1Mesa 25.0.1 Release Notes / 2025-03-05 2====================================== 3 4Mesa 25.0.1 is a bug fix release which fixes bugs found since the 25.0.0 release. 5 6Mesa 25.0.1 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 25.0.1 implements the Vulkan 1.4 API, but the version reported by 14the apiVersion property of the VkPhysicalDeviceProperties struct 15depends on the particular driver being used. 16 17SHA checksums 18------------- 19 20:: 21 22 TBD. 23 24 25New features 26------------ 27 28- None 29 30 31Bug fixes 32--------- 33 34- Zink: Kopper's present thread causes Wayland protocol races 35- GLmatrix needs aligned malloc 36- Lavapipe crashes if no Position is output in mesh shader 37- [RADV/aco][regression][bisected] - Avowed (2457220) - GPU hangs near Watermill outside of Dawnshore 38- radv/sqtt: assertion "layout transition marker should be only emitted inside a barrier marker" 39- [radv] Glitchy ground geometry regression in Total War Warhammer III on RX 7600 40 41 42Changes 43------- 44 45Benjamin Lee (4): 46 47- panfrost: remove NIR_PASS_V usage for noperspective lowering 48- panfrost: fix large int32->float16 conversions 49- panfrost: fix condition in bi_nir_is_replicated 50- panfrost/va: remove swizzle mod from LDEXP 51 52Caio Oliveira (1): 53 54- brw: Fix size in assembler when compacting 55 56Daniel Schürmann (5): 57 58- aco/scheduler: always respect min_waves on GFX10+ 59- aco/insert_exec_mask: Don't immediately set exec to zero in break/continue blocks 60- aco/insert_exec_mask: don't restore exec in continue_or_break blocks 61- aco/ssa_elimination: insert parallelcopies for p_phi immediately before branch 62- aco/assembler: Fix short jumps over chained branches 63 64Dave Airlie (1): 65 66- vulkan/wsi/x11: don't use update_region for damage if not created 67 68David Rosca (2): 69 70- frontends/va: Set AV1 max_width/height to surface size 71- radeonsi/vcn: Set all pic params for H264 encode references 72 73Dylan Baker (2): 74 75- iris: Correctly set NOS for geometry shader state changes 76- iris: fix handling of GL_*_VERTEX_CONVENTION 77 78Emmanuel Gil Peyrot (1): 79 80- panvk: Initialize out array with the correct length 81 82Eric Engestrom (8): 83 84- docs: add sha sum for 25.0.0 85- .pick_status.json: Update to b331713f20148852370a4fae5c2830d46801eb3b 86- .pick_status.json: Update to 55c476efed01121b3a64a58c304aae8ef9a79475 87- .pick_status.json: Mark b85c94fc891fe9d73b3a032aea8a6a71b8e6173b as denominated 88- .pick_status.json: Update to 4348253db5232b7be4db0a0ff47b31d51bc8f534 89- .pick_status.json: Update to fbc55afbdfc93a82c69f1cd6a1f4abbed96cfd19 90- .pick_status.json: Mark 5461ed5808421a8ffb79bdaa1449265f3e8f40a5 as denominated 91- .pick_status.json: Update to 45e771f4fbe4245b252c6360e55776080f0bf458 92 93Erik Faye-Lund (1): 94 95- mesa/main: wire up glapi bits for EXT_multi_draw_indirect 96 97Faith Ekstrand (12): 98 99- nak: Only use suld.constant on Ampere+ 100- zink: Use the correct array size for signal_values[] 101- zink: Use persistent semaphores for PIPE_FD_TYPE_SYNCOBJ 102- nvk: Don't bind a fragment shading rate image pre-Turing 103- nvk: Do not set INVALIDATE_SKED_CACHES pre-MaxwellB 104- nak/qmd: Add a nak_get_qmd_cbuf_desc_layout() helper 105- nvk: Handle pre-Turing dispatch indirect commands 106- nvk: Only support deviceGeneratedCommandsMultiDrawIndirectCount on Turing+ 107- nvk: Only support compute shader derivatives on Turing+ 108- zink: Don't present to Wayland surfaces asynchronously 109- egl/dri2: Rework get_wl_surface_proxy() 110- egl/wayland: Pass the original wl_surface to kopper 111 112Georg Lehmann (1): 113 114- aco/insert_exec: fix continue_or_break on gfx6-7 115 116Gert Wollny (1): 117 118- r600/sfn: gather info and set lowering 64 bit after nir_lower_io 119 120Guilherme Gallo (3): 121 122- ci/lava: Drop the repeating quotes on lava-test-case 123- ci/lava: Propagate errors in SSH tests 124- ci/lava: Add U-Boot action timeout for rockchip DUTs 125 126Hans-Kristian Arntzen (1): 127 128- radv: Always set 0 dispatch offset for indirect CS. 129 130Hyunjun Ko (1): 131 132- anv: Do not support the tiling of DRM modifier if DECODE_DST 133 134Iago Toral Quiroga (1): 135 136- pan/va: fix FAU validation 137 138James Hogan (5): 139 140- mesa: Consider NumViews to reuse FBO attachments 141- mesa: Handle GL_FRAMEBUFFER_INCOMPLETE_VIEW_TARGETS_OVR 142- mesa: Check views don't exceed GL_MAX_ARRAY_TEXTURE_LAYERS 143- mesa: OVR_multiview framebuffer attachment parameters 144- mesa: Handle getting GL_MAX_VIEWS_OVR 145 146Job Noorman (1): 147 148- ir3/ra: prevent reusing parent interval of reloaded sources 149 150Juan A. Suarez Romero (2): 151 152- v3dv: duplicate key for texel_buffer cache 153- broadcom/simulator: use string copy instead of memcpy 154 155Karol Herbst (3): 156 157- rusticl/mem: set num_samples and num_mip_levels to 0 when importing from GL 158- rusticl/platform: advertise all extensions supported by all devices 159- intel/brw, lp: enable lower_pack_64_4x16 160 161Kevin Chuang (2): 162 163- anv/bvh: Fix encoder handling sparse buffer 164- anv/bvh: Fix copy shader handling sparse buffer 165 166Konstantin Seurer (1): 167 168- llvmpipe: Skip draw_mesh if the ms did not write gl_Position 169 170Lars-Ivar Hesselberg Simonsen (2): 171 172- panfrost: Use RUN_COMPUTE over RUN_COMPUTE_INDIRECT 173- panvk: Use RUN_COMPUTE over RUN_COMPUTE_INDIRECT 174 175Lionel Landwerlin (2): 176 177- anv: fix missing 3DSTATE_PS:Kernel0MaximumPolysperThread programming 178- vulkan/runtime: ensure robustness state is fully initialized 179 180Lorenzo Rossi (1): 181 182- nvk: Fix MSAA sparse residency lowering crash 183 184Marek Olšák (1): 185 186- mesa: allocate GLmatrix aligned to 16 bytes 187 188Mary Guillemard (1): 189 190- pan/bi: Disallow FAU special page 3 and WARP_ID on message instructions 191 192Mike Blumenkrantz (6): 193 194- zink: wait on tc fence before checking for fd semaphore 195- zink: always fully unwrap contexts 196- zink: clamp UBO sizes instead of asserting 197- llvmpipe: pass layer count to rast clear 198- gallium: fix pipe_framebuffer_state::view_mask 199- mesa: avoid creating incomplete surfaces when multiview goes out of range 200 201Natalie Vock (1): 202 203- radv/rt: Don't allocate the traversal shader in a capture/replay range 204 205Patrick Lerda (3): 206 207- r600: fix evergreen_emit_vertex_buffers() related cl regression 208- r600: fix the indirect draw 8-bits path 209- r600: fix emit_image_size() range base compatibility 210 211Paulo Zanoni (1): 212 213- brw: extend the NOP+WHILE workaround 214 215Peyton Lee (1): 216 217- radeonsi/vpe: check reduction ratio 218 219Pierre-Eric Pelloux-Prayer (2): 220 221- tc: add missing TC_SENTINEL for TC_END_BATCH 222- mesa/st: call _mesa_glthread_finish before _mesa_make_current 223 224Rhys Perry (1): 225 226- ac/nir: fix tess factor optimization when workgroup barriers are reduced 227 228Roland Scheidegger (1): 229 230- llvmpipe: Fix alpha-to-coverage without dithering 231 232Samuel Pitoiset (3): 233 234- radv/video: fix adding the query pool BO to the cmdbuf list 235- radv: fix missing SQTT barriers for fbfetch color/depth decompressions 236- radv: fix re-emitting fragment output state when resetting gfx pipeline state 237 238Tapani Pälli (2): 239 240- iris: wait for imported fences to be available in iris_fence_await 241- iris: remove dead code that cannot get hit anymore 242 243Yiwei Zhang (2): 244 245- venus: fix image format cache miss with AHB usage query 246- venus: relax the requirement for sync2 247 248Yogesh Mohan Marimuthu (1): 249 250- winsys/amdgpu: same_queue variable should be set if there is only one queue 251