMatrix3x4dvEXT)
#define glProgramUniformMatrix3x4dv MANGLE(ProgramUniformMatrix3x4dv)
#define glProgramUniformMatrix3x4fvEXT MANGLE(ProgramUniformMatrix3x4fvEXT)
#define glProgramUniformMatrix3x4fv MANGLE(ProgramUniformMatrix3x4fv)
#define glProgramUniformMatrix4dvEXT MANGLE(ProgramUniformMatrix4dvEXT)
#define glProgramUniformMatrix4dv MANGLE(ProgramUniformMatrix4dv)
#define glProgramUniformMatrix4fvEXT MANGLE(ProgramUniformMatrix4fvEXT)
#define glProgramUniformMatrix4fv MANGLE(ProgramUniformMatrix4fv)
#define glProgramUniformMatrix4x2dvEXT MANGLE(ProgramUniformMatrix4x2dvEXT)
#define glProgramUniformMatrix4x2dv MANGLE(ProgramUniformMatrix4x2dv)
#define glProgramUniformMatrix4x2fvEXT MANGLE(ProgramUniformMatrix4x2fvEXT)
#define glProgramUniformMatrix4x2fv MANGLE(ProgramUniformMatrix4x2fv)
#define glProgramUniformMatrix4x3dvEXT MANGLE(ProgramUniformMatrix4x3dvEXT)
#define glProgramUniformMatrix4x3dv MANGLE(ProgramUniformMatrix4x3dv)
#define glProgramUniformMatrix4x3fvEXT MANGLE(ProgramUniformMatrix4x3fvEXT)
#define glProgramUniformMatrix4x3fv MANGLE(ProgramUniformMatrix4x3fv)
#define glProgramUniformui64NV MANGLE(ProgramUniformui64NV)
#define glProgramUniformui64vNV MANGLE(ProgramUniformui64vNV)
#define glProgramVertexLimitNV MANGLE(ProgramVertexLimitNV)
@@ -1274,6 +1475,7 @@
#define glPushClientAttrib MANGLE(PushClientAttrib)
#define glPushMatrix MANGLE(PushMatrix)
#define glPushName MANGLE(PushName)
#define glQueryCounter MANGLE(QueryCounter)
#define glRasterPos2d MANGLE(RasterPos2d)
#define glRasterPos2dv MANGLE(RasterPos2dv)
#define glRasterPos2f MANGLE(RasterPos2f)
@@ -1300,6 +1502,7 @@
#define glRasterPos4sv MANGLE(RasterPos4sv)
#define glReadBuffer MANGLE(ReadBuffer)
#define glReadInstrumentsSGIX MANGLE(ReadInstrumentsSGIX)
#define glReadnPixelsARB MANGLE(ReadnPixelsARB)
#define glReadPixels MANGLE(ReadPixels)
#define glRectd MANGLE(Rectd)
#define glRectdv MANGLE(Rectdv)
@@ -1310,6 +1513,7 @@
#define glRects MANGLE(Rects)
#define glRectsv MANGLE(Rectsv)
#define glReferencePlaneSGIX MANGLE(ReferencePlaneSGIX)
#define glReleaseShaderCompiler MANGLE(ReleaseShaderCompiler)
#define glRenderbufferStorageEXT MANGLE(RenderbufferStorageEXT)
#define glRenderbufferStorage MANGLE(RenderbufferStorage)
#define glRenderbufferStorageMultisampleCoverageNV MANGLE(RenderbufferStorageMultisampleCoverageNV)
@@ -1345,6 +1549,7 @@
#define glResetMinmaxEXT MANGLE(ResetMinmaxEXT)
#define glResetMinmax MANGLE(ResetMinmax)
#define glResizeBuffersMESA MANGLE(ResizeBuffersMESA)
#define glResumeTransformFeedback MANGLE(ResumeTransformFeedback)
#define glResumeTransformFeedbackNV MANGLE(ResumeTransformFeedbackNV)
#define glRotated MANGLE(Rotated)
#define glRotatef MANGLE(Rotatef)
@@ -1357,8 +1562,17 @@
#define glSampleMaskSGIS MANGLE(SampleMaskSGIS)
#define glSamplePatternEXT MANGLE(SamplePatternEXT)
#define glSamplePatternSGIS MANGLE(SamplePatternSGIS)
#define glSamplerParameterf MANGLE(SamplerParameterf)
#define glSamplerParameterfv MANGLE(SamplerParameterfv)
#define glSamplerParameterIiv MANGLE(SamplerParameterIiv)
#define glSamplerParameteri MANGLE(SamplerParameteri)
#define glSamplerParameterIuiv MANGLE(SamplerParameterIuiv)
#define glSamplerParameteriv MANGLE(SamplerParameteriv)
#define glScaled MANGLE(Scaled)
#define glScalef MANGLE(Scalef)
#define glScissorArrayv MANGLE(ScissorArrayv)
#define glScissorIndexed MANGLE(ScissorIndexed)
#define glScissorIndexedv MANGLE(ScissorIndexedv)
#define glScissor MANGLE(Scissor)
#define glSecondaryColor3bEXT MANGLE(SecondaryColor3bEXT)
#define glSecondaryColor3b MANGLE(SecondaryColor3b)
@@ -1395,6 +1609,8 @@
#define glSecondaryColor3usvEXT MANGLE(SecondaryColor3usvEXT)
#define glSecondaryColor3usv MANGLE(SecondaryColor3usv)
#define glSecondaryColorFormatNV MANGLE(SecondaryColorFormatNV)
#define glSecondaryColorP3ui MANGLE(SecondaryColorP3ui)
#define glSecondaryColorP3uiv MANGLE(SecondaryColorP3uiv)
#define glSecondaryColorPointerEXT MANGLE(SecondaryColorPointerEXT)
#define glSecondaryColorPointerListIBM MANGLE(SecondaryColorPointerListIBM)
#define glSecondaryColorPointer MANGLE(SecondaryColorPointer)
@@ -1408,6 +1624,7 @@
#define glSetInvariantEXT MANGLE(SetInvariantEXT)
#define glSetLocalConstantEXT MANGLE(SetLocalConstantEXT)
#define glShadeModel MANGLE(ShadeModel)
#define glShaderBinary MANGLE(ShaderBinary)
#define glShaderOp1EXT MANGLE(ShaderOp1EXT)
#define glShaderOp2EXT MANGLE(ShaderOp2EXT)
#define glShaderOp3EXT MANGLE(ShaderOp3EXT)
@@ -1509,6 +1726,14 @@
#define glTexCoord4s MANGLE(TexCoord4s)
#define glTexCoord4sv MANGLE(TexCoord4sv)
#define glTexCoordFormatNV MANGLE(TexCoordFormatNV)
#define glTexCoordP1ui MANGLE(TexCoordP1ui)
#define glTexCoordP1uiv MANGLE(TexCoordP1uiv)
#define glTexCoordP2ui MANGLE(TexCoordP2ui)
#define glTexCoordP2uiv MANGLE(TexCoordP2uiv)
#define glTexCoordP3ui MANGLE(TexCoordP3ui)
#define glTexCoordP3uiv MANGLE(TexCoordP3uiv)
#define glTexCoordP4ui MANGLE(TexCoordP4ui)
#define glTexCoordP4uiv MANGLE(TexCoordP4uiv)
#define glTexCoordPointerEXT MANGLE(TexCoordPointerEXT)
#define glTexCoordPointerListIBM MANGLE(TexCoordPointerListIBM)
#define glTexCoordPointer MANGLE(TexCoordPointer)
@@ -1569,73 +1794,108 @@
#define glTextureSubImage3DEXT MANGLE(TextureSubImage3DEXT)
#define glTrackMatrixNV MANGLE(TrackMatrixNV)
#define glTransformFeedbackAttribsNV MANGLE(TransformFeedbackAttribsNV)
#define glTransformFeedbackStreamAttribsNV MANGLE(TransformFeedbackStreamAttribsNV)
#define glTransformFeedbackVaryingsEXT MANGLE(TransformFeedbackVaryingsEXT)
#define glTransformFeedbackVaryings MANGLE(TransformFeedbackVaryings)
#define glTransformFeedbackVaryingsNV MANGLE(TransformFeedbackVaryingsNV)
#define glTranslated MANGLE(Translated)
#define glTranslatef MANGLE(Translatef)
#define glUniform1d MANGLE(Uniform1d)
#define glUniform1dv MANGLE(Uniform1dv)
#define glUniform1fARB MANGLE(Uniform1fARB)
#define glUniform1f MANGLE(Uniform1f)
#define glUniform1fvARB MANGLE(Uniform1fvARB)
#define glUniform1fv MANGLE(Uniform1fv)
#define glUniform1i64NV MANGLE(Uniform1i64NV)
#define glUniform1i64vNV MANGLE(Uniform1i64vNV)
#define glUniform1iARB MANGLE(Uniform1iARB)
#define glUniform1i MANGLE(Uniform1i)
#define glUniform1ivARB MANGLE(Uniform1ivARB)
#define glUniform1iv MANGLE(Uniform1iv)
#define glUniform1ui64NV MANGLE(Uniform1ui64NV)
#define glUniform1ui64vNV MANGLE(Uniform1ui64vNV)
#define glUniform1uiEXT MANGLE(Uniform1uiEXT)
#define glUniform1ui MANGLE(Uniform1ui)
#define glUniform1uivEXT MANGLE(Uniform1uivEXT)
#define glUniform1uiv MANGLE(Uniform1uiv)
#define glUniform2d MANGLE(Uniform2d)
#define glUniform2dv MANGLE(Uniform2dv)
#define glUniform2fARB MANGLE(Uniform2fARB)
#define glUniform2f MANGLE(Uniform2f)
#define glUniform2fvARB MANGLE(Uniform2fvARB)
#define glUniform2fv MANGLE(Uniform2fv)
#define glUniform2i64NV MANGLE(Uniform2i64NV)
#define glUniform2i64vNV MANGLE(Uniform2i64vNV)
#define glUniform2iARB MANGLE(Uniform2iARB)
#define glUniform2i MANGLE(Uniform2i)
#define glUniform2ivARB MANGLE(Uniform2ivARB)
#define glUniform2iv MANGLE(Uniform2iv)
#define glUniform2ui64NV MANGLE(Uniform2ui64NV)
#define glUniform2ui64vNV MANGLE(Uniform2ui64vNV)
#define glUniform2uiEXT MANGLE(Uniform2uiEXT)
#define glUniform2ui MANGLE(Uniform2ui)
#define glUniform2uivEXT MANGLE(Uniform2uivEXT)
#define glUniform2uiv MANGLE(Uniform2uiv)
#define glUniform3d MANGLE(Uniform3d)
#define glUniform3dv MANGLE(Uniform3dv)
#define glUniform3fARB MANGLE(Uniform3fARB)
#define glUniform3f MANGLE(Uniform3f)
#define glUniform3fvARB MANGLE(Uniform3fvARB)
#define glUniform3fv MANGLE(Uniform3fv)
#define glUniform3i64NV MANGLE(Uniform3i64NV)
#define glUniform3i64vNV MANGLE(Uniform3i64vNV)
#define glUniform3iARB MANGLE(Uniform3iARB)
#define glUniform3i MANGLE(Uniform3i)
#define glUniform3ivARB MANGLE(Uniform3ivARB)
#define glUniform3iv MANGLE(Uniform3iv)
#define glUniform3ui64NV MANGLE(Uniform3ui64NV)
#define glUniform3ui64vNV MANGLE(Uniform3ui64vNV)
#define glUniform3uiEXT MANGLE(Uniform3uiEXT)
#define glUniform3ui MANGLE(Uniform3ui)
#define glUniform3uivEXT MANGLE(Uniform3uivEXT)
#define glUniform3uiv MANGLE(Uniform3uiv)
#define glUniform4d MANGLE(Uniform4d)
#define glUniform4dv MANGLE(Uniform4dv)
#define glUniform4fARB MANGLE(Uniform4fARB)
#define glUniform4f MANGLE(Uniform4f)
#define glUniform4fvARB MANGLE(Uniform4fvARB)
#define glUniform4fv MANGLE(Uniform4fv)
#define glUniform4i64NV MANGLE(Uniform4i64NV)
#define glUniform4i64vNV MANGLE(Uniform4i64vNV)
#define glUniform4iARB MANGLE(Uniform4iARB)
#define glUniform4i MANGLE(Uniform4i)
#define glUniform4ivARB MANGLE(Uniform4ivARB)
#define glUniform4iv MANGLE(Uniform4iv)
#define glUniform4ui64NV MANGLE(Uniform4ui64NV)
#define glUniform4ui64vNV MANGLE(Uniform4ui64vNV)
#define glUniform4uiEXT MANGLE(Uniform4uiEXT)
#define glUniform4ui MANGLE(Uniform4ui)
#define glUniform4uivEXT MANGLE(Uniform4uivEXT)
#define glUniform4uiv MANGLE(Uniform4uiv)
#define glUniformBlockBinding MANGLE(UniformBlockBinding)
#define glUniformBufferEXT MANGLE(UniformBufferEXT)
#define glUniformMatrix2dv MANGLE(UniformMatrix2dv)
#define glUniformMatrix2fvARB MANGLE(UniformMatrix2fvARB)
#define glUniformMatrix2fv MANGLE(UniformMatrix2fv)
#define glUniformMatrix2x3dv MANGLE(UniformMatrix2x3dv)
#define glUniformMatrix2x3fv MANGLE(UniformMatrix2x3fv)
#define glUniformMatrix2x4dv MANGLE(UniformMatrix2x4dv)
#define glUniformMatrix2x4fv MANGLE(UniformMatrix2x4fv)
#define glUniformMatrix3dv MANGLE(UniformMatrix3dv)
#define glUniformMatrix3fvARB MANGLE(UniformMatrix3fvARB)
#define glUniformMatrix3fv MANGLE(UniformMatrix3fv)
#define glUniformMatrix3x2dv MANGLE(UniformMatrix3x2dv)
#define glUniformMatrix3x2fv MANGLE(UniformMatrix3x2fv)
#define glUniformMatrix3x4dv MANGLE(UniformMatrix3x4dv)
#define glUniformMatrix3x4fv MANGLE(UniformMatrix3x4fv)
#define glUniformMatrix4dv MANGLE(UniformMatrix4dv)
#define glUniformMatrix4fvARB MANGLE(UniformMatrix4fvARB)
#define glUniformMatrix4fv MANGLE(UniformMatrix4fv)
#define glUniformMatrix4x2dv MANGLE(UniformMatrix4x2dv)
#define glUniformMatrix4x2fv MANGLE(UniformMatrix4x2fv)
#define glUniformMatrix4x3dv MANGLE(UniformMatrix4x3dv)
#define glUniformMatrix4x3fv MANGLE(UniformMatrix4x3fv)
#define glUniformSubroutinesuiv MANGLE(UniformSubroutinesuiv)
#define glUniformui64NV MANGLE(Uniformui64NV)
#define glUniformui64vNV MANGLE(Uniformui64vNV)
#define glUnlockArraysEXT MANGLE(UnlockArraysEXT)
@@ -1646,9 +1906,11 @@
#define glUpdateObjectBufferATI MANGLE(UpdateObjectBufferATI)
#define glUseProgram MANGLE(UseProgram)
#define glUseProgramObjectARB MANGLE(UseProgramObjectARB)
#define glUseProgramStages MANGLE(UseProgramStages)
#define glUseShaderProgramEXT MANGLE(UseShaderProgramEXT)
#define glValidateProgramARB MANGLE(ValidateProgramARB)
#define glValidateProgram MANGLE(ValidateProgram)
#define glValidateProgramPipeline MANGLE(ValidateProgramPipeline)
#define glVariantArrayObjectATI MANGLE(VariantArrayObjectATI)
#define glVariantbvEXT MANGLE(VariantbvEXT)
#define glVariantdvEXT MANGLE(VariantdvEXT)
@@ -1659,6 +1921,16 @@
#define glVariantubvEXT MANGLE(VariantubvEXT)
#define glVariantuivEXT MANGLE(VariantuivEXT)
#define glVariantusvEXT MANGLE(VariantusvEXT)
#define glVDPAUFiniNV MANGLE(VDPAUFiniNV)
#define glVDPAUGetSurfaceivNV MANGLE(VDPAUGetSurfaceivNV)
#define glVDPAUInitNV MANGLE(VDPAUInitNV)
#define glVDPAUIsSurfaceNV MANGLE(VDPAUIsSurfaceNV)
#define glVDPAUMapSurfacesNV MANGLE(VDPAUMapSurfacesNV)
#define glVDPAURegisterOutputSurfaceNV MANGLE(VDPAURegisterOutputSurfaceNV)
#define glVDPAURegisterVideoSurfaceNV MANGLE(VDPAURegisterVideoSurfaceNV)
#define glVDPAUSurfaceAccessNV MANGLE(VDPAUSurfaceAccessNV)
#define glVDPAUUnmapSurfacesNV MANGLE(VDPAUUnmapSurfacesNV)
#define glVDPAUUnregisterSurfaceNV MANGLE(VDPAUUnregisterSurfaceNV)
#define glVertex2d MANGLE(Vertex2d)
#define glVertex2dv MANGLE(Vertex2dv)
#define glVertex2f MANGLE(Vertex2f)
@@ -1692,6 +1964,7 @@
#define glVertexArrayParameteriAPPLE MANGLE(VertexArrayParameteriAPPLE)
#define glVertexArrayRangeAPPLE MANGLE(VertexArrayRangeAPPLE)
#define glVertexArrayRangeNV MANGLE(VertexArrayRangeNV)
#define glVertexArrayVertexAttribLOffsetEXT MANGLE(VertexArrayVertexAttribLOffsetEXT)
#define glVertexAttrib1dARB MANGLE(VertexAttrib1dARB)
#define glVertexAttrib1d MANGLE(VertexAttrib1d)
#define glVertexAttrib1dNV MANGLE(VertexAttrib1dNV)
@@ -1800,6 +2073,7 @@
#define glVertexAttrib4usv MANGLE(VertexAttrib4usv)
#define glVertexAttribArrayObjectATI MANGLE(VertexAttribArrayObjectATI)
#define glVertexAttribDivisorARB MANGLE(VertexAttribDivisorARB)
#define glVertexAttribDivisor MANGLE(VertexAttribDivisor)
#define glVertexAttribFormatNV MANGLE(VertexAttribFormatNV)
#define glVertexAttribI1iEXT MANGLE(VertexAttribI1iEXT)
#define glVertexAttribI1i MANGLE(VertexAttribI1i)
@@ -1844,6 +2118,49 @@
#define glVertexAttribIFormatNV MANGLE(VertexAttribIFormatNV)
#define glVertexAttribIPointerEXT MANGLE(VertexAttribIPointerEXT)
#define glVertexAttribIPointer MANGLE(VertexAttribIPointer)
#define glVertexAttribL1dEXT MANGLE(VertexAttribL1dEXT)
#define glVertexAttribL1d MANGLE(VertexAttribL1d)
#define glVertexAttribL1dvEXT MANGLE(VertexAttribL1dvEXT)
#define glVertexAttribL1dv MANGLE(VertexAttribL1dv)
#define glVertexAttribL1i64NV MANGLE(VertexAttribL1i64NV)
#define glVertexAttribL1i64vNV MANGLE(VertexAttribL1i64vNV)
#define glVertexAttribL1ui64NV MANGLE(VertexAttribL1ui64NV)
#define glVertexAttribL1ui64vNV MANGLE(VertexAttribL1ui64vNV)
#define glVertexAttribL2dEXT MANGLE(VertexAttribL2dEXT)
#define glVertexAttribL2d MANGLE(VertexAttribL2d)
#define glVertexAttribL2dvEXT MANGLE(VertexAttribL2dvEXT)
#define glVertexAttribL2dv MANGLE(VertexAttribL2dv)
#define glVertexAttribL2i64NV MANGLE(VertexAttribL2i64NV)
#define glVertexAttribL2i64vNV MANGLE(VertexAttribL2i64vNV)
#define glVertexAttribL2ui64NV MANGLE(VertexAttribL2ui64NV)
#define glVertexAttribL2ui64vNV MANGLE(VertexAttribL2ui64vNV)
#define glVertexAttribL3dEXT MANGLE(VertexAttribL3dEXT)
#define glVertexAttribL3d MANGLE(VertexAttribL3d)
#define glVertexAttribL3dvEXT MANGLE(VertexAttribL3dvEXT)
#define glVertexAttribL3dv MANGLE(VertexAttribL3dv)
#define glVertexAttribL3i64NV MANGLE(VertexAttribL3i64NV)
#define glVertexAttribL3i64vNV MANGLE(VertexAttribL3i64vNV)
#define glVertexAttribL3ui64NV MANGLE(VertexAttribL3ui64NV)
#define glVertexAttribL3ui64vNV MANGLE(VertexAttribL3ui64vNV)
#define glVertexAttribL4dEXT MANGLE(VertexAttribL4dEXT)
#define glVertexAttribL4d MANGLE(VertexAttribL4d)
#define glVertexAttribL4dvEXT MANGLE(VertexAttribL4dvEXT)
#define glVertexAttribL4dv MANGLE(VertexAttribL4dv)
#define glVertexAttribL4i64NV MANGLE(VertexAttribL4i64NV)
#define glVertexAttribL4i64vNV MANGLE(VertexAttribL4i64vNV)
#define glVertexAttribL4ui64NV MANGLE(VertexAttribL4ui64NV)
#define glVertexAttribL4ui64vNV MANGLE(VertexAttribL4ui64vNV)
#define glVertexAttribLFormatNV MANGLE(VertexAttribLFormatNV)
#define glVertexAttribLPointerEXT MANGLE(VertexAttribLPointerEXT)
#define glVertexAttribLPointer MANGLE(VertexAttribLPointer)
#define glVertexAttribP1ui MANGLE(VertexAttribP1ui)
#define glVertexAttribP1uiv MANGLE(VertexAttribP1uiv)
#define glVertexAttribP2ui MANGLE(VertexAttribP2ui)
#define glVertexAttribP2uiv MANGLE(VertexAttribP2uiv)
#define glVertexAttribP3ui MANGLE(VertexAttribP3ui)
#define glVertexAttribP3uiv MANGLE(VertexAttribP3uiv)
#define glVertexAttribP4ui MANGLE(VertexAttribP4ui)
#define glVertexAttribP4uiv MANGLE(VertexAttribP4uiv)
#define glVertexAttribPointerARB MANGLE(VertexAttribPointerARB)
#define glVertexAttribPointer MANGLE(VertexAttribPointer)
#define glVertexAttribPointerNV MANGLE(VertexAttribPointerNV)
@@ -1868,6 +2185,12 @@
#define glVertexBlendEnvfATI MANGLE(VertexBlendEnvfATI)
#define glVertexBlendEnviATI MANGLE(VertexBlendEnviATI)
#define glVertexFormatNV MANGLE(VertexFormatNV)
#define glVertexP2ui MANGLE(VertexP2ui)
#define glVertexP2uiv MANGLE(VertexP2uiv)
#define glVertexP3ui MANGLE(VertexP3ui)
#define glVertexP3uiv MANGLE(VertexP3uiv)
#define glVertexP4ui MANGLE(VertexP4ui)
#define glVertexP4uiv MANGLE(VertexP4uiv)
#define glVertexPointerEXT MANGLE(VertexPointerEXT)
#define glVertexPointerListIBM MANGLE(VertexPointerListIBM)
#define glVertexPointer MANGLE(VertexPointer)
@@ -1913,6 +2236,9 @@
#define glVideoCaptureStreamParameterdvNV MANGLE(VideoCaptureStreamParameterdvNV)
#define glVideoCaptureStreamParameterfvNV MANGLE(VideoCaptureStreamParameterfvNV)
#define glVideoCaptureStreamParameterivNV MANGLE(VideoCaptureStreamParameterivNV)
#define glViewportArrayv MANGLE(ViewportArrayv)
#define glViewportIndexedf MANGLE(ViewportIndexedf)
#define glViewportIndexedfv MANGLE(ViewportIndexedfv)
#define glViewport MANGLE(Viewport)
#define glWaitSync MANGLE(WaitSync)
#define glWeightbvARB MANGLE(WeightbvARB)

View File

@@ -29,9 +29,9 @@ extern "C" {
*/
/* Header file version number, required by OpenGL ABI for Linux */
/* glext.h last updated $Date: 2010-08-03 01:30:25 -0700 (Tue, 03 Aug 2010) $ */
/* glext.h last updated $Date: 2010-11-03 18:59:30 -0700 (Wed, 03 Nov 2010) $ */
/* Current version at http://www.opengl.org/registry/ */
#define GL_GLEXT_VERSION 64
#define GL_GLEXT_VERSION 66
/* Function declaration macros - to move into glplatform.h */
#if defined(_WIN32) && !defined(APIENTRY) && !defined(__CYGWIN__) && !defined(__SCITECH_SNAP__)
@@ -4840,7 +4840,7 @@ extern "C" {
#endif
#ifndef GL_AMD_seamless_cubemap_per_texture
/* reuse GL_TEXTURE_CUBE_MAP_SEAMLESS_ARB */
/* reuse GL_TEXTURE_CUBE_MAP_SEAMLESS */
#endif
#ifndef GL_AMD_conservative_depth
@@ -4925,6 +4925,8 @@ extern "C" {
#define GL_MIN_FRAGMENT_INTERPOLATION_OFFSET_NV 0x8E5B
#define GL_MAX_FRAGMENT_INTERPOLATION_OFFSET_NV 0x8E5C
#define GL_FRAGMENT_PROGRAM_INTERPOLATION_OFFSET_BITS_NV 0x8E5D
#define GL_MIN_PROGRAM_TEXTURE_GATHER_OFFSET_NV 0x8E5E
#define GL_MAX_PROGRAM_TEXTURE_GATHER_OFFSET_NV 0x8E5F
#define GL_MAX_PROGRAM_SUBROUTINE_PARAMETERS_NV 0x8F44
#define GL_MAX_PROGRAM_SUBROUTINE_NUM_NV 0x8F45
#endif
@@ -5019,6 +5021,11 @@ extern "C" {
#ifndef GL_AMD_transform_feedback3_lines_triangles
#endif
#ifndef GL_AMD_depth_clamp_separate
#define GL_DEPTH_CLAMP_NEAR_AMD 0x901E
#define GL_DEPTH_CLAMP_FAR_AMD 0x901F
#endif
/*************************************************************/
@@ -8765,8 +8772,8 @@ GLAPI void APIENTRY glProgramParameter4dNV (GLenum target, GLuint index, GLdoubl
GLAPI void APIENTRY glProgramParameter4dvNV (GLenum target, GLuint index, const GLdouble *v);
GLAPI void APIENTRY glProgramParameter4fNV (GLenum target, GLuint index, GLfloat x, GLfloat y, GLfloat z, GLfloat w);
GLAPI void APIENTRY glProgramParameter4fvNV (GLenum target, GLuint index, const GLfloat *v);
GLAPI void APIENTRY glProgramParameters4dvNV (GLenum target, GLuint index, GLuint count, const GLdouble *v);
GLAPI void APIENTRY glProgramParameters4fvNV (GLenum target, GLuint index, GLuint count, const GLfloat *v);
GLAPI void APIENTRY glProgramParameters4dvNV (GLenum target, GLuint index, GLsizei count, const GLdouble *v);
GLAPI void APIENTRY glProgramParameters4fvNV (GLenum target, GLuint index, GLsizei count, const GLfloat *v);
GLAPI void APIENTRY glRequestResidentProgramsNV (GLsizei n, const GLuint *programs);
GLAPI void APIENTRY glTrackMatrixNV (GLenum target, GLuint address, GLenum matrix, GLenum transform);
GLAPI void APIENTRY glVertexAttribPointerNV (GLuint index, GLint fsize, GLenum type, GLsizei stride, const GLvoid *pointer);
@@ -8830,8 +8837,8 @@ typedef void (APIENTRYP PFNGLPROGRAMPARAMETER4DNVPROC) (GLenum target, GLuint in
typedef void (APIENTRYP PFNGLPROGRAMPARAMETER4DVNVPROC) (GLenum target, GLuint index, const GLdouble *v);
typedef void (APIENTRYP PFNGLPROGRAMPARAMETER4FNVPROC) (GLenum target, GLuint index, GLfloat x, GLfloat y, GLfloat z, GLfloat w);
typedef void (APIENTRYP PFNGLPROGRAMPARAMETER4FVNVPROC) (GLenum target, GLuint index, const GLfloat *v);
typedef void (APIENTRYP PFNGLPROGRAMPARAMETERS4DVNVPROC) (GLenum target, GLuint index, GLuint count, const GLdouble *v);
typedef void (APIENTRYP PFNGLPROGRAMPARAMETERS4FVNVPROC) (GLenum target, GLuint index, GLuint count, const GLfloat *v);
typedef void (APIENTRYP PFNGLPROGRAMPARAMETERS4DVNVPROC) (GLenum target, GLuint index, GLsizei count, const GLdouble *v);
typedef void (APIENTRYP PFNGLPROGRAMPARAMETERS4FVNVPROC) (GLenum target, GLuint index, GLsizei count, const GLfloat *v);
typedef void (APIENTRYP PFNGLREQUESTRESIDENTPROGRAMSNVPROC) (GLsizei n, const GLuint *programs);
typedef void (APIENTRYP PFNGLTRACKMATRIXNVPROC) (GLenum target, GLuint address, GLenum matrix, GLenum transform);
typedef void (APIENTRYP PFNGLVERTEXATTRIBPOINTERNVPROC) (GLuint index, GLint fsize, GLenum type, GLsizei stride, const GLvoid *pointer);
@@ -11020,6 +11027,10 @@ typedef void (APIENTRYP PFNGLVDPAUUNMAPSURFACESNVPROC) (GLsizei numSurface, cons
#define GL_AMD_transform_feedback3_lines_triangles 1
#endif
#ifndef GL_AMD_depth_clamp_separate
#define GL_AMD_depth_clamp_separate 1
#endif
#ifdef __cplusplus
}

View File

@@ -1,8 +1,8 @@
/* $Revision: 6822 $ on $Date:: 2008-10-30 05:14:19 -0400 #$ */
/* $Revision: 9203 $ on $Date:: 2009-10-07 02:21:52 -0700 #$ */
/*------------------------------------------------------------------------
*
* OpenVG 1.0.1 Reference Implementation
* OpenVG 1.1 Reference Implementation
* -------------------------------------
*
* Copyright (c) 2008 The Khronos Group Inc.
@@ -28,7 +28,7 @@
*
*//**
* \file
* \brief OpenVG 1.0.1 API.
* \brief OpenVG 1.1 API.
*//*-------------------------------------------------------------------*/
#ifndef _OPENVG_H
@@ -42,6 +42,7 @@ extern "C" {
#define OPENVG_VERSION_1_0 1
#define OPENVG_VERSION_1_0_1 1
#define OPENVG_VERSION_1_1 2
#ifndef VG_MAXSHORT
#define VG_MAXSHORT 0x7FFF
@@ -55,10 +56,12 @@ extern "C" {
#define VG_MAX_ENUM 0x7FFFFFFF
#endif
typedef long VGHandle;
typedef VGuint VGHandle;
typedef VGHandle VGPath;
typedef VGHandle VGImage;
typedef VGHandle VGMaskLayer;
typedef VGHandle VGFont;
typedef VGHandle VGPaint;
#define VG_INVALID_HANDLE ((VGHandle)0)
@@ -96,6 +99,10 @@ typedef enum {
/* Scissoring rectangles */
VG_SCISSOR_RECTS = 0x1106,
/* Color Transformation */
VG_COLOR_TRANSFORM = 0x1170,
VG_COLOR_TRANSFORM_VALUES = 0x1171,
/* Stroke parameters */
VG_STROKE_LINE_WIDTH = 0x1110,
VG_STROKE_CAP_STYLE = 0x1111,
@@ -111,6 +118,9 @@ typedef enum {
/* Color for vgClear */
VG_CLEAR_COLOR = 0x1121,
/* Glyph origin */
VG_GLYPH_ORIGIN = 0x1122,
/* Enable/disable alpha masking and scissoring */
VG_MASKING = 0x1130,
VG_SCISSORING = 0x1131,
@@ -165,6 +175,7 @@ typedef enum {
VG_MATRIX_IMAGE_USER_TO_SURFACE = 0x1401,
VG_MATRIX_FILL_PAINT_TO_USER = 0x1402,
VG_MATRIX_STROKE_PAINT_TO_USER = 0x1403,
VG_MATRIX_GLYPH_USER_TO_SURFACE = 0x1404,
VG_MATRIX_MODE_FORCE_SIZE = VG_MAX_ENUM
} VGMatrixMode;
@@ -365,6 +376,8 @@ typedef enum {
VG_lL_8 = 10,
VG_A_8 = 11,
VG_BW_1 = 12,
VG_A_1 = 13,
VG_A_4 = 14,
/* {A,X}RGB channel ordering */
VG_sXRGB_8888 = 0 | (1 << 6),
@@ -448,6 +461,12 @@ typedef enum {
VG_BLEND_MODE_FORCE_SIZE = VG_MAX_ENUM
} VGBlendMode;
typedef enum {
VG_FONT_NUM_GLYPHS = 0x2F00,
VG_FONT_PARAM_TYPE_FORCE_SIZE = VG_MAX_ENUM
} VGFontParamType;
typedef enum {
VG_IMAGE_FORMAT_QUERY = 0x2100,
VG_PATH_DATATYPE_QUERY = 0x2101,
@@ -541,8 +560,22 @@ VG_API_CALL void VG_API_ENTRY vgShear(VGfloat shx, VGfloat shy) VG_API_EXIT;
VG_API_CALL void VG_API_ENTRY vgRotate(VGfloat angle) VG_API_EXIT;
/* Masking and Clearing */
VG_API_CALL void VG_API_ENTRY vgMask(VGImage mask, VGMaskOperation operation,
VGint x, VGint y, VGint width, VGint height) VG_API_EXIT;
VG_API_CALL void VG_API_ENTRY vgMask(VGHandle mask, VGMaskOperation operation,
VGint x, VGint y,
VGint width, VGint height) VG_API_EXIT;
VG_API_CALL void VG_API_ENTRY vgRenderToMask(VGPath path,
VGbitfield paintModes,
VGMaskOperation operation) VG_API_EXIT;
VG_API_CALL VGMaskLayer VG_API_ENTRY vgCreateMaskLayer(VGint width, VGint height) VG_API_EXIT;
VG_API_CALL void VG_API_ENTRY vgDestroyMaskLayer(VGMaskLayer maskLayer) VG_API_EXIT;
VG_API_CALL void VG_API_ENTRY vgFillMaskLayer(VGMaskLayer maskLayer,
VGint x, VGint y,
VGint width, VGint height,
VGfloat value) VG_API_EXIT;
VG_API_CALL void VG_API_ENTRY vgCopyMask(VGMaskLayer maskLayer,
VGint dx, VGint dy,
VGint sx, VGint sy,
VGint width, VGint height) VG_API_EXIT;
VG_API_CALL void VG_API_ENTRY vgClear(VGint x, VGint y, VGint width, VGint height) VG_API_EXIT;
/* Paths */
@@ -636,6 +669,33 @@ VG_API_CALL void VG_API_ENTRY vgCopyPixels(VGint dx, VGint dy,
VGint sx, VGint sy,
VGint width, VGint height) VG_API_EXIT;
/* Text */
VG_API_CALL VGFont VG_API_ENTRY vgCreateFont(VGint glyphCapacityHint) VG_API_EXIT;
VG_API_CALL void VG_API_ENTRY vgDestroyFont(VGFont font) VG_API_EXIT;
VG_API_CALL void VG_API_ENTRY vgSetGlyphToPath(VGFont font,
VGuint glyphIndex,
VGPath path,
VGboolean isHinted,
const VGfloat glyphOrigin [2],
const VGfloat escapement[2]) VG_API_EXIT;
VG_API_CALL void VG_API_ENTRY vgSetGlyphToImage(VGFont font,
VGuint glyphIndex,
VGImage image,
const VGfloat glyphOrigin [2],
const VGfloat escapement[2]) VG_API_EXIT;
VG_API_CALL void VG_API_ENTRY vgClearGlyph(VGFont font,VGuint glyphIndex) VG_API_EXIT;
VG_API_CALL void VG_API_ENTRY vgDrawGlyph(VGFont font,
VGuint glyphIndex,
VGbitfield paintModes,
VGboolean allowAutoHinting) VG_API_EXIT;
VG_API_CALL void VG_API_ENTRY vgDrawGlyphs(VGFont font,
VGint glyphCount,
const VGuint *glyphIndices,
const VGfloat *adjustments_x,
const VGfloat *adjustments_y,
VGbitfield paintModes,
VGboolean allowAutoHinting) VG_API_EXIT;
/* Image Filters */
VG_API_CALL void VG_API_ENTRY vgColorMatrix(VGImage dst, VGImage src,
const VGfloat * matrix) VG_API_EXIT;

View File

@@ -1,233 +1,233 @@
/* $Revision: 6810 $ on $Date:: 2008-10-29 10:31:37 -0400 #$ */
/*------------------------------------------------------------------------
*
* VG extensions Reference Implementation
* -------------------------------------
*
* Copyright (c) 2008 The Khronos Group Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and /or associated documentation files
* (the "Materials "), to deal in the Materials without restriction,
* including without limitation the rights to use, copy, modify, merge,
* publish, distribute, sublicense, and/or sell copies of the Materials,
* and to permit persons to whom the Materials are furnished to do so,
* subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Materials.
*
* THE MATERIALS ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,
* DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
* OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE MATERIALS OR
* THE USE OR OTHER DEALINGS IN THE MATERIALS.
*
*//**
* \file
* \brief VG extensions
*//*-------------------------------------------------------------------*/
#ifndef _VGEXT_H
#define _VGEXT_H
#ifdef __cplusplus
extern "C" {
#endif
#include <VG/openvg.h>
#include <VG/vgu.h>
#ifndef VG_API_ENTRYP
# define VG_API_ENTRYP VG_API_ENTRY*
#endif
#ifndef VGU_API_ENTRYP
# define VGU_API_ENTRYP VGU_API_ENTRY*
#endif
/*-------------------------------------------------------------------------------
* KHR extensions
*------------------------------------------------------------------------------*/
typedef enum {
#ifndef VG_KHR_iterative_average_blur
VG_MAX_AVERAGE_BLUR_DIMENSION_KHR = 0x116B,
VG_AVERAGE_BLUR_DIMENSION_RESOLUTION_KHR = 0x116C,
VG_MAX_AVERAGE_BLUR_ITERATIONS_KHR = 0x116D,
#endif
VG_PARAM_TYPE_KHR_FORCE_SIZE = VG_MAX_ENUM
} VGParamTypeKHR;
#ifndef VG_KHR_EGL_image
#define VG_KHR_EGL_image 1
/* VGEGLImageKHR is an opaque handle to an EGLImage */
typedef void* VGeglImageKHR;
#ifdef VG_VGEXT_PROTOTYPES
VG_API_CALL VGImage VG_API_ENTRY vgCreateEGLImageTargetKHR(VGeglImageKHR image);
#endif
typedef VGImage (VG_API_ENTRYP PFNVGCREATEEGLIMAGETARGETKHRPROC) (VGeglImageKHR image);
#endif
#ifndef VG_KHR_iterative_average_blur
#define VG_KHR_iterative_average_blur 1
#ifdef VG_VGEXT_PROTOTYPES
VG_API_CALL void vgIterativeAverageBlurKHR(VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGTilingMode tilingMode);
#endif
typedef void (VG_API_ENTRYP PFNVGITERATIVEAVERAGEBLURKHRPROC) (VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGTilingMode tilingMode);
#endif
#ifndef VG_KHR_advanced_blending
#define VG_KHR_advanced_blending 1
typedef enum {
VG_BLEND_OVERLAY_KHR = 0x2010,
VG_BLEND_HARDLIGHT_KHR = 0x2011,
VG_BLEND_SOFTLIGHT_SVG_KHR = 0x2012,
VG_BLEND_SOFTLIGHT_KHR = 0x2013,
VG_BLEND_COLORDODGE_KHR = 0x2014,
VG_BLEND_COLORBURN_KHR = 0x2015,
VG_BLEND_DIFFERENCE_KHR = 0x2016,
VG_BLEND_SUBTRACT_KHR = 0x2017,
VG_BLEND_INVERT_KHR = 0x2018,
VG_BLEND_EXCLUSION_KHR = 0x2019,
VG_BLEND_LINEARDODGE_KHR = 0x201a,
VG_BLEND_LINEARBURN_KHR = 0x201b,
VG_BLEND_VIVIDLIGHT_KHR = 0x201c,
VG_BLEND_LINEARLIGHT_KHR = 0x201d,
VG_BLEND_PINLIGHT_KHR = 0x201e,
VG_BLEND_HARDMIX_KHR = 0x201f,
VG_BLEND_CLEAR_KHR = 0x2020,
VG_BLEND_DST_KHR = 0x2021,
VG_BLEND_SRC_OUT_KHR = 0x2022,
VG_BLEND_DST_OUT_KHR = 0x2023,
VG_BLEND_SRC_ATOP_KHR = 0x2024,
VG_BLEND_DST_ATOP_KHR = 0x2025,
VG_BLEND_XOR_KHR = 0x2026,
VG_BLEND_MODE_KHR_FORCE_SIZE= VG_MAX_ENUM
} VGBlendModeKHR;
#endif
#ifndef VG_KHR_parametric_filter
#define VG_KHR_parametric_filter 1
typedef enum {
VG_PF_OBJECT_VISIBLE_FLAG_KHR = (1 << 0),
VG_PF_KNOCKOUT_FLAG_KHR = (1 << 1),
VG_PF_OUTER_FLAG_KHR = (1 << 2),
VG_PF_INNER_FLAG_KHR = (1 << 3),
VG_PF_TYPE_KHR_FORCE_SIZE = VG_MAX_ENUM
} VGPfTypeKHR;
typedef enum {
VGU_IMAGE_IN_USE_ERROR = 0xF010,
VGU_ERROR_CODE_KHR_FORCE_SIZE = VG_MAX_ENUM
} VGUErrorCodeKHR;
#ifdef VG_VGEXT_PROTOTYPES
VG_API_CALL void VG_API_ENTRY vgParametricFilterKHR(VGImage dst,VGImage src,VGImage blur,VGfloat strength,VGfloat offsetX,VGfloat offsetY,VGbitfield filterFlags,VGPaint highlightPaint,VGPaint shadowPaint);
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguDropShadowKHR(VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGfloat distance,VGfloat angle,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint shadowColorRGBA);
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguGlowKHR(VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint glowColorRGBA) ;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguBevelKHR(VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGfloat distance,VGfloat angle,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint highlightColorRGBA,VGuint shadowColorRGBA);
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguGradientGlowKHR(VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGfloat distance,VGfloat angle,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint stopsCount,const VGfloat* glowColorRampStops);
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguGradientBevelKHR(VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGfloat distance,VGfloat angle,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint stopsCount,const VGfloat* bevelColorRampStops);
#endif
typedef void (VG_API_ENTRYP PFNVGPARAMETRICFILTERKHRPROC) (VGImage dst,VGImage src,VGImage blur,VGfloat strength,VGfloat offsetX,VGfloat offsetY,VGbitfield filterFlags,VGPaint highlightPaint,VGPaint shadowPaint);
typedef VGUErrorCode (VGU_API_ENTRYP PFNVGUDROPSHADOWKHRPROC) (VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGfloat distance,VGfloat angle,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint shadowColorRGBA);
typedef VGUErrorCode (VGU_API_ENTRYP PFNVGUGLOWKHRPROC) (VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint glowColorRGBA);
typedef VGUErrorCode (VGU_API_ENTRYP PFNVGUBEVELKHRPROC) (VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGfloat distance,VGfloat angle,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint highlightColorRGBA,VGuint shadowColorRGBA);
typedef VGUErrorCode (VGU_API_ENTRYP PFNVGUGRADIENTGLOWKHRPROC) (VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGfloat distance,VGfloat angle,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint stopsCount,const VGfloat* glowColorRampStops);
typedef VGUErrorCode (VGU_API_ENTRYP PFNVGUGRADIENTBEVELKHRPROC) (VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGfloat distance,VGfloat angle,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint stopsCount,const VGfloat* bevelColorRampStops);
#endif
/*-------------------------------------------------------------------------------
* NDS extensions
*------------------------------------------------------------------------------*/
#ifndef VG_NDS_paint_generation
#define VG_NDS_paint_generation 1
typedef enum {
VG_PAINT_COLOR_RAMP_LINEAR_NDS = 0x1A10,
VG_COLOR_MATRIX_NDS = 0x1A11,
VG_PAINT_COLOR_TRANSFORM_LINEAR_NDS = 0x1A12,
VG_PAINT_PARAM_TYPE_NDS_FORCE_SIZE = VG_MAX_ENUM
} VGPaintParamTypeNds;
typedef enum {
VG_DRAW_IMAGE_COLOR_MATRIX_NDS = 0x1F10,
VG_IMAGE_MODE_NDS_FORCE_SIZE = VG_MAX_ENUM
} VGImageModeNds;
#endif
#ifndef VG_NDS_projective_geometry
#define VG_NDS_projective_geometry 1
typedef enum {
VG_CLIP_MODE_NDS = 0x1180,
VG_CLIP_LINES_NDS = 0x1181,
VG_MAX_CLIP_LINES_NDS = 0x1182,
VG_PARAM_TYPE_NDS_FORCE_SIZE = VG_MAX_ENUM
} VGParamTypeNds;
typedef enum {
VG_CLIPMODE_NONE_NDS = 0x3000,
VG_CLIPMODE_CLIP_CLOSED_NDS = 0x3001,
VG_CLIPMODE_CLIP_OPEN_NDS = 0x3002,
VG_CLIPMODE_CULL_NDS = 0x3003,
VG_CLIPMODE_NDS_FORCE_SIZE = VG_MAX_ENUM
} VGClipModeNds;
typedef enum {
VG_RQUAD_TO_NDS = ( 13 << 1 ),
VG_RCUBIC_TO_NDS = ( 14 << 1 ),
VG_PATH_SEGMENT_NDS_FORCE_SIZE = VG_MAX_ENUM
} VGPathSegmentNds;
typedef enum {
VG_RQUAD_TO_ABS_NDS = (VG_RQUAD_TO_NDS | VG_ABSOLUTE),
VG_RQUAD_TO_REL_NDS = (VG_RQUAD_TO_NDS | VG_RELATIVE),
VG_RCUBIC_TO_ABS_NDS = (VG_RCUBIC_TO_NDS | VG_ABSOLUTE),
VG_RCUBIC_TO_REL_NDS = (VG_RCUBIC_TO_NDS | VG_RELATIVE),
VG_PATH_COMMAND_NDS_FORCE_SIZE = VG_MAX_ENUM
} VGPathCommandNds;
#ifdef VG_VGEXT_PROTOTYPES
VG_API_CALL void VG_API_ENTRY vgProjectiveMatrixNDS(VGboolean enable) ;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguTransformClipLineNDS(const VGfloat Ain,const VGfloat Bin,const VGfloat Cin,const VGfloat* matrix,const VGboolean inverse,VGfloat* Aout,VGfloat* Bout,VGfloat* Cout);
#endif
typedef void (VG_API_ENTRYP PFNVGPROJECTIVEMATRIXNDSPROC) (VGboolean enable) ;
typedef VGUErrorCode (VGU_API_ENTRYP PFNVGUTRANSFORMCLIPLINENDSPROC) (const VGfloat Ain,const VGfloat Bin,const VGfloat Cin,const VGfloat* matrix,const VGboolean inverse,VGfloat* Aout,VGfloat* Bout,VGfloat* Cout);
#endif
#ifdef __cplusplus
} /* extern "C" */
#endif
#endif /* _VGEXT_H */
/* $Revision: 6810 $ on $Date:: 2008-10-29 07:31:37 -0700 #$ */
/*------------------------------------------------------------------------
*
* VG extensions Reference Implementation
* -------------------------------------
*
* Copyright (c) 2008 The Khronos Group Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and /or associated documentation files
* (the "Materials "), to deal in the Materials without restriction,
* including without limitation the rights to use, copy, modify, merge,
* publish, distribute, sublicense, and/or sell copies of the Materials,
* and to permit persons to whom the Materials are furnished to do so,
* subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Materials.
*
* THE MATERIALS ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,
* DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
* OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE MATERIALS OR
* THE USE OR OTHER DEALINGS IN THE MATERIALS.
*
*//**
* \file
* \brief VG extensions
*//*-------------------------------------------------------------------*/
#ifndef _VGEXT_H
#define _VGEXT_H
#ifdef __cplusplus
extern "C" {
#endif
#include <VG/openvg.h>
#include <VG/vgu.h>
#ifndef VG_API_ENTRYP
# define VG_API_ENTRYP VG_API_ENTRY*
#endif
#ifndef VGU_API_ENTRYP
# define VGU_API_ENTRYP VGU_API_ENTRY*
#endif
/*-------------------------------------------------------------------------------
* KHR extensions
*------------------------------------------------------------------------------*/
typedef enum {
#ifndef VG_KHR_iterative_average_blur
VG_MAX_AVERAGE_BLUR_DIMENSION_KHR = 0x116B,
VG_AVERAGE_BLUR_DIMENSION_RESOLUTION_KHR = 0x116C,
VG_MAX_AVERAGE_BLUR_ITERATIONS_KHR = 0x116D,
#endif
VG_PARAM_TYPE_KHR_FORCE_SIZE = VG_MAX_ENUM
} VGParamTypeKHR;
#ifndef VG_KHR_EGL_image
#define VG_KHR_EGL_image 1
/* VGEGLImageKHR is an opaque handle to an EGLImage */
typedef void* VGeglImageKHR;
#ifdef VG_VGEXT_PROTOTYPES
VG_API_CALL VGImage VG_API_ENTRY vgCreateEGLImageTargetKHR(VGeglImageKHR image);
#endif
typedef VGImage (VG_API_ENTRYP PFNVGCREATEEGLIMAGETARGETKHRPROC) (VGeglImageKHR image);
#endif
#ifndef VG_KHR_iterative_average_blur
#define VG_KHR_iterative_average_blur 1
#ifdef VG_VGEXT_PROTOTYPES
VG_API_CALL void vgIterativeAverageBlurKHR(VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGTilingMode tilingMode);
#endif
typedef void (VG_API_ENTRYP PFNVGITERATIVEAVERAGEBLURKHRPROC) (VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGTilingMode tilingMode);
#endif
#ifndef VG_KHR_advanced_blending
#define VG_KHR_advanced_blending 1
typedef enum {
VG_BLEND_OVERLAY_KHR = 0x2010,
VG_BLEND_HARDLIGHT_KHR = 0x2011,
VG_BLEND_SOFTLIGHT_SVG_KHR = 0x2012,
VG_BLEND_SOFTLIGHT_KHR = 0x2013,
VG_BLEND_COLORDODGE_KHR = 0x2014,
VG_BLEND_COLORBURN_KHR = 0x2015,
VG_BLEND_DIFFERENCE_KHR = 0x2016,
VG_BLEND_SUBTRACT_KHR = 0x2017,
VG_BLEND_INVERT_KHR = 0x2018,
VG_BLEND_EXCLUSION_KHR = 0x2019,
VG_BLEND_LINEARDODGE_KHR = 0x201a,
VG_BLEND_LINEARBURN_KHR = 0x201b,
VG_BLEND_VIVIDLIGHT_KHR = 0x201c,
VG_BLEND_LINEARLIGHT_KHR = 0x201d,
VG_BLEND_PINLIGHT_KHR = 0x201e,
VG_BLEND_HARDMIX_KHR = 0x201f,
VG_BLEND_CLEAR_KHR = 0x2020,
VG_BLEND_DST_KHR = 0x2021,
VG_BLEND_SRC_OUT_KHR = 0x2022,
VG_BLEND_DST_OUT_KHR = 0x2023,
VG_BLEND_SRC_ATOP_KHR = 0x2024,
VG_BLEND_DST_ATOP_KHR = 0x2025,
VG_BLEND_XOR_KHR = 0x2026,
VG_BLEND_MODE_KHR_FORCE_SIZE= VG_MAX_ENUM
} VGBlendModeKHR;
#endif
#ifndef VG_KHR_parametric_filter
#define VG_KHR_parametric_filter 1
typedef enum {
VG_PF_OBJECT_VISIBLE_FLAG_KHR = (1 << 0),
VG_PF_KNOCKOUT_FLAG_KHR = (1 << 1),
VG_PF_OUTER_FLAG_KHR = (1 << 2),
VG_PF_INNER_FLAG_KHR = (1 << 3),
VG_PF_TYPE_KHR_FORCE_SIZE = VG_MAX_ENUM
} VGPfTypeKHR;
typedef enum {
VGU_IMAGE_IN_USE_ERROR = 0xF010,
VGU_ERROR_CODE_KHR_FORCE_SIZE = VG_MAX_ENUM
} VGUErrorCodeKHR;
#ifdef VG_VGEXT_PROTOTYPES
VG_API_CALL void VG_API_ENTRY vgParametricFilterKHR(VGImage dst,VGImage src,VGImage blur,VGfloat strength,VGfloat offsetX,VGfloat offsetY,VGbitfield filterFlags,VGPaint highlightPaint,VGPaint shadowPaint);
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguDropShadowKHR(VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGfloat distance,VGfloat angle,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint shadowColorRGBA);
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguGlowKHR(VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint glowColorRGBA) ;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguBevelKHR(VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGfloat distance,VGfloat angle,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint highlightColorRGBA,VGuint shadowColorRGBA);
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguGradientGlowKHR(VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGfloat distance,VGfloat angle,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint stopsCount,const VGfloat* glowColorRampStops);
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguGradientBevelKHR(VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGfloat distance,VGfloat angle,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint stopsCount,const VGfloat* bevelColorRampStops);
#endif
typedef void (VG_API_ENTRYP PFNVGPARAMETRICFILTERKHRPROC) (VGImage dst,VGImage src,VGImage blur,VGfloat strength,VGfloat offsetX,VGfloat offsetY,VGbitfield filterFlags,VGPaint highlightPaint,VGPaint shadowPaint);
typedef VGUErrorCode (VGU_API_ENTRYP PFNVGUDROPSHADOWKHRPROC) (VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGfloat distance,VGfloat angle,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint shadowColorRGBA);
typedef VGUErrorCode (VGU_API_ENTRYP PFNVGUGLOWKHRPROC) (VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint glowColorRGBA);
typedef VGUErrorCode (VGU_API_ENTRYP PFNVGUBEVELKHRPROC) (VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGfloat distance,VGfloat angle,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint highlightColorRGBA,VGuint shadowColorRGBA);
typedef VGUErrorCode (VGU_API_ENTRYP PFNVGUGRADIENTGLOWKHRPROC) (VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGfloat distance,VGfloat angle,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint stopsCount,const VGfloat* glowColorRampStops);
typedef VGUErrorCode (VGU_API_ENTRYP PFNVGUGRADIENTBEVELKHRPROC) (VGImage dst,VGImage src,VGfloat dimX,VGfloat dimY,VGuint iterative,VGfloat strength,VGfloat distance,VGfloat angle,VGbitfield filterFlags,VGbitfield allowedQuality,VGuint stopsCount,const VGfloat* bevelColorRampStops);
#endif
/*-------------------------------------------------------------------------------
* NDS extensions
*------------------------------------------------------------------------------*/
#ifndef VG_NDS_paint_generation
#define VG_NDS_paint_generation 1
typedef enum {
VG_PAINT_COLOR_RAMP_LINEAR_NDS = 0x1A10,
VG_COLOR_MATRIX_NDS = 0x1A11,
VG_PAINT_COLOR_TRANSFORM_LINEAR_NDS = 0x1A12,
VG_PAINT_PARAM_TYPE_NDS_FORCE_SIZE = VG_MAX_ENUM
} VGPaintParamTypeNds;
typedef enum {
VG_DRAW_IMAGE_COLOR_MATRIX_NDS = 0x1F10,
VG_IMAGE_MODE_NDS_FORCE_SIZE = VG_MAX_ENUM
} VGImageModeNds;
#endif
#ifndef VG_NDS_projective_geometry
#define VG_NDS_projective_geometry 1
typedef enum {
VG_CLIP_MODE_NDS = 0x1180,
VG_CLIP_LINES_NDS = 0x1181,
VG_MAX_CLIP_LINES_NDS = 0x1182,
VG_PARAM_TYPE_NDS_FORCE_SIZE = VG_MAX_ENUM
} VGParamTypeNds;
typedef enum {
VG_CLIPMODE_NONE_NDS = 0x3000,
VG_CLIPMODE_CLIP_CLOSED_NDS = 0x3001,
VG_CLIPMODE_CLIP_OPEN_NDS = 0x3002,
VG_CLIPMODE_CULL_NDS = 0x3003,
VG_CLIPMODE_NDS_FORCE_SIZE = VG_MAX_ENUM
} VGClipModeNds;
typedef enum {
VG_RQUAD_TO_NDS = ( 13 << 1 ),
VG_RCUBIC_TO_NDS = ( 14 << 1 ),
VG_PATH_SEGMENT_NDS_FORCE_SIZE = VG_MAX_ENUM
} VGPathSegmentNds;
typedef enum {
VG_RQUAD_TO_ABS_NDS = (VG_RQUAD_TO_NDS | VG_ABSOLUTE),
VG_RQUAD_TO_REL_NDS = (VG_RQUAD_TO_NDS | VG_RELATIVE),
VG_RCUBIC_TO_ABS_NDS = (VG_RCUBIC_TO_NDS | VG_ABSOLUTE),
VG_RCUBIC_TO_REL_NDS = (VG_RCUBIC_TO_NDS | VG_RELATIVE),
VG_PATH_COMMAND_NDS_FORCE_SIZE = VG_MAX_ENUM
} VGPathCommandNds;
#ifdef VG_VGEXT_PROTOTYPES
VG_API_CALL void VG_API_ENTRY vgProjectiveMatrixNDS(VGboolean enable) ;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguTransformClipLineNDS(const VGfloat Ain,const VGfloat Bin,const VGfloat Cin,const VGfloat* matrix,const VGboolean inverse,VGfloat* Aout,VGfloat* Bout,VGfloat* Cout);
#endif
typedef void (VG_API_ENTRYP PFNVGPROJECTIVEMATRIXNDSPROC) (VGboolean enable) ;
typedef VGUErrorCode (VGU_API_ENTRYP PFNVGUTRANSFORMCLIPLINENDSPROC) (const VGfloat Ain,const VGfloat Bin,const VGfloat Cin,const VGfloat* matrix,const VGboolean inverse,VGfloat* Aout,VGfloat* Bout,VGfloat* Cout);
#endif
#ifdef __cplusplus
} /* extern "C" */
#endif
#endif /* _VGEXT_H */

View File

@@ -1,92 +1,92 @@
/* $Revision: 6810 $ on $Date:: 2008-10-29 10:31:37 -0400 #$ */
/*------------------------------------------------------------------------
*
* VG platform specific header Reference Implementation
* ----------------------------------------------------
*
* Copyright (c) 2008 The Khronos Group Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and /or associated documentation files
* (the "Materials "), to deal in the Materials without restriction,
* including without limitation the rights to use, copy, modify, merge,
* publish, distribute, sublicense, and/or sell copies of the Materials,
* and to permit persons to whom the Materials are furnished to do so,
* subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Materials.
*
* THE MATERIALS ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,
* DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
* OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE MATERIALS OR
* THE USE OR OTHER DEALINGS IN THE MATERIALS.
*
*//**
* \file
* \brief VG platform specific header
*//*-------------------------------------------------------------------*/
#ifndef _VGPLATFORM_H
#define _VGPLATFORM_H
#include <KHR/khrplatform.h>
#ifdef __cplusplus
extern "C" {
#endif
#ifndef VG_API_CALL
#if defined(OPENVG_STATIC_LIBRARY)
# define VG_API_CALL
#else
# define VG_API_CALL KHRONOS_APICALL
#endif /* defined OPENVG_STATIC_LIBRARY */
#endif /* ifndef VG_API_CALL */
#ifndef VGU_API_CALL
#if defined(OPENVG_STATIC_LIBRARY)
# define VGU_API_CALL
#else
# define VGU_API_CALL KHRONOS_APICALL
#endif /* defined OPENVG_STATIC_LIBRARY */
#endif /* ifndef VGU_API_CALL */
#ifndef VG_API_ENTRY
#define VG_API_ENTRY
#endif
#ifndef VG_API_EXIT
#define VG_API_EXIT
#endif
#ifndef VGU_API_ENTRY
#define VGU_API_ENTRY
#endif
#ifndef VGU_API_EXIT
#define VGU_API_EXIT
#endif
typedef float VGfloat;
typedef signed char VGbyte;
typedef unsigned char VGubyte;
typedef signed short VGshort;
typedef signed int VGint;
typedef unsigned int VGuint;
typedef unsigned int VGbitfield;
#ifndef VG_VGEXT_PROTOTYPES
#define VG_VGEXT_PROTOTYPES
#endif
#ifdef __cplusplus
} /* extern "C" */
#endif
#endif /* _VGPLATFORM_H */
/* $Revision: 6810 $ on $Date:: 2008-10-29 07:31:37 -0700 #$ */
/*------------------------------------------------------------------------
*
* VG platform specific header Reference Implementation
* ----------------------------------------------------
*
* Copyright (c) 2008 The Khronos Group Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and /or associated documentation files
* (the "Materials "), to deal in the Materials without restriction,
* including without limitation the rights to use, copy, modify, merge,
* publish, distribute, sublicense, and/or sell copies of the Materials,
* and to permit persons to whom the Materials are furnished to do so,
* subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Materials.
*
* THE MATERIALS ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,
* DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
* OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE MATERIALS OR
* THE USE OR OTHER DEALINGS IN THE MATERIALS.
*
*//**
* \file
* \brief VG platform specific header
*//*-------------------------------------------------------------------*/
#ifndef _VGPLATFORM_H
#define _VGPLATFORM_H
#include <KHR/khrplatform.h>
#ifdef __cplusplus
extern "C" {
#endif
#ifndef VG_API_CALL
#if defined(OPENVG_STATIC_LIBRARY)
# define VG_API_CALL
#else
# define VG_API_CALL KHRONOS_APICALL
#endif /* defined OPENVG_STATIC_LIBRARY */
#endif /* ifndef VG_API_CALL */
#ifndef VGU_API_CALL
#if defined(OPENVG_STATIC_LIBRARY)
# define VGU_API_CALL
#else
# define VGU_API_CALL KHRONOS_APICALL
#endif /* defined OPENVG_STATIC_LIBRARY */
#endif /* ifndef VGU_API_CALL */
#ifndef VG_API_ENTRY
#define VG_API_ENTRY
#endif
#ifndef VG_API_EXIT
#define VG_API_EXIT
#endif
#ifndef VGU_API_ENTRY
#define VGU_API_ENTRY
#endif
#ifndef VGU_API_EXIT
#define VGU_API_EXIT
#endif
typedef float VGfloat;
typedef signed char VGbyte;
typedef unsigned char VGubyte;
typedef signed short VGshort;
typedef signed int VGint;
typedef unsigned int VGuint;
typedef unsigned int VGbitfield;
#ifndef VG_VGEXT_PROTOTYPES
#define VG_VGEXT_PROTOTYPES
#endif
#ifdef __cplusplus
} /* extern "C" */
#endif
#endif /* _VGPLATFORM_H */

View File

@@ -1,130 +1,131 @@
/* $Revision: 6810 $ on $Date:: 2008-10-29 10:31:37 -0400 #$ */
/*------------------------------------------------------------------------
*
* VGU 1.0.1 Reference Implementation
* -------------------------------------
*
* Copyright (c) 2008 The Khronos Group Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and /or associated documentation files
* (the "Materials "), to deal in the Materials without restriction,
* including without limitation the rights to use, copy, modify, merge,
* publish, distribute, sublicense, and/or sell copies of the Materials,
* and to permit persons to whom the Materials are furnished to do so,
* subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Materials.
*
* THE MATERIALS ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,
* DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
* OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE MATERIALS OR
* THE USE OR OTHER DEALINGS IN THE MATERIALS.
*
*//**
* \file
* \brief VGU 1.0.1 API.
*//*-------------------------------------------------------------------*/
#ifndef _VGU_H
#define _VGU_H
#ifdef __cplusplus
extern "C" {
#endif
#include <VG/openvg.h>
#define VGU_VERSION_1_0 1
#ifndef VGU_API_CALL
# error VGU_API_CALL must be defined
#endif
#ifndef VGU_API_ENTRY
# error VGU_API_ENTRY must be defined
#endif
#ifndef VGU_API_EXIT
# error VGU_API_EXIT must be defined
#endif
typedef enum {
VGU_NO_ERROR = 0,
VGU_BAD_HANDLE_ERROR = 0xF000,
VGU_ILLEGAL_ARGUMENT_ERROR = 0xF001,
VGU_OUT_OF_MEMORY_ERROR = 0xF002,
VGU_PATH_CAPABILITY_ERROR = 0xF003,
VGU_BAD_WARP_ERROR = 0xF004,
VGU_ERROR_CODE_FORCE_SIZE = VG_MAX_ENUM
} VGUErrorCode;
typedef enum {
VGU_ARC_OPEN = 0xF100,
VGU_ARC_CHORD = 0xF101,
VGU_ARC_PIE = 0xF102,
VGU_ARC_TYPE_FORCE_SIZE = VG_MAX_ENUM
} VGUArcType;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguLine(VGPath path,
VGfloat x0, VGfloat y0,
VGfloat x1, VGfloat y1) VGU_API_EXIT;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguPolygon(VGPath path,
const VGfloat * points, VGint count,
VGboolean closed) VGU_API_EXIT;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguRect(VGPath path,
VGfloat x, VGfloat y,
VGfloat width, VGfloat height) VGU_API_EXIT;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguRoundRect(VGPath path,
VGfloat x, VGfloat y,
VGfloat width, VGfloat height,
VGfloat arcWidth, VGfloat arcHeight) VGU_API_EXIT;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguEllipse(VGPath path,
VGfloat cx, VGfloat cy,
VGfloat width, VGfloat height) VGU_API_EXIT;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguArc(VGPath path,
VGfloat x, VGfloat y,
VGfloat width, VGfloat height,
VGfloat startAngle, VGfloat angleExtent,
VGUArcType arcType) VGU_API_EXIT;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguComputeWarpQuadToSquare(VGfloat sx0, VGfloat sy0,
VGfloat sx1, VGfloat sy1,
VGfloat sx2, VGfloat sy2,
VGfloat sx3, VGfloat sy3,
VGfloat * matrix) VGU_API_EXIT;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguComputeWarpSquareToQuad(VGfloat dx0, VGfloat dy0,
VGfloat dx1, VGfloat dy1,
VGfloat dx2, VGfloat dy2,
VGfloat dx3, VGfloat dy3,
VGfloat * matrix) VGU_API_EXIT;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguComputeWarpQuadToQuad(VGfloat dx0, VGfloat dy0,
VGfloat dx1, VGfloat dy1,
VGfloat dx2, VGfloat dy2,
VGfloat dx3, VGfloat dy3,
VGfloat sx0, VGfloat sy0,
VGfloat sx1, VGfloat sy1,
VGfloat sx2, VGfloat sy2,
VGfloat sx3, VGfloat sy3,
VGfloat * matrix) VGU_API_EXIT;
#ifdef __cplusplus
} /* extern "C" */
#endif
#endif /* #ifndef _VGU_H */
/* $Revision: 6810 $ on $Date:: 2008-10-29 07:31:37 -0700 #$ */
/*------------------------------------------------------------------------
*
* VGU 1.1 Reference Implementation
* -------------------------------------
*
* Copyright (c) 2008 The Khronos Group Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and /or associated documentation files
* (the "Materials "), to deal in the Materials without restriction,
* including without limitation the rights to use, copy, modify, merge,
* publish, distribute, sublicense, and/or sell copies of the Materials,
* and to permit persons to whom the Materials are furnished to do so,
* subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Materials.
*
* THE MATERIALS ARE PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,
* DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
* OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE MATERIALS OR
* THE USE OR OTHER DEALINGS IN THE MATERIALS.
*
*//**
* \file
* \brief VGU 1.1 API.
*//*-------------------------------------------------------------------*/
#ifndef _VGU_H
#define _VGU_H
#ifdef __cplusplus
extern "C" {
#endif
#include <VG/openvg.h>
#define VGU_VERSION_1_0 1
#define VGU_VERSION_1_1 2
#ifndef VGU_API_CALL
# error VGU_API_CALL must be defined
#endif
#ifndef VGU_API_ENTRY
# error VGU_API_ENTRY must be defined
#endif
#ifndef VGU_API_EXIT
# error VGU_API_EXIT must be defined
#endif
typedef enum {
VGU_NO_ERROR = 0,
VGU_BAD_HANDLE_ERROR = 0xF000,
VGU_ILLEGAL_ARGUMENT_ERROR = 0xF001,
VGU_OUT_OF_MEMORY_ERROR = 0xF002,
VGU_PATH_CAPABILITY_ERROR = 0xF003,
VGU_BAD_WARP_ERROR = 0xF004,
VGU_ERROR_CODE_FORCE_SIZE = VG_MAX_ENUM
} VGUErrorCode;
typedef enum {
VGU_ARC_OPEN = 0xF100,
VGU_ARC_CHORD = 0xF101,
VGU_ARC_PIE = 0xF102,
VGU_ARC_TYPE_FORCE_SIZE = VG_MAX_ENUM
} VGUArcType;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguLine(VGPath path,
VGfloat x0, VGfloat y0,
VGfloat x1, VGfloat y1) VGU_API_EXIT;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguPolygon(VGPath path,
const VGfloat * points, VGint count,
VGboolean closed) VGU_API_EXIT;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguRect(VGPath path,
VGfloat x, VGfloat y,
VGfloat width, VGfloat height) VGU_API_EXIT;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguRoundRect(VGPath path,
VGfloat x, VGfloat y,
VGfloat width, VGfloat height,
VGfloat arcWidth, VGfloat arcHeight) VGU_API_EXIT;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguEllipse(VGPath path,
VGfloat cx, VGfloat cy,
VGfloat width, VGfloat height) VGU_API_EXIT;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguArc(VGPath path,
VGfloat x, VGfloat y,
VGfloat width, VGfloat height,
VGfloat startAngle, VGfloat angleExtent,
VGUArcType arcType) VGU_API_EXIT;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguComputeWarpQuadToSquare(VGfloat sx0, VGfloat sy0,
VGfloat sx1, VGfloat sy1,
VGfloat sx2, VGfloat sy2,
VGfloat sx3, VGfloat sy3,
VGfloat * matrix) VGU_API_EXIT;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguComputeWarpSquareToQuad(VGfloat dx0, VGfloat dy0,
VGfloat dx1, VGfloat dy1,
VGfloat dx2, VGfloat dy2,
VGfloat dx3, VGfloat dy3,
VGfloat * matrix) VGU_API_EXIT;
VGU_API_CALL VGUErrorCode VGU_API_ENTRY vguComputeWarpQuadToQuad(VGfloat dx0, VGfloat dy0,
VGfloat dx1, VGfloat dy1,
VGfloat dx2, VGfloat dy2,
VGfloat dx3, VGfloat dy3,
VGfloat sx0, VGfloat sy0,
VGfloat sx1, VGfloat sy1,
VGfloat sx2, VGfloat sy2,
VGfloat sx3, VGfloat sy3,
VGfloat * matrix) VGU_API_EXIT;
#ifdef __cplusplus
} /* extern "C" */
#endif
#endif /* #ifndef _VGU_H */

View File

@@ -4,7 +4,6 @@ SConscript('mapi/vgapi/SConscript')
if env['platform'] == 'windows':
SConscript('egl/main/SConscript')
SConscript('talloc/SConscript')
SConscript('glsl/SConscript')
SConscript('mapi/glapi/SConscript')

View File

@@ -750,7 +750,7 @@ dri2_create_screen(_EGLDisplay *disp)
if (dri2_dpy->dri2->base.version >= 2)
api_mask = dri2_dpy->dri2->getAPIMask(dri2_dpy->dri_screen);
else
api_mask = __DRI_API_OPENGL;
api_mask = 1 << __DRI_API_OPENGL;
disp->ClientAPIsMask = 0;
if (api_mask & (1 <<__DRI_API_OPENGL))
@@ -899,10 +899,20 @@ const int i915_chip_ids[] = {
0x29b2, /* PCI_CHIP_Q35_G */
0x29c2, /* PCI_CHIP_G33_G */
0x29d2, /* PCI_CHIP_Q33_G */
0xa001, /* PCI_CHIP_IGD_G */
0xa011, /* Pineview */
};
const int i965_chip_ids[] = {
0x0042, /* PCI_CHIP_ILD_G */
0x0046, /* PCI_CHIP_ILM_G */
0x0102, /* PCI_CHIP_SANDYBRIDGE_GT1 */
0x0106, /* PCI_CHIP_SANDYBRIDGE_M_GT1 */
0x010a, /* PCI_CHIP_SANDYBRIDGE_S */
0x0112, /* PCI_CHIP_SANDYBRIDGE_GT2 */
0x0116, /* PCI_CHIP_SANDYBRIDGE_M_GT2 */
0x0122, /* PCI_CHIP_SANDYBRIDGE_GT2_PLUS */
0x0126, /* PCI_CHIP_SANDYBRIDGE_M_GT2_PLUS */
0x29a2, /* PCI_CHIP_I965_G */
0x2992, /* PCI_CHIP_I965_Q */
0x2982, /* PCI_CHIP_I965_G_1 */
@@ -914,11 +924,440 @@ const int i965_chip_ids[] = {
0x2e12, /* PCI_CHIP_Q45_G */
0x2e22, /* PCI_CHIP_G45_G */
0x2e32, /* PCI_CHIP_G41_G */
0x2e42, /* PCI_CHIP_B43_G */
0x2e92, /* PCI_CHIP_B43_G1 */
};
const int r100_chip_ids[] = {
0x4C57, /* PCI_CHIP_RADEON_LW */
0x4C58, /* PCI_CHIP_RADEON_LX */
0x4C59, /* PCI_CHIP_RADEON_LY */
0x4C5A, /* PCI_CHIP_RADEON_LZ */
0x5144, /* PCI_CHIP_RADEON_QD */
0x5145, /* PCI_CHIP_RADEON_QE */
0x5146, /* PCI_CHIP_RADEON_QF */
0x5147, /* PCI_CHIP_RADEON_QG */
0x5159, /* PCI_CHIP_RADEON_QY */
0x515A, /* PCI_CHIP_RADEON_QZ */
0x5157, /* PCI_CHIP_RV200_QW */
0x5158, /* PCI_CHIP_RV200_QX */
0x515E, /* PCI_CHIP_RN50_515E */
0x5969, /* PCI_CHIP_RN50_5969 */
0x4136, /* PCI_CHIP_RS100_4136 */
0x4336, /* PCI_CHIP_RS100_4336 */
0x4137, /* PCI_CHIP_RS200_4137 */
0x4337, /* PCI_CHIP_RS200_4337 */
0x4237, /* PCI_CHIP_RS250_4237 */
0x4437, /* PCI_CHIP_RS250_4437 */
};
const int r200_chip_ids[] = {
0x5148, /* PCI_CHIP_R200_QH */
0x514C, /* PCI_CHIP_R200_QL */
0x514D, /* PCI_CHIP_R200_QM */
0x4242, /* PCI_CHIP_R200_BB */
0x4243, /* PCI_CHIP_R200_BC */
0x4966, /* PCI_CHIP_RV250_If */
0x4967, /* PCI_CHIP_RV250_Ig */
0x4C64, /* PCI_CHIP_RV250_Ld */
0x4C66, /* PCI_CHIP_RV250_Lf */
0x4C67, /* PCI_CHIP_RV250_Lg */
0x5960, /* PCI_CHIP_RV280_5960 */
0x5961, /* PCI_CHIP_RV280_5961 */
0x5962, /* PCI_CHIP_RV280_5962 */
0x5964, /* PCI_CHIP_RV280_5964 */
0x5965, /* PCI_CHIP_RV280_5965 */
0x5C61, /* PCI_CHIP_RV280_5C61 */
0x5C63, /* PCI_CHIP_RV280_5C63 */
0x5834, /* PCI_CHIP_RS300_5834 */
0x5835, /* PCI_CHIP_RS300_5835 */
0x7834, /* PCI_CHIP_RS350_7834 */
0x7835, /* PCI_CHIP_RS350_7835 */
};
const int r300_chip_ids[] = {
0x4144, /* PCI_CHIP_R300_AD */
0x4145, /* PCI_CHIP_R300_AE */
0x4146, /* PCI_CHIP_R300_AF */
0x4147, /* PCI_CHIP_R300_AG */
0x4E44, /* PCI_CHIP_R300_ND */
0x4E45, /* PCI_CHIP_R300_NE */
0x4E46, /* PCI_CHIP_R300_NF */
0x4E47, /* PCI_CHIP_R300_NG */
0x4E48, /* PCI_CHIP_R350_NH */
0x4E49, /* PCI_CHIP_R350_NI */
0x4E4B, /* PCI_CHIP_R350_NK */
0x4148, /* PCI_CHIP_R350_AH */
0x4149, /* PCI_CHIP_R350_AI */
0x414A, /* PCI_CHIP_R350_AJ */
0x414B, /* PCI_CHIP_R350_AK */
0x4E4A, /* PCI_CHIP_R360_NJ */
0x4150, /* PCI_CHIP_RV350_AP */
0x4151, /* PCI_CHIP_RV350_AQ */
0x4152, /* PCI_CHIP_RV350_AR */
0x4153, /* PCI_CHIP_RV350_AS */
0x4154, /* PCI_CHIP_RV350_AT */
0x4155, /* PCI_CHIP_RV350_AU */
0x4156, /* PCI_CHIP_RV350_AV */
0x4E50, /* PCI_CHIP_RV350_NP */
0x4E51, /* PCI_CHIP_RV350_NQ */
0x4E52, /* PCI_CHIP_RV350_NR */
0x4E53, /* PCI_CHIP_RV350_NS */
0x4E54, /* PCI_CHIP_RV350_NT */
0x4E56, /* PCI_CHIP_RV350_NV */
0x5460, /* PCI_CHIP_RV370_5460 */
0x5462, /* PCI_CHIP_RV370_5462 */
0x5464, /* PCI_CHIP_RV370_5464 */
0x5B60, /* PCI_CHIP_RV370_5B60 */
0x5B62, /* PCI_CHIP_RV370_5B62 */
0x5B63, /* PCI_CHIP_RV370_5B63 */
0x5B64, /* PCI_CHIP_RV370_5B64 */
0x5B65, /* PCI_CHIP_RV370_5B65 */
0x3150, /* PCI_CHIP_RV380_3150 */
0x3152, /* PCI_CHIP_RV380_3152 */
0x3154, /* PCI_CHIP_RV380_3154 */
0x3155, /* PCI_CHIP_RV380_3155 */
0x3E50, /* PCI_CHIP_RV380_3E50 */
0x3E54, /* PCI_CHIP_RV380_3E54 */
0x4A48, /* PCI_CHIP_R420_JH */
0x4A49, /* PCI_CHIP_R420_JI */
0x4A4A, /* PCI_CHIP_R420_JJ */
0x4A4B, /* PCI_CHIP_R420_JK */
0x4A4C, /* PCI_CHIP_R420_JL */
0x4A4D, /* PCI_CHIP_R420_JM */
0x4A4E, /* PCI_CHIP_R420_JN */
0x4A4F, /* PCI_CHIP_R420_JO */
0x4A50, /* PCI_CHIP_R420_JP */
0x4A54, /* PCI_CHIP_R420_JT */
0x5548, /* PCI_CHIP_R423_UH */
0x5549, /* PCI_CHIP_R423_UI */
0x554A, /* PCI_CHIP_R423_UJ */
0x554B, /* PCI_CHIP_R423_UK */
0x5550, /* PCI_CHIP_R423_5550 */
0x5551, /* PCI_CHIP_R423_UQ */
0x5552, /* PCI_CHIP_R423_UR */
0x5554, /* PCI_CHIP_R423_UT */
0x5D57, /* PCI_CHIP_R423_5D57 */
0x554C, /* PCI_CHIP_R430_554C */
0x554D, /* PCI_CHIP_R430_554D */
0x554E, /* PCI_CHIP_R430_554E */
0x554F, /* PCI_CHIP_R430_554F */
0x5D48, /* PCI_CHIP_R430_5D48 */
0x5D49, /* PCI_CHIP_R430_5D49 */
0x5D4A, /* PCI_CHIP_R430_5D4A */
0x5D4C, /* PCI_CHIP_R480_5D4C */
0x5D4D, /* PCI_CHIP_R480_5D4D */
0x5D4E, /* PCI_CHIP_R480_5D4E */
0x5D4F, /* PCI_CHIP_R480_5D4F */
0x5D50, /* PCI_CHIP_R480_5D50 */
0x5D52, /* PCI_CHIP_R480_5D52 */
0x4B49, /* PCI_CHIP_R481_4B49 */
0x4B4A, /* PCI_CHIP_R481_4B4A */
0x4B4B, /* PCI_CHIP_R481_4B4B */
0x4B4C, /* PCI_CHIP_R481_4B4C */
0x564A, /* PCI_CHIP_RV410_564A */
0x564B, /* PCI_CHIP_RV410_564B */
0x564F, /* PCI_CHIP_RV410_564F */
0x5652, /* PCI_CHIP_RV410_5652 */
0x5653, /* PCI_CHIP_RV410_5653 */
0x5657, /* PCI_CHIP_RV410_5657 */
0x5E48, /* PCI_CHIP_RV410_5E48 */
0x5E4A, /* PCI_CHIP_RV410_5E4A */
0x5E4B, /* PCI_CHIP_RV410_5E4B */
0x5E4C, /* PCI_CHIP_RV410_5E4C */
0x5E4D, /* PCI_CHIP_RV410_5E4D */
0x5E4F, /* PCI_CHIP_RV410_5E4F */
0x5A41, /* PCI_CHIP_RS400_5A41 */
0x5A42, /* PCI_CHIP_RS400_5A42 */
0x5A61, /* PCI_CHIP_RC410_5A61 */
0x5A62, /* PCI_CHIP_RC410_5A62 */
0x5954, /* PCI_CHIP_RS480_5954 */
0x5955, /* PCI_CHIP_RS480_5955 */
0x5974, /* PCI_CHIP_RS482_5974 */
0x5975, /* PCI_CHIP_RS482_5975 */
0x7100, /* PCI_CHIP_R520_7100 */
0x7101, /* PCI_CHIP_R520_7101 */
0x7102, /* PCI_CHIP_R520_7102 */
0x7103, /* PCI_CHIP_R520_7103 */
0x7104, /* PCI_CHIP_R520_7104 */
0x7105, /* PCI_CHIP_R520_7105 */
0x7106, /* PCI_CHIP_R520_7106 */
0x7108, /* PCI_CHIP_R520_7108 */
0x7109, /* PCI_CHIP_R520_7109 */
0x710A, /* PCI_CHIP_R520_710A */
0x710B, /* PCI_CHIP_R520_710B */
0x710C, /* PCI_CHIP_R520_710C */
0x710E, /* PCI_CHIP_R520_710E */
0x710F, /* PCI_CHIP_R520_710F */
0x7140, /* PCI_CHIP_RV515_7140 */
0x7141, /* PCI_CHIP_RV515_7141 */
0x7142, /* PCI_CHIP_RV515_7142 */
0x7143, /* PCI_CHIP_RV515_7143 */
0x7144, /* PCI_CHIP_RV515_7144 */
0x7145, /* PCI_CHIP_RV515_7145 */
0x7146, /* PCI_CHIP_RV515_7146 */
0x7147, /* PCI_CHIP_RV515_7147 */
0x7149, /* PCI_CHIP_RV515_7149 */
0x714A, /* PCI_CHIP_RV515_714A */
0x714B, /* PCI_CHIP_RV515_714B */
0x714C, /* PCI_CHIP_RV515_714C */
0x714D, /* PCI_CHIP_RV515_714D */
0x714E, /* PCI_CHIP_RV515_714E */
0x714F, /* PCI_CHIP_RV515_714F */
0x7151, /* PCI_CHIP_RV515_7151 */
0x7152, /* PCI_CHIP_RV515_7152 */
0x7153, /* PCI_CHIP_RV515_7153 */
0x715E, /* PCI_CHIP_RV515_715E */
0x715F, /* PCI_CHIP_RV515_715F */
0x7180, /* PCI_CHIP_RV515_7180 */
0x7181, /* PCI_CHIP_RV515_7181 */
0x7183, /* PCI_CHIP_RV515_7183 */
0x7186, /* PCI_CHIP_RV515_7186 */
0x7187, /* PCI_CHIP_RV515_7187 */
0x7188, /* PCI_CHIP_RV515_7188 */
0x718A, /* PCI_CHIP_RV515_718A */
0x718B, /* PCI_CHIP_RV515_718B */
0x718C, /* PCI_CHIP_RV515_718C */
0x718D, /* PCI_CHIP_RV515_718D */
0x718F, /* PCI_CHIP_RV515_718F */
0x7193, /* PCI_CHIP_RV515_7193 */
0x7196, /* PCI_CHIP_RV515_7196 */
0x719B, /* PCI_CHIP_RV515_719B */
0x719F, /* PCI_CHIP_RV515_719F */
0x7200, /* PCI_CHIP_RV515_7200 */
0x7210, /* PCI_CHIP_RV515_7210 */
0x7211, /* PCI_CHIP_RV515_7211 */
0x71C0, /* PCI_CHIP_RV530_71C0 */
0x71C1, /* PCI_CHIP_RV530_71C1 */
0x71C2, /* PCI_CHIP_RV530_71C2 */
0x71C3, /* PCI_CHIP_RV530_71C3 */
0x71C4, /* PCI_CHIP_RV530_71C4 */
0x71C5, /* PCI_CHIP_RV530_71C5 */
0x71C6, /* PCI_CHIP_RV530_71C6 */
0x71C7, /* PCI_CHIP_RV530_71C7 */
0x71CD, /* PCI_CHIP_RV530_71CD */
0x71CE, /* PCI_CHIP_RV530_71CE */
0x71D2, /* PCI_CHIP_RV530_71D2 */
0x71D4, /* PCI_CHIP_RV530_71D4 */
0x71D5, /* PCI_CHIP_RV530_71D5 */
0x71D6, /* PCI_CHIP_RV530_71D6 */
0x71DA, /* PCI_CHIP_RV530_71DA */
0x71DE, /* PCI_CHIP_RV530_71DE */
0x7281, /* PCI_CHIP_RV560_7281 */
0x7283, /* PCI_CHIP_RV560_7283 */
0x7287, /* PCI_CHIP_RV560_7287 */
0x7290, /* PCI_CHIP_RV560_7290 */
0x7291, /* PCI_CHIP_RV560_7291 */
0x7293, /* PCI_CHIP_RV560_7293 */
0x7297, /* PCI_CHIP_RV560_7297 */
0x7280, /* PCI_CHIP_RV570_7280 */
0x7288, /* PCI_CHIP_RV570_7288 */
0x7289, /* PCI_CHIP_RV570_7289 */
0x728B, /* PCI_CHIP_RV570_728B */
0x728C, /* PCI_CHIP_RV570_728C */
0x7240, /* PCI_CHIP_R580_7240 */
0x7243, /* PCI_CHIP_R580_7243 */
0x7244, /* PCI_CHIP_R580_7244 */
0x7245, /* PCI_CHIP_R580_7245 */
0x7246, /* PCI_CHIP_R580_7246 */
0x7247, /* PCI_CHIP_R580_7247 */
0x7248, /* PCI_CHIP_R580_7248 */
0x7249, /* PCI_CHIP_R580_7249 */
0x724A, /* PCI_CHIP_R580_724A */
0x724B, /* PCI_CHIP_R580_724B */
0x724C, /* PCI_CHIP_R580_724C */
0x724D, /* PCI_CHIP_R580_724D */
0x724E, /* PCI_CHIP_R580_724E */
0x724F, /* PCI_CHIP_R580_724F */
0x7284, /* PCI_CHIP_R580_7284 */
0x793F, /* PCI_CHIP_RS600_793F */
0x7941, /* PCI_CHIP_RS600_7941 */
0x7942, /* PCI_CHIP_RS600_7942 */
0x791E, /* PCI_CHIP_RS690_791E */
0x791F, /* PCI_CHIP_RS690_791F */
0x796C, /* PCI_CHIP_RS740_796C */
0x796D, /* PCI_CHIP_RS740_796D */
0x796E, /* PCI_CHIP_RS740_796E */
0x796F, /* PCI_CHIP_RS740_796F */
};
const int r600_chip_ids[] = {
0x9400, /* PCI_CHIP_R600_9400 */
0x9401, /* PCI_CHIP_R600_9401 */
0x9402, /* PCI_CHIP_R600_9402 */
0x9403, /* PCI_CHIP_R600_9403 */
0x9405, /* PCI_CHIP_R600_9405 */
0x940A, /* PCI_CHIP_R600_940A */
0x940B, /* PCI_CHIP_R600_940B */
0x940F, /* PCI_CHIP_R600_940F */
0x94C0, /* PCI_CHIP_RV610_94C0 */
0x94C1, /* PCI_CHIP_RV610_94C1 */
0x94C3, /* PCI_CHIP_RV610_94C3 */
0x94C4, /* PCI_CHIP_RV610_94C4 */
0x94C5, /* PCI_CHIP_RV610_94C5 */
0x94C6, /* PCI_CHIP_RV610_94C6 */
0x94C7, /* PCI_CHIP_RV610_94C7 */
0x94C8, /* PCI_CHIP_RV610_94C8 */
0x94C9, /* PCI_CHIP_RV610_94C9 */
0x94CB, /* PCI_CHIP_RV610_94CB */
0x94CC, /* PCI_CHIP_RV610_94CC */
0x94CD, /* PCI_CHIP_RV610_94CD */
0x9580, /* PCI_CHIP_RV630_9580 */
0x9581, /* PCI_CHIP_RV630_9581 */
0x9583, /* PCI_CHIP_RV630_9583 */
0x9586, /* PCI_CHIP_RV630_9586 */
0x9587, /* PCI_CHIP_RV630_9587 */
0x9588, /* PCI_CHIP_RV630_9588 */
0x9589, /* PCI_CHIP_RV630_9589 */
0x958A, /* PCI_CHIP_RV630_958A */
0x958B, /* PCI_CHIP_RV630_958B */
0x958C, /* PCI_CHIP_RV630_958C */
0x958D, /* PCI_CHIP_RV630_958D */
0x958E, /* PCI_CHIP_RV630_958E */
0x958F, /* PCI_CHIP_RV630_958F */
0x9500, /* PCI_CHIP_RV670_9500 */
0x9501, /* PCI_CHIP_RV670_9501 */
0x9504, /* PCI_CHIP_RV670_9504 */
0x9505, /* PCI_CHIP_RV670_9505 */
0x9506, /* PCI_CHIP_RV670_9506 */
0x9507, /* PCI_CHIP_RV670_9507 */
0x9508, /* PCI_CHIP_RV670_9508 */
0x9509, /* PCI_CHIP_RV670_9509 */
0x950F, /* PCI_CHIP_RV670_950F */
0x9511, /* PCI_CHIP_RV670_9511 */
0x9515, /* PCI_CHIP_RV670_9515 */
0x9517, /* PCI_CHIP_RV670_9517 */
0x9519, /* PCI_CHIP_RV670_9519 */
0x95C0, /* PCI_CHIP_RV620_95C0 */
0x95C2, /* PCI_CHIP_RV620_95C2 */
0x95C4, /* PCI_CHIP_RV620_95C4 */
0x95C5, /* PCI_CHIP_RV620_95C5 */
0x95C6, /* PCI_CHIP_RV620_95C6 */
0x95C7, /* PCI_CHIP_RV620_95C7 */
0x95C9, /* PCI_CHIP_RV620_95C9 */
0x95CC, /* PCI_CHIP_RV620_95CC */
0x95CD, /* PCI_CHIP_RV620_95CD */
0x95CE, /* PCI_CHIP_RV620_95CE */
0x95CF, /* PCI_CHIP_RV620_95CF */
0x9590, /* PCI_CHIP_RV635_9590 */
0x9591, /* PCI_CHIP_RV635_9591 */
0x9593, /* PCI_CHIP_RV635_9593 */
0x9595, /* PCI_CHIP_RV635_9595 */
0x9596, /* PCI_CHIP_RV635_9596 */
0x9597, /* PCI_CHIP_RV635_9597 */
0x9598, /* PCI_CHIP_RV635_9598 */
0x9599, /* PCI_CHIP_RV635_9599 */
0x959B, /* PCI_CHIP_RV635_959B */
0x9610, /* PCI_CHIP_RS780_9610 */
0x9611, /* PCI_CHIP_RS780_9611 */
0x9612, /* PCI_CHIP_RS780_9612 */
0x9613, /* PCI_CHIP_RS780_9613 */
0x9614, /* PCI_CHIP_RS780_9614 */
0x9615, /* PCI_CHIP_RS780_9615 */
0x9616, /* PCI_CHIP_RS780_9616 */
0x9710, /* PCI_CHIP_RS880_9710 */
0x9711, /* PCI_CHIP_RS880_9711 */
0x9712, /* PCI_CHIP_RS880_9712 */
0x9713, /* PCI_CHIP_RS880_9713 */
0x9714, /* PCI_CHIP_RS880_9714 */
0x9715, /* PCI_CHIP_RS880_9715 */
0x9440, /* PCI_CHIP_RV770_9440 */
0x9441, /* PCI_CHIP_RV770_9441 */
0x9442, /* PCI_CHIP_RV770_9442 */
0x9443, /* PCI_CHIP_RV770_9443 */
0x9444, /* PCI_CHIP_RV770_9444 */
0x9446, /* PCI_CHIP_RV770_9446 */
0x944A, /* PCI_CHIP_RV770_944A */
0x944B, /* PCI_CHIP_RV770_944B */
0x944C, /* PCI_CHIP_RV770_944C */
0x944E, /* PCI_CHIP_RV770_944E */
0x9450, /* PCI_CHIP_RV770_9450 */
0x9452, /* PCI_CHIP_RV770_9452 */
0x9456, /* PCI_CHIP_RV770_9456 */
0x945A, /* PCI_CHIP_RV770_945A */
0x945B, /* PCI_CHIP_RV770_945B */
0x945E, /* PCI_CHIP_RV770_945E */
0x9460, /* PCI_CHIP_RV790_9460 */
0x9462, /* PCI_CHIP_RV790_9462 */
0x946A, /* PCI_CHIP_RV770_946A */
0x946B, /* PCI_CHIP_RV770_946B */
0x947A, /* PCI_CHIP_RV770_947A */
0x947B, /* PCI_CHIP_RV770_947B */
0x9480, /* PCI_CHIP_RV730_9480 */
0x9487, /* PCI_CHIP_RV730_9487 */
0x9488, /* PCI_CHIP_RV730_9488 */
0x9489, /* PCI_CHIP_RV730_9489 */
0x948A, /* PCI_CHIP_RV730_948A */
0x948F, /* PCI_CHIP_RV730_948F */
0x9490, /* PCI_CHIP_RV730_9490 */
0x9491, /* PCI_CHIP_RV730_9491 */
0x9495, /* PCI_CHIP_RV730_9495 */
0x9498, /* PCI_CHIP_RV730_9498 */
0x949C, /* PCI_CHIP_RV730_949C */
0x949E, /* PCI_CHIP_RV730_949E */
0x949F, /* PCI_CHIP_RV730_949F */
0x9540, /* PCI_CHIP_RV710_9540 */
0x9541, /* PCI_CHIP_RV710_9541 */
0x9542, /* PCI_CHIP_RV710_9542 */
0x954E, /* PCI_CHIP_RV710_954E */
0x954F, /* PCI_CHIP_RV710_954F */
0x9552, /* PCI_CHIP_RV710_9552 */
0x9553, /* PCI_CHIP_RV710_9553 */
0x9555, /* PCI_CHIP_RV710_9555 */
0x9557, /* PCI_CHIP_RV710_9557 */
0x955F, /* PCI_CHIP_RV710_955F */
0x94A0, /* PCI_CHIP_RV740_94A0 */
0x94A1, /* PCI_CHIP_RV740_94A1 */
0x94A3, /* PCI_CHIP_RV740_94A3 */
0x94B1, /* PCI_CHIP_RV740_94B1 */
0x94B3, /* PCI_CHIP_RV740_94B3 */
0x94B4, /* PCI_CHIP_RV740_94B4 */
0x94B5, /* PCI_CHIP_RV740_94B5 */
0x94B9, /* PCI_CHIP_RV740_94B9 */
0x68E0, /* PCI_CHIP_CEDAR_68E0 */
0x68E1, /* PCI_CHIP_CEDAR_68E1 */
0x68E4, /* PCI_CHIP_CEDAR_68E4 */
0x68E5, /* PCI_CHIP_CEDAR_68E5 */
0x68E8, /* PCI_CHIP_CEDAR_68E8 */
0x68E9, /* PCI_CHIP_CEDAR_68E9 */
0x68F1, /* PCI_CHIP_CEDAR_68F1 */
0x68F8, /* PCI_CHIP_CEDAR_68F8 */
0x68F9, /* PCI_CHIP_CEDAR_68F9 */
0x68FE, /* PCI_CHIP_CEDAR_68FE */
0x68C0, /* PCI_CHIP_REDWOOD_68C0 */
0x68C1, /* PCI_CHIP_REDWOOD_68C1 */
0x68C8, /* PCI_CHIP_REDWOOD_68C8 */
0x68C9, /* PCI_CHIP_REDWOOD_68C9 */
0x68D8, /* PCI_CHIP_REDWOOD_68D8 */
0x68D9, /* PCI_CHIP_REDWOOD_68D9 */
0x68DA, /* PCI_CHIP_REDWOOD_68DA */
0x68DE, /* PCI_CHIP_REDWOOD_68DE */
0x68A0, /* PCI_CHIP_JUNIPER_68A0 */
0x68A1, /* PCI_CHIP_JUNIPER_68A1 */
0x68A8, /* PCI_CHIP_JUNIPER_68A8 */
0x68A9, /* PCI_CHIP_JUNIPER_68A9 */
0x68B0, /* PCI_CHIP_JUNIPER_68B0 */
0x68B8, /* PCI_CHIP_JUNIPER_68B8 */
0x68B9, /* PCI_CHIP_JUNIPER_68B9 */
0x68BE, /* PCI_CHIP_JUNIPER_68BE */
0x6880, /* PCI_CHIP_CYPRESS_6880 */
0x6888, /* PCI_CHIP_CYPRESS_6888 */
0x6889, /* PCI_CHIP_CYPRESS_6889 */
0x688A, /* PCI_CHIP_CYPRESS_688A */
0x6898, /* PCI_CHIP_CYPRESS_6898 */
0x6899, /* PCI_CHIP_CYPRESS_6899 */
0x689E, /* PCI_CHIP_CYPRESS_689E */
0x689C, /* PCI_CHIP_HEMLOCK_689C */
0x689D, /* PCI_CHIP_HEMLOCK_689D */
};
const struct dri2_driver_map driver_map[] = {
{ 0x8086, "i915", i915_chip_ids, ARRAY_SIZE(i915_chip_ids) },
{ 0x8086, "i965", i965_chip_ids, ARRAY_SIZE(i965_chip_ids) },
{ 0x1002, "radeon", r100_chip_ids, ARRAY_SIZE(r100_chip_ids) },
{ 0x1002, "r200", r200_chip_ids, ARRAY_SIZE(r200_chip_ids) },
{ 0x1002, "r300", r300_chip_ids, ARRAY_SIZE(r300_chip_ids) },
{ 0x1002, "r600", r600_chip_ids, ARRAY_SIZE(r600_chip_ids) },
};
static char *
@@ -1209,7 +1648,11 @@ dri2_make_current(_EGLDriver *drv, _EGLDisplay *disp, _EGLSurface *dsurf,
dri2_destroy_surface(drv, disp, old_dsurf);
dri2_destroy_surface(drv, disp, old_rsurf);
if (old_ctx) {
dri2_dpy->core->unbindContext(dri2_egl_context(old_ctx)->dri_context);
/* unbind the old context only when there is no new context bound */
if (!ctx) {
__DRIcontext *old_cctx = dri2_egl_context(old_ctx)->dri_context;
dri2_dpy->core->unbindContext(old_cctx);
}
/* no destroy? */
_eglPutContext(old_ctx);
}

View File

@@ -118,38 +118,39 @@ _eglFindArray(_EGLArray *array, void *elem)
/**
* Filter an array and return the filtered data. The returned data pointer
* should be freed.
* Filter an array and return the number of filtered elements.
*/
void **
_eglFilterArray(_EGLArray *array, EGLint *size,
EGLint
_eglFilterArray(_EGLArray *array, void **data, EGLint size,
_EGLArrayForEach filter, void *filter_data)
{
void **data;
EGLint count = 0, i;
if (!array) {
*size = 0;
return malloc(0);
}
data = malloc(array->Size * sizeof(array->Elements[0]));
if (!data)
return NULL;
if (!array)
return 0;
if (filter) {
for (i = 0; i < array->Size; i++) {
if (filter(array->Elements[i], filter_data))
data[count++] = array->Elements[i];
if (filter(array->Elements[i], filter_data)) {
if (data && count < size)
data[count] = array->Elements[i];
count++;
}
if (data && count >= size)
break;
}
}
else {
memcpy(data, array->Elements, array->Size * sizeof(array->Elements[0]));
if (data) {
count = (size < array->Size) ? size : array->Size;
memcpy(data, array->Elements, count * sizeof(array->Elements[0]));
}
else {
count = array->Size;
}
}
*size = count;
return data;
return count;
}

View File

@@ -37,8 +37,8 @@ void *
_eglFindArray(_EGLArray *array, void *elem);
PUBLIC void **
_eglFilterArray(_EGLArray *array, EGLint *size,
PUBLIC EGLint
_eglFilterArray(_EGLArray *array, void **data, EGLint size,
_EGLArrayForEach filter, void *filter_data);

View File

@@ -697,11 +697,22 @@ _eglChooseConfig(_EGLDriver *drv, _EGLDisplay *disp, const EGLint *attrib_list,
if (!_eglParseConfigAttribList(&criteria, disp, attrib_list))
return _eglError(EGL_BAD_ATTRIBUTE, "eglChooseConfig");
configList = (_EGLConfig **) _eglFilterArray(disp->Configs, &count,
/* get the number of matched configs */
count = _eglFilterArray(disp->Configs, NULL, 0,
(_EGLArrayForEach) _eglMatchConfig, (void *) &criteria);
if (!count) {
*num_configs = count;
return EGL_TRUE;
}
configList = malloc(sizeof(*configList) * count);
if (!configList)
return _eglError(EGL_BAD_ALLOC, "eglChooseConfig(out of memory)");
/* get the matched configs */
_eglFilterArray(disp->Configs, (void **) configList, count,
(_EGLArrayForEach) _eglMatchConfig, (void *) &criteria);
/* perform sorting of configs */
if (configs && count) {
_eglSortConfigs((const _EGLConfig **) configList, count,

View File

@@ -4,7 +4,7 @@
#include "egltypedefs.h"
#include "eglapi.h"
#include <stddef.h>
/**
* Define an inline driver typecast function.

View File

@@ -52,7 +52,6 @@ if env['drm']:
# Needed by some state trackers
SConscript('winsys/sw/null/SConscript')
SConscript('state_trackers/python/SConscript')
if env['platform'] != 'embedded':
SConscript('state_trackers/vega/SConscript')
@@ -116,7 +115,7 @@ if env['dri']:
'targets/dri-r600/SConscript',
])
if env['xorg']:
if env['xorg'] and env['drm']:
SConscript([
#'targets/xorg-i915/SConscript',
#'targets/xorg-i965/SConscript',

View File

@@ -128,12 +128,12 @@ C_SOURCES = \
util/u_linkage.c \
util/u_network.c \
util/u_math.c \
util/u_mempool.c \
util/u_mm.c \
util/u_rect.c \
util/u_ringbuffer.c \
util/u_sampler.c \
util/u_simple_shaders.c \
util/u_slab.c \
util/u_snprintf.c \
util/u_staging.c \
util/u_surface.c \
@@ -142,8 +142,7 @@ C_SOURCES = \
util/u_tile.c \
util/u_transfer.c \
util/u_resource.c \
util/u_upload_mgr.c \
target-helpers/wrap_screen.c
util/u_upload_mgr.c
# Disabling until pipe-video branch gets merged in
#vl/vl_bitstream_parser.c \
@@ -186,7 +185,7 @@ GALLIVM_SOURCES = \
draw/draw_pt_fetch_shade_pipeline_llvm.c
GALLIVM_CPP_SOURCES = \
gallivm/lp_bld_misc.cpp
gallivm/lp_bld_misc.cpp
GENERATED_SOURCES = \
indices/u_indices_gen.c \
@@ -204,9 +203,6 @@ CPP_SOURCES += \
endif
LIBRARY_DEFINES += -D__STDC_CONSTANT_MACROS
include ../Makefile.template

View File

@@ -175,13 +175,13 @@ source = [
'util/u_linkage.c',
'util/u_network.c',
'util/u_math.c',
'util/u_mempool.c',
'util/u_mm.c',
'util/u_rect.c',
'util/u_resource.c',
'util/u_ringbuffer.c',
'util/u_sampler.c',
'util/u_simple_shaders.c',
'util/u_slab.c',
'util/u_snprintf.c',
'util/u_staging.c',
'util/u_surface.c',
@@ -196,7 +196,6 @@ source = [
#'vl/vl_compositor.c',
#'vl/vl_csc.c',
#'vl/vl_shader_build.c',
'target-helpers/wrap_screen.c',
]
if env['llvm']:

View File

@@ -595,7 +595,7 @@ enum pipe_error cso_set_vertex_samplers(struct cso_context *ctx,
error = temp;
}
for ( ; i < ctx->nr_samplers; i++) {
for ( ; i < ctx->nr_vertex_samplers; i++) {
temp = cso_single_vertex_sampler( ctx, i, NULL );
if (temp != PIPE_OK)
error = temp;

View File

@@ -63,19 +63,32 @@ draw_get_option_use_llvm(void)
}
#endif
struct draw_context *draw_create( struct pipe_context *pipe )
/**
* Create new draw module context.
*/
struct draw_context *
draw_create(struct pipe_context *pipe)
{
return draw_create_gallivm(pipe, NULL);
}
/**
* Create new draw module context with gallivm state for LLVM JIT.
*/
struct draw_context *
draw_create_gallivm(struct pipe_context *pipe, struct gallivm_state *gallivm)
{
struct draw_context *draw = CALLOC_STRUCT( draw_context );
if (draw == NULL)
goto fail;
#if HAVE_LLVM
if(draw_get_option_use_llvm())
{
lp_build_init();
assert(lp_build_engine);
draw->engine = lp_build_engine;
draw->llvm = draw_llvm_create(draw);
if (draw_get_option_use_llvm() && gallivm) {
draw->llvm = draw_llvm_create(draw, gallivm);
}
#endif
@@ -91,6 +104,8 @@ fail:
return NULL;
}
boolean draw_init(struct draw_context *draw)
{
/*

View File

@@ -48,10 +48,15 @@ struct draw_vertex_shader;
struct draw_geometry_shader;
struct draw_fragment_shader;
struct tgsi_sampler;
struct gallivm_state;
struct draw_context *draw_create( struct pipe_context *pipe );
struct draw_context *
draw_create_gallivm(struct pipe_context *pipe, struct gallivm_state *gallivm);
void draw_destroy( struct draw_context *draw );
void draw_flush(struct draw_context *draw);

File diff suppressed because it is too large Load Diff

View File

@@ -103,41 +103,41 @@ struct draw_jit_context
};
#define draw_jit_context_vs_constants(_builder, _ptr) \
lp_build_struct_get(_builder, _ptr, 0, "vs_constants")
#define draw_jit_context_vs_constants(_gallivm, _ptr) \
lp_build_struct_get(_gallivm, _ptr, 0, "vs_constants")
#define draw_jit_context_gs_constants(_builder, _ptr) \
lp_build_struct_get(_builder, _ptr, 1, "gs_constants")
#define draw_jit_context_gs_constants(_gallivm, _ptr) \
lp_build_struct_get(_gallivm, _ptr, 1, "gs_constants")
#define draw_jit_context_planes(_builder, _ptr) \
lp_build_struct_get(_builder, _ptr, 2, "planes")
#define draw_jit_context_planes(_gallivm, _ptr) \
lp_build_struct_get(_gallivm, _ptr, 2, "planes")
#define draw_jit_context_viewport(_builder, _ptr) \
lp_build_struct_get(_builder, _ptr, 3, "viewport")
#define draw_jit_context_viewport(_gallivm, _ptr) \
lp_build_struct_get(_gallivm, _ptr, 3, "viewport")
#define DRAW_JIT_CTX_TEXTURES 4
#define draw_jit_context_textures(_builder, _ptr) \
lp_build_struct_get_ptr(_builder, _ptr, DRAW_JIT_CTX_TEXTURES, "textures")
#define draw_jit_context_textures(_gallivm, _ptr) \
lp_build_struct_get_ptr(_gallivm, _ptr, DRAW_JIT_CTX_TEXTURES, "textures")
#define draw_jit_header_id(_builder, _ptr) \
lp_build_struct_get_ptr(_builder, _ptr, 0, "id")
#define draw_jit_header_id(_gallivm, _ptr) \
lp_build_struct_get_ptr(_gallivm, _ptr, 0, "id")
#define draw_jit_header_clip(_builder, _ptr) \
lp_build_struct_get_ptr(_builder, _ptr, 1, "clip")
#define draw_jit_header_clip(_gallivm, _ptr) \
lp_build_struct_get_ptr(_gallivm, _ptr, 1, "clip")
#define draw_jit_header_data(_builder, _ptr) \
lp_build_struct_get_ptr(_builder, _ptr, 2, "data")
#define draw_jit_header_data(_gallivm, _ptr) \
lp_build_struct_get_ptr(_gallivm, _ptr, 2, "data")
#define draw_jit_vbuffer_stride(_builder, _ptr) \
lp_build_struct_get(_builder, _ptr, 0, "stride")
#define draw_jit_vbuffer_stride(_gallivm, _ptr) \
lp_build_struct_get(_gallivm, _ptr, 0, "stride")
#define draw_jit_vbuffer_max_index(_builder, _ptr) \
lp_build_struct_get(_builder, _ptr, 1, "max_index")
#define draw_jit_vbuffer_max_index(_gallivm, _ptr) \
lp_build_struct_get(_gallivm, _ptr, 1, "max_index")
#define draw_jit_vbuffer_offset(_builder, _ptr) \
lp_build_struct_get(_builder, _ptr, 2, "buffer_offset")
#define draw_jit_vbuffer_offset(_gallivm, _ptr) \
lp_build_struct_get(_gallivm, _ptr, 2, "buffer_offset")
typedef int
@@ -246,21 +246,19 @@ struct draw_llvm {
struct draw_jit_context jit_context;
struct gallivm_state *gallivm;
struct draw_llvm_variant_list_item vs_variants_list;
int nr_variants;
LLVMModuleRef module;
LLVMExecutionEngineRef engine;
LLVMModuleProviderRef provider;
LLVMTargetDataRef target;
LLVMPassManagerRef pass;
/* LLVM JIT builder types */
LLVMTypeRef context_ptr_type;
LLVMTypeRef vertex_header_ptr_type;
LLVMTypeRef buffer_ptr_type;
LLVMTypeRef vb_ptr_type;
LLVMTypeRef vertex_header_ptr_type;
};
static INLINE struct llvm_vertex_shader *
llvm_vertex_shader(struct draw_vertex_shader *vs)
{
@@ -269,7 +267,7 @@ llvm_vertex_shader(struct draw_vertex_shader *vs)
struct draw_llvm *
draw_llvm_create(struct draw_context *draw);
draw_llvm_create(struct draw_context *draw, struct gallivm_state *gallivm);
void
draw_llvm_destroy(struct draw_llvm *llvm);
@@ -286,7 +284,7 @@ struct draw_llvm_variant_key *
draw_llvm_make_variant_key(struct draw_llvm *llvm, char *store);
LLVMValueRef
draw_llvm_translate_from(LLVMBuilderRef builder,
draw_llvm_translate_from(struct gallivm_state *gallivm,
LLVMValueRef vbuffer,
enum pipe_format from_format);

View File

@@ -32,6 +32,7 @@
#include "pipe/p_defines.h"
#include "pipe/p_shader_tokens.h"
#include "gallivm/lp_bld_const.h"
#include "gallivm/lp_bld_debug.h"
#include "gallivm/lp_bld_type.h"
#include "gallivm/lp_bld_sample.h"
@@ -84,12 +85,13 @@ struct draw_llvm_sampler_soa
*/
static LLVMValueRef
draw_llvm_texture_member(const struct lp_sampler_dynamic_state *base,
LLVMBuilderRef builder,
struct gallivm_state *gallivm,
unsigned unit,
unsigned member_index,
const char *member_name,
boolean emit_load)
{
LLVMBuilderRef builder = gallivm->builder;
struct draw_llvm_sampler_dynamic_state *state =
(struct draw_llvm_sampler_dynamic_state *)base;
LLVMValueRef indices[4];
@@ -99,13 +101,13 @@ draw_llvm_texture_member(const struct lp_sampler_dynamic_state *base,
debug_assert(unit < PIPE_MAX_VERTEX_SAMPLERS);
/* context[0] */
indices[0] = LLVMConstInt(LLVMInt32Type(), 0, 0);
indices[0] = lp_build_const_int32(gallivm, 0);
/* context[0].textures */
indices[1] = LLVMConstInt(LLVMInt32Type(), DRAW_JIT_CTX_TEXTURES, 0);
indices[1] = lp_build_const_int32(gallivm, DRAW_JIT_CTX_TEXTURES);
/* context[0].textures[unit] */
indices[2] = LLVMConstInt(LLVMInt32Type(), unit, 0);
indices[2] = lp_build_const_int32(gallivm, unit);
/* context[0].textures[unit].member */
indices[3] = LLVMConstInt(LLVMInt32Type(), member_index, 0);
indices[3] = lp_build_const_int32(gallivm, member_index);
ptr = LLVMBuildGEP(builder, state->context_ptr, indices, Elements(indices), "");
@@ -132,10 +134,10 @@ draw_llvm_texture_member(const struct lp_sampler_dynamic_state *base,
#define DRAW_LLVM_TEXTURE_MEMBER(_name, _index, _emit_load) \
static LLVMValueRef \
draw_llvm_texture_##_name( const struct lp_sampler_dynamic_state *base, \
LLVMBuilderRef builder, \
struct gallivm_state *gallivm, \
unsigned unit) \
{ \
return draw_llvm_texture_member(base, builder, unit, _index, #_name, _emit_load ); \
return draw_llvm_texture_member(base, gallivm, unit, _index, #_name, _emit_load ); \
}
@@ -165,7 +167,7 @@ draw_llvm_sampler_soa_destroy(struct lp_build_sampler_soa *sampler)
*/
static void
draw_llvm_sampler_soa_emit_fetch_texel(const struct lp_build_sampler_soa *base,
LLVMBuilderRef builder,
struct gallivm_state *gallivm,
struct lp_type type,
unsigned unit,
unsigned num_coords,
@@ -180,7 +182,7 @@ draw_llvm_sampler_soa_emit_fetch_texel(const struct lp_build_sampler_soa *base,
assert(unit < PIPE_MAX_VERTEX_SAMPLERS);
lp_build_sample_soa(builder,
lp_build_sample_soa(gallivm,
&sampler->dynamic_state.static_state[unit],
&sampler->dynamic_state.base,
type,

View File

@@ -3,6 +3,7 @@
#include "draw_llvm.h"
#include "gallivm/lp_bld_const.h"
#include "gallivm/lp_bld_struct.h"
#include "gallivm/lp_bld_format.h"
#include "gallivm/lp_bld_debug.h"
@@ -16,272 +17,279 @@
#define DRAW_DBG 0
static LLVMValueRef
from_64_float(LLVMBuilderRef builder, LLVMValueRef val)
from_64_float(struct gallivm_state *gallivm, LLVMValueRef val)
{
LLVMValueRef bc = LLVMBuildBitCast(builder, val,
LLVMPointerType(LLVMDoubleType(), 0) , "");
LLVMValueRef l = LLVMBuildLoad(builder, bc, "");
return LLVMBuildFPTrunc(builder, l, LLVMFloatType(), "");
LLVMValueRef bc = LLVMBuildBitCast(gallivm->builder, val,
LLVMPointerType(LLVMDoubleTypeInContext(gallivm->context), 0) , "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, bc, "");
return LLVMBuildFPTrunc(gallivm->builder, l, LLVMFloatTypeInContext(gallivm->context), "");
}
static LLVMValueRef
from_32_float(LLVMBuilderRef builder, LLVMValueRef val)
from_32_float(struct gallivm_state *gallivm, LLVMValueRef val)
{
LLVMValueRef bc = LLVMBuildBitCast(builder, val,
LLVMPointerType(LLVMFloatType(), 0) , "");
return LLVMBuildLoad(builder, bc, "");
LLVMValueRef bc = LLVMBuildBitCast(gallivm->builder, val,
LLVMPointerType(LLVMFloatTypeInContext(gallivm->context), 0) , "");
return LLVMBuildLoad(gallivm->builder, bc, "");
}
static INLINE LLVMValueRef
from_8_uscaled(LLVMBuilderRef builder, LLVMValueRef val)
from_8_uscaled(struct gallivm_state *gallivm, LLVMValueRef val)
{
LLVMValueRef l = LLVMBuildLoad(builder, val, "");
return LLVMBuildUIToFP(builder, l, LLVMFloatType(), "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, val, "");
return LLVMBuildUIToFP(gallivm->builder, l, LLVMFloatTypeInContext(gallivm->context), "");
}
static INLINE LLVMValueRef
from_16_uscaled(LLVMBuilderRef builder, LLVMValueRef val)
from_16_uscaled(struct gallivm_state *gallivm, LLVMValueRef val)
{
LLVMValueRef bc = LLVMBuildBitCast(builder, val,
LLVMPointerType(LLVMIntType(16), 0) , "");
LLVMValueRef l = LLVMBuildLoad(builder, bc, "");
return LLVMBuildUIToFP(builder, l, LLVMFloatType(), "");
LLVMValueRef bc = LLVMBuildBitCast(gallivm->builder, val,
LLVMPointerType(LLVMIntTypeInContext(gallivm->context, 16), 0) , "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, bc, "");
return LLVMBuildUIToFP(gallivm->builder, l, LLVMFloatTypeInContext(gallivm->context), "");
}
static INLINE LLVMValueRef
from_32_uscaled(LLVMBuilderRef builder, LLVMValueRef val)
from_32_uscaled(struct gallivm_state *gallivm, LLVMValueRef val)
{
LLVMValueRef bc = LLVMBuildBitCast(builder, val,
LLVMPointerType(LLVMIntType(32), 0) , "");
LLVMValueRef l = LLVMBuildLoad(builder, bc, "");
return LLVMBuildUIToFP(builder, l, LLVMFloatType(), "");
LLVMValueRef bc = LLVMBuildBitCast(gallivm->builder, val,
LLVMPointerType(LLVMIntTypeInContext(gallivm->context, 32), 0) , "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, bc, "");
return LLVMBuildUIToFP(gallivm->builder, l, LLVMFloatTypeInContext(gallivm->context), "");
}
static INLINE LLVMValueRef
from_8_sscaled(LLVMBuilderRef builder, LLVMValueRef val)
from_8_sscaled(struct gallivm_state *gallivm, LLVMValueRef val)
{
LLVMValueRef l = LLVMBuildLoad(builder, val, "");
return LLVMBuildSIToFP(builder, l, LLVMFloatType(), "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, val, "");
return LLVMBuildSIToFP(gallivm->builder, l, LLVMFloatTypeInContext(gallivm->context), "");
}
static INLINE LLVMValueRef
from_16_sscaled(LLVMBuilderRef builder, LLVMValueRef val)
from_16_sscaled(struct gallivm_state *gallivm, LLVMValueRef val)
{
LLVMValueRef bc = LLVMBuildBitCast(builder, val,
LLVMPointerType(LLVMIntType(16), 0) , "");
LLVMValueRef l = LLVMBuildLoad(builder, bc, "");
return LLVMBuildSIToFP(builder, l, LLVMFloatType(), "");
LLVMValueRef bc = LLVMBuildBitCast(gallivm->builder, val,
LLVMPointerType(LLVMIntTypeInContext(gallivm->context, 16), 0) , "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, bc, "");
return LLVMBuildSIToFP(gallivm->builder, l, LLVMFloatTypeInContext(gallivm->context), "");
}
static INLINE LLVMValueRef
from_32_sscaled(LLVMBuilderRef builder, LLVMValueRef val)
from_32_sscaled(struct gallivm_state *gallivm, LLVMValueRef val)
{
LLVMValueRef bc = LLVMBuildBitCast(builder, val,
LLVMPointerType(LLVMIntType(32), 0) , "");
LLVMValueRef l = LLVMBuildLoad(builder, bc, "");
return LLVMBuildSIToFP(builder, l, LLVMFloatType(), "");
LLVMValueRef bc = LLVMBuildBitCast(gallivm->builder, val,
LLVMPointerType(LLVMIntTypeInContext(gallivm->context, 32), 0) , "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, bc, "");
return LLVMBuildSIToFP(gallivm->builder, l, LLVMFloatTypeInContext(gallivm->context), "");
}
static INLINE LLVMValueRef
from_8_unorm(LLVMBuilderRef builder, LLVMValueRef val)
from_8_unorm(struct gallivm_state *gallivm, LLVMValueRef val)
{
LLVMValueRef l = LLVMBuildLoad(builder, val, "");
LLVMValueRef uscaled = LLVMBuildUIToFP(builder, l, LLVMFloatType(), "");
return LLVMBuildFDiv(builder, uscaled,
LLVMConstReal(LLVMFloatType(), 255.), "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, val, "");
LLVMValueRef uscaled = LLVMBuildUIToFP(gallivm->builder, l, LLVMFloatTypeInContext(gallivm->context), "");
return LLVMBuildFDiv(gallivm->builder, uscaled,
lp_build_const_float(gallivm, 255.), "");
}
static INLINE LLVMValueRef
from_16_unorm(LLVMBuilderRef builder, LLVMValueRef val)
from_16_unorm(struct gallivm_state *gallivm, LLVMValueRef val)
{
LLVMValueRef bc = LLVMBuildBitCast(builder, val,
LLVMPointerType(LLVMIntType(16), 0) , "");
LLVMValueRef l = LLVMBuildLoad(builder, bc, "");
LLVMValueRef uscaled = LLVMBuildUIToFP(builder, l, LLVMFloatType(), "");
return LLVMBuildFDiv(builder, uscaled,
LLVMConstReal(LLVMFloatType(), 65535.), "");
LLVMValueRef bc = LLVMBuildBitCast(gallivm->builder, val,
LLVMPointerType(LLVMIntTypeInContext(gallivm->context, 16), 0) , "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, bc, "");
LLVMValueRef uscaled = LLVMBuildUIToFP(gallivm->builder, l, LLVMFloatTypeInContext(gallivm->context), "");
return LLVMBuildFDiv(gallivm->builder, uscaled,
lp_build_const_float(gallivm, 65535.), "");
}
static INLINE LLVMValueRef
from_32_unorm(LLVMBuilderRef builder, LLVMValueRef val)
from_32_unorm(struct gallivm_state *gallivm, LLVMValueRef val)
{
LLVMValueRef bc = LLVMBuildBitCast(builder, val,
LLVMPointerType(LLVMIntType(32), 0) , "");
LLVMValueRef l = LLVMBuildLoad(builder, bc, "");
LLVMValueRef uscaled = LLVMBuildUIToFP(builder, l, LLVMFloatType(), "");
LLVMValueRef bc = LLVMBuildBitCast(gallivm->builder, val,
LLVMPointerType(LLVMIntTypeInContext(gallivm->context, 32), 0) , "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, bc, "");
LLVMValueRef uscaled = LLVMBuildUIToFP(gallivm->builder, l, LLVMFloatTypeInContext(gallivm->context), "");
return LLVMBuildFDiv(builder, uscaled,
LLVMConstReal(LLVMFloatType(), 4294967295.), "");
return LLVMBuildFDiv(gallivm->builder, uscaled,
lp_build_const_float(gallivm, 4294967295.), "");
}
static INLINE LLVMValueRef
from_8_snorm(LLVMBuilderRef builder, LLVMValueRef val)
from_8_snorm(struct gallivm_state *gallivm, LLVMValueRef val)
{
LLVMValueRef l = LLVMBuildLoad(builder, val, "");
LLVMValueRef uscaled = LLVMBuildSIToFP(builder, l, LLVMFloatType(), "");
return LLVMBuildFDiv(builder, uscaled,
LLVMConstReal(LLVMFloatType(), 127.0), "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, val, "");
LLVMValueRef uscaled = LLVMBuildSIToFP(gallivm->builder, l, LLVMFloatTypeInContext(gallivm->context), "");
return LLVMBuildFDiv(gallivm->builder, uscaled,
lp_build_const_float(gallivm, 127.0), "");
}
static INLINE LLVMValueRef
from_16_snorm(LLVMBuilderRef builder, LLVMValueRef val)
from_16_snorm(struct gallivm_state *gallivm, LLVMValueRef val)
{
LLVMValueRef bc = LLVMBuildBitCast(builder, val,
LLVMPointerType(LLVMIntType(16), 0) , "");
LLVMValueRef l = LLVMBuildLoad(builder, bc, "");
LLVMValueRef uscaled = LLVMBuildSIToFP(builder, l, LLVMFloatType(), "");
return LLVMBuildFDiv(builder, uscaled,
LLVMConstReal(LLVMFloatType(), 32767.0f), "");
LLVMValueRef bc = LLVMBuildBitCast(gallivm->builder, val,
LLVMPointerType(LLVMIntTypeInContext(gallivm->context, 16), 0) , "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, bc, "");
LLVMValueRef uscaled = LLVMBuildSIToFP(gallivm->builder, l, LLVMFloatTypeInContext(gallivm->context), "");
return LLVMBuildFDiv(gallivm->builder, uscaled,
lp_build_const_float(gallivm, 32767.0f), "");
}
static INLINE LLVMValueRef
from_32_snorm(LLVMBuilderRef builder, LLVMValueRef val)
from_32_snorm(struct gallivm_state *gallivm, LLVMValueRef val)
{
LLVMValueRef bc = LLVMBuildBitCast(builder, val,
LLVMPointerType(LLVMIntType(32), 0) , "");
LLVMValueRef l = LLVMBuildLoad(builder, bc, "");
LLVMValueRef uscaled = LLVMBuildSIToFP(builder, l, LLVMFloatType(), "");
LLVMValueRef bc = LLVMBuildBitCast(gallivm->builder, val,
LLVMPointerType(LLVMIntTypeInContext(gallivm->context, 32), 0) , "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, bc, "");
LLVMValueRef uscaled = LLVMBuildSIToFP(gallivm->builder, l, LLVMFloatTypeInContext(gallivm->context), "");
return LLVMBuildFDiv(builder, uscaled,
LLVMConstReal(LLVMFloatType(), 2147483647.0), "");
return LLVMBuildFDiv(gallivm->builder, uscaled,
lp_build_const_float(gallivm, 2147483647.0), "");
}
static INLINE LLVMValueRef
from_32_fixed(LLVMBuilderRef builder, LLVMValueRef val)
from_32_fixed(struct gallivm_state *gallivm, LLVMValueRef val)
{
LLVMValueRef bc = LLVMBuildBitCast(builder, val,
LLVMPointerType(LLVMIntType(32), 0) , "");
LLVMValueRef l = LLVMBuildLoad(builder, bc, "");
LLVMValueRef uscaled = LLVMBuildSIToFP(builder, l, LLVMFloatType(), "");
LLVMValueRef bc = LLVMBuildBitCast(gallivm->builder, val,
LLVMPointerType(LLVMIntTypeInContext(gallivm->context, 32), 0) , "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, bc, "");
LLVMValueRef uscaled = LLVMBuildSIToFP(gallivm->builder, l, LLVMFloatTypeInContext(gallivm->context), "");
return LLVMBuildFDiv(builder, uscaled,
LLVMConstReal(LLVMFloatType(), 65536.0), "");
return LLVMBuildFDiv(gallivm->builder, uscaled,
lp_build_const_float(gallivm, 65536.0), "");
}
static LLVMValueRef
to_64_float(LLVMBuilderRef builder, LLVMValueRef fp)
to_64_float(struct gallivm_state *gallivm, LLVMValueRef fp)
{
LLVMValueRef l = LLVMBuildLoad(builder, fp, "");
return LLVMBuildFPExt(builder, l, LLVMDoubleType(), "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, fp, "");
return LLVMBuildFPExt(gallivm->builder, l, LLVMDoubleTypeInContext(gallivm->context), "");
}
static LLVMValueRef
to_32_float(LLVMBuilderRef builder, LLVMValueRef fp)
to_32_float(struct gallivm_state *gallivm, LLVMValueRef fp)
{
return LLVMBuildLoad(builder, fp, "");
return LLVMBuildLoad(gallivm->builder, fp, "");
}
static INLINE LLVMValueRef
to_8_uscaled(LLVMBuilderRef builder, LLVMValueRef fp)
to_8_uscaled(struct gallivm_state *gallivm, LLVMValueRef fp)
{
LLVMValueRef l = LLVMBuildLoad(builder, fp, "");
return LLVMBuildFPToUI(builder, l, LLVMIntType(8), "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, fp, "");
return LLVMBuildFPToUI(gallivm->builder, l, LLVMIntTypeInContext(gallivm->context, 8), "");
}
static INLINE LLVMValueRef
to_16_uscaled(LLVMBuilderRef builder, LLVMValueRef fp)
to_16_uscaled(struct gallivm_state *gallivm, LLVMValueRef fp)
{
LLVMValueRef l = LLVMBuildLoad(builder, fp, "");
return LLVMBuildFPToUI(builder, l, LLVMIntType(16), "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, fp, "");
return LLVMBuildFPToUI(gallivm->builder, l, LLVMIntTypeInContext(gallivm->context, 16), "");
}
static INLINE LLVMValueRef
to_32_uscaled(LLVMBuilderRef builder, LLVMValueRef fp)
to_32_uscaled(struct gallivm_state *gallivm, LLVMValueRef fp)
{
LLVMValueRef l = LLVMBuildLoad(builder, fp, "");
return LLVMBuildFPToUI(builder, l, LLVMIntType(32), "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, fp, "");
return LLVMBuildFPToUI(gallivm->builder, l, LLVMIntTypeInContext(gallivm->context, 32), "");
}
static INLINE LLVMValueRef
to_8_sscaled(LLVMBuilderRef builder, LLVMValueRef fp)
to_8_sscaled(struct gallivm_state *gallivm, LLVMValueRef fp)
{
LLVMValueRef l = LLVMBuildLoad(builder, fp, "");
return LLVMBuildFPToSI(builder, l, LLVMIntType(8), "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, fp, "");
return LLVMBuildFPToSI(gallivm->builder, l, LLVMIntTypeInContext(gallivm->context, 8), "");
}
static INLINE LLVMValueRef
to_16_sscaled(LLVMBuilderRef builder, LLVMValueRef fp)
to_16_sscaled(struct gallivm_state *gallivm, LLVMValueRef fp)
{
LLVMValueRef l = LLVMBuildLoad(builder, fp, "");
return LLVMBuildFPToSI(builder, l, LLVMIntType(16), "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, fp, "");
return LLVMBuildFPToSI(gallivm->builder, l, LLVMIntTypeInContext(gallivm->context, 16), "");
}
static INLINE LLVMValueRef
to_32_sscaled(LLVMBuilderRef builder, LLVMValueRef fp)
to_32_sscaled(struct gallivm_state *gallivm, LLVMValueRef fp)
{
LLVMValueRef l = LLVMBuildLoad(builder, fp, "");
return LLVMBuildFPToSI(builder, l, LLVMIntType(32), "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, fp, "");
return LLVMBuildFPToSI(gallivm->builder, l, LLVMIntTypeInContext(gallivm->context, 32), "");
}
static INLINE LLVMValueRef
to_8_unorm(LLVMBuilderRef builder, LLVMValueRef fp)
to_8_unorm(struct gallivm_state *gallivm, LLVMValueRef fp)
{
LLVMValueRef l = LLVMBuildLoad(builder, fp, "");
LLVMValueRef uscaled = LLVMBuildFPToUI(builder, l, LLVMIntType(8), "");
return LLVMBuildFMul(builder, uscaled,
LLVMConstReal(LLVMFloatType(), 255.), "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, fp, "");
LLVMValueRef uscaled = LLVMBuildFPToUI(gallivm->builder, l,
LLVMIntTypeInContext(gallivm->context, 8), "");
return LLVMBuildFMul(gallivm->builder, uscaled,
lp_build_const_float(gallivm, 255.), "");
}
static INLINE LLVMValueRef
to_16_unorm(LLVMBuilderRef builder, LLVMValueRef fp)
to_16_unorm(struct gallivm_state *gallivm, LLVMValueRef fp)
{
LLVMValueRef l = LLVMBuildLoad(builder, fp, "");
LLVMValueRef uscaled = LLVMBuildFPToUI(builder, l, LLVMIntType(32), "");
return LLVMBuildFMul(builder, uscaled,
LLVMConstReal(LLVMFloatType(), 65535.), "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, fp, "");
LLVMValueRef uscaled = LLVMBuildFPToUI(gallivm->builder, l,
LLVMIntTypeInContext(gallivm->context, 32), "");
return LLVMBuildFMul(gallivm->builder, uscaled,
lp_build_const_float(gallivm, 65535.), "");
}
static INLINE LLVMValueRef
to_32_unorm(LLVMBuilderRef builder, LLVMValueRef fp)
to_32_unorm(struct gallivm_state *gallivm, LLVMValueRef fp)
{
LLVMValueRef l = LLVMBuildLoad(builder, fp, "");
LLVMValueRef uscaled = LLVMBuildFPToUI(builder, l, LLVMIntType(32), "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, fp, "");
LLVMValueRef uscaled = LLVMBuildFPToUI(gallivm->builder, l,
LLVMIntTypeInContext(gallivm->context, 32), "");
return LLVMBuildFMul(builder, uscaled,
LLVMConstReal(LLVMFloatType(), 4294967295.), "");
return LLVMBuildFMul(gallivm->builder, uscaled,
lp_build_const_float(gallivm, 4294967295.), "");
}
static INLINE LLVMValueRef
to_8_snorm(LLVMBuilderRef builder, LLVMValueRef val)
to_8_snorm(struct gallivm_state *gallivm, LLVMValueRef val)
{
LLVMValueRef l = LLVMBuildLoad(builder, val, "");
LLVMValueRef uscaled = LLVMBuildFPToSI(builder, l, LLVMIntType(8), "");
return LLVMBuildFMul(builder, uscaled,
LLVMConstReal(LLVMFloatType(), 127.0), "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, val, "");
LLVMValueRef uscaled = LLVMBuildFPToSI(gallivm->builder, l,
LLVMIntTypeInContext(gallivm->context, 8), "");
return LLVMBuildFMul(gallivm->builder, uscaled,
lp_build_const_float(gallivm, 127.0), "");
}
static INLINE LLVMValueRef
to_16_snorm(LLVMBuilderRef builder, LLVMValueRef fp)
to_16_snorm(struct gallivm_state *gallivm, LLVMValueRef fp)
{
LLVMValueRef l = LLVMBuildLoad(builder, fp, "");
LLVMValueRef uscaled = LLVMBuildFPToSI(builder, l, LLVMIntType(16), "");
return LLVMBuildFMul(builder, uscaled,
LLVMConstReal(LLVMFloatType(), 32767.0f), "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, fp, "");
LLVMValueRef uscaled = LLVMBuildFPToSI(gallivm->builder, l,
LLVMIntTypeInContext(gallivm->context, 16), "");
return LLVMBuildFMul(gallivm->builder, uscaled,
lp_build_const_float(gallivm, 32767.0f), "");
}
static INLINE LLVMValueRef
to_32_snorm(LLVMBuilderRef builder, LLVMValueRef fp)
to_32_snorm(struct gallivm_state *gallivm, LLVMValueRef fp)
{
LLVMValueRef l = LLVMBuildLoad(builder, fp, "");
LLVMValueRef uscaled = LLVMBuildFPToSI(builder, l, LLVMIntType(32), "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, fp, "");
LLVMValueRef uscaled = LLVMBuildFPToSI(gallivm->builder, l,
LLVMIntTypeInContext(gallivm->context, 32), "");
return LLVMBuildFMul(builder, uscaled,
LLVMConstReal(LLVMFloatType(), 2147483647.0), "");
return LLVMBuildFMul(gallivm->builder, uscaled,
lp_build_const_float(gallivm, 2147483647.0), "");
}
static INLINE LLVMValueRef
to_32_fixed(LLVMBuilderRef builder, LLVMValueRef fp)
to_32_fixed(struct gallivm_state *gallivm, LLVMValueRef fp)
{
LLVMValueRef l = LLVMBuildLoad(builder, fp, "");
LLVMValueRef uscaled = LLVMBuildFPToSI(builder, l, LLVMIntType(32), "");
LLVMValueRef l = LLVMBuildLoad(gallivm->builder, fp, "");
LLVMValueRef uscaled = LLVMBuildFPToSI(gallivm->builder, l,
LLVMIntTypeInContext(gallivm->context, 32), "");
return LLVMBuildFMul(builder, uscaled,
LLVMConstReal(LLVMFloatType(), 65536.0), "");
return LLVMBuildFMul(gallivm->builder, uscaled,
lp_build_const_float(gallivm, 65536.0), "");
}
typedef LLVMValueRef (*from_func)(LLVMBuilderRef, LLVMValueRef);
typedef LLVMValueRef (*to_func)(LLVMBuilderRef, LLVMValueRef);
typedef LLVMValueRef (*from_func)(struct gallivm_state *, LLVMValueRef);
typedef LLVMValueRef (*to_func)(struct gallivm_state *, LLVMValueRef);
/* so that underneath can avoid function calls which are prohibited
* for static initialization we need this conversion */
@@ -294,21 +302,21 @@ enum ll_type {
};
static INLINE LLVMTypeRef
ll_type_to_llvm(enum ll_type type)
ll_type_to_llvm(struct gallivm_state *gallivm, enum ll_type type)
{
switch (type) {
case LL_Double:
return LLVMDoubleType();
return LLVMDoubleTypeInContext(gallivm->context);
case LL_Float:
return LLVMFloatType();
return LLVMFloatTypeInContext(gallivm->context);
case LL_Int32:
return LLVMInt32Type();
return LLVMInt32TypeInContext(gallivm->context);
case LL_Int16:
return LLVMIntType(16);
return LLVMIntTypeInContext(gallivm->context, 16);
case LL_Int8:
return LLVMIntType(8);
return LLVMIntTypeInContext(gallivm->context, 8);
}
return LLVMIntType(8);
return LLVMIntTypeInContext(gallivm->context, 8);
}
static INLINE int
@@ -414,42 +422,42 @@ struct draw_llvm_translate {
static LLVMValueRef
fetch(LLVMBuilderRef builder,
fetch(struct gallivm_state *gallivm,
LLVMValueRef ptr, int val_size, int nr_components,
from_func func)
{
int i;
int offset = 0;
LLVMValueRef res = LLVMConstNull(
LLVMVectorType(LLVMFloatType(), 4));
LLVMValueRef res =
LLVMConstNull(LLVMVectorType(LLVMFloatTypeInContext(gallivm->context), 4));
LLVMValueRef defaults[4];
defaults[0] = LLVMConstReal(LLVMFloatType(), 0);
defaults[1] = LLVMConstReal(LLVMFloatType(), 0);
defaults[2] = LLVMConstReal(LLVMFloatType(), 0);
defaults[3] = LLVMConstReal(LLVMFloatType(), 1);
defaults[0] =
defaults[1] =
defaults[2] = lp_build_const_float(gallivm, 0.0);
defaults[3] = lp_build_const_float(gallivm, 1.0);
for (i = 0; i < nr_components; ++i) {
LLVMValueRef src_index = LLVMConstInt(LLVMInt32Type(), offset, 0);
LLVMValueRef dst_index = LLVMConstInt(LLVMInt32Type(), i, 0);
LLVMValueRef src_index = lp_build_const_int32(gallivm, offset);
LLVMValueRef dst_index = lp_build_const_int32(gallivm, i);
LLVMValueRef src_tmp;
LLVMValueRef component;
src_tmp = LLVMBuildGEP(builder, ptr, &src_index, 1, "src_tmp");
src_tmp = LLVMBuildGEP(gallivm->builder, ptr, &src_index, 1, "src_tmp");
/* convert src_tmp to float */
component = func(builder, src_tmp);
component = func(gallivm, src_tmp);
/* vec.comp = component */
res = LLVMBuildInsertElement(builder,
res = LLVMBuildInsertElement(gallivm->builder,
res,
component,
dst_index, "");
offset += val_size;
}
for (; i < 4; ++i) {
LLVMValueRef dst_index = LLVMConstInt(LLVMInt32Type(), i, 0);
res = LLVMBuildInsertElement(builder,
LLVMValueRef dst_index = lp_build_const_int32(gallivm, i);
res = LLVMBuildInsertElement(gallivm->builder,
res,
defaults[i],
dst_index, "");
@@ -459,7 +467,7 @@ fetch(LLVMBuilderRef builder,
LLVMValueRef
draw_llvm_translate_from(LLVMBuilderRef builder,
draw_llvm_translate_from(struct gallivm_state *gallivm,
LLVMValueRef vbuffer,
enum pipe_format from_format)
{
@@ -476,7 +484,7 @@ draw_llvm_translate_from(LLVMBuilderRef builder,
for (i = 0; i < Elements(translates); ++i) {
if (translates[i].format == from_format) {
/*LLVMTypeRef type = ll_type_to_llvm(translates[i].type);*/
return fetch(builder,
return fetch(gallivm,
vbuffer,
ll_type_size(translates[i].type),
translates[i].num_components,
@@ -493,6 +501,6 @@ draw_llvm_translate_from(LLVMBuilderRef builder,
*/
format_desc = util_format_description(from_format);
zero = LLVMConstNull(LLVMInt32Type());
return lp_build_fetch_rgba_aos(builder, format_desc, type, vbuffer, zero, zero, zero);
zero = LLVMConstNull(LLVMInt32TypeInContext(gallivm->context));
return lp_build_fetch_rgba_aos(gallivm, format_desc, type, vbuffer, zero, zero, zero);
}

View File

@@ -406,6 +406,7 @@ aaline_create_texture(struct aaline_stage *aaline)
texTemp.width0 = 1 << MAX_TEXTURE_LEVEL;
texTemp.height0 = 1 << MAX_TEXTURE_LEVEL;
texTemp.depth0 = 1;
texTemp.array_size = 1;
texTemp.bind = PIPE_BIND_SAMPLER_VIEW;
aaline->texture = screen->resource_create(screen, &texTemp);
@@ -441,10 +442,10 @@ aaline_create_texture(struct aaline_stage *aaline)
/* This texture is new, no need to flush.
*/
transfer = pipe->get_transfer(pipe,
aaline->texture,
u_subresource(0, level),
PIPE_TRANSFER_WRITE,
&box);
aaline->texture,
level,
PIPE_TRANSFER_WRITE,
&box);
data = pipe->transfer_map(pipe, transfer);
if (data == NULL)

View File

@@ -163,6 +163,7 @@ static void interp( const struct clip_stage *clip,
*/
static void emit_poly( struct draw_stage *stage,
struct vertex_header **inlist,
const boolean *edgeflags,
unsigned n,
const struct prim_header *origPrim)
{
@@ -181,6 +182,9 @@ static void emit_poly( struct draw_stage *stage,
edge_last = DRAW_PIPE_EDGE_FLAG_1;
}
if (!edgeflags[0])
edge_first = 0;
/* later stages may need the determinant, but only the sign matters */
header.det = origPrim->det;
header.flags = DRAW_PIPE_RESET_STIPPLE | edge_first | edge_middle;
@@ -199,7 +203,11 @@ static void emit_poly( struct draw_stage *stage,
header.v[2] = inlist[0]; /* the provoking vertex */
}
if (i == n-1)
if (!edgeflags[i-1]) {
header.flags &= ~edge_middle;
}
if (i == n - 1 && edgeflags[i])
header.flags |= edge_last;
if (0) {
@@ -248,15 +256,33 @@ do_clip_tri( struct draw_stage *stage,
unsigned tmpnr = 0;
unsigned n = 3;
unsigned i;
boolean aEdges[MAX_CLIPPED_VERTICES];
boolean bEdges[MAX_CLIPPED_VERTICES];
boolean *inEdges = aEdges;
boolean *outEdges = bEdges;
inlist[0] = header->v[0];
inlist[1] = header->v[1];
inlist[2] = header->v[2];
/*
* Note: at this point we can't just use the per-vertex edge flags.
* We have to observe the edge flag bits set in header->flags which
* were set during primitive decomposition. Put those flags into
* an edge flags array which parallels the vertex array.
* Later, in the 'unfilled' pipeline stage we'll draw the edge if both
* the header.flags bit is set AND the per-vertex edgeflag field is set.
*/
inEdges[0] = !!(header->flags & DRAW_PIPE_EDGE_FLAG_0);
inEdges[1] = !!(header->flags & DRAW_PIPE_EDGE_FLAG_1);
inEdges[2] = !!(header->flags & DRAW_PIPE_EDGE_FLAG_2);
while (clipmask && n >= 3) {
const unsigned plane_idx = ffs(clipmask)-1;
const boolean is_user_clip_plane = plane_idx >= 6;
const float *plane = clipper->plane[plane_idx];
struct vertex_header *vert_prev = inlist[0];
boolean *edge_prev = &inEdges[0];
float dp_prev = dot4( vert_prev->clip, plane );
unsigned outcount = 0;
@@ -266,9 +292,11 @@ do_clip_tri( struct draw_stage *stage,
if (n >= MAX_CLIPPED_VERTICES)
return;
inlist[n] = inlist[0]; /* prevent rotation of vertices */
inEdges[n] = inEdges[0];
for (i = 1; i <= n; i++) {
struct vertex_header *vert = inlist[i];
boolean *edge = &inEdges[i];
float dp = dot4( vert->clip, plane );
@@ -276,11 +304,13 @@ do_clip_tri( struct draw_stage *stage,
assert(outcount < MAX_CLIPPED_VERTICES);
if (outcount >= MAX_CLIPPED_VERTICES)
return;
outEdges[outcount] = *edge_prev;
outlist[outcount++] = vert_prev;
}
if (DIFFERENT_SIGNS(dp, dp_prev)) {
struct vertex_header *new_vert;
boolean *new_edge;
assert(tmpnr < MAX_CLIPPED_VERTICES + 1);
if (tmpnr >= MAX_CLIPPED_VERTICES + 1)
@@ -290,6 +320,8 @@ do_clip_tri( struct draw_stage *stage,
assert(outcount < MAX_CLIPPED_VERTICES);
if (outcount >= MAX_CLIPPED_VERTICES)
return;
new_edge = &outEdges[outcount];
outlist[outcount++] = new_vert;
if (IS_NEGATIVE(dp)) {
@@ -299,10 +331,22 @@ do_clip_tri( struct draw_stage *stage,
float t = dp / (dp - dp_prev);
interp( clipper, new_vert, t, vert, vert_prev );
/* Force edgeflag true in this case:
/* Whether or not to set edge flag for the new vert depends
* on whether it's a user-defined clipping plane. We're
* copying NVIDIA's behaviour here.
*/
new_vert->edgeflag = 1;
} else {
if (is_user_clip_plane) {
/* we want to see an edge along the clip plane */
*new_edge = TRUE;
new_vert->edgeflag = TRUE;
}
else {
/* we don't want to see an edge along the frustum clip plane */
*new_edge = *edge_prev;
new_vert->edgeflag = FALSE;
}
}
else {
/* Coming back in.
*/
float t = dp_prev / (dp_prev - dp);
@@ -311,10 +355,12 @@ do_clip_tri( struct draw_stage *stage,
/* Copy starting vert's edgeflag:
*/
new_vert->edgeflag = vert_prev->edgeflag;
*new_edge = *edge_prev;
}
}
vert_prev = vert;
edge_prev = edge;
dp_prev = dp;
}
@@ -325,6 +371,12 @@ do_clip_tri( struct draw_stage *stage,
outlist = tmp;
n = outcount;
}
{
boolean *tmp = inEdges;
inEdges = outEdges;
outEdges = tmp;
}
}
/* If flat-shading, copy provoking vertex color to polygon vertex[0]
@@ -353,7 +405,7 @@ do_clip_tri( struct draw_stage *stage,
/* Emit the polygon as triangles to the setup stage:
*/
emit_poly( stage, inlist, n, header );
emit_poly( stage, inlist, inEdges, n, header );
}
}

View File

@@ -393,8 +393,8 @@ pstip_update_texture(struct pstip_stage *pstip)
*/
pipe->flush( pipe, PIPE_FLUSH_TEXTURE_CACHE, NULL );
transfer = pipe_get_transfer(pipe, pstip->texture, 0, 0, 0,
PIPE_TRANSFER_WRITE, 0, 0, 32, 32);
transfer = pipe_get_transfer(pipe, pstip->texture, 0, 0,
PIPE_TRANSFER_WRITE, 0, 0, 32, 32);
data = pipe->transfer_map(pipe, transfer);
/*
@@ -440,6 +440,7 @@ pstip_create_texture(struct pstip_stage *pstip)
texTemp.width0 = 32;
texTemp.height0 = 32;
texTemp.depth0 = 1;
texTemp.array_size = 1;
texTemp.bind = PIPE_BIND_SAMPLER_VIEW;
pstip->texture = screen->resource_create(screen, &texTemp);

View File

@@ -286,7 +286,6 @@ struct draw_context
#ifdef HAVE_LLVM
struct draw_llvm *llvm;
LLVMExecutionEngineRef engine;
#endif
struct pipe_sampler_view *sampler_views[PIPE_MAX_VERTEX_SAMPLERS];

View File

@@ -34,6 +34,7 @@
#include "draw/draw_pt.h"
#include "draw/draw_vs.h"
#include "draw/draw_llvm.h"
#include "gallivm/lp_bld_init.h"
struct llvm_middle_end {
@@ -72,19 +73,18 @@ llvm_middle_end_prepare( struct draw_pt_middle_end *middle,
struct draw_llvm_variant_list_item *li;
unsigned i;
unsigned instance_id_index = ~0;
unsigned out_prim = (draw->gs.geometry_shader ?
draw->gs.geometry_shader->output_primitive :
in_prim);
const unsigned out_prim = (draw->gs.geometry_shader ?
draw->gs.geometry_shader->output_primitive :
in_prim);
/* Add one to num_outputs because the pipeline occasionally tags on
* an additional texcoord, eg for AA lines.
*/
unsigned nr = MAX2( shader->base.info.num_inputs,
shader->base.info.num_outputs + 1 );
const unsigned nr = MAX2( shader->base.info.num_inputs,
shader->base.info.num_outputs + 1 );
/* Scan for instanceID system value.
* XXX but we never use instance_id_index?!
*/
for (i = 0; i < shader->base.info.num_inputs; i++) {
if (shader->base.info.input_semantic_name[i] == TGSI_SEMANTIC_INSTANCEID) {
@@ -133,9 +133,10 @@ llvm_middle_end_prepare( struct draw_pt_middle_end *middle,
key = draw_llvm_make_variant_key(fpme->llvm, store);
/* Search shader's list of variants for the key */
li = first_elem(&shader->variants);
while(!at_end(&shader->variants, li)) {
if(memcmp(&li->base->key, key, shader->variant_key_size) == 0) {
while (!at_end(&shader->variants, li)) {
if (memcmp(&li->base->key, key, shader->variant_key_size) == 0) {
variant = li->base;
break;
}
@@ -143,10 +144,16 @@ llvm_middle_end_prepare( struct draw_pt_middle_end *middle,
}
if (variant) {
/* found the variant, move to head of global list (for LRU) */
move_to_head(&fpme->llvm->vs_variants_list, &variant->list_item_global);
}
else {
/* Need to create new variant */
unsigned i;
/* First check if we've created too many variants. If so, free
* 25% of the LRU to avoid using too much memory.
*/
if (fpme->llvm->nr_variants >= DRAW_MAX_SHADER_VARIANTS) {
/*
* XXX: should we flush here ?
@@ -422,7 +429,7 @@ draw_pt_fetch_pipeline_or_emit_llvm(struct draw_context *draw)
{
struct llvm_middle_end *fpme = 0;
if (!draw->engine)
if (!draw->llvm->gallivm->engine)
return NULL;
fpme = CALLOC_STRUCT( llvm_middle_end );

View File

@@ -258,9 +258,10 @@ vsplit_segment_fan_linear(struct vsplit_frontend *vsplit, unsigned flags,
boolean use_spoken = ((flags & DRAW_SPLIT_BEFORE) != 0);
unsigned nr = 0, i;
assert(icount + !!use_spoken <= vsplit->segment_size);
assert(icount <= vsplit->segment_size);
if (use_spoken) {
/* replace istart by i0 */
vsplit->fetch_elts[nr++] = i0;
for (i = 1 ; i < icount; i++)
vsplit->fetch_elts[nr++] = istart + i;

View File

@@ -33,6 +33,7 @@
#include "draw_context.h"
#include "draw_private.h"
#include "draw_vertex.h"
struct draw_context;
@@ -48,7 +49,7 @@ struct draw_varient_input
struct draw_varient_output
{
enum pipe_format format; /* output format */
enum attrib_emit format; /* output format */
unsigned vs_output:8; /* which vertex shader output is this? */
unsigned offset:24; /* offset into output vertex */
};

View File

@@ -384,7 +384,7 @@ static void emit_store_R8G8B8A8_UNORM( struct aos_compilation *cp,
static boolean emit_output( struct aos_compilation *cp,
struct x86_reg ptr,
struct x86_reg dataXMM,
unsigned format )
enum attrib_emit format )
{
switch (format) {
case EMIT_1F:
@@ -422,7 +422,7 @@ boolean aos_emit_outputs( struct aos_compilation *cp )
unsigned i;
for (i = 0; i < cp->vaos->base.key.nr_outputs; i++) {
unsigned format = cp->vaos->base.key.element[i].out.format;
enum attrib_emit format = cp->vaos->base.key.element[i].out.format;
unsigned offset = cp->vaos->base.key.element[i].out.offset;
unsigned vs_output = cp->vaos->base.key.element[i].out.vs_output;

View File

@@ -65,19 +65,7 @@ static void
vs_llvm_delete( struct draw_vertex_shader *dvs )
{
struct llvm_vertex_shader *shader = llvm_vertex_shader(dvs);
struct pipe_fence_handle *fence = NULL;
struct draw_llvm_variant_list_item *li;
struct pipe_context *pipe = dvs->draw->pipe;
/*
* XXX: This might be not neccessary at all.
*/
pipe->flush(pipe, 0, &fence);
if (fence) {
pipe->screen->fence_finish(pipe->screen, fence, 0);
pipe->screen->fence_reference(pipe->screen, &fence, NULL);
}
li = first_elem(&shader->variants);
while(!at_end(&shader->variants, li)) {

View File

@@ -55,4 +55,33 @@
#endif
/**
* Redefine these LLVM entrypoints as invalid macros to make sure we
* don't accidentally use them. We need to use the functions which
* take an explicit LLVMContextRef parameter.
*/
#define LLVMInt1Type ILLEGAL_LLVM_FUNCTION
#define LLVMInt8Type ILLEGAL_LLVM_FUNCTION
#define LLVMInt16Type ILLEGAL_LLVM_FUNCTION
#define LLVMInt32Type ILLEGAL_LLVM_FUNCTION
#define LLVMInt64Type ILLEGAL_LLVM_FUNCTION
#define LLVMIntType ILLEGAL_LLVM_FUNCTION
#define LLVMFloatType ILLEGAL_LLVM_FUNCTION
#define LLVMDoubleType ILLEGAL_LLVM_FUNCTION
#define LLVMX86FP80Type ILLEGAL_LLVM_FUNCTION
#define LLVMFP128Type ILLEGAL_LLVM_FUNCTION
#define LLVMPPCFP128Type ILLEGAL_LLVM_FUNCTION
#define LLVMStructType ILLEGAL_LLVM_FUNCTION
#define LLVMVoidType ILLEGAL_LLVM_FUNCTION
#define LLVMLabelType ILLEGAL_LLVM_FUNCTION
#define LLVMOpaqueType ILLEGAL_LLVM_FUNCTION
#define LLVMUnionType ILLEGAL_LLVM_FUNCTION
#define LLVMMDString ILLEGAL_LLVM_FUNCTION
#define LLVMMDNode ILLEGAL_LLVM_FUNCTION
#define LLVMConstString ILLEGAL_LLVM_FUNCTION
#define LLVMConstStruct ILLEGAL_LLVM_FUNCTION
#define LLVMAppendBasicBlock ILLEGAL_LLVM_FUNCTION
#define LLVMInsertBasicBlock ILLEGAL_LLVM_FUNCTION
#define LLVMCreateBuilder ILLEGAL_LLVM_FUNCTION
#endif /* LP_BLD_H */

File diff suppressed because it is too large Load Diff

View File

@@ -56,20 +56,21 @@ lp_assert(int condition, const char *msg)
* \param msg a string to print if the assertion fails.
*/
LLVMValueRef
lp_build_assert(LLVMBuilderRef builder, LLVMValueRef condition,
lp_build_assert(struct gallivm_state *gallivm,
LLVMValueRef condition,
const char *msg)
{
LLVMModuleRef module;
LLVMBuilderRef builder = gallivm->builder;
LLVMContextRef context = gallivm->context;
LLVMModuleRef module = gallivm->module;
LLVMTypeRef arg_types[2];
LLVMValueRef msg_string, assert_func, params[2], r;
module = LLVMGetGlobalParent(LLVMGetBasicBlockParent(
LLVMGetInsertBlock(builder)));
msg_string = lp_build_const_string_variable(module, context,
msg, strlen(msg) + 1);
msg_string = lp_build_const_string_variable(module, msg, strlen(msg) + 1);
arg_types[0] = LLVMInt32Type();
arg_types[1] = LLVMPointerType(LLVMInt8Type(), 0);
arg_types[0] = LLVMInt32TypeInContext(context);
arg_types[1] = LLVMPointerType(LLVMInt8TypeInContext(context), 0);
/* lookup the lp_assert function */
assert_func = LLVMGetNamedFunction(module, "lp_assert");
@@ -77,12 +78,12 @@ lp_build_assert(LLVMBuilderRef builder, LLVMValueRef condition,
/* Create the assertion function if not found */
if (!assert_func) {
LLVMTypeRef func_type =
LLVMFunctionType(LLVMVoidType(), arg_types, 2, 0);
LLVMFunctionType(LLVMVoidTypeInContext(context), arg_types, 2, 0);
assert_func = LLVMAddFunction(module, "lp_assert", func_type);
LLVMSetFunctionCallConv(assert_func, LLVMCCallConv);
LLVMSetLinkage(assert_func, LLVMExternalLinkage);
LLVMAddGlobalMapping(lp_build_engine, assert_func,
LLVMAddGlobalMapping(gallivm->engine, assert_func,
func_to_pointer((func_pointer)lp_assert));
}
assert(assert_func);

View File

@@ -30,10 +30,12 @@
#include "lp_bld.h"
#include "lp_bld_init.h"
LLVMValueRef
lp_build_assert(LLVMBuilderRef builder, LLVMValueRef condition,
lp_build_assert(struct gallivm_state *gallivm,
LLVMValueRef condition,
const char *msg);

View File

@@ -40,6 +40,7 @@
LLVMValueRef
lp_build_or(struct lp_build_context *bld, LLVMValueRef a, LLVMValueRef b)
{
LLVMBuilderRef builder = bld->gallivm->builder;
const struct lp_type type = bld->type;
LLVMValueRef res;
@@ -48,14 +49,14 @@ lp_build_or(struct lp_build_context *bld, LLVMValueRef a, LLVMValueRef b)
/* can't do bitwise ops on floating-point values */
if (type.floating) {
a = LLVMBuildBitCast(bld->builder, a, bld->int_vec_type, "");
b = LLVMBuildBitCast(bld->builder, b, bld->int_vec_type, "");
a = LLVMBuildBitCast(builder, a, bld->int_vec_type, "");
b = LLVMBuildBitCast(builder, b, bld->int_vec_type, "");
}
res = LLVMBuildOr(bld->builder, a, b, "");
res = LLVMBuildOr(builder, a, b, "");
if (type.floating) {
res = LLVMBuildBitCast(bld->builder, res, bld->vec_type, "");
res = LLVMBuildBitCast(builder, res, bld->vec_type, "");
}
return res;
@@ -68,6 +69,7 @@ lp_build_or(struct lp_build_context *bld, LLVMValueRef a, LLVMValueRef b)
LLVMValueRef
lp_build_and(struct lp_build_context *bld, LLVMValueRef a, LLVMValueRef b)
{
LLVMBuilderRef builder = bld->gallivm->builder;
const struct lp_type type = bld->type;
LLVMValueRef res;
@@ -76,14 +78,14 @@ lp_build_and(struct lp_build_context *bld, LLVMValueRef a, LLVMValueRef b)
/* can't do bitwise ops on floating-point values */
if (type.floating) {
a = LLVMBuildBitCast(bld->builder, a, bld->int_vec_type, "");
b = LLVMBuildBitCast(bld->builder, b, bld->int_vec_type, "");
a = LLVMBuildBitCast(builder, a, bld->int_vec_type, "");
b = LLVMBuildBitCast(builder, b, bld->int_vec_type, "");
}
res = LLVMBuildAnd(bld->builder, a, b, "");
res = LLVMBuildAnd(builder, a, b, "");
if (type.floating) {
res = LLVMBuildBitCast(bld->builder, res, bld->vec_type, "");
res = LLVMBuildBitCast(builder, res, bld->vec_type, "");
}
return res;
@@ -96,6 +98,7 @@ lp_build_and(struct lp_build_context *bld, LLVMValueRef a, LLVMValueRef b)
LLVMValueRef
lp_build_andnot(struct lp_build_context *bld, LLVMValueRef a, LLVMValueRef b)
{
LLVMBuilderRef builder = bld->gallivm->builder;
const struct lp_type type = bld->type;
LLVMValueRef res;
@@ -104,15 +107,15 @@ lp_build_andnot(struct lp_build_context *bld, LLVMValueRef a, LLVMValueRef b)
/* can't do bitwise ops on floating-point values */
if (type.floating) {
a = LLVMBuildBitCast(bld->builder, a, bld->int_vec_type, "");
b = LLVMBuildBitCast(bld->builder, b, bld->int_vec_type, "");
a = LLVMBuildBitCast(builder, a, bld->int_vec_type, "");
b = LLVMBuildBitCast(builder, b, bld->int_vec_type, "");
}
res = LLVMBuildNot(bld->builder, b, "");
res = LLVMBuildAnd(bld->builder, a, res, "");
res = LLVMBuildNot(builder, b, "");
res = LLVMBuildAnd(builder, a, res, "");
if (type.floating) {
res = LLVMBuildBitCast(bld->builder, res, bld->vec_type, "");
res = LLVMBuildBitCast(builder, res, bld->vec_type, "");
}
return res;
@@ -125,6 +128,7 @@ lp_build_andnot(struct lp_build_context *bld, LLVMValueRef a, LLVMValueRef b)
LLVMValueRef
lp_build_shl(struct lp_build_context *bld, LLVMValueRef a, LLVMValueRef b)
{
LLVMBuilderRef builder = bld->gallivm->builder;
const struct lp_type type = bld->type;
LLVMValueRef res;
@@ -133,7 +137,7 @@ lp_build_shl(struct lp_build_context *bld, LLVMValueRef a, LLVMValueRef b)
assert(lp_check_value(type, a));
assert(lp_check_value(type, b));
res = LLVMBuildShl(bld->builder, a, b, "");
res = LLVMBuildShl(builder, a, b, "");
return res;
}
@@ -145,6 +149,7 @@ lp_build_shl(struct lp_build_context *bld, LLVMValueRef a, LLVMValueRef b)
LLVMValueRef
lp_build_shr(struct lp_build_context *bld, LLVMValueRef a, LLVMValueRef b)
{
LLVMBuilderRef builder = bld->gallivm->builder;
const struct lp_type type = bld->type;
LLVMValueRef res;
@@ -154,9 +159,9 @@ lp_build_shr(struct lp_build_context *bld, LLVMValueRef a, LLVMValueRef b)
assert(lp_check_value(type, b));
if (type.sign) {
res = LLVMBuildAShr(bld->builder, a, b, "");
res = LLVMBuildAShr(builder, a, b, "");
} else {
res = LLVMBuildLShr(bld->builder, a, b, "");
res = LLVMBuildLShr(builder, a, b, "");
}
return res;
@@ -169,7 +174,7 @@ lp_build_shr(struct lp_build_context *bld, LLVMValueRef a, LLVMValueRef b)
LLVMValueRef
lp_build_shl_imm(struct lp_build_context *bld, LLVMValueRef a, unsigned imm)
{
LLVMValueRef b = lp_build_const_int_vec(bld->type, imm);
LLVMValueRef b = lp_build_const_int_vec(bld->gallivm, bld->type, imm);
assert(imm <= bld->type.width);
return lp_build_shl(bld, a, b);
}
@@ -181,7 +186,7 @@ lp_build_shl_imm(struct lp_build_context *bld, LLVMValueRef a, unsigned imm)
LLVMValueRef
lp_build_shr_imm(struct lp_build_context *bld, LLVMValueRef a, unsigned imm)
{
LLVMValueRef b = lp_build_const_int_vec(bld->type, imm);
LLVMValueRef b = lp_build_const_int_vec(bld->gallivm, bld->type, imm);
assert(imm <= bld->type.width);
return lp_build_shr(bld, a, b);
}

View File

@@ -39,6 +39,7 @@
#include "lp_bld_type.h"
#include "lp_bld_const.h"
#include "lp_bld_init.h"
unsigned
@@ -211,31 +212,31 @@ lp_const_eps(struct lp_type type)
LLVMValueRef
lp_build_undef(struct lp_type type)
lp_build_undef(struct gallivm_state *gallivm, struct lp_type type)
{
LLVMTypeRef vec_type = lp_build_vec_type(type);
LLVMTypeRef vec_type = lp_build_vec_type(gallivm, type);
return LLVMGetUndef(vec_type);
}
LLVMValueRef
lp_build_zero(struct lp_type type)
lp_build_zero(struct gallivm_state *gallivm, struct lp_type type)
{
if (type.length == 1) {
if (type.floating)
return LLVMConstReal(LLVMFloatType(), 0.0);
return lp_build_const_float(gallivm, 0.0);
else
return LLVMConstInt(LLVMIntType(type.width), 0, 0);
return LLVMConstInt(LLVMIntTypeInContext(gallivm->context, type.width), 0, 0);
}
else {
LLVMTypeRef vec_type = lp_build_vec_type(type);
LLVMTypeRef vec_type = lp_build_vec_type(gallivm, type);
return LLVMConstNull(vec_type);
}
}
LLVMValueRef
lp_build_one(struct lp_type type)
lp_build_one(struct gallivm_state *gallivm, struct lp_type type)
{
LLVMTypeRef elem_type;
LLVMValueRef elems[LP_MAX_VECTOR_LENGTH];
@@ -243,7 +244,7 @@ lp_build_one(struct lp_type type)
assert(type.length <= LP_MAX_VECTOR_LENGTH);
elem_type = lp_build_elem_type(type);
elem_type = lp_build_elem_type(gallivm, type);
if(type.floating)
elems[0] = LLVMConstReal(elem_type, 1.0);
@@ -283,10 +284,11 @@ lp_build_one(struct lp_type type)
* Build constant-valued element from a scalar value.
*/
LLVMValueRef
lp_build_const_elem(struct lp_type type,
lp_build_const_elem(struct gallivm_state *gallivm,
struct lp_type type,
double val)
{
LLVMTypeRef elem_type = lp_build_elem_type(type);
LLVMTypeRef elem_type = lp_build_elem_type(gallivm, type);
LLVMValueRef elem;
if(type.floating) {
@@ -306,15 +308,15 @@ lp_build_const_elem(struct lp_type type,
* Build constant-valued vector from a scalar value.
*/
LLVMValueRef
lp_build_const_vec(struct lp_type type,
lp_build_const_vec(struct gallivm_state *gallivm, struct lp_type type,
double val)
{
if (type.length == 1) {
return lp_build_const_elem(type, val);
return lp_build_const_elem(gallivm, type, val);
} else {
LLVMValueRef elems[LP_MAX_VECTOR_LENGTH];
unsigned i;
elems[0] = lp_build_const_elem(type, val);
elems[0] = lp_build_const_elem(gallivm, type, val);
for(i = 1; i < type.length; ++i)
elems[i] = elems[0];
return LLVMConstVector(elems, type.length);
@@ -323,10 +325,10 @@ lp_build_const_vec(struct lp_type type,
LLVMValueRef
lp_build_const_int_vec(struct lp_type type,
long long val)
lp_build_const_int_vec(struct gallivm_state *gallivm, struct lp_type type,
long long val)
{
LLVMTypeRef elem_type = lp_build_int_elem_type(type);
LLVMTypeRef elem_type = lp_build_int_elem_type(gallivm, type);
LLVMValueRef elems[LP_MAX_VECTOR_LENGTH];
unsigned i;
@@ -343,7 +345,8 @@ lp_build_const_int_vec(struct lp_type type,
LLVMValueRef
lp_build_const_aos(struct lp_type type,
lp_build_const_aos(struct gallivm_state *gallivm,
struct lp_type type,
double r, double g, double b, double a,
const unsigned char *swizzle)
{
@@ -355,7 +358,7 @@ lp_build_const_aos(struct lp_type type,
assert(type.length % 4 == 0);
assert(type.length <= LP_MAX_VECTOR_LENGTH);
elem_type = lp_build_elem_type(type);
elem_type = lp_build_elem_type(gallivm, type);
if(swizzle == NULL)
swizzle = default_swizzle;
@@ -386,10 +389,11 @@ lp_build_const_aos(struct lp_type type,
* @param mask TGSI_WRITEMASK_xxx
*/
LLVMValueRef
lp_build_const_mask_aos(struct lp_type type,
lp_build_const_mask_aos(struct gallivm_state *gallivm,
struct lp_type type,
unsigned mask)
{
LLVMTypeRef elem_type = LLVMIntType(type.width);
LLVMTypeRef elem_type = LLVMIntTypeInContext(gallivm->context, type.width);
LLVMValueRef masks[LP_MAX_VECTOR_LENGTH];
unsigned i, j;

View File

@@ -39,6 +39,7 @@
#include "pipe/p_compiler.h"
#include "gallivm/lp_bld.h"
#include "gallivm/lp_bld_init.h"
@@ -73,44 +74,71 @@ lp_const_eps(struct lp_type type);
LLVMValueRef
lp_build_undef(struct lp_type type);
lp_build_undef(struct gallivm_state *gallivm, struct lp_type type);
LLVMValueRef
lp_build_zero(struct lp_type type);
lp_build_zero(struct gallivm_state *gallivm, struct lp_type type);
LLVMValueRef
lp_build_one(struct lp_type type);
lp_build_one(struct gallivm_state *gallivm, struct lp_type type);
LLVMValueRef
lp_build_const_elem(struct lp_type type,
lp_build_const_elem(struct gallivm_state *gallivm, struct lp_type type,
double val);
LLVMValueRef
lp_build_const_vec(struct lp_type type, double val);
lp_build_const_vec(struct gallivm_state *gallivm, struct lp_type type,
double val);
LLVMValueRef
lp_build_const_int_vec(struct lp_type type, long long val);
lp_build_const_int_vec(struct gallivm_state *gallivm,
struct lp_type type, long long val);
LLVMValueRef
lp_build_const_aos(struct lp_type type,
lp_build_const_aos(struct gallivm_state *gallivm, struct lp_type type,
double r, double g, double b, double a,
const unsigned char *swizzle);
LLVMValueRef
lp_build_const_mask_aos(struct lp_type type,
lp_build_const_mask_aos(struct gallivm_state *gallivm,
struct lp_type type,
unsigned mask);
static INLINE LLVMValueRef
lp_build_const_int32(int i)
lp_build_const_int32(struct gallivm_state *gallivm, int i)
{
return LLVMConstInt(LLVMInt32Type(), i, 0);
return LLVMConstInt(LLVMInt32TypeInContext(gallivm->context), i, 0);
}
static INLINE LLVMValueRef
lp_build_const_float(struct gallivm_state *gallivm, float x)
{
return LLVMConstReal(LLVMFloatTypeInContext(gallivm->context), x);
}
/** Return constant-valued pointer to int */
static INLINE LLVMValueRef
lp_build_const_int_pointer(struct gallivm_state *gallivm, const void *ptr)
{
LLVMTypeRef int_type;
LLVMValueRef v;
/* int type large enough to hold a pointer */
int_type = LLVMIntTypeInContext(gallivm->context, 8 * sizeof(void *));
v = LLVMConstInt(int_type, (unsigned long long) ptr, 0);
v = LLVMBuildIntToPtr(gallivm->builder, v,
LLVMPointerType(int_type, 0),
"cast int to ptr");
return v;
}

View File

@@ -89,12 +89,13 @@
* return { i32, i32, i32, i32 } where each value is in [0, 2^dst_width-1].
*/
LLVMValueRef
lp_build_clamped_float_to_unsigned_norm(LLVMBuilderRef builder,
lp_build_clamped_float_to_unsigned_norm(struct gallivm_state *gallivm,
struct lp_type src_type,
unsigned dst_width,
LLVMValueRef src)
{
LLVMTypeRef int_vec_type = lp_build_int_vec_type(src_type);
LLVMBuilderRef builder = gallivm->builder;
LLVMTypeRef int_vec_type = lp_build_int_vec_type(gallivm, src_type);
LLVMValueRef res;
unsigned mantissa;
@@ -122,10 +123,11 @@ lp_build_clamped_float_to_unsigned_norm(LLVMBuilderRef builder,
scale = (double)mask/ubound;
bias = (double)(1ULL << (mantissa - dst_width));
res = LLVMBuildFMul(builder, src, lp_build_const_vec(src_type, scale), "");
res = LLVMBuildFAdd(builder, res, lp_build_const_vec(src_type, bias), "");
res = LLVMBuildFMul(builder, src, lp_build_const_vec(gallivm, src_type, scale), "");
res = LLVMBuildFAdd(builder, res, lp_build_const_vec(gallivm, src_type, bias), "");
res = LLVMBuildBitCast(builder, res, int_vec_type, "");
res = LLVMBuildAnd(builder, res, lp_build_const_int_vec(src_type, mask), "");
res = LLVMBuildAnd(builder, res,
lp_build_const_int_vec(gallivm, src_type, mask), "");
}
else if (dst_width == (mantissa + 1)) {
/*
@@ -138,7 +140,8 @@ lp_build_clamped_float_to_unsigned_norm(LLVMBuilderRef builder,
scale = (double)((1ULL << dst_width) - 1);
res = LLVMBuildFMul(builder, src, lp_build_const_vec(src_type, scale), "");
res = LLVMBuildFMul(builder, src,
lp_build_const_vec(gallivm, src_type, scale), "");
res = LLVMBuildFPToSI(builder, res, int_vec_type, "");
}
else {
@@ -166,7 +169,8 @@ lp_build_clamped_float_to_unsigned_norm(LLVMBuilderRef builder,
LLVMValueRef lshifted;
LLVMValueRef rshifted;
res = LLVMBuildFMul(builder, src, lp_build_const_vec(src_type, scale), "");
res = LLVMBuildFMul(builder, src,
lp_build_const_vec(gallivm, src_type, scale), "");
res = LLVMBuildFPToSI(builder, res, int_vec_type, "");
/*
@@ -177,7 +181,8 @@ lp_build_clamped_float_to_unsigned_norm(LLVMBuilderRef builder,
*/
if (lshift) {
lshifted = LLVMBuildShl(builder, res,
lp_build_const_int_vec(src_type, lshift), "");
lp_build_const_int_vec(gallivm, src_type,
lshift), "");
} else {
lshifted = res;
}
@@ -186,7 +191,8 @@ lp_build_clamped_float_to_unsigned_norm(LLVMBuilderRef builder,
* Align the most significant bit to the right.
*/
rshifted = LLVMBuildAShr(builder, res,
lp_build_const_int_vec(src_type, rshift), "");
lp_build_const_int_vec(gallivm, src_type, rshift),
"");
/*
* Subtract the MSB to the LSB, therefore re-scaling from
@@ -206,13 +212,14 @@ lp_build_clamped_float_to_unsigned_norm(LLVMBuilderRef builder,
* return {float, float, float, float} with values in range [0, 1].
*/
LLVMValueRef
lp_build_unsigned_norm_to_float(LLVMBuilderRef builder,
lp_build_unsigned_norm_to_float(struct gallivm_state *gallivm,
unsigned src_width,
struct lp_type dst_type,
LLVMValueRef src)
{
LLVMTypeRef vec_type = lp_build_vec_type(dst_type);
LLVMTypeRef int_vec_type = lp_build_int_vec_type(dst_type);
LLVMBuilderRef builder = gallivm->builder;
LLVMTypeRef vec_type = lp_build_vec_type(gallivm, dst_type);
LLVMTypeRef int_vec_type = lp_build_int_vec_type(gallivm, dst_type);
LLVMValueRef bias_;
LLVMValueRef res;
unsigned mantissa;
@@ -230,7 +237,8 @@ lp_build_unsigned_norm_to_float(LLVMBuilderRef builder,
if (src_width == 8) {
scale = 1.0/255.0;
res = LLVMBuildSIToFP(builder, src, vec_type, "");
res = LLVMBuildFMul(builder, res, lp_build_const_vec(dst_type, scale), "");
res = LLVMBuildFMul(builder, res,
lp_build_const_vec(gallivm, dst_type, scale), "");
return res;
}
@@ -247,10 +255,11 @@ lp_build_unsigned_norm_to_float(LLVMBuilderRef builder,
if(src_width > mantissa) {
int shift = src_width - mantissa;
res = LLVMBuildLShr(builder, res, lp_build_const_int_vec(dst_type, shift), "");
res = LLVMBuildLShr(builder, res,
lp_build_const_int_vec(gallivm, dst_type, shift), "");
}
bias_ = lp_build_const_vec(dst_type, bias);
bias_ = lp_build_const_vec(gallivm, dst_type, bias);
res = LLVMBuildOr(builder,
res,
@@ -259,7 +268,7 @@ lp_build_unsigned_norm_to_float(LLVMBuilderRef builder,
res = LLVMBuildBitCast(builder, res, vec_type, "");
res = LLVMBuildFSub(builder, res, bias_, "");
res = LLVMBuildFMul(builder, res, lp_build_const_vec(dst_type, scale), "");
res = LLVMBuildFMul(builder, res, lp_build_const_vec(gallivm, dst_type, scale), "");
return res;
}
@@ -272,12 +281,13 @@ lp_build_unsigned_norm_to_float(LLVMBuilderRef builder,
* to the lp_type union.
*/
void
lp_build_conv(LLVMBuilderRef builder,
lp_build_conv(struct gallivm_state *gallivm,
struct lp_type src_type,
struct lp_type dst_type,
const LLVMValueRef *src, unsigned num_srcs,
LLVMValueRef *dst, unsigned num_dsts)
{
LLVMBuilderRef builder = gallivm->builder;
struct lp_type tmp_type;
LLVMValueRef tmp[LP_MAX_VECTOR_LENGTH];
unsigned num_tmps;
@@ -342,12 +352,12 @@ lp_build_conv(LLVMBuilderRef builder,
int32_type.length /= 4;
int32_type.sign = 1;
src_vec_type = lp_build_vec_type(src_type);
dst_vec_type = lp_build_vec_type(dst_type);
int16_vec_type = lp_build_vec_type(int16_type);
int32_vec_type = lp_build_vec_type(int32_type);
src_vec_type = lp_build_vec_type(gallivm, src_type);
dst_vec_type = lp_build_vec_type(gallivm, dst_type);
int16_vec_type = lp_build_vec_type(gallivm, int16_type);
int32_vec_type = lp_build_vec_type(gallivm, int32_type);
const_255f = lp_build_const_vec(src_type, 255.0f);
const_255f = lp_build_const_vec(gallivm, src_type, 255.0f);
a = LLVMBuildFMul(builder, src[0], const_255f, "");
b = LLVMBuildFMul(builder, src[1], const_255f, "");
@@ -357,14 +367,14 @@ lp_build_conv(LLVMBuilderRef builder,
{
struct lp_build_context bld;
bld.builder = builder;
bld.gallivm = gallivm;
bld.type = src_type;
bld.vec_type = src_vec_type;
bld.int_elem_type = lp_build_elem_type(int32_type);
bld.int_elem_type = lp_build_elem_type(gallivm, int32_type);
bld.int_vec_type = int32_vec_type;
bld.undef = lp_build_undef(src_type);
bld.zero = lp_build_zero(src_type);
bld.one = lp_build_one(src_type);
bld.undef = lp_build_undef(gallivm, src_type);
bld.zero = lp_build_zero(gallivm, src_type);
bld.one = lp_build_one(gallivm, src_type);
src_int0 = lp_build_iround(&bld, a);
src_int1 = lp_build_iround(&bld, b);
@@ -372,9 +382,9 @@ lp_build_conv(LLVMBuilderRef builder,
src_int3 = lp_build_iround(&bld, d);
}
/* relying on clamping behavior of sse2 intrinsics here */
lo = lp_build_pack2(builder, int32_type, int16_type, src_int0, src_int1);
hi = lp_build_pack2(builder, int32_type, int16_type, src_int2, src_int3);
dst[i] = lp_build_pack2(builder, int16_type, dst_type, lo, hi);
lo = lp_build_pack2(gallivm, int32_type, int16_type, src_int0, src_int1);
hi = lp_build_pack2(gallivm, int32_type, int16_type, src_int2, src_int3);
dst[i] = lp_build_pack2(gallivm, int16_type, dst_type, lo, hi);
}
return;
}
@@ -391,13 +401,13 @@ lp_build_conv(LLVMBuilderRef builder,
double dst_max = lp_const_max(dst_type);
LLVMValueRef thres;
lp_build_context_init(&bld, builder, tmp_type);
lp_build_context_init(&bld, gallivm, tmp_type);
if(src_min < dst_min) {
if(dst_min == 0.0)
thres = bld.zero;
else
thres = lp_build_const_vec(src_type, dst_min);
thres = lp_build_const_vec(gallivm, src_type, dst_min);
for(i = 0; i < num_tmps; ++i)
tmp[i] = lp_build_max(&bld, tmp[i], thres);
}
@@ -406,7 +416,7 @@ lp_build_conv(LLVMBuilderRef builder,
if(dst_max == 1.0)
thres = bld.one;
else
thres = lp_build_const_vec(src_type, dst_max);
thres = lp_build_const_vec(gallivm, src_type, dst_max);
for(i = 0; i < num_tmps; ++i)
tmp[i] = lp_build_min(&bld, tmp[i], thres);
}
@@ -422,7 +432,7 @@ lp_build_conv(LLVMBuilderRef builder,
else if(tmp_type.floating) {
if(!dst_type.fixed && !dst_type.sign && dst_type.norm) {
for(i = 0; i < num_tmps; ++i) {
tmp[i] = lp_build_clamped_float_to_unsigned_norm(builder,
tmp[i] = lp_build_clamped_float_to_unsigned_norm(gallivm,
tmp_type,
dst_type.width,
tmp[i]);
@@ -434,14 +444,14 @@ lp_build_conv(LLVMBuilderRef builder,
LLVMTypeRef tmp_vec_type;
if (dst_scale != 1.0) {
LLVMValueRef scale = lp_build_const_vec(tmp_type, dst_scale);
LLVMValueRef scale = lp_build_const_vec(gallivm, tmp_type, dst_scale);
for(i = 0; i < num_tmps; ++i)
tmp[i] = LLVMBuildFMul(builder, tmp[i], scale, "");
}
/* Use an equally sized integer for intermediate computations */
tmp_type.floating = FALSE;
tmp_vec_type = lp_build_vec_type(tmp_type);
tmp_vec_type = lp_build_vec_type(gallivm, tmp_type);
for(i = 0; i < num_tmps; ++i) {
#if 0
if(dst_type.sign)
@@ -461,7 +471,8 @@ lp_build_conv(LLVMBuilderRef builder,
/* FIXME: compensate different offsets too */
if(src_shift > dst_shift) {
LLVMValueRef shift = lp_build_const_int_vec(tmp_type, src_shift - dst_shift);
LLVMValueRef shift = lp_build_const_int_vec(gallivm, tmp_type,
src_shift - dst_shift);
for(i = 0; i < num_tmps; ++i)
if(src_type.sign)
tmp[i] = LLVMBuildAShr(builder, tmp[i], shift, "");
@@ -485,7 +496,7 @@ lp_build_conv(LLVMBuilderRef builder,
new_type.width = dst_type.width;
new_type.length = dst_type.length;
lp_build_resize(builder, tmp_type, new_type, tmp, num_srcs, tmp, num_dsts);
lp_build_resize(gallivm, tmp_type, new_type, tmp, num_srcs, tmp, num_dsts);
tmp_type = new_type;
num_tmps = num_dsts;
@@ -501,7 +512,7 @@ lp_build_conv(LLVMBuilderRef builder,
else if(!src_type.floating && dst_type.floating) {
if(!src_type.fixed && !src_type.sign && src_type.norm) {
for(i = 0; i < num_tmps; ++i) {
tmp[i] = lp_build_unsigned_norm_to_float(builder,
tmp[i] = lp_build_unsigned_norm_to_float(gallivm,
src_type.width,
dst_type,
tmp[i]);
@@ -515,7 +526,7 @@ lp_build_conv(LLVMBuilderRef builder,
/* Use an equally sized integer for intermediate computations */
tmp_type.floating = TRUE;
tmp_type.sign = TRUE;
tmp_vec_type = lp_build_vec_type(tmp_type);
tmp_vec_type = lp_build_vec_type(gallivm, tmp_type);
for(i = 0; i < num_tmps; ++i) {
#if 0
if(dst_type.sign)
@@ -529,7 +540,7 @@ lp_build_conv(LLVMBuilderRef builder,
}
if (src_scale != 1.0) {
LLVMValueRef scale = lp_build_const_vec(tmp_type, 1.0/src_scale);
LLVMValueRef scale = lp_build_const_vec(gallivm, tmp_type, 1.0/src_scale);
for(i = 0; i < num_tmps; ++i)
tmp[i] = LLVMBuildFMul(builder, tmp[i], scale, "");
}
@@ -541,7 +552,7 @@ lp_build_conv(LLVMBuilderRef builder,
/* FIXME: compensate different offsets too */
if(src_shift < dst_shift) {
LLVMValueRef shift = lp_build_const_int_vec(tmp_type, dst_shift - src_shift);
LLVMValueRef shift = lp_build_const_int_vec(gallivm, tmp_type, dst_shift - src_shift);
for(i = 0; i < num_tmps; ++i)
tmp[i] = LLVMBuildShl(builder, tmp[i], shift, "");
}
@@ -565,7 +576,7 @@ lp_build_conv(LLVMBuilderRef builder,
* This is basically a very trimmed down version of lp_build_conv.
*/
void
lp_build_conv_mask(LLVMBuilderRef builder,
lp_build_conv_mask(struct gallivm_state *gallivm,
struct lp_type src_type,
struct lp_type dst_type,
const LLVMValueRef *src, unsigned num_srcs,
@@ -599,11 +610,11 @@ lp_build_conv_mask(LLVMBuilderRef builder,
if(src_type.width > dst_type.width) {
assert(num_dsts == 1);
dst[0] = lp_build_pack(builder, src_type, dst_type, TRUE, src, num_srcs);
dst[0] = lp_build_pack(gallivm, src_type, dst_type, TRUE, src, num_srcs);
}
else if(src_type.width < dst_type.width) {
assert(num_srcs == 1);
lp_build_unpack(builder, src_type, dst_type, src[0], dst, num_dsts);
lp_build_unpack(gallivm, src_type, dst_type, src[0], dst, num_dsts);
}
else {
assert(num_srcs == num_dsts);

View File

@@ -44,27 +44,27 @@ struct lp_type;
LLVMValueRef
lp_build_clamped_float_to_unsigned_norm(LLVMBuilderRef builder,
lp_build_clamped_float_to_unsigned_norm(struct gallivm_state *gallivm,
struct lp_type src_type,
unsigned dst_width,
LLVMValueRef src);
LLVMValueRef
lp_build_unsigned_norm_to_float(LLVMBuilderRef builder,
lp_build_unsigned_norm_to_float(struct gallivm_state *gallivm,
unsigned src_width,
struct lp_type dst_type,
LLVMValueRef src);
void
lp_build_conv(LLVMBuilderRef builder,
lp_build_conv(struct gallivm_state *gallivm,
struct lp_type src_type,
struct lp_type dst_type,
const LLVMValueRef *srcs, unsigned num_srcs,
LLVMValueRef *dsts, unsigned num_dsts);
void
lp_build_conv_mask(LLVMBuilderRef builder,
lp_build_conv_mask(struct gallivm_state *gallivm,
struct lp_type src_type,
struct lp_type dst_type,
const LLVMValueRef *src, unsigned num_srcs,

View File

@@ -42,6 +42,7 @@
#define GALLIVM_DEBUG_NO_OPT (1 << 3)
#define GALLIVM_DEBUG_PERF (1 << 4)
#define GALLIVM_DEBUG_NO_BRILINEAR (1 << 5)
#define GALLIVM_DEBUG_GC (1 << 6)
#ifdef DEBUG

View File

@@ -34,6 +34,7 @@
#include "util/u_debug.h"
#include "util/u_memory.h"
#include "lp_bld_init.h"
#include "lp_bld_type.h"
#include "lp_bld_flow.h"
@@ -51,25 +52,25 @@
* be used elsewhere.
*/
LLVMBasicBlockRef
lp_build_insert_new_block(LLVMBuilderRef builder, const char *name)
lp_build_insert_new_block(struct gallivm_state *gallivm, const char *name)
{
LLVMBasicBlockRef current_block;
LLVMBasicBlockRef next_block;
LLVMBasicBlockRef new_block;
/* get current basic block */
current_block = LLVMGetInsertBlock(builder);
current_block = LLVMGetInsertBlock(gallivm->builder);
/* check if there's another block after this one */
next_block = LLVMGetNextBasicBlock(current_block);
if (next_block) {
/* insert the new block before the next block */
new_block = LLVMInsertBasicBlock(next_block, name);
new_block = LLVMInsertBasicBlockInContext(gallivm->context, next_block, name);
}
else {
/* append new block after current block */
LLVMValueRef function = LLVMGetBasicBlockParent(current_block);
new_block = LLVMAppendBasicBlock(function, name);
new_block = LLVMAppendBasicBlockInContext(gallivm->context, function, name);
}
return new_block;
@@ -82,12 +83,11 @@ lp_build_insert_new_block(LLVMBuilderRef builder, const char *name)
*/
void
lp_build_flow_skip_begin(struct lp_build_skip_context *skip,
LLVMBuilderRef builder)
struct gallivm_state *gallivm)
{
skip->builder = builder;
skip->gallivm = gallivm;
/* create new basic block */
skip->block = lp_build_insert_new_block(skip->builder, "skip");
skip->block = lp_build_insert_new_block(gallivm, "skip");
}
@@ -101,12 +101,12 @@ lp_build_flow_skip_cond_break(struct lp_build_skip_context *skip,
{
LLVMBasicBlockRef new_block;
new_block = lp_build_insert_new_block(skip->builder, "");
new_block = lp_build_insert_new_block(skip->gallivm, "");
/* if cond is true, goto skip->block, else goto new_block */
LLVMBuildCondBr(skip->builder, cond, skip->block, new_block);
LLVMBuildCondBr(skip->gallivm->builder, cond, skip->block, new_block);
LLVMPositionBuilderAtEnd(skip->builder, new_block);
LLVMPositionBuilderAtEnd(skip->gallivm->builder, new_block);
}
@@ -114,8 +114,8 @@ void
lp_build_flow_skip_end(struct lp_build_skip_context *skip)
{
/* goto block */
LLVMBuildBr(skip->builder, skip->block);
LLVMPositionBuilderAtEnd(skip->builder, skip->block);
LLVMBuildBr(skip->gallivm->builder, skip->block);
LLVMPositionBuilderAtEnd(skip->gallivm->builder, skip->block);
}
@@ -125,7 +125,7 @@ lp_build_flow_skip_end(struct lp_build_skip_context *skip)
void
lp_build_mask_check(struct lp_build_mask_context *mask)
{
LLVMBuilderRef builder = mask->skip.builder;
LLVMBuilderRef builder = mask->skip.gallivm->builder;
LLVMValueRef value;
LLVMValueRef cond;
@@ -152,27 +152,27 @@ lp_build_mask_check(struct lp_build_mask_context *mask)
*/
void
lp_build_mask_begin(struct lp_build_mask_context *mask,
LLVMBuilderRef builder,
struct gallivm_state *gallivm,
struct lp_type type,
LLVMValueRef value)
{
memset(mask, 0, sizeof *mask);
mask->reg_type = LLVMIntType(type.width * type.length);
mask->var = lp_build_alloca(builder,
lp_build_int_vec_type(type),
mask->reg_type = LLVMIntTypeInContext(gallivm->context, type.width * type.length);
mask->var = lp_build_alloca(gallivm,
lp_build_int_vec_type(gallivm, type),
"execution_mask");
LLVMBuildStore(builder, value, mask->var);
LLVMBuildStore(gallivm->builder, value, mask->var);
lp_build_flow_skip_begin(&mask->skip, builder);
lp_build_flow_skip_begin(&mask->skip, gallivm);
}
LLVMValueRef
lp_build_mask_value(struct lp_build_mask_context *mask)
{
return LLVMBuildLoad(mask->skip.builder, mask->var, "");
return LLVMBuildLoad(mask->skip.gallivm->builder, mask->var, "");
}
@@ -185,10 +185,10 @@ void
lp_build_mask_update(struct lp_build_mask_context *mask,
LLVMValueRef value)
{
value = LLVMBuildAnd(mask->skip.builder,
value = LLVMBuildAnd(mask->skip.gallivm->builder,
lp_build_mask_value(mask),
value, "");
LLVMBuildStore(mask->skip.builder, value, mask->var);
LLVMBuildStore(mask->skip.gallivm->builder, value, mask->var);
}
@@ -205,13 +205,17 @@ lp_build_mask_end(struct lp_build_mask_context *mask)
void
lp_build_loop_begin(LLVMBuilderRef builder,
LLVMValueRef start,
struct lp_build_loop_state *state)
lp_build_loop_begin(struct lp_build_loop_state *state,
struct gallivm_state *gallivm,
LLVMValueRef start)
{
state->block = lp_build_insert_new_block(builder, "loop_begin");
LLVMBuilderRef builder = gallivm->builder;
state->counter_var = lp_build_alloca(builder, LLVMTypeOf(start), "loop_counter");
state->block = lp_build_insert_new_block(gallivm, "loop_begin");
state->counter_var = lp_build_alloca(gallivm, LLVMTypeOf(start), "loop_counter");
state->gallivm = gallivm;
LLVMBuildStore(builder, start, state->counter_var);
@@ -224,12 +228,12 @@ lp_build_loop_begin(LLVMBuilderRef builder,
void
lp_build_loop_end_cond(LLVMBuilderRef builder,
lp_build_loop_end_cond(struct lp_build_loop_state *state,
LLVMValueRef end,
LLVMValueRef step,
LLVMIntPredicate llvm_cond,
struct lp_build_loop_state *state)
LLVMIntPredicate llvm_cond)
{
LLVMBuilderRef builder = state->gallivm->builder;
LLVMValueRef next;
LLVMValueRef cond;
LLVMBasicBlockRef after_block;
@@ -243,7 +247,7 @@ lp_build_loop_end_cond(LLVMBuilderRef builder,
cond = LLVMBuildICmp(builder, llvm_cond, next, end, "");
after_block = lp_build_insert_new_block(builder, "loop_end");
after_block = lp_build_insert_new_block(state->gallivm, "loop_end");
LLVMBuildCondBr(builder, cond, after_block, state->block);
@@ -254,12 +258,11 @@ lp_build_loop_end_cond(LLVMBuilderRef builder,
void
lp_build_loop_end(LLVMBuilderRef builder,
lp_build_loop_end(struct lp_build_loop_state *state,
LLVMValueRef end,
LLVMValueRef step,
struct lp_build_loop_state *state)
LLVMValueRef step)
{
lp_build_loop_end_cond(builder, end, step, LLVMIntNE, state);
lp_build_loop_end_cond(state, end, step, LLVMIntNE);
}
@@ -296,24 +299,27 @@ lp_build_loop_end(LLVMBuilderRef builder,
*/
void
lp_build_if(struct lp_build_if_state *ifthen,
LLVMBuilderRef builder,
struct gallivm_state *gallivm,
LLVMValueRef condition)
{
LLVMBasicBlockRef block = LLVMGetInsertBlock(builder);
LLVMBasicBlockRef block = LLVMGetInsertBlock(gallivm->builder);
memset(ifthen, 0, sizeof *ifthen);
ifthen->builder = builder;
ifthen->gallivm = gallivm;
ifthen->condition = condition;
ifthen->entry_block = block;
/* create endif/merge basic block for the phi functions */
ifthen->merge_block = lp_build_insert_new_block(builder, "endif-block");
ifthen->merge_block = lp_build_insert_new_block(gallivm, "endif-block");
/* create/insert true_block before merge_block */
ifthen->true_block = LLVMInsertBasicBlock(ifthen->merge_block, "if-true-block");
ifthen->true_block =
LLVMInsertBasicBlockInContext(gallivm->context,
ifthen->merge_block,
"if-true-block");
/* successive code goes into the true block */
LLVMPositionBuilderAtEnd(builder, ifthen->true_block);
LLVMPositionBuilderAtEnd(gallivm->builder, ifthen->true_block);
}
@@ -323,14 +329,19 @@ lp_build_if(struct lp_build_if_state *ifthen,
void
lp_build_else(struct lp_build_if_state *ifthen)
{
LLVMBuilderRef builder = ifthen->gallivm->builder;
/* Append an unconditional Br(anch) instruction on the true_block */
LLVMBuildBr(ifthen->builder, ifthen->merge_block);
LLVMBuildBr(builder, ifthen->merge_block);
/* create/insert false_block before the merge block */
ifthen->false_block = LLVMInsertBasicBlock(ifthen->merge_block, "if-false-block");
ifthen->false_block =
LLVMInsertBasicBlockInContext(ifthen->gallivm->context,
ifthen->merge_block,
"if-false-block");
/* successive code goes into the else block */
LLVMPositionBuilderAtEnd(ifthen->builder, ifthen->false_block);
LLVMPositionBuilderAtEnd(builder, ifthen->false_block);
}
@@ -340,28 +351,30 @@ lp_build_else(struct lp_build_if_state *ifthen)
void
lp_build_endif(struct lp_build_if_state *ifthen)
{
LLVMBuilderRef builder = ifthen->gallivm->builder;
/* Insert branch to the merge block from current block */
LLVMBuildBr(ifthen->builder, ifthen->merge_block);
LLVMBuildBr(builder, ifthen->merge_block);
/*
* Now patch in the various branch instructions.
*/
/* Insert the conditional branch instruction at the end of entry_block */
LLVMPositionBuilderAtEnd(ifthen->builder, ifthen->entry_block);
LLVMPositionBuilderAtEnd(builder, ifthen->entry_block);
if (ifthen->false_block) {
/* we have an else clause */
LLVMBuildCondBr(ifthen->builder, ifthen->condition,
LLVMBuildCondBr(builder, ifthen->condition,
ifthen->true_block, ifthen->false_block);
}
else {
/* no else clause */
LLVMBuildCondBr(ifthen->builder, ifthen->condition,
LLVMBuildCondBr(builder, ifthen->condition,
ifthen->true_block, ifthen->merge_block);
}
/* Resume building code at end of the ifthen->merge_block */
LLVMPositionBuilderAtEnd(ifthen->builder, ifthen->merge_block);
LLVMPositionBuilderAtEnd(builder, ifthen->merge_block);
}
@@ -381,15 +394,16 @@ lp_build_endif(struct lp_build_if_state *ifthen)
* - http://www.llvm.org/docs/tutorial/OCamlLangImpl7.html#memory
*/
LLVMValueRef
lp_build_alloca(LLVMBuilderRef builder,
lp_build_alloca(struct gallivm_state *gallivm,
LLVMTypeRef type,
const char *name)
{
LLVMBuilderRef builder = gallivm->builder;
LLVMBasicBlockRef current_block = LLVMGetInsertBlock(builder);
LLVMValueRef function = LLVMGetBasicBlockParent(current_block);
LLVMBasicBlockRef first_block = LLVMGetEntryBasicBlock(function);
LLVMValueRef first_instr = LLVMGetFirstInstruction(first_block);
LLVMBuilderRef first_builder = LLVMCreateBuilder();
LLVMBuilderRef first_builder = LLVMCreateBuilderInContext(gallivm->context);
LLVMValueRef res;
if (first_instr) {
@@ -422,16 +436,17 @@ lp_build_alloca(LLVMBuilderRef builder,
* - http://www.llvm.org/docs/tutorial/OCamlLangImpl7.html#memory
*/
LLVMValueRef
lp_build_array_alloca(LLVMBuilderRef builder,
lp_build_array_alloca(struct gallivm_state *gallivm,
LLVMTypeRef type,
LLVMValueRef count,
const char *name)
{
LLVMBuilderRef builder = gallivm->builder;
LLVMBasicBlockRef current_block = LLVMGetInsertBlock(builder);
LLVMValueRef function = LLVMGetBasicBlockParent(current_block);
LLVMBasicBlockRef first_block = LLVMGetEntryBasicBlock(function);
LLVMValueRef first_instr = LLVMGetFirstInstruction(first_block);
LLVMBuilderRef first_builder = LLVMCreateBuilder();
LLVMBuilderRef first_builder = LLVMCreateBuilderInContext(gallivm->context);
LLVMValueRef res;
if (first_instr) {

View File

@@ -47,7 +47,7 @@ struct lp_type;
*/
struct lp_build_skip_context
{
LLVMBuilderRef builder;
struct gallivm_state *gallivm;
/** Block to skip to */
LLVMBasicBlockRef block;
@@ -55,7 +55,7 @@ struct lp_build_skip_context
void
lp_build_flow_skip_begin(struct lp_build_skip_context *ctx,
LLVMBuilderRef builder);
struct gallivm_state *gallivm);
void
lp_build_flow_skip_cond_break(struct lp_build_skip_context *ctx,
@@ -77,7 +77,7 @@ struct lp_build_mask_context
void
lp_build_mask_begin(struct lp_build_mask_context *mask,
LLVMBuilderRef builder,
struct gallivm_state *gallivm,
struct lp_type type,
LLVMValueRef value);
@@ -107,31 +107,28 @@ lp_build_mask_end(struct lp_build_mask_context *mask);
*/
struct lp_build_loop_state
{
LLVMBasicBlockRef block;
LLVMValueRef counter_var;
LLVMValueRef counter;
LLVMBasicBlockRef block;
LLVMValueRef counter_var;
LLVMValueRef counter;
struct gallivm_state *gallivm;
};
void
lp_build_loop_begin(LLVMBuilderRef builder,
LLVMValueRef start,
struct lp_build_loop_state *state);
lp_build_loop_begin(struct lp_build_loop_state *state,
struct gallivm_state *gallivm,
LLVMValueRef start);
void
lp_build_loop_end(LLVMBuilderRef builder,
lp_build_loop_end(struct lp_build_loop_state *state,
LLVMValueRef end,
LLVMValueRef step,
struct lp_build_loop_state *state);
LLVMValueRef step);
void
lp_build_loop_end_cond(LLVMBuilderRef builder,
lp_build_loop_end_cond(struct lp_build_loop_state *state,
LLVMValueRef end,
LLVMValueRef step,
LLVMIntPredicate cond,
struct lp_build_loop_state *state);
LLVMIntPredicate cond);
@@ -140,7 +137,7 @@ lp_build_loop_end_cond(LLVMBuilderRef builder,
*/
struct lp_build_if_state
{
LLVMBuilderRef builder;
struct gallivm_state *gallivm;
LLVMValueRef condition;
LLVMBasicBlockRef entry_block;
LLVMBasicBlockRef true_block;
@@ -151,7 +148,7 @@ struct lp_build_if_state
void
lp_build_if(struct lp_build_if_state *ctx,
LLVMBuilderRef builder,
struct gallivm_state *gallivm,
LLVMValueRef condition);
void
@@ -161,15 +158,15 @@ void
lp_build_endif(struct lp_build_if_state *ctx);
LLVMBasicBlockRef
lp_build_insert_new_block(LLVMBuilderRef builder, const char *name);
lp_build_insert_new_block(struct gallivm_state *gallivm, const char *name);
LLVMValueRef
lp_build_alloca(LLVMBuilderRef builder,
lp_build_alloca(struct gallivm_state *gallivm,
LLVMTypeRef type,
const char *name);
LLVMValueRef
lp_build_array_alloca(LLVMBuilderRef builder,
lp_build_array_alloca(struct gallivm_state *gallivm,
LLVMTypeRef type,
LLVMValueRef count,
const char *name);

View File

@@ -35,6 +35,7 @@
*/
#include "gallivm/lp_bld.h"
#include "gallivm/lp_bld_init.h"
#include "pipe/p_format.h"
@@ -53,12 +54,12 @@ lp_build_format_swizzle_aos(const struct util_format_description *desc,
LLVMValueRef unswizzled);
LLVMValueRef
lp_build_pack_rgba_aos(LLVMBuilderRef builder,
lp_build_pack_rgba_aos(struct gallivm_state *gallivm,
const struct util_format_description *desc,
LLVMValueRef rgba);
LLVMValueRef
lp_build_fetch_rgba_aos(LLVMBuilderRef builder,
lp_build_fetch_rgba_aos(struct gallivm_state *gallivm,
const struct util_format_description *format_desc,
struct lp_type type,
LLVMValueRef base_ptr,
@@ -78,20 +79,20 @@ lp_build_format_swizzle_soa(const struct util_format_description *format_desc,
LLVMValueRef swizzled_out[4]);
void
lp_build_unpack_rgba_soa(LLVMBuilderRef builder,
lp_build_unpack_rgba_soa(struct gallivm_state *gallivm,
const struct util_format_description *format_desc,
struct lp_type type,
LLVMValueRef packed,
LLVMValueRef rgba_out[4]);
void
lp_build_rgba8_to_f32_soa(LLVMBuilderRef builder,
lp_build_rgba8_to_f32_soa(struct gallivm_state *gallivm,
struct lp_type dst_type,
LLVMValueRef packed,
LLVMValueRef *rgba);
void
lp_build_fetch_rgba_soa(LLVMBuilderRef builder,
lp_build_fetch_rgba_soa(struct gallivm_state *gallivm,
const struct util_format_description *format_desc,
struct lp_type type,
LLVMValueRef base_ptr,
@@ -106,7 +107,7 @@ lp_build_fetch_rgba_soa(LLVMBuilderRef builder,
LLVMValueRef
lp_build_fetch_subsampled_rgba_aos(LLVMBuilderRef builder,
lp_build_fetch_subsampled_rgba_aos(struct gallivm_state *gallivm,
const struct util_format_description *format_desc,
unsigned n,
LLVMValueRef base_ptr,

View File

@@ -36,6 +36,7 @@
#include "util/u_format.h"
#include "util/u_memory.h"
#include "util/u_math.h"
#include "util/u_pointer.h"
#include "util/u_string.h"
#include "lp_bld_arit.h"
@@ -145,10 +146,11 @@ format_matches_type(const struct util_format_description *desc,
* @return RGBA in a float[4] or ubyte[4] or ushort[4] vector.
*/
static INLINE LLVMValueRef
lp_build_unpack_arith_rgba_aos(LLVMBuilderRef builder,
lp_build_unpack_arith_rgba_aos(struct gallivm_state *gallivm,
const struct util_format_description *desc,
LLVMValueRef packed)
{
LLVMBuilderRef builder = gallivm->builder;
LLVMValueRef shifted, casted, scaled, masked;
LLVMValueRef shifts[4];
LLVMValueRef masks[4];
@@ -167,21 +169,21 @@ lp_build_unpack_arith_rgba_aos(LLVMBuilderRef builder,
/* Do the intermediate integer computations with 32bit integers since it
* matches floating point size */
assert (LLVMTypeOf(packed) == LLVMInt32Type());
assert (LLVMTypeOf(packed) == LLVMInt32TypeInContext(gallivm->context));
/* Broadcast the packed value to all four channels
* before: packed = BGRA
* after: packed = {BGRA, BGRA, BGRA, BGRA}
*/
packed = LLVMBuildInsertElement(builder,
LLVMGetUndef(LLVMVectorType(LLVMInt32Type(), 4)),
LLVMGetUndef(LLVMVectorType(LLVMInt32TypeInContext(gallivm->context), 4)),
packed,
LLVMConstNull(LLVMInt32Type()),
LLVMConstNull(LLVMInt32TypeInContext(gallivm->context)),
"");
packed = LLVMBuildShuffleVector(builder,
packed,
LLVMGetUndef(LLVMVectorType(LLVMInt32Type(), 4)),
LLVMConstNull(LLVMVectorType(LLVMInt32Type(), 4)),
LLVMGetUndef(LLVMVectorType(LLVMInt32TypeInContext(gallivm->context), 4)),
LLVMConstNull(LLVMVectorType(LLVMInt32TypeInContext(gallivm->context), 4)),
"");
/* Initialize vector constants */
@@ -194,9 +196,9 @@ lp_build_unpack_arith_rgba_aos(LLVMBuilderRef builder,
unsigned bits = desc->channel[i].size;
if (desc->channel[i].type == UTIL_FORMAT_TYPE_VOID) {
shifts[i] = LLVMGetUndef(LLVMInt32Type());
masks[i] = LLVMConstNull(LLVMInt32Type());
scales[i] = LLVMConstNull(LLVMFloatType());
shifts[i] = LLVMGetUndef(LLVMInt32TypeInContext(gallivm->context));
masks[i] = LLVMConstNull(LLVMInt32TypeInContext(gallivm->context));
scales[i] = LLVMConstNull(LLVMFloatTypeInContext(gallivm->context));
}
else {
unsigned long long mask = (1ULL << bits) - 1;
@@ -207,15 +209,15 @@ lp_build_unpack_arith_rgba_aos(LLVMBuilderRef builder,
needs_uitofp = TRUE;
}
shifts[i] = LLVMConstInt(LLVMInt32Type(), shift, 0);
masks[i] = LLVMConstInt(LLVMInt32Type(), mask, 0);
shifts[i] = lp_build_const_int32(gallivm, shift);
masks[i] = lp_build_const_int32(gallivm, mask);
if (desc->channel[i].normalized) {
scales[i] = LLVMConstReal(LLVMFloatType(), 1.0/mask);
scales[i] = lp_build_const_float(gallivm, 1.0 / mask);
normalized = TRUE;
}
else
scales[i] = LLVMConstReal(LLVMFloatType(), 1.0);
scales[i] = lp_build_const_float(gallivm, 1.0);
}
shift += bits;
@@ -230,9 +232,9 @@ lp_build_unpack_arith_rgba_aos(LLVMBuilderRef builder,
if (!needs_uitofp) {
/* UIToFP can't be expressed in SSE2 */
casted = LLVMBuildSIToFP(builder, masked, LLVMVectorType(LLVMFloatType(), 4), "");
casted = LLVMBuildSIToFP(builder, masked, LLVMVectorType(LLVMFloatTypeInContext(gallivm->context), 4), "");
} else {
casted = LLVMBuildUIToFP(builder, masked, LLVMVectorType(LLVMFloatType(), 4), "");
casted = LLVMBuildUIToFP(builder, masked, LLVMVectorType(LLVMFloatTypeInContext(gallivm->context), 4), "");
}
/* At this point 'casted' may be a vector of floats such as
@@ -258,10 +260,11 @@ lp_build_unpack_arith_rgba_aos(LLVMBuilderRef builder,
* a time is rarely if ever needed.
*/
LLVMValueRef
lp_build_pack_rgba_aos(LLVMBuilderRef builder,
lp_build_pack_rgba_aos(struct gallivm_state *gallivm,
const struct util_format_description *desc,
LLVMValueRef rgba)
{
LLVMBuilderRef builder = gallivm->builder;
LLVMTypeRef type;
LLVMValueRef packed = NULL;
LLVMValueRef swizzles[4];
@@ -276,7 +279,7 @@ lp_build_pack_rgba_aos(LLVMBuilderRef builder,
assert(desc->block.width == 1);
assert(desc->block.height == 1);
type = LLVMIntType(desc->block.bits);
type = LLVMIntTypeInContext(gallivm->context, desc->block.bits);
/* Unswizzle the color components into the source vector. */
for (i = 0; i < 4; ++i) {
@@ -285,13 +288,13 @@ lp_build_pack_rgba_aos(LLVMBuilderRef builder,
break;
}
if (j < 4)
swizzles[i] = LLVMConstInt(LLVMInt32Type(), j, 0);
swizzles[i] = lp_build_const_int32(gallivm, j);
else
swizzles[i] = LLVMGetUndef(LLVMInt32Type());
swizzles[i] = LLVMGetUndef(LLVMInt32TypeInContext(gallivm->context));
}
unswizzled = LLVMBuildShuffleVector(builder, rgba,
LLVMGetUndef(LLVMVectorType(LLVMFloatType(), 4)),
LLVMGetUndef(LLVMVectorType(LLVMFloatTypeInContext(gallivm->context), 4)),
LLVMConstVector(swizzles, 4), "");
normalized = FALSE;
@@ -300,8 +303,8 @@ lp_build_pack_rgba_aos(LLVMBuilderRef builder,
unsigned bits = desc->channel[i].size;
if (desc->channel[i].type == UTIL_FORMAT_TYPE_VOID) {
shifts[i] = LLVMGetUndef(LLVMInt32Type());
scales[i] = LLVMGetUndef(LLVMFloatType());
shifts[i] = LLVMGetUndef(LLVMInt32TypeInContext(gallivm->context));
scales[i] = LLVMGetUndef(LLVMFloatTypeInContext(gallivm->context));
}
else {
unsigned mask = (1 << bits) - 1;
@@ -309,14 +312,14 @@ lp_build_pack_rgba_aos(LLVMBuilderRef builder,
assert(desc->channel[i].type == UTIL_FORMAT_TYPE_UNSIGNED);
assert(bits < 32);
shifts[i] = LLVMConstInt(LLVMInt32Type(), shift, 0);
shifts[i] = lp_build_const_int32(gallivm, shift);
if (desc->channel[i].normalized) {
scales[i] = LLVMConstReal(LLVMFloatType(), mask);
scales[i] = lp_build_const_float(gallivm, mask);
normalized = TRUE;
}
else
scales[i] = LLVMConstReal(LLVMFloatType(), 1.0);
scales[i] = lp_build_const_float(gallivm, 1.0);
}
shift += bits;
@@ -327,14 +330,15 @@ lp_build_pack_rgba_aos(LLVMBuilderRef builder,
else
scaled = unswizzled;
casted = LLVMBuildFPToSI(builder, scaled, LLVMVectorType(LLVMInt32Type(), 4), "");
casted = LLVMBuildFPToSI(builder, scaled, LLVMVectorType(LLVMInt32TypeInContext(gallivm->context), 4), "");
shifted = LLVMBuildShl(builder, casted, LLVMConstVector(shifts, 4), "");
/* Bitwise or all components */
for (i = 0; i < 4; ++i) {
if (desc->channel[i].type == UTIL_FORMAT_TYPE_UNSIGNED) {
LLVMValueRef component = LLVMBuildExtractElement(builder, shifted, LLVMConstInt(LLVMInt32Type(), i, 0), "");
LLVMValueRef component = LLVMBuildExtractElement(builder, shifted,
lp_build_const_int32(gallivm, i), "");
if (packed)
packed = LLVMBuildOr(builder, packed, component, "");
else
@@ -343,7 +347,7 @@ lp_build_pack_rgba_aos(LLVMBuilderRef builder,
}
if (!packed)
packed = LLVMGetUndef(LLVMInt32Type());
packed = LLVMGetUndef(LLVMInt32TypeInContext(gallivm->context));
if (desc->block.bits < 32)
packed = LLVMBuildTrunc(builder, packed, type, "");
@@ -364,7 +368,7 @@ lp_build_pack_rgba_aos(LLVMBuilderRef builder,
* \return a 4 element vector with the pixel's RGBA values.
*/
LLVMValueRef
lp_build_fetch_rgba_aos(LLVMBuilderRef builder,
lp_build_fetch_rgba_aos(struct gallivm_state *gallivm,
const struct util_format_description *format_desc,
struct lp_type type,
LLVMValueRef base_ptr,
@@ -372,13 +376,14 @@ lp_build_fetch_rgba_aos(LLVMBuilderRef builder,
LLVMValueRef i,
LLVMValueRef j)
{
LLVMBuilderRef builder = gallivm->builder;
unsigned num_pixels = type.length / 4;
struct lp_build_context bld;
assert(type.length <= LP_MAX_VECTOR_LENGTH);
assert(type.length % 4 == 0);
lp_build_context_init(&bld, builder, type);
lp_build_context_init(&bld, gallivm, type);
/*
* Trivial case
@@ -397,13 +402,14 @@ lp_build_fetch_rgba_aos(LLVMBuilderRef builder,
* scaling or converting.
*/
packed = lp_build_gather(builder, type.length/4,
packed = lp_build_gather(gallivm, type.length/4,
format_desc->block.bits, type.width*4,
base_ptr, offset);
assert(format_desc->block.bits <= type.width * type.length);
packed = LLVMBuildBitCast(builder, packed, lp_build_vec_type(type), "");
packed = LLVMBuildBitCast(gallivm->builder, packed,
lp_build_vec_type(gallivm, type), "");
return lp_build_format_swizzle_aos(format_desc, &bld, packed);
}
@@ -435,11 +441,12 @@ lp_build_fetch_rgba_aos(LLVMBuilderRef builder,
for (k = 0; k < num_pixels; ++k) {
LLVMValueRef packed;
packed = lp_build_gather_elem(builder, num_pixels,
packed = lp_build_gather_elem(gallivm, num_pixels,
format_desc->block.bits, 32,
base_ptr, offset, k);
tmps[k] = lp_build_unpack_arith_rgba_aos(builder, format_desc,
tmps[k] = lp_build_unpack_arith_rgba_aos(gallivm,
format_desc,
packed);
}
@@ -455,7 +462,7 @@ lp_build_fetch_rgba_aos(LLVMBuilderRef builder,
__FUNCTION__, format_desc->short_name);
}
lp_build_conv(builder,
lp_build_conv(gallivm,
lp_float32_vec4_type(),
type,
tmps, num_pixels, &res, 1);
@@ -476,14 +483,14 @@ lp_build_fetch_rgba_aos(LLVMBuilderRef builder,
tmp_type.length = num_pixels * 4;
tmp_type.norm = TRUE;
tmp = lp_build_fetch_subsampled_rgba_aos(builder,
tmp = lp_build_fetch_subsampled_rgba_aos(gallivm,
format_desc,
num_pixels,
base_ptr,
offset,
i, j);
lp_build_conv(builder,
lp_build_conv(gallivm,
tmp_type, type,
&tmp, 1, &tmp, 1);
@@ -505,15 +512,16 @@ lp_build_fetch_rgba_aos(LLVMBuilderRef builder,
* or incentive to optimize.
*/
LLVMModuleRef module = LLVMGetGlobalParent(LLVMGetBasicBlockParent(LLVMGetInsertBlock(builder)));
LLVMModuleRef module = LLVMGetGlobalParent(LLVMGetBasicBlockParent(LLVMGetInsertBlock(gallivm->builder)));
char name[256];
LLVMTypeRef i8t = LLVMInt8Type();
LLVMTypeRef i8t = LLVMInt8TypeInContext(gallivm->context);
LLVMTypeRef pi8t = LLVMPointerType(i8t, 0);
LLVMTypeRef i32t = LLVMInt32Type();
LLVMTypeRef i32t = LLVMInt32TypeInContext(gallivm->context);
LLVMValueRef function;
LLVMValueRef tmp_ptr;
LLVMValueRef tmp;
LLVMValueRef res;
LLVMValueRef callee;
unsigned k;
util_snprintf(name, sizeof name, "util_format_%s_fetch_rgba_8unorm",
@@ -529,27 +537,38 @@ lp_build_fetch_rgba_aos(LLVMBuilderRef builder,
function = LLVMGetNamedFunction(module, name);
if (!function) {
/*
* Function to call looks like:
* fetch(uint8_t *dst, const uint8_t *src, unsigned i, unsigned j)
*/
LLVMTypeRef ret_type;
LLVMTypeRef arg_types[4];
LLVMTypeRef function_type;
ret_type = LLVMVoidType();
ret_type = LLVMVoidTypeInContext(gallivm->context);
arg_types[0] = pi8t;
arg_types[1] = pi8t;
arg_types[3] = arg_types[2] = LLVMIntType(sizeof(unsigned) * 8);
function_type = LLVMFunctionType(ret_type, arg_types, Elements(arg_types), 0);
arg_types[2] = i32t;
arg_types[3] = i32t;
function_type = LLVMFunctionType(ret_type, arg_types,
Elements(arg_types), 0);
function = LLVMAddFunction(module, name, function_type);
LLVMSetFunctionCallConv(function, LLVMCCallConv);
LLVMSetLinkage(function, LLVMExternalLinkage);
assert(LLVMIsDeclaration(function));
LLVMAddGlobalMapping(lp_build_engine, function,
func_to_pointer((func_pointer)format_desc->fetch_rgba_8unorm));
}
tmp_ptr = lp_build_alloca(builder, i32t, "");
/* make const pointer for the C fetch_rgba_float function */
callee = lp_build_const_int_pointer(gallivm,
func_to_pointer((func_pointer) format_desc->fetch_rgba_8unorm));
/* cast the callee pointer to the function's type */
function = LLVMBuildBitCast(builder, callee,
LLVMTypeOf(function), "cast callee");
tmp_ptr = lp_build_alloca(gallivm, i32t, "");
res = LLVMGetUndef(LLVMVectorType(i32t, num_pixels));
@@ -559,11 +578,11 @@ lp_build_fetch_rgba_aos(LLVMBuilderRef builder,
*/
for (k = 0; k < num_pixels; ++k) {
LLVMValueRef index = LLVMConstInt(LLVMInt32Type(), k, 0);
LLVMValueRef index = lp_build_const_int32(gallivm, k);
LLVMValueRef args[4];
args[0] = LLVMBuildBitCast(builder, tmp_ptr, pi8t, "");
args[1] = lp_build_gather_elem_ptr(builder, num_pixels,
args[1] = lp_build_gather_elem_ptr(gallivm, num_pixels,
base_ptr, offset, k);
if (num_pixels == 1) {
@@ -610,13 +629,16 @@ lp_build_fetch_rgba_aos(LLVMBuilderRef builder,
LLVMModuleRef module = LLVMGetGlobalParent(LLVMGetBasicBlockParent(LLVMGetInsertBlock(builder)));
char name[256];
LLVMTypeRef f32t = LLVMFloatType();
LLVMTypeRef f32t = LLVMFloatTypeInContext(gallivm->context);
LLVMTypeRef f32x4t = LLVMVectorType(f32t, 4);
LLVMTypeRef pf32t = LLVMPointerType(f32t, 0);
LLVMTypeRef pi8t = LLVMPointerType(LLVMInt8TypeInContext(gallivm->context), 0);
LLVMTypeRef i32t = LLVMInt32TypeInContext(gallivm->context);
LLVMValueRef function;
LLVMValueRef tmp_ptr;
LLVMValueRef tmps[LP_MAX_VECTOR_LENGTH/4];
LLVMValueRef res;
LLVMValueRef callee;
unsigned k;
util_snprintf(name, sizeof name, "util_format_%s_fetch_rgba_float",
@@ -632,27 +654,43 @@ lp_build_fetch_rgba_aos(LLVMBuilderRef builder,
function = LLVMGetNamedFunction(module, name);
if (!function) {
/*
* Function to call looks like:
* fetch(float *dst, const uint8_t *src, unsigned i, unsigned j)
*/
LLVMTypeRef ret_type;
LLVMTypeRef arg_types[4];
LLVMTypeRef function_type;
ret_type = LLVMVoidType();
ret_type = LLVMVoidTypeInContext(gallivm->context);
arg_types[0] = pf32t;
arg_types[1] = LLVMPointerType(LLVMInt8Type(), 0);
arg_types[3] = arg_types[2] = LLVMIntType(sizeof(unsigned) * 8);
function_type = LLVMFunctionType(ret_type, arg_types, Elements(arg_types), 0);
arg_types[1] = pi8t;
arg_types[2] = i32t;
arg_types[3] = i32t;
function_type = LLVMFunctionType(ret_type, arg_types,
Elements(arg_types), 0);
function = LLVMAddFunction(module, name, function_type);
LLVMSetFunctionCallConv(function, LLVMCCallConv);
LLVMSetLinkage(function, LLVMExternalLinkage);
assert(LLVMIsDeclaration(function));
LLVMAddGlobalMapping(lp_build_engine, function,
func_to_pointer((func_pointer)format_desc->fetch_rgba_float));
}
tmp_ptr = lp_build_alloca(builder, f32x4t, "");
/* Note: we're using this casting here instead of LLVMAddGlobalMapping()
* to work around a bug in LLVM 2.6.
*/
/* make const pointer for the C fetch_rgba_float function */
callee = lp_build_const_int_pointer(gallivm,
func_to_pointer((func_pointer) format_desc->fetch_rgba_float));
/* cast the callee pointer to the function's type */
function = LLVMBuildBitCast(builder, callee,
LLVMTypeOf(function), "cast callee");
tmp_ptr = lp_build_alloca(gallivm, f32x4t, "");
/*
* Invoke format_desc->fetch_rgba_float() for each pixel and insert the result
@@ -663,7 +701,7 @@ lp_build_fetch_rgba_aos(LLVMBuilderRef builder,
LLVMValueRef args[4];
args[0] = LLVMBuildBitCast(builder, tmp_ptr, pf32t, "");
args[1] = lp_build_gather_elem_ptr(builder, num_pixels,
args[1] = lp_build_gather_elem_ptr(gallivm, num_pixels,
base_ptr, offset, k);
if (num_pixels == 1) {
@@ -671,7 +709,7 @@ lp_build_fetch_rgba_aos(LLVMBuilderRef builder,
args[3] = j;
}
else {
LLVMValueRef index = LLVMConstInt(LLVMInt32Type(), k, 0);
LLVMValueRef index = lp_build_const_int32(gallivm, k);
args[2] = LLVMBuildExtractElement(builder, i, index, "");
args[3] = LLVMBuildExtractElement(builder, j, index, "");
}
@@ -681,7 +719,7 @@ lp_build_fetch_rgba_aos(LLVMBuilderRef builder,
tmps[k] = LLVMBuildLoad(builder, tmp_ptr, "");
}
lp_build_conv(builder,
lp_build_conv(gallivm,
lp_float32_vec4_type(),
type,
tmps, num_pixels, &res, 1);
@@ -690,5 +728,5 @@ lp_build_fetch_rgba_aos(LLVMBuilderRef builder,
}
assert(0);
return lp_build_undef(type);
return lp_build_undef(gallivm, type);
}

View File

@@ -97,12 +97,13 @@ lp_build_format_swizzle_soa(const struct util_format_description *format_desc,
* \param rgba_out returns the SoA R,G,B,A vectors
*/
void
lp_build_unpack_rgba_soa(LLVMBuilderRef builder,
lp_build_unpack_rgba_soa(struct gallivm_state *gallivm,
const struct util_format_description *format_desc,
struct lp_type type,
LLVMValueRef packed,
LLVMValueRef rgba_out[4])
{
LLVMBuilderRef builder = gallivm->builder;
struct lp_build_context bld;
LLVMValueRef inputs[4];
unsigned start;
@@ -116,7 +117,7 @@ lp_build_unpack_rgba_soa(LLVMBuilderRef builder,
assert(type.floating);
assert(type.width == 32);
lp_build_context_init(&bld, builder, type);
lp_build_context_init(&bld, gallivm, type);
/* Decode the input vector components */
start = 0;
@@ -129,7 +130,7 @@ lp_build_unpack_rgba_soa(LLVMBuilderRef builder,
switch(format_desc->channel[chan].type) {
case UTIL_FORMAT_TYPE_VOID:
input = lp_build_undef(type);
input = lp_build_undef(gallivm, type);
break;
case UTIL_FORMAT_TYPE_UNSIGNED:
@@ -138,7 +139,7 @@ lp_build_unpack_rgba_soa(LLVMBuilderRef builder,
*/
if (start) {
input = LLVMBuildLShr(builder, input, lp_build_const_int_vec(type, start), "");
input = LLVMBuildLShr(builder, input, lp_build_const_int_vec(gallivm, type, start), "");
}
/*
@@ -147,7 +148,7 @@ lp_build_unpack_rgba_soa(LLVMBuilderRef builder,
if (stop < format_desc->block.bits) {
unsigned mask = ((unsigned long long)1 << width) - 1;
input = LLVMBuildAnd(builder, input, lp_build_const_int_vec(type, mask), "");
input = LLVMBuildAnd(builder, input, lp_build_const_int_vec(gallivm, type, mask), "");
}
/*
@@ -156,14 +157,15 @@ lp_build_unpack_rgba_soa(LLVMBuilderRef builder,
if (type.floating) {
if(format_desc->channel[chan].normalized)
input = lp_build_unsigned_norm_to_float(builder, width, type, input);
input = lp_build_unsigned_norm_to_float(gallivm, width, type, input);
else
input = LLVMBuildSIToFP(builder, input, lp_build_vec_type(type), "");
input = LLVMBuildSIToFP(builder, input,
lp_build_vec_type(gallivm, type), "");
}
else {
/* FIXME */
assert(0);
input = lp_build_undef(type);
input = lp_build_undef(gallivm, type);
}
break;
@@ -175,7 +177,7 @@ lp_build_unpack_rgba_soa(LLVMBuilderRef builder,
if (stop < type.width) {
unsigned bits = type.width - stop;
LLVMValueRef bits_val = lp_build_const_int_vec(type, bits);
LLVMValueRef bits_val = lp_build_const_int_vec(gallivm, type, bits);
input = LLVMBuildShl(builder, input, bits_val, "");
}
@@ -185,7 +187,7 @@ lp_build_unpack_rgba_soa(LLVMBuilderRef builder,
if (format_desc->channel[chan].size < type.width) {
unsigned bits = type.width - format_desc->channel[chan].size;
LLVMValueRef bits_val = lp_build_const_int_vec(type, bits);
LLVMValueRef bits_val = lp_build_const_int_vec(gallivm, type, bits);
input = LLVMBuildAShr(builder, input, bits_val, "");
}
@@ -194,17 +196,17 @@ lp_build_unpack_rgba_soa(LLVMBuilderRef builder,
*/
if (type.floating) {
input = LLVMBuildSIToFP(builder, input, lp_build_vec_type(type), "");
input = LLVMBuildSIToFP(builder, input, lp_build_vec_type(gallivm, type), "");
if (format_desc->channel[chan].normalized) {
double scale = 1.0 / ((1 << (format_desc->channel[chan].size - 1)) - 1);
LLVMValueRef scale_val = lp_build_const_vec(type, scale);
LLVMValueRef scale_val = lp_build_const_vec(gallivm, type, scale);
input = LLVMBuildFMul(builder, input, scale_val, "");
}
}
else {
/* FIXME */
assert(0);
input = lp_build_undef(type);
input = lp_build_undef(gallivm, type);
}
break;
@@ -214,32 +216,32 @@ lp_build_unpack_rgba_soa(LLVMBuilderRef builder,
assert(start == 0);
assert(stop == 32);
assert(type.width == 32);
input = LLVMBuildBitCast(builder, input, lp_build_vec_type(type), "");
input = LLVMBuildBitCast(builder, input, lp_build_vec_type(gallivm, type), "");
}
else {
/* FIXME */
assert(0);
input = lp_build_undef(type);
input = lp_build_undef(gallivm, type);
}
break;
case UTIL_FORMAT_TYPE_FIXED:
if (type.floating) {
double scale = 1.0 / ((1 << (format_desc->channel[chan].size/2)) - 1);
LLVMValueRef scale_val = lp_build_const_vec(type, scale);
input = LLVMBuildSIToFP(builder, input, lp_build_vec_type(type), "");
LLVMValueRef scale_val = lp_build_const_vec(gallivm, type, scale);
input = LLVMBuildSIToFP(builder, input, lp_build_vec_type(gallivm, type), "");
input = LLVMBuildFMul(builder, input, scale_val, "");
}
else {
/* FIXME */
assert(0);
input = lp_build_undef(type);
input = lp_build_undef(gallivm, type);
}
break;
default:
assert(0);
input = lp_build_undef(type);
input = lp_build_undef(gallivm, type);
break;
}
@@ -253,16 +255,17 @@ lp_build_unpack_rgba_soa(LLVMBuilderRef builder,
void
lp_build_rgba8_to_f32_soa(LLVMBuilderRef builder,
lp_build_rgba8_to_f32_soa(struct gallivm_state *gallivm,
struct lp_type dst_type,
LLVMValueRef packed,
LLVMValueRef *rgba)
{
LLVMValueRef mask = lp_build_const_int_vec(dst_type, 0xff);
LLVMBuilderRef builder = gallivm->builder;
LLVMValueRef mask = lp_build_const_int_vec(gallivm, dst_type, 0xff);
unsigned chan;
packed = LLVMBuildBitCast(builder, packed,
lp_build_int_vec_type(dst_type), "");
lp_build_int_vec_type(gallivm, dst_type), "");
/* Decode the input vector components */
for (chan = 0; chan < 4; ++chan) {
@@ -274,12 +277,12 @@ lp_build_rgba8_to_f32_soa(LLVMBuilderRef builder,
if (start)
input = LLVMBuildLShr(builder, input,
lp_build_const_int_vec(dst_type, start), "");
lp_build_const_int_vec(gallivm, dst_type, start), "");
if (stop < 32)
input = LLVMBuildAnd(builder, input, mask, "");
input = lp_build_unsigned_norm_to_float(builder, 8, dst_type, input);
input = lp_build_unsigned_norm_to_float(gallivm, 8, dst_type, input);
rgba[chan] = input;
}
@@ -303,7 +306,7 @@ lp_build_rgba8_to_f32_soa(LLVMBuilderRef builder,
* be in [0, block_width-1] and j will be in [0, block_height-1].
*/
void
lp_build_fetch_rgba_soa(LLVMBuilderRef builder,
lp_build_fetch_rgba_soa(struct gallivm_state *gallivm,
const struct util_format_description *format_desc,
struct lp_type type,
LLVMValueRef base_ptr,
@@ -312,6 +315,7 @@ lp_build_fetch_rgba_soa(LLVMBuilderRef builder,
LLVMValueRef j,
LLVMValueRef rgba_out[4])
{
LLVMBuilderRef builder = gallivm->builder;
if (format_desc->layout == UTIL_FORMAT_LAYOUT_PLAIN &&
(format_desc->colorspace == UTIL_FORMAT_COLORSPACE_RGB ||
@@ -334,7 +338,7 @@ lp_build_fetch_rgba_soa(LLVMBuilderRef builder,
* gather the texels from the texture
* Ex: packed = {BGRA, BGRA, BGRA, BGRA}.
*/
packed = lp_build_gather(builder,
packed = lp_build_gather(gallivm,
type.length,
format_desc->block.bits,
type.width,
@@ -343,7 +347,7 @@ lp_build_fetch_rgba_soa(LLVMBuilderRef builder,
/*
* convert texels to float rgba
*/
lp_build_unpack_rgba_soa(builder,
lp_build_unpack_rgba_soa(gallivm,
format_desc,
type,
packed, rgba_out);
@@ -364,10 +368,10 @@ lp_build_fetch_rgba_soa(LLVMBuilderRef builder,
tmp_type.length = type.length * 4;
tmp_type.norm = TRUE;
tmp = lp_build_fetch_rgba_aos(builder, format_desc, tmp_type,
tmp = lp_build_fetch_rgba_aos(gallivm, format_desc, tmp_type,
base_ptr, offset, i, j);
lp_build_rgba8_to_f32_soa(builder,
lp_build_rgba8_to_f32_soa(gallivm,
type,
tmp,
rgba_out);
@@ -397,23 +401,24 @@ lp_build_fetch_rgba_soa(LLVMBuilderRef builder,
tmp_type.length = 4;
for (chan = 0; chan < 4; ++chan) {
rgba_out[chan] = lp_build_undef(type);
rgba_out[chan] = lp_build_undef(gallivm, type);
}
/* loop over number of pixels */
for(k = 0; k < type.length; ++k) {
LLVMValueRef index = LLVMConstInt(LLVMInt32Type(), k, 0);
LLVMValueRef index = lp_build_const_int32(gallivm, k);
LLVMValueRef offset_elem;
LLVMValueRef i_elem, j_elem;
LLVMValueRef tmp;
offset_elem = LLVMBuildExtractElement(builder, offset, index, "");
offset_elem = LLVMBuildExtractElement(builder, offset,
index, "");
i_elem = LLVMBuildExtractElement(builder, i, index, "");
j_elem = LLVMBuildExtractElement(builder, j, index, "");
/* Get a single float[4]={R,G,B,A} pixel */
tmp = lp_build_fetch_rgba_aos(builder, format_desc, tmp_type,
tmp = lp_build_fetch_rgba_aos(gallivm, format_desc, tmp_type,
base_ptr, offset_elem,
i_elem, j_elem);
@@ -422,7 +427,7 @@ lp_build_fetch_rgba_soa(LLVMBuilderRef builder,
* position = 'index'.
*/
for (chan = 0; chan < 4; ++chan) {
LLVMValueRef chan_val = LLVMConstInt(LLVMInt32Type(), chan, 0),
LLVMValueRef chan_val = lp_build_const_int32(gallivm, chan),
tmp_chan = LLVMBuildExtractElement(builder, tmp, chan_val, "");
rgba_out[chan] = LLVMBuildInsertElement(builder, rgba_out[chan],
tmp_chan, index, "");

View File

@@ -43,6 +43,7 @@
#include "lp_bld_conv.h"
#include "lp_bld_gather.h"
#include "lp_bld_format.h"
#include "lp_bld_init.h"
#include "lp_bld_logic.h"
/**
@@ -51,7 +52,7 @@
* @param i is a <n x i32> vector with the x pixel coordinate (0 or 1)
*/
static void
uyvy_to_yuv_soa(LLVMBuilderRef builder,
uyvy_to_yuv_soa(struct gallivm_state *gallivm,
unsigned n,
LLVMValueRef packed,
LLVMValueRef i,
@@ -59,6 +60,7 @@ uyvy_to_yuv_soa(LLVMBuilderRef builder,
LLVMValueRef *u,
LLVMValueRef *v)
{
LLVMBuilderRef builder = gallivm->builder;
struct lp_type type;
LLVMValueRef mask;
@@ -86,25 +88,25 @@ uyvy_to_yuv_soa(LLVMBuilderRef builder,
LLVMValueRef sel, tmp, tmp2;
struct lp_build_context bld32;
lp_build_context_init(&bld32, builder, type);
lp_build_context_init(&bld32, gallivm, type);
tmp = LLVMBuildLShr(builder, packed, lp_build_const_int_vec(type, 8), "");
tmp2 = LLVMBuildLShr(builder, tmp, lp_build_const_int_vec(type, 16), "");
sel = lp_build_compare(builder, type, PIPE_FUNC_EQUAL, i, lp_build_const_int_vec(type, 0));
tmp = LLVMBuildLShr(builder, packed, lp_build_const_int_vec(gallivm, type, 8), "");
tmp2 = LLVMBuildLShr(builder, tmp, lp_build_const_int_vec(gallivm, type, 16), "");
sel = lp_build_compare(gallivm, type, PIPE_FUNC_EQUAL, i, lp_build_const_int_vec(gallivm, type, 0));
*y = lp_build_select(&bld32, sel, tmp, tmp2);
} else
#endif
{
LLVMValueRef shift;
shift = LLVMBuildMul(builder, i, lp_build_const_int_vec(type, 16), "");
shift = LLVMBuildAdd(builder, shift, lp_build_const_int_vec(type, 8), "");
shift = LLVMBuildMul(builder, i, lp_build_const_int_vec(gallivm, type, 16), "");
shift = LLVMBuildAdd(builder, shift, lp_build_const_int_vec(gallivm, type, 8), "");
*y = LLVMBuildLShr(builder, packed, shift, "");
}
*u = packed;
*v = LLVMBuildLShr(builder, packed, lp_build_const_int_vec(type, 16), "");
*v = LLVMBuildLShr(builder, packed, lp_build_const_int_vec(gallivm, type, 16), "");
mask = lp_build_const_int_vec(type, 0xff);
mask = lp_build_const_int_vec(gallivm, type, 0xff);
*y = LLVMBuildAnd(builder, *y, mask, "y");
*u = LLVMBuildAnd(builder, *u, mask, "u");
@@ -118,7 +120,7 @@ uyvy_to_yuv_soa(LLVMBuilderRef builder,
* @param i is a <n x i32> vector with the x pixel coordinate (0 or 1)
*/
static void
yuyv_to_yuv_soa(LLVMBuilderRef builder,
yuyv_to_yuv_soa(struct gallivm_state *gallivm,
unsigned n,
LLVMValueRef packed,
LLVMValueRef i,
@@ -126,6 +128,7 @@ yuyv_to_yuv_soa(LLVMBuilderRef builder,
LLVMValueRef *u,
LLVMValueRef *v)
{
LLVMBuilderRef builder = gallivm->builder;
struct lp_type type;
LLVMValueRef mask;
@@ -153,23 +156,23 @@ yuyv_to_yuv_soa(LLVMBuilderRef builder,
LLVMValueRef sel, tmp;
struct lp_build_context bld32;
lp_build_context_init(&bld32, builder, type);
lp_build_context_init(&bld32, gallivm, type);
tmp = LLVMBuildLShr(builder, packed, lp_build_const_int_vec(type, 16), "");
sel = lp_build_compare(builder, type, PIPE_FUNC_EQUAL, i, lp_build_const_int_vec(type, 0));
tmp = LLVMBuildLShr(builder, packed, lp_build_const_int_vec(gallivm, type, 16), "");
sel = lp_build_compare(gallivm, type, PIPE_FUNC_EQUAL, i, lp_build_const_int_vec(gallivm, type, 0));
*y = lp_build_select(&bld32, sel, packed, tmp);
} else
#endif
{
LLVMValueRef shift;
shift = LLVMBuildMul(builder, i, lp_build_const_int_vec(type, 16), "");
shift = LLVMBuildMul(builder, i, lp_build_const_int_vec(gallivm, type, 16), "");
*y = LLVMBuildLShr(builder, packed, shift, "");
}
*u = LLVMBuildLShr(builder, packed, lp_build_const_int_vec(type, 8), "");
*v = LLVMBuildLShr(builder, packed, lp_build_const_int_vec(type, 24), "");
*u = LLVMBuildLShr(builder, packed, lp_build_const_int_vec(gallivm, type, 8), "");
*v = LLVMBuildLShr(builder, packed, lp_build_const_int_vec(gallivm, type, 24), "");
mask = lp_build_const_int_vec(type, 0xff);
mask = lp_build_const_int_vec(gallivm, type, 0xff);
*y = LLVMBuildAnd(builder, *y, mask, "y");
*u = LLVMBuildAnd(builder, *u, mask, "u");
@@ -178,11 +181,12 @@ yuyv_to_yuv_soa(LLVMBuilderRef builder,
static INLINE void
yuv_to_rgb_soa(LLVMBuilderRef builder,
yuv_to_rgb_soa(struct gallivm_state *gallivm,
unsigned n,
LLVMValueRef y, LLVMValueRef u, LLVMValueRef v,
LLVMValueRef *r, LLVMValueRef *g, LLVMValueRef *b)
{
LLVMBuilderRef builder = gallivm->builder;
struct lp_type type;
struct lp_build_context bld;
@@ -203,7 +207,7 @@ yuv_to_rgb_soa(LLVMBuilderRef builder,
type.width = 32;
type.length = n;
lp_build_context_init(&bld, builder, type);
lp_build_context_init(&bld, gallivm, type);
assert(lp_check_value(type, y));
assert(lp_check_value(type, u));
@@ -213,17 +217,17 @@ yuv_to_rgb_soa(LLVMBuilderRef builder,
* Constants
*/
c0 = lp_build_const_int_vec(type, 0);
c8 = lp_build_const_int_vec(type, 8);
c16 = lp_build_const_int_vec(type, 16);
c128 = lp_build_const_int_vec(type, 128);
c255 = lp_build_const_int_vec(type, 255);
c0 = lp_build_const_int_vec(gallivm, type, 0);
c8 = lp_build_const_int_vec(gallivm, type, 8);
c16 = lp_build_const_int_vec(gallivm, type, 16);
c128 = lp_build_const_int_vec(gallivm, type, 128);
c255 = lp_build_const_int_vec(gallivm, type, 255);
cy = lp_build_const_int_vec(type, 298);
cug = lp_build_const_int_vec(type, -100);
cub = lp_build_const_int_vec(type, 516);
cvr = lp_build_const_int_vec(type, 409);
cvg = lp_build_const_int_vec(type, -208);
cy = lp_build_const_int_vec(gallivm, type, 298);
cug = lp_build_const_int_vec(gallivm, type, -100);
cub = lp_build_const_int_vec(gallivm, type, 516);
cvr = lp_build_const_int_vec(gallivm, type, 409);
cvg = lp_build_const_int_vec(gallivm, type, -208);
/*
* y -= 16;
@@ -276,10 +280,11 @@ yuv_to_rgb_soa(LLVMBuilderRef builder,
static LLVMValueRef
rgb_to_rgba_aos(LLVMBuilderRef builder,
rgb_to_rgba_aos(struct gallivm_state *gallivm,
unsigned n,
LLVMValueRef r, LLVMValueRef g, LLVMValueRef b)
{
LLVMBuilderRef builder = gallivm->builder;
struct lp_type type;
LLVMValueRef a;
LLVMValueRef rgba;
@@ -298,9 +303,9 @@ rgb_to_rgba_aos(LLVMBuilderRef builder,
*/
r = r;
g = LLVMBuildShl(builder, g, lp_build_const_int_vec(type, 8), "");
b = LLVMBuildShl(builder, b, lp_build_const_int_vec(type, 16), "");
a = lp_build_const_int_vec(type, 0xff000000);
g = LLVMBuildShl(builder, g, lp_build_const_int_vec(gallivm, type, 8), "");
b = LLVMBuildShl(builder, b, lp_build_const_int_vec(gallivm, type, 16), "");
a = lp_build_const_int_vec(gallivm, type, 0xff000000);
rgba = r;
rgba = LLVMBuildOr(builder, rgba, g, "");
@@ -308,7 +313,7 @@ rgb_to_rgba_aos(LLVMBuilderRef builder,
rgba = LLVMBuildOr(builder, rgba, a, "");
rgba = LLVMBuildBitCast(builder, rgba,
LLVMVectorType(LLVMInt8Type(), 4*n), "");
LLVMVectorType(LLVMInt8TypeInContext(gallivm->context), 4*n), "");
return rgba;
}
@@ -318,7 +323,7 @@ rgb_to_rgba_aos(LLVMBuilderRef builder,
* Convert from <n x i32> packed UYVY to <4n x i8> RGBA AoS
*/
static LLVMValueRef
uyvy_to_rgba_aos(LLVMBuilderRef builder,
uyvy_to_rgba_aos(struct gallivm_state *gallivm,
unsigned n,
LLVMValueRef packed,
LLVMValueRef i)
@@ -327,9 +332,9 @@ uyvy_to_rgba_aos(LLVMBuilderRef builder,
LLVMValueRef r, g, b;
LLVMValueRef rgba;
uyvy_to_yuv_soa(builder, n, packed, i, &y, &u, &v);
yuv_to_rgb_soa(builder, n, y, u, v, &r, &g, &b);
rgba = rgb_to_rgba_aos(builder, n, r, g, b);
uyvy_to_yuv_soa(gallivm, n, packed, i, &y, &u, &v);
yuv_to_rgb_soa(gallivm, n, y, u, v, &r, &g, &b);
rgba = rgb_to_rgba_aos(gallivm, n, r, g, b);
return rgba;
}
@@ -339,7 +344,7 @@ uyvy_to_rgba_aos(LLVMBuilderRef builder,
* Convert from <n x i32> packed YUYV to <4n x i8> RGBA AoS
*/
static LLVMValueRef
yuyv_to_rgba_aos(LLVMBuilderRef builder,
yuyv_to_rgba_aos(struct gallivm_state *gallivm,
unsigned n,
LLVMValueRef packed,
LLVMValueRef i)
@@ -348,9 +353,9 @@ yuyv_to_rgba_aos(LLVMBuilderRef builder,
LLVMValueRef r, g, b;
LLVMValueRef rgba;
yuyv_to_yuv_soa(builder, n, packed, i, &y, &u, &v);
yuv_to_rgb_soa(builder, n, y, u, v, &r, &g, &b);
rgba = rgb_to_rgba_aos(builder, n, r, g, b);
yuyv_to_yuv_soa(gallivm, n, packed, i, &y, &u, &v);
yuv_to_rgb_soa(gallivm, n, y, u, v, &r, &g, &b);
rgba = rgb_to_rgba_aos(gallivm, n, r, g, b);
return rgba;
}
@@ -360,7 +365,7 @@ yuyv_to_rgba_aos(LLVMBuilderRef builder,
* Convert from <n x i32> packed RG_BG to <4n x i8> RGBA AoS
*/
static LLVMValueRef
rgbg_to_rgba_aos(LLVMBuilderRef builder,
rgbg_to_rgba_aos(struct gallivm_state *gallivm,
unsigned n,
LLVMValueRef packed,
LLVMValueRef i)
@@ -368,8 +373,8 @@ rgbg_to_rgba_aos(LLVMBuilderRef builder,
LLVMValueRef r, g, b;
LLVMValueRef rgba;
uyvy_to_yuv_soa(builder, n, packed, i, &g, &r, &b);
rgba = rgb_to_rgba_aos(builder, n, r, g, b);
uyvy_to_yuv_soa(gallivm, n, packed, i, &g, &r, &b);
rgba = rgb_to_rgba_aos(gallivm, n, r, g, b);
return rgba;
}
@@ -379,7 +384,7 @@ rgbg_to_rgba_aos(LLVMBuilderRef builder,
* Convert from <n x i32> packed GR_GB to <4n x i8> RGBA AoS
*/
static LLVMValueRef
grgb_to_rgba_aos(LLVMBuilderRef builder,
grgb_to_rgba_aos(struct gallivm_state *gallivm,
unsigned n,
LLVMValueRef packed,
LLVMValueRef i)
@@ -387,8 +392,8 @@ grgb_to_rgba_aos(LLVMBuilderRef builder,
LLVMValueRef r, g, b;
LLVMValueRef rgba;
yuyv_to_yuv_soa(builder, n, packed, i, &g, &r, &b);
rgba = rgb_to_rgba_aos(builder, n, r, g, b);
yuyv_to_yuv_soa(gallivm, n, packed, i, &g, &r, &b);
rgba = rgb_to_rgba_aos(gallivm, n, r, g, b);
return rgba;
}
@@ -401,7 +406,7 @@ grgb_to_rgba_aos(LLVMBuilderRef builder,
* @return a <4*n x i8> vector with the pixel RGBA values in AoS
*/
LLVMValueRef
lp_build_fetch_subsampled_rgba_aos(LLVMBuilderRef builder,
lp_build_fetch_subsampled_rgba_aos(struct gallivm_state *gallivm,
const struct util_format_description *format_desc,
unsigned n,
LLVMValueRef base_ptr,
@@ -417,26 +422,26 @@ lp_build_fetch_subsampled_rgba_aos(LLVMBuilderRef builder,
assert(format_desc->block.width == 2);
assert(format_desc->block.height == 1);
packed = lp_build_gather(builder, n, 32, 32, base_ptr, offset);
packed = lp_build_gather(gallivm, n, 32, 32, base_ptr, offset);
(void)j;
switch (format_desc->format) {
case PIPE_FORMAT_UYVY:
rgba = uyvy_to_rgba_aos(builder, n, packed, i);
rgba = uyvy_to_rgba_aos(gallivm, n, packed, i);
break;
case PIPE_FORMAT_YUYV:
rgba = yuyv_to_rgba_aos(builder, n, packed, i);
rgba = yuyv_to_rgba_aos(gallivm, n, packed, i);
break;
case PIPE_FORMAT_R8G8_B8G8_UNORM:
rgba = rgbg_to_rgba_aos(builder, n, packed, i);
rgba = rgbg_to_rgba_aos(gallivm, n, packed, i);
break;
case PIPE_FORMAT_G8R8_G8B8_UNORM:
rgba = grgb_to_rgba_aos(builder, n, packed, i);
rgba = grgb_to_rgba_aos(gallivm, n, packed, i);
break;
default:
assert(0);
rgba = LLVMGetUndef(LLVMVectorType(LLVMInt8Type(), 4*n));
rgba = LLVMGetUndef(LLVMVectorType(LLVMInt8TypeInContext(gallivm->context), 4*n));
break;
}

View File

@@ -31,6 +31,7 @@
#include "lp_bld_const.h"
#include "lp_bld_format.h"
#include "lp_bld_gather.h"
#include "lp_bld_init.h"
/**
@@ -39,7 +40,7 @@
* @sa lp_build_gather()
*/
LLVMValueRef
lp_build_gather_elem_ptr(LLVMBuilderRef builder,
lp_build_gather_elem_ptr(struct gallivm_state *gallivm,
unsigned length,
LLVMValueRef base_ptr,
LLVMValueRef offsets,
@@ -48,17 +49,17 @@ lp_build_gather_elem_ptr(LLVMBuilderRef builder,
LLVMValueRef offset;
LLVMValueRef ptr;
assert(LLVMTypeOf(base_ptr) == LLVMPointerType(LLVMInt8Type(), 0));
assert(LLVMTypeOf(base_ptr) == LLVMPointerType(LLVMInt8TypeInContext(gallivm->context), 0));
if (length == 1) {
assert(i == 0);
offset = offsets;
} else {
LLVMValueRef index = LLVMConstInt(LLVMInt32Type(), i, 0);
offset = LLVMBuildExtractElement(builder, offsets, index, "");
LLVMValueRef index = lp_build_const_int32(gallivm, i);
offset = LLVMBuildExtractElement(gallivm->builder, offsets, index, "");
}
ptr = LLVMBuildGEP(builder, base_ptr, &offset, 1, "");
ptr = LLVMBuildGEP(gallivm->builder, base_ptr, &offset, 1, "");
return ptr;
}
@@ -70,7 +71,7 @@ lp_build_gather_elem_ptr(LLVMBuilderRef builder,
* @sa lp_build_gather()
*/
LLVMValueRef
lp_build_gather_elem(LLVMBuilderRef builder,
lp_build_gather_elem(struct gallivm_state *gallivm,
unsigned length,
unsigned src_width,
unsigned dst_width,
@@ -78,23 +79,23 @@ lp_build_gather_elem(LLVMBuilderRef builder,
LLVMValueRef offsets,
unsigned i)
{
LLVMTypeRef src_type = LLVMIntType(src_width);
LLVMTypeRef src_type = LLVMIntTypeInContext(gallivm->context, src_width);
LLVMTypeRef src_ptr_type = LLVMPointerType(src_type, 0);
LLVMTypeRef dst_elem_type = LLVMIntType(dst_width);
LLVMTypeRef dst_elem_type = LLVMIntTypeInContext(gallivm->context, dst_width);
LLVMValueRef ptr;
LLVMValueRef res;
assert(LLVMTypeOf(base_ptr) == LLVMPointerType(LLVMInt8Type(), 0));
assert(LLVMTypeOf(base_ptr) == LLVMPointerType(LLVMInt8TypeInContext(gallivm->context), 0));
ptr = lp_build_gather_elem_ptr(builder, length, base_ptr, offsets, i);
ptr = LLVMBuildBitCast(builder, ptr, src_ptr_type, "");
res = LLVMBuildLoad(builder, ptr, "");
ptr = lp_build_gather_elem_ptr(gallivm, length, base_ptr, offsets, i);
ptr = LLVMBuildBitCast(gallivm->builder, ptr, src_ptr_type, "");
res = LLVMBuildLoad(gallivm->builder, ptr, "");
assert(src_width <= dst_width);
if (src_width > dst_width)
res = LLVMBuildTrunc(builder, res, dst_elem_type, "");
res = LLVMBuildTrunc(gallivm->builder, res, dst_elem_type, "");
if (src_width < dst_width)
res = LLVMBuildZExt(builder, res, dst_elem_type, "");
res = LLVMBuildZExt(gallivm->builder, res, dst_elem_type, "");
return res;
}
@@ -112,7 +113,7 @@ lp_build_gather_elem(LLVMBuilderRef builder,
* @param offsets vector with offsets
*/
LLVMValueRef
lp_build_gather(LLVMBuilderRef builder,
lp_build_gather(struct gallivm_state *gallivm,
unsigned length,
unsigned src_width,
unsigned dst_width,
@@ -123,24 +124,24 @@ lp_build_gather(LLVMBuilderRef builder,
if (length == 1) {
/* Scalar */
return lp_build_gather_elem(builder, length,
return lp_build_gather_elem(gallivm, length,
src_width, dst_width,
base_ptr, offsets, 0);
} else {
/* Vector */
LLVMTypeRef dst_elem_type = LLVMIntType(dst_width);
LLVMTypeRef dst_elem_type = LLVMIntTypeInContext(gallivm->context, dst_width);
LLVMTypeRef dst_vec_type = LLVMVectorType(dst_elem_type, length);
unsigned i;
res = LLVMGetUndef(dst_vec_type);
for (i = 0; i < length; ++i) {
LLVMValueRef index = LLVMConstInt(LLVMInt32Type(), i, 0);
LLVMValueRef index = lp_build_const_int32(gallivm, i);
LLVMValueRef elem;
elem = lp_build_gather_elem(builder, length,
elem = lp_build_gather_elem(gallivm, length,
src_width, dst_width,
base_ptr, offsets, i);
res = LLVMBuildInsertElement(builder, res, elem, index, "");
res = LLVMBuildInsertElement(gallivm->builder, res, elem, index, "");
}
}

View File

@@ -34,14 +34,14 @@
LLVMValueRef
lp_build_gather_elem_ptr(LLVMBuilderRef builder,
lp_build_gather_elem_ptr(struct gallivm_state *gallivm,
unsigned length,
LLVMValueRef base_ptr,
LLVMValueRef offsets,
unsigned i);
LLVMValueRef
lp_build_gather_elem(LLVMBuilderRef builder,
lp_build_gather_elem(struct gallivm_state *gallivm,
unsigned length,
unsigned src_width,
unsigned dst_width,
@@ -50,7 +50,7 @@ lp_build_gather_elem(LLVMBuilderRef builder,
unsigned i);
LLVMValueRef
lp_build_gather(LLVMBuilderRef builder,
lp_build_gather(struct gallivm_state *gallivm,
unsigned length,
unsigned src_width,
unsigned dst_width,

View File

@@ -29,6 +29,7 @@
#include "pipe/p_compiler.h"
#include "util/u_cpu_detect.h"
#include "util/u_debug.h"
#include "util/u_memory.h"
#include "lp_bld_debug.h"
#include "lp_bld_init.h"
@@ -45,6 +46,7 @@ static const struct debug_named_value lp_bld_debug_flags[] = {
{ "nopt", GALLIVM_DEBUG_NO_OPT, NULL },
{ "perf", GALLIVM_DEBUG_PERF, NULL },
{ "no_brilinear", GALLIVM_DEBUG_NO_BRILINEAR, NULL },
{ "gc", GALLIVM_DEBUG_GC, NULL },
DEBUG_NAMED_VALUE_END
};
@@ -52,11 +54,7 @@ DEBUG_GET_ONCE_FLAGS_OPTION(gallivm_debug, "GALLIVM_DEBUG", lp_bld_debug_flags,
#endif
LLVMModuleRef lp_build_module = NULL;
LLVMExecutionEngineRef lp_build_engine = NULL;
LLVMModuleProviderRef lp_build_provider = NULL;
LLVMTargetDataRef lp_build_target = NULL;
LLVMPassManagerRef lp_build_pass = NULL;
static boolean gallivm_initialized = FALSE;
/*
@@ -82,6 +80,19 @@ enum LLVM_CodeGenOpt_Level {
};
/**
* LLVM 2.6 permits only one ExecutionEngine to be created. This is it.
*/
static LLVMExecutionEngineRef GlobalEngine = NULL;
/**
* Same gallivm state shared by all contexts.
*/
static struct gallivm_state *GlobalGallivm = NULL;
extern void
lp_register_oprofile_jit_event_listener(LLVMExecutionEngineRef EE);
@@ -89,6 +100,288 @@ extern void
lp_set_target_options(void);
/**
* Create the LLVM (optimization) pass manager and install
* relevant optimization passes.
* \return TRUE for success, FALSE for failure
*/
static boolean
create_pass_manager(struct gallivm_state *gallivm)
{
assert(!gallivm->passmgr);
gallivm->passmgr = LLVMCreateFunctionPassManager(gallivm->provider);
if (!gallivm->passmgr)
return FALSE;
LLVMAddTargetData(gallivm->target, gallivm->passmgr);
if ((gallivm_debug & GALLIVM_DEBUG_NO_OPT) == 0) {
/* These are the passes currently listed in llvm-c/Transforms/Scalar.h,
* but there are more on SVN.
* TODO: Add more passes.
*/
LLVMAddCFGSimplificationPass(gallivm->passmgr);
if (HAVE_LLVM >= 0x207 && sizeof(void*) == 4) {
/* For LLVM >= 2.7 and 32-bit build, use this order of passes to
* avoid generating bad code.
* Test with piglit glsl-vs-sqrt-zero test.
*/
LLVMAddConstantPropagationPass(gallivm->passmgr);
LLVMAddPromoteMemoryToRegisterPass(gallivm->passmgr);
}
else {
LLVMAddPromoteMemoryToRegisterPass(gallivm->passmgr);
LLVMAddConstantPropagationPass(gallivm->passmgr);
}
if (util_cpu_caps.has_sse4_1) {
/* FIXME: There is a bug in this pass, whereby the combination
* of fptosi and sitofp (necessary for trunc/floor/ceil/round
* implementation) somehow becomes invalid code.
*/
LLVMAddInstructionCombiningPass(gallivm->passmgr);
}
LLVMAddGVNPass(gallivm->passmgr);
}
else {
/* We need at least this pass to prevent the backends to fail in
* unexpected ways.
*/
LLVMAddPromoteMemoryToRegisterPass(gallivm->passmgr);
}
return TRUE;
}
/**
* Free gallivm object's LLVM allocations, but not the gallivm object itself.
*/
static void
free_gallivm_state(struct gallivm_state *gallivm)
{
#if HAVE_LLVM >= 0x207 /* XXX or 0x208? */
/* This leads to crashes w/ some versions of LLVM */
LLVMModuleRef mod;
char *error;
if (gallivm->engine && gallivm->provider)
LLVMRemoveModuleProvider(gallivm->engine, gallivm->provider,
&mod, &error);
#endif
#if 0
/* XXX this seems to crash with all versions of LLVM */
if (gallivm->provider)
LLVMDisposeModuleProvider(gallivm->provider);
#endif
if (gallivm->passmgr)
LLVMDisposePassManager(gallivm->passmgr);
#if HAVE_LLVM >= 0x207
if (gallivm->module)
LLVMDisposeModule(gallivm->module);
#endif
#if 0
/* Don't free the exec engine, it's a global/singleton */
if (gallivm->engine)
LLVMDisposeExecutionEngine(gallivm->engine);
#endif
#if 0
/* Don't free the TargetData, it's owned by the exec engine */
LLVMDisposeTargetData(gallivm->target);
#endif
if (gallivm->context)
LLVMContextDispose(gallivm->context);
if (gallivm->builder)
LLVMDisposeBuilder(gallivm->builder);
gallivm->engine = NULL;
gallivm->target = NULL;
gallivm->module = NULL;
gallivm->provider = NULL;
gallivm->passmgr = NULL;
gallivm->context = NULL;
gallivm->builder = NULL;
}
/**
* Allocate gallivm LLVM objects.
* \return TRUE for success, FALSE for failure
*/
static boolean
init_gallivm_state(struct gallivm_state *gallivm)
{
assert(gallivm_initialized);
assert(!gallivm->context);
assert(!gallivm->module);
assert(!gallivm->provider);
gallivm->context = LLVMContextCreate();
if (!gallivm->context)
goto fail;
gallivm->module = LLVMModuleCreateWithNameInContext("gallivm",
gallivm->context);
if (!gallivm->module)
goto fail;
gallivm->provider =
LLVMCreateModuleProviderForExistingModule(gallivm->module);
if (!gallivm->provider)
goto fail;
if (!GlobalEngine) {
/* We can only create one LLVMExecutionEngine (w/ LLVM 2.6 anyway) */
enum LLVM_CodeGenOpt_Level optlevel;
char *error = NULL;
if (gallivm_debug & GALLIVM_DEBUG_NO_OPT) {
optlevel = None;
}
else {
optlevel = Default;
}
if (LLVMCreateJITCompiler(&GlobalEngine, gallivm->provider,
(unsigned) optlevel, &error)) {
_debug_printf("%s\n", error);
LLVMDisposeMessage(error);
goto fail;
}
#if defined(DEBUG) || defined(PROFILE)
lp_register_oprofile_jit_event_listener(GlobalEngine);
#endif
}
gallivm->engine = GlobalEngine;
LLVMAddModuleProvider(gallivm->engine, gallivm->provider);//new
gallivm->target = LLVMGetExecutionEngineTargetData(gallivm->engine);
if (!gallivm->target)
goto fail;
if (!create_pass_manager(gallivm))
goto fail;
gallivm->builder = LLVMCreateBuilderInContext(gallivm->context);
if (!gallivm->builder)
goto fail;
return TRUE;
fail:
free_gallivm_state(gallivm);
return FALSE;
}
struct callback
{
garbage_collect_callback_func func;
void *cb_data;
};
#define MAX_CALLBACKS 32
static struct callback Callbacks[MAX_CALLBACKS];
static unsigned NumCallbacks = 0;
/**
* Register a function with gallivm which will be called when we
* do garbage collection.
*/
void
gallivm_register_garbage_collector_callback(garbage_collect_callback_func func,
void *cb_data)
{
unsigned i;
for (i = 0; i < NumCallbacks; i++) {
if (Callbacks[i].func == func && Callbacks[i].cb_data == cb_data) {
/* already in list: no-op */
return;
}
}
assert(NumCallbacks < MAX_CALLBACKS);
if (NumCallbacks < MAX_CALLBACKS) {
Callbacks[NumCallbacks].func = func;
Callbacks[NumCallbacks].cb_data = cb_data;
NumCallbacks++;
}
}
/**
* Remove a callback.
*/
void
gallivm_remove_garbage_collector_callback(garbage_collect_callback_func func,
void *cb_data)
{
unsigned i;
for (i = 0; i < NumCallbacks; i++) {
if (Callbacks[i].func == func && Callbacks[i].cb_data == cb_data) {
/* found, now remove it */
NumCallbacks--;
for ( ; i < NumCallbacks; i++) {
Callbacks[i] = Callbacks[i + 1];
}
return;
}
}
}
/**
* Call the callback functions (which are typically in the
* draw module and llvmpipe driver.
*/
static void
call_garbage_collector_callbacks(void)
{
unsigned i;
for (i = 0; i < NumCallbacks; i++) {
Callbacks[i].func(Callbacks[i].cb_data);
}
}
/**
* Other gallium components using gallivm should call this periodically
* to let us do garbage collection (or at least try to free memory
* accumulated by the LLVM libraries).
*/
void
gallivm_garbage_collect(struct gallivm_state *gallivm)
{
if (gallivm->context) {
if (gallivm_debug & GALLIVM_DEBUG_GC)
debug_printf("***** Doing LLVM garbage collection\n");
call_garbage_collector_callbacks();
free_gallivm_state(gallivm);
init_gallivm_state(gallivm);
}
}
void
lp_build_init(void)
{
@@ -102,60 +395,9 @@ lp_build_init(void)
LLVMLinkInJIT();
if (!lp_build_module)
lp_build_module = LLVMModuleCreateWithName("gallivm");
if (!lp_build_provider)
lp_build_provider = LLVMCreateModuleProviderForExistingModule(lp_build_module);
if (!lp_build_engine) {
enum LLVM_CodeGenOpt_Level optlevel;
char *error = NULL;
if (gallivm_debug & GALLIVM_DEBUG_NO_OPT) {
optlevel = None;
}
else {
optlevel = Default;
}
if (LLVMCreateJITCompiler(&lp_build_engine, lp_build_provider,
(unsigned)optlevel, &error)) {
_debug_printf("%s\n", error);
LLVMDisposeMessage(error);
assert(0);
}
#if defined(DEBUG) || defined(PROFILE)
lp_register_oprofile_jit_event_listener(lp_build_engine);
#endif
}
if (!lp_build_target)
lp_build_target = LLVMGetExecutionEngineTargetData(lp_build_engine);
if (!lp_build_pass) {
lp_build_pass = LLVMCreateFunctionPassManager(lp_build_provider);
LLVMAddTargetData(lp_build_target, lp_build_pass);
if ((gallivm_debug & GALLIVM_DEBUG_NO_OPT) == 0) {
/* These are the passes currently listed in llvm-c/Transforms/Scalar.h,
* but there are more on SVN. */
/* TODO: Add more passes */
LLVMAddCFGSimplificationPass(lp_build_pass);
LLVMAddPromoteMemoryToRegisterPass(lp_build_pass);
LLVMAddConstantPropagationPass(lp_build_pass);
LLVMAddInstructionCombiningPass(lp_build_pass);
LLVMAddGVNPass(lp_build_pass);
} else {
/* We need at least this pass to prevent the backends to fail in
* unexpected ways.
*/
LLVMAddPromoteMemoryToRegisterPass(lp_build_pass);
}
}
util_cpu_detect();
gallivm_initialized = TRUE;
#if 0
/* For simulating less capable machines */
@@ -166,6 +408,39 @@ lp_build_init(void)
}
/**
* Create a new gallivm_state object.
* Note that we return a singleton.
*/
struct gallivm_state *
gallivm_create(void)
{
if (!GlobalGallivm) {
GlobalGallivm = CALLOC_STRUCT(gallivm_state);
if (GlobalGallivm) {
if (!init_gallivm_state(GlobalGallivm)) {
FREE(GlobalGallivm);
GlobalGallivm = NULL;
}
}
}
return GlobalGallivm;
}
/**
* Destroy a gallivm_state object.
*/
void
gallivm_destroy(struct gallivm_state *gallivm)
{
/* No-op: don't destroy the singleton */
(void) gallivm;
}
/*
* Hack to allow the linking of release LLVM static libraries on a debug build.
*

View File

@@ -30,24 +30,53 @@
#define LP_BLD_INIT_H
#include "pipe/p_compiler.h"
#include "lp_bld.h"
#include <llvm-c/ExecutionEngine.h>
extern LLVMModuleRef lp_build_module;
extern LLVMExecutionEngineRef lp_build_engine;
extern LLVMModuleProviderRef lp_build_provider;
extern LLVMTargetDataRef lp_build_target;
extern LLVMPassManagerRef lp_build_pass;
struct gallivm_state
{
LLVMModuleRef module;
LLVMExecutionEngineRef engine;
LLVMModuleProviderRef provider;
LLVMTargetDataRef target;
LLVMPassManagerRef passmgr;
LLVMContextRef context;
LLVMBuilderRef builder;
};
void
lp_build_init(void);
extern void
lp_func_delete_body(LLVMValueRef func);
void
gallivm_garbage_collect(struct gallivm_state *gallivm);
typedef void (*garbage_collect_callback_func)(void *cb_data);
void
gallivm_register_garbage_collector_callback(garbage_collect_callback_func func,
void *cb_data);
void
gallivm_remove_garbage_collector_callback(garbage_collect_callback_func func,
void *cb_data);
struct gallivm_state *
gallivm_create(void);
void
gallivm_destroy(struct gallivm_state *gallivm);
extern LLVMValueRef
lp_build_load_volatile(LLVMBuilderRef B, LLVMValueRef PointerVal,
const char *Name);

View File

@@ -46,6 +46,7 @@
#include "util/u_debug.h"
#include "lp_bld_const.h"
#include "lp_bld_intr.h"
@@ -136,12 +137,13 @@ lp_build_intrinsic_binary(LLVMBuilderRef builder,
LLVMValueRef
lp_build_intrinsic_map(LLVMBuilderRef builder,
lp_build_intrinsic_map(struct gallivm_state *gallivm,
const char *name,
LLVMTypeRef ret_type,
LLVMValueRef *args,
unsigned num_args)
{
LLVMBuilderRef builder = gallivm->builder;
LLVMTypeRef ret_elem_type = LLVMGetElementType(ret_type);
unsigned n = LLVMGetVectorSize(ret_type);
unsigned i, j;
@@ -151,7 +153,7 @@ lp_build_intrinsic_map(LLVMBuilderRef builder,
res = LLVMGetUndef(ret_type);
for(i = 0; i < n; ++i) {
LLVMValueRef index = LLVMConstInt(LLVMInt32Type(), i, 0);
LLVMValueRef index = lp_build_const_int32(gallivm, i);
LLVMValueRef arg_elems[LP_MAX_FUNC_ARGS];
LLVMValueRef res_elem;
for(j = 0; j < num_args; ++j)
@@ -165,17 +167,17 @@ lp_build_intrinsic_map(LLVMBuilderRef builder,
LLVMValueRef
lp_build_intrinsic_map_unary(LLVMBuilderRef builder,
lp_build_intrinsic_map_unary(struct gallivm_state *gallivm,
const char *name,
LLVMTypeRef ret_type,
LLVMValueRef a)
{
return lp_build_intrinsic_map(builder, name, ret_type, &a, 1);
return lp_build_intrinsic_map(gallivm, name, ret_type, &a, 1);
}
LLVMValueRef
lp_build_intrinsic_map_binary(LLVMBuilderRef builder,
lp_build_intrinsic_map_binary(struct gallivm_state *gallivm,
const char *name,
LLVMTypeRef ret_type,
LLVMValueRef a,
@@ -186,7 +188,7 @@ lp_build_intrinsic_map_binary(LLVMBuilderRef builder,
args[0] = a;
args[1] = b;
return lp_build_intrinsic_map(builder, name, ret_type, args, 2);
return lp_build_intrinsic_map(gallivm, name, ret_type, args, 2);
}

View File

@@ -38,6 +38,7 @@
#include "gallivm/lp_bld.h"
#include "gallivm/lp_bld_init.h"
/**
@@ -77,7 +78,7 @@ lp_build_intrinsic_binary(LLVMBuilderRef builder,
LLVMValueRef
lp_build_intrinsic_map(LLVMBuilderRef builder,
lp_build_intrinsic_map(struct gallivm_state *gallivm,
const char *name,
LLVMTypeRef ret_type,
LLVMValueRef *args,
@@ -85,14 +86,14 @@ lp_build_intrinsic_map(LLVMBuilderRef builder,
LLVMValueRef
lp_build_intrinsic_map_unary(LLVMBuilderRef builder,
lp_build_intrinsic_map_unary(struct gallivm_state *gallivm,
const char *name,
LLVMTypeRef ret_type,
LLVMValueRef a);
LLVMValueRef
lp_build_intrinsic_map_binary(LLVMBuilderRef builder,
lp_build_intrinsic_map_binary(struct gallivm_state *gallivm,
const char *name,
LLVMTypeRef ret_type,
LLVMValueRef a,

View File

@@ -39,6 +39,7 @@
#include "lp_bld_type.h"
#include "lp_bld_const.h"
#include "lp_bld_init.h"
#include "lp_bld_intr.h"
#include "lp_bld_debug.h"
#include "lp_bld_logic.h"
@@ -70,13 +71,14 @@
* The result values will be 0 for false or ~0 for true.
*/
LLVMValueRef
lp_build_compare(LLVMBuilderRef builder,
lp_build_compare(struct gallivm_state *gallivm,
const struct lp_type type,
unsigned func,
LLVMValueRef a,
LLVMValueRef b)
{
LLVMTypeRef int_vec_type = lp_build_int_vec_type(type);
LLVMBuilderRef builder = gallivm->builder;
LLVMTypeRef int_vec_type = lp_build_int_vec_type(gallivm, type);
LLVMValueRef zeros = LLVMConstNull(int_vec_type);
LLVMValueRef ones = LLVMConstAllOnes(int_vec_type);
LLVMValueRef cond;
@@ -115,7 +117,7 @@ lp_build_compare(LLVMBuilderRef builder,
if(type.width * type.length == 128) {
if(type.floating && util_cpu_caps.has_sse) {
/* float[4] comparison */
LLVMTypeRef vec_type = lp_build_vec_type(type);
LLVMTypeRef vec_type = lp_build_vec_type(gallivm, type);
LLVMValueRef args[3];
unsigned cc;
boolean swap;
@@ -144,7 +146,7 @@ lp_build_compare(LLVMBuilderRef builder,
break;
default:
assert(0);
return lp_build_undef(type);
return lp_build_undef(gallivm, type);
}
if(swap) {
@@ -156,7 +158,7 @@ lp_build_compare(LLVMBuilderRef builder,
args[1] = b;
}
args[2] = LLVMConstInt(LLVMInt8Type(), cc, 0);
args[2] = LLVMConstInt(LLVMInt8TypeInContext(gallivm->context), cc, 0);
res = lp_build_intrinsic(builder,
"llvm.x86.sse.cmp.ps",
vec_type,
@@ -185,7 +187,7 @@ lp_build_compare(LLVMBuilderRef builder,
const char *pcmpgt;
LLVMValueRef args[2];
LLVMValueRef res;
LLVMTypeRef vec_type = lp_build_vec_type(type);
LLVMTypeRef vec_type = lp_build_vec_type(gallivm, type);
switch (type.width) {
case 8:
@@ -202,14 +204,14 @@ lp_build_compare(LLVMBuilderRef builder,
break;
default:
assert(0);
return lp_build_undef(type);
return lp_build_undef(gallivm, type);
}
/* There are no unsigned comparison instructions. So flip the sign bit
* so that the results match.
*/
if (table[func].gt && !type.sign) {
LLVMValueRef msb = lp_build_const_int_vec(type, (unsigned long long)1 << (type.width - 1));
LLVMValueRef msb = lp_build_const_int_vec(gallivm, type, (unsigned long long)1 << (type.width - 1));
a = LLVMBuildXor(builder, a, msb, "");
b = LLVMBuildXor(builder, b, msb, "");
}
@@ -270,7 +272,7 @@ lp_build_compare(LLVMBuilderRef builder,
break;
default:
assert(0);
return lp_build_undef(type);
return lp_build_undef(gallivm, type);
}
#if HAVE_LLVM >= 0x0207
@@ -289,7 +291,7 @@ lp_build_compare(LLVMBuilderRef builder,
debug_printf("%s: warning: using slow element-wise float"
" vector comparison\n", __FUNCTION__);
for (i = 0; i < type.length; ++i) {
LLVMValueRef index = LLVMConstInt(LLVMInt32Type(), i, 0);
LLVMValueRef index = lp_build_const_int32(gallivm, i);
cond = LLVMBuildFCmp(builder, op,
LLVMBuildExtractElement(builder, a, index, ""),
LLVMBuildExtractElement(builder, b, index, ""),
@@ -326,7 +328,7 @@ lp_build_compare(LLVMBuilderRef builder,
break;
default:
assert(0);
return lp_build_undef(type);
return lp_build_undef(gallivm, type);
}
#if HAVE_LLVM >= 0x0207
@@ -348,7 +350,7 @@ lp_build_compare(LLVMBuilderRef builder,
}
for(i = 0; i < type.length; ++i) {
LLVMValueRef index = LLVMConstInt(LLVMInt32Type(), i, 0);
LLVMValueRef index = lp_build_const_int32(gallivm, i);
cond = LLVMBuildICmp(builder, op,
LLVMBuildExtractElement(builder, a, index, ""),
LLVMBuildExtractElement(builder, b, index, ""),
@@ -379,7 +381,7 @@ lp_build_cmp(struct lp_build_context *bld,
LLVMValueRef a,
LLVMValueRef b)
{
return lp_build_compare(bld->builder, bld->type, func, a, b);
return lp_build_compare(bld->gallivm, bld->type, func, a, b);
}
@@ -392,6 +394,7 @@ lp_build_select_bitwise(struct lp_build_context *bld,
LLVMValueRef a,
LLVMValueRef b)
{
LLVMBuilderRef builder = bld->gallivm->builder;
struct lp_type type = bld->type;
LLVMValueRef res;
@@ -403,25 +406,25 @@ lp_build_select_bitwise(struct lp_build_context *bld,
}
if(type.floating) {
LLVMTypeRef int_vec_type = lp_build_int_vec_type(type);
a = LLVMBuildBitCast(bld->builder, a, int_vec_type, "");
b = LLVMBuildBitCast(bld->builder, b, int_vec_type, "");
LLVMTypeRef int_vec_type = lp_build_int_vec_type(bld->gallivm, type);
a = LLVMBuildBitCast(builder, a, int_vec_type, "");
b = LLVMBuildBitCast(builder, b, int_vec_type, "");
}
a = LLVMBuildAnd(bld->builder, a, mask, "");
a = LLVMBuildAnd(builder, a, mask, "");
/* This often gets translated to PANDN, but sometimes the NOT is
* pre-computed and stored in another constant. The best strategy depends
* on available registers, so it is not a big deal -- hopefully LLVM does
* the right decision attending the rest of the program.
*/
b = LLVMBuildAnd(bld->builder, b, LLVMBuildNot(bld->builder, mask, ""), "");
b = LLVMBuildAnd(builder, b, LLVMBuildNot(builder, mask, ""), "");
res = LLVMBuildOr(bld->builder, a, b, "");
res = LLVMBuildOr(builder, a, b, "");
if(type.floating) {
LLVMTypeRef vec_type = lp_build_vec_type(type);
res = LLVMBuildBitCast(bld->builder, res, vec_type, "");
LLVMTypeRef vec_type = lp_build_vec_type(bld->gallivm, type);
res = LLVMBuildBitCast(builder, res, vec_type, "");
}
return res;
@@ -440,6 +443,8 @@ lp_build_select(struct lp_build_context *bld,
LLVMValueRef a,
LLVMValueRef b)
{
LLVMBuilderRef builder = bld->gallivm->builder;
LLVMContextRef lc = bld->gallivm->context;
struct lp_type type = bld->type;
LLVMValueRef res;
@@ -450,8 +455,8 @@ lp_build_select(struct lp_build_context *bld,
return a;
if (type.length == 1) {
mask = LLVMBuildTrunc(bld->builder, mask, LLVMInt1Type(), "");
res = LLVMBuildSelect(bld->builder, mask, a, b, "");
mask = LLVMBuildTrunc(builder, mask, LLVMInt1TypeInContext(lc), "");
res = LLVMBuildSelect(builder, mask, a, b, "");
}
else if (util_cpu_caps.has_sse4_1 &&
type.width * type.length == 128 &&
@@ -465,34 +470,34 @@ lp_build_select(struct lp_build_context *bld,
if (type.floating &&
type.width == 64) {
intrinsic = "llvm.x86.sse41.blendvpd";
arg_type = LLVMVectorType(LLVMDoubleType(), 2);
arg_type = LLVMVectorType(LLVMDoubleTypeInContext(lc), 2);
} else if (type.floating &&
type.width == 32) {
intrinsic = "llvm.x86.sse41.blendvps";
arg_type = LLVMVectorType(LLVMFloatType(), 4);
arg_type = LLVMVectorType(LLVMFloatTypeInContext(lc), 4);
} else {
intrinsic = "llvm.x86.sse41.pblendvb";
arg_type = LLVMVectorType(LLVMInt8Type(), 16);
arg_type = LLVMVectorType(LLVMInt8TypeInContext(lc), 16);
}
if (arg_type != bld->int_vec_type) {
mask = LLVMBuildBitCast(bld->builder, mask, arg_type, "");
mask = LLVMBuildBitCast(builder, mask, arg_type, "");
}
if (arg_type != bld->vec_type) {
a = LLVMBuildBitCast(bld->builder, a, arg_type, "");
b = LLVMBuildBitCast(bld->builder, b, arg_type, "");
a = LLVMBuildBitCast(builder, a, arg_type, "");
b = LLVMBuildBitCast(builder, b, arg_type, "");
}
args[0] = b;
args[1] = a;
args[2] = mask;
res = lp_build_intrinsic(bld->builder, intrinsic,
res = lp_build_intrinsic(builder, intrinsic,
arg_type, args, Elements(args));
if (arg_type != bld->vec_type) {
res = LLVMBuildBitCast(bld->builder, res, bld->vec_type, "");
res = LLVMBuildBitCast(builder, res, bld->vec_type, "");
}
}
else {
@@ -514,6 +519,7 @@ lp_build_select_aos(struct lp_build_context *bld,
LLVMValueRef a,
LLVMValueRef b)
{
LLVMBuilderRef builder = bld->gallivm->builder;
const struct lp_type type = bld->type;
const unsigned n = type.length;
unsigned i, j;
@@ -544,7 +550,7 @@ lp_build_select_aos(struct lp_build_context *bld,
/*
* Shuffle.
*/
LLVMTypeRef elem_type = LLVMInt32Type();
LLVMTypeRef elem_type = LLVMInt32TypeInContext(bld->gallivm->context);
LLVMValueRef shuffles[LP_MAX_VECTOR_LENGTH];
for(j = 0; j < n; j += 4)
@@ -553,7 +559,7 @@ lp_build_select_aos(struct lp_build_context *bld,
(mask & (1 << i) ? 0 : n) + j + i,
0);
return LLVMBuildShuffleVector(bld->builder, a, b, LLVMConstVector(shuffles, n), "");
return LLVMBuildShuffleVector(builder, a, b, LLVMConstVector(shuffles, n), "");
}
else {
#if 0
@@ -567,9 +573,9 @@ lp_build_select_aos(struct lp_build_context *bld,
cond_vec[j + i] = LLVMConstInt(elem_type,
mask & (1 << i) ? 1 : 0, 0);
return LLVMBuildSelect(bld->builder, LLVMConstVector(cond_vec, n), a, b, "");
return LLVMBuildSelect(builder, LLVMConstVector(cond_vec, n), a, b, "");
#else
LLVMValueRef mask_vec = lp_build_const_mask_aos(type, mask);
LLVMValueRef mask_vec = lp_build_const_mask_aos(bld->gallivm, type, mask);
return lp_build_select(bld, mask_vec, a, b);
#endif
}

View File

@@ -47,7 +47,7 @@ struct lp_build_context;
LLVMValueRef
lp_build_compare(LLVMBuilderRef builder,
lp_build_compare(struct gallivm_state *gallivm,
const struct lp_type type,
unsigned func,
LLVMValueRef a,

View File

@@ -133,6 +133,19 @@ lp_set_target_options(void)
#endif
#endif
/*
* LLVM revision 123367 switched the default stack alignment to 16 bytes on
* Linux (and several other Unices in later revisions), to match recent gcc
* versions.
*
* However our drivers can be loaded by old binary applications, still
* maintaining a 4 bytes stack alignment. Therefore we must tell LLVM here
* to only assume a 4 bytes alignment for backwards compatibility.
*/
#if defined(PIPE_ARCH_X86)
llvm::StackAlignment = 4;
#endif
#if defined(DEBUG) || defined(PROFILE)
llvm::NoFramePointerElim = true;
#endif
@@ -144,6 +157,7 @@ lp_set_target_options(void)
llvm::UnsafeFPMath = true;
#endif
#if HAVE_LLVM < 0x0209
/*
* LLVM will generate MMX instructions for vectors <= 64 bits, leading to
* innefficient code, and in 32bit systems, to the corruption of the FPU
@@ -152,16 +166,27 @@ lp_set_target_options(void)
* See also:
* - http://llvm.org/bugs/show_bug.cgi?id=3287
* - http://l4.me.uk/post/2009/06/07/llvm-wrinkle-3-configuration-what-configuration/
*
* The -disable-mmx global option can be specified only once since we
* dynamically link against LLVM it will reside in a separate shared object,
* which may or not be delete when this shared object is, so we use the
* llvm::DisablePrettyStackTrace variable (which we set below and should
* reside in the same shared library) to determine whether the -disable-mmx
* option has been set or not.
*
* Thankfully this ugly hack is not necessary on LLVM 2.9 onwards.
*/
static boolean first = TRUE;
if (first) {
if (!llvm::DisablePrettyStackTrace) {
static boolean first = TRUE;
static const char* options[] = {
"prog",
"-disable-mmx"
};
assert(first);
llvm::cl::ParseCommandLineOptions(2, const_cast<char**>(options));
first = FALSE;
}
#endif
/*
* By default LLVM adds a signal handler to output a pretty stack trace.

View File

@@ -72,6 +72,7 @@
#include "lp_bld_type.h"
#include "lp_bld_const.h"
#include "lp_bld_init.h"
#include "lp_bld_intr.h"
#include "lp_bld_arit.h"
#include "lp_bld_pack.h"
@@ -81,7 +82,8 @@
* Build shuffle vectors that match PUNPCKLxx and PUNPCKHxx instructions.
*/
static LLVMValueRef
lp_build_const_unpack_shuffle(unsigned n, unsigned lo_hi)
lp_build_const_unpack_shuffle(struct gallivm_state *gallivm,
unsigned n, unsigned lo_hi)
{
LLVMValueRef elems[LP_MAX_VECTOR_LENGTH];
unsigned i, j;
@@ -92,8 +94,8 @@ lp_build_const_unpack_shuffle(unsigned n, unsigned lo_hi)
/* TODO: cache results in a static table */
for(i = 0, j = lo_hi*n/2; i < n; i += 2, ++j) {
elems[i + 0] = LLVMConstInt(LLVMInt32Type(), 0 + j, 0);
elems[i + 1] = LLVMConstInt(LLVMInt32Type(), n + j, 0);
elems[i + 0] = lp_build_const_int32(gallivm, 0 + j);
elems[i + 1] = lp_build_const_int32(gallivm, n + j);
}
return LLVMConstVector(elems, n);
@@ -104,7 +106,7 @@ lp_build_const_unpack_shuffle(unsigned n, unsigned lo_hi)
* Build shuffle vectors that match PACKxx instructions.
*/
static LLVMValueRef
lp_build_const_pack_shuffle(unsigned n)
lp_build_const_pack_shuffle(struct gallivm_state *gallivm, unsigned n)
{
LLVMValueRef elems[LP_MAX_VECTOR_LENGTH];
unsigned i;
@@ -112,7 +114,7 @@ lp_build_const_pack_shuffle(unsigned n)
assert(n <= LP_MAX_VECTOR_LENGTH);
for(i = 0; i < n; ++i)
elems[i] = LLVMConstInt(LLVMInt32Type(), 2*i, 0);
elems[i] = lp_build_const_int32(gallivm, 2*i);
return LLVMConstVector(elems, n);
}
@@ -124,7 +126,7 @@ lp_build_const_pack_shuffle(unsigned n)
* Matches the PUNPCKLxx and PUNPCKHxx SSE instructions.
*/
LLVMValueRef
lp_build_interleave2(LLVMBuilderRef builder,
lp_build_interleave2(struct gallivm_state *gallivm,
struct lp_type type,
LLVMValueRef a,
LLVMValueRef b,
@@ -132,9 +134,9 @@ lp_build_interleave2(LLVMBuilderRef builder,
{
LLVMValueRef shuffle;
shuffle = lp_build_const_unpack_shuffle(type.length, lo_hi);
shuffle = lp_build_const_unpack_shuffle(gallivm, type.length, lo_hi);
return LLVMBuildShuffleVector(builder, a, b, shuffle, "");
return LLVMBuildShuffleVector(gallivm->builder, a, b, shuffle, "");
}
@@ -145,13 +147,14 @@ lp_build_interleave2(LLVMBuilderRef builder,
* values themselves.
*/
void
lp_build_unpack2(LLVMBuilderRef builder,
lp_build_unpack2(struct gallivm_state *gallivm,
struct lp_type src_type,
struct lp_type dst_type,
LLVMValueRef src,
LLVMValueRef *dst_lo,
LLVMValueRef *dst_hi)
{
LLVMBuilderRef builder = gallivm->builder;
LLVMValueRef msb;
LLVMTypeRef dst_vec_type;
@@ -162,24 +165,24 @@ lp_build_unpack2(LLVMBuilderRef builder,
if(dst_type.sign && src_type.sign) {
/* Replicate the sign bit in the most significant bits */
msb = LLVMBuildAShr(builder, src, lp_build_const_int_vec(src_type, src_type.width - 1), "");
msb = LLVMBuildAShr(builder, src, lp_build_const_int_vec(gallivm, src_type, src_type.width - 1), "");
}
else
/* Most significant bits always zero */
msb = lp_build_zero(src_type);
msb = lp_build_zero(gallivm, src_type);
/* Interleave bits */
#ifdef PIPE_ARCH_LITTLE_ENDIAN
*dst_lo = lp_build_interleave2(builder, src_type, src, msb, 0);
*dst_hi = lp_build_interleave2(builder, src_type, src, msb, 1);
*dst_lo = lp_build_interleave2(gallivm, src_type, src, msb, 0);
*dst_hi = lp_build_interleave2(gallivm, src_type, src, msb, 1);
#else
*dst_lo = lp_build_interleave2(builder, src_type, msb, src, 0);
*dst_hi = lp_build_interleave2(builder, src_type, msb, src, 1);
*dst_lo = lp_build_interleave2(gallivm, src_type, msb, src, 0);
*dst_hi = lp_build_interleave2(gallivm, src_type, msb, src, 1);
#endif
/* Cast the result into the new type (twice as wide) */
dst_vec_type = lp_build_vec_type(dst_type);
dst_vec_type = lp_build_vec_type(gallivm, dst_type);
*dst_lo = LLVMBuildBitCast(builder, *dst_lo, dst_vec_type, "");
*dst_hi = LLVMBuildBitCast(builder, *dst_hi, dst_vec_type, "");
@@ -193,7 +196,7 @@ lp_build_unpack2(LLVMBuilderRef builder,
* values themselves.
*/
void
lp_build_unpack(LLVMBuilderRef builder,
lp_build_unpack(struct gallivm_state *gallivm,
struct lp_type src_type,
struct lp_type dst_type,
LLVMValueRef src,
@@ -218,7 +221,7 @@ lp_build_unpack(LLVMBuilderRef builder,
tmp_type.length /= 2;
for(i = num_tmps; i--; ) {
lp_build_unpack2(builder, src_type, tmp_type, dst[i], &dst[2*i + 0], &dst[2*i + 1]);
lp_build_unpack2(gallivm, src_type, tmp_type, dst[i], &dst[2*i + 0], &dst[2*i + 1]);
}
src_type = tmp_type;
@@ -247,16 +250,17 @@ lp_build_unpack(LLVMBuilderRef builder,
* lp_build_packs2 instead.
*/
LLVMValueRef
lp_build_pack2(LLVMBuilderRef builder,
lp_build_pack2(struct gallivm_state *gallivm,
struct lp_type src_type,
struct lp_type dst_type,
LLVMValueRef lo,
LLVMValueRef hi)
{
LLVMBuilderRef builder = gallivm->builder;
#if HAVE_LLVM < 0x0207
LLVMTypeRef src_vec_type = lp_build_vec_type(src_type);
LLVMTypeRef src_vec_type = lp_build_vec_type(gallivm, src_type);
#endif
LLVMTypeRef dst_vec_type = lp_build_vec_type(dst_type);
LLVMTypeRef dst_vec_type = lp_build_vec_type(gallivm, dst_type);
LLVMValueRef shuffle;
LLVMValueRef res = NULL;
@@ -318,7 +322,7 @@ lp_build_pack2(LLVMBuilderRef builder,
lo = LLVMBuildBitCast(builder, lo, dst_vec_type, "");
hi = LLVMBuildBitCast(builder, hi, dst_vec_type, "");
shuffle = lp_build_const_pack_shuffle(dst_type.length);
shuffle = lp_build_const_pack_shuffle(gallivm, dst_type.length);
res = LLVMBuildShuffleVector(builder, lo, hi, shuffle, "");
@@ -334,7 +338,7 @@ lp_build_pack2(LLVMBuilderRef builder,
* destination type.
*/
LLVMValueRef
lp_build_packs2(LLVMBuilderRef builder,
lp_build_packs2(struct gallivm_state *gallivm,
struct lp_type src_type,
struct lp_type dst_type,
LLVMValueRef lo,
@@ -360,14 +364,14 @@ lp_build_packs2(LLVMBuilderRef builder,
if(clamp) {
struct lp_build_context bld;
unsigned dst_bits = dst_type.sign ? dst_type.width - 1 : dst_type.width;
LLVMValueRef dst_max = lp_build_const_int_vec(src_type, ((unsigned long long)1 << dst_bits) - 1);
lp_build_context_init(&bld, builder, src_type);
LLVMValueRef dst_max = lp_build_const_int_vec(gallivm, src_type, ((unsigned long long)1 << dst_bits) - 1);
lp_build_context_init(&bld, gallivm, src_type);
lo = lp_build_min(&bld, lo, dst_max);
hi = lp_build_min(&bld, hi, dst_max);
/* FIXME: What about lower bound? */
}
return lp_build_pack2(builder, src_type, dst_type, lo, hi);
return lp_build_pack2(gallivm, src_type, dst_type, lo, hi);
}
@@ -377,13 +381,13 @@ lp_build_packs2(LLVMBuilderRef builder,
* TODO: Handle saturation consistently.
*/
LLVMValueRef
lp_build_pack(LLVMBuilderRef builder,
lp_build_pack(struct gallivm_state *gallivm,
struct lp_type src_type,
struct lp_type dst_type,
boolean clamped,
const LLVMValueRef *src, unsigned num_srcs)
{
LLVMValueRef (*pack2)(LLVMBuilderRef builder,
LLVMValueRef (*pack2)(struct gallivm_state *gallivm,
struct lp_type src_type,
struct lp_type dst_type,
LLVMValueRef lo,
@@ -419,7 +423,8 @@ lp_build_pack(LLVMBuilderRef builder,
num_srcs /= 2;
for(i = 0; i < num_srcs; ++i)
tmp[i] = pack2(builder, src_type, tmp_type, tmp[2*i + 0], tmp[2*i + 1]);
tmp[i] = pack2(gallivm, src_type, tmp_type,
tmp[2*i + 0], tmp[2*i + 1]);
src_type = tmp_type;
}
@@ -437,12 +442,13 @@ lp_build_pack(LLVMBuilderRef builder,
* intrinsics that do saturation.
*/
void
lp_build_resize(LLVMBuilderRef builder,
lp_build_resize(struct gallivm_state *gallivm,
struct lp_type src_type,
struct lp_type dst_type,
const LLVMValueRef *src, unsigned num_srcs,
LLVMValueRef *dst, unsigned num_dsts)
{
LLVMBuilderRef builder = gallivm->builder;
LLVMValueRef tmp[LP_MAX_VECTOR_LENGTH];
unsigned i;
@@ -482,7 +488,7 @@ lp_build_resize(LLVMBuilderRef builder,
* Register width remains constant -- use vector packing intrinsics
*/
tmp[0] = lp_build_pack(builder, src_type, dst_type, TRUE, src, num_srcs);
tmp[0] = lp_build_pack(gallivm, src_type, dst_type, TRUE, src, num_srcs);
}
else {
/*
@@ -490,11 +496,11 @@ lp_build_resize(LLVMBuilderRef builder,
*/
assert(src_type.length == dst_type.length);
tmp[0] = lp_build_undef(dst_type);
tmp[0] = lp_build_undef(gallivm, dst_type);
for (i = 0; i < dst_type.length; ++i) {
LLVMValueRef index = LLVMConstInt(LLVMInt32Type(), i, 0);
LLVMValueRef index = lp_build_const_int32(gallivm, i);
LLVMValueRef val = LLVMBuildExtractElement(builder, src[0], index, "");
val = LLVMBuildTrunc(builder, val, lp_build_elem_type(dst_type), "");
val = LLVMBuildTrunc(builder, val, lp_build_elem_type(gallivm, dst_type), "");
tmp[0] = LLVMBuildInsertElement(builder, tmp[0], val, index, "");
}
}
@@ -510,7 +516,7 @@ lp_build_resize(LLVMBuilderRef builder,
/*
* Register width remains constant -- use vector unpack intrinsics
*/
lp_build_unpack(builder, src_type, dst_type, src[0], tmp, num_dsts);
lp_build_unpack(gallivm, src_type, dst_type, src[0], tmp, num_dsts);
}
else {
/*
@@ -518,15 +524,15 @@ lp_build_resize(LLVMBuilderRef builder,
*/
assert(src_type.length == dst_type.length);
tmp[0] = lp_build_undef(dst_type);
tmp[0] = lp_build_undef(gallivm, dst_type);
for (i = 0; i < dst_type.length; ++i) {
LLVMValueRef index = LLVMConstInt(LLVMInt32Type(), i, 0);
LLVMValueRef index = lp_build_const_int32(gallivm, i);
LLVMValueRef val = LLVMBuildExtractElement(builder, src[0], index, "");
if (src_type.sign && dst_type.sign) {
val = LLVMBuildSExt(builder, val, lp_build_elem_type(dst_type), "");
val = LLVMBuildSExt(builder, val, lp_build_elem_type(gallivm, dst_type), "");
} else {
val = LLVMBuildZExt(builder, val, lp_build_elem_type(dst_type), "");
val = LLVMBuildZExt(builder, val, lp_build_elem_type(gallivm, dst_type), "");
}
tmp[0] = LLVMBuildInsertElement(builder, tmp[0], val, index, "");
}

View File

@@ -46,7 +46,7 @@ struct lp_type;
LLVMValueRef
lp_build_interleave2(LLVMBuilderRef builder,
lp_build_interleave2(struct gallivm_state *gallivm,
struct lp_type type,
LLVMValueRef a,
LLVMValueRef b,
@@ -54,7 +54,7 @@ lp_build_interleave2(LLVMBuilderRef builder,
void
lp_build_unpack2(LLVMBuilderRef builder,
lp_build_unpack2(struct gallivm_state *gallivm,
struct lp_type src_type,
struct lp_type dst_type,
LLVMValueRef src,
@@ -63,7 +63,7 @@ lp_build_unpack2(LLVMBuilderRef builder,
void
lp_build_unpack(LLVMBuilderRef builder,
lp_build_unpack(struct gallivm_state *gallivm,
struct lp_type src_type,
struct lp_type dst_type,
LLVMValueRef src,
@@ -71,7 +71,7 @@ lp_build_unpack(LLVMBuilderRef builder,
LLVMValueRef
lp_build_packs2(LLVMBuilderRef builder,
lp_build_packs2(struct gallivm_state *gallivm,
struct lp_type src_type,
struct lp_type dst_type,
LLVMValueRef lo,
@@ -79,7 +79,7 @@ lp_build_packs2(LLVMBuilderRef builder,
LLVMValueRef
lp_build_pack2(LLVMBuilderRef builder,
lp_build_pack2(struct gallivm_state *gallivm,
struct lp_type src_type,
struct lp_type dst_type,
LLVMValueRef lo,
@@ -87,7 +87,7 @@ lp_build_pack2(LLVMBuilderRef builder,
LLVMValueRef
lp_build_pack(LLVMBuilderRef builder,
lp_build_pack(struct gallivm_state *gallivm,
struct lp_type src_type,
struct lp_type dst_type,
boolean clamped,
@@ -95,7 +95,7 @@ lp_build_pack(LLVMBuilderRef builder,
void
lp_build_resize(LLVMBuilderRef builder,
lp_build_resize(struct gallivm_state *gallivm,
struct lp_type src_type,
struct lp_type dst_type,
const LLVMValueRef *src, unsigned num_srcs,

View File

@@ -31,6 +31,8 @@
#include "util/u_memory.h"
#include "util/u_string.h"
#include "lp_bld_const.h"
#include "lp_bld_init.h"
#include "lp_bld_const.h"
#include "lp_bld_printf.h"
@@ -65,12 +67,14 @@ lp_get_printf_arg_count(const char *fmt)
}
LLVMValueRef
lp_build_const_string_variable(LLVMModuleRef module, const char *str, int len)
lp_build_const_string_variable(LLVMModuleRef module,
LLVMContextRef context,
const char *str, int len)
{
LLVMValueRef string = LLVMAddGlobal(module, LLVMArrayType(LLVMInt8Type(), len + 1), "");
LLVMValueRef string = LLVMAddGlobal(module, LLVMArrayType(LLVMInt8TypeInContext(context), len + 1), "");
LLVMSetGlobalConstant(string, TRUE);
LLVMSetLinkage(string, LLVMInternalLinkage);
LLVMSetInitializer(string, LLVMConstString(str, len + 1, TRUE));
LLVMSetInitializer(string, LLVMConstStringInContext(context, str, len + 1, TRUE));
return string;
}
@@ -83,15 +87,18 @@ lp_build_const_string_variable(LLVMModuleRef module, const char *str, int len)
* LLVMValueRef.
*/
LLVMValueRef
lp_build_printf(LLVMBuilderRef builder, const char *fmt, ...)
lp_build_printf(struct gallivm_state *gallivm, const char *fmt, ...)
{
va_list arglist;
int i = 0;
int argcount = lp_get_printf_arg_count(fmt);
LLVMModuleRef module = LLVMGetGlobalParent(LLVMGetBasicBlockParent(LLVMGetInsertBlock(builder)));
LLVMBuilderRef builder = gallivm->builder;
LLVMContextRef context = gallivm->context;
LLVMModuleRef module = gallivm->module;
LLVMValueRef params[50];
LLVMValueRef fmtarg = lp_build_const_string_variable(module, fmt, strlen(fmt) + 1);
LLVMValueRef int0 = LLVMConstInt(LLVMInt32Type(), 0, 0);
LLVMValueRef fmtarg = lp_build_const_string_variable(module, context,
fmt, strlen(fmt) + 1);
LLVMValueRef int0 = lp_build_const_int32(gallivm, 0);
LLVMValueRef index[2];
LLVMValueRef func_printf = LLVMGetNamedFunction(module, "printf");
@@ -100,7 +107,7 @@ lp_build_printf(LLVMBuilderRef builder, const char *fmt, ...)
index[0] = index[1] = int0;
if (!func_printf) {
LLVMTypeRef printf_type = LLVMFunctionType(LLVMIntType(32), NULL, 0, 1);
LLVMTypeRef printf_type = LLVMFunctionType(LLVMIntTypeInContext(context, 32), NULL, 0, 1);
func_printf = LLVMAddFunction(module, "printf", printf_type);
}
@@ -113,7 +120,7 @@ lp_build_printf(LLVMBuilderRef builder, const char *fmt, ...)
/* printf wants doubles, so lets convert so that
* we can actually print them */
if (LLVMGetTypeKind(type) == LLVMFloatTypeKind)
val = LLVMBuildFPExt(builder, val, LLVMDoubleType(), "");
val = LLVMBuildFPExt(builder, val, LLVMDoubleTypeInContext(context), "");
params[i] = val;
}
va_end(arglist);
@@ -127,16 +134,18 @@ lp_build_printf(LLVMBuilderRef builder, const char *fmt, ...)
* Print a float[4] vector.
*/
LLVMValueRef
lp_build_print_vec4(LLVMBuilderRef builder, const char *msg, LLVMValueRef vec)
lp_build_print_vec4(struct gallivm_state *gallivm,
const char *msg, LLVMValueRef vec)
{
LLVMBuilderRef builder = gallivm->builder;
char format[1000];
LLVMValueRef x, y, z, w;
x = LLVMBuildExtractElement(builder, vec, lp_build_const_int32(0), "");
y = LLVMBuildExtractElement(builder, vec, lp_build_const_int32(1), "");
z = LLVMBuildExtractElement(builder, vec, lp_build_const_int32(2), "");
w = LLVMBuildExtractElement(builder, vec, lp_build_const_int32(3), "");
x = LLVMBuildExtractElement(builder, vec, lp_build_const_int32(gallivm, 0), "");
y = LLVMBuildExtractElement(builder, vec, lp_build_const_int32(gallivm, 1), "");
z = LLVMBuildExtractElement(builder, vec, lp_build_const_int32(gallivm, 2), "");
w = LLVMBuildExtractElement(builder, vec, lp_build_const_int32(gallivm, 3), "");
util_snprintf(format, sizeof(format), "%s %%f %%f %%f %%f\n", msg);
return lp_build_printf(builder, format, x, y, z, w);
return lp_build_printf(gallivm, format, x, y, z, w);
}

View File

@@ -31,12 +31,19 @@
#include "pipe/p_compiler.h"
#include "lp_bld.h"
#include "lp_bld_init.h"
LLVMValueRef lp_build_const_string_variable(LLVMModuleRef module, const char *str, int len);
LLVMValueRef lp_build_printf(LLVMBuilderRef builder, const char *fmt, ...);
LLVMValueRef lp_build_const_string_variable(LLVMModuleRef module,
LLVMContextRef context,
const char *str, int len);
LLVMValueRef lp_build_printf(struct gallivm_state *gallivm,
const char *fmt, ...);
LLVMValueRef
lp_build_print_vec4(LLVMBuilderRef builder, const char *msg, LLVMValueRef vec);
lp_build_print_vec4(struct gallivm_state *gallivm,
const char *msg, LLVMValueRef vec);
#endif

View File

@@ -28,6 +28,7 @@
#include "lp_bld_type.h"
#include "lp_bld_arit.h"
#include "lp_bld_const.h"
#include "lp_bld_swizzle.h"
#include "lp_bld_quad.h"
@@ -81,15 +82,15 @@ LLVMValueRef
lp_build_scalar_ddx(struct lp_build_context *bld,
LLVMValueRef a)
{
LLVMTypeRef i32t = LLVMInt32Type();
LLVMValueRef idx_left = LLVMConstInt(i32t, LP_BLD_QUAD_TOP_LEFT, 0);
LLVMValueRef idx_right = LLVMConstInt(i32t, LP_BLD_QUAD_TOP_RIGHT, 0);
LLVMValueRef a_left = LLVMBuildExtractElement(bld->builder, a, idx_left, "left");
LLVMValueRef a_right = LLVMBuildExtractElement(bld->builder, a, idx_right, "right");
LLVMBuilderRef builder = bld->gallivm->builder;
LLVMValueRef idx_left = lp_build_const_int32(bld->gallivm, LP_BLD_QUAD_TOP_LEFT);
LLVMValueRef idx_right = lp_build_const_int32(bld->gallivm, LP_BLD_QUAD_TOP_RIGHT);
LLVMValueRef a_left = LLVMBuildExtractElement(builder, a, idx_left, "left");
LLVMValueRef a_right = LLVMBuildExtractElement(builder, a, idx_right, "right");
if (bld->type.floating)
return LLVMBuildFSub(bld->builder, a_right, a_left, "ddx");
return LLVMBuildFSub(builder, a_right, a_left, "ddx");
else
return LLVMBuildSub(bld->builder, a_right, a_left, "ddx");
return LLVMBuildSub(builder, a_right, a_left, "ddx");
}
@@ -97,13 +98,13 @@ LLVMValueRef
lp_build_scalar_ddy(struct lp_build_context *bld,
LLVMValueRef a)
{
LLVMTypeRef i32t = LLVMInt32Type();
LLVMValueRef idx_top = LLVMConstInt(i32t, LP_BLD_QUAD_TOP_LEFT, 0);
LLVMValueRef idx_bottom = LLVMConstInt(i32t, LP_BLD_QUAD_BOTTOM_LEFT, 0);
LLVMValueRef a_top = LLVMBuildExtractElement(bld->builder, a, idx_top, "top");
LLVMValueRef a_bottom = LLVMBuildExtractElement(bld->builder, a, idx_bottom, "bottom");
LLVMBuilderRef builder = bld->gallivm->builder;
LLVMValueRef idx_top = lp_build_const_int32(bld->gallivm, LP_BLD_QUAD_TOP_LEFT);
LLVMValueRef idx_bottom = lp_build_const_int32(bld->gallivm, LP_BLD_QUAD_BOTTOM_LEFT);
LLVMValueRef a_top = LLVMBuildExtractElement(builder, a, idx_top, "top");
LLVMValueRef a_bottom = LLVMBuildExtractElement(builder, a, idx_bottom, "bottom");
if (bld->type.floating)
return LLVMBuildFSub(bld->builder, a_bottom, a_top, "ddy");
return LLVMBuildFSub(builder, a_bottom, a_top, "ddy");
else
return LLVMBuildSub(bld->builder, a_bottom, a_top, "ddy");
return LLVMBuildSub(builder, a_bottom, a_top, "ddy");
}

View File

@@ -134,7 +134,7 @@ lp_sampler_static_state(struct lp_sampler_static_state *state,
state->min_img_filter = sampler->min_img_filter;
state->mag_img_filter = sampler->mag_img_filter;
if (view->last_level && sampler->max_lod > 0.0f) {
if (view->u.tex.last_level && sampler->max_lod > 0.0f) {
state->min_mip_filter = sampler->min_mip_filter;
} else {
state->min_mip_filter = PIPE_TEX_MIPFILTER_NONE;
@@ -155,7 +155,7 @@ lp_sampler_static_state(struct lp_sampler_static_state *state,
state->apply_min_lod = 1;
}
if (sampler->max_lod < (float)view->last_level) {
if (sampler->max_lod < (float)view->u.tex.last_level) {
state->apply_max_lod = 1;
}
}
@@ -190,7 +190,8 @@ lp_build_rho(struct lp_build_sample_context *bld,
struct lp_build_context *float_size_bld = &bld->float_size_bld;
struct lp_build_context *float_bld = &bld->float_bld;
const unsigned dims = bld->dims;
LLVMTypeRef i32t = LLVMInt32Type();
LLVMBuilderRef builder = bld->gallivm->builder;
LLVMTypeRef i32t = LLVMInt32TypeInContext(bld->gallivm->context);
LLVMValueRef index0 = LLVMConstInt(i32t, 0, 0);
LLVMValueRef index1 = LLVMConstInt(i32t, 1, 0);
LLVMValueRef index2 = LLVMConstInt(i32t, 2, 0);
@@ -211,21 +212,21 @@ lp_build_rho(struct lp_build_sample_context *bld,
rho_x = float_size_bld->undef;
rho_y = float_size_bld->undef;
rho_x = LLVMBuildInsertElement(bld->builder, rho_x, dsdx, index0, "");
rho_y = LLVMBuildInsertElement(bld->builder, rho_y, dsdy, index0, "");
rho_x = LLVMBuildInsertElement(builder, rho_x, dsdx, index0, "");
rho_y = LLVMBuildInsertElement(builder, rho_y, dsdy, index0, "");
dtdx = ddx[1];
dtdy = ddy[1];
rho_x = LLVMBuildInsertElement(bld->builder, rho_x, dtdx, index1, "");
rho_y = LLVMBuildInsertElement(bld->builder, rho_y, dtdy, index1, "");
rho_x = LLVMBuildInsertElement(builder, rho_x, dtdx, index1, "");
rho_y = LLVMBuildInsertElement(builder, rho_y, dtdy, index1, "");
if (dims >= 3) {
drdx = ddx[2];
drdy = ddy[2];
rho_x = LLVMBuildInsertElement(bld->builder, rho_x, drdx, index2, "");
rho_y = LLVMBuildInsertElement(bld->builder, rho_y, drdy, index2, "");
rho_x = LLVMBuildInsertElement(builder, rho_x, drdx, index2, "");
rho_y = LLVMBuildInsertElement(builder, rho_y, drdy, index2, "");
}
}
@@ -245,13 +246,13 @@ lp_build_rho(struct lp_build_sample_context *bld,
if (dims >= 2) {
LLVMValueRef rho_s, rho_t, rho_r;
rho_s = LLVMBuildExtractElement(bld->builder, rho_vec, index0, "");
rho_t = LLVMBuildExtractElement(bld->builder, rho_vec, index1, "");
rho_s = LLVMBuildExtractElement(builder, rho_vec, index0, "");
rho_t = LLVMBuildExtractElement(builder, rho_vec, index1, "");
rho = lp_build_max(float_bld, rho_s, rho_t);
if (dims >= 3) {
rho_r = LLVMBuildExtractElement(bld->builder, rho_vec, index0, "");
rho_r = LLVMBuildExtractElement(builder, rho_vec, index0, "");
rho = lp_build_max(float_bld, rho, rho_r);
}
}
@@ -304,19 +305,19 @@ lp_build_brilinear_lod(struct lp_build_context *bld,
double post_offset = 1 - factor;
if (0) {
lp_build_printf(bld->builder, "lod = %f\n", lod);
lp_build_printf(bld->gallivm, "lod = %f\n", lod);
}
lod = lp_build_add(bld, lod,
lp_build_const_vec(bld->type, pre_offset));
lp_build_const_vec(bld->gallivm, bld->type, pre_offset));
lp_build_ifloor_fract(bld, lod, out_lod_ipart, &lod_fpart);
lod_fpart = lp_build_mul(bld, lod_fpart,
lp_build_const_vec(bld->type, factor));
lp_build_const_vec(bld->gallivm, bld->type, factor));
lod_fpart = lp_build_add(bld, lod_fpart,
lp_build_const_vec(bld->type, post_offset));
lp_build_const_vec(bld->gallivm, bld->type, post_offset));
/*
* It's not necessary to clamp lod_fpart since:
@@ -327,8 +328,8 @@ lp_build_brilinear_lod(struct lp_build_context *bld,
*out_lod_fpart = lod_fpart;
if (0) {
lp_build_printf(bld->builder, "lod_ipart = %i\n", *out_lod_ipart);
lp_build_printf(bld->builder, "lod_fpart = %f\n\n", *out_lod_fpart);
lp_build_printf(bld->gallivm, "lod_ipart = %i\n", *out_lod_ipart);
lp_build_printf(bld->gallivm, "lod_fpart = %f\n\n", *out_lod_fpart);
}
}
@@ -363,7 +364,7 @@ lp_build_brilinear_rho(struct lp_build_context *bld,
* part will not need any post adjustments.
*/
rho = lp_build_mul(bld, rho,
lp_build_const_vec(bld->type, pre_factor));
lp_build_const_vec(bld->gallivm, bld->type, pre_factor));
/* ipart = ifloor(log2(rho)) */
lod_ipart = lp_build_extract_exponent(bld, rho, 0);
@@ -372,10 +373,10 @@ lp_build_brilinear_rho(struct lp_build_context *bld,
lod_fpart = lp_build_extract_mantissa(bld, rho);
lod_fpart = lp_build_mul(bld, lod_fpart,
lp_build_const_vec(bld->type, factor));
lp_build_const_vec(bld->gallivm, bld->type, factor));
lod_fpart = lp_build_add(bld, lod_fpart,
lp_build_const_vec(bld->type, post_offset));
lp_build_const_vec(bld->gallivm, bld->type, post_offset));
/*
* Like lp_build_brilinear_lod, it's not necessary to clamp lod_fpart since:
@@ -413,6 +414,7 @@ lp_build_lod_selector(struct lp_build_sample_context *bld,
LLVMValueRef *out_lod_fpart)
{
LLVMBuilderRef builder = bld->gallivm->builder;
struct lp_build_context *float_bld = &bld->float_bld;
LLVMValueRef lod;
@@ -424,17 +426,17 @@ lp_build_lod_selector(struct lp_build_sample_context *bld,
* This is hit during mipmap generation.
*/
LLVMValueRef min_lod =
bld->dynamic_state->min_lod(bld->dynamic_state, bld->builder, unit);
bld->dynamic_state->min_lod(bld->dynamic_state, bld->gallivm, unit);
lod = min_lod;
}
else {
LLVMValueRef sampler_lod_bias =
bld->dynamic_state->lod_bias(bld->dynamic_state, bld->builder, unit);
LLVMValueRef index0 = LLVMConstInt(LLVMInt32Type(), 0, 0);
bld->dynamic_state->lod_bias(bld->dynamic_state, bld->gallivm, unit);
LLVMValueRef index0 = lp_build_const_int32(bld->gallivm, 0);
if (explicit_lod) {
lod = LLVMBuildExtractElement(bld->builder, explicit_lod,
lod = LLVMBuildExtractElement(builder, explicit_lod,
index0, "");
}
else {
@@ -479,27 +481,27 @@ lp_build_lod_selector(struct lp_build_sample_context *bld,
/* add shader lod bias */
if (lod_bias) {
lod_bias = LLVMBuildExtractElement(bld->builder, lod_bias,
lod_bias = LLVMBuildExtractElement(builder, lod_bias,
index0, "");
lod = LLVMBuildFAdd(bld->builder, lod, lod_bias, "shader_lod_bias");
lod = LLVMBuildFAdd(builder, lod, lod_bias, "shader_lod_bias");
}
}
/* add sampler lod bias */
if (bld->static_state->lod_bias_non_zero)
lod = LLVMBuildFAdd(bld->builder, lod, sampler_lod_bias, "sampler_lod_bias");
lod = LLVMBuildFAdd(builder, lod, sampler_lod_bias, "sampler_lod_bias");
/* clamp lod */
if (bld->static_state->apply_max_lod) {
LLVMValueRef max_lod =
bld->dynamic_state->max_lod(bld->dynamic_state, bld->builder, unit);
bld->dynamic_state->max_lod(bld->dynamic_state, bld->gallivm, unit);
lod = lp_build_min(float_bld, lod, max_lod);
}
if (bld->static_state->apply_min_lod) {
LLVMValueRef min_lod =
bld->dynamic_state->min_lod(bld->dynamic_state, bld->builder, unit);
bld->dynamic_state->min_lod(bld->dynamic_state, bld->gallivm, unit);
lod = lp_build_max(float_bld, lod, min_lod);
}
@@ -542,10 +544,10 @@ lp_build_nearest_mip_level(struct lp_build_sample_context *bld,
struct lp_build_context *int_bld = &bld->int_bld;
LLVMValueRef last_level, level;
LLVMValueRef zero = LLVMConstInt(LLVMInt32Type(), 0, 0);
LLVMValueRef zero = lp_build_const_int32(bld->gallivm, 0);
last_level = bld->dynamic_state->last_level(bld->dynamic_state,
bld->builder, unit);
bld->gallivm, unit);
/* convert float lod to integer */
level = lod_ipart;
@@ -568,7 +570,7 @@ lp_build_linear_mip_levels(struct lp_build_sample_context *bld,
LLVMValueRef *level0_out,
LLVMValueRef *level1_out)
{
LLVMBuilderRef builder = bld->builder;
LLVMBuilderRef builder = bld->gallivm->builder;
struct lp_build_context *int_bld = &bld->int_bld;
struct lp_build_context *float_bld = &bld->float_bld;
LLVMValueRef last_level;
@@ -579,7 +581,7 @@ lp_build_linear_mip_levels(struct lp_build_sample_context *bld,
*level1_out = lp_build_add(int_bld, lod_ipart, int_bld->one);
last_level = bld->dynamic_state->last_level(bld->dynamic_state,
bld->builder, unit);
bld->gallivm, unit);
/*
* Clamp both lod_ipart and lod_ipart + 1 to [0, last_level], with the
@@ -630,11 +632,13 @@ LLVMValueRef
lp_build_get_mipmap_level(struct lp_build_sample_context *bld,
LLVMValueRef level)
{
LLVMBuilderRef builder = bld->gallivm->builder;
LLVMValueRef indexes[2], data_ptr;
indexes[0] = LLVMConstInt(LLVMInt32Type(), 0, 0);
indexes[0] = lp_build_const_int32(bld->gallivm, 0);
indexes[1] = level;
data_ptr = LLVMBuildGEP(bld->builder, bld->data_array, indexes, 2, "");
data_ptr = LLVMBuildLoad(bld->builder, data_ptr, "");
data_ptr = LLVMBuildGEP(builder, bld->data_array, indexes, 2, "");
data_ptr = LLVMBuildLoad(builder, data_ptr, "");
return data_ptr;
}
@@ -643,7 +647,7 @@ LLVMValueRef
lp_build_get_const_mipmap_level(struct lp_build_sample_context *bld,
int level)
{
LLVMValueRef lvl = LLVMConstInt(LLVMInt32Type(), level, 0);
LLVMValueRef lvl = lp_build_const_int32(bld->gallivm, level);
return lp_build_get_mipmap_level(bld, lvl);
}
@@ -657,6 +661,7 @@ lp_build_minify(struct lp_build_context *bld,
LLVMValueRef base_size,
LLVMValueRef level)
{
LLVMBuilderRef builder = bld->gallivm->builder;
assert(lp_check_value(bld->type, base_size));
assert(lp_check_value(bld->type, level));
@@ -666,7 +671,7 @@ lp_build_minify(struct lp_build_context *bld,
}
else {
LLVMValueRef size =
LLVMBuildLShr(bld->builder, base_size, level, "minify");
LLVMBuildLShr(builder, base_size, level, "minify");
assert(bld->type.sign);
size = lp_build_max(bld, size, bld->one);
return size;
@@ -682,11 +687,12 @@ static LLVMValueRef
lp_build_get_level_stride_vec(struct lp_build_sample_context *bld,
LLVMValueRef stride_array, LLVMValueRef level)
{
LLVMBuilderRef builder = bld->gallivm->builder;
LLVMValueRef indexes[2], stride;
indexes[0] = LLVMConstInt(LLVMInt32Type(), 0, 0);
indexes[0] = lp_build_const_int32(bld->gallivm, 0);
indexes[1] = level;
stride = LLVMBuildGEP(bld->builder, stride_array, indexes, 2, "");
stride = LLVMBuildLoad(bld->builder, stride, "");
stride = LLVMBuildGEP(builder, stride_array, indexes, 2, "");
stride = LLVMBuildLoad(builder, stride, "");
stride = lp_build_broadcast_scalar(&bld->int_coord_bld, stride);
return stride;
}
@@ -747,21 +753,21 @@ lp_build_extract_image_sizes(struct lp_build_sample_context *bld,
LLVMValueRef *out_depth)
{
const unsigned dims = bld->dims;
LLVMTypeRef i32t = LLVMInt32Type();
LLVMTypeRef i32t = LLVMInt32TypeInContext(bld->gallivm->context);
*out_width = lp_build_extract_broadcast(bld->builder,
*out_width = lp_build_extract_broadcast(bld->gallivm,
size_type,
coord_type,
size,
LLVMConstInt(i32t, 0, 0));
if (dims >= 2) {
*out_height = lp_build_extract_broadcast(bld->builder,
*out_height = lp_build_extract_broadcast(bld->gallivm,
size_type,
coord_type,
size,
LLVMConstInt(i32t, 1, 0));
if (dims == 3) {
*out_depth = lp_build_extract_broadcast(bld->builder,
*out_depth = lp_build_extract_broadcast(bld->gallivm,
size_type,
coord_type,
size,
@@ -812,7 +818,7 @@ static LLVMValueRef
lp_build_cube_ima(struct lp_build_context *coord_bld, LLVMValueRef coord)
{
/* ima = -0.5 / abs(coord); */
LLVMValueRef negHalf = lp_build_const_vec(coord_bld->type, -0.5);
LLVMValueRef negHalf = lp_build_const_vec(coord_bld->gallivm, coord_bld->type, -0.5);
LLVMValueRef absCoord = lp_build_abs(coord_bld, coord);
LLVMValueRef ima = lp_build_div(coord_bld, negHalf, absCoord);
return ima;
@@ -831,7 +837,7 @@ lp_build_cube_coord(struct lp_build_context *coord_bld,
LLVMValueRef coord, LLVMValueRef ima)
{
/* return negate(coord) * ima * sign + 0.5; */
LLVMValueRef half = lp_build_const_vec(coord_bld->type, 0.5);
LLVMValueRef half = lp_build_const_vec(coord_bld->gallivm, coord_bld->type, 0.5);
LLVMValueRef res;
assert(negate_coord == +1 || negate_coord == -1);
@@ -859,12 +865,14 @@ lp_build_cube_face(struct lp_build_sample_context *bld,
LLVMValueRef major_coord,
unsigned pos_face, unsigned neg_face)
{
LLVMValueRef cmp = LLVMBuildFCmp(bld->builder, LLVMRealUGE,
struct gallivm_state *gallivm = bld->gallivm;
LLVMBuilderRef builder = gallivm->builder;
LLVMValueRef cmp = LLVMBuildFCmp(builder, LLVMRealUGE,
major_coord,
bld->float_bld.zero, "");
LLVMValueRef pos = LLVMConstInt(LLVMInt32Type(), pos_face, 0);
LLVMValueRef neg = LLVMConstInt(LLVMInt32Type(), neg_face, 0);
LLVMValueRef res = LLVMBuildSelect(bld->builder, cmp, pos, neg, "");
LLVMValueRef pos = lp_build_const_int32(gallivm, pos_face);
LLVMValueRef neg = lp_build_const_int32(gallivm, neg_face);
LLVMValueRef res = LLVMBuildSelect(builder, cmp, pos, neg, "");
return res;
}
@@ -884,9 +892,10 @@ lp_build_cube_lookup(struct lp_build_sample_context *bld,
{
struct lp_build_context *float_bld = &bld->float_bld;
struct lp_build_context *coord_bld = &bld->coord_bld;
LLVMBuilderRef builder = bld->gallivm->builder;
LLVMValueRef rx, ry, rz;
LLVMValueRef arx, ary, arz;
LLVMValueRef c25 = LLVMConstReal(LLVMFloatType(), 0.25);
LLVMValueRef c25 = lp_build_const_float(bld->gallivm, 0.25);
LLVMValueRef arx_ge_ary, arx_ge_arz;
LLVMValueRef ary_ge_arx, ary_ge_arz;
LLVMValueRef arx_ge_ary_arz, ary_ge_arx_arz;
@@ -911,17 +920,17 @@ lp_build_cube_lookup(struct lp_build_sample_context *bld,
/*
* Compare sign/magnitude of rx,ry,rz to determine face
*/
arx_ge_ary = LLVMBuildFCmp(bld->builder, LLVMRealUGE, arx, ary, "");
arx_ge_arz = LLVMBuildFCmp(bld->builder, LLVMRealUGE, arx, arz, "");
ary_ge_arx = LLVMBuildFCmp(bld->builder, LLVMRealUGE, ary, arx, "");
ary_ge_arz = LLVMBuildFCmp(bld->builder, LLVMRealUGE, ary, arz, "");
arx_ge_ary = LLVMBuildFCmp(builder, LLVMRealUGE, arx, ary, "");
arx_ge_arz = LLVMBuildFCmp(builder, LLVMRealUGE, arx, arz, "");
ary_ge_arx = LLVMBuildFCmp(builder, LLVMRealUGE, ary, arx, "");
ary_ge_arz = LLVMBuildFCmp(builder, LLVMRealUGE, ary, arz, "");
arx_ge_ary_arz = LLVMBuildAnd(bld->builder, arx_ge_ary, arx_ge_arz, "");
ary_ge_arx_arz = LLVMBuildAnd(bld->builder, ary_ge_arx, ary_ge_arz, "");
arx_ge_ary_arz = LLVMBuildAnd(builder, arx_ge_ary, arx_ge_arz, "");
ary_ge_arx_arz = LLVMBuildAnd(builder, ary_ge_arx, ary_ge_arz, "");
rx_pos = LLVMBuildFCmp(bld->builder, LLVMRealUGE, rx, float_bld->zero, "");
ry_pos = LLVMBuildFCmp(bld->builder, LLVMRealUGE, ry, float_bld->zero, "");
rz_pos = LLVMBuildFCmp(bld->builder, LLVMRealUGE, rz, float_bld->zero, "");
rx_pos = LLVMBuildFCmp(builder, LLVMRealUGE, rx, float_bld->zero, "");
ry_pos = LLVMBuildFCmp(builder, LLVMRealUGE, ry, float_bld->zero, "");
rz_pos = LLVMBuildFCmp(builder, LLVMRealUGE, rz, float_bld->zero, "");
{
struct lp_build_if_state if_ctx;
@@ -929,11 +938,11 @@ lp_build_cube_lookup(struct lp_build_sample_context *bld,
LLVMValueRef face_t_var;
LLVMValueRef face_var;
face_s_var = lp_build_alloca(bld->builder, bld->coord_bld.vec_type, "face_s_var");
face_t_var = lp_build_alloca(bld->builder, bld->coord_bld.vec_type, "face_t_var");
face_var = lp_build_alloca(bld->builder, bld->int_bld.vec_type, "face_var");
face_s_var = lp_build_alloca(bld->gallivm, bld->coord_bld.vec_type, "face_s_var");
face_t_var = lp_build_alloca(bld->gallivm, bld->coord_bld.vec_type, "face_t_var");
face_var = lp_build_alloca(bld->gallivm, bld->int_bld.vec_type, "face_var");
lp_build_if(&if_ctx, bld->builder, arx_ge_ary_arz);
lp_build_if(&if_ctx, bld->gallivm, arx_ge_ary_arz);
{
/* +/- X face */
LLVMValueRef sign = lp_build_sgn(float_bld, rx);
@@ -943,17 +952,17 @@ lp_build_cube_lookup(struct lp_build_sample_context *bld,
*face = lp_build_cube_face(bld, rx,
PIPE_TEX_FACE_POS_X,
PIPE_TEX_FACE_NEG_X);
LLVMBuildStore(bld->builder, *face_s, face_s_var);
LLVMBuildStore(bld->builder, *face_t, face_t_var);
LLVMBuildStore(bld->builder, *face, face_var);
LLVMBuildStore(builder, *face_s, face_s_var);
LLVMBuildStore(builder, *face_t, face_t_var);
LLVMBuildStore(builder, *face, face_var);
}
lp_build_else(&if_ctx);
{
struct lp_build_if_state if_ctx2;
ary_ge_arx_arz = LLVMBuildAnd(bld->builder, ary_ge_arx, ary_ge_arz, "");
ary_ge_arx_arz = LLVMBuildAnd(builder, ary_ge_arx, ary_ge_arz, "");
lp_build_if(&if_ctx2, bld->builder, ary_ge_arx_arz);
lp_build_if(&if_ctx2, bld->gallivm, ary_ge_arx_arz);
{
/* +/- Y face */
LLVMValueRef sign = lp_build_sgn(float_bld, ry);
@@ -963,9 +972,9 @@ lp_build_cube_lookup(struct lp_build_sample_context *bld,
*face = lp_build_cube_face(bld, ry,
PIPE_TEX_FACE_POS_Y,
PIPE_TEX_FACE_NEG_Y);
LLVMBuildStore(bld->builder, *face_s, face_s_var);
LLVMBuildStore(bld->builder, *face_t, face_t_var);
LLVMBuildStore(bld->builder, *face, face_var);
LLVMBuildStore(builder, *face_s, face_s_var);
LLVMBuildStore(builder, *face_t, face_t_var);
LLVMBuildStore(builder, *face, face_var);
}
lp_build_else(&if_ctx2);
{
@@ -977,18 +986,18 @@ lp_build_cube_lookup(struct lp_build_sample_context *bld,
*face = lp_build_cube_face(bld, rz,
PIPE_TEX_FACE_POS_Z,
PIPE_TEX_FACE_NEG_Z);
LLVMBuildStore(bld->builder, *face_s, face_s_var);
LLVMBuildStore(bld->builder, *face_t, face_t_var);
LLVMBuildStore(bld->builder, *face, face_var);
LLVMBuildStore(builder, *face_s, face_s_var);
LLVMBuildStore(builder, *face_t, face_t_var);
LLVMBuildStore(builder, *face, face_var);
}
lp_build_endif(&if_ctx2);
}
lp_build_endif(&if_ctx);
*face_s = LLVMBuildLoad(bld->builder, face_s_var, "face_s");
*face_t = LLVMBuildLoad(bld->builder, face_t_var, "face_t");
*face = LLVMBuildLoad(bld->builder, face_var, "face");
*face_s = LLVMBuildLoad(builder, face_s_var, "face_s");
*face_t = LLVMBuildLoad(builder, face_t_var, "face_t");
*face = LLVMBuildLoad(builder, face_var, "face");
}
}
@@ -1011,6 +1020,7 @@ lp_build_sample_partial_offset(struct lp_build_context *bld,
LLVMValueRef *out_offset,
LLVMValueRef *out_subcoord)
{
LLVMBuilderRef builder = bld->gallivm->builder;
LLVMValueRef offset;
LLVMValueRef subcoord;
@@ -1028,14 +1038,14 @@ lp_build_sample_partial_offset(struct lp_build_context *bld,
*/
#if 0
LLVMValueRef block_width = lp_build_const_int_vec(bld->type, block_length);
subcoord = LLVMBuildURem(bld->builder, coord, block_width, "");
coord = LLVMBuildUDiv(bld->builder, coord, block_width, "");
subcoord = LLVMBuildURem(builder, coord, block_width, "");
coord = LLVMBuildUDiv(builder, coord, block_width, "");
#else
unsigned logbase2 = util_unsigned_logbase2(block_length);
LLVMValueRef block_shift = lp_build_const_int_vec(bld->type, logbase2);
LLVMValueRef block_mask = lp_build_const_int_vec(bld->type, block_length - 1);
subcoord = LLVMBuildAnd(bld->builder, coord, block_mask, "");
coord = LLVMBuildLShr(bld->builder, coord, block_shift, "");
LLVMValueRef block_shift = lp_build_const_int_vec(bld->gallivm, bld->type, logbase2);
LLVMValueRef block_mask = lp_build_const_int_vec(bld->gallivm, bld->type, block_length - 1);
subcoord = LLVMBuildAnd(builder, coord, block_mask, "");
coord = LLVMBuildLShr(builder, coord, block_shift, "");
#endif
}
@@ -1071,7 +1081,8 @@ lp_build_sample_offset(struct lp_build_context *bld,
LLVMValueRef x_stride;
LLVMValueRef offset;
x_stride = lp_build_const_vec(bld->type, format_desc->block.bits/8);
x_stride = lp_build_const_vec(bld->gallivm, bld->type,
format_desc->block.bits/8);
lp_build_sample_partial_offset(bld,
format_desc->block.width,

View File

@@ -105,64 +105,64 @@ struct lp_sampler_dynamic_state
/** Obtain the base texture width (returns int32) */
LLVMValueRef
(*width)( const struct lp_sampler_dynamic_state *state,
LLVMBuilderRef builder,
struct gallivm_state *gallivm,
unsigned unit);
/** Obtain the base texture height (returns int32) */
LLVMValueRef
(*height)( const struct lp_sampler_dynamic_state *state,
LLVMBuilderRef builder,
struct gallivm_state *gallivm,
unsigned unit);
/** Obtain the base texture depth (returns int32) */
LLVMValueRef
(*depth)( const struct lp_sampler_dynamic_state *state,
LLVMBuilderRef builder,
struct gallivm_state *gallivm,
unsigned unit);
/** Obtain the number of mipmap levels minus one (returns int32) */
LLVMValueRef
(*last_level)( const struct lp_sampler_dynamic_state *state,
LLVMBuilderRef builder,
struct gallivm_state *gallivm,
unsigned unit);
/** Obtain stride in bytes between image rows/blocks (returns int32) */
LLVMValueRef
(*row_stride)( const struct lp_sampler_dynamic_state *state,
LLVMBuilderRef builder,
struct gallivm_state *gallivm,
unsigned unit);
/** Obtain stride in bytes between image slices (returns int32) */
LLVMValueRef
(*img_stride)( const struct lp_sampler_dynamic_state *state,
LLVMBuilderRef builder,
struct gallivm_state *gallivm,
unsigned unit);
/** Obtain pointer to array of pointers to mimpap levels */
LLVMValueRef
(*data_ptr)( const struct lp_sampler_dynamic_state *state,
LLVMBuilderRef builder,
struct gallivm_state *gallivm,
unsigned unit);
/** Obtain texture min lod (returns float) */
LLVMValueRef
(*min_lod)(const struct lp_sampler_dynamic_state *state,
LLVMBuilderRef builder, unsigned unit);
struct gallivm_state *gallivm, unsigned unit);
/** Obtain texture max lod (returns float) */
LLVMValueRef
(*max_lod)(const struct lp_sampler_dynamic_state *state,
LLVMBuilderRef builder, unsigned unit);
struct gallivm_state *gallivm, unsigned unit);
/** Obtain texture lod bias (returns float) */
LLVMValueRef
(*lod_bias)(const struct lp_sampler_dynamic_state *state,
LLVMBuilderRef builder, unsigned unit);
struct gallivm_state *gallivm, unsigned unit);
/** Obtain texture border color (returns ptr to float[4]) */
LLVMValueRef
(*border_color)(const struct lp_sampler_dynamic_state *state,
LLVMBuilderRef builder, unsigned unit);
struct gallivm_state *gallivm, unsigned unit);
};
@@ -171,7 +171,7 @@ struct lp_sampler_dynamic_state
*/
struct lp_build_sample_context
{
LLVMBuilderRef builder;
struct gallivm_state *gallivm;
const struct lp_sampler_static_state *static_state;
@@ -385,7 +385,7 @@ lp_build_sample_offset(struct lp_build_context *bld,
void
lp_build_sample_soa(LLVMBuilderRef builder,
lp_build_sample_soa(struct gallivm_state *gallivm,
const struct lp_sampler_static_state *static_state,
struct lp_sampler_dynamic_state *dynamic_state,
struct lp_type fp_type,
@@ -399,7 +399,7 @@ lp_build_sample_soa(LLVMBuilderRef builder,
LLVMValueRef texel_out[4]);
void
lp_build_sample_nop(struct lp_type type,
lp_build_sample_nop(struct gallivm_state *gallivm, struct lp_type type,
LLVMValueRef texel_out[4]);

View File

@@ -52,6 +52,7 @@
#include "lp_bld_flow.h"
#include "lp_bld_gather.h"
#include "lp_bld_format.h"
#include "lp_bld_init.h"
#include "lp_bld_sample.h"
#include "lp_bld_sample_aos.h"
#include "lp_bld_quad.h"
@@ -82,6 +83,7 @@ lp_build_sample_wrap_nearest_int(struct lp_build_sample_context *bld,
LLVMValueRef *out_i)
{
struct lp_build_context *int_coord_bld = &bld->int_coord_bld;
LLVMBuilderRef builder = bld->gallivm->builder;
LLVMValueRef length_minus_one;
length_minus_one = lp_build_sub(int_coord_bld, length, int_coord_bld->one);
@@ -89,12 +91,12 @@ lp_build_sample_wrap_nearest_int(struct lp_build_sample_context *bld,
switch(wrap_mode) {
case PIPE_TEX_WRAP_REPEAT:
if(is_pot)
coord = LLVMBuildAnd(bld->builder, coord, length_minus_one, "");
coord = LLVMBuildAnd(builder, coord, length_minus_one, "");
else {
/* Add a bias to the texcoord to handle negative coords */
LLVMValueRef bias = lp_build_mul_imm(int_coord_bld, length, 1024);
coord = LLVMBuildAdd(bld->builder, coord, bias, "");
coord = LLVMBuildURem(bld->builder, coord, length, "");
coord = LLVMBuildAdd(builder, coord, bias, "");
coord = LLVMBuildURem(builder, coord, length, "");
}
break;
@@ -147,6 +149,7 @@ lp_build_sample_wrap_linear_int(struct lp_build_sample_context *bld,
LLVMValueRef *i1)
{
struct lp_build_context *int_coord_bld = &bld->int_coord_bld;
LLVMBuilderRef builder = bld->gallivm->builder;
LLVMValueRef length_minus_one;
LLVMValueRef lmask, umask, mask;
@@ -195,39 +198,39 @@ lp_build_sample_wrap_linear_int(struct lp_build_sample_context *bld,
switch(wrap_mode) {
case PIPE_TEX_WRAP_REPEAT:
if (is_pot) {
coord0 = LLVMBuildAnd(bld->builder, coord0, length_minus_one, "");
coord0 = LLVMBuildAnd(builder, coord0, length_minus_one, "");
}
else {
/* Add a bias to the texcoord to handle negative coords */
LLVMValueRef bias = lp_build_mul_imm(int_coord_bld, length, 1024);
coord0 = LLVMBuildAdd(bld->builder, coord0, bias, "");
coord0 = LLVMBuildURem(bld->builder, coord0, length, "");
coord0 = LLVMBuildAdd(builder, coord0, bias, "");
coord0 = LLVMBuildURem(builder, coord0, length, "");
}
mask = lp_build_compare(bld->builder, int_coord_bld->type,
mask = lp_build_compare(bld->gallivm, int_coord_bld->type,
PIPE_FUNC_NOTEQUAL, coord0, length_minus_one);
*offset0 = lp_build_mul(int_coord_bld, coord0, stride);
*offset1 = LLVMBuildAnd(bld->builder,
*offset1 = LLVMBuildAnd(builder,
lp_build_add(int_coord_bld, *offset0, stride),
mask, "");
break;
case PIPE_TEX_WRAP_CLAMP_TO_EDGE:
lmask = lp_build_compare(int_coord_bld->builder, int_coord_bld->type,
lmask = lp_build_compare(int_coord_bld->gallivm, int_coord_bld->type,
PIPE_FUNC_GEQUAL, coord0, int_coord_bld->zero);
umask = lp_build_compare(int_coord_bld->builder, int_coord_bld->type,
umask = lp_build_compare(int_coord_bld->gallivm, int_coord_bld->type,
PIPE_FUNC_LESS, coord0, length_minus_one);
coord0 = lp_build_select(int_coord_bld, lmask, coord0, int_coord_bld->zero);
coord0 = lp_build_select(int_coord_bld, umask, coord0, length_minus_one);
mask = LLVMBuildAnd(bld->builder, lmask, umask, "");
mask = LLVMBuildAnd(builder, lmask, umask, "");
*offset0 = lp_build_mul(int_coord_bld, coord0, stride);
*offset1 = lp_build_add(int_coord_bld,
*offset0,
LLVMBuildAnd(bld->builder, stride, mask, ""));
LLVMBuildAnd(builder, stride, mask, ""));
break;
case PIPE_TEX_WRAP_CLAMP:
@@ -263,7 +266,7 @@ lp_build_sample_image_nearest(struct lp_build_sample_context *bld,
LLVMValueRef *colors_hi)
{
const unsigned dims = bld->dims;
LLVMBuilderRef builder = bld->builder;
LLVMBuilderRef builder = bld->gallivm->builder;
struct lp_build_context i32, h16, u8n;
LLVMTypeRef i32_vec_type, h16_vec_type, u8n_vec_type;
LLVMValueRef i32_c8;
@@ -273,13 +276,13 @@ lp_build_sample_image_nearest(struct lp_build_sample_context *bld,
LLVMValueRef x_offset, offset;
LLVMValueRef x_subcoord, y_subcoord, z_subcoord;
lp_build_context_init(&i32, builder, lp_type_int_vec(32));
lp_build_context_init(&h16, builder, lp_type_ufixed(16));
lp_build_context_init(&u8n, builder, lp_type_unorm(8));
lp_build_context_init(&i32, bld->gallivm, lp_type_int_vec(32));
lp_build_context_init(&h16, bld->gallivm, lp_type_ufixed(16));
lp_build_context_init(&u8n, bld->gallivm, lp_type_unorm(8));
i32_vec_type = lp_build_vec_type(i32.type);
h16_vec_type = lp_build_vec_type(h16.type);
u8n_vec_type = lp_build_vec_type(u8n.type);
i32_vec_type = lp_build_vec_type(bld->gallivm, i32.type);
h16_vec_type = lp_build_vec_type(bld->gallivm, h16.type);
u8n_vec_type = lp_build_vec_type(bld->gallivm, u8n.type);
lp_build_extract_image_sizes(bld,
bld->int_size_type,
@@ -317,7 +320,7 @@ lp_build_sample_image_nearest(struct lp_build_sample_context *bld,
r = LLVMBuildFPToSI(builder, r, i32_vec_type, "");
/* compute floor (shift right 8) */
i32_c8 = lp_build_const_int_vec(i32.type, 8);
i32_c8 = lp_build_const_int_vec(bld->gallivm, i32.type, 8);
s_ipart = LLVMBuildAShr(builder, s, i32_c8, "");
if (dims >= 2)
t_ipart = LLVMBuildAShr(builder, t, i32_c8, "");
@@ -325,7 +328,8 @@ lp_build_sample_image_nearest(struct lp_build_sample_context *bld,
r_ipart = LLVMBuildAShr(builder, r, i32_c8, "");
/* get pixel, row, image strides */
x_stride = lp_build_const_vec(bld->int_coord_bld.type,
x_stride = lp_build_const_vec(bld->gallivm,
bld->int_coord_bld.type,
bld->format_desc->block.bits/8);
/* Do texcoord wrapping, compute texel offset */
@@ -387,7 +391,7 @@ lp_build_sample_image_nearest(struct lp_build_sample_context *bld,
* Given the format is a rgba8, just read the pixels as is,
* without any swizzling. Swizzling will be done later.
*/
rgba8 = lp_build_gather(bld->builder,
rgba8 = lp_build_gather(bld->gallivm,
bld->texel_type.length,
bld->format_desc->block.bits,
bld->texel_type.width,
@@ -396,7 +400,7 @@ lp_build_sample_image_nearest(struct lp_build_sample_context *bld,
rgba8 = LLVMBuildBitCast(builder, rgba8, u8n_vec_type, "");
}
else {
rgba8 = lp_build_fetch_rgba_aos(bld->builder,
rgba8 = lp_build_fetch_rgba_aos(bld->gallivm,
bld->format_desc,
u8n.type,
data_ptr, offset,
@@ -405,7 +409,7 @@ lp_build_sample_image_nearest(struct lp_build_sample_context *bld,
}
/* Expand one 4*rgba8 to two 2*rgba16 */
lp_build_unpack2(builder, u8n.type, h16.type,
lp_build_unpack2(bld->gallivm, u8n.type, h16.type,
rgba8,
colors_lo, colors_hi);
}
@@ -429,7 +433,7 @@ lp_build_sample_image_linear(struct lp_build_sample_context *bld,
LLVMValueRef *colors_hi)
{
const unsigned dims = bld->dims;
LLVMBuilderRef builder = bld->builder;
LLVMBuilderRef builder = bld->gallivm->builder;
struct lp_build_context i32, h16, u8n;
LLVMTypeRef i32_vec_type, h16_vec_type, u8n_vec_type;
LLVMValueRef i32_c8, i32_c128, i32_c255;
@@ -450,13 +454,13 @@ lp_build_sample_image_linear(struct lp_build_sample_context *bld,
unsigned i, j, k;
unsigned numj, numk;
lp_build_context_init(&i32, builder, lp_type_int_vec(32));
lp_build_context_init(&h16, builder, lp_type_ufixed(16));
lp_build_context_init(&u8n, builder, lp_type_unorm(8));
lp_build_context_init(&i32, bld->gallivm, lp_type_int_vec(32));
lp_build_context_init(&h16, bld->gallivm, lp_type_ufixed(16));
lp_build_context_init(&u8n, bld->gallivm, lp_type_unorm(8));
i32_vec_type = lp_build_vec_type(i32.type);
h16_vec_type = lp_build_vec_type(h16.type);
u8n_vec_type = lp_build_vec_type(u8n.type);
i32_vec_type = lp_build_vec_type(bld->gallivm, i32.type);
h16_vec_type = lp_build_vec_type(bld->gallivm, h16.type);
u8n_vec_type = lp_build_vec_type(bld->gallivm, u8n.type);
lp_build_extract_image_sizes(bld,
bld->int_size_type,
@@ -494,7 +498,7 @@ lp_build_sample_image_linear(struct lp_build_sample_context *bld,
r = LLVMBuildFPToSI(builder, r, i32_vec_type, "");
/* subtract 0.5 (add -128) */
i32_c128 = lp_build_const_int_vec(i32.type, -128);
i32_c128 = lp_build_const_int_vec(bld->gallivm, i32.type, -128);
s = LLVMBuildAdd(builder, s, i32_c128, "");
if (dims >= 2) {
t = LLVMBuildAdd(builder, t, i32_c128, "");
@@ -504,7 +508,7 @@ lp_build_sample_image_linear(struct lp_build_sample_context *bld,
}
/* compute floor (shift right 8) */
i32_c8 = lp_build_const_int_vec(i32.type, 8);
i32_c8 = lp_build_const_int_vec(bld->gallivm, i32.type, 8);
s_ipart = LLVMBuildAShr(builder, s, i32_c8, "");
if (dims >= 2)
t_ipart = LLVMBuildAShr(builder, t, i32_c8, "");
@@ -512,7 +516,7 @@ lp_build_sample_image_linear(struct lp_build_sample_context *bld,
r_ipart = LLVMBuildAShr(builder, r, i32_c8, "");
/* compute fractional part (AND with 0xff) */
i32_c255 = lp_build_const_int_vec(i32.type, 255);
i32_c255 = lp_build_const_int_vec(bld->gallivm, i32.type, 255);
s_fpart = LLVMBuildAnd(builder, s, i32_c255, "");
if (dims >= 2)
t_fpart = LLVMBuildAnd(builder, t, i32_c255, "");
@@ -520,7 +524,7 @@ lp_build_sample_image_linear(struct lp_build_sample_context *bld,
r_fpart = LLVMBuildAnd(builder, r, i32_c255, "");
/* get pixel, row and image strides */
x_stride = lp_build_const_vec(bld->int_coord_bld.type,
x_stride = lp_build_const_vec(bld->gallivm, bld->int_coord_bld.type,
bld->format_desc->block.bits/8);
y_stride = row_stride_vec;
z_stride = img_stride_vec;
@@ -612,7 +616,7 @@ lp_build_sample_image_linear(struct lp_build_sample_context *bld,
r_fpart = LLVMBuildBitCast(builder, r_fpart, h16_vec_type, "");
{
LLVMTypeRef elem_type = LLVMInt32Type();
LLVMTypeRef elem_type = LLVMInt32TypeInContext(bld->gallivm->context);
LLVMValueRef shuffles_lo[LP_MAX_VECTOR_LENGTH];
LLVMValueRef shuffles_hi[LP_MAX_VECTOR_LENGTH];
LLVMValueRef shuffle_lo;
@@ -685,7 +689,7 @@ lp_build_sample_image_linear(struct lp_build_sample_context *bld,
* Given the format is a rgba8, just read the pixels as is,
* without any swizzling. Swizzling will be done later.
*/
rgba8 = lp_build_gather(bld->builder,
rgba8 = lp_build_gather(bld->gallivm,
bld->texel_type.length,
bld->format_desc->block.bits,
bld->texel_type.width,
@@ -694,7 +698,7 @@ lp_build_sample_image_linear(struct lp_build_sample_context *bld,
rgba8 = LLVMBuildBitCast(builder, rgba8, u8n_vec_type, "");
}
else {
rgba8 = lp_build_fetch_rgba_aos(bld->builder,
rgba8 = lp_build_fetch_rgba_aos(bld->gallivm,
bld->format_desc,
u8n.type,
data_ptr, offset[k][j][i],
@@ -703,7 +707,7 @@ lp_build_sample_image_linear(struct lp_build_sample_context *bld,
}
/* Expand one 4*rgba8 to two 2*rgba16 */
lp_build_unpack2(builder, u8n.type, h16.type,
lp_build_unpack2(bld->gallivm, u8n.type, h16.type,
rgba8,
&neighbors_lo[k][j][i], &neighbors_hi[k][j][i]);
}
@@ -790,7 +794,7 @@ lp_build_sample_mipmap(struct lp_build_sample_context *bld,
LLVMValueRef colors_lo_var,
LLVMValueRef colors_hi_var)
{
LLVMBuilderRef builder = bld->builder;
LLVMBuilderRef builder = bld->gallivm->builder;
LLVMValueRef size0;
LLVMValueRef size1;
LLVMValueRef row_stride0_vec;
@@ -802,7 +806,6 @@ lp_build_sample_mipmap(struct lp_build_sample_context *bld,
LLVMValueRef colors0_lo, colors0_hi;
LLVMValueRef colors1_lo, colors1_hi;
/* sample the first mipmap level */
lp_build_mipmap_level_sizes(bld, ilevel0,
&size0,
@@ -829,8 +832,8 @@ lp_build_sample_mipmap(struct lp_build_sample_context *bld,
LLVMBuildStore(builder, colors0_hi, colors_hi_var);
if (mip_filter == PIPE_TEX_MIPFILTER_LINEAR) {
LLVMValueRef h16_scale = LLVMConstReal(LLVMFloatType(), 256.0);
LLVMTypeRef i32_type = LLVMIntType(32);
LLVMValueRef h16_scale = lp_build_const_float(bld->gallivm, 256.0);
LLVMTypeRef i32_type = LLVMIntTypeInContext(bld->gallivm->context, 32);
struct lp_build_if_state if_ctx;
LLVMValueRef need_lerp;
@@ -842,11 +845,11 @@ lp_build_sample_mipmap(struct lp_build_sample_context *bld,
lod_fpart, LLVMConstNull(i32_type),
"need_lerp");
lp_build_if(&if_ctx, builder, need_lerp);
lp_build_if(&if_ctx, bld->gallivm, need_lerp);
{
struct lp_build_context h16_bld;
lp_build_context_init(&h16_bld, builder, lp_type_ufixed(16));
lp_build_context_init(&h16_bld, bld->gallivm, lp_type_ufixed(16));
/* sample the second mipmap level */
lp_build_mipmap_level_sizes(bld, ilevel1,
@@ -885,7 +888,7 @@ lp_build_sample_mipmap(struct lp_build_sample_context *bld,
int i;
assert(h16_bld.type.length <= Elements(shuffles));
for (i = 0; i < h16_bld.type.length; i++)
shuffles[i] = lp_build_const_int32(2 * (i & 1));
shuffles[i] = lp_build_const_int32(bld->gallivm, 2 * (i & 1));
shuffle = LLVMConstVector(shuffles, h16_bld.type.length);
lod_fpart = LLVMBuildShuffleVector(builder,
lod_fpart, lod_fpart,
@@ -925,7 +928,7 @@ lp_build_sample_aos(struct lp_build_sample_context *bld,
LLVMValueRef texel_out[4])
{
struct lp_build_context *int_bld = &bld->int_bld;
LLVMBuilderRef builder = bld->builder;
LLVMBuilderRef builder = bld->gallivm->builder;
const unsigned mip_filter = bld->static_state->min_mip_filter;
const unsigned min_filter = bld->static_state->min_img_filter;
const unsigned mag_filter = bld->static_state->mag_img_filter;
@@ -936,8 +939,7 @@ lp_build_sample_aos(struct lp_build_sample_context *bld,
LLVMValueRef unswizzled[4];
LLVMValueRef face_ddx[4], face_ddy[4];
struct lp_build_context h16_bld;
LLVMTypeRef i32t = LLVMInt32Type();
LLVMValueRef i32t_zero = LLVMConstInt(i32t, 0, 0);
LLVMValueRef i32t_zero = lp_build_const_int32(bld->gallivm, 0);
/* we only support the common/simple wrap modes at this time */
assert(lp_is_simple_wrap_mode(bld->static_state->wrap_s));
@@ -948,7 +950,7 @@ lp_build_sample_aos(struct lp_build_sample_context *bld,
/* make 16-bit fixed-pt builder context */
lp_build_context_init(&h16_bld, builder, lp_type_ufixed(16));
lp_build_context_init(&h16_bld, bld->gallivm, lp_type_ufixed(16));
/* cube face selection, compute pre-face coords, etc. */
if (bld->static_state->target == PIPE_TEXTURE_CUBE) {
@@ -1026,8 +1028,8 @@ lp_build_sample_aos(struct lp_build_sample_context *bld,
* Get/interpolate texture colors.
*/
packed_lo = lp_build_alloca(builder, h16_bld.vec_type, "packed_lo");
packed_hi = lp_build_alloca(builder, h16_bld.vec_type, "packed_hi");
packed_lo = lp_build_alloca(bld->gallivm, h16_bld.vec_type, "packed_lo");
packed_hi = lp_build_alloca(bld->gallivm, h16_bld.vec_type, "packed_hi");
if (min_filter == mag_filter) {
/* no need to distinquish between minification and magnification */
@@ -1048,7 +1050,7 @@ lp_build_sample_aos(struct lp_build_sample_context *bld,
minify = LLVMBuildICmp(builder, LLVMIntSGE,
lod_ipart, int_bld->zero, "");
lp_build_if(&if_ctx, builder, minify);
lp_build_if(&if_ctx, bld->gallivm, minify);
{
/* Use the minification filter */
lp_build_sample_mipmap(bld,
@@ -1073,7 +1075,7 @@ lp_build_sample_aos(struct lp_build_sample_context *bld,
* combine the values stored in 'packed_lo' and 'packed_hi' variables
* into 'packed'
*/
packed = lp_build_pack2(builder,
packed = lp_build_pack2(bld->gallivm,
h16_bld.type, lp_type_unorm(8),
LLVMBuildLoad(builder, packed_lo, ""),
LLVMBuildLoad(builder, packed_hi, ""));
@@ -1081,7 +1083,7 @@ lp_build_sample_aos(struct lp_build_sample_context *bld,
/*
* Convert to SoA and swizzle.
*/
lp_build_rgba8_to_f32_soa(builder,
lp_build_rgba8_to_f32_soa(bld->gallivm,
bld->texel_type,
packed, unswizzled);

View File

@@ -84,6 +84,7 @@ lp_build_sample_texel_soa(struct lp_build_sample_context *bld,
const struct lp_sampler_static_state *static_state = bld->static_state;
const unsigned dims = bld->dims;
struct lp_build_context *int_coord_bld = &bld->int_coord_bld;
LLVMBuilderRef builder = bld->gallivm->builder;
LLVMValueRef offset;
LLVMValueRef i, j;
LLVMValueRef use_border = NULL;
@@ -95,7 +96,7 @@ lp_build_sample_texel_soa(struct lp_build_sample_context *bld,
LLVMValueRef b1, b2;
b1 = lp_build_cmp(int_coord_bld, PIPE_FUNC_LESS, x, int_coord_bld->zero);
b2 = lp_build_cmp(int_coord_bld, PIPE_FUNC_GEQUAL, x, width);
use_border = LLVMBuildOr(bld->builder, b1, b2, "b1_or_b2");
use_border = LLVMBuildOr(builder, b1, b2, "b1_or_b2");
}
if (dims >= 2 &&
@@ -106,11 +107,11 @@ lp_build_sample_texel_soa(struct lp_build_sample_context *bld,
b1 = lp_build_cmp(int_coord_bld, PIPE_FUNC_LESS, y, int_coord_bld->zero);
b2 = lp_build_cmp(int_coord_bld, PIPE_FUNC_GEQUAL, y, height);
if (use_border) {
use_border = LLVMBuildOr(bld->builder, use_border, b1, "ub_or_b1");
use_border = LLVMBuildOr(bld->builder, use_border, b2, "ub_or_b2");
use_border = LLVMBuildOr(builder, use_border, b1, "ub_or_b1");
use_border = LLVMBuildOr(builder, use_border, b2, "ub_or_b2");
}
else {
use_border = LLVMBuildOr(bld->builder, b1, b2, "b1_or_b2");
use_border = LLVMBuildOr(builder, b1, b2, "b1_or_b2");
}
}
@@ -122,11 +123,11 @@ lp_build_sample_texel_soa(struct lp_build_sample_context *bld,
b1 = lp_build_cmp(int_coord_bld, PIPE_FUNC_LESS, z, int_coord_bld->zero);
b2 = lp_build_cmp(int_coord_bld, PIPE_FUNC_GEQUAL, z, depth);
if (use_border) {
use_border = LLVMBuildOr(bld->builder, use_border, b1, "ub_or_b1");
use_border = LLVMBuildOr(bld->builder, use_border, b2, "ub_or_b2");
use_border = LLVMBuildOr(builder, use_border, b1, "ub_or_b1");
use_border = LLVMBuildOr(builder, use_border, b2, "ub_or_b2");
}
else {
use_border = LLVMBuildOr(bld->builder, b1, b2, "b1_or_b2");
use_border = LLVMBuildOr(builder, b1, b2, "b1_or_b2");
}
}
@@ -148,7 +149,7 @@ lp_build_sample_texel_soa(struct lp_build_sample_context *bld,
offset = lp_build_andnot(&bld->int_coord_bld, offset, use_border);
}
lp_build_fetch_rgba_soa(bld->builder,
lp_build_fetch_rgba_soa(bld->gallivm,
bld->format_desc,
bld->texel_type,
data_ptr, offset,
@@ -174,12 +175,12 @@ lp_build_sample_texel_soa(struct lp_build_sample_context *bld,
/* select texel color or border color depending on use_border */
LLVMValueRef border_color_ptr =
bld->dynamic_state->border_color(bld->dynamic_state,
bld->builder, unit);
bld->gallivm, unit);
int chan;
for (chan = 0; chan < 4; chan++) {
LLVMValueRef border_chan =
lp_build_array_get(bld->builder, border_color_ptr,
lp_build_const_int32(chan));
lp_build_array_get(bld->gallivm, border_color_ptr,
lp_build_const_int32(bld->gallivm, chan));
LLVMValueRef border_chan_vec =
lp_build_broadcast_scalar(&bld->float_vec_bld, border_chan);
texel_out[chan] = lp_build_select(&bld->texel_bld, use_border,
@@ -205,7 +206,7 @@ lp_build_coord_mirror(struct lp_build_sample_context *bld,
lp_build_ifloor_fract(coord_bld, coord, &flr, &fract);
/* isOdd = flr & 1 */
isOdd = LLVMBuildAnd(bld->builder, flr, int_coord_bld->one, "");
isOdd = LLVMBuildAnd(bld->gallivm->builder, flr, int_coord_bld->one, "");
/* make coord positive or negative depending on isOdd */
coord = lp_build_set_sign(coord_bld, fract, isOdd);
@@ -239,7 +240,8 @@ lp_build_sample_wrap_linear(struct lp_build_sample_context *bld,
{
struct lp_build_context *coord_bld = &bld->coord_bld;
struct lp_build_context *int_coord_bld = &bld->int_coord_bld;
LLVMValueRef half = lp_build_const_vec(coord_bld->type, 0.5);
LLVMBuilderRef builder = bld->gallivm->builder;
LLVMValueRef half = lp_build_const_vec(bld->gallivm, coord_bld->type, 0.5);
LLVMValueRef length_minus_one = lp_build_sub(int_coord_bld, length, int_coord_bld->one);
LLVMValueRef coord0, coord1, weight;
@@ -253,18 +255,18 @@ lp_build_sample_wrap_linear(struct lp_build_sample_context *bld,
/* repeat wrap */
if (is_pot) {
coord1 = lp_build_add(int_coord_bld, coord0, int_coord_bld->one);
coord0 = LLVMBuildAnd(bld->builder, coord0, length_minus_one, "");
coord1 = LLVMBuildAnd(bld->builder, coord1, length_minus_one, "");
coord0 = LLVMBuildAnd(builder, coord0, length_minus_one, "");
coord1 = LLVMBuildAnd(builder, coord1, length_minus_one, "");
}
else {
/* Add a bias to the texcoord to handle negative coords */
LLVMValueRef bias = lp_build_mul_imm(int_coord_bld, length, 1024);
LLVMValueRef mask;
coord0 = LLVMBuildAdd(bld->builder, coord0, bias, "");
coord0 = LLVMBuildURem(bld->builder, coord0, length, "");
mask = lp_build_compare(bld->builder, int_coord_bld->type,
coord0 = LLVMBuildAdd(builder, coord0, bias, "");
coord0 = LLVMBuildURem(builder, coord0, length, "");
mask = lp_build_compare(bld->gallivm, int_coord_bld->type,
PIPE_FUNC_NOTEQUAL, coord0, length_minus_one);
coord1 = LLVMBuildAnd(bld->builder,
coord1 = LLVMBuildAnd(builder,
lp_build_add(int_coord_bld, coord0, int_coord_bld->one),
mask, "");
}
@@ -318,7 +320,7 @@ lp_build_sample_wrap_linear(struct lp_build_sample_context *bld,
}
/* was: clamp to [-0.5, length + 0.5], then sub 0.5 */
coord = lp_build_sub(coord_bld, coord, half);
min = lp_build_const_vec(coord_bld->type, -1.0F);
min = lp_build_const_vec(bld->gallivm, coord_bld->type, -1.0F);
coord = lp_build_clamp(coord_bld, coord, min, length_f);
/* convert to int, compute lerp weight */
lp_build_ifloor_fract(coord_bld, coord, &coord0, &weight);
@@ -437,6 +439,7 @@ lp_build_sample_wrap_nearest(struct lp_build_sample_context *bld,
{
struct lp_build_context *coord_bld = &bld->coord_bld;
struct lp_build_context *int_coord_bld = &bld->int_coord_bld;
LLVMBuilderRef builder = bld->gallivm->builder;
LLVMValueRef length_minus_one = lp_build_sub(int_coord_bld, length, int_coord_bld->one);
LLVMValueRef icoord;
@@ -445,12 +448,12 @@ lp_build_sample_wrap_nearest(struct lp_build_sample_context *bld,
coord = lp_build_mul(coord_bld, coord, length_f);
icoord = lp_build_ifloor(coord_bld, coord);
if (is_pot)
icoord = LLVMBuildAnd(bld->builder, icoord, length_minus_one, "");
icoord = LLVMBuildAnd(builder, icoord, length_minus_one, "");
else {
/* Add a bias to the texcoord to handle negative coords */
LLVMValueRef bias = lp_build_mul_imm(int_coord_bld, length, 1024);
icoord = LLVMBuildAdd(bld->builder, icoord, bias, "");
icoord = LLVMBuildURem(bld->builder, icoord, length, "");
icoord = LLVMBuildAdd(builder, icoord, bias, "");
icoord = LLVMBuildURem(builder, icoord, length, "");
}
break;
@@ -830,7 +833,7 @@ lp_build_sample_mipmap(struct lp_build_sample_context *bld,
LLVMValueRef lod_fpart,
LLVMValueRef *colors_out)
{
LLVMBuilderRef builder = bld->builder;
LLVMBuilderRef builder = bld->gallivm->builder;
LLVMValueRef size0;
LLVMValueRef size1;
LLVMValueRef row_stride0_vec;
@@ -878,7 +881,7 @@ lp_build_sample_mipmap(struct lp_build_sample_context *bld,
bld->float_bld.zero,
"need_lerp");
lp_build_if(&if_ctx, builder, need_lerp);
lp_build_if(&if_ctx, bld->gallivm, need_lerp);
{
/* sample the second mipmap level */
lp_build_mipmap_level_sizes(bld, ilevel1,
@@ -934,7 +937,7 @@ lp_build_sample_general(struct lp_build_sample_context *bld,
LLVMValueRef *colors_out)
{
struct lp_build_context *int_bld = &bld->int_bld;
LLVMBuilderRef builder = bld->builder;
LLVMBuilderRef builder = bld->gallivm->builder;
const unsigned mip_filter = bld->static_state->min_mip_filter;
const unsigned min_filter = bld->static_state->min_img_filter;
const unsigned mag_filter = bld->static_state->mag_img_filter;
@@ -942,8 +945,7 @@ lp_build_sample_general(struct lp_build_sample_context *bld,
LLVMValueRef ilevel0, ilevel1 = NULL;
LLVMValueRef face_ddx[4], face_ddy[4];
LLVMValueRef texels[4];
LLVMTypeRef i32t = LLVMInt32Type();
LLVMValueRef i32t_zero = LLVMConstInt(i32t, 0, 0);
LLVMValueRef i32t_zero = lp_build_const_int32(bld->gallivm, 0);
unsigned chan;
/*
@@ -1030,7 +1032,7 @@ lp_build_sample_general(struct lp_build_sample_context *bld,
*/
for (chan = 0; chan < 4; ++chan) {
texels[chan] = lp_build_alloca(builder, bld->texel_bld.vec_type, "");
texels[chan] = lp_build_alloca(bld->gallivm, bld->texel_bld.vec_type, "");
lp_build_name(texels[chan], "sampler%u_texel_%c_var", unit, "xyzw"[chan]);
}
@@ -1053,7 +1055,7 @@ lp_build_sample_general(struct lp_build_sample_context *bld,
minify = LLVMBuildICmp(builder, LLVMIntSGE,
lod_ipart, int_bld->zero, "");
lp_build_if(&if_ctx, builder, minify);
lp_build_if(&if_ctx, bld->gallivm, minify);
{
/* Use the minification filter */
lp_build_sample_mipmap(bld, unit,
@@ -1092,6 +1094,7 @@ lp_build_sample_compare(struct lp_build_sample_context *bld,
LLVMValueRef texel[4])
{
struct lp_build_context *texel_bld = &bld->texel_bld;
LLVMBuilderRef builder = bld->gallivm->builder;
LLVMValueRef res;
const unsigned chan = 0;
@@ -1100,11 +1103,10 @@ lp_build_sample_compare(struct lp_build_sample_context *bld,
/* debug code */
if (0) {
LLVMValueRef indx = lp_build_const_int32(0);
LLVMValueRef coord = LLVMBuildExtractElement(bld->builder, p, indx, "");
LLVMValueRef tex = LLVMBuildExtractElement(bld->builder,
texel[chan], indx, "");
lp_build_printf(bld->builder, "shadow compare coord %f to texture %f\n",
LLVMValueRef indx = lp_build_const_int32(bld->gallivm, 0);
LLVMValueRef coord = LLVMBuildExtractElement(builder, p, indx, "");
LLVMValueRef tex = LLVMBuildExtractElement(builder, texel[chan], indx, "");
lp_build_printf(bld->gallivm, "shadow compare coord %f to texture %f\n",
coord, tex);
}
@@ -1126,10 +1128,10 @@ lp_build_sample_compare(struct lp_build_sample_context *bld,
* For debugging.
*/
void
lp_build_sample_nop(struct lp_type type,
lp_build_sample_nop(struct gallivm_state *gallivm, struct lp_type type,
LLVMValueRef texel_out[4])
{
LLVMValueRef one = lp_build_one(type);
LLVMValueRef one = lp_build_one(gallivm, type);
unsigned chan;
for (chan = 0; chan < 4; chan++) {
@@ -1147,7 +1149,7 @@ lp_build_sample_nop(struct lp_type type,
* \param ddy partial derivatives of (s,t,r,q) with respect to y
*/
void
lp_build_sample_soa(LLVMBuilderRef builder,
lp_build_sample_soa(struct gallivm_state *gallivm,
const struct lp_sampler_static_state *static_state,
struct lp_sampler_dynamic_state *dynamic_state,
struct lp_type type,
@@ -1162,8 +1164,8 @@ lp_build_sample_soa(LLVMBuilderRef builder,
{
unsigned dims = texture_dims(static_state->target);
struct lp_build_sample_context bld;
LLVMTypeRef i32t = LLVMInt32Type();
LLVMTypeRef i32t = LLVMInt32TypeInContext(gallivm->context);
LLVMBuilderRef builder = gallivm->builder;
LLVMValueRef s;
LLVMValueRef t;
LLVMValueRef r;
@@ -1178,7 +1180,7 @@ lp_build_sample_soa(LLVMBuilderRef builder,
/* Setup our build context */
memset(&bld, 0, sizeof bld);
bld.builder = builder;
bld.gallivm = gallivm;
bld.static_state = static_state;
bld.dynamic_state = dynamic_state;
bld.format_desc = util_format_description(static_state->format);
@@ -1195,22 +1197,22 @@ lp_build_sample_soa(LLVMBuilderRef builder,
float_vec_type = lp_type_float_vec(32);
lp_build_context_init(&bld.float_bld, builder, bld.float_type);
lp_build_context_init(&bld.float_vec_bld, builder, float_vec_type);
lp_build_context_init(&bld.int_bld, builder, bld.int_type);
lp_build_context_init(&bld.coord_bld, builder, bld.coord_type);
lp_build_context_init(&bld.int_coord_bld, builder, bld.int_coord_type);
lp_build_context_init(&bld.int_size_bld, builder, bld.int_size_type);
lp_build_context_init(&bld.float_size_bld, builder, bld.float_size_type);
lp_build_context_init(&bld.texel_bld, builder, bld.texel_type);
lp_build_context_init(&bld.float_bld, gallivm, bld.float_type);
lp_build_context_init(&bld.float_vec_bld, gallivm, float_vec_type);
lp_build_context_init(&bld.int_bld, gallivm, bld.int_type);
lp_build_context_init(&bld.coord_bld, gallivm, bld.coord_type);
lp_build_context_init(&bld.int_coord_bld, gallivm, bld.int_coord_type);
lp_build_context_init(&bld.int_size_bld, gallivm, bld.int_size_type);
lp_build_context_init(&bld.float_size_bld, gallivm, bld.float_size_type);
lp_build_context_init(&bld.texel_bld, gallivm, bld.texel_type);
/* Get the dynamic state */
bld.width = dynamic_state->width(dynamic_state, builder, unit);
bld.height = dynamic_state->height(dynamic_state, builder, unit);
bld.depth = dynamic_state->depth(dynamic_state, builder, unit);
bld.row_stride_array = dynamic_state->row_stride(dynamic_state, builder, unit);
bld.img_stride_array = dynamic_state->img_stride(dynamic_state, builder, unit);
bld.data_array = dynamic_state->data_ptr(dynamic_state, builder, unit);
bld.width = dynamic_state->width(dynamic_state, gallivm, unit);
bld.height = dynamic_state->height(dynamic_state, gallivm, unit);
bld.depth = dynamic_state->depth(dynamic_state, gallivm, unit);
bld.row_stride_array = dynamic_state->row_stride(dynamic_state, gallivm, unit);
bld.img_stride_array = dynamic_state->img_stride(dynamic_state, gallivm, unit);
bld.data_array = dynamic_state->data_ptr(dynamic_state, gallivm, unit);
/* Note that data_array is an array[level] of pointers to texture images */
s = coords[0];
@@ -1236,7 +1238,7 @@ lp_build_sample_soa(LLVMBuilderRef builder,
if (0) {
/* For debug: no-op texture sampling */
lp_build_sample_nop(bld.texel_type, texel_out);
lp_build_sample_nop(gallivm, bld.texel_type, texel_out);
}
else if (util_format_fits_8unorm(bld.format_desc) &&
lp_is_simple_wrap_mode(static_state->wrap_s) &&

View File

@@ -37,12 +37,13 @@
#include "util/u_debug.h"
#include "util/u_memory.h"
#include "lp_bld_const.h"
#include "lp_bld_debug.h"
#include "lp_bld_struct.h"
LLVMValueRef
lp_build_struct_get_ptr(LLVMBuilderRef builder,
lp_build_struct_get_ptr(struct gallivm_state *gallivm,
LLVMValueRef ptr,
unsigned member,
const char *name)
@@ -51,16 +52,16 @@ lp_build_struct_get_ptr(LLVMBuilderRef builder,
LLVMValueRef member_ptr;
assert(LLVMGetTypeKind(LLVMTypeOf(ptr)) == LLVMPointerTypeKind);
assert(LLVMGetTypeKind(LLVMGetElementType(LLVMTypeOf(ptr))) == LLVMStructTypeKind);
indices[0] = LLVMConstInt(LLVMInt32Type(), 0, 0);
indices[1] = LLVMConstInt(LLVMInt32Type(), member, 0);
member_ptr = LLVMBuildGEP(builder, ptr, indices, Elements(indices), "");
indices[0] = lp_build_const_int32(gallivm, 0);
indices[1] = lp_build_const_int32(gallivm, member);
member_ptr = LLVMBuildGEP(gallivm->builder, ptr, indices, Elements(indices), "");
lp_build_name(member_ptr, "%s.%s_ptr", LLVMGetValueName(ptr), name);
return member_ptr;
}
LLVMValueRef
lp_build_struct_get(LLVMBuilderRef builder,
lp_build_struct_get(struct gallivm_state *gallivm,
LLVMValueRef ptr,
unsigned member,
const char *name)
@@ -69,15 +70,15 @@ lp_build_struct_get(LLVMBuilderRef builder,
LLVMValueRef res;
assert(LLVMGetTypeKind(LLVMTypeOf(ptr)) == LLVMPointerTypeKind);
assert(LLVMGetTypeKind(LLVMGetElementType(LLVMTypeOf(ptr))) == LLVMStructTypeKind);
member_ptr = lp_build_struct_get_ptr(builder, ptr, member, name);
res = LLVMBuildLoad(builder, member_ptr, "");
member_ptr = lp_build_struct_get_ptr(gallivm, ptr, member, name);
res = LLVMBuildLoad(gallivm->builder, member_ptr, "");
lp_build_name(res, "%s.%s", LLVMGetValueName(ptr), name);
return res;
}
LLVMValueRef
lp_build_array_get_ptr(LLVMBuilderRef builder,
lp_build_array_get_ptr(struct gallivm_state *gallivm,
LLVMValueRef ptr,
LLVMValueRef index)
{
@@ -85,9 +86,9 @@ lp_build_array_get_ptr(LLVMBuilderRef builder,
LLVMValueRef element_ptr;
assert(LLVMGetTypeKind(LLVMTypeOf(ptr)) == LLVMPointerTypeKind);
assert(LLVMGetTypeKind(LLVMGetElementType(LLVMTypeOf(ptr))) == LLVMArrayTypeKind);
indices[0] = LLVMConstInt(LLVMInt32Type(), 0, 0);
indices[0] = lp_build_const_int32(gallivm, 0);
indices[1] = index;
element_ptr = LLVMBuildGEP(builder, ptr, indices, Elements(indices), "");
element_ptr = LLVMBuildGEP(gallivm->builder, ptr, indices, Elements(indices), "");
#ifdef DEBUG
lp_build_name(element_ptr, "&%s[%s]",
LLVMGetValueName(ptr), LLVMGetValueName(index));
@@ -97,7 +98,7 @@ lp_build_array_get_ptr(LLVMBuilderRef builder,
LLVMValueRef
lp_build_array_get(LLVMBuilderRef builder,
lp_build_array_get(struct gallivm_state *gallivm,
LLVMValueRef ptr,
LLVMValueRef index)
{
@@ -105,8 +106,8 @@ lp_build_array_get(LLVMBuilderRef builder,
LLVMValueRef res;
assert(LLVMGetTypeKind(LLVMTypeOf(ptr)) == LLVMPointerTypeKind);
assert(LLVMGetTypeKind(LLVMGetElementType(LLVMTypeOf(ptr))) == LLVMArrayTypeKind);
element_ptr = lp_build_array_get_ptr(builder, ptr, index);
res = LLVMBuildLoad(builder, element_ptr, "");
element_ptr = lp_build_array_get_ptr(gallivm, ptr, index);
res = LLVMBuildLoad(gallivm->builder, element_ptr, "");
#ifdef DEBUG
lp_build_name(res, "%s[%s]", LLVMGetValueName(ptr), LLVMGetValueName(index));
#endif
@@ -115,7 +116,7 @@ lp_build_array_get(LLVMBuilderRef builder,
void
lp_build_array_set(LLVMBuilderRef builder,
lp_build_array_set(struct gallivm_state *gallivm,
LLVMValueRef ptr,
LLVMValueRef index,
LLVMValueRef value)
@@ -123,8 +124,8 @@ lp_build_array_set(LLVMBuilderRef builder,
LLVMValueRef element_ptr;
assert(LLVMGetTypeKind(LLVMTypeOf(ptr)) == LLVMPointerTypeKind);
assert(LLVMGetTypeKind(LLVMGetElementType(LLVMTypeOf(ptr))) == LLVMArrayTypeKind);
element_ptr = lp_build_array_get_ptr(builder, ptr, index);
LLVMBuildStore(builder, value, element_ptr);
element_ptr = lp_build_array_get_ptr(gallivm, ptr, index);
LLVMBuildStore(gallivm->builder, value, element_ptr);
}

View File

@@ -38,7 +38,7 @@
#include "gallivm/lp_bld.h"
#include <llvm-c/Target.h>
#include "gallivm/lp_bld_init.h"
#include "util/u_debug.h"
#include "util/u_memory.h"
@@ -57,7 +57,7 @@
* Get value pointer to a structure member.
*/
LLVMValueRef
lp_build_struct_get_ptr(LLVMBuilderRef builder,
lp_build_struct_get_ptr(struct gallivm_state *gallivm,
LLVMValueRef ptr,
unsigned member,
const char *name);
@@ -66,7 +66,7 @@ lp_build_struct_get_ptr(LLVMBuilderRef builder,
* Get the value of a structure member.
*/
LLVMValueRef
lp_build_struct_get(LLVMBuilderRef builder,
lp_build_struct_get(struct gallivm_state *gallivm,
LLVMValueRef ptr,
unsigned member,
const char *name);
@@ -75,7 +75,7 @@ lp_build_struct_get(LLVMBuilderRef builder,
* Get value pointer to an array element.
*/
LLVMValueRef
lp_build_array_get_ptr(LLVMBuilderRef builder,
lp_build_array_get_ptr(struct gallivm_state *gallivm,
LLVMValueRef ptr,
LLVMValueRef index);
@@ -83,7 +83,7 @@ lp_build_array_get_ptr(LLVMBuilderRef builder,
* Get the value of an array element.
*/
LLVMValueRef
lp_build_array_get(LLVMBuilderRef builder,
lp_build_array_get(struct gallivm_state *gallivm,
LLVMValueRef ptr,
LLVMValueRef index);
@@ -91,7 +91,7 @@ lp_build_array_get(LLVMBuilderRef builder,
* Set the value of an array element.
*/
void
lp_build_array_set(LLVMBuilderRef builder,
lp_build_array_set(struct gallivm_state *gallivm,
LLVMValueRef ptr,
LLVMValueRef index,
LLVMValueRef value);

View File

@@ -37,12 +37,13 @@
#include "lp_bld_type.h"
#include "lp_bld_const.h"
#include "lp_bld_init.h"
#include "lp_bld_logic.h"
#include "lp_bld_swizzle.h"
LLVMValueRef
lp_build_broadcast(LLVMBuilderRef builder,
lp_build_broadcast(struct gallivm_state *gallivm,
LLVMTypeRef vec_type,
LLVMValueRef scalar)
{
@@ -52,8 +53,8 @@ lp_build_broadcast(LLVMBuilderRef builder,
res = LLVMGetUndef(vec_type);
for(i = 0; i < n; ++i) {
LLVMValueRef index = LLVMConstInt(LLVMInt32Type(), i, 0);
res = LLVMBuildInsertElement(builder, res, scalar, index, "");
LLVMValueRef index = lp_build_const_int32(gallivm, i);
res = LLVMBuildInsertElement(gallivm->builder, res, scalar, index, "");
}
return res;
@@ -67,6 +68,7 @@ LLVMValueRef
lp_build_broadcast_scalar(struct lp_build_context *bld,
LLVMValueRef scalar)
{
LLVMBuilderRef builder = bld->gallivm->builder;
const struct lp_type type = bld->type;
assert(lp_check_elem_type(type, LLVMTypeOf(scalar)));
@@ -82,17 +84,17 @@ lp_build_broadcast_scalar(struct lp_build_context *bld,
struct lp_type i32_vec_type = lp_type_int_vec(32);
i32_vec_type.length = type.length;
res = LLVMBuildInsertElement(bld->builder, bld->undef, scalar,
LLVMConstInt(LLVMInt32Type(), 0, 0), "");
res = LLVMBuildShuffleVector(bld->builder, res, bld->undef,
lp_build_const_int_vec(i32_vec_type, 0), "");
res = LLVMBuildInsertElement(builder, bld->undef, scalar,
lp_build_const_int32(bld->gallivm, 0), "");
res = LLVMBuildShuffleVector(builder, res, bld->undef,
lp_build_const_int_vec(bld->gallivm, i32_vec_type, 0), "");
#else
/* XXX: The above path provokes a bug in LLVM 2.6 */
unsigned i;
res = bld->undef;
for(i = 0; i < type.length; ++i) {
LLVMValueRef index = LLVMConstInt(LLVMInt32Type(), i, 0);
res = LLVMBuildInsertElement(bld->builder, res, scalar, index, "");
LLVMValueRef index = lp_build_const_int32(bld->gallivm, i);
res = LLVMBuildInsertElement(builder, res, scalar, index, "");
}
#endif
return res;
@@ -104,13 +106,13 @@ lp_build_broadcast_scalar(struct lp_build_context *bld,
* Combined extract and broadcast (or a mere shuffle when the two types match)
*/
LLVMValueRef
lp_build_extract_broadcast(LLVMBuilderRef builder,
lp_build_extract_broadcast(struct gallivm_state *gallivm,
struct lp_type src_type,
struct lp_type dst_type,
LLVMValueRef vector,
LLVMValueRef index)
{
LLVMTypeRef i32t = LLVMInt32Type();
LLVMTypeRef i32t = LLVMInt32TypeInContext(gallivm->context);
LLVMValueRef res;
assert(src_type.floating == dst_type.floating);
@@ -132,8 +134,8 @@ lp_build_extract_broadcast(LLVMBuilderRef builder,
* Broadcast scalar -> vector.
*/
res = lp_build_broadcast(builder,
lp_build_vec_type(dst_type),
res = lp_build_broadcast(gallivm,
lp_build_vec_type(gallivm, dst_type),
vector);
}
}
@@ -144,16 +146,16 @@ lp_build_extract_broadcast(LLVMBuilderRef builder,
*/
LLVMValueRef shuffle;
shuffle = lp_build_broadcast(builder,
shuffle = lp_build_broadcast(gallivm,
LLVMVectorType(i32t, dst_type.length),
index);
res = LLVMBuildShuffleVector(builder, vector,
LLVMGetUndef(lp_build_vec_type(dst_type)),
res = LLVMBuildShuffleVector(gallivm->builder, vector,
LLVMGetUndef(lp_build_vec_type(gallivm, dst_type)),
shuffle, "");
}
else {
LLVMValueRef scalar;
scalar = LLVMBuildExtractElement(builder, vector, index, "");
scalar = LLVMBuildExtractElement(gallivm->builder, vector, index, "");
if (dst_type.length == 1) {
/*
* Trivial extract scalar from vector.
@@ -166,8 +168,8 @@ lp_build_extract_broadcast(LLVMBuilderRef builder,
* General case of different sized vectors.
*/
res = lp_build_broadcast(builder,
lp_build_vec_type(dst_type),
res = lp_build_broadcast(gallivm,
lp_build_vec_type(gallivm, dst_type),
vector);
}
}
@@ -185,6 +187,7 @@ lp_build_swizzle_scalar_aos(struct lp_build_context *bld,
LLVMValueRef a,
unsigned channel)
{
LLVMBuilderRef builder = bld->gallivm->builder;
const struct lp_type type = bld->type;
const unsigned n = type.length;
unsigned i, j;
@@ -199,14 +202,14 @@ lp_build_swizzle_scalar_aos(struct lp_build_context *bld,
/*
* Shuffle.
*/
LLVMTypeRef elem_type = LLVMInt32Type();
LLVMTypeRef elem_type = LLVMInt32TypeInContext(bld->gallivm->context);
LLVMValueRef shuffles[LP_MAX_VECTOR_LENGTH];
for(j = 0; j < n; j += 4)
for(i = 0; i < 4; ++i)
shuffles[j + i] = LLVMConstInt(elem_type, j + channel, 0);
return LLVMBuildShuffleVector(bld->builder, a, bld->undef, LLVMConstVector(shuffles, n), "");
return LLVMBuildShuffleVector(builder, a, bld->undef, LLVMConstVector(shuffles, n), "");
}
else {
/*
@@ -226,8 +229,9 @@ lp_build_swizzle_scalar_aos(struct lp_build_context *bld,
};
unsigned i;
a = LLVMBuildAnd(bld->builder, a,
lp_build_const_mask_aos(type, 1 << channel), "");
a = LLVMBuildAnd(builder, a,
lp_build_const_mask_aos(bld->gallivm,
type, 1 << channel), "");
/*
* Build a type where each element is an integer that cover the four
@@ -239,7 +243,7 @@ lp_build_swizzle_scalar_aos(struct lp_build_context *bld,
type4.width *= 4;
type4.length /= 4;
a = LLVMBuildBitCast(bld->builder, a, lp_build_vec_type(type4), "");
a = LLVMBuildBitCast(builder, a, lp_build_vec_type(bld->gallivm, type4), "");
for(i = 0; i < 2; ++i) {
LLVMValueRef tmp = NULL;
@@ -250,16 +254,16 @@ lp_build_swizzle_scalar_aos(struct lp_build_context *bld,
#endif
if(shift > 0)
tmp = LLVMBuildLShr(bld->builder, a, lp_build_const_int_vec(type4, shift*type.width), "");
tmp = LLVMBuildLShr(builder, a, lp_build_const_int_vec(bld->gallivm, type4, shift*type.width), "");
if(shift < 0)
tmp = LLVMBuildShl(bld->builder, a, lp_build_const_int_vec(type4, -shift*type.width), "");
tmp = LLVMBuildShl(builder, a, lp_build_const_int_vec(bld->gallivm, type4, -shift*type.width), "");
assert(tmp);
if(tmp)
a = LLVMBuildOr(bld->builder, a, tmp, "");
a = LLVMBuildOr(builder, a, tmp, "");
}
return LLVMBuildBitCast(bld->builder, a, lp_build_vec_type(type), "");
return LLVMBuildBitCast(builder, a, lp_build_vec_type(bld->gallivm, type), "");
}
}
@@ -269,6 +273,7 @@ lp_build_swizzle_aos(struct lp_build_context *bld,
LLVMValueRef a,
const unsigned char swizzles[4])
{
LLVMBuilderRef builder = bld->gallivm->builder;
const struct lp_type type = bld->type;
const unsigned n = type.length;
unsigned i, j;
@@ -303,8 +308,8 @@ lp_build_swizzle_aos(struct lp_build_context *bld,
/*
* Shuffle.
*/
LLVMValueRef undef = LLVMGetUndef(lp_build_elem_type(type));
LLVMTypeRef i32t = LLVMInt32Type();
LLVMValueRef undef = LLVMGetUndef(lp_build_elem_type(bld->gallivm, type));
LLVMTypeRef i32t = LLVMInt32TypeInContext(bld->gallivm->context);
LLVMValueRef shuffles[LP_MAX_VECTOR_LENGTH];
LLVMValueRef aux[LP_MAX_VECTOR_LENGTH];
@@ -326,13 +331,13 @@ lp_build_swizzle_aos(struct lp_build_context *bld,
case PIPE_SWIZZLE_ZERO:
shuffle = type.length + 0;
if (!aux[0]) {
aux[0] = lp_build_const_elem(type, 0.0);
aux[0] = lp_build_const_elem(bld->gallivm, type, 0.0);
}
break;
case PIPE_SWIZZLE_ONE:
shuffle = type.length + 1;
if (!aux[1]) {
aux[1] = lp_build_const_elem(type, 1.0);
aux[1] = lp_build_const_elem(bld->gallivm, type, 1.0);
}
break;
}
@@ -346,7 +351,7 @@ lp_build_swizzle_aos(struct lp_build_context *bld,
}
}
return LLVMBuildShuffleVector(bld->builder, a,
return LLVMBuildShuffleVector(builder, a,
LLVMConstVector(aux, n),
LLVMConstVector(shuffles, n), "");
} else {
@@ -387,8 +392,8 @@ lp_build_swizzle_aos(struct lp_build_context *bld,
type4.width *= 4;
type4.length /= 4;
a = LLVMBuildBitCast(bld->builder, a, lp_build_vec_type(type4), "");
res = LLVMBuildBitCast(bld->builder, res, lp_build_vec_type(type4), "");
a = LLVMBuildBitCast(builder, a, lp_build_vec_type(bld->gallivm, type4), "");
res = LLVMBuildBitCast(builder, res, lp_build_vec_type(bld->gallivm, type4), "");
/*
* Mask and shift the channels, trying to group as many channels in the
@@ -414,23 +419,24 @@ lp_build_swizzle_aos(struct lp_build_context *bld,
if (0)
debug_printf("shift = %i, mask = 0x%08llx\n", shift, mask);
masked = LLVMBuildAnd(bld->builder, a,
lp_build_const_int_vec(type4, mask), "");
masked = LLVMBuildAnd(builder, a,
lp_build_const_int_vec(bld->gallivm, type4, mask), "");
if (shift > 0) {
shifted = LLVMBuildShl(bld->builder, masked,
lp_build_const_int_vec(type4, shift*type.width), "");
shifted = LLVMBuildShl(builder, masked,
lp_build_const_int_vec(bld->gallivm, type4, shift*type.width), "");
} else if (shift < 0) {
shifted = LLVMBuildLShr(bld->builder, masked,
lp_build_const_int_vec(type4, -shift*type.width), "");
shifted = LLVMBuildLShr(builder, masked,
lp_build_const_int_vec(bld->gallivm, type4, -shift*type.width), "");
} else {
shifted = masked;
}
res = LLVMBuildOr(bld->builder, res, shifted, "");
res = LLVMBuildOr(builder, res, shifted, "");
}
}
return LLVMBuildBitCast(bld->builder, res, lp_build_vec_type(type), "");
return LLVMBuildBitCast(builder, res,
lp_build_vec_type(bld->gallivm, type), "");
}
}

View File

@@ -45,7 +45,7 @@ struct lp_build_context;
LLVMValueRef
lp_build_broadcast(LLVMBuilderRef builder,
lp_build_broadcast(struct gallivm_state *gallivm,
LLVMTypeRef vec_type,
LLVMValueRef scalar);
@@ -56,7 +56,7 @@ lp_build_broadcast_scalar(struct lp_build_context *bld,
LLVMValueRef
lp_build_extract_broadcast(LLVMBuilderRef builder,
lp_build_extract_broadcast(struct gallivm_state *gallivm,
struct lp_type src_type,
struct lp_type dst_type,
LLVMValueRef vector,

View File

@@ -46,6 +46,7 @@ struct tgsi_shader_info;
struct lp_type;
struct lp_build_context;
struct lp_build_mask_context;
struct gallivm_state;
enum lp_build_tex_modifier {
@@ -141,7 +142,7 @@ struct lp_build_sampler_soa
void
(*emit_fetch_texel)( const struct lp_build_sampler_soa *sampler,
LLVMBuilderRef builder,
struct gallivm_state *gallivm,
struct lp_type type,
unsigned unit,
unsigned num_coords,
@@ -174,7 +175,7 @@ lp_build_tgsi_info(const struct tgsi_token *tokens,
void
lp_build_tgsi_soa(LLVMBuilderRef builder,
lp_build_tgsi_soa(struct gallivm_state *gallivm,
const struct tgsi_token *tokens,
struct lp_type type,
struct lp_build_mask_context *mask,
@@ -187,7 +188,7 @@ lp_build_tgsi_soa(LLVMBuilderRef builder,
void
lp_build_tgsi_aos(LLVMBuilderRef builder,
lp_build_tgsi_aos(struct gallivm_state *gallivm,
const struct tgsi_token *tokens,
struct lp_type type,
const unsigned char swizzles[4],

View File

@@ -151,6 +151,7 @@ emit_fetch(
const struct tgsi_full_instruction *inst,
unsigned src_op)
{
LLVMBuilderRef builder = bld->base.gallivm->builder;
struct lp_type type = bld->base.type;
const struct tgsi_full_src_register *reg = &inst->Src[src_op];
LLVMValueRef res;
@@ -175,14 +176,12 @@ emit_fetch(
LLVMValueRef scalar;
LLVMValueRef swizzle;
index = LLVMConstInt(LLVMInt32Type(),
reg->Register.Index*4 + chan,
0);
index = lp_build_const_int32(bld->base.gallivm, reg->Register.Index * 4 + chan);
scalar_ptr = LLVMBuildGEP(bld->base.builder, bld->consts_ptr,
scalar_ptr = LLVMBuildGEP(builder, bld->consts_ptr,
&index, 1, "");
scalar = LLVMBuildLoad(bld->base.builder, scalar_ptr, "");
scalar = LLVMBuildLoad(builder, scalar_ptr, "");
lp_build_name(scalar, "const[%u].%c", reg->Register.Index, "xyzw"[chan]);
@@ -190,9 +189,9 @@ emit_fetch(
* NOTE: constants array is always assumed to be RGBA
*/
swizzle = LLVMConstInt(LLVMInt32Type(), chan, 0);
swizzle = lp_build_const_int32(bld->base.gallivm, chan);
res = LLVMBuildInsertElement(bld->base.builder, res, scalar, swizzle, "");
res = LLVMBuildInsertElement(builder, res, scalar, swizzle, "");
}
/*
@@ -206,14 +205,14 @@ emit_fetch(
unsigned i;
for (chan = 0; chan < 4; ++chan) {
shuffles[chan] = LLVMConstInt(LLVMInt32Type(), chan, 0);
shuffles[chan] = lp_build_const_int32(bld->base.gallivm, chan);
}
for (i = 4; i < type.length; ++i) {
shuffles[i] = shuffles[i % 4];
}
res = LLVMBuildShuffleVector(bld->base.builder,
res = LLVMBuildShuffleVector(builder,
res, bld->base.undef,
LLVMConstVector(shuffles, type.length),
"");
@@ -234,7 +233,7 @@ emit_fetch(
{
LLVMValueRef temp_ptr;
temp_ptr = bld->temps[reg->Register.Index];
res = LLVMBuildLoad(bld->base.builder, temp_ptr, "");
res = LLVMBuildLoad(builder, temp_ptr, "");
if (!res)
return bld->base.undef;
}
@@ -281,6 +280,7 @@ emit_store(
unsigned index,
LLVMValueRef value)
{
LLVMBuilderRef builder = bld->base.gallivm->builder;
const struct tgsi_full_dst_register *reg = &inst->Dst[index];
LLVMValueRef mask = NULL;
LLVMValueRef ptr;
@@ -299,7 +299,7 @@ emit_store(
break;
case TGSI_SAT_MINUS_PLUS_ONE:
value = lp_build_max(&bld->base, value, lp_build_const_vec(bld->base.type, -1.0));
value = lp_build_max(&bld->base, value, lp_build_const_vec(bld->base.gallivm, bld->base.type, -1.0));
value = lp_build_min(&bld->base, value, bld->base.one);
break;
@@ -344,20 +344,20 @@ emit_store(
assert(inst->Predicate.Index < LP_MAX_TGSI_PREDS);
pred = LLVMBuildLoad(bld->base.builder,
pred = LLVMBuildLoad(builder,
bld->preds[inst->Predicate.Index], "");
/*
* Convert the value to an integer mask.
*/
pred = lp_build_compare(bld->base.builder,
pred = lp_build_compare(bld->base.gallivm,
bld->base.type,
PIPE_FUNC_NOTEQUAL,
pred,
bld->base.zero);
if (inst->Predicate.Negate) {
pred = LLVMBuildNot(bld->base.builder, pred, "");
pred = LLVMBuildNot(builder, pred, "");
}
pred = swizzle_aos(bld, pred,
@@ -367,7 +367,7 @@ emit_store(
inst->Predicate.SwizzleW);
if (mask) {
mask = LLVMBuildAnd(bld->base.builder, mask, pred, "");
mask = LLVMBuildAnd(builder, mask, pred, "");
} else {
mask = pred;
}
@@ -380,10 +380,11 @@ emit_store(
if (reg->Register.WriteMask != TGSI_WRITEMASK_XYZW) {
LLVMValueRef writemask;
writemask = lp_build_const_mask_aos(bld->base.type, reg->Register.WriteMask);
writemask = lp_build_const_mask_aos(bld->base.gallivm, bld->base.type,
reg->Register.WriteMask);
if (mask) {
mask = LLVMBuildAnd(bld->base.builder, mask, writemask, "");
mask = LLVMBuildAnd(builder, mask, writemask, "");
} else {
mask = writemask;
}
@@ -392,12 +393,12 @@ emit_store(
if (mask) {
LLVMValueRef orig_value;
orig_value = LLVMBuildLoad(bld->base.builder, ptr, "");
orig_value = LLVMBuildLoad(builder, ptr, "");
value = lp_build_select(&bld->base,
mask, value, orig_value);
}
LLVMBuildStore(bld->base.builder, value, ptr);
LLVMBuildStore(builder, value, ptr);
}
@@ -454,7 +455,8 @@ emit_declaration(
struct lp_build_tgsi_aos_context *bld,
const struct tgsi_full_declaration *decl)
{
LLVMTypeRef vec_type = lp_build_vec_type(bld->base.type);
struct gallivm_state *gallivm = bld->base.gallivm;
LLVMTypeRef vec_type = lp_build_vec_type(bld->base.gallivm, bld->base.type);
unsigned first = decl->Range.First;
unsigned last = decl->Range.Last;
@@ -465,31 +467,26 @@ emit_declaration(
case TGSI_FILE_TEMPORARY:
assert(idx < LP_MAX_TGSI_TEMPS);
if (bld->indirect_files & (1 << TGSI_FILE_TEMPORARY)) {
LLVMValueRef array_size = LLVMConstInt(LLVMInt32Type(),
last + 1, 0);
bld->temps_array = lp_build_array_alloca(bld->base.builder,
LLVMValueRef array_size = lp_build_const_int32(gallivm, last + 1);
bld->temps_array = lp_build_array_alloca(bld->base.gallivm,
vec_type, array_size, "");
} else {
bld->temps[idx] = lp_build_alloca(bld->base.builder,
vec_type, "");
bld->temps[idx] = lp_build_alloca(gallivm, vec_type, "");
}
break;
case TGSI_FILE_OUTPUT:
bld->outputs[idx] = lp_build_alloca(bld->base.builder,
vec_type, "");
bld->outputs[idx] = lp_build_alloca(gallivm, vec_type, "");
break;
case TGSI_FILE_ADDRESS:
assert(idx < LP_MAX_TGSI_ADDRS);
bld->addr[idx] = lp_build_alloca(bld->base.builder,
vec_type, "");
bld->addr[idx] = lp_build_alloca(gallivm, vec_type, "");
break;
case TGSI_FILE_PREDICATE:
assert(idx < LP_MAX_TGSI_PREDS);
bld->preds[idx] = lp_build_alloca(bld->base.builder,
vec_type, "");
bld->preds[idx] = lp_build_alloca(gallivm, vec_type, "");
break;
default:
@@ -644,7 +641,7 @@ emit_instruction(
src0 = emit_fetch(bld, inst, 0);
src1 = emit_fetch(bld, inst, 1);
src2 = emit_fetch(bld, inst, 2);
tmp1 = lp_build_const_vec(bld->base.type, 0.5);
tmp1 = lp_build_const_vec(bld->base.gallivm, bld->base.type, 0.5);
tmp0 = lp_build_cmp(&bld->base, PIPE_FUNC_GREATER, src2, tmp1);
dst0 = lp_build_select(&bld->base, tmp0, src0, src1);
break;
@@ -1039,7 +1036,7 @@ emit_instruction(
void
lp_build_tgsi_aos(LLVMBuilderRef builder,
lp_build_tgsi_aos(struct gallivm_state *gallivm,
const struct tgsi_token *tokens,
struct lp_type type,
const unsigned char swizzles[4],
@@ -1058,8 +1055,8 @@ lp_build_tgsi_aos(LLVMBuilderRef builder,
/* Setup build context */
memset(&bld, 0, sizeof bld);
lp_build_context_init(&bld.base, builder, type);
lp_build_context_init(&bld.int_bld, builder, lp_int_type(type));
lp_build_context_init(&bld.base, gallivm, type);
lp_build_context_init(&bld.int_bld, gallivm, lp_int_type(type));
for (chan = 0; chan < 4; ++chan) {
bld.swizzles[chan] = swizzles[chan];
@@ -1131,7 +1128,7 @@ lp_build_tgsi_aos(LLVMBuilderRef builder,
imm[swizzle] = parse.FullToken.FullImmediate.u[chan].Float;
}
bld.immediates[num_immediates] =
lp_build_const_aos(type,
lp_build_const_aos(gallivm, type,
imm[0], imm[1], imm[2], imm[3],
NULL);
num_immediates++;
@@ -1156,7 +1153,7 @@ lp_build_tgsi_aos(LLVMBuilderRef builder,
}
if (0) {
LLVMBasicBlockRef block = LLVMGetInsertBlock(builder);
LLVMBasicBlockRef block = LLVMGetInsertBlock(gallivm->builder);
LLVMValueRef function = LLVMGetBasicBlockParent(block);
debug_printf("11111111111111111111111111111 \n");
tgsi_dump(tokens, 0);
@@ -1167,7 +1164,7 @@ lp_build_tgsi_aos(LLVMBuilderRef builder,
if (0) {
LLVMModuleRef module = LLVMGetGlobalParent(
LLVMGetBasicBlockParent(LLVMGetInsertBlock(bld.base.builder)));
LLVMGetBasicBlockParent(LLVMGetInsertBlock(gallivm->builder)));
LLVMDumpModule(module);
}

View File

@@ -51,6 +51,7 @@
#include "lp_bld_arit.h"
#include "lp_bld_bitarit.h"
#include "lp_bld_gather.h"
#include "lp_bld_init.h"
#include "lp_bld_logic.h"
#include "lp_bld_swizzle.h"
#include "lp_bld_flow.h"
@@ -144,6 +145,18 @@ struct lp_build_tgsi_soa_context
*/
LLVMValueRef temps_array;
/* We allocate/use this array of output if (1 << TGSI_FILE_OUTPUT) is
* set in the indirect_files field.
* The outputs[] array above is unused then.
*/
LLVMValueRef outputs_array;
/* We allocate/use this array of inputs if (1 << TGSI_FILE_INPUT) is
* set in the indirect_files field.
* The inputs[] array above is unused then.
*/
LLVMValueRef inputs_array;
const struct tgsi_shader_info *info;
/** bitmask indicating which register files are accessed indirectly */
unsigned indirect_files;
@@ -163,22 +176,24 @@ static void lp_exec_mask_init(struct lp_exec_mask *mask, struct lp_build_context
mask->loop_stack_size = 0;
mask->call_stack_size = 0;
mask->int_vec_type = lp_build_int_vec_type(mask->bld->type);
mask->int_vec_type = lp_build_int_vec_type(bld->gallivm, mask->bld->type);
mask->exec_mask = mask->ret_mask = mask->break_mask = mask->cont_mask = mask->cond_mask =
LLVMConstAllOnes(mask->int_vec_type);
}
static void lp_exec_mask_update(struct lp_exec_mask *mask)
{
LLVMBuilderRef builder = mask->bld->gallivm->builder;
if (mask->loop_stack_size) {
/*for loops we need to update the entire mask at runtime */
LLVMValueRef tmp;
assert(mask->break_mask);
tmp = LLVMBuildAnd(mask->bld->builder,
tmp = LLVMBuildAnd(builder,
mask->cont_mask,
mask->break_mask,
"maskcb");
mask->exec_mask = LLVMBuildAnd(mask->bld->builder,
mask->exec_mask = LLVMBuildAnd(builder,
mask->cond_mask,
tmp,
"maskfull");
@@ -186,7 +201,7 @@ static void lp_exec_mask_update(struct lp_exec_mask *mask)
mask->exec_mask = mask->cond_mask;
if (mask->call_stack_size) {
mask->exec_mask = LLVMBuildAnd(mask->bld->builder,
mask->exec_mask = LLVMBuildAnd(builder,
mask->exec_mask,
mask->ret_mask,
"callmask");
@@ -200,13 +215,15 @@ static void lp_exec_mask_update(struct lp_exec_mask *mask)
static void lp_exec_mask_cond_push(struct lp_exec_mask *mask,
LLVMValueRef val)
{
LLVMBuilderRef builder = mask->bld->gallivm->builder;
assert(mask->cond_stack_size < LP_MAX_TGSI_NESTING);
if (mask->cond_stack_size == 0) {
assert(mask->cond_mask == LLVMConstAllOnes(mask->int_vec_type));
}
mask->cond_stack[mask->cond_stack_size++] = mask->cond_mask;
assert(LLVMTypeOf(val) == mask->int_vec_type);
mask->cond_mask = LLVMBuildAnd(mask->bld->builder,
mask->cond_mask = LLVMBuildAnd(builder,
mask->cond_mask,
val,
"");
@@ -215,6 +232,7 @@ static void lp_exec_mask_cond_push(struct lp_exec_mask *mask,
static void lp_exec_mask_cond_invert(struct lp_exec_mask *mask)
{
LLVMBuilderRef builder = mask->bld->gallivm->builder;
LLVMValueRef prev_mask;
LLVMValueRef inv_mask;
@@ -224,9 +242,9 @@ static void lp_exec_mask_cond_invert(struct lp_exec_mask *mask)
assert(prev_mask == LLVMConstAllOnes(mask->int_vec_type));
}
inv_mask = LLVMBuildNot(mask->bld->builder, mask->cond_mask, "");
inv_mask = LLVMBuildNot(builder, mask->cond_mask, "");
mask->cond_mask = LLVMBuildAnd(mask->bld->builder,
mask->cond_mask = LLVMBuildAnd(builder,
inv_mask,
prev_mask, "");
lp_exec_mask_update(mask);
@@ -241,6 +259,8 @@ static void lp_exec_mask_cond_pop(struct lp_exec_mask *mask)
static void lp_exec_bgnloop(struct lp_exec_mask *mask)
{
LLVMBuilderRef builder = mask->bld->gallivm->builder;
if (mask->loop_stack_size == 0) {
assert(mask->loop_block == NULL);
assert(mask->cont_mask == LLVMConstAllOnes(mask->int_vec_type));
@@ -256,25 +276,26 @@ static void lp_exec_bgnloop(struct lp_exec_mask *mask)
mask->loop_stack[mask->loop_stack_size].break_var = mask->break_var;
++mask->loop_stack_size;
mask->break_var = lp_build_alloca(mask->bld->builder, mask->int_vec_type, "");
LLVMBuildStore(mask->bld->builder, mask->break_mask, mask->break_var);
mask->break_var = lp_build_alloca(mask->bld->gallivm, mask->int_vec_type, "");
LLVMBuildStore(builder, mask->break_mask, mask->break_var);
mask->loop_block = lp_build_insert_new_block(mask->bld->builder, "bgnloop");
LLVMBuildBr(mask->bld->builder, mask->loop_block);
LLVMPositionBuilderAtEnd(mask->bld->builder, mask->loop_block);
mask->loop_block = lp_build_insert_new_block(mask->bld->gallivm, "bgnloop");
LLVMBuildBr(builder, mask->loop_block);
LLVMPositionBuilderAtEnd(builder, mask->loop_block);
mask->break_mask = LLVMBuildLoad(mask->bld->builder, mask->break_var, "");
mask->break_mask = LLVMBuildLoad(builder, mask->break_var, "");
lp_exec_mask_update(mask);
}
static void lp_exec_break(struct lp_exec_mask *mask)
{
LLVMValueRef exec_mask = LLVMBuildNot(mask->bld->builder,
LLVMBuilderRef builder = mask->bld->gallivm->builder;
LLVMValueRef exec_mask = LLVMBuildNot(builder,
mask->exec_mask,
"break");
mask->break_mask = LLVMBuildAnd(mask->bld->builder,
mask->break_mask = LLVMBuildAnd(builder,
mask->break_mask,
exec_mask, "break_full");
@@ -283,11 +304,12 @@ static void lp_exec_break(struct lp_exec_mask *mask)
static void lp_exec_continue(struct lp_exec_mask *mask)
{
LLVMValueRef exec_mask = LLVMBuildNot(mask->bld->builder,
LLVMBuilderRef builder = mask->bld->gallivm->builder;
LLVMValueRef exec_mask = LLVMBuildNot(builder,
mask->exec_mask,
"");
mask->cont_mask = LLVMBuildAnd(mask->bld->builder,
mask->cont_mask = LLVMBuildAnd(builder,
mask->cont_mask,
exec_mask, "");
@@ -295,11 +317,14 @@ static void lp_exec_continue(struct lp_exec_mask *mask)
}
static void lp_exec_endloop(struct lp_exec_mask *mask)
static void lp_exec_endloop(struct gallivm_state *gallivm,
struct lp_exec_mask *mask)
{
LLVMBuilderRef builder = mask->bld->gallivm->builder;
LLVMBasicBlockRef endloop;
LLVMTypeRef reg_type = LLVMIntType(mask->bld->type.width*
mask->bld->type.length);
LLVMTypeRef reg_type = LLVMIntTypeInContext(gallivm->context,
mask->bld->type.width *
mask->bld->type.length);
LLVMValueRef i1cond;
assert(mask->break_mask);
@@ -315,21 +340,21 @@ static void lp_exec_endloop(struct lp_exec_mask *mask)
* Unlike the continue mask, the break_mask must be preserved across loop
* iterations
*/
LLVMBuildStore(mask->bld->builder, mask->break_mask, mask->break_var);
LLVMBuildStore(builder, mask->break_mask, mask->break_var);
/* i1cond = (mask == 0) */
i1cond = LLVMBuildICmp(
mask->bld->builder,
builder,
LLVMIntNE,
LLVMBuildBitCast(mask->bld->builder, mask->exec_mask, reg_type, ""),
LLVMBuildBitCast(builder, mask->exec_mask, reg_type, ""),
LLVMConstNull(reg_type), "");
endloop = lp_build_insert_new_block(mask->bld->builder, "endloop");
endloop = lp_build_insert_new_block(mask->bld->gallivm, "endloop");
LLVMBuildCondBr(mask->bld->builder,
LLVMBuildCondBr(builder,
i1cond, mask->loop_block, endloop);
LLVMPositionBuilderAtEnd(mask->bld->builder, endloop);
LLVMPositionBuilderAtEnd(builder, endloop);
assert(mask->loop_stack_size);
--mask->loop_stack_size;
@@ -351,10 +376,12 @@ static void lp_exec_mask_store(struct lp_exec_mask *mask,
LLVMValueRef val,
LLVMValueRef dst)
{
LLVMBuilderRef builder = mask->bld->gallivm->builder;
/* Mix the predicate and execution mask */
if (mask->has_mask) {
if (pred) {
pred = LLVMBuildAnd(mask->bld->builder, pred, mask->exec_mask, "");
pred = LLVMBuildAnd(builder, pred, mask->exec_mask, "");
} else {
pred = mask->exec_mask;
}
@@ -363,14 +390,14 @@ static void lp_exec_mask_store(struct lp_exec_mask *mask,
if (pred) {
LLVMValueRef real_val, dst_val;
dst_val = LLVMBuildLoad(mask->bld->builder, dst, "");
dst_val = LLVMBuildLoad(builder, dst, "");
real_val = lp_build_select(mask->bld,
pred,
val, dst_val);
LLVMBuildStore(mask->bld->builder, real_val, dst);
LLVMBuildStore(builder, real_val, dst);
} else
LLVMBuildStore(mask->bld->builder, val, dst);
LLVMBuildStore(builder, val, dst);
}
static void lp_exec_mask_call(struct lp_exec_mask *mask,
@@ -386,6 +413,7 @@ static void lp_exec_mask_call(struct lp_exec_mask *mask,
static void lp_exec_mask_ret(struct lp_exec_mask *mask, int *pc)
{
LLVMBuilderRef builder = mask->bld->gallivm->builder;
LLVMValueRef exec_mask;
if (mask->call_stack_size == 0) {
@@ -393,11 +421,11 @@ static void lp_exec_mask_ret(struct lp_exec_mask *mask, int *pc)
*pc = -1;
return;
}
exec_mask = LLVMBuildNot(mask->bld->builder,
exec_mask = LLVMBuildNot(builder,
mask->exec_mask,
"ret");
mask->ret_mask = LLVMBuildAnd(mask->bld->builder,
mask->ret_mask = LLVMBuildAnd(builder,
mask->ret_mask,
exec_mask, "ret_full");
@@ -429,16 +457,39 @@ get_temp_ptr(struct lp_build_tgsi_soa_context *bld,
unsigned index,
unsigned chan)
{
LLVMBuilderRef builder = bld->base.gallivm->builder;
assert(chan < 4);
if (bld->indirect_files & (1 << TGSI_FILE_TEMPORARY)) {
LLVMValueRef lindex = lp_build_const_int32(index * 4 + chan);
return LLVMBuildGEP(bld->base.builder, bld->temps_array, &lindex, 1, "");
LLVMValueRef lindex = lp_build_const_int32(bld->base.gallivm, index * 4 + chan);
return LLVMBuildGEP(builder, bld->temps_array, &lindex, 1, "");
}
else {
return bld->temps[index][chan];
}
}
/**
* Return pointer to a output register channel (src or dest).
* Note that indirect addressing cannot be handled here.
* \param index which output register
* \param chan which channel of the output register.
*/
static LLVMValueRef
get_output_ptr(struct lp_build_tgsi_soa_context *bld,
unsigned index,
unsigned chan)
{
LLVMBuilderRef builder = bld->base.gallivm->builder;
assert(chan < 4);
if (bld->indirect_files & (1 << TGSI_FILE_OUTPUT)) {
LLVMValueRef lindex = lp_build_const_int32(bld->base.gallivm,
index * 4 + chan);
return LLVMBuildGEP(builder, bld->outputs_array, &lindex, 1, "");
}
else {
return bld->outputs[index][chan];
}
}
/**
* Gather vector.
@@ -450,6 +501,7 @@ build_gather(struct lp_build_tgsi_soa_context *bld,
LLVMValueRef base_ptr,
LLVMValueRef indexes)
{
LLVMBuilderRef builder = bld->base.gallivm->builder;
LLVMValueRef res = bld->base.undef;
unsigned i;
@@ -457,14 +509,14 @@ build_gather(struct lp_build_tgsi_soa_context *bld,
* Loop over elements of index_vec, load scalar value, insert it into 'res'.
*/
for (i = 0; i < bld->base.type.length; i++) {
LLVMValueRef ii = LLVMConstInt(LLVMInt32Type(), i, 0);
LLVMValueRef index = LLVMBuildExtractElement(bld->base.builder,
LLVMValueRef ii = lp_build_const_int32(bld->base.gallivm, i);
LLVMValueRef index = LLVMBuildExtractElement(builder,
indexes, ii, "");
LLVMValueRef scalar_ptr = LLVMBuildGEP(bld->base.builder, base_ptr,
LLVMValueRef scalar_ptr = LLVMBuildGEP(builder, base_ptr,
&index, 1, "gather_ptr");
LLVMValueRef scalar = LLVMBuildLoad(bld->base.builder, scalar_ptr, "");
LLVMValueRef scalar = LLVMBuildLoad(builder, scalar_ptr, "");
res = LLVMBuildInsertElement(bld->base.builder, res, scalar, ii, "");
res = LLVMBuildInsertElement(builder, res, scalar, ii, "");
}
return res;
@@ -482,13 +534,14 @@ emit_mask_scatter(struct lp_build_tgsi_soa_context *bld,
struct lp_exec_mask *mask,
LLVMValueRef pred)
{
LLVMBuilderRef builder = bld->base.builder;
struct gallivm_state *gallivm = bld->base.gallivm;
LLVMBuilderRef builder = gallivm->builder;
unsigned i;
/* Mix the predicate and execution mask */
if (mask->has_mask) {
if (pred) {
pred = LLVMBuildAnd(mask->bld->builder, pred, mask->exec_mask, "");
pred = LLVMBuildAnd(builder, pred, mask->exec_mask, "");
}
else {
pred = mask->exec_mask;
@@ -499,7 +552,7 @@ emit_mask_scatter(struct lp_build_tgsi_soa_context *bld,
* Loop over elements of index_vec, store scalar value.
*/
for (i = 0; i < bld->base.type.length; i++) {
LLVMValueRef ii = LLVMConstInt(LLVMInt32Type(), i, 0);
LLVMValueRef ii = lp_build_const_int32(gallivm, i);
LLVMValueRef index = LLVMBuildExtractElement(builder, indexes, ii, "");
LLVMValueRef scalar_ptr = LLVMBuildGEP(builder, base_ptr, &index, 1, "scatter_ptr");
LLVMValueRef val = LLVMBuildExtractElement(builder, values, ii, "scatter_val");
@@ -507,7 +560,7 @@ emit_mask_scatter(struct lp_build_tgsi_soa_context *bld,
LLVMBuildExtractElement(builder, pred, ii, "scatter_pred") : NULL;
if (0)
lp_build_printf(builder, "scatter %d: val %f at %d %p\n",
lp_build_printf(gallivm, "scatter %d: val %f at %d %p\n",
ii, val, index, scalar_ptr);
if (scalar_pred) {
@@ -534,6 +587,7 @@ get_indirect_index(struct lp_build_tgsi_soa_context *bld,
unsigned reg_file, unsigned reg_index,
const struct tgsi_src_register *indirect_reg)
{
LLVMBuilderRef builder = bld->base.gallivm->builder;
struct lp_build_context *uint_bld = &bld->uint_bld;
/* always use X component of address register */
unsigned swizzle = indirect_reg->SwizzleX;
@@ -544,21 +598,22 @@ get_indirect_index(struct lp_build_tgsi_soa_context *bld,
assert(bld->indirect_files & (1 << reg_file));
base = lp_build_const_int_vec(uint_bld->type, reg_index);
base = lp_build_const_int_vec(bld->base.gallivm, uint_bld->type, reg_index);
assert(swizzle < 4);
rel = LLVMBuildLoad(bld->base.builder,
rel = LLVMBuildLoad(builder,
bld->addr[indirect_reg->Index][swizzle],
"load addr reg");
/* for indexing we want integers */
rel = LLVMBuildFPToSI(bld->base.builder,
rel = LLVMBuildFPToSI(builder,
rel,
uint_bld->vec_type, "");
index = lp_build_add(uint_bld, base, rel);
max_index = lp_build_const_int_vec(uint_bld->type,
max_index = lp_build_const_int_vec(bld->base.gallivm,
uint_bld->type,
bld->info->file_max[reg_file]);
assert(!uint_bld->type.sign);
@@ -578,6 +633,8 @@ emit_fetch(
unsigned src_op,
const unsigned chan_index )
{
struct gallivm_state *gallivm = bld->base.gallivm;
LLVMBuilderRef builder = gallivm->builder;
struct lp_build_context *uint_bld = &bld->uint_bld;
const struct tgsi_full_src_register *reg = &inst->Src[src_op];
const unsigned swizzle =
@@ -603,7 +660,7 @@ emit_fetch(
case TGSI_FILE_CONSTANT:
if (reg->Register.Indirect) {
LLVMValueRef swizzle_vec =
lp_build_const_int_vec(uint_bld->type, swizzle);
lp_build_const_int_vec(bld->base.gallivm, uint_bld->type, swizzle);
LLVMValueRef index_vec; /* index into the const buffer */
/* index_vec = indirect_index * 4 + swizzle */
@@ -617,11 +674,11 @@ emit_fetch(
LLVMValueRef index; /* index into the const buffer */
LLVMValueRef scalar, scalar_ptr;
index = lp_build_const_int32(reg->Register.Index*4 + swizzle);
index = lp_build_const_int32(gallivm, reg->Register.Index*4 + swizzle);
scalar_ptr = LLVMBuildGEP(bld->base.builder, bld->consts_ptr,
scalar_ptr = LLVMBuildGEP(builder, bld->consts_ptr,
&index, 1, "");
scalar = LLVMBuildLoad(bld->base.builder, scalar_ptr, "");
scalar = LLVMBuildLoad(builder, scalar_ptr, "");
res = lp_build_broadcast_scalar(&bld->base, scalar);
}
@@ -633,16 +690,49 @@ emit_fetch(
break;
case TGSI_FILE_INPUT:
res = bld->inputs[reg->Register.Index][swizzle];
if (reg->Register.Indirect) {
LLVMValueRef swizzle_vec =
lp_build_const_int_vec(gallivm, uint_bld->type, swizzle);
LLVMValueRef length_vec =
lp_build_const_int_vec(gallivm, uint_bld->type, bld->base.type.length);
LLVMValueRef index_vec; /* index into the const buffer */
LLVMValueRef inputs_array;
LLVMTypeRef float4_ptr_type;
/* index_vec = (indirect_index * 4 + swizzle) * length */
index_vec = lp_build_shl_imm(uint_bld, indirect_index, 2);
index_vec = lp_build_add(uint_bld, index_vec, swizzle_vec);
index_vec = lp_build_mul(uint_bld, index_vec, length_vec);
/* cast inputs_array pointer to float* */
float4_ptr_type = LLVMPointerType(LLVMFloatTypeInContext(gallivm->context), 0);
inputs_array = LLVMBuildBitCast(builder, bld->inputs_array,
float4_ptr_type, "");
/* Gather values from the temporary register array */
res = build_gather(bld, inputs_array, index_vec);
} else {
if (bld->indirect_files & (1 << TGSI_FILE_INPUT)) {
LLVMValueRef lindex = lp_build_const_int32(gallivm,
reg->Register.Index * 4 + swizzle);
LLVMValueRef input_ptr = LLVMBuildGEP(builder,
bld->inputs_array, &lindex, 1, "");
res = LLVMBuildLoad(builder, input_ptr, "");
}
else {
res = bld->inputs[reg->Register.Index][swizzle];
}
}
assert(res);
break;
case TGSI_FILE_TEMPORARY:
if (reg->Register.Indirect) {
LLVMValueRef swizzle_vec =
lp_build_const_int_vec(uint_bld->type, swizzle);
lp_build_const_int_vec(bld->base.gallivm, uint_bld->type, swizzle);
LLVMValueRef length_vec =
lp_build_const_int_vec(uint_bld->type, bld->base.type.length);
lp_build_const_int_vec(bld->base.gallivm, uint_bld->type,
bld->base.type.length);
LLVMValueRef index_vec; /* index into the const buffer */
LLVMValueRef temps_array;
LLVMTypeRef float4_ptr_type;
@@ -653,8 +743,8 @@ emit_fetch(
index_vec = lp_build_mul(uint_bld, index_vec, length_vec);
/* cast temps_array pointer to float* */
float4_ptr_type = LLVMPointerType(LLVMFloatType(), 0);
temps_array = LLVMBuildBitCast(uint_bld->builder, bld->temps_array,
float4_ptr_type = LLVMPointerType(LLVMFloatTypeInContext(bld->base.gallivm->context), 0);
temps_array = LLVMBuildBitCast(builder, bld->temps_array,
float4_ptr_type, "");
/* Gather values from the temporary register array */
@@ -663,7 +753,7 @@ emit_fetch(
else {
LLVMValueRef temp_ptr;
temp_ptr = get_temp_ptr(bld, reg->Register.Index, swizzle);
res = LLVMBuildLoad(bld->base.builder, temp_ptr, "");
res = LLVMBuildLoad(builder, temp_ptr, "");
if (!res)
return bld->base.undef;
}
@@ -733,6 +823,7 @@ emit_fetch_predicate(
const struct tgsi_full_instruction *inst,
LLVMValueRef *pred)
{
LLVMBuilderRef builder = bld->base.gallivm->builder;
unsigned index;
unsigned char swizzles[4];
LLVMValueRef unswizzled[4] = {NULL, NULL, NULL, NULL};
@@ -762,7 +853,7 @@ emit_fetch_predicate(
* in the swizzles
*/
if (!unswizzled[swizzle]) {
value = LLVMBuildLoad(bld->base.builder,
value = LLVMBuildLoad(builder,
bld->preds[index][swizzle], "");
/*
@@ -772,13 +863,13 @@ emit_fetch_predicate(
* is needlessly causing two comparisons due to storing the intermediate
* result as float vector instead of an integer mask vector.
*/
value = lp_build_compare(bld->base.builder,
value = lp_build_compare(bld->base.gallivm,
bld->base.type,
PIPE_FUNC_NOTEQUAL,
value,
bld->base.zero);
if (inst->Predicate.Negate) {
value = LLVMBuildNot(bld->base.builder, value, "");
value = LLVMBuildNot(builder, value, "");
}
unswizzled[swizzle] = value;
@@ -803,6 +894,8 @@ emit_store(
LLVMValueRef pred,
LLVMValueRef value)
{
struct gallivm_state *gallivm = bld->base.gallivm;
LLVMBuilderRef builder = gallivm->builder;
const struct tgsi_full_dst_register *reg = &inst->Dst[index];
struct lp_build_context *uint_bld = &bld->uint_bld;
LLVMValueRef indirect_index = NULL;
@@ -817,7 +910,7 @@ emit_store(
break;
case TGSI_SAT_MINUS_PLUS_ONE:
value = lp_build_max(&bld->base, value, lp_build_const_vec(bld->base.type, -1.0));
value = lp_build_max(&bld->base, value, lp_build_const_vec(bld->base.gallivm, bld->base.type, -1.0));
value = lp_build_min(&bld->base, value, bld->base.one);
break;
@@ -836,27 +929,21 @@ emit_store(
switch( reg->Register.File ) {
case TGSI_FILE_OUTPUT:
lp_exec_mask_store(&bld->exec_mask, pred, value,
bld->outputs[reg->Register.Index][chan_index]);
break;
case TGSI_FILE_TEMPORARY:
if (reg->Register.Indirect) {
LLVMBuilderRef builder = bld->base.builder;
LLVMValueRef chan_vec =
lp_build_const_int_vec(uint_bld->type, chan_index);
lp_build_const_int_vec(gallivm, uint_bld->type, chan_index);
LLVMValueRef length_vec =
lp_build_const_int_vec(uint_bld->type, bld->base.type.length);
lp_build_const_int_vec(gallivm, uint_bld->type, bld->base.type.length);
LLVMValueRef index_vec; /* indexes into the temp registers */
LLVMValueRef temps_array;
LLVMValueRef outputs_array;
LLVMValueRef pixel_offsets;
LLVMTypeRef float_ptr_type;
int i;
/* build pixel offset vector: {0, 1, 2, 3, ...} */
pixel_offsets = uint_bld->undef;
pixel_offsets = uint_bld->undef;
for (i = 0; i < bld->base.type.length; i++) {
LLVMValueRef ii = lp_build_const_int32(i);
LLVMValueRef ii = lp_build_const_int32(gallivm, i);
pixel_offsets = LLVMBuildInsertElement(builder, pixel_offsets,
ii, ii, "");
}
@@ -867,7 +954,51 @@ emit_store(
index_vec = lp_build_mul(uint_bld, index_vec, length_vec);
index_vec = lp_build_add(uint_bld, index_vec, pixel_offsets);
float_ptr_type = LLVMPointerType(LLVMFloatType(), 0);
float_ptr_type =
LLVMPointerType(LLVMFloatTypeInContext(gallivm->context), 0);
outputs_array = LLVMBuildBitCast(builder, bld->outputs_array,
float_ptr_type, "");
/* Scatter store values into temp registers */
emit_mask_scatter(bld, outputs_array, index_vec, value,
&bld->exec_mask, pred);
}
else {
LLVMValueRef out_ptr = get_output_ptr(bld, reg->Register.Index,
chan_index);
lp_exec_mask_store(&bld->exec_mask, pred, value, out_ptr);
}
break;
case TGSI_FILE_TEMPORARY:
if (reg->Register.Indirect) {
LLVMValueRef chan_vec =
lp_build_const_int_vec(gallivm, uint_bld->type, chan_index);
LLVMValueRef length_vec =
lp_build_const_int_vec(gallivm, uint_bld->type,
bld->base.type.length);
LLVMValueRef index_vec; /* indexes into the temp registers */
LLVMValueRef temps_array;
LLVMValueRef pixel_offsets;
LLVMTypeRef float_ptr_type;
int i;
/* build pixel offset vector: {0, 1, 2, 3, ...} */
pixel_offsets = uint_bld->undef;
for (i = 0; i < bld->base.type.length; i++) {
LLVMValueRef ii = lp_build_const_int32(gallivm, i);
pixel_offsets = LLVMBuildInsertElement(builder, pixel_offsets,
ii, ii, "");
}
/* index_vec = (indirect_index * 4 + chan_index) * length + offsets */
index_vec = lp_build_shl_imm(uint_bld, indirect_index, 2);
index_vec = lp_build_add(uint_bld, index_vec, chan_vec);
index_vec = lp_build_mul(uint_bld, index_vec, length_vec);
index_vec = lp_build_add(uint_bld, index_vec, pixel_offsets);
float_ptr_type =
LLVMPointerType(LLVMFloatTypeInContext(gallivm->context), 0);
temps_array = LLVMBuildBitCast(builder, bld->temps_array,
float_ptr_type, "");
@@ -884,7 +1015,7 @@ emit_store(
case TGSI_FILE_ADDRESS:
lp_exec_mask_store(&bld->exec_mask, pred, value,
bld->addr[reg->Indirect.Index][chan_index]);
bld->addr[reg->Register.Index][chan_index]);
break;
case TGSI_FILE_PREDICATE:
@@ -908,6 +1039,7 @@ emit_tex( struct lp_build_tgsi_soa_context *bld,
enum lp_build_tex_modifier modifier,
LLVMValueRef *texel)
{
LLVMBuilderRef builder = bld->base.gallivm->builder;
unsigned unit;
LLVMValueRef lod_bias, explicit_lod;
LLVMValueRef oow = NULL;
@@ -973,13 +1105,12 @@ emit_tex( struct lp_build_tgsi_soa_context *bld,
}
if (modifier == LP_BLD_TEX_MODIFIER_EXPLICIT_DERIV) {
LLVMTypeRef i32t = LLVMInt32Type();
LLVMValueRef index0 = LLVMConstInt(i32t, 0, 0);
LLVMValueRef index0 = lp_build_const_int32(bld->base.gallivm, 0);
for (i = 0; i < num_coords; i++) {
LLVMValueRef src1 = emit_fetch( bld, inst, 1, i );
LLVMValueRef src2 = emit_fetch( bld, inst, 2, i );
ddx[i] = LLVMBuildExtractElement(bld->base.builder, src1, index0, "");
ddy[i] = LLVMBuildExtractElement(bld->base.builder, src2, index0, "");
ddx[i] = LLVMBuildExtractElement(builder, src1, index0, "");
ddy[i] = LLVMBuildExtractElement(builder, src2, index0, "");
}
unit = inst->Src[3].Register.Index;
} else {
@@ -995,7 +1126,7 @@ emit_tex( struct lp_build_tgsi_soa_context *bld,
}
bld->sampler->emit_fetch_texel(bld->sampler,
bld->base.builder,
bld->base.gallivm,
bld->base.type,
unit, num_coords, coords,
ddx, ddy,
@@ -1050,6 +1181,7 @@ emit_kil(
const struct tgsi_full_instruction *inst,
int pc)
{
LLVMBuilderRef builder = bld->base.gallivm->builder;
const struct tgsi_full_src_register *reg = &inst->Src[0];
LLVMValueRef terms[NUM_CHANNELS];
LLVMValueRef mask;
@@ -1081,7 +1213,7 @@ emit_kil(
chan_mask = lp_build_cmp(&bld->base, PIPE_FUNC_GEQUAL, terms[chan_index], bld->base.zero);
if(mask)
mask = LLVMBuildAnd(bld->base.builder, mask, chan_mask, "");
mask = LLVMBuildAnd(builder, mask, chan_mask, "");
else
mask = chan_mask;
}
@@ -1107,13 +1239,14 @@ emit_kilp(struct lp_build_tgsi_soa_context *bld,
const struct tgsi_full_instruction *inst,
int pc)
{
LLVMBuilderRef builder = bld->base.gallivm->builder;
LLVMValueRef mask;
/* For those channels which are "alive", disable fragment shader
* execution.
*/
if (bld->exec_mask.has_mask) {
mask = LLVMBuildNot(bld->base.builder, bld->exec_mask.exec_mask, "kilp");
mask = LLVMBuildNot(builder, bld->exec_mask.exec_mask, "kilp");
}
else {
LLVMValueRef zero = LLVMConstNull(bld->base.int_vec_type);
@@ -1134,38 +1267,39 @@ emit_kilp(struct lp_build_tgsi_soa_context *bld,
static void
emit_dump_temps(struct lp_build_tgsi_soa_context *bld)
{
LLVMBuilderRef builder = bld->base.builder;
struct gallivm_state *gallivm = bld->base.gallivm;
LLVMBuilderRef builder = gallivm->builder;
LLVMValueRef temp_ptr;
LLVMValueRef i0 = lp_build_const_int32(0);
LLVMValueRef i1 = lp_build_const_int32(1);
LLVMValueRef i2 = lp_build_const_int32(2);
LLVMValueRef i3 = lp_build_const_int32(3);
LLVMValueRef i0 = lp_build_const_int32(gallivm, 0);
LLVMValueRef i1 = lp_build_const_int32(gallivm, 1);
LLVMValueRef i2 = lp_build_const_int32(gallivm, 2);
LLVMValueRef i3 = lp_build_const_int32(gallivm, 3);
int index;
int n = bld->info->file_max[TGSI_FILE_TEMPORARY];
for (index = 0; index < n; index++) {
LLVMValueRef idx = lp_build_const_int32(index);
LLVMValueRef idx = lp_build_const_int32(gallivm, index);
LLVMValueRef v[4][4], res;
int chan;
lp_build_printf(builder, "TEMP[%d]:\n", idx);
lp_build_printf(gallivm, "TEMP[%d]:\n", idx);
for (chan = 0; chan < 4; chan++) {
temp_ptr = get_temp_ptr(bld, index, chan);
res = LLVMBuildLoad(bld->base.builder, temp_ptr, "");
res = LLVMBuildLoad(builder, temp_ptr, "");
v[chan][0] = LLVMBuildExtractElement(builder, res, i0, "");
v[chan][1] = LLVMBuildExtractElement(builder, res, i1, "");
v[chan][2] = LLVMBuildExtractElement(builder, res, i2, "");
v[chan][3] = LLVMBuildExtractElement(builder, res, i3, "");
}
lp_build_printf(builder, " X: %f %f %f %f\n",
lp_build_printf(gallivm, " X: %f %f %f %f\n",
v[0][0], v[0][1], v[0][2], v[0][3]);
lp_build_printf(builder, " Y: %f %f %f %f\n",
lp_build_printf(gallivm, " Y: %f %f %f %f\n",
v[1][0], v[1][1], v[1][2], v[1][3]);
lp_build_printf(builder, " Z: %f %f %f %f\n",
lp_build_printf(gallivm, " Z: %f %f %f %f\n",
v[2][0], v[2][1], v[2][2], v[2][3]);
lp_build_printf(builder, " W: %f %f %f %f\n",
lp_build_printf(gallivm, " W: %f %f %f %f\n",
v[3][0], v[3][1], v[3][2], v[3][3]);
}
}
@@ -1177,6 +1311,7 @@ emit_declaration(
struct lp_build_tgsi_soa_context *bld,
const struct tgsi_full_declaration *decl)
{
struct gallivm_state *gallivm = bld->base.gallivm;
LLVMTypeRef vec_type = bld->base.vec_type;
const unsigned first = decl->Range.First;
const unsigned last = decl->Range.Last;
@@ -1187,39 +1322,31 @@ emit_declaration(
switch (decl->Declaration.File) {
case TGSI_FILE_TEMPORARY:
assert(idx < LP_MAX_TGSI_TEMPS);
if (bld->indirect_files & (1 << TGSI_FILE_TEMPORARY)) {
/* ignore 'first' - we want to index into a 0-based array */
LLVMValueRef array_size = LLVMConstInt(LLVMInt32Type(),
last*4 + 4, 0);
bld->temps_array = lp_build_array_alloca(bld->base.builder,
vec_type, array_size,
"temporary");
idx = last;
} else {
if (!(bld->indirect_files & (1 << TGSI_FILE_TEMPORARY))) {
for (i = 0; i < NUM_CHANNELS; i++)
bld->temps[idx][i] = lp_build_alloca(bld->base.builder,
vec_type, "temp");
bld->temps[idx][i] = lp_build_alloca(gallivm, vec_type, "temp");
}
break;
case TGSI_FILE_OUTPUT:
for (i = 0; i < NUM_CHANNELS; i++)
bld->outputs[idx][i] = lp_build_alloca(bld->base.builder,
vec_type, "output");
if (!(bld->indirect_files & (1 << TGSI_FILE_OUTPUT))) {
for (i = 0; i < NUM_CHANNELS; i++)
bld->outputs[idx][i] = lp_build_alloca(gallivm,
vec_type, "output");
}
break;
case TGSI_FILE_ADDRESS:
assert(idx < LP_MAX_TGSI_ADDRS);
for (i = 0; i < NUM_CHANNELS; i++)
bld->addr[idx][i] = lp_build_alloca(bld->base.builder,
vec_type, "addr");
bld->addr[idx][i] = lp_build_alloca(gallivm, vec_type, "addr");
break;
case TGSI_FILE_PREDICATE:
assert(idx < LP_MAX_TGSI_PREDS);
for (i = 0; i < NUM_CHANNELS; i++)
bld->preds[idx][i] = lp_build_alloca(bld->base.builder,
vec_type, "predicate");
bld->preds[idx][i] = lp_build_alloca(gallivm, vec_type,
"predicate");
break;
default:
@@ -1545,7 +1672,7 @@ emit_instruction(
src0 = emit_fetch( bld, inst, 0, chan_index );
src1 = emit_fetch( bld, inst, 1, chan_index );
src2 = emit_fetch( bld, inst, 2, chan_index );
tmp1 = lp_build_const_vec(bld->base.type, 0.5);
tmp1 = lp_build_const_vec(bld->base.gallivm, bld->base.type, 0.5);
tmp0 = lp_build_cmp( &bld->base, PIPE_FUNC_GREATER, src2, tmp1);
dst0[chan_index] = lp_build_select( &bld->base, tmp0, src0, src1 );
}
@@ -2057,7 +2184,7 @@ emit_instruction(
break;
case TGSI_OPCODE_ENDLOOP:
lp_exec_endloop(&bld->exec_mask);
lp_exec_endloop(bld->base.gallivm, &bld->exec_mask);
break;
case TGSI_OPCODE_ENDSUB:
@@ -2190,7 +2317,7 @@ emit_instruction(
void
lp_build_tgsi_soa(LLVMBuilderRef builder,
lp_build_tgsi_soa(struct gallivm_state *gallivm,
const struct tgsi_token *tokens,
struct lp_type type,
struct lp_build_mask_context *mask,
@@ -2218,9 +2345,9 @@ lp_build_tgsi_soa(LLVMBuilderRef builder,
/* Setup build context */
memset(&bld, 0, sizeof bld);
lp_build_context_init(&bld.base, builder, type);
lp_build_context_init(&bld.uint_bld, builder, lp_uint_type(type));
lp_build_context_init(&bld.elem_bld, builder, lp_elem_type(type));
lp_build_context_init(&bld.base, gallivm, type);
lp_build_context_init(&bld.uint_bld, gallivm, lp_uint_type(type));
lp_build_context_init(&bld.elem_bld, gallivm, lp_elem_type(type));
bld.mask = mask;
bld.pos = pos;
bld.inputs = inputs;
@@ -2239,6 +2366,51 @@ lp_build_tgsi_soa(LLVMBuilderRef builder,
lp_exec_mask_init(&bld.exec_mask, &bld.base);
if (bld.indirect_files & (1 << TGSI_FILE_TEMPORARY)) {
LLVMValueRef array_size =
lp_build_const_int32(gallivm,
info->file_max[TGSI_FILE_TEMPORARY] * 4 + 4);
bld.temps_array = lp_build_array_alloca(gallivm,
bld.base.vec_type, array_size,
"temp_array");
}
if (bld.indirect_files & (1 << TGSI_FILE_OUTPUT)) {
LLVMValueRef array_size =
lp_build_const_int32(gallivm,
info->file_max[TGSI_FILE_OUTPUT] * 4 + 4);
bld.outputs_array = lp_build_array_alloca(gallivm,
bld.base.vec_type, array_size,
"output_array");
}
/* If we have indirect addressing in inputs we need to copy them into
* our alloca array to be able to iterate over them */
if (bld.indirect_files & (1 << TGSI_FILE_INPUT)) {
unsigned index, chan;
LLVMTypeRef vec_type = bld.base.vec_type;
LLVMValueRef array_size =
lp_build_const_int32(gallivm, info->file_max[TGSI_FILE_INPUT]*4 + 4);
bld.inputs_array = lp_build_array_alloca(gallivm,
vec_type, array_size,
"input_array");
assert(info->num_inputs <= info->file_max[TGSI_FILE_INPUT] + 1);
for (index = 0; index < info->num_inputs; ++index) {
for (chan = 0; chan < NUM_CHANNELS; ++chan) {
LLVMValueRef lindex =
lp_build_const_int32(gallivm, index * 4 + chan);
LLVMValueRef input_ptr =
LLVMBuildGEP(gallivm->builder, bld.inputs_array,
&lindex, 1, "");
LLVMValueRef value = bld.inputs[index][chan];
if (value)
LLVMBuildStore(gallivm->builder, value, input_ptr);
}
}
}
tgsi_parse_init( &parse, tokens );
while( !tgsi_parse_end_of_tokens( &parse ) ) {
@@ -2284,7 +2456,7 @@ lp_build_tgsi_soa(LLVMBuilderRef builder,
assert(num_immediates < LP_MAX_TGSI_IMMEDIATES);
for( i = 0; i < size; ++i )
bld.immediates[num_immediates][i] =
lp_build_const_vec(type, parse.FullToken.FullImmediate.u[i].Float);
lp_build_const_vec(gallivm, type, parse.FullToken.FullImmediate.u[i].Float);
for( i = size; i < 4; ++i )
bld.immediates[num_immediates][i] = bld.base.undef;
num_immediates++;
@@ -2308,8 +2480,20 @@ lp_build_tgsi_soa(LLVMBuilderRef builder,
opcode_info->mnemonic);
}
/* If we have indirect addressing in outputs we need to copy our alloca array
* to the outputs slots specified by the called */
if (bld.indirect_files & (1 << TGSI_FILE_OUTPUT)) {
unsigned index, chan;
assert(info->num_outputs <= info->file_max[TGSI_FILE_OUTPUT] + 1);
for (index = 0; index < info->num_outputs; ++index) {
for (chan = 0; chan < NUM_CHANNELS; ++chan) {
bld.outputs[index][chan] = get_output_ptr(&bld, index, chan);
}
}
}
if (0) {
LLVMBasicBlockRef block = LLVMGetInsertBlock(builder);
LLVMBasicBlockRef block = LLVMGetInsertBlock(gallivm->builder);
LLVMValueRef function = LLVMGetBasicBlockParent(block);
debug_printf("11111111111111111111111111111 \n");
tgsi_dump(tokens, 0);
@@ -2320,7 +2504,7 @@ lp_build_tgsi_soa(LLVMBuilderRef builder,
if (0) {
LLVMModuleRef module = LLVMGetGlobalParent(
LLVMGetBasicBlockParent(LLVMGetInsertBlock(bld.base.builder)));
LLVMGetBasicBlockParent(LLVMGetInsertBlock(gallivm->builder)));
LLVMDumpModule(module);
}

View File

@@ -30,34 +30,35 @@
#include "lp_bld_type.h"
#include "lp_bld_const.h"
#include "lp_bld_init.h"
LLVMTypeRef
lp_build_elem_type(struct lp_type type)
lp_build_elem_type(struct gallivm_state *gallivm, struct lp_type type)
{
if (type.floating) {
switch(type.width) {
case 32:
return LLVMFloatType();
return LLVMFloatTypeInContext(gallivm->context);
break;
case 64:
return LLVMDoubleType();
return LLVMDoubleTypeInContext(gallivm->context);
break;
default:
assert(0);
return LLVMFloatType();
return LLVMFloatTypeInContext(gallivm->context);
}
}
else {
return LLVMIntType(type.width);
return LLVMIntTypeInContext(gallivm->context, type.width);
}
}
LLVMTypeRef
lp_build_vec_type(struct lp_type type)
lp_build_vec_type(struct gallivm_state *gallivm,struct lp_type type)
{
LLVMTypeRef elem_type = lp_build_elem_type(type);
LLVMTypeRef elem_type = lp_build_elem_type(gallivm, type);
if (type.length == 1)
return elem_type;
else
@@ -149,16 +150,16 @@ lp_check_value(struct lp_type type, LLVMValueRef val)
LLVMTypeRef
lp_build_int_elem_type(struct lp_type type)
lp_build_int_elem_type(struct gallivm_state *gallivm, struct lp_type type)
{
return LLVMIntType(type.width);
return LLVMIntTypeInContext(gallivm->context, type.width);
}
LLVMTypeRef
lp_build_int_vec_type(struct lp_type type)
lp_build_int_vec_type(struct gallivm_state *gallivm, struct lp_type type)
{
LLVMTypeRef elem_type = lp_build_int_elem_type(type);
LLVMTypeRef elem_type = lp_build_int_elem_type(gallivm, type);
if (type.length == 1)
return elem_type;
else
@@ -170,7 +171,7 @@ lp_build_int_vec_type(struct lp_type type)
* Build int32[4] vector type
*/
LLVMTypeRef
lp_build_int32_vec4_type(void)
lp_build_int32_vec4_type(struct gallivm_state *gallivm)
{
struct lp_type t;
LLVMTypeRef type;
@@ -182,7 +183,7 @@ lp_build_int32_vec4_type(void)
t.width = 32; /* 32-bit int */
t.length = 4; /* 4 elements per vector */
type = lp_build_int_elem_type(t);
type = lp_build_int_elem_type(gallivm, t);
return LLVMVectorType(type, t.length);
}
@@ -383,15 +384,15 @@ lp_dump_llvmtype(LLVMTypeRef t)
void
lp_build_context_init(struct lp_build_context *bld,
LLVMBuilderRef builder,
struct gallivm_state *gallivm,
struct lp_type type)
{
bld->builder = builder;
bld->gallivm = gallivm;
bld->type = type;
bld->int_elem_type = lp_build_int_elem_type(type);
bld->int_elem_type = lp_build_int_elem_type(gallivm, type);
if (type.floating)
bld->elem_type = lp_build_elem_type(type);
bld->elem_type = lp_build_elem_type(gallivm, type);
else
bld->elem_type = bld->int_elem_type;
@@ -406,5 +407,5 @@ lp_build_context_init(struct lp_build_context *bld,
bld->undef = LLVMGetUndef(bld->vec_type);
bld->zero = LLVMConstNull(bld->vec_type);
bld->one = lp_build_one(type);
bld->one = lp_build_one(gallivm, type);
}

Some files were not shown because too many files have changed in this diff Show More