Compare commits

..

400 Commits

Author SHA1 Message Date
Ian Romanick
1659d87afe docs: Add 8.0.3 release notes
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
2012-05-18 16:11:34 -07:00
Ian Romanick
60dffb92de mesa: Bump version number to 8.0.3
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
2012-05-18 16:08:35 -07:00
Eric Anholt
b5b65af5fa mesa: Check for framebuffer completeness before looking at the rb.
Otherwise, an incomplete framebuffer could have a NULL
_ColorReadBuffer and we'd deref that.

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 5a827d9a2b)
2012-05-18 16:08:34 -07:00
Eric Anholt
e0e8f1efad glsl: Fix assertion failure on handling switch on uint expressions.
Fixes piglit glsl-1.30/execution/switch/fs-uint.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 9c4e9ce051)
2012-05-18 16:08:34 -07:00
Eric Anholt
91874c3252 glsl: Reject non-scalar switch expressions.
The comment quotes spec saying that only scalar integers are allowed,
but we only checked for integer.

Fixes piglit switch-expression-const-ivec2.vert

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit bbbc7c7d56)
2012-05-18 16:08:34 -07:00
Eric Anholt
2d105870c9 glsl: Let the constructor figure out the types of switch-related expressions.
I noticed this while unindenting the code.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 5d6ea16dfe)
2012-05-18 16:08:34 -07:00
Eric Anholt
912acd045c glsl: Fix indentation of switch code.
I managed to completely trash it in 22d81f15.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 5462f3679a)
2012-05-18 16:08:34 -07:00
Eric Anholt
1f90817e3c i965/vs: Fix up swizzle for dereference_array of matrices.
Fixes assertion failure in piglit:
vs-mat2-struct-assignment.shader_test
vs-mat2-array-assignment.shader_test

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit aa02884c4f)
2012-05-18 16:08:34 -07:00
Eric Anholt
1974502223 mesa: Throw error on glGetActiveUniform inside Begin/End.
Fixes piglit GL_ARB_shader_objeccts/getactiveuniform-beginend.

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit ef691885c9)
2012-05-18 16:08:34 -07:00
Brian Paul
0bf439e5da mesa: fix glMaterial / dlist bug
When glColorMaterial() is used to latch glColor commands to a material
attribute, glMaterial calls to change that material should become no-ops.
This failed to work properly when the glMaterial call was inside a
display list.

This removes the Material function from the vbo_attrib_tmp.h template
file.  We have separate/different implementations for the "save" and
"exec" cases now.

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit 1bc16bf98a)
2012-05-18 16:08:34 -07:00
Brian Paul
d28e968edb mesa: fix/add error check in _mesa_ColorMaterial()
_mesa_material_bitmask() will record a GL error and return 0 if
face or mode are illegal.  Return early in that case.

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit f272490546)
2012-05-18 16:08:34 -07:00
Kenneth Graunke
9f67d73830 glsl: Fix broken constant expression handling for <, <=, >, and >=.
We were looping over all the vector components, but only dealing with
the first one.  This was masked by the fact that constant expression
handling on built-ins went through custom code for the lessThan()
/function/ rather than the ir_binop_less expression operator.

NOTE: This is a candidate for all release branches.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Signed-off-by: Olivier Galibert <galibert@pobox.com>
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit f72e9b2041)
2012-05-18 16:08:34 -07:00
Chad Versace
84f537e72b intel: Disable ARB_framebuffer_object in ES contexts
This patch removes ARB_framebuffer_object from the GLES1 and GLES2
extension lists in intel_extensions_es.c.

Fixes a crash in the Android browser on Ice Cream Sandwich.

The Android browser crashed because it did the following, which is legal
in GLES2 but not in ARB_framebuffer_object.
    glGenFramebuffers(1, &fb);
    glBindFramebuffer(GL_FRAMEBUFFER, fb);
    // render render render...
    glDeleteFramebuffers(1, &fb);
    // go do other stuff...
    glBindFramebuffer(GL_FRAMEBUFFER, fb);
    // This bind unexpectedly failed, and the app panics.

The semantics of glBindFramebuffer specified by ARB_framebuffer_object (a
desktop GL extension) and GLES2 specs are incompatible. The ideal solution
to fix this is to create separate API entry points for glBindFramebuffer,
one for GL and the other for GLES2. But, until that work is complete,
disabling ARB_framebuffer_object in GLES2 contexts safely fixes the problem.

Likewise, the semantics of glBindFramebuffer in ARB_framebuffer_object and
of glBindFramebufferOES in OES_framebuffer_object (a GLES1 extension) are
incompatible. Even though the functions have different names, the semantic
difference still results in a bug because both API calls are implemented
by a single function, _mesa_BindFramebufferEXT, which handles the semantic
difference incorrectly. Again, disabling ARB_framebuffer_object in GLES1
contexts safely fixes this problem.

According to the ARB_framebuffer_object spec, the extension is an
amalgamation of
    EXT_framebuffer_object
    EXT_framebuffer_blit
    EXT_packed_depth_stencil
    EXT_framebuffer_multisample
By disabling this extension, however, no functionality is removed from
GLES1 and GLES2 contexts because 1) the first three extensions are
explicitly enabled in Intel's ES extension lists and 2) no functionality
of the last extension is exposed in an ES context.

Note: This is a candidate for the 8.0 branch.
See-also: http://www.mail-archive.com/mesa-dev@lists.freedesktop.org/msg21006.html
CC: Charles Johnson <charles.f.johnson@intel.com>
CC: Sean Kelley <sean.v.kelley@intel.com>
Reviewed-by: Ian Romanick <idr@freedesktop.org>
Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit 1c0f5d8324)
2012-05-18 16:08:34 -07:00
Jordan Justen
3eaa27789d mesa: Add primitive restart support to glArrayElement
When primitive restart is enabled, and glArrayElement is called
with the restart index value, then call glPrimitiveRestartNV.

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Brian Paul<brianp@vmware.com>
(cherry picked from commit 5795d3b5ae)
2012-05-18 16:08:34 -07:00
Vinson Lee
6e8c30304e mesa: Fix memory leak in generate_mipmap_compressed.
Fixes Coverity resource leak defect.

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit b688700edc)
2012-05-18 16:08:34 -07:00
Han Shen(沈涵)
3e5f17b5f9 bin/mklib: remove '-m32' for arm linux
-m32 is not a valid option for ARM.

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit b352d676e4)
2012-05-18 16:08:34 -07:00
Kenneth Graunke
6b9363a316 intel: Remove pointless software fallback for glBitmap on Gen6.
We already have a meta path below that works just fine; no apparent
regressions in oglconform.

NOTE: This is a candidate for the 8.0 branch.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=46834
Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit 3cd7bee48f)
2012-05-18 16:08:34 -07:00
Eric Anholt
4d9ff6af59 intel: Return success when asked to allocate a 0-width/height renderbuffer.
It seems silly that GL lets you allocate these given that they're
framebuffer attachment incomplete, but the webgl conformance tests
actually go looking to see if the getters on 0-width/height
depth/stencil renderbuffers return good values.  By failing out here,
they all got smashed to 0, which turned out to be correct for all the
getters they tested except for GL_RENDERBUFFER_INTERNAL_FORMAT.  Now,
by succeeding but not making a miptree, that one also returns the
expected value.

Acked-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit b7406404ab)
2012-05-18 16:08:34 -07:00
Brian Paul
110111bc5b mesa: do more teximage error checking for generic compressed formats
When glTexImage or glCopyTexImage is called with internalFormat being a
generic compressed format (like GL_COMPRESSED_RGB) we need to do the same
error checks as for specific compressed formats.  In particular, check if
the texture target is compatible with the format.  None of the texture
compression formats we support so far work with GL_TEXTURE_1D, for example.

See also https://bugs.freedesktop.org/show_bug.cgi?id=49124

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit a36581ccc0)
2012-05-18 16:08:34 -07:00
Brian Paul
2aaaa18853 st/mesa: no-op glCopyPixels if source region is out of bounds
If the source region for a glCopyPixels is completely outside the
source buffer bounds, no-op the copy.  Fixes a failed assertion.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
(cherry picked from commit 99ed0b2f9b)
2012-05-18 16:08:34 -07:00
Vinson Lee
7e824b728b ir_to_mesa: Fix uninitialized member in add_uniform_to_shader.
Fix uninitialized scalar field defect reported by Coverity.

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 10ec14865a)
2012-05-18 16:08:34 -07:00
Chad Versace
fc65fc5035 main: Fix memory leak in _mesa_make_extension_string()
I forgot to free the string returned by strdup().

Note: This is a candidate for the stable branches.
CC: Johannes Obermayr <johannesobermayr@gmx.de>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit 36fef005b1)
2012-05-18 16:08:33 -07:00
Vinson Lee
c9088b5251 swrast: Fix memory leaks in blit_linear.
Fixes Coverity resource leak defects.

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 431b458f24)
2012-05-18 16:08:33 -07:00
Vinson Lee
8537544be9 linker: Fix memory leak in count_uniform_size::visit_field.
Fixes a Coverity resource leak defect.

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 4123d0b321)
2012-05-18 16:08:33 -07:00
Kenneth Graunke
b4693383bf i965: Avoid explicit accumulator operands in SIMD16 mode on Gen7.
According to the BSpec ISA volume's "Accumulator Register" section:

"[DevIVB] SIMD16 execution on dwords is not allowed when accumulator is
 explicit source or destination operand."

Fixes piglit tests:
- fs-multiply-const-ivec4
- fs-multiply-const-uvec4
- fs-multiply-ivec4-const
- fs-multiply-uvec4-const

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 01044fce6b)
2012-05-18 16:08:33 -07:00
Dave Airlie
5eb2e5bdc7 intel: fix TFP at 16-bpp
don't ask why I had to debug this.

tested to fix g-s and kwin at 16-bpp on Ironlake.

Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit 7b6424143d)
2012-05-18 16:08:33 -07:00
Dave Airlie
ca8be6fe2b drisw: fix image stride calculation for 16-bit.
If you ran g-s in 16-bpp we'd do a bunch of memory corruption.

now it just misrenders for some other reasons.

applies to stable.

Signed-off-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit d6c5ad52b2)
2012-05-18 16:08:33 -07:00
Dave Airlie
aa17a8b81e glx/drisw: avoid segfaults when we fail to get visual
piglit glx-tfp segfaults on llvmpipe when run vs a 16-bit radeon screen,

it now fails instead of segfaulting, much prettier.

Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit 2c778375a1)
2012-05-18 16:08:33 -07:00
Vinson Lee
b304eca8df mesa: Fix memory leak in _mesa_get_uniform_location.
Fixes Coverity resource leak defect.

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 0f3aa9f4bd)
2012-05-18 16:08:33 -07:00
Brian Paul
97bff96352 st/mesa: set MaxUnrollIterations = 255
The default was 32 for the EmitNoLoops=0 case.  This allows the oZone3D
soft shadows test to work properly with the vmware driver.  Jose reported
that SM3 supports up to 255 loop iterations.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: José Fonseca <jfonseca@vmware.com>
(cherry picked from commit 20836c8185)
2012-05-18 16:08:33 -07:00
Brian Paul
9405567c28 glsl: propagate MaxUnrollIterations to the optimizer's loop unroller
Instead of the hard-coded value of 32.  Note that MaxUnrollIterations
defaults to 32 so there's no net change.  But the gallium state tracker
can override this.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 7feabfe23d)
2012-05-18 16:08:33 -07:00
Jeremy Huddleston
e69758260b darwin: Address a build failure on Leopard and earlier OS versions
<https://trac.macports.org/ticket/34499>

Regression-from: 51691f0767
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 27b821bc95)
2012-05-18 11:33:05 -07:00
Anuj Phogat
869c34527a intel: Fix a case when mapping large texture fails
This is a squash of:

    intel: Fix a case when mapping large texture fails

    This patch handles a case when mapping a large texture fails
    in drm_intel_gem_bo_map_gtt(). These changes avoid assertion
    failure later in the driver as reported in following bugs:

    https://bugs.freedesktop.org/show_bug.cgi?id=44970
    https://bugs.freedesktop.org/show_bug.cgi?id=46303

    Testing: No regressions in piglit quick.tests

    Signed-off-by: Anuj Phogat <anuj.phogat@gmail.com>
    (cherry picked from commit cdcfd5d1d6)

and:

    intel: fix un-blanced map_refcount issue

    This is a regression introduced by commit cdcfd5, which forget to
    increase the map_refcount for successfully-mapped region. Thus caused a
    wrong non-blanced map_refcount.

    This would fix the regression found in the two following webglc testcase
    on Pineview platform:
       texture-npot.html
       gl-max-texture-dimensions.html

    Cc: Anuj Phogat <anuj.phogat@gmail.com>
    Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
    Reviewed-by: Anuj Phogat <anuj.phogat@gmail.com>
    (cherry picked from commit 9cb777eb71)
2012-05-17 23:38:07 -07:00
Anuj Phogat
7e7f099108 mesa: Fix the cause of piglit test fbo-array failure
Handle the special case of glFramebufferTextureLayer() for which we pass
teximage = 0 internally in framebuffer_texture(). This patch makes failing
piglit test fbo-array, fbo-depth-array to pass.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=47126

V4: Removed the duplicated code.
Note: This is a candidate for the stable branches.

Signed-off-by: Anuj Phogat <anuj.phogat@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit cc5b0ffae0)
2012-05-17 22:13:49 -07:00
Kenneth Graunke
cc2413c889 drirc: Add force_glsl_extensions_warn workaround for Unigine Heaven.
Unfortunately, Unigine Heaven 3.0 still needs this.

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 60218b604a)
2012-05-17 22:13:49 -07:00
Anuj Phogat
1ebdf22224 mesa: Fix valid texture target test in _mesa_GetTexLevelParameteriv()
_mesa_max_texture_levels() is also used to test valid texture target
in _mesa_GetTexLevelParameteriv(). GL_TEXTURE_CUBE_MAP is not allowed
as texture target in glGetTexLevelParameter(). So, this should throw
GL_INVALID_ENUM error.

Few other functions which use _mesa_max_texture_levels() like
getcompressedteximage_error_check() and getteximage_error_check()
also don't accept GL_TEXTURE_CUBE_MAP.

Above fix makes piglit fbo-cubemap test to fail. This is because of
incorrect texture target passed to _mesa_max_texture_levels() in
framebuffer_texture(). Fixing that as well

Note: This is a candidate for the stable branches

Signed-off-by: Anuj Phogat <anuj.phogat@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit a9523af0e8)
2012-05-17 22:13:49 -07:00
Kurt Roeckx
a3f6e8f431 i915: Fix i830 polygon stipple from PBOs.
This is a direct port of the i915 patch in
a856da6324.

Fixes glean's pbo test.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=41372
Reviewed-by: Eric Anholt <eric@anholt.net>
NOTE: This is a candidate for release branches.
(cherry picked from commit 489ac8e73a)
2012-05-17 22:13:49 -07:00
Kurt Roeckx
5696077656 i915: Compute maximum number of verts using the actual batchbuffer size.
We were looking at the size of batch.map for how big the batchbuffer
was, but on 865 we just use a single-page batchbuffer due to hardware
limits.

v2: Removed check for sizeof map < bo->size, since that's always false.
    [change by anholt]
NOTE: This is a candidate for release branches.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=41495
(cherry picked from commit 024ece7523)
2012-05-17 22:13:49 -07:00
Chris Wilson
815d6e3f2f i830: Compute initial number of vertices from remaining batch space
In order to prevent an overflow of the batch buffer when emitting
triangles, we need to limit the initial primitive to fit within the
current batch. To do we need to measure the remaining space and thence
compute the maximum number of vertices that fit into that space.

Reported-by: Kurt Roeckx <kurt@roeckx.be>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=41495
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Eric Anholt <eric@anholt.net>
NOTE: This is a candidate for release branches.
(cherry picked from commit 33b07893e9)
2012-05-17 22:13:49 -07:00
Alban Browaeys
c9aa4607c5 dri/i915: Fix off-by-one in i830 clip region size.
The hardware, like i915, uses an inclusive bounds on min and max for
the drawing rectangle, but we were providing a number for exclusive.
The number of bits used by the hardware only covers this value going
up to the maximum size, so when we programmed 2048 as the maximum
inclusive X, it saw a maximum X of 0 and clipped all rendering.  This
caused rendering failures in gnome-shell.

Fixes piglit fbo-maxsize.

v2: dropped changes to the blitter, which does use an exclusive x2, y2.
    [change by anholt]

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=45558
Reviewed-by: Eric Anholt <eric@anholt.net>
NOTE: This is a candidate for release branches.
(cherry picked from commit 7d13a6e64b)
2012-05-17 22:13:49 -07:00
Eric Anholt
a164f23d09 i915: Fix piglit fbo-nodepth-test on i830.
This is a direct port of fc4fba52cf from
i915, and fixes GPU hangs when running piglit.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=41372
Reviewed-by: Eric Anholt <eric@anholt.net>
NOTE: This is a candidate for release branches.
(cherry picked from commit 4d4f2daefa)
2012-05-17 21:53:27 -07:00
Eric Anholt
31aaf56f28 mesa: Add missing error check for first < 0 in glDrawArraysInstanced().
Fixes piglit GL_ARB_draw_instanced/negative-arrays-first-negative.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 767ba60831)
2012-05-17 21:53:12 -07:00
Eric Anholt
9bbd435900 mesa: Fix display lists for draw_elements_base_vertex with draw_instanced.
Fixes piglit GL_ARB_draw_elements_base_vertex/dlist-arb_draw_instanced

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 871518dbf8)
2012-05-17 21:53:06 -07:00
Eric Anholt
fe64ad0cba mesa: Fix display list handling for GL_ARB_draw_instanced.
When you called them in a display list compile before, you would just
end up calling through NULL.

Fixes piglit GL_ARB_draw_instanced/dlist.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 3c69a18b6a)
2012-05-17 21:52:55 -07:00
Paul Berry
ae02489a6d i915: Initialize swrast_texture_image structure fields.
Commit 980f6f1 (mesa: move gl_texture_image::Width/Height/DepthScale
fields to swrast) moved the initialization of the Width, Height, and
DepthScale fields to _swrast_alloc_texture_image_buffer().  However,
i915 doesn't call this function because it performs its own buffer
allocation.  As a result, the Width, Height, and DepthScale fields
weren't getting initialized properly, and some operations requiring
swrast would fail.

This patch ensures that Width, Height, and DepthScale are properly
initialized by separating the code that sets them into a new function,
_swrast_init_texture_image(), which is called by
intel_alloc_texture_image_buffer() as well as
_swrast_alloc_texture_image_buffer().  It also moves the
initialization of _IsPowerOfTwo into this function.

Fixes piglit test fbo/fbo-cubemap on i915.

Partially fixes https://bugs.freedesktop.org/show_bug.cgi?id=41216

This is a candidate for the 8.0 branch.

Reviewed-and-tested-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 80513ec8b4)
2012-05-17 21:52:48 -07:00
Anuj Phogat
00a182d66e mesa: fix issues with texture border and array textures
For a 1D texture array, the border only applies to the width.  For a 2D
texture array the border applies to the width and height but not the depth.
Sucha cases were  not handled correctly in _mesa_init_teximage_fields().

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 b5c409363c)
2012-05-17 21:29:41 -07:00
Brian Paul
88cbc43d1f mesa: add missing return after _mesa_error() in update_array()
NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: José Fonseca <jfonseca@vmware.com>
(cherry picked from commit e14b357367)
2012-05-17 21:29:35 -07:00
Kenneth Graunke
d342246e2b i965: Make the dummy fragment shader work in SIMD16 mode.
If you're resorting to the dummy shader, you've probably already turned
off SIMD16 mode.  But if you didn't, it would die in a fire.

We could either fail to compile in SIMD16 mode...or just fix it.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit df5963c256)
2012-05-17 21:29:27 -07:00
Kenneth Graunke
51e1111bd6 i965: Fix GPU hangs in the dummy fragment shader.
The dummy FB write failed to specify EOT and a message length, causing
the GPU to hang.  Now we can enjoy "everyone's favorite color" again.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 393b42240f)
2012-05-17 21:29:17 -07:00
Kenneth Graunke
b85c5d9344 vbo: Eliminate short-circuiting in invalid-start case.
Now that we have a index_range_invalid flag, we can just use that rather
than calling vbo_validated_drawrangeelements directly and returning.

NOTE: This is a candidate for release branches.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 6e738d35c5)
2012-05-17 21:29:05 -07:00
Kenneth Graunke
eb2eb97baf vbo: Rework checking of 'end' against _MaxElement.
This failed to take basevertex into account:

If basevertex < 0:
   (end + basevertex) might actually be in-bounds while 'end' is not.
   We would have clamped in this case when we probably shouldn't.
   This could break application drawing.

If basevertex > 0:
   'end' might be in-bounds while (end + basevertex) might not.
   We would have failed to clamp in this place.  There's a comment
   indicating the TNL module depends on max_index being in-bounds;
   if so, it would likely break horribly.

Rather than trying to clamp correctly in the face of basevertex, simply
delete the clamping code and indicate that we don't have a valid range.
This causes _tnl_vbo_draw_prims to use vbo_get_minmax_indices() to
compute the actual bounds, which is much safer.

NOTE: This is a candidate for release branches.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
(cherry picked from commit 112b02c324)
2012-05-17 21:28:57 -07:00
Kenneth Graunke
7032a56c85 vbo: Ignore invalid element ranges which are outside VBO bounds.
Some applications, such as Regnum Online, appear to pass invalid
start/end values to glDrawRangeElements.  In particular, the 'start'
index sometimes exceeds the maximum array element.  This is clearly
invalid behavior, and although the spec isn't clear, seems to result
in undefined, implementation-specific behavior.

This patch takes the conservative approach and simply ignores the range,
while issuing a warning indicating that the application is broken and
should be fixed.

NOTE: This is a candidate for release branches.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=45214
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=44701
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=41152
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=40361
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=28138
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Brian Paul <brianp@vmware.com> [v1]
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
(cherry picked from commit f00c97b23f)
2012-05-17 21:28:48 -07:00
Kenneth Graunke
3aaf3c85e4 vbo: Remove pedantic warning about 'end' beind out of bounds.
The application supplied [start, end] range is merely a conservative
hint of the ranges of index values inside the index buffer.  There is no
requirement that all vertices in the range [start, end] be referenced.

Passing an 'end' value larger than the maximum legal index is perfectly
acceptible; applications can legally pass 0xffffffff when they don't
have a tighter bound readily available.

Thus, the warning doesn't indicate a correctness issue; it could only
indicate a performance issue.  However, it does not even do that.

glDrawRangeElements is designed to optimize non-VBO vertex data uploads
by providing an upper bound on the size of buffers a driver would need
to allocate.  With VBOs, the data is already in an uploaded buffer, so
the range doesn't help.

The clincher is: we only know _MaxElement for VBOs.  For user-space
arrays, we just set it to 2,000,000,000 (see mesa/main/varray.h:63.)
So we can only check this in the case where it is not useful.

Many applications, including the Unigine demos, currently trigger this
warning, which suggests the applications are buggy when they're actually
fine.  Eliminating the warning should confuse users less while not
actually losing any benefit to application developers.

NOTE: This is a candidate for release branches.

Suggested-by: Jose Fonseca <jfonseca@vmware.com>
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
(cherry picked from commit f9be8543aa)
2012-05-17 21:28:38 -07:00
Brian Paul
0a24f0868f mesa: add missing texture integer test in glTexSubImage()
If the texture format is integer, the incoming user data must also be
integer (and similarly for non-integer textures).

NOTE: This is a candidate for the stable branches.

Reviewed-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 6f3d918409)
2012-05-17 21:28:29 -07:00
Eric Anholt
a325146f11 mesa: Include the multisample enables under GL_MULTISAMPLE_BIT attrib as well.
Fixes (with the previous commit) piglit GL_ARB_multisample/pushpop.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit b631b471d8)
2012-05-17 21:28:20 -07:00
Eric Anholt
64c69a5b86 mesa: Fix push/pop of multisample coverage invert.
In the table of of push/pop attributes, this one doesn't fall under
the enable group.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 4b6e45c7d6)
2012-05-17 21:28:13 -07:00
Mathias Fröhlich
8a7cb5d21c glsl: Avoid excessive loop unrolling.
Avoid unrollong loops that are either nested loops or
where the loop body times the unroll count is huge.

The change is far from being perfect but it extends the
loop unrolling decision heuristic by some additional
safeguard. In particular this cuts down compilation of
a shader precomputing atmospheric scattering integral
tables containing two nesting levels in a loop from
something way beyond some minutes (I never waited for
it to finish) to some fractions of a second.

This fixes piglit tests glsl-fs-unroll-explosion and
glsl-vs-unroll-explosion on r600g.

Reviewed-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Mathias Fröhlich <Mathias.Froehlich@web.de>
(cherry picked from commit 67007080b7)
2012-05-17 21:27:59 -07:00
Eric Anholt
865ad64e06 i965/fs: Implement GL_CLAMP behavior on texture rectangles on gen6+.
We were doing saturate-based clamping on the [0,width] or [0,height]
coordinate, which meant only the first pixel was addressable.

Fixes piglit ARB_texture_rectangle/texwrap-RECT-bordercolor

NOTE: This is a candidate for the 8.0 release branch.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 7c857a6b15)
2012-05-17 21:27:48 -07:00
Eric Anholt
0b2ffc647d i965/fs: Move GL_CLAMP handling to coordinate setup.
We should be able to merge self-move instruction into the MRF move
anyway, and this simplifies things for the next commit.

NOTE: This is a candidate for the 8.0 release branch.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 07e621c523)
2012-05-17 21:27:41 -07:00
Jon TURNEY
0d1049ff26 Have __glImageSize handle format GL_DEPTH_STENCIL_NV the same way as the server does
There is a mismatch between the way the X server and GLX library
calculates the image size for format GL_DEPTH_STENCIL(|_NV|_EXT)

See https://bugs.freedesktop.org/show_bug.cgi?id=30102

Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk>
(cherry picked from commit 8937c166ef)
2012-05-17 17:23:57 -07:00
Roland Scheidegger
78b66adc1a mesa: check_index_bounds off-by-one fix
in check_index_bounds the comparison needs to be "greater equal" since
contrary to the name _MaxElement is the count of the array (this matches
similar code in vbo_exec_DrawRangeElementsBaseVertex).

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 1f4a853b1e)
2012-05-17 17:23:37 -07:00
Brian Paul
d7b1a7bfb1 vbo: fix node_attrsz[] usage in vbo_bind_vertex_list()
The node_attrsz[] array is initially copied from the node->attrsz[]
array but some values get rewritten.  Thereafter, we need to use the
node_attrsz[] values.

Fixes a bug when replaying a display list that uses generic vertex
array[16] (at least).

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit e53557996e)
2012-05-17 17:23:18 -07:00
Dave Airlie
4d5b40a5a2 mesa/format_unpack: add LUMINANCE 8/16 UINT/INT
This just copies what the LUMINANCE_ALPHA bits do.

Fixes piglit tests on softpipe complaining about missing unpack.

Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit fef395c0c3)
2012-05-17 17:23:07 -07:00
Brian Paul
1e987b6e7a mesa: add BGR888 code in _mesa_format_matches_format_and_type()
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 966720f507)
2012-05-17 17:22:41 -07:00
Brian Paul
3158636f4c mesa: fix error in _mesa_format_matches_format_and_type() for RGB888
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit b019228d6b)
2012-05-17 17:22:35 -07:00
Brian Paul
a341475398 mesa: remove LSB-first pixel packing check in glReadPixels
GL_UNPACK_LSB_FIRST only applies to bitmap data, not glReadPixels.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 343100d1fc)
2012-05-17 17:22:30 -07:00
Eric Anholt
d953370670 mesa: Fix handling of glCopyBufferSubData() for src == dst.
Fixes piglit ARB_copy_buffer-overlap, on swrast, which previously
assertion failed.

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 531e44415b)
2012-05-17 17:22:15 -07:00
Alexander von Gluck
8e8c974fff glsl: Don't use newlocale on Haiku
NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 873f3ae92e)
2012-05-17 17:21:55 -07:00
Alexander von Gluck
a0f010db1c mesa: Don't use newlocale on Haiku
NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 02a1f9f14d)
2012-05-17 17:21:44 -07:00
Alexander von Gluck
9de17f5f4d svga: fix typedef conflicts on Haiku
NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 32876a452f)
2012-05-17 17:21:29 -07:00
Alexander von Gluck
abd1431707 llvmpipe: fix symbol conflict on Haiku
NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 1550b0668e)
2012-05-17 17:21:15 -07:00
Matt Turner
c2fdcc1e12 Remove -ffast-math from default CFLAGS
Fixes glsl-const-folding-01. inversesqrt(1.0) != 1.0 was evaluating as
true.

Signed-off-by: Matt Turner <mattst88@gmail.com>
(cherry picked from commit 2fdbbeca43)
2012-05-17 17:20:32 -07:00
Alex Deucher
f1b6c69137 radeon: fix fog coordinate emit
Noticed by dungeon on phoronix:
http://phoronix.com/forums/showthread.php?65408-Radeon-R100-R200-Mesa-Driver-Sees-Attention&p=247018#post247018

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Note: this is a candidate for the stable branches.
(cherry picked from commit e77c495d09)
2012-05-17 17:19:52 -07:00
Alex Deucher
e72cbdf395 r200: fix fog coordinate emit
Noticed by dungeon on phoronix:
http://phoronix.com/forums/showthread.php?65408-Radeon-R100-R200-Mesa-Driver-Sees-Attention&p=247018#post247018

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Note: this is a candidate for the stable branches.
(cherry picked from commit afdd6f8c34)
2012-05-17 17:19:36 -07:00
Eugeni Dodonov
bf7407f631 intel: add PCI IDs for Ivy Bridge GT2 server variant
Those IDs are used by Bromolow.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Eugeni Dodonov <eugeni.dodonov@intel.com>

(cherry picked from commit a45247fb1b)

Conflicts:
	include/pci_ids/i965_pci_ids.h
2012-05-17 08:16:09 -07:00
Jeremy Huddleston
9724c8d13c 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 18:59:02 -07:00
Jeremy Huddleston
7e624edba4 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 18:59:00 -07:00
Dylan Noblesmith
03354a3c4a i965: fix typo
Noticed by clang:

brw_wm_surface_state.c:330:30: warning: initializer overrides prior
initialization of this subobject [-Winitializer-overrides]
      [MESA_FORMAT_Z24_S8] = 0,
                             ^
brw_wm_surface_state.c:326:30: note: previous initialization is here
      [MESA_FORMAT_Z24_S8] = 0,
                             ^

No functionality change, since the array is declared static so
it was zero-initialized by default.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit dd32df3829)
2012-05-16 10:58:49 -07:00
Brian Paul
0a25c4c384 swrast: include s_fragprog.h to silence warnings 2012-05-14 09:59:41 -06:00
Yuanhan Liu
fa68a8bae3 i965: fix wrong cube/3D texture layout
Fix wrong cube/3D texture layout for the tailing levels whose width or
height is smaller than the align unit.

From 965 B-spec http://intellinuxgraphics.org/VOL_1_graphics_core.pdf at
page 135:
   All of the LOD=0 q-planes are stacked vertically, then below that,
   the LOD=1 qplanes are stacked two-wide, then the LOD=2 qplanes are
   stacked four-wide below that, and so on.

Thus we should always inrease pack_x_nr, which results to the pitch of LODn
may greater than the pitch of LOD0. So we should refactor mt->total_width
when needed.

This would fix the following webgl test case on all gen4 platforms:
  conformance/textures/texture-size-cube-maps.html

NOTE: This is a candidate for stable release branches.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
(cherry picked from commit f939776cb2)
2012-05-09 15:18:42 +08:00
Brian Paul
064c324d8d mesa: bump version to 8.0.2 in configs/default 2012-05-07 08:59:05 -06:00
Brian Paul
8700db8c87 mesa/gdi: remove clear_color() function
Setup the clearing color in the clear() function.

Reviewed-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit 9c53fc593e)
2012-05-07 08:59:04 -06:00
Brian Paul
53f88f8606 mesa/gdi: remove wmesa_set_renderbuffer_funcs() function
The code is no longer relevant.
Note: this driver is probably broken now.  There's no implementation
of ctx->Driver.Map/UnmapRenderbuffer().
(cherry picked from commit 4a1c660599)
2012-05-07 08:59:04 -06:00
Tom Stellard
0558ac1fd7 r300/compiler: Copy all instruction attributes during local transfoms
Instruction attributes like WriteALUResult and ALUResultCompare
were being discarded during the some of the local transformations.

This fixes the following piglit tests:

glsl1-inequality (vec2, pass)
loopfunc
fs-any-bvec2-using-if
fs-op-ne-bvec2-bvec2-using-if
fs-op-ne-ivec2-ivec2-using-if
fs-op-ne-mat2-mat2-using-if
fs-op-ne-vec2-vec2-using-if
fs-op-ne-mat2x3-mat2x3-using-if
fs-op-ne-mat2x4-mat2x4-using-if

https://bugs.freedesktop.org/show_bug.cgi?id=45921

(cherry-picked from commit 73249239cf)
2012-05-06 19:05:47 -04:00
Tom Stellard
649a8952df r300/compiler: Clear loop registers in vertex shaders w/o loops
The loop registers weren't being cleared, so any shader that was
executed after a shader containing loops was at risk of having a loop
randomly inserted into it.

This fixes over one hundred piglit tests, although these test
only failed during full piglit runs and would pass if
run individually.  The exact number of piglit tests that this patch
fixes will vary depending on the version of piglit and the order the
tests are run.

(cherry-picked from commit 4a269a8dc0)
2012-05-06 19:05:17 -04:00
Jeremy Huddleston
f36e638c76 darwin: Use ASL for logging
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 51691f0767)
2012-05-03 11:00:57 -07:00
Jeremy Huddleston
f818673acb darwin: Make reported errors more user-friendly
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit cf5db0a418)
2012-05-03 11:00:53 -07:00
Jeremy Huddleston
8010ff17ae darwin: Fix an error message
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 244dc05214)
2012-05-03 11:00:48 -07:00
Eric Anholt
93e94cbb48 intel: Fix rendering from textures after RenderTexture().
There's a serious trap for drivers: RenderTexture() does not indicate
that the texture is currently bound to the draw buffer, despite
FinishRenderTexture() signaling that the texture is just now being
unbound from the draw buffer.

We were acting as if RenderTexture() *was* the start of rendering and
that we could make texturing incoherent with the current contents of
the renderbuffer.  This caused intel oglconform sRGB
Mipmap.1D_textures to fail, because we got a call to TexImage() and
thus RenderTexture() on a texture bound to a framebuffer that wasn't
the draw buffer, so we skipped validating the new image into the
texture object used for rendering.

We can't (easily) make RenderTexture() indicate the start of drawing,
because both our driver and gallium are using it as the moment to set
up the renderbuffer wrapper used for things like MapRenderbuffer().
Instead, postpone the setup of the workaround render target miptree
until update_renderbuffer time, so that we no longer need to skip
validation of miptrees used as render targets.  As a bonus, this
should make GL_NV_texture_barrier possible.

(This also fixes a regression in the gen4 small-mipmap rendering since
3b38b33c16, which switched
set_draw_offset from image->mt to irb->mt but didn't move the irb->mt
replacement up before set_draw_offset).

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=44961
NOTE: This is a candidate for the 8.0 branch.
2012-04-30 11:46:42 -07:00
Eric Anholt
4e172532d1 i965/fs: Jump from discard statements to the end of the program when done.
From the GLSL 1.30 spec:

     The discard keyword is only allowed within fragment shaders. It
     can be used within a fragment shader to abandon the operation on
     the current fragment. This keyword causes the fragment to be
     discarded and no updates to any buffers will occur. Control flow
     exits the shader, and subsequent implicit or explicit derivatives
     are undefined when this control flow is non-uniform (meaning
     different fragments within the primitive take different control
     paths).

v2: Don't emit the final HALT if no other HALTs were emitted.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> (v1)
2012-04-30 11:28:01 -07:00
Dylan Noblesmith
e995b41a16 egl-static: fix printf warning
Noticed by clang:

egl_st.c:57:50: warning: field precision should have type 'int',
but argument has type 'size_t' (aka 'unsigned long') [-Wformat]
      ret = util_snprintf(path, sizeof(path), "%.*s/%s" UTIL_DL_EXT,
                                               ~~^~

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 2bb91274e2)
2012-04-29 11:22:57 +00:00
Dylan Noblesmith
fca35d2e5e st/vega: fix uninitialized values
C still treats array arguments exactly like pointer arguments.
By sheer coincidence, this still worked fine on 64-bit
machines where 2 * sizeof(float) == sizeof(void*), but not
on 32-bit.

Noticed by clang:

text.c:76:51: warning: sizeof on array function parameter will
return size of 'const VGfloat *' (aka 'const float *') instead of
'const VGfloat [2]' [-Wsizeof-array-argument]
   memcpy(glyph->glyph_origin, glyphOrigin, sizeof(glyphOrigin));

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 6a491b5728)
2012-04-29 11:22:47 +00:00
Dylan Noblesmith
4dd228a0a9 egl: fix uninitialized values
Noticed by clang:

eglimage.c:48:28: warning: argument to 'sizeof' in 'memset' call is
the same expression as the destination; did you mean to dereference
it? [-Wsizeof-pointer-memaccess]
   memset(attrs, 0, sizeof(attrs));
          ~~~~~            ^~~~~

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 071501a681)
2012-04-29 11:22:34 +00:00
Dylan Noblesmith
108d544cae util: fix uninitialized table
Most of the 256 values in the 'generic_to_slot' table were supposed to
be initialized with the default value 0xff, but were left at zero
(from CALLOC_STRUCT()) instead.

Noticed by clang:

u_linkage.h:60:31: warning: argument to 'sizeof' in 'memset' call is the same expression as the destination;
      did you mean to provide an explicit length? [-Wsizeof-pointer-memaccess]
   memset(table, 0xff, sizeof(table));
          ~~~~~               ^~~~~

Also fix a signed/unsigned comparison and a comment typo here.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 520521e380)
2012-04-29 11:22:23 +00:00
Dylan Noblesmith
546abd2503 util: fix undefined behavior
container_of() can legally return anything, even invalid addresses
that cause segfaults, when 'sample' is an uninitialized pointer.

Bug exposed by clang.

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit ccff749712)
2012-04-29 11:22:15 +00:00
Kenneth Graunke
bcc5caf642 i965/fs: Fix FB writes that tried to use the non-existent m16 register.
A little analysis shows that the worst-case value for "nr" is 17:
- base_mrf = 2                       ... 2
- header present (say gen == 5)      ... 4
- aa_dest_stencil_reg (stencil test) ... 5
- SIMD16 mode: += 4 * reg_width      ... 13
- source_depth_to_render_target      ... 15
- dest_depth_reg                     ... 17

This resulted in us setting base_mrf to 2 and mlen to 15.  In other
words, we'd try to use m2..m16.  But m16 doesn't exist pre-Gen6.  Also,
the instruction scheduler data structures use arrays of size 16, so this
would cause us to access them out of bounds.

While the debugger system routine may need m0 and m1, we don't use it
today, so the simplest solution is just to move base_mrf back to 1.
That way, our worst case message fits in m1..m15, which is legal.

An alternative would be to fail on SIMD16 in this case, but that seems
a bit unfortunate if there's no real need to reserve m0 and m1.

Fixes new piglit test shaders/depth-test-and-write on Ironlake,
as well as gzdoom.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=48218
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit aa429ea73c0931d5cfa2c263fe005ead8dc32ddf)
2012-04-27 16:53:08 -07:00
Jeremy Huddleston
69d8a25d42 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 00:27:07 -07:00
Jonas Maebe
6095a17534 apple: Fix a use after free
Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit c60ffd2840)
2012-04-23 16:18:19 -07:00
Jonas Maebe
bb30e76328 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 20:41:04 -07:00
Brian Paul
49ed43b6de mesa: add a couple fast-paths to fast_read_rgba_pixels_memcpy()
Accelerates a few glReadPixels cases for WebGL.
See https://bugs.freedesktop.org/show_bug.cgi?id=48545

v2: Per Jose, use bit twiddling for the swizzle case instead of ubyte
arrays (it's about 44% faster).

Note: This is a candidate for the 8.0 branch.

Reviewed-by: José Fonseca <jfonseca@vmware.com>

(cherry picked from commit a5e95a419e)
2012-04-17 17:41:09 -06:00
Yuanhan Liu
9f150ffe8e i915: set SPRITE_POINT_ENABLE bit correctly
When SPRITE_POINT_ENABLE bit is set, the texture coord would be
replaced, and this is only needed when we called something like
glTexEnvi(GL_POINT_SPRITE, GL_COORD_REPLACE, GL_TRUE).

And more,  we currently handle varying inputs as texture coord,
we would be careful when setting this bit and set it just when
needed, or you will find the value of varying input is not right
and changed.

Thus we do set SPRITE_POINT_ENABLE bit only when all enabled tex
coord units need do CoordReplace. Or fallback is needed to make
sure the rendering is right.

With handling the bit setup at i915_update_sprite_point_enable(),
we don't need the relative code at i915Enable then.

This patch would _really_ fix the webglc point-size.html test case and
of course, not regress piglit point-sprite and glean-pointSprite
testcase.

NOTE: This is a candidate for stable release branches.

v2: fallback just when all enabled tex coord units need do
    CoordReplace (Eric)
v3: move the sprite point validate code at I915InvalidateState (Eric)
v4: sprite point enable bit update based on _NEW_PROGRAM, too
    add relative _NEW-state comments to show what state is being used(Eric)

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
(cherry picked from commit c653287549)
2012-04-09 10:07:56 +08:00
Kenneth Graunke
edeb3976d4 i965: Actually upload sampler state pointers for the VS unit on Gen6.
We already program all the sampler state correctly, we just didn't give
the GPU a pointer to it for the VS stage.  Thus, any texturing other
than texelFetch() wouldn't work.

Fixes piglit test vs-textureLod-miplevels and 99 of oglconform's
glsl-bif-tex subtests.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 2a80a1e2a7)
2012-04-05 18:38:31 -07:00
Marek Olšák
a30790a9a3 r300g/swtcl: fix crash when back color is present in vertex shader
The shader transformation code sometimes produced invalid TGSI.
(cherry picked from commit 75f8990547)
2012-04-04 05:04:00 +02:00
Marek Olšák
3d436f6c37 r300g/swtcl: fix polygon offset
(cherry picked from commit c3481f3410)
2012-04-04 05:03:51 +02:00
Marek Olšák
8d40c2f5e2 r300g/swtcl: don't expose shader subroutine support
RET in the main function doesn't work. This should be fixed in Draw, but meh.
(cherry picked from commit 3b8fe06eb6)
2012-04-04 05:03:46 +02:00
Marek Olšák
d146c50358 r300g/swtcl: don't enter u_vbuf_mgr
(cherry picked from commit da2123051c)
2012-04-04 05:03:39 +02:00
Marek Olšák
1709144338 r300g/swtcl: don't print an error when getting ClipVertex
Draw can do it just fine.
(cherry picked from commit 5ce0598a03)
2012-04-04 05:03:34 +02:00
Chad Versace
54f7391664 glsl: Fix Android build
The build was broken by the line below, added in commit 4f82fed4.
  s_expression.cpp:26: #include <limits>

Mesa's half of the fix is to add 'external/astl/include' to the include
path. The other half of the fix requires implementing
numeric_limits<float>::infinity() in astl, for which I have patches
submitted upstream for review.

Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit 5497cc428f)
2012-03-23 16:28:38 -07:00
Dylan Noblesmith
89e796aef5 intel: fix null dereference processing HiZ buffer
Or technically, a near-null dereference.

https://bugs.freedesktop.org/show_bug.cgi?id=46303
https://bugs.freedesktop.org/show_bug.cgi?id=46739

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit 8d9decb75f)
2012-03-23 19:01:58 +00:00
Jakob Bornecrantz
0bf0ba44de docs: Add 8.0.2 md5sums
Signed-off-by: Jakob Bornecrantz <jakob@vmware.com>
2012-03-21 19:19:59 +00:00
Jakob Bornecrantz
5f7204c3bb docs: Add 8.0.2 release notes
Signed-off-by: Jakob Bornecrantz <jakob@vmware.com>
2012-03-21 18:01:20 +00:00
Jakob Bornecrantz
dc20396a14 mesa: Bump version number to 8.0.2
Signed-off-by: Jakob Bornecrantz <jakob@vmware.com>
2012-03-21 17:55:53 +00:00
Jakob Bornecrantz
770f785a6f mesa: Include mesa ES mapi generated files
Signed-off-by: Jakob Bornecrantz <jakob@vmware.com>
2012-03-21 17:55:53 +00:00
Brian Paul
4b52be53a4 docs: add links to 8.0.1 release notes 2012-03-21 17:55:53 +00:00
Daniel Vetter
e26aa8c660 i965: fixup W-tile offset computation to take swizzling into account
There's even a comment in the code containing the right swizzling
computations!

Previously this has not been noticed because we need to manually
enabled swizzling on snb/ivb (kernel 3.4 will do that) and we
don't use the separate stencil on ilk (where the bios enables
swizzling). This fixes

piglit ./bin/fbo-stencil  readpixels GL_DEPTH32F_STENCIL8 -auto

on recent drm-intel-next kernels.

Also remove the comment about ivb, it's stale now.

Swizzling detection is done by allocating a temporary x-tiled
buffer object. Unfortunately kernels before v3.2 lie on snb/ivb
because they claim that swizzling is enable, but it isn't. The
kernel commit that fixes this for backport to pre-v3.2 is

commit acc83eb5a1e0ae7dbbf89ca2a1a943ade224bb84
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Mon Sep 12 20:49:16 2011 +0200

    drm/i915: fix swizzling on gen6+

But if the kernel doesn't lie, this now works on swizzling and
not swizzling machines.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit f172eae8b2)
2012-03-20 12:20:04 -07:00
Eugeni Dodonov
2f5182cfcf intel: check for LLC support when reading maps
This checks for advertised LLC support by the GPU instead of relying on
the GPU generation for detection.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
(cherry picked from commit 84e5f1c635)
2012-03-20 12:19:58 -07:00
Eugeni Dodonov
7fe667a18d intel: verify if hardware has LLC support
Rely on libdrm HAS_LLC parameter to verify if hardware supports it. In
case the libdrm version does not supports this check, fallback to older
way of detecting it which assumed that GPUs newer than GEN6 have it.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
(cherry picked from commit 7def293204)
2012-03-20 12:19:38 -07:00
Kenneth Graunke
3d036f3f0a i965: Fix Gen6+ dynamic state upper bound on older kernels.
Kernels prior to 271d81b84171d84723357ae6d172ec16b0d8139c (March 2011)
don't support relocations outside of the target buffer object.  Rather
than guarding this with a I915_PARAM_HAS_RELAXED_DELTA check, just
smash the bound to 0xfffff001 like we do on Ironlake.

This effectively gives us no upper bound check, just like we did prior
to commit 271d81b84171d84723357ae6d172ec16b0d8139c.

Daniel Vetter would also like to mention that this relies on the guard
page at the end of the GTT.

Fixes a regression since 271d81b84171d84723357ae6d172ec16b0d8139c.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=46766
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
(cherry picked from commit b2ace06cbb)
2012-03-20 12:16:19 -07:00
Dylan Noblesmith
b056fc0741 mesa: add back glGetnUniform*v() overflow error reporting
The error was removed in:

commit 719909698c
Author: Ian Romanick <ian.d.romanick@intel.com>
Date:   Tue Oct 18 16:01:49 2011 -0700

    mesa: Rewrite the way uniforms are tracked and handled

The GL_ARB_robustness spec doesn't say the implementation
should truncate the output, so just return after setting
the required error like it did before the above commit.

Also fixup an old comment and add an assert.

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit b536ac6b2b)
2012-03-20 07:49:53 +00:00
Yuanhan Liu
adcb180671 i915: fallback for NPOT cubemap texture
Although some hardware support NPOT cubemap, but it seems we don't know
the right layout for NPOT cubemap. Thus seems we need do fallback for
other platforms as well.

See comments inline the code for more detailed info.

v2: give a more detailed info about why we need fallback for other
    platfroms as well.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=46666

NOTE: This is a candidate for stable release branches.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
(cherry picked from commit 40c995c1fd)
2012-03-19 10:15:48 +08:00
Jeremy Huddleston
d982036c3a darwin: Link against libxcb
Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 90a51753c4)
2012-03-16 17:14:31 -07:00
Jeremy Huddleston
63c8f7142c darwin: Build create_context.c
Fixes a build regression from: 588042a8ec

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit f9e1295cff)
2012-03-16 17:14:26 -07:00
Jeremy Huddleston
485d1c491a darwin: config file cleanups
Set our default compiler based on what our installed XCode prefers

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
(cherry picked from commit 61f6aff5d9)
2012-03-16 17:14:21 -07:00
Yuanhan Liu
7b1fbc6889 i965: handle gl_PointCoord for Gen4 and Gen5 platforms
This patch add the support of gl_PointCoord gl builtin variable for
platform gen4 and gen5(ILK).

Unlike gen6+, we don't have a hardware support of gl_PointCoord, means
hardware will not calculate the interpolation coefficient for you.
Instead, you should handle it yourself in sf shader stage.

But badly, gl_PointCoord is a FS instead of VS builtin variable, thus
it's not included in c.vue_map generated in VS stage. Thus the current
code doesn't aware of this attribute. And to handle it correctly, we
need add it to c.vue_map manually to let SF shader generate the needed
interpolation coefficient for FS shader. SF stage has it's own copy of
vue_map, thus I think it's safe to do it manually.

Since handling gl_PointCoord for gen4 and gen5 platforms is somehow a
little special, I added a lot of comments and hope I didn't overdo it ;)

v2: add a /* _NEW_BUFFERS */ comment to note the state flag dependency
    and also add the _NEW_BUFFERS dirty mask (Eric).

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=45975
Piglit: glsl-fs-pointcoord and fbo-gl_pointcoord

NOTE: This is a candidate for stable release branches.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 43af02ac73)
2012-03-16 10:44:01 +08:00
Yuanhan Liu
7f8ac0e70f i915: move the FALLBACK_DRAW_OFFSET check outside the drawing rect check
We have to do fallback when the 'Clipped Drawing Rectangle X/Y Max'
exceed the hardware's limit no matter the drawing rectangle offset
changed or not.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=46665

NOTE: This is a candidate for stable release branches.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit cf2f9ef015)
2012-03-16 10:37:43 +08:00
Yuanhan Liu
5cfc7d1167 i915: fix wrong rendering of gl_PointSize on Pineview
The current code would ignore the point size specified by gl_PointSize
builtin variable in vertex shader on Pineview. This patch servers as
fixing that.

This patch fixes the following issues on Pineview:
webglc: https://cvs.khronos.org/svn/repos/registry/trunk/public/webgl/sdk/tests/conformance/rendering/point-size.html
piglit: glsl-vs-point-size

NOTE: This is a candidate for stable release branches.

v2: pick Eric's nice tip for fixing this issue in hardware rendering.
v3: the last arg of EMIT_ATTR specify the size in _byte_. (Eric)

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 058fc6521e)
2012-03-16 10:36:57 +08:00
Yuanhan Liu
fae3a31bbb tnl: let _TNL_ATTRIB_POINTSIZE do not depend on ctx->VertexProgram._Enabled
We may specify the point size in a glsl vertex shader.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=46311
piglit: glsl-vs-point-size

NOTE: This is a candidate for stable release branches.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 9962280c33)
2012-03-16 10:35:42 +08:00
Brian Paul
b9f8cb9e0b mesa: fix GL_LUMINANCE handling in glGetTexImage
There are several cases in which we need to explicity "rebase" colors
(ex: set G=B=0) when getting GL_LUMINANCE textures:
1. If the luminance texture is actually stored as rgba
2. If getting a luminance texture, but returning rgba
3. If getting an rgba texture, but returning luminance

Fixes https://bugs.freedesktop.org/show_bug.cgi?id=46679

Also fixes the new piglit getteximage-luminance test.

Reviewed-by: José Fonseca <jfonseca@vmware.com>
(cherry picked from commit f5d0ced242)
2012-03-14 17:32:17 -06:00
Brian Paul
aabbf5adac mesa: use _mesa_rebase_rgba_float/uint() in glGetTexImage code
NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: José Fonseca <jfonseca@vmware.com>
2012-03-14 17:26:06 -06:00
Brian Paul
83728cf4ce mesa: use _mesa_rebase_rgba_float/uint() in glReadPixels code
See the comments for _mesa_rebase_rgba_float() for details.
Fixes https://bugs.freedesktop.org/show_bug.cgi?id=46679

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: José Fonseca <jfonseca@vmware.com>
(cherry picked from commit ad897fff77)
2012-03-14 17:23:59 -06:00
Brian Paul
8836517250 mesa: add _mesa_rebase_rgba_float/uint() functions
These will be used by glReadPixels() and glGetTexImage() to fix issues
with reading GL_LUMINANCE and other formats.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: José Fonseca <jfonseca@vmware.com>
(cherry picked from commit 774c402765)
2012-03-14 17:23:47 -06:00
Brian Paul
bc9d4ae6c7 util: add mutex lock in u_debug_memory.c code
The linked list of memory allocations was not protected by a mutex.
This lead to sporadic failures with multi-threaded apps.

Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
2012-03-14 12:27:57 +00:00
Brian Paul
437ed1faaf svga: add null vs pointer check in update_need_pipeline()
Based on a patch submitted by Vic Lee.  The other part of his patch
which checked the fs pointer wasn't needed.

This fixes a crash when clear() is called before any VS or FS is set.
But this can only happen when the driver is used without the Mesa
state tracker.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: José Fonseca <jfonseca@vmware.com>
(cherry picked from commit 9917988223)
2012-03-09 08:44:06 -07:00
Zack Rusin
fa9efdbab8 svga: fix the rasterizer state resets
draw module calls back into the driver and sets certain parts
of the state to whatever it needs, unfortunately unless you
get the ordering of calls to draw just right you'll end up
reseting your own driver state. That's what was happening to us
draw module would under certain conditions reset our own driver
state.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 831de96db8)
2012-03-09 08:43:58 -07:00
Zack Rusin
151d32dd45 svga: Fix stencil op mapping
We were inverting the meaning of the stencil op flags: in svga/d3d
the normal incr/decr wraps and the SAT ops clamp.
This fixes piglit failures (at least stencil-twoside and stencil-wrap).
We should backport this everywhere we can.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 5d9bfc4d3f)
2012-03-09 08:43:41 -07:00
Zack Rusin
1fae49b0f5 svga: fix a crash happening before setting fragment shaders.
In certain situations API's will call pipe->clear which doesn't
require fragment shader, but then we'd try to verify the pipeline
and assume fragment shader was always set. This was leading to
crash when API would just call simple clear's before anything else.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: José Fonseca <jfonseca@vmware.com>
(cherry picked from commit cff0eac702)
2012-03-09 08:43:22 -07:00
Neil Roberts
9664fb70dc mesa: Don't disable fast path for normalized types
Mesa has a fast path for the generic fallback when using glReadPixels
for RGBA data which uses memcpy.  However it was really difficult to
hit this case because it would not be used if any transferOps are
enabled.  Any type apart from floating point or non-normalized integer
types (so any of the common types) would force enabling clamping so
the fast path could not be used.  This patch makes it ignore clamping
when determining whether to use the fast path if the data type of the
buffer is an unsigned normalized type because in that case clamping
will not have any effect anyway.

https://bugs.freedesktop.org/show_bug.cgi?id=46631

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit d9c4209777)
2012-03-05 12:53:37 -07:00
José Fonseca
a30809878f svga: Clamp advertised PIPE_SHADER_CAP_MAX_TEMPS to SVGA3D_TEMPREG_MAX.
Some backends may advertise more temps than SVGA3D_TEMPREG_MAX, but the
driver is hardwired to only support up to the value defined by
SVGA3D_TEMPREG_MAX, so clamp to it.

Reviewed-by: Brian Paul <brianp@vmware.com>
2012-03-01 08:15:53 +00:00
Yuanhan Liu
9f44387e48 mesa: let GL3 buf obj queries not depend on opengl major version
While the ARB_map_buffer_range extension spec says nothing about these
queries -- they were added in GL 3.0 --, it seems like this could be an
error in the extension spec.  This is one of the extensions, like
ARB_framebuffer_object, that "back ports" OpenGL 3.0 functionality to
previous versions.  These extensions are supposed to provide identical
functionality to OpenGL 3.0.  The other cases of mismatches have been
determined to be bugs in the extension specs.

And tools like apitrace rely on such queries to function properly.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Signed-off-by: José Fonseca <jfonseca@vmware.com>
Acked-by: Brian Paul <brianp@vmware.com>
Acked-by: Ian Romanick <ian.d.romanick@intel.com>
2012-03-01 08:15:36 +00:00
Kenneth Graunke
16cc79f975 i965: Only set Last Render Target Select on the last FB write.
Fixes GPU hangs in OilRush, Trine, and Amnesia: The Dark Descent,
which all use MRT (multiple render targets).

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=38720
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=40059
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=45216
Reviewed-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 172bb92db1)
2012-02-28 11:26:05 -08:00
Tom Stellard
82043eb72d r300/compiler: Fix bug when lowering KILP on r300 cards
KILP instruction inside IF blocks were being lowered to an unconditional
KIL.  Since r300 doesn't support branching, when the IF's were lowered
to conditional moves, the KIL would always be executed.  This is not a
problem with the mesa state tracker, because the GLSL compiler handles
lowering IF's, but this bug was appearing in the VDPAU state tracker,
which does not use the GLSL compiler.

(cherry picked from commit 342cac7166)
2012-02-26 20:51:21 -05:00
Marek Olšák
2845a0be81 gallium/rtasm: properly detect SSE and SSE2
This should fix crashes on ancient processors.
(cherry picked from commit 74d303521e)
2012-02-27 02:06:59 +01:00
Ian Romanick
d38a295289 docs: Add 8.0.1 release md5sums
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
2012-02-16 23:25:14 -08:00
Ian Romanick
fe77fd3983 docs: Add 8.0.1 release notes
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
2012-02-16 18:55:18 -08:00
Ian Romanick
b695078937 mesa: Bump version number to 8.0.1
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
2012-02-16 18:54:28 -08:00
Simon Farnsworth
106ea10d1b r600g: Use a fake reloc to sleep for fences
r300g is able to sleep until a fence completes rather than busywait because
it creates a special buffer object and relocation that stays busy until the
CS containing the fence is finished.

Copy the idea into r600g, and use it to sleep if the user asked for an
infinite wait, falling back to busywaiting if the user provided a timeout.

Signed-off-by: Simon Farnsworth <simon.farnsworth@onelan.co.uk>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 8cd03b933c)

Conflicts:

	src/gallium/drivers/r600/r600_pipe.c
2012-02-16 17:49:23 +01:00
Alex Deucher
fca1a33c96 r600g: 128 bit formats require tile_type = 1 on cayman
Noticed by taiu on IRC.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

(cherry picked from commit 5e1495b2d9)

Conflicts:

	src/gallium/drivers/r600/evergreen_state.c
2012-02-16 17:48:13 +01:00
Alex Deucher
036d999265 r600g: fix tex tile_type offset for cayman
Noticed by taiu on IRC.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

(cherry picked from commit acca690c25)

Conflicts:

	src/gallium/drivers/r600/evergreen_state.c
2012-02-16 17:45:41 +01:00
Kenneth Graunke
e3943cf1cc i965: Emit Ivybridge VS workaround flushes.
I recently discovered this text in the BSpec.  It seems wise to comply,
though I haven't observed it to fix anything yet.

Fixes a regression in glean/fbo since 28cfa1fa21.

NOTE: This is a candidate for stable release branches.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=45221
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 709f50928e)
2012-02-16 00:09:31 -08:00
Kenneth Graunke
0aadb240e1 i965/fs: Take # of components into account in try_rewrite_rhs_to_dst.
Commit dc7f449d1a introduced a new method
for avoiding MOVs: try to rewrite the destination of the instruction
that produced the RHS so it writes into the LHS.

Unfortunately, this is not safe for swizzled texturing operations, as
they return a set of four contiguous registers.  Consider the following:

(assign (x)
        (var_ref vec_ctor_x)
        (swiz x (tex vec4 (var_ref m_sampY) (var_ref m_cordY) 0 1 ())))

In this case, the source and destination registers are equal, since
reg_offset is 0 for both.  Yet, this is only a partial move: the texture
operation generates four registers, and the LHS only covers one.

Fixes color distortion in XBMC when using GLSL shaders.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=44333
Reviewed-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 4b27406820)
2012-02-15 17:17:35 -08:00
Kenneth Graunke
740123fff7 i965/fs: Add a new fs_inst::regs_written function.
Certain instructions write more than one register.  Texturing, for
example, returns 4 registers.  (We set rlen to 4 even for TXS and float
shadow sampling.)  Some math functions return 2.  Most return 1.

The next commit introduces a use of this function.

Reviewed-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 8ab02b5118)
2012-02-15 17:17:16 -08:00
Ian Romanick
ff1d945032 swrast: Only avoid empty _TexEnvPrograms
If the generated shader for _TexEnvProgram is empty, force the use of
the fixed-function code.  Otherwise, go ahead and use the shader.
This works around a mysterious issue on i915 where fixed-function
software fallbacks are not working correctly.

This isn't really the fix we want, but it works around the issue.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=45872
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=45876
(cherry picked from commit 3e22d4e5fc)
2012-02-15 14:51:37 -08:00
Ian Romanick
efca49fd51 glapi: Include GLES2 headers for ES2 extension functions
This fixes build errors like

In file included from glapi_dispatch.c:91:
../../../src/mapi/glapi/glapitemp.h:4641: error: no previous prototype for
'glDrawBuffersNV'

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Tested-by: Lucas Stach <dev@lynxeye.de>
(cherry picked from commit 8f3be33985)
2012-02-15 14:51:26 -08:00
Anuj Phogat
d45a5fde45 meta: Avoid FBO resizing/reallocating in decompress_texture_image
Reallocate/resize decompress FBO only if texture image width/height is
greater than existing decompress FBO width/height.

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 66bf25f1a2)
2012-02-15 14:48:59 -08:00
Paul Berry
e55f2d97f6 i915: Fix type of "specoffset" variable.
Commit 2e5a1a2 (intel: Convert from GLboolean to 'bool' from
stdbool.h.) converted the "specoffset" local variable (in
intel_tris.c) from a GLboolean to a bool.  However, GLboolean was the
wrong type for specoffset--it should have been a GLuint (to match the
declaration of specoffset in struct intel_context).

This patch changes specoffset to the proper type.

Fixes piglit test general/two-sided-lighting-separate-specular.

This is a candidate for stable branches.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=45917
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 6b0a07f9ce)
2012-02-15 14:48:20 -08:00
Mathias Fröhlich
6e09d3cff2 state_stracker: Fix access to uninitialized memory.
Fix an access to uninitialized memory pointed out by valgrind in
glsl_to_tgsi_visitor::simplify_cmp(void).

Note: This is a candidate for the 8.0 branch.
Signed-off-by: Mathias Fröhlich <Mathias.Froehlich@web.de>
(cherry picked from commit 1d01429c6a)
2012-02-15 14:45:06 -08:00
Eric Anholt
99f9c9789a i965/fs: Enable register spilling on gen7 too.
It turns out the same messages work on gen7, we were just being paranoid.

Fixes the penumbra shadows mode of Lightsmark since the register
allocation fix.

NOTE: This is a candidate for release branches.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 93831a54c7)
2012-02-14 17:53:14 -08:00
Eric Anholt
a63d79dd40 i965: Report the failure message when failing to compile the fragment shader.
We just abort later, but at least this should result in more
informative bug reports.

NOTE: This is a candidate for release branches.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit a7f46eadea)
2012-02-14 17:53:07 -08:00
Anuj Phogat
ff7ccb1cf1 meta: Add pixel store/pack operations in decompress_texture_image
This patch adds the pixel store operations in decompress_texture_image().
decompress_texture_image() is used in glGetTexImage() for compressed
textures with unsigned, normalized values.

It also fixes the failures in intel oglconform pxstore-gettex due to
following sub test cases:

 - Test all mipmaps with byte swapping enabled
 - Test all small mipmaps with all allowable alignment values
 - Test subimage packing for all mipmap levels

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=40864

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 4042702591)
2012-02-14 17:52:50 -08:00
Dave Airlie
9e98d38e58 st/mesa: only resolve if number of samples is > 1
Marek: this fixes a firefox crash and maybe even:
https://bugs.freedesktop.org/show_bug.cgi?id=45943

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Marek Olšák <maraeo@gmail.com>
(cherry picked from commit 094eeff199)
2012-02-14 17:52:32 -08:00
Brian Paul
a2186a2ea6 swrast: fix span color type selection
Fixes a regression from commit 660ed923de.
The basic idea is to look at the format of the dest renderbuffer and
choose either GLubyte or GLfloat for colors.  The previous code used
_mesa_format_to_type_and_comps() which could return a bunch types other
than ubyte/float.

Determine the datatype at renderbuffer mapping time to avoid frequent
calls to the format query functions.

NOTE: This is a candidate for the 8.0 branch.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=45578
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=45577
(cherry picked from commit bd1ae51b13)
2012-02-14 17:51:39 -08:00
Kenneth Graunke
be1377c33c i965: Fix border color on Ironlake.
Ironlake appears to check our pointer against the General State Base
Address upper bound, rather than ignoring the zero bound as it ought.

Unfortunately, since we leave GSBA set to zero, there is no logical
upper bound.  Set it to the maximum possible value, which should work
since our virtual addresses only go up to 2GB.

+94 piglits.

NOTE: This is a candidate for stable release branches.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=28924
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 3340b47c22)
2012-02-14 17:51:22 -08:00
Brian Paul
9489ce6e06 mesa: fix proxy texture target initialization
The mapping from TEXTURE_x_INDEX to GL_TEXTURE_x was broken in
alloc_proxy_textures() because the elements in the targets[] array
were in the wrong order.

This didn't actually cause any failures since we never really use the
proxy texture's Target field.  But let's get it right.

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit d925b0d4a7)
2012-02-14 17:50:11 -08:00
Chad Versace
c1dd6ddfd0 i965: Remove file i965/junk, accidentally added in 7b36c68
(cherry picked from commit b44c459cc3)
2012-02-14 11:07:24 -08:00
Eric Anholt
3d3bd0e917 i965: Fix HiZ change compiler warning.
(cherry picked from commit 94866ffbb8)
2012-02-14 11:07:19 -08:00
Chad Versace
e1f9820b47 i965: Rewrite the HiZ op
This is a combination of 4 commits. The first commit rewrites the HiZ op,
and remaining three fix bugs introduced by the rewrite.

======== commit 1 ========

i965: Rewrite the HiZ op

The HiZ op was implemented as a meta-op. This patch reimplements it by
emitting a special HiZ batch. This fixes several known bugs, and likely
a lot of undiscovered ones too.

==== Why the HiZ meta-op needed to die ====

The HiZ op was implemented as a meta-op, which caused lots of trouble. All
other meta-ops occur as a result of some GL call (for example, glClear and
glGenerateMipmap), but the HiZ meta-op was special. It was called in
places that Mesa (in particular, the vbo and swrast modules) did not
expect---and were not prepared for---state changes to occur (for example:
glDraw; glCallList; within glBegin/End blocks; and within
swrast_prepare_render as a result of intel_miptree_map).

In an attempt to work around these unexpected state changes, I added two
hooks in i965:
  - A hook for glDraw, located in brw_predraw_resolve_buffers (which is
    called in the glDraw path). This hook detected if a predraw resolve
    meta-op had occurred, and would hackishly repropagate some GL state
    if necessary. This ensured that the meta-op state changes would not
    intefere with the vbo module's subsequent execution of glDraw.
  - A hook for glBegin, implemented by brwPrepareExecBegin. This hook
    resolved all buffers before entering
    a glBegin/End block, thus preventing an infinitely recurring call to
    vbo_exec_FlushVertices. The vbo module calls vbo_exec_FlushVertices to
    flush its vertex queue in response to GL state changes.

Unfortunately, these hooks were not sufficient. The meta-op state changes
still interacted badly with glPopAttrib (as discovered in bug 44927) and
with swrast rendering (as discovered by debugging gen6's swrast fallback
for glBitmap). I expect there are more undiscovered bugs. Rather than play
whack-a-mole in a minefield, the sane approach is to replace the HiZ
meta-op with something safer.

==== How it was killed ====

This patch consists of several logical components:
  1. Rewrite the HiZ op by replacing function gen6_resolve_slice with
     gen6_hiz_exec and gen7_hiz_exec. The new functions do not call
     a meta-op, but instead manually construct and emit a batch to "draw"
     the HiZ op's rectangle primitive. The new functions alter no GL
     state.
  2. Add fields to brw_context::hiz for the new HiZ op.
  3. Emit a workaround flush when toggling 3DSTATE_VS.VsFunctionEnable.
  4. Kill all dead HiZ code:
     - the function gen6_resolve_slice
     - the dirty flag BRW_NEW_HIZ
     - the dead fields in brw_context::hiz
     - the state packet manipulation triggered by the now removed
       brw_context::hiz::op
     - the meta-op workaround in brw_predraw_resolve_buffers (discussed
       above)
     - the meta-op workaround brwPrepareExecBegin (discussed above)

Note: This is a candidate for the 8.0 branch.
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Acked-by: Paul Berry <stereotype441@gmail.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=43327
Reported-by: xunx.fang@intel.com
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=44927
Reported-by: chao.a.chen@intel.com
Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit 7b36c68ba6)

======== commit 2 ========

i965/gen7: Fix GPU hangs from the HiZ op.

The wm max threads is in the same dword as the dispatch enable.  The
hardware gets super angry if you set max threads to 0, even if you
aren't dispatching threads.
(cherry picked from commit e5b225afbd)

======== commit 3 ========

i965/gen7: Fix the length of the DS state packet in the HiZ op.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit cdcfaa64e3)

======== commit 4 ========

i965/gen7: Fix the length of the MULTISAMPLE state packet in the HiZ op.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit a7750c9fb5)
2012-02-14 10:50:42 -08:00
Brian Paul
65526d54aa docs: remove link to the GLSL compiler page
The page is terribly out of date.
2012-02-09 18:08:02 -07:00
Brian Paul
6aa9ce2687 docs: add VMware link 2012-02-09 18:07:47 -07:00
Brian Paul
14cf3dd826 docs: update info about supported systems, GPUs, APIs
Add link to Intel's Linux graphics page, etc.
2012-02-09 18:07:22 -07:00
Brian Paul
7aef839760 docs: add news item for 8.0 release 2012-02-09 15:52:13 -07:00
Ian Romanick
fb56b0972d docs: Add 8.0 MD5 checksums
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
2012-02-09 14:28:58 -08:00
Ian Romanick
f9c9933f9c mesa: Bump version number to 8.0 (final)
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
2012-02-09 14:26:15 -08:00
Anuj Phogat
b2b5d6b8fb mesa: fix maximum allowed proxy texture size condition
width, height parameter in glTexImage2D() includes: texture image
width + 2 * border (if any). So when doing the texture size check
in _mesa_test_proxy_teximage() width and height should not exceed
maximum supported size for target texture type + 2 * border.
i.e. 1 << (ctx->Const.MaxTextureLevels - 1) + 2 * border

Texture border is anyway stripped out before it is given to intel
or gallium drivers.

This patch fixes Intel oglconform test case:
max_values negative.textureSize.textureCube
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=44970

Note: This is a candidate for mesa 8.0 branch.

Signed-off-by: Anuj Phogat <anuj.phogat@gmail.com>
Reviewed-by: Ian Romanick <idr@freedesktop.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit ea228d97f8)
2012-02-09 13:16:07 -08:00
Eric Anholt
e86d90eb20 dri: Add Unigine Tropics as an app that requires the GLSL warn workaround.
I wasn't seeing it be needed because of the previous bug.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
(cherry picked from commit b8c9252570)
2012-02-07 16:18:32 -08:00
Eric Anholt
1531b94471 dri: Fix typo in xml file that made all applications use the workaround.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
(cherry picked from commit 4dd2743d45)
2012-02-07 16:18:32 -08:00
Eric Anholt
b5efe0881e dri: Add a default drirc to be installed to provide application workarounds.
Specifially, this being present works around a bug in Unigine
Sanctuary on i965 which previously resulted in bad rendering.

NOTE: This is a candidate for the 8.0 branch.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit ff2497dca3)

NOTE: Compared to ff2497d this does not install the default drirc.
The pre-automake build system is sufficiently braindamaged to make
this exceptionally difficult.
2012-02-07 16:17:33 -08:00
Chih-Wei Huang
73e15679ce vbo: fix a building error
Signed-off-by: Marek Olšák <maraeo@gmail.com>

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit f8be4f33d3)
2012-02-07 10:26:17 -08:00
Eric Anholt
974a67b41e glsl: Add error case for switch() with two default cases.
Fixes piglit switch-case-duplicated.vert.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 57e44371a5)
2012-02-07 10:26:17 -08:00
Eric Anholt
83075bd0fe glsl: Throw an error when faced with a duplicated switch() case label.
The error message I chose matches gcc's error.  Fixes piglit
switch-case-duplicated.vert.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 140632190c)
2012-02-07 10:26:17 -08:00
Eric Anholt
d799a7b585 glsl: Add other missing error location information for switch statements.
NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 01a5a2c9d7)
2012-02-07 10:26:17 -08:00
Eric Anholt
2b4df494b1 glsl: Add missing location info to case labels.
Otherwise, the upcoming error messages said the location was 0:0(0).

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 663dcbbffe)
2012-02-07 10:26:17 -08:00
Eric Anholt
728bda08d8 glsl: Throw the required error when a case label is a non-constant.
It's not quite spelled out in the spec text, but the grammar indicates
that only constant values are allowed as switch() case labels (and
only constant values make sense, anyway).

Fixes piglit glsl-1.30/compiler/switch-statement/switch-case-uniform-int.vert.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 2c3e10e719)
2012-02-07 10:26:16 -08:00
Eric Anholt
f775d9aa84 glsl: Save and restore the whole switch state for nesting.
This stuffs them all in a struct for sanity.  Fixes piglit
glsl-1.30/execution/switch/fs-uniform-nested.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 22d81f154f)
2012-02-07 10:26:16 -08:00
Eric Anholt
6887ec766b mesa: Fix the error message function names for glFlushMappedBufferRange().
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 27af00eac8)
2012-02-07 10:26:16 -08:00
Eric Anholt
6d82dc18f1 mesa: Fix bad-enum/no-buffer error handling for buffer object functions.
For all the extension entrypoints using the get_buffer() helper, they
wanted the same error handling.  In some cases, the error was doing
the same error return whether target was a bad enum, or a user buffer
wasn't bound.

(Actually, GL_ARB_map_buffer_range doesn't specify the error for a zero
buffer being bound for MapBufferRange, though it does for
FlushMappedBufferRange.  This appears to be an oversight).

Fixes piglit GL_ARB_copy_buffer/negative-bound-zero.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit f20fb80a91)
2012-02-07 10:26:16 -08:00
Carl Worth
02962ea086 glsl: Avoid ralloc_stealing a long-lived object to a short-lived parent
In commit 6ecee54a9a a call to
talloc_reference was replaced with a call to talloc_steal. This was in
preparation for moving to ralloc which doesn't support reference
counting.

The justification for talloc_steal within token_list_append in that
commit is that the tokens are being copied already. But the copies are
shallow, so this does not work.

Fortunately, the lifetime of these tokens is easy to understand. A
token list for "replacements" is created and stored in a hash table
when a function-like macro is defined. This list will live until the
macro is #undefed (if ever).

Meanwhile, a shallow copy of the list is created when the macro is
used and the list expanded. This copy is short-lived, so is unsuitable
as a new parent.

So we can just let the original, longer-lived owner continue to own
the underlying objects and things will work.

This fixes bug #45082:

	"ralloc.c:78: get_header: Assertion `info->canary == 0x5A1106'
	failed." when using a macro in GLSL
	https://bugs.freedesktop.org/show_bug.cgi?id=45082

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>

NOTE: This is a candidate for stable release branches.
(cherry picked from commit cd2e2187cb)
2012-02-07 10:26:16 -08:00
Brian Paul
358389fe59 mesa: reference shared state in glPushAttrib(GL_TEXTURE_BIT)
This fixes a dangling texture object pointer bug hit via wglShareLists().
When we push the GL_TEXTURE_BIT state we may push references to the default
texture objects which are owned by the gl_shared_state object.  We don't
want to accidentally delete that shared state while the attribute stack
references shared objects.  So keep a reference to it.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: José Fonseca <jfonseca@vmware.com>
(cherry picked from commit a1471e4877)
2012-02-07 10:23:43 -08:00
Brian Paul
2ed8367d72 mesa: use new _mesa_reference_shared_state() function
This cleans up the reference counting of shared context state.
The next patch will use this to fix an actual bug.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: José Fonseca <jfonseca@vmware.com>
(cherry picked from commit 361cd53a77)
2012-02-07 10:23:43 -08:00
Eric Anholt
f723df005c i965: Add a driconf option to force GLSL extension behavior to "warn".
This can be used to work around broken application behavior, like in
Unigine where it attempts to use texture arrays without declaring
either "#extension GL_EXT_texture_array : enable" or "#version 130".

NOTE: This is a candidate for the 8.0 branch.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 642247883f)
2012-02-07 10:20:07 -08:00
Eric Anholt
92d842c073 mesa: Add a flag for forcing all GLSL extensions to "warn".
NOTE: This is a candidate for the 8.0 branch.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit b9e27cc142)
2012-02-07 10:20:07 -08:00
Eric Anholt
1b5e151ffa i965/vs: Avoid allocating registers in to the gen7 MRF hack region.
This is the corresponding fix to the previous one for the FS, but I
don't have a particular test for it.

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit 9195191e50)
2012-02-07 10:20:07 -08:00
Chad Versace
20da01fecd swrast: Fix fixed-function fragment processing
On i965, _mesa_ir_link_shader is never called. As a consequence, the
current fragment program (ctx->FragmentProgram->_Current) exists but is
invalid because it has no instructions. Yet swrast continued to attempt to
use the empty program.

To avoid using the empty program, this patch 1) defines a new function,
_swrast_use_fragment_program, which checks if the current fragment program
exists and differs from the fixed function fragment program, and, when
appropriate, 2) replaces checks of the form
    if (ctx->FragmentProgram->_Current == NULL)
with
    if (_swrast_use_fragment_program(ctx))

Fixes the following oglconform regressions on i965/gen6:
    api-fogcoord(basic.allCases.log)
    api-mtexcoord(basic.allCases.log)
    api-seccolor(basic.allCases.log)
    api-texcoord(basic.allCases.log)
    blend-separate(basic.allCases)
    colorsum(basic.allCases.log)

The tests were ran with the GLXFBConfig:
    visual  x   bf lv rg d st  colorbuffer  sr ax dp st accumbuffer ms  cav
  id dep cl sp  sz l  ci b ro  r  g  b  a F gb bf th cl  r  g  b  a ns b eat
----------------------------------------------------------------------------
0x021 24 tc  0  32  0 r  y .   8  8  8  8 .  .  0 24 8  0  0  0  0  0 0 None

(Note: I originally believed that the hunk in
_swrast_update_fragment_program was unnecessary. But it is required to fix
blend-separate.)

Note: This is a candidate for the 8.0 branch.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=43327
Reveiwed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Ian Romanick <idr@freedesktop.org>
Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit 1c0f1dd42a)
2012-02-07 10:20:07 -08:00
Eric Anholt
203ef2a12b mesa: Don't round-trip integer texture data through a floating point temp.
This was losing bits of precision.  Fixes (with the previous commits):
piglit EXT_texture_integer/getteximage-clamping
piglit EXT_texture_integer/getteximage-clamping GL_ARB_texture_rg
oglc advanced.mipmap.upload

Regresses oglc negative.typeFormatMismatch.teximage from fail to
abort, because it's been hitting texstore for a format/type combo that
shouldn't happen.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 5f65598cc7)
2012-02-07 10:20:07 -08:00
Eric Anholt
05ff4d209d mesa: When unpacking signed integer pixel data, don't clamp to 0.
In the core, we always treat spans of int/uint data as uint, so this
extract function was truncating storage of integer pixel data to a n
int texture to (0, max_int) instead of (min_int, max_int).  There is
probably missing code for handling truncation on conversion between
pixel formats, still, but this does improve things.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit dadbec1e90)
2012-02-07 10:20:07 -08:00
Eric Anholt
c1ccb52c72 mesa: Add clamping for packing of integer data.
Mostly fixes piglit EXT_texture_integer/getteximage-clamping.  The
remaining failure involves precision loss on storing of int32 texture
data (something I knew was an issue, but wasn't trying to test).

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 8b97bb02fb)
2012-02-07 10:20:07 -08:00
Eric Anholt
35af090907 mesa: Add missing format unpack for some integer texture formats.
This cut and paste is pretty awful.  I'm tempted to do a lot of this
using preprocessor tricks for customizing the parameter type from a
template function, but that's just a different sort of hideous.

Fixes 8 Intel oglconform int-textures cases.

NOTE: This is a candidate for the 8.0 branch.
v2: Add alpha formats, too.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit de24ccabd6)
2012-02-07 10:20:06 -08:00
Eric Anholt
80cd02f517 i965: Don't allow rendering to non-GL_RED/RG/RGBA integer textures.
Fixes piglit EXT_texture_integer/fbo-blending.
(cherry picked from commit 3a8cf3357a)
2012-02-07 10:20:06 -08:00
Eric Anholt
0749290d69 intel: Pass the gl_renderbuffer to render_target_supported() vtable method.
I'm going to want to go looking at it for an integer texture fix.

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit 796f44d779)
2012-02-07 10:20:06 -08:00
Eric Anholt
f62c8648d3 intel: Make a renderbuffer wrapping a texture have the same _BaseFormat.
Otherwise, when you asked for the _BaseFormat of an rb wrapping a
GL_RGB texture, you got GL_RGBA because that's what we were storing
the texture data as.

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit 7cac88679b)
2012-02-07 10:20:06 -08:00
Eric Anholt
869728bd99 intel: Simplify intel_renderbuffer_update_wrapper() by passing in the image.
NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit b73f5df648)
2012-02-07 10:20:05 -08:00
Eric Anholt
1100a19da8 intel: Drop intel_wrap_miptree().
Most of this function was just calling
intel_renderbuffer_update_wrapper(), which was called immediately
afterwards in the only caller.

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit 74484c5d41)
2012-02-07 10:20:05 -08:00
Yuanhan Liu
f811d501b6 i965: fix inverted point sprite origin when rendering to FBO
When rendering to FBO, rendering is inverted. At the same time, we would
also make sure the point sprite origin is inverted. Or, we will get an
inverted result correspoinding to rendering to the default winsys FBO.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=44613

NOTE: This is a candidate for stable release branches.

v2: add the simliar logic to ivb, too (comments from Ian)
    simplify the logic operation (comments from Brian)

v3: pick a better comment from Eric
    use != for the logic instead of ^ (comments from Ian)

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>
(cherry picked from commit eaf360e5bf)
2012-02-07 10:09:02 -08:00
Eric Anholt
ecd0d46070 i965/fs: Fix rendering corruption in unigine tropics.
We were allocating registers into the MRF hack region, resulting in
sparkly renering in a few of the scenes.  We could do better
allocation by making an MRF class, having MRFs conflict with the
corresponding GRFs, and tracking the live intervals of the "MRF"s and
setting up the conflicts.  But this is way easier for the moment.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit e910241e97)
2012-02-07 10:07:41 -08:00
Ian Romanick
cf4a7c41f6 intel: Avoid divide by zero for very small linear blits
If size is small (such as 1),

   pitch = ROUND_DOWN_TO(MIN2(size, (1 << 15) - 1), 4);

makes pitch = 0.  Then

   height = size / pitch;

causes a division-by-zero exception.  If pitch is zero, set height to
1 and avoid the division.

This fixes piglit's bin/getteximage-formats test and glean's
bufferObject test.

NOTE: This is a candidate for the 8.0 release branch.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=44971
(cherry picked from commit d59466279e)
2012-02-07 10:02:23 -08:00
Ian Romanick
74a5f03066 intel: Remove num_mapped_regions assertion from _intel_batchbuffer_flush
There are cases where a buffer can be mapped while another buffer is
flushed.  This can happen in the CopyPixels meta-op path for piglit's
fbo-mipmap-copypix.  After some discussion with Eric, it seems this
assertion is no longer necessary, and it has always been too strict.

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=43328
Cc: Eric Anholt <eric@anholt.net>
Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit 65b096aedd)
2012-02-07 10:02:18 -08:00
Ian Romanick
65b500857e intel: FBOs with texture border are unsupported
FBOs differ from textures in a significant way.  With textures, we can
strip the border and get correct rendering except when the application
fetches texels outside [0,1].

With an FBO, the pixel at (0,0) is in the border.  The
ARB_framebuffer_object spec says:

    "If the attached image is a texture image, then the window
    coordinates (x[w], y[w]) correspond to the texel (i, j, k), from
    figure 3.10 as follows:

                           i = (x[w] - b)

                           j = (y[w] - b)

                           k = (layer - b)

    where <b> is the texture image's border width..."

Since the border doesn't exist, we can never render any pixels in the
correct location.  Just mark these FBOs FRAMEBUFFER_UNSUPPORTED.

NOTE: This is a candidate for the 8.0 branch.

Signed-off-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=42336
(cherry picked from commit 87b4c9b322)
2012-02-03 11:08:40 -07:00
Carl Worth
c231482725 glcpp: Fix so that trailing punctuation does not prevent macro expansion
The trick here is that flex always chooses the rule that matches the most
text. So with a input text of "two:" which we want to be lexed as an
IDENTIFIER token "two" followed by an OTHER token ":" the previous OTHER
rule would match longer as a single token of "two:" which we don't want.

We prevent this by forcing the OTHER pattern to never match any
characters that appear in other constructs, (no letters, numbers, #,
_, whitespace, nor any punctuation that appear in CPP operators).

Fixes bug #44764:

	GLSL preprocessor doesn't replace defines ending with ":"
	https://bugs.freedesktop.org/show_bug.cgi?id=44764

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>

NOTE: This is a candidate for stable release branches.
(cherry picked from commit 7ab1c7f792)
2012-02-03 11:08:26 -07:00
Ville Syrjala
50a8b9971e gallium/dri: Handle xserver that doesn't send needless DRI2 invalidate events
Ever since xserver commit 531869448d07e00ae241120b59f3aaaa5709d59c,
the server no longer sends invalidate events to clients, unless they
have performed a GetBuffers request since the drawable was last
invalidated.

If the drawable gets invalidated immediately after the GetBuffers
request was processed by the X server, it's possible that Xlib
will process the invalidate event while waiting for the GetBuffers
reply. So the server, thinking the client knows that the buffers
are invalid, is waiting for another GetBuffers request before
sending any more invalidate events. The client, on the other hand,
believes the buffers to be valid, and thus is expecting to receive
another invalidate event before it has to send another GetBuffers
request. The end result is that the client never again sends
a GetBuffers request.

To avoid this problem, take a snapshot of the lastStamp before
doing GetBuffers, and retry if the snapshot and the current
lastStamp no longer match after the GetBuffers reply has been
processed.

Signed-off-by: Ville Syrjälä <syrjala@sci.fi>
Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit 0fcc518964)
2012-02-03 12:05:36 +00:00
Brian Paul
a4cb2fdb64 gallium/postprocess: move declarations before code
To fix MSVC build.
(cherry picked from commit 8cbe699c0d)
2012-02-02 15:22:17 -07:00
Lauri Kasanen
1c403f4999 gallium/postprocess: Just to be safe, reference all buffers from outside
Even though it should be safe to use them for one frame, better be sure.
Suggested by Michael Dänzer.

NOTE: This is a candidate for the 8.0 stable branch.

Signed-off-by: Lauri Kasanen <cand@gmx.com>
(cherry picked from commit 81938d2137)
2012-02-02 17:09:29 -05:00
Lauri Kasanen
ad83ddc868 gallium/postprocess: Fix depth logic
This prevents a possible lapse of the depth buffer - the situation where
the app and pp have different depth buffers.

NOTE: This is a candidate for the 8.0 stable branch.

Signed-off-by: Lauri Kasanen <cand@gmx.com>
(cherry picked from commit c5976017e3)
2012-02-02 17:09:16 -05:00
Michel Dänzer
ac089040d7 gallium/postprocess: Proper reference counting of pp_jimenezmlaa depth buffer.
Fixes https://bugs.freedesktop.org/show_bug.cgi?id=40776

NOTE: This is a candidate for the stable branches.
(cherry picked from commit 7219af5ec1)
2012-02-02 17:08:06 -05:00
Paul Berry
0e08205421 i965/gen6: Fix segfault in transform feedback to DYNAMIC_DRAW buffers.
When storing data in a buffer of type DYNAMIC_DRAW, we don't create a
drm_intel_bo for it; instead we store the data in system memory and
defer allocation of the GPU buffer until it is needed.  Therefore, in
brw_update_sol_surface(), we can't just consult the "buffer" field of
the intel_buffer_object structure; we need to call
intel_bufferobj_buffer() to ensure that the deferred allocation
occurs.

This parallels a similar fix for gen7 (see commit ba6f4c9).

Fixes piglit test EXT_transform_feedback/buffer-usage on gen6.

This is a candidate for the 8.0 release branch.

Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 6bc08ee569)
2012-02-02 11:39:28 -07:00
Paul Berry
cc27a42b6f i965/vs: Fix bogus assertion in emit_block_move()
i965 processes assignments of whole structures using
vec4_visitor::emit_block_move, a recursive function which visits each
element of a structure or array (to arbitrary nesting depth) and
copies it from the source to the destination.  Then it increments the
source and destination register numbers so that further recursive
invocations will copy the rest of the structure.  In addition, it sets
the swizzle field for the source register to an appropriate value of
swizzle_for_size(...) for the size of each element being copied, so
that later optimization passes won't be fooled into thinking that
unused vector elements are live.

This all works fine.  However, emit_block_move also contains an
assertion to verify, before setting the swizzle field for the source
register, that the source register doesn't already contain a
nontrivial swizzle.  The intention is to make sure that the caller of
emit_block_move hasn't already done some swizzling of the data before
the call, which emit_block_move would then counteract when it
overwrites the swizzle field.  But the assertion is at the lowest
level of nesting of emit_block_move, which means that after the first
element is copied, instead of checking the swizzle field set by the
caller, it checks the swizzle field used when moving the previous
element.  That means that if the structure contains elements of
different vector sizes (which therefore require different swizzles),
the assertion will erroneously fire.

This patch moves the assertion from emit_block_move to the calling
function, vec4_visitor::visit(ir_assignment *).  Since the caller is
non-recursive, the assertion will only happen once, and won't be
fooled by emit_block_move's modification of the swizzle field.

This patch also reverts commit fe006a7 (i965/vs: Fix swizzle related
assertion), which attempted to fix the bug by making the assertion
more lenient, but only worked properly for structures, arrays, and
matrices in which each constituent vector is the same size.

This fixes the problem described in comment 9 of
https://bugs.freedesktop.org/show_bug.cgi?id=40865.  Unfortunately, it
doesn't fix the whole bug, since the test in question is also failing
due to lack of register spilling support in the VS.

Fixes piglit test vs-assign-varied-struct.  No piglit regressions on
Sandy Bridge.

This is a candidate for the 8.0 release branch.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=40865#c9
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit e2274aa739)
2012-02-02 11:37:21 -07:00
Eric Anholt
67937502f5 mesa: Fix display list handling for GL_EXT_framebuffer_multisample.
From the extension spec:

    Added to section 5.4, as part of the discussion of which commands
    are not compiled into display lists:

    "Certain commands, when called while compiling a display list, are
    not compiled into the display list but are executed immediately.
    These are: ..., RenderbufferStorageMultisampleEXT..."

Fixes piglit EXT_framebuffer_multisample/dlist.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 3d8c27f882)
2012-02-02 11:28:25 -07:00
Eric Anholt
1a77654e6a mesa: Fix display list handling for EXT_framebuffer_object.
Noticed when handling a similar problem in EXT_framebuffer_multisample.

From the EXT_framebuffer_object spec:

    Added to section 5.4, as part of the discussion of which commands
    are not compiled into display lists:

    "Certain commands, when called while compiling a display list, are
    not compiled into the display list but are executed immediately.
    These are: ..., GenFramebuffersEXT, BindFramebufferEXT,
    DeleteFramebuffersEXT, CheckFramebufferStatusEXT,
    GenRenderbuffersEXT, BindRenderbufferEXT, DeleteRenderbuffersEXT,
    RenderbufferStorageEXT, FramebufferTexture1DEXT,
    FramebufferTexture2DEXT, FramebufferTexture3DEXT,
    FramebufferRenderbufferEXT, GenerateMipmapEXT..."

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 25dd80555d)
2012-02-02 11:28:17 -07:00
Anuj Phogat
b38640082c mesa: fix maximum allowed proxy texture size condition
width, height parameter in glTexImage2D() includes: texture image
width + 2 * border (if any). So when doing the texture size check
in _mesa_test_proxy_teximage() width and height should not exceed
maximum supported size for target texture type.
i.e. 1 << (ctx->Const.MaxTextureLevels - 1)

Texture border is anyway stripped out before it is given to intel
or gallium drivers.

This patch fixes Intel oglconform test case: max_values
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=44970

Note: This is a candidate for mesa 8.0 branch.

Signed-off-by: Anuj Phogat <anuj.phogat@gmail.com>
Reviewed-by: Ian Romanick <idr@freedesktop.org>
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
(cherry picked from commit 15986d21eb)
2012-02-02 11:21:20 -07:00
Anuj Phogat
736f1e53e4 mesa: set clamp bit in glGetTexImage for GL_UNSIGNED_NORMALIZED
Color clamping should be enabled in glGetTexImage if texture dataType is
GL_UNSIGNED_NORMALIZED and format is GL_LUMINANCE or GL_LUMINANCE_ALPHA

Fixes 2 Intel oglconform test cases: pxconv-gettex and pxtrans-gettex
https://bugs.freedesktop.org/show_bug.cgi?id=40864

NOTE: This is a candidate for the 8.0 branch

Signed-off-by: Anuj Phogat <anuj.phogat@gmail.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 5665b5cc31)
2012-02-02 11:21:07 -07:00
Ian Romanick
d45d250ad1 mesa: Fix copy-and-paste error in _mesa_pack_rgba_span_float
GL_RG_INTEGER only has two components, not three.  I'll be surprised
if anyone ever tries to glReadPixels(..., GL_SHORT, GL_RG_INTEGER,
...).  This was found by inspection.

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 2e8f8cb383)
2012-02-02 10:17:49 -07:00
Ian Romanick
4635e26db2 mesa: Fix copy-and-paste bug in do_row_3D
Several of the half-float cases used 4 as the texel size when it
should have been some smaller value.

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=43324
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=43325
(cherry picked from commit 5c341b7df3)
2012-02-02 10:17:49 -07:00
Ian Romanick
65b9c1dee6 mesa: Convert colors if span ChanType and renderbuffer data type don't match
This is a partial revert of f9874fe.  It turns out that the types
don't always match.  Specifically, this can happen when doing
glCopyPixels from a float FBO to a RGBA8 FBO.

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=45429
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 660ed923de)
2012-02-02 10:17:49 -07:00
Ian Romanick
c34947dbb1 mesa: Set the gl_array_object::ARBsemantics flag at the right time
With 0963990 the flag was only set when Bind created the object.  In
all cases where ::ARBsemantics could be true, this path never
happened.  Instead, add a _Used flag to track whether a VAO has ever
been bound.  On the first Bind, set the _Used flag, and set the
ARBsemantics flag to the correct value.

NOTE: This is a candidate for release branches.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=45423
(cherry picked from commit e06b1c65bc)
2012-02-02 10:17:49 -07:00
Ian Romanick
9da7b58b39 mesa: Add unpack_uint_z_row support for floating-point depth buffers
This is a hack, and it will result in incorrect rendering.  However,
it does eliminate spurious warnings in several piglit CopyPixels tests
that involve floating-point depth buffers.

The real solution is to add a zf field to SWspan to store float Z
values.  When a float depth buffer is involved, swrast should also
populate the zf field.  I'll consider this post-8.0 work.

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit af1477b088)
2012-02-02 10:17:48 -07:00
José Fonseca
5ac4c8cf53 mapi/glapi: Never use a generic no-op entry-point on Windows.
When GLAPIENTRY is __stdcall (ie Windows), the stack is popped by the
callee making the number/type of arguments significant, therefore
using a generic no-op causes stack corruption for many entry-points.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
2012-02-02 15:43:10 +00:00
Ian Romanick
7f5d3f7ed2 meta: Fallback for glBlitFramebuffer from a multisample surface
NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=44818
(cherry picked from commit b48d4b64e9)
2012-02-01 09:23:32 -07:00
Eric Anholt
442dc31fa7 intel: Fix accum buffer mapping since the swrast rework.
A pure swrast-allocated buffer gets an irb of NULL, so we segfaulted
in the clear-accum test.  Just look at the swrast renderbuffer pointer
for handling swrast rbs.
(cherry picked from commit 42e9936ce6)

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=45428
2012-01-31 09:39:10 -07:00
Brian Paul
3ad7f44926 osmesa: set RefCount = 1 in new_osmesa_renderbuffer()
This was lost during the renderbuffer overhaul work.  Fixes a failed
refcount assertion.
(cherry picked from commit 3fc6e4e025)
2012-01-31 07:51:04 -07:00
Vinson Lee
9f0088e906 osmesa: Fix osmesa_context.DataType type.
Fixes these GCC warnings.
osmesa.c: In function ‘osmesa_renderbuffer_storage’:
osmesa.c:417: warning: comparison is always false due to limited range of data type
osmesa.c:423: warning: comparison is always false due to limited range of data type
osmesa.c:431: warning: comparison is always false due to limited range of data type
osmesa.c:437: warning: comparison is always false due to limited range of data type
osmesa.c:447: warning: comparison is always false due to limited range of data type
osmesa.c:453: warning: comparison is always false due to limited range of data type
osmesa.c:463: warning: comparison is always false due to limited range of data type
osmesa.c:466: warning: comparison is always false due to limited range of data type
osmesa.c:476: warning: comparison is always false due to limited range of data type
osmesa.c:479: warning: comparison is always false due to limited range of data type

Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 136791ebc1)
2012-01-31 07:50:51 -07:00
Brian Paul
f893fde63b vega: memset data array to zero to silence uninitialized var warnings
(cherry picked from commit 6386f80dbd)
2012-01-31 07:04:25 -07:00
Brian Paul
6bb4823f7d softpipe: move var initialization to silence warning
(cherry picked from commit 3e01c3f3ba)
2012-01-31 06:55:25 -07:00
Vinson Lee
ddd2503750 softpipe: Silence unused variable warning on non-LLVM builds.
Fix this GCC warning with non-LLVM builds.
sp_screen.c: In function ‘softpipe_get_shader_param’:
sp_screen.c:141:28: warning: unused variable ‘sp_screen’ [-Wunused-variable]

Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 8e543cc098)
2012-01-31 06:55:07 -07:00
Benjamin Franzke
5f60d134e6 st/dri: Support 24bit formats in dri2_allocate_buffer
Prior commit 576161289d,
the parameter format was bpp, thus both 24bit and 32bit formats were
requested with format set to 32. Handle 24bit seperately now.

Fixes RGBX formats in wayland platform for egl_dri2 (EGL_ALPHA_SIZE=0).

Note: This is a candidate for the 8.0 branch.
(cherry picked from commit c72d7df168)
2012-01-30 17:22:44 +01:00
Christian König
caebd7929d st/xvmc: remove xorg-server dependency
Fixing a circular build dependency.

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Christian König <deathsimple@vodafone.de>
(cherry picked from commit c2e2b58a58)
2012-01-29 00:06:59 +01:00
Ian Romanick
0dddf4c575 mesa: Bump version number to 8.0-rc2
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
2012-01-27 18:02:43 -08:00
Kenneth Graunke
4aa158d09a i965/vs: Use the sampler for VS pull constant loading on Ivybridge.
Substantially increases performance in GLBenchmark PRO:
- 320x240   => 3.28x
- 1920x1080 => 1.47x
- 2560x1440 => 1.27x

The LD message ignores the sampler unit index and SAMPLER_STATE pointer,
instead relying on hard-wired default state.  Thus, there's no need to
worry about running out of sampler units or providing SAMPLER_STATE;
this small patch should be all that's required.

NOTE: This is a candidate for release branches.
      (It requires the preceding commit to compile.)

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 259b65e2e7)
2012-01-27 18:02:43 -08:00
Kenneth Graunke
38b76cf831 i965: Expose brw_set_sampler_message for use outside brw_eu_emit.c.
brw_SAMPLE is full of complex workarounds for original Broadwater
hardware, and I'd rather avoid all that for my next Ivybridge patch.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 5f4575d42f)
2012-01-27 18:02:43 -08:00
Paul Berry
73ff415b59 i965: Set pitch of pull constant buffers to 16.
We always access pull constant buffers using the message types "OWord
Block Read" or "OWord Dual Block Read".  According to the Sandy Bridge
PRM, Vol 4 Part 1, pages 214 and 218, when using these messages:

    "the surface pitch is ignored, the surface is treated as a
    1-dimensional surface.  An element size (pitch) of 16 bytes is
    used to determine the size of the buffer for out-of-bounds
    checking if using the surface state model."

Previously we were setting the pitch for pull constant buffers to the
size of the whole constant buffer--this made no sense and would have
led to incorrect behavior if it were not for the fact that the pitch
is ignored.

For clarity, this patch sets the pitch for pull constant buffers to 16
bytes, consistent with the hardware's behavior.

v2: Clarify the meaning of the ignored values by writing them as (16 - 1).

Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit fcd5af4a91)
2012-01-27 18:02:42 -08:00
Ian Romanick
58e0c4f872 mesa: Don't resurrect deleted ARB VAOs in glPopClientAttrib
When ARB VAOs are used, glPopClientAttrib does not resurrect a deleted
VAO or VBO.  This difference between the two spec is, unfortunately,
not very well spelled out in the specs.

Fixes oglc vao(advanced.pushPop.deleteVAO) and
vao(advanced.pushPop.deleteVBO) tests.

NOTE: This is a candidate for release branches.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 34c353ce46)
2012-01-27 18:02:42 -08:00
Ian Romanick
eea63b7621 mesa: Rename gl_array_object::VBOonly to ::ARBsemantics
There are more differences between Apple and ARB than just requiring
that all arrays be stored in VBOs.  Additional uses will be added in
following commits.

Also, set the flag at Bind time instead of Gen time.  The ARB_vao spec
specifies that behavior.

NOTE: This is a candidate for release branches.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 0963990153)
2012-01-27 18:02:42 -08:00
Ian Romanick
85a52bf7b5 swrast: Use fixed-function processing instead _TexEnvProgram for DrawPixels
This is a hack to work around drivers such as i965 that:

    - Set _MaintainTexEnvProgram to generate GLSL IR for
      fixed-function fragment processing.
    - Don't call _mesa_ir_link_shader to generate Mesa IR from the
      GLSL IR.
    - May use swrast to handle glDrawPixels.

Since _mesa_ir_link_shader is never called, there is no Mesa IR to
execute.  Instead do regular fixed-function processing.

Even on platforms that don't need this, the software fixed-function
code is much faster than the software shader code.

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=44749
(cherry picked from commit 9be3be3c66)
2012-01-27 18:02:42 -08:00
Ian Romanick
05b7f13af1 mesa: Make sure _TexEnvProgram points at the current ff fragment program
At least one place, the _mesa_need_secondary_color function in
state.h, uses this to make decisions.  The next patch in this series
will add another dependency.  Ideally, this field would go away and be
replace by a flag or something.

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 34db7a8c1e)
2012-01-27 18:02:42 -08:00
Brian Paul
93d5799e8b swrast: fix uninitialized variable warning
(cherry picked from commit ba151a333b)
2012-01-27 18:02:42 -08:00
Brian Paul
4f8b00d8ef swrast: make rowStride variable signed in put_z32_values()
As with commit aed5c8299f
(cherry picked from commit cf386f0a2b)
2012-01-27 18:02:42 -08:00
Brian Paul
b5a5a1c615 intel: use swrast code to map/unmap renderbuffers for swrast rendering
(cherry picked from commit 89bb19adb0)
2012-01-27 18:02:42 -08:00
Brian Paul
4d72e190c0 mesa: update comments, fix whitespace in dd.h
(cherry picked from commit 4baf90353d)
2012-01-27 18:02:42 -08:00
Brian Paul
e442906c87 swrast: remove unused StoreTexel code
No longer needed since we do all rendering to texture with the buffer
mapping and pixel packing functions.
(cherry picked from commit 9403cc3aba)
2012-01-27 18:02:42 -08:00
Brian Paul
12370e050a swrast: use Map/UnmapTextureImage() in framebuffer map/unmap code
When we're actually rendering into a texture, map the texture image
instead of the corresponding renderbuffer.  Before, we just copied
a pointer from the texture image to the renderbuffer.  This change
will make the code usable by hardware drivers.
(cherry picked from commit 1caf698191)
2012-01-27 18:02:41 -08:00
Brian Paul
5163fed919 mesa: remove ctx->Driver.Map/UnmapTexture() hooks
No longer used anywhere.
(cherry picked from commit 56d83ac4bf)
2012-01-27 18:02:41 -08:00
Brian Paul
acfbf24335 nouveau: stop calling ctx->Driver.Map/UnmapTexture()
And remove unused nouveau_texture_map/unmap()
(cherry picked from commit 8b8a54afd9)
2012-01-27 18:02:41 -08:00
Brian Paul
765892865f tnl: remove ctx->Driver.Map/UnmapTexture() calls
ctx->Driver.MapTexture() always points to _swrast_map_texture().
We're already reaching into swrast from t_vb_program.c anyway.
This will let us remove the ctx->Driver.Map/UnmapTexture() functions.
(cherry picked from commit 4bbab2275f)
2012-01-27 18:02:41 -08:00
Brian Paul
9daa974d4d swrast: move some renderbuffer functions to s_renderbuffer.c
(cherry picked from commit bde356a158)
2012-01-27 18:02:41 -08:00
Brian Paul
05adf4d6e4 intel: remove intel_span_supports_format()
It always returned True.
(cherry picked from commit 1839a7fc9f)
2012-01-27 18:02:41 -08:00
Brian Paul
eca54b1eb4 swrast: remove a few extra _mesa_get_format_bytes() calls
(cherry picked from commit 6c1e27ba21)
2012-01-27 18:02:41 -08:00
Brian Paul
8e5cf4cb45 mesa: update comments for gl_renderbuffer
(cherry picked from commit 1da7d6c919)
2012-01-27 18:02:41 -08:00
Brian Paul
1fd179b984 mesa/swrast/drivers: remove obsolete gl_renderbuffer fields
This removes the last of the legacy fields from gl_renderbuffer.
(cherry picked from commit 1888dd52a3)
2012-01-27 18:02:41 -08:00
Brian Paul
843194d89f dri/swrast: use swrast_renderbuffer type
(cherry picked from commit becbb64313)
2012-01-27 18:02:41 -08:00
Brian Paul
7c95a4ec22 swrast: use swrast_renderbuffer instead of gl_renderbuffer
(cherry picked from commit 0c1862851f)
2012-01-27 18:02:41 -08:00
Brian Paul
9567178ca2 osmesa: use swrast_renderbuffer
(cherry picked from commit 797c18be1f)
2012-01-27 18:02:40 -08:00
Brian Paul
83602e8342 radeon: derive radeon_renderbuffer from swrast_renderbuffer
(cherry picked from commit c080202db5)
2012-01-27 18:02:40 -08:00
Brian Paul
b31bfae0db intel: derive intel_renderbuffer from swrast_renderbuffer
Drivers that rely on swrast need to do this, as with swrast_texture_image.
(cherry picked from commit 9f8ed9d662)
2012-01-27 18:02:40 -08:00
Brian Paul
49ab5ad142 xlib: derive xmesa_renderbuffer from swrast_renderbuffer
(cherry picked from commit d16e71eeb4)
2012-01-27 18:02:40 -08:00
Brian Paul
5916b15247 swrast: allocate swrast_renderbuffers instead of gl_renderbuffers
(cherry picked from commit f2479530b8)
2012-01-27 18:02:40 -08:00
Brian Paul
a36cc604cf swrast: new swrast_renderbuffer type
This will let us move the swrast-specific fields out of gl_renderbuffer.
(cherry picked from commit 34988272d9)
2012-01-27 18:02:40 -08:00
Brian Paul
73ed8cec4a intel: use intel_rb_format() to get renderbuffer format
This will make future changes cleaner and less invasive.
(cherry picked from commit 924de7dc96)
2012-01-27 18:02:40 -08:00
Brian Paul
c6e56a69d8 dri/swrast: rename swrast_renderbuffer to dri_swrast_renderbuffer
To prevent name collision with future swrast_renderbuffer in the swrast
module.
(cherry picked from commit 1048d55d5f)
2012-01-27 18:02:40 -08:00
Brian Paul
d980b8350f swrast: use stencil packing function in s_stencil.c
(cherry picked from commit c45771905f)
2012-01-27 18:02:40 -08:00
Brian Paul
143999bdd8 swrast: use color packing functions in s_span.c
(cherry picked from commit 881ef2a9db)
2012-01-27 18:02:39 -08:00
Brian Paul
ae3bd76e31 swrast: remove s_spantemp.h
(cherry picked from commit 8696a52102)
2012-01-27 18:02:39 -08:00
Brian Paul
86fc256131 xlib: remove xm_span.c and related code
(cherry picked from commit b0f0d7a811)
2012-01-27 18:02:39 -08:00
Brian Paul
5cb5aa4a84 mesa: remove gl_renderbuffer::Wrapped
There's no such thing as renderbuffer wrappers anymore.
(cherry picked from commit 59a5b5a193)
2012-01-27 18:02:39 -08:00
Brian Paul
f0ad188c9f swrast: rewrite, simplify the the render-to-texture code
(cherry picked from commit ab331140c6)
2012-01-27 18:02:39 -08:00
Brian Paul
8786555199 mesa: rename gl_renderbuffer::Data to Buffer
To better indicate that this pointer to the malloc'd memory.
(cherry picked from commit 7a36345f70)
2012-01-27 18:02:39 -08:00
Brian Paul
f29212ecba mesa: move freeing of software renderbuffers into swrast
(cherry picked from commit f6a3979a04)
2012-01-27 18:02:39 -08:00
Brian Paul
8e5e735424 mesa: remove gl_renderbuffer::DataType
(cherry picked from commit f9874feef4)
2012-01-27 18:02:39 -08:00
Brian Paul
f5367c4503 mesa: remove gl_renderbuffer:RowStride field
(cherry picked from commit 1e1b5cb01a)
2012-01-27 18:02:39 -08:00
Brian Paul
d60bcbe31d mesa: finally, remove the GetRow/PutRow/etc functions
(cherry picked from commit 82846fea4d)
2012-01-27 18:02:38 -08:00
Brian Paul
64deeddce8 dri: remove all the obsolete spantmp files
(cherry picked from commit 304f7a1327)
2012-01-27 18:02:38 -08:00
Brian Paul
514fcec7da radeon: remove obsolete GetRow/PutRow code
(cherry picked from commit a4c6dedb27)
2012-01-27 18:02:38 -08:00
Brian Paul
9b5e402837 nouveau: remove obsolete GetRow/PutRow code
(cherry picked from commit f892debdc2)
2012-01-27 18:02:38 -08:00
Brian Paul
57df49908e intel: remove most of the span Get/PutRow code
(cherry picked from commit 41869c4942)
2012-01-27 18:02:38 -08:00
Brian Paul
38e4496ac3 dri/swrast: remove obsolete GetRow/PutRow code
This is a squash of:

    dri/swrast: remove obsolete GetRow/PutRow code
    (cherry picked from commit cb5fa9ea62)

and

    dri/swrast: remove obsolete swrast_span.c file from source list
    (cherry picked from commit a9bf149e7f)

    Conflicts:

	src/mesa/drivers/dri/swrast/Makefile.sources
2012-01-27 18:02:14 -08:00
Brian Paul
39d978a60c osmesa: remove obsolete GetRow/PutRow code
(cherry picked from commit 2873555a76)
2012-01-27 17:43:22 -08:00
Brian Paul
e4b42d618c xlib: remove obsolete GetRow/PutRow code
(cherry picked from commit 2e80c7e5bf)
2012-01-27 17:43:22 -08:00
Brian Paul
d85b4bae0e mesa: remove obsolete PutRow, etc assignments
(cherry picked from commit 0d2f0c8bb8)
2012-01-27 17:43:22 -08:00
Brian Paul
3dd8937492 swrast: remove Get/PutRow()-related code
(cherry picked from commit d65bbfa947)
2012-01-27 17:43:22 -08:00
Brian Paul
1f0cc3faf2 st/mesa: remove gl_renderbuffer::GetPointer stuff
(cherry picked from commit a4a566a610)
2012-01-27 17:43:22 -08:00
Brian Paul
f00ba21915 swrast: stop using Put/GetRow/Values() in swrast code
All color buffer rendering is now done by accessing mapped renderbuffer
memory.  We're now able to get rid of all the GetRow/PutRow stuff.
(cherry picked from commit 0ff817f200)
2012-01-27 17:43:22 -08:00
Brian Paul
c8562a6ed3 swrast: use gl_renderbuffer::StrideInBytes in depth/stencil code
(cherry picked from commit b766d4bb43)
2012-01-27 17:43:22 -08:00
Brian Paul
48173180b5 mesa: use gl_renderbuffer::Map for all depth/stencil accesses
Instead of using the obsolete gl_renderbuffer::Data field.
Color buffer are still accessed through GetRow/PutRow().
(cherry picked from commit 7d1ddec921)
2012-01-27 17:43:22 -08:00
Brian Paul
349d5a8a38 intel: make intel_renderbuffer_map/unmap() static
(cherry picked from commit 14da67d9b9)
2012-01-27 17:43:21 -08:00
Brian Paul
fee5f10241 mesa: add new gl_renderbuffer fields
These are temporary, actually, but they'll make follow-on work easier to
implement in a step-by-step manner.  Eventually the Map and RowStrideBytes
fields will go into a new swrast_renderbuffer type, but adding that type
now would involve touching a _lot_ of code that'll eventually be removed.

The fields marked as obsolete will go away completely at some point.
(cherry picked from commit 827c1d66f6)
2012-01-27 17:43:21 -08:00
Brian Paul
83eb5e4372 swrast: flush pending rendering before unmapping buffers
(cherry picked from commit fc9f74839d)
2012-01-27 17:43:21 -08:00
Brian Paul
5f6676b0ee swrast: new assertions in _swrast_pixel_address()
(cherry picked from commit 33257803d9)
2012-01-27 17:43:21 -08:00
Brian Paul
7c6f54811c swrast: use _swrast_pixel_address() in more places
(cherry picked from commit e34a54ff45)
2012-01-27 17:43:21 -08:00
Brian Paul
a42dcf1614 swrast: s/Data/Map/ in swrast_texture_image
To indicate that it points to mapped texture memory.
(cherry picked from commit bd3c10c0f0)
2012-01-27 17:43:21 -08:00
Brian Paul
573e1478eb swrast: remove gl_renderbuffer::DataType check in DrawPixels()
The field will be going away so update this code.
(cherry picked from commit ecb8594c18)
2012-01-27 17:43:21 -08:00
Brian Paul
490057ae81 swrast: remove gl_renderbuffer::DataType assertions
This field will go away, so remove some uses of it.
(cherry picked from commit 7726be1c1b)
2012-01-27 17:43:21 -08:00
Brian Paul
d2d1419ba0 st/mesa: remove gl_renderbuffer:DataType assignments
That field is only used by swrast code so there's no reason to mess
with it in the gallium state tracker.

This also lets us remove the unused st_format_data() type function and
related code.
(cherry picked from commit ca6d86d26b)
2012-01-27 17:43:21 -08:00
Brian Paul
0941b808a6 swrast: make _swrast_get_values(), _swrast_get_row() static
They were only called from in s_span.c
(cherry picked from commit ff57b0f037)
2012-01-27 17:43:21 -08:00
Brian Paul
ac6074de17 swrast: remove dstType param from _swrast_read_rgba_span()
It was always GL_FLOAT.
(cherry picked from commit 267fb17884)
2012-01-27 17:43:21 -08:00
Brian Paul
289a7e74b3 swrast: remove unused _swrast_put_row()
(cherry picked from commit 64be85540f)
2012-01-27 17:43:21 -08:00
Chad Versace
889f756c28 swrast: Fix unsigned promotion in pointer arithmetic
When rowstride was negatie, unsigned promotion caused a segfault here:

299│    if (rb->Format == MESA_FORMAT_S8) {
300│       const GLuint rowStride = rb->RowStride;
301│       for (i = 0; i < count; i++) {
302│          if (x[i] >= 0 && y[i] >= 0 && x[i] < w && y[i] < h) {
303├>            stencil[i] = *(map + y[i] * rowStride + x[i]);
304│          }
305│       }
306│    }

Fixes segfault in oglconform
separatestencil-neu(NonPolygon.BothFacesBitmapCoreAPI),
though test still fails.

Note: This is a candidate for the stable branches.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=43327
Reviewed-by: Brian Paul <brianp@vmware.com>
Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit aed5c8299f)
2012-01-27 17:43:21 -08:00
Eric Anholt
e10c47adff mesa: Fix CopyTex{Sub,}Image error checks for integer vs non-integer.
Fixes Intel oglconform negative.typeFormatMismatch.copyteximage.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 422b18794e)
2012-01-27 11:29:27 -08:00
Eric Anholt
dcaf26ac91 mesa: Add missing integer R/RG cases to _mesa_is_color_format().
This is part of fixing Intel oglconform
negative.typeFormatMismatch.copyteximage.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit ee9804af14)
2012-01-27 11:29:27 -08:00
Eric Anholt
9a33b5a269 meta: Detect currently-unsupported integer CopyTexSubImage and complain.
This code is unprepared for handling integer (particularly, the
baseFormat of the TexFormat comes out as GL_RGBA, not GL_RGBA_INTEGER,
so the direct call of Driver.ReadPixels crashes due to the int vs
non-int error checking not having happened).  I'm frankly tempted to
convert this code to MapRenderbuffer/MapTexImage rather than doing it
as meta ops, now that we have that support.

Improves the remaining crash in Intel oglconform for int-textures to
just a rendering failure.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 6bf0f6ae86)
2012-01-27 11:29:27 -08:00
Eric Anholt
61be81ae5c mesa: Add the remaining from/to types for GL_EXT_texture_integer (and R/RG).
This aborts and crashes in intel oglconform's int-textures into being
just rendering failures.  Clamping isn't handled yet.

v2: Add missing "break".
v3: Drop the int/uint distinction, since they don't need different clamping.
NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Brian Paul <brianp@vmware.com> (v2)
(cherry picked from commit d6c58545a1)
2012-01-27 11:29:27 -08:00
Eric Anholt
46c146116f mesa: Add support for glGetTexImage on GL_TEXTURE_1D_ARRAY
Similarly to how we handle this in texstore, we have to remap height
to depth so that we MapTextureImage each image layer individually.

Fixes part of Intel oglconform's int-textures advanced.fbo.rtt

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 08acd4bd61)
2012-01-27 11:29:27 -08:00
Eric Anholt
e695c7e5b8 mesa: Add support for glGetTexImage() from integer textures.
This is a step toward fixing Intel oglconform's
int-textures advanced.fbo.rtt.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 9f1e64d085)
2012-01-27 11:29:27 -08:00
Eric Anholt
bc973082fc i965/gen6+: Work around GPU hangs with logic ops on integer textures.
This doesn't result in correct rendering -- GL requires that logic ops
work, while the hardware specs say it doesn't do them.  I'm not sure
how we would want to handle this.

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit f6e82cd2a1)
2012-01-27 11:29:26 -08:00
Eric Anholt
f92503e171 i965/gen6+: Disable blending, alpha test, and dither on integer FBOs.
Fixes GPU hangs and some rendering failures in piglit
EXT_texture_integer/fbo-blending

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit 04b4880d7c)
2012-01-27 11:29:26 -08:00
Jeremy Huddleston
e406659b8b 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 09:05:22 -08:00
Ander Conselvan de Oliveira
19f88670b5 gbm: fix copy & paste error in gbm_bo_get_handle documentation
(cherry picked from commit 33f8a3cfbe)
2012-01-25 12:24:51 -05:00
Rob Bradford
dcb1a3bc27 gbm: Add documentation for the public facing API
(cherry picked from commit baab68e1a6)
2012-01-25 12:24:48 -05:00
Scott Moreau
b064078581 Complete ARGB8888 naming convention format renames missed
(cherry picked from commit e0897009f8)
2012-01-25 11:25:31 -05:00
Benjamin Franzke
252a1d8635 st/mesa: Fix recurring surfaceless contexts
A current incomplete framebuffer was incorrectly used as a
st_framebuffer. When accessing st_framebuffer childs bad things happen:
e.g. st_framebuffer::iface was used to check whether its an incomplete
fb, instead we need to compare st_framebuffer::Base against
mesa_get_incomplete_framebuffer.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=44919
Note: This is a candidate for the 8.0 branch.

Reviewed-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit 36fb83e4a8)
2012-01-25 10:13:19 +01:00
Kenneth Graunke
8e42dadf70 i965: Fix border color on Sandybridge and Ivybridge.
While reading through the simulator, I found some interesting code that
looks like it checks the sampler default color pointer against the bound
set in STATE_BASE_ADDRESS.  On failure, it appears to program it to the
base address itself.

So I decided to try programming a legitimate bound, and lo and behold,
border color worked.

+92 piglits on Sandybridge.  Also fixes Lightsmark on Ivybridge.

NOTE: This is a candidate for stable release branches.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=28924
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=38868
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit c25e5300cb)
2012-01-24 15:39:47 -08:00
Vinson Lee
0b6003af71 glsl: Fix 'control reaches end of non-void function' warning.
Fix this GCC warning on non-debug builds.
glsl_types.cpp: In member function 'gl_texture_index
glsl_type::sampler_index() const':
glsl_types.cpp:157: warning: control reaches end of non-void function

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit e9bcf4d56b)
2012-01-24 15:39:47 -08:00
Chad Versace
d8f66afa5d mesa: Loosen glBlitFramebuffer restrictions on depthstencil buffers (v2)
This loosens the format validation in glBlitFramebuffer. When blitting
depth bits, don't require an exact match between the depth formats; only
require that the two formats have the same number of depth bits and the
same depth datatype (float vs uint). Ditto for stencil.

Between S8_Z24 buffers, the EXT_framebuffer_blit spec allows
glBlitFramebuffer to blit the depth and stencil bits separately. So I see
no reason to prevent blitting the depth bits between X8_Z24 and S8_Z24 or
the stencil bits between S8 and S8_Z24. However, we of course don't want
to allow blitting from Z32 to Z32_FLOAT.

Fixes Piglit fbo/fbo-blit-d24s8 on Intel drivers with separate stencil
enabled.

The problem was that, on Intel drivers with separate stencil, the default
framebuffer has separate depth and stencil buffers with formats X8_Z24 and
S8. The test attempts to blit the depth bits from a S8_Z24 buffer into the
default framebuffer.

v2: Check that depth datatypes match.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=44665
Note: This is a candidate for the 8.0 branch.
Reported-by: Xunx Fang <xunx.fang@intel.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit f74d8aacbf)
2012-01-24 15:39:47 -08:00
Paul Berry
09b6308a2d glsl: Fix isinf() for non-C99-compliant compilers.
Commit ede60bc467 (glsl: Add isinf() and
isnan() builtins) uses "+INF" in the .ir file to represent infinity.
This worked on C99-compliant compilers, since the s-expression reader
uses strtod() to read numbers, and C99 requires strtod() to understand
"+INF".  However, it didn't work on non-C99-compliant compilers such
as MSVC.

This patch modifies the s-expression reader to explicitly check for
"+INF" rather than relying on strtod() to support it.

This is a candidate for the 8.0 branch.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=44767
Tested-by: Morgan Armand <morgan.devel@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit 4f82fed493)
2012-01-24 15:39:47 -08:00
Ian Romanick
30e9bfd84a mesa: Set default access flags based on the run-time API
The default access flags for OpenGL ES (via GL_OES_map_buffer) and
desktop OpenGL are different.  The code previously tried to handle
this, but the decision was made at compile time.  Since the same
driver binary can be used for both OpenGL ES and desktop OpenGL, the
decision must be made at run-time.

This should fix bug #44433.  It appears that the test case does
various map and unmap operations and inspects the state of the buffer
object around each.  When it sees that GL_BUFFER_ACCESS does not match
its expectations, it fails.

NOTE: This is a candidate for release branches.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=44433
(cherry picked from commit f0ea46790f)
2012-01-24 15:39:47 -08:00
Kenneth Graunke
aa00ccdc02 i965: Fix disassembly of data port writes on Ivybridge.
msg_type moved by a bit, so the message type was being disassembled
incorrectly.  In particular, render target writes were showing up as
"OWORD block write".

NOTE: This is a candidate for stable release branches.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit dcdfd1905c)
2012-01-24 15:39:47 -08:00
Kenneth Graunke
a62a4f77e8 i965: Fix disassembly of sampler messages on Ivybridge.
Compared to sampler_gen5, simd_mode shifted by a bit and msg_type grew
by a bit.  So we were printing slightly incorrect numbers.

NOTE: This is a candidate for stable release branches.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit a608be5d33)
2012-01-24 15:39:47 -08:00
Kenneth Graunke
2fae4d26f3 i965/vs: Take attributes into account when deciding urb_entry_size.
Both the VF and VS share space in the URB.  First, the VF stores
attributes (shader inputs) there.  The VS then reads the attributes,
executes, and reuses the space to store varyings (shader outputs).

Thus, we need to calculate the amount of URB space necessary for inputs,
outputs, and pick whichever is greater.

The old VS backend correctly did this (brw_vs_emit.c:408), but the new
VS backend only considered outputs.

Fixes vertex scrambling in GLBenchmark PRO on Ivybridge.

NOTE: This is a candidate for the 8.0 branch.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=41318
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 2e712e41db)
2012-01-24 15:39:47 -08:00
Ian Romanick
759da0cb56 intel: Set depth to 6 for cubemaps
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=41216
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=43212
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=43250
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Cc: Jin Yang <jin.a.yang@intel.com>
(cherry picked from commit 8a47242755)
2012-01-24 15:39:47 -08:00
Eric Anholt
4dbc544bb1 i965: Add support for Z16 depth formats.
This is a squash of the following two commits.  The first caused a
regression on SNB systems without a HiZ capable DDX, and the second
fixes that regression.

    i965: Add support for Z16 depth formats.

    v2: Don't flag the format as being HiZ ready (there's DRI2 handshake
        pain to go through).

    Fixes piglit gl-3.0-required-sized-texture-formats

    NOTE: This is a candidate for the 8.0 branch.

    Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
    Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
    (cherry picked from commit 2f868f1ddd)

    intel/gen6: Some framebuffers having separate depthstencil should be unsupported

    When the framebuffer has separate depth and stencil buffers, and HiZ is
    not enabled on the depth buffer, mark the framebuffer as unsupported. This
    happens when trying to create a framebuffer with Z16/S8 because we haven't
    enabled HiZ on Z16 yet.

    Fixes gles2conform test stencil8.

    Note: This is a candiate for the 8.0 branch.
    Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=44948
    Reviewed-and-tested-by: Ian Romanick <ian.d.romanick@intel.com>
    Reviewed--by: Eric Anholt <eric@anholt.net>
    Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
    (cherry picked from commit ba5252e590)
2012-01-24 15:39:47 -08:00
Eric Anholt
c256fd094b i965/gen7: Set up surface horizontal alignment field.
This is required for Z16 support for texturing, which is the first
thing to have a horizontal alignment of 8.  Renderbuffers don't need
it, since they're always set up as the only mip level, but do it for
completeness anyway.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit f0d5c92a4c)
2012-01-24 15:39:47 -08:00
Eric Anholt
f0662ee609 i965/gen7: Remove stale comment.
This field is actually set up above.

NOTE: This is a candidate for the 8.0 branch, to avoid conflicts.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit fc767ff590)
2012-01-24 15:39:47 -08:00
Eric Anholt
99c0aeb3db glsl: Fix leak of linked uniform names at relink/free of the shader_program.
NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit aad3a46ff4)
2012-01-24 15:39:47 -08:00
Eric Anholt
7861ccbe23 glsl: Fix leak of LinkedTransformFeedback.Varyings.
I copy-and-pasted the thing I was allocating for as the context, so
the first time it would be NULL (root of a ralloc context) and they'd
chain off each other from then on.

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit 5a0f395bcf)
2012-01-24 15:39:47 -08:00
Eric Anholt
bffefc3d27 mesa: Fix leak of uniform storage records on shader program link/free.
NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit 0f68d88034)
2012-01-24 15:39:47 -08:00
Eric Anholt
531948947e i965: Fix leak of the program cache BO on context destroy.
NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit cbd464a117)
2012-01-24 15:39:46 -08:00
Eric Anholt
decf80c621 i965/vs: Fix leak of an empty hash_table structure per compile.
This statement got duplicated above, probably in a rebase resolution,
so we never freed the extra one.

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit 7f278e15ad)
2012-01-24 15:39:46 -08:00
Eric Anholt
812a8eade5 i965: Fix refcount leak of the gl_program structure.
Fixes a leak of almost 200kb on a minimal shader_runner program
(algebraic-add-add-1).

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit b2be486962)
2012-01-24 15:39:46 -08:00
Eric Anholt
230f6e7ddc mesa: Make the register allocator allocation take a ralloc context.
This fixes a memory leak on i965 context destruction.

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit b972744c78)
2012-01-24 15:39:46 -08:00
Brian Paul
655b36d1d5 mesa: use GL_MAP_INVALIDATE_RANGE_BIT in glTexImage paths
Update the dd.h docs to indicate that GL_MAP_INVALIDATE_RANGE_BIT
can be used with GL_MAP_WRITE_BIT when mapping renderbuffers and
texture images.

Pass the flag when mapping texture images for glTexImage, glTexSubImage,
etc.  It's up to drivers whether to actually make use of the flag.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 64fdfefb9d)
2012-01-24 15:39:46 -08:00
Brian Paul
7a11d201e7 mesa: try RGBA_FLOAT16 before RGBA_FLOAT32 when choosing A,L,LA,I formats
To try to use less tex memory and maybe get better performance.
Spotted by Roland Scheidegger.

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

Reviewed-by: José Fonseca <jfonseca@vmware.com>
(cherry picked from commit 1d7048f12e)
2012-01-24 15:39:46 -08:00
Brian Paul
dd73100c24 mesa: fix tex format selection for GL_R32F and other R/G float formats
The i965 driver advertises GL_ARB_texture_float and GL_ARB_texture_rg
support but the ctx->TextureFormatSupported[] table entries for
MESA_FORMAT_R_FLOAT32 and MESA_FORMAT_RGBA_FLOAT32 are false on gen 4
hardware.  So the case for GL_R32F would fail and we'd print an
implementation error.

This patch adds more Mesa tex format options for GL_R32F and other R/G
formats so we fall back to 16-bit formats when 32-bit formats aren't
available.

Eric made the same fix in commit 6216a5b4 for the non R/G formats.

v2: try 16-bit formats before 32-bit formats and try RG formats before
RGBA where possible.

This should fix https://bugs.freedesktop.org/show_bug.cgi?id=44039

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

Reviewed-by: José Fonseca <jfonseca@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
(cherry picked from commit 7628696004)
2012-01-24 15:39:46 -08:00
Kenneth Graunke
5f25e0a39c i965: Bump Ivybridge's fake MRF range to g112-127 instead of g111-126.
When I originally implemented the hack to use GRFs 111+ as fake MRFs, I
did so purely to avoid rewriting all the code that dealt with MRFs.
However, it turns out that a similar hack is actually required.

Newly discovered language in the BSpec indicates that SEND instructions
with EOT set "should" use g112-g127 as their source registers.  Based on
assertions in the simulator, this is actually a requirement on certain
platforms.

Since we're faking MRFs already, we may as well use the officially
sanctioned range.  My guess is that we avoided this issue because we
seldom use m0: URB writes in the new VS backend start at m1, and RT
writes in the new FS backend start at m2.

NOTE: This is a candidate for stable release branches.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 5acc7f38d4)
2012-01-24 15:39:46 -08:00
Anuj Phogat
f48e6748a7 intel: Return if pointer to intel_context is null
It is better to test if(intel == NULL) and simply return in that case.

Signed-off-by: Anuj Phogat <anuj.phogat@gmail.com>
Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit ce1c949b16)
2012-01-24 15:39:46 -08:00
Eric Anholt
68a9cd6fa1 intel: Fix warnings of undefined ffs().
For some reason these started showing up with the automake conversion.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit ccf0d31a21)
2012-01-24 15:39:46 -08:00
Chad Versace
907495465a i965: Comment gen6_hiz_get_framebuffer_enum()
Make the comments precise. Explain why each branch is needed and correct.
Document the potential pitfall in the true-branch.

Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit e13c99a004)
2012-01-24 15:39:46 -08:00
Christoph Bumiller
daf5ee5ef1 nvc0: fix some limit cap values
NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit ab69d584f9)
2012-01-24 20:51:58 +01:00
Christoph Bumiller
1a499d761a mesa: allocate transform_feedback_info::Outputs array dynamically
The nvc0 gallium driver is advertising 128 MAX_INTERLEAVED_COMPS
which made it always assert in the linker when TFB was used since
the Outputs array was smaller than that maximum.

v2: added assertions

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Paul Berry <stereotype441@gmail.com>
(cherry picked from commit d540af554a)
2012-01-24 20:51:43 +01:00
Christoph Bumiller
f97ee64606 nv50/ir: make use of TGSI_INTERPOLATE_COLOR
Flat SHADE_MODEL still overrides any non-flat interpolation
qualifier, but pulling that state out of the rasterizer cso
isn't really worth the effort, is it ?

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit af0ce1dba8)
2012-01-24 20:50:02 +01:00
Christoph Bumiller
4ac2e2f159 nvc0: fix submission of VertexID and EdgeFlag in push mode
NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit 7b6881932a)
2012-01-24 20:49:46 +01:00
Vadim Girlin
f5b787b9e3 r600g: fix interpolation with clipvertex
Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit 5a84cc4ebc)
2012-01-24 11:06:59 +00:00
Vadim Girlin
84d2bb4dfc r600g: fix VS fog export
Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit 13daa059c0)
2012-01-23 10:02:46 +00:00
Alex Deucher
e1a02333a0 r600g: fix typo in evergreen register
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 5e576efef2)
2012-01-23 09:46:46 +00:00
Vadim Girlin
20457cfcb8 r600g: take into account kcache banks for bank swizzle check
Due to the changes for multiple kcache banks support, now we are assigning
final SRCx_SEL values for kcache access at the later stage, when building the
bytecode. So we need to take into account kcache banks to distinguish
the constants with the same address but different bank index.

Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit 840a342cd0)
2012-01-23 09:30:14 +00:00
Vadim Girlin
fceca6191a r600g: implement clip vertex v2
Clip planes are uploaded as a constant buffer and used by the vertex
shader to produce corresponding clip distances for hw clipping.

Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit 54e8dcaad6)

Conflicts:

	src/gallium/drivers/r600/r600_state_common.c
2012-01-23 09:29:47 +00:00
Vadim Girlin
a71013d120 r600g: improve kcache line sets handling v2
Add support for multiple kcache banks (constant buffers).
Lock the required lines only.
Allow up to 4 kcache line sets in the alu clause by using ALU_EXTENDED on eg+.

Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit d649bf51ec)
2012-01-23 09:28:53 +00:00
Vadim Girlin
cbdf7de014 r600g: implement clip distances
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit 91d4729696)
2012-01-23 09:28:30 +00:00
Vadim Girlin
daefd7f20b r600g: implement two-sided lighting (v3)
v2: select the colors in the pixel shader

v3: fix rs state creation for pre-evergreen

Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit 725a820b92)
2012-01-23 09:28:24 +00:00
Dave Airlie
4b6dc4c1b3 r600g: srgb mode is only valid on certain format types.
"If set, forces degamma on XYZ if format is
FMT_8_8_8_8, FMT_BC1, FMT_BC2, or FMT_BC3"

Don't claim support for sRGB on any other formts.

This fixes glean texture_srgb.

Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit a9d8809f16)
2012-01-23 09:16:17 +00:00
Vadim Girlin
0d6f7b181f r600g: make INTERP_LOAD_P0 vector-only
Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit 8b1471f8ca)
2012-01-23 09:14:37 +00:00
Dave Airlie
6988699e44 r600g: fixup AR handling (v5)
So it appears R600s (except rv670) do AR handling different using a different
opcode. This patch fixes up r600g to work properly on r600.

This fixes ~100 piglit tests here (in GLSL1.30 mode) on rv610.

v3: add index_mode as per the docs.

This still fails any dst relative tests for some reason I can't quite see yet,
but it passes a lot more tests than without.

v4: add a nop after dst.rel this could be improved using a second pass,
where we only insert nops if two instructions are sure to collide.
The docs say r600, rv610, rv630 needs this, and not rv670, rs780, rs880,
need AMD to confirm rv620, rv635.

v5: add is_nop_inst.

NOTE: This is a candidate for stable branches.

Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit c96b983403)
2012-01-23 09:14:19 +00:00
Alex Deucher
533651b679 r600g: add workaround for original R600 PS setup
The original R600 requires the UNCACHED_FIRST_INST bit
to be set in the PS.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Note: this is candidate for the stable branches.
(cherry picked from commit 46ce25722b)
2012-01-23 09:14:07 +00:00
Dave Airlie
ee6a817f80 r600g: add missing r32 uint/sint fbo formats.
Fixes the GL3 required formats test.

Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit 5250bd00c0)
2012-01-23 09:13:09 +00:00
Kenneth Graunke
1b33ae3a7d i965: Remove the INTEL_OLD_VS option.
Now that we no longer generate Mesa IR from GLSL IR, it's impossible to
use the old vertex shader backend for GLSL programs.  There's simply no
Mesa IR to codegen from.

Any attempt to do so would result in immediate GPU hangs, presumably due
to the driver uploading an empty program with no EOT message.

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
(cherry picked from commit bdedd03b70)
2012-01-19 11:15:31 -08:00
Kenneth Graunke
41c1a7311e mesa: Support GL_VERTEX_ATTRIB_ARRAY_INTEGER in GL 3.0 contexts.
According to Table 6.8 (Page 348) in the OpenGL 3.0 specification,
glGetVertexAttribiv supports GL_VERTEX_ATTRIB_ARRAY_INTEGER.

NOTE: This is a candidate for the 8.0 branch.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
(cherry picked from commit d56ad273c0)
2012-01-19 11:15:20 -08:00
Chad Versace
d090099f7d i965/gen5: Fix rendering of depth buffers without stencil [v2]
Fixes the following OGLConform tests on gen5:
    depth-stencil(misc.state_on.depth_int)
    fbo_db_ARBfp(basic.OnlyDepthBuffDrawBufferRender)

The problem was that, if the depth buffer's Mesa format was X8_Z24, then
we emitted the hardware format D24_UNORM_X8. But, on gen5, D24_UNORM_S8
must be emitted.

This bug was introduced by:
    commit d84a180417
    Author: Eric Anholt <eric@anholt.net>
    i965: Base HW depth format setup based on MESA_FORMAT, not bpp.

v2: Deref 'intel' directly. Move the branch for newer chipset to top.
    Quote the PRM. As requested by Ken.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=43408
Note: This is a candidate for the 8.0 branch.
Reported-by: Xunx Fang <xunx.fang@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit a6dd4bf5fc)
2012-01-19 11:14:55 -08:00
Anuj Phogat
8ac4470041 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)
2012-01-19 11:13:07 -08:00
Eric Anholt
6382f98b9a intel: Drop the version override code now that we don't have any left.
Fixes a compiler warning.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit a14582d7e2)
2012-01-19 11:12:35 -08:00
Chad Versace
274e8b9234 i965: Fix gen6,gen7 when used with a non-HiZ capable DDX
Nothing works if HiZ is enabled and the DDX is incapable of HiZ (that is,
the DDX version is < 2.16).

The problem is that the refactoring that eliminated
intel_renderbuffer::stencil_rb broke the recovery path in
intel_verify_dri2_has_hiz().  Specifically, it broke line
intel_context.c:1445, which allocates the region for
DRI_BUFFER_DEPTH_STENCIL. That allocation was creating a separate stencil
miptree, despite the buffer being a packed depthstencil buffer. Havoc
ensued.

This patch introduces a bool flag that prevents allocation of that stencil
miptree.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=44103
Tested-by: Ian Romanick <idr@freedesktop.org>
Note: This is a candidate for the 8.0 branch.
Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit 7e08bf08d1)
2012-01-19 11:10:52 -08:00
Anuj Phogat
a12606cb54 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)
2012-01-19 11:10:30 -08:00
Marek Olšák
33f5c3946d mesa: update compute_version for GL3
only check ARB_fbo, add shader_texture_lod as a requirement

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
(cherry picked from commit 5596db7411)
2012-01-19 11:09:35 -08:00
Vinson Lee
a07afa11d2 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 11:06:02 -08:00
Paul Berry
9e0cec45ae 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 11:04:11 -08:00
Jakob Bornecrantz
c85402aba9 mesa: Bump version number to 8.0-rc1
Signed-off-by: Jakob Bornecrantz <jakob@vmware.com>
2012-01-13 22:23:28 +01:00
Brian Paul
9c81e4eed1 docs: add links to xf86-video-vmware wiki pages
(cherry picked from commit 548526f2e9)
2012-01-13 13:46:57 -07:00
Eric Anholt
daa2545508 mesa: Throw the required error for glCopyPixels from multisample FBO.
Fixes piglit EXT_framebuffer_multisample/negative-copypixels.

Reviewed-by: Brian Paul <brianp@vmware.com>
NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit 9be6654c1f)
2012-01-13 10:23:41 -08:00
Eric Anholt
baaa30ad52 mesa: Throw the required error for glCopyTex{Sub,}Image from multisample FBO.
Fixes piglit EXT_framebuffer_multisample/negative-copyteximage.

Reviewed-by: Brian Paul <brianp@vmware.com>
NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit 6950a4faf6)
2012-01-13 10:23:35 -08:00
Eric Anholt
b178514e24 mesa: Throw the required error for glReadPixels() from a multisampled FBO.
Fixes piglit EXT_framebuffer_multisample-negative-readpixels.

Reviewed-by: Brian Paul <brianp@vmware.com>
NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit 86b7c6707f)
2012-01-13 10:23:11 -08:00
Eric Anholt
48e72b6605 mesa: Avoid short-circuiting realloc of renderbuffers to new sample count.
Fixes piglit EXT_framebuffer_multisample/renderbuffer-samples.

Reviewed-by: Brian Paul <brianp@vmware.com>
NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit 0e8d156c3c)
2012-01-13 10:23:00 -08:00
Eric Anholt
89fdeab1a2 meta: Add GL_RED/GL_RG support to meta CopyTexImage.
Fixes some _mesa_problem()s in oglconform.

Reviewed-by: Brian Paul <brianp@vmware.com>
NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit f83756f80f)
2012-01-13 10:22:58 -08:00
Eric Anholt
504eaa1212 i965/gen7: Fix depth buffer rendering to tile offsets.
Previously, we were saying that everything from the starting tile to
region width+height was part of the limits of our depthbuffer, even if
the tile was near the bottom of the depthbuffer.  This mean that our
range was not clipping to buffer buonds if the start tile was anything
but the start of the buffer.

In bebc91f0f3, this was changed to
saying that we're just rendering to a region of the size of the
renderbuffer.  This is great -- we get a range that should actually
match what we want.  However, the hardware's range checking occurs
after the X/Y offset addition, so we were clipping out rendering to
small depth mip levels when an X/Y offset was present.  Just add
tile_x/y to the width in that case -- the WM won't produce negative
x/y values pre-offset, so we just need to get the left/bottom sides of
the region to cover our buffer.

Fixes the following Piglit regressions on gen7:
    spec/ARB_depth_buffer_float/fbo-clear-formats
    spec/ARB_depth_texture/fbo-clear-formats
    spec/EXT_packed_depth_stencil/fbo-clear-formats

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit e6d6a10c5a)
2012-01-13 10:22:54 -08:00
Neil Roberts
399b9799de gen6_hiz: Don't bind GL_DRAW_FRAMEBUFFER on GLES
When using Mesa with a GLES API, calling _mesa_FramebufferRenderbuffer
with GL_DRAW_FRAMEBUFFER will report a 'user error' because
get_framebuffer_target validates that this enum from the framebuffer
blit extension is only used on GL. To work around it this patch makes
it use the GL_FRAMEBUFFER enum instead in that case.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=43418
Note: This is a candidate for the 8.0 branch.
Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
(cherry picked from commit 9462b84478)
2012-01-13 10:08:13 -08:00
Brian Paul
2c1ee157c7 docs: new page describing how to build, install VMware SVGA3D guest driver
(cherry picked from commit 27915708ed)
2012-01-13 10:57:45 -07:00
Brian Paul
8b7f6de8b9 mesa: s/GLushort/GLubyte/ in pack_ubyte_AL44()
The AL44 format occupies one byte, not two.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: José Fonseca <jfonseca@vmware.com>
(cherry picked from commit b0af16abf1)
2012-01-13 10:16:33 -07:00
Brian Paul
eb8063361e osmesa: fix renderbuffer format selection
The gl_renderbuffer::Format field wasn't always set properly.  This
didn't matter much in the past but with the recent swrast/renderbuffer
mapping changes, core Mesa will be directly touching OSMesa colorbuffers
so using the right MESA_FORMAT_x value is important.

Unfortunately, there aren't MESA_FORMATs for all the possible OSmesa
format/type combinations, such as GL_FLOAT / OSMESA_ARGB.  If anyone
runs into these we can add new Mesa formats.

v2: add warnings for unsupported formats, fix ARGB_REV mix-up.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: José Fonseca <jfonseca@vmware.com>
(cherry picked from commit 540a8b2cfd)
2012-01-13 10:16:24 -07:00
Brian Paul
b26682e12e docs: freshen up the introduction page with Mesa 8.0 info, etc
(cherry picked from commit 0c14bbbc86)
2012-01-13 10:16:14 -07:00
Brian Paul
deff0244ed osmesa: fix glReadPixels, etc
Needed to implement the Map/UnmapRenderbuffer() driver hooks.
This fixes glRead/Draw/CopyPixels, etc.

See https://bugs.freedesktop.org/show_bug.cgi?id=44723

Note: This is a candidate for the 8.0 branch.

Tested-by: Kevin Hobbs <hobbsk@ohiou.edu>
(cherry picked from commit cb254b75d7)
2012-01-13 10:15:51 -07:00
Brian Paul
830688b36a intel: move declaration before code
(cherry picked from commit 062a4b601e)
2012-01-13 10:15:42 -07:00
Brian Paul
cd56917a50 intel: fix mapping of malloc'd renderbuffers
This fixes accum buffer operations.  The accumulation buffer is the
only malloc-based renderbuffer for the intel drivers.

v2: apply x/y offset to returned pointer

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: Eric Anholt <eric@anholt.net>
(cherry picked from commit 6dbdc03956)
2012-01-13 10:15:30 -07:00
Brian Paul
867fdbd36d mesa: remove incorrect (float) cast in mipmap do_row()
The array holds GLuint values so remove the float cast.
Note, however, that to compute the average of four GLuints we really
want to do (a+b+c+d)/4 but that could overflow.  This change doesn't
address that for now.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: José Fonseca <jfonseca@vmware.com>
(cherry picked from commit 003dd8adf3)
2012-01-13 10:15:07 -07:00
Brian Paul
d7a4eb331b swrast: use BITFIELD64_BIT() macro to fix MSVC warnings
NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: José Fonseca <jfonseca@vmware.com>
(cherry picked from commit 706400f0a7)
2012-01-13 10:14:59 -07:00
Brian Paul
ee5b35c568 mesa: fix ir_variable declaration
ir_variable is a class, not a struct.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: José Fonseca <jfonseca@vmware.com>
(cherry picked from commit decd018b99)
2012-01-13 10:14:46 -07:00
Brian Paul
a748a9f8ee mesa: fix incorrect float vs. int values in a few places
In the first case, the newImage[] array contains GLuint values.
In the second case, the parameter type is GLuint, but the maxDepth
value is never used in this case (GL_FLOAT_32_UNSIGNED_INT_24_8_REV).
Pass ~OU just to be safe.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: José Fonseca <jfonseca@vmware.com>
(cherry picked from commit a240c998ac)
2012-01-13 10:14:34 -07:00
Brian Paul
f1ff449120 meta: fix incorrect argument order in setup_texture_coords() call
And pass integer width, height values.

NOTE: This is a candidate for the 8.0 branch.

Reviewed-by: José Fonseca <jfonseca@vmware.com>
(cherry picked from commit 2f0fa456e3)
2012-01-13 10:14:25 -07:00
Brian Paul
af33e16eab sofpipe: remove extraneous semicolon
(cherry picked from commit 0c57323de8)
2012-01-13 10:14:07 -07:00
Brian Paul
89a1e7caf4 st/mesa: fix struct vs. class compilation warning
glsl_to_tgsi_visitor is earlier defined as a class, not a struct.
Fixes MSVC warning.

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit 9f2963b631)
2012-01-13 10:13:58 -07:00
Brian Paul
ad77e2fb78 configs: fix, simplify RADEON_LIBS, RADEON_CFLAGS
Fixes build problems with the r200, radeon drivers.

NOTE: This is a candidate for the 8.0 branch.
(cherry picked from commit c7188ece0e)
2012-01-13 10:13:41 -07:00
Dave Airlie
1e51d4b4f6 softpipe: bump max texture array layers to 256.
This as per GL3 specification.

Signed-off-by: Dave Airlie <airlied@redhat.com>
2012-01-13 08:29:58 +00:00
Dave Airlie
0b9e6d2017 r600g: don't advertise integers yet on r600.
Still some work to be done before this is finished.

This is a candidate for 8.0 branch.

Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit 3e044bcc4b)
2012-01-13 08:28:52 +00:00
Thomas Hellstrom
57665cb09e configure: Add the svga gallium driver to the default gallium drivers
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Jakob Bornecrantz <jakob@vmware.com>
2012-01-12 13:34:14 +01:00
Thomas Hellstrom
372b7f1d35 st/xa: Bump version to 1.0.0 according to the README
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Jakob Bornecrantz <jakob@vmware.com>
2012-01-12 13:33:41 +01:00
José Fonseca
4b86bd1347 svga: Fix user clip planes.
Dirty flags also need to be updated in face of recent interface change.

Fixes regression in compiz.

Reviewed-by: Brian Paul <brianp@vmware.com>
2012-01-12 13:33:24 +01:00
Thomas Hellstrom
7465c5d976 gallium/svga: Pass the SVGA3D_SURFACE_HINT_RENDERTARGET flag to the device
Some hardware versions rely on it to render correctly.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
Reviewed-by: Jakob Bornecrantz <jakob@vmware.com>
2012-01-12 11:28:07 +01:00
Brian Paul
9489ae8938 mesa: remove const qualifier from fProg to silence warning
The args to _mesa_reference_shader_program() can't be const.
(cherry picked from commit 459a44460e)
2012-01-11 18:24:34 -07:00
Brian Paul
b8af8b83cf mesa: include uniforms.h to silence warning, remove unused var
(cherry picked from commit fe1b38960b)
2012-01-11 18:24:26 -07:00
13591 changed files with 994813 additions and 5549077 deletions

View File

View File

View File

View File

View File

View File

View File

View File

View File

View File

View File

@@ -1,2 +0,0 @@
# Vendored code
src/amd/vulkan/radix_sort/*

View File

@@ -1,10 +0,0 @@
# The following files are opted into `ninja clang-format` and
# enforcement in the CI.
src/gallium/drivers/i915
src/gallium/drivers/r300/compiler/*
src/gallium/targets/teflon/**/*
src/amd/vulkan/**/*
src/amd/compiler/**/*
src/egl/**/*
src/etnaviv/isa/**/*

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

@@ -1,44 +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,y,yy}]
indent_style = space
indent_size = 3
max_line_length = 78
[{Makefile*,*.mk}]
indent_style = tab
[*.py]
indent_style = space
indent_size = 4
[*.yml]
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
[*.ps1]
indent_style = space
indent_size = 2
[*.rs]
indent_style = space
indent_size = 4

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)

View File

@@ -1,70 +0,0 @@
# List of commits to ignore when using `git blame`.
# Enable with:
# git config blame.ignoreRevsFile .git-blame-ignore-revs
#
# Per git-blame(1):
# Ignore revisions listed in the file, one unabbreviated object name
# per line, in git-blame. Whitespace and comments beginning with # are
# ignored.
#
# Please keep these in chronological order :)
#
# You can add a new commit with the following command:
# git log -1 --pretty=format:'%n# %s%n%H%n' >> .git-blame-ignore-revs $COMMIT
# pvr: Fix clang-format error.
0ad5b0a74ef73f5fcbe1406ad9d57fe5dc00a5b1
# panfrost: Fix up some formatting for clang-format
a4705afe63412498d13ded73cba969c66be67907
# asahi: clang-format the world again
26c51bb8d8a33098b1990425a391f56ffba5728c
# perfetto: Add a .clang-format for the directory.
da78d5d729b1800136dd713b68492cb339993f4a
# panfrost/winsys: Clang-format
c90f036516a5376002be6550a917e8bad6a8a3b8
# panfrost: Re-run clang-format
4ccf174009af6732cbffa5d8ebb4687da7517505
# panvk: Clang-format
c7bf3b69ebc8f2252dbf724a4de638e6bb2ac402
# pan/mdg: Fix icky formatting
133af0d6c945d3aaca8989edd15283a2b7dcc6c7
# mapi: clang-format _glapi_add_dispatch()
30332529663268a6406e910848e906e725e6fda7
# radv: reformat according to its .clang-format
8b319c6db8bd93603b18bd783eb75225fcfd51b7
# aco: reformat according to its .clang-format
6b21653ab4d3a67e711fe10e3d403128b6d26eb2
# egl: re-format using clang-format
2f670d89db038d5a29f6b72732fd7ad63dfaf4c6
# panfrost: clang-format the tree
0afd691f29683f6e9dde60f79eca094373521806
# aco: Format.
1e2639026fec7069806449f9ba2a124ce4eb5569
# radv: Format.
59c501ca353f8ec9d2717c98af2bfa1a1dbf4d75
# pvr: clang-format fixes
953c04ebd39c52d457301bdd8ac803949001da2d
# freedreno: Re-indent
2d439343ea1aee146d4ce32800992cd389bd505d
# ir3: Reformat source with clang-format
177138d8cb0b4f6a42ef0a1f8593e14d79f17c54
# ir3: reformat after refactoring in previous commit
8ae5b27ee0331a739d14b42e67586784d6840388

11
.gitattributes vendored
View File

@@ -1,7 +1,4 @@
*.csv eol=crlf
* text=auto
*.jpg binary
*.png binary
*.gif binary
*.ico binary
*.cl gitlab-language=c
*.dsp -crlf
*.dsw -crlf
*.sln -crlf
*.vcproj -crlf

View File

@@ -1,60 +0,0 @@
name: macOS-CI
on: push
permissions:
contents: read
jobs:
macOS-CI:
strategy:
matrix:
glx_option: ['dri', 'xlib']
runs-on: macos-11
env:
GALLIUM_DUMP_CPU: true
MESON_EXEC: /Users/runner/Library/Python/3.11/bin/meson
steps:
- name: Checkout
uses: actions/checkout@v3
- name: Install Dependencies
run: |
cat > Brewfile <<EOL
brew "bison"
brew "expat"
brew "gettext"
brew "libx11"
brew "libxcb"
brew "libxdamage"
brew "libxext"
brew "molten-vk"
brew "ninja"
brew "pkg-config"
brew "python@3.10"
EOL
brew update
brew bundle --verbose
- name: Install Mako and meson
run: pip3 install --user mako meson
- name: Configure
run: |
cat > native_config <<EOL
[binaries]
llvm-config = '/usr/local/opt/llvm/bin/llvm-config'
EOL
$MESON_EXEC . build --native-file=native_config -Dmoltenvk-dir=$(brew --prefix molten-vk) -Dbuild-tests=true -Dosmesa=true -Dgallium-drivers=swrast,zink -Dglx=${{ matrix.glx_option }}
- name: Build
run: $MESON_EXEC compile -C build
- name: Test
run: $MESON_EXEC test -C build --print-errorlogs
- name: Install
run: $MESON_EXEC install -C build --destdir $PWD/install
- name: 'Upload Artifact'
if: always()
uses: actions/upload-artifact@v3
with:
name: macos-${{ matrix.glx_option }}-result
path: |
build/meson-logs/
install/
retention-days: 5

37
.gitignore vendored
View File

@@ -1,7 +1,34 @@
.cache
.vscode*
*.a
*.dll
*.exe
*.ilk
*.o
*.obj
*.os
*.pc
*.pdb
*.pyc
*.pyo
*.out
/build
.venv/
*.so
*.so.*
*.sw[a-z]
*.tar
*.tar.bz2
*.tar.gz
*.zip
*~
depend
depend.bak
lib
lib64
configure
autom4te.cache
aclocal.m4
config.log
config.status
cscope*
.scon*
config.py
build
manifest.txt
.dir-locals.el

View File

@@ -1,375 +0,0 @@
# Types of CI pipelines:
# | pipeline name | context | description |
# |----------------------|-----------|-------------------------------------------------------------|
# | merge pipeline | mesa/mesa | pipeline running for an MR; if it passes the MR gets merged |
# | pre-merge pipeline | mesa/mesa | same as above, except its status doesn't affect the MR |
# | post-merge pipeline | mesa/mesa | pipeline immediately after merging |
# | fork pipeline | fork | pipeline running in a user fork |
# | scheduled pipeline | mesa/mesa | nightly pipelines, running every morning at 4am UTC |
# | direct-push pipeline | mesa/mesa | when commits are pushed directly to mesa/mesa, bypassing Marge and its gating pipeline |
#
# Note that the release branches maintained by the release manager fall under
# the "direct push" category.
#
# "context" indicates the permissions that the jobs get; notably, any
# container created in mesa/mesa gets pushed immediately for everyone to use
# as soon as the image tag change is merged.
#
# Merge pipelines contain all jobs that must pass before the MR can be merged.
# Pre-merge pipelines contain the exact same jobs as merge pipelines.
# Post-merge pipelines contain *only* the `pages` job that deploys the new
# version of the website.
# Fork pipelines contain everything.
# Scheduled pipelines only contain the container+build jobs, and some extra
# test jobs (typically "full" variants of pre-merge jobs that only run 1/X
# test cases), but not a repeat of the merge pipeline jobs.
# Direct-push pipelines contain the same jobs as merge pipelines.
workflow:
rules:
# do not duplicate pipelines on merge pipelines
- if: $CI_COMMIT_BRANCH && $CI_OPEN_MERGE_REQUESTS && $CI_PIPELINE_SOURCE == "push"
when: never
# tag pipelines are disabled as it's too late to run all the tests by
# then, the release has been made based on the staging pipelines results
- if: $CI_COMMIT_TAG
when: never
# merge pipeline
- if: &is-merge-attempt $GITLAB_USER_LOGIN == "marge-bot" && $CI_PIPELINE_SOURCE == "merge_request_event"
variables:
MESA_CI_PERFORMANCE_ENABLED: 1
CI_TRON_JOB_PRIORITY_TAG: "" # Empty tags are ignored by gitlab
JOB_PRIORITY: 75
# fast-fail in merge pipelines: stop early if we get this many unexpected fails/crashes
DEQP_RUNNER_MAX_FAILS: 40
# post-merge pipeline
- if: &is-post-merge $GITLAB_USER_LOGIN == "marge-bot" && $CI_PIPELINE_SOURCE == "push"
# Pre-merge pipeline
- if: &is-pre-merge $CI_PIPELINE_SOURCE == "merge_request_event"
# Push to a branch on a fork
- if: &is-fork-push $CI_PROJECT_NAMESPACE != "mesa" && $CI_PIPELINE_SOURCE == "push"
# nightly pipeline
- if: &is-scheduled-pipeline $CI_PIPELINE_SOURCE == "schedule"
variables:
JOB_PRIORITY: 45
# (some) nightly builds perform LTO, so they take much longer than the
# short timeout allowed in other pipelines.
# Note: 0 = infinity = gitlab's job `timeout:` applies, which is 1h
BUILD_JOB_TIMEOUT_OVERRIDE: 0
# pipeline for direct pushes that bypassed the CI
- if: &is-direct-push $CI_PROJECT_NAMESPACE == "mesa" && $CI_PIPELINE_SOURCE == "push" && $CI_COMMIT_REF_NAME == $CI_DEFAULT_BRANCH
variables:
JOB_PRIORITY: 70
# pipeline for direct pushes from release maintainer
- if: &is-staging-push $CI_PROJECT_NAMESPACE == "mesa" && $CI_PIPELINE_SOURCE == "push" && $CI_COMMIT_REF_NAME =~ /^staging\//
variables:
JOB_PRIORITY: 70
variables:
FDO_UPSTREAM_REPO: mesa/mesa
MESA_TEMPLATES_COMMIT: &ci-templates-commit e195d80f35b45cc73668be3767b923fd76c70ed5
CI_PRE_CLONE_SCRIPT: |-
set -o xtrace
wget -q -O download-git-cache.sh ${CI_PROJECT_URL}/-/raw/${CI_COMMIT_SHA}/.gitlab-ci/download-git-cache.sh
bash download-git-cache.sh
rm download-git-cache.sh
set +o xtrace
S3_JWT_FILE: /s3_jwt
S3_JWT_FILE_SCRIPT: |-
echo -n '${S3_JWT}' > '${S3_JWT_FILE}' &&
unset CI_JOB_JWT S3_JWT # Unsetting vulnerable env variables
S3_HOST: s3.freedesktop.org
# This bucket is used to fetch ANDROID prebuilts and images
S3_ANDROID_BUCKET: mesa-rootfs
# This bucket is used to fetch the kernel image
S3_KERNEL_BUCKET: mesa-rootfs
# Bucket for git cache
S3_GITCACHE_BUCKET: git-cache
# Bucket for the pipeline artifacts pushed to S3
S3_ARTIFACTS_BUCKET: artifacts
# Buckets for traces
S3_TRACIE_RESULTS_BUCKET: mesa-tracie-results
S3_TRACIE_PUBLIC_BUCKET: mesa-tracie-public
S3_TRACIE_PRIVATE_BUCKET: mesa-tracie-private
# per-pipeline artifact storage on MinIO
PIPELINE_ARTIFACTS_BASE: ${S3_HOST}/${S3_ARTIFACTS_BUCKET}/${CI_PROJECT_PATH}/${CI_PIPELINE_ID}
# per-job artifact storage on MinIO
JOB_ARTIFACTS_BASE: ${PIPELINE_ARTIFACTS_BASE}/${CI_JOB_ID}
# reference images stored for traces
PIGLIT_REPLAY_REFERENCE_IMAGES_BASE: "${S3_HOST}/${S3_TRACIE_RESULTS_BUCKET}/$FDO_UPSTREAM_REPO"
# For individual CI farm status see .ci-farms folder
# Disable farm with `git mv .ci-farms{,-disabled}/$farm_name`
# Re-enable farm with `git mv .ci-farms{-disabled,}/$farm_name`
# NEVER MIX FARM MAINTENANCE WITH ANY OTHER CHANGE IN THE SAME MERGE REQUEST!
ARTIFACTS_BASE_URL: https://${CI_PROJECT_ROOT_NAMESPACE}.${CI_PAGES_DOMAIN}/-/${CI_PROJECT_NAME}/-/jobs/${CI_JOB_ID}/artifacts
# Python scripts for structured logger
PYTHONPATH: "$PYTHONPATH:$CI_PROJECT_DIR/install"
# No point in continuing once the device is lost
MESA_VK_ABORT_ON_DEVICE_LOSS: 1
# Avoid the wall of "Unsupported SPIR-V capability" warnings in CI job log, hiding away useful output
MESA_SPIRV_LOG_LEVEL: error
# Default priority for non-merge pipelines
CI_TRON_JOB_PRIORITY_TAG: ci-tron:priority:low
JOB_PRIORITY: 50
DATA_STORAGE_PATH: data_storage
default:
id_tokens:
S3_JWT:
aud: https://s3.freedesktop.org
before_script:
- |
if [ -z "${KERNEL_IMAGE_BASE:-}" ]; then
export KERNEL_IMAGE_BASE="https://${S3_HOST}/${S3_KERNEL_BUCKET}/${KERNEL_REPO}/${EXTERNAL_KERNEL_TAG:-$KERNEL_TAG}"
fi
- >
export SCRIPTS_DIR=$(mktemp -d) &&
curl -L -s --retry 4 -f --retry-all-errors --retry-delay 60 -O --output-dir "${SCRIPTS_DIR}" "${CI_PROJECT_URL}/-/raw/${CI_COMMIT_SHA}/.gitlab-ci/setup-test-env.sh" &&
. ${SCRIPTS_DIR}/setup-test-env.sh
- eval "$S3_JWT_FILE_SCRIPT"
after_script:
# Work around https://gitlab.com/gitlab-org/gitlab/-/issues/20338
- find -name '*.log' -exec mv {} {}.txt \;
# Retry when job fails. Failed jobs can be found in the Mesa CI Daily Reports:
# https://gitlab.freedesktop.org/mesa/mesa/-/issues/?sort=created_date&state=opened&label_name%5B%5D=CI%20daily
retry:
max: 1
# Ignore runner_unsupported, stale_schedule, archived_failure, or
# unmet_prerequisites
when:
- api_failure
- runner_system_failure
- script_failure
- job_execution_timeout
- scheduler_failure
- data_integrity_failure
- unknown_failure
stages:
- sanity
- container
- git-archive
- build-for-tests
- build-only
- code-validation
- amd
- amd-postmerge
- intel
- intel-postmerge
- nouveau
- nouveau-postmerge
- arm
- arm-postmerge
- broadcom
- broadcom-postmerge
- freedreno
- freedreno-postmerge
- etnaviv
- etnaviv-postmerge
- software-renderer
- software-renderer-postmerge
- layered-backends
- layered-backends-postmerge
- performance
- deploy
include:
- project: 'freedesktop/ci-templates'
ref: 16bc29078de5e0a067ff84a1a199a3760d3b3811
file:
- '/templates/ci-fairy.yml'
- project: 'freedesktop/ci-templates'
ref: *ci-templates-commit
file:
- '/templates/alpine.yml'
- '/templates/debian.yml'
- '/templates/fedora.yml'
- local: '.gitlab-ci/image-tags.yml'
- local: '.gitlab-ci/lava/lava-gitlab-ci.yml'
- local: '.gitlab-ci/container/gitlab-ci.yml'
- local: '.gitlab-ci/build/gitlab-ci.yml'
- local: '.gitlab-ci/test/gitlab-ci.yml'
- local: '.gitlab-ci/farm-rules.yml'
- local: '.gitlab-ci/test-source-dep.yml'
- local: 'docs/gitlab-ci.yml'
- local: 'src/**/ci/gitlab-ci.yml'
# Rules applied to every job in the pipeline
.common-rules:
rules:
- if: *is-fork-push
when: manual
.never-post-merge-rules:
rules:
- if: *is-post-merge
when: never
.container+build-rules:
rules:
- !reference [.common-rules, rules]
# Run when re-enabling a disabled farm, but not when disabling it
- !reference [.disable-farm-mr-rules, rules]
# Never run immediately after merging, as we just ran everything
- !reference [.never-post-merge-rules, rules]
# Build everything in merge pipelines, if any files affecting the pipeline
# were changed
- if: *is-merge-attempt
changes: &all_paths
- VERSION
- bin/git_sha1_gen.py
- bin/install_megadrivers.py
- bin/symbols-check.py
- bin/ci/**/*
# GitLab CI
- .gitlab-ci.yml
- .gitlab-ci/**/*
- .ci-farms/*
# Meson
- meson*
- build-support/**/*
- subprojects/**/*
# clang format
- .clang-format
- .clang-format-include
- .clang-format-ignore
# Source code
- include/**/*
- src/**/*
when: on_success
# Same as above, but for pre-merge pipelines
- if: *is-pre-merge
changes:
*all_paths
when: manual
# Skip everything for pre-merge and merge pipelines which don't change
# anything in the build
- if: *is-merge-attempt
when: never
- if: *is-pre-merge
when: never
# Build everything after someone bypassed the CI
- if: *is-direct-push
when: on_success
# Build everything when pushing to staging branches
- if: *is-staging-push
when: on_success
# Build everything in scheduled pipelines
- if: *is-scheduled-pipeline
when: on_success
# Allow building everything in fork pipelines, but build nothing unless
# manually triggered
- when: manual
.ci-deqp-artifacts:
artifacts:
name: "${CI_PROJECT_NAME}_${CI_JOB_NAME}"
when: always
untracked: false
paths:
# Watch out! Artifacts are relative to the build dir.
# https://gitlab.com/gitlab-org/gitlab-ce/commit/8788fb925706cad594adf6917a6c5f6587dd1521
- artifacts
- _build/meson-logs/*.txt
- _build/meson-logs/strace
- _build/.ninja_log
# Git archive
make git archive:
extends:
- .fdo.ci-fairy
stage: git-archive
rules:
- !reference [.scheduled_pipeline-rules, rules]
# ensure we are running on packet
tags:
- packet.net
script:
# Compactify the .git directory
- git gc --aggressive
# Download & cache the perfetto subproject as well.
- rm -rf subprojects/perfetto ; mkdir -p subprojects/perfetto && curl --fail https://android.googlesource.com/platform/external/perfetto/+archive/$(grep 'revision =' subprojects/perfetto.wrap | cut -d ' ' -f3).tar.gz | tar zxf - -C subprojects/perfetto
# compress the current folder
- tar -cvzf ../$CI_PROJECT_NAME.tar.gz .
- s3_upload ../$CI_PROJECT_NAME.tar.gz "https://$S3_HOST/git-cache/$CI_PROJECT_NAMESPACE/$CI_PROJECT_NAME/"
# Sanity checks of MR settings and commit logs
sanity:
extends:
- .fdo.ci-fairy
stage: sanity
rules:
- if: *is-pre-merge
when: on_success
- when: never
variables:
GIT_STRATEGY: none
script:
# ci-fairy check-commits --junit-xml=check-commits.xml
- ci-fairy check-merge-request --require-allow-collaboration --junit-xml=check-merge-request.xml
- |
set -eu
image_tags=(
ALPINE_X86_64_BUILD_TAG
ALPINE_X86_64_LAVA_SSH_TAG
DEBIAN_BASE_TAG
DEBIAN_BUILD_TAG
DEBIAN_PYUTILS_TAG
DEBIAN_TEST_ANDROID_TAG
DEBIAN_TEST_GL_TAG
DEBIAN_TEST_VK_TAG
FEDORA_X86_64_BUILD_TAG
KERNEL_ROOTFS_TAG
KERNEL_TAG
PKG_REPO_REV
WINDOWS_X64_BUILD_TAG
WINDOWS_X64_MSVC_TAG
WINDOWS_X64_TEST_TAG
)
for var in "${image_tags[@]}"
do
if [ "$(echo -n "${!var}" | wc -c)" -gt 20 ]
then
echo "$var is too long; please make sure it is at most 20 chars."
exit 1
fi
done
artifacts:
when: on_failure
reports:
junit: check-*.xml
tags:
- placeholder-job
mr-label-maker-test:
extends:
- .fdo.ci-fairy
stage: sanity
rules:
- !reference [.mr-label-maker-rules, rules]
variables:
GIT_STRATEGY: fetch
timeout: 10m
script:
- set -eu
- python3 -m venv .venv
- source .venv/bin/activate
- pip install git+https://gitlab.freedesktop.org/freedesktop/mr-label-maker
- mr-label-maker --dry-run --mr $CI_MERGE_REQUEST_IID
# Jobs that need to pass before spending hardware resources on further testing
.required-for-hardware-jobs:
needs:
- job: rustfmt
optional: true
- job: yaml-toml-shell-py-test
optional: true

View File

@@ -1,33 +0,0 @@
[flake8]
exclude = .venv*,
# PEP 8 Style Guide limits line length to 79 characters
max-line-length = 159
ignore =
# continuation line under-indented for hanging indent
E121
# continuation line over-indented for hanging indent
E126,
# continuation line under-indented for visual indent
E128,
# whitespace before ':'
E203,
# missing whitespace around arithmetic operator
E226,
# missing whitespace after ','
E231,
# expected 2 blank lines, found 1
E302,
# too many blank lines
E303,
# imported but unused
F401,
# f-string is missing placeholders
F541,
# local variable assigned to but never used
F841,
# line break before binary operator
W503,
# line break after binary operator
W504,

View File

@@ -1,87 +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.
# This test checks the driver's reported conformance version against the
# version of the CTS we're running. This check fails every few months
# and everyone has to go and bump the number in every driver.
# Running this check only makes sense while preparing a conformance
# submission, so skip it in the regular CI.
dEQP-VK.api.driver_properties.conformance_version
# Exclude this test which might fail when a new extension is implemented.
dEQP-VK.info.device_extensions
# 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
# These tests all read from the front buffer after a swap. Given that we
# run piglit tests in parallel in Mesa CI, and don't have a compositor
# running, the frontbuffer reads may end up with undefined results from
# windows overlapping us.
#
# Piglit does mark these tests as not to be run in parallel, but deqp-runner
# doesn't respect that. We need to extend deqp-runner to allow some tests to be
# marked as single-threaded and run after the rayon loop if we want to support
# them.
#
# Note that "glx-" tests don't appear in x11-skips.txt because they can be
# run even if PIGLIT_PLATFORM=gbm (for example)
glx@glx-copy-sub-buffer.*
# A majority of the tests introduced in CTS 1.3.7.0 are experiencing failures and flakes.
# Disable these tests until someone with a more deeper understanding of EGL examines them.
#
# Note: on sc8280xp/a690 I get identical results (same passes and fails)
# between freedreno, zink, and llvmpipe, so I believe this is either a
# deqp bug or egl/wayland bug, rather than driver issue.
#
# With llvmpipe, the failing tests have the error message:
#
# "Illegal sampler view creation without bind flag"
#
# which might be a hint. (But some passing tests also have the same
# error message.)
#
# more context from David Heidelberg on IRC: the deqp commit where these
# started failing is: https://github.com/KhronosGroup/VK-GL-CTS/commit/79b25659bcbced0cfc2c3fe318951c585f682abe
# prior to that they were skipping.
wayland-dEQP-EGL.functional.color_clears.single_context.gles1.other
wayland-dEQP-EGL.functional.color_clears.single_context.gles2.other
wayland-dEQP-EGL.functional.color_clears.single_context.gles3.other
wayland-dEQP-EGL.functional.color_clears.multi_context.gles1.other
wayland-dEQP-EGL.functional.color_clears.multi_context.gles2.other
wayland-dEQP-EGL.functional.color_clears.multi_context.gles3.other
wayland-dEQP-EGL.functional.color_clears.multi_context.gles1_gles2.other
wayland-dEQP-EGL.functional.color_clears.multi_context.gles1_gles2_gles3.other
wayland-dEQP-EGL.functional.color_clears.multi_thread.gles1.other
wayland-dEQP-EGL.functional.color_clears.multi_thread.gles2.other
wayland-dEQP-EGL.functional.color_clears.multi_thread.gles3.other
wayland-dEQP-EGL.functional.color_clears.multi_thread.gles1_gles2.other
wayland-dEQP-EGL.functional.color_clears.multi_thread.gles1_gles2_gles3.other
# Seems to be the same is as wayland-dEQP-EGL.functional.color_clears.*
wayland-dEQP-EGL.functional.render.single_context.gles2.other
wayland-dEQP-EGL.functional.render.single_context.gles3.other
wayland-dEQP-EGL.functional.render.multi_context.gles2.other
wayland-dEQP-EGL.functional.render.multi_context.gles3.other
wayland-dEQP-EGL.functional.render.multi_context.gles2_gles3.other
wayland-dEQP-EGL.functional.render.multi_thread.gles2.other
wayland-dEQP-EGL.functional.render.multi_thread.gles3.other
wayland-dEQP-EGL.functional.render.multi_thread.gles2_gles3.other
# These test the loader more than the implementation and are broken because the
# Vulkan loader in Debian is too old
dEQP-VK.api.get_device_proc_addr.non_enabled
dEQP-VK.api.version_check.unavailable_entry_points

View File

@@ -1,7 +0,0 @@
# Unlike zink which does support it, ANGLE relies on a waiver to not implement
# capturing individual array elements (see waivers.xml and gles3-waivers.txt in the CTS)
dEQP-GLES3.functional.transform_feedback.array_element.*
dEQP-GLES3.functional.transform_feedback.random.*
dEQP-GLES31.functional.program_interface_query.transform_feedback_varying.*_array_element
dEQP-GLES31.functional.program_interface_query.transform_feedback_varying.type.*.array.*
KHR-GLES31.core.program_interface_query.transform-feedback-types

View File

@@ -1,119 +0,0 @@
version: 1
# Rules to match for a machine to qualify
target:
id: '{{ ci_runner_description }}'
timeouts:
first_console_activity: # This limits the time it can take to receive the first console log
minutes: {{ timeout_first_console_activity_minutes | default(0, true) }}
seconds: {{ timeout_first_console_activity_seconds | default(0, true) }}
retries: {{ timeout_first_console_activity_retries }}
console_activity: # Reset every time we receive a message from the logs
minutes: {{ timeout_console_activity_minutes | default(0, true) }}
seconds: {{ timeout_console_activity_seconds | default(0, true) }}
retries: {{ timeout_console_activity_retries }}
boot_cycle:
minutes: {{ timeout_boot_minutes | default(0, true) }}
seconds: {{ timeout_boot_seconds | default(0, true) }}
retries: {{ timeout_boot_retries }}
overall: # Maximum time the job can take, not overrideable by the "continue" deployment
minutes: {{ timeout_overall_minutes | default(0, true) }}
seconds: {{ timeout_overall_seconds | default(0, true) }}
retries: 0
# no retries possible here
watchdogs:
boot:
minutes: {{ timeout_boot_wd_minutes | default(0, true) }}
seconds: {{ timeout_boot_wd_seconds | default(0, true) }}
retries: {{ timeout_boot_wd_retries | default(0, true) }}
console_patterns:
session_end:
regex: >-
{{ session_end_regex }}
{% if session_reboot_regex %}
session_reboot:
regex: >-
{{ session_reboot_regex }}
{% endif %}
job_success:
regex: >-
{{ job_success_regex }}
{% if job_warn_regex %}
job_warn:
regex: >-
{{ job_warn_regex }}
{% endif %}
{% if boot_wd_start_regex and boot_wd_stop_regex %}
watchdogs:
boot:
start:
regex: >-
{{ boot_wd_start_regex }}
reset:
regex: >-
{{ boot_wd_reset_regex | default(boot_wd_start_regex, true) }}
stop:
regex: >-
{{ boot_wd_stop_regex }}
{% endif %}
# Environment to deploy
deployment:
# Initial boot
start:
storage:
http:
- path: "/b2c-extra-args"
data: >
b2c.pipefail b2c.poweroff_delay={{ poweroff_delay }}
b2c.minio="gateway,{{ '{{' }} minio_url }},{{ '{{' }} job_bucket_access_key }},{{ '{{' }} job_bucket_secret_key }}"
b2c.volume="{{ '{{' }} job_bucket }}-results,mirror=gateway/{{ '{{' }} job_bucket }},pull_on=pipeline_start,push_on=changes,overwrite{% for excl in job_volume_exclusions %},exclude={{ excl }}{% endfor %},remove,expiration=pipeline_end,preserve"
{% for volume in volumes %}
b2c.volume={{ volume }}
{% endfor %}
b2c.run_service="--privileged --tls-verify=false --pid=host docker://{{ '{{' }} fdo_proxy_registry }}/gfx-ci/ci-tron/telegraf:latest" b2c.hostname=dut-{{ '{{' }} machine.full_name }}
b2c.run="-ti --tls-verify=false docker://{{ '{{' }} fdo_proxy_registry }}/gfx-ci/ci-tron/machine-registration:latest check"
b2c.run="-v {{ '{{' }} job_bucket }}-results:{{ working_dir }} -w {{ working_dir }} {% for mount_volume in mount_volumes %} -v {{ mount_volume }}{% endfor %} --tls-verify=false docker://{{ local_container }} {{ container_cmd | replace('"', '\\\"') }}"
kernel:
{% if kernel_url %}
url: '{{ kernel_url }}'
{% endif %}
# NOTE: b2c.cache_device should not be here, but this works around
# a limitation of b2c which will be removed in the next release
cmdline: >
SALAD.machine_id={{ '{{' }} machine_id }}
console={{ '{{' }} local_tty_device }},115200
b2c.cache_device=auto b2c.ntp_peer=10.42.0.1
b2c.extra_args_url={{ '{{' }} job.http.url }}/b2c-extra-args
{% if kernel_cmdline_extras is defined %}
{{ kernel_cmdline_extras }}
{% endif %}
{% if initramfs_url or firmware_url %}
initramfs:
{% if firmware_url %}
- url: '{{ firmware_url }}'
{% endif %}
{% if initramfs_url %}
- url: '{{ initramfs_url }}'
{% endif %}
{% endif %}
{% if dtb_url %}
dtb:
url: '{{ dtb_url }}'
{% if dtb_match %}
format:
archive:
match: "{{ dtb_match }}"
{% endif %}
{% endif %}

View File

@@ -1,55 +0,0 @@
#!/usr/bin/env python3
# Copyright © 2022 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 (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.
from jinja2 import Environment, FileSystemLoader
from os import environ, path
# Pass all the environment variables prefixed by B2C_
values = {
key.removeprefix("B2C_").lower(): environ[key]
for key in environ if key.startswith("B2C_")
}
env = Environment(loader=FileSystemLoader(path.dirname(values['job_template'])),
trim_blocks=True, lstrip_blocks=True)
template = env.get_template(path.basename(values['job_template']))
values['ci_job_id'] = environ['CI_JOB_ID']
values['ci_runner_description'] = environ['CI_RUNNER_DESCRIPTION']
values['job_volume_exclusions'] = [excl for excl in values['job_volume_exclusions'].split(",") if excl]
values['working_dir'] = environ['CI_PROJECT_DIR']
# Use the gateway's pull-through registry caches to reduce load on fd.o.
values['local_container'] = environ['IMAGE_UNDER_TEST']
values['local_container'] = values['local_container'].replace(
'registry.freedesktop.org',
'{{ fdo_proxy_registry }}'
)
if 'kernel_cmdline_extras' not in values:
values['kernel_cmdline_extras'] = ''
with open(path.splitext(path.basename(values['job_template']))[0], "w") as f:
f.write(template.render(values))

View File

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

View File

@@ -1,15 +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
export CURRENT_SECTION=dut_boot
# 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

@@ -1,17 +0,0 @@
#!/bin/bash
# shellcheck disable=SC2086 # we want word splitting
if [ -z "$BM_POE_INTERFACE" ]; then
echo "Must supply the PoE Interface to power down"
exit 1
fi
if [ -z "$BM_POE_ADDRESS" ]; then
echo "Must supply the PoE Switch host"
exit 1
fi
SNMP_KEY="1.3.6.1.4.1.9.9.402.1.2.1.1.1.$BM_POE_INTERFACE"
SNMP_OFF="i 4"
snmpset -v2c -r 3 -t 30 -cmesaci "$BM_POE_ADDRESS" "$SNMP_KEY" $SNMP_OFF

View File

@@ -1,22 +0,0 @@
#!/bin/bash
# shellcheck disable=SC2086 # we want word splitting
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
set -ex
SNMP_KEY="1.3.6.1.4.1.9.9.402.1.2.1.1.1.$BM_POE_INTERFACE"
SNMP_ON="i 1"
SNMP_OFF="i 4"
snmpset -v2c -r 3 -t 10 -cmesaci "$BM_POE_ADDRESS" "$SNMP_KEY" $SNMP_OFF
sleep 3s
snmpset -v2c -r 3 -t 10 -cmesaci "$BM_POE_ADDRESS" "$SNMP_KEY" $SNMP_ON

View File

@@ -1,129 +0,0 @@
#!/bin/bash
# shellcheck disable=SC1091 # The relative paths in this file only become valid at runtime.
# shellcheck disable=SC2034
# shellcheck disable=SC2086 # we want word splitting
# 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
CI_INSTALL=$CI_PROJECT_DIR/install
# Runner config checks
if [ -z "$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
. "${SCRIPTS_DIR}/setup-test-env.sh"
section_start prepare_rootfs "Preparing rootfs components"
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
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
$BM_KERNEL -o /tftp/vmlinuz
elif [ -n "${EXTERNAL_KERNEL_TAG}" ]; then
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
"${FDO_HTTP_CACHE_URI:-}${KERNEL_IMAGE_BASE}/${DEBIAN_ARCH}/${BM_KERNEL}" -o /tftp/vmlinuz
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
"${FDO_HTTP_CACHE_URI:-}${KERNEL_IMAGE_BASE}/${DEBIAN_ARCH}/modules.tar.zst" -o modules.tar.zst
tar --keep-directory-symlink --zstd -xf modules.tar.zst -C "/nfs/"
rm modules.tar.zst &
else
cp /baremetal-files/"$BM_KERNEL" /tftp/vmlinuz
fi
echo "$BM_CMDLINE" > /tftp/cmdline
set +e
STRUCTURED_LOG_FILE=results/job_detail.json
python3 $CI_INSTALL/custom_logger.py ${STRUCTURED_LOG_FILE} --update dut_job_type "${DEVICE_TYPE}"
python3 $CI_INSTALL/custom_logger.py ${STRUCTURED_LOG_FILE} --update farm "${FARM}"
python3 $CI_INSTALL/custom_logger.py ${STRUCTURED_LOG_FILE} --create-dut-job dut_name "${CI_RUNNER_DESCRIPTION}"
python3 $CI_INSTALL/custom_logger.py ${STRUCTURED_LOG_FILE} --update-dut-time submit "${CI_JOB_STARTED_AT}"
section_end prepare_rootfs
python3 $BM/cros_servo_run.py \
--cpu $BM_SERIAL \
--ec $BM_SERIAL_EC \
--test-timeout ${TEST_PHASE_TIMEOUT_MINUTES:-20}
ret=$?
section_start dut_cleanup "Cleaning up after job"
python3 $CI_INSTALL/custom_logger.py ${STRUCTURED_LOG_FILE} --close-dut-job
python3 $CI_INSTALL/custom_logger.py ${STRUCTURED_LOG_FILE} --close
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/
section_end dut_cleanup
exit $ret

View File

@@ -1,206 +0,0 @@
#!/usr/bin/env python3
#
# Copyright © 2020 Google LLC
# SPDX-License-Identifier: MIT
import argparse
import datetime
import math
import os
import re
import sys
from custom_logger import CustomLogger
from serial_buffer import SerialBuffer
ANSI_ESCAPE="\x1b[0K"
ANSI_COLOUR="\x1b[0;36m"
ANSI_RESET="\x1b[0m"
SECTION_START="start"
SECTION_END="end"
class CrosServoRun:
def __init__(self, cpu, ec, test_timeout, logger):
self.cpu_ser = SerialBuffer(
cpu, "results/serial.txt", ": ")
# Merge the EC serial into the cpu_ser's line stream so that we can
# effectively poll on both at the same time and not have to worry about
self.ec_ser = SerialBuffer(
ec, "results/serial-ec.txt", " EC: ", line_queue=self.cpu_ser.line_queue)
self.test_timeout = test_timeout
self.logger = logger
def close(self):
self.ec_ser.close()
self.cpu_ser.close()
def ec_write(self, s):
print("EC> %s" % s)
self.ec_ser.serial.write(s.encode())
def cpu_write(self, s):
print("> %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)
self.logger.update_status_fail(message)
def get_rel_timestamp(self):
now = datetime.datetime.now(tz=datetime.UTC)
then_env = os.getenv("CI_JOB_STARTED_AT")
if not then_env:
return ""
delta = now - datetime.datetime.fromisoformat(then_env)
return f"[{math.floor(delta.seconds / 60):02}:{(delta.seconds % 60):02}]"
def get_cur_timestamp(self):
return str(int(datetime.datetime.timestamp(datetime.datetime.now())))
def print_gitlab_section(self, action, name, description, collapse=True):
assert action in [SECTION_START, SECTION_END]
out = ANSI_ESCAPE + "section_" + action + ":"
out += self.get_cur_timestamp() + ":"
out += name
if action == "start" and collapse:
out += "[collapsed=true]"
out += "\r" + ANSI_ESCAPE + ANSI_COLOUR
out += self.get_rel_timestamp() + " " + description + ANSI_RESET
print(out)
def boot_section(self, action):
self.print_gitlab_section(action, "dut_boot", "Booting hardware device", True)
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")
bootloader_done = False
self.logger.create_job_phase("boot")
self.boot_section(SECTION_START)
tftp_failures = 0
# 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.cpu_ser.lines(timeout=120, phase="bootloader"):
if re.search("load_archive: loading locale_en.bin", line):
self.cpu_write("\016")
bootloader_done = True
break
# 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. Currently mostly visible on google-freedreno-cheza-14.
if re.search("R8152: Bulk read error 0xffffffbf", line):
tftp_failures += 1
if tftp_failures >= 10:
self.print_error(
"Detected intermittent tftp failure, restarting run.")
return 1
# If the board has a netboot firmware and we made it to booting the
# kernel, proceed to processing of the test run.
if re.search("Booting Linux", line):
bootloader_done = True
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, abandoning run.")
return 1
if not bootloader_done:
self.print_error("Failed to make it through bootloader, abandoning run.")
return 1
self.logger.create_job_phase("test")
for line in self.cpu_ser.lines(timeout=self.test_timeout, phase="test"):
if re.search("---. end Kernel panic", line):
return 1
# 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, abandoning run.")
return 1
# 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, abandoning run.")
return 1
# 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, abandoning run.")
return 1
if re.search("coreboot.*bootblock starting", line):
self.print_error(
"Detected spontaneous reboot, abandoning run.")
return 1
if re.search("arm-smmu 5040000.iommu: TLB sync timed out -- SMMU may be deadlocked", line):
self.print_error("Detected cheza MMU fail, abandoning run.")
return 1
result = re.search(r"hwci: mesa: (\S*), exit_code: (\d+)", line)
if result:
status = result.group(1)
exit_code = int(result.group(2))
if status == "pass":
self.logger.update_dut_job("status", "pass")
else:
self.logger.update_status_fail("test fail")
self.logger.update_dut_job("exit_code", exit_code)
return exit_code
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)
parser.add_argument(
'--test-timeout', type=int, help='Test phase timeout (minutes)', required=True)
args = parser.parse_args()
logger = CustomLogger("results/job_detail.json")
logger.update_dut_time("start", None)
servo = CrosServoRun(args.cpu, args.ec, args.test_timeout * 60, logger)
retval = servo.run()
# power down the CPU on the device
servo.ec_write("power off\n")
logger.update_dut_time("end", None)
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

@@ -1,31 +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";;
*) exit
esac
done
shift $((OPTIND -1))
echo "Waiting for $FILE to say one of following strings"
cat "$STRINGS"
while ! grep -E -wf "$STRINGS" "$FILE"; do
sleep 2
done
if grep -E -wf "$ERRORS" "$FILE"; then
exit 1
fi

View File

@@ -1,171 +0,0 @@
#!/bin/bash
# shellcheck disable=SC1091 # The relative paths in this file only become valid at runtime.
# shellcheck disable=SC2034
# shellcheck disable=SC2086 # we want word splitting
. "$SCRIPTS_DIR"/setup-test-env.sh
BM=$CI_PROJECT_DIR/install/bare-metal
CI_COMMON=$CI_PROJECT_DIR/install/common
if [ -z "$BM_SERIAL" ] && [ -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 $BM_CMDLINE | grep -q "root=/dev/nfs"; then
BM_FASTBOOT_NFSROOT=1
fi
section_start prepare_rootfs "Preparing rootfs components"
set -ex
# Clear out any previous run's artifacts.
rm -rf results/
mkdir -p results/
if [ -n "$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 . | \
grep -E -v "external/(openglcts|vulkancts|amber|glslang|spirv-tools)" |
grep -E -v "traces-db|apitrace|renderdoc" | \
grep -E -v $EXCLUDE_FILTER | \
cpio -H newc -o | \
xz --check=crc32 -T4 - > $CI_PROJECT_DIR/rootfs.cpio.gz
popd
fi
if echo "$BM_KERNEL $BM_DTB" | grep -q http; then
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
"$BM_KERNEL" -o kernel
# FIXME: modules should be supplied too
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
"$BM_DTB" -o dtb
cat kernel dtb > Image.gz-dtb
elif [ -n "${EXTERNAL_KERNEL_TAG}" ]; then
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
"${FDO_HTTP_CACHE_URI:-}${KERNEL_IMAGE_BASE}/${DEBIAN_ARCH}/${BM_KERNEL}" -o kernel
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
"${FDO_HTTP_CACHE_URI:-}${KERNEL_IMAGE_BASE}/${DEBIAN_ARCH}/modules.tar.zst" -o modules.tar.zst
if [ -n "$BM_DTB" ]; then
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
"${FDO_HTTP_CACHE_URI:-}${KERNEL_IMAGE_BASE}/${DEBIAN_ARCH}/${BM_DTB}.dtb" -o dtb
fi
cat kernel dtb > Image.gz-dtb || echo "No DTB available, using pure kernel."
rm kernel
tar --keep-directory-symlink --zstd -xf modules.tar.zst -C "$BM_ROOTFS/"
rm modules.tar.zst &
else
cat /baremetal-files/"$BM_KERNEL" /baremetal-files/"$BM_DTB".dtb > Image.gz-dtb
cp /baremetal-files/"$BM_DTB".dtb dtb
fi
export PATH=$BM:$PATH
mkdir -p artifacts
mkbootimg.py \
--kernel Image.gz-dtb \
--ramdisk rootfs.cpio.gz \
--dtb dtb \
--cmdline "$BM_CMDLINE" \
$BM_MKBOOT_PARAMS \
--header_version 2 \
-o artifacts/fastboot.img
rm Image.gz-dtb dtb
# Start background command for talking to serial if we have one.
if [ -n "$BM_SERIAL_SCRIPT" ]; then
$BM_SERIAL_SCRIPT > results/serial-output.txt &
while [ ! -e results/serial-output.txt ]; do
sleep 1
done
fi
section_end prepare_rootfs
set +e
$BM/fastboot_run.py \
--dev="$BM_SERIAL" \
--test-timeout ${TEST_PHASE_TIMEOUT_MINUTES:-20} \
--fbserial="$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

@@ -1,159 +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 subprocess
import re
from serial_buffer import SerialBuffer
import sys
import threading
class FastbootRun:
def __init__(self, args, test_timeout):
self.powerup = args.powerup
self.ser = SerialBuffer(
args.dev, "results/serial-output.txt", "R SERIAL> ")
self.fastboot = "fastboot boot -s {ser} artifacts/fastboot.img".format(
ser=args.fbserial)
self.test_timeout = test_timeout
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, timeout=60):
print("Running '{}'".format(cmd))
try:
return subprocess.call(cmd, shell=True, timeout=timeout)
except subprocess.TimeoutExpired:
self.print_error("timeout, abandoning run.")
return 1
def run(self):
if ret := self.logged_system(self.powerup):
return ret
fastboot_ready = False
for line in self.ser.lines(timeout=2 * 60, phase="bootloader"):
if re.search("[Ff]astboot: [Pp]rocessing 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, abandoning run.")
return 1
if not fastboot_ready:
self.print_error(
"Failed to get to fastboot prompt, abandoning run.")
return 1
if ret := self.logged_system(self.fastboot):
return ret
print_more_lines = -1
for line in self.ser.lines(timeout=self.test_timeout, phase="test"):
if print_more_lines == 0:
return 1
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, abandoning run.")
return 1
# db820c sometimes wedges around iommu fault recovery
if re.search("watchdog: BUG: soft lockup - CPU.* stuck", line):
self.print_error(
"Detected kernel soft lockup, abandoning run.")
return 1
# 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, abandoning run.")
return 1
# 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, abandoning run.")
if print_more_lines == -1:
print_more_lines = 30
result = re.search(r"hwci: mesa: (\S*), exit_code: (\d+)", line)
if result:
status = result.group(1)
exit_code = int(result.group(2))
return exit_code
self.print_error(
"Reached the end of the CPU serial log without finding a result, abandoning run.")
return 1
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)
parser.add_argument('--test-timeout', type=int,
help='Test phase timeout (minutes)', required=True)
args = parser.parse_args()
fastboot = FastbootRun(args, args.test_timeout * 60)
retval = fastboot.run()
fastboot.close()
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

@@ -1,569 +0,0 @@
#!/usr/bin/env python3
#
# Copyright 2015, The Android Open Source Project
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""Creates the boot image."""
from argparse import (ArgumentParser, ArgumentTypeError,
FileType, RawDescriptionHelpFormatter)
from hashlib import sha1
from os import fstat
from struct import pack
import array
import collections
import os
import re
import subprocess
import tempfile
# Constant and structure definition is in
# system/tools/mkbootimg/include/bootimg/bootimg.h
BOOT_MAGIC = 'ANDROID!'
BOOT_MAGIC_SIZE = 8
BOOT_NAME_SIZE = 16
BOOT_ARGS_SIZE = 512
BOOT_EXTRA_ARGS_SIZE = 1024
BOOT_IMAGE_HEADER_V1_SIZE = 1648
BOOT_IMAGE_HEADER_V2_SIZE = 1660
BOOT_IMAGE_HEADER_V3_SIZE = 1580
BOOT_IMAGE_HEADER_V3_PAGESIZE = 4096
BOOT_IMAGE_HEADER_V4_SIZE = 1584
BOOT_IMAGE_V4_SIGNATURE_SIZE = 4096
VENDOR_BOOT_MAGIC = 'VNDRBOOT'
VENDOR_BOOT_MAGIC_SIZE = 8
VENDOR_BOOT_NAME_SIZE = BOOT_NAME_SIZE
VENDOR_BOOT_ARGS_SIZE = 2048
VENDOR_BOOT_IMAGE_HEADER_V3_SIZE = 2112
VENDOR_BOOT_IMAGE_HEADER_V4_SIZE = 2128
VENDOR_RAMDISK_TYPE_NONE = 0
VENDOR_RAMDISK_TYPE_PLATFORM = 1
VENDOR_RAMDISK_TYPE_RECOVERY = 2
VENDOR_RAMDISK_TYPE_DLKM = 3
VENDOR_RAMDISK_NAME_SIZE = 32
VENDOR_RAMDISK_TABLE_ENTRY_BOARD_ID_SIZE = 16
VENDOR_RAMDISK_TABLE_ENTRY_V4_SIZE = 108
# Names with special meaning, mustn't be specified in --ramdisk_name.
VENDOR_RAMDISK_NAME_BLOCKLIST = {b'default'}
PARSER_ARGUMENT_VENDOR_RAMDISK_FRAGMENT = '--vendor_ramdisk_fragment'
def filesize(f):
if f is None:
return 0
try:
return fstat(f.fileno()).st_size
except OSError:
return 0
def update_sha(sha, f):
if f:
sha.update(f.read())
f.seek(0)
sha.update(pack('I', filesize(f)))
else:
sha.update(pack('I', 0))
def pad_file(f, padding):
pad = (padding - (f.tell() & (padding - 1))) & (padding - 1)
f.write(pack(str(pad) + 'x'))
def get_number_of_pages(image_size, page_size):
"""calculates the number of pages required for the image"""
return (image_size + page_size - 1) // page_size
def get_recovery_dtbo_offset(args):
"""calculates the offset of recovery_dtbo image in the boot image"""
num_header_pages = 1 # header occupies a page
num_kernel_pages = get_number_of_pages(filesize(args.kernel), args.pagesize)
num_ramdisk_pages = get_number_of_pages(filesize(args.ramdisk),
args.pagesize)
num_second_pages = get_number_of_pages(filesize(args.second), args.pagesize)
dtbo_offset = args.pagesize * (num_header_pages + num_kernel_pages +
num_ramdisk_pages + num_second_pages)
return dtbo_offset
def write_header_v3_and_above(args):
if args.header_version > 3:
boot_header_size = BOOT_IMAGE_HEADER_V4_SIZE
else:
boot_header_size = BOOT_IMAGE_HEADER_V3_SIZE
args.output.write(pack(f'{BOOT_MAGIC_SIZE}s', BOOT_MAGIC.encode()))
# kernel size in bytes
args.output.write(pack('I', filesize(args.kernel)))
# ramdisk size in bytes
args.output.write(pack('I', filesize(args.ramdisk)))
# os version and patch level
args.output.write(pack('I', (args.os_version << 11) | args.os_patch_level))
args.output.write(pack('I', boot_header_size))
# reserved
args.output.write(pack('4I', 0, 0, 0, 0))
# version of boot image header
args.output.write(pack('I', args.header_version))
args.output.write(pack(f'{BOOT_ARGS_SIZE + BOOT_EXTRA_ARGS_SIZE}s',
args.cmdline))
if args.header_version >= 4:
# The signature used to verify boot image v4.
args.output.write(pack('I', BOOT_IMAGE_V4_SIGNATURE_SIZE))
pad_file(args.output, BOOT_IMAGE_HEADER_V3_PAGESIZE)
def write_vendor_boot_header(args):
if filesize(args.dtb) == 0:
raise ValueError('DTB image must not be empty.')
if args.header_version > 3:
vendor_ramdisk_size = args.vendor_ramdisk_total_size
vendor_boot_header_size = VENDOR_BOOT_IMAGE_HEADER_V4_SIZE
else:
vendor_ramdisk_size = filesize(args.vendor_ramdisk)
vendor_boot_header_size = VENDOR_BOOT_IMAGE_HEADER_V3_SIZE
args.vendor_boot.write(pack(f'{VENDOR_BOOT_MAGIC_SIZE}s',
VENDOR_BOOT_MAGIC.encode()))
# version of boot image header
args.vendor_boot.write(pack('I', args.header_version))
# flash page size
args.vendor_boot.write(pack('I', args.pagesize))
# kernel physical load address
args.vendor_boot.write(pack('I', args.base + args.kernel_offset))
# ramdisk physical load address
args.vendor_boot.write(pack('I', args.base + args.ramdisk_offset))
# ramdisk size in bytes
args.vendor_boot.write(pack('I', vendor_ramdisk_size))
args.vendor_boot.write(pack(f'{VENDOR_BOOT_ARGS_SIZE}s',
args.vendor_cmdline))
# kernel tags physical load address
args.vendor_boot.write(pack('I', args.base + args.tags_offset))
# asciiz product name
args.vendor_boot.write(pack(f'{VENDOR_BOOT_NAME_SIZE}s', args.board))
# header size in bytes
args.vendor_boot.write(pack('I', vendor_boot_header_size))
# dtb size in bytes
args.vendor_boot.write(pack('I', filesize(args.dtb)))
# dtb physical load address
args.vendor_boot.write(pack('Q', args.base + args.dtb_offset))
if args.header_version > 3:
vendor_ramdisk_table_size = (args.vendor_ramdisk_table_entry_num *
VENDOR_RAMDISK_TABLE_ENTRY_V4_SIZE)
# vendor ramdisk table size in bytes
args.vendor_boot.write(pack('I', vendor_ramdisk_table_size))
# number of vendor ramdisk table entries
args.vendor_boot.write(pack('I', args.vendor_ramdisk_table_entry_num))
# vendor ramdisk table entry size in bytes
args.vendor_boot.write(pack('I', VENDOR_RAMDISK_TABLE_ENTRY_V4_SIZE))
# bootconfig section size in bytes
args.vendor_boot.write(pack('I', filesize(args.vendor_bootconfig)))
pad_file(args.vendor_boot, args.pagesize)
def write_header(args):
if args.header_version > 4:
raise ValueError(
f'Boot header version {args.header_version} not supported')
if args.header_version in {3, 4}:
return write_header_v3_and_above(args)
ramdisk_load_address = ((args.base + args.ramdisk_offset)
if filesize(args.ramdisk) > 0 else 0)
second_load_address = ((args.base + args.second_offset)
if filesize(args.second) > 0 else 0)
args.output.write(pack(f'{BOOT_MAGIC_SIZE}s', BOOT_MAGIC.encode()))
# kernel size in bytes
args.output.write(pack('I', filesize(args.kernel)))
# kernel physical load address
args.output.write(pack('I', args.base + args.kernel_offset))
# ramdisk size in bytes
args.output.write(pack('I', filesize(args.ramdisk)))
# ramdisk physical load address
args.output.write(pack('I', ramdisk_load_address))
# second bootloader size in bytes
args.output.write(pack('I', filesize(args.second)))
# second bootloader physical load address
args.output.write(pack('I', second_load_address))
# kernel tags physical load address
args.output.write(pack('I', args.base + args.tags_offset))
# flash page size
args.output.write(pack('I', args.pagesize))
# version of boot image header
args.output.write(pack('I', args.header_version))
# os version and patch level
args.output.write(pack('I', (args.os_version << 11) | args.os_patch_level))
# asciiz product name
args.output.write(pack(f'{BOOT_NAME_SIZE}s', args.board))
args.output.write(pack(f'{BOOT_ARGS_SIZE}s', args.cmdline))
sha = sha1()
update_sha(sha, args.kernel)
update_sha(sha, args.ramdisk)
update_sha(sha, args.second)
if args.header_version > 0:
update_sha(sha, args.recovery_dtbo)
if args.header_version > 1:
update_sha(sha, args.dtb)
img_id = pack('32s', sha.digest())
args.output.write(img_id)
args.output.write(pack(f'{BOOT_EXTRA_ARGS_SIZE}s', args.extra_cmdline))
if args.header_version > 0:
if args.recovery_dtbo:
# recovery dtbo size in bytes
args.output.write(pack('I', filesize(args.recovery_dtbo)))
# recovert dtbo offset in the boot image
args.output.write(pack('Q', get_recovery_dtbo_offset(args)))
else:
# Set to zero if no recovery dtbo
args.output.write(pack('I', 0))
args.output.write(pack('Q', 0))
# Populate boot image header size for header versions 1 and 2.
if args.header_version == 1:
args.output.write(pack('I', BOOT_IMAGE_HEADER_V1_SIZE))
elif args.header_version == 2:
args.output.write(pack('I', BOOT_IMAGE_HEADER_V2_SIZE))
if args.header_version > 1:
if filesize(args.dtb) == 0:
raise ValueError('DTB image must not be empty.')
# dtb size in bytes
args.output.write(pack('I', filesize(args.dtb)))
# dtb physical load address
args.output.write(pack('Q', args.base + args.dtb_offset))
pad_file(args.output, args.pagesize)
return img_id
class AsciizBytes:
"""Parses a string and encodes it as an asciiz bytes object.
>>> AsciizBytes(bufsize=4)('foo')
b'foo\\x00'
>>> AsciizBytes(bufsize=4)('foob')
Traceback (most recent call last):
...
argparse.ArgumentTypeError: Encoded asciiz length exceeded: max 4, got 5
"""
def __init__(self, bufsize):
self.bufsize = bufsize
def __call__(self, arg):
arg_bytes = arg.encode() + b'\x00'
if len(arg_bytes) > self.bufsize:
raise ArgumentTypeError(
'Encoded asciiz length exceeded: '
f'max {self.bufsize}, got {len(arg_bytes)}')
return arg_bytes
class VendorRamdiskTableBuilder:
"""Vendor ramdisk table builder.
Attributes:
entries: A list of VendorRamdiskTableEntry namedtuple.
ramdisk_total_size: Total size in bytes of all ramdisks in the table.
"""
VendorRamdiskTableEntry = collections.namedtuple( # pylint: disable=invalid-name
'VendorRamdiskTableEntry',
['ramdisk_path', 'ramdisk_size', 'ramdisk_offset', 'ramdisk_type',
'ramdisk_name', 'board_id'])
def __init__(self):
self.entries = []
self.ramdisk_total_size = 0
self.ramdisk_names = set()
def add_entry(self, ramdisk_path, ramdisk_type, ramdisk_name, board_id):
# Strip any trailing null for simple comparison.
stripped_ramdisk_name = ramdisk_name.rstrip(b'\x00')
if stripped_ramdisk_name in VENDOR_RAMDISK_NAME_BLOCKLIST:
raise ValueError(
f'Banned vendor ramdisk name: {stripped_ramdisk_name}')
if stripped_ramdisk_name in self.ramdisk_names:
raise ValueError(
f'Duplicated vendor ramdisk name: {stripped_ramdisk_name}')
self.ramdisk_names.add(stripped_ramdisk_name)
if board_id is None:
board_id = array.array(
'I', [0] * VENDOR_RAMDISK_TABLE_ENTRY_BOARD_ID_SIZE)
else:
board_id = array.array('I', board_id)
if len(board_id) != VENDOR_RAMDISK_TABLE_ENTRY_BOARD_ID_SIZE:
raise ValueError('board_id size must be '
f'{VENDOR_RAMDISK_TABLE_ENTRY_BOARD_ID_SIZE}')
with open(ramdisk_path, 'rb') as f:
ramdisk_size = filesize(f)
self.entries.append(self.VendorRamdiskTableEntry(
ramdisk_path, ramdisk_size, self.ramdisk_total_size, ramdisk_type,
ramdisk_name, board_id))
self.ramdisk_total_size += ramdisk_size
def write_ramdisks_padded(self, fout, alignment):
for entry in self.entries:
with open(entry.ramdisk_path, 'rb') as f:
fout.write(f.read())
pad_file(fout, alignment)
def write_entries_padded(self, fout, alignment):
for entry in self.entries:
fout.write(pack('I', entry.ramdisk_size))
fout.write(pack('I', entry.ramdisk_offset))
fout.write(pack('I', entry.ramdisk_type))
fout.write(pack(f'{VENDOR_RAMDISK_NAME_SIZE}s',
entry.ramdisk_name))
fout.write(entry.board_id)
pad_file(fout, alignment)
def write_padded_file(f_out, f_in, padding):
if f_in is None:
return
f_out.write(f_in.read())
pad_file(f_out, padding)
def parse_int(x):
return int(x, 0)
def parse_os_version(x):
match = re.search(r'^(\d{1,3})(?:\.(\d{1,3})(?:\.(\d{1,3}))?)?', x)
if match:
a = int(match.group(1))
b = c = 0
if match.lastindex >= 2:
b = int(match.group(2))
if match.lastindex == 3:
c = int(match.group(3))
# 7 bits allocated for each field
assert a < 128
assert b < 128
assert c < 128
return (a << 14) | (b << 7) | c
return 0
def parse_os_patch_level(x):
match = re.search(r'^(\d{4})-(\d{2})(?:-(\d{2}))?', x)
if match:
y = int(match.group(1)) - 2000
m = int(match.group(2))
# 7 bits allocated for the year, 4 bits for the month
assert 0 <= y < 128
assert 0 < m <= 12
return (y << 4) | m
return 0
def parse_vendor_ramdisk_type(x):
type_dict = {
'none': VENDOR_RAMDISK_TYPE_NONE,
'platform': VENDOR_RAMDISK_TYPE_PLATFORM,
'recovery': VENDOR_RAMDISK_TYPE_RECOVERY,
'dlkm': VENDOR_RAMDISK_TYPE_DLKM,
}
if x.lower() in type_dict:
return type_dict[x.lower()]
return parse_int(x)
def get_vendor_boot_v4_usage():
return """vendor boot version 4 arguments:
--ramdisk_type {none,platform,recovery,dlkm}
specify the type of the ramdisk
--ramdisk_name NAME
specify the name of the ramdisk
--board_id{0..15} NUMBER
specify the value of the board_id vector, defaults to 0
--vendor_ramdisk_fragment VENDOR_RAMDISK_FILE
path to the vendor ramdisk file
These options can be specified multiple times, where each vendor ramdisk
option group ends with a --vendor_ramdisk_fragment option.
Each option group appends an additional ramdisk to the vendor boot image.
"""
def parse_vendor_ramdisk_args(args, args_list):
"""Parses vendor ramdisk specific arguments.
Args:
args: An argparse.Namespace object. Parsed results are stored into this
object.
args_list: A list of argument strings to be parsed.
Returns:
A list argument strings that are not parsed by this method.
"""
parser = ArgumentParser(add_help=False)
parser.add_argument('--ramdisk_type', type=parse_vendor_ramdisk_type,
default=VENDOR_RAMDISK_TYPE_NONE)
parser.add_argument('--ramdisk_name',
type=AsciizBytes(bufsize=VENDOR_RAMDISK_NAME_SIZE),
required=True)
for i in range(VENDOR_RAMDISK_TABLE_ENTRY_BOARD_ID_SIZE):
parser.add_argument(f'--board_id{i}', type=parse_int, default=0)
parser.add_argument(PARSER_ARGUMENT_VENDOR_RAMDISK_FRAGMENT, required=True)
unknown_args = []
vendor_ramdisk_table_builder = VendorRamdiskTableBuilder()
if args.vendor_ramdisk is not None:
vendor_ramdisk_table_builder.add_entry(
args.vendor_ramdisk.name, VENDOR_RAMDISK_TYPE_PLATFORM, b'', None)
while PARSER_ARGUMENT_VENDOR_RAMDISK_FRAGMENT in args_list:
idx = args_list.index(PARSER_ARGUMENT_VENDOR_RAMDISK_FRAGMENT) + 2
vendor_ramdisk_args = args_list[:idx]
args_list = args_list[idx:]
ramdisk_args, extra_args = parser.parse_known_args(vendor_ramdisk_args)
ramdisk_args_dict = vars(ramdisk_args)
unknown_args.extend(extra_args)
ramdisk_path = ramdisk_args.vendor_ramdisk_fragment
ramdisk_type = ramdisk_args.ramdisk_type
ramdisk_name = ramdisk_args.ramdisk_name
board_id = [ramdisk_args_dict[f'board_id{i}']
for i in range(VENDOR_RAMDISK_TABLE_ENTRY_BOARD_ID_SIZE)]
vendor_ramdisk_table_builder.add_entry(ramdisk_path, ramdisk_type,
ramdisk_name, board_id)
if len(args_list) > 0:
unknown_args.extend(args_list)
args.vendor_ramdisk_total_size = (vendor_ramdisk_table_builder
.ramdisk_total_size)
args.vendor_ramdisk_table_entry_num = len(vendor_ramdisk_table_builder
.entries)
args.vendor_ramdisk_table_builder = vendor_ramdisk_table_builder
return unknown_args
def parse_cmdline():
version_parser = ArgumentParser(add_help=False)
version_parser.add_argument('--header_version', type=parse_int, default=0)
if version_parser.parse_known_args()[0].header_version < 3:
# For boot header v0 to v2, the kernel commandline field is split into
# two fields, cmdline and extra_cmdline. Both fields are asciiz strings,
# so we minus one here to ensure the encoded string plus the
# null-terminator can fit in the buffer size.
cmdline_size = BOOT_ARGS_SIZE + BOOT_EXTRA_ARGS_SIZE - 1
else:
cmdline_size = BOOT_ARGS_SIZE + BOOT_EXTRA_ARGS_SIZE
parser = ArgumentParser(formatter_class=RawDescriptionHelpFormatter,
epilog=get_vendor_boot_v4_usage())
parser.add_argument('--kernel', type=FileType('rb'),
help='path to the kernel')
parser.add_argument('--ramdisk', type=FileType('rb'),
help='path to the ramdisk')
parser.add_argument('--second', type=FileType('rb'),
help='path to the second bootloader')
parser.add_argument('--dtb', type=FileType('rb'), help='path to the dtb')
dtbo_group = parser.add_mutually_exclusive_group()
dtbo_group.add_argument('--recovery_dtbo', type=FileType('rb'),
help='path to the recovery DTBO')
dtbo_group.add_argument('--recovery_acpio', type=FileType('rb'),
metavar='RECOVERY_ACPIO', dest='recovery_dtbo',
help='path to the recovery ACPIO')
parser.add_argument('--cmdline', type=AsciizBytes(bufsize=cmdline_size),
default='', help='kernel command line arguments')
parser.add_argument('--vendor_cmdline',
type=AsciizBytes(bufsize=VENDOR_BOOT_ARGS_SIZE),
default='',
help='vendor boot kernel command line arguments')
parser.add_argument('--base', type=parse_int, default=0x10000000,
help='base address')
parser.add_argument('--kernel_offset', type=parse_int, default=0x00008000,
help='kernel offset')
parser.add_argument('--ramdisk_offset', type=parse_int, default=0x01000000,
help='ramdisk offset')
parser.add_argument('--second_offset', type=parse_int, default=0x00f00000,
help='second bootloader offset')
parser.add_argument('--dtb_offset', type=parse_int, default=0x01f00000,
help='dtb offset')
parser.add_argument('--os_version', type=parse_os_version, default=0,
help='operating system version')
parser.add_argument('--os_patch_level', type=parse_os_patch_level,
default=0, help='operating system patch level')
parser.add_argument('--tags_offset', type=parse_int, default=0x00000100,
help='tags offset')
parser.add_argument('--board', type=AsciizBytes(bufsize=BOOT_NAME_SIZE),
default='', help='board name')
parser.add_argument('--pagesize', type=parse_int,
choices=[2**i for i in range(11, 15)], default=2048,
help='page size')
parser.add_argument('--id', action='store_true',
help='print the image ID on standard output')
parser.add_argument('--header_version', type=parse_int, default=0,
help='boot image header version')
parser.add_argument('-o', '--output', type=FileType('wb'),
help='output file name')
parser.add_argument('--gki_signing_algorithm',
help='GKI signing algorithm to use')
parser.add_argument('--gki_signing_key',
help='path to RSA private key file')
parser.add_argument('--gki_signing_signature_args',
help='other hash arguments passed to avbtool')
parser.add_argument('--gki_signing_avbtool_path',
help='path to avbtool for boot signature generation')
parser.add_argument('--vendor_boot', type=FileType('wb'),
help='vendor boot output file name')
parser.add_argument('--vendor_ramdisk', type=FileType('rb'),
help='path to the vendor ramdisk')
parser.add_argument('--vendor_bootconfig', type=FileType('rb'),
help='path to the vendor bootconfig file')
args, extra_args = parser.parse_known_args()
if args.vendor_boot is not None and args.header_version > 3:
extra_args = parse_vendor_ramdisk_args(args, extra_args)
if len(extra_args) > 0:
raise ValueError(f'Unrecognized arguments: {extra_args}')
if args.header_version < 3:
args.extra_cmdline = args.cmdline[BOOT_ARGS_SIZE-1:]
args.cmdline = args.cmdline[:BOOT_ARGS_SIZE-1] + b'\x00'
assert len(args.cmdline) <= BOOT_ARGS_SIZE
assert len(args.extra_cmdline) <= BOOT_EXTRA_ARGS_SIZE
return args
def add_boot_image_signature(args, pagesize):
"""Adds the boot image signature.
Note that the signature will only be verified in VTS to ensure a
generic boot.img is used. It will not be used by the device
bootloader at boot time. The bootloader should only verify
the boot vbmeta at the end of the boot partition (or in the top-level
vbmeta partition) via the Android Verified Boot process, when the
device boots.
"""
args.output.flush() # Flush the buffer for signature calculation.
# Appends zeros if the signing key is not specified.
if not args.gki_signing_key or not args.gki_signing_algorithm:
zeros = b'\x00' * BOOT_IMAGE_V4_SIGNATURE_SIZE
args.output.write(zeros)
pad_file(args.output, pagesize)
return
avbtool = 'avbtool' # Used from otatools.zip or Android build env.
# We need to specify the path of avbtool in build/core/Makefile.
# Because avbtool is not guaranteed to be in $PATH there.
if args.gki_signing_avbtool_path:
avbtool = args.gki_signing_avbtool_path
# Need to specify a value of --partition_size for avbtool to work.
# We use 64 MB below, but avbtool will not resize the boot image to
# this size because --do_not_append_vbmeta_image is also specified.
avbtool_cmd = [
avbtool, 'add_hash_footer',
'--partition_name', 'boot',
'--partition_size', str(64 * 1024 * 1024),
'--image', args.output.name,
'--algorithm', args.gki_signing_algorithm,
'--key', args.gki_signing_key,
'--salt', 'd00df00d'] # TODO: use a hash of kernel/ramdisk as the salt.
# Additional arguments passed to avbtool.
if args.gki_signing_signature_args:
avbtool_cmd += args.gki_signing_signature_args.split()
# Outputs the signed vbmeta to a separate file, then append to boot.img
# as the boot signature.
with tempfile.TemporaryDirectory() as temp_out_dir:
boot_signature_output = os.path.join(temp_out_dir, 'boot_signature')
avbtool_cmd += ['--do_not_append_vbmeta_image',
'--output_vbmeta_image', boot_signature_output]
subprocess.check_call(avbtool_cmd)
with open(boot_signature_output, 'rb') as boot_signature:
if filesize(boot_signature) > BOOT_IMAGE_V4_SIGNATURE_SIZE:
raise ValueError(
f'boot sigature size is > {BOOT_IMAGE_V4_SIGNATURE_SIZE}')
write_padded_file(args.output, boot_signature, pagesize)
def write_data(args, pagesize):
write_padded_file(args.output, args.kernel, pagesize)
write_padded_file(args.output, args.ramdisk, pagesize)
write_padded_file(args.output, args.second, pagesize)
if args.header_version > 0 and args.header_version < 3:
write_padded_file(args.output, args.recovery_dtbo, pagesize)
if args.header_version == 2:
write_padded_file(args.output, args.dtb, pagesize)
if args.header_version >= 4:
add_boot_image_signature(args, pagesize)
def write_vendor_boot_data(args):
if args.header_version > 3:
builder = args.vendor_ramdisk_table_builder
builder.write_ramdisks_padded(args.vendor_boot, args.pagesize)
write_padded_file(args.vendor_boot, args.dtb, args.pagesize)
builder.write_entries_padded(args.vendor_boot, args.pagesize)
write_padded_file(args.vendor_boot, args.vendor_bootconfig,
args.pagesize)
else:
write_padded_file(args.vendor_boot, args.vendor_ramdisk, args.pagesize)
write_padded_file(args.vendor_boot, args.dtb, args.pagesize)
def main():
args = parse_cmdline()
if args.vendor_boot is not None:
if args.header_version not in {3, 4}:
raise ValueError(
'--vendor_boot not compatible with given header version')
if args.header_version == 3 and args.vendor_ramdisk is None:
raise ValueError('--vendor_ramdisk missing or invalid')
write_vendor_boot_header(args)
write_vendor_boot_data(args)
if args.output is not None:
if args.second is not None and args.header_version > 2:
raise ValueError(
'--second not compatible with given header version')
img_id = write_header(args)
if args.header_version > 2:
write_data(args, BOOT_IMAGE_HEADER_V3_PAGESIZE)
else:
write_data(args, args.pagesize)
if args.id and img_id is not None:
print('0x' + ''.join(f'{octet:02x}' for octet in img_id))
if __name__ == '__main__':
main()

View File

@@ -1,16 +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.$((${BM_POE_BASE:-0} + BM_POE_INTERFACE))"
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.$((${BM_POE_BASE:-0} + 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

@@ -1,236 +0,0 @@
#!/bin/bash
# shellcheck disable=SC1091
# shellcheck disable=SC2034
# shellcheck disable=SC2059
# shellcheck disable=SC2086 # we want word splitting
. "$SCRIPTS_DIR"/setup-test-env.sh
# Boot script for devices attached to a PoE switch, using NFS for the root
# filesystem.
# We're run from the root of the repo, make a helper var for our paths
BM=$CI_PROJECT_DIR/install/bare-metal
CI_COMMON=$CI_PROJECT_DIR/install/common
CI_INSTALL=$CI_PROJECT_DIR/install
# 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 "$BM_POE_INTERFACE" ]; then
echo "Must set BM_POE_INTERFACE in your gitlab-runner config.toml [[runners]] environment"
echo "This is the PoE switch interface where the device is connected."
exit 1
fi
if [ -z "$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 "$BM_BOOTFS" ] && { [ -z "$BM_KERNEL" ] || [ -z "$BM_DTB" ]; } ; then
echo "Must set /boot files for the TFTP boot in the job's variables or set kernel and dtb"
exit 1
fi
if [ -z "$BM_CMDLINE" ]; then
echo "Must set BM_CMDLINE to your board's kernel command line arguments"
exit 1
fi
section_start prepare_rootfs "Preparing rootfs components"
set -ex
date +'%F %T'
# 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/
date +'%F %T'
# If BM_BOOTFS is an URL, download it
if echo $BM_BOOTFS | grep -q http; then
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
"${FDO_HTTP_CACHE_URI:-}$BM_BOOTFS" -o /tmp/bootfs.tar
BM_BOOTFS=/tmp/bootfs.tar
fi
date +'%F %T'
# If BM_BOOTFS is a file, assume it is a tarball and uncompress it
if [ -f "${BM_BOOTFS}" ]; then
mkdir -p /tmp/bootfs
tar xf $BM_BOOTFS -C /tmp/bootfs
BM_BOOTFS=/tmp/bootfs
fi
# If BM_KERNEL and BM_DTS is present
if [ -n "${EXTERNAL_KERNEL_TAG}" ]; then
if [ -z "${BM_KERNEL}" ] || [ -z "${BM_DTB}" ]; then
echo "This machine cannot be tested with external kernel since BM_KERNEL or BM_DTB missing!"
exit 1
fi
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
"${FDO_HTTP_CACHE_URI:-}${KERNEL_IMAGE_BASE}/${DEBIAN_ARCH}/${BM_KERNEL}" -o "${BM_KERNEL}"
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
"${FDO_HTTP_CACHE_URI:-}${KERNEL_IMAGE_BASE}/${DEBIAN_ARCH}/${BM_DTB}.dtb" -o "${BM_DTB}.dtb"
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
"${FDO_HTTP_CACHE_URI:-}${KERNEL_IMAGE_BASE}/${DEBIAN_ARCH}/modules.tar.zst" -o modules.tar.zst
fi
date +'%F %T'
# Install kernel modules (it could be either in /lib/modules or
# /usr/lib/modules, but we want to install in the latter)
if [ -n "${EXTERNAL_KERNEL_TAG}" ]; then
tar --keep-directory-symlink --zstd -xf modules.tar.zst -C /nfs/
rm modules.tar.zst &
elif [ -n "${BM_BOOTFS}" ]; then
[ -d $BM_BOOTFS/usr/lib/modules ] && rsync -a $BM_BOOTFS/usr/lib/modules/ /nfs/usr/lib/modules/
[ -d $BM_BOOTFS/lib/modules ] && rsync -a $BM_BOOTFS/lib/modules/ /nfs/lib/modules/
else
echo "No modules!"
fi
date +'%F %T'
# Install kernel image + bootloader files
if [ -n "${EXTERNAL_KERNEL_TAG}" ] || [ -z "$BM_BOOTFS" ]; then
mv "${BM_KERNEL}" "${BM_DTB}.dtb" /tftp/
else # BM_BOOTFS
rsync -aL --delete $BM_BOOTFS/boot/ /tftp/
fi
date +'%F %T'
# Set up the pxelinux config for Jetson Nano
mkdir -p /tftp/pxelinux.cfg
cat <<EOF >/tftp/pxelinux.cfg/default-arm-tegra210-p3450-0000
PROMPT 0
TIMEOUT 30
DEFAULT primary
MENU TITLE jetson nano boot options
LABEL primary
MENU LABEL CI kernel on TFTP
LINUX Image
FDT tegra210-p3450-0000.dtb
APPEND \${cbootargs} $BM_CMDLINE
EOF
# Set up the pxelinux config for Jetson TK1
cat <<EOF >/tftp/pxelinux.cfg/default-arm-tegra124-jetson-tk1
PROMPT 0
TIMEOUT 30
DEFAULT primary
MENU TITLE jetson TK1 boot options
LABEL primary
MENU LABEL CI kernel on TFTP
LINUX zImage
FDT tegra124-jetson-tk1.dtb
APPEND \${cbootargs} $BM_CMDLINE
EOF
# Create the rootfs in the NFS directory
. $BM/rootfs-setup.sh /nfs
date +'%F %T'
echo "$BM_CMDLINE" > /tftp/cmdline.txt
# Add some options in config.txt, if defined
if [ -n "$BM_BOOTCONFIG" ]; then
printf "$BM_BOOTCONFIG" >> /tftp/config.txt
fi
section_end prepare_rootfs
set +e
STRUCTURED_LOG_FILE=results/job_detail.json
python3 $CI_INSTALL/custom_logger.py ${STRUCTURED_LOG_FILE} --update dut_job_type "${DEVICE_TYPE}"
python3 $CI_INSTALL/custom_logger.py ${STRUCTURED_LOG_FILE} --update farm "${FARM}"
ATTEMPTS=3
first_attempt=True
while [ $((ATTEMPTS--)) -gt 0 ]; do
section_start dut_boot "Booting hardware device ..."
python3 $CI_INSTALL/custom_logger.py ${STRUCTURED_LOG_FILE} --create-dut-job dut_name "${CI_RUNNER_DESCRIPTION}"
# Update subtime time to CI_JOB_STARTED_AT only for the first run
if [ "$first_attempt" = "True" ]; then
python3 $CI_INSTALL/custom_logger.py ${STRUCTURED_LOG_FILE} --update-dut-time submit "${CI_JOB_STARTED_AT}"
else
python3 $CI_INSTALL/custom_logger.py ${STRUCTURED_LOG_FILE} --update-dut-time submit
fi
python3 $BM/poe_run.py \
--dev="$BM_SERIAL" \
--powerup="$BM_POWERUP" \
--powerdown="$BM_POWERDOWN" \
--boot-timeout-seconds ${BOOT_PHASE_TIMEOUT_SECONDS:-300} \
--test-timeout-minutes ${TEST_PHASE_TIMEOUT_MINUTES:-$((CI_JOB_TIMEOUT/60 - ${TEST_SETUP_AND_UPLOAD_MARGIN_MINUTES:-5}))}
ret=$?
if [ $ret -eq 2 ]; then
python3 $CI_INSTALL/custom_logger.py ${STRUCTURED_LOG_FILE} --close-dut-job
first_attempt=False
error "Device failed to boot; will retry"
else
# We're no longer in dut_boot by this point
unset CURRENT_SECTION
ATTEMPTS=0
fi
done
section_start dut_cleanup "Cleaning up after job"
python3 $CI_INSTALL/custom_logger.py ${STRUCTURED_LOG_FILE} --close-dut-job
python3 $CI_INSTALL/custom_logger.py ${STRUCTURED_LOG_FILE} --close
set -e
date +'%F %T'
# Bring artifacts back from the NFS dir to the build dir where gitlab-runner
# will look for them.
cp -Rp /nfs/results/. results/
date +'%F %T'
section_end dut_cleanup
exit $ret

View File

@@ -1,134 +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
import sys
import threading
from custom_logger import CustomLogger
from serial_buffer import SerialBuffer
class PoERun:
def __init__(self, args, boot_timeout, test_timeout, logger):
self.powerup = args.powerup
self.powerdown = args.powerdown
self.ser = SerialBuffer(
args.dev, "results/serial-output.txt", ": ")
self.boot_timeout = boot_timeout
self.test_timeout = test_timeout
self.logger = logger
def print_error(self, message):
RED = '\033[0;31m'
NO_COLOR = '\033[0m'
print(RED + message + NO_COLOR)
self.logger.update_status_fail(message)
def logged_system(self, cmd):
print("Running '{}'".format(cmd))
return os.system(cmd)
def run(self):
if self.logged_system(self.powerup) != 0:
self.logger.update_status_fail("powerup failed")
return 1
boot_detected = False
self.logger.create_job_phase("boot")
for line in self.ser.lines(timeout=self.boot_timeout, phase="bootloader"):
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
self.logger.create_job_phase("test")
for line in self.ser.lines(timeout=self.test_timeout, phase="test"):
if re.search("---. end Kernel panic", line):
self.logger.update_status_fail("kernel panic")
return 1
# Binning memory problems
if re.search("binner overflow mem", line):
self.print_error("Memory overflow in the binner; GPU hang")
return 1
if re.search("nouveau 57000000.gpu: bus: MMIO read of 00000000 FAULT at 137000", line):
self.print_error("nouveau jetson boot bug, abandoning run.")
return 1
# network fail on tk1
if re.search("NETDEV WATCHDOG:.* transmit queue 0 timed out", line):
self.print_error("nouveau jetson tk1 network fail, abandoning run.")
return 1
result = re.search(r"hwci: mesa: (\S*), exit_code: (\d+)", line)
if result:
status = result.group(1)
exit_code = int(result.group(2))
if status == "pass":
self.logger.update_dut_job("status", "pass")
else:
self.logger.update_status_fail("test fail")
self.logger.update_dut_job("exit_code", exit_code)
return exit_code
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('--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(
'--boot-timeout-seconds', type=int, help='Boot phase timeout (seconds)', required=True)
parser.add_argument(
'--test-timeout-minutes', type=int, help='Test phase timeout (minutes)', required=True)
args = parser.parse_args()
logger = CustomLogger("results/job_detail.json")
logger.update_dut_time("start", None)
poe = PoERun(args, args.boot_timeout_seconds, args.test_timeout_minutes * 60, logger)
retval = poe.run()
poe.logged_system(args.powerdown)
logger.update_dut_time("end", None)
sys.exit(retval)
if __name__ == '__main__':
main()

View File

@@ -1,34 +0,0 @@
#!/usr/bin/env bash
# shellcheck disable=SC2086 # we want word splitting
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 $rootfs_dst/
date +'%F %T'
# Make JWT token available as file in the bare-metal storage to enable access
# to MinIO
cp "${S3_JWT_FILE}" "${rootfs_dst}${S3_JWT_FILE}"
date +'%F %T'
cp "$SCRIPTS_DIR/setup-test-env.sh" "$rootfs_dst/"
set +x
# Pass through relevant env vars from the gitlab job to the baremetal init script
echo "Variables passed through:"
"$CI_COMMON"/generate-env.sh | tee $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/
date +'%F %T'

View File

@@ -1,186 +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, UTC
import queue
import serial
import threading
import time
class SerialBuffer:
def __init__(self, dev, filename, prefix, timeout=None, line_queue=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()
# allow multiple SerialBuffers to share a line queue so you can merge
# servo's CPU and EC streams into one thing to watch the boot/test
# progress on.
if line_queue:
self.line_queue = line_queue
else:
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")
ts = datetime.now(tz=UTC)
ts_str = f"{ts.hour:02}:{ts.minute:02}:{ts.second:02}.{int(ts.microsecond / 1000):03}"
print("{endc}{time}{prefix}{line}".format(
time=ts_str, prefix=self.prefix, line=line, endc='\033[0m'), flush=True, end='')
self.line_queue.put(line)
line = bytearray()
def lines(self, timeout=None, phase=None):
start_time = time.monotonic()
while True:
read_timeout = None
if timeout:
read_timeout = timeout - (time.monotonic() - start_time)
if read_timeout <= 0:
print("read timeout waiting for serial during {}".format(phase))
self.close()
break
try:
line = self.line_queue.get(timeout=read_timeout)
except queue.Empty:
print("read timeout waiting for serial during {}".format(phase))
self.close()
break
if line == self.sentinel:
print("End of serial output")
self.lines_thread.join()
break
yield line
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 +0,0 @@
../bin/ci

View File

@@ -1,906 +0,0 @@
# Shared between windows and Linux
.build-common:
extends: .container+build-rules
# Cancel job if a newer commit is pushed to the same branch
interruptible: true
variables:
# Build jobs don't take more than 1-3 minutes. 5-8 min max on a fresh runner
# without a populated ccache.
# These jobs are never slow, either they finish within reasonable time or
# something has gone wrong and the job will never terminate, so we should
# instead timeout so that the retry mechanism can kick in.
# A few exception are made, see overrides in the rest of this file.
BUILD_JOB_TIMEOUT: 15m
RUN_MESON_TESTS: "true"
timeout: 1h
# We don't want to download any previous job's artifacts
dependencies: []
artifacts:
name: "${CI_PROJECT_NAME}_${CI_JOB_NAME}"
when: always
paths:
- _build/meson-logs/*.txt
- _build/meson-logs/strace
- _build/.ninja_log
- artifacts
# Just Linux
.build-linux:
extends: .build-common
variables:
CCACHE_COMPILERCHECK: "content"
CCACHE_COMPRESS: "true"
CCACHE_DIR: /cache/mesa/ccache
# Use ccache transparently, and print stats before/after
before_script:
- !reference [default, before_script]
- |
export PATH="/usr/lib/ccache:$PATH"
export CCACHE_BASEDIR="$PWD"
if test -x /usr/bin/ccache; then
section_start ccache_before "ccache stats before build"
ccache --show-stats
section_end ccache_before
fi
after_script:
- if test -x /usr/bin/ccache; then ccache --show-stats | grep "Hits:"; fi
- !reference [default, after_script]
.build-windows:
extends:
- .build-common
- .windows-docker-tags
cache:
key: ${CI_JOB_NAME}
paths:
- subprojects/packagecache
.meson-build:
extends:
- .build-linux
- .use-debian/x86_64_build
stage: build-only
script:
- &meson-build timeout --verbose ${BUILD_JOB_TIMEOUT_OVERRIDE:-$BUILD_JOB_TIMEOUT} .gitlab-ci/meson/build.sh
# Make sure this list stays the same as all the jobs with
# `stage: build-for-tests`, except for the windows job as
# explained below.
.build-for-tests-jobs:
- job: debian-testing
optional: true
- job: debian-testing-asan
optional: true
- job: debian-testing-ubsan
optional: true
- job: debian-build-testing
optional: true
- job: debian-arm32
optional: true
- job: debian-arm32-asan
optional: true
- job: debian-arm64
optional: true
- job: debian-arm64-asan
optional: true
# Windows runners don't have more than one build right now, so there is
# no need to wait on the "first one" to be done.
# - job: windows-msvc
# optional: true
- job: python-artifacts
optional: true
debian-testing:
extends:
- .meson-build
- .ci-deqp-artifacts
stage: build-for-tests
variables:
BUILD_JOB_TIMEOUT: 30m
UNWIND: "enabled"
DRI_LOADERS: >
-D egl=enabled
-D gbm=enabled
-D glvnd=disabled
-D glx=dri
-D platforms=x11,wayland
GALLIUM_ST: >
-D gallium-nine=false
-D gallium-rusticl=true
-D gallium-va=enabled
GALLIUM_DRIVERS: "llvmpipe,softpipe,virgl,radeonsi,zink,iris,svga"
VULKAN_DRIVERS: "swrast,amd,intel,virtio"
BUILDTYPE: "debugoptimized"
EXTRA_OPTION: >
-D intel-elk=false
-D spirv-to-dxil=true
-D tools=drm-shim
-D valgrind=disabled
S3_ARTIFACT_NAME: mesa-x86_64-default-${BUILDTYPE}
RUN_MESON_TESTS: "false" # debian-build-testing already runs these
script:
- *meson-build
- .gitlab-ci/prepare-artifacts.sh
artifacts:
reports:
junit: artifacts/ci_scripts_report.xml
debian-testing-asan:
extends:
- debian-testing
stage: build-for-tests
variables:
BUILD_JOB_TIMEOUT: 30m
VULKAN_DRIVERS: "swrast"
GALLIUM_DRIVERS: "llvmpipe,softpipe"
C_ARGS: >
-Wno-error=stringop-truncation
EXTRA_OPTION: >
-D b_sanitize=address
-D gallium-va=false
-D gallium-nine=false
-D gallium-rusticl=false
-D mesa-clc=system
-D tools=dlclose-skip
-D valgrind=disabled
S3_ARTIFACT_NAME: ""
ARTIFACTS_DEBUG_SYMBOLS: 1
RUN_MESON_TESTS: "false" # just too slow
# Do a host build for mesa-clc (asan complains not being loaded as
# the first library)
HOST_BUILD_OPTIONS: >
-D build-tests=false
-D enable-glcpp-tests=false
-D gallium-opencl=disabled
-D gallium-rusticl=false
-D gallium-nine=false
-D gallium-drivers=
-D glx=disabled
-D install-mesa-clc=true
-D mesa-clc=enabled
-D platforms=
-D video-codecs=
-D vulkan-drivers=
debian-testing-msan:
# https://github.com/google/sanitizers/wiki/MemorySanitizerLibcxxHowTo
# msan cannot fully work until it's used together with msan libc
extends:
- debian-clang
# `needs:` inherited from debian-clang
variables:
BUILD_JOB_TIMEOUT: 30m
# l_undef is incompatible with msan
EXTRA_OPTION:
-D b_sanitize=memory
-D b_lundef=false
-D mesa-clc=system
S3_ARTIFACT_NAME: ""
ARTIFACTS_DEBUG_SYMBOLS: 1
# Don't run all the tests yet:
# GLSL has some issues in sexpression reading.
# gtest has issues in its test initialization.
MESON_TEST_ARGS: "--suite glcpp --suite format"
GALLIUM_DRIVERS: "freedreno,iris,nouveau,r300,r600,llvmpipe,softpipe,svga,v3d,vc4,virgl,etnaviv,panfrost,lima,zink,radeonsi,tegra,d3d12,crocus"
VULKAN_DRIVERS: intel,amd,broadcom,virtio
RUN_MESON_TESTS: "false" # just too slow
# Do a host build for mesa-clc (msan complains about uninitialized
# values in the LLVM libs)
HOST_BUILD_OPTIONS: >
-D build-tests=false
-D enable-glcpp-tests=false
-D gallium-opencl=disabled
-D gallium-drivers=
-D vulkan-drivers=
-D video-codecs=
-D glx=disabled
-D platforms=
-D mesa-clc=enabled
-D install-mesa-clc=true
debian-testing-ubsan:
extends:
- debian-testing
stage: build-for-tests
timeout: 40m
variables:
BUILD_JOB_TIMEOUT: 30m
C_ARGS: >
-Wno-error=stringop-overflow
-Wno-error=stringop-truncation
CPP_ARGS: >
-Wno-error=array-bounds
GALLIUM_DRIVERS: "swrast"
VULKAN_DRIVERS: "swrast"
EXTRA_OPTION: >
-D b_sanitize=undefined
-D mesa-clc=system
-D gallium-rusticl=false
-D gallium-va=false
-D gallium-nine=false
S3_ARTIFACT_NAME: ""
ARTIFACTS_DEBUG_SYMBOLS: 1
RUN_MESON_TESTS: "false" # just too slow
HOST_BUILD_OPTIONS: >
-D build-tests=false
-D enable-glcpp-tests=false
-D gallium-opencl=disabled
-D gallium-rusticl=false
-D gallium-drivers=
-D vulkan-drivers=
-D video-codecs=
-D glx=disabled
-D platforms=
-D mesa-clc=enabled
-D install-mesa-clc=true
debian-build-testing:
extends: .meson-build
stage: build-for-tests
variables:
UNWIND: "enabled"
DRI_LOADERS: >
-D glx=dri
-D gbm=enabled
-D egl=enabled
-D glvnd=disabled
-D platforms=x11,wayland
GALLIUM_ST: >
-D gallium-extra-hud=true
-D gallium-vdpau=enabled
-D gallium-va=enabled
-D gallium-xa=enabled
-D gallium-nine=true
-D gallium-rusticl=false
GALLIUM_DRIVERS: "i915,iris,nouveau,r300,r600,freedreno,llvmpipe,softpipe,svga,v3d,vc4,virgl,etnaviv,panfrost,lima,zink,d3d12,asahi,crocus"
VULKAN_DRIVERS: "intel_hasvk,imagination-experimental,microsoft-experimental,nouveau,swrast"
BUILDTYPE: "debugoptimized"
EXTRA_OPTION: >
-D spirv-to-dxil=true
-D osmesa=true
-D tools=drm-shim,etnaviv,freedreno,glsl,intel,intel-ui,nir,nouveau,lima,panfrost,asahi
-D perfetto=true
S3_ARTIFACT_NAME: debian-build-testing
script:
- *meson-build
- .gitlab-ci/prepare-artifacts.sh
shader-db:
stage: code-validation
extends:
- .use-debian/x86_64_build
rules:
- !reference [.never-post-merge-rules, rules]
- !reference [.core-rules, rules]
# Keep this list in sync with the drivers tested in run-shader-db.sh
- !reference [.freedreno-common-rules, rules]
- !reference [.intel-common-rules, rules]
- !reference [.lima-rules, rules]
- !reference [.v3d-rules, rules]
- !reference [.vc4-rules, rules]
- !reference [.nouveau-rules, rules]
- !reference [.r300-rules, rules]
# Also run if this job's own config or script changes
- changes:
- .gitlab-ci/build/gitlab-ci.yml
- .gitlab-ci/run-shader-db.sh
needs:
- debian-build-testing
variables:
S3_ARTIFACT_NAME: debian-build-testing
before_script:
- !reference [.download_s3, before_script]
script: |
.gitlab-ci/run-shader-db.sh
artifacts:
paths:
- shader-db
timeout: 15m
tags:
- kvm # FIXME: this is a hack, should not be needed
# Test a release build with -Werror so new warnings don't sneak in.
debian-release:
extends: .meson-build
needs:
- !reference [.meson-build, needs]
- !reference [.build-for-tests-jobs]
variables:
UNWIND: "enabled"
C_ARGS: >
-Wno-error=stringop-overread
DRI_LOADERS: >
-D glx=dri
-D gbm=enabled
-D egl=enabled
-D glvnd=disabled
-D platforms=x11,wayland
GALLIUM_ST: >
-D gallium-extra-hud=true
-D gallium-vdpau=enabled
-D gallium-va=enabled
-D gallium-xa=enabled
-D gallium-nine=false
-D gallium-rusticl=false
-D llvm=enabled
GALLIUM_DRIVERS: "i915,iris,nouveau,r300,freedreno,llvmpipe,softpipe,svga,v3d,vc4,virgl,etnaviv,panfrost,lima,zink,d3d12,asahi,crocus"
VULKAN_DRIVERS: "swrast,intel_hasvk,imagination-experimental,microsoft-experimental"
EXTRA_OPTION: >
-D spirv-to-dxil=true
-D osmesa=true
-D tools=all
-D mesa-clc=enabled
-D intel-rt=enabled
-D imagination-srv=true
BUILDTYPE: "release"
S3_ARTIFACT_NAME: "mesa-x86_64-default-${BUILDTYPE}"
script:
- *meson-build
- 'if [ -n "$MESA_CI_PERFORMANCE_ENABLED" ]; then .gitlab-ci/prepare-artifacts.sh; fi'
alpine-build-testing:
extends:
- .meson-build
- .use-alpine/x86_64_build
needs:
- !reference [.use-alpine/x86_64_build, needs]
- !reference [.build-for-tests-jobs]
variables:
BUILD_JOB_TIMEOUT: 30m
BUILDTYPE: "release"
C_ARGS: >
-Wno-error=cpp
-Wno-error=array-bounds
-Wno-error=stringop-overflow
-Wno-error=stringop-overread
DRI_LOADERS: >
-D glx=disabled
-D gbm=enabled
-D egl=enabled
-D glvnd=disabled
-D platforms=wayland
GALLIUM_DRIVERS: "crocus,etnaviv,freedreno,iris,lima,nouveau,panfrost,r300,r600,radeonsi,svga,llvmpipe,softpipe,tegra,v3d,vc4,virgl,zink"
GALLIUM_ST: >
-D gallium-extra-hud=true
-D gallium-vdpau=disabled
-D gallium-va=enabled
-D gallium-xa=disabled
-D gallium-nine=true
-D gallium-rusticl=false
-D gles1=disabled
-D gles2=enabled
-D llvm=enabled
-D llvm-orcjit=true
-D microsoft-clc=disabled
-D shared-llvm=enabled
UNWIND: "disabled"
VULKAN_DRIVERS: "amd,asahi,broadcom,freedreno,intel,imagination-experimental"
fedora-release:
extends:
- .meson-build
- .use-fedora/x86_64_build
needs:
- !reference [.use-fedora/x86_64_build, needs]
- !reference [.build-for-tests-jobs]
variables:
BUILDTYPE: "release"
# array-bounds are pure non-LTO gcc buggy warning
# maybe-uninitialized is misfiring in nir_lower_gs_intrinsics.c, and
# a "maybe" warning should never be an error anyway.
C_ARGS: >
-Wno-error=stringop-overflow
-Wno-error=stringop-overread
-Wno-error=array-bounds
-Wno-error=maybe-uninitialized
CPP_ARGS: >
-Wno-error=dangling-reference
-Wno-error=overloaded-virtual
DRI_LOADERS: >
-D glx=dri
-D gbm=enabled
-D egl=enabled
-D glvnd=enabled
-D platforms=x11,wayland
EXTRA_OPTION: >
-D osmesa=true
-D tools=drm-shim,etnaviv,freedreno,glsl,intel,nir,nouveau,lima,panfrost,imagination
-D vulkan-layers=device-select,overlay
-D intel-rt=enabled
-D imagination-srv=true
-D teflon=true
GALLIUM_DRIVERS: "crocus,etnaviv,freedreno,i915,iris,lima,nouveau,panfrost,r300,r600,radeonsi,svga,llvmpipe,softpipe,tegra,v3d,vc4,virgl,zink"
GALLIUM_ST: >
-D gallium-extra-hud=true
-D gallium-vdpau=enabled
-D gallium-va=enabled
-D gallium-xa=enabled
-D gallium-nine=false
-D gallium-rusticl=true
-D gles1=disabled
-D gles2=enabled
-D llvm=enabled
-D microsoft-clc=disabled
-D shared-llvm=enabled
UNWIND: "disabled"
VULKAN_DRIVERS: "amd,asahi,broadcom,freedreno,imagination-experimental,intel,intel_hasvk"
debian-android:
extends:
- .android-variables
- .meson-cross
- .use-debian/android_build
- .ci-deqp-artifacts
needs:
- !reference [.use-debian/android_build, needs]
- !reference [.build-for-tests-jobs]
variables:
BUILDTYPE: debug
UNWIND: "disabled"
C_ARGS: >
-Wno-error=asm-operand-widths
-Wno-error=constant-conversion
-Wno-error=enum-conversion
-Wno-error=initializer-overrides
-Wno-error=sometimes-uninitialized
CPP_ARGS: >
-Wno-error=c99-designator
-Wno-error=unused-variable
-Wno-error=unused-but-set-variable
-Wno-error=self-assign
DRI_LOADERS: >
-D glx=disabled
-D gbm=disabled
-D egl=enabled
-D glvnd=disabled
-D platforms=android
FORCE_FALLBACK_FOR: llvm
EXTRA_OPTION: >
-D android-stub=true
-D platform-sdk-version=${ANDROID_SDK_VERSION}
-D cpp_rtti=false
-D valgrind=disabled
-D android-libbacktrace=disabled
-D mesa-clc=system
GALLIUM_ST: >
-D gallium-vdpau=disabled
-D gallium-va=disabled
-D gallium-xa=disabled
-D gallium-nine=false
-D gallium-rusticl=false
PKG_CONFIG_LIBDIR: "/disable/non/android/system/pc/files"
HOST_BUILD_OPTIONS: >
-D build-tests=false
-D enable-glcpp-tests=false
-D gallium-opencl=disabled
-D gallium-drivers=
-D vulkan-drivers=
-D video-codecs=
-D glx=disabled
-D platforms=
-D mesa-clc=enabled
-D install-mesa-clc=true
ARTIFACTS_DEBUG_SYMBOLS: 1
S3_ARTIFACT_NAME: mesa-x86_64-android-${BUILDTYPE}
script:
- export CROSS=aarch64-linux-android
- export GALLIUM_DRIVERS=etnaviv,freedreno,lima,panfrost,vc4,v3d
- export VULKAN_DRIVERS=freedreno,broadcom,virtio
- *meson-build
# x86_64 build:
# Can't do AMD drivers because they require LLVM, which is currently
# problematic in our Android builds.
- export CROSS=x86_64-linux-android
- export GALLIUM_DRIVERS=iris,virgl,zink,softpipe,llvmpipe,swrast
- export VULKAN_DRIVERS=intel,virtio,swrast
- .gitlab-ci/create-llvm-meson-wrap-file.sh
- *meson-build
- .gitlab-ci/prepare-artifacts.sh
.meson-cross:
extends:
- .meson-build
variables:
UNWIND: "disabled"
DRI_LOADERS: >
-D glx=dri
-D gbm=enabled
-D egl=enabled
-D platforms=x11,wayland
-D osmesa=false
GALLIUM_ST: >
-D gallium-vdpau=disabled
-D gallium-va=disabled
-D gallium-xa=disabled
-D gallium-nine=false
.meson-arm:
extends:
- .meson-cross
- .use-debian/arm64_build
needs:
- debian/arm64_build
variables:
VULKAN_DRIVERS: "asahi,broadcom,freedreno"
GALLIUM_DRIVERS: "etnaviv,freedreno,lima,nouveau,panfrost,llvmpipe,softpipe,tegra,v3d,vc4,zink"
BUILDTYPE: "debugoptimized"
tags:
- aarch64
debian-arm32:
extends:
- .meson-arm
- .ci-deqp-artifacts
stage: build-for-tests
variables:
CROSS: armhf
DRI_LOADERS:
-D glvnd=disabled
# remove asahi & llvmpipe from the .meson-arm list because here we have llvm=disabled
VULKAN_DRIVERS: "broadcom,freedreno"
GALLIUM_DRIVERS: "etnaviv,freedreno,lima,nouveau,panfrost,softpipe,tegra,v3d,vc4,zink"
EXTRA_OPTION: >
-D llvm=disabled
-D valgrind=disabled
-D gallium-rusticl=false
S3_ARTIFACT_NAME: mesa-arm32-default-${BUILDTYPE}
# The strip command segfaults, failing to strip the binary and leaving
# tempfiles in our artifacts.
ARTIFACTS_DEBUG_SYMBOLS: 1
script:
- *meson-build
- .gitlab-ci/prepare-artifacts.sh
debian-arm32-asan:
extends:
- debian-arm32
stage: build-for-tests
variables:
BUILD_JOB_TIMEOUT: 30m
GALLIUM_DRIVERS: "etnaviv"
VULKAN_DRIVERS: ""
DRI_LOADERS:
-D glvnd=disabled
EXTRA_OPTION: >
-D llvm=disabled
-D b_sanitize=address
-D valgrind=disabled
-D tools=dlclose-skip
-D gallium-rusticl=false
ARTIFACTS_DEBUG_SYMBOLS: 1
RUN_MESON_TESTS: "false" # just too slow
S3_ARTIFACT_NAME: mesa-arm32-asan-${BUILDTYPE}
debian-arm64:
extends:
- .meson-arm
- .ci-deqp-artifacts
stage: build-for-tests
variables:
C_ARGS: >
-Wno-error=array-bounds
-Wno-error=stringop-truncation
GALLIUM_DRIVERS: "etnaviv,freedreno,lima,panfrost,v3d,vc4,zink"
VULKAN_DRIVERS: "broadcom,freedreno,panfrost"
DRI_LOADERS:
-D glvnd=disabled
EXTRA_OPTION: >
-D valgrind=disabled
-D imagination-srv=true
-D freedreno-kmds=msm,virtio
-D teflon=true
GALLIUM_ST:
-D gallium-rusticl=true
RUN_MESON_TESTS: "false" # run by debian-arm64-build-testing
S3_ARTIFACT_NAME: mesa-arm64-default-${BUILDTYPE}
script:
- *meson-build
- .gitlab-ci/prepare-artifacts.sh
debian-arm64-asan:
extends:
- debian-arm64
stage: build-for-tests
variables:
BUILD_JOB_TIMEOUT: 30m
VULKAN_DRIVERS: "broadcom,freedreno"
GALLIUM_DRIVERS: "freedreno,vc4,v3d"
DRI_LOADERS:
-D glvnd=disabled
EXTRA_OPTION: >
-D b_sanitize=address
-D valgrind=disabled
-D tools=dlclose-skip
-D gallium-rusticl=false
ARTIFACTS_DEBUG_SYMBOLS: 1
RUN_MESON_TESTS: "false" # just too slow
S3_ARTIFACT_NAME: mesa-arm64-asan-${BUILDTYPE}
debian-arm64-ubsan:
extends:
- debian-arm64
stage: build-for-tests
variables:
BUILD_JOB_TIMEOUT: 30m
VULKAN_DRIVERS: "broadcom"
GALLIUM_DRIVERS: "v3d,vc4"
C_ARGS: >
-Wno-error=array-bounds
-Wno-error=stringop-overflow
-Wno-error=stringop-truncation
CPP_ARGS: >
-Wno-error=array-bounds
-fno-var-tracking-assignments
DRI_LOADERS:
-D glvnd=disabled
EXTRA_OPTION: >
-D b_sanitize=undefined
-D gallium-rusticl=false
ARTIFACTS_DEBUG_SYMBOLS: 1
RUN_MESON_TESTS: "false" # just too slow
S3_ARTIFACT_NAME: mesa-arm64-ubsan-${BUILDTYPE}
debian-arm64-build-test:
extends:
- .meson-arm
- .ci-deqp-artifacts
needs:
- !reference [.meson-arm, needs]
- !reference [.build-for-tests-jobs]
variables:
VULKAN_DRIVERS: "amd,asahi,imagination-experimental,nouveau"
DRI_LOADERS:
-D glvnd=disabled
EXTRA_OPTION: >
-D tools=panfrost,imagination
-D perfetto=true
debian-arm64-release:
extends:
- debian-arm64
stage: build-only
needs:
- !reference [debian-arm64, needs]
- !reference [.build-for-tests-jobs]
variables:
BUILDTYPE: release
S3_ARTIFACT_NAME: mesa-arm64-default-${BUILDTYPE}
C_ARGS: >
-Wno-error=array-bounds
-Wno-error=stringop-truncation
-Wno-error=stringop-overread
script:
- *meson-build
- 'if [ -n "$MESA_CI_PERFORMANCE_ENABLED" ]; then .gitlab-ci/prepare-artifacts.sh; fi'
debian-no-libdrm:
extends:
- .meson-arm
stage: build-only
needs:
- !reference [.meson-arm, needs]
- !reference [.build-for-tests-jobs]
variables:
VULKAN_DRIVERS: freedreno
GALLIUM_DRIVERS: "zink,llvmpipe"
BUILDTYPE: release
C_ARGS: >
-Wno-error=array-bounds
-Wno-error=stringop-truncation
-Wno-error=stringop-overread
EXTRA_OPTION: >
-D freedreno-kmds=kgsl
-D glx=disabled
-D gbm=disabled
-D egl=disabled
debian-clang:
extends: .meson-build
needs:
- !reference [.meson-build, needs]
- !reference [.build-for-tests-jobs]
variables:
BUILDTYPE: debug
UNWIND: "enabled"
C_ARGS: >
-Wno-error=constant-conversion
-Wno-error=enum-conversion
-Wno-error=initializer-overrides
-Wno-error=sometimes-uninitialized
-Werror=misleading-indentation
CPP_ARGS: >
-Wno-error=c99-designator
-Wno-error=overloaded-virtual
-Wno-error=tautological-constant-out-of-range-compare
-Wno-error=unused-private-field
-Wno-error=vla-cxx-extension
DRI_LOADERS: >
-D glx=dri
-D gbm=enabled
-D egl=enabled
-D glvnd=enabled
-D platforms=x11,wayland
GALLIUM_ST: >
-D gallium-extra-hud=true
-D gallium-vdpau=enabled
-D gallium-va=enabled
-D gallium-xa=enabled
-D gallium-nine=true
-D gles1=enabled
-D gles2=enabled
-D llvm=enabled
-D microsoft-clc=disabled
-D shared-llvm=enabled
-D shared-glapi=enabled
GALLIUM_DRIVERS: "iris,nouveau,r300,r600,freedreno,llvmpipe,softpipe,svga,v3d,vc4,virgl,etnaviv,panfrost,lima,zink,radeonsi,tegra,d3d12,crocus,i915,asahi"
VULKAN_DRIVERS: intel,amd,freedreno,broadcom,virtio,swrast,panfrost,imagination-experimental,microsoft-experimental,nouveau
EXTRA_OPTION:
-D spirv-to-dxil=true
-D osmesa=true
-D imagination-srv=true
-D tools=drm-shim,etnaviv,freedreno,glsl,intel,intel-ui,nir,nouveau,lima,panfrost,asahi,imagination
-D vulkan-layers=device-select,overlay
-D build-aco-tests=true
-D mesa-clc=enabled
-D intel-rt=enabled
-D imagination-srv=true
-D teflon=true
CC: clang-${LLVM_VERSION}
CXX: clang++-${LLVM_VERSION}
debian-clang-release:
extends: debian-clang
# `needs:` inherited from debian-clang
variables:
BUILD_JOB_TIMEOUT: 30m
BUILDTYPE: "release"
DRI_LOADERS: >
-D glx=xlib
-D platforms=x11,wayland
GALLIUM_ST: >
-D gallium-extra-hud=true
-D gallium-vdpau=enabled
-D gallium-va=enabled
-D gallium-xa=enabled
-D gallium-nine=true
-D gles1=disabled
-D gles2=disabled
-D llvm=enabled
-D microsoft-clc=disabled
-D shared-llvm=enabled
-D shared-glapi=disabled
windows-msvc:
extends:
- .build-windows
- .use-windows_build_msvc
- .windows-build-rules
stage: build-for-tests
script:
- pwsh -ExecutionPolicy RemoteSigned .\.gitlab-ci\windows\mesa_build.ps1
artifacts:
paths:
- _build/meson-logs/*.txt
- _install/
debian-vulkan:
extends: .meson-build
needs:
- !reference [.meson-build, needs]
- !reference [.build-for-tests-jobs]
variables:
BUILD_JOB_TIMEOUT: 30m
BUILDTYPE: debug
UNWIND: "disabled"
DRI_LOADERS: >
-D glx=disabled
-D gbm=disabled
-D egl=disabled
-D opengl=false
-D gles1=disabled
-D gles2=disabled
-D glvnd=disabled
-D platforms=x11,wayland
-D osmesa=false
GALLIUM_ST: >
-D gallium-vdpau=disabled
-D gallium-va=disabled
-D gallium-xa=disabled
-D gallium-nine=false
-D gallium-rusticl=false
-D b_sanitize=undefined
-D c_args=-fno-sanitize-recover=all
-D cpp_args=-fno-sanitize-recover=all
UBSAN_OPTIONS: "print_stacktrace=1"
VULKAN_DRIVERS: amd,asahi,broadcom,freedreno,intel,intel_hasvk,panfrost,virtio,imagination-experimental,microsoft-experimental,nouveau
EXTRA_OPTION: >
-D vulkan-layers=device-select,overlay
-D build-aco-tests=true
-D intel-rt=disabled
-D imagination-srv=true
debian-x86_32:
extends:
- .meson-cross
- .use-debian/x86_32_build
needs:
- !reference [.use-debian/x86_32_build, needs]
- !reference [.build-for-tests-jobs]
variables:
BUILDTYPE: debug
CROSS: i386
VULKAN_DRIVERS: intel,amd,swrast,virtio,panfrost
GALLIUM_DRIVERS: "iris,nouveau,r300,r600,radeonsi,llvmpipe,softpipe,virgl,zink,crocus,d3d12,panfrost"
DRI_LOADERS:
-D glvnd=disabled
EXTRA_OPTION: >
-D vulkan-layers=device-select,overlay
-D mesa-clc=system
C_LINK_ARGS: >
-Wl,--no-warn-rwx-segments
CPP_LINK_ARGS: >
-Wl,--no-warn-rwx-segments
HOST_BUILD_OPTIONS: >
-D build-tests=false
-D enable-glcpp-tests=false
-D gallium-opencl=disabled
-D gallium-drivers=
-D vulkan-drivers=
-D video-codecs=
-D glx=disabled
-D platforms=
-D mesa-clc=enabled
-D install-mesa-clc=true
# While s390 is dead, s390x is very much alive, and one of the last major
# big-endian platforms, so it provides useful coverage.
# In case of issues with this job, contact @ajax
debian-s390x:
extends:
- .meson-cross
- .use-debian/s390x_build
needs:
- !reference [.use-debian/s390x_build, needs]
- !reference [.build-for-tests-jobs]
tags:
- kvm
variables:
BUILDTYPE: debug
CROSS: s390x
GALLIUM_DRIVERS: "llvmpipe,virgl,zink"
VULKAN_DRIVERS: "swrast,virtio"
DRI_LOADERS:
-D glvnd=disabled
debian-ppc64el:
extends:
- .meson-cross
- .use-debian/ppc64el_build
needs:
- !reference [.use-debian/ppc64el_build, needs]
- !reference [.build-for-tests-jobs]
variables:
BUILDTYPE: debug
CROSS: ppc64el
GALLIUM_DRIVERS: "nouveau,radeonsi,llvmpipe,softpipe,virgl,zink"
VULKAN_DRIVERS: "amd,swrast"
DRI_LOADERS:
-D glvnd=disabled
# This job emits our scripts into artifacts so they can be reused for
# job submission to hardware devices.
python-artifacts:
stage: build-for-tests
extends:
- .use-debian/x86_64_pyutils
- .build-common
variables:
GIT_STRATEGY: fetch
S3_ARTIFACT_NAME: mesa-python-ci-artifacts
timeout: 10m
script:
- .gitlab-ci/prepare-artifacts-python.sh
tags:
- placeholder-job

View File

@@ -1,35 +0,0 @@
#!/usr/bin/env bash
# shellcheck disable=SC2035
# shellcheck disable=SC2061
# shellcheck disable=SC2086 # we want word splitting
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_DIR/first.devcore; then
echo 1 > $i
echo "Saved to the job artifacts at /first.devcore"
exit 0
fi
done
i915_error_states=$(find /sys/devices/ -path */drm/card*/error)
for i in $i915_error_states; do
tmpfile=$(mktemp)
cp "$i" "$tmpfile"
filesize=$(stat --printf="%s" "$tmpfile")
# Does the file contain "No error state collected" ?
if [ "$filesize" = 25 ]; then
rm "$tmpfile"
else
echo "Found an i915 error state at $i size=$filesize."
if cp "$tmpfile" $RESULTS_DIR/first.i915_error_state; then
rm "$tmpfile"
echo 1 > "$i"
echo "Saved to the job artifacts at /first.i915_error_state"
exit 0
fi
fi
done
sleep 10
done

View File

@@ -1,139 +0,0 @@
#!/bin/bash
VARS=(
ACO_DEBUG
ARTIFACTS_BASE_URL
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
S3_JWT_FILE
CI_JOB_STARTED_AT
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
CURRENT_SECTION
DEQP_BIN_DIR
DEQP_FORCE_ASAN
DEQP_FRACTION
DEQP_RUNNER_MAX_FAILS
DEQP_SUITE
DEQP_TEMP_DIR
DEVICE_NAME
DRIVER_NAME
EGL_PLATFORM
ETNA_MESA_DEBUG
FDO_CI_CONCURRENT
FDO_HTTP_CACHE_URI
FDO_UPSTREAM_REPO
FD_MESA_DEBUG
FLAKES_CHANNEL
FLUSTER_CODECS
FLUSTER_FRACTION
FLUSTER_VECTORS_VERSION
FREEDRENO_HANGCHECK_MS
GALLIUM_DRIVER
GALLIVM_PERF
GPU_VERSION
GTEST
GTEST_FAILS
GTEST_FRACTION
GTEST_RUNNER_OPTIONS
GTEST_SKIPS
HWCI_FREQ_MAX
HWCI_KERNEL_MODULES
HWCI_KVM
HWCI_START_WESTON
HWCI_START_XORG
HWCI_TEST_ARGS
HWCI_TEST_SCRIPT
IR3_SHADER_DEBUG
JOB_ARTIFACTS_BASE
JOB_RESULTS_PATH
JOB_ROOTFS_OVERLAY_PATH
KERNEL_IMAGE_BASE
KERNEL_IMAGE_NAME
LD_LIBRARY_PATH
LIBGL_ALWAYS_SOFTWARE
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_LOADER_DRIVER_OVERRIDE
MESA_SPIRV_LOG_LEVEL
MESA_TEMPLATES_COMMIT
MESA_VK_ABORT_ON_DEVICE_LOSS
MESA_VK_IGNORE_CONFORMANCE_WARNING
S3_HOST
S3_RESULTS_UPLOAD
NIR_DEBUG
PAN_I_WANT_A_BROKEN_VULKAN_DRIVER
PAN_MESA_DEBUG
PANVK_DEBUG
PIGLIT_FRACTION
PIGLIT_NO_WINDOW
PIGLIT_OPTIONS
PIGLIT_PLATFORM
PIGLIT_PROFILES
PIGLIT_REPLAY_ANGLE_TAG
PIGLIT_REPLAY_ARTIFACTS_BASE_URL
PIGLIT_REPLAY_DEVICE_NAME
PIGLIT_REPLAY_EXTRA_ARGS
PIGLIT_REPLAY_LOOP_TIMES
PIGLIT_REPLAY_REFERENCE_IMAGES_BASE
PIGLIT_REPLAY_SUBCOMMAND
PIGLIT_RESULTS
PIGLIT_RUNNER_OPTIONS
PIGLIT_TESTS
PIGLIT_TRACES_FILE
PIPELINE_ARTIFACTS_BASE
RADEON_DEBUG
RADV_DEBUG
RADV_PERFTEST
SKQP_ASSETS_DIR
SKQP_BACKENDS
STORAGE_FORK_HOST_PATH
STORAGE_MAINLINE_HOST_PATH
TU_DEBUG
USE_ANGLE
VIRGL_HOST_API
VIRGL_RENDER_SERVER
WAFFLE_PLATFORM
VK_DRIVER
ZINK_DESCRIPTORS
ZINK_DEBUG
LVP_POISON_MEMORY
# Dead code within Mesa CI, but required by virglrender CI
# (because they include our files in their CI)
VK_DRIVER_FILES
)
for var in "${VARS[@]}"; do
if [ -n "${!var+x}" ]; then
echo "export $var=${!var@Q}"
fi
done

View File

@@ -1,25 +0,0 @@
#!/bin/sh
# Very early init, used to make sure devices and network are set up and
# reachable.
set -ex
cd /
findmnt --mountpoint /proc || mount -t proc none /proc
findmnt --mountpoint /sys || mount -t sysfs none /sys
mount -t debugfs none /sys/kernel/debug
findmnt --mountpoint /dev || mount -t devtmpfs none /dev
mkdir -p /dev/pts
mount -t devpts devpts /dev/pts
mkdir /dev/shm
mount -t tmpfs -o noexec,nodev,nosuid tmpfs /dev/shm
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 _ in 1 2 3; do sntp -sS pool.ntp.org && break || sleep 2; done || true

View File

@@ -1,249 +0,0 @@
#!/bin/bash
# shellcheck disable=SC1090
# shellcheck disable=SC1091
# shellcheck disable=SC2086 # we want word splitting
# shellcheck disable=SC2155
# Second-stage init, used to set up devices and our job environment before
# running tests.
shopt -s extglob
# Make sure to kill itself and all the children process from this script on
# exiting, since any console output may interfere with LAVA signals handling,
# which based on the log console.
cleanup() {
if [ "$BACKGROUND_PIDS" = "" ]; then
return 0
fi
set +x
echo "Killing all child processes"
for pid in $BACKGROUND_PIDS
do
kill "$pid" 2>/dev/null || true
done
# Sleep just a little to give enough time for subprocesses to be gracefully
# killed. Then apply a SIGKILL if necessary.
sleep 5
for pid in $BACKGROUND_PIDS
do
kill -9 "$pid" 2>/dev/null || true
done
BACKGROUND_PIDS=
set -x
}
trap cleanup INT TERM EXIT
# Space separated values with the PIDS of the processes started in the
# background by this script
BACKGROUND_PIDS=
for path in '/dut-env-vars.sh' '/set-job-env-vars.sh' './set-job-env-vars.sh'; do
[ -f "$path" ] && source "$path"
done
. "$SCRIPTS_DIR"/setup-test-env.sh
# Flush out anything which might be stuck in a serial buffer
echo
echo
echo
section_switch init_stage2 "Pre-testing hardware setup"
set -ex
# Set up any devices required by the jobs
[ -z "$HWCI_KERNEL_MODULES" ] || {
echo -n $HWCI_KERNEL_MODULES | xargs -d, -n1 /usr/sbin/modprobe
}
# Set up ZRAM
HWCI_ZRAM_SIZE=2G
if /sbin/zramctl --find --size $HWCI_ZRAM_SIZE -a zstd; then
mkswap /dev/zram0
swapon /dev/zram0
echo "zram: $HWCI_ZRAM_SIZE activated"
else
echo "zram: skipping, not supported"
fi
#
# Load the KVM module specific to the detected CPU virtualization extensions:
# - vmx for Intel VT
# - svm for AMD-V
#
# Additionally, download the kernel image to boot the VM via HWCI_TEST_SCRIPT.
#
if [ "$HWCI_KVM" = "true" ]; then
unset KVM_KERNEL_MODULE
{
grep -qs '\bvmx\b' /proc/cpuinfo && KVM_KERNEL_MODULE=kvm_intel
} || {
grep -qs '\bsvm\b' /proc/cpuinfo && KVM_KERNEL_MODULE=kvm_amd
}
{
[ -z "${KVM_KERNEL_MODULE}" ] && \
echo "WARNING: Failed to detect CPU virtualization extensions"
} || \
modprobe ${KVM_KERNEL_MODULE}
mkdir -p /lava-files
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
-o "/lava-files/${KERNEL_IMAGE_NAME}" \
"${KERNEL_IMAGE_BASE}/amd64/${KERNEL_IMAGE_NAME}"
fi
# Fix prefix confusion: the build installs to $CI_PROJECT_DIR, but we expect
# it in /install
ln -sf $CI_PROJECT_DIR/install /install
export LD_LIBRARY_PATH=/install/lib
export LIBGL_DRIVERS_PATH=/install/lib/dri
# https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/22495#note_1876691
# The navi21 boards seem to have trouble with ld.so.cache, so try explicitly
# telling it to look in /usr/local/lib.
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/lib
# 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=$(python3 -c "import sys;print(\":\".join(sys.path))")
# If we need to specify a driver, it means several drivers could pick up this gpu;
# ensure that the other driver can't accidentally be used
if [ -n "$MESA_LOADER_DRIVER_OVERRIDE" ]; then
rm /install/lib/dri/!($MESA_LOADER_DRIVER_OVERRIDE)_dri.so
fi
ls -1 /install/lib/dri/*_dri.so || true
if [ "$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 > $GPU_AUTOSUSPEND || true
# Lock Intel GPU frequency to 70% of the maximum allowed by hardware
# and enable throttling detection & reporting.
# Additionally, set the upper limit for CPU scaling frequency to 65% of the
# maximum permitted, as an additional measure to mitigate thermal throttling.
/install/common/intel-gpu-freq.sh -s 70% --cpu-set-max 65% -g all -d
fi
# Start a little daemon to capture sysfs records and produce a JSON file
KDL_PATH=/install/common/kdl.sh
if [ -x "$KDL_PATH" ]; then
echo "launch kdl.sh!"
$KDL_PATH &
BACKGROUND_PIDS="$! $BACKGROUND_PIDS"
else
echo "kdl.sh not found!"
fi
# Increase freedreno hangcheck timer because it's right at the edge of the
# spilling tests timing out (and some traces, too)
if [ -n "$FREEDRENO_HANGCHECK_MS" ]; then
echo $FREEDRENO_HANGCHECK_MS | tee -a /sys/kernel/debug/dri/128/hangcheck_period_ms
fi
# Start a little daemon to capture the first devcoredump we encounter. (They
# expire after 5 minutes, so we poll for them).
CAPTURE_DEVCOREDUMP=/install/common/capture-devcoredump.sh
if [ -x "$CAPTURE_DEVCOREDUMP" ]; then
$CAPTURE_DEVCOREDUMP &
BACKGROUND_PIDS="$! $BACKGROUND_PIDS"
fi
ARCH=$(uname -m)
export VK_DRIVER_FILES="/install/share/vulkan/icd.d/${VK_DRIVER}_icd.$ARCH.json"
# 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 "$HWCI_START_XORG" ]; then
echo "touch /xorg-started; sleep 100000" > /xorg-script
env \
xinit /bin/sh /xorg-script -- /usr/bin/Xorg -noreset -s 0 -dpms -logfile "$RESULTS_DIR/Xorg.0.log" &
BACKGROUND_PIDS="$! $BACKGROUND_PIDS"
# Wait for xorg to be ready for connections.
for _ in 1 2 3 4 5; do
if [ -e /xorg-started ]; then
break
fi
sleep 5
done
export DISPLAY=:0
fi
if [ -n "$HWCI_START_WESTON" ]; then
WESTON_X11_SOCK="/tmp/.X11-unix/X0"
if [ -n "$HWCI_START_XORG" ]; then
echo "Please consider dropping HWCI_START_XORG and instead using Weston XWayland for testing."
WESTON_X11_SOCK="/tmp/.X11-unix/X1"
fi
export WAYLAND_DISPLAY=wayland-0
# Display server is Weston Xwayland when HWCI_START_XORG is not set or Xorg when it's
export DISPLAY=:0
mkdir -p /tmp/.X11-unix
env \
weston -Bheadless-backend.so --use-gl -Swayland-0 --xwayland --idle-time=0 &
BACKGROUND_PIDS="$! $BACKGROUND_PIDS"
while [ ! -S "$WESTON_X11_SOCK" ]; do sleep 1; done
fi
set +x
section_end init_stage2
echo "Running ${HWCI_TEST_SCRIPT} ${HWCI_TEST_ARGS} ..."
set +e
$HWCI_TEST_SCRIPT ${HWCI_TEST_ARGS:-}; EXIT_CODE=$?
set -e
section_start post_test_cleanup "Cleaning up after testing, uploading results"
set -x
# Make sure that capture-devcoredump is done before we start trying to tar up
# artifacts -- if it's writing while tar is reading, tar will throw an error and
# kill the job.
cleanup
# upload artifacts
if [ -n "$S3_RESULTS_UPLOAD" ]; then
tar --zstd -cf results.tar.zst results/;
s3_upload results.tar.zst https://"$S3_RESULTS_UPLOAD"/
fi
# We still need to echo the hwci: mesa message, as some scripts rely on it, such
# as the python ones inside the bare-metal folder
[ ${EXIT_CODE} -eq 0 ] && RESULT=pass || RESULT=fail
set +x
section_end post_test_cleanup
# Print the final result; both bare-metal and LAVA look for this string to get
# the result of our run, so try really hard to get it out rather than losing
# the run. The device gets shut down right at this point, and a630 seems to
# enjoy corrupting the last line of serial output before shutdown.
for _ in $(seq 0 3); do echo "hwci: mesa: $RESULT, exit_code: $EXIT_CODE"; sleep 1; echo; done
exit $EXIT_CODE

View File

@@ -1,769 +0,0 @@
#!/usr/bin/env bash
# shellcheck disable=SC2013
# shellcheck disable=SC2015
# shellcheck disable=SC2034
# shellcheck disable=SC2046
# shellcheck disable=SC2059
# shellcheck disable=SC2086 # we want word splitting
# shellcheck disable=SC2154
# shellcheck disable=SC2155
# shellcheck disable=SC2162
# shellcheck disable=SC2229
#
# This is an utility script to manage Intel GPU frequencies.
# It can be used for debugging performance problems or trying to obtain a stable
# frequency while benchmarking.
#
# Note the Intel i915 GPU driver allows to change the minimum, maximum and boost
# frequencies in steps of 50 MHz via:
#
# /sys/class/drm/card<n>/<freq_info>
#
# Where <n> is the DRM card index and <freq_info> one of the following:
#
# - gt_max_freq_mhz (enforced maximum freq)
# - gt_min_freq_mhz (enforced minimum freq)
# - gt_boost_freq_mhz (enforced boost freq)
#
# The hardware capabilities can be accessed via:
#
# - gt_RP0_freq_mhz (supported maximum freq)
# - gt_RPn_freq_mhz (supported minimum freq)
# - gt_RP1_freq_mhz (most efficient freq)
#
# The current frequency can be read from:
# - gt_act_freq_mhz (the actual GPU freq)
# - gt_cur_freq_mhz (the last requested freq)
#
# Also note that in addition to GPU management, the script offers the
# possibility to adjust CPU operating frequencies. However, this is currently
# limited to just setting the maximum scaling frequency as percentage of the
# maximum frequency allowed by the hardware.
#
# Copyright (C) 2022 Collabora Ltd.
# Author: Cristian Ciocaltea <cristian.ciocaltea@collabora.com>
#
# SPDX-License-Identifier: MIT
#
#
# Constants
#
# GPU
DRM_FREQ_SYSFS_PATTERN="/sys/class/drm/card%d/gt_%s_freq_mhz"
ENF_FREQ_INFO="max min boost"
CAP_FREQ_INFO="RP0 RPn RP1"
ACT_FREQ_INFO="act cur"
THROTT_DETECT_SLEEP_SEC=2
THROTT_DETECT_PID_FILE_PATH=/tmp/thrott-detect.pid
# CPU
CPU_SYSFS_PREFIX=/sys/devices/system/cpu
CPU_PSTATE_SYSFS_PATTERN="${CPU_SYSFS_PREFIX}/intel_pstate/%s"
CPU_FREQ_SYSFS_PATTERN="${CPU_SYSFS_PREFIX}/cpu%s/cpufreq/%s_freq"
CAP_CPU_FREQ_INFO="cpuinfo_max cpuinfo_min"
ENF_CPU_FREQ_INFO="scaling_max scaling_min"
ACT_CPU_FREQ_INFO="scaling_cur"
#
# Global variables.
#
unset INTEL_DRM_CARD_INDEX
unset GET_ACT_FREQ GET_ENF_FREQ GET_CAP_FREQ
unset SET_MIN_FREQ SET_MAX_FREQ
unset MONITOR_FREQ
unset CPU_SET_MAX_FREQ
unset DETECT_THROTT
unset DRY_RUN
#
# Simple printf based stderr logger.
#
log() {
local msg_type=$1
shift
printf "%s: %s: " "${msg_type}" "${0##*/}" >&2
printf "$@" >&2
printf "\n" >&2
}
#
# Helper to print sysfs path for the given card index and freq info.
#
# arg1: Frequency info sysfs name, one of *_FREQ_INFO constants above
# arg2: Video card index, defaults to INTEL_DRM_CARD_INDEX
#
print_freq_sysfs_path() {
printf ${DRM_FREQ_SYSFS_PATTERN} "${2:-${INTEL_DRM_CARD_INDEX}}" "$1"
}
#
# Helper to set INTEL_DRM_CARD_INDEX for the first identified Intel video card.
#
identify_intel_gpu() {
local i=0 vendor path
while [ ${i} -lt 16 ]; do
[ -c "/dev/dri/card$i" ] || {
i=$((i + 1))
continue
}
path=$(print_freq_sysfs_path "" ${i})
path=${path%/*}/device/vendor
[ -r "${path}" ] && read vendor < "${path}" && \
[ "${vendor}" = "0x8086" ] && INTEL_DRM_CARD_INDEX=$i && return 0
i=$((i + 1))
done
return 1
}
#
# Read the specified freq info from sysfs.
#
# arg1: Flag (y/n) to also enable printing the freq info.
# arg2...: Frequency info sysfs name(s), see *_FREQ_INFO constants above
# return: Global variable(s) FREQ_${arg} containing the requested information
#
read_freq_info() {
local var val info path print=0 ret=0
[ "$1" = "y" ] && print=1
shift
while [ $# -gt 0 ]; do
info=$1
shift
var=FREQ_${info}
path=$(print_freq_sysfs_path "${info}")
[ -r ${path} ] && read ${var} < ${path} || {
log ERROR "Failed to read freq info from: %s" "${path}"
ret=1
continue
}
[ -n "${var}" ] || {
log ERROR "Got empty freq info from: %s" "${path}"
ret=1
continue
}
[ ${print} -eq 1 ] && {
eval val=\$${var}
printf "%6s: %4s MHz\n" "${info}" "${val}"
}
done
return ${ret}
}
#
# Display requested info.
#
print_freq_info() {
local req_freq
[ -n "${GET_CAP_FREQ}" ] && {
printf "* Hardware capabilities\n"
read_freq_info y ${CAP_FREQ_INFO}
printf "\n"
}
[ -n "${GET_ENF_FREQ}" ] && {
printf "* Enforcements\n"
read_freq_info y ${ENF_FREQ_INFO}
printf "\n"
}
[ -n "${GET_ACT_FREQ}" ] && {
printf "* Actual\n"
read_freq_info y ${ACT_FREQ_INFO}
printf "\n"
}
}
#
# Helper to print frequency value as requested by user via '-s, --set' option.
# arg1: user requested freq value
#
compute_freq_set() {
local val
case "$1" in
+)
val=${FREQ_RP0}
;;
-)
val=${FREQ_RPn}
;;
*%)
val=$((${1%?} * FREQ_RP0 / 100))
# Adjust freq to comply with 50 MHz increments
val=$((val / 50 * 50))
;;
*[!0-9]*)
log ERROR "Cannot set freq to invalid value: %s" "$1"
return 1
;;
"")
log ERROR "Cannot set freq to unspecified value"
return 1
;;
*)
# Adjust freq to comply with 50 MHz increments
val=$(($1 / 50 * 50))
;;
esac
printf "%s" "${val}"
}
#
# Helper for set_freq().
#
set_freq_max() {
log INFO "Setting GPU max freq to %s MHz" "${SET_MAX_FREQ}"
read_freq_info n min || return $?
[ ${SET_MAX_FREQ} -gt ${FREQ_RP0} ] && {
log ERROR "Cannot set GPU max freq (%s) to be greater than hw max freq (%s)" \
"${SET_MAX_FREQ}" "${FREQ_RP0}"
return 1
}
[ ${SET_MAX_FREQ} -lt ${FREQ_RPn} ] && {
log ERROR "Cannot set GPU max freq (%s) to be less than hw min freq (%s)" \
"${SET_MIN_FREQ}" "${FREQ_RPn}"
return 1
}
[ ${SET_MAX_FREQ} -lt ${FREQ_min} ] && {
log ERROR "Cannot set GPU max freq (%s) to be less than min freq (%s)" \
"${SET_MAX_FREQ}" "${FREQ_min}"
return 1
}
[ -z "${DRY_RUN}" ] || return 0
if ! printf "%s" ${SET_MAX_FREQ} | tee $(print_freq_sysfs_path max) \
$(print_freq_sysfs_path boost) > /dev/null;
then
log ERROR "Failed to set GPU max frequency"
return 1
fi
}
#
# Helper for set_freq().
#
set_freq_min() {
log INFO "Setting GPU min freq to %s MHz" "${SET_MIN_FREQ}"
read_freq_info n max || return $?
[ ${SET_MIN_FREQ} -gt ${FREQ_max} ] && {
log ERROR "Cannot set GPU min freq (%s) to be greater than max freq (%s)" \
"${SET_MIN_FREQ}" "${FREQ_max}"
return 1
}
[ ${SET_MIN_FREQ} -lt ${FREQ_RPn} ] && {
log ERROR "Cannot set GPU min freq (%s) to be less than hw min freq (%s)" \
"${SET_MIN_FREQ}" "${FREQ_RPn}"
return 1
}
[ -z "${DRY_RUN}" ] || return 0
if ! printf "%s" ${SET_MIN_FREQ} > $(print_freq_sysfs_path min);
then
log ERROR "Failed to set GPU min frequency"
return 1
fi
}
#
# Set min or max or both GPU frequencies to the user indicated values.
#
set_freq() {
# Get hw max & min frequencies
read_freq_info n RP0 RPn || return $?
[ -z "${SET_MAX_FREQ}" ] || {
SET_MAX_FREQ=$(compute_freq_set "${SET_MAX_FREQ}")
[ -z "${SET_MAX_FREQ}" ] && return 1
}
[ -z "${SET_MIN_FREQ}" ] || {
SET_MIN_FREQ=$(compute_freq_set "${SET_MIN_FREQ}")
[ -z "${SET_MIN_FREQ}" ] && return 1
}
#
# Ensure correct operation order, to avoid setting min freq
# to a value which is larger than max freq.
#
# E.g.:
# crt_min=crt_max=600; new_min=new_max=700
# > operation order: max=700; min=700
#
# crt_min=crt_max=600; new_min=new_max=500
# > operation order: min=500; max=500
#
if [ -n "${SET_MAX_FREQ}" ] && [ -n "${SET_MIN_FREQ}" ]; then
[ ${SET_MAX_FREQ} -lt ${SET_MIN_FREQ} ] && {
log ERROR "Cannot set GPU max freq to be less than min freq"
return 1
}
read_freq_info n min || return $?
if [ ${SET_MAX_FREQ} -lt ${FREQ_min} ]; then
set_freq_min || return $?
set_freq_max
else
set_freq_max || return $?
set_freq_min
fi
elif [ -n "${SET_MAX_FREQ}" ]; then
set_freq_max
elif [ -n "${SET_MIN_FREQ}" ]; then
set_freq_min
else
log "Unexpected call to set_freq()"
return 1
fi
}
#
# Helper for detect_throttling().
#
get_thrott_detect_pid() {
[ -e ${THROTT_DETECT_PID_FILE_PATH} ] || return 0
local pid
read pid < ${THROTT_DETECT_PID_FILE_PATH} || {
log ERROR "Failed to read pid from: %s" "${THROTT_DETECT_PID_FILE_PATH}"
return 1
}
local proc_path=/proc/${pid:-invalid}/cmdline
[ -r ${proc_path} ] && grep -qs "${0##*/}" ${proc_path} && {
printf "%s" "${pid}"
return 0
}
# Remove orphaned PID file
rm -rf ${THROTT_DETECT_PID_FILE_PATH}
return 1
}
#
# Control detection and reporting of GPU throttling events.
# arg1: start - run throttle detector in background
# stop - stop throttle detector process, if any
# status - verify if throttle detector is running
#
detect_throttling() {
local pid
pid=$(get_thrott_detect_pid)
case "$1" in
status)
printf "Throttling detector is "
[ -z "${pid}" ] && printf "not running\n" && return 0
printf "running (pid=%s)\n" ${pid}
;;
stop)
[ -z "${pid}" ] && return 0
log INFO "Stopping throttling detector (pid=%s)" "${pid}"
kill ${pid}; sleep 1; kill -0 ${pid} 2>/dev/null && kill -9 ${pid}
rm -rf ${THROTT_DETECT_PID_FILE_PATH}
;;
start)
[ -n "${pid}" ] && {
log WARN "Throttling detector is already running (pid=%s)" ${pid}
return 0
}
(
read_freq_info n RPn || exit $?
while true; do
sleep ${THROTT_DETECT_SLEEP_SEC}
read_freq_info n act min cur || exit $?
#
# The throttling seems to occur when act freq goes below min.
# However, it's necessary to exclude the idle states, where
# act freq normally reaches RPn and cur goes below min.
#
[ ${FREQ_act} -lt ${FREQ_min} ] && \
[ ${FREQ_act} -gt ${FREQ_RPn} ] && \
[ ${FREQ_cur} -ge ${FREQ_min} ] && \
printf "GPU throttling detected: act=%s min=%s cur=%s RPn=%s\n" \
${FREQ_act} ${FREQ_min} ${FREQ_cur} ${FREQ_RPn}
done
) &
pid=$!
log INFO "Started GPU throttling detector (pid=%s)" ${pid}
printf "%s\n" ${pid} > ${THROTT_DETECT_PID_FILE_PATH} || \
log WARN "Failed to write throttle detector PID file"
;;
esac
}
#
# Retrieve the list of online CPUs.
#
get_online_cpus() {
local path cpu_index
printf "0"
for path in $(grep 1 ${CPU_SYSFS_PREFIX}/cpu*/online); do
cpu_index=${path##*/cpu}
printf " %s" ${cpu_index%%/*}
done
}
#
# Helper to print sysfs path for the given CPU index and freq info.
#
# arg1: Frequency info sysfs name, one of *_CPU_FREQ_INFO constants above
# arg2: CPU index
#
print_cpu_freq_sysfs_path() {
printf ${CPU_FREQ_SYSFS_PATTERN} "$2" "$1"
}
#
# Read the specified CPU freq info from sysfs.
#
# arg1: CPU index
# arg2: Flag (y/n) to also enable printing the freq info.
# arg3...: Frequency info sysfs name(s), see *_CPU_FREQ_INFO constants above
# return: Global variable(s) CPU_FREQ_${arg} containing the requested information
#
read_cpu_freq_info() {
local var val info path cpu_index print=0 ret=0
cpu_index=$1
[ "$2" = "y" ] && print=1
shift 2
while [ $# -gt 0 ]; do
info=$1
shift
var=CPU_FREQ_${info}
path=$(print_cpu_freq_sysfs_path "${info}" ${cpu_index})
[ -r ${path} ] && read ${var} < ${path} || {
log ERROR "Failed to read CPU freq info from: %s" "${path}"
ret=1
continue
}
[ -n "${var}" ] || {
log ERROR "Got empty CPU freq info from: %s" "${path}"
ret=1
continue
}
[ ${print} -eq 1 ] && {
eval val=\$${var}
printf "%6s: %4s Hz\n" "${info}" "${val}"
}
done
return ${ret}
}
#
# Helper to print freq. value as requested by user via '--cpu-set-max' option.
# arg1: user requested freq value
#
compute_cpu_freq_set() {
local val
case "$1" in
+)
val=${CPU_FREQ_cpuinfo_max}
;;
-)
val=${CPU_FREQ_cpuinfo_min}
;;
*%)
val=$((${1%?} * CPU_FREQ_cpuinfo_max / 100))
;;
*[!0-9]*)
log ERROR "Cannot set CPU freq to invalid value: %s" "$1"
return 1
;;
"")
log ERROR "Cannot set CPU freq to unspecified value"
return 1
;;
*)
log ERROR "Cannot set CPU freq to custom value; use +, -, or % instead"
return 1
;;
esac
printf "%s" "${val}"
}
#
# Adjust CPU max scaling frequency.
#
set_cpu_freq_max() {
local target_freq res=0
case "${CPU_SET_MAX_FREQ}" in
+)
target_freq=100
;;
-)
target_freq=1
;;
*%)
target_freq=${CPU_SET_MAX_FREQ%?}
;;
*)
log ERROR "Invalid CPU freq"
return 1
;;
esac
local pstate_info=$(printf "${CPU_PSTATE_SYSFS_PATTERN}" max_perf_pct)
[ -e "${pstate_info}" ] && {
log INFO "Setting intel_pstate max perf to %s" "${target_freq}%"
if ! printf "%s" "${target_freq}" > "${pstate_info}";
then
log ERROR "Failed to set intel_pstate max perf"
res=1
fi
}
local cpu_index
for cpu_index in $(get_online_cpus); do
read_cpu_freq_info ${cpu_index} n ${CAP_CPU_FREQ_INFO} || { res=$?; continue; }
target_freq=$(compute_cpu_freq_set "${CPU_SET_MAX_FREQ}")
tf_res=$?
[ -z "${target_freq}" ] && { res=$tf_res; continue; }
log INFO "Setting CPU%s max scaling freq to %s Hz" ${cpu_index} "${target_freq}"
[ -n "${DRY_RUN}" ] && continue
if ! printf "%s" ${target_freq} > $(print_cpu_freq_sysfs_path scaling_max ${cpu_index});
then
res=1
log ERROR "Failed to set CPU%s max scaling frequency" ${cpu_index}
fi
done
return ${res}
}
#
# Show help message.
#
print_usage() {
cat <<EOF
Usage: ${0##*/} [OPTION]...
A script to manage Intel GPU frequencies. Can be used for debugging performance
problems or trying to obtain a stable frequency while benchmarking.
Note Intel GPUs only accept specific frequencies, usually multiples of 50 MHz.
Options:
-g, --get [act|enf|cap|all]
Get frequency information: active (default), enforced,
hardware capabilities or all of them.
-s, --set [{min|max}=]{FREQUENCY[%]|+|-}
Set min or max frequency to the given value (MHz).
Append '%' to interpret FREQUENCY as % of hw max.
Use '+' or '-' to set frequency to hardware max or min.
Omit min/max prefix to set both frequencies.
-r, --reset Reset frequencies to hardware defaults.
-m, --monitor [act|enf|cap|all]
Monitor the indicated frequencies via 'watch' utility.
See '-g, --get' option for more details.
-d|--detect-thrott [start|stop|status]
Start (default operation) the throttling detector
as a background process. Use 'stop' or 'status' to
terminate the detector process or verify its status.
--cpu-set-max [FREQUENCY%|+|-}
Set CPU max scaling frequency as % of hw max.
Use '+' or '-' to set frequency to hardware max or min.
-r, --reset Reset frequencies to hardware defaults.
--dry-run See what the script will do without applying any
frequency changes.
-h, --help Display this help text and exit.
EOF
}
#
# Parse user input for '-g, --get' option.
# Returns 0 if a value has been provided, otherwise 1.
#
parse_option_get() {
local ret=0
case "$1" in
act) GET_ACT_FREQ=1;;
enf) GET_ENF_FREQ=1;;
cap) GET_CAP_FREQ=1;;
all) GET_ACT_FREQ=1; GET_ENF_FREQ=1; GET_CAP_FREQ=1;;
-*|"")
# No value provided, using default.
GET_ACT_FREQ=1
ret=1
;;
*)
print_usage
exit 1
;;
esac
return ${ret}
}
#
# Validate user input for '-s, --set' option.
# arg1: input value to be validated
# arg2: optional flag indicating input is restricted to %
#
validate_option_set() {
case "$1" in
+|-|[0-9]%|[0-9][0-9]%)
return 0
;;
*[!0-9]*|"")
print_usage
exit 1
;;
esac
[ -z "$2" ] || { print_usage; exit 1; }
}
#
# Parse script arguments.
#
[ $# -eq 0 ] && { print_usage; exit 1; }
while [ $# -gt 0 ]; do
case "$1" in
-g|--get)
parse_option_get "$2" && shift
;;
-s|--set)
shift
case "$1" in
min=*)
SET_MIN_FREQ=${1#min=}
validate_option_set "${SET_MIN_FREQ}"
;;
max=*)
SET_MAX_FREQ=${1#max=}
validate_option_set "${SET_MAX_FREQ}"
;;
*)
SET_MIN_FREQ=$1
validate_option_set "${SET_MIN_FREQ}"
SET_MAX_FREQ=${SET_MIN_FREQ}
;;
esac
;;
-r|--reset)
RESET_FREQ=1
SET_MIN_FREQ="-"
SET_MAX_FREQ="+"
;;
-m|--monitor)
MONITOR_FREQ=act
parse_option_get "$2" && MONITOR_FREQ=$2 && shift
;;
-d|--detect-thrott)
DETECT_THROTT=start
case "$2" in
start|stop|status)
DETECT_THROTT=$2
shift
;;
esac
;;
--cpu-set-max)
shift
CPU_SET_MAX_FREQ=$1
validate_option_set "${CPU_SET_MAX_FREQ}" restricted
;;
--dry-run)
DRY_RUN=1
;;
-h|--help)
print_usage
exit 0
;;
*)
print_usage
exit 1
;;
esac
shift
done
#
# Main
#
RET=0
identify_intel_gpu || {
log INFO "No Intel GPU detected"
exit 0
}
[ -n "${SET_MIN_FREQ}${SET_MAX_FREQ}" ] && { set_freq || RET=$?; }
print_freq_info
[ -n "${DETECT_THROTT}" ] && detect_throttling ${DETECT_THROTT}
[ -n "${CPU_SET_MAX_FREQ}" ] && { set_cpu_freq_max || RET=$?; }
[ -n "${MONITOR_FREQ}" ] && {
log INFO "Entering frequency monitoring mode"
sleep 2
exec watch -d -n 1 "$0" -g "${MONITOR_FREQ}"
}
exit ${RET}

View File

@@ -1,18 +0,0 @@
#!/usr/bin/env bash
# shellcheck disable=SC1091 # the path is created in build-kdl and
# here is check if exist
# shellcheck disable=SC2086 # we want the arguments to be expanded
if ! [ -f /ci-kdl/bin/activate ]; then
echo -e "ci-kdl not installed; not monitoring temperature"
exit 0
fi
KDL_ARGS="
--output-file=${RESULTS_DIR}/kdl.json
--log-level=WARNING
--num-samples=-1
"
source /ci-kdl/bin/activate
exec /ci-kdl/bin/ci-kdl ${KDL_ARGS}

View File

@@ -1,81 +0,0 @@
#!/usr/bin/env bash
# shellcheck disable=SC1091
# When changing this file, you need to bump the following
# .gitlab-ci/image-tags.yml tags:
# ALPINE_X86_64_BUILD_TAG
set -e
. .gitlab-ci/setup-test-env.sh
set -o xtrace
EPHEMERAL=(
)
DEPS=(
bash
bison
ccache
"clang${LLVM_VERSION}-dev"
cmake
clang-dev
coreutils
curl
flex
gcc
g++
git
gettext
glslang
graphviz
linux-headers
"llvm${LLVM_VERSION}-static"
"llvm${LLVM_VERSION}-dev"
meson
mold
musl-dev
expat-dev
elfutils-dev
libclc-dev
libdrm-dev
libva-dev
libpciaccess-dev
zlib-dev
python3-dev
py3-clang
py3-cparser
py3-mako
py3-packaging
py3-pip
py3-ply
py3-yaml
vulkan-headers
spirv-tools-dev
spirv-llvm-translator-dev
util-macros
wayland-dev
wayland-protocols
)
apk --no-cache add "${DEPS[@]}" "${EPHEMERAL[@]}"
pip3 install --break-system-packages sphinx===5.1.1 hawkmoth===0.16.0
. .gitlab-ci/container/container_pre_build.sh
############### Uninstall the build software
# too many vendor binarise, just keep the ones we need
find /usr/share/clc \
\( -type f -o -type l \) \
! -name 'spirv-mesa3d-.spv' \
! -name 'spirv64-mesa3d-.spv' \
-delete
apk del "${EPHEMERAL[@]}"
. .gitlab-ci/container/container_post_build.sh

View File

@@ -1,32 +0,0 @@
#!/usr/bin/env bash
# This is a ci-templates build script to generate a container for LAVA SSH client.
# shellcheck disable=SC1091
set -e
. .gitlab-ci/setup-test-env.sh
set -o xtrace
EPHEMERAL=(
)
# We only need these very basic packages to run the tests.
DEPS=(
openssh-client # for ssh
iputils # for ping
bash
curl
)
apk --no-cache add "${DEPS[@]}" "${EPHEMERAL[@]}"
. .gitlab-ci/container/container_pre_build.sh
############### Uninstall the build software
apk del "${EPHEMERAL[@]}"
. .gitlab-ci/container/container_post_build.sh

View File

@@ -1,62 +0,0 @@
#!/usr/bin/env bash
set -e
set -o xtrace
# Fetch the arm-built rootfs image and unpack it in our x86_64 container (saves
# network transfer, disk usage, and runtime on test jobs)
# shellcheck disable=SC2154 # arch is assigned in previous scripts
if curl --fail -X HEAD -s "${ARTIFACTS_PREFIX}/${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
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
"${ARTIFACTS_URL}"/lava-rootfs.tar.zst -o rootfs.tar.zst
mkdir -p /rootfs-"$arch"
tar -C /rootfs-"$arch" '--exclude=./dev/*' --zstd -xf rootfs.tar.zst
rm rootfs.tar.zst
if [[ $arch == "arm64" ]]; then
mkdir -p /baremetal-files
pushd /baremetal-files
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
-O "${KERNEL_IMAGE_BASE}"/arm64/Image
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
-O "${KERNEL_IMAGE_BASE}"/arm64/Image.gz
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
-O "${KERNEL_IMAGE_BASE}"/arm64/cheza-kernel
DEVICE_TREES=""
DEVICE_TREES="$DEVICE_TREES apq8016-sbc-usb-host.dtb"
DEVICE_TREES="$DEVICE_TREES apq8096-db820c.dtb"
DEVICE_TREES="$DEVICE_TREES tegra210-p3450-0000.dtb"
DEVICE_TREES="$DEVICE_TREES imx8mq-nitrogen.dtb"
for DTB in $DEVICE_TREES; do
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
-O "${KERNEL_IMAGE_BASE}/arm64/$DTB"
done
popd
elif [[ $arch == "armhf" ]]; then
mkdir -p /baremetal-files
pushd /baremetal-files
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
-O "${KERNEL_IMAGE_BASE}"/armhf/zImage
DEVICE_TREES=""
DEVICE_TREES="$DEVICE_TREES imx6q-cubox-i.dtb"
DEVICE_TREES="$DEVICE_TREES tegra124-jetson-tk1.dtb"
for DTB in $DEVICE_TREES; do
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
-O "${KERNEL_IMAGE_BASE}/armhf/$DTB"
done
popd
fi

View File

@@ -1,119 +0,0 @@
#!/usr/bin/env bash
set -exu
# If CI vars are not set, assign an empty value, this prevents -u to fail
: "${CI:=}"
: "${CI_PROJECT_PATH:=}"
# Early check for required env variables, relies on `set -u`
: "$ANDROID_NDK_VERSION"
: "$ANDROID_SDK_VERSION"
: "$ANDROID_LLVM_VERSION"
: "$ANDROID_LLVM_ARTIFACT_NAME"
: "$S3_JWT_FILE"
: "$S3_HOST"
: "$S3_ANDROID_BUCKET"
# Check for CI if the auth file used later on is non-empty
if [ -n "$CI" ] && [ ! -s "${S3_JWT_FILE}" ]; then
echo "Error: ${S3_JWT_FILE} is empty." 1>&2
exit 1
fi
if curl -s -o /dev/null -I -L -f --retry 4 --retry-delay 15 "https://${S3_HOST}/${S3_ANDROID_BUCKET}/${CI_PROJECT_PATH}/${ANDROID_LLVM_ARTIFACT_NAME}.tar.zst"; then
echo "Artifact ${ANDROID_LLVM_ARTIFACT_NAME}.tar.zst already exists, skip re-building."
# Download prebuilt LLVM libraries for Android when they have not changed,
# to save some time
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
-o "/${ANDROID_LLVM_ARTIFACT_NAME}.tar.zst" "https://${S3_HOST}/${S3_ANDROID_BUCKET}/${CI_PROJECT_PATH}/${ANDROID_LLVM_ARTIFACT_NAME}.tar.zst"
tar -C / --zstd -xf "/${ANDROID_LLVM_ARTIFACT_NAME}.tar.zst"
rm "/${ANDROID_LLVM_ARTIFACT_NAME}.tar.zst"
exit
fi
# Install some dependencies needed to build LLVM
EPHEMERAL=(
ninja-build
unzip
)
apt-get update
apt-get install -y --no-install-recommends --no-remove "${EPHEMERAL[@]}"
ANDROID_NDK="android-ndk-${ANDROID_NDK_VERSION}"
ANDROID_NDK_ROOT="/${ANDROID_NDK}"
if [ ! -d "$ANDROID_NDK_ROOT" ];
then
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
-o "${ANDROID_NDK}.zip" \
"https://dl.google.com/android/repository/${ANDROID_NDK}-linux.zip"
unzip -d / "${ANDROID_NDK}.zip" "$ANDROID_NDK/source.properties" "$ANDROID_NDK/build/cmake/*" "$ANDROID_NDK/toolchains/llvm/*"
rm "${ANDROID_NDK}.zip"
fi
if [ ! -d "/llvm-project" ];
then
mkdir "/llvm-project"
pushd "/llvm-project"
git init
git remote add origin https://github.com/llvm/llvm-project.git
git fetch --depth 1 origin "$ANDROID_LLVM_VERSION"
git checkout FETCH_HEAD
popd
fi
pushd "/llvm-project"
# Checkout again the intended version, just in case of a pre-existing full clone
git checkout "$ANDROID_LLVM_VERSION" || true
LLVM_INSTALL_PREFIX="/${ANDROID_LLVM_ARTIFACT_NAME}"
rm -rf build/
cmake -GNinja -S llvm -B build/ \
-DCMAKE_TOOLCHAIN_FILE="${ANDROID_NDK_ROOT}/build/cmake/android.toolchain.cmake" \
-DANDROID_ABI=x86_64 \
-DANDROID_PLATFORM="android-${ANDROID_SDK_VERSION}" \
-DANDROID_NDK="${ANDROID_NDK_ROOT}" \
-DCMAKE_ANDROID_ARCH_ABI=x86_64 \
-DCMAKE_ANDROID_NDK="${ANDROID_NDK_ROOT}" \
-DCMAKE_BUILD_TYPE=MinSizeRel \
-DCMAKE_SYSTEM_NAME=Android \
-DCMAKE_SYSTEM_VERSION="${ANDROID_SDK_VERSION}" \
-DCMAKE_INSTALL_PREFIX="${LLVM_INSTALL_PREFIX}" \
-DCMAKE_CXX_FLAGS="-march=x86-64 --target=x86_64-linux-android${ANDROID_SDK_VERSION} -fno-rtti" \
-DLLVM_HOST_TRIPLE="x86_64-linux-android${ANDROID_SDK_VERSION}" \
-DLLVM_TARGETS_TO_BUILD=X86 \
-DLLVM_BUILD_LLVM_DYLIB=OFF \
-DLLVM_BUILD_TESTS=OFF \
-DLLVM_BUILD_EXAMPLES=OFF \
-DLLVM_BUILD_DOCS=OFF \
-DLLVM_BUILD_TOOLS=OFF \
-DLLVM_ENABLE_RTTI=OFF \
-DLLVM_BUILD_INSTRUMENTED_COVERAGE=OFF \
-DLLVM_NATIVE_TOOL_DIR="${ANDROID_NDK_ROOT}/toolchains/llvm/prebuilt/linux-x86_64/bin" \
-DLLVM_ENABLE_PIC=False \
-DLLVM_OPTIMIZED_TABLEGEN=ON
ninja "-j${FDO_CI_CONCURRENT:-4}" -C build/ install
popd
rm -rf /llvm-project
tar --zstd -cf "${ANDROID_LLVM_ARTIFACT_NAME}.tar.zst" "$LLVM_INSTALL_PREFIX"
# If run in CI upload the tar.zst archive to S3 to avoid rebuilding it if the
# version does not change, and delete it.
# The file is not deleted for non-CI because it can be useful in local runs.
if [ -n "$CI" ]; then
s3_upload "${ANDROID_LLVM_ARTIFACT_NAME}.tar.zst" "https://${S3_HOST}/${S3_ANDROID_BUCKET}/${CI_PROJECT_PATH}/"
rm "${ANDROID_LLVM_ARTIFACT_NAME}.tar.zst"
fi
rm -rf "$LLVM_INSTALL_PREFIX"
apt-get purge -y "${EPHEMERAL[@]}"

View File

@@ -1,103 +0,0 @@
#!/usr/bin/env bash
# When changing this file, you need to bump the following
# .gitlab-ci/image-tags.yml tags:
# KERNEL_ROOTFS_TAG
set -uex
uncollapsed_section_start angle "Building angle"
ANGLE_REV="76025caa1a059f464a2b0e8f879dbd4746f092b9"
SCRIPTS_DIR="$(pwd)/.gitlab-ci"
ANGLE_PATCH_DIR="${SCRIPTS_DIR}/container/patches"
# DEPOT tools
git clone --depth 1 https://chromium.googlesource.com/chromium/tools/depot_tools.git /depot-tools
export PATH=/depot-tools:$PATH
export DEPOT_TOOLS_UPDATE=0
mkdir /angle-build
mkdir /angle
pushd /angle-build
git init
git remote add origin https://chromium.googlesource.com/angle/angle.git
git fetch --depth 1 origin "$ANGLE_REV"
git checkout FETCH_HEAD
angle_patch_files=(
build-angle_deps_Make-more-sources-conditional.patch
)
for patch in "${angle_patch_files[@]}"; do
echo "Apply patch to ANGLE from ${patch}"
GIT_COMMITTER_DATE="$(LC_TIME=C date -d@0)" git am < "${ANGLE_PATCH_DIR}/${patch}"
done
{
echo "ANGLE base version $ANGLE_REV"
echo "The following local patches are applied on top:"
git log --reverse --oneline $ANGLE_REV.. --format='- %s'
} > /angle/version
# source preparation
gclient config --name REPLACE-WITH-A-DOT --unmanaged \
--custom-var='angle_enable_cl=False' \
--custom-var='angle_enable_cl_testing=False' \
--custom-var='angle_enable_vulkan_validation_layers=False' \
--custom-var='angle_enable_wgpu=False' \
--custom-var='build_allow_regenerate=False' \
--custom-var='build_angle_deqp_tests=False' \
--custom-var='build_angle_perftests=False' \
--custom-var='build_with_catapult=False' \
--custom-var='build_with_swiftshader=False' \
https://chromium.googlesource.com/angle/angle.git
sed -e 's/REPLACE-WITH-A-DOT/./;' -i .gclient
gclient sync -j"${FDO_CI_CONCURRENT:-4}"
mkdir -p out/Release
echo '
angle_build_all=false
angle_build_tests=false
angle_enable_cl=false
angle_enable_cl_testing=false
angle_enable_gl=false
angle_enable_gl_desktop_backend=false
angle_enable_null=false
angle_enable_swiftshader=false
angle_enable_trace=false
angle_enable_wgpu=false
angle_enable_vulkan=true
angle_enable_vulkan_api_dump_layer=false
angle_enable_vulkan_validation_layers=false
angle_has_frame_capture=false
angle_has_histograms=false
angle_use_custom_libvulkan=false
angle_egl_extension="so.1"
angle_glesv2_extension="so.2"
build_angle_deqp_tests=false
dcheck_always_on=true
enable_expensive_dchecks=false
is_debug=false
' > out/Release/args.gn
if [[ "$DEBIAN_ARCH" = "arm64" ]]; then
build/linux/sysroot_scripts/install-sysroot.py --arch=arm64
fi
gn gen out/Release
# depot_tools overrides ninja with a version that doesn't work. We want
# ninja with FDO_CI_CONCURRENT anyway.
/usr/local/bin/ninja -C out/Release/ libEGL libGLESv2
rm -f out/Release/libvulkan.so* out/Release/*.so.TOC
cp out/Release/lib*.so* /angle/
ln -s libEGL.so.1 /angle/libEGL.so
ln -s libGLESv2.so.2 /angle/libGLESv2.so
rm -rf out
popd
rm -rf /depot-tools
rm -rf /angle-build
section_end angle

View File

@@ -1,29 +0,0 @@
#!/usr/bin/env bash
# shellcheck disable=SC2086 # we want word splitting
# When changing this file, you need to bump the following
# .gitlab-ci/image-tags.yml tags:
# DEBIAN_TEST_GL_TAG
# DEBIAN_TEST_VK_TAG
# KERNEL_ROOTFS_TAG
set -uex
uncollapsed_section_start apitrace "Building apitrace"
APITRACE_VERSION="0a6506433e1f9f7b69757b4e5730326970c4321a"
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:-}
cmake --build _build --parallel --target apitrace eglretrace
mkdir build
cp _build/apitrace build
cp _build/eglretrace build
${STRIP_CMD:-strip} build/*
find . -not -path './build' -not -path './build/*' -delete
popd
section_end apitrace

View File

@@ -1,23 +0,0 @@
#!/usr/bin/env bash
# shellcheck disable=SC2086 # we want word splitting
uncollapsed_section_start bindgen "Building bindgen"
BINDGEN_VER=0.65.1
CBINDGEN_VER=0.26.0
# bindgen
RUSTFLAGS='-L native=/usr/local/lib' cargo install \
bindgen-cli --version ${BINDGEN_VER} \
--locked \
-j ${FDO_CI_CONCURRENT:-4} \
--root /usr/local
# cbindgen
RUSTFLAGS='-L native=/usr/local/lib' cargo install \
cbindgen --version ${CBINDGEN_VER} \
--locked \
-j ${FDO_CI_CONCURRENT:-4} \
--root /usr/local
section_end bindgen

View File

@@ -1,55 +0,0 @@
#!/usr/bin/env bash
# shellcheck disable=SC2086 # we want word splitting
# When changing this file, you need to bump the following
# .gitlab-ci/image-tags.yml tags:
# DEBIAN_BASE_TAG
# DEBIAN_TEST_GL_TAG
# DEBIAN_TEST_VK_TAG
# KERNEL_ROOTFS_TAG
set -uex
uncollapsed_section_start crosvm "Building crosvm"
git config --global user.email "mesa@example.com"
git config --global user.name "Mesa CI"
CROSVM_VERSION=2118fbb57ca26b495a9aa407845c7729d697a24b
git clone --single-branch -b main --no-checkout https://chromium.googlesource.com/crosvm/crosvm /platform/crosvm
pushd /platform/crosvm
git checkout "$CROSVM_VERSION"
git submodule update --init
VIRGLRENDERER_VERSION=57a2b82e0958f08d02ade8400786e1ca0935c9b1
rm -rf third_party/virglrenderer
git clone --single-branch -b main --no-checkout https://gitlab.freedesktop.org/virgl/virglrenderer.git third_party/virglrenderer
pushd third_party/virglrenderer
git checkout "$VIRGLRENDERER_VERSION"
meson setup build/ -D libdir=lib -D render-server-worker=process -D venus=true ${EXTRA_MESON_ARGS:-}
meson install -C build
popd
cargo update -p pkg-config@0.3.26 --precise 0.3.27
RUSTFLAGS='-L native=/usr/local/lib' cargo install \
bindgen-cli \
--locked \
-j ${FDO_CI_CONCURRENT:-4} \
--root /usr/local \
--version 0.65.1 \
${EXTRA_CARGO_ARGS:-}
CROSVM_USE_SYSTEM_MINIGBM=1 CROSVM_USE_SYSTEM_VIRGLRENDERER=1 RUSTFLAGS='-L native=/usr/local/lib' cargo install \
-j ${FDO_CI_CONCURRENT:-4} \
--locked \
--features 'default-no-sandbox gpu x virgl_renderer' \
--path . \
--root /usr/local \
${EXTRA_CARGO_ARGS:-}
popd
rm -rf /platform/crosvm
section_end crosvm

View File

@@ -1,100 +0,0 @@
#!/usr/bin/env bash
# shellcheck disable=SC2086 # we want word splitting
# When changing this file, you need to bump the following
# .gitlab-ci/image-tags.yml tags:
# DEBIAN_TEST_ANDROID_TAG
# DEBIAN_BASE_TAG
# KERNEL_ROOTFS_TAG
set -uex
uncollapsed_section_start deqp-runner "Building deqp-runner"
DEQP_RUNNER_VERSION=0.20.3
commits_to_backport=(
)
patch_files=(
)
DEQP_RUNNER_GIT_URL="${DEQP_RUNNER_GIT_URL:-https://gitlab.freedesktop.org/mesa/deqp-runner.git}"
if [ -n "${DEQP_RUNNER_GIT_TAG:-}" ]; then
DEQP_RUNNER_GIT_CHECKOUT="$DEQP_RUNNER_GIT_TAG"
elif [ -n "${DEQP_RUNNER_GIT_REV:-}" ]; then
DEQP_RUNNER_GIT_CHECKOUT="$DEQP_RUNNER_GIT_REV"
else
DEQP_RUNNER_GIT_CHECKOUT="v$DEQP_RUNNER_VERSION"
fi
BASE_PWD=$PWD
mkdir -p /deqp-runner
pushd /deqp-runner
mkdir deqp-runner-git
pushd deqp-runner-git
git init
git remote add origin "$DEQP_RUNNER_GIT_URL"
git fetch --depth 1 origin "$DEQP_RUNNER_GIT_CHECKOUT"
git checkout FETCH_HEAD
for commit in "${commits_to_backport[@]}"
do
PATCH_URL="https://gitlab.freedesktop.org/mesa/deqp-runner/-/commit/$commit.patch"
echo "Backport deqp-runner commit $commit from $PATCH_URL"
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 $PATCH_URL | git am
done
for patch in "${patch_files[@]}"
do
echo "Apply patch to deqp-runner from $patch"
git am "$BASE_PWD/.gitlab-ci/container/patches/$patch"
done
if [ -z "${RUST_TARGET:-}" ]; then
RUST_TARGET=""
fi
if [[ "$RUST_TARGET" != *-android ]]; then
# When CC (/usr/lib/ccache/gcc) variable is set, the rust compiler uses
# this variable when cross-compiling arm32 and build fails for zsys-sys.
# So unset the CC variable when cross-compiling for arm32.
SAVEDCC=${CC:-}
if [ "$RUST_TARGET" = "armv7-unknown-linux-gnueabihf" ]; then
unset CC
fi
cargo install --locked \
-j ${FDO_CI_CONCURRENT:-4} \
--root /usr/local \
${EXTRA_CARGO_ARGS:-} \
--path .
CC=$SAVEDCC
else
cargo install --locked \
-j ${FDO_CI_CONCURRENT:-4} \
--root /usr/local --version 2.10.0 \
cargo-ndk
rustup target add $RUST_TARGET
RUSTFLAGS='-C target-feature=+crt-static' cargo ndk --target $RUST_TARGET build --release
mv target/$RUST_TARGET/release/deqp-runner /deqp-runner
cargo uninstall --locked \
--root /usr/local \
cargo-ndk
fi
popd
rm -rf deqp-runner-git
popd
# remove unused test runners to shrink images for the Mesa CI build (not kernel,
# which chooses its own deqp branch)
if [ -z "${DEQP_RUNNER_GIT_TAG:-}${DEQP_RUNNER_GIT_REV:-}" ]; then
rm -f /usr/local/bin/igt-runner
fi
section_end deqp-runner

View File

@@ -1,321 +0,0 @@
#!/usr/bin/env bash
# shellcheck disable=SC2086 # we want word splitting
# When changing this file, you need to bump the following
# .gitlab-ci/image-tags.yml tags:
# DEBIAN_TEST_ANDROID_TAG
# DEBIAN_TEST_GL_TAG
# DEBIAN_TEST_VK_TAG
# KERNEL_ROOTFS_TAG
set -ue -o pipefail
# shellcheck disable=SC2153
deqp_api=${DEQP_API,,}
uncollapsed_section_start deqp-$deqp_api "Building dEQP $DEQP_API"
set -x
# See `deqp_build_targets` below for which release is used to produce which
# binary. Unless this comment has bitrotten:
# - the commit from the main branch produces the deqp tools and `deqp-vk`,
# - the VK release produces `deqp-vk`,
# - the GL release produces `glcts`, and
# - the GLES release produces `deqp-gles*` and `deqp-egl`
DEQP_MAIN_COMMIT=a9988483c0864d7190e5e6264ccead95423dfd00
DEQP_VK_VERSION=1.4.1.1
DEQP_GL_VERSION=4.6.5.0
DEQP_GLES_VERSION=3.2.11.0
# Patches to VulkanCTS may come from commits in their repo (listed in
# cts_commits_to_backport) or patch files stored in our repo (in the patch
# directory `$OLDPWD/.gitlab-ci/container/patches/` listed in cts_patch_files).
# Both list variables would have comments explaining the reasons behind the
# patches.
# shellcheck disable=SC2034
main_cts_commits_to_backport=(
# If you find yourself wanting to add something in here, consider whether
# bumping DEQP_MAIN_COMMIT is not a better solution :)
)
# shellcheck disable=SC2034
main_cts_patch_files=(
)
# shellcheck disable=SC2034
vk_cts_commits_to_backport=(
)
# shellcheck disable=SC2034
vk_cts_patch_files=(
)
# shellcheck disable=SC2034
gl_cts_commits_to_backport=(
# Add #include <cmath> in deMath.h when being compiled by C++
71808fe7d0a640dfd703e845d93ba1c5ab751055
# Revert "Add #include <cmath> in deMath.h when being compiled by C++ compiler"
# This also adds an alternative fix along with the revert.
6164879a0acce258637d261592a9c395e564b361
)
# shellcheck disable=SC2034
gl_cts_patch_files=(
build-deqp-gl_Build-Don-t-build-Vulkan-utilities-for-GL-builds.patch
)
if [ "${DEQP_TARGET}" = 'android' ]; then
gl_cts_patch_files+=(
build-deqp-gl_Allow-running-on-Android-from-the-command-line.patch
build-deqp-gl_Android-prints-to-stdout-instead-of-logcat.patch
)
fi
# shellcheck disable=SC2034
# GLES builds also EGL
gles_cts_commits_to_backport=(
# Add #include <cmath> in deMath.h when being compiled by C++
71808fe7d0a640dfd703e845d93ba1c5ab751055
# Revert "Add #include <cmath> in deMath.h when being compiled by C++ compiler"
# This also adds an alternative fix along with the revert.
6164879a0acce258637d261592a9c395e564b361
)
# shellcheck disable=SC2034
gles_cts_patch_files=(
build-deqp-gl_Build-Don-t-build-Vulkan-utilities-for-GL-builds.patch
)
if [ "${DEQP_TARGET}" = 'android' ]; then
gles_cts_patch_files+=(
build-deqp-gles_Allow-running-on-Android-from-the-command-line.patch
build-deqp-gles_Android-prints-to-stdout-instead-of-logcat.patch
)
fi
### Careful editing anything below this line
git config --global user.email "mesa@example.com"
git config --global user.name "Mesa CI"
# shellcheck disable=SC2153
case "${DEQP_API}" in
tools) DEQP_VERSION="$DEQP_MAIN_COMMIT";;
*-main) DEQP_VERSION="$DEQP_MAIN_COMMIT";;
VK) DEQP_VERSION="vulkan-cts-$DEQP_VK_VERSION";;
GL) DEQP_VERSION="opengl-cts-$DEQP_GL_VERSION";;
GLES) DEQP_VERSION="opengl-es-cts-$DEQP_GLES_VERSION";;
*) echo "Unexpected DEQP_API value: $DEQP_API"; exit 1;;
esac
mkdir -p /VK-GL-CTS
pushd /VK-GL-CTS
[ -e .git ] || {
git init
git remote add origin https://github.com/KhronosGroup/VK-GL-CTS.git
}
git fetch --depth 1 origin "$DEQP_VERSION"
git checkout FETCH_HEAD
DEQP_COMMIT=$(git rev-parse FETCH_HEAD)
if [ "$DEQP_VERSION" = "$DEQP_MAIN_COMMIT" ]; then
git fetch origin main
if ! git merge-base --is-ancestor "$DEQP_MAIN_COMMIT" origin/main; then
echo "VK-GL-CTS commit $DEQP_MAIN_COMMIT is not a commit from the main branch."
exit 1
fi
fi
mkdir -p /deqp-$deqp_api
if [ "$DEQP_VERSION" = "$DEQP_MAIN_COMMIT" ]; then
prefix="main"
else
prefix="$deqp_api"
fi
cts_commits_to_backport="${prefix}_cts_commits_to_backport[@]"
for commit in "${!cts_commits_to_backport}"
do
PATCH_URL="https://github.com/KhronosGroup/VK-GL-CTS/commit/$commit.patch"
echo "Apply patch to ${DEQP_API} CTS from $PATCH_URL"
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 $PATCH_URL | \
GIT_COMMITTER_DATE=$(LC_TIME=C date -d@0) git am -
done
cts_patch_files="${prefix}_cts_patch_files[@]"
for patch in "${!cts_patch_files}"
do
echo "Apply patch to ${DEQP_API} CTS from $patch"
GIT_COMMITTER_DATE=$(LC_TIME=C date -d@0) git am < $OLDPWD/.gitlab-ci/container/patches/$patch
done
{
if [ "$DEQP_VERSION" = "$DEQP_MAIN_COMMIT" ]; then
commit_desc=$(git show --no-patch --format='commit %h on %ci' --abbrev=10 "$DEQP_COMMIT")
echo "dEQP $DEQP_API at $commit_desc"
else
echo "dEQP $DEQP_API version $DEQP_VERSION"
fi
if [ "$(git rev-parse HEAD)" != "$DEQP_COMMIT" ]; then
echo "The following local patches are applied on top:"
git log --reverse --oneline "$DEQP_COMMIT".. --format='- %s'
fi
} > /deqp-$deqp_api/deqp-$deqp_api-version
# --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
if [[ "$DEQP_API" = tools ]]; then
# Save the testlog stylesheets:
cp doc/testlog-stylesheet/testlog.{css,xsl} /deqp-$deqp_api
fi
popd
deqp_build_targets=()
case "${DEQP_API}" in
VK|VK-main)
deqp_build_targets+=(deqp-vk)
;;
GL)
deqp_build_targets+=(glcts)
;;
GLES)
deqp_build_targets+=(deqp-gles{2,3,31})
deqp_build_targets+=(glcts) # needed for gles*-khr tests
# deqp-egl also comes from this build, but it is handled separately below.
;;
tools)
deqp_build_targets+=(testlog-to-xml)
deqp_build_targets+=(testlog-to-csv)
deqp_build_targets+=(testlog-to-junit)
;;
esac
OLD_IFS="$IFS"
IFS=";"
CMAKE_SBT="${deqp_build_targets[*]}"
IFS="$OLD_IFS"
pushd /deqp-$deqp_api
if [ "${DEQP_API}" = 'GLES' ]; then
if [ "${DEQP_TARGET}" = 'android' ]; then
cmake -S /VK-GL-CTS -B . -G Ninja \
-DDEQP_TARGET=android \
-DCMAKE_BUILD_TYPE=Release \
-DSELECTED_BUILD_TARGETS="deqp-egl" \
${EXTRA_CMAKE_ARGS:-}
ninja modules/egl/deqp-egl
mv modules/egl/deqp-egl{,-android}
else
# 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 \
-DSELECTED_BUILD_TARGETS="deqp-egl" \
${EXTRA_CMAKE_ARGS:-}
ninja modules/egl/deqp-egl
mv modules/egl/deqp-egl{,-x11}
cmake -S /VK-GL-CTS -B . -G Ninja \
-DDEQP_TARGET=wayland \
-DCMAKE_BUILD_TYPE=Release \
-DSELECTED_BUILD_TARGETS="deqp-egl" \
${EXTRA_CMAKE_ARGS:-}
ninja modules/egl/deqp-egl
mv modules/egl/deqp-egl{,-wayland}
fi
fi
cmake -S /VK-GL-CTS -B . -G Ninja \
-DDEQP_TARGET=${DEQP_TARGET} \
-DCMAKE_BUILD_TYPE=Release \
-DSELECTED_BUILD_TARGETS="${CMAKE_SBT}" \
${EXTRA_CMAKE_ARGS:-}
# Make sure `default` doesn't silently stop detecting one of the platforms we care about
if [ "${DEQP_TARGET}" = 'default' ]; then
grep -q DEQP_SUPPORT_WAYLAND=1 build.ninja
grep -q DEQP_SUPPORT_X11=1 build.ninja
grep -q DEQP_SUPPORT_XCB=1 build.ninja
fi
ninja "${deqp_build_targets[@]}"
if [ "$DEQP_API" != tools ]; then
# Copy out the mustpass lists we want.
mkdir -p mustpass
if [ "${DEQP_API}" = 'VK' ] || [ "${DEQP_API}" = 'VK-main' ]; then
for mustpass in $(< /VK-GL-CTS/external/vulkancts/mustpass/main/vk-default.txt) ; do
cat /VK-GL-CTS/external/vulkancts/mustpass/main/$mustpass \
>> mustpass/vk-main.txt
done
fi
if [ "${DEQP_API}" = 'GL' ]; then
cp \
/VK-GL-CTS/external/openglcts/data/gl_cts/data/mustpass/gl/khronos_mustpass/main/*-main.txt \
mustpass/
cp \
/VK-GL-CTS/external/openglcts/data/gl_cts/data/mustpass/gl/khronos_mustpass_single/main/*-single.txt \
mustpass/
fi
if [ "${DEQP_API}" = 'GLES' ]; then
cp \
/VK-GL-CTS/external/openglcts/data/gl_cts/data/mustpass/gles/aosp_mustpass/main/*.txt \
mustpass/
cp \
/VK-GL-CTS/external/openglcts/data/gl_cts/data/mustpass/egl/aosp_mustpass/main/egl-main.txt \
mustpass/
cp \
/VK-GL-CTS/external/openglcts/data/gl_cts/data/mustpass/gles/khronos_mustpass/main/*-main.txt \
mustpass/
fi
# Compress the caselists, since Vulkan's in particular are gigantic; higher
# compression levels provide no real measurable benefit.
zstd -1 --rm mustpass/*.txt
fi
if [ "$DEQP_API" = tools ]; then
# Save *some* executor utils, but otherwise strip things down
# to reduct deqp build size:
mv executor/testlog-to-* .
rm -rf executor
fi
# Remove other mustpass files, since we saved off the ones we wanted to conventient locations above.
rm -rf external/**/mustpass/
rm -rf external/vulkancts/modules/vulkan/vk-main*
rm -rf external/vulkancts/modules/vulkan/vk-default
rm -rf external/openglcts/modules/cts-runner
rm -rf modules/internal
rm -rf execserver
rm -rf framework
find . -depth \( -iname '*cmake*' -o -name '*ninja*' -o -name '*.o' -o -name '*.a' \) -exec rm -rf {} \;
if [ "${DEQP_API}" = 'VK' ] || [ "${DEQP_API}" = 'VK-main' ]; then
${STRIP_CMD:-strip} external/vulkancts/modules/vulkan/deqp-vk
fi
if [ "${DEQP_API}" = 'GL' ] || [ "${DEQP_API}" = 'GLES' ]; then
${STRIP_CMD:-strip} external/openglcts/modules/glcts
fi
if [ "${DEQP_API}" = 'GLES' ]; then
${STRIP_CMD:-strip} modules/*/deqp-*
fi
du -sh ./*
popd
section_end deqp-$deqp_api

View File

@@ -1,19 +0,0 @@
#!/usr/bin/env bash
# shellcheck disable=SC2086 # we want word splitting
# When changing this file, you need to bump the following
# .gitlab-ci/image-tags.yml tags:
# DEBIAN_BUILD_TAG
set -uex
uncollapsed_section_start directx-headers "Building directx-headers"
git clone https://github.com/microsoft/DirectX-Headers -b v1.614.1 --depth 1
pushd DirectX-Headers
meson setup build --backend=ninja --buildtype=release -Dbuild-test=false ${EXTRA_MESON_ARGS:-}
meson install -C build
popd
rm -rf DirectX-Headers
section_end directx-headers

View File

@@ -1,39 +0,0 @@
#!/usr/bin/env bash
# shellcheck disable=SC1091 # The relative paths in this file only become valid at runtime.
# shellcheck disable=SC2034 # Variables are used in scripts called from here
# shellcheck disable=SC2086 # we want word splitting
# Install fluster in /usr/local.
FLUSTER_REVISION="e997402978f62428fffc8e5a4a709690d9ca9bc5"
git clone https://github.com/fluendo/fluster.git --single-branch --no-checkout
pushd fluster || exit
git checkout ${FLUSTER_REVISION}
popd || exit
if [ "${SKIP_UPDATE_FLUSTER_VECTORS}" != 1 ]; then
# Download the necessary vectors: H264, H265 and VP9
# When updating FLUSTER_REVISION, make sure to update the vectors if necessary or
# fluster-runner will report Missing results.
fluster/fluster.py download \
JVT-AVC_V1 JVT-FR-EXT JVT-MVC JVT-SVC_V1 \
JCT-VC-3D-HEVC JCT-VC-HEVC_V1 JCT-VC-MV-HEVC JCT-VC-RExt JCT-VC-SCC JCT-VC-SHVC \
VP9-TEST-VECTORS-HIGH VP9-TEST-VECTORS
# Build fluster vectors archive and upload it
tar --zstd -cf "vectors.tar.zst" fluster/resources/
s3_upload vectors.tar.zst "https://${S3_PATH_FLUSTER}/"
touch /lava-files/done
s3_upload /lava-files/done "https://${S3_PATH_FLUSTER}/"
# Don't include the vectors in the rootfs
rm -fr fluster/resources/*
fi
mkdir -p "${ROOTFS}/usr/local/"
mv fluster "${ROOTFS}/usr/local/"

View File

@@ -1,23 +0,0 @@
#!/bin/bash
# When changing this file, you need to bump the following
# .gitlab-ci/image-tags.yml tags:
# DEBIAN_TEST_VK_TAG
# KERNEL_ROOTFS_TAG
set -ex
uncollapsed_section_start fossilize "Building fossilize"
git clone https://github.com/ValveSoftware/Fossilize.git
cd Fossilize
git checkout b43ee42bbd5631ea21fe9a2dee4190d5d875c327
git submodule update --init
mkdir build
cd build
cmake -S .. -B . -G Ninja -DCMAKE_BUILD_TYPE=Release
ninja -C . install
cd ../..
rm -rf Fossilize
section_end fossilize

View File

@@ -1,23 +0,0 @@
#!/usr/bin/env bash
set -ex
uncollapsed_section_start gfxreconstruct "Building gfxreconstruct"
GFXRECONSTRUCT_VERSION=761837794a1e57f918a85af7000b12e531b178ae
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 -DCMAKE_INSTALL_PREFIX:PATH=/gfxreconstruct/build -DBUILD_WERROR=OFF
cmake --build _build --parallel --target tools/{replay,info}/install/strip
find . -not -path './build' -not -path './build/*' -delete
popd
section_end gfxreconstruct

View File

@@ -1,32 +0,0 @@
#!/usr/bin/env bash
# shellcheck disable=SC1091 # the path is created by the script
set -ex
uncollapsed_section_start kdl "Building kdl"
KDL_REVISION="cbbe5fd54505fd03ee34f35bfd16794f0c30074f"
KDL_CHECKOUT_DIR="/tmp/ci-kdl.git"
mkdir -p ${KDL_CHECKOUT_DIR}
pushd ${KDL_CHECKOUT_DIR}
git init
git remote add origin https://gitlab.freedesktop.org/gfx-ci/ci-kdl.git
git fetch --depth 1 origin ${KDL_REVISION}
git checkout FETCH_HEAD
popd
# Run venv in a subshell, so we don't accidentally leak the venv state into
# calling scripts
(
python3 -m venv /ci-kdl
source /ci-kdl/bin/activate &&
pushd ${KDL_CHECKOUT_DIR} &&
pip install -r requirements.txt &&
pip install . &&
popd
)
rm -rf ${KDL_CHECKOUT_DIR}
section_end kdl

View File

@@ -1,35 +0,0 @@
#!/usr/bin/env bash
set -uex
uncollapsed_section_start libclc "Building libclc"
export LLVM_CONFIG="llvm-config-${LLVM_VERSION:?"llvm unset!"}"
LLVM_TAG="llvmorg-15.0.7"
$LLVM_CONFIG --version
git config --global user.email "mesa@example.com"
git config --global user.name "Mesa CI"
git clone \
https://github.com/llvm/llvm-project \
--depth 1 \
-b "${LLVM_TAG}" \
/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
section_end libclc

View File

@@ -1,21 +0,0 @@
#!/usr/bin/env bash
# Script used for Android and Fedora builds (Debian builds get their libdrm version
# from https://gitlab.freedesktop.org/gfx-ci/ci-deb-repo - see PKG_REPO_REV)
# shellcheck disable=SC2086 # we want word splitting
set -uex
uncollapsed_section_start libdrm "Building libdrm"
export LIBDRM_VERSION=libdrm-2.4.122
curl -L -O --retry 4 -f --retry-all-errors --retry-delay 60 \
https://dri.freedesktop.org/libdrm/"$LIBDRM_VERSION".tar.xz
tar -xvf "$LIBDRM_VERSION".tar.xz && rm "$LIBDRM_VERSION".tar.xz
cd "$LIBDRM_VERSION"
meson setup build -D vc4=disabled -D freedreno=disabled -D etnaviv=disabled ${EXTRA_MESON_ARGS:-}
meson install -C build
cd ..
rm -rf "$LIBDRM_VERSION"
section_end libdrm

View File

@@ -1,30 +0,0 @@
#!/usr/bin/env bash
set -ex
uncollapsed_section_start llvm-spirv "Building LLVM-SPIRV-Translator"
if [ "${LLVM_VERSION:?llvm version not set}" -ge 18 ]; then
VER="${LLVM_VERSION}.1.0"
else
VER="${LLVM_VERSION}.0.0"
fi
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
-O "https://github.com/KhronosGroup/SPIRV-LLVM-Translator/archive/refs/tags/v${VER}.tar.gz"
tar -xvf "v${VER}.tar.gz" && rm "v${VER}.tar.gz"
mkdir "SPIRV-LLVM-Translator-${VER}/build"
pushd "SPIRV-LLVM-Translator-${VER}/build"
cmake .. -G Ninja -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=/usr
ninja
ninja install
# For some reason llvm-spirv is not installed by default
ninja llvm-spirv
cp tools/llvm-spirv/llvm-spirv /usr/bin/
popd
du -sh "SPIRV-LLVM-Translator-${VER}"
rm -rf "SPIRV-LLVM-Translator-${VER}"
section_end llvm-spirv

View File

@@ -1,32 +0,0 @@
#!/usr/bin/env bash
set -ex
# When changing this file, you need to bump the following
# .gitlab-ci/image-tags.yml tags:
# ALPINE_X86_64_BUILD_TAG
# DEBIAN_BASE_TAG
# DEBIAN_BUILD_TAG
# FEDORA_X86_64_BUILD_TAG
# KERNEL_ROOTFS_TAG
uncollapsed_section_start mold "Building mold"
MOLD_VERSION="2.32.0"
git clone -b v"$MOLD_VERSION" --single-branch --depth 1 https://github.com/rui314/mold.git
pushd mold
cmake -DCMAKE_BUILD_TYPE=Release -D BUILD_TESTING=OFF -D MOLD_LTO=ON
cmake --build . --parallel "${FDO_CI_CONCURRENT:-4}"
cmake --install . --strip
# Always use mold from now on
find /usr/bin \( -name '*-ld' -o -name 'ld' \) \
-exec ln -sf /usr/local/bin/ld.mold {} \; \
-exec ls -l {} +
popd
rm -rf mold
section_end mold

View File

@@ -1,29 +0,0 @@
#!/usr/bin/env bash
# When changing this file, you need to bump the following
# .gitlab-ci/image-tags.yml tags:
# DEBIAN_TEST_GL_TAG
set -ex -o pipefail
uncollapsed_section_start ninetests "Building Nine tests"
### Careful editing anything below this line
git config --global user.email "mesa@example.com"
git config --global user.name "Mesa CI"
git clone https://github.com/axeldavy/Xnine.git /Xnine
mkdir /Xnine/build
pushd /Xnine/build
git checkout c64753d224c08006bcdcfa7880ada826f27164b1
cmake .. -DBUILD_TESTS=1 -DWITH_DRI3=1 -DD3DADAPTER9_LOCATION=/install/lib/d3d/d3dadapter9.so
make
mkdir -p /NineTests/
mv NineTests/NineTests /NineTests/
popd
rm -rf /Xnine
section_end ninetests

View File

@@ -1,38 +0,0 @@
#!/bin/bash
# shellcheck disable=SC2086 # we want word splitting
set -uex
uncollapsed_section_start piglit "Building piglit"
# When changing this file, you need to bump the following
# .gitlab-ci/image-tags.yml tags:
# DEBIAN_TEST_GL_TAG
# DEBIAN_TEST_VK_TAG
# KERNEL_ROOTFS_TAG
REV="68658566da1c9cd6a378b5ca36999617e26440e7"
git clone https://gitlab.freedesktop.org/mesa/piglit.git --single-branch --no-checkout /piglit
pushd /piglit
git checkout "$REV"
patch -p1 <$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 . -depth \( -name .git -o -name '*ninja*' -o -iname '*cmake*' -o -name '*.[chao]' \) \
! -name 'include_test.h' -exec rm -rf {} \;
rm -rf target_api
if [ "${PIGLIT_BUILD_TARGETS:-}" = "piglit_replayer" ]; then
find . -depth \
! -regex "^\.$" \
! -regex "^\.\/piglit.*" \
! -regex "^\.\/framework.*" \
! -regex "^\.\/bin$" \
! -regex "^\.\/bin\/replayer\.py" \
! -regex "^\.\/templates.*" \
! -regex "^\.\/tests$" \
! -regex "^\.\/tests\/replay\.py" \
-exec rm -rf {} \; 2>/dev/null
fi
popd
section_end piglit

View File

@@ -1,40 +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
uncollapsed_section_start rust "Building Rust toolchain"
# 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
# Pick a specific snapshot from rustup so the compiler doesn't drift on us.
RUST_VERSION=1.78.0-2024-05-02
# 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.
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
--proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- \
--default-toolchain $RUST_VERSION \
--profile minimal \
-y
rustup component add clippy rustfmt
# 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
section_end rust

View File

@@ -1,18 +0,0 @@
#!/usr/bin/env bash
# When changing this file, you need to bump the following
# .gitlab-ci/image-tags.yml tags:
# DEBIAN_BUILD_TAG
set -ex
uncollapsed_section_start shader-db "Building shader-db"
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
section_end shader-db

View File

@@ -1,104 +0,0 @@
#!/usr/bin/env bash
# SPDX-License-Identifier: MIT
#
# Copyright © 2022 Collabora Limited
# Author: Guilherme Gallo <guilherme.gallo@collabora.com>
#
# When changing this file, you need to bump the following
# .gitlab-ci/image-tags.yml tags:
# KERNEL_ROOTFS_TAG
set -uex
uncollapsed_section_start skqp "Building skqp"
SKQP_BRANCH=android-cts-12.1_r5
SCRIPT_DIR="$(pwd)/.gitlab-ci/container"
SKQP_PATCH_DIR="${SCRIPT_DIR}/patches"
BASE_ARGS_GN_FILE="${SCRIPT_DIR}/build-skqp_base.gn"
case "$DEBIAN_ARCH" in
amd64)
SKQP_ARCH=x64
;;
armhf)
SKQP_ARCH=arm
;;
arm64)
SKQP_ARCH=arm64
;;
esac
SKIA_DIR=${SKIA_DIR:-$(mktemp -d)}
SKQP_OUT_DIR=${SKIA_DIR}/out/${SKQP_ARCH}
SKQP_INSTALL_DIR=${SKQP_INSTALL_DIR:-/skqp}
SKQP_ASSETS_DIR="${SKQP_INSTALL_DIR}/assets"
SKQP_BINARIES=(skqp list_gpu_unit_tests list_gms)
create_gn_args() {
# gn can be configured to cross-compile skia and its tools
# It is important to set the target_cpu to guarantee the intended target
# machine
cp "${BASE_ARGS_GN_FILE}" "${SKQP_OUT_DIR}"/args.gn
echo "target_cpu = \"${SKQP_ARCH}\"" >> "${SKQP_OUT_DIR}"/args.gn
}
download_skia_source() {
if [ -z ${SKIA_DIR+x} ]
then
return 1
fi
# Skia cloned from https://android.googlesource.com/platform/external/skqp
# has all needed assets tracked on git-fs
SKQP_REPO=https://android.googlesource.com/platform/external/skqp
git clone --branch "${SKQP_BRANCH}" --depth 1 "${SKQP_REPO}" "${SKIA_DIR}"
}
download_skia_source
pushd "${SKIA_DIR}"
# Apply all skqp patches for Mesa CI
cat "${SKQP_PATCH_DIR}"/build-skqp_*.patch |
patch -p1
# hack for skqp see the clang
pushd /usr/bin/
ln -s "../lib/llvm-${LLVM_VERSION}/bin/clang" clang
ln -s "../lib/llvm-${LLVM_VERSION}/bin/clang++" clang++
popd
# Fetch some needed build tools needed to build skia/skqp.
# Basically, it clones repositories with commits SHAs from ${SKIA_DIR}/DEPS
# directory.
python tools/git-sync-deps
mkdir -p "${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}"
# Strip binary, since gn is not stripping it even when `is_debug == false`
${STRIP_CMD:-strip} "${SKQP_OUT_DIR}/${BINARY}"
install -m 0755 "${SKQP_OUT_DIR}/${BINARY}" "${SKQP_INSTALL_DIR}"
done
# Move assets to the target directory, which will reside in rootfs.
mv platform_tools/android/apps/skqp/src/main/assets/ "${SKQP_ASSETS_DIR}"
popd
rm -Rf "${SKIA_DIR}"
set +ex
section_end skqp

View File

@@ -1,64 +0,0 @@
cc = "clang"
cxx = "clang++"
extra_cflags = [
"-Wno-error",
"-DSK_ENABLE_DUMP_GPU",
"-DSK_BUILD_FOR_SKQP"
]
extra_cflags_cc = [
"-Wno-error",
# skqp build process produces a lot of compilation warnings, silencing
# most of them to remove clutter and avoid the CI job log to exceed the
# maximum size
# GCC flags
"-Wno-redundant-move",
"-Wno-suggest-override",
"-Wno-class-memaccess",
"-Wno-deprecated-copy",
"-Wno-uninitialized",
# Clang flags
"-Wno-macro-redefined",
"-Wno-anon-enum-enum-conversion",
"-Wno-suggest-destructor-override",
"-Wno-return-std-move-in-c++11",
"-Wno-extra-semi-stmt",
"-Wno-reserved-identifier",
"-Wno-bitwise-instead-of-logical",
"-Wno-reserved-identifier",
"-Wno-psabi",
"-Wno-unused-but-set-variable",
"-Wno-sizeof-array-div",
"-Wno-string-concatenation",
"-Wno-unsafe-buffer-usage",
"-Wno-switch-default",
"-Wno-cast-function-type-strict",
"-Wno-format",
"-Wno-enum-constexpr-conversion",
]
cc_wrapper = "ccache"
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

@@ -1,29 +0,0 @@
#!/usr/bin/env bash
# shellcheck disable=SC2086 # we want word splitting
# When changing this file, you need to bump the following
# .gitlab-ci/image-tags.yml tags:
# KERNEL_ROOTFS_TAG
set -uex
uncollapsed_section_start va-tools "Building va-tools"
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 2.18.1 \
--depth 1 \
/va-utils
pushd /va-utils
# Too old libva in Debian 11. TODO: when this PR gets in, refer to the patch.
curl --fail -L https://github.com/intel/libva-utils/pull/329.patch | git am
meson setup build -D tests=true -Dprefix=/va ${EXTRA_MESON_ARGS:-}
meson install -C build
popd
rm -rf /va-utils
section_end va-tools

View File

@@ -1,51 +0,0 @@
#!/bin/bash
# When changing this file, you need to bump the following
# .gitlab-ci/image-tags.yml tags:
# DEBIAN_TEST_VK_TAG
# KERNEL_ROOTFS_TAG
set -ex
uncollapsed_section_start vkd3d-proton "Building vkd3d-proton"
VKD3D_PROTON_COMMIT="b121e6d746341e0aaba7663e3d85f3194e8e20e1"
VKD3D_PROTON_DST_DIR="/vkd3d-proton-tests"
VKD3D_PROTON_SRC_DIR="/vkd3d-proton-src"
VKD3D_PROTON_BUILD_DIR="/vkd3d-proton-build"
function build_arch {
local arch="$1"
meson setup \
-Denable_tests=true \
--buildtype release \
--prefix "$VKD3D_PROTON_DST_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/d3d12"
}
git clone https://github.com/HansKristian-Work/vkd3d-proton.git --single-branch -b master --no-checkout "$VKD3D_PROTON_SRC_DIR"
pushd "$VKD3D_PROTON_SRC_DIR"
git checkout "$VKD3D_PROTON_COMMIT"
git submodule update --init --recursive
git submodule update --recursive
build_arch 64
build_arch 86
mkdir "$VKD3D_PROTON_DST_DIR/tests"
cp \
"tests/test-runner.sh" \
"tests/d3d12_tests.h" \
"$VKD3D_PROTON_DST_DIR/tests/"
popd
rm -rf "$VKD3D_PROTON_BUILD_DIR"
rm -rf "$VKD3D_PROTON_SRC_DIR"
section_end vkd3d-proton

View File

@@ -1,25 +0,0 @@
#!/usr/bin/env bash
# When changing this file, you need to bump the following
# .gitlab-ci/image-tags.yml tags:
# DEBIAN_TEST_GL_TAG
# KERNEL_ROOTFS_TAG
set -uex
uncollapsed_section_start vulkan-validation "Building Vulkan validation layers"
VALIDATION_TAG="snapshot-2024wk39"
git clone -b "$VALIDATION_TAG" --single-branch --depth 1 https://github.com/KhronosGroup/Vulkan-ValidationLayers.git
pushd Vulkan-ValidationLayers
# we don't need to build SPIRV-Tools tools
sed -i scripts/known_good.json -e 's/SPIRV_SKIP_EXECUTABLES=OFF/SPIRV_SKIP_EXECUTABLES=ON/'
python3 scripts/update_deps.py --dir external --config release --generator Ninja --optional tests
cmake -G Ninja -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=/usr -DBUILD_TESTS=OFF -DBUILD_WERROR=OFF -C external/helper.cmake -S . -B build
ninja -C build
cmake --install build --strip
popd
rm -rf Vulkan-ValidationLayers
section_end vulkan-validation

View File

@@ -1,36 +0,0 @@
#!/usr/bin/env bash
# shellcheck disable=SC2086 # we want word splitting
set -uex
uncollapsed_section_start wayland "Building Wayland"
# When changing this file, you need to bump the following
# .gitlab-ci/image-tags.yml tags:
# DEBIAN_BUILD_TAG
# DEBIAN_TEST_ANDROID_TAG
# DEBIAN_TEST_GL_TAG
# DEBIAN_TEST_VK_TAG
# FEDORA_X86_64_BUILD_TAG
# KERNEL_ROOTFS_TAG
export LIBWAYLAND_VERSION="1.21.0"
export WAYLAND_PROTOCOLS_VERSION="1.38"
git clone https://gitlab.freedesktop.org/wayland/wayland
cd wayland
git checkout "$LIBWAYLAND_VERSION"
meson setup -Ddocumentation=false -Ddtd_validation=false -Dlibraries=true _build ${EXTRA_MESON_ARGS:-}
meson install -C _build
cd ..
rm -rf wayland
git clone https://gitlab.freedesktop.org/wayland/wayland-protocols
cd wayland-protocols
git checkout "$WAYLAND_PROTOCOLS_VERSION"
meson setup -Dtests=false _build ${EXTRA_MESON_ARGS:-}
meson install -C _build
cd ..
rm -rf wayland-protocols
section_end wayland

View File

@@ -1,24 +0,0 @@
#!/usr/bin/env bash
# When changing this file, check if the *_BUIL_TAG tags in
# .gitlab-ci/image-tags.yml need updating.
set -eu
# Early check for required env variables, relies on `set -u`
: "$S3_JWT_FILE_SCRIPT"
if [ -z "$1" ]; then
echo "usage: $(basename "$0") <CONTAINER_CI_JOB_NAME>" 1>&2
exit 1
fi
CONTAINER_CI_JOB_NAME="$1"
# Tasks to perform before executing the script of a container job
eval "$S3_JWT_FILE_SCRIPT"
unset S3_JWT_FILE_SCRIPT
trap 'rm -f ${S3_JWT_FILE}' EXIT INT TERM
bash ".gitlab-ci/container/${CONTAINER_CI_JOB_NAME}.sh"

View File

@@ -1,14 +0,0 @@
#!/usr/bin/env bash
if test -f /etc/debian_version; then
apt-get autoremove -y --purge
fi
# Clean up any build cache
rm -rf /root/.cache
rm -rf /root/.cargo
rm -rf /.cargo
if test -x /usr/bin/ccache; then
ccache --show-stats
fi

View File

@@ -1,44 +0,0 @@
#!/bin/sh
if test -x /usr/bin/ccache; then
if test -f /etc/debian_version; then
CCACHE_PATH=/usr/lib/ccache
elif test -f /etc/alpine-release; then
CCACHE_PATH=/usr/lib/ccache/bin
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/$CI_PROJECT_NAME/ccache"
export PATH="$CCACHE_PATH:$PATH"
# CMake ignores $PATH, so we have to force CC/GCC to the ccache versions.
export CC="${CCACHE_PATH}/gcc"
export CXX="${CCACHE_PATH}/g++"
ccache --show-stats
fi
# Make a wrapper script for ninja to always include the -j flags
{
echo '#!/bin/sh -x'
# shellcheck disable=SC2016
echo '/usr/bin/ninja -j${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}"
# make wget to try more than once, when download fails or timeout
echo -e "retry_connrefused = on\n" \
"read_timeout = 300\n" \
"tries = 4\n" \
"retry_on_host_error = on\n" \
"retry_on_http_error = 429,500,502,503,504\n" \
"wait_retry = 32" >> /etc/wgetrc

View File

@@ -1,37 +0,0 @@
#!/bin/bash
ndk=$1
arch=$2
cpu_family=$3
cpu=$4
cross_file="/cross_file-$arch.txt"
sdk_version=$5
# armv7 has the toolchain split between two names.
arch2=${6:-$2}
# Note that we disable C++ exceptions, because Mesa doesn't use exceptions,
# and allowing it in code generation means we get unwind symbols that break
# the libEGL and driver symbol tests.
cat > "$cross_file" <<EOF
[binaries]
ar = '$ndk/toolchains/llvm/prebuilt/linux-x86_64/bin/llvm-ar'
c = ['ccache', '$ndk/toolchains/llvm/prebuilt/linux-x86_64/bin/${arch2}${sdk_version}-clang', '-fno-exceptions', '-fno-unwind-tables', '-fno-asynchronous-unwind-tables']
cpp = ['ccache', '$ndk/toolchains/llvm/prebuilt/linux-x86_64/bin/${arch2}${sdk_version}-clang++', '-fno-exceptions', '-fno-unwind-tables', '-fno-asynchronous-unwind-tables', '--start-no-unused-arguments', '-static-libstdc++', '--end-no-unused-arguments']
c_ld = 'lld'
cpp_ld = 'lld'
strip = '$ndk/toolchains/llvm/prebuilt/linux-x86_64/bin/llvm-strip'
pkg-config = ['/usr/bin/pkgconf']
[host_machine]
system = 'android'
cpu_family = '$cpu_family'
cpu = '$cpu'
endian = 'little'
[properties]
needs_exe_wrapper = true
pkg_config_libdir = '/usr/local/lib/${arch2}/pkgconfig/:/${ndk}/toolchains/llvm/prebuilt/linux-x86_64/sysroot/usr/lib/${arch2}/pkgconfig/'
EOF

View File

@@ -1,40 +0,0 @@
#!/bin/sh
# shellcheck disable=SC2086 # we want word splitting
# 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"
sdk_version="$6"
sysroot=$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/$arch/$sdk_version
sharedlibdir=$sysroot/usr/lib/$arch
includedir=$sysroot/usr/include
Name: zlib
Description: zlib compression library
Version: $version
Requires:
Libs: -L$sysroot/usr/lib/$arch/$sdk_version $libs
Cflags: -I$sysroot/usr/include $cflags
EOF
done

View File

@@ -1,54 +0,0 @@
#!/bin/bash
arch=$1
cross_file="/cross_file-$arch.txt"
meson env2mfile --cross --debarch "$arch" -o "$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"
# Rely on qemu-user being configured in binfmt_misc on the host
# shellcheck disable=SC1003 # how this sed doesn't seems to work for me locally
sed -i -e '/\[properties\]/a\' -e "needs_exe_wrapper = False" "$cross_file"
# Add a line for rustc, which meson env2mfile is missing.
cc=$(sed -n "s|^c\s*=\s*\[?'\(.*\)'\]?|\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
# shellcheck disable=SC1003 # how this sed doesn't seems to work for me locally
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 [[ "$arch" = "arm64" ]]; then
GCC_ARCH="aarch64-linux-gnu"
DE_CPU="DE_CPU_ARM_64"
elif [[ "$arch" = "armhf" ]]; then
GCC_ARCH="arm-linux-gnueabihf"
DE_CPU="DE_CPU_ARM"
fi
if [[ -n "$GCC_ARCH" ]]; then
{
echo "set(CMAKE_SYSTEM_NAME Linux)";
echo "set(CMAKE_SYSTEM_PROCESSOR arm)";
echo "set(CMAKE_C_COMPILER /usr/lib/ccache/$GCC_ARCH-gcc)";
echo "set(CMAKE_CXX_COMPILER /usr/lib/ccache/$GCC_ARCH-g++)";
echo "set(CMAKE_CXX_FLAGS_INIT \"-Wno-psabi\")"; # makes ABI warnings quiet for ARMv7
echo "set(ENV{PKG_CONFIG} \"/usr/bin/$GCC_ARCH-pkgconf\")";
echo "set(DE_CPU $DE_CPU)";
} > "$toolchain_file"
fi

View File

@@ -1,95 +0,0 @@
#!/usr/bin/env bash
# shellcheck disable=SC2086 # we want word splitting
set -e
. .gitlab-ci/setup-test-env.sh
set -o xtrace
export DEBIAN_FRONTEND=noninteractive
: "${LLVM_VERSION:?llvm version not set!}"
# Ephemeral packages (installed for this script and removed again at the end)
EPHEMERAL=(
)
DEPS=(
"crossbuild-essential-$arch"
"pkgconf:$arch"
"libasan8:$arch"
"libdrm-dev:$arch"
"libelf-dev:$arch"
"libexpat1-dev:$arch"
"libffi-dev:$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:$arch"
"libwayland-dev:$arch"
)
dpkg --add-architecture $arch
echo "deb [trusted=yes] https://gitlab.freedesktop.org/gfx-ci/ci-deb-repo/-/raw/${PKG_REPO_REV}/ ${FDO_DISTRIBUTION_VERSION%-*} main" | tee /etc/apt/sources.list.d/gfx-ci_.list
apt-get update
apt-get install -y --no-remove "${DEPS[@]}" "${EPHEMERAL[@]}" \
$EXTRA_LOCAL_PACKAGES
if [[ $arch != "armhf" ]]; then
# We don't need clang-format for the crossbuilds, but the installed amd64
# package will conflict with libclang. Uninstall clang-format (and its
# problematic dependency) to fix.
apt-get remove -y "clang-format-${LLVM_VERSION}" "libclang-cpp${LLVM_VERSION}" \
"llvm-${LLVM_VERSION}-runtime" "llvm-${LLVM_VERSION}-linker-tools"
# llvm-*-tools:$arch conflicts with python3:amd64. Install dependencies only
# with apt-get, then force-install llvm-*-{dev,tools}:$arch with dpkg to get
# around this.
apt-get install -y --no-remove --no-install-recommends \
"libclang-cpp${LLVM_VERSION}:$arch" \
"libgcc-s1:$arch" \
"libtinfo-dev:$arch" \
"libz3-dev:$arch" \
"llvm-${LLVM_VERSION}:$arch" \
zlib1g
fi
. .gitlab-ci/container/create-cross-file.sh $arch
. .gitlab-ci/container/container_pre_build.sh
# dependencies where we want a specific version
MULTIARCH_PATH=$(dpkg-architecture -A $arch -qDEB_TARGET_MULTIARCH)
export EXTRA_MESON_ARGS="--cross-file=/cross_file-${arch}.txt -D libdir=lib/${MULTIARCH_PATH}"
. .gitlab-ci/container/build-wayland.sh
. .gitlab-ci/container/build-directx-headers.sh
apt-get purge -y "${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_VERSION}"-{dev,tools}:"$arch"
dpkg -i --force-depends llvm-"${LLVM_VERSION}"-*_"${arch}".deb
rm llvm-"${LLVM_VERSION}"-*_"${arch}".deb
fi

View File

@@ -1,108 +0,0 @@
#!/usr/bin/env bash
# shellcheck disable=SC2086 # we want word splitting
# When changing this file, you need to bump the following
# .gitlab-ci/image-tags.yml tags:
# DEBIAN_BUILD_TAG
set -e
. .gitlab-ci/setup-test-env.sh
set -x
EPHEMERAL=(
autoconf
rdfind
unzip
)
apt-get install -y --no-remove "${EPHEMERAL[@]}"
# Fetch the NDK and extract just the toolchain we want.
ndk="android-ndk-${ANDROID_NDK_VERSION}"
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
-o $ndk.zip https://dl.google.com/android/repository/$ndk-linux.zip
unzip -d / $ndk.zip "$ndk/source.properties" "$ndk/build/cmake/*" "$ndk/toolchains/llvm/*"
rm $ndk.zip
# Since it was packed as a zip file, symlinks/hardlinks got turned into
# duplicate files. Turn them into hardlinks to save on container space.
rdfind -makehardlinks true -makeresultsfile false /${ndk}/
# Drop some large tools we won't use in this build.
find /${ndk}/ -type f \( -iname '*clang-check*' -o -iname '*clang-tidy*' -o -iname '*lldb*' \) -exec rm -f {} \;
sh .gitlab-ci/container/create-android-ndk-pc.sh /$ndk zlib.pc "" "-lz" "1.2.3" $ANDROID_SDK_VERSION
sh .gitlab-ci/container/create-android-cross-file.sh /$ndk x86_64-linux-android x86_64 x86_64 $ANDROID_SDK_VERSION
sh .gitlab-ci/container/create-android-cross-file.sh /$ndk i686-linux-android x86 x86 $ANDROID_SDK_VERSION
sh .gitlab-ci/container/create-android-cross-file.sh /$ndk aarch64-linux-android aarch64 armv8 $ANDROID_SDK_VERSION
sh .gitlab-ci/container/create-android-cross-file.sh /$ndk arm-linux-androideabi arm armv7hl $ANDROID_SDK_VERSION armv7a-linux-androideabi
for arch in \
x86_64-linux-android \
i686-linux-android \
aarch64-linux-android \
arm-linux-androideabi ; do
EXTRA_MESON_ARGS="--cross-file=/cross_file-$arch.txt --libdir=lib/$arch -Dnouveau=disabled -Dintel=disabled" \
. .gitlab-ci/container/build-libdrm.sh
done
rm -rf $LIBDRM_VERSION
export LIBELF_VERSION=libelf-0.8.13
curl -L --retry 4 -f --retry-all-errors --retry-delay 60 \
-O https://fossies.org/linux/misc/old/$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 [ "${arch}" == 'arm-linux-androideabi' ]
then
ccarch=armv7a-linux-androideabi
fi
export CC=/${ndk}/toolchains/llvm/prebuilt/linux-x86_64/bin/llvm-ar
export CC=/${ndk}/toolchains/llvm/prebuilt/linux-x86_64/bin/${ccarch}${ANDROID_SDK_VERSION}-clang
export CXX=/${ndk}/toolchains/llvm/prebuilt/linux-x86_64/bin/${ccarch}${ANDROID_SDK_VERSION}-clang++
export LD=/${ndk}/toolchains/llvm/prebuilt/linux-x86_64/bin/${arch}-ld
export RANLIB=/${ndk}/toolchains/llvm/prebuilt/linux-x86_64/bin/llvm-ranlib
# The configure script doesn't know about android, but doesn't really use the host anyway it
# seems
./configure --host=x86_64-linux-gnu --disable-nls --disable-shared \
--libdir=/usr/local/lib/${arch}
make install
make distclean
unset CC
unset CC
unset CXX
unset LD
unset RANLIB
done
cd ..
rm -rf $LIBELF_VERSION
# Build LLVM libraries for Android only if necessary, uploading a copy to S3
# to avoid rebuilding it in a future run if the version does not change.
bash .gitlab-ci/container/build-android-x86_64-llvm.sh
apt-get purge -y "${EPHEMERAL[@]}"

View File

@@ -1,112 +0,0 @@
#!/usr/bin/env bash
# shellcheck disable=SC2086 # we want word splitting
set -e
. .gitlab-ci/setup-test-env.sh
set -o xtrace
: "${LLVM_VERSION:?llvm version not set}"
apt-get -y install ca-certificates curl gnupg2
sed -i -e 's/http:\/\/deb/https:\/\/deb/g' /etc/apt/sources.list.d/*
echo "deb [trusted=yes] https://gitlab.freedesktop.org/gfx-ci/ci-deb-repo/-/raw/${PKG_REPO_REV}/ ${FDO_DISTRIBUTION_VERSION%-*} main" | tee /etc/apt/sources.list.d/gfx-ci_.list
. .gitlab-ci/container/debian/maybe-add-llvm-repo.sh
# Ephemeral packages (installed for this script and removed again at the end)
EPHEMERAL=(
libssl-dev
)
DEPS=(
apt-utils
android-libext4-utils
autoconf
automake
bc
bison
ccache
cmake
curl
"clang-${LLVM_VERSION}"
fastboot
flatbuffers-compiler
flex
g++
git
glslang-tools
kmod
"libclang-${LLVM_VERSION}-dev"
"libclang-cpp${LLVM_VERSION}-dev"
"libclang-common-${LLVM_VERSION}-dev"
libasan8
libdrm-dev
libelf-dev
libexpat1-dev
libflatbuffers-dev
"libllvm${LLVM_VERSION}"
libvulkan-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
libxtensor-dev
libxxf86vm-dev
libwayland-dev
libwayland-egl-backend-dev
"llvm-${LLVM_VERSION}-dev"
ninja-build
meson
openssh-server
pkgconf
python3-mako
python3-pil
python3-pip
python3-pycparser
python3-requests
python3-setuptools
u-boot-tools
xz-utils
zlib1g-dev
zstd
)
apt-get update
apt-get -y install "${DEPS[@]}" "${EPHEMERAL[@]}"
pip3 install --break-system-packages git+http://gitlab.freedesktop.org/freedesktop/ci-templates@ffe4d1b10aab7534489f0c4bbc4c5899df17d3f2
arch=armhf
. .gitlab-ci/container/cross_build.sh
. .gitlab-ci/container/container_pre_build.sh
. .gitlab-ci/container/build-mold.sh
. .gitlab-ci/container/build-wayland.sh
. .gitlab-ci/container/build-llvm-spirv.sh
. .gitlab-ci/container/build-libclc.sh
. .gitlab-ci/container/install-meson.sh
. .gitlab-ci/container/build-rust.sh
. .gitlab-ci/container/build-bindgen.sh
apt-get purge -y "${EPHEMERAL[@]}"
. .gitlab-ci/container/container_post_build.sh

View File

@@ -1,4 +0,0 @@
#!/usr/bin/env bash
DEBIAN_ARCH=arm64 \
. .gitlab-ci/container/debian/test-base.sh

View File

@@ -1,5 +0,0 @@
#!/usr/bin/env bash
. .gitlab-ci/container/debian/test-gl.sh
. .gitlab-ci/container/strip-rootfs.sh

View File

@@ -1,5 +0,0 @@
#!/usr/bin/env bash
. .gitlab-ci/container/debian/test-vk.sh
. .gitlab-ci/container/strip-rootfs.sh

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