Kernel: (drm-intel-nightly) 164a4cb4c1431a0689f85507868356fae24da638
Bug detailed description:
It fails on Sandybridge/Ivybridge/Haswell with mesa master branch.It doesn't happen on 9.2 branch.
Bisect shows:c3c9a8c85758796a26b48e484286e6b6f5a5299a is the first bad commit.
Author: Eric Anholt <email@example.com>
AuthorDate: Tue Oct 22 12:27:06 2013 -0700
Commit: Eric Anholt <firstname.lastname@example.org>
CommitDate: Fri Oct 25 16:45:30 2013 -0700
i965: Try not to reverse-schedule things when doing LIFO scheduling.
The LIFO plan was simple: Take the most recently made available
instructions, and pick those first.
But because of the order we were pushing things onto our list of
available-to-schedule instructions, it meant that when a set of
instructions was made available at the same time (for example, everything
at the start of the program that didn't depend on other instructions) we'd
schedule them in reverse order.
If you had 10 texture calls in a row in your program, each with
independent argument setup, we'd set up the last texture call's args and
execute it first, even though we wouldn't be able to consume its results
until we'd finished the other 9 texture calls (assuming consumption of
texture results happens near each texture call, and combines it with
another texture result, which is normal for a convolution shader).
To fix this, walk the list for doing LIFO in the order that instructions
were originally generated in the program, but choose to push
newly-made-available instructions to the other end of the list instead.
total instructions in shared programs: 1587242 -> 1586290 (-0.06%)
instructions in affected programs: 7801 -> 6849 (-12.20%)
Thanks to Chia-I Wu for pointing out the bug in my first version of the
patch that made it a huge loss.
Reviewed-by: Matt Turner <email@example.com>
Running GTF (Graphics Test Framework)
egl_vendor: Mesa Project
ConfigID: 14 RGBA bits( 8, 8, 8, 8) Depth bits: 0 Stencil bits: 0
arg: /GFX/Test/Ogles3conform/ogles3conform/GTF_ES/glsl/GTF/GTF -width=64 -height=64 -run=/GFX/Test/Ogles3conform/ogles3conform/GTF_ES/glsl/GTF/GL3Tests/uniform_buffer_object/uniform_buffer_object_max_uniform_block_size.test
gl_renderer: Mesa DRI Intel(R) Sandybridge Desktop x86/MMX/SSE2
gl_shading_language_version: OpenGL ES GLSL ES 3.0
gl_version: OpenGL ES 3.0 Mesa 10.0.0-devel (git-7f76368)
gl_vendor: Intel Open Source Technology Center
gl_extensions: GL_EXT_blend_minmax GL_EXT_multi_draw_arrays GL_EXT_texture_filter_anisotropic GL_EXT_texture_compression_dxt1 GL_EXT_texture_format_BGRA8888 GL_OES_compressed_ETC1_RGB8_texture GL_OES_depth24 GL_OES_element_index_uint GL_OES_fbo_render_mipmap GL_OES_mapbuffer GL_OES_rgb8_rgba8 GL_OES_standard_derivatives GL_OES_stencil8 GL_OES_texture_3D GL_OES_texture_npot GL_OES_EGL_image GL_OES_depth_texture GL_OES_packed_depth_stencil GL_EXT_texture_type_2_10_10_10_REV GL_OES_get_program_binary GL_APPLE_texture_max_level GL_EXT_discard_framebuffer GL_EXT_read_format_bgra GL_NV_fbo_color_attachments GL_OES_EGL_image_external GL_OES_vertex_array_object GL_ANGLE_texture_compression_dxt3 GL_ANGLE_texture_compression_dxt5 GL_EXT_texture_rg GL_EXT_unpack_subimage GL_NV_draw_buffers GL_NV_read_buffer GL_EXT_map_buffer_range GL_OES_depth_texture_cube_map GL_OES_surfaceless_context GL_EXT_color_buffer_float GL_EXT_shader_integer_mix
Zero epsilon: 0.000122, RGBA epsilon: ( 0.004044, 0.004044, 0.004044, 0.004044)
2. ./GTF -width=64 -height=64 -run=GL3Tests/uniform_buffer_object/uniform_buffer_object_max_uniform_block_size.test
I noticed this on Friday (not sure how I didn't notice before pushing) and have been working on a fix.
Author: Eric Anholt <firstname.lastname@example.org>
Date: Mon Oct 28 15:17:07 2013 -0700
i965/fs: Prefer more-critical instructions of the same age in LIFO scheduling.