Mesa 24.2.1 Release Notes / 2024-08-28

Mesa 24.2.1 is a bug fix release which fixes bugs found since the 24.2.0 release.

Mesa 24.2.1 implements the OpenGL 4.6 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 4.6. OpenGL 4.6 is only available if requested at context creation. Compatibility contexts may report a lower version depending on each driver.

Mesa 24.2.1 implements the Vulkan 1.3 API, but the version reported by the apiVersion property of the VkPhysicalDeviceProperties struct depends on the particular driver being used.

SHA checksums

SHA256: fc9a495f3a9af906838be89367564e10ef335e058f88965ad49ccc3e9a3b420b  mesa-24.2.1.tar.xz
SHA512: 3b77e5faec51b67583131123b0cc010b52325ea308e4075323102aa999d9c9fbb65b873eb537ed4f577b5a0811e7f096e7e101510cb50326ea5c439b4b468380  mesa-24.2.1.tar.xz

New features

  • None

Bug fixes

  • Artifacts on Black Myth Wukong

  • DCS world radeon rx 7900 flicker

  • radv/aco: Scheduler bug in ACO somewhere in parallel-gs

  • Once human black artifacts consume the screen

  • [ANV] Breaking Limit benchmark does not run

  • Possible race condition when generating enums.h

  • [radeonsi] regression with bindless textures (bisected)

  • [anv] possible regression found on 982106e6768e6f69d71710e314325c813591452d

  • xe2: Non-compiler assertion failure replaying cp2077 fossil

  • Mesa doesn’t build with the following config options

  • Intel: Enable Compression on depth buffers (Xe2)

  • Confidential issue #11711

  • Spectacle crashes recording h264 video with 24.2-rc3

Changes

Boris Brezillon (1):

  • panvk: Adjust RGB component order for fixed-function blending

Connor Abbott (5):

  • tu: Fix off-by-one in UBO CP_LOAD_STATE size

  • ir3, tu: Use a UBO for VS primitive params on a750+

  • Revert “tu/a750: Disable HW binning when there is GS”

  • tu: Fix passthrough D24S8 attachments

  • tu: Treat partially-bound depth/stencil attachments as passthrough

Dave Airlie (3):

  • radv/video: fix reporting video format props for encode.

  • llvmpipe: handle stride properly on lvp udmabuf imports

  • llvmpipe: make sure to duplicate the fd handle before giving out

David Heidelberg (3):

  • etnaviv: rename enums_h appropriately

  • etnaviv: build dependency for the etnaviv tests

  • bin/gen_release_notes: adjust checksums section

David Rosca (3):

  • frontends/va: Fix use after free with EFC

  • radeonsi: Don’t allow DCC for encode in is_video_target_buffer_supported

  • frontends/va: Fix leaks with multiple coded buffer segments

Eric Engestrom (13):

  • docs: add release notes for 24.2.0

  • docs: add sha256sum for 24.2.0

  • .pick_status.json: Update to 2c47ad7774a7d0fe47cf870676c3e2390bca5b50

  • .pick_status.json: Mark 77f783462a9758b22e211c72a57ec7c36d6c09fd as denominated

  • .pick_status.json: Update to 2b2b66f497bf8c5f91067752995a5c1003255a6f

  • .pick_status.json: Update to 47a2ab6f3fb41bef9062182588f69c217d7e6541

  • .pick_status.json: Mark a33ef21106a3b34cb359f0272a103c8b8066fbe6 as denominated

  • .pick_status.json: Update to c5156257d93ae2a9e54ba5adc73b7342c9bef96e

  • .pick_status.json: Update to 81e3930ec0a4d586752e59c5172e08e2edf5e4d5

  • .pick_status.json: Update to 4a8f3181baa1eca48d44e5270962835040d0f743

  • .pick_status.json: Update to 7392e3306ef376121097cf03b388dc75bff127c0

  • vc4: Add missing libvc4_neon build dependencies

  • .pick_status.json: Update to 64ca0fd2f2896faa2ec85e454e2ecba2f68390c8

Faith Ekstrand (6):

  • vulkan: Add null descriptor bits to vk_pipeline_robustness_state

  • nvk: Plumb the whole vk_pipeline_robustness_state through to nvk_ubo/ssbo_addr_format

  • nvk: Enable shader bounds checking when nullDescriptor is enabled

  • nouveau/mme: Fix add64 of immediates on Fermi

  • nvk: Disable conditional rendering around CopyQueryPoolResults

  • nil,nvk: Disable modifiers for B10G11R11_UFLOAT and E5B9G9R9_UFLOAT

Francisco Jerez (1):

  • intel/brw/gfx12.5+: Fix IR of sub-dword atomic LSC operations.

Friedrich Vock (1):

  • aco: Fix 1D->2D dispatch conversion on <gfx9

GKraats (2):

  • i915g: Screen corruption with ENOBUFS caused by fence register shortage

  • i915g: fix count of buffers at i915_drm_batchbuffer_validate_buffers

Guilherme Gallo (1):

  • ci/a618: Fix zink-tu-a618-full rules

Ian Romanick (2):

  • anv: Protect against OOB access to anv_state_pool::buckets

  • anv: Larger memory pools for huge shaders

Jianxun Zhang (2):

  • Revert “anv: Disable PAT-based compression on depth images (xe2)”

  • Revert “iris: Disable PAT-based compression on depth surfaces (xe2)”

Job Noorman (4):

  • ir3: update merge set affinity in shared RA

  • ir3: fix clearing merge sets after shared RA

  • ir3: fix wrong dstn used in postsched

  • ir3/legalize: handle scalar ALU WAR hazards for a0.x

José Roberto de Souza (3):

  • intel/isl/gfx20: Alow hierarchial depth buffer write through for multi sampled surfaces

  • anv/gfx20: Enable depth buffer write through for multi sampled images

  • iris/gfx20: Enable depth buffer write through for multi sampled images

Karol Herbst (3):

  • rusticl/mem: do not check against image base alignment for 1Dbuffer images

  • rusticl/device: limit CL_DEVICE_IMAGE_MAX_BUFFER_SIZE more aggressively

  • vtn: ignore volatile on functions for now

Kenneth Graunke (2):

  • intel/brw: Pass opcode to brw_swsb_encode/decode

  • intel/brw: Fix Xe2+ SWSB encoding/decoding for DPAS instructions

Konstantin (2):

  • radv: Handle instruction encodings > 8 bytes when splitting disassembly

  • radv: Handle repeated instructions when splitting disassembly

Lepton Wu (1):

  • egl/android: Fix wrong pipe format for RGB_565

Lionel Landwerlin (10):

  • vulkan/runtime: fix GetBufferMemoryRequirements2 for maintenance4

  • anv: fix extended buffer flags usages

  • anv: only set 3DSTATE_CLIP::MaximumVPIndex once

  • anv: optimize CLIP::MaximumVPIndex setting

  • anv: move conditional render predicate after gfx_flush_state

  • anv: don’t miss workaround for indirect draws

  • anv: explicitly disable BT pool allocations at device init

  • anv: always use workaround_address, not workaround_bo

  • nir/divergence: add missing load_constant_base_ptr

  • brw: switch mesh/task URB fence prior to EOT to GPU

Mary Guillemard (2):

  • panvk: Fix NULL deref on model name when device isn’t supported

  • panvk: Fix viewport calculation

Matt Turner (1):

  • nir: Skip opt_if_merge when next_if has block ending in a jump

Mauro Rossi (1):

  • nvk: Fix regression observed on Kepler

Mike Blumenkrantz (6):

  • glx/dri2: strdup driver name

  • zink: bail on choose_pdev immediately if no devices are available

  • st/pbo: reject vs/fs pbo ops if rowstride < width

  • zink: don’t skip cbuf store ops if resolve is set

  • tc: set resolve on renderpass info if blit terminates the renderpass

  • dril: add zink stub

Nanley Chery (4):

  • anv: Add want_hiz_wt_for_image()

  • iris: Add and use want_hiz_wt_for_res

  • iris: Invalidate state cache for some depth fast clears

  • intel/isl: Fix packing of SINT formats

Pavel Ondračka (1):

  • r300: fix RGB10_A2 CONSTANT_COLOR blending

Rhys Perry (7):

  • aco: split selection_control_remove into rarely_taken and never_taken

  • aco: only remove branch jumping over SMEM/barrier if it’s never taken

  • aco: ignore exec and literals when mitigating VALUMaskWriteHazard

  • aco: also consider VALU reads for VALUMaskWriteHazard

  • aco: don’t consider sa_sdst=0 before SALU write to fix VALUMaskWriteHazard

  • aco: check SALU writing lanemask later for VALUMaskWriteHazard

  • aco: preserve bitsets after a lane mask is written

Rob Clark (1):

  • nir/opt_loop: Don’t peel initial break if loop ends in break

Rohan Garg (5):

  • anv: program a custom byte stride on Xe2 for indirect draws

  • anv,iris: prefix the argument format with XI for a upcoming refactor

  • anv: refactor indirect draw support into it’s own function

  • anv: dispatch indirect draws with a count buffer through the XI hardware on ARL+

  • anv: migrate indirect mesh draws to indirect draws on ARL+

Sagar Ghuge (3):

  • intel/compiler: Ray query requires write-back register

  • intel/compiler: Adjust trace ray control field on Xe2

  • intel/compiler: Fix indirect offset in GS input read for Xe2+

Samuel Pitoiset (1):

  • aco: fix bogus assert in RT prolog on GFX11+

Sviatoslav Peleshko (3):

  • brw,elk: Fix opening flags on dumping shader binaries

  • anv: Release correct BO in anv_cmd_buffer_set_ray_query_buffer

  • anv: Add full subgroups WA for the shaders with barriers in Breaking Limit

Tapani Pälli (1):

  • gbm: depend on libdrm indepedent of dri2 setting

Timothy Arceri (4):

  • nir: create validate_tex_src_texture_deref() helper

  • nir: add nir_tex_src_{sampler,texture}_deref_intrinsic

  • glsl: make use of new tex src deref intrinsic

  • nir/glsl: set deref cast mode during function inlining

Valentine Burley (2):

  • android: Extract version from llvm-project instead of hardcoding it

  • llvmpipe: Only use udmabuf with libdrm

Yiwei Zhang (1):

  • venus: workaround cacheline overflush issue on Intel JSL

bbhtt (1):

  • pipe_loader_drm: Fix virtgpu_drm header path