Compare commits

..

355 Commits

Author SHA1 Message Date
Jonas Maebe
1bdf31e8da darwin: do not create double-buffered offscreen pixel formats
http://xquartz.macosforge.org/trac/ticket/536

Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com>
(cherry picked from commit 5fdf1f784b)
2012-09-24 16:07:14 -07:00
Jeremy Huddleston
ecf0bef4cd darwin: Eliminate a possible race condition while destroying a surface
Introduced by: c60ffd2840
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit d65bd195ec)
2012-05-16 18:59:30 -07:00
Jeremy Huddleston
60fa18b90f darwin: Unlock our mutex before destroying it
http://xquartz.macosforge.org/trac/ticket/575

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit a73a800b32)
2012-05-16 18:59:28 -07:00
Jeremy Huddleston
62058968e7 darwin: Eliminate a pthread mutex leak
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
Tested-by: Charles Davis <cdavis@mines.edu>
(cherry picked from commit 1a33c1b2b8)
2012-04-24 00:27:44 -07:00
Jonas Maebe
1fa6c87c88 apple: Fix a use after free
Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit c60ffd2840)
2012-04-23 16:18:36 -07:00
Jonas Maebe
00d310a130 glapi: Correct size of allocated _glapi_table struct
The __glapi_gentable_set_remaining_noop() routine treats the _glapi_struct
as an array of _glapi_get_dispatch_table_size() pointers, so we have to
allocate _glapi_get_dispatch_table_size()*sizeof(void*) bytes rather
than sizeof(struct _glapi_struct) bytes.

Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 8d09f4d0cc)
2012-04-22 20:41:36 -07:00
Jeremy Huddleston
906f670f1a configure.ac: Don't use $CLANG since it will collide with the static analyzer.
We just prefix the $CLANG environment variable in configure.ac with acv_mesa_

Found by: tinderbox
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit b728eefb06)
2012-01-27 09:05:10 -08:00
Anuj Phogat
9ae2499be2 mesa: Add condition in glGetTexImage for zero size textures
TestMipMaps() function in src/OGLconform/textureNPOT.c calls glTexImage2D()
with width = 0. Texture with zero size skips miptree allocation due to a
condition in function _mesa_store_teximage3d(). While calling glGetTexImage()
it results in assertion failure in intel_map_texture_image() due to null mt
pointer.

This patch fixes the issue by detecting the zero size texture early in
glGetTexImage and glGetCompressedTexImage functions. In such a case function
simply returns doing nothing.
Verified that below mentioned bug is fixed by this patch.

https://bugs.freedesktop.org/show_bug.cgi?id=42334

NOTE: This is a candidate for stable branches

Signed-off-by: Anuj Phogat <anuj.phogat@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit f1a9a9bcd1)

Conflicts:

	src/mesa/main/teximage.h
2012-01-19 12:56:15 -08:00
Anuj Phogat
9b3ac17991 intel: Fix segfault in glXSwapBuffers with no bound context
Calling glXSwapBuffers with no bound context causes segmentation
fault in function intelDRI2Flush. All the gl calls should be
ignored after setting the current context to null. So the contents
of framebuffer stay unchanged. But the driver should not seg fault.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=44614

Reported-by: Yi Sun <yi.sun@intel.com>
Signed-off-by: Anuj Phogat <anuj.phogat@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Tested-by: Yi Sun <yi.sun@intel.com>
(cherry picked from commit dd7220652e)

Conflicts:

	src/mesa/drivers/dri/intel/intel_screen.c
2012-01-19 12:56:15 -08:00
Vinson Lee
d7d47d6c0b i965: Fix Coverity wrong sizeof argument defect.
NOTE: This is a candidate for stable release branches.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=42542
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 3e18ad7fd7)
2012-01-19 12:56:15 -08:00
Paul Berry
29f07f6461 i965 gen4-6: Fix off-by-one errors brw_create_constant_surface()
Commit 9bdc44a528 (i965: Replace struct
with bit shifting for WM pull constant surfaces) accidentally
introduced off-by-one errors into the calculation of the surface
width, height, and depth.  This patch restores the correct
computation.

The reason this wasn't noticed by Piglit tests is that the size of our
constant surfaces is always less than 2^20, therefore the off-by-one
error was causing the "depth" field of the surface to be set to all
1's.  The hardware interpreted this as an extremely large surface, so
overflow checking was effectively disabled.

No Piglit regressions on Sandy Bridge.

NOTE: This is a candidate for the 7.11 and 8.0 branches.

Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit f6f43bd5a2)
2012-01-19 12:56:15 -08:00
Brian Paul
1f18137d28 glw: s/lighgrey/lightgrey/
http://bugs.freedesktop.org/show_bug.cgi?id=44844
2012-01-16 15:07:23 -07:00
Ian Romanick
7722a54e60 glsl: Don't use base type for bit-not when there's an error
Other parts of the compiler assume that expressions will have
well-formed types or the error type.  Just using the type of the thing
being operated on can cause expressions like ~3.14 or ~false to not
have a well-formed type.  This could then result in an assertion
failure in the context epxression handler.

If there is an error processing the expression, set the type of the IR
expression to error.

Fixes piglit's bit-not-0[789].frag tests.

NOTE: This is a candidate for the 7.11 branch.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=42755
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Cc: Vinson Lee <vlee@vmware.com>
(cherry picked from commit 3946448951)
2012-01-04 12:44:49 -08:00
Ian Romanick
bc507471db glsl: Clamp vector indices when lowering to swizzles
This prevents other code from seeing a swizzle of the 16th component
of a vector, for example.

NOTE: This is a candidate for the 7.11 branch.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=42517
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Paul Berry <stereotype441@gmail.com>
Tested-by: Christian Holler <choller@mozilla.com>
(cherry picked from commit 6f5c737970)
2012-01-03 19:23:15 -08:00
Ian Romanick
42d1a7b3de glsl: Silence "ast_to_hir.cpp:1984:25: warning: comparison of unsigned expression >= 0 is always true"
ast_type_qualifier::location should have been a signed integer from
the beginning, and the giant comment in
apply_type_qualifier_to_variable explains why.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=40207
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit f0dd21ac88)
2011-12-22 17:01:53 -08:00
Marcin Slusarz
237d6fa97a nv50/nvc0: fix crash when channel allocation fails 2011-12-11 17:00:42 +01:00
Christoph Bumiller
0e9b79c22a nv50,nvc0: use screen instead of context for flush notifier
Context may become NULL and we still have to be able to flush
pending fences.

(Cherry picked from commit 3069a7eaa5)
Fixes crash on screen shutdown: https://bugs.freedesktop.org/show_bug.cgi?id=43428
2011-12-11 16:58:09 +01:00
Brian Paul
7a18f005ed gallivm: fix build with llvm 3.0svn
https://bugs.freedesktop.org/show_bug.cgi?id=41065
(cherry picked from commit 51002968c9)
2011-12-05 08:58:18 -07:00
Tobias Droste
95a3724c69 gallivm: fix build with LLVM 3.0svn
LLVM 3.0svn added SubtargetInfo as additional parameter to
createMCDisassembler() and createMCInstPrinter().
See revision 139237 of LLVM.

Signed-off-by: Tobias Droste <tdroste@gmx.de>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 1795372fee)
2011-12-05 08:58:12 -07:00
Tobias Droste
98e8fb84de gallivm: fix build with LLVM 3.0svn
LLVM 3.0svn moved TargetRegistry.h and TargetSelect.h.
See revision 138450 of LLVM.

Signed-off-by: Tobias Droste <tdroste@gmx.de>
(cherry picked from commit 4a468de2d7)
2011-12-05 08:58:06 -07:00
Tobias Droste
5b5c436c65 llvmpipe: fix build with LLVM 3.0svn
LLVM 3.0svn introduced a new type system. It defines a new way to create
named structs and removes the (now not needed) LLVMInvalidateStructLayout
function.  See revision 134829 of LLVM.

Signed-off-by: Tobias Droste <tdroste@gmx.de>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 3143e95353)
2011-12-05 08:58:00 -07:00
Vinson Lee
970df845ef gallivm: Rename createAsmInfo to createMCAsmInfo with llvm-3.0.
llvm-3.0svn r135219 renamed createAsmInfo to createMCAsmInfo in
include/llvm/Target/TargetRegistry.h.
(cherry picked from commit 9228bfb375)
2011-12-05 08:57:49 -07:00
Vinson Lee
32f4a83fe8 gallivm: Re-enable LLVMUnionTypeKind case for llvm-2.7 only.
LLVMUnionTypeKind is not in llvm-2.6, llvm-2.8, llvm-2.9, or llvm-3.0svn.
(cherry picked from commit 1844ae7e7e)
2011-12-05 08:57:42 -07:00
Vinson Lee
2c1998b3b9 gallivm: Remove LLVMOpaqueKindType case with llvm-3.0.
llvm-3.0svn r134829 removed LLVMOpaqueKindType from enum LLVMTypeKind in
include/llvm-c/Core.h.
(cherry picked from commit e4189f2e2e)
2011-12-05 08:57:11 -07:00
Vinson Lee
a3337a66ef gallivm: Rename TargetInstrDesc to MCInstrDesc when using llvm-3.0.
llvm-3.0svn revision 134021 renamed TargetInstrDesc to MCInstrDesc.
(cherry picked from commit b61e56756c)
2011-12-05 08:57:00 -07:00
Brian Paul
25ba654229 docs: update news, relnotes list for 7.11.2 release 2011-11-28 08:24:19 -07:00
Ian Romanick
733478975d docs: Add 7.11.2 release md5sums 2011-11-27 14:06:59 -08:00
Ian Romanick
ea72e3cda8 docs: Add 7.11.2 release notes 2011-11-27 14:04:48 -08:00
Ian Romanick
835205e295 mesa: Bump version to 7.11.2 (final) 2011-11-27 14:04:05 -08:00
Chad Versace
2a45566587 intel: Fix separate stencil in builtin DRI2 backend
intelAllocateBuffer() was oblivious to separate stencil buffers.  This
patch fixes it to allocate a non-tiled stencil buffer with special pitch,
just as the DDX does.

Without this, any app that attempted to create an EGL surface with stencil
bits would crash. Of course, this affected only environments that used the
builtin DRI2 backend, such as Android and Wayland.

Fixes GLBenchmark2.1 on Android on gen7.

Note: This is a candidate for the 7.11 branch.
Tested-by: Louie Tsaie <louie.tsai@intel.com>
Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit 79653c12d6)

Conflicts:

	src/mesa/drivers/dri/intel/intel_screen.c
2011-11-21 16:11:50 -08:00
Chad Versace
68060d6bfd intel: Fix region dimensions for stencil buffers received from DDX
I changed the dimensions of the stencil buffer's region, as allocated by
the DDX, at xf86-video-intel commit
   commit 3e55f3e88b40471706d5cd45c4df4010f8675c75
   dri: Do not tile stencil buffer
But I forgot to make the analogous update to the Intel DRI2 glue in Mesa.
This patch makes that update.

Surprisingly, the mismatch did not cause any bugs. But the mismatch, if
left unfixed, *would* create bugs in the next commit.

Note: This is a candidate for the 7.11 branch.
Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit 50b3356078)
2011-11-21 16:10:49 -08:00
Chad Versace
fd35fab340 intel: Simplify stencil detiling arithmetic
When calculating the y offset needed for detiling window system stencil
buffers, replace the term
   region->height * 2 + region->height % 2 - 1
with
   rb->Height - 1 .

The two terms are incidentally equivalent due to some out-of-date,
incorrect code in the Intel DRI2 glue for DDX. (See
intel_process_dri2_buffer_with_separate_stencil(), line ``buffer_height /=
2;``).

Note: This is a candidate for the 7.11 branch (only the intel_span.c hunk).
Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit dc4c3a31c6)

Conflicts:

	src/mesa/drivers/dri/intel/intel_fbo.c
	src/mesa/drivers/dri/intel/intel_span.c
2011-11-21 16:10:22 -08:00
Brian Paul
658d994473 mesa: use format string in _mesa_error() call to silence warning
(cherry picked from commit 5485192fc8)
2011-11-21 19:49:48 +01:00
Brian Paul
8a8ccc00b5 docs: update news.html and relnotes.html for 7.11.1 release 2011-11-21 09:16:38 -07:00
Brian Paul
21364bd09d mesa: fix format selection for meta CopyTexSubImage()
When we do a glReadPixels into the temporary buffer, we don't want to
use GL_LUMINANCE, GL_LUMINANCE_ALPHA or GL_INTENSITY since they will
compute L=R+G+B which is not what we want.

This bug has existed all along but was only exposed by the elimination
of the driver hook for glCopyTexImage() in
5874890c26.

Fixes https://bugs.freedesktop.org/show_bug.cgi?id=39604
Tested-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit e4fdc95277)

(cherry-pick note: This patch and the one before it combine to fix
CopyTexImage from RGBA window to LA texture.  This was believed to
already be fixed by commit 296e6b9, but these changes were also
necessary.)
2011-11-18 12:18:19 -08:00
Brian Paul
6b2c771ddd mesa: stop using ctx->Driver.CopyTexImage1D/2D() hooks
(cherry picked from commit 5874890c26)
2011-11-18 12:12:22 -08:00
Ian Romanick
22ff4ec969 mesa: set version string to 7.11.2-devel 2011-11-18 11:06:33 -08:00
Ian Romanick
228da884c9 docs: Add 7.11.1 release md5sums 2011-11-17 12:12:34 -08:00
Ian Romanick
c0009739bf docs: Add 7.11.1 release notes 2011-11-17 11:45:06 -08:00
Ian Romanick
bb7d993953 mesa: set version string to 7.11.1 2011-11-17 11:44:08 -08:00
Eric Anholt
172de77b12 glsl: Fix gl_NormalMatrix swizzle setup to match i965's invariants.
A driver trying to set up builtin uniforms is faced with a problem:
How do I walk the ir_variable structure (representing an array of
structs, or array of matrices, or struct, or whatever), and set up
driver structures so that dereference of that uniform gets the
corresponding ParameterValues[] entry.  The rule in general is that
each corresponding vector-sized field of an array of structs is one
builtin uniform state slot.  i965 relied on another invariant: each
state slot has a number of unique channel swizzles corresponding to
the number of elements in the field's vector, to avoid needing to walk
the glsl_type in parallel to get at vector_elements.

All of the builtin uniforms followed this behavior, except for
gl_NormalMatrix.  That's a mat3 (so 3 vec3s), but it was swizzled as 3
vec4s.

Fixes piglit glsl-fs-normalmatrix.
Reviewed-by: Paul Berry <stereotype441@gmail.com>
(cherry picked from commit cc4ddc3a1e)
2011-11-10 10:36:52 -08:00
Kenneth Graunke
6b151886fd mesa/get: Move MAX_LIGHTS from GL/ES2 to GL/ES1.
It's required for ES 1.0 and 1.1, and isn't specified for ES 2.

While the comment says Mesa depends on it internally, removing it from
ES2 doesn't seem to regress any Piglit or ES2 conformance tests.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 5785cd2bf5)
2011-11-10 10:21:14 -08:00
José Fonseca
436db5df9c docs: Update llvmpipe docs.
Recommend LLVM 2.9, it has been working quite well, and unlike earlier
versions, it works out-of-the-box without patches.

Update Windows instructions.
2011-11-05 11:03:11 +00:00
Yuanhan Liu
5459781715 intel: don't call unmap pbo if pbo is not mapped
The PBO only needs to be unmapped if one of the previous calls to
_mesa_validate_pbo_* succeeded.  In this case, pixels will be
non-NULL.  Various paths through _mesa_unmap_texmiage_pbo can hit
assertion failures or segfaults if the buffer is not mapped.

To work around this, move the call to _mesa_unmap_teximage_pbo inside
the last 'if (pixels)' block.

NOTE: this is just for 7.11 stable branch

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=42268
2011-11-03 11:55:39 -07:00
Michel Dänzer
b95767a57a r300g: Fix queries on big endian hosts.
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Corbin Simpson <MostAwesomeDude@gmail.com>
2011-11-02 22:29:14 +01:00
Michel Dänzer
f60e81ecb2 gallium/util: Add macros for converting from little endian to CPU byte order.
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
(cherry picked from commit 4a3be16fd2)
2011-11-02 22:29:07 +01:00
Marek Olšák
d15ce8dd29 r300g: don't call u_trim_pipe_prim in r300_swtcl_draw_vbo
This was dead code anyway.
(cherry picked from commit 21e3c585f7)
2011-11-02 22:27:05 +01:00
Tom Fogal
449b301eec Only use gcc visibility support with gcc4+.
I had a colleague hitting issues compiling with an old gcc3.2
system.  These patches got them through.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry-picked from commit cbb2b4149b)
2011-11-02 12:58:34 -06:00
Adam Jackson
4464ee1a9a glx: Don't enable INTEL_swap_event unconditionally
DRI2 supports this now - and already enables it explicitly - but drisw
does not and should not.  Otherwise toolkits like clutter will only ever
SwapBuffers once and wait forever for an event that's not coming.

Signed-off-by: Adam Jackson <ajax@redhat.com>
(cherry picked from commit 25620eb1d2)
2011-10-28 20:41:45 -04:00
Kenneth Graunke
439628318b i965: Apply post-sync non-zero workaround to homebrew workaround.
In commit 3e5d3626, Eric added a homebrew workaround to fix GPU hangs in
the Mesa "engine" demo and oglc's api-texcoord test.

Unfortunately, his PIPE_CONTROL contains a Depth Stall, which
necessitates the post-sync non-zero workaround,

Fixes GPU hangs in Civilization 4, PlaneShift, Minecraft, Neverwinter
Nights, 3DMMES, and hopefully Heroes of Newerth as well.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=40324
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=41096
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
Tested-by: Joel <k00_fol@k.kth.se> (Neverwinter Nights)
Tested-by: brot <brot@minad.de> (Minecraft)
Tested-by: Eric Anholt <eric@anholt.net> (3DMMES)
Tested-by: Kenneth Graunke <kenneth@whitecape.org> (Civ 4 & PlaneShift)

(cherry-picked from commit 3cc0a7be23)
2011-10-27 10:37:40 -07:00
Marek Olšák
00c44de1a6 r600g: set correct tiling flags in depth info
The kernel currently overwrites the flags, but if we stopped doing that,
this would break badly.
(cherry picked from commit faa16dc456)

BTW, this may be an actual fix for very old kernels.
https://bugs.freedesktop.org/show_bug.cgi?id=42175

Conflicts:

	src/gallium/drivers/r600/evergreen_state.c
	src/gallium/drivers/r600/r600_state.c
2011-10-26 14:51:30 +02:00
Brian Paul
e4f88bcad3 mesa: fix format/type check in unpack_image() for bitmaps
Passing type == GL_BITMAP returns 0 while error values return -1.
This fixes glPolygonStipple being compiled into display lists.
(cherry picked from commit 2ce8c3553b)
2011-10-25 18:41:29 -07:00
Brian Paul
0092375714 mesa: fix incorrect error code in _mesa_FramebufferTexture1D/3DEXT()
The spec says GL_INVALID_OPERATION is generated when texture!=0 and
textarget is not a legal value.  We had this right for the 2D function.
(cherry picked from commit ccecc08f79)
2011-10-25 18:41:29 -07:00
Ben Widawsky
c4d37ed43a intel: GetBuffer fix
After copy buffer on preGEN6, it is necessary to wait for the blit to
complete before returning data to the user.

This should fix the piglit test: copy_buffer_coherency (pre-GEN6).

Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit fa351bd2e0)
2011-10-25 18:41:29 -07:00
Eric Anholt
242a18dc46 mesa: Don't error on glFeedbackBuffer(size = 0, buffer = NULL)
The existing error result doesn't appear in the GL 2.1 or 3.2
compatibility specs, and triggers an unexpected GL error in Intel's
oglconform when it tries to reset the feedback state after usage so
that the "diff the state at error time vs. context init time" code
doesn't generate spurious diffs.  The unexpected GL error then
translates into testcase failure.  Brian wants the safety check on
buffer = NULL, though, so that people can't as easily set up a broken
buffer.
(cherry picked from commit 07e5295b6f)
2011-10-25 18:41:29 -07:00
Ian Romanick
0abee468f8 mesa: Advertise GL_OES_compressed_paletted_texture in OpenGL ES1.x
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Tested-by: Jin Yang <jin.a.yang@intel.com>
(cherry picked from commit 24a113093b)
2011-10-25 18:41:29 -07:00
Ian Romanick
72ea656dad mesa: Remove redundant compressed paletted texture error checks
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Tested-by: Jin Yang <jin.a.yang@intel.com>
(cherry picked from commit 13757f7080)
2011-10-25 18:41:29 -07:00
Ian Romanick
30b199f867 mesa: Refactor compressed texture error checks to work with paletted textures
This code was really broken before.  A lot of the error checks were
done much later (too late), and some of the error checks would fail.
The underlying problem is that Mesa doesn't ever keep compressed paletted
textures in their original format.  The textures are immediately
converted to some RGB or RGBA format.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=39991
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Tested-by: Jin Yang <jin.a.yang@intel.com>
(cherry picked from commit 3ebbfc8372)
2011-10-25 18:41:29 -07:00
Ian Romanick
9b242a03d5 mesa: Add _mesa_cpal_compressed_format_type
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Tested-by: Jin Yang <jin.a.yang@intel.com>
(cherry picked from commit b433e7ba07)
2011-10-25 18:41:29 -07:00
Ian Romanick
8317a43a75 mesa: Refactor expected texture size check in cpal_get_info
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Tested-by: Jin Yang <jin.a.yang@intel.com>
(cherry picked from commit a2cab751be)
2011-10-25 18:41:29 -07:00
Ian Romanick
625fdf58c6 mesa: Add GL_OES_compressed_paletted_texture formats to _mesa_base_tex_format
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Tested-by: Jin Yang <jin.a.yang@intel.com>
(cherry picked from commit fc0fa16be3)
2011-10-25 18:41:29 -07:00
Ian Romanick
5f776f2a7d mesa: Add GL_OES_compressed_paletted_texture formats to _mesa_is_compressed_format
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Tested-by: Jin Yang <jin.a.yang@intel.com>
(cherry picked from commit a454c835fa)
2011-10-25 18:41:29 -07:00
Chia-I Wu
d2e633a8bd intel: fix GLESv1 support
Add intelInitExtensionsES1 to enable required and optional GLESv1
extensions.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 6b9e4b6ca7)
2011-10-25 18:41:29 -07:00
Chia-I Wu
135501c201 intel: rename intel_extensions_es2.c to intel_extensions_es.c
We'd like to add intelInitExtensionsES1 to it later.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 820789ac69)
2011-10-25 18:41:29 -07:00
Jeremy Huddleston
681f4820ab 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 16:26:17 -07:00
Yuanhan Liu
4b7ad91990 mesa: handle PBO access error in display list mode
Simply generate GL_INVALID_OPERATION error at display list mode. As
explained by Brian, we are going to access PBO data at compile time.
No need to defer the error at execution time.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 46d5fb576a)
2011-10-24 15:54:31 -07:00
Yuanhan Liu
cd114ba503 mesa: handle the pbo case for save_Bitmap
Wrap _mesa_unpack_bitmap to handle the case that data is stored in pixel
buffer object.

This would make calling Bitmap with data stored in PBO by display list work.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 02b801c1ed)
2011-10-24 15:54:31 -07:00
Yuanhan Liu
b8cc97ed26 mesa: fix inverted pbo test error at _mesa_GetnCompressedTexImageARB
It seems like a typo.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 403cf7c56f)
2011-10-24 15:54:31 -07:00
Yuanhan Liu
53a3e743ae mesa: generate error if pbo offset is not aligned with the size of specified type
v2: quote the spec; explicitly exclude the GL_BITMAP case to make code
    more readable. (comments from Ian)

v3: Cast the offset by GLintptr to remove the compile warning(comments
    from Brian).

    I also found that I should use _mesa_sizeof_packed_type() instead,
    as it includes packed pixel type, like GL_UNSIGNED_SHORT_5_6_5.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 9024d8af0a)
2011-10-24 15:54:31 -07:00
Yuanhan Liu
a31eec0d10 i965: setup address rounding enable bits
The patch(based on the reading of the emulator) came from while I was
trying to fix the oglc pbo texImage.1PBODefaults fail. This case
generates a texture with the width and height equal to window's width
and height respectively, then try to texture it on the whole window.
So, it's exactly one texel for one pixel.  And, the min filter and mag
filter are GL_LINEAR. It runs with swrast OK, as expected. But it failed
with i965 driver.

Well, you can't tell the difference from the screen, as the error is
quite tiny. From my digging, it seems that there are some tiny error
happened while getting tex address. This will break the one texel for
one pixel rule in this case. Thus the linear result is taken, with tiny
error.

This patch would fix all oglc pbo subcase fail with the same issue on
both ILK, SNB and IVB.

v2: comments from Ian, make the address_round filed assignment consistent.
    (the sampler is alread memset to 0 by the xxx_update_samper_state
     caller, so need to assign 0 first)

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
(cherry picked from commit 76669381c0)
2011-10-24 15:54:31 -07:00
Yuanhan Liu
6c85918159 mesa: add a function to do the image data copy stuff for save_CompressedTex(Sub)Image
Introuduce a simple function called copy_data to do the image data copy
stuff for all the save_CompressedTex*Image function. The function check
the NULL data case to avoid some potential segfault. This also would
make the code a bit simpler and less redundance.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit e9edcf8b1d)
2011-10-24 15:54:31 -07:00
Yuanhan Liu
eaad245418 intel: fix the wrong code to detect null texture.
There is already comments show how to detect a null texture. Fix the
code to match the comments.

This would fix the oglc divzero(basic.texQOrWEqualsZero) and
divzero(basic.texTrivialPrim) test case fail.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 1a662e7c18)
2011-10-24 15:54:31 -07:00
Yuanhan Liu
0a77734c62 mesa: fix error handling for glMaterial*
Trigger GL_INVALID_ENUM error if the face paramter is not a valid value.

Trigger GL_INVALID_VALUE error if the GL_SHININESS value is out side
[0, ctx->Constant.MaxShiniess].

v2: fix the max shininess value.

v3: suggested by Brian, move the face check into glMaterialfv function
    to reduce code duplicate. Also, refactor the error message.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit a11b4c1e7a)
2011-10-24 15:54:31 -07:00
Yuanhan Liu
8d0d4381da mesa: fix error handling for glMapBufferRange
Accroding the man page, GL_INVALID_VALUE would generated if access has any
bits set other than those valid defined bits.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 099af9e9df)
2011-10-24 15:54:31 -07:00
Brian Paul
a0eefc9bd0 mesa: generate GL_INVALID_OPERATION in glIsEnabledIndex() between Begin/End
(cherry picked from commit 386ec5e80e)
2011-10-24 15:54:31 -07:00
Yuanhan Liu
2af0708c85 mesa: fix error handling for glSelectBuffer
According the man page, trigger a GL_INVALID_VALUE if size < 0.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 57b8f13aa4)
2011-10-24 15:54:31 -07:00
Yuanhan Liu
8fb8b8528d mesa: fix error handling for glPixelZoom
According the man page, GL_INVALID_OPERATION should generated if
glPixelZoom is executed between the execution of glBegin and the
corresponding execution of glEnd.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 7a9a8bbabd)
2011-10-24 15:54:30 -07:00
Yuanhan Liu
0ebdfa31bc mesa: fix error handling for glIsEnabled
According the man page, GL_INVALID_OPERATION should be generated if
glIsEnabled is executed betwwen the execution of glBegin and the
correspoding execution of glEnd.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 6a98802243)
2011-10-24 15:54:30 -07:00
Yuanhan Liu
ab68b00453 mesa: fix error handling for glTexEnv
Fix error handling while calling glTexEnv with invalid texture
environment parameters.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit b020b111a8)
2011-10-24 15:54:30 -07:00
Yuanhan Liu
a895cce693 mesa: fix error handling for some glGet* functions
According to the man page, it should trigger a GL_INVALID_OPERATION
while calling some glGet* functions inside glBegin and glEnd.

This patch dose handle the following functions:
 glGetBooleanv
 glGetFloatv
 glGetIntegerv
 glGetInteger64v
 glGetDoublev

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit f1ddde5c16)
2011-10-24 15:54:30 -07:00
Yuanhan Liu
23a3753a28 mesa: fix error handling for glEvalMesh1/2D
According man page, trigger error when calling glEvalMesh1/2D inside
glBegin/glEnd.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 21b2895bd0)
2011-10-24 15:54:30 -07:00
Brian Paul
b53035e10d mesa: fix error handling for dlist image unpacking
When compiling glDrawPixels, glTexImage(), etc. and we're copying
the user's image we need to be careful about GL error checking.
Previously, we were incorrectly generating GL_OUT_OF_MEMORY in
unpack_image() if width <= 0 or height <= 0 or for invalid format/type
values.  We now check those arguments in unpack_image() and return NULL
if there's a bad value.  The command will get compiled with the
arguments as-is and image=NULL.  Later, when the command is executed the
correct errors will be generated.

This issue was reported by Yuanhan Liu <yuanhan.liu@linux.intel.com>

Reviewed-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
(cherry picked from commit 6fd6efa7bf)
2011-10-24 15:54:30 -07:00
Adam Jackson
b2fbf8225b drisw: Remove cargo culting that breaks GLX 1.3 ctors
Signed-off-by: Adam Jackson <ajax@redhat.com>
(cherry picked from commit d44f821213)
2011-10-24 13:56:46 -04:00
Jeremy Huddleston
bf7b347c10 apple: Use the correct (OpenGL.framework) glViewport and glScissor during init
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 9f2abbee62)
2011-10-21 00:35:34 -07:00
Jeremy Huddleston
7e90db0ddc apple: Silence some debug spew
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 098ecfad83)
2011-10-21 00:35:29 -07:00
Marek Olšák
521819f29c r300g: don't return NULL in resource_from_handle if the resource is too small
The DDX may allocate a buffer with a too small size.
Instead of failing, let's pretend everything's alright.

Such bugs should be fixed in the DDX, of course.

NOTE: This is a candidate for the stable branches.
(cherry picked from commit a04f8c3612)

Conflicts:

	src/gallium/drivers/r300/r300_texture.c
	src/gallium/drivers/r300/r300_texture.h
	src/gallium/drivers/r300/r300_texture_desc.c
2011-10-21 00:31:06 +02:00
Marek Olšák
8b4315cb47 pb_bufmgr_cache: flush cache when create_buffer fails and try again
NOTE: This is a candidate for the stable branches.
(cherry picked from commit 39d7de69b1)
2011-10-20 23:47:51 +02:00
Neil Roberts
986319cd20 meta: Fix saving the active program
When saving the active program in _mesa_meta_begin, it was actually
saving the fragment program instead. This means that if the
application binds a program that only has a vertex shader then when
the meta saved state is restored it will forget the bound program.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=41969
Reviewed-by: Chad Versace <chad@chad-versace.us>
(cherry picked from commit 5625f78cd7)
2011-10-18 11:15:05 -07:00
Yuanhan Liu
c3fd76ce09 i965: fix the constant interp bitmask for flat mode
Fix the constant interpolation enable bit mask for flat light mode.
FRAG_BIT_COL0 attribute bit might be 0, in which case we need to
shift one more bit right.

This would fix the oglc specularColor test fail on both Sandybridge and
Ivybridge.

v2: move the constant interp bitmask setup code into for(; attr <
FRAG_ATTRIB_MAX; attr++) loop suggested by Eric.

Also fixes the Civilization 4 intro videos.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit cd6b8421ca)
2011-10-17 14:22:03 -07:00
Marcin Slusarz
08fa61dab6 nouveau: fix fence hang
If there is not enough space in pushbuffer for fence emission
(nouveau_fence_emit -> nv50_screen_fence_emit -> MARK_RING),
the pushbuffer is flushed, which through flush_notify ->
nv50_default_flush_notify -> nouveau_fence_update marks currently
emitting fence as flushed. But actual emission is done after this mark.
So later when there is a need to wait on this fence and pushbuffer
was not flushed in between, fence wait will never finish causing
application to hang.

To fix this, introduce new fence state between AVAILABLE and EMITTED,
set it before emission and handle it everywhere.

Additionally obtain fence sequence numbers after possible flush in
MARK_RING, because we want to emit fences in correct order.

Reviewed-by: Christoph Bumiller <e0425955@student.tuwien.ac.at>

(commit 9849f366cb in master)
2011-10-17 23:01:41 +02:00
Marcin Slusarz
8d1f1eae93 nouveau: fix crash during fence emission
Fence emission can flush the push buffer, which through flush_notify
unreferences recently emitted fence. If ref count is increased after
fence emission, unreference deletes the fence, which causes SIGSEGV.

Backtrace:
nouveau_fence_del
nouveau_fence_ref
nouveau_fence_next
nouveau_pushbuf_flush
MARK_RING
nv50_screen_fence_emit
nouveau_fence_emit
nv50_flush

This bug manifested as an assertion failure in nouveau_fence.c, because
SIGSEGV handler tried to shutdown the application and used messed up
fence.

This issue was reported by Maxim Levitsky.

(commit e1e03ce492 in master)
2011-10-17 23:00:39 +02:00
Marek Olšák
b9cc9166cf Revert "r300g: fix rendering with a non-zero index bias in draw_elements_immediate"
This reverts commit b9c7773e0d.

It breaks more things than it fixes.
2011-10-16 03:21:02 +02:00
Kenneth Graunke
07210d5c77 intel: Depth format fixes
This is a squash of:

    intel: Recognize all depth formats in get_teximage_readbuffer.

    The existing code was missing GL_DEPTH_COMPONENT32, resulting in it
    wrongly returning the color buffer instead of the depth buffer.

    Fixes an issue in PlaneShift 0.5.7 when casting spells.  The game calls
    CopyTexSubImage2D on buffers with a GL_DEPTH_COMPONENT32 internal
    format, which (prior to this patch) resulted in an attempt to copy
    ARGB8888 to X8_Z24.

    Instead of adding the missing enumeration directly, convert the code to
    use _mesa_is_depth_format() and _mesa_is_depthstencil_format() as these
    should catch any newly added depth formats in the future.

    Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
    Reviewed-by: Eric Anholt <eric@anholt.net>
    (cherry-picked from commit 440224ab73)

And:

    i915: Fix depth texturing since 86e62b2357

    The 965 driver already had the X8_Z24 case, but 915 was missing it.

    Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
    (cherry picked from commit 6aae729d6e)
2011-10-14 17:28:45 -07:00
Eric Anholt
57a6e6092f intel: Mark MESA_FORMAT_X8_Z24 as always supported.
This prevents developer surprise at seeing a GL_DEPTH_COMPONENT
texture have stencil bits, and avoids the metaops path accidentally
copying stencil bits around in glCopyTexImage(GL_DEPTH_COMPONENT) (and
being broken because swrast's glReadPixels(GL_UNSIGNED_INT_24_8) is
broken).

Acked-by: Chad Versace <chad@chad-versace.us>

(cherry-picked from commit 86e62b2357)
2011-10-14 17:28:45 -07:00
Chris Wilson
5b09cf5c57 i915: out-of-bounds write in calc_live_regs()
From a Coverity defect report.

src/mesa/drivers/dri/i915/i915_fragprog.c
   301  /*
   302   * TODO: consider moving this into core
   303   */
   304  static bool calc_live_regs( struct i915_fragment_program *p )
   305  {
   306      const struct gl_fragment_program *program = &p->FragProg;
   307      GLuint regsUsed = 0xffff0000;
-> 308      uint8_t live_components[16] = { 0, };
   309      GLint i;
   310
   311      for (i = program->Base.NumInstructions - 1; i >= 0; i--) {
   312          struct prog_instruction *inst =
&program->Base.Instructions[i];
   313          int opArgs = _mesa_num_inst_src_regs(inst->Opcode);
   314          int a;
   315
   316          /* Register is written to: unmark as live for this and
preceeding ops */
   317          if (inst->DstReg.File == PROGRAM_TEMPORARY) {
-> 318              if (inst->DstReg.Index > 16)
   319                 return false;
   320
-> 321              live_components[inst->DstReg.Index] &= ~inst->DstReg.WriteMask;
   322              if (live_components[inst->DstReg.Index] == 0)
   323                  regsUsed &= ~(1 << inst->DstReg.Index);
   324          }
   325
   326          for (a = 0; a < opArgs; a++) {
   327              /* Register is read from: mark as live for this and preceeding ops */
   328              if (inst->SrcReg[a].File == PROGRAM_TEMPORARY) {
   329                  unsigned c;
   330
   331                  if (inst->SrcReg[a].Index > 16)
   332                     return false;
   333
   334                  regsUsed |= 1 << inst->SrcReg[a].Index;
   335
   336                  for (c = 0; c < 4; c++) {
   337                      const unsigned field = GET_SWZ(inst->SrcReg[a].Swizzle, c);
   338
   339                      if (field <= SWIZZLE_W)
   340                          live_components[inst->SrcReg[a].Index] |= (1U << field);
   341                  }
   342              }
   343          }
   344
   345          p->usedRegs[i] = regsUsed;
   346      }

Reported-by: Vinson Lee <vlee@vmware.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=40022
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
(cherry picked from commit 67582e6eef)
2011-10-14 17:28:45 -07:00
Carl Worth
eb0fd67f6a glcpp: Add a test for #elif with an undefined macro.
As written, this test correctly raises an error for #elif being used
with an undefined macro (and not as an argument to "defined"). If the
preceding #if were '#if 1' then this diagnositc would correctly be
hidden. That allows code such as the following to not raise an error:

	#ifndef MAYBE_UNDEFINED
	#elif MAYBE_UNDEFINED < 5
	...
	#endif

So this test case is working as expected already. We add it here just
to improve test coverage.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Carl Worth <cworth@cworth.org>
(cherry picked from commit 201485bae0)
2011-10-14 17:28:45 -07:00
Carl Worth
c6dfde2136 glcpp: Raise error if defining any macro containing two consecutive underscores
The specification reserves any macro name containing two consecutive
underscores, (anywhere within the name). Previously, we only raised
this error for macro names that started with two underscores.

Fix the implementation to check for two underscores anywhere, and also
update the corresponding 086-reserved-macro-names test.

This also fixes the following two piglit tests:

	spec/glsl-1.30/preprocessor/reserved/double-underscore-02.frag
	spec/glsl-1.30/preprocessor/reserved/double-underscore-03.frag

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Carl Worth <cworth@cworth.org>
(cherry picked from commit c4aaf7943c)
2011-10-14 17:28:45 -07:00
Carl Worth
71bd5d424c glcpp: Implement token pasting for non-function-like macros
This is as simple as abstracting one existing block of code into a
function call and then adding a single call to that function for the
case of a non-function-like macro.

This fixes the recently-added 097-paste-with-non-function-macro test
as well as the following piglit tests:

	spec/glsl-1.30/preprocessor/concat/concat-01.frag
	spec/glsl-1.30/preprocessor/concat/concat-02.frag

Also, the concat-04.frag test now passes for the right reason. The
test is intended to fail the compilation, but before this commit it
was failing compilation (and hence passing the test) for the wrong
reason.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Carl Worth <cworth@cworth.org>
(cherry picked from commit 28842c2331)
2011-10-14 17:28:45 -07:00
Carl Worth
33e1019d95 glcpp: Test a non-function-like macro using the token paste operator
Apparently we never implemented this, (but we've got a GLSL 1.30 test
in piglit that is exercising this case).

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Carl Worth <cworth@cworth.org>
(cherry picked from commit 7bb3403e01)
2011-10-14 17:28:45 -07:00
Carl Worth
ab94d6f902 glcpp: Fix two (or more) successive applications of token pasting
There was already a loop here to look for multiple token pastes, but
it was mistakenly incrementing the iterator counter after performing
one paste.

Instead, leave the loop iterator in place to coalesce as many tokens
as necessary into one.

This fixes the recently add 096-paste-twice test as well as the
following piglit test:

	spec/glsl-1.30/preprocessor/concat/concat-03.frag

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Carl Worth <cworth@cworth.org>
(cherry picked from commit 3c01a58944)
2011-10-14 17:28:45 -07:00
Ian Romanick
5becf89a17 i915: Only emit program errors when INTEL_DEBUG=wm or INTEL_DEBUG=fallbacks
This makes piglit a lot more happy.  The errors are logged when
INTEL_DEBUG=fallbacks because the application is about to hit a big
software fallback.  We frequently ask people to run applications that
are hitting software fallbacks with INTEL_DEBUG=fallbacks so the we
can help them debug the reason for the software fallback.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 0290a018a5)
2011-10-14 17:28:45 -07:00
Ian Romanick
13a2d4a985 i915: Fail without crashing if a Mesa IR program uses too many registers
This can only happen in GLSL shaders because assembly shaders that use
too many temps are rejected by core Mesa.  It is easiest to make this
happen with shaders that contain flow-control that could not be lowered.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 3bb2f0dde1)
2011-10-14 17:28:45 -07:00
Ian Romanick
f2d166583d ir_to_mesa: Emit warnings instead of errors for IR that can't be lowered
Rely on the driver to do the right thing.  This probably means falling
back to software.  Page 88 of the OpenGL 2.1 spec specifically says:

    "A shader should not fail to compile, and a program object should
    not fail to link due to lack of instruction space or lack of
    temporary variables. Implementations should ensure that all valid
    shaders and program objects may be successfully compiled, linked
    and executed."

There is no provision for saying "No" to a valid shader that is
difficult for the hardware to handle, so stop doing that.

On i915 this causes a large number of piglit tests to change from FAIL
to WARN.  The warning is because the driver still emits messages to
stderr like "i915_program_error: Unsupported opcode: BGNLOOP".

It also fixes ES2 conformance CorrectFull_frag and CorrectParse1_frag
on i915 (and probably other hardware that can't handle loops).

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 322c3bf9dc)
2011-10-14 17:28:42 -07:00
Ian Romanick
49d2c552a5 ir_to_mesa: Use Add linker_error instead of fail_link
The functions were almost identical.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 8aadd89d07)
2011-10-14 17:28:03 -07:00
Ian Romanick
a864b15b83 mesa: Ensure that gl_shader_program::InfoLog is never NULL
This prevents assertion failures in ralloc_strcat.  The ralloc_free in
_mesa_free_shader_program_data can be omitted because freeing the
gl_shader_program in _mesa_delete_shader_program will take care of
this automatically.

A bunch of this code could use a refactor to use ralloc a bit more
effectively.  A bunch of the things that are allocated with malloc and
owned by the gl_shader_program should be allocated with ralloc (using
the gl_shader_program as the context).

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 89193933cb)
2011-10-14 17:27:17 -07:00
Ian Romanick
e458c3ddbb linker: Make linker_{error,warning} generally available
linker_warning is a new function.  It's identical to linker_error
except that it doesn't set LinkStatus=false and it prepends "warning: "
on messages instead of "error: ".

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 379a32f42e)
2011-10-14 17:27:17 -07:00
Ian Romanick
b8a46f910d linker: Make linker_error set LinkStatus to false
Remove the other places that set LinkStatus to false since they all
immediately follow a call to linker_error.  The function linker_error
was previously known as linker_error_printf.  The name was changed
because it may seem surprising that a printf function will set an
error flag.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 586e741ac1)
2011-10-14 17:27:17 -07:00
Kenneth Graunke
eca2a91a9b i965: Fix inconsistent indentation in brw_eu_emit.c.
Most of these functions used three spaces for the first level of
indentation, but four spaces for the next level.  One used tabs and then
three spaces.  Some used 3/4 in a then block but 3/3 in the else block.

Normally I try to avoid field days like this, but since the functions
were so inconsistent, even internally, it was making it difficult to
edit without introducing spurious whitespace changes.

So, just get it over with.  git diff -b shows 0 lines changed.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit b861479f83)
2011-10-13 16:35:40 -07:00
Kenneth Graunke
1f083e1839 i965: Allow SIMD16 color writes on Ivybridge.
Again, the check was needlessly specific: this works fine on Gen7.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 7db874bf4c4273d2d46218b1490d312fe2654284)
2011-10-13 14:06:11 -07:00
Kenneth Graunke
9bbf2a343f i965/fs: Allow SIMD16 with control flow on Ivybridge.
The check was designed to forbid it on old generations (Gen5/Ironlake),
not on new ones.  It just works on Gen7/Ivybridge.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit ae5da817e2aeb9f9447fdd6d2eb4b22d6f8f6a87)
2011-10-13 14:06:11 -07:00
Kenneth Graunke
38dfedccb2 i965: Emit depth stalls and flushes before changing depth state on Gen6+.
Fixes OpenArena on Gen7.  Technically, adding only the first depth stall
fixes it, but the documentation says to do all three, and the Windows
driver seems to do it.

Not observed to fix anything on Gen6 yet.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=38863
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 02c4dc807e91640c69c8addc3c797300a3c536ad)
2011-10-13 14:06:11 -07:00
Kenneth Graunke
1e0e116d6d i965: Fix incorrect maximum PS thread count shift on Ivybridge.
At one point, the documentation said that max thread count in 3DSTATE_PS
was at bit offset 23, but it's actually 24 on Ivybridge.  Not only did
this halve our thread count, it caused us to write 1 into a bit 23, which
is marked as MBZ (must be zero).  Furthermore, it made us write an even
number into this field, which is apparently not allowed.  Apparently we
were just lucky it worked.

NOTE: This is a candidate for the 7.11 branch.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 556e7eea80de778b44a37d51cb757ce32221d1e3)
2011-10-13 14:06:11 -07:00
Eric Anholt
aaadd4c111 i965: Fix polygon stipple offset state flagging.
_NEW_WINDOW_POS wasn't a real Mesa state flag, but we were missing
_NEW_BUFFERS to update the stipple offset when FBO binding or window
size changed, and _NEW_POLYGON to update when stippling gets enabled.

Fixes oglconform's tristrip test.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
(cherry picked from commit d598851d401f7f34d623c9cfbd85d7f5faccd7c2)
2011-10-13 13:59:06 -07:00
Eric Anholt
0d31b130bb i965: Add missing _NEW_POLYGON flag to polygon stipple upload.
Because we skip the pattern upload when stippling is disabled, we need
to check again when it might have been turned on.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
(cherry picked from commit e19541aa2ad05f687c859001b62713209787c9c8)
2011-10-13 13:58:57 -07:00
Kenneth Graunke
e4b1dce9ec i965: Use proper texture alignment units for cubemaps on Gen5+.
In particular, S3TC compressed textures need align_h == 4.

Fixes skybox errors in Quake 4 and FEAR.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=34628
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit e0e688ca5441e2c8bc59ec7488bc1bc4ba196602)
2011-10-13 13:58:51 -07:00
Kenneth Graunke
0f87fe948a i965/gen5+: Fix incorrect miptree layout for non-power-of-two cubemaps.
For power-of-two sizes, h0 == mt->height0 since it's already a multiple
of two.  However, for NPOT, they're different; h1 should be computed
based on the original size.

Fixes piglit test "cubemap npot" and oglconform test "textureNPOT".

NOTE: This is a candidate for stable release branches.

Reviewed-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit bebc19448f45dbe8c3b016d440403f52e1036e15)
2011-10-13 13:58:45 -07:00
Eric Anholt
f484fc7476 i965/fs: Respect ARB_color_buffer_float clamping.
This was done in the old codegen path, but not the new one.  Caught by
piglit fbo tests after the conversion to GLSL ff_fragment_shader.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit da53ca641106e47f1d74386d8dc0f7eebeec5225)
2011-10-13 13:58:37 -07:00
Marek Olšák
b9c7773e0d r300g: fix rendering with a non-zero index bias in draw_elements_immediate
NOTE: This is a candidate for the stable branches.
(cherry picked from commit 5506f6ef96)
2011-10-04 17:48:33 +02:00
Paul Berry
7d2ff4ae77 glsl: improve the accuracy of the asin() builtin function.
The previous formula for asin(x) was algebraically equivalent to:

sign(x)*(pi/2 - sqrt(1-|x|)*(A + B|x| + C|x|^2))

where A, B, and C were arbitrary constants determined by a curve fit.

This formula had a worst case absolute error of 0.00448, an unbounded
worst case relative error, and a discontinuity near x=0.

Changed the formula to:

sign(x)*(pi/2 - sqrt(1-|x|)*(pi/2 + (pi/4-1)|x| + A|x|^2 + B|x|^3))

where A and B are arbitrary constants determined by a curve fit.  This
has a worst case absolute error of 0.00039, a worst case relative
error of 0.000405, and no discontinuities.

I don't expect a significant performance degradation, since the extra
multiply-accumulate should be fast compared to the sqrt() computation.

Fixes piglit tests {vs,fs}-asin-float and {vs,fs}-atan-*
(cherry picked from commit d4c80f5f85)
2011-10-02 21:39:30 +02:00
Paul Berry
1bbf124ff8 glsl hierarchical visitor: Do not overwrite base_ir for parameter lists.
This patch fixes a bug in ir_hirearchical_visitor: when traversing an
exec_list representing the formal or actual parameters of a function,
it modified base_ir to point to each parameter in turn, rather than
leaving it as a pointer to the enclosing statement.  This was a
problem, since base_ir is used by visitor classes to locate the
statement containing the node being visited (usually so that
additional statements can be inserted before or after it).  Without
this fix, visitors might attempt to insert statements into parameter
lists.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit cc81eb09b9)
2011-10-02 19:57:57 +02:00
Eric Anholt
ca7560765c glsl: When assiging from a whole array, mark it as used.
Fixes piglit link-uniform-array-size.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 407a1001ae)
2011-10-02 19:57:57 +02:00
Eric Anholt
878d701da4 glsl: When assigning to a whole array, mark the array as accessed.
The vs-varying-array-mat2-col-row-wr test writes a mat2[3] constant to
a mat2[3] varying out array, and also statically accesses element 1 of
it on the VS and FS sides.  At link time it would get trimmed down to
just 2 elements, and then codegen of the VS would end up generating
assignments to the unallocated last entry of the array.  On the new
i965 VS backend, that happened to land on the vertex position.

Some issues remain in this test on softpipe, i965/old-vs and
i965/new-vs on visual inspection, but i965 is passing because only one
green pixel is probed, not the whole split green/red quad.
2011-10-02 19:57:56 +02:00
Paul Berry
c19b963ad6 glsl: Remove field array_lvalue from ir_variable.
The array_lvalue field was attempting to enforce the restriction that
whole arrays can't be used on the left-hand side of an assignment in
GLSL 1.10 or GLSL ES, and can't be used as out or inout parameters in
GLSL 1.10.

However, it was buggy (it didn't work properly for built-in arrays),
and it was clumsy (it unnecessarily kept track on a
variable-by-variable basis, and it didn't cover the GLSL ES case).

This patch removes the array_lvalue field completely in favor of
explicit checks in ast_parameter_declarator::hir() (this check is
added) and in do_assignment (this check was already present).

This causes a benign behavioral change: when the user attempts to pass
an array as an out or inout parameter of a function in GLSL 1.10, the
error is now flagged at the time the function definition is
encountered, rather than at the time of invocation.  Previously we
allowed such functions to be defined, and only flagged the error if
they were invoked.

Fixes Piglit tests
spec/glsl-1.10/compiler/qualifiers/fn-{out,inout}-array-prohibited*
and
spec/glsl-1.20/compiler/assignment-operators/assign-builtin-array-allowed.vert.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 00792e3586)
2011-10-02 19:57:08 +02:00
Eric Anholt
95185c7fe2 glsl: Clarify error message about whole-array assignment in GLSL 1.10.
Previously, it would produce:

    Failed to compile FS: 0:6(7): error: non-lvalue in assignment

and now it produces:

    Failed to compile FS: 0:5(7): error: whole array assignment is not
    allowed in GLSL 1.10 or GLSL ES 1.00.

Also, add spec quotation to the two places we have code for array
lvalues in GLSL 1.10.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 525cec98a5)
2011-10-02 19:56:53 +02:00
Paul Berry
e1221a8811 glsl: Rework oversize array check for gl_TexCoord.
The check now applies both when explicitly declaring the size of
gl_TexCoord and when implicitly setting the size of gl_TexCoord by
accessing it using integral constant expressions.

This is prep work for adding similar size checks to gl_ClipDistance.

Fixes piglit tests texcoord/implicit-access-max.{frag,vert}.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 93b9758d01)
2011-10-02 19:21:49 +02:00
Paul Berry
f732b5a999 glsl: Fix type error when lowering integer divisions
This patch fixes a bug when lowering an integer division:

  x/y

to a multiplication by a reciprocal:

  int(float(x)*reciprocal(float(y)))

If x was a plain int and y was an ivecN, the lowering pass
incorrectly assigned the type of the product to be float, when in fact
it should be vecN.  This caused mesa to abort with an IR validation
error.

Fixes piglit tests {fs,vs}-op-div-int-ivec{2,3,4}.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit af501e2b29)
2011-10-02 19:19:49 +02:00
Paul Berry
0129d5297b glsl: Perform implicit type conversions on function call out parameters.
When an out parameter undergoes an implicit type conversion, we need
to store it in a temporary, and then after the call completes, convert
the resulting value.  In other words, we convert code like the
following:

void f(out int x);
float value;
f(value);

Into IR that's equivalent to this:

void f(out int x);
float value;
int out_parameter_conversion;
f(out_parameter_conversion);
value = float(out_parameter_conversion);

This transformation needs to happen during ast-to-IR convertion (as
opposed to, say, a lowering pass), because it is invalid IR for formal
and actual parameters to have types that don't match.

Fixes piglit tests
spec/glsl-1.20/compiler/qualifiers/out-conversion-int-to-float.vert and
spec/glsl-1.20/execution/qualifiers/vs-out-conversion-*.shader_test,
and bug 39651.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=39651

Reviewed-by: Chad Versace <chad@chad-versace.us>
(cherry picked from commit 67b5a3267d)
2011-10-02 19:19:08 +02:00
Paul Berry
27f00df2b7 glsl: Check array size is const before asserting that no IR was generated.
process_array_type() contains an assertion to verify that no IR
instructions are generated while processing the expression that
specifies the size of the array.  This assertion needs to happen
_after_ checking whether the expression is constant.  Otherwise we may
crash on an illegal shader rather than reporting an error.

Fixes piglit tests array-size-non-builtin-function.vert and
array-size-with-side-effect.vert.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit d4144a123b)
2011-10-02 19:17:42 +02:00
Paul Berry
8dcfe15a9a glsl: Constant-fold built-in functions before outputting IR
Rearranged the logic for converting the ast for a function call to
hir, so that we constant fold before emitting any IR.  Previously we
would emit some IR, and then only later detect whether we could
constant fold.  The unnecessary IR would usually get cleaned up by a
later optimization step, however in the case of a builtin function
being used to compute an array size, it was causing an assertion.

Fixes Piglit test array-size-constant-relational.vert.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=38625
(cherry picked from commit 789ee6516b)
2011-10-02 19:16:47 +02:00
Paul Berry
2c0e00de23 glsl: Emit function signatures at toplevel, even for built-ins.
The ast-to-hir conversion needs to emit function signatures in two
circumstances: when a function declaration (or definition) is
encountered, and when a built-in function is encountered.

To avoid emitting a function signature in an illegal place (such as
inside a function), emit_function() checked whether we were inside a
function definition, and if so, emitted the signature before the
function definition.

However, this didn't cover the case of emitting function signatures
for built-in functions when those built-in functions are called from
inside the constant integer expression that specifies the length of a
global array.  This failed because when processing an array length, we
are emitting IR into a dummy exec_list (see process_array_type() in
ast_to_hir.cpp).  process_array_type() later checks (via an assertion)
that no instructions were emitted to the dummy exec_list, based on the
reasonable assumption that we shouldn't need to emit instructions to
calculate the value of a constant.

This patch changes emit_function() so that it emits function
signatures at toplevel in all cases.

This partially fixes bug 38625
(https://bugs.freedesktop.org/show_bug.cgi?id=38625).  The remainder
of the fix is in the patch that follows.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 0d81b0e184)
2011-10-02 19:03:22 +02:00
Paul Berry
1895de7a32 Revert "glsl: Skip processing the first function's body in do_dead_functions()."
opt_dead_functions contained a shortcut to skip processing the first
function's body, based on the assumption that IR functions are
topologically sorted, with callees always coming before their callers
(therefore the first function cannot contain any calls).

This assumption turns out not to be true in general.  For example, the
following code snippet gets translated to IR that violates this
assumption:

    void f();
    void g();
    void f() { g(); }
    void g() { ... }

In practice, the shortcut didn't cause bugs because of a coincidence
of the circumstances in which opt_dead_functions is called:

(a) we do inlining right before dead function elimination, and
    inlining (when successful) eliminates all calls.

(b) for user-defined functions, inlining is always successful, because
    previous optimization passes (during compilation) have reduced
    them to a form that is eligible for inlining.

(c) the function that appears first in the IR can't possibly call a
    built-in function, because built-in functions are always emitted
    before the function that calls them.

It seems unnecessarily fragile to have opt_dead_functions depend on
these coincidences.  And the next patch in this series will break (c).
So I'm reverting the shortcut.  The consequence will be a slight
increase in link time for complex shaders.

This reverts commit c75427f4c8.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 482338842d)
2011-10-02 19:02:30 +02:00
Paul Berry
7dc636dd77 glsl: improve the accuracy of the atan(x,y) builtin function.
The previous formula for atan(x,y) returned a value of +/- pi whenever
|x|<0.0001, and used a formula based on atan(y/x) otherwise.  This
broke in cases where both x and y were small (e.g. atan(1e-5, 1e-5)).

This patch modifies the formula so that it returns a value of +/- pi
whenever |x|<1e-8*|y|, and uses the formula based on atan(y/x)
otherwise.
(cherry picked from commit b1b4ea0b36)
2011-10-02 19:00:31 +02:00
Paul Berry
e42b822fec glsl: improve the accuracy of the radians() builtin function
The constant used in the radians() function didn't have enough
precision, causing a relative error of 1.676e-5, which is far worse
than the precision of 32-bit floats.  This patch reduces the relative
error to 1.14e-9, which is the best we can do in 32 bits.

Fixes piglit tests {fs,vs}-radians-{float,vec2,vec3,vec4}.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit fe33c886a7)
2011-10-02 18:50:28 +02:00
Paul Berry
0501cee136 glsl: Lower break instructions when necessary at the end of a loop.
Normally lower_jumps.cpp doesn't need to lower a break instruction
that occurs at the end of a loop, because all back-ends can produce
proper GPU instructions for a break instruction in this "canonical"
location.  However, if other break instructions within the loop are
already being lowered, then a break instruction at the end of the loop
needs to be lowered too, since after the optimization is complete a
new conditional break will be inserted at the end of the loop.

Without this patch, lower_jumps.cpp may require multiple passes in
order to lower all jumps.  This results in sub-optimal output because
lower_jumps.cpp produces a brand new set of temporary variables each
time it is run, and the redundant temporary variables are not
guaranteed to be eliminated by later optimization passes.

Fixes unit test test_lower_breaks_6.
(cherry picked from commit 067c9d7bd7)

Conflicts:

	src/glsl/lower_jumps.cpp
2011-10-02 18:48:27 +02:00
Paul Berry
38ae26b709 glsl: In lower_jumps.cpp, lower both branches of a conditional.
Previously, lower_jumps.cpp would break out of its loop after lowering
a jump instruction in just the then- or else-branch of a conditional,
and it would fail to lower a jump instruction occurring in the other
branch.

Without this patch, lower_jumps.cpp may require multiple passes in
order to lower all jumps.  This results in sub-optimal output because
lower_jumps.cpp produces a brand new set of temporary variables each
time it is run, and the redundant temporary variables are not
guaranteed to be eliminated by later optimization passes.

Fixes unit test test_lower_returns_4.
(cherry picked from commit e71b4ab8a6)
2011-10-02 18:45:10 +02:00
Paul Berry
de798938d4 glsl: Use foreach_list in lower_jumps.cpp
The visitor class in lower_jumps.cpp never removes or replaces the
instruction being visited, but it frequently alters or removes the
instructions that follow it.  Therefore, to make sure the altered IR
is visited, it needs to iterate through exec_lists using foreach_list
rather than visit_exec_list().

Without this patch, lower_jumps.cpp may require multiple passes in
order to lower all jumps.  This results in sub-optimal output because
lower_jumps.cpp produces a brand new set of temporary variables each
time it is run, and the redundant temporary variables are not
guaranteed to be eliminated by later optimization passes.

Also, certain invariants assumed by lower_jumps.cpp may fail to hold,
causing assertion failures.

Fixes unit tests test_lower_pulled_out_jump,
test_lower_unified_returns, test_lower_guarded_conditional_break,
test_lower_return_non_void_at_end_of_loop, and test_lower_returns_3.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 382cee91a4)
2011-10-02 18:44:50 +02:00
Paul Berry
acd2a03ffb glsl: lower unconditional returns and continues in loops.
Previously, lower_jumps.cpp would only lower return and continue
statements that appeared inside conditionals.  This patch makes it
lower unconditional returns and continue statements that occur inside
a loop.

Such unconditional flow control statements would be unlikely to be
explicitly coded by a reasonable user, however they might arise as a
result of other optimizations.

Without this patch, lower_jumps.cpp might not lower certain return and
continue statements, causing some backends to fail.

Fixes unit tests test_lower_return_void_at_end_of_loop and
test_remove_continue_at_end_of_loop.
(cherry picked from commit 03145ba655)

Conflicts:

	src/glsl/lower_jumps.cpp
2011-10-02 18:44:31 +02:00
Paul Berry
d1786cea1c glsl: Refactor logic for determining whether to lower return statements.
Previously, do_lower_jumps.cpp determined whether to lower return
statements in ir_lower_jumps_visitor::should_lower_jumps().  Moved
this logic to ir_lower_jumps_visitor::visit(ir_function_signature *),
so that it can be used in determining whether to lower a return
statement at the end of a function.
(cherry picked from commit dbaa2e627e)
2011-10-02 18:43:00 +02:00
Paul Berry
934c7a0661 glsl: Lower unconditional return statements.
Previously, lower_jumps.cpp only lowered return statements that
appeared inside of an if statement.

Without this patch, lower_jumps.cpp might not lower certain return
statements, causing some back-ends to fail (as in bug #36669).

Fixes unit test test_lower_returns_1.
(cherry picked from commit afc9a50fba)
2011-10-02 18:35:00 +02:00
Brian Paul
2ba0d0a5e8 mesa: add _NEW_CURRENT_ATTRIB in _mesa_program_state_flags()
If color material mode is enabled, constant buffer entries related
to the material coefficients will depend on glColor.  So add
_NEW_CURRENT_ATTRIB to the bitset returned for material-related
constants in _mesa_program_state_flags().

This fixes a bug exercised by the new piglit draw-arrays-colormaterial
test.

Note: This is a candidate for the 7.11 branch.
(cherry picked from commit 57169c4694)
2011-10-02 18:10:40 +02:00
Marek Olšák
1cf8f9599c r600g: add index_bias to index buffer bounds
This fixes ARB_draw_elements_base_vertex with max_index != ~0.

NOTE: This is a candidate for the 7.11 branch.
(cherry picked from commit 44afac04ea)
2011-10-02 18:10:19 +02:00
Brian Paul
2781baaa64 meta: fix broken sRGB mipmap generation
If we're generating a mipmap for an sRGB texture we need to bypass
sRGB->linear conversion.  Otherwise the destination mipmap level
(drawn with a textured quad) will have the wrong colors.
If we can't turn of sRGB->linear conversion (GL_EXT_texture_sRGB_decode)
we need to use the software fallback for mipmap generation.

Note: This is a candidate for the 7.11 branch.
(cherry picked from commit 1e939f5374)
2011-10-02 18:09:22 +02:00
Brian Paul
a74400ca30 mesa: fix PACK_COLOR_5551(), PACK_COLOR_1555() macros
The 1-bit alpha channel was incorrectly encoded.  Previously, any non-zero
alpha value for the ubyte alpha value would set A=1.  Instead, use the
most significant bit of the ubyte alpha to determine the A bit.  This is
consistent with the other channels and other OpenGL implementations.

Note: This is a candidate for the 7.11 branch.

Reviewed-by: Michel Dänzer <michel@daenzer.net>
(cherry picked from commit 4731a598f0)
2011-10-02 18:08:31 +02:00
Tom Stellard
a5e2074fdd r300/compiler: Fix regalloc for values with multiple writers
https://bugs.freedesktop.org/show_bug.cgi?id=40062
https://bugs.freedesktop.org/show_bug.cgi?id=36939

Note: This is a candidate for the 7.11 branch.
(applied diff manually from 2d1004d9aa)
2011-10-02 18:07:53 +02:00
Brian Paul
fad6e2ea5a meta: fix/add checks for GL_EXT_framebuffer_sRGB
This fixes spurious GL errors when the GL_EXT_framebuffer_sRGB extension
is not supported.

Note: This is a candidate for the 7.11 branch
(cherry picked from commit 6e423253e7)
2011-10-02 18:05:03 +02:00
Vadim Girlin
a73c667069 r600g: fix replace_gpr_with_pv_ps
Instructions with 3 source operands have no write mask, so we may replace their
destinations with PV/PS in the next group even if their dst.write is 0.

Note: This is a candidate for the 7.11 branch.

Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit fdb62ef3f5)
2011-10-02 18:04:38 +02:00
Vadim Girlin
e87f79c8a4 r600g: fix check_and_set_bank_swizzle
Need to do full check when not all bank swizzles in the group are forced
(e.g. when trying to merge interp_* group with the next instruction)

Note: This is a candidate for the 7.11 branch.

Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 6ba68c7654)

Conflicts:

	src/gallium/drivers/r600/r600_asm.c
2011-10-02 18:04:22 +02:00
Chad Versace
fa8cfbfb64 x86-64: Fix compile error with clang
Remove the 'f' suffix from a float literal.
    - .float 0.0f+1.0
    + .float 1.0

This fixes the following compile error with clang:
    error: unexpected token in directive
    .float 0.0f+1.0
              ^

Note: This is a candidate for the stable branches.
Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Chad Versace <chad@chad-versace.us>
(cherry picked from commit 9cd64ec35a)
2011-10-02 18:00:27 +02:00
Brian Paul
446a67b74e swrast: don't try to do depth testing if there's no depth buffer
Fixes piglit hiz-depth-stencil-test-fbo-d0-s8 crash.
See http://bugs.freedesktop.org/show_bug.cgi?id=37907

NOTE: This is a candidate for the 7.11 branch.
(cherry picked from commit 37a64baea8)
2011-10-02 18:00:13 +02:00
Kenneth Graunke
d5c84929a7 mesa: In validate_program(), initialize errMsg for safety.
validate_program relies on validate_shader_program to fill in errMsg;
empirically, there exist cases where that doesn't happen.

While tracking those down may be worthwhile, initializing the string so
we don't try to ralloc_strdup random garbage also seems wise.

Fixes issues caught by valgrind while running some test case.

NOTE: This is a candidate for stable release branches.

Reviewed-by: Chad Versace <chad@chad-versace.us>
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit db726b048e)
2011-10-02 18:00:03 +02:00
Christopher James Halse Rogers
dd9b78e212 glx/dri2: Paper over errors in DRI2Connect when indirect
DRI2 will throw BadRequest for this when the client is not local, but
DRI2 is an implementation detail and not something callers should have
to know about.  Silently swallow errors in this case, and just propagate
the failure through DRI2Connect's return code.

Note: This is a candidate for the stable release branches.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=28125
Signed-off-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com>
(cherry picked from commit fbc2fcf685)
2011-10-02 17:59:54 +02:00
Chia-I Wu
2cadae90c0 glsl: empty declarations should be valid
Unlike C++, empty declarations such as

  float;

should be valid.  The spec is not explicit about this actually.

Some apps that generate their shader sources may rely on this.  This was
noted when porting one of them to Linux from Windows.

Reviewed-by: Chad Versace <chad@chad-versace.us>

Note: this is a candidate for the 7.11 branch.
(cherry picked from commit 547212d963)
2011-10-02 17:59:26 +02:00
Vadim Girlin
ffb0f94136 r600g: take into account force_add_cf in pops
When we have two ENDIFs in a row, we shouldn't modify the pop_count
for the same alu clause twice.

Fixes https://bugs.freedesktop.org/show_bug.cgi?id=38163

Note: this is a candidate for the 7.11 branch.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 2bde0cc95d)
2011-10-02 17:58:50 +02:00
Vadim Girlin
badd2900ea r600g: use backend mask for occlusion queries
Use backend_map kernel query if supported, otherwise analyze ZPASS_DONE
results to get the mask.

Fixes lockups with predicated rendering due to incorrect query buffer
initialization on some cards.

Note: this is a candidate for the 7.11 branch.

Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 6eb94fc344)
2011-10-02 17:58:40 +02:00
Chad Versace
6c032dd837 glsl: Fix conversions in array constructors
Array constructors obey narrower conversion rules than other constructors
[1] --- they use the implicit conversion rules [2] instead of the scalar
constructor conversions [3].  But process_array_constructor() was
incorrectly applying the broader rules.

[1] GLSL 1.50 spec, Section 5.4.4 Array Constructors, page 52 (58 of pdf)
[2] GLSL 1.50 spec, Section 4.1.10 Implicit Conversions, page 25 (31 of pdf)
[3] GLSL 1.50 spec, Section 5.4.1 Conversion, page 48 (54 of pdf)

To fix this, first check (with glsl_type::can_be_implicitly_converted_to)
if an implicit conversion is legal before performing the conversion.

Fixes:
piglit:spec/glsl-1.20/compiler/structure-and-array-operations/array-ctor-implicit-conversion-bool-float.vert
piglit:spec/glsl-1.20/compiler/structure-and-array-operations/array-ctor-implicit-conversion-bvec*-vec*.vert

Note: This is a candidate for the 7.10 and 7.11 branches.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Chad Versace <chad@chad-versace.us>
(cherry picked from commit a5ab9398e3)
2011-10-02 17:58:28 +02:00
Chad Versace
70c5be6c91 glsl: Remove ir_function.cpp:type_compare()
The function is no longer used and has been replaced by
glsl_type::can_implicitly_convert_to().

Note: This is a candidate for the 7.10 and 7.11 branches.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Chad Versace <chad@chad-versace.us>
(cherry picked from commit 6efe1a8495)
2011-10-02 17:00:59 +02:00
Chad Versace
3b92831fab glsl: Fix implicit conversions in non-constructor function calls
Context
-------
In ast_function_expression::hir(), parameter_lists_match() checks if the
function call's actual parameter list matches the signature's parameter
list, where the match may require implicit conversion of some arguments.
To check if an implicit conversion exists between individual arguments,
type_compare() is used.

Problems
--------
type_compare() allowed the following illegal implicit conversions:
    bool -> float
    bvecN -> vecN

    int -> uint
    ivecN -> uvecN

    uint -> int
    uvecN -> ivecN

Change
------
type_compare() is buggy, so replace it with glsl_type::can_be_implicitly_converted_to().
This comprises a rewrite of parameter_lists_match().

Fixes piglit:spec/glsl-1.20/compiler/built-in-functions/outerProduct-bvec*.vert

Note: This is a candidate for the 7.10 and 7.11 branches.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Chad Versace <chad@chad-versace.us>
(cherry picked from commit 8b3627fd7b)
2011-10-02 17:00:45 +02:00
Chad Versace
e6d07585f8 glsl: Add method glsl_type::can_implicitly_convert_to()
This method checks if a source type is identical to or can be implicitly
converted to a target type according to the GLSL 1.20 spec, Section 4.1.10
Implicit Conversions.

The following commits use the method for a bugfix:
    glsl: Fix implicit conversions in non-constructor function calls
    glsl: Fix implicit conversions in array constructors

Note: This is a candidate for the 7.10 and 7.11 branches.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Chad Versace <chad@chad-versace.us>
(cherry picked from commit 200e4972c1)
2011-10-02 16:57:32 +02:00
Brian Paul
4bd0f04531 mesa: add missing breaks for GL_TEXTURE_CUBE_MAP_SEAMLESS queries
And fix indentation.

NOTE: This is a candidate for the 7.11 branch.
(cherry picked from commit dc1f32deae)
2011-10-02 16:53:57 +02:00
Alex Deucher
45716cffbe r600g: fix up vs export handling
Certain attributes (position, psize, etc.) don't
count as params; they are handled separately by the hw.
However, the VS is required to export at least one param
and r600_shader_from_tgsi() takes care of adding a dummy
export if there is none.  Make sure the VS param export
count in the SPI properly accounts for this.

Note: This is a candidate for the 7.11 branch.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit dc1c0ca22a)
2011-10-02 16:53:44 +02:00
Marek Olšák
ae633fa0ef configure.ac: fix xlib-based softpipe build
Tested-by: Jon TURNEY <jon.turney@dronecode.org.uk>

NOTE: This is a candidate for the 7.11 branch.
(cherry picked from commit c6f59fcd00)

Conflicts:

	configure.ac
2011-10-02 16:53:21 +02:00
Kenneth Graunke
25861dc7f3 glsl: Avoid massive ralloc_strndup overhead in S-Expression parsing.
When parsing S-Expressions, we need to store nul-terminated strings for
Symbol nodes.  Prior to this patch, we called ralloc_strndup each time
we constructed a new s_symbol.  It turns out that this is obscenely
expensive.

Instead, copy the whole buffer before parsing and overwrite it to
contain \0 bytes at the appropriate locations.  Since atoms are
separated by whitespace, (), or ;, we can safely overwrite the character
after a Symbol.  While much of the buffer may be unused, copying the
whole buffer is simple and guaranteed to provide enough space.

Prior to this, running piglit-run.py -t glsl tests/quick.tests with GLSL
1.30 enabled took just over 10 minutes on my machine.  Now it takes 5.

NOTE: This is a candidate for stable release branches (because it will
      make running comparison tests so much less irritating.)

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 3875526926)
2011-10-02 16:50:39 +02:00
Marcin Baczyński
3fc660e896 configure: allow C{,XX}FLAGS override
NOTE: This is a candidate for the 7.11 branch.

Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit ff2efdf599)
2011-10-02 16:49:56 +02:00
Marcin Baczyński
0fbf3562d1 configure: fix gcc version check
NOTE: This is a candidate for the 7.11 branch.

Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit fa013419de)
2011-10-02 16:49:49 +02:00
Vadim Girlin
13476840a6 st/mesa: flush bitmap cache on query and conditional render boundaries
Bitmap caching shouldn't affect the results of the queries and
conditional render.

NOTE: This is a candidate for the 7.11 branch.

Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 4f4855b249)
2011-10-02 16:49:34 +02:00
Henri Verbeet
5336f7f5a5 mesa: Fix a couple of TexEnv unit limits.
NOTE: This is a candidate for the 7.11 branch.

Signed-off-by: Henri Verbeet <hverbeet@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit bfe284fd26)
2011-10-02 16:46:10 +02:00
Henri Verbeet
a470104763 mesa: Use the Elements macro for the sampler index assert in validate_samplers().
This is probably nicer if the array size ever changes.

NOTE: This is a candidate for the 7.11 branch.

Signed-off-by: Henri Verbeet <hverbeet@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 4744195628)
2011-10-02 16:46:02 +02:00
Henri Verbeet
338cf7128c mesa: Allow sampling from units >= MAX_TEXTURE_UNITS in shaders.
The total number of units used by a shader is limited to MAX_TEXTURE_UNITS,
but the actual indices are only limited by MAX_COMBINED_TEXTURE_IMAGE_UNITS,
since they're shared between vertex and fragment shaders.

NOTE: This is a candidate for the 7.11 branch.

Signed-off-by: Henri Verbeet <hverbeet@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 86adc2b29e)
2011-10-02 16:45:54 +02:00
Henri Verbeet
36dab12726 mesa: Check the texture against all units in unbind_texobj_from_texunits().
NOTE: This is a candidate for the 7.11 branch.

Signed-off-by: Henri Verbeet <hverbeet@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 2e35d90fb9)
2011-10-02 16:45:46 +02:00
Brian Paul
dc253d3100 mesa: fix texstore addressing bugs for depth/stencil formats
Using GLuint pointers worked when the pixel size was four bytes
or the row stride was a multiple of four but was otherwise broken.
Fixes failures found with the piglit fbo-stencil test.

This helps to fix https://bugs.freedesktop.org/show_bug.cgi?id=38729

NOTE: This is a candidate for the 7.11 branch.
(cherry picked from commit b786db0654)
2011-10-02 16:43:08 +02:00
Brian Paul
9f362a587e softpipe: add missing stencil format case in convert_quad_stencil()
Part of the fix for https://bugs.freedesktop.org/show_bug.cgi?id=38729

NOTE: This is a candidate for the 7.11 branch
(cherry picked from commit 057a107d44)

Conflicts:

	src/gallium/drivers/softpipe/sp_quad_depth_test.c
2011-10-02 16:42:54 +02:00
Henri Verbeet
8b5257a96f r600g: Support the PIPE_FORMAT_R16_FLOAT colorformat.
NOTE: This is a candidate for the 7.11 branch.
(cherry picked from commit 39fecd3229)
2011-10-02 16:38:09 +02:00
Eric Anholt
94e12df164 glsl: Allow ir_assignment() constructor to not specify condition.
We almost never want to specify a condition, and when we do we're
already thinking about it (because we're writing a lowering pass
generating the condition), so a default argument should make the code
more pleasant to read.

NOTE: This is a candidate for the 7.11 branch (we want to be able to
cherry-pick future code).
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit e617a53a74)
2011-10-02 16:31:16 +02:00
Eric Anholt
ac6a24001b mesa: Don't skip glGetProgramLocalParam4dvARB if there was already an error.
Like the previous commit, but fixes
ARB_vertex_program/getlocal4d-with-error.

v2: Move the success case line into the conditional, use ASSIGN_4V more.
(cherry picked from commit c9aac11713)
2011-10-02 15:33:05 +02:00
Eric Anholt
b7e69912fa mesa: Throw an error when starting conditional render on an active query.
From the NV_conditional_render spec:

    BeginQuery sets the active query object name for the query type given by
    <target> to <id>.  If BeginQuery is called with an <id> of zero, if the
    active query object name for <target> is non-zero, if <id> is the active
    query object name for any query type, or if <id> is the active query
    object for condtional rendering (Section 2.X), the error INVALID OPERATION
    is generated.

Fixes piglit nv_conditional_render-begin-while-active.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit fd17de2123)
2011-10-02 15:33:05 +02:00
Eric Anholt
268a2c1a8a mesa: Throw an error instead of asserting for condrender with query == 0.
From the NV_conditional_render spec:

    BeginQuery sets the active query object name for the query type given by
    <target> to <id>.  If BeginQuery is called with an <id> of zero, if the
    active query object name for <target> is non-zero, if <id> is the active
    query object name for any query type, or if <id> is the active query
    object for condtional rendering (Section 2.X), the error INVALID OPERATION
    is generated.

Fixes piglit nv_conditional_render-begin-zero.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 7371224c06)
2011-10-02 15:33:05 +02:00
Eric Anholt
28b95b2b01 mesa: Add support for Begin/EndConditionalRender in display lists.
Fixes piglit nv_conditional_render-dlist.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 8899f6e93c)
2011-10-02 15:33:05 +02:00
Ian Romanick
0a6ef34741 mesa: Make _mesa_get_compressed_formats match the texture compression specs
The implementation deviated slightly from the GL_EXT_texture_sRGB spec
and from other implementations.  A giant comment block was added to
justify the somewhat odd behavior of this function.

In addition, the interface had unnecessary cruft.  The 'all' parameter
was false at all callers, so it has been removed.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit b189d1635d)
2011-10-02 15:33:04 +02:00
Ian Romanick
545ecf6542 mesa: Return the correct internal fmt when a generic compressed fmt was used
If an application requests a generic compressed format for a texture
and the driver does not pick a specific compressed format, return the
generic base format (e.g., GL_RGBA) for the GL_TEXTURE_INTERNAL_FORMAT
query.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=3165
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 143b65f761)
2011-10-02 15:33:04 +02:00
Ian Romanick
a88f4914a3 mesa: Add utility function to get base format from a GL compressed format
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 09916e877f)
2011-10-02 15:33:04 +02:00
Eric Anholt
ad2b567f51 mesa: Fix glGetUniform() type conversions.
We were primarily failing to convert in the NativeIntegers case, which
this fixes.  However, we were also just truncating float uniforms when
converting to integer, which does not appear to be the correct
behavior.  Note, however, that the NVIDIA drivers also truncate
instead of rounding.

GL_DOUBLE return type is dropped because it was never used and
completely broken.  It can be added when there's test code.

Fixes piglit ARB_shader_objects/getuniform

v2: This is a rewrite of my previous glGetUniform patch, which Ken
    pointed out missed storage_type-based conversions to integer,
    which was totally broken still thanks to a typo in the testcase.
v3: Quote the spec justifying the rounding behavior.

Acked-by: Kenneth Graunke <kenneth@whitecape.org>
Acked-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 9fa41f0742)

Conflicts:

	src/mesa/main/uniforms.c
2011-10-02 15:33:04 +02:00
Marek Olšák
37c2c9688c u_vbuf_mgr: fix uploading with a non-zero index bias
Also don't rely on pipe_draw_info being set correctly.

NOTE: This is a candidate for the 7.11 branch.
(cherry picked from commit 60a77cf316)
2011-10-02 15:33:04 +02:00
Marek Olšák
012b2057e8 u_vbuf_mgr: rework user buffer uploads
- first determine the buffer range to upload for each buffer by walking over
  vertex elements
- take buffer_offset into account
- take src_offset into account
- take src_format into account in more places
- don't just blindly upload (stride*count) bytes

NOTE: This is a candidate for the 7.11 branch.
(cherry picked from commit cd9bbb3935)
2011-10-02 15:33:03 +02:00
Marek Olšák
eb2a04b467 u_vbuf_mgr: remove unused flag U_VBUF_UPLOAD_FLUSHED
(cherry picked from commit 315300e444)
2011-10-02 15:33:03 +02:00
Marek Olšák
060e22c212 u_vbuf_mgr: s/u_vbuf_mgr_/u_vbuf_
(cherry picked from commit 28fb798911)

Conflicts:

	src/gallium/auxiliary/util/u_vbuf_mgr.c
	src/gallium/drivers/r600/r600_state_common.c
2011-10-02 15:33:03 +02:00
Marek Olšák
2330b7267c u_vbuf_mgr: fix max_index computation for large src_offset
NOTE: This is a candidate for the 7.11 branch.
2011-10-02 15:33:02 +02:00
Marek Olšák
7a7377a090 u_vbuf_mgr: don't take per-instance attribs into acc. when computing max index
NOTE: This is a candidate for the 7.11 branch.
2011-10-02 15:33:02 +02:00
Marek Olšák
d79d69c933 u_vbuf_mgr: cleanup original vs real vertex buffer arrays
It can now override both buffer offsets and strides in additions to resources.
Overriding buffer offsets was kinda hackish and could cause issues with
non-native vertex formats.
2011-10-02 15:33:02 +02:00
Henri Verbeet
d072260921 mesa: Also set the remaining draw buffers to GL_NONE when updating just the first buffer in _mesa_drawbuffers().
Without this we'd miss the last update in a sequence like {COLOR0, COLOR1},
{COLOR0}, {COLOR0, COLOR1}. I originally had a patch for this that called
updated_drawbuffers() when the buffer count changed, but later realized that
was wrong. The ARB_draw_buffers spec explicitly says "The draw buffer for
output colors beyond <n> is set to NONE.", and this is queryable state.
This fixes piglit arb_draw_buffers-state_change.

NOTE: This is a candidate for the 7.11 branch.

Signed-off-by: Henri Verbeet <hverbeet@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit a4d72189b2)
2011-10-02 15:33:02 +02:00
Michel Dänzer
942c1b8de1 glx/dri2: Don't call X server for SwapBuffers when there's no back buffer.
As already done in dri2CopySubBuffer().

Should fix:

https://bugs.freedesktop.org/show_bug.cgi?id=36371
https://bugs.freedesktop.org/show_bug.cgi?id=40533

Might fix:

https://bugs.freedesktop.org/show_bug.cgi?id=32589

Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
(cherry picked from commit d8c443ddde)
2011-10-02 03:54:02 +02:00
Kristian Høgsberg
0502da1ade glx: Don't flush twice if we fallback to dri2CopySubBuffer
The flush extensions flush call indicates end of frame and should only
be called once per frame.  However, in the dri2SwapBuffer fallback
path, we call flush and then call dri2CopySubBuffer, which also calls
flush.  Refactor the code to only call flush once.
(cherry picked from commit 4a7667b96b)
2011-10-02 03:53:55 +02:00
Michel Dänzer
bb38f931a8 st/mesa: Finalize texture on render-to-texture.
This makes sure that stObj->pt exists and is up to date.

Fixes https://bugs.freedesktop.org/show_bug.cgi?id=39193 and piglit
fbo-incomplete-texture-03.

Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Brian Paul <brianp@vmware.com>

NOTE: This is a candidate for the 7.11 branch.
(cherry picked from commit de414f4915)
2011-10-02 03:45:43 +02:00
Brian Paul
726ce042f8 st/mesa: Convert size assertions to conditionals in st_texture_image_copy.
Prevents potential assertion failures in piglit fbo-incomplete-texture-03 test.

NOTE: This is a candidate for the 7.11 branch.
(cherry picked from commit 4beb8f9e9d)
2011-10-02 03:45:36 +02:00
Eric Anholt
4ebd2c7c09 mesa: Don't skip glGetProgramEnvParam4dvARB if there was already an error.
Fixes a bug caught by oglconform, and now piglit
ARB_vertex_program/getenv4d-with-error.  The wrapping of an existing
GL function made it so that we couldn't distinguish an error in
looking up our arguments from an existing error.  Instead, make a
helper function to choose the param, and use it from multiple callers.

v2: Move the success case line into the conditional, use COPY_4V more.
(cherry picked from commit e9d563e3ff)
2011-10-02 03:25:03 +02:00
Marcin Slusarz
e7794048ca nouveau: fix nouveau_fence leak
(commit 96054375b1 in master)
2011-09-13 15:36:01 +02:00
Carl Simonson
e20346bfc8 i830: Add missing vtable entry for i830 from the hiz work.
(cherry picked from commit 09eeb0ff27)
2011-09-10 18:33:13 -07:00
David Reveman
2217a70aaf i915g: Fix off-by-one in scissors. 2011-08-26 17:43:20 -07:00
Tobias Droste
6c1a9a327d r300/compiler: simplify code in peephole_add_presub_add
Signed-off-by: Tobias Droste <tdroste@gmx.de>
Signed-off-by: Marek Olšák <maraeo@gmail.com>
(cherry picked from commit 84f8548dfc)
2011-08-07 18:36:17 +02:00
Marek Olšák
e33f306d67 r300/compiler: remove an unused-but-set variable and simplify the code
(cherry picked from commit ed5e95ada6)
2011-08-07 18:36:10 +02:00
Marek Olšák
f69357d77a r300/compiler: fix a warning that a variable may be uninitialized
(cherry picked from commit 2ce6c3ea6e)
2011-08-07 18:36:03 +02:00
Marek Olšák
d8a0c1b4bc winsys/radeon: fix space checking
We should remove the relocations which caused a validation failure
from the list, so that the kernel receives only the validated ones.

NOTE: This is a candidate for the 7.11 branch.
(cherry picked from commit 64ab39b035)

Conflicts:

	src/gallium/winsys/radeon/drm/radeon_drm_cs.c
2011-08-07 18:35:06 +02:00
Marek Olšák
f396e43b7d vbo: do not call _mesa_max_buffer_index in debug builds
That code drops performance in Unigine Heaven and Tropics
by a factor of 10. That's too crazy even for a debug build.

NOTE: This is a candidate for the 7.11 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit c251d83d91)
2011-08-07 18:28:51 +02:00
Brian Paul
aedfd07fb2 docs: news item for 7.11 release 2011-08-02 10:38:17 -06:00
Brian Paul
9e669d9f28 docs: add 7.11 md5 sums 2011-08-02 10:38:03 -06:00
Marc Pignat
4258e9b3a5 drisw: Fix 24bpp software rendering, take 2
This patch add the support for 24bpp in the dri/swrast implementation.
See http://bugs.freedesktop.org/show_bug.cgi?id=23525

Signed-off-by: Marc Pignat <marc at pignat.org>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit cfec000e75)
2011-08-02 10:03:15 -06:00
Ian Romanick
de8f22af28 mesa: Bump version to 7.11 (final) 2011-07-31 22:48:10 -07:00
Ian Romanick
6ee71bab94 docs: More bits of 7.11 release notes 2011-07-31 22:47:26 -07:00
Jeremy Huddleston
6c72801c2b darwin: Use machine/endian.h to determine endianness
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 5b3c719983)
2011-07-31 09:47:56 -07:00
Jeremy Huddleston
0d2c369535 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 09:47:51 -07:00
Ian Romanick
fffd20cdc1 Merge remote-tracking branch 'origin/7.11' into 7.11 2011-07-28 16:18:51 -07:00
Ian Romanick
fad610fec9 mesa: Bump version to 7.11-rc4 2011-07-28 16:08:42 -07:00
Vadim Girlin
9ca791d380 r600g: fix vs export count
Fixes https://bugs.freedesktop.org/show_bug.cgi?id=39572

Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
2011-07-28 19:05:59 -04:00
Kenneth Graunke
69720cb0c4 i965: Remove the now unused intel_renderbuffer::draw_offset field.
The previous commit removed the last use of this field.

Reviewed-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit f73caddd33)
2011-07-28 14:23:08 -07:00
Kenneth Graunke
8589ca000a i965: Check actual tile offsets in Gen4 miptree workaround.
The purpose of the (irb->draw_offset & 4095) != 0 check was to ensure
that we don't have XYy offsets into a tile, since Gen4 hardware doesn't
support that.  However, it's insufficient: there are cases where
draw_offset & 4095 is 0 but we still have a Y-offset.  This leads to an
assertion failure in brw_update_renderbuffer_surface with tile_y != 0.

Instead, simply call intel_renderbuffer_tile_offsets to compute the
actual X/Y offsets and check if either are non-zero.  This makes both
the workaround and the assertion check the same things.

Fixes piglit test fbo-generatemipmap-formats, and should also fix
bugs #34009 and #39487.

NOTE: This is a candidate for stable release branches.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=34009
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=39487
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Chad Versace <chad@chad-versace.us>
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 15c0bc5eef)
2011-07-28 14:22:59 -07:00
Kenneth Graunke
dcbd00e73c i965/gen4: Fix message parameter loading for 1D TXD sampling.
We were neglecting to load dvdx and dvdy.  v is not optional.

Fixes glslparsertests tex-grad-0[12345].frag on Broadwater/Crestline.
(We still need an execution test using sampler1D.)

NOTE: This is a candidate for the 7.11 branch.

Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 3e1fd13f60)
2011-07-28 14:22:48 -07:00
Fredrik Höglund
bba1600531 st/mesa: fix the texture format in st_context_teximage
Commit 1a339b6c71 made
st_ChooseTextureFormat map GL_RGBA with type GL_UNSIGNED_BYTE
to PIPE_FORMAT_A8B8G8R8_UNORM.

The image format for ARGB pixmaps is PIPE_FORMAT_B8G8R8A8_UNORM
however. This mismatch caused the texture to be recreated in
st_finalize_texture.

NOTE: This is a candidate for the 7.11 branch.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=39209
Signed-off-by: Fredrik Höglund <fredrik@kde.org>
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Signed-off-by: Brian Paul <brianp@vmware.com>
2011-07-28 11:56:45 -07:00
Ian Romanick
f4c55ea016 mesa: Ensure that r300 compiler files only appear once in the tarballs
Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com>
Tested-by: Jeremy Huddleston <jeremyhu@apple.com>
Cc: Andreas Radke <a.radke@arcor.de>
2011-07-28 11:47:46 -07:00
Ian Romanick
aa05fbe14d mesa: Bump version to 7.11-rc3 2011-07-28 11:47:45 -07:00
Ian Romanick
0a7574f62c mesa: Use --dereference to avoid symlinks in tarballs 2011-07-28 11:47:45 -07:00
Eric Anholt
1098f9228d i965/fs: Fix MRT drawing since the m0->m2 move for shader debug.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 3daa2d97eb)
2011-07-28 11:47:45 -07:00
Eric Anholt
5238303790 i965: Fix many of the trivial WebGL demos that broke due to IB optimization.
The index buffer state emit only occurred if there was an IB in place
and we were in either a new batch or a new IB state.  But because we
only flagged new IB state if IB state changed from the last IB state
we calculated, we could simply never emit IB state after batchbuffer
wraps if the first draw didn't use the IB and we didn't actually
change the IB.

Fixes piglit glx-multi-context-ib-1.
(cherry picked from commit 818db3848b)
2011-07-28 11:47:45 -07:00
Eric Anholt
a03974ffb9 i965: Emit texture cache flushes on gen6 along with render cache flushes.
It turns out that internally the texture cache gets flushed in a
couple of cases, particularly around 2D operations mixed with 3D.  In
almost all cases one of those happens between rendering to an
FBO-attached texture and rendering from that texture.  However, as of
the next patch, glean tfbo (and the new fbo-flushing-2 test) would
manage to get stale texture values because one of those flushes didn't
occur.  The intention of this code was always to get the render cache
cleared and ready to be used from the sampler cache (and it does on <=
gen4), so this just catches gen5 up.

This patch was also tested to fix fbo-flushing on gen7.
(cherry picked from commit 185868c9c2)
2011-07-28 11:47:45 -07:00
Paul Berry
2d64d34cb9 i965: vs optimization fix: Check val.{negate,abs} in accumulator_contains()
When emitting a MAC instruction in a vertex shader, brw_vs_emit()
calls accumulator_contains() to determine whether the accumulator
already contains the appropriate addend; if it does, then we can avoid
emitting an unnecessary MOV instruction.

However, accumulator_contains() wasn't checking the val.negate or
val.abs flags.  As a result, if the desired value was the negation, or
the absolute value, of what was already in the accumulator, we would
generate an incorrect shader.

Fixes piglit test vs-refract-vec4-vec4-float.

Tested on Gen5 and Gen6.

Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit d92463d5dc)
2011-07-28 11:47:45 -07:00
Kenneth Graunke
66b41af391 i965/gen7: Fix shadow sampling in the old brw_wm_emit backend.
On Ivybridge, the shadow comparitor goes in the first slot, rather than
at the end.  It's not necessary to send u, v, and r.

Fixes tests texturing/texdepth and glean/fbo.

NOTE: This is a candidate for the 7.11 branch.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 572f631895)
2011-07-28 11:47:45 -07:00
Kenneth Graunke
67aa20d9d5 i965/fs: Clear result before visiting shadow comparitor and LOD info.
Commit 53c89c67f3 ("i965: Avoid generating
MOVs for assignments of expressions.") added the line "this->result =
reg_undef" all over the code.  Unfortunately, since Eric developed his
patch before I landed Ivybridge support, he missed adding it to
fs_visitor::emit_texture_gen7() after rebasing.

Furthermore, since I developed TXD support before Eric's patch, I
neglected to add it to the gradient handling when I rebased.

Neglecting to set this causes the visitor to use this->result as storage
rather than generating a new temporary.  These missing statements
resulted in the same register being used to store several different
values.

Fixes the following piglit tests on Ivybridge:
- glsl-fs-shadow2dproj.shader_test
- glsl-fs-shadow2dproj-bias.shader_test

NOTE: This is a candidate for the 7.11 branch.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 156cef0fba)
2011-07-28 11:47:45 -07:00
Ian Romanick
7d6c37b0c2 glsl: Treat ir_dereference_array of non-var as a constant for lowering
Previously the code would just look at deref->array->type to see if it
was a constant.  This isn't good enough because deref->array might be
another ir_dereference_array... of a constant.  As a result,
deref->array->type wouldn't be a constant, but
deref->variable_referenced() would return NULL.  The unchecked NULL
pointer would shortly lead to a segfault.

Instead just look at the return of deref->variable_referenced().  If
it's NULL, assume that either a constant or some other form of
anonymous temporary storage is being dereferenced.

This is a bit hinkey because most drivers treat constant arrays as
uniforms, but the lowering pass treats them as temporaries.  This
keeps the behavior of the old code, so this change isn't making things
worse.

Fixes i965 piglit:

    vs-temp-array-mat[234]-index-col-rd
    vs-temp-array-mat[234]-index-col-row-rd
    vs-uniform-array-mat[234]-index-col-rd
    vs-uniform-array-mat[234]-index-col-row-rd

Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 156f85336f)
2011-07-28 11:47:45 -07:00
Ian Romanick
32c7224edb i965: When emitting a src/dst read of an output, keep the swizzle and neg
Fixes i965 piglit vs-varying-array-mat[234]-row-rd.

Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 1d3f09f159)
2011-07-28 11:47:45 -07:00
Ian Romanick
788acd552e i965: When emitting a src/dst write of an output, keep the write mask
Fixes i965 piglit:

    vs-varying-array-mat[234]-col-row-wr
    vs-varying-array-mat[234]-index-col-row-wr
    vs-varying-array-mat[234]-index-row-wr
    vs-varying-array-mat[234]-row-wr
    vs-varying-mat[234]-col-row-wr
    vs-varying-mat[234]-row-wr

Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 337e2dfad0)
2011-07-28 11:47:45 -07:00
Ian Romanick
a2b8802ed5 prog_optimize: Set unused regs to PROGRAM_UNDEFINED after CMP->MOV conversion
Leaving the unused registers with other values caused assertion
failures and other problems in places that blindly iterate over all
sources.

brw_vs_emit.c:1381: get_src_reg: Assertion `c->regs[file][index].nr !=
0' failed.

Fixes i965 piglit:

    vs-uniform-array-mat[234]-col-row-rd
    vs-uniform-array-mat[234]-index-col-row-rd
    vs-uniform-array-mat[234]-index-row-rd
    vs-uniform-mat[234]-col-row-rd

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit fbeb68e880)
2011-07-28 11:47:45 -07:00
Ian Romanick
8b41ae0b2a ir_to_mesa: Copy reladdr in src_reg(dst_reg) constructor
Fixes i965 piglit:

    vs-temp-array-mat[234]-col-row-wr
    vs-temp-array-mat[234]-index-col-row-wr
    vs-temp-array-mat[234]-index-row-wr
    vs-temp-mat[234]-col-row-wr

Fixes swrast piglit:

    fs-temp-array-mat[234]-col-row-wr
    fs-temp-array-mat[234]-index-col-row-wr
    fs-temp-array-mat[234]-index-row-wr
    fs-temp-mat[234]-col-row-wr
    vs-temp-array-mat[234]-col-row-wr
    vs-temp-array-mat[234]-index-col-row-wr
    vs-temp-array-mat[234]-index-row-wr
    vs-temp-mat[234]-col-row-wr

Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit f7cd9a858c)
2011-07-28 11:47:44 -07:00
Ian Romanick
b00b5fe267 ir_to_mesa: Add each relative address to the previous
This fixes many cases of accessing arrays of matrices using
non-constant indices at each level.

Fixes i965 piglit:

    vs-temp-array-mat[234]-index-col-rd
    vs-temp-array-mat[234]-index-col-row-rd
    vs-temp-array-mat[234]-index-col-wr
    vs-uniform-array-mat[234]-index-col-rd

Fixes swrast piglit:

    fs-temp-array-mat[234]-index-col-rd
    fs-temp-array-mat[234]-index-col-row-rd
    fs-temp-array-mat[234]-index-col-wr
    fs-uniform-array-mat[234]-index-col-rd
    fs-uniform-array-mat[234]-index-col-row-rd
    fs-varying-array-mat[234]-index-col-rd
    fs-varying-array-mat[234]-index-col-row-rd
    vs-temp-array-mat[234]-index-col-rd
    vs-temp-array-mat[234]-index-col-row-rd
    vs-temp-array-mat[234]-index-col-wr
    vs-uniform-array-mat[234]-index-col-rd
    vs-uniform-array-mat[234]-index-col-row-rd
    vs-varying-array-mat[234]-index-col-rd
    vs-varying-array-mat[234]-index-col-row-rd
    vs-varying-array-mat[234]-index-col-wr

Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit d6e1a8f714)
2011-07-28 11:47:44 -07:00
Ian Romanick
3a5a7cfd2a glsl: When lowering non-constant vector indexing, respect existing conditions
If the non-constant index was in the LHS of an assignment, any
existing condititon on that assignment would be lost.

Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 601428d2bb)
2011-07-28 11:47:44 -07:00
Ian Romanick
aad12d15bb glsl: When lowering non-constant array indexing, respect existing conditions
If the non-constant index was in the LHS of an assignment, any
existing condititon on that assignment would be lost.

Fixes i965 piglit:

    fs-temp-array-mat[234]-col-row-wr
    fs-temp-array-mat[234]-index-col-row-wr
    fs-temp-array-mat[234]-index-col-wr
    fs-temp-array-mat[234]-index-row-wr
    vs-varying-array-mat[234]-index-col-wr

Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 5f83dfe5b7)
2011-07-28 11:47:44 -07:00
Ian Romanick
977db7cc65 glsl: Rework lowering of non-constant array indexing
The previous implementation could easily get tricked if the LHS of an
assignment included a non-constant index that was "inside" another
dereference.  For example:

    mat4 m[2];
    m[0][i] = vec4(0.0);

Due to the way it tracked whether the array was being assigned, it
would think that the non-constant index was in an r-value.  The new
code fixes that by tracking l-values and r-values differently.  The
index is also replaced by cloning the IR and replacing the index
variable instead of the odd way it was done before.

v2: Apply some simplifications suggested by Eric Anholt.  Making
assignment_generator::rvalue be ir_dereference instead of ir_rvalue
simplified the code a bit.

Fixes i965 piglit fs-temp-array-mat[234]-index-wr and
vs-varying-array-mat[234]-index-wr.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=34691
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 1731ac3086)

To make bisects work, this also squashes in:

glsl: Correctly return progress from lower_variable_index_to_cond_assign

lower_variable_index_to_cond_assign runs until it can't make any more
progress.  It then returns the result of the last pass which will
always be false.  This caused the lowering loop in
_mesa_ir_link_shader to end before doing one last round of
lower_if_to_cond_assign.  This caused several if-statements (resulting
from lower_variable_index_to_cond_assign) to be left in the IR.

In addition to this change, lower_variable_index_to_cond_assign should
take a flag indicating whether or not it should even generate
if-statements.  This is easily controlled by
switch_generator::linear_sequence_max_length.  This would generate
much better code on architectures without any flow contol.

Fixes i915 piglit regressions glsl-texcoord-array and
glsl-fs-vec4-indexing-temp-src.

Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit c1e591eed4)
2011-07-28 11:47:44 -07:00
Ian Romanick
d31c1c33ed glsl: Split out part of variable_index_to_cond_assign_visitor::needs_lowering
Other code will soon need to know if an array needs lowering based
exclusively on the storage mode.

Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit d2296e784a)
2011-07-28 11:47:44 -07:00
Ian Romanick
3481a5a1e6 glsl: Move is_array_or_matrix outside visitor class
There's no reason for it to be there, and another class that may not
have access to the visitor will need it soon.

Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 8d5f3cef79)
2011-07-28 11:47:44 -07:00
Marek Olšák
a26b6cc001 configure.ac: add DLOPEN_LIBS to xlib build
Otherwise xlib-based llvmpipe fails to link.

NOTE: This is a candidate for the 7.11 branch.
(cherry picked from commit 0aed27ee37)
2011-07-28 00:29:07 +02:00
Benjamin Franzke
7d992b8471 wayland-drm: Add copyright notice to protocol
Fixes build since wayland 986703ac7365bc87a5501714adb9fc73157c62b7.
(cherry picked from commit 79dcfb266a)
2011-07-27 10:08:01 +02:00
Tobias Droste
0bfa310049 egl/gallium: fix build without softpipe and llvmpipe
Signed-off-by: Tobias Droste <tdroste@gmx.de>
Acked-by: Jakob Bornecrantz <wallbraker@gmail.com>
Reviewed-by: Marek Olšák <maraeo@gmail.com>
(cherry picked from commit d4d5e3a336)
2011-07-27 09:43:03 +02:00
Benjamin Franzke
7555eb7478 Fix broken merge in cherry-pick from 42cdf407
Was cherry-picked to 337102684b.
2011-07-27 09:25:54 +02:00
Bryan Cain
fd461c5888 util: enable S3TC support when the force_s3tc_enable env var is set to "true"
NOTE: This is a candidate for the 7.10 and 7.11 branches.
2011-07-26 13:23:18 -05:00
Bryan Cain
208bae4251 st/mesa: respect force_s3tc_enable environment variable
NOTE: This is a candidate for the 7.10 and 7.11 branches.
2011-07-26 13:23:15 -05:00
Marek Olšák
13d12b35e9 configure.ac: do not check for llvm-config if llvm is disabled 2011-07-25 23:54:28 +02:00
Benjamin Franzke
337102684b configure: Move gbm before egl in SRC_DIRS
egl_dri2 built into libEGL depends on libgbm.

Fixes https://bugs.freedesktop.org/show_bug.cgi?id=39515
(cherry picked from commit 42cdf4074e)
2011-07-25 09:47:39 +02:00
Marek Olšák
b305c956be configure.ac: check for libdrm_radeon only when building classic
(cherry picked from commit 50e32fefb1)
2011-07-23 16:00:27 +02:00
Ian Romanick
7c98381ed4 glsl: Reject shaders that contain static recursion
The GLSL 1.20 and later specs say:

    "Recursion is not allowed, not even statically. Static recursion is
    present if the static function call graph of the program contains
    cycles."

Recursion is detected and rejected both a compile-time and at
link-time.  The complie-time check happens to detect some cases that
may be removed by various optimization passes.  The spec doesn't seem
to allow this, but other vendors (e.g., NVIDIA) appear to only check
at link-time after all optimizations.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=33885
Reviewed-by: Paul Berry <stereotype441@gmail.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 02c5ae1b3f)

This also squashes in the following commit to make sure that bisects
in scons builds work:

glsl: Add ir_function_detect_recursion.cpp to SConscript.
(cherry picked from commit 76bccaff0c)
2011-07-23 01:55:25 -07:00
Ian Romanick
0ce6506d6e glsl: Make prototype_string publicly available
Also clarify the documentation for one of the parameters.

Reviewed-by: Paul Berry <stereotype441@gmail.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 1ad3ba4ad9)
2011-07-23 01:53:59 -07:00
Stéphane Marchesin
7fc66f0bed Revert "i915: Eliminate redundant CONSTANTS updates"
This reverts commit 87641cffd9.
(cherry picked from commit 3c0c624879)
2011-07-23 01:53:40 -07:00
Chia-I Wu
e24c3575e4 u_vbuf_mgr: restore buffer offsets
u_vbuf_upload_buffers modifies the buffer offsets.  If they are not
restored, and any of the vertex formats is not supported natively, the
next u_vbuf_mgr_draw_begin call will translate the vertex buffers with
incorrect buffer offsets.
(cherry picked from commit afc160e1c8)

Signed-off-by: Marek Olšák <maraeo@gmail.com>
2011-07-21 22:18:56 +02:00
Marek Olšák
2e50a93cf3 r600g: more valgrind fixes
(cherry picked from commit dc9d789d1b)
2011-07-21 15:03:03 +02:00
Marek Olšák
2c29c24543 r600g: zero memory of ioctl parameters
Fixes valgrind warning.
(cherry picked from commit daf6604435)
2011-07-21 15:02:53 +02:00
Marek Olšák
4cbaa2283c configure.ac: Check for the respective libdrm_* when building gallium drivers
In a rare case of building gallium only, we need to
check if the required packages are available

libdrm_[intel|nouveau] - gallium[i915 i965|nouveau]

v2: r300g and r600g do not need libdrm_radeon

Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Signed-off-by: Marek Olšák <maraeo@gmail.com>
(cherry picked from commit c2426bbf86)

Conflicts:

	configure.ac
2011-07-21 14:57:19 +02:00
Marek Olšák
e29d44eacd mesa: GLES2 should return different error enums for invalid fbo queries
ES 2.0.25 page 127 says:

  If the value of FRAMEBUFFER_ATTACHMENT_OBJECT_TYPE is NONE, then
  querying any other pname will generate INVALID_ENUM.

See also:
b9e9df78a0

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

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 000896c0bb)
2011-07-21 14:54:09 +02:00
Andrew Randrianasulu
f29208aa9b dri/nouveau: nv10: fix vertex format for GL_UNSIGNED_BYTE
Broken accidentally in f4efc256fd,
the switch to rnn headers.

NV10TCL_VTXFMT_TYPE_BYTE_RGBA became U8_UNORM but B8G8R8A8_UNORM
was used instead.
2011-07-21 10:33:33 +02:00
David Heidelberger
7af4e18dcd nvfx: handle PIPE_CAP_SM3
Signed-off-by: David Heidelberger <d.okias@gmail.com>
2011-07-21 10:33:03 +02:00
Eric Anholt
344db29ede i965: Apply a homebrew workaround for GPU hang in OGLC api-texcoord.
The behavior of flushes in the hardware is a maze of twisty passages,
and strangely the VS constants appear to be loaded during a pipeline
flush instead of at the time of the packet emit according to the
simulator.  On moving the STATE_BASE_ADDRESS packet to where it really
needed to live (in order for data loads by other packets to be
correct), we sometimes no longer got a flush between those packets
where we apparently needed it.  This replicates the flushes implied by
a STATE_BASE_ADDRESS update, fixing the GPU hangs in OGLC and the
"engine" demo.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=36821
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=39257
Tested-by: Keith Packard <keithp@keithp.com> (bzflag and etracer fixed)
Acked-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 3e5d36267d)
2011-07-20 12:50:06 -07:00
Eric Anholt
680c468f61 i965: Enable the PIPE_CONTROL workaround workaround out of paranoia.
There's scary stuff going on in PIPE_CONTROL internals, and if the
BSpec says to do this to make PIPE_CONTROL work, I'll go ahead and do
it because we'll probably never be able to debug it after the fact.

v2: Use stall at scoreboard instead of depth stall, as noted by Ken.
(cherry picked from commit 407785d0e9)
2011-07-20 12:50:06 -07:00
Eric Anholt
7ee4cb453b i965: Avoid kernel BUG_ON if we happen to wait on the pipe_control w/a BO.
For this and occlusion queries, we're trying to avoid setting
I915_GEM_DOMAIN_RENDER for the write domain, because the data written
is definitely not going through the render cache, but we do need to
tell the kernel that the object has been written.  However, with using
I915_GEM_DOMAIN_GTT, the kernel on retiring the batchbuffer sees that
the w/a BO has a write domain of GTT, and puts it on the flushing
list.  If something tries to wait for that BO to finish rendering
(such as the AUB dumper reading the contents of BOs), we get into
wait_request (since obj->active) but with a 0 seqno (since the object
is on the flushing list, not actually on a ringbuffer), and BUG_ONs.

To avoid the kernel bug (which I'm hoping to delete soon anyway), just
use I915_GEM_DOMAIN_INSTRUCTION like occlusion queries do.  This
doesn't result in more flushing, because we invalidate INSTRUCTION on
every batchbuffer now that we're state streaming, anyway.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Tested-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit dc7422405f)
2011-07-20 12:50:06 -07:00
Eric Anholt
a152ab6df1 i915: Simplify intel_wpos_* with a helper function.
(cherry picked from commit cb5e0ba2aa)
2011-07-20 12:50:06 -07:00
Eric Anholt
a99914509e i915: Include gl_FragCoord.w data, not just xyz.
Fixes piglit fragcoord_w test.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=34323
(cherry picked from commit fceda4342c)
2011-07-20 12:50:06 -07:00
Eric Anholt
891073fea3 i915: Fix incorrect depth scaling when enabling/disabling depth buffers.
We were updating our new viewport using the old buffers' _WindowMap.m.
We can do less math and avoid using that deprecated matrix by just
folding the viewport calculation right in to the driver.

Fixes piglit fbo-depthtex.
(cherry picked from commit debf751aea)
2011-07-20 12:50:06 -07:00
Eric Anholt
dcaec739ea i915: Make stencil test for no-stencil handling match depth test.
i915_update_draw_buffers() already handles the fallback bit for
missing stencil region, so here we just need to handle whether the GL
thinks we have stencil data or not (and disable the test if so).
(cherry picked from commit 79fee3a76b)
2011-07-20 12:50:06 -07:00
Eric Anholt
ea241750f7 i915: Disable the depth test whenever we don't have a depth buffer.
We were disabling it once at the moment we changed draw buffers, but
later enabling of depth test could turn it back on.  Fixes
fbo-nodepth-test.

Note that ctx->DrawBuffer has to be checked because during context
create we get called while it's still unset.  However, we know we'll
get an intel_draw_buffer() after that, so it's safe to make a silly
choice at this point.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=30080
(cherry picked from commit fc4fba52cf)
2011-07-20 12:50:06 -07:00
Eric Anholt
9efd6dc677 i915: Remove i965 paths from i915_update_drawbuffer() and i830's too.
Reviewed-by: Chad Versace <chad@chad-versace.us>
(cherry picked from commit 4c47fce92e)
2011-07-20 12:50:06 -07:00
Eric Anholt
c6ddeeed7a i965: Remove i915 paths from brw_update_draw_buffers().
Reviewed-by: Chad Versace <chad@chad-versace.us>
(cherry picked from commit 94efc350b4)
2011-07-20 12:50:05 -07:00
Eric Anholt
108e807b7b i965: Remove unused region calculations in brw_update_draw_buffer().
Reviewed-by: Chad Versace <chad@chad-versace.us>
(cherry picked from commit c68270a26b)
2011-07-20 12:50:05 -07:00
Eric Anholt
421bca32fb i965: Remove empty brw_set_draw_region.
Reviewed-by: Chad Versace <chad@chad-versace.us>
(cherry picked from commit 15af0f54b8)
2011-07-20 12:50:05 -07:00
Eric Anholt
0f0ab15a46 i965: Remove FALLBACK() from brw_update_draw_region().
The 965 driver doesn't use these for deciding on fallbacks.

Reviewed-by: Chad Versace <chad@chad-versace.us>
(cherry picked from commit dd898c3e89)
2011-07-20 12:50:05 -07:00
Eric Anholt
b3c2438d4e intel: Move intel_draw_buffers() code into each driver.
The illusion of shared code here wasn't fooling anybody.  It was
tempting to keep i830 and i915 still shared, but I think I actually
want to make them diverge shortly.

Reviewed-by: Chad Versace <chad@chad-versace.us>
(cherry picked from commit f34ec6169d)
2011-07-20 12:50:05 -07:00
Brian Paul
25287008b6 glsl: silence warning in linker.cpp
(cherry picked from commit 4470ff2ebf)
2011-07-20 09:02:22 -07:00
Jørgen Lind
62b282e749 Make it possible to use gbm with c++
NOTE: This is a candiate for 7.11
(cherry picked from commit 496bf3822a)
2011-07-20 09:02:18 -07:00
Paul Berry
9aa8c02ae2 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.

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

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 3097715d41)
2011-07-20 08:15:32 -07:00
Paul Berry
b0ecde7f39 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.

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

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 9c4445de6e)
2011-07-20 08:15:21 -07:00
Marek Olšák
fe70a40e47 prog_optimize: fix a warning that a variable may be uninitialized
(cherry picked from commit dade65505b)
Signed-off-by: Brian Paul <brianp@vmware.com>
2011-07-20 08:01:47 -06:00
Ian Romanick
73b68316f4 mesa: Bump version to 7.11-rc2 2011-07-19 16:39:57 -07:00
Brian Paul
7ba7531929 mesa: remove depend files from tarballs
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
2011-07-19 16:39:57 -07:00
Kenneth Graunke
2826e3a000 glsl: Correctly handle function matching when there are multiple inexact matches
This is a squash cherry pick commit of:

    glsl: Find the "closest" signature when there are multiple matches.

    Previously, ir_function::matching_signature had a fatal bug: if a
    function had more than one non-exact match, it would simply return NULL.

    This occured, for example, when looking for max(uvec3, uvec3):
    - max(vec3, vec3)   -> score 1 (found first)
    - max(ivec3, ivec3) -> score 1 (found second...used to return NULL here)
    - max(uvec3, uvec3) -> score 0 (exact match...the right answer)

    This did not occur for max(ivec3, ivec3) since the second match found
    was an exact match.

    The new behavior is to return a match with the lowest score.  If there
    is an exact match, that will be returned.  Otherwise, a match with the
    least number of implicit conversions is chosen.

    Fixes piglit tests max-uvec3.vert and glsl-inexact-overloads.shader_test.

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

    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 60eb63a855)

    glsl: Suppress warning from matching_signature change.

    gcc isn't smart enough to see that we only look at matched_score after
    we've initialized it (because match != NULL happens at the same time)
    (cherry picked from commit b043409adf)

    glsl: Reject ambiguous function calls (multiple inexact matches).

    According to the GLSL 1.20 specification, "it is a semantic error if
    there are multiple ways to apply [implicit] conversions [...] such that
    the call can be made to match multiple signatures."

    Fixes a regression caused by 60eb63a855,
    which implemented the wrong policy of finding a "closest" match.
    However, this is not a revert, since the original code failed to
    continue looking for an exact match once it found two inexact matches.

    It's OK to have multiple inexact matches if there's also an exact match.

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

    Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=38971
    Reviewed-by: Eric Anholt <eric@anholt.net>
    Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
    (cherry picked from commit 7304909d65)
2011-07-19 16:39:56 -07:00
Paul Berry
f80ae99cbd 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)
2011-07-19 16:39:56 -07:00
Paul Berry
ae11fb02dd 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)
2011-07-19 16:39:56 -07:00
Ian Romanick
a90b88f354 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)
2011-07-19 16:39:56 -07:00
Ian Romanick
0e699cc0e8 configure.ac: Make --{without,with}-gallium-drivers work as expected
This version is mostly Dan's post to the mesa-dev mailing list on
6/22/2011.

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

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Dan Nicholson <dbn.lists@gmail.com>
(cherry picked from commit db311b45be)
2011-07-19 16:39:56 -07:00
Kenneth Graunke
e6e7c456de i965/gen7: Add support for gl_PointCoord.
This is exactly analogous to Eric's Gen6 change in commit
6861a70177.  His explanation:

"This is just like PointSprite overrides, but it's always on for that
 attribute."

Fixes glsl-fs-pointcoord and gtf/point_sprites.

Signed-off-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>

(cherry-picked from commit 186e37c754)
2011-07-19 16:39:56 -07:00
Kenneth Graunke
31f4ab790f i965/gen7: Fix point sprite texture coordinate overrides.
This is exactly analogous to Eric's Gen6 change in commit
f304bb8a5d.  His explanation:

"We were assuming that the input attribute n to the FS was
 FRAG_ATTRIB_TEXn, which happened to be true often enough for our
 testcases."

Signed-off-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>

(cherry-picked from commit 147d010295)
2011-07-19 16:39:56 -07:00
Kenneth Graunke
9f978104d8 i965/gen7: Refactor SF setup a bit to handle overrides in one place.
This is exactly analogous to Eric's Gen6 change in commit
e7280b16d6.

Signed-off-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>

(cherry-picked from commit 5edb3ddf41)
2011-07-19 16:39:55 -07:00
Kenneth Graunke
9d78935100 i965/gen7: Remove gratuitous dirty flags from WM and PS state.
Commit b46dc45cee claimed that
NEW_POLYGONSTIPPLE is gratuitous, but somehow just changed comments
and whitespace instead of actually removing the flag.

While we're at it, 3DSTATE_PS doesn't appear to need NEW_LINE or
NEW_POLYGON either (those are in 3DSTATE_WM).  Also, 3DSTATE_WM
doesn't appear to need BRW_NEW_NR_WM_SURFACES or BRW_NEW_CURBE_OFFSETS
either (those are in 3DSTATE_PS).

Reviewed-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>

(cherry-picked from commit 57b57f6d1c)
2011-07-19 16:39:55 -07:00
Henri Verbeet
d469ebaa0a 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 23:29:16 +02:00
Chad Versace
f5fa4606ea intel: Fix stencil buffer to be W tiled
Until now, the stencil buffer was allocated as a Y tiled buffer, because
in several locations the PRM states that it is. However, it is actually
W tiled. From the PRM, 2011 Sandy Bridge, Volume 1, Part 2, Section
4.5.2.1 W-Major Format:
    W-Major Tile Format is used for separate stencil.

The GTT is incapable of W fencing, so we allocate the stencil buffer with
I915_TILING_NONE and decode the tile's layout in software.

This fix touches the following portions of code:
    - In intel_allocate_renderbuffer_storage(), allocate the stencil
      buffer with I915_TILING_NONE.
    - In intel_verify_dri2_has_hiz(), verify that the stencil buffer is
      not tiled.
    - In the stencil buffer's span functions, the tile's layout must be
      decoded in software.

This commit mutually depends on the xf86-video-intel commit
    dri: Do not tile stencil buffer
    Author: Chad Versace <chad@chad-versace.us>
    Date:   Mon Jul 18 00:38:00 2011 -0700

On Gen6 with separate stencil enabled, fixes the following Piglit tests:
    bugs/fdo23670-drawpix_stencil
    general/stencil-drawpixels
    spec/EXT_framebuffer_object/fbo-stencil-GL_STENCIL_INDEX16-copypixels
    spec/EXT_framebuffer_object/fbo-stencil-GL_STENCIL_INDEX16-drawpixels
    spec/EXT_framebuffer_object/fbo-stencil-GL_STENCIL_INDEX16-readpixels
    spec/EXT_framebuffer_object/fbo-stencil-GL_STENCIL_INDEX1-copypixels
    spec/EXT_framebuffer_object/fbo-stencil-GL_STENCIL_INDEX1-drawpixels
    spec/EXT_framebuffer_object/fbo-stencil-GL_STENCIL_INDEX1-readpixels
    spec/EXT_framebuffer_object/fbo-stencil-GL_STENCIL_INDEX4-copypixels
    spec/EXT_framebuffer_object/fbo-stencil-GL_STENCIL_INDEX4-drawpixels
    spec/EXT_framebuffer_object/fbo-stencil-GL_STENCIL_INDEX4-readpixels
    spec/EXT_framebuffer_object/fbo-stencil-GL_STENCIL_INDEX8-copypixels
    spec/EXT_framebuffer_object/fbo-stencil-GL_STENCIL_INDEX8-drawpixels
    spec/EXT_framebuffer_object/fbo-stencil-GL_STENCIL_INDEX8-readpixels
    spec/EXT_packed_depth_stencil/fbo-stencil-GL_DEPTH24_STENCIL8-copypixels
    spec/EXT_packed_depth_stencil/fbo-stencil-GL_DEPTH24_STENCIL8-readpixels
    spec/EXT_packed_depth_stencil/readpixels-24_8

Note: This is a candidate for the 7.11 branch.

Signed-off-by: Chad Versace <chad@chad-versace.us>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Acked-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit f7dbcba280)
2011-07-19 13:15:00 -07:00
Vadim Girlin
89ed95ad3d r600g: fix corner case checks for the queries 2011-07-18 09:00:42 -04:00
Christoph Bumiller
3b605cb0d6 nv50,nvc0: add correct storage type for Z32_FLOAT 2011-07-18 13:48:19 +02:00
Christoph Bumiller
9bfb79923f nv50,nvc0: don't advertise unaligned texture format support
Because we don't support them.
For instance, R32G32B32 is not R32G32B32X32 as was assumed.

Add support for R8G8B8X8_UNORM instead of R8G8B8_UNORM surfaces.
2011-07-18 13:45:52 +02:00
Vadim Girlin
7d8a04643f r600g: fix queries and predication
Use all zpass data for predication instead of the last block only.
Use query buffer as a ring instead of reusing the same area
for each new BeginQuery. All query buffer offsets are in bytes
to simplify offsets math.
2011-07-15 15:43:48 -04:00
Alex Deucher
3065bae508 r600c/g: add new NI pci ids
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2011-07-15 10:56:31 -04:00
Chia-I Wu
443ff6024d targets/egl-static: fix a linking error
rbug is always linked in and it needs libpthread.
(cherry picked from commit 5fe5d236c2)
2011-07-14 11:56:39 +08:00
Eric Anholt
a20a950829 i915: Add support for gl_FragData[0] for output color.
We advertised ARB_draw_buffers, but either fell back to software when
using this output, or assertion failed.  Fixes glsl-fs-fragdata-1, and
failures in some webgl conformance tests.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=39024
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=34906
(cherry picked from commit 556a47a262)
2011-07-13 13:14:09 -07:00
Eric Anholt
9279c1e556 i915: Fix NPOT compressed textures on 915.
We were failing at rounding, misplacing the non-baselevels.  Fixes:
3DFX_texture_compression_FXT1/fbo-generate-mipmaps
ARB_texture_compression/fbo-generate-mipmaps
EXT_texture_compression_s3tc/fbo-generate-mipmaps

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit f2fd0d6304)
2011-07-13 13:14:09 -07:00
Eric Anholt
b8f722a82e i915: Fix map/unmap mismatches from leaving INTEL_FALLBACK during TNL.
The first rendering after context create didn't know of the color
buffer yet, triggering a sw fallback.  The intel_prepare_render() from
intelSpanRenderStart then found the buffer and turned off fallbacks,
but intelSpanRenderFinish was never called and things were left
mapped.  By checking buffers before making the call on whether to do
the fallback pipeline or not, we avoid the fallback change inside of
the rendering pipeline.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=31561
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 6e6b388604)
2011-07-13 13:14:09 -07:00
Eric Anholt
9304645a07 i965: Fix fp-dst-aliasing-[12].vpfp.
There's no pretty way to avoid the overwriting of the src operands, so
just use a temporary destination and rely on the MOV optimization.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 46a7639174)
2011-07-13 13:14:09 -07:00
Eric Anholt
c3b3719096 i965: Fix fp-lit-src-equals-dst.
We were stomping over the source for the body of the LIT instruction
when doing the MOV of 1.0 to the uninteresting channels.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit e3ea5bc08e)
2011-07-13 13:14:09 -07:00
Eric Anholt
55a75856fb intel: Remove gratuitous context checks in intel_delete_renderbuffer().
Even if we don't have a current context, if we're freeing the rb we
should free its region (and BO).  The renderbuffer unreference checks
appear to be just cargo-cult from the region unreference code.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=30217
Reviewed-by: Chad Versace <chad@chad-versace.us>
(cherry picked from commit 007c2d6cd2)
2011-07-13 13:14:09 -07:00
Eric Anholt
e3e99be131 intel: Allow intel_region_reference() with *dst != NULL.
This should help us avoid leaking regions in region reference code by
making the API more predictable.

Reviewed-by: Chad Versace <chad@chad-versace.us>
(cherry picked from commit 036b74a7f8)
(cherry picked from commit d8f65c07e9)
2011-07-13 13:12:55 -07:00
Eric Anholt
5a7d1c9710 glsl: Fix make clean for dricore.
(cherry picked from commit abbbd14dd4)
2011-07-13 13:08:23 -07:00
Eric Anholt
6ac5554298 i965: Reissue PIPELINE_POINTERS and BINDING_TABLE_POINTERS on SBA change.
This was a requirement we didn't run into until we started using
STATE_BASE_ADDRESS for instruction data.
(cherry picked from commit a09c5c2e30)
2011-07-13 13:08:23 -07:00
Eric Anholt
9eace71048 i965/gen6: Fix scissors using invalid STATE_BASE_ADDRESS.
The scissor state was incorrectly in a .prepare function instead of
.emit, so the packet would end up in the batch before the
STATE_BASE_ADDRESS.  It appears that this doesn't actually hurt, as
the scissor address gets dereferenced according to the current SBA at
draw time.
(cherry picked from commit cd7bfd5d44)
2011-07-13 13:08:23 -07:00
Stéphane Marchesin
8d9202c162 i915g: don't try to check if a NULL buffer is busy. 2011-07-13 12:08:10 -07:00
Alex Deucher
ef9f16f632 r600g: emit SQ_LDS_RESOURCE_MGMT
Need to be initialized to a reasonable value as
compute code may change it.

Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=39119

NOTE: This is a candidate for the 7.11 branch.

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
2011-07-12 12:04:57 -04:00
Brian Paul
d739434af8 glx: add a few missing glXChooseFBConfig() attributes
Fixes https://bugs.freedesktop.org/show_bug.cgi?id=38842

NOTE: This is a candidate for the 7.11 branch.
(cherry picked from commit d60880db35)
2011-07-12 09:46:50 -06:00
Brian Paul
72f2bd2a41 glext.h: update to version 71
(cherry picked from commit bb0d5cae00)
2011-07-12 09:46:42 -06:00
Benjamin Franzke
b0549fab5c configure: Require libudev for drm & wayland egl platforms
NOTE: This is a candidate for the 7.11 branch.
(cherry picked from commit 7ed1826e2e)
2011-07-12 09:58:11 +02:00
Benjamin Franzke
ac88916978 configure: Fix typo in gbm check for egl drm platform
NOTE: This is a candidate for the 7.11 branch.
(cherry picked from commit 9b8cd49930)
2011-07-12 09:58:11 +02:00
Benjamin Franzke
336b2c7fbd configure: Enable st/gbm if st/egl has drm platform
NOTE: This is a candidate for the 7.11 branch.
(cherry picked from commit b18b2994ef)
2011-07-12 09:58:11 +02:00
Benjamin Franzke
a8907c6005 egl_dri2: Fix compilation if udev devel files are not installed
NOTE: This is a candidate for the 7.11 branch.
(cherry picked from commit b2d6375e6a)
2011-07-12 09:58:11 +02:00
Benjamin Franzke
89af428aea egl: Fix Terminate with shared gbm screens
NOTE: This is a candidate for the 7.11 branch.
(cherry picked from commit 992680c8b4)
2011-07-12 09:58:11 +02:00
Marek Olšák
8a77029f4c 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 23:03:34 +02:00
Marek Olšák
928bf189ff 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 23:03:25 +02:00
Vadim Girlin
a9e34ada26 r600g: LIT: clamp negative src.y to 0
Fixes https://bugs.freedesktop.org/show_bug.cgi?id=39083

Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
2011-07-10 13:30:20 -04:00
Eric Anholt
804995807d i965/gen4: Fix GPU hangs since the program streaming change.
This was tricky.  We were doing a use-before-initialize of
grf_reg_count, but the value usually got overwritten anyway -- when we
didn't have to do a relocation (typical), or on gen5 when we didn't
have relocations at all.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=38771
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit d03fdc4cde)
2011-07-09 07:52:35 -07:00
Ian Romanick
b033f050fd mesa: Fix the parsers build rule so that 'make tarballs' can work
You'd think that with all the commit messages about adding stuff to
tarballs or fixing 'make tarballs' that someone would have noticed
that it was completely broken for 4 months (3158cc7).
2011-07-08 18:47:21 -07:00
Ian Romanick
c66982f7dc mesa: Bump version to 7.11-rc1 2011-07-08 18:26:39 -07:00
Ian Romanick
530c68d616 glsl: Fix depth unbalancing problem in if-statement flattening
Previously, if max_depth were 1, the following code would see the
first if-statement (correctly) not get flattened, but the second
if-statement would (incorrectly) get flattened:

void main()
{
    if (a)
        gl_Position = vec4(0);

    if (b)
        gl_Position = vec4(1);
}

This is because the visit_leave(ir_if*) method would not decrement the
depth before returning on the first if-statement.

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

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit d2c6cef18a)
2011-07-08 16:02:20 -07:00
Vadim Girlin
576f489dad r600g: introduce r600_bc_src_toggle_neg helper and fix SUB & LRP
SUB & LRP instructions should toggle NEG bit instead of setting it,
otherwise e.g. "SUB a,b,-1" is translated as "ADD a,b,-1"

Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
2011-07-08 17:23:54 -04:00
Vadim Girlin
270de51f1d r600g: introduce r600_bc_src_set_abs helper and fix LOG
LOG instruction should use absolute values of source operand.

Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
2011-07-08 17:23:34 -04:00
Vadim Girlin
57fe695a17 r600g: RSQ: clear NEG for operand
Need to clear NEG bit because it applies after ABS, e.g. "RSQ ..., -1"
uses -|1| as operand.

Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
2011-07-08 17:23:21 -04:00
Vadim Girlin
189303fb30 r600g: LIT: swap MUL_LIT operands to fix 0^0
For 0^0 case result of "LOG_CLAMPED ...,0" is -MAX_FLOAT, and then result of
"MUL_LIT ...,0,-MAX_FLOAT,..." is -MAX_FLOAT instead of 0 because of special
src1 checks for -MAX_FLOAT. So swap src0/1:
"MUL_LIT ...,-MAX_FLOAT,0,..." to get expected 0, then result of
"EXP_IEEE ...,0" is 1 as expected for LIT.

Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
2011-07-08 17:23:07 -04:00
Brian Paul
c4da12e74f glsl: use casts to silence warning
(cherry picked from commit 7eb7d67d50)
2011-07-08 08:03:56 -06:00
Brian Paul
8428b48673 gallivm: Fix build with llvm-3.0
LLVM 3.0svn changes pretty rapidly. The change in
Target->createMCInstPrinter() signature which inspired commits
40ae214067 and
92e29dc5b0 has been reverted.

Signed-off-by: Gustaw Smolarczyk <wielkiegie@gmail.com>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit fc98444bd5)

Conflicts:

	src/gallium/auxiliary/gallivm/lp_bld_debug.cpp
2011-07-08 08:03:40 -06:00
Marek Olšák
b0a4f34ea8 st/mesa: handle float formats in st_format_datatype
NOTE: This is a candidate for the 7.11 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 7de28e80dc)
2011-07-08 13:24:50 +02:00
Marek Olšák
efd0ffd1b0 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 13:24:42 +02:00
Stéphane Marchesin
dc062db95d i915g: Improve flushing using heuristics. 2011-07-08 00:28:29 -07:00
Stéphane Marchesin
b292ef8f88 i915g: Move back to the old method for target format fixup.
Conflicts:

	src/gallium/drivers/i915/i915_state_emit.c
2011-07-08 00:27:59 -07:00
Eric Anholt
d3bfa9bb4a intel: Fix use of freed buffer if glBitmap is called after a swap.
Regions looked up from the framebuffer are invalid after
intel_prepare_render().

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=30266
Tested-by: Thomas Jones <thomas.jones@utoronto.ca>
(cherry picked from commit 066bee64e1)
2011-07-07 15:05:24 -07:00
Paul Berry
98af042079 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 14:23:08 -07:00
Ian Romanick
127bd9d5b6 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)
2011-07-06 17:06:29 -07:00
Ian Romanick
b8972db223 glsl: Don't choke when printing an anonymous function parameter
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: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 174cef7fee)
2011-07-06 17:06:19 -07:00
Ian Romanick
f28cf18609 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 17:05:50 -07:00
Ian Romanick
42cd6192a2 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 17:05:36 -07:00
Vadim Girlin
1ae00c5960 r600g: fix buffer overflow check in r600_query_begin 2011-07-05 16:16:54 -04:00
Vadim Girlin
65d0d69c91 r600g: fix bo map usage flags in r600_query_begin 2011-07-05 16:16:40 -04:00
Vadim Girlin
433afb7352 r600g: reduce flushes for queries 2011-07-05 16:16:23 -04:00
Vadim Girlin
f70c2f8521 r600g: fix buffer offset in r600_query_begin 2011-07-05 16:16:05 -04:00
Chia-I Wu
e4cef07b87 egl: add copyright notices
The list of copyright holders could be incomplete.  Please update
directly or notify me if your name is missing.
(cherry picked from commit f2001df508)
2011-07-02 18:25:28 +09:00
Vadim Girlin
2196feb47f r600g: fix check for empty cs 2011-06-30 16:40:45 -04:00
Chia-I Wu
b90c710c6c target/egl-static: fix a compiler warning
(cherry picked from commit 3e3df5fcd1)
2011-06-30 15:01:17 +09:00
Chia-I Wu
bd1ceb5c5b targets/egl-static: fix library search order
Use

  $(MKLIB) -ldflags '-L$(TOP)/$(LIB_DIR)'

instead of

  $(MKLIB) -L$(TOP)/$(LIB_DIR)

to make sure the local library path appears before system's.
(cherry picked from commit 24137afb31)
2011-06-30 15:01:17 +09:00
Chia-I Wu
567778e49a targets/gbm: attemp to fix unresolved symbols
Move system libraries (usually .so) out of --start-group / --end-group
pair.  Add possiblly missing archives, defines, and shared libraries.
(cherry picked from commit 56ec8e17d3)
2011-06-30 15:01:17 +09:00
Chia-I Wu
0fafcc6919 targets/egl-static: do not use DRI_LIB_DEPS
It brings in libraries that are not necessarily needed.
(cherry picked from commit 1e9f0b1736)
2011-06-30 15:01:17 +09:00
Chia-I Wu
29574af377 egl: fix EGL_MATCH_NATIVE_PIXMAP
EGL_MATCH_NATIVE_PIXMAP is valid for eglChooseConfig, but invalid for
eglGetConfigAttrib.
(cherry picked from commit 8ea5330200)
2011-06-30 15:01:17 +09:00
Chia-I Wu
0eb780262c st/egl: update fbdev backend
Considering fbdev as an in-kernel window system,

 - opening a device opens a connection
 - there is only one window: the framebuffer
 - fb_var_screeninfo decides window position, size, and even color format
 - there is no pixmap

Now EGL is built on top of this window system.  So we should have

 - the fd as the handle of the native display
 - reject all but one native window: NULL
 - no pixmap support

modeset support is still around, but it should be removed soon.
(cherry picked from commit aa281dd392)
2011-06-30 15:01:16 +09:00
Chia-I Wu
e43a096f0c st/d3d1x: fix for st/egl native.h interface change
The interface was changed in 73df31eedd.
(cherry picked from commit 3a07d9594a)
2011-06-30 15:01:16 +09:00
Chia-I Wu
52aa06a2cd st/egl: fix a compile error
It is triggered when --with-driver=xlib is specified.
(cherry picked from commit ed47d65c7c)
2011-06-30 15:01:16 +09:00
Chia-I Wu
5d1561b4ab st/egl: reorganize backend initialization
Remove set_event_handler() and pass the event handler with
native_get_XXX_platform().  Add init_screen() so that the pipe screen is
created later.  This way we don't need to pass user_data to
create_display().
(cherry picked from commit 73df31eedd)
2011-06-30 15:01:16 +09:00
Kenneth Graunke
a8d7f36d65 i965/gen7: Add missing ! to brw->gs.prog_active assertion.
A typo in commit c173541d97 accidentally removed the !.
It's supposed to assert that there is _not_ an active GS program.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=38762

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry-picked from commit 5ddc518401)
2011-06-29 11:00:16 -07:00
Emil Velikov
82ebfa6387 st/mesa: Use correct internal target
Commit 1a339b6c(st/mesa: prefer native texture formats when possible)
introduced two new arguments to the st_choose_format() functions.
This patch fixes the order and passes the correct internal_target
rather than GL_NONE

NOTE: This is a candidate for the 7.11 branch
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 9b5c538726)
2011-06-29 07:20:08 -06:00
Andre Maasikas
ee416c6ffe st/mesa: fix overwriting gl_format with pipe_format since 9d380f48
fixes assert later on in texcompress2/r600g

Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 19789e403c)
2011-06-29 06:59:46 -06:00
Marek Olšák
ebc884d3dd 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)
2011-06-25 20:05:08 +02:00
Brian Paul
9383cfb4ba Revert "Fix 24bpp software rendering"
This reverts commit c0c0bb6cb1.
2011-06-25 06:20:32 -06:00
10986 changed files with 1229559 additions and 3386291 deletions

View File

@@ -1,18 +0,0 @@
((nil . ((show-trailing-whitespace . t)))
(prog-mode
(indent-tabs-mode . nil)
(tab-width . 8)
(c-basic-offset . 3)
(c-file-style . "stroustrup")
(fill-column . 78)
(eval . (progn
(c-set-offset 'case-label '0)
(c-set-offset 'innamespace '0)
(c-set-offset 'inline-open '0)))
(whitespace-style face indentation)
(whitespace-line-column . 79)
(eval ignore-errors
(require 'whitespace)
(whitespace-mode 1)))
(makefile-mode (indent-tabs-mode . t))
)

View File

@@ -1,48 +0,0 @@
# To use this config on you editor, follow the instructions at:
# http://editorconfig.org
root = true
[*]
charset = utf-8
insert_final_newline = true
tab_width = 8
[*.{c,h,cpp,hpp,cc,hh}]
indent_style = space
indent_size = 3
max_line_length = 78
[{Makefile*,*.mk}]
indent_style = tab
[{*.py,SCons*}]
indent_style = space
indent_size = 4
[*.pl]
indent_style = space
indent_size = 4
[*.m4]
indent_style = space
indent_size = 2
[*.yml]
indent_style = space
indent_size = 2
[*.html]
indent_style = space
indent_size = 2
[*.rst]
indent_style = space
indent_size = 3
[*.patch]
trim_trailing_whitespace = false
[{meson.build,meson_options.txt}]
indent_style = space
indent_size = 2

10
.emacs-dirvars Normal file
View File

@@ -0,0 +1,10 @@
;; -*- emacs-lisp -*-
;;
;; This file is processed by the dirvars emacs package. Each variable
;; setting below is performed when this dirvars file is loaded.
;;
indent-tabs-mode: nil
tab-width: 8
c-basic-offset: 3
kde-emacs-after-parent-string: ""
evaluate: (c-set-offset 'inline-open '0)

10
.gitattributes vendored
View File

@@ -1,6 +1,4 @@
*.csv eol=crlf
* text=auto
*.jpg binary
*.png binary
*.gif binary
*.ico binary
*.dsp -crlf
*.dsw -crlf
*.sln -crlf
*.vcproj -crlf

26
.gitignore vendored
View File

@@ -1,4 +1,28 @@
*.a
*.dll
*.exe
*.ilk
*.o
*.obj
*.os
*.pc
*.pdb
*.pyc
*.pyo
*.out
*.so
*.sw[a-z]
*~
depend
depend.bak
lib
lib64
configure
autom4te.cache
aclocal.m4
config.log
config.status
cscope*
.scon*
config.py
build
.dir-locals.el

File diff suppressed because it is too large Load Diff

View File

@@ -1,17 +0,0 @@
# Note: skips lists for CI are just a list of lines that, when
# non-zero-length and not starting with '#', will regex match to
# delete lines from the test list. Be careful.
# These are tremendously slow (pushing toward a minute), and aren't
# reliable to be run in parallel with other tests due to CPU-side timing.
dEQP-GLES[0-9]*.functional.flush_finish.*
# piglit: WGL is Windows-only
wgl@.*
# These are sensitive to CPU timing, and would need to be run in isolation
# on the system rather than in parallel with other tests.
glx@glx_arb_sync_control@timing.*
# This test is not built with waffle, while we do build tests with waffle
spec@!opengl 1.1@windowoverlap

View File

@@ -1,2 +0,0 @@
[*.sh]
indent_size = 2

View File

@@ -1,26 +0,0 @@
#!/bin/sh
# This test script groups together a bunch of fast dEQP variant runs
# to amortize the cost of rebooting the board.
set -ex
EXIT=0
# Run reset tests without parallelism:
if ! env \
DEQP_RESULTS_DIR=results/reset \
FDO_CI_CONCURRENT=1 \
DEQP_CASELIST_FILTER='.*reset.*' \
/install/deqp-runner.sh; then
EXIT=1
fi
# Then run everything else with parallelism:
if ! env \
DEQP_RESULTS_DIR=results/nonrobustness \
DEQP_CASELIST_INV_FILTER='.*reset.*' \
/install/deqp-runner.sh; then
EXIT=1
fi

View File

@@ -1,13 +0,0 @@
#!/bin/sh
# Init entrypoint for bare-metal devices; calls common init code.
# First stage: very basic setup to bring up network and /dev etc
/init-stage1.sh
# Second stage: run jobs
test $? -eq 0 && /init-stage2.sh
# Wait until the job would have timed out anyway, so we don't spew a "init
# exited" panic.
sleep 6000

View File

@@ -1,101 +0,0 @@
#!/bin/bash
# Boot script for Chrome OS devices attached to a servo debug connector, using
# NFS and TFTP to boot.
# We're run from the root of the repo, make a helper var for our paths
BM=$CI_PROJECT_DIR/install/bare-metal
CI_COMMON=$CI_PROJECT_DIR/install/common
# Runner config checks
if [ -z "$BM_SERIAL" ]; then
echo "Must set BM_SERIAL in your gitlab-runner config.toml [[runners]] environment"
echo "This is the CPU serial device."
exit 1
fi
if [ -z "$BM_SERIAL_EC" ]; then
echo "Must set BM_SERIAL in your gitlab-runner config.toml [[runners]] environment"
echo "This is the EC serial device for controlling board power"
exit 1
fi
if [ ! -d /nfs ]; then
echo "NFS rootfs directory needs to be mounted at /nfs by the gitlab runner"
exit 1
fi
if [ ! -d /tftp ]; then
echo "TFTP directory for this board needs to be mounted at /tftp by the gitlab runner"
exit 1
fi
# job config checks
if [ -z "$BM_KERNEL" ]; then
echo "Must set BM_KERNEL to your board's kernel FIT image"
exit 1
fi
if [ -z "$BM_ROOTFS" ]; then
echo "Must set BM_ROOTFS to your board's rootfs directory in the job's variables"
exit 1
fi
if [ -z "$BM_CMDLINE" ]; then
echo "Must set BM_CMDLINE to your board's kernel command line arguments"
exit 1
fi
set -ex
# Clear out any previous run's artifacts.
rm -rf results/
mkdir -p results
# Create the rootfs in the NFS directory. rm to make sure it's in a pristine
# state, since it's volume-mounted on the host.
rsync -a --delete $BM_ROOTFS/ /nfs/
mkdir -p /nfs/results
. $BM/rootfs-setup.sh /nfs
# Put the kernel/dtb image and the boot command line in the tftp directory for
# the board to find. For normal Mesa development, we build the kernel and
# store it in the docker container that this script is running in.
#
# However, container builds are expensive, so when you're hacking on the
# kernel, it's nice to be able to skip the half hour container build and plus
# moving that container to the runner. So, if BM_KERNEL is a URL, fetch it
# instead of looking in the container. Note that the kernel build should be
# the output of:
#
# make Image.lzma
#
# mkimage \
# -A arm64 \
# -f auto \
# -C lzma \
# -d arch/arm64/boot/Image.lzma \
# -b arch/arm64/boot/dts/qcom/sdm845-cheza-r3.dtb \
# cheza-image.img
rm -rf /tftp/*
if echo "$BM_KERNEL" | grep -q http; then
apt install -y wget
wget $BM_KERNEL -O /tftp/vmlinuz
else
cp $BM_KERNEL /tftp/vmlinuz
fi
echo "$BM_CMDLINE" > /tftp/cmdline
set +e
python3 $BM/cros_servo_run.py \
--cpu $BM_SERIAL \
--ec $BM_SERIAL_EC
ret=$?
set -e
# Bring artifacts back from the NFS dir to the build dir where gitlab-runner
# will look for them.
cp -Rp /nfs/results/. results/
exit $ret

View File

@@ -1,194 +0,0 @@
#!/usr/bin/env python3
#
# Copyright © 2020 Google LLC
#
# Permission is hereby granted, free of charge, to any person obtaining a
# copy of this software and associated documentation files (the "Software"),
# to deal in the Software without restriction, including without limitation
# the rights to use, copy, modify, merge, publish, distribute, sublicense,
# and/or sell copies of the Software, and to permit persons to whom the
# Software is furnished to do so, subject to the following conditions:
#
# The above copyright notice and this permission notice (including the next
# paragraph) shall be included in all copies or substantial portions of the
# Software.
#
# THE SOFTWARE IS 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 SOFTWARE OR THE USE OR OTHER DEALINGS
# IN THE SOFTWARE.
import argparse
import queue
import re
from serial_buffer import SerialBuffer
import sys
import threading
class CrosServoRun:
def __init__(self, cpu, ec):
# Merged FIFO for the two serial buffers, fed by threads.
self.serial_queue = queue.Queue()
self.sentinel = object()
self.threads_done = 0
self.ec_ser = SerialBuffer(
ec, "results/serial-ec.txt", "R SERIAL-EC> ")
self.cpu_ser = SerialBuffer(
cpu, "results/serial.txt", "R SERIAL-CPU> ")
self.iter_feed_ec = threading.Thread(
target=self.iter_feed_queue, daemon=True, args=(self.ec_ser.lines(),))
self.iter_feed_ec.start()
self.iter_feed_cpu = threading.Thread(
target=self.iter_feed_queue, daemon=True, args=(self.cpu_ser.lines(),))
self.iter_feed_cpu.start()
def close(self):
self.ec_ser.close()
self.cpu_ser.close()
self.iter_feed_ec.join()
self.iter_feed_cpu.join()
# Feed lines from our serial queues into the merged queue, marking when our
# input is done.
def iter_feed_queue(self, it):
for i in it:
self.serial_queue.put(i)
self.serial_queue.put(self.sentinel)
# Return the next line from the queue, counting how many threads have
# terminated and joining when done
def get_serial_queue_line(self):
line = self.serial_queue.get()
if line == self.sentinel:
self.threads_done = self.threads_done + 1
if self.threads_done == 2:
self.iter_feed_cpu.join()
self.iter_feed_ec.join()
return line
# Returns an iterator for getting the next line.
def serial_queue_lines(self):
return iter(self.get_serial_queue_line, self.sentinel)
def ec_write(self, s):
print("W SERIAL-EC> %s" % s)
self.ec_ser.serial.write(s.encode())
def cpu_write(self, s):
print("W SERIAL-CPU> %s" % s)
self.cpu_ser.serial.write(s.encode())
def print_error(self, message):
RED = '\033[0;31m'
NO_COLOR = '\033[0m'
print(RED + message + NO_COLOR)
def run(self):
# Flush any partial commands in the EC's prompt, then ask for a reboot.
self.ec_write("\n")
self.ec_write("reboot\n")
# This is emitted right when the bootloader pauses to check for input.
# Emit a ^N character to request network boot, because we don't have a
# direct-to-netboot firmware on cheza.
for line in self.serial_queue_lines():
if re.search("load_archive: loading locale_en.bin", line):
self.cpu_write("\016")
break
# The Cheza boards have issues with failing to bring up power to
# the system sometimes, possibly dependent on ambient temperature
# in the farm.
if re.search("POWER_GOOD not seen in time", line):
self.print_error("Detected intermittent poweron failure, restarting run...")
return 2
tftp_failures = 0
for line in self.serial_queue_lines():
if re.search("---. end Kernel panic", line):
return 1
# The Cheza firmware seems to occasionally get stuck looping in
# this error state during TFTP booting, possibly based on amount of
# network traffic around it, but it'll usually recover after a
# reboot.
if re.search("R8152: Bulk read error 0xffffffbf", line):
tftp_failures += 1
if tftp_failures >= 100:
self.print_error("Detected intermittent tftp failure, restarting run...")
return 2
# There are very infrequent bus errors during power management transitions
# on cheza, which we don't expect to be the case on future boards.
if re.search("Kernel panic - not syncing: Asynchronous SError Interrupt", line):
self.print_error("Detected cheza power management bus error, restarting run...")
return 2
# If the network device dies, it's probably not graphics's fault, just try again.
if re.search("NETDEV WATCHDOG", line):
self.print_error(
"Detected network device failure, restarting run...")
return 2
# These HFI response errors started appearing with the introduction
# of piglit runs. CosmicPenguin says:
#
# "message ID 106 isn't a thing, so likely what happened is that we
# got confused when parsing the HFI queue. If it happened on only
# one run, then memory corruption could be a possible clue"
#
# Given that it seems to trigger randomly near a GPU fault and then
# break many tests after that, just restart the whole run.
if re.search("a6xx_hfi_send_msg.*Unexpected message id .* on the response queue", line):
self.print_error("Detected cheza power management bus error, restarting run...")
return 2
if re.search("coreboot.*bootblock starting", line):
self.print_error(
"Detected spontaneous reboot, restarting run...")
return 2
result = re.search("hwci: mesa: (\S*)", line)
if result:
if result.group(1) == "pass":
return 0
else:
return 1
self.print_error("Reached the end of the CPU serial log without finding a result")
return 1
def main():
parser = argparse.ArgumentParser()
parser.add_argument('--cpu', type=str,
help='CPU Serial device', required=True)
parser.add_argument(
'--ec', type=str, help='EC Serial device', required=True)
args = parser.parse_args()
servo = CrosServoRun(args.cpu, args.ec)
while True:
retval = servo.run()
if retval != 2:
break
# power down the CPU on the device
servo.ec_write("power off\n")
servo.close()
sys.exit(retval)
if __name__ == '__main__':
main()

View File

@@ -1,10 +0,0 @@
#!/bin/bash
relay=$1
if [ -z "$relay" ]; then
echo "Must supply a relay arg"
exit 1
fi
$CI_PROJECT_DIR/install/bare-metal/eth008-power-relay.py $ETH_HOST $ETH_PORT off $relay

View File

@@ -1,28 +0,0 @@
#!/usr/bin/python3
import sys
import socket
host = sys.argv[1]
port = sys.argv[2]
mode = sys.argv[3]
relay = sys.argv[4]
msg = None
if mode == "on":
msg = b'\x20'
else:
msg = b'\x21'
msg += int(relay).to_bytes(1, 'big')
msg += b'\x00'
c = socket.create_connection((host, int(port)))
c.sendall(msg)
data = c.recv(1)
c.close()
if data[0] == b'\x01':
print('Command failed')
sys.exit(1)

View File

@@ -1,12 +0,0 @@
#!/bin/bash
relay=$1
if [ -z "$relay" ]; then
echo "Must supply a relay arg"
exit 1
fi
$CI_PROJECT_DIR/install/bare-metal/eth008-power-relay.py $ETH_HOST $ETH_PORT off $relay
sleep 5
$CI_PROJECT_DIR/install/bare-metal/eth008-power-relay.py $ETH_HOST $ETH_PORT on $relay

View File

@@ -1,30 +0,0 @@
#!/bin/bash
set -e
STRINGS=$(mktemp)
ERRORS=$(mktemp)
trap "rm $STRINGS; rm $ERRORS;" EXIT
FILE=$1
shift 1
while getopts "f:e:" opt; do
case $opt in
f) echo "$OPTARG" >> $STRINGS;;
e) echo "$OPTARG" >> $STRINGS ; echo "$OPTARG" >> $ERRORS;;
esac
done
shift $((OPTIND -1))
echo "Waiting for $FILE to say one of following strings"
cat $STRINGS
while ! egrep -wf $STRINGS $FILE; do
sleep 2
done
if egrep -wf $ERRORS $FILE; then
exit 1
fi

View File

@@ -1,148 +0,0 @@
#!/bin/bash
BM=$CI_PROJECT_DIR/install/bare-metal
CI_COMMON=$CI_PROJECT_DIR/install/common
if [ -z "$BM_SERIAL" -a -z "$BM_SERIAL_SCRIPT" ]; then
echo "Must set BM_SERIAL OR BM_SERIAL_SCRIPT in your gitlab-runner config.toml [[runners]] environment"
echo "BM_SERIAL:"
echo " This is the serial device to talk to for waiting for fastboot to be ready and logging from the kernel."
echo "BM_SERIAL_SCRIPT:"
echo " This is a shell script to talk to for waiting for fastboot to be ready and logging from the kernel."
exit 1
fi
if [ -z "$BM_POWERUP" ]; then
echo "Must set BM_POWERUP in your gitlab-runner config.toml [[runners]] environment"
echo "This is a shell script that should reset the device and begin its boot sequence"
echo "such that it pauses at fastboot."
exit 1
fi
if [ -z "$BM_POWERDOWN" ]; then
echo "Must set BM_POWERDOWN in your gitlab-runner config.toml [[runners]] environment"
echo "This is a shell script that should power off the device."
exit 1
fi
if [ -z "$BM_FASTBOOT_SERIAL" ]; then
echo "Must set BM_FASTBOOT_SERIAL in your gitlab-runner config.toml [[runners]] environment"
echo "This must be the a stable-across-resets fastboot serial number."
exit 1
fi
if [ -z "$BM_KERNEL" ]; then
echo "Must set BM_KERNEL to your board's kernel vmlinuz or Image.gz in the job's variables:"
exit 1
fi
if [ -z "$BM_DTB" ]; then
echo "Must set BM_DTB to your board's DTB file in the job's variables:"
exit 1
fi
if [ -z "$BM_ROOTFS" ]; then
echo "Must set BM_ROOTFS to your board's rootfs directory in the job's variables:"
exit 1
fi
if echo $BM_CMDLINE | grep -q "root=/dev/nfs"; then
BM_FASTBOOT_NFSROOT=1
fi
set -ex
# Clear out any previous run's artifacts.
rm -rf results/
mkdir -p results/
if [ -n "$BM_FASTBOOT_NFSROOT" ]; then
# Create the rootfs in the NFS directory. rm to make sure it's in a pristine
# state, since it's volume-mounted on the host.
rsync -a --delete $BM_ROOTFS/ /nfs/
mkdir -p /nfs/results
. $BM/rootfs-setup.sh /nfs
# Root on NFS, no need for an inintramfs.
rm -f rootfs.cpio.gz
touch rootfs.cpio
gzip rootfs.cpio
else
# Create the rootfs in a temp dir
rsync -a --delete $BM_ROOTFS/ rootfs/
. $BM/rootfs-setup.sh rootfs
# Finally, pack it up into a cpio rootfs. Skip the vulkan CTS since none of
# these devices use it and it would take up space in the initrd.
if [ -n "$PIGLIT_PROFILES" ]; then
EXCLUDE_FILTER="deqp|arb_gpu_shader5|arb_gpu_shader_fp64|arb_gpu_shader_int64|glsl-4.[0123456]0|arb_tessellation_shader"
else
EXCLUDE_FILTER="piglit|python"
fi
pushd rootfs
find -H | \
egrep -v "external/(openglcts|vulkancts|amber|glslang|spirv-tools)" |
egrep -v "traces-db|apitrace|renderdoc" | \
egrep -v $EXCLUDE_FILTER | \
cpio -H newc -o | \
xz --check=crc32 -T4 - > $CI_PROJECT_DIR/rootfs.cpio.gz
popd
fi
# Make the combined kernel image and dtb for passing to fastboot. For normal
# Mesa development, we build the kernel and store it in the docker container
# that this script is running in.
#
# However, container builds are expensive, so when you're hacking on the
# kernel, it's nice to be able to skip the half hour container build and plus
# moving that container to the runner. So, if BM_KERNEL+BM_DTB are URLs,
# fetch them instead of looking in the container.
if echo "$BM_KERNEL $BM_DTB" | grep -q http; then
apt install -y wget
wget $BM_KERNEL -O kernel
wget $BM_DTB -O dtb
cat kernel dtb > Image.gz-dtb
rm kernel dtb
else
cat $BM_KERNEL $BM_DTB > Image.gz-dtb
fi
mkdir -p artifacts
abootimg \
--create artifacts/fastboot.img \
-k Image.gz-dtb \
-r rootfs.cpio.gz \
-c cmdline="$BM_CMDLINE"
rm Image.gz-dtb
export PATH=$BM:$PATH
# Start background command for talking to serial if we have one.
if [ -n "$BM_SERIAL_SCRIPT" ]; then
$BM_SERIAL_SCRIPT > results/serial-output.txt &
while [ ! -e results/serial-output.txt ]; do
sleep 1
done
fi
set +e
$BM/fastboot_run.py \
--dev="$BM_SERIAL" \
--fbserial="$BM_FASTBOOT_SERIAL" \
--powerup="$BM_POWERUP" \
--powerdown="$BM_POWERDOWN"
ret=$?
set -e
if [ -n "$BM_FASTBOOT_NFSROOT" ]; then
# Bring artifacts back from the NFS dir to the build dir where gitlab-runner
# will look for them.
cp -Rp /nfs/results/. results/
fi
exit $ret

View File

@@ -1,146 +0,0 @@
#!/usr/bin/env python3
#
# Copyright © 2020 Google LLC
#
# Permission is hereby granted, free of charge, to any person obtaining a
# copy of this software and associated documentation files (the "Software"),
# to deal in the Software without restriction, including without limitation
# the rights to use, copy, modify, merge, publish, distribute, sublicense,
# and/or sell copies of the Software, and to permit persons to whom the
# Software is furnished to do so, subject to the following conditions:
#
# The above copyright notice and this permission notice (including the next
# paragraph) shall be included in all copies or substantial portions of the
# Software.
#
# THE SOFTWARE IS 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 SOFTWARE OR THE USE OR OTHER DEALINGS
# IN THE SOFTWARE.
import argparse
import os
import re
from serial_buffer import SerialBuffer
import sys
import threading
class FastbootRun:
def __init__(self, args):
self.powerup = args.powerup
# We would like something like a 1 minute timeout, but the piglit traces
# jobs stall out for long periods of time.
self.ser = SerialBuffer(args.dev, "results/serial-output.txt", "R SERIAL> ", timeout=600)
self.fastboot="fastboot boot -s {ser} artifacts/fastboot.img".format(ser=args.fbserial)
def close(self):
self.ser.close()
def print_error(self, message):
RED = '\033[0;31m'
NO_COLOR = '\033[0m'
print(RED + message + NO_COLOR)
def logged_system(self, cmd):
print("Running '{}'".format(cmd))
return os.system(cmd)
def run(self):
if self.logged_system(self.powerup) != 0:
return 1
fastboot_ready = False
for line in self.ser.lines():
if re.search("fastboot: processing commands", line) or \
re.search("Listening for fastboot command on", line):
fastboot_ready = True
break
if re.search("data abort", line):
self.print_error("Detected crash during boot, restarting run...")
return 2
if not fastboot_ready:
self.print_error("Failed to get to fastboot prompt, restarting run...")
return 2
if self.logged_system(self.fastboot) != 0:
return 1
print_more_lines = -1
for line in self.ser.lines():
if print_more_lines == 0:
return 2
if print_more_lines > 0:
print_more_lines -= 1
if re.search("---. end Kernel panic", line):
return 1
# The db820c boards intermittently reboot. Just restart the run
# when if we see a reboot after we got past fastboot.
if re.search("PON REASON", line):
self.print_error("Detected spontaneous reboot, restarting run...")
return 2
# db820c sometimes wedges around iommu fault recovery
if re.search("watchdog: BUG: soft lockup - CPU.* stuck", line):
self.print_error(
"Detected kernel soft lockup, restarting run...")
return 2
# If the network device dies, it's probably not graphics's fault, just try again.
if re.search("NETDEV WATCHDOG", line):
self.print_error(
"Detected network device failure, restarting run...")
return 2
# A3xx recovery doesn't quite work. Sometimes the GPU will get
# wedged and recovery will fail (because power can't be reset?)
# This assumes that the jobs are sufficiently well-tested that GPU
# hangs aren't always triggered, so just try again. But print some
# more lines first so that we get better information on the cause
# of the hang. Once a hang happens, it's pretty chatty.
if "[drm:adreno_recover] *ERROR* gpu hw init failed: -22" in line:
self.print_error(
"Detected GPU hang, restarting run...")
if print_more_lines == -1:
print_more_lines = 30
result = re.search("hwci: mesa: (\S*)", line)
if result:
if result.group(1) == "pass":
return 0
else:
return 1
self.print_error("Reached the end of the CPU serial log without finding a result, restarting run...")
return 2
def main():
parser = argparse.ArgumentParser()
parser.add_argument('--dev', type=str, help='Serial device (otherwise reading from serial-output.txt)')
parser.add_argument('--powerup', type=str, help='shell command for rebooting', required=True)
parser.add_argument('--powerdown', type=str, help='shell command for powering off', required=True)
parser.add_argument('--fbserial', type=str, help='fastboot serial number of the board', required=True)
args = parser.parse_args()
fastboot = FastbootRun(args)
while True:
retval = fastboot.run()
fastboot.close()
if retval != 2:
break
fastboot = FastbootRun(args)
fastboot.logged_system(args.powerdown)
sys.exit(retval)
if __name__ == '__main__':
main()

View File

@@ -1,10 +0,0 @@
#!/bin/bash
relay=$1
if [ -z "$relay" ]; then
echo "Must supply a relay arg"
exit 1
fi
$CI_PROJECT_DIR/install/bare-metal/google-power-relay.py off $relay

View File

@@ -1,19 +0,0 @@
#!/usr/bin/python3
import sys
import serial
mode = sys.argv[1]
relay = sys.argv[2]
# our relays are "off" means "board is powered".
mode_swap = {
"on" : "off",
"off" : "on",
}
mode = mode_swap[mode]
ser = serial.Serial('/dev/ttyACM0', 115200, timeout=2)
command = "relay {} {}\n\r".format(mode, relay)
ser.write(command.encode())
ser.close()

View File

@@ -1,12 +0,0 @@
#!/bin/bash
relay=$1
if [ -z "$relay" ]; then
echo "Must supply a relay arg"
exit 1
fi
$CI_PROJECT_DIR/install/bare-metal/google-power-relay.py off $relay
sleep 5
$CI_PROJECT_DIR/install/bare-metal/google-power-relay.py on $relay

View File

@@ -1,17 +0,0 @@
#!/bin/bash
if [ -z "$BM_POE_INTERFACE" ]; then
echo "Must supply the PoE Interface to power up"
exit 1
fi
if [ -z "$BM_POE_ADDRESS" ]; then
echo "Must supply the PoE Switch host"
exit 1
fi
SNMP_KEY="SNMPv2-SMI::mib-2.105.1.1.1.3.1.`expr 48 + $BM_POE_INTERFACE`"
SNMP_ON="i 1"
SNMP_OFF="i 2"
flock /var/run/poe.lock -c "snmpset -v2c -r 3 -t 30 -cmesaci $BM_POE_ADDRESS $SNMP_KEY $SNMP_OFF"

View File

@@ -1,19 +0,0 @@
#!/bin/bash
if [ -z "$BM_POE_INTERFACE" ]; then
echo "Must supply the PoE Interface to power up"
exit 1
fi
if [ -z "$BM_POE_ADDRESS" ]; then
echo "Must supply the PoE Switch host"
exit 1
fi
SNMP_KEY="SNMPv2-SMI::mib-2.105.1.1.1.3.1.`expr 48 + $BM_POE_INTERFACE`"
SNMP_ON="i 1"
SNMP_OFF="i 2"
flock /var/run/poe.lock -c "snmpset -v2c -r 3 -t 30 -cmesaci $BM_POE_ADDRESS $SNMP_KEY $SNMP_OFF"
sleep 3s
flock /var/run/poe.lock -c "snmpset -v2c -r 3 -t 30 -cmesaci $BM_POE_ADDRESS $SNMP_KEY $SNMP_ON"

View File

@@ -1,147 +0,0 @@
#!/bin/bash
# Boot script for devices attached to a PoE switch, using NFS for the root
# filesystem.
# We're run from the root of the repo, make a helper var for our paths
BM=$CI_PROJECT_DIR/install/bare-metal
CI_COMMON=$CI_PROJECT_DIR/install/common
# Runner config checks
if [ -z "$BM_SERIAL" ]; then
echo "Must set BM_SERIAL in your gitlab-runner config.toml [[runners]] environment"
echo "This is the serial port to listen the device."
exit 1
fi
if [ -z "$BM_POE_ADDRESS" ]; then
echo "Must set BM_POE_ADDRESS in your gitlab-runner config.toml [[runners]] environment"
echo "This is the PoE switch address to connect for powering up/down devices."
exit 1
fi
if [ -z "$BM_POE_USERNAME" ]; then
echo "Must set BM_POE_USERNAME in your gitlab-runner config.toml [[runners]] environment"
echo "This is the PoE switch username."
exit 1
fi
if [ -z "$BM_POE_PASSWORD" ]; then
echo "Must set BM_POE_PASSWORD in your gitlab-runner config.toml [[runners]] environment"
echo "This is the PoE switch password."
exit 1
fi
if [ -z "$BM_POE_INTERFACE" ]; then
echo "Must set BM_POE_INTERFACE in your gitlab-runner config.toml [[runners]] environment"
echo "This is the PoE switch interface where the device is connected."
exit 1
fi
if [ -z "$BM_POWERUP" ]; then
echo "Must set BM_POWERUP in your gitlab-runner config.toml [[runners]] environment"
echo "This is a shell script that should power up the device and begin its boot sequence."
exit 1
fi
if [ -z "$BM_POWERDOWN" ]; then
echo "Must set BM_POWERDOWN in your gitlab-runner config.toml [[runners]] environment"
echo "This is a shell script that should power off the device."
exit 1
fi
if [ ! -d /nfs ]; then
echo "NFS rootfs directory needs to be mounted at /nfs by the gitlab runner"
exit 1
fi
if [ ! -d /tftp ]; then
echo "TFTP directory for this board needs to be mounted at /tftp by the gitlab runner"
exit 1
fi
# job config checks
if [ -z "$BM_ROOTFS" ]; then
echo "Must set BM_ROOTFS to your board's rootfs directory in the job's variables"
exit 1
fi
if [ -z "$BM_BOOTFS" ]; then
echo "Must set /boot files for the TFTP boot in the job's variables"
exit 1
fi
if [ -z "$BM_CMDLINE" ]; then
echo "Must set BM_CMDLINE to your board's kernel command line arguments"
exit 1
fi
if [ -z "$BM_BOOTCONFIG" ]; then
echo "Must set BM_BOOTCONFIG to your board's required boot configuration arguments"
exit 1
fi
set -ex
# Clear out any previous run's artifacts.
rm -rf results/
mkdir -p results
# Create the rootfs in the NFS directory. rm to make sure it's in a pristine
# state, since it's volume-mounted on the host.
rsync -a --delete $BM_ROOTFS/ /nfs/
# If BM_BOOTFS is an URL, download it
if echo $BM_BOOTFS | grep -q http; then
apt install -y wget
wget ${FDO_HTTP_CACHE_URI:-}$BM_BOOTFS -O /tmp/bootfs.tar
BM_BOOTFS=/tmp/bootfs.tar
fi
# If BM_BOOTFS is a file, assume it is a tarball and uncompress it
if [ -f $BM_BOOTFS ]; then
mkdir -p /tmp/bootfs
tar xf $BM_BOOTFS -C /tmp/bootfs
BM_BOOTFS=/tmp/bootfs
fi
# Install kernel modules (it could be either in /lib/modules or
# /usr/lib/modules, but we want to install in the latter)
[ -d $BM_BOOTFS/usr/lib/modules ] && rsync -a --delete $BM_BOOTFS/usr/lib/modules/ /nfs/usr/lib/modules/
[ -d $BM_BOOTFS/lib/modules ] && rsync -a --delete $BM_BOOTFS/lib/modules/ /nfs/usr/lib/modules/
# Install kernel image + bootloader files
rsync -a --delete $BM_BOOTFS/boot/ /tftp/
# Create the rootfs in the NFS directory
mkdir -p /nfs/results
. $BM/rootfs-setup.sh /nfs
echo "$BM_CMDLINE" > /tftp/cmdline.txt
# Add some required options in config.txt
printf "$BM_BOOTCONFIG" >> /tftp/config.txt
set +e
ATTEMPTS=2
while [ $((ATTEMPTS--)) -gt 0 ]; do
python3 $BM/poe_run.py \
--dev="$BM_SERIAL" \
--powerup="$BM_POWERUP" \
--powerdown="$BM_POWERDOWN" \
--timeout="${BM_POE_TIMEOUT:-60}"
ret=$?
if [ $ret -eq 2 ]; then
echo "Did not detect boot sequence, retrying..."
else
ATTEMPTS=0
fi
done
set -e
# Bring artifacts back from the NFS dir to the build dir where gitlab-runner
# will look for them.
cp -Rp /nfs/results/. results/
exit $ret

View File

@@ -1,96 +0,0 @@
#!/usr/bin/env python3
#
# Copyright © 2020 Igalia, S.L.
#
# Permission is hereby granted, free of charge, to any person obtaining a
# copy of this software and associated documentation files (the "Software"),
# to deal in the Software without restriction, including without limitation
# the rights to use, copy, modify, merge, publish, distribute, sublicense,
# and/or sell copies of the Software, and to permit persons to whom the
# Software is furnished to do so, subject to the following conditions:
#
# The above copyright notice and this permission notice (including the next
# paragraph) shall be included in all copies or substantial portions of the
# Software.
#
# THE SOFTWARE IS 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 SOFTWARE OR THE USE OR OTHER DEALINGS
# IN THE SOFTWARE.
import argparse
import os
import re
from serial_buffer import SerialBuffer
import sys
import threading
class PoERun:
def __init__(self, args):
self.powerup = args.powerup
self.powerdown = args.powerdown
self.ser = SerialBuffer(args.dev, "results/serial-output.txt", "", args.timeout)
def print_error(self, message):
RED = '\033[0;31m'
NO_COLOR = '\033[0m'
print(RED + message + NO_COLOR)
def logged_system(self, cmd):
print("Running '{}'".format(cmd))
return os.system(cmd)
def run(self):
if self.logged_system(self.powerup) != 0:
return 1
boot_detected = False
for line in self.ser.lines():
if re.search("Booting Linux", line):
boot_detected = True
break
if not boot_detected:
self.print_error("Something wrong; couldn't detect the boot start up sequence")
return 2
for line in self.ser.lines():
if re.search("---. end Kernel panic", line):
return 1
# Binning memory problems
if re.search("binner overflow mem", line):
self.print_error("Memory overflow in the binner; GPU hang")
return 1
result = re.search("hwci: mesa: (\S*)", line)
if result:
if result.group(1) == "pass":
return 0
else:
return 1
self.print_error("Reached the end of the CPU serial log without finding a result")
return 2
def main():
parser = argparse.ArgumentParser()
parser.add_argument('--dev', type=str, help='Serial device to monitor', required=True)
parser.add_argument('--powerup', type=str, help='shell command for rebooting', required=True)
parser.add_argument('--powerdown', type=str, help='shell command for powering off', required=True)
parser.add_argument('--timeout', type=int, default=60,
help='time in seconds to wait for activity', required=False)
args = parser.parse_args()
poe = PoERun(args)
retval = poe.run()
poe.logged_system(args.powerdown)
sys.exit(retval)
if __name__ == '__main__':
main()

View File

@@ -1,29 +0,0 @@
#!/bin/bash
rootfs_dst=$1
mkdir -p $rootfs_dst/results
# Set up the init script that brings up the system.
cp $BM/bm-init.sh $rootfs_dst/init
cp $CI_COMMON/init*.sh $rootfs_dst/
# Make JWT token available as file in the bare-metal storage to enable access
# to MinIO
cp "${CI_JOB_JWT_FILE}" "${rootfs_dst}${CI_JOB_JWT_FILE}"
cp $CI_COMMON/capture-devcoredump.sh $rootfs_dst/
set +x
# Pass through relevant env vars from the gitlab job to the baremetal init script
"$CI_COMMON"/generate-env.sh > $rootfs_dst/set-job-env-vars.sh
chmod +x $rootfs_dst/set-job-env-vars.sh
echo "Variables passed through:"
cat $rootfs_dst/set-job-env-vars.sh
set -x
# Add the Mesa drivers we built, and make a consistent symlink to them.
mkdir -p $rootfs_dst/$CI_PROJECT_DIR
rsync -aH --delete $CI_PROJECT_DIR/install/ $rootfs_dst/$CI_PROJECT_DIR/install/

View File

@@ -1,162 +0,0 @@
#!/usr/bin/env python3
#
# Copyright © 2020 Google LLC
#
# Permission is hereby granted, free of charge, to any person obtaining a
# copy of this software and associated documentation files (the "Software"),
# to deal in the Software without restriction, including without limitation
# the rights to use, copy, modify, merge, publish, distribute, sublicense,
# and/or sell copies of the Software, and to permit persons to whom the
# Software is furnished to do so, subject to the following conditions:
#
# The above copyright notice and this permission notice (including the next
# paragraph) shall be included in all copies or substantial portions of the
# Software.
#
# THE SOFTWARE IS 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 SOFTWARE OR THE USE OR OTHER DEALINGS
# IN THE SOFTWARE.
import argparse
from datetime import datetime, timezone
import queue
import serial
import threading
import time
class SerialBuffer:
def __init__(self, dev, filename, prefix, timeout = None):
self.filename = filename
self.dev = dev
if dev:
self.f = open(filename, "wb+")
self.serial = serial.Serial(dev, 115200, timeout=timeout)
else:
self.f = open(filename, "rb")
self.serial = None
self.byte_queue = queue.Queue()
self.line_queue = queue.Queue()
self.prefix = prefix
self.timeout = timeout
self.sentinel = object()
self.closing = False
if self.dev:
self.read_thread = threading.Thread(
target=self.serial_read_thread_loop, daemon=True)
else:
self.read_thread = threading.Thread(
target=self.serial_file_read_thread_loop, daemon=True)
self.read_thread.start()
self.lines_thread = threading.Thread(
target=self.serial_lines_thread_loop, daemon=True)
self.lines_thread.start()
def close(self):
self.closing = True
if self.serial:
self.serial.cancel_read()
self.read_thread.join()
self.lines_thread.join()
if self.serial:
self.serial.close()
# Thread that just reads the bytes from the serial device to try to keep from
# buffer overflowing it. If nothing is received in 1 minute, it finalizes.
def serial_read_thread_loop(self):
greet = "Serial thread reading from %s\n" % self.dev
self.byte_queue.put(greet.encode())
while not self.closing:
try:
b = self.serial.read()
if len(b) == 0:
break
self.byte_queue.put(b)
except Exception as err:
print(self.prefix + str(err))
break
self.byte_queue.put(self.sentinel)
# Thread that just reads the bytes from the file of serial output that some
# other process is appending to.
def serial_file_read_thread_loop(self):
greet = "Serial thread reading from %s\n" % self.filename
self.byte_queue.put(greet.encode())
while not self.closing:
line = self.f.readline()
if line:
self.byte_queue.put(line)
else:
time.sleep(0.1)
self.byte_queue.put(self.sentinel)
# Thread that processes the stream of bytes to 1) log to stdout, 2) log to
# file, 3) add to the queue of lines to be read by program logic
def serial_lines_thread_loop(self):
line = bytearray()
while True:
bytes = self.byte_queue.get(block=True)
if bytes == self.sentinel:
self.read_thread.join()
self.line_queue.put(self.sentinel)
break
# Write our data to the output file if we're the ones reading from
# the serial device
if self.dev:
self.f.write(bytes)
self.f.flush()
for b in bytes:
line.append(b)
if b == b'\n'[0]:
line = line.decode(errors="replace")
time = datetime.now().strftime('%y-%m-%d %H:%M:%S')
print("{endc}{time} {prefix}{line}".format(
time=time, prefix=self.prefix, line=line, endc='\033[0m'), flush=True, end='')
self.line_queue.put(line)
line = bytearray()
def get_line(self):
line = self.line_queue.get()
if line == self.sentinel:
self.lines_thread.join()
return line
def lines(self):
return iter(self.get_line, self.sentinel)
def main():
parser = argparse.ArgumentParser()
parser.add_argument('--dev', type=str, help='Serial device')
parser.add_argument('--file', type=str,
help='Filename for serial output', required=True)
parser.add_argument('--prefix', type=str,
help='Prefix for logging serial to stdout', nargs='?')
args = parser.parse_args()
ser = SerialBuffer(args.dev, args.file, args.prefix or "")
for line in ser.lines():
# We're just using this as a logger, so eat the produced lines and drop
# them
pass
if __name__ == '__main__':
main()

View File

@@ -1,41 +0,0 @@
#!/usr/bin/python3
# Copyright © 2020 Christian Gmeiner
#
# Permission is hereby granted, free of charge, to any person obtaining a
# copy of this software and associated documentation files (the "Software"),
# to deal in the Software without restriction, including without limitation
# the rights to use, copy, modify, merge, publish, distribute, sublicense,
# and/or sell copies of the Software, and to permit persons to whom the
# Software is furnished to do so, subject to the following conditions:
#
# The above copyright notice and this permission notice (including the next
# paragraph) shall be included in all copies or substantial portions of the
# Software.
#
# THE SOFTWARE IS 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 SOFTWARE OR THE USE OR OTHER DEALINGS
# IN THE SOFTWARE.
#
# Tiny script to read bytes from telnet, and write the output to stdout, with a
# buffer in between so we don't lose serial output from its buffer.
#
import sys
import telnetlib
host=sys.argv[1]
port=sys.argv[2]
tn = telnetlib.Telnet(host, port, 1000000)
while True:
bytes = tn.read_some()
sys.stdout.buffer.write(bytes)
sys.stdout.flush()
tn.close()

View File

@@ -1,14 +0,0 @@
#!/bin/sh
while true; do
devcds=`find /sys/devices/virtual/devcoredump/ -name data 2>/dev/null`
for i in $devcds; do
echo "Found a devcoredump at $i."
if cp $i /results/first.devcore; then
echo 1 > $i
echo "Saved to the job artifacts at /first.devcore"
exit 0
fi
done
sleep 10
done

View File

@@ -1,111 +0,0 @@
#!/bin/bash
for var in \
ASAN_OPTIONS \
BASE_SYSTEM_FORK_HOST_PREFIX \
BASE_SYSTEM_MAINLINE_HOST_PREFIX \
CI_COMMIT_BRANCH \
CI_COMMIT_REF_NAME \
CI_COMMIT_TITLE \
CI_JOB_ID \
CI_JOB_JWT_FILE \
CI_JOB_NAME \
CI_JOB_URL \
CI_MERGE_REQUEST_SOURCE_BRANCH_NAME \
CI_MERGE_REQUEST_TITLE \
CI_NODE_INDEX \
CI_NODE_TOTAL \
CI_PAGES_DOMAIN \
CI_PIPELINE_ID \
CI_PIPELINE_URL \
CI_PROJECT_DIR \
CI_PROJECT_NAME \
CI_PROJECT_PATH \
CI_PROJECT_ROOT_NAMESPACE \
CI_RUNNER_DESCRIPTION \
CI_SERVER_URL \
CROSVM_GALLIUM_DRIVER \
CROSVM_GPU_ARGS \
CROSVM_TEST_SCRIPT \
DEQP_BIN_DIR \
DEQP_CASELIST_FILTER \
DEQP_CASELIST_INV_FILTER \
DEQP_CONFIG \
DEQP_EXPECTED_RENDERER \
DEQP_FRACTION \
DEQP_HEIGHT \
DEQP_RESULTS_DIR \
DEQP_RUNNER_OPTIONS \
DEQP_SUITE \
DEQP_TEMP_DIR \
DEQP_VARIANT \
DEQP_VER \
DEQP_WIDTH \
DEVICE_NAME \
DRIVER_NAME \
EGL_PLATFORM \
ETNA_MESA_DEBUG \
FDO_CI_CONCURRENT \
FDO_UPSTREAM_REPO \
FD_MESA_DEBUG \
FLAKES_CHANNEL \
GALLIUM_DRIVER \
GALLIVM_PERF \
GPU_VERSION \
GTEST \
GTEST_FAILS \
GTEST_FRACTION \
GTEST_RESULTS_DIR \
GTEST_RUNNER_OPTIONS \
GTEST_SKIPS \
HWCI_FREQ_MAX \
HWCI_KERNEL_MODULES \
HWCI_START_XORG \
HWCI_TEST_SCRIPT \
IR3_SHADER_DEBUG \
JOB_ARTIFACTS_BASE \
JOB_RESULTS_PATH \
JOB_ROOTFS_OVERLAY_PATH \
LD_LIBRARY_PATH \
LP_NUM_THREADS \
MESA_BASE_TAG \
MESA_BUILD_PATH \
MESA_DEBUG \
MESA_GLES_VERSION_OVERRIDE \
MESA_GLSL_VERSION_OVERRIDE \
MESA_GL_VERSION_OVERRIDE \
MESA_IMAGE \
MESA_IMAGE_PATH \
MESA_IMAGE_TAG \
MESA_TEMPLATES_COMMIT \
MESA_VK_IGNORE_CONFORMANCE_WARNING \
MINIO_HOST \
NIR_DEBUG \
PAN_I_WANT_A_BROKEN_VULKAN_DRIVER \
PAN_MESA_DEBUG \
PIGLIT_FRACTION \
PIGLIT_NO_WINDOW \
PIGLIT_OPTIONS \
PIGLIT_PLATFORM \
PIGLIT_PROFILES \
PIGLIT_REPLAY_ARTIFACTS_BASE_URL \
PIGLIT_REPLAY_DESCRIPTION_FILE \
PIGLIT_REPLAY_DEVICE_NAME \
PIGLIT_REPLAY_EXTRA_ARGS \
PIGLIT_REPLAY_REFERENCE_IMAGES_BASE \
PIGLIT_REPLAY_SUBCOMMAND \
PIGLIT_RESULTS \
PIGLIT_TESTS \
PIPELINE_ARTIFACTS_BASE \
SKQP_ASSETS_DIR \
SKQP_BACKENDS \
TU_DEBUG \
VIRGL_HOST_API \
VK_CPU \
VK_DRIVER \
VK_ICD_FILENAMES \
; do
if [ -n "${!var+x}" ]; then
echo "export $var=${!var@Q}"
fi
done

View File

@@ -1,22 +0,0 @@
#!/bin/sh
# Very early init, used to make sure devices and network are set up and
# reachable.
set -ex
cd /
mount -t proc none /proc
mount -t sysfs none /sys
mount -t devtmpfs none /dev || echo possibly already mounted
mkdir -p /dev/pts
mount -t devpts devpts /dev/pts
mount -t tmpfs tmpfs /tmp
echo "nameserver 8.8.8.8" > /etc/resolv.conf
[ -z "$NFS_SERVER_IP" ] || echo "$NFS_SERVER_IP caching-proxy" >> /etc/hosts
# Set the time so we can validate certificates before we fetch anything;
# however as not all DUTs have network, make this non-fatal.
for i in 1 2 3; do sntp -sS pool.ntp.org && break || sleep 2; done || true

View File

@@ -1,78 +0,0 @@
#!/bin/sh
# Second-stage init, used to set up devices and our job environment before
# running tests.
. /set-job-env-vars.sh
set -ex
# Set up any devices required by the jobs
[ -z "$HWCI_KERNEL_MODULES" ] || (echo -n $HWCI_KERNEL_MODULES | xargs -d, -n1 /usr/sbin/modprobe)
# Fix prefix confusion: the build installs to $CI_PROJECT_DIR, but we expect
# it in /install
ln -sf $CI_PROJECT_DIR/install /install
export LD_LIBRARY_PATH=/install/lib
export LIBGL_DRIVERS_PATH=/install/lib/dri
# Store Mesa's disk cache under /tmp, rather than sending it out over NFS.
export XDG_CACHE_HOME=/tmp
# Make sure Python can find all our imports
export PYTHONPATH=$(python3 -c "import sys;print(\":\".join(sys.path))")
if [ "$HWCI_FREQ_MAX" = "true" ]; then
# Ensure initialization of the DRM device (needed by MSM)
head -0 /dev/dri/renderD128
# Disable GPU frequency scaling
DEVFREQ_GOVERNOR=`find /sys/devices -name governor | grep gpu || true`
test -z "$DEVFREQ_GOVERNOR" || echo performance > $DEVFREQ_GOVERNOR || true
# Disable CPU frequency scaling
echo performance | tee -a /sys/devices/system/cpu/cpufreq/policy*/scaling_governor || true
# Disable GPU runtime power management
GPU_AUTOSUSPEND=`find /sys/devices -name autosuspend_delay_ms | grep gpu | head -1`
test -z "$GPU_AUTOSUSPEND" || echo -1 > $GPU_AUTOSUSPEND || true
fi
# Start a little daemon to capture the first devcoredump we encounter. (They
# expire after 5 minutes, so we poll for them).
./capture-devcoredump.sh &
# If we want Xorg to be running for the test, then we start it up before the
# HWCI_TEST_SCRIPT because we need to use xinit to start X (otherwise
# without using -displayfd you can race with Xorg's startup), but xinit will eat
# your client's return code
if [ -n "$HWCI_START_XORG" ]; then
echo "touch /xorg-started; sleep 100000" > /xorg-script
env \
xinit /bin/sh /xorg-script -- /usr/bin/Xorg -noreset -s 0 -dpms -logfile /Xorg.0.log &
# Wait for xorg to be ready for connections.
for i in 1 2 3 4 5; do
if [ -e /xorg-started ]; then
break
fi
sleep 5
done
export DISPLAY=:0
fi
RESULT=fail
if sh $HWCI_TEST_SCRIPT; then
RESULT=pass
rm -rf results/trace/$PIGLIT_REPLAY_DEVICE_NAME
fi
# upload artifacts
MINIO=$(cat /proc/cmdline | tr ' ' '\n' | grep minio_results | cut -d '=' -f 2 || true)
if [ -n "$MINIO" ]; then
tar -czf results.tar.gz results/;
ci-fairy minio login --token-file "${CI_JOB_JWT_FILE}";
ci-fairy minio cp results.tar.gz minio://"$MINIO"/results.tar.gz;
fi
echo "hwci: mesa: $RESULT"

View File

@@ -1,21 +0,0 @@
#!/bin/sh
set -ex
_XORG_SCRIPT="/xorg-script"
_FLAG_FILE="/xorg-started"
echo "touch ${_FLAG_FILE}; sleep 100000" > "${_XORG_SCRIPT}"
if [ "x$1" != "x" ]; then
export LD_LIBRARY_PATH="${1}/lib"
export LIBGL_DRIVERS_PATH="${1}/lib/dri"
fi
xinit /bin/sh "${_XORG_SCRIPT}" -- /usr/bin/Xorg vt45 -noreset -s 0 -dpms -logfile /Xorg.0.log &
# Wait for xorg to be ready for connections.
for i in 1 2 3 4 5; do
if [ -e "${_FLAG_FILE}" ]; then
break
fi
sleep 5
done

View File

@@ -1,57 +0,0 @@
CONFIG_LOCALVERSION_AUTO=y
CONFIG_DEBUG_KERNEL=y
# abootimg with a 'dummy' rootfs fails with root=/dev/nfs
CONFIG_BLK_DEV_INITRD=n
CONFIG_DEVFREQ_GOV_PERFORMANCE=y
CONFIG_DEVFREQ_GOV_POWERSAVE=y
CONFIG_DEVFREQ_GOV_USERSPACE=y
CONFIG_DEVFREQ_GOV_PASSIVE=y
CONFIG_DEVFREQ_GOV_SIMPLE_ONDEMAND=y
CONFIG_DRM=y
CONFIG_DRM_ETNAVIV=y
CONFIG_DRM_ROCKCHIP=y
CONFIG_DRM_PANFROST=y
CONFIG_DRM_LIMA=y
CONFIG_DRM_PANEL_SIMPLE=y
CONFIG_PWM_CROS_EC=y
CONFIG_BACKLIGHT_PWM=y
CONFIG_ROCKCHIP_CDN_DP=n
CONFIG_SPI_ROCKCHIP=y
CONFIG_PWM_ROCKCHIP=y
CONFIG_PHY_ROCKCHIP_DP=y
CONFIG_DWMAC_ROCKCHIP=y
CONFIG_MFD_RK808=y
CONFIG_REGULATOR_RK808=y
CONFIG_RTC_DRV_RK808=y
CONFIG_COMMON_CLK_RK808=y
CONFIG_REGULATOR_FAN53555=y
CONFIG_REGULATOR=y
CONFIG_REGULATOR_VCTRL=y
CONFIG_KASAN=n
CONFIG_KASAN_INLINE=n
CONFIG_STACKTRACE=n
CONFIG_TMPFS=y
CONFIG_PROVE_LOCKING=n
CONFIG_DEBUG_LOCKDEP=n
CONFIG_SOFTLOCKUP_DETECTOR=n
CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC=n
CONFIG_FW_LOADER_COMPRESS=y
CONFIG_USB_USBNET=y
CONFIG_NETDEVICES=y
CONFIG_USB_NET_DRIVERS=y
CONFIG_USB_RTL8152=y
CONFIG_USB_NET_AX8817X=y
CONFIG_USB_NET_SMSC95XX=y

View File

@@ -1,159 +0,0 @@
CONFIG_LOCALVERSION_AUTO=y
CONFIG_DEBUG_KERNEL=y
# abootimg with a 'dummy' rootfs fails with root=/dev/nfs
CONFIG_BLK_DEV_INITRD=n
CONFIG_DEVFREQ_GOV_PERFORMANCE=y
CONFIG_DEVFREQ_GOV_POWERSAVE=y
CONFIG_DEVFREQ_GOV_USERSPACE=y
CONFIG_DEVFREQ_GOV_PASSIVE=y
CONFIG_DRM=y
CONFIG_DRM_ROCKCHIP=y
CONFIG_DRM_PANFROST=y
CONFIG_DRM_LIMA=y
CONFIG_DRM_PANEL_SIMPLE=y
CONFIG_DRM_PANEL_EDP=y
CONFIG_DRM_MSM=y
CONFIG_DRM_I2C_ADV7511=y
CONFIG_PWM_CROS_EC=y
CONFIG_BACKLIGHT_PWM=y
CONFIG_ROCKCHIP_CDN_DP=n
CONFIG_SPI_ROCKCHIP=y
CONFIG_PWM_ROCKCHIP=y
CONFIG_PHY_ROCKCHIP_DP=y
CONFIG_DWMAC_ROCKCHIP=y
CONFIG_STMMAC_ETH=y
CONFIG_TYPEC_FUSB302=y
CONFIG_TYPEC=y
CONFIG_TYPEC_TCPM=y
# MSM platform bits
# For CONFIG_QCOM_LMH
CONFIG_OF=y
CONFIG_QCOM_COMMAND_DB=y
CONFIG_QCOM_RPMHPD=y
CONFIG_QCOM_RPMPD=y
CONFIG_SDM_GPUCC_845=y
CONFIG_SDM_VIDEOCC_845=y
CONFIG_SDM_DISPCC_845=y
CONFIG_SDM_LPASSCC_845=y
CONFIG_SDM_CAMCC_845=y
CONFIG_RESET_QCOM_PDC=y
CONFIG_DRM_TI_SN65DSI86=y
CONFIG_I2C_QCOM_GENI=y
CONFIG_SPI_QCOM_GENI=y
CONFIG_PHY_QCOM_QUSB2=y
CONFIG_PHY_QCOM_QMP=y
CONFIG_QCOM_CLK_APCC_MSM8996=y
CONFIG_QCOM_LLCC=y
CONFIG_QCOM_LMH=y
CONFIG_QCOM_SPMI_TEMP_ALARM=y
CONFIG_QCOM_WDT=y
CONFIG_POWER_RESET_QCOM_PON=y
CONFIG_RTC_DRV_PM8XXX=y
CONFIG_INTERCONNECT=y
CONFIG_INTERCONNECT_QCOM=y
CONFIG_INTERCONNECT_QCOM_SDM845=y
CONFIG_INTERCONNECT_QCOM_MSM8916=y
CONFIG_INTERCONNECT_QCOM_OSM_L3=y
CONFIG_INTERCONNECT_QCOM_SC7180=y
CONFIG_CRYPTO_DEV_QCOM_RNG=y
CONFIG_SC_DISPCC_7180=y
CONFIG_SC_GPUCC_7180=y
# db410c ethernet
CONFIG_USB_RTL8152=y
# db820c ethernet
CONFIG_ATL1C=y
CONFIG_ARCH_ALPINE=n
CONFIG_ARCH_BCM2835=n
CONFIG_ARCH_BCM_IPROC=n
CONFIG_ARCH_BERLIN=n
CONFIG_ARCH_BRCMSTB=n
CONFIG_ARCH_EXYNOS=n
CONFIG_ARCH_K3=n
CONFIG_ARCH_LAYERSCAPE=n
CONFIG_ARCH_LG1K=n
CONFIG_ARCH_HISI=n
CONFIG_ARCH_MVEBU=n
CONFIG_ARCH_SEATTLE=n
CONFIG_ARCH_SYNQUACER=n
CONFIG_ARCH_RENESAS=n
CONFIG_ARCH_R8A774A1=n
CONFIG_ARCH_R8A774C0=n
CONFIG_ARCH_R8A7795=n
CONFIG_ARCH_R8A7796=n
CONFIG_ARCH_R8A77965=n
CONFIG_ARCH_R8A77970=n
CONFIG_ARCH_R8A77980=n
CONFIG_ARCH_R8A77990=n
CONFIG_ARCH_R8A77995=n
CONFIG_ARCH_STRATIX10=n
CONFIG_ARCH_TEGRA=n
CONFIG_ARCH_SPRD=n
CONFIG_ARCH_THUNDER=n
CONFIG_ARCH_THUNDER2=n
CONFIG_ARCH_UNIPHIER=n
CONFIG_ARCH_VEXPRESS=n
CONFIG_ARCH_XGENE=n
CONFIG_ARCH_ZX=n
CONFIG_ARCH_ZYNQMP=n
# Strip out some stuff we don't need for graphics testing, to reduce
# the build.
CONFIG_CAN=n
CONFIG_WIRELESS=n
CONFIG_RFKILL=n
CONFIG_WLAN=n
CONFIG_REGULATOR_FAN53555=y
CONFIG_REGULATOR=y
CONFIG_REGULATOR_VCTRL=y
CONFIG_KASAN=n
CONFIG_KASAN_INLINE=n
CONFIG_STACKTRACE=n
CONFIG_TMPFS=y
CONFIG_PROVE_LOCKING=n
CONFIG_DEBUG_LOCKDEP=n
CONFIG_SOFTLOCKUP_DETECTOR=y
CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC=y
CONFIG_DETECT_HUNG_TASK=y
CONFIG_FW_LOADER_COMPRESS=y
CONFIG_FW_LOADER_USER_HELPER=n
CONFIG_USB_USBNET=y
CONFIG_NETDEVICES=y
CONFIG_USB_NET_DRIVERS=y
CONFIG_USB_RTL8152=y
CONFIG_USB_NET_AX8817X=y
CONFIG_USB_NET_SMSC95XX=y
# For amlogic
CONFIG_MESON_GXL_PHY=y
CONFIG_MDIO_BUS_MUX_MESON_G12A=y
CONFIG_DRM_MESON=y
# For Mediatek
CONFIG_DRM_MEDIATEK=y
CONFIG_PWM_MEDIATEK=y
CONFIG_DRM_MEDIATEK_HDMI=y
CONFIG_GNSS=y
CONFIG_GNSS_MTK_SERIAL=y
CONFIG_HW_RANDOM=y
CONFIG_HW_RANDOM_MTK=y
CONFIG_MTK_DEVAPC=y
CONFIG_PWM_MTK_DISP=y
CONFIG_MTK_CMDQ=y

View File

@@ -1,48 +0,0 @@
#!/bin/bash
set -e
set -o xtrace
# Fetch the arm-built rootfs image and unpack it in our x86 container (saves
# network transfer, disk usage, and runtime on test jobs)
if wget -q --method=HEAD "${ARTIFACTS_PREFIX}/${FDO_UPSTREAM_REPO}/${ARTIFACTS_SUFFIX}/${arch}/done"; then
ARTIFACTS_URL="${ARTIFACTS_PREFIX}/${FDO_UPSTREAM_REPO}/${ARTIFACTS_SUFFIX}/${arch}"
else
ARTIFACTS_URL="${ARTIFACTS_PREFIX}/${CI_PROJECT_PATH}/${ARTIFACTS_SUFFIX}/${arch}"
fi
wget ${ARTIFACTS_URL}/lava-rootfs.tgz -O rootfs.tgz
mkdir -p /rootfs-$arch
tar -C /rootfs-$arch '--exclude=./dev/*' -zxf rootfs.tgz
rm rootfs.tgz
if [[ $arch == "arm64" ]]; then
mkdir -p /baremetal-files
pushd /baremetal-files
wget ${ARTIFACTS_URL}/Image
wget ${ARTIFACTS_URL}/Image.gz
wget ${ARTIFACTS_URL}/cheza-kernel
DEVICE_TREES="apq8016-sbc.dtb apq8096-db820c.dtb"
for DTB in $DEVICE_TREES; do
wget ${ARTIFACTS_URL}/$DTB
done
popd
elif [[ $arch == "armhf" ]]; then
mkdir -p /baremetal-files
pushd /baremetal-files
wget ${ARTIFACTS_URL}/zImage
DEVICE_TREES="imx6q-cubox-i.dtb"
for DTB in $DEVICE_TREES; do
wget ${ARTIFACTS_URL}/$DTB
done
popd
fi

View File

@@ -1,18 +0,0 @@
#!/bin/bash
set -ex
APITRACE_VERSION="170424754bb46002ba706e16ee5404b61988d74a"
git clone https://github.com/apitrace/apitrace.git --single-branch --no-checkout /apitrace
pushd /apitrace
git checkout "$APITRACE_VERSION"
git submodule update --init --depth 1 --recursive
cmake -S . -B _build -G Ninja -DCMAKE_BUILD_TYPE=Release -DENABLE_GUI=False -DENABLE_WAFFLE=on $EXTRA_CMAKE_ARGS
ninja -C _build
mkdir build
cp _build/apitrace build
cp _build/eglretrace build
${STRIP_CMD:-strip} build/*
find . -not -path './build' -not -path './build/*' -delete
popd

View File

@@ -1,36 +0,0 @@
#!/bin/bash
set -ex
CROSVM_VERSION=d2b6a64dd31c92a284a905c0f2483d0b222b1220
git clone --single-branch -b for-mesa-ci --no-checkout https://gitlab.freedesktop.org/tomeu/crosvm.git /platform/crosvm
pushd /platform/crosvm
git checkout "$CROSVM_VERSION"
git submodule update --init
VIRGLRENDERER_VERSION=e420a5aab92de8fb42fad50762f0ac3b5fcb3bfb
rm -rf third_party/virglrenderer
git clone --single-branch -b master --no-checkout https://gitlab.freedesktop.org/virgl/virglrenderer.git third_party/virglrenderer
pushd third_party/virglrenderer
git checkout "$VIRGLRENDERER_VERSION"
meson build/ $EXTRA_MESON_ARGS
ninja -C build install
popd
RUSTFLAGS='-L native=/usr/local/lib' cargo install \
bindgen \
-j ${FDO_CI_CONCURRENT:-4} \
--root /usr/local \
$EXTRA_CARGO_ARGS
RUSTFLAGS='-L native=/usr/local/lib' cargo install \
-j ${FDO_CI_CONCURRENT:-4} \
--locked \
--features 'default-no-sandbox gpu x virgl_renderer virgl_renderer_next' \
--path . \
--root /usr/local \
$EXTRA_CARGO_ARGS
popd
rm -rf $PLATFORM2_ROOT $AOSP_EXTERNAL_ROOT/minijail $THIRD_PARTY_ROOT/adhd $THIRD_PARTY_ROOT/rust-vmm /platform/crosvm

View File

@@ -1,24 +0,0 @@
#!/bin/sh
set -ex
if [ -n "${DEQP_RUNNER_GIT_TAG}${DEQP_RUNNER_GIT_REV}" ]; then
# Build and install from source
DEQP_RUNNER_CARGO_ARGS="--git ${DEQP_RUNNER_GIT_URL:-https://gitlab.freedesktop.org/anholt/deqp-runner.git}"
if [ -n "${DEQP_RUNNER_GIT_TAG}" ]; then
DEQP_RUNNER_CARGO_ARGS="--tag ${DEQP_RUNNER_GIT_TAG} ${DEQP_RUNNER_CARGO_ARGS}"
else
DEQP_RUNNER_CARGO_ARGS="--rev ${DEQP_RUNNER_GIT_REV} ${DEQP_RUNNER_CARGO_ARGS}"
fi
DEQP_RUNNER_CARGO_ARGS="${DEQP_RUNNER_CARGO_ARGS} ${EXTRA_CARGO_ARGS}"
else
# Install from package registry
DEQP_RUNNER_CARGO_ARGS="--version 0.11.0 ${EXTRA_CARGO_ARGS} -- deqp-runner"
fi
cargo install --locked \
-j ${FDO_CI_CONCURRENT:-4} \
--root /usr/local \
${DEQP_RUNNER_CARGO_ARGS}

View File

@@ -1,86 +0,0 @@
#!/bin/bash
set -ex
git config --global user.email "mesa@example.com"
git config --global user.name "Mesa CI"
git clone \
https://github.com/KhronosGroup/VK-GL-CTS.git \
-b vulkan-cts-1.2.7.2 \
--depth 1 \
/VK-GL-CTS
pushd /VK-GL-CTS
# --insecure is due to SSL cert failures hitting sourceforge for zlib and
# libpng (sigh). The archives get their checksums checked anyway, and git
# always goes through ssh or https.
python3 external/fetch_sources.py --insecure
mkdir -p /deqp
# Save the testlog stylesheets:
cp doc/testlog-stylesheet/testlog.{css,xsl} /deqp
popd
pushd /deqp
# When including EGL/X11 testing, do that build first and save off its
# deqp-egl binary.
cmake -S /VK-GL-CTS -B . -G Ninja \
-DDEQP_TARGET=x11_egl_glx \
-DCMAKE_BUILD_TYPE=Release \
$EXTRA_CMAKE_ARGS
ninja modules/egl/deqp-egl
cp /deqp/modules/egl/deqp-egl /deqp/modules/egl/deqp-egl-x11
cmake -S /VK-GL-CTS -B . -G Ninja \
-DDEQP_TARGET=${DEQP_TARGET:-x11_glx} \
-DCMAKE_BUILD_TYPE=Release \
$EXTRA_CMAKE_ARGS
ninja
mv /deqp/modules/egl/deqp-egl-x11 /deqp/modules/egl/deqp-egl
# Copy out the mustpass lists we want.
mkdir /deqp/mustpass
for mustpass in $(< /VK-GL-CTS/external/vulkancts/mustpass/master/vk-default.txt) ; do
cat /VK-GL-CTS/external/vulkancts/mustpass/master/$mustpass \
>> /deqp/mustpass/vk-master.txt
done
cp \
/deqp/external/openglcts/modules/gl_cts/data/mustpass/gles/aosp_mustpass/3.2.6.x/*.txt \
/deqp/mustpass/.
cp \
/deqp/external/openglcts/modules/gl_cts/data/mustpass/egl/aosp_mustpass/3.2.6.x/egl-master.txt \
/deqp/mustpass/.
cp \
/deqp/external/openglcts/modules/gl_cts/data/mustpass/gles/khronos_mustpass/3.2.6.x/*-master.txt \
/deqp/mustpass/.
cp \
/deqp/external/openglcts/modules/gl_cts/data/mustpass/gl/khronos_mustpass/4.6.1.x/*-master.txt \
/deqp/mustpass/.
# Save *some* executor utils, but otherwise strip things down
# to reduct deqp build size:
mkdir /deqp/executor.save
cp /deqp/executor/testlog-to-* /deqp/executor.save
rm -rf /deqp/executor
mv /deqp/executor.save /deqp/executor
# Remove other mustpass files, since we saved off the ones we wanted to conventient locations above.
rm -rf /deqp/external/openglcts/modules/gl_cts/data/mustpass
rm -rf /deqp/external/vulkancts/modules/vulkan/vk-master*
rm -rf /deqp/external/vulkancts/modules/vulkan/vk-default
rm -rf /deqp/external/openglcts/modules/cts-runner
rm -rf /deqp/modules/internal
rm -rf /deqp/execserver
rm -rf /deqp/framework
find -iname '*cmake*' -o -name '*ninja*' -o -name '*.o' -o -name '*.a' | xargs rm -rf
${STRIP_CMD:-strip} external/vulkancts/modules/vulkan/deqp-vk
${STRIP_CMD:-strip} external/openglcts/modules/glcts
${STRIP_CMD:-strip} modules/*/deqp-*
du -sh *
rm -rf /VK-GL-CTS
popd

View File

@@ -1,14 +0,0 @@
#!/bin/bash
set -ex
git clone https://github.com/ValveSoftware/Fossilize.git
cd Fossilize
git checkout 72088685d90bc814d14aad5505354ffa8a642789
git submodule update --init
mkdir build
cd build
cmake -S .. -B . -G Ninja -DCMAKE_BUILD_TYPE=Release
ninja -C . install
cd ../..
rm -rf Fossilize

View File

@@ -1,19 +0,0 @@
#!/bin/bash
set -ex
GFXRECONSTRUCT_VERSION=3738decc2f4f9ff183818e5ab213a75a79fb7ab1
git clone https://github.com/LunarG/gfxreconstruct.git --single-branch -b master --no-checkout /gfxreconstruct
pushd /gfxreconstruct
git checkout "$GFXRECONSTRUCT_VERSION"
git submodule update --init
git submodule update
cmake -S . -B _build -G Ninja -DCMAKE_BUILD_TYPE=Release
ninja -C _build gfxrecon-replay gfxrecon-info
mkdir -p build/bin
install _build/tools/replay/gfxrecon-replay build/bin
install _build/tools/info/gfxrecon-info build/bin
strip build/bin/*
find . -not -path './build' -not -path './build/*' -delete
popd

View File

@@ -1,16 +0,0 @@
#!/bin/bash
set -ex
PARALLEL_DEQP_RUNNER_VERSION=6596b71cf37a7efb4d54acd48c770ed2d4ad6b7e
git clone https://gitlab.freedesktop.org/mesa/parallel-deqp-runner --single-branch -b master --no-checkout /parallel-deqp-runner
pushd /parallel-deqp-runner
git checkout "$PARALLEL_DEQP_RUNNER_VERSION"
meson . _build
ninja -C _build hang-detection
mkdir -p build/bin
install _build/hang-detection build/bin
strip build/bin/*
find . -not -path './build' -not -path './build/*' -delete
popd

View File

@@ -1,51 +0,0 @@
#!/bin/bash
set -ex
mkdir -p kernel
wget -qO- ${KERNEL_URL} | tar -xj --strip-components=1 -C kernel
pushd kernel
# The kernel doesn't like the gold linker (or the old lld in our debians).
# Sneak in some override symlinks during kernel build until we can update
# debian (they'll get blown away by the rm of the kernel dir at the end).
mkdir -p ld-links
for i in /usr/bin/*-ld /usr/bin/ld; do
i=`basename $i`
ln -sf /usr/bin/$i.bfd ld-links/$i
done
export PATH=`pwd`/ld-links:$PATH
export LOCALVERSION="`basename $KERNEL_URL`"
./scripts/kconfig/merge_config.sh ${DEFCONFIG} ../.gitlab-ci/container/${KERNEL_ARCH}.config
make ${KERNEL_IMAGE_NAME}
for image in ${KERNEL_IMAGE_NAME}; do
cp arch/${KERNEL_ARCH}/boot/${image} /lava-files/.
done
if [[ -n ${DEVICE_TREES} ]]; then
make dtbs
cp ${DEVICE_TREES} /lava-files/.
fi
if [[ ${DEBIAN_ARCH} = "amd64" ]]; then
make modules
INSTALL_MOD_PATH=/lava-files/rootfs-${DEBIAN_ARCH}/ make modules_install
fi
if [[ ${DEBIAN_ARCH} = "arm64" ]]; then
make Image.lzma
mkimage \
-f auto \
-A arm \
-O linux \
-d arch/arm64/boot/Image.lzma \
-C lzma\
-b arch/arm64/boot/dts/qcom/sdm845-cheza-r3.dtb \
/lava-files/cheza-kernel
KERNEL_IMAGE_NAME+=" cheza-kernel"
fi
popd
rm -rf kernel

View File

@@ -1,30 +0,0 @@
#!/bin/bash
set -ex
export LLVM_CONFIG="llvm-config-11"
$LLVM_CONFIG --version
git config --global user.email "mesa@example.com"
git config --global user.name "Mesa CI"
git clone \
https://github.com/llvm/llvm-project \
--depth 1 \
-b llvmorg-12.0.0-rc3 \
/llvm-project
mkdir /libclc
pushd /libclc
cmake -S /llvm-project/libclc -B . -G Ninja -DLLVM_CONFIG=$LLVM_CONFIG -DLIBCLC_TARGETS_TO_BUILD="spirv-mesa3d-;spirv64-mesa3d-" -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=/usr -DLLVM_SPIRV=/usr/bin/llvm-spirv
ninja
ninja install
popd
# workaroud cmake vs debian packaging.
mkdir -p /usr/lib/clc
ln -s /usr/share/clc/spirv64-mesa3d-.spv /usr/lib/clc/
ln -s /usr/share/clc/spirv-mesa3d-.spv /usr/lib/clc/
du -sh *
rm -rf /libclc /llvm-project

View File

@@ -1,14 +0,0 @@
#!/bin/bash
set -ex
export LIBDRM_VERSION=libdrm-2.4.109
wget https://dri.freedesktop.org/libdrm/$LIBDRM_VERSION.tar.xz
tar -xvf $LIBDRM_VERSION.tar.xz && rm $LIBDRM_VERSION.tar.xz
cd $LIBDRM_VERSION
meson build -D vc4=false -D freedreno=false -D etnaviv=false $EXTRA_MESON_ARGS
ninja -C build install
cd ..
rm -rf $LIBDRM_VERSION

View File

@@ -1,23 +0,0 @@
#!/bin/bash
set -ex
git clone https://gitlab.freedesktop.org/mesa/piglit.git --single-branch --no-checkout /piglit
pushd /piglit
git checkout af1785f31f65622d9b1ca1c08c75cf140bc7ed22
patch -p1 <$OLDPWD/.gitlab-ci/piglit/disable-vs_in.diff
cmake -S . -B . -G Ninja -DCMAKE_BUILD_TYPE=Release $PIGLIT_OPTS $EXTRA_CMAKE_ARGS
ninja $PIGLIT_BUILD_TARGETS
find -name .git -o -name '*ninja*' -o -iname '*cmake*' -o -name '*.[chao]' | xargs rm -rf
rm -rf target_api
if [ "x$PIGLIT_BUILD_TARGETS" = "xpiglit_replayer" ]; then
find ! -regex "^\.$" \
! -regex "^\.\/piglit.*" \
! -regex "^\.\/framework.*" \
! -regex "^\.\/bin$" \
! -regex "^\.\/bin\/replayer\.py" \
! -regex "^\.\/templates.*" \
! -regex "^\.\/tests$" \
! -regex "^\.\/tests\/replay\.py" 2>/dev/null | xargs rm -rf
fi
popd

View File

@@ -1,31 +0,0 @@
#!/bin/bash
# Note that this script is not actually "building" rust, but build- is the
# convention for the shared helpers for putting stuff in our containers.
set -ex
# cargo (and rustup) wants to store stuff in $HOME/.cargo, and binaries in
# $HOME/.cargo/bin. Make bin a link to a public bin directory so the commands
# are just available to all build jobs.
mkdir -p $HOME/.cargo
ln -s /usr/local/bin $HOME/.cargo/bin
# For rust in Mesa, we use rustup to install. This lets us pick an arbitrary
# version of the compiler, rather than whatever the container's Debian comes
# with.
#
# Pick the rust compiler (1.48) available in Debian stable, and pick a specific
# snapshot from rustup so the compiler doesn't drift on us.
wget https://sh.rustup.rs -O - | \
sh -s -- -y --default-toolchain 1.49.0-2020-12-31
# Set up a config script for cross compiling -- cargo needs your system cc for
# linking in cross builds, but doesn't know what you want to use for system cc.
cat > /root/.cargo/config <<EOF
[target.armv7-unknown-linux-gnueabihf]
linker = "arm-linux-gnueabihf-gcc"
[target.aarch64-unknown-linux-gnu]
linker = "aarch64-linux-gnu-gcc"
EOF

View File

@@ -1,82 +0,0 @@
#!/bin/bash
create_gn_args() {
cp "${BASE_ARGS_GN_FILE}" "${SKQP_OUT_DIR}"/args.gn
echo "target_cpu = \"${SKQP_ARCH}\"" >> "${SKQP_OUT_DIR}"/args.gn
}
download_skqp_models() (
# The download_model.py script needs a checksum file to know what models
# version to download.
# This is the most recent commit available in the skia repository with a
# valid files.checksum
SKIA_LAST_SKQP_CUT_COMMIT_SHA=ccf5f0d75b6a6b54756f2c62d57e3730eed8aa45
git fetch origin "${SKIA_LAST_SKQP_CUT_COMMIT_SHA}:refs/remotes/origin/${SKIA_LAST_SKQP_CUT_COMMIT_SHA}"
git checkout "${SKIA_LAST_SKQP_CUT_COMMIT_SHA}" -- \
platform_tools/android/apps/skqp/src/main/assets/files.checksum
# The following patch transforms download_model.py from python2 to python3.
git apply "${DOWNLOAD_MODEL_PATCH_FILE}"
python3 tools/skqp/download_model.py
# Copy resources from skia to skqp directory
python3 tools/skqp/setup_resources
)
set -ex
SCRIPT_DIR=$(realpath "$(dirname "$0")")
FETCH_GN_PATCH_FILE="${SCRIPT_DIR}/build-skqp_fetch-gn.patch"
BASE_ARGS_GN_FILE="${SCRIPT_DIR}/build-skqp_base.gn"
DOWNLOAD_MODEL_PATCH_FILE="${SCRIPT_DIR}/build-skqp_download_model.patch"
SKQP_ARCH=${SKQP_ARCH:-x64}
SKIA_DIR=${SKIA_DIR:-$(mktemp -d)}
SKQP_DIR=${SKQP_DIR:-$(mktemp -d)}
SKQP_OUT_DIR=${SKIA_DIR}/out/${SKQP_ARCH}
SKQP_INSTALL_DIR=/skqp
SKQP_ASSETS_DIR="${SKQP_INSTALL_DIR}/assets"
# Build list_gpu_unit_tests to update the unittests.txt file properly to the
# target hardware.
SKQP_BINARIES=(skqp list_gpu_unit_tests)
# Using a recent release version to mitigate instability during test phase
SKIA_COMMIT_SHA="canvaskit/0.32.0"
git clone 'https://skia.googlesource.com/skia/' \
--single-branch \
-b "${SKIA_COMMIT_SHA}" \
"${SKIA_DIR}"
pushd "${SKIA_DIR}"
git apply "${FETCH_GN_PATCH_FILE}"
# Fetch some needed build tools needed to build skia/skqp
# Basically, it clones repositories with commits SHAs from
# ${SKIA_DIR}/DEPS directory
python3 tools/git-sync-deps
mkdir -p "${SKQP_OUT_DIR}"
mkdir -p "${SKQP_INSTALL_DIR}"
create_gn_args
# Build and install skqp binaries
bin/gn gen "${SKQP_OUT_DIR}"
for BINARY in "${SKQP_BINARIES[@]}"
do
/usr/bin/ninja -C "${SKQP_OUT_DIR}" "${BINARY}"
install -m 0755 "${SKQP_OUT_DIR}/${BINARY}" "${SKQP_INSTALL_DIR}"
done
# Acquire assets and move them to the target directory.
download_skqp_models
mv platform_tools/android/apps/skqp/src/main/assets/ "${SKQP_ASSETS_DIR}"
popd
rm -Rf "${SKQP_DIR}"
rm -Rf "${SKIA_DIR}"
set +ex

View File

@@ -1,25 +0,0 @@
cc = "gcc"
cxx = "g++"
extra_cflags = [ "-DSK_ENABLE_DUMP_GPU", "-DSK_BUILD_FOR_SKQP" ]
extra_cflags_cc = [ "-static", "-Wno-error", "-Wno-macro-redefined", "-Wno-suggest-destructor-override", "-Wno-suggest-override" ]
is_debug = false
skia_enable_fontmgr_android = false
skia_enable_fontmgr_empty = true
skia_enable_pdf = false
skia_enable_skottie = false
skia_skqp_global_error_tolerance = 8
skia_tools_require_resources = true
skia_use_dng_sdk = false
skia_use_expat = true
skia_use_icu = false
skia_use_libheif = false
skia_use_lua = false
skia_use_piex = false
skia_use_vulkan = true
target_os = "linux"

View File

@@ -1,22 +0,0 @@
diff --git a/tools/skqp/download_model.py b/tools/skqp/download_model.py
index fb0020e481..a5d8a03754 100755
--- a/tools/skqp/download_model.py
+++ b/tools/skqp/download_model.py
@@ -10,7 +10,7 @@ import os
import shutil
import sys
import tempfile
-import urllib2
+import urllib.request, urllib.error, urllib.parse
def checksum(path):
if not os.path.exists(path):
@@ -33,7 +33,7 @@ def download(md5, path):
pass # ignore race condition
url = 'https://storage.googleapis.com/skia-skqp-assets/' + md5
with open(path, 'wb') as o:
- shutil.copyfileobj(urllib2.urlopen(url), o)
+ shutil.copyfileobj(urllib.request.urlopen(url), o)
def tmp(prefix):
fd, path = tempfile.mkstemp(prefix=prefix)

View File

@@ -1,13 +0,0 @@
diff --git a/bin/fetch-gn b/bin/fetch-gn
index b4bb14c630..59c4591a30 100755
--- a/bin/fetch-gn
+++ b/bin/fetch-gn
@@ -23,7 +23,7 @@ os.chdir(os.path.join(os.path.dirname(__file__), os.pardir))
gnzip = os.path.join(tempfile.mkdtemp(), 'gn.zip')
with open(gnzip, 'wb') as f:
OS = {'darwin': 'mac', 'linux': 'linux', 'linux2': 'linux', 'win32': 'windows'}[sys.platform]
- cpu = {'amd64': 'amd64', 'arm64': 'arm64', 'x86_64': 'amd64'}[platform.machine().lower()]
+ cpu = {'amd64': 'amd64', 'arm64': 'arm64', 'x86_64': 'amd64', 'aarch64': 'arm64'}[platform.machine().lower()]
rev = 'd62642c920e6a0d1756316d225a90fd6faa9e21e'
url = 'https://chrome-infra-packages.appspot.com/dl/gn/gn/{}-{}/+/git_revision:{}'.format(

View File

@@ -1,17 +0,0 @@
#!/bin/bash
set -ex
git config --global user.email "mesa@example.com"
git config --global user.name "Mesa CI"
git clone \
https://github.com/intel/libva-utils.git \
-b 2.13.0 \
--depth 1 \
/va-utils
pushd /va-utils
meson build -D tests=true -Dprefix=/va $EXTRA_MESON_ARGS
ninja -C build install
popd
rm -rf /va-utils

View File

@@ -1,43 +0,0 @@
#!/bin/bash
set -ex
VKD3D_PROTON_VERSION="2.3.1"
VKD3D_PROTON_COMMIT="3ed3526332f53d7d35cf1b685fa8096b01f26ff0"
VKD3D_PROTON_DST_DIR="/vkd3d-proton-tests"
VKD3D_PROTON_SRC_DIR="/vkd3d-proton-src"
VKD3D_PROTON_BUILD_DIR="/vkd3d-proton-$VKD3D_PROTON_VERSION"
function build_arch {
local arch="$1"
shift
meson "$@" \
-Denable_tests=true \
--buildtype release \
--prefix "$VKD3D_PROTON_BUILD_DIR" \
--strip \
--bindir "x${arch}" \
--libdir "x${arch}" \
"$VKD3D_PROTON_BUILD_DIR/build.${arch}"
ninja -C "$VKD3D_PROTON_BUILD_DIR/build.${arch}" install
install -D -m755 -t "${VKD3D_PROTON_DST_DIR}/x${arch}/bin" "$VKD3D_PROTON_BUILD_DIR/build.${arch}/tests/"*.exe
}
git clone https://github.com/HansKristian-Work/vkd3d-proton.git --single-branch -b "v$VKD3D_PROTON_VERSION" --no-checkout "$VKD3D_PROTON_SRC_DIR"
pushd "$VKD3D_PROTON_SRC_DIR"
git checkout "$VKD3D_PROTON_COMMIT"
git submodule update --init --recursive
git submodule update --recursive
build_arch 64 --cross-file build-win64.txt
build_arch 86 --cross-file build-win32.txt
cp "setup_vkd3d_proton.sh" "$VKD3D_PROTON_BUILD_DIR/setup_vkd3d_proton.sh"
chmod +x "$VKD3D_PROTON_BUILD_DIR/setup_vkd3d_proton.sh"
popd
"$VKD3D_PROTON_BUILD_DIR"/setup_vkd3d_proton.sh install
rm -rf "$VKD3D_PROTON_BUILD_DIR"
rm -rf "$VKD3D_PROTON_SRC_DIR"

View File

@@ -1,22 +0,0 @@
#!/bin/bash
set -ex
export LIBWAYLAND_VERSION="1.18.0"
export WAYLAND_PROTOCOLS_VERSION="1.24"
git clone https://gitlab.freedesktop.org/wayland/wayland
cd wayland
git checkout "$LIBWAYLAND_VERSION"
meson -Ddocumentation=false -Ddtd_validation=false -Dlibraries=true _build
ninja -C _build install
cd ..
rm -rf wayland
git clone https://gitlab.freedesktop.org/wayland/wayland-protocols
cd wayland-protocols
git checkout "$WAYLAND_PROTOCOLS_VERSION"
meson _build
ninja -C _build install
cd ..
rm -rf wayland-protocols

View File

@@ -1,10 +0,0 @@
#!/bin/sh
if test -f /etc/debian_version; then
apt-get autoremove -y --purge
fi
# Clean up any build cache for rust.
rm -rf /.cargo
ccache --show-stats

View File

@@ -1,36 +0,0 @@
#!/bin/sh
if test -f /etc/debian_version; then
CCACHE_PATH=/usr/lib/ccache
else
CCACHE_PATH=/usr/lib64/ccache
fi
# Common setup among container builds before we get to building code.
export CCACHE_COMPILERCHECK=content
export CCACHE_COMPRESS=true
export CCACHE_DIR=/cache/mesa/ccache
export PATH=$CCACHE_PATH:$PATH
# CMake ignores $PATH, so we have to force CC/GCC to the ccache versions.
export CC="${CCACHE_PATH}/gcc"
export CXX="${CCACHE_PATH}/g++"
# Force linkers to gold, since it's so much faster for building. We can't use
# lld because we're on old debian and it's buggy. ming fails meson builds
# with it with "meson.build:21:0: ERROR: Unable to determine dynamic linker"
find /usr/bin -name \*-ld -o -name ld | \
grep -v mingw | \
xargs -n 1 -I '{}' ln -sf '{}.gold' '{}'
ccache --show-stats
# Make a wrapper script for ninja to always include the -j flags
echo '#!/bin/sh -x' > /usr/local/bin/ninja
echo '/usr/bin/ninja -j${FDO_CI_CONCURRENT:-4} "$@"' >> /usr/local/bin/ninja
chmod +x /usr/local/bin/ninja
# Set MAKEFLAGS so that all make invocations in container builds include the
# flags (doesn't apply to non-container builds, but we don't run make there)
export MAKEFLAGS="-j${FDO_CI_CONCURRENT:-4}"

View File

@@ -1,35 +0,0 @@
#!/bin/bash
ndk=$1
arch=$2
cpu_family=$3
cpu=$4
cross_file="/cross_file-$arch.txt"
# armv7 has the toolchain split between two names.
arch2=${5:-$2}
# Note that we disable C++ exceptions, because Mesa doesn't use exceptions,
# and allowing it in code generation means we get unwind symbols that break
# the libEGL and driver symbol tests.
cat >$cross_file <<EOF
[binaries]
ar = '$ndk/toolchains/llvm/prebuilt/linux-x86_64/bin/$arch-ar'
c = ['ccache', '$ndk/toolchains/llvm/prebuilt/linux-x86_64/bin/${arch2}29-clang', '-fno-exceptions', '-fno-unwind-tables', '-fno-asynchronous-unwind-tables']
cpp = ['ccache', '$ndk/toolchains/llvm/prebuilt/linux-x86_64/bin/${arch2}29-clang++', '-fno-exceptions', '-fno-unwind-tables', '-fno-asynchronous-unwind-tables']
c_ld = 'lld'
cpp_ld = 'lld'
strip = '$ndk/toolchains/llvm/prebuilt/linux-x86_64/bin/$arch-strip'
pkgconfig = ['/usr/bin/pkg-config']
[host_machine]
system = 'linux'
cpu_family = '$cpu_family'
cpu = '$cpu'
endian = 'little'
[properties]
needs_exe_wrapper = true
EOF

View File

@@ -1,38 +0,0 @@
#!/bin/sh
# Makes a .pc file in the Android NDK for meson to find its libraries.
set -ex
ndk="$1"
pc="$2"
cflags="$3"
libs="$4"
version="$5"
sysroot=$ndk/toolchains/llvm/prebuilt/linux-x86_64/sysroot
for arch in \
x86_64-linux-android \
i686-linux-android \
aarch64-linux-android \
arm-linux-androideabi; do
pcdir=$sysroot/usr/lib/$arch/pkgconfig
mkdir -p $pcdir
cat >$pcdir/$pc <<EOF
prefix=$sysroot
exec_prefix=$sysroot
libdir=$sysroot/usr/lib/$arch/29
sharedlibdir=$sysroot/usr/lib/$arch
includedir=$sysroot/usr/include
Name: zlib
Description: zlib compression library
Version: $version
Requires:
Libs: -L$sysroot/usr/lib/$arch/29 $libs
Cflags: -I$sysroot/usr/include $cflags
EOF
done

View File

@@ -1,51 +0,0 @@
#!/bin/bash
arch=$1
cross_file="/cross_file-$arch.txt"
/usr/share/meson/debcrossgen --arch $arch -o "$cross_file"
# Explicitly set ccache path for cross compilers
sed -i "s|/usr/bin/\([^-]*\)-linux-gnu\([^-]*\)-g|/usr/lib/ccache/\\1-linux-gnu\\2-g|g" "$cross_file"
if [ "$arch" = "i386" ]; then
# Work around a bug in debcrossgen that should be fixed in the next release
sed -i "s|cpu_family = 'i686'|cpu_family = 'x86'|g" "$cross_file"
fi
# Rely on qemu-user being configured in binfmt_misc on the host
sed -i -e '/\[properties\]/a\' -e "needs_exe_wrapper = False" "$cross_file"
# Add a line for rustc, which debcrossgen is missing.
cc=`sed -n 's|c = .\(.*\).|\1|p' < $cross_file`
if [[ "$arch" = "arm64" ]]; then
rust_target=aarch64-unknown-linux-gnu
elif [[ "$arch" = "armhf" ]]; then
rust_target=armv7-unknown-linux-gnueabihf
elif [[ "$arch" = "i386" ]]; then
rust_target=i686-unknown-linux-gnu
elif [[ "$arch" = "ppc64el" ]]; then
rust_target=powerpc64le-unknown-linux-gnu
elif [[ "$arch" = "s390x" ]]; then
rust_target=s390x-unknown-linux-gnu
else
echo "Needs rustc target mapping"
fi
sed -i -e '/\[binaries\]/a\' -e "rust = ['rustc', '--target=$rust_target', '-C', 'linker=$cc']" "$cross_file"
# Set up cmake cross compile toolchain file for dEQP builds
toolchain_file="/toolchain-$arch.cmake"
if [[ "$arch" = "arm64" ]]; then
GCC_ARCH="aarch64-linux-gnu"
DE_CPU="DE_CPU_ARM_64"
CMAKE_ARCH=arm
elif [[ "$arch" = "armhf" ]]; then
GCC_ARCH="arm-linux-gnueabihf"
DE_CPU="DE_CPU_ARM"
CMAKE_ARCH=arm
fi
if [[ -n "$GCC_ARCH" ]]; then
echo "set(CMAKE_SYSTEM_NAME Linux)" > "$toolchain_file"
echo "set(CMAKE_SYSTEM_PROCESSOR arm)" >> "$toolchain_file"
echo "set(CMAKE_C_COMPILER /usr/lib/ccache/$GCC_ARCH-gcc)" >> "$toolchain_file"
echo "set(CMAKE_CXX_COMPILER /usr/lib/ccache/$GCC_ARCH-g++)" >> "$toolchain_file"
echo "set(ENV{PKG_CONFIG} \"/usr/bin/$GCC_ARCH-pkg-config\")" >> "$toolchain_file"
echo "set(DE_CPU $DE_CPU)" >> "$toolchain_file"
fi

View File

@@ -1,276 +0,0 @@
#!/bin/bash
set -ex
if [ $DEBIAN_ARCH = arm64 ]; then
ARCH_PACKAGES="firmware-qcom-media
libfontconfig1
libgl1
libglu1-mesa
"
elif [ $DEBIAN_ARCH = amd64 ]; then
ARCH_PACKAGES="firmware-amd-graphics
libelf1
libllvm11
libva2
libva-drm2
"
fi
INSTALL_CI_FAIRY_PACKAGES="git
python3-dev
python3-pip
python3-setuptools
python3-wheel
"
apt-get -y install --no-install-recommends \
$ARCH_PACKAGES \
$INSTALL_CI_FAIRY_PACKAGES \
ca-certificates \
firmware-realtek \
initramfs-tools \
libasan6 \
libexpat1 \
libpng16-16 \
libpython3.9 \
libsensors5 \
libvulkan1 \
libwaffle-1-0 \
libx11-6 \
libx11-xcb1 \
libxcb-dri2-0 \
libxcb-dri3-0 \
libxcb-glx0 \
libxcb-present0 \
libxcb-randr0 \
libxcb-shm0 \
libxcb-sync1 \
libxcb-xfixes0 \
libxdamage1 \
libxext6 \
libxfixes3 \
libxkbcommon0 \
libxrender1 \
libxshmfence1 \
libxxf86vm1 \
netcat-openbsd \
python3 \
python3-lxml \
python3-mako \
python3-numpy \
python3-packaging \
python3-pil \
python3-renderdoc \
python3-requests \
python3-simplejson \
python3-yaml \
sntp \
strace \
waffle-utils \
wget \
xinit \
xserver-xorg-core \
xz-utils
# Needed for ci-fairy, this revision is able to upload files to
# MinIO and doesn't depend on git
pip3 install git+http://gitlab.freedesktop.org/freedesktop/ci-templates@34f4ade99434043f88e164933f570301fd18b125
apt-get purge -y \
$INSTALL_CI_FAIRY_PACKAGES
passwd root -d
chsh -s /bin/sh
cat > /init <<EOF
#!/bin/sh
export PS1=lava-shell:
exec sh
EOF
chmod +x /init
#######################################################################
# Strip the image to a small minimal system without removing the debian
# toolchain.
# xz compress firmware so it doesn't waste RAM at runtime on ramdisk systems
find /lib/firmware -type f -print0 | \
xargs -0r -P4 -n4 xz -T1 -C crc32
# Copy timezone file and remove tzdata package
rm -rf /etc/localtime
cp /usr/share/zoneinfo/Etc/UTC /etc/localtime
UNNEEDED_PACKAGES="
libfdisk1
"
export DEBIAN_FRONTEND=noninteractive
# Removing unused packages
for PACKAGE in ${UNNEEDED_PACKAGES}
do
echo ${PACKAGE}
if ! apt-get remove --purge --yes "${PACKAGE}"
then
echo "WARNING: ${PACKAGE} isn't installed"
fi
done
apt-get autoremove --yes || true
# Dropping logs
rm -rf /var/log/*
# Dropping documentation, localization, i18n files, etc
rm -rf /usr/share/doc/*
rm -rf /usr/share/locale/*
rm -rf /usr/share/X11/locale/*
rm -rf /usr/share/man
rm -rf /usr/share/i18n/*
rm -rf /usr/share/info/*
rm -rf /usr/share/lintian/*
rm -rf /usr/share/common-licenses/*
rm -rf /usr/share/mime/*
# Dropping reportbug scripts
rm -rf /usr/share/bug
# Drop udev hwdb not required on a stripped system
rm -rf /lib/udev/hwdb.bin /lib/udev/hwdb.d/*
# Drop all gconv conversions && binaries
rm -rf usr/bin/iconv
rm -rf usr/sbin/iconvconfig
rm -rf usr/lib/*/gconv/
# Remove libusb database
rm -rf usr/sbin/update-usbids
rm -rf var/lib/usbutils/usb.ids
rm -rf usr/share/misc/usb.ids
rm -rf /root/.pip
#######################################################################
# Crush into a minimal production image to be deployed via some type of image
# updating system.
# IMPORTANT: The Debian system is not longer functional at this point,
# for example, apt and dpkg will stop working
UNNEEDED_PACKAGES="apt libapt-pkg6.0 "\
"ncurses-bin ncurses-base libncursesw6 libncurses6 "\
"perl-base "\
"debconf libdebconfclient0 "\
"e2fsprogs e2fslibs libfdisk1 "\
"insserv "\
"udev "\
"init-system-helpers "\
"bash "\
"cpio "\
"xz-utils "\
"passwd "\
"libsemanage1 libsemanage-common "\
"libsepol1 "\
"gpgv "\
"hostname "\
"adduser "\
"debian-archive-keyring "\
"libegl1-mesa-dev "\
"libegl-mesa0 "\
"libgl1-mesa-dev "\
"libgl1-mesa-dri "\
"libglapi-mesa "\
"libgles2-mesa-dev "\
"libglx-mesa0 "\
"mesa-common-dev "\
# Removing unneeded packages
for PACKAGE in ${UNNEEDED_PACKAGES}
do
echo "Forcing removal of ${PACKAGE}"
if ! dpkg --purge --force-remove-essential --force-depends "${PACKAGE}"
then
echo "WARNING: ${PACKAGE} isn't installed"
fi
done
# Show what's left package-wise before dropping dpkg itself
COLUMNS=300 dpkg-query -W --showformat='${Installed-Size;10}\t${Package}\n' | sort -k1,1n
# Drop dpkg
dpkg --purge --force-remove-essential --force-depends dpkg
# No apt or dpkg, no need for its configuration archives
rm -rf etc/apt
rm -rf etc/dpkg
# Drop directories not part of ostree
# Note that /var needs to exist as ostree bind mounts the deployment /var over
# it
rm -rf var/* opt srv share
# ca-certificates are in /etc drop the source
rm -rf usr/share/ca-certificates
# No bash, no need for completions
rm -rf usr/share/bash-completion
# No zsh, no need for comletions
rm -rf usr/share/zsh/vendor-completions
# drop gcc python helpers
rm -rf usr/share/gcc
# Drop sysvinit leftovers
rm -rf etc/init.d
rm -rf etc/rc[0-6S].d
# Drop upstart helpers
rm -rf etc/init
# Various xtables helpers
rm -rf usr/lib/xtables
# Drop all locales
# TODO: only remaining locale is actually "C". Should we really remove it?
rm -rf usr/lib/locale/*
# partition helpers
rm -rf usr/sbin/*fdisk
# local compiler
rm -rf usr/bin/localedef
# Systemd dns resolver
find usr etc -name '*systemd-resolve*' -prune -exec rm -r {} \;
# Systemd network configuration
find usr etc -name '*networkd*' -prune -exec rm -r {} \;
# systemd ntp client
find usr etc -name '*timesyncd*' -prune -exec rm -r {} \;
# systemd hw database manager
find usr etc -name '*systemd-hwdb*' -prune -exec rm -r {} \;
# No need for fuse
find usr etc -name '*fuse*' -prune -exec rm -r {} \;
# lsb init function leftovers
rm -rf usr/lib/lsb
# Only needed when adding libraries
rm -rf usr/sbin/ldconfig*
# Games, unused
rmdir usr/games
# Remove pam module to authenticate against a DB
# plus libdb-5.3.so that is only used by this pam module
rm -rf usr/lib/*/security/pam_userdb.so
rm -rf usr/lib/*/libdb-5.3.so
# remove NSS support for nis, nisplus and hesiod
rm -rf usr/lib/*/libnss_hesiod*
rm -rf usr/lib/*/libnss_nis*

View File

@@ -1,79 +0,0 @@
#!/bin/bash
set -e
set -o xtrace
export DEBIAN_FRONTEND=noninteractive
# Ephemeral packages (installed for this script and removed again at the end)
STABLE_EPHEMERAL=" \
"
dpkg --add-architecture $arch
apt-get update
apt-get install -y --no-remove \
$STABLE_EPHEMERAL \
crossbuild-essential-$arch \
libelf-dev:$arch \
libexpat1-dev:$arch \
libpciaccess-dev:$arch \
libstdc++6:$arch \
libvulkan-dev:$arch \
libx11-dev:$arch \
libx11-xcb-dev:$arch \
libxcb-dri2-0-dev:$arch \
libxcb-dri3-dev:$arch \
libxcb-glx0-dev:$arch \
libxcb-present-dev:$arch \
libxcb-randr0-dev:$arch \
libxcb-shm0-dev:$arch \
libxcb-xfixes0-dev:$arch \
libxdamage-dev:$arch \
libxext-dev:$arch \
libxrandr-dev:$arch \
libxshmfence-dev:$arch \
libxxf86vm-dev:$arch \
wget
if [[ $arch != "armhf" ]]; then
if [[ $arch == "s390x" ]]; then
LLVM=9
else
LLVM=11
fi
# llvm-*-tools:$arch conflicts with python3:amd64. Install dependencies only
# with apt-get, then force-install llvm-*-{dev,tools}:$arch with dpkg to get
# around this.
apt-get install -y --no-remove \
libclang-cpp${LLVM}:$arch \
libffi-dev:$arch \
libgcc-s1:$arch \
libtinfo-dev:$arch \
libz3-dev:$arch \
llvm-${LLVM}:$arch \
zlib1g
fi
. .gitlab-ci/container/create-cross-file.sh $arch
. .gitlab-ci/container/container_pre_build.sh
# dependencies where we want a specific version
EXTRA_MESON_ARGS="--cross-file=/cross_file-${arch}.txt -D libdir=lib/$(dpkg-architecture -A $arch -qDEB_TARGET_MULTIARCH)"
. .gitlab-ci/container/build-libdrm.sh
apt-get purge -y \
$STABLE_EPHEMERAL
. .gitlab-ci/container/container_post_build.sh
# This needs to be done after container_post_build.sh, or apt-get breaks in there
if [[ $arch != "armhf" ]]; then
apt-get download llvm-${LLVM}-{dev,tools}:$arch
dpkg -i --force-depends llvm-${LLVM}-*_${arch}.deb
rm llvm-${LLVM}-*_${arch}.deb
fi

View File

@@ -1,106 +0,0 @@
#!/bin/bash
set -ex
EPHEMERAL="\
autoconf \
rdfind \
unzip \
"
apt-get install -y --no-remove $EPHEMERAL
# Fetch the NDK and extract just the toolchain we want.
ndk=android-ndk-r21d
wget -O $ndk.zip https://dl.google.com/android/repository/$ndk-linux-x86_64.zip
unzip -d / $ndk.zip "$ndk/toolchains/llvm/*"
rm $ndk.zip
# Since it was packed as a zip file, symlinks/hardlinks got turned into
# duplicate files. Turn them into hardlinks to save on container space.
rdfind -makehardlinks true -makeresultsfile false /android-ndk-r21d/
# Drop some large tools we won't use in this build.
find /android-ndk-r21d/ -type f | egrep -i "clang-check|clang-tidy|lldb" | xargs rm -f
sh .gitlab-ci/container/create-android-ndk-pc.sh /$ndk zlib.pc "" "-lz" "1.2.3"
sh .gitlab-ci/container/create-android-cross-file.sh /$ndk x86_64-linux-android x86_64 x86_64
sh .gitlab-ci/container/create-android-cross-file.sh /$ndk i686-linux-android x86 x86
sh .gitlab-ci/container/create-android-cross-file.sh /$ndk aarch64-linux-android arm armv8
sh .gitlab-ci/container/create-android-cross-file.sh /$ndk arm-linux-androideabi arm armv7hl armv7a-linux-androideabi
# Not using build-libdrm.sh because we don't want its cleanup after building
# each arch. Fetch and extract now.
export LIBDRM_VERSION=libdrm-2.4.109
wget https://dri.freedesktop.org/libdrm/$LIBDRM_VERSION.tar.xz
tar -xf $LIBDRM_VERSION.tar.xz && rm $LIBDRM_VERSION.tar.xz
for arch in \
x86_64-linux-android \
i686-linux-android \
aarch64-linux-android \
arm-linux-androideabi ; do
cd $LIBDRM_VERSION
rm -rf build-$arch
meson build-$arch \
--cross-file=/cross_file-$arch.txt \
--libdir=lib/$arch \
-Dlibkms=false \
-Dnouveau=false \
-Dvc4=false \
-Detnaviv=false \
-Dfreedreno=false \
-Dintel=false \
-Dcairo-tests=false \
-Dvalgrind=false
ninja -C build-$arch install
cd ..
done
rm -rf $LIBDRM_VERSION
export LIBELF_VERSION=libelf-0.8.13
wget https://fossies.org/linux/misc/old/$LIBELF_VERSION.tar.gz
# Not 100% sure who runs the mirror above so be extra careful
if ! echo "4136d7b4c04df68b686570afa26988ac ${LIBELF_VERSION}.tar.gz" | md5sum -c -; then
echo "Checksum failed"
exit 1
fi
tar -xf ${LIBELF_VERSION}.tar.gz
cd $LIBELF_VERSION
# Work around a bug in the original configure not enabling __LIBELF64.
autoreconf
for arch in \
x86_64-linux-android \
i686-linux-android \
aarch64-linux-android \
arm-linux-androideabi ; do
ccarch=${arch}
if [ "${arch}" == 'arm-linux-androideabi' ]
then
ccarch=armv7a-linux-androideabi
fi
export CC=/android-ndk-r21d/toolchains/llvm/prebuilt/linux-x86_64/bin/${arch}-ar
export CC=/android-ndk-r21d/toolchains/llvm/prebuilt/linux-x86_64/bin/${ccarch}29-clang
export CXX=/android-ndk-r21d/toolchains/llvm/prebuilt/linux-x86_64/bin/${ccarch}29-clang++
export LD=/android-ndk-r21d/toolchains/llvm/prebuilt/linux-x86_64/bin/${arch}-ld
export RANLIB=/android-ndk-r21d/toolchains/llvm/prebuilt/linux-x86_64/bin/${arch}-ranlib
# The configure script doesn't know about android, but doesn't really use the host anyway it
# seems
./configure --host=x86_64-linux-gnu --disable-nls --disable-shared \
--libdir=/usr/local/lib/${arch}
make install
make distclean
done
cd ..
rm -rf $LIBELF_VERSION
apt-get purge -y $EPHEMERAL

View File

@@ -1,71 +0,0 @@
#!/bin/bash
set -e
set -o xtrace
apt-get -y install ca-certificates
sed -i -e 's/http:\/\/deb/https:\/\/deb/g' /etc/apt/sources.list
echo 'deb https://deb.debian.org/debian buster main' >/etc/apt/sources.list.d/buster.list
apt-get update
apt-get -y install \
abootimg \
autoconf \
automake \
bc \
bison \
ccache \
cmake \
debootstrap \
fastboot \
flex \
g++ \
git \
kmod \
libasan6 \
libdrm-dev \
libelf-dev \
libexpat1-dev \
libx11-dev \
libx11-xcb-dev \
libxcb-dri2-0-dev \
libxcb-dri3-dev \
libxcb-glx0-dev \
libxcb-present-dev \
libxcb-randr0-dev \
libxcb-shm0-dev \
libxcb-xfixes0-dev \
libxdamage-dev \
libxext-dev \
libxrandr-dev \
libxshmfence-dev \
libxxf86vm-dev \
llvm-11-dev \
meson \
pkg-config \
python3-mako \
python3-pil \
python3-pip \
python3-requests \
python3-setuptools \
u-boot-tools \
wget \
xz-utils \
zlib1g-dev
# Not available anymore in bullseye
apt-get install -y --no-remove -t buster \
android-sdk-ext4-utils
pip3 install git+http://gitlab.freedesktop.org/freedesktop/ci-templates@34f4ade99434043f88e164933f570301fd18b125
arch=armhf
. .gitlab-ci/container/cross_build.sh
. .gitlab-ci/container/container_pre_build.sh
# dependencies where we want a specific version
EXTRA_MESON_ARGS=
. .gitlab-ci/container/build-libdrm.sh
. .gitlab-ci/container/container_post_build.sh

View File

@@ -1,33 +0,0 @@
#!/bin/bash
set -e
set -o xtrace
############### Install packages for baremetal testing
apt-get install -y ca-certificates
sed -i -e 's/http:\/\/deb/https:\/\/deb/g' /etc/apt/sources.list
apt-get update
apt-get install -y --no-remove \
abootimg \
cpio \
fastboot \
netcat \
procps \
python3-distutils \
python3-minimal \
python3-serial \
rsync \
snmp \
wget
# setup SNMPv2 SMI MIB
wget https://raw.githubusercontent.com/net-snmp/net-snmp/master/mibs/SNMPv2-SMI.txt \
-O /usr/share/snmp/mibs/SNMPv2-SMI.txt
arch=arm64 . .gitlab-ci/container/baremetal_build.sh
arch=armhf . .gitlab-ci/container/baremetal_build.sh
# This firmware file from Debian bullseye causes hangs
wget https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/plain/qcom/a530_pfp.fw?id=d5f9eea5a251d43412b07f5295d03e97b89ac4a5 \
-O /rootfs-arm64/lib/firmware/qcom/a530_pfp.fw

View File

@@ -1,5 +0,0 @@
#!/bin/bash
arch=i386
. .gitlab-ci/container/cross_build.sh

View File

@@ -1,5 +0,0 @@
#!/bin/bash
arch=ppc64el
. .gitlab-ci/container/cross_build.sh

View File

@@ -1,5 +0,0 @@
#!/bin/bash
arch=s390x
. .gitlab-ci/container/cross_build.sh

View File

@@ -1,81 +0,0 @@
#!/bin/bash
set -e
set -o xtrace
export DEBIAN_FRONTEND=noninteractive
apt-get install -y ca-certificates
sed -i -e 's/http:\/\/deb/https:\/\/deb/g' /etc/apt/sources.list
# Ephemeral packages (installed for this script and removed again at
# the end)
STABLE_EPHEMERAL=" \
python3-pip \
python3-setuptools \
"
apt-get update
apt-get install -y --no-remove \
$STABLE_EPHEMERAL \
bison \
ccache \
dpkg-cross \
flex \
g++ \
g++-mingw-w64-x86-64 \
gcc \
git \
glslang-tools \
kmod \
libclang-11-dev \
libclang-9-dev \
libclc-dev \
libelf-dev \
libepoxy-dev \
libexpat1-dev \
libgtk-3-dev \
libllvm11 \
libllvm9 \
libomxil-bellagio-dev \
libpciaccess-dev \
libunwind-dev \
libva-dev \
libvdpau-dev \
libvulkan-dev \
libx11-dev \
libx11-xcb-dev \
libxext-dev \
libxml2-utils \
libxrandr-dev \
libxrender-dev \
libxshmfence-dev \
libxvmc-dev \
libxxf86vm-dev \
libz-mingw-w64-dev \
make \
meson \
pkg-config \
python3-mako \
python3-pil \
python3-requests \
qemu-user \
valgrind \
wget \
wine64 \
x11proto-dri2-dev \
x11proto-gl-dev \
x11proto-randr-dev \
xz-utils \
zlib1g-dev
# Needed for ci-fairy, this revision is able to upload files to MinIO
pip3 install git+http://gitlab.freedesktop.org/freedesktop/ci-templates@34f4ade99434043f88e164933f570301fd18b125
############### Uninstall ephemeral packages
apt-get purge -y $STABLE_EPHEMERAL
. .gitlab-ci/container/container_post_build.sh

View File

@@ -1,107 +0,0 @@
#!/bin/bash
set -e
set -o xtrace
export DEBIAN_FRONTEND=noninteractive
# Ephemeral packages (installed for this script and removed again at the end)
STABLE_EPHEMERAL=" \
autoconf \
automake \
autotools-dev \
bzip2 \
libtool \
python3-pip \
"
# We need multiarch for Wine
dpkg --add-architecture i386
apt-get update
apt-get install -y --no-remove \
$STABLE_EPHEMERAL \
check \
clang \
cmake \
libasan6 \
libarchive-dev \
libclang-cpp11-dev \
libgbm-dev \
libglvnd-dev \
libllvmspirvlib-dev \
liblua5.3-dev \
libxcb-dri2-0-dev \
libxcb-dri3-dev \
libxcb-glx0-dev \
libxcb-present-dev \
libxcb-randr0-dev \
libxcb-shm0-dev \
libxcb-sync-dev \
libxcb-xfixes0-dev \
libxcb1-dev \
libxml2-dev \
llvm-11-dev \
llvm-9-dev \
ocl-icd-opencl-dev \
procps \
spirv-tools \
strace \
time \
wine \
wine32
. .gitlab-ci/container/container_pre_build.sh
# Debian's pkg-config wrapers for mingw are broken, and there's no sign that
# they're going to be fixed, so we'll just have to fix it ourselves
# https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=930492
cat >/usr/local/bin/x86_64-w64-mingw32-pkg-config <<EOF
#!/bin/sh
PKG_CONFIG_LIBDIR=/usr/x86_64-w64-mingw32/lib/pkgconfig pkg-config \$@
EOF
chmod +x /usr/local/bin/x86_64-w64-mingw32-pkg-config
# dependencies where we want a specific version
export XORG_RELEASES=https://xorg.freedesktop.org/releases/individual
export XORGMACROS_VERSION=util-macros-1.19.0
wget $XORG_RELEASES/util/$XORGMACROS_VERSION.tar.bz2
tar -xvf $XORGMACROS_VERSION.tar.bz2 && rm $XORGMACROS_VERSION.tar.bz2
cd $XORGMACROS_VERSION; ./configure; make install; cd ..
rm -rf $XORGMACROS_VERSION
. .gitlab-ci/container/build-libdrm.sh
. .gitlab-ci/container/build-wayland.sh
pushd /usr/local
git clone https://gitlab.freedesktop.org/mesa/shader-db.git --depth 1
rm -rf shader-db/.git
cd shader-db
make
popd
git clone https://github.com/microsoft/DirectX-Headers -b v1.0.1 --depth 1
pushd DirectX-Headers
mkdir build
cd build
meson .. --backend=ninja --buildtype=release -Dbuild-test=false
ninja
ninja install
popd
rm -rf DirectX-Headers
pip3 install git+https://git.lavasoftware.org/lava/lavacli@3db3ddc45e5358908bc6a17448059ea2340492b7
############### Uninstall the build software
apt-get purge -y \
$STABLE_EPHEMERAL
. .gitlab-ci/container/container_post_build.sh

View File

@@ -1,70 +0,0 @@
#!/bin/bash
set -e
set -o xtrace
export DEBIAN_FRONTEND=noninteractive
apt-get install -y ca-certificates
sed -i -e 's/http:\/\/deb/https:\/\/deb/g' /etc/apt/sources.list
# Ephemeral packages (installed for this script and removed again at
# the end)
STABLE_EPHEMERAL=" \
cargo \
python3-dev \
python3-pip \
python3-setuptools \
python3-wheel \
"
apt-get update
apt-get dist-upgrade -y
apt-get install -y --no-remove \
git \
git-lfs \
libasan6 \
libexpat1 \
libllvm11 \
libllvm9 \
liblz4-1 \
libpng16-16 \
libpython3.9 \
libvulkan1 \
libwayland-client0 \
libwayland-server0 \
libxcb-ewmh2 \
libxcb-randr0 \
libxcb-xfixes0 \
libxkbcommon0 \
libxrandr2 \
libxrender1 \
python3-mako \
python3-numpy \
python3-packaging \
python3-pil \
python3-requests \
python3-six \
python3-yaml \
vulkan-tools \
waffle-utils \
xauth \
xvfb \
zlib1g
apt-get install -y --no-install-recommends \
$STABLE_EPHEMERAL
# Needed for ci-fairy, this revision is able to upload files to MinIO
# and doesn't depend on git
pip3 install git+http://gitlab.freedesktop.org/freedesktop/ci-templates@34f4ade99434043f88e164933f570301fd18b125
############### Build dEQP runner
. .gitlab-ci/container/build-deqp-runner.sh
rm -rf ~/.cargo
apt-get purge -y $STABLE_EPHEMERAL
apt-get autoremove -y --purge

View File

@@ -1,122 +0,0 @@
#!/bin/bash
set -e
set -o xtrace
export DEBIAN_FRONTEND=noninteractive
# Ephemeral packages (installed for this script and removed again at the end)
STABLE_EPHEMERAL=" \
autoconf \
automake \
bc \
bison \
bzip2 \
ccache \
clang-11 \
cmake \
flex \
g++ \
glslang-tools \
libasound2-dev \
libcap-dev \
libclang-cpp11-dev \
libelf-dev \
libexpat1-dev \
libfdt-dev \
libgbm-dev \
libgles2-mesa-dev \
libllvmspirvlib-dev \
libpciaccess-dev \
libpng-dev \
libudev-dev \
libvulkan-dev \
libwaffle-dev \
libx11-xcb-dev \
libxcb-dri2-0-dev \
libxext-dev \
libxkbcommon-dev \
libxrender-dev \
llvm-11-dev \
llvm-spirv \
make \
meson \
ocl-icd-opencl-dev \
patch \
pkg-config \
python3-distutils \
xz-utils \
"
apt-get install -y --no-remove \
$STABLE_EPHEMERAL \
clinfo \
iptables \
libclang-common-11-dev \
libclang-cpp11 \
libcap2 \
libegl1 \
libepoxy-dev \
libfdt1 \
libllvmspirvlib11 \
libxcb-shm0 \
ocl-icd-libopencl1 \
python3-lxml \
python3-renderdoc \
python3-simplejson \
spirv-tools \
sysvinit-core \
wget
. .gitlab-ci/container/container_pre_build.sh
############### Build libdrm
. .gitlab-ci/container/build-libdrm.sh
############### Build Wayland
. .gitlab-ci/container/build-wayland.sh
############### Build Crosvm
. .gitlab-ci/container/build-rust.sh
. .gitlab-ci/container/build-crosvm.sh
rm -rf /root/.cargo
rm -rf /root/.rustup
############### Build kernel
export DEFCONFIG="arch/x86/configs/x86_64_defconfig"
export KERNEL_IMAGE_NAME=bzImage
export KERNEL_ARCH=x86_64
export DEBIAN_ARCH=amd64
mkdir -p /lava-files/
. .gitlab-ci/container/build-kernel.sh
############### Build libclc
. .gitlab-ci/container/build-libclc.sh
############### Build piglit
PIGLIT_OPTS="-DPIGLIT_BUILD_CL_TESTS=ON -DPIGLIT_BUILD_DMA_BUF_TESTS=ON" . .gitlab-ci/container/build-piglit.sh
############### Build dEQP GL
DEQP_TARGET=surfaceless . .gitlab-ci/container/build-deqp.sh
############### Build apitrace
. .gitlab-ci/container/build-apitrace.sh
############### Uninstall the build software
ccache --show-stats
apt-get purge -y \
$STABLE_EPHEMERAL
apt-get autoremove -y --purge

View File

@@ -1,164 +0,0 @@
#!/bin/bash
set -e
set -o xtrace
export DEBIAN_FRONTEND=noninteractive
# Ephemeral packages (installed for this script and removed again at the end)
STABLE_EPHEMERAL=" \
ccache \
cmake \
g++ \
g++-mingw-w64-i686-posix \
g++-mingw-w64-x86-64-posix \
glslang-tools \
libexpat1-dev \
libgbm-dev \
libgles2-mesa-dev \
liblz4-dev \
libpciaccess-dev \
libudev-dev \
libvulkan-dev \
libwaffle-dev \
libx11-xcb-dev \
libxcb-ewmh-dev \
libxcb-keysyms1-dev \
libxkbcommon-dev \
libxrandr-dev \
libxrender-dev \
libzstd-dev \
meson \
mingw-w64-i686-dev \
mingw-w64-tools \
mingw-w64-x86-64-dev \
p7zip \
patch \
pkg-config \
python3-distutils \
wget \
xz-utils \
"
apt-get install -y --no-remove \
$STABLE_EPHEMERAL \
libxcb-shm0 \
python3-lxml \
python3-simplejson \
xinit \
xserver-xorg-video-amdgpu \
xserver-xorg-video-ati
# We need multiarch for Wine
dpkg --add-architecture i386
apt-get update
apt-get install -y --no-remove \
wine \
wine32 \
wine64
function setup_wine() {
export WINEDEBUG="-all"
export WINEPREFIX="$1"
# We don't want crash dialogs
cat >crashdialog.reg <<EOF
Windows Registry Editor Version 5.00
[HKEY_CURRENT_USER\Software\Wine\WineDbg]
"ShowCrashDialog"=dword:00000000
EOF
# Set the wine prefix and disable the crash dialog
wine regedit crashdialog.reg
rm crashdialog.reg
# An immediate wine command may fail with: "${WINEPREFIX}: Not a
# valid wine prefix." and that is just spit because of checking
# the existance of the system.reg file, which fails. Just giving
# it a bit more of time for it to be created solves the problem
# ...
while ! test -f "${WINEPREFIX}/system.reg"; do sleep 1; done
}
############### Install DXVK
DXVK_VERSION="1.8.1"
setup_wine "/dxvk-wine64"
wget "https://github.com/doitsujin/dxvk/releases/download/v${DXVK_VERSION}/dxvk-${DXVK_VERSION}.tar.gz"
tar xzpf dxvk-"${DXVK_VERSION}".tar.gz
dxvk-"${DXVK_VERSION}"/setup_dxvk.sh install
rm -rf dxvk-"${DXVK_VERSION}"
rm dxvk-"${DXVK_VERSION}".tar.gz
############### Install Windows' apitrace binaries
APITRACE_VERSION="10.0"
APITRACE_VERSION_DATE=""
wget "https://github.com/apitrace/apitrace/releases/download/${APITRACE_VERSION}/apitrace-${APITRACE_VERSION}${APITRACE_VERSION_DATE}-win64.7z"
7zr x "apitrace-${APITRACE_VERSION}${APITRACE_VERSION_DATE}-win64.7z" \
"apitrace-${APITRACE_VERSION}${APITRACE_VERSION_DATE}-win64/bin/apitrace.exe" \
"apitrace-${APITRACE_VERSION}${APITRACE_VERSION_DATE}-win64/bin/d3dretrace.exe"
mv "apitrace-${APITRACE_VERSION}${APITRACE_VERSION_DATE}-win64" /apitrace-msvc-win64
rm "apitrace-${APITRACE_VERSION}${APITRACE_VERSION_DATE}-win64.7z"
# Add the apitrace path to the registry
wine \
reg add "HKEY_LOCAL_MACHINE\System\CurrentControlSet\Control\Session Manager\Environment" \
/v Path \
/t REG_EXPAND_SZ \
/d "C:\windows\system32;C:\windows;C:\windows\system32\wbem;Z:\apitrace-msvc-win64\bin" \
/f
############### Building ...
. .gitlab-ci/container/container_pre_build.sh
############### Build libdrm
. .gitlab-ci/container/build-libdrm.sh
############### Build Wayland
. .gitlab-ci/container/build-wayland.sh
############### Build parallel-deqp-runner's hang-detection tool
. .gitlab-ci/container/build-hang-detection.sh
############### Build piglit
PIGLIT_BUILD_TARGETS="piglit_replayer" . .gitlab-ci/container/build-piglit.sh
############### Build Fossilize
. .gitlab-ci/container/build-fossilize.sh
############### Build dEQP VK
. .gitlab-ci/container/build-deqp.sh
############### Build gfxreconstruct
. .gitlab-ci/container/build-gfxreconstruct.sh
############### Build VKD3D-Proton
setup_wine "/vkd3d-proton-wine64"
. .gitlab-ci/container/build-vkd3d-proton.sh
############### Uninstall the build software
ccache --show-stats
apt-get purge -y \
$STABLE_EPHEMERAL
apt-get autoremove -y --purge

View File

@@ -1,98 +0,0 @@
#!/bin/bash
set -e
set -o xtrace
EPHEMERAL="
autoconf
automake
bzip2
git
libtool
pkgconfig(epoxy)
pkgconfig(gbm)
unzip
wget
xz
"
dnf install -y --setopt=install_weak_deps=False \
bison \
ccache \
clang-devel \
flex \
gcc \
gcc-c++ \
gettext \
kernel-headers \
llvm-devel \
meson \
"pkgconfig(dri2proto)" \
"pkgconfig(expat)" \
"pkgconfig(glproto)" \
"pkgconfig(libclc)" \
"pkgconfig(libelf)" \
"pkgconfig(libglvnd)" \
"pkgconfig(libomxil-bellagio)" \
"pkgconfig(libselinux)" \
"pkgconfig(libva)" \
"pkgconfig(pciaccess)" \
"pkgconfig(vdpau)" \
"pkgconfig(vulkan)" \
"pkgconfig(x11)" \
"pkgconfig(x11-xcb)" \
"pkgconfig(xcb)" \
"pkgconfig(xcb-dri2)" \
"pkgconfig(xcb-dri3)" \
"pkgconfig(xcb-glx)" \
"pkgconfig(xcb-present)" \
"pkgconfig(xcb-randr)" \
"pkgconfig(xcb-sync)" \
"pkgconfig(xcb-xfixes)" \
"pkgconfig(xdamage)" \
"pkgconfig(xext)" \
"pkgconfig(xfixes)" \
"pkgconfig(xrandr)" \
"pkgconfig(xshmfence)" \
"pkgconfig(xxf86vm)" \
"pkgconfig(zlib)" \
python-unversioned-command \
python3-devel \
python3-mako \
python3-devel \
python3-mako \
vulkan-headers \
$EPHEMERAL
. .gitlab-ci/container/container_pre_build.sh
# dependencies where we want a specific version
export XORG_RELEASES=https://xorg.freedesktop.org/releases/individual
export XORGMACROS_VERSION=util-macros-1.19.0
wget $XORG_RELEASES/util/$XORGMACROS_VERSION.tar.bz2
tar -xvf $XORGMACROS_VERSION.tar.bz2 && rm $XORGMACROS_VERSION.tar.bz2
cd $XORGMACROS_VERSION; ./configure; make install; cd ..
rm -rf $XORGMACROS_VERSION
. .gitlab-ci/container/build-libdrm.sh
. .gitlab-ci/container/build-wayland.sh
pushd /usr/local
git clone https://gitlab.freedesktop.org/mesa/shader-db.git --depth 1
rm -rf shader-db/.git
cd shader-db
make
popd
############### Uninstall the build software
dnf remove -y $EPHEMERAL
. .gitlab-ci/container/container_post_build.sh

View File

@@ -1,237 +0,0 @@
#!/bin/bash
set -e
set -o xtrace
export DEBIAN_FRONTEND=noninteractive
check_minio()
{
MINIO_PATH="${MINIO_HOST}/mesa-lava/$1/${DISTRIBUTION_TAG}/${DEBIAN_ARCH}"
if wget -q --method=HEAD "https://${MINIO_PATH}/done"; then
exit
fi
}
# If remote files are up-to-date, skip rebuilding them
check_minio "${FDO_UPSTREAM_REPO}"
check_minio "${CI_PROJECT_PATH}"
. .gitlab-ci/container/container_pre_build.sh
# Install rust, which we'll be using for deqp-runner. It will be cleaned up at the end.
. .gitlab-ci/container/build-rust.sh
if [[ "$DEBIAN_ARCH" = "arm64" ]]; then
GCC_ARCH="aarch64-linux-gnu"
KERNEL_ARCH="arm64"
DEFCONFIG="arch/arm64/configs/defconfig"
DEVICE_TREES="arch/arm64/boot/dts/rockchip/rk3399-gru-kevin.dtb"
DEVICE_TREES+=" arch/arm64/boot/dts/amlogic/meson-gxl-s805x-libretech-ac.dtb"
DEVICE_TREES+=" arch/arm64/boot/dts/allwinner/sun50i-h6-pine-h64.dtb"
DEVICE_TREES+=" arch/arm64/boot/dts/amlogic/meson-gxm-khadas-vim2.dtb"
DEVICE_TREES+=" arch/arm64/boot/dts/qcom/apq8016-sbc.dtb"
DEVICE_TREES+=" arch/arm64/boot/dts/qcom/apq8096-db820c.dtb"
DEVICE_TREES+=" arch/arm64/boot/dts/amlogic/meson-g12b-a311d-khadas-vim3.dtb"
DEVICE_TREES+=" arch/arm64/boot/dts/mediatek/mt8183-kukui-jacuzzi-juniper-sku16.dtb"
DEVICE_TREES+=" arch/arm64/boot/dts/qcom/sc7180-trogdor-lazor-limozeen-nots.dtb"
KERNEL_IMAGE_NAME="Image"
elif [[ "$DEBIAN_ARCH" = "armhf" ]]; then
GCC_ARCH="arm-linux-gnueabihf"
KERNEL_ARCH="arm"
DEFCONFIG="arch/arm/configs/multi_v7_defconfig"
DEVICE_TREES="arch/arm/boot/dts/rk3288-veyron-jaq.dtb"
DEVICE_TREES+=" arch/arm/boot/dts/sun8i-h3-libretech-all-h3-cc.dtb"
DEVICE_TREES+=" arch/arm/boot/dts/imx6q-cubox-i.dtb"
KERNEL_IMAGE_NAME="zImage"
. .gitlab-ci/container/create-cross-file.sh armhf
else
GCC_ARCH="x86_64-linux-gnu"
KERNEL_ARCH="x86_64"
DEFCONFIG="arch/x86/configs/x86_64_defconfig"
DEVICE_TREES=""
KERNEL_IMAGE_NAME="bzImage"
ARCH_PACKAGES="libva-dev"
fi
# Determine if we're in a cross build.
if [[ -e /cross_file-$DEBIAN_ARCH.txt ]]; then
EXTRA_MESON_ARGS="--cross-file /cross_file-$DEBIAN_ARCH.txt"
EXTRA_CMAKE_ARGS="-DCMAKE_TOOLCHAIN_FILE=/toolchain-$DEBIAN_ARCH.cmake"
if [ $DEBIAN_ARCH = arm64 ]; then
RUST_TARGET="aarch64-unknown-linux-gnu"
elif [ $DEBIAN_ARCH = armhf ]; then
RUST_TARGET="armv7-unknown-linux-gnueabihf"
fi
rustup target add $RUST_TARGET
export EXTRA_CARGO_ARGS="--target $RUST_TARGET"
export ARCH=${KERNEL_ARCH}
export CROSS_COMPILE="${GCC_ARCH}-"
fi
apt-get update
apt-get install -y --no-remove \
${ARCH_PACKAGES} \
automake \
bc \
cmake \
debootstrap \
git \
glslang-tools \
libdrm-dev \
libegl1-mesa-dev \
libfontconfig-dev \
libgbm-dev \
libgl-dev \
libgles2-mesa-dev \
libglu1-mesa-dev \
libglx-dev \
libpng-dev \
libssl-dev \
libudev-dev \
libvulkan-dev \
libwaffle-dev \
libwayland-dev \
libx11-xcb-dev \
libxcb-dri2-0-dev \
libxkbcommon-dev \
ninja-build \
patch \
python-is-python3 \
python3-distutils \
python3-mako \
python3-numpy \
python3-serial \
unzip \
wget
if [[ "$DEBIAN_ARCH" = "armhf" ]]; then
apt-get install -y --no-remove \
libegl1-mesa-dev:armhf \
libelf-dev:armhf \
libgbm-dev:armhf \
libgles2-mesa-dev:armhf \
libpng-dev:armhf \
libudev-dev:armhf \
libvulkan-dev:armhf \
libwaffle-dev:armhf \
libwayland-dev:armhf \
libx11-xcb-dev:armhf \
libxkbcommon-dev:armhf
fi
############### Building
STRIP_CMD="${GCC_ARCH}-strip"
mkdir -p /lava-files/rootfs-${DEBIAN_ARCH}
############### Build apitrace
. .gitlab-ci/container/build-apitrace.sh
mkdir -p /lava-files/rootfs-${DEBIAN_ARCH}/apitrace
mv /apitrace/build /lava-files/rootfs-${DEBIAN_ARCH}/apitrace
rm -rf /apitrace
############### Build dEQP runner
. .gitlab-ci/container/build-deqp-runner.sh
mkdir -p /lava-files/rootfs-${DEBIAN_ARCH}/usr/bin
mv /usr/local/bin/*-runner /lava-files/rootfs-${DEBIAN_ARCH}/usr/bin/.
############### Build dEQP
DEQP_TARGET=surfaceless . .gitlab-ci/container/build-deqp.sh
mv /deqp /lava-files/rootfs-${DEBIAN_ARCH}/.
############### Build SKQP
if [[ "$DEBIAN_ARCH" = "arm64" ]]; then
SKQP_ARCH="arm64" . .gitlab-ci/container/build-skqp.sh
mv /skqp /lava-files/rootfs-${DEBIAN_ARCH}/.
fi
############### Build piglit
PIGLIT_OPTS="-DPIGLIT_BUILD_DMA_BUF_TESTS=ON" . .gitlab-ci/container/build-piglit.sh
mv /piglit /lava-files/rootfs-${DEBIAN_ARCH}/.
############### Build libva tests
if [[ "$DEBIAN_ARCH" = "amd64" ]]; then
. .gitlab-ci/container/build-va-tools.sh
mv /va/bin/* /lava-files/rootfs-${DEBIAN_ARCH}/usr/bin/
fi
############### Build libdrm
EXTRA_MESON_ARGS+=" -D prefix=/libdrm"
. .gitlab-ci/container/build-libdrm.sh
############### Build kernel
. .gitlab-ci/container/build-kernel.sh
############### Delete rust, since the tests won't be compiling anything.
rm -rf /root/.cargo
rm -rf /root/.rustup
############### Create rootfs
set +e
if ! debootstrap \
--variant=minbase \
--arch=${DEBIAN_ARCH} \
--components main,contrib,non-free \
bullseye \
/lava-files/rootfs-${DEBIAN_ARCH}/ \
http://deb.debian.org/debian; then
cat /lava-files/rootfs-${DEBIAN_ARCH}/debootstrap/debootstrap.log
exit 1
fi
set -e
cp .gitlab-ci/container/create-rootfs.sh /lava-files/rootfs-${DEBIAN_ARCH}/.
chroot /lava-files/rootfs-${DEBIAN_ARCH} sh /create-rootfs.sh
rm /lava-files/rootfs-${DEBIAN_ARCH}/create-rootfs.sh
############### Install the built libdrm
# Dependencies pulled during the creation of the rootfs may overwrite
# the built libdrm. Hence, we add it after the rootfs has been already
# created.
mkdir -p /lava-files/rootfs-${DEBIAN_ARCH}/usr/lib/$GCC_ARCH
find /libdrm/ -name lib\*\.so\* | xargs cp -t /lava-files/rootfs-${DEBIAN_ARCH}/usr/lib/$GCC_ARCH/.
mkdir -p /lava-files/rootfs-${DEBIAN_ARCH}/libdrm/
cp -Rp /libdrm/share /lava-files/rootfs-${DEBIAN_ARCH}/libdrm/share
rm -rf /libdrm
if [ ${DEBIAN_ARCH} = arm64 ]; then
# Make a gzipped copy of the Image for db410c.
gzip -k /lava-files/Image
KERNEL_IMAGE_NAME+=" Image.gz"
fi
du -ah /lava-files/rootfs-${DEBIAN_ARCH} | sort -h | tail -100
pushd /lava-files/rootfs-${DEBIAN_ARCH}
tar czf /lava-files/lava-rootfs.tgz .
popd
. .gitlab-ci/container/container_post_build.sh
############### Upload the files!
ci-fairy minio login --token-file "${CI_JOB_JWT_FILE}"
FILES_TO_UPLOAD="lava-rootfs.tgz \
$KERNEL_IMAGE_NAME"
if [[ -n $DEVICE_TREES ]]; then
FILES_TO_UPLOAD="$FILES_TO_UPLOAD $(basename -a $DEVICE_TREES)"
fi
for f in $FILES_TO_UPLOAD; do
ci-fairy minio cp /lava-files/$f \
minio://${MINIO_PATH}/$f
done
touch /lava-files/done
ci-fairy minio cp /lava-files/done minio://${MINIO_PATH}/done

View File

@@ -1,100 +0,0 @@
CONFIG_LOCALVERSION_AUTO=y
CONFIG_DEBUG_KERNEL=y
CONFIG_PWM=y
CONFIG_PM_DEVFREQ=y
CONFIG_OF=y
CONFIG_CROS_EC=y
# abootimg with a 'dummy' rootfs fails with root=/dev/nfs
CONFIG_BLK_DEV_INITRD=n
CONFIG_DEVFREQ_GOV_PERFORMANCE=y
CONFIG_DEVFREQ_GOV_POWERSAVE=y
CONFIG_DEVFREQ_GOV_USERSPACE=y
CONFIG_DEVFREQ_GOV_PASSIVE=y
CONFIG_DRM=y
CONFIG_DRM_PANEL_SIMPLE=y
CONFIG_PWM_CROS_EC=y
CONFIG_BACKLIGHT_PWM=y
# Strip out some stuff we don't need for graphics testing, to reduce
# the build.
CONFIG_CAN=n
CONFIG_WIRELESS=n
CONFIG_RFKILL=n
CONFIG_WLAN=n
CONFIG_REGULATOR_FAN53555=y
CONFIG_REGULATOR=y
CONFIG_REGULATOR_VCTRL=y
CONFIG_KASAN=n
CONFIG_KASAN_INLINE=n
CONFIG_STACKTRACE=n
CONFIG_TMPFS=y
CONFIG_PROVE_LOCKING=n
CONFIG_DEBUG_LOCKDEP=n
CONFIG_SOFTLOCKUP_DETECTOR=y
CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC=y
CONFIG_DETECT_HUNG_TASK=y
CONFIG_USB_USBNET=y
CONFIG_NETDEVICES=y
CONFIG_USB_NET_DRIVERS=y
CONFIG_USB_RTL8152=y
CONFIG_USB_NET_AX8817X=y
CONFIG_USB_NET_SMSC95XX=y
CONFIG_USB_GADGET=y
CONFIG_USB_ETH=y
CONFIG_FW_LOADER_COMPRESS=y
# options for AMD devices
CONFIG_X86_AMD_PLATFORM_DEVICE=y
CONFIG_ACPI_VIDEO=y
CONFIG_X86_AMD_FREQ_SENSITIVITY=y
CONFIG_PINCTRL=y
CONFIG_PINCTRL_AMD=y
CONFIG_DRM_AMDGPU=m
CONFIG_DRM_AMDGPU_SI=y
CONFIG_DRM_AMDGPU_USERPTR=y
CONFIG_DRM_AMD_ACP=n
CONFIG_ACPI_WMI=y
CONFIG_MXM_WMI=y
CONFIG_PARPORT=y
CONFIG_PARPORT_PC=y
CONFIG_PARPORT_SERIAL=y
CONFIG_SERIAL_8250_DW=y
CONFIG_CHROME_PLATFORMS=y
#options for Intel devices
CONFIG_MFD_INTEL_LPSS_PCI=y
#options for KVM guests
CONFIG_FUSE_FS=y
CONFIG_HYPERVISOR_GUEST=y
CONFIG_KVM=y
CONFIG_KVM_GUEST=y
CONFIG_VIRT_DRIVERS=y
CONFIG_VIRTIO_FS=y
CONFIG_DRM_VIRTIO_GPU=y
CONFIG_SERIAL_8250_CONSOLE=y
CONFIG_VIRTIO_NET=y
CONFIG_VIRTIO_CONSOLE=y
CONFIG_PARAVIRT=y
CONFIG_VIRTIO_BLK=y
CONFIG_VIRTUALIZATION=y
CONFIG_VIRTIO=y
CONFIG_VIRTIO_PCI=y
CONFIG_VIRTIO_MMIO=y
CONFIG_VIRTIO_MMIO_CMDLINE_DEVICES=y
CONFIG_CRYPTO_DEV_VIRTIO=y
CONFIG_HW_RANDOM_VIRTIO=y
CONFIG_BLK_MQ_VIRTIO=y
CONFIG_TUN=y

View File

@@ -1 +0,0 @@
lp_test_arit

View File

@@ -1 +0,0 @@
lp_test_format

View File

@@ -1,38 +0,0 @@
#!/bin/sh
set -e
export DEQP_TEMP_DIR="$1"
mount -t proc none /proc
mount -t sysfs none /sys
mkdir -p /dev/pts
mount -t devpts devpts /dev/pts
mount -t tmpfs tmpfs /tmp
. $DEQP_TEMP_DIR/crosvm-env.sh
# .gitlab-ci.yml script variable is using relative paths to install directory,
# so change to that dir before running `crosvm-script`
cd "${CI_PROJECT_DIR}"
# The exception is the dEQP binary, since it needs to run from the directory
# it's in
if [ -d "${DEQP_BIN_DIR}" ]
then
cd "${DEQP_BIN_DIR}"
fi
dmesg --level crit,err,warn -w >> $DEQP_TEMP_DIR/stderr &
set +e
stdbuf -oL sh $DEQP_TEMP_DIR/crosvm-script.sh 2>> $DEQP_TEMP_DIR/stderr >> $DEQP_TEMP_DIR/stdout
echo $? > $DEQP_TEMP_DIR/exit_code
set -e
sync
sleep 1
poweroff -d -n -f || true
sleep 1 # Just in case init would exit before the kernel shuts down the VM

View File

@@ -1,58 +0,0 @@
#!/bin/sh
set -ex
# This script can be called concurrently, pass arguments and env in a
# per-instance tmp dir
DEQP_TEMP_DIR=$(mktemp -d /tmp.XXXXXXXXXX)
export DEQP_TEMP_DIR
# The dEQP binary needs to run from the directory it's in
if [ -n "${1##*.sh}" ] && [ -z "${1##*"deqp"*}" ]; then
DEQP_BIN_DIR=$(dirname "$1")
export DEQP_BIN_DIR
fi
# Securely pass the current variables to the crosvm environment
CI_COMMON="$CI_PROJECT_DIR"/install/common
echo "Variables passed through:"
"${CI_COMMON}"/generate-env.sh | tee ${DEQP_TEMP_DIR}/crosvm-env.sh
CROSVM_KERNEL_ARGS="quiet console=null root=my_root rw rootfstype=virtiofs init=$CI_PROJECT_DIR/install/crosvm-init.sh ip=192.168.30.2::192.168.30.1:255.255.255.0:crosvm:eth0 -- $DEQP_TEMP_DIR"
# Set the crosvm-script as the arguments of the current script.
echo "$@" > $DEQP_TEMP_DIR/crosvm-script.sh
unset DISPLAY
unset XDG_RUNTIME_DIR
/usr/sbin/iptables-legacy -w -t nat -A POSTROUTING -o eth0 -j MASQUERADE
echo 1 > /proc/sys/net/ipv4/ip_forward
# Send output from guest to host
touch $DEQP_TEMP_DIR/stderr $DEQP_TEMP_DIR/stdout
tail -f $DEQP_TEMP_DIR/stderr >> /dev/stderr &
ERR_TAIL_PID=$!
tail -f $DEQP_TEMP_DIR/stdout >> /dev/stdout &
OUT_TAIL_PID=$!
trap "exit \$exit_code" INT TERM
trap "exit_code=\$?; kill $ERR_TAIL_PID $OUT_TAIL_PID; rm -rf $DEQP_TEMP_DIR" EXIT
# We aren't testing LLVMPipe here, so we don't need to validate NIR on the host
NIR_DEBUG="novalidate" LIBGL_ALWAYS_SOFTWARE="true" GALLIUM_DRIVER="$CROSVM_GALLIUM_DRIVER" crosvm run \
--gpu "$CROSVM_GPU_ARGS" \
-m 4096 \
-c 2 \
--disable-sandbox \
--shared-dir /:my_root:type=fs:writeback=true:timeout=60:cache=always \
--host_ip=192.168.30.1 --netmask=255.255.255.0 --mac "AA:BB:CC:00:00:12" \
-p "$CROSVM_KERNEL_ARGS" \
/lava-files/bzImage > $DEQP_TEMP_DIR/crosvm 2>&1
RET=$(cat $DEQP_TEMP_DIR/exit_code || true)
# Got no exit code from the script, show crosvm output to help with debugging
[ -n "$RET" ] || cat $DEQP_TEMP_DIR/crosvm || true
exit ${RET:-1}

View File

@@ -1,216 +0,0 @@
#!/bin/sh
set -ex
# Needed so configuration files can contain paths to files in /install
ln -sf $CI_PROJECT_DIR/install /install
if [ -z "$GPU_VERSION" ]; then
echo 'GPU_VERSION must be set to something like "llvmpipe" or "freedreno-a630" (the name used in .gitlab-ci/gpu-version-*.txt)'
exit 1
fi
INSTALL=`pwd`/install
# Set up the driver environment.
export LD_LIBRARY_PATH=`pwd`/install/lib/
export EGL_PLATFORM=surfaceless
export VK_ICD_FILENAMES=`pwd`/install/share/vulkan/icd.d/"$VK_DRIVER"_icd.${VK_CPU:-`uname -m`}.json
RESULTS=`pwd`/${DEQP_RESULTS_DIR:-results}
mkdir -p $RESULTS
HANG_DETECTION_CMD=""
if [ -z "$DEQP_SUITE" ]; then
if [ -z "$DEQP_VER" ]; then
echo 'DEQP_SUITE must be set to the name of your deqp-gpu_version.toml, or DEQP_VER must be set to something like "gles2", "gles31-khr" or "vk" for the test run'
exit 1
fi
DEQP_WIDTH=${DEQP_WIDTH:-256}
DEQP_HEIGHT=${DEQP_HEIGHT:-256}
DEQP_CONFIG=${DEQP_CONFIG:-rgba8888d24s8ms0}
DEQP_VARIANT=${DEQP_VARIANT:-master}
DEQP_OPTIONS="$DEQP_OPTIONS --deqp-surface-width=$DEQP_WIDTH --deqp-surface-height=$DEQP_HEIGHT"
DEQP_OPTIONS="$DEQP_OPTIONS --deqp-surface-type=${DEQP_SURFACE_TYPE:-pbuffer}"
DEQP_OPTIONS="$DEQP_OPTIONS --deqp-gl-config-name=$DEQP_CONFIG"
DEQP_OPTIONS="$DEQP_OPTIONS --deqp-visibility=hidden"
if [ "$DEQP_VER" = "vk" -a -z "$VK_DRIVER" ]; then
echo 'VK_DRIVER must be to something like "radeon" or "intel" for the test run'
exit 1
fi
# Generate test case list file.
if [ "$DEQP_VER" = "vk" ]; then
MUSTPASS=/deqp/mustpass/vk-$DEQP_VARIANT.txt
DEQP=/deqp/external/vulkancts/modules/vulkan/deqp-vk
HANG_DETECTION_CMD="/parallel-deqp-runner/build/bin/hang-detection"
elif [ "$DEQP_VER" = "gles2" -o "$DEQP_VER" = "gles3" -o "$DEQP_VER" = "gles31" -o "$DEQP_VER" = "egl" ]; then
MUSTPASS=/deqp/mustpass/$DEQP_VER-$DEQP_VARIANT.txt
DEQP=/deqp/modules/$DEQP_VER/deqp-$DEQP_VER
elif [ "$DEQP_VER" = "gles2-khr" -o "$DEQP_VER" = "gles3-khr" -o "$DEQP_VER" = "gles31-khr" -o "$DEQP_VER" = "gles32-khr" ]; then
MUSTPASS=/deqp/mustpass/$DEQP_VER-$DEQP_VARIANT.txt
DEQP=/deqp/external/openglcts/modules/glcts
else
MUSTPASS=/deqp/mustpass/$DEQP_VER-$DEQP_VARIANT.txt
DEQP=/deqp/external/openglcts/modules/glcts
fi
cp $MUSTPASS /tmp/case-list.txt
# If the caselist is too long to run in a reasonable amount of time, let the job
# specify what fraction (1/n) of the caselist we should run. Note: N~M is a gnu
# sed extension to match every nth line (first line is #1).
if [ -n "$DEQP_FRACTION" ]; then
sed -ni 1~$DEQP_FRACTION"p" /tmp/case-list.txt
fi
# If the job is parallel at the gitab job level, take the corresponding fraction
# of the caselist.
if [ -n "$CI_NODE_INDEX" ]; then
sed -ni $CI_NODE_INDEX~$CI_NODE_TOTAL"p" /tmp/case-list.txt
fi
if [ -n "$DEQP_CASELIST_FILTER" ]; then
sed -ni "/$DEQP_CASELIST_FILTER/p" /tmp/case-list.txt
fi
if [ -n "$DEQP_CASELIST_INV_FILTER" ]; then
sed -ni "/$DEQP_CASELIST_INV_FILTER/!p" /tmp/case-list.txt
fi
if [ ! -s /tmp/case-list.txt ]; then
echo "Caselist generation failed"
exit 1
fi
fi
if [ -e "$INSTALL/$GPU_VERSION-fails.txt" ]; then
DEQP_RUNNER_OPTIONS="$DEQP_RUNNER_OPTIONS --baseline $INSTALL/$GPU_VERSION-fails.txt"
fi
# Default to an empty known flakes file if it doesn't exist.
touch $INSTALL/$GPU_VERSION-flakes.txt
if [ -n "$VK_DRIVER" ] && [ -e "$INSTALL/$VK_DRIVER-skips.txt" ]; then
DEQP_SKIPS="$DEQP_SKIPS $INSTALL/$VK_DRIVER-skips.txt"
fi
if [ -n "$GALLIUM_DRIVER" ] && [ -e "$INSTALL/$GALLIUM_DRIVER-skips.txt" ]; then
DEQP_SKIPS="$DEQP_SKIPS $INSTALL/$GALLIUM_DRIVER-skips.txt"
fi
if [ -n "$DRIVER_NAME" ] && [ -e "$INSTALL/$DRIVER_NAME-skips.txt" ]; then
DEQP_SKIPS="$DEQP_SKIPS $INSTALL/$DRIVER_NAME-skips.txt"
fi
if [ -e "$INSTALL/$GPU_VERSION-skips.txt" ]; then
DEQP_SKIPS="$DEQP_SKIPS $INSTALL/$GPU_VERSION-skips.txt"
fi
set +e
report_load() {
echo "System load: $(cut -d' ' -f1-3 < /proc/loadavg)"
echo "# of CPU cores: $(cat /proc/cpuinfo | grep processor | wc -l)"
}
# wrapper to supress +x to avoid spamming the log
quiet() {
set +x
"$@"
set -x
}
if [ "$GALLIUM_DRIVER" = "virpipe" ]; then
# deqp is to use virpipe, and virgl_test_server llvmpipe
export GALLIUM_DRIVER="$GALLIUM_DRIVER"
VTEST_ARGS="--use-egl-surfaceless"
if [ "$VIRGL_HOST_API" = "GLES" ]; then
VTEST_ARGS="$VTEST_ARGS --use-gles"
fi
GALLIUM_DRIVER=llvmpipe \
virgl_test_server $VTEST_ARGS >$RESULTS/vtest-log.txt 2>&1 &
sleep 1
fi
if [ -z "$DEQP_SUITE" ]; then
if [ -n "$DEQP_EXPECTED_RENDERER" ]; then
export DEQP_RUNNER_OPTIONS="$DEQP_RUNNER_OPTIONS --renderer-check "$DEQP_EXPECTED_RENDERER""
fi
if [ $DEQP_VER != vk -a $DEQP_VER != egl ]; then
export DEQP_RUNNER_OPTIONS="$DEQP_RUNNER_OPTIONS --version-check `cat $INSTALL/VERSION | sed 's/[() ]/./g'`"
fi
deqp-runner \
run \
--deqp $DEQP \
--output $RESULTS \
--caselist /tmp/case-list.txt \
--skips $INSTALL/all-skips.txt $DEQP_SKIPS \
--flakes $INSTALL/$GPU_VERSION-flakes.txt \
--testlog-to-xml /deqp/executor/testlog-to-xml \
--jobs ${FDO_CI_CONCURRENT:-4} \
$DEQP_RUNNER_OPTIONS \
-- \
$DEQP_OPTIONS
else
deqp-runner \
suite \
--suite $INSTALL/deqp-$DEQP_SUITE.toml \
--output $RESULTS \
--skips $INSTALL/all-skips.txt $DEQP_SKIPS \
--flakes $INSTALL/$GPU_VERSION-flakes.txt \
--testlog-to-xml /deqp/executor/testlog-to-xml \
--fraction-start $CI_NODE_INDEX \
--fraction $CI_NODE_TOTAL \
--jobs ${FDO_CI_CONCURRENT:-4} \
$DEQP_RUNNER_OPTIONS
fi
DEQP_EXITCODE=$?
quiet report_load
# Remove all but the first 50 individual XML files uploaded as artifacts, to
# save fd.o space when you break everything.
find $RESULTS -name \*.xml | \
sort -n |
sed -n '1,+49!p' | \
xargs rm -f
# If any QPA XMLs are there, then include the XSL/CSS in our artifacts.
find $RESULTS -name \*.xml \
-exec cp /deqp/testlog.css /deqp/testlog.xsl "$RESULTS/" ";" \
-quit
deqp-runner junit \
--testsuite dEQP \
--results $RESULTS/failures.csv \
--output $RESULTS/junit.xml \
--limit 50 \
--template "See https://$CI_PROJECT_ROOT_NAMESPACE.pages.freedesktop.org/-/$CI_PROJECT_NAME/-/jobs/$CI_JOB_ID/artifacts/results/{{testcase}}.xml"
# Report the flakes to the IRC channel for monitoring (if configured):
if [ -n "$FLAKES_CHANNEL" ]; then
python3 $INSTALL/report-flakes.py \
--host irc.oftc.net \
--port 6667 \
--results $RESULTS/results.csv \
--known-flakes $INSTALL/$GPU_VERSION-flakes.txt \
--channel "$FLAKES_CHANNEL" \
--runner "$CI_RUNNER_DESCRIPTION" \
--job "$CI_JOB_ID" \
--url "$CI_JOB_URL" \
--branch "${CI_MERGE_REQUEST_SOURCE_BRANCH_NAME:-$CI_COMMIT_BRANCH}" \
--branch-title "${CI_MERGE_REQUEST_TITLE:-$CI_COMMIT_TITLE}"
fi
exit $DEQP_EXITCODE

View File

@@ -1 +0,0 @@
../docs/ci

View File

@@ -1,36 +0,0 @@
#!/bin/bash
set +e
set -o xtrace
# if we run this script outside of gitlab-ci for testing, ensure
# we got meaningful variables
CI_PROJECT_DIR=${CI_PROJECT_DIR:-$(mktemp -d)/mesa}
if [[ -e $CI_PROJECT_DIR/.git ]]
then
echo "Repository already present, skip cache download"
exit
fi
TMP_DIR=$(mktemp -d)
echo "Downloading archived master..."
/usr/bin/wget -O $TMP_DIR/mesa.tar.gz \
https://${MINIO_HOST}/git-cache/${FDO_UPSTREAM_REPO}/mesa.tar.gz
# check wget error code
if [[ $? -ne 0 ]]
then
echo "Repository cache not available"
exit
fi
set -e
rm -rf "$CI_PROJECT_DIR"
echo "Extracting tarball into '$CI_PROJECT_DIR'..."
mkdir -p "$CI_PROJECT_DIR"
tar xzf "$TMP_DIR/mesa.tar.gz" -C "$CI_PROJECT_DIR"
rm -rf "$TMP_DIR"
chmod a+w "$CI_PROJECT_DIR"

View File

@@ -1,20 +0,0 @@
#!/bin/sh
set -ex
if [ -z "$VK_DRIVER" ]; then
echo 'VK_DRIVER must be to something like "radeon" or "intel" for the test run'
exit 1
fi
INSTALL=`pwd`/install
# Set up the driver environment.
export LD_LIBRARY_PATH=`pwd`/install/lib/
export VK_ICD_FILENAMES=`pwd`/install/share/vulkan/icd.d/"$VK_DRIVER"_icd.x86_64.json
# To store Fossilize logs on failure.
RESULTS=`pwd`/results
mkdir -p results
"$INSTALL/fossils/fossils.sh" "$INSTALL/fossils.yml" "$RESULTS"

View File

@@ -1,10 +0,0 @@
fossils-db:
repo: "https://gitlab.freedesktop.org/hakzsam/fossils-db"
commit: "5626cedcb58bd95a7b79a9664651818aea92b21c"
fossils:
- path: sascha-willems/database.foz
- path: parallel-rdp/small_subgroup.foz
- path: parallel-rdp/small_uber_subgroup.foz
- path: parallel-rdp/subgroup.foz
- path: parallel-rdp/uber_subgroup.foz

View File

@@ -1,77 +0,0 @@
#!/usr/bin/env bash
FOSSILS_SCRIPT_DIR="$(dirname "$(readlink -f "$0")")"
FOSSILS_YAML="$(readlink -f "$1")"
FOSSILS_RESULTS="$2"
clone_fossils_db()
{
local repo="$1"
local commit="$2"
rm -rf fossils-db
git clone --no-checkout "$repo" fossils-db
(cd fossils-db; git reset "$commit" || git reset "origin/$commit")
}
query_fossils_yaml()
{
python3 "$FOSSILS_SCRIPT_DIR/query_fossils_yaml.py" \
--file "$FOSSILS_YAML" "$@"
}
create_clean_git()
{
rm -rf .clean_git
cp -R .git .clean_git
}
restore_clean_git()
{
rm -rf .git
cp -R .clean_git .git
}
fetch_fossil()
{
local fossil="${1//,/?}"
echo -n "[fetch_fossil] Fetching $1... "
local output=$(git lfs pull -I "$fossil" 2>&1)
local ret=0
if [[ $? -ne 0 || ! -f "$1" ]]; then
echo "ERROR"
echo "$output"
ret=1
else
echo "OK"
fi
restore_clean_git
return $ret
}
if [[ -n "$(query_fossils_yaml fossils_db_repo)" ]]; then
clone_fossils_db "$(query_fossils_yaml fossils_db_repo)" \
"$(query_fossils_yaml fossils_db_commit)"
cd fossils-db
else
echo "Warning: No fossils-db entry in $FOSSILS_YAML, assuming fossils-db is current directory"
fi
# During git operations various git objects get created which
# may take up significant space. Store a clean .git instance,
# which we restore after various git operations to keep our
# storage consumption low.
create_clean_git
for fossil in $(query_fossils_yaml fossils)
do
fetch_fossil "$fossil" || exit $?
fossilize-replay --num-threads 4 $fossil 1>&2 2> $FOSSILS_RESULTS/fossil_replay.txt
if [ $? != 0 ]; then
echo "Replay of $fossil failed"
grep "pipeline crashed or hung" $FOSSILS_RESULTS/fossil_replay.txt
exit 1
fi
rm $fossil
done
exit $ret

View File

@@ -1,69 +0,0 @@
#!/usr/bin/python3
# Copyright (c) 2019 Collabora Ltd
# Copyright (c) 2020 Valve Corporation
#
# Permission is hereby granted, free of charge, to any person obtaining a
# copy of this software and associated documentation files (the "Software"),
# to deal in the Software without restriction, including without limitation
# the rights to use, copy, modify, merge, publish, distribute, sublicense,
# and/or sell copies of the Software, and to permit persons to whom the
# Software is 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 Software.
#
# THE SOFTWARE IS 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 SOFTWARE OR THE USE OR
# OTHER DEALINGS IN THE SOFTWARE.
#
# SPDX-License-Identifier: MIT
import argparse
import yaml
def cmd_fossils_db_repo(args):
with open(args.file, 'r') as f:
y = yaml.safe_load(f)
print(y['fossils-db']['repo'])
def cmd_fossils_db_commit(args):
with open(args.file, 'r') as f:
y = yaml.safe_load(f)
print(y['fossils-db']['commit'])
def cmd_fossils(args):
with open(args.file, 'r') as f:
y = yaml.safe_load(f)
fossils = list(y['fossils'])
if len(fossils) == 0:
return
print('\n'.join((t['path'] for t in fossils)))
def main():
parser = argparse.ArgumentParser()
parser.add_argument('--file', required=True,
help='the name of the yaml file')
subparsers = parser.add_subparsers(help='sub-command help')
parser_fossils_db_repo = subparsers.add_parser('fossils_db_repo')
parser_fossils_db_repo.set_defaults(func=cmd_fossils_db_repo)
parser_fossils_db_commit = subparsers.add_parser('fossils_db_commit')
parser_fossils_db_commit.set_defaults(func=cmd_fossils_db_commit)
parser_fossils = subparsers.add_parser('fossils')
parser_fossils.set_defaults(func=cmd_fossils)
args = parser.parse_args()
args.func(args)
if __name__ == "__main__":
main()

View File

@@ -1,70 +0,0 @@
#!/bin/sh
set -ex
INSTALL=`pwd`/install
# Set up the driver environment.
export LD_LIBRARY_PATH=`pwd`/install/lib/
export LIBVA_DRIVERS_PATH=`pwd`/install/lib/dri/
# libva spams driver open info by default, and that happens per testcase.
export LIBVA_MESSAGING_LEVEL=1
if [ -e "$INSTALL/$GPU_VERSION-fails.txt" ]; then
GTEST_RUNNER_OPTIONS="$GTEST_RUNNER_OPTIONS --baseline $INSTALL/$GPU_VERSION-fails.txt"
fi
# Default to an empty known flakes file if it doesn't exist.
touch $INSTALL/$GPU_VERSION-flakes.txt
if [ -n "$GALLIUM_DRIVER" ] && [ -e "$INSTALL/$GALLIUM_DRIVER-skips.txt" ]; then
GTEST_SKIPS="$GTEST_SKIPS --skips $INSTALL/$GALLIUM_DRIVER-skips.txt"
fi
if [ -n "$DRIVER_NAME" ] && [ -e "$INSTALL/$DRIVER_NAME-skips.txt" ]; then
GTEST_SKIPS="$GTEST_SKIPS --skips $INSTALL/$DRIVER_NAME-skips.txt"
fi
if [ -e "$INSTALL/$GPU_VERSION-skips.txt" ]; then
GTEST_SKIPS="$GTEST_SKIPS --skips $INSTALL/$GPU_VERSION-skips.txt"
fi
set +e
gtest-runner \
run \
--gtest $GTEST \
--output ${GTEST_RESULTS_DIR:-results} \
--jobs ${FDO_CI_CONCURRENT:-4} \
$GTEST_SKIPS \
--flakes $INSTALL/$GPU_VERSION-flakes.txt \
--fraction-start ${CI_NODE_INDEX:-1} \
--fraction $((${CI_NODE_TOTAL:-1} * ${GTEST_FRACTION:-1})) \
--env "LD_PRELOAD=$TEST_LD_PRELOAD" \
$GTEST_RUNNER_OPTIONS
GTEST_EXITCODE=$?
deqp-runner junit \
--testsuite gtest \
--results $RESULTS/failures.csv \
--output $RESULTS/junit.xml \
--limit 50 \
--template "See https://$CI_PROJECT_ROOT_NAMESPACE.pages.freedesktop.org/-/$CI_PROJECT_NAME/-/jobs/$CI_JOB_ID/artifacts/results/{{testcase}}.xml"
# Report the flakes to the IRC channel for monitoring (if configured):
if [ -n "$FLAKES_CHANNEL" ]; then
python3 $INSTALL/report-flakes.py \
--host irc.oftc.net \
--port 6667 \
--results $RESULTS/results.csv \
--known-flakes $INSTALL/$GPU_VERSION-flakes.txt \
--channel "$FLAKES_CHANNEL" \
--runner "$CI_RUNNER_DESCRIPTION" \
--job "$CI_JOB_ID" \
--url "$CI_JOB_URL" \
--branch "${CI_MERGE_REQUEST_SOURCE_BRANCH_NAME:-$CI_COMMIT_BRANCH}" \
--branch-title "${CI_MERGE_REQUEST_TITLE:-$CI_COMMIT_TITLE}"
fi
exit $GTEST_EXITCODE

View File

@@ -1,21 +0,0 @@
variables:
DEBIAN_X86_BUILD_BASE_IMAGE: "debian/x86_build-base"
DEBIAN_BASE_TAG: "2022-01-02-deqp-runner"
DEBIAN_X86_BUILD_IMAGE_PATH: "debian/x86_build"
DEBIAN_BUILD_TAG: "2021-12-31-keep-cmake"
DEBIAN_X86_TEST_BASE_IMAGE: "debian/x86_test-base"
DEBIAN_X86_TEST_IMAGE_PATH: "debian/x86_test-gl"
DEBIAN_X86_TEST_GL_TAG: "2022-01-18-kernel"
DEBIAN_X86_TEST_VK_TAG: "2022-01-02-deqp-runner"
FEDORA_X86_BUILD_TAG: "2021-12-31-refactor"
KERNEL_ROOTFS_TAG: "2022-01-18-kernel"
WINDOWS_X64_BUILD_PATH: "windows/x64_build"
WINDOWS_X64_BUILD_TAG: "2022-20-02-base_split"
WINDOWS_X64_TEST_PATH: "windows/x64_test"
WINDOWS_X64_TEST_TAG: "2022-20-02-base_split"

View File

@@ -1,120 +0,0 @@
.lava-test:
extends:
- .ci-run-policy
# Cancel job if a newer commit is pushed to the same branch
interruptible: true
variables:
GIT_STRATEGY: none # testing doesn't build anything from source
FDO_CI_CONCURRENT: 6 # should be replaced by per-machine definitions
DEQP_VER: gles2
# proxy used to cache data locally
FDO_HTTP_CACHE_URI: "http://caching-proxy/cache/?uri="
# base system generated by the container build job, shared between many pipelines
BASE_SYSTEM_HOST_PREFIX: "${MINIO_HOST}/mesa-lava"
BASE_SYSTEM_MAINLINE_HOST_PATH: "${BASE_SYSTEM_HOST_PREFIX}/${FDO_UPSTREAM_REPO}/${DISTRIBUTION_TAG}/${ARCH}"
BASE_SYSTEM_FORK_HOST_PATH: "${BASE_SYSTEM_HOST_PREFIX}/${CI_PROJECT_PATH}/${DISTRIBUTION_TAG}/${ARCH}"
# per-job build artifacts
MESA_BUILD_PATH: "${PIPELINE_ARTIFACTS_BASE}/mesa-${ARCH}.tar.gz"
JOB_ROOTFS_OVERLAY_PATH: "${JOB_ARTIFACTS_BASE}/job-rootfs-overlay.tar.gz"
JOB_RESULTS_PATH: "${JOB_ARTIFACTS_BASE}/results.tar.gz"
PIGLIT_NO_WINDOW: 1
VISIBILITY_GROUP: "Collabora+fdo"
script:
- ./artifacts/lava/lava-submit.sh
artifacts:
name: "mesa_${CI_JOB_NAME}"
when: always
paths:
- results/
exclude:
- results/*.shader_cache
after_script:
- wget -q "https://${JOB_RESULTS_PATH}" -O- | tar -xz
.lava-test:armhf:
variables:
ARCH: armhf
KERNEL_IMAGE_NAME: zImage
KERNEL_IMAGE_TYPE: "zimage"
BOOT_METHOD: u-boot
HWCI_TEST_SCRIPT: "/install/deqp-runner.sh"
extends:
- .use-debian/arm_build # for same $MESA_ARTIFACTS_TAG as in kernel+rootfs_armhf
- .use-debian/x86_build
- .lava-test
- .use-kernel+rootfs-arm
needs:
- kernel+rootfs_armhf
- debian/x86_build
- debian-armhf
.lava-test:arm64:
variables:
ARCH: arm64
KERNEL_IMAGE_NAME: Image
KERNEL_IMAGE_TYPE: "image"
BOOT_METHOD: u-boot
HWCI_TEST_SCRIPT: "/install/deqp-runner.sh"
extends:
- .use-debian/arm_build # for same $MESA_ARTIFACTS_TAG as in kernel+rootfs_arm64
- .use-debian/x86_build
- .lava-test
- .use-kernel+rootfs-arm
dependencies:
- debian-arm64
needs:
- kernel+rootfs_arm64
- debian/x86_build
- debian-arm64
.lava-test:amd64:
variables:
ARCH: amd64
KERNEL_IMAGE_NAME: bzImage
KERNEL_IMAGE_TYPE: "zimage"
BOOT_METHOD: u-boot
HWCI_TEST_SCRIPT: "/install/deqp-runner.sh"
extends:
- .use-debian/x86_build-base # for same $MESA_ARTIFACTS_BASE_TAG as in kernel+rootfs_amd64
- .use-debian/x86_build
- .lava-test
- .use-kernel+rootfs-amd64
needs:
- kernel+rootfs_amd64
- debian-testing
.lava-traces-base:
variables:
HWCI_TEST_SCRIPT: "/install/piglit/piglit-traces.sh"
artifacts:
reports:
junit: results/junit.xml
.lava-piglit:
variables:
PIGLIT_REPLAY_DEVICE_NAME: "gl-${GPU_VERSION}"
PIGLIT_RESULTS: "${GPU_VERSION}-${PIGLIT_PROFILES}"
HWCI_TEST_SCRIPT: "/install/piglit/piglit-runner.sh"
.lava-piglit-traces:amd64:
extends:
- .lava-test:amd64
- .lava-piglit
- .lava-traces-base
.lava-piglit-traces:armhf:
extends:
- .lava-test:armhf
- .lava-piglit
- .lava-traces-base
.lava-piglit-traces:arm64:
extends:
- .lava-test:arm64
- .lava-piglit
- .lava-traces-base
.lava-piglit:amd64:
extends:
- .lava-test:amd64
- .lava-piglit

View File

@@ -1,44 +0,0 @@
#!/bin/bash
set -e
set -x
# Try to use the kernel and rootfs built in mainline first, so we're more
# likely to hit cache
if wget -q --method=HEAD "https://${BASE_SYSTEM_MAINLINE_HOST_PATH}/done"; then
BASE_SYSTEM_HOST_PATH="${BASE_SYSTEM_MAINLINE_HOST_PATH}"
else
BASE_SYSTEM_HOST_PATH="${BASE_SYSTEM_FORK_HOST_PATH}"
fi
rm -rf results
mkdir -p results/job-rootfs-overlay/
cp artifacts/ci-common/capture-devcoredump.sh results/job-rootfs-overlay/
cp artifacts/ci-common/init-*.sh results/job-rootfs-overlay/
artifacts/ci-common/generate-env.sh > results/job-rootfs-overlay/set-job-env-vars.sh
tar zcf job-rootfs-overlay.tar.gz -C results/job-rootfs-overlay/ .
ci-fairy minio login --token-file "${CI_JOB_JWT_FILE}"
ci-fairy minio cp job-rootfs-overlay.tar.gz "minio://${JOB_ROOTFS_OVERLAY_PATH}"
touch results/lava.log
tail -f results/lava.log &
artifacts/lava/lava_job_submitter.py \
--dump-yaml \
--pipeline-info "$CI_JOB_NAME: $CI_PIPELINE_URL on $CI_COMMIT_REF_NAME ${CI_NODE_INDEX}/${CI_NODE_TOTAL}" \
--base-system-url-prefix "https://${BASE_SYSTEM_HOST_PATH}" \
--mesa-build-url "${FDO_HTTP_CACHE_URI:-}https://${MESA_BUILD_PATH}" \
--job-rootfs-overlay-url "${FDO_HTTP_CACHE_URI:-}https://${JOB_ROOTFS_OVERLAY_PATH}" \
--job-artifacts-base ${JOB_ARTIFACTS_BASE} \
--job-timeout ${JOB_TIMEOUT:-30} \
--first-stage-init artifacts/ci-common/init-stage1.sh \
--ci-project-dir ${CI_PROJECT_DIR} \
--device-type ${DEVICE_TYPE} \
--dtb ${DTB} \
--jwt-file "${CI_JOB_JWT_FILE}" \
--kernel-image-name ${KERNEL_IMAGE_NAME} \
--kernel-image-type "${KERNEL_IMAGE_TYPE}" \
--boot-method ${BOOT_METHOD} \
--visibility-group ${VISIBILITY_GROUP} \
--lava-tags "${LAVA_TAGS}" >> results/lava.log

View File

@@ -1,373 +0,0 @@
#!/usr/bin/env python3
#
# Copyright (C) 2020, 2021 Collabora Limited
# Author: Gustavo Padovan <gustavo.padovan@collabora.com>
#
# Permission is hereby granted, free of charge, to any person obtaining a
# copy of this software and associated documentation files (the "Software"),
# to deal in the Software without restriction, including without limitation
# the rights to use, copy, modify, merge, publish, distribute, sublicense,
# and/or sell copies of the Software, and to permit persons to whom the
# Software is furnished to do so, subject to the following conditions:
#
# The above copyright notice and this permission notice (including the next
# paragraph) shall be included in all copies or substantial portions of the
# Software.
#
# THE SOFTWARE IS 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 SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
# SOFTWARE.
"""Send a job to LAVA, track it and collect log back"""
import argparse
import pathlib
import sys
import time
import traceback
import urllib.parse
import xmlrpc
from datetime import datetime, timedelta
import lavacli
import yaml
from lavacli.utils import loader
# Timeout in minutes to decide if the device from the dispatched LAVA job has
# hung or not due to the lack of new log output.
DEVICE_HANGING_TIMEOUT_MIN = 5
# How many seconds the script should wait before try a new polling iteration to
# check if the dispatched LAVA job is running or waiting in the job queue.
WAIT_FOR_DEVICE_POLLING_TIME_SEC = 10
# How many seconds to wait between log output LAVA RPC calls.
LOG_POLLING_TIME_SEC = 5
# How many retries should be made when a timeout happen.
NUMBER_OF_RETRIES_TIMEOUT_DETECTION = 2
def print_log(msg):
print("{}: {}".format(datetime.now(), msg))
def fatal_err(msg):
print_log(msg)
sys.exit(1)
def hide_sensitive_data(yaml_data, hide_tag="HIDEME"):
out_data = ""
for line in yaml_data.splitlines(True):
if hide_tag in line:
continue
out_data += line
return out_data
def generate_lava_yaml(args):
# General metadata and permissions, plus also inexplicably kernel arguments
values = {
'job_name': 'mesa: {}'.format(args.pipeline_info),
'device_type': args.device_type,
'visibility': { 'group': [ args.visibility_group ] },
'priority': 75,
'context': {
'extra_nfsroot_args': ' init=/init rootwait minio_results={}'.format(args.job_artifacts_base)
},
'timeouts': {
'job': {
'minutes': args.job_timeout
}
},
}
if args.lava_tags:
values['tags'] = args.lava_tags.split(',')
# URLs to our kernel rootfs to boot from, both generated by the base
# container build
deploy = {
'timeout': { 'minutes': 10 },
'to': 'tftp',
'os': 'oe',
'kernel': {
'url': '{}/{}'.format(args.base_system_url_prefix, args.kernel_image_name),
},
'nfsrootfs': {
'url': '{}/lava-rootfs.tgz'.format(args.base_system_url_prefix),
'compression': 'gz',
}
}
if args.kernel_image_type:
deploy['kernel']['type'] = args.kernel_image_type
if args.dtb:
deploy['dtb'] = {
'url': '{}/{}.dtb'.format(args.base_system_url_prefix, args.dtb)
}
# always boot over NFS
boot = {
'timeout': { 'minutes': 25 },
'method': args.boot_method,
'commands': 'nfs',
'prompts': ['lava-shell:'],
}
# skeleton test definition: only declaring each job as a single 'test'
# since LAVA's test parsing is not useful to us
test = {
'timeout': { 'minutes': args.job_timeout },
'failure_retry': 1,
'definitions': [ {
'name': 'mesa',
'from': 'inline',
'path': 'inline/mesa.yaml',
'repository': {
'metadata': {
'name': 'mesa',
'description': 'Mesa test plan',
'os': [ 'oe' ],
'scope': [ 'functional' ],
'format': 'Lava-Test Test Definition 1.0',
},
'parse': {
'pattern': r'hwci: (?P<test_case_id>\S*):\s+(?P<result>(pass|fail))'
},
'run': {
},
},
} ],
}
# job execution script:
# - inline .gitlab-ci/common/init-stage1.sh
# - fetch and unpack per-pipeline build artifacts from build job
# - fetch and unpack per-job environment from lava-submit.sh
# - exec .gitlab-ci/common/init-stage2.sh
init_lines = []
with open(args.first_stage_init, 'r') as init_sh:
init_lines += [ x.rstrip() for x in init_sh if not x.startswith('#') and x.rstrip() ]
with open(args.jwt_file) as jwt_file:
init_lines += [
"set +x",
f'echo -n "{jwt_file.read()}" > "{args.jwt_file}" # HIDEME',
"set -x",
]
init_lines += [
'mkdir -p {}'.format(args.ci_project_dir),
'wget -S --progress=dot:giga -O- {} | tar -xz -C {}'.format(args.mesa_build_url, args.ci_project_dir),
'wget -S --progress=dot:giga -O- {} | tar -xz -C /'.format(args.job_rootfs_overlay_url),
f'echo "export CI_JOB_JWT_FILE={args.jwt_file}" >> /set-job-env-vars.sh',
'exec /init-stage2.sh',
]
test['definitions'][0]['repository']['run']['steps'] = init_lines
values['actions'] = [
{ 'deploy': deploy },
{ 'boot': boot },
{ 'test': test },
]
return yaml.dump(values, width=10000000)
def setup_lava_proxy():
config = lavacli.load_config("default")
uri, usr, tok = (config.get(key) for key in ("uri", "username", "token"))
uri_obj = urllib.parse.urlparse(uri)
uri_str = "{}://{}:{}@{}{}".format(uri_obj.scheme, usr, tok, uri_obj.netloc, uri_obj.path)
transport = lavacli.RequestsTransport(
uri_obj.scheme,
config.get("proxy"),
config.get("timeout", 120.0),
config.get("verify_ssl_cert", True),
)
proxy = xmlrpc.client.ServerProxy(
uri_str, allow_none=True, transport=transport)
print_log("Proxy for {} created.".format(config['uri']))
return proxy
def _call_proxy(fn, *args):
retries = 60
for n in range(1, retries + 1):
try:
return fn(*args)
except xmlrpc.client.ProtocolError as err:
if n == retries:
traceback.print_exc()
fatal_err("A protocol error occurred (Err {} {})".format(err.errcode, err.errmsg))
else:
time.sleep(15)
pass
except xmlrpc.client.Fault as err:
traceback.print_exc()
fatal_err("FATAL: Fault: {} (code: {})".format(err.faultString, err.faultCode))
def get_job_results(proxy, job_id, test_suite, test_case):
# Look for infrastructure errors and retry if we see them.
results_yaml = _call_proxy(proxy.results.get_testjob_results_yaml, job_id)
results = yaml.load(results_yaml, Loader=loader(False))
for res in results:
metadata = res['metadata']
if not 'result' in metadata or metadata['result'] != 'fail':
continue
if 'error_type' in metadata and metadata['error_type'] == "Infrastructure":
print_log("LAVA job {} failed with Infrastructure Error. Retry.".format(job_id))
return False
if 'case' in metadata and metadata['case'] == "validate":
print_log("LAVA job {} failed validation (possible download error). Retry.".format(job_id))
return False
results_yaml = _call_proxy(proxy.results.get_testcase_results_yaml, job_id, test_suite, test_case)
results = yaml.load(results_yaml, Loader=loader(False))
if not results:
fatal_err("LAVA: no result for test_suite '{}', test_case '{}'".format(test_suite, test_case))
print_log("LAVA: result for test_suite '{}', test_case '{}': {}".format(test_suite, test_case, results[0]['result']))
if results[0]['result'] != 'pass':
fatal_err("FAIL")
return True
def wait_until_job_is_started(proxy, job_id):
print_log(f"Waiting for job {job_id} to start.")
current_state = "Submitted"
waiting_states = ["Submitted", "Scheduling", "Scheduled"]
while current_state in waiting_states:
job_state = _call_proxy(proxy.scheduler.job_state, job_id)
current_state = job_state["job_state"]
time.sleep(WAIT_FOR_DEVICE_POLLING_TIME_SEC)
print_log(f"Job {job_id} started.")
def follow_job_execution(proxy, job_id):
line_count = 0
finished = False
last_time_logs = datetime.now()
while not finished:
(finished, data) = _call_proxy(proxy.scheduler.jobs.logs, job_id, line_count)
logs = yaml.load(str(data), Loader=loader(False))
if logs:
# Reset the timeout
last_time_logs = datetime.now()
for line in logs:
print("{} {}".format(line["dt"], line["msg"]))
line_count += len(logs)
else:
time_limit = timedelta(minutes=DEVICE_HANGING_TIMEOUT_MIN)
if datetime.now() - last_time_logs > time_limit:
print_log("LAVA job {} doesn't advance (machine got hung?). Retry.".format(job_id))
return False
# `proxy.scheduler.jobs.logs` does not block, even when there is no
# new log to be fetched. To avoid dosing the LAVA dispatcher
# machine, let's add a sleep to save them some stamina.
time.sleep(LOG_POLLING_TIME_SEC)
return True
def show_job_data(proxy, job_id):
show = _call_proxy(proxy.scheduler.jobs.show, job_id)
for field, value in show.items():
print("{}\t: {}".format(field, value))
def validate_job(proxy, job_file):
try:
return _call_proxy(proxy.scheduler.jobs.validate, job_file, True)
except:
return False
def submit_job(proxy, job_file):
return _call_proxy(proxy.scheduler.jobs.submit, job_file)
def main(args):
proxy = setup_lava_proxy()
yaml_file = generate_lava_yaml(args)
if args.dump_yaml:
print(hide_sensitive_data(generate_lava_yaml(args)))
if args.validate_only:
ret = validate_job(proxy, yaml_file)
if not ret:
fatal_err("Error in LAVA job definition")
print("LAVA job definition validated successfully")
return
retry_count = NUMBER_OF_RETRIES_TIMEOUT_DETECTION
while retry_count >= 0:
job_id = submit_job(proxy, yaml_file)
print_log("LAVA job id: {}".format(job_id))
wait_until_job_is_started(proxy, job_id)
if not follow_job_execution(proxy, job_id):
print_log(f"Job {job_id} has timed out. Cancelling it.")
# Cancel the job as it is considered unreachable by Mesa CI.
proxy.scheduler.jobs.cancel(job_id)
retry_count -= 1
continue
show_job_data(proxy, job_id)
if get_job_results(proxy, job_id, "0_mesa", "mesa") == True:
break
def create_parser():
parser = argparse.ArgumentParser("LAVA job submitter")
parser.add_argument("--pipeline-info")
parser.add_argument("--base-system-url-prefix")
parser.add_argument("--mesa-build-url")
parser.add_argument("--job-rootfs-overlay-url")
parser.add_argument("--job-artifacts-base")
parser.add_argument("--job-timeout", type=int)
parser.add_argument("--first-stage-init")
parser.add_argument("--ci-project-dir")
parser.add_argument("--device-type")
parser.add_argument("--dtb", nargs='?', default="")
parser.add_argument("--kernel-image-name")
parser.add_argument("--kernel-image-type", nargs='?', default="")
parser.add_argument("--boot-method")
parser.add_argument("--lava-tags", nargs='?', default="")
parser.add_argument("--jwt-file", type=pathlib.Path)
parser.add_argument("--validate-only", action='store_true')
parser.add_argument("--dump-yaml", action='store_true')
parser.add_argument("--visibility-group")
return parser
if __name__ == "__main__":
# given that we proxy from DUT -> LAVA dispatcher -> LAVA primary -> us ->
# GitLab runner -> GitLab primary -> user, safe to say we don't need any
# more buffering
sys.stdout.reconfigure(line_buffering=True)
sys.stderr.reconfigure(line_buffering=True)
parser = create_parser()
parser.set_defaults(func=main)
args = parser.parse_args()
args.func(args)

View File

@@ -1,81 +0,0 @@
#!/bin/bash
set -e
set -o xtrace
CROSS_FILE=/cross_file-"$CROSS".txt
# We need to control the version of llvm-config we're using, so we'll
# tweak the cross file or generate a native file to do so.
if test -n "$LLVM_VERSION"; then
LLVM_CONFIG="llvm-config-${LLVM_VERSION}"
echo -e "[binaries]\nllvm-config = '`which $LLVM_CONFIG`'" > native.file
if [ -n "$CROSS" ]; then
sed -i -e '/\[binaries\]/a\' -e "llvm-config = '`which $LLVM_CONFIG`'" $CROSS_FILE
fi
$LLVM_CONFIG --version
else
rm -f native.file
touch native.file
fi
# cross-xfail-$CROSS, if it exists, contains a list of tests that are expected
# to fail for the $CROSS configuration, one per line. you can then mark those
# tests in their meson.build with:
#
# test(...,
# should_fail: meson.get_cross_property('xfail', '').contains(t),
# )
#
# where t is the name of the test, and the '' is the string to search when
# not cross-compiling (which is empty, because for amd64 everything is
# expected to pass).
if [ -n "$CROSS" ]; then
CROSS_XFAIL=.gitlab-ci/cross-xfail-"$CROSS"
if [ -s "$CROSS_XFAIL" ]; then
sed -i \
-e '/\[properties\]/a\' \
-e "xfail = '$(tr '\n' , < $CROSS_XFAIL)'" \
"$CROSS_FILE"
fi
fi
# Only use GNU time if available, not any shell built-in command
case $CI_JOB_NAME in
# strace and wine don't seem to mix well
# ASAN leak detection is incompatible with strace
debian-mingw32-x86_64|*-asan*)
if test -f /usr/bin/time; then
MESON_TEST_ARGS+=--wrapper=$PWD/.gitlab-ci/meson/time.sh
fi
;;
*)
if test -f /usr/bin/time -a -f /usr/bin/strace; then
MESON_TEST_ARGS+=--wrapper=$PWD/.gitlab-ci/meson/time-strace.sh
fi
;;
esac
rm -rf _build
meson _build --native-file=native.file \
--wrap-mode=nofallback \
${CROSS+--cross "$CROSS_FILE"} \
-D prefix=`pwd`/install \
-D libdir=lib \
-D buildtype=${BUILDTYPE:-debug} \
-D build-tests=true \
-D c_args="$(echo -n $C_ARGS)" \
-D cpp_args="$(echo -n $CPP_ARGS)" \
-D libunwind=${UNWIND} \
${DRI_LOADERS} \
${GALLIUM_ST} \
-D gallium-drivers=${GALLIUM_DRIVERS:-[]} \
-D vulkan-drivers=${VULKAN_DRIVERS:-[]} \
-D werror=true \
${EXTRA_OPTION}
cd _build
meson configure
ninja
LC_ALL=C.UTF-8 meson test --num-processes ${FDO_CI_CONCURRENT:-4} ${MESON_TEST_ARGS}
ninja install
cd ..

View File

@@ -1,27 +0,0 @@
#!/bin/sh
STRACEDIR=meson-logs/strace/$(for i in $@; do basename -z -- $i; echo -n _; done)
mkdir -p $STRACEDIR
# If the test times out, meson sends SIGTERM to this process.
# Simply exec'ing "time" would result in no output from that in this case.
# Instead, we need to run "time" in the background, catch the signals and
# propagate them to the actual test process.
/usr/bin/time -v strace -ff -tt -T -o $STRACEDIR/log "$@" &
TIMEPID=$!
STRACEPID=$(ps --ppid $TIMEPID -o pid=)
TESTPID=$(ps --ppid $STRACEPID -o pid=)
if test "x$TESTPID" != x; then
trap 'kill -TERM $TESTPID; wait $TIMEPID; exit $?' TERM
fi
wait $TIMEPID
EXITCODE=$?
# Only keep strace logs if the test timed out
rm -rf $STRACEDIR &
exit $EXITCODE

View File

@@ -1,17 +0,0 @@
#!/bin/sh
# If the test times out, meson sends SIGTERM to this process.
# Simply exec'ing "time" would result in no output from that in this case.
# Instead, we need to run "time" in the background, catch the signals and
# propagate them to the actual test process.
/usr/bin/time -v "$@" &
TIMEPID=$!
TESTPID=$(ps --ppid $TIMEPID -o pid=)
if test "x$TESTPID" != x; then
trap 'kill -TERM $TESTPID; wait $TIMEPID; exit $?' TERM
fi
wait $TIMEPID
exit $?

View File

@@ -1,36 +0,0 @@
diff --git a/generated_tests/CMakeLists.txt b/generated_tests/CMakeLists.txt
index 738526546..6f89048cd 100644
--- a/generated_tests/CMakeLists.txt
+++ b/generated_tests/CMakeLists.txt
@@ -206,11 +206,6 @@ piglit_make_generated_tests(
templates/gen_variable_index_write_tests/vs.shader_test.mako
templates/gen_variable_index_write_tests/fs.shader_test.mako
templates/gen_variable_index_write_tests/helpers.mako)
-piglit_make_generated_tests(
- vs_in_fp64.list
- gen_vs_in_fp64.py
- templates/gen_vs_in_fp64/columns.shader_test.mako
- templates/gen_vs_in_fp64/regular.shader_test.mako)
piglit_make_generated_tests(
shader_framebuffer_fetch_tests.list
gen_shader_framebuffer_fetch_tests.py)
@@ -279,7 +274,6 @@ add_custom_target(gen-gl-tests
gen_extensions_defined.list
vp-tex.list
variable_index_write_tests.list
- vs_in_fp64.list
gpu_shader4_tests.list
)
diff --git a/tests/sanity.py b/tests/sanity.py
index 12f1614c9..9019087e2 100644
--- a/tests/sanity.py
+++ b/tests/sanity.py
@@ -100,7 +100,6 @@ shader_tests = (
'spec/arb_tessellation_shader/execution/barrier-patch.shader_test',
'spec/arb_tessellation_shader/execution/built-in-functions/tcs-any-bvec4-using-if.shader_test',
'spec/arb_tessellation_shader/execution/sanity.shader_test',
- 'spec/arb_vertex_attrib_64bit/execution/vs_in/vs-input-uint_uvec4-double_dmat3x4_array2-position.shader_test',
'spec/glsl-1.50/execution/geometry-basic.shader_test',
'spec/oes_viewport_array/viewport-gs-write-simple.shader_test',
)

View File

@@ -1,108 +0,0 @@
#!/bin/sh
set -ex
if [ -z "$GPU_VERSION" ]; then
echo 'GPU_VERSION must be set to something like "llvmpipe" or "freedreno-a630" (the name used in your ci/gpu-version-*.txt)'
exit 1
fi
INSTALL=`pwd`/install
# Set up the driver environment.
export LD_LIBRARY_PATH=`pwd`/install/lib/
export EGL_PLATFORM=surfaceless
export VK_ICD_FILENAMES=`pwd`/install/share/vulkan/icd.d/"$VK_DRIVER"_icd.${VK_CPU:-`uname -m`}.json
RESULTS=`pwd`/${PIGLIT_RESULTS_DIR:-results}
mkdir -p $RESULTS
if [ "$GALLIUM_DRIVER" = "virpipe" ]; then
# deqp is to use virpipe, and virgl_test_server llvmpipe
export GALLIUM_DRIVER="$GALLIUM_DRIVER"
VTEST_ARGS="--use-egl-surfaceless"
if [ "$VIRGL_HOST_API" = "GLES" ]; then
VTEST_ARGS="$VTEST_ARGS --use-gles"
fi
GALLIUM_DRIVER=llvmpipe \
GALLIVM_PERF="nopt" \
virgl_test_server $VTEST_ARGS >$RESULTS/vtest-log.txt 2>&1 &
sleep 1
fi
if [ -n "$PIGLIT_FRACTION" -o -n "$CI_NODE_INDEX" ]; then
FRACTION=`expr ${PIGLIT_FRACTION:-1} \* ${CI_NODE_TOTAL:-1}`
PIGLIT_RUNNER_OPTIONS="$PIGLIT_RUNNER_OPTIONS --fraction $FRACTION"
fi
# If the job is parallel at the gitab job level, take the corresponding fraction
# of the caselist.
if [ -n "$CI_NODE_INDEX" ]; then
PIGLIT_RUNNER_OPTIONS="$PIGLIT_RUNNER_OPTIONS --fraction-start ${CI_NODE_INDEX}"
fi
if [ -e "$INSTALL/$GPU_VERSION-fails.txt" ]; then
PIGLIT_RUNNER_OPTIONS="$PIGLIT_RUNNER_OPTIONS --baseline $INSTALL/$GPU_VERSION-fails.txt"
fi
# Default to an empty known flakes file if it doesn't exist.
touch $INSTALL/$GPU_VERSION-flakes.txt
if [ -n "$VK_DRIVER" ] && [ -e "$INSTALL/$VK_DRIVER-skips.txt" ]; then
PIGLIT_SKIPS="$PIGLIT_SKIPS $INSTALL/$VK_DRIVER-skips.txt"
fi
if [ -n "$GALLIUM_DRIVER" ] && [ -e "$INSTALL/$GALLIUM_DRIVER-skips.txt" ]; then
PIGLIT_SKIPS="$PIGLIT_SKIPS $INSTALL/$GALLIUM_DRIVER-skips.txt"
fi
if [ -n "$DRIVER_NAME" ] && [ -e "$INSTALL/$DRIVER_NAME-skips.txt" ]; then
PIGLIT_SKIPS="$PIGLIT_SKIPS $INSTALL/$DRIVER_NAME-skips.txt"
fi
if [ -e "$INSTALL/$GPU_VERSION-skips.txt" ]; then
PIGLIT_SKIPS="$PIGLIT_SKIPS $INSTALL/$GPU_VERSION-skips.txt"
fi
set +e
piglit-runner \
run \
--piglit-folder /piglit \
--output $RESULTS \
--jobs ${FDO_CI_CONCURRENT:-4} \
--skips $INSTALL/all-skips.txt $PIGLIT_SKIPS \
--flakes $INSTALL/$GPU_VERSION-flakes.txt \
--profile $PIGLIT_PROFILES \
--process-isolation \
$PIGLIT_RUNNER_OPTIONS \
-v -v
PIGLIT_EXITCODE=$?
deqp-runner junit \
--testsuite $PIGLIT_PROFILES \
--results $RESULTS/failures.csv \
--output $RESULTS/junit.xml \
--limit 50 \
--template "See https://$CI_PROJECT_ROOT_NAMESPACE.pages.freedesktop.org/-/$CI_PROJECT_NAME/-/jobs/$CI_JOB_ID/artifacts/results/{{testcase}}.xml"
# Report the flakes to the IRC channel for monitoring (if configured):
if [ -n "$FLAKES_CHANNEL" ]; then
python3 $INSTALL/report-flakes.py \
--host irc.oftc.net \
--port 6667 \
--results $RESULTS/results.csv \
--known-flakes $INSTALL/$GPU_VERSION-flakes.txt \
--channel "$FLAKES_CHANNEL" \
--runner "$CI_RUNNER_DESCRIPTION" \
--job "$CI_JOB_ID" \
--url "$CI_JOB_URL" \
--branch "${CI_MERGE_REQUEST_SOURCE_BRANCH_NAME:-$CI_COMMIT_BRANCH}" \
--branch-title "${CI_MERGE_REQUEST_TITLE:-$CI_COMMIT_TITLE}"
fi
exit $PIGLIT_EXITCODE

View File

@@ -1,234 +0,0 @@
#!/bin/sh
set -ex
INSTALL=$(realpath -s "$PWD"/install)
MINIO_ARGS="--credentials=/tmp/.minio_credentials"
RESULTS=$(realpath -s "$PWD"/results)
mkdir -p "$RESULTS"
# Set up the driver environment.
# Modifiying here directly LD_LIBRARY_PATH may cause problems when
# using a command wrapper. Hence, we will just set it when running the
# command.
export __LD_LIBRARY_PATH="$LD_LIBRARY_PATH:$INSTALL/lib/"
# Sanity check to ensure that our environment is sufficient to make our tests
# run against the Mesa built by CI, rather than any installed distro version.
MESA_VERSION=$(head -1 "$INSTALL/VERSION" | sed 's/\./\\./g')
print_red() {
RED='\033[0;31m'
NC='\033[0m' # No Color
printf "${RED}"
"$@"
printf "${NC}"
}
# wrapper to supress +x to avoid spamming the log
quiet() {
set +x
"$@"
set -x
}
if [ "$VK_DRIVER" ]; then
### VULKAN ###
# Set the Vulkan driver to use.
export VK_ICD_FILENAMES="$INSTALL/share/vulkan/icd.d/${VK_DRIVER}_icd.x86_64.json"
# Set environment for Wine.
export WINEDEBUG="-all"
export WINEPREFIX="/dxvk-wine64"
export WINEESYNC=1
# Set environment for DXVK.
export DXVK_LOG_LEVEL="none"
export DXVK_STATE_CACHE=0
# Set environment for gfxreconstruct executables.
export PATH="/gfxreconstruct/build/bin:$PATH"
SANITY_MESA_VERSION_CMD="vulkaninfo"
HANG_DETECTION_CMD="/parallel-deqp-runner/build/bin/hang-detection"
# Set up the Window System Interface (WSI)
if [ ${TEST_START_XORG:-0} -eq 1 ]; then
"$INSTALL"/common/start-x.sh "$INSTALL"
export DISPLAY=:0
else
# Run vulkan against the host's running X server (xvfb doesn't
# have DRI3 support).
# Set the DISPLAY env variable in each gitlab-runner's
# configuration file:
# https://docs.gitlab.com/runner/configuration/advanced-configuration.html#the-runners-section
quiet printf "%s%s\n" "Running against the hosts' X server. " \
"DISPLAY is \"$DISPLAY\"."
fi
else
### GL/ES ###
# Set environment for apitrace executable.
export PATH="/apitrace/build:$PATH"
# Our rootfs may not have "less", which apitrace uses during
# apitrace dump
export PAGER=cat
SANITY_MESA_VERSION_CMD="wflinfo"
HANG_DETECTION_CMD=""
# Set up the platform windowing system.
if [ "x$EGL_PLATFORM" = "xsurfaceless" ]; then
# Use the surfaceless EGL platform.
export DISPLAY=
export WAFFLE_PLATFORM="surfaceless_egl"
SANITY_MESA_VERSION_CMD="$SANITY_MESA_VERSION_CMD --platform surfaceless_egl --api gles2"
if [ "x$GALLIUM_DRIVER" = "xvirpipe" ]; then
# piglit is to use virpipe, and virgl_test_server llvmpipe
export GALLIUM_DRIVER="$GALLIUM_DRIVER"
LD_LIBRARY_PATH="$__LD_LIBRARY_PATH" \
GALLIUM_DRIVER=llvmpipe \
VTEST_USE_EGL_SURFACELESS=1 \
VTEST_USE_GLES=1 \
virgl_test_server >"$RESULTS"/vtest-log.txt 2>&1 &
sleep 1
fi
elif [ "x$PIGLIT_PLATFORM" = "xgbm" ]; then
SANITY_MESA_VERSION_CMD="$SANITY_MESA_VERSION_CMD --platform gbm --api gl"
elif [ "x$PIGLIT_PLATFORM" = "xmixed_glx_egl" ]; then
# It is assumed that you have already brought up your X server before
# calling this script.
SANITY_MESA_VERSION_CMD="$SANITY_MESA_VERSION_CMD --platform glx --api gl"
else
SANITY_MESA_VERSION_CMD="$SANITY_MESA_VERSION_CMD --platform glx --api gl --profile core"
RUN_CMD_WRAPPER="xvfb-run --server-args=\"-noreset\" sh -c"
fi
fi
if [ "$ZINK_USE_LAVAPIPE" ]; then
export VK_ICD_FILENAMES="$INSTALL/share/vulkan/icd.d/lvp_icd.x86_64.json"
fi
# If the job is parallel at the gitlab job level, will take the corresponding
# fraction of the caselist.
if [ -n "$CI_NODE_INDEX" ]; then
USE_CASELIST=1
fi
replay_minio_upload_images() {
find "$RESULTS/$__PREFIX" -type f -name "*.png" -printf "%P\n" \
| while read -r line; do
__TRACE="${line%-*-*}"
if grep -q "^$__PREFIX/$__TRACE: pass$" ".gitlab-ci/piglit/$PIGLIT_RESULTS.txt.orig"; then
if [ "x$CI_PROJECT_PATH" != "x$FDO_UPSTREAM_REPO" ]; then
continue
fi
__MINIO_PATH="$PIGLIT_REPLAY_REFERENCE_IMAGES_BASE"
__DESTINATION_FILE_PATH="${line##*-}"
if wget -q --method=HEAD "https://${__MINIO_PATH}/${__DESTINATION_FILE_PATH}" 2>/dev/null; then
continue
fi
else
__MINIO_PATH="$JOB_ARTIFACTS_BASE"
__DESTINATION_FILE_PATH="$__MINIO_TRACES_PREFIX/${line##*-}"
fi
ci-fairy minio cp $MINIO_ARGS "$RESULTS/$__PREFIX/$line" \
"minio://${__MINIO_PATH}/${__DESTINATION_FILE_PATH}"
done
}
SANITY_MESA_VERSION_CMD="$SANITY_MESA_VERSION_CMD | tee /tmp/version.txt | grep \"Mesa $MESA_VERSION\(\s\|$\)\""
if [ -d results ]; then
cd results && rm -rf ..?* .[!.]* *
fi
cd /piglit
if [ -n "$USE_CASELIST" ]; then
PIGLIT_TESTS=$(printf "%s" "$PIGLIT_TESTS")
PIGLIT_GENTESTS="./piglit print-cmd $PIGLIT_TESTS replay --format \"{name}\" > /tmp/case-list.txt"
RUN_GENTESTS="export LD_LIBRARY_PATH=$__LD_LIBRARY_PATH; $PIGLIT_GENTESTS"
eval $RUN_GENTESTS
sed -ni $CI_NODE_INDEX~$CI_NODE_TOTAL"p" /tmp/case-list.txt
PIGLIT_TESTS="--test-list /tmp/case-list.txt"
fi
PIGLIT_OPTIONS=$(printf "%s" "$PIGLIT_OPTIONS")
PIGLIT_TESTS=$(printf "%s" "$PIGLIT_TESTS")
PIGLIT_CMD="./piglit run --timeout 300 -j${FDO_CI_CONCURRENT:-4} $PIGLIT_OPTIONS $PIGLIT_TESTS replay "$(/usr/bin/printf "%q" "$RESULTS")
RUN_CMD="export LD_LIBRARY_PATH=$__LD_LIBRARY_PATH; $SANITY_MESA_VERSION_CMD && $HANG_DETECTION_CMD $PIGLIT_CMD"
if [ "$RUN_CMD_WRAPPER" ]; then
RUN_CMD="set +e; $RUN_CMD_WRAPPER "$(/usr/bin/printf "%q" "$RUN_CMD")"; set -e"
fi
ci-fairy minio login $MINIO_ARGS --token-file "${CI_JOB_JWT_FILE}"
eval $RUN_CMD
if [ $? -ne 0 ]; then
printf "%s\n" "Found $(cat /tmp/version.txt), expected $MESA_VERSION"
fi
ARTIFACTS_BASE_URL="https://${CI_PROJECT_ROOT_NAMESPACE}.${CI_PAGES_DOMAIN}/-/${CI_PROJECT_NAME}/-/jobs/${CI_JOB_ID}/artifacts"
./piglit summary aggregate "$RESULTS" -o junit.xml
PIGLIT_RESULTS="${PIGLIT_RESULTS:-replay}"
RESULTSFILE="$RESULTS/$PIGLIT_RESULTS.txt"
mkdir -p .gitlab-ci/piglit
./piglit summary console "$RESULTS"/results.json.bz2 \
| tee ".gitlab-ci/piglit/$PIGLIT_RESULTS.txt.orig" \
| head -n -1 | grep -v ": pass" \
| sed '/^summary:/Q' \
> $RESULTSFILE
__PREFIX="trace/$PIGLIT_REPLAY_DEVICE_NAME"
__MINIO_PATH="$PIGLIT_REPLAY_ARTIFACTS_BASE_URL"
__MINIO_TRACES_PREFIX="traces"
if [ "x$PIGLIT_REPLAY_SUBCOMMAND" != "xprofile" ]; then
quiet replay_minio_upload_images
fi
if [ ! -s $RESULTSFILE ]; then
exit 0
fi
./piglit summary html --exclude-details=pass \
"$RESULTS"/summary "$RESULTS"/results.json.bz2
find "$RESULTS"/summary -type f -name "*.html" -print0 \
| xargs -0 sed -i 's%<img src="file://'"${RESULTS}"'.*-\([0-9a-f]*\)\.png%<img src="https://'"${JOB_ARTIFACTS_BASE}"'/traces/\1.png%g'
find "$RESULTS"/summary -type f -name "*.html" -print0 \
| xargs -0 sed -i 's%<img src="file://%<img src="https://'"${PIGLIT_REPLAY_REFERENCE_IMAGES_BASE}"'/%g'
quiet print_red echo "Failures in traces:"
cat $RESULTSFILE
quiet print_red echo "Review the image changes and get the new checksums at: ${ARTIFACTS_BASE_URL}/results/summary/problems.html"
exit 1

View File

@@ -1,75 +0,0 @@
#!/bin/bash
set -e
set -o xtrace
VERSION=`head -1 install/VERSION`
ROOTDIR=`pwd`
if [ -d results ]; then
cd results && rm -rf ..?* .[!.]* *
fi
cd /piglit
export OCL_ICD_VENDORS=$ROOTDIR/install/etc/OpenCL/vendors/
set +e
unset DISPLAY
export LD_LIBRARY_PATH=$ROOTDIR/install/lib
clinfo
# If the job is parallel at the gitlab job level, will take the corresponding
# fraction of the caselist.
if [ -n "$CI_NODE_INDEX" ]; then
if [ "$PIGLIT_PROFILES" != "${PIGLIT_PROFILES% *}" ]; then
echo "Can't parallelize piglit with multiple profiles"
exit 1
fi
USE_CASELIST=1
fi
if [ -n "$USE_CASELIST" ]; then
./piglit print-cmd $PIGLIT_TESTS $PIGLIT_PROFILES --format "{name}" > /tmp/case-list.txt
sed -ni $CI_NODE_INDEX~$CI_NODE_TOTAL"p" /tmp/case-list.txt
PIGLIT_TESTS="--test-list /tmp/case-list.txt"
fi
./piglit run -c -j${FDO_CI_CONCURRENT:-4} $PIGLIT_OPTIONS $PIGLIT_TESTS $PIGLIT_PROFILES $ROOTDIR/results
retVal=$?
if [ $retVal -ne 0 ]; then
echo "Found $(cat /tmp/version.txt), expected $VERSION"
fi
set -e
PIGLIT_RESULTS=${PIGLIT_RESULTS:-$PIGLIT_PROFILES}
mkdir -p .gitlab-ci/piglit
./piglit summary console $ROOTDIR/results \
| tee ".gitlab-ci/piglit/$PIGLIT_RESULTS.txt.orig" \
| head -n -1 \
| grep -v ": pass" \
| sed '/^summary:/Q' \
> .gitlab-ci/piglit/$PIGLIT_RESULTS.txt
if [ -n "$USE_CASELIST" ]; then
# Just filter the expected results based on the tests that were actually
# executed, and switch to the version with no summary
cat .gitlab-ci/piglit/$PIGLIT_RESULTS.txt.orig | sed '/^summary:/Q' | rev \
| cut -f2- -d: | rev | sed "s/$/:/g" > /tmp/executed.txt
grep -F -f /tmp/executed.txt $ROOTDIR/install/$PIGLIT_RESULTS.txt \
> .gitlab-ci/piglit/$PIGLIT_RESULTS.txt.baseline || true
else
cp $ROOTDIR/install/$PIGLIT_RESULTS.txt .gitlab-ci/piglit/$PIGLIT_RESULTS.txt.baseline
fi
if diff -q .gitlab-ci/piglit/$PIGLIT_RESULTS.txt{.baseline,}; then
exit 0
fi
./piglit summary html --exclude-details=pass $ROOTDIR/results/summary $ROOTDIR/results
echo Unexpected change in results:
diff -u .gitlab-ci/piglit/$PIGLIT_RESULTS.txt{.baseline,}
exit 1

View File

@@ -1,57 +0,0 @@
#!/bin/bash
set -e
set -o xtrace
CROSS_FILE=/cross_file-"$CROSS".txt
# Delete unused bin and includes from artifacts to save space.
rm -rf install/bin install/include
# Strip the drivers in the artifacts to cut 80% of the artifacts size.
if [ -n "$CROSS" ]; then
STRIP=`sed -n -E "s/strip\s*=\s*'(.*)'/\1/p" "$CROSS_FILE"`
if [ -z "$STRIP" ]; then
echo "Failed to find strip command in cross file"
exit 1
fi
else
STRIP="strip"
fi
if [ -z "$ARTIFACTS_DEBUG_SYMBOLS" ]; then
find install -name \*.so -exec $STRIP {} \;
fi
# Test runs don't pull down the git tree, so put the dEQP helper
# script and associated bits there.
echo "$(cat VERSION) (git-$(git rev-parse HEAD | cut -b -10))" > install/VERSION
cp -Rp .gitlab-ci/bare-metal install/
cp -Rp .gitlab-ci/common install/
cp -Rp .gitlab-ci/piglit install/
cp -Rp .gitlab-ci/fossils.yml install/
cp -Rp .gitlab-ci/fossils install/
cp -Rp .gitlab-ci/fossilize-runner.sh install/
cp -Rp .gitlab-ci/crosvm-init.sh install/
cp -Rp .gitlab-ci/*.txt install/
cp -Rp .gitlab-ci/report-flakes.py install/
cp -Rp .gitlab-ci/vkd3d-proton install/
cp -Rp .gitlab-ci/*-runner.sh install/
find . -path \*/ci/\*.txt \
-o -path \*/ci/\*.toml \
-o -path \*/ci/\*traces\*.yml \
| xargs -I '{}' cp -p '{}' install/
# Tar up the install dir so that symlinks and hardlinks aren't each
# packed separately in the zip file.
mkdir -p artifacts/
tar -cf artifacts/install.tar install
cp -Rp .gitlab-ci/common artifacts/ci-common
cp -Rp .gitlab-ci/lava artifacts/
if [ -n "$MINIO_ARTIFACT_NAME" ]; then
# Pass needed files to the test stage
MINIO_ARTIFACT_NAME="$MINIO_ARTIFACT_NAME.tar.gz"
gzip -c artifacts/install.tar > ${MINIO_ARTIFACT_NAME}
ci-fairy minio login --token-file "${CI_JOB_JWT_FILE}"
ci-fairy minio cp ${MINIO_ARTIFACT_NAME} minio://${PIPELINE_ARTIFACTS_BASE}/${MINIO_ARTIFACT_NAME}
fi

Some files were not shown because too many files have changed in this diff Show More