1Mesa 24.0.4 Release Notes / 2024-03-27 2====================================== 3 4Mesa 24.0.4 is a bug fix release which fixes bugs found since the 24.0.3 release. 5 6Mesa 24.0.4 implements the OpenGL 4.6 API, but the version reported by 7glGetString(GL_VERSION) or glGetIntegerv(GL_MAJOR_VERSION) / 8glGetIntegerv(GL_MINOR_VERSION) depends on the particular driver being used. 9Some drivers don't support all the features required in OpenGL 4.6. OpenGL 104.6 is **only** available if requested at context creation. 11Compatibility contexts may report a lower version depending on each driver. 12 13Mesa 24.0.4 implements the Vulkan 1.3 API, but the version reported by 14the apiVersion property of the VkPhysicalDeviceProperties struct 15depends on the particular driver being used. 16 17SHA256 checksum 18--------------- 19 20:: 21 22 90febd30a098cbcd97ff62ecc3dcf5c93d76f7fa314de944cfce81951ba745f0 mesa-24.0.4.tar.xz 23 24 25New features 26------------ 27 28- None 29 30 31Bug fixes 32--------- 33 34- nvk: dota 2 crashes after ~5 seconds in game 35- VAAPI: Incorrect HEVC block size reported with radeonsi 36- radv: WWE 2K24 has very quirky DCC issues on RDNA2 37- RUSTICL creating a shared reference to mutable static is discouraged and will become a hard error 38- KiCAD 3D Viewer - rounded pads rendered incorrectly (texture mapping or stencil test error) 39- OpenSCAD rendering incorrect and inconsistent on radeonsi 40- [radv] Half-Life Alyx renders solid black for reflective surfaces 41- [RX 7900 XTX] Helldivers 2 cause GPU reset 42- radeon: Crash in radeon_bo_can_reclaim_slab 43- RV530 renders improperly at non 4:3 resolutions. 44- anv: new cooperative matrix failures with CTS 1.3.8.0 45- \`[gfxhub0] no-retry page fault` triggered by \`AMD_TEST=testdmaperf` on gfx90c APU 46 47 48Changes 49------- 50 51Boris Brezillon (1): 52 53- panvk: Disable global offset on varying and non-VS attribute descriptors 54 55Caio Oliveira (2): 56 57- intel/brw: Use helper to create accumulator register 58- intel/brw: Fix validation of accumulator register 59 60Charlie Turner (1): 61 62- {vulkan,radv,anv}/video: fix issue in H264 scaling lists derivation 63 64Corentin Noël (2): 65 66- st_pbo/compute: Use the correct structure type when allocating a specialized key 67- zink: Make sure to initialize all the fields of VkMemoryBarrier 68 69Dave Airlie (1): 70 71- radv/video: fix h265 decode with unaligned w/h 72 73David Rosca (1): 74 75- radv/video: Set maxActiveReferencePictures to 16 for H264/5 76 77Eric Engestrom (5): 78 79- docs: add sha256sum for 24.0.3 80- .pick_status.json: Update to 9b6d6c1d2d0c8a517e974abbf7b75a47a607f6ec 81- .pick_status.json: Update to eac703f69128d5aa6879c9becbad627ce08a7920 82- .pick_status.json: Update to 912e203a534be8b70b3ef8bf00294e9c962e385a 83- .pick_status.json: Update to c0875d21563257442fd91aab5740248b0fd96a5c 84 85Faith Ekstrand (2): 86 87- nir/builder: Correctly handle decl_reg or undef as the first instruction 88- nir/gather_types: Support unstructured control-flow 89 90Francisco Jerez (1): 91 92- intel/eu/xe2+: Translate brw_reg fields in REG_SIZE units to physical 512b GRF units during codegen. 93 94Friedrich Vock (2): 95 96- radv: Only enable SEs that the device reports 97- radeonsi: Only enable SEs that the device reports 98 99Gert Wollny (2): 100 101- nir-to-spirv: Cast SSBO input pointer when needed 102- nir_to_spirv: Allow LOD for external images 103 104Hyunjun Ko (1): 105 106- anv/video: fix scan order for scaling lists on H265 decoding. 107 108Iván Briano (2): 109 110- compiler/types: fix serialization of cooperative matrix 111- intel/cmat: fix stride calculation in cmat load/store 112 113Jordan Justen (1): 114 115- intel/compiler/fs: Restore SIMD32 restriction for ray_queries on Xe2 116 117Karol Herbst (2): 118 119- rusticl/kernel: assign sampler locations before DCEing variables 120- nouveau: call glsl_type_singleton_init_or_ref earlier 121 122Kenneth Graunke (1): 123 124- intel/brw: Fix opt_split_sends() to allow for FIXED_GRF send sources 125 126Konstantin Seurer (1): 127 128- zink: Handle aoa derefs of images 129 130Lionel Landwerlin (6): 131 132- intel/fs: fixup sampler header message 133- anv: return unsupported for FSR images on Gfx12.0 134- anv: ignore descriptor alignment for inline uniforms 135- blorp: handle a few allocation failure cases 136- anv: fix block pool allocation failure 137- anv: fix bitfield checks in gfx runtime flushing 138 139Lucas Stach (1): 140 141- etnaviv: fix fixpoint conversion of negative values 142 143Marek Olšák (8): 144 145- amd/registers: add correct gfx11.x enums for BINNING_MODE 146- radeonsi: disable binning correctly on gfx11.5 147- radeonsi/gfx11: fix programming of PA_SC_BINNER_CNTL_1.MAX_ALLOC_COUNT 148- radeonsi/gfx10.3: add a GPU hang workaround for legacy tess+GS 149- radeonsi/gfx11: add missing DCC_RD_POLICY setting 150- ac/llvm: fix SSBO bounds checking by using raw instead of struct opcodes 151- radeonsi: fix the DMA compute shader 152- r300: port scanout pitch alignment from the DDX to fix DRI3 153 154Mary Guillemard (1): 155 156- nvk: Always copy conditional rendering value before compare 157 158Matthew Waters (1): 159 160- teximage: allow glCopyTex{Sub}Image[123]D into R/RG textures with OpenGL ES 2.0 161 162Mike Blumenkrantz (13): 163 164- zink: destroy batch states after copy context 165- mesa: force rendertarget usage on required-renderable formats 166- zink: try getting sparse page size again without storage bit on fail 167- zink: set the sparse format usage flags directly based on queried props 168- zink: rename optimal_key in update_gfx_program_optimal() 169- zink: use the sanitized key in update_gfx_program_optimal() 170- zink: always sync and replace separable progs even with ZINK_DEBUG=noopt 171- zink: add even more strict checks for separate shader usage 172- glx: only print zink failure-to-load messages if explicitly requested 173- zink: iterate all the modes when doing separate shader fixups 174- zink: do io fixup on patch variables too 175- zink: defer present barrier to flush if a clear is pending 176- zink: clamp swapchain renderarea instead of asserting 177 178Patrick Lerda (1): 179 180- ac/llvm,radeonsi: fix memory leaks triggered by ac_nir_translate() errors 181 182Paulo Zanoni (1): 183 184- anv: don't leak device->vma_samplers 185 186Philipp Zabel (1): 187 188- rusticl: work around reference-to-mutable-static warnings 189 190Pierre-Eric Pelloux-Prayer (2): 191 192- winsys/radeon: pass priv instead NULL to radeon_bo_can_reclaim 193- radeonsi: preserve alpha if needed in kill_ps_outputs_cb 194 195Rhys Perry (4): 196 197- aco: don't reuse misaligned attribute destination VGPRs in VS prologs 198- radv: use dual_color_blend_by_location with Half-Life Alyx 199- aco/cssa: reset equal_anc_out if merging fails 200- aco/gfx11: fix scratch ST mode assembly 201 202Ruijing Dong (3): 203 204- radeonsi/vcn: add enc surface alignment caps 205- frontends/va: add surface alignment attribute 206- radeonsi/vcn: update to use correct padding size. 207 208Samuel Pitoiset (7): 209 210- ac/nir: fix exporting NGG streamout outputs with implicit PrimId from VS/TES 211- radv: disable binning correctly on GFX11.5 212- radv: fix programming of PA_SC_BINNER_CNTL_1.MAX_ALLOC_COUNT on GFX11 213- radv: fix occlusion queries with MSAA and no attachments 214- radv: add radv_force_pstate_peak_gfx11_dgpu and enable it for Helldivers 2 215- radv: add a workaround for null IBO on GFX6 216- radv: invalidate L2 metadata for VK_ACCESS_2_MEMORY_READ_BIT 217 218Yusuf Khan (1): 219 220- nvk: fix valve segfault from setting a descriptor set from NULL 221