Mesa 10.5.1 Release Notes / March 13, 2015

Mesa 10.5.1 is a bug fix release which fixes bugs found since the 10.5.0 release.

Mesa 10.5.1 implements the OpenGL 3.3 API, but the version reported by glGetString(GL_VERSION) or glGetIntegerv(GL_MAJOR_VERSION) / glGetIntegerv(GL_MINOR_VERSION) depends on the particular driver being used. Some drivers don't support all the features required in OpenGL 3.3. OpenGL 3.3 is only available if requested at context creation because compatibility contexts are not supported.

SHA256 checksums

TBD

New features

None

Bug fixes

This list is likely incomplete.

• Bug 79202 - valgrind errors in glsl-fs-uniform-array-loop-unroll.shader_test; random code generation
• Bug 84613 - [G965, bisected] piglit regressions : glslparsertest.glsl2
• Bug 86747 - Noise in Football Manager 2014 textures
• Bug 86974 - INTEL_DEBUG=shader_time always asserts in fs_generator::generate_code() when Mesa is built with --enable-debug (= with asserts)
• Bug 88246 - Commit 2881b12 causes 43 DrawElements test regressions
• Bug 88793 - [BDW/BSW Bisected]Piglit/shaders_glsl-max-varyings fails
• Bug 88883 - ir-a2xx.c: variable changed in assert statement
• Bug 88885 - Transform feedback uses incorrect interleaving if a previous draw did not write gl_Position
• Bug 89095 - [SNB/IVB/BYT Bisected]Webglc conformance/glsl/functions/glsl-function-mix-float.html fails
• Bug 89156 - r300g: GL_COMPRESSED_RED_RGTC1 / ATI1N support broken
• Bug 89224 - Incorrect rendering of Unigine Valley running in VM on VMware Workstation
• Bug 89292 - [regression,bisected] incomplete screenshots in some cases
• Bug 89311 - [regression, bisected] dEQP: Added entry points for glCompressedTextureSubImage*D.
• Bug 89312 - [regression, bisected] main: Added entry points for CopyTextureSubImage*D. (d6b7c40cecfe01)
• Bug 89315 - [HSW, regression, bisected] i965/fs: Emit MAD instructions when possible.
• Bug 89317 - [HSW, regression, bisected] i965: Add LINTERP/CINTERP to can_do_cmod() (d91390634)
• Bug 89416 - UE4Editor crash after load project
• Bug 89430 - [g965][bisected] arb_copy_image-targets gl_texture* tests fail

Changes

Andrey Sudnik (1):

• i965/vec4: Don't lose the saturate modifier in copy propagation.

Chris Forbes (1):

• i965/gs: Check newly-generated GS-out VUE map against correct stage

Daniel Stone (1):

• egl: Take alpha bits into account when selecting GBM formats

Emil Velikov (5):

• docs: Add sha256 sums for the 10.5.0 release
• egl/main: no longer export internal function
• cherry-ignore: ignore a few more commits picked without -x
• mapi: fix commit 90411b56f6bc817e229d8801ac0adad6d4e3fb7a
• Update version to 10.5.1

Frank Henigman (1):

• intel: fix EGLImage renderbuffer _BaseFormat

Iago Toral Quiroga (1):

• i965: Fix out-of-bounds accesses into pull_constant_loc array

Ian Romanick (1):

• i965/fs/nir: Use emit_math for nir_op_fpow

Ilia Mirkin (3):

• freedreno: move fb state copy after checking for size change
• freedreno/ir3: fix array count returned by TXQ
• freedreno/ir3: get the # of miplevels from getinfo

Jason Ekstrand (2):

• meta/TexSubImage: Stash everything other than PIXEL_TRANSFER/store in meta_begin
• main/base_tex_format: Properly handle STENCIL_INDEX1/4/16

Kenneth Graunke (8):

• i965: Split Gen4-5 BlitFramebuffer code; prefer BLT over Meta.
• glsl: Mark array access when copying to a temporary for the ?: operator.
• i965/fs: Set force_writemask_all on shader_time instructions.
• i965/fs: Set smear on shader_time diff register.
• i965/fs: Make emit_shader_time_write return rather than emit.
• i965/fs: Make get_timestamp() pass back the MOV rather than emitting it.
• i965/fs: Make emit_shader_time_end() insert before EOT.
• i965/fs: Don't issue FB writes for bound but unwritten color targets.

Laura Ekstrand (2):

• main: Fix target checking for CompressedTexSubImage*D.
• main: Fix target checking for CopyTexSubImage*D.

Marc-Andre Lureau (1):

• gallium/auxiliary/indices: fix start param

Marek Olšák (3):

• r300g: fix RGTC1 and LATC1 SNORM formats
• r300g: fix a crash when resolving into an sRGB texture
• r300g: fix sRGB->sRGB blits

Matt Turner (12):

• i965/vec4: Fix implementation of i2b.
• mesa: Indent break statements and add a missing one.
• mesa: Free memory allocated for luminance in readpixels.
• mesa: Correct backwards NULL check.
• i965: Consider scratch writes to have side effects.
• i965/fs: Don't use backend_visitor::instructions after creating the CFG.
• r300g: Use PATH_MAX instead of limiting ourselves to 100 chars.
• r300g: Check return value of snprintf().
• i965/fs: Don't propagate cmod to inst with different type.
• i965: Tell intel_get_memcpy() which direction the memcpy() is going.
• Revert SHA1 additions.
• i965: Avoid applying negate to wrong MAD source.

Neil Roberts (4):

• meta: In pbo_{Get,}TexSubImage don't repeatedly rebind the source tex
• Revert "common: Fix PBOs for 1D_ARRAY."
• meta: Allow GL_UN/PACK_IMAGE_HEIGHT in _mesa_meta_pbo_Get/TexSubImage
• meta: Fix the y offset for 1D_ARRAY in _mesa_meta_pbo_TexSubImage

Rob Clark (11):

• freedreno/ir3: fix silly typo for binning pass shaders
• freedreno/a2xx: fix increment in assert
• freedreno/a4xx: bit of cleanup
• freedreno: update generated headers
• freedreno/a4xx: set PC_PRIM_VTX_CNTL.VAROUT properly
• freedreno: update generated headers
• freedreno/a4xx: aniso filtering
• freedreno/ir3: fix up cat6 instruction encodings
• freedreno/ir3: add support for memory (cat6) instructions
• freedreno/ir3: handle flat bypass for a4xx
• freedreno/ir3: fix failed assert in grouping

Stefan Dösinger (1):

• r300g: Fix the ATI1N swizzle (RGTC1 and LATC1)