Lines Matching +full:opencl +full:- +full:icd +full:- +full:loader

1 Mesa 20.2.0 Release Notes / 2020-09-28
20 ---------------
24     63f0359575d558ef98dd78adffc0df4c66b76964ebf603b778b7004964191d30  mesa-20.2.0.tar.xz
28 ------------
30 - GL_ARB_compute_variable_group_size on Iris.
32 - GL_ARB_gpu_shader5 on llvmpipe
34 - GL_ARB_post_depth_coverage on llvmpipe
36 - GLES 3.2 on llvmpipe
38 - GL_EXT_shader_group_vote on GLES3.
40 - GL_EXT_texture_shadow_lod on llvmpipe
42 - VK_AMD_texture_gather_bias_lod on RADV.
44 - VK_AMD_gpu_shader_half_float on RADV/ACO.
46 - VK_AMD_gpu_shader_int16 on RADV/ACO.
48 - VK_EXT_extended_dynamic_state on ANV and RADV.
50 - VK_EXT_image_robustness on RADV.
52 - VK_EXT_private_data on ANV and RADV.
54 - VK_EXT_custom_border_color on ANV and RADV.
56 - VK_EXT_pipeline_creation_cache_control on ANV and RADV.
58 - VK_EXT_shader_demote_to_helper_invocation on RADV/LLVM.
60 - VK_EXT_subgroup_size_control on RADV/ACO.
62 - VK_GOOGLE_user_type on ANV and RADV.
64 - VK_KHR_shader_subgroup_extended_types on RADV/ACO.
66 - GL_ARB_gl_spirv on nvc0/nir.
68 - GL_ARB_spirv_extensions on nvc0/nir.
70 - RADV now uses ACO per default as backend
72 - RADV_DEBUG=llvm option to enable LLVM backend for RADV
74 - VK_EXT_image_robustness for ANV
76 - VK_EXT_shader_atomic_float on ANV
78 - VK_EXT_4444_formats on ANV and RADV.
80 - VK_KHR_memory_model on RADV.
82 - GL 4.5 on llvmpipe
84 - EGL_KHR_swap_buffers_with_damage on X11 (DRI3)
88 ---------
90 - [Regression][Bisected][20.2][radeonsi] American Truck Simulator continually allocates memory unti…
91 - anv: dEQP-VK.robustness.robustness2.* failures on gen12
92 - [RADV] Problems reading primitive ID in fragment shader after tessellation
93 - Massive memory leak (at least AMD, others unknown)
94 - Substance Painter 6.1.3 black glitches on Radeon RX570
95 - vkCmdCopyImage broadcasts subsample 0 of MSAA src into all subsamples of dst on RADV
96 - Crash in ruvd_end_frame when calling vaBeginPicture/vaEndPicture without rendering anything
97 - X-Plane 11 Installer crashes on startup since `glsl: declare gl_Layer/gl_ViewportIndex/gl_Viewpor…
98 - Horizon Zero Dawn graphics corruption with with radv
99 - Amber test opt_peel_loop_initial_if: Assertion failed
100 - Dirt Rally: Flickering glitches on certain foliage since Mesa 20.1.0 caused by MSAA
101 - [BRW] WRC 5 asserts with gallium nine and iris.
102 - radv: Corruption in "The Surge 2"
103 - [RADV] Detroit: Become Human Demo game lock-ups with RADV
104 - Road Redemption certain  graphic effects rendered white color
105 - vulkan/wsi/x11: deadlock with Xwayland when compositor holds multiple buffers
106 - [RADV/ACO] Death Stranding cause a GPU hung (*ERROR* Waiting for fences timed out!)
107 - lp_bld_init.c:172:7: error: implicit declaration of function ‘LLVMAddConstantPropagationPass’; di…
108 - Intel Vulkan driver crash with alpha-to-coverage
109 - EGL_KHR_swap_buffers_with_damage support on X11
110 - radv: blitting 3D images with linear filter
111 - [ACO] Compiling pipelines from RPCS3's shader interpreter spins forever in ACO code
112 - Intel Vulkan driver assertion with small xfb buffer
113 - [spirv-fuzz] SPIR-V parsing failed "src->type->type == dest->type->type"
114 - radeonsi: radeonsi crashes in Chrome on chromeos
115 - [RADV] commit d19bc94e4eb94 broke gamescope with Navi
116 - 4e3a7dcf6ee4946c46ae8b35e7883a49859ef6fb breaks Gamescope showing windows properly.
117 - anv: crashes in CTS test dEQP-VK.subgroups.*.framebuffer.*_tess_eval
118 - Intel Vuikan (anv) crash in copy_non_dynamic_state() when using validation layer
119 - Mafia 3: Trees get rendered incorrectly
120 - radv: dEQP-VK.synchronization.op.multi_queue.timeline_semaphore.write_clear_attachments_*_concurr…
121 - Crash on GTA 5 through proton 5.0.9 and GE versions
122 - Mesa 20.2.0-rc1 fails to build for AMD
123 - Assertion failure compiling shader from Zigguart
124 - Panfrost locks for waiting fence when running Source engine games
125 - ci: `-Dtools=panfrost` should be build-tested
126 - panfrost: Register allocation fails for Firefox WebRender shaders
127 - VRAM leak with vuilkan external memory + opengl memory objects
128 - [vulkan/build] Recent build system changes made VK_EXT_acquire_xlib_display unnecessarily depend …
129 - ci: Capture devcoredumps on chezas
130 - Possible array out of bounds in brw_vec4_nir.cpp
131 - freedreno/a6xx: incorrect rendering in asphalt 9
132 - [tgl][bisected][regression][iris] failure on dEQP-EGL.functional.wide_color.pbuffer_8888_colorspa…
133 - Multiply defined symbols compiling with gcc@10.1.0
134 - shrinking descriptor pool on intel+vulkan
135 - dEQP-VK.renderpass2.dedicated_allocation.attachment.1.12 fails on NAVI14
136 - turnip: binning and indirect dependency
137 - Amber test leads to NIR validation failed after nir_opt_if (on spirv-fuzz shader)
138 - Unable to compile mesa-git from b559d26c
139 - Ambient light too bright with ACO in AC: Odyssey
140 - Multiple issues with Detroit Become Human
141 - ci: Capture artifacts in baremetal mode
142 - turnip/ir3: fine derivatives
143 - panfrost: regression: Major stuttering and low compositor FPS with glmark2
144 - khr_debug-push-pop-group_gl: ../src/util/simple_mtx.h:86: simple_mtx_lock: Assertion \`c != _SIMP…
145 - freedreno/a6xx: skai/skqp fails
146 - SPIR-V parsing fails in src/compiler/spirv/spirv_to_nir.c
147 - SPIR-V parsing fails in src/compiler/spirv/vtn_cfg.c
148 - Weird GLSL bug
149 - iris driver is broken in Freedesktop 19.08
150 - LLVM not properly shutdown in `si_pipe.c`?
151 - Panfrost: add current status to docs/features.txt
152 - Opengl incorrect rendering on yuzu Amd
153 - RADV: VK_ACCESS_MEMORY_READ/WRITE_BIT is not implemented
154 - [bisected][regression][all platforms] multiple deqp-gles31/glescts/piglit failures
155 - 7406ea37, "ac/surface: require that gfx8 doesn't have DCC in order to be displayable", breaks Gam…
156 - vkGetSemaphoreCounterValue doesn't update without vkWaitSemaphores calls on Intel UHD 620
157 - [RADV] System crash when playing XCOM Chimera Squad because of commit #7a5e6fd2
158 - [RADV] Non-precise occlusion queries return non-zero when all fragments are discarded
159 - [DXVK] Project Cars rendering problems
160 - ADDRLIB ODR Violation
161 - Build fails with current mesa from git "undefinierter Verweis auf »nir_lower_clip_disable«"
162 - KDE Compositor stuttering after Check for window destruction in dri3_wait_for_event_locked
163 - Add fallthrough to prevent errors caused by missing break
164 - i965/20.1: gray rendering with torcs racing
165 - glBindBufferRange call seems to be ignored by one of two shader-programs on radeon cards
166 - [bisected][g33] piglit.spec.ext_framebuffer_object.fbo-cubemap failure
167 - Increase GL_MAX_COMPUTE_SHADER_STORAGE_BLOCKS to greater value.
168 - nir: st_nir_lower_builtin fails for gl_LightSource[i]
169 - Sometimes VLC player process gets stuck in memory after closure if video output used is Auto or O…
170 - Double unlock in rbug_context.c
171 - Double copy for TexSubImage
172 - [v3d] corruption when GS omits some vertices
173 - Iris crashes when reading from multisampled front buffer on platforms without front buffer
174 - freedreno: subway surfers crash when repeatedly toggling fullscreen
175 - [RADV/GFX8] Performance drop in DOOM Eternal when "Present from compute" is enabled
176 - freedreno: multiple applications crash on a5xx
177 - Use-after-free crash innv50_ir::GCRA::RIG_Node::init()
178 - intel: Sample mask writes need to be honored in Vulkan
179 - [RADV] - Path of Exile (238960) - Map outline, landscape and markers are missing with the Vulkan …
180 - ASTC texture decompression fails when using software fallback
181 - [i965][iris][regression][bisected] multiple piglit and glcts failures on all platforms
182 - please publish GPG keyring used to sign new releases
183 - [BISECTED] compiling shader causes crash
184 - Missing render Information on Stellaris
185 - freedreno/ir3: allow copy-propagate from array
186 - Zink + GALLIUM_HUD SIGSEGV
187 - piglit spec@egl_ext_device_base@conformance fails LLVM 11 Git assertion since "llvmpipe/fs: add c…
188 - llvmpipe: 1x1 framebuffer with a 2x2 viewport
189 - [regression] nir build failure
190 - ci: need to end baremetal tests after kernel panic/instaboot
191 - If-statement body is executed for false condition
192 - freedreno/a6xx: broken rendering in playcanvas "after the flood"
193 - [regression] performance drop on Dota 2, CS:GO, and gfxbench GL benchmarks on ICL/Iris
194 - [amd] C++ ODR violatation for union GB_ADDR_CONFIG
195 - Zink reports incorrect amount of video memory
196 - [RADV/LLVM]: void llvm::ICmpInst::AssertOK(): Assertion \`getOperand(0)->getType() == getOperand(…
197 - glsl-1.50-gs-max-output hangs on Navi10 + NGG
198 - anv: Runs out of binding tables with PPSSPP during long runs
199 - Segfault in Panfrost with waypipe
200 - ci: Use rsync instead of rm -rf ; cp for baremetal rootfs
201 - i965: Rendering problems replaying a trace of "Refunct" after mesa-20.1.0-rc1 release [bisected]
202 - Panfrost (rk3399 NanoPi M4) hang/crash on playing video on Kodi/X11
203 - gallium/winsys/radeon/drm fails assertion on 32bit
204 - NIR validation failed after glsl to nir, before function inline, wrong {src,dst}->type ?
205 - nir/spirv asin() function not precise enough
206 - Mesa 20.0.7 / 20.1.0-rc4 regression, extremally long shader compilation time in NIR
207 - Android build error after 689acc73
208 - freedreno/a6xx: gpu hangs in google earth
209 - Mesa-git build fails on Fedora Rawhide
210 - Doom Eternal 1.1 performs very poorly on RADV
211 - iris/i965: possible regression in 20.0.5 due to changes in buffer manager sharing across screens …
212 - iris/i965: possible regression in 20.0.5 due to changes in buffer manager sharing across screens …
213 - Incorrect _NetBSD__ macro inside execmem.c
214 - Possible invalid sizeof in device.c
215 - YUV FP16 lowering validation failing
216 - GLSL compiler assertion is_float() failed in glsl/ir_validate.cpp, visit_leave on specific WebGL …
217 - [RADV] - Doom Eternal (782330) & Metro Exodus (412020) - Title requires 'RADV_DEBUG=zerovram' to …
218 - [RADV] - Doom Eternal (782330) & Metro Exodus (412020) - Title requires 'RADV_DEBUG=zerovram' to …
219 - mesa trunk master vulkan overlay-layer meson.build warning empty configuration_data() object
220 - [meson] increase minimum required version
221 - Kicad fails to render 3D PCB models.
222 - freedreno: minetest: alpha channel issue on a6xx
223 - Reproduceable i915 gpu hang Intel Iris Plus Graphics (Ice Lake 8x8 GT2)
224 - 7 Days to Die - "Reflection Quality" setting broken, results in environment rendered black
225 - glsl: regression affecting shader compilation time
226 - freedreno: glamor issue with x11 desktops
227 - finish converting from fnv1a to xxhash
228 - Hang in iris_dri in kitty
229 - Setting twice value to output_stream in radv_nir_to_llvm.c
230 - Overwriting value of `jit_tex->sample_stride` in lp_setup.c
231 - [AMDGPU][OpenGL] apitrace of kernel/firmware crash that requires a reboot
232 - Flickering in Superposition benchmark
233 - Double lock in fbobject.c
234 - Possible typo in aco_insert_waitcnt.cpp
235 - [bisected] Steam crashes when newest Iris built with LTO
236 - Freeing null pointer inside radv_amdgpu_cs.c
237 - Duplicated sub expression in radv_nir_to_llvm.c
238 - i965/vec4: opt_cse_local cause the out of bound array access
239 - NIR: Regression on shader using 8/16-bit integers
240 - ACO: Compiler segfault on 8/16-bit integers.
241 - lp_bld_intr.c:70:16: error: use of undeclared identifier 'LLVMFixedVectorTypeKind'; did you mean …
242 - recent seqno changes causing surfaceflinger crash
243 - [radeonsi] [glthread] Crash with glthread enabled
244 - Deadlock in anv_timelines_wait()
245 - [gles3] supertuxkart: some textures are incorrect
246 - post_version.py does not work with release candidates
247 - post_version.py does not work with release candidates
248 - radv regression on android
249 - ogl: Set mesa_glthread=true as default on the RPCS3 emulator
250 - [iris] android deqp dEQP-EGL.functional.robustness.negative_context#invalid_notification_strategy…
251 - zink: conditional rendering
252 - [RadeonSI] Glitches on VEGA8 + RX 560X after MR 4863
253 - RadeonSI OpenGL broken for GFX8 after unify code for overriding offset
254 - freedreno/turnip: Don't request fragcoord components we don't use
255 - Make check fails in ANV
256 - src\util\meson.build:294:4: ERROR: Program or command 'winepath' not found or not executable
257 - Please add Zink to features.txt
258 - llvmpipe: assert triggers in LLVM
259 - debug builds are massively broken on Windows
260 - ci: Report flakes on IRC from baremetal tests
261 - heavy glitches on amd ryzen 5 since version 20.x
262 - zink asserts with 32-bit boolean
263 - OpenGL: Surviving Mars black screen late-game (possible shader problem)
264 - Kerbal Space Program (KSP) hangs entire Navi system
265 - Dirt: Showdown bad performance and broken rendering with enabled advanced lightning
266 - gravit & Firefox WebGL broken since 3dc2ccc14c0e035368fea6ae3cce8c481f3c4ad2 "ac/surface: replace…
267 - mesa 20.0.5 causing kitty to crash
268 - radeonsi: "Torchlight II" trace showing regression on mesa-20.0.6 [bisected]
269 - [RADV/LLVM/ACO/Regression] After mesa commit a3dc7fffbb7be0f1b2ac478b16d3acc5662dff66 all games s…
270 - Android building error after commit 2ab45f41
271 - freedreno/a6xx: pubg rendering glitches
272 - iris: Crash when trying to capture window in OBS Studio
273 - lp_test_format failure with llvm-11
277 -------
281 - egl: Limit the EGL ver for android
285 - glx: Fix build and warnings with -Dglx=dri -Dglx-direct=false
289 - v3d/tex: only look up the 2nd texture gather offset for 1d non-arrays
290 - v3d/tex: set up default values for Configuration Parameter 1 if possible
291 - v3d/tex: use TMUSLOD register if possible
292 - v3d: moving v3d simulator to src/broadcom
293 - v3d/tex: handle correctly coordinates for cube/cubearrays images
294 - vulkan/util: add struct vk_pipeline_cache_header
295 - nir/lower_tex: handle query lod with nir_lower_tex_packing_16 at lower_tex_packing
296 - v3d/packet: fix typo on Set InstanceID/PrimitiveID packet
297 - v3d: set instance id to 0 at start of tile
301 - pan/mdg: Track more types
302 - pan/mdg: Be a bit more pedantic in invert passes
303 - panfrost: Enumify bifrost blend types
304 - pan/bi: Add texture indices to IR
305 - pan/bi: Pipe multiple textures through
306 - pan/bi: Pack round opcodes (FMA, either 16 or 32)
307 - pan/bit: Add framework forinterpreting double vs float
308 - pan/bit: Interpret ROUND
309 - pan/bit: Add round tests
310 - panfrost: Fix texture field size
311 - panfrost: Fix size of bifrost sampler descriptor
312 - panfrost: Fix sampler wrap/filter field orders
313 - panfrost: Fix norm coords on bifrost sampler
314 - panfrost: Fix tiled texture "stride"s on Bifrost
315 - pan/decode: Don't crash on missing payload
316 - pan/bi: Enable lower_mediump_outputs NIR pass
317 - panfrost: Update Bifrost fields in mali_shader_meta
318 - pan/bi: Lower for now sincos
319 - pan/mdg: Ingest actual isub ops
320 - pan/mdg: Rename .one to .sat_signed
321 - pan/mdg: Move constant switch opts to algebraic pass
322 - pan/mdg: Drop forever todo
323 - pan/mdg: Drop `opt` in name of midgard_opt_cull_dead_branch
324 - pan/mdg: Enable nir_opt_algebraic_distribute_src_mods
325 - panfrost: Update dEQP expectation list
326 - panfrost: Setup gl_FragCoord as sysval on Bifrost
327 - pan/bi: Add clause type for gl_FragCoord.zw load
328 - pan/bi: Abort on unknown op packing
329 - pan/bi: Abort on unhandled intrinsics
330 - pan/bi: Futureproof COMBINE lowering against non-u32
331 - pan/bi: Print bad instruction on src packing fail
332 - pan/bi: Passthrough direct ld_var addresses
333 - pan/bi: Lower gl_FragCoord
334 - pan/bi: Set clause type for gl_FragCoord.z
335 - pan/bi: Fix double-abs flipping
336 - pan/bi: Fix missing swizzle
337 - pan/bi: Fix incorrectly flipped swizzle
338 - pan/bi: Disable CSEL4 emit for now
339 - pan/bi: Fix DISCARD ops in disasm
340 - pan/bi: Structify DISCARD
341 - pan/bi: Remove BI_GENERIC
342 - pan/bi: Unwrap BRANCH into CONDITIONAL class
343 - pan/bi: Handle discard_if in NIR->BIR naively
344 - pan/bi: Emit discard (not if)
345 - pan/bi: Add float-only mode to condition fusing
346 - pan/bi: Fuse conditions into discard_if
347 - pan/bi: Handle discard/branch in get_component_count
348 - pan/bi: Pack ADD.DISCARD
349 - pan/bi: Structify ADD ICMP 16
350 - pan/bi: Pack ADD ICMP 32
351 - pan/bi: Pack ADD ICMP 16
352 - pan/bi: Don't pack ICMP on FMA
353 - pan/bit: Add swizzles to round tests
354 - pan/bit: Add more 16-bit fmod tests
355 - pan/bit: Add ICMP tests
356 - pan/bi: Rename BI_ISUB to BI_IMATH
357 - pan/bi: Use IMATH for nir_op_iadd
358 - pan/bi: Pack FMA IADD/ISUB 32
359 - pan/bi: Pack ADD IADD/ISUB for 8/16/32
360 - pan/bi: Add SUB.v2i16/SUB.v4i8 opcodes to disasm
361 - pan/bi: Don't schedule <32-bit IMATH to FMA
362 - pan/bit: Interpret IMATH
363 - pan/bit: Interpret v4i8 ops
364 - pan/bit: Remove test names
365 - pan/bit: Use swizzle helper for round
366 - pan/bit: Factor out identity swizzle helper
367 - pan/bit: Add IMATH packing tests
368 - pan/decode: Fix flags_hi printing
369 - pan/mdg: Explain helper invocations dataflow theory
370 - pan/mdg: Analyze helper invocation termination
371 - pan/mdg: Analyze helper execution requirements
372 - pan/mdg: Use the helper invo analyze passes
373 - pan/mdg: Use analysis to set .cont/.last flags
374 - pan/mdg: Remove texture_op_count
375 - pan/mdg: Set types for derivatives
376 - pan/mdg: Fix derivative swizzle
377 - panfrost: Run dEQP-GLES3.functional.shaders.derivate.* on CI
378 - pan/decode: Use a page table for tracking mmaps
379 - pan/decode: Fix min/max_tile_coord mixup
380 - pan/mfbd: Add format codes for PIPE_FORMAT_B5G5R5A1_UNORM
381 - panfrost: Switch formats to table
382 - panfrost: Fix Z24 vs Z32 mixup
383 - panfrost: Enable AFBC for Z24X8
384 - nir: Add fsat_signed opcode
385 - nir: Add fclamp_pos opcode
386 - panfrost: Add modifier detection helpers
387 - pan/mdg: Remove .pos propagation pass
388 - pan/mdg: Drop nir_lower_to_source_mods
389 - pan/mdg: Prepare for modifier helpers
390 - pan/mdg: Ingest fsat_signed/fclamp_pos
391 - pan/mdg: Apply abs/neg modifiers
392 - pan/mdg: Treat inot as a modifier
393 - pan/mdg: Remove invert optimizations
394 - pan/mdg: Use helpers for branch/discard inversion
395 - pan/mdg: Apply outmods
396 - pan/mdg: Emit fcsel when beneficial
397 - pan/mdg: Optimize pipelining logic
398 - pan/mdg: Precompute mir_special_index
399 - pan/mdg: Optimize liveness computation in DCE
400 - pan/mdg: Handle comparisons in fp16 path
401 - pan/mdg: Fix constant combining crash
402 - pan/mdg: Remove mir_*size routines
403 - pan/mdg: Remove mir_get_alu_src
404 - pan/mdg: Include more types
405 - pan/mdg: Handle dest up/lower correctly with swizzles
406 - pan/mdg: Respect !32-bit sizes in RA
407 - pan/mdg: Explain ld/st sign/zero extension
408 - pan/mdg: Add abs/neg/shift modifiers to IR
409 - pan/mdg: Use src_types to determine size in scheduling
410 - pan/mdg: Use type to determine triviality of a move
411 - pan/mdg: Identify scalar integer mods
412 - pan/mdg: Promote imov to fmov on a NIR level
413 - pan/mdg: Remove promote_float pass
414 - pan/mdg: Defer modifier packing until emit time
415 - pan/mdg: Remove redundant redundancy
416 - pan/mdg: Streamline dest_override handling
417 - pan/mdg: Implement b2f16
418 - pan/mdg: Don't generate conversions for fp16 LUTs
419 - pan/mdg: Ignore dest.type when offseting load swizzle
420 - pan/lcra: Remove unused alignment parameters
421 - pan/lcra: Allow per-variable bounds to be set
422 - pan/mdg: Use type size to determine alignment
423 - pan/mdg: Eliminate load_64
424 - pan/mdg: Set RA bounds for fp16
425 - pan/mdg: Print mask when dest=0
426 - pan/mdg: Round up bytemasks when spilling
427 - pan/mdg: Print constant vectors less wrong
428 - pan/mdg: Factor out mir_adjust_constant
429 - pan/mdg: Only combine 16-bit constants to lower half
430 - pan/mdg: Separately pack constants to the upper half
431 - pan/mdg: Fix type checking issues with compute
432 - pan/mdg: Pack barriers correctly
433 - pan/mdg: Use shifts instead of division for RA sizes
434 - pan/mdg: Implement vector constant printing for 8-bit
435 - pan/mdg: Implement condense_writemask for 8-bit
436 - pan/mdg: Pack 8-bit swizzles in 16-bit ops
437 - panfrost: Guard experimental fp16 behind debug flag
438 - panfrost: Keep cached BOs mmap'd
439 - panfrost: Remove deadcode
440 - panfrost: Fill in SCALED formats to format table
441 - panfrost: Don't set PIPE_CAP_VERTEX_BUFFER_STRIDE_4BYTE_ALIGNED_ONLY
442 - panfrost: Don't zero staging buffer for tiling
443 - panfrost: Allow bpp24 tiling
444 - panfrost: Allow tiling on RECT textures
445 - panfrost: Limit blend shader work count
446 - panfrost: Remove dated comment about leaks
447 - panfrost: Disable tib read/write when colourmask = 0x0
448 - panfrost: Avoid redundant shader executions with mask=0x0
449 - panfrost: Don't set CAN_DISCARD for MFBD
450 - panfrost: Fix transform feedback types
451 - pan/mdg: Cleanup comments that look like division
452 - pan/mdg: Eliminate expand_writemask division
453 - pan/mdg: Eliminate 64-bit swizzle packing division
454 - pan/mdg: Avoid division in printing helpers
455 - pan/mdg: Eliminate remaining divisions from compiler
456 - panfrost: Fix dated comment
457 - panfrost: Use _mesa_roundevenf when packing clear colours
458 - panfrost: Handle !independent_blend for blend shaders
459 - pan/mdg: Add pack_colour_32 opcode
460 - pan/mdg: Lower shifts to 32-bit
461 - pan/mdg: Ensure we don't DCE into impossible masks
462 - pan/mdg: Allow DCE on ld_color_buffer masks
463 - panfrost: Add debug print before query flushes
464 - panfrost: Only run batch debug when specifically asked
465 - nir: Add un/pack_32_4x8 opcodes
466 - util: Add SATURATE macro
467 - util/format: Use SATURATE
468 - mesa: Use SATURATE
469 - mesa/swrast: Use SATURATE
470 - gallium/draw: Use SATURATE
471 - glsl: Use SATURATE
472 - panfrost: Use SATURATE
473 - softpipe: Use SATURATE
474 - intel: Use SATURATE
475 - i965: Use SATURATE
476 - iris: Use SATURATE
477 - etnaviv: Use SATURATE
478 - nouveau: Use SATURATE
479 - pan/decode: Fix unused variable warning
480 - pan/decode: Fix tiler warning
481 - pan/decode: Dump missing field on Bifrost
482 - pan/decode: Dump unknown2
483 - panfrost: Fix Bifrost blending with depth-only FBO
484 - panfrost: Adjust null_rt for Bifrost
485 - panfrost: Tweak zsbuf magic numbers for Bifrost
486 - panfrost: Tweak Bifrost colour buffer magic
487 - panfrost: Force Z/S tiling on Bifrost
488 - panfrost: Share MRT blend flag calculation with Bifrost
489 - panfrost: Set unk2 to accomodate blending
490 - panfrost: Identify Bifrost texture format swizzle
491 - panfrost: Ensure nonlinear strides are 16-aligned
492 - panfrost: Document Midgard Inf/NaN suppress bit
493 - panfrost: Add defines for bifrost unk1 flags
494 - panfrost: Identify MALI_BIFROST_EARLY_Z flag
495 - panfrost: Set MALI_BIFROST_EARLY_Z as necessary
496 - pan/decode: Decode Bifrost shader flags
497 - pan/bi: Add TEX.vtx opcode for vertex texturing
498 - pan/bi: Also add compact vertex texturing
499 - pan/bi: Document compute_lod bit for compact tex
500 - pan/bi: Allow vertex txl with lod=0 as compact
501 - pan/bi: Add f16 TEXC.vtx op
502 - pan/bi: Pack compact vertex texturing
503 - pan/bi: Add CSEL.16 packing tests
504 - pan/bi: Suppress inf/nan for now
505 - panfrost: Don't generate gl_FragCoord varying on Bifrost
506 - panfrost: Set reads_frag_coord as a sysval
507 - panfrost: Preload gl_FragCoord on Bifrost
508 - pan/bi: Remove FMA? parameter from get_src
509 - pan/bi: Remove comment about old scheduler design
510 - pan/bi: Move bi_registers to common IR structures
511 - pan/bi: Move bi_registers to bi_bundle
512 - pan/bi: Drop `struct` from bi_registers
513 - pan/bi: Add FILE* argument to bi_print_registers
514 - pan/bi: Move bi_flip_ports out of port assignment
515 - pan/bi: Document constant count invariant
516 - pan/bi: Disassemble pos=0xe
517 - pan/bi: Add MUL.i32 to disasm
518 - pan/bi: Remove more artefacts of 2-pass scheduling
519 - pan/bi: Add bi_layout.c for clause layout helpers
520 - pan/bi: Add helper to measure clause size
521 - pan/bi: Remove schedule_barrier
522 - pan/bi: Allow printing branches without targets
523 - pan/bi: Fix emit_if successor assignment
524 - pan/bi: Only rewrite COMBINE dest if not SSA
525 - pan/bi: Fix CONVERT component counting
526 - pan/bi: Fix branch condition typesize
527 - pan/bi: Passthrough ZERO in branch packing
528 - pan/bi: Add branch constant field to IR
529 - pan/bi: Pack branch offset constants
530 - pan/bi: Set branch_constant if there is a branch
531 - pan/bi: Assign constant port for branch offsets
532 - pan/bi: Preliminary branch packing
533 - pan/bi: Link clauses back to their blocks
534 - pan/bi: Add bi_foreach_clause_in_block_from{_rev} helpers
535 - pan/bi: Measure distance between blocks
536 - pan/bi: Pack proper clause offsets
537 - pan/bi: Set branch_conditional if b2b is set
538 - pan/bi: Set back-to-back bit more accurately
539 - pan/bi: Set branch conditional bit
540 - pan/bi: Pack unconditional branch
541 - pan/bi: Defer block naming until after emit
542 - pan/bi: Add bi_foreach_block_from_rev helper
543 - pan/bi: Measure backwards branches as well
544 - pan/bi: Allow two successors in header packing
545 - pan/bi: Passthrough deps of the branch target
546 - panfrost: Disable QUAD_STRIP/POLYGON on Bifrost
547 - panfrost: Add GPU IDs for G31/G52
548 - panfrost: Probe G31/G52 if PAN_MESA_DEBUG=bifrost
549 - pan/mdg: Handle un/pack opcodes as moves
550 - pan/mdg: Add pack_unorm_4x8 via 8-bit
551 - pan/mdg: Treat packs "specially"
552 - pan/mdg: Handle bitsize for packs
553 - pan/mdg: Print 8-bit constants
554 - pan/mdg: Drop the u8 from the colorbuf op names
555 - pan/mdg: Implement raw colourbuf loads on T720
556 - panfrost: Add theory for new framebuffer lowering
557 - panfrost: Determine unpacked type for formats
558 - panfrost: Add quirks for blend shader types
559 - panfrost: Determine load classes for formats
560 - panfrost: Determine classes for stores
561 - panfrost: Stub out lowering boilerplate
562 - panfrost: Un/pack pure 32-bit
563 - panfrost: Un/pack pure 16-bit
564 - panfrost: Un/pack pure 8-bit
565 - panfrost: Un/pack 8-bit UNORM
566 - panfrost: Flesh out dispatch
567 - panfrost: Un/pack UNORM 4
568 - panfrost: Un/pack RGB565 and RGB5A1
569 - panfrost: Un/pack RGB10_A2_UNORM
570 - panfrost: Un/pack RGB10_A2_UINT
571 - panfrost: Un/pack R11G11B10
572 - panfrost: Un/pack sRGB via NIR
573 - panfrost: Switch to pan_lower_framebuffer
574 - panfrost: Conditionally allow fp16 blending
575 - panfrost: Account for differing types in blend lower
576 - panfrost: Let Gallium pack colours
577 - panfrost: Check for large tilebuffer requirements
578 - panfrost: Add separate_stencil BO to batch
579 - panfrost: Use internal_format throughout
580 - panfrost: Update fails list
581 - pan/mdg: Handle 16-bit ld_vary
582 - pan/mdg: Fuse f2f16 into load_interpolated_input
583 - panfrost: Fix PRESENT flag mix-up
584 - panfrost: Permit AFBC of RGB8
585 - panfrost: Use VTX tag for vertex texturing
586 - panfrost: Don't flush explicitly when mipmapping
587 - panfrost: Remove unused nir_lower_framebuffer pass
588 - pan/mdg: Disassemble out-of-order bits
589 - pan/mdg: Add quirk for missing out-of-order support
590 - pan/mdg: Enable out-of-order execution after texture ops
591 - nir: Fold f2f16(b2f32(x)) to b2f16(x)
592 - pan/mdg: Don't double-replicate blend on T720
593 - pan/mdg: Distinguish blend shaders in internal shader-db
594 - pan/mdg: Add roundmode enum
595 - pan/mdg: Add opcode roundmode property
596 - pan/mdg: Lower roundmodes
597 - pan/mdg: Implement \*_rtz conversions with roundmode
598 - pan/mdg: Fold roundmode into applicable instructions
599 - pan/mdg: Handle f2u8
600 - pan/mdg: Allow f2u8 and friends thru
601 - pan/mdg: Handle regular nir_intrinsic_load_output
602 - panfrost: Passthrough NATIVE loads/stores
603 - pan/bi: Handle SEL with vec3 16-bit
604 - pan/bi: Fix SEL.16 swizzle
605 - pan/bi: Pack second argument of F32_TO_F16
606 - pan/bi: Passthrough second argument of F32_TO_F16
607 - pan/bi: Handle vectorized load_const
608 - panfrost: Update MALI_EARLY_Z description
609 - panfrost: Document MALI_WRITES_GLOBAL bit
610 - panfrost: Handle writes_memory correctly
611 - panfrost: Readd MIDGARD_SHADERLESS quirk to t760
612 - panfrost: Explicitly convert to 32-bit for logic-ops
613 - pan/bi: Disassemble gl_PointCoord reads.
614 - panfrost: Prefer sysval for gl_PointCoord on Bifrost
615 - panfrost: Fix gl_PointSize out of GL_POINTS
616 - panfrost: Mark point sprites as todo on Bifrost
617 - pan/mdg: Legalize inverts with constants
618 - pan/mdg: Ensure ld_vary_16 is aligned
619 - panfrost: Ensure we have ro before using it
620 - nir: Remove nir_intrinsic_output_u8_as_fp16_pan
621 - pan/mdg: Avoid fusing ld_vary_16 with non-zero component
622 - panfrost: Calculate varying size by format
623 - panfrost: Add panfrost_streamout_offset helper
624 - panfrost: Introduce bitfields for tracking varyings
625 - panfrost: Determine varying buffer presence
626 - panfrost: Emit unlinked varyings
627 - panfrost: Emit special varyings
628 - panfrost: Emit xfb records
629 - panfrost: Add helper to determine if we are capturing
630 - panfrost: Add high-level varying emit
631 - panfrost: Use new varying linking
632 - panfrost: Remove unused routines
633 - panfrost: Allow R/RG/RGB varyings
634 - panfrost: Only store varying formats
635 - panfrost: Use shader_info harder
636 - panfrost: Override varying format to minimal precision
637 - panfrost: Demote mediump varyings to fp16
638 - pan/mdg: Explicitly type 64-bit uniform moves
639 - pan/mdg: Analyze types for 64-bitness in RA
640 - pan/mdg: Prefer type over regmode for schedule constraints
641 - pan/mdg: Precolour blend inputs
642 - panfrost: Merge bifrost_bo/midgard_bo
643 - panfrost: Update sampler view in Bifrost path
644 - panfrost: Fix level_2
645 - panfrost: Correctly calculate tiled stride
646 - panfrost: Enable AFBC for RGB565
647 - panfrost: Simplify AFBC format check
648 - pan/mdg: Factor out unit check
649 - pan/mdg: Allow scheduling "x + x" to multipliers
650 - pan/mdg: Canonicalize (x * 2.0) to (x + x)
651 - pan/mdg: Reassociate adds for multiply-by-two
652 - nir: Propagate \*2*16 conversions into vectors
653 - panfrost: Specify stack_shift on SFBD
654 - pan/mdg: Defer nir_fuse_io_16 until after opts
655 - pan/mdg: Don't assign destination in writeout block to r1
656 - pan/mdg: Remove bundle interference code
657 - pan/mdg: Schedule writeout to VLUT
658 - pan/mdg: Defer smul, vlut until after writeout moves
659 - pan/mdg: Allow Z/S writes to use any 2nd stage unit
660 - pan/mdg: Prioritize non-moves on VADD/VLUT
661 - pan/mdg: Skip r1.w write where possible
662 - pan/mdg: Schedule based on liveness
663 - pan/mdg: Respect type/mask in mir_lower_special_reads
664 - pan/mdg: Fix indirect UBO swizzles
665 - pan/decode: Fix MSAA texture decoding
666 - pan/decode: Identify layered MSAA flag
667 - pan/mdg: Allow ignoring move mode
668 - pan/mdg: Handle GLSL_SAMPLER_DIM_MS
669 - pan/mdg: Handle nir_tex_src_ms_index
670 - pan/mdg: Handle nir_texop_txf_ms
671 - pan/mdg: Use _VTX tag for texelFetch in frag shaders
672 - panfrost: Set depth to sample_count for MSAA 2D
673 - panfrost: Identify layer_stride
674 - panfrost: Allocate space for multisampling
675 - panfrost: Index texture by sample
676 - panfrost: Include pointer for each sample
677 - panfrost: Set layer_stride for multisampled rendering
678 - panfrost: Don't advertise MSAA 2x
679 - panfrost: Identify coverage_mask
680 - panfrost: Pass sample_mask to the hardware
681 - panfrost: Implement alpha-to-coverage
682 - panfrost: Identify depth/stencil layer strides
683 - panfrost: Set depth/stencil_layer_stride accordingly
684 - panfrost: Enable MSAA if we render to such a surface
685 - panfrost: Save sample_mask before blitting
686 - panfrost: Expose MSAA 4x
687 - glsl: Handle 16-bit types in loop analysis
688 - docs/features: Track Panfrost
689 - panfrost: Introduce pan_pool struct
690 - panfrost: Allocate pool BOs against the pool
691 - panfrost: Track the device through the pool
692 - panfrost: Expose pool-based allocation API
693 - panfrost: Move debug flags into the device
694 - panfrost: Drop Gallium-local pan_bo_create wrapper
695 - panfrost: Move pool routines to common code
696 - panfrost: Factor out scoreboarding state
697 - panfrost: Pass polygon_list to tiler init function
698 - panfrost: Drop batch from scoreboard routines
699 - panfrost: Move scoreboarding routines to common
700 - panfrost: Handle PIPE_FORMAT_X24S8_UINT
701 - panfrost: Handle PIPE_FORMAT_S8_UINT
702 - panfrost: Move panfrost_translate_texture_type
703 - panfrost: Report blend shader work count
704 - panfrost: Clamp pure int pixels
705 - panfrost: Generate shader variants on framebuffer bind
706 - panfrost: Always use SOFTWARE for pure formats
707 - panfrost: Extend fetched framebuffer results
708 - panfrost: Fix fence leak
709 - panfrost: Fix write to free'd memory
710 - panfrost: Add a sparse array to map GEM handles to BOs
711 - panfrost: Index BOs from the BO map sparse array
712 - panfrost: Merge PAN_BO_IMPORTED/PAN_BO_EXPORTED
713 - panfrost: Remove PAN_BO_COHERENT_LOCAL
714 - panfrost: Remove PAN_BO_DONT_REUSE
715 - panfrost: Remove panfrost_bo_access type
716 - panfrost: Compact unused BO flag bits
717 - panfrost: Add format codes for new compressed textures
718 - panfrost: Pipe in compressed texture feature mask
719 - panfrost: Filter compressed texture formats
720 - panfrost: Map PIPE_{DXT, RGTC, BPTC} to MALI_BCn
721 - docs/features: Update ASTC entries for Panfrost
722 - pan/mdg: Bump compiler RT maximum
723 - pan/mdg: Identify per-sample interpolation mode
724 - pan/mdg: Implement gl_SampleID
725 - panfrost: Force Z/S writeback
726 - panfrost: Expose panfrost_get_blend_shader
727 - panfrost: Add MALI_PER_SAMPLE bit
728 - panfrost: Include sample count in payload estimates
729 - panfrost: Identify zs_samples field
730 - panfrost: Add rectangle subtraction algorithm
731 - panfrost: Handle per-sample shading
732 - panfrost: Set zs_samples as necessary
733 - panfrost: Track surfaces drawn per-batch
734 - panfrost: Extract panfrost_batch_reserve_framebuffer
735 - panfrost: Use Midgard-specific reloads
736 - panfrost: Call util_blitter_save_fragment_constant_buffer_slot
737 - panfrost: Overhaul tilebuffer allocations
738 - panfrost: Set PIPE_CAP_MIXED_COLORBUFFER_FORMATS
739 - panfrost: Fix sRGB clear colour packing
740 - panfrost: Implement Z32F_S8 blits
741 - panfrost: Abort on unsupported blit
742 - panfrost: Avoid integer underflow in rt_count_1
743 - panfrost: Honour cso->compare_mode
744 - panfrost: Fix faults with RASTERIZER_DISCARD
745 - panfrost: Report CAPs more honestly
746 - panfrost: Enable Chromium
747 - panfrost: Revert "Disable frame throttling"
748 - docs/features: Mark trivial missed feature
749 - panfrost: Enable FP16 by default
750 - panfrost: Avoid wait=true flushing all batches
751 - panfrost: Remove wait parameter to flush_all_batches
752 - panfrost: Skip specifying in_syncs
753 - panfrost: Allocate syncobjs in panfrost_flush
754 - panfrost: Remove unused batch_fence->signaled
755 - panfrost: Remove unused batch_fence->ctx
756 - pan/bit: Update f32->f16 convert test
757 - pan/bit: Remove BI_SHIFT stub
758 - pan/mdg: Mask spills from texture write
759 - pan/mdg: Test for SSA before chasing addresses
760 - docs/features: Add GL_EXT_multisampled_render_to_texture
761 - panfrost: Add MSAA mode selection field
762 - panfrost: Implement EXT_multisampled_render_to_texture
763 - panfrost: Set STRIDE_4BYTE_ALIGNED_ONLY
764 - panfrost: Fix WRITES_GLOBAL bit
765 - pan/mdg: Ensure barrier op is set on texture
766 - panfrost: Fix blend leak for render targets 5-8
767 - panfrost: Free cloned NIR shader
768 - panfrost: Free NIR of blit shaders
769 - panfrost: Free hash_to_temp map
770 - pan/mdg: Free previous liveness
771 - panfrost: Use memctx for sysvals
772 - panfrost: Free batch->dependencies
773 - pan/mdg: Fix discard encoding
774 - pan/mdg: Fix perspective combination
775 - pan/bit: Set d3d=true for CMP tests
779 - nir/ lower_int_to_float: Handle umax and umin
783 - .mailmap: add an alias for Iago Toral Quiroga
784 - .mailmap: add an alias for Andres Gomez
785 - gitlab-ci: update tracie README after changes in main script
786 - scripts: remove unittest.mock dependency when not used
787 - gitlab-ci: create always the "results" directory with tracie
788 - gitlab-ci: correct tracie behavior with replay errors
789 - gitlab-ci: build gfxreconstruct from the "dev" branch
790 - gitlab-ci: get the last frame from a gfxr trace using gfxrecon-info
791 - gitlab-ci/traces: updated paths and checksums for POLARIS10 traces
792 - gitlab-ci: Test AMD's Raven with traces
796 - egl/x11: Free memory allocated for reply structures on error
800 - glsl_type: don't serialize padding bytes from glsl_struct_field
801 - i965/vec4: Ignore swizzle of VGRF for use by var_range_end()
802 - glsl: fix crash on glsl macro redefinition
806 - drirc: Enable glthread for rpcs3
810 - intel/devinfo: Add is_dg1 to device info
811 - intel/l3: Add DG1 L3 configuration
812 - intel/ehl: Use GEN11_URB_MIN_MAX_ENTRIES in device info
813 - intel/ehl: Use macro GEN11_LP_FEATURES in device info
814 - intel/ehl: Rename gen_device_info struct
815 - intel/ehl: Add new PCI-IDs
817 Arcady Goldmints-Orlov (4):
819 - anv: increase minUniformBufferOffsetAlignment to 64
820 - intel/compiler: fix alignment assert in nir_emit_intrinsic
821 - nir/spirv/glsl450: increase asin(x) precision
822 - intel/compiler: Always apply sample mask on Vulkan.
826 - st/nine: Set correctly blend max_rt
827 - gallium/util: Fix leak in the live shader cache
828 - ttn: Add new allow_disk_cache parameter
829 - ttn: Implement disk cache
830 - st/nine: Enable ttn cache
831 - radeonsi: Enable tgsi to nir disk cache
832 - st/nine: Add checks for pure device
833 - st/nine: Return error when setting invalid depth buffer
834 - st/nine: Do not return invalidcall on getrenderstate
835 - st/nine: Pass more adapter formats for CheckDepthStencilMatch
836 - st/nine: Improve return error code in CheckDeviceFormat
837 - st/nine: Fix uninitialized variable in BEM()
838 - st/nine: Fix a crash if the state is not initialized
839 - st/nine: Add missing NULL checks
840 - st/nine: Increase available GPU memory
841 - st/nine: Retry allocations after freeing some space
842 - st/nine: Improve pDestRect handling
843 - st/nine: Ignore pDirtyRegion
844 - st/nine: Handle full pSourceRect better
848 - radv: Fix implicit sync with recent allocation changes.
849 - radv: Extend tiling flags to 64-bit.
850 - radv: Provide a better error for permission issues with priorities.
851 - radv: Support VK_PIPELINE_COMPILE_REQUIRED_EXT.
852 - radv: Support VK_PIPELINE_CREATE_EARLY_RETURN_ON_FAILURE_BIT_EXT.
853 - radv: Support VK_PIPELINE_CACHE_CREATE_EXTERNALLY_SYNCHRONIZED_BIT_EXT.
854 - radv: Expose VK_EXT_pipeline_creation_cache_control.
855 - radv/winsys: Finish mapping for sparse residency.
856 - radv/winsys:  Remove extra sizeof multiply.
857 - radv: Handle failing to create .cache dir.
858 - radv: Remove dead code.
859 - radv: Do not close fd -1 when NULL-winsys creation fails.
860 - radv: Implement vkGetSwapchainGrallocUsage2ANDROID.
861 - frontend/dri: Implement mapping individual planes.
862 - util/format: Add VK_FORMAT_D16_UNORM_S8_UINT.
863 - util/format: Use correct pipe format for VK_FORMAT_G8_B8_R8_3PLANE_420_UNORM.
864 - util/format: Add more multi-planar formats.
865 - gallium/dri: Remove lowered_yuv tracking for plane mapping.
866 - radeonsi: Explicitly map Z16_UNORM_S8_UINT to None for GFX10.
867 - amd/common,radeonsi: Move gfx10_format_table to common.
868 - radeonsi: Define gfx10_format in the common header.
869 - radv: Include gfx10_format_table.h only from a single source file.
870 - radv: Use common gfx10_format_table.h
871 - radv: Use ac_surface to determine fmask enable.
872 - radv: Pass no_metadata_planes info in to ac_surface.
873 - radv: Enforce the contiguous memory for DCC layers in ac_surface.
874 - radv: Rely on ac_surface for avoiding cmask for linear images.
875 - radv: Use offsets in surface struct.
876 - radv: Disable DCC in ac_surface.
877 - radv: Disable HTILE in ac_surface.
878 - radv: Allocate values/predicates at the end of the image.
879 - amd/common: Add total alignment calculation.
880 - radv: Use ac_surface to allocate aux surfaces.
881 - vulkan/wsi/x11: Ensure we create at least minImageCount images.
882 - radv/winsys: Deal with realloc failures in BO lists.
883 - radv: Handle mmap failures.
884 - radv/winsys: Distinguish device/host memory errors.
885 - radv: Make radv_alloc_shader_memory static.
886 - turnip: semaphore support.
887 - meson: Do not require shader cache for radv.
888 - amd/addrlib: fix another C++ one definition rule violation
889 - radv: Set handle types in Android semaphore/fence import.
890 - radv: Always enable PERFECT_ZPASS_COUNTS.
891 - Revert "radv: add support for MRTs compaction to avoid holes"
892 - radv: Use correct semaphore handle type for Android import.
893 - amd/llvm: Mark pointer function arguments as 32-byte aligned.
894 - amd/common: Cache intra-tile addresses for retile map.
895 - amd/addrlib: Clean up unused colorFlags argument
896 - amd/registers: add RLC_PERFMON_CLK_CNTL for pre-GFX10
897 - radeonsi: Inhibit clock-gating for perf counters.
898 - meson: Add mising git_sha1.h dependency.
899 - amd: Add detection of timeline semaphore support.
900 - radv/winsys: Add binary syncobj ABI changes for timeline semaphores.
901 - radv: Add thread for timeline syncobj submission.
902 - radv: Add winsys support for submitting timeline syncobj.
903 - radv: Add winsys functions for timeline syncobj.
904 - radv: Add timeline syncobj for timeline semaphores.
905 - radv: Fix uninitialized variable in renderpass.
906 - vulkan/wsi/x11: report device-group present rectangles with prime.
907 - vulkan/wsi: Convert usage of -1 to UINT32_MAX.
908 - radv: Fix host->host signalling with legacy timeline semaphores.
909 - mesa/st: Actually free the driver part of memory objects on destruction.
910 - radv: Don't use both DCC and CMASK for single sample images.
911 - radv: Fix assert that is too strict.
912 - radv: Do not consider layouts fast-clearable on compute queue.
913 - radv: When importing an image, redo the layout based on the metadata.
914 - radv: Use getter instead of setter to extract value.
915 - driconf: Support selection by Vulkan applicationName.
916 - radv: Override the uniform buffer offset alignment for World War Z.
917 - radv: Fix handling of attribs 16-31.
918 - radv: Remove conformance warnings with ACO.
919 - radv: Update CTS version.
920 - radv: Fix 3d blits.
921 - radv: Fix threading issue with submission refcounts.
922 - radv: Avoid deadlock on bo_list.
923 - spirv: Deal with glslang not setting NonUniform on constructors.
924 - radeonsi: Work around Wasteland 2 bug.
925 - spirv: Deal with glslang bug not setting the decoration for stores.
926 - ac/surface: Fix depth import on GFX6-GFX8.
927 - st/mesa: Deal with empty textures/buffers in semaphore wait/signal.
931 - nir: use bitfield_insert instead of bfi in nir_lower_double_ops
932 - nvir: bump max encoding size of instructions
933 - nvir: introduce OP_LOP3_LUT
934 - nvir: introduce OP_WARPSYNC
935 - nvir: introduce OP_BREV with lowering to EXTBF_REV for current GPUs
936 - nvir: introduce OP_SHF
937 - nvir: introduce OP_BMSK
938 - nvir: introduce OP_SGXT
939 - nvir: introduce OP_FINAL
940 - nvir: add constant folding for OP_PERMT
941 - nvir: run replaceZero() before replaceCvt()
942 - nvir/nir: fix fragment program output when using MRT
943 - nvir/nir: move nir options to codegen
944 - nvir/nir: flesh out options
945 - nvir/nir: turn on lower_rotate
946 - nvir/nir: implement nir_op_extract_u8
947 - nvir/nir: implement nir_op_extract_i8
948 - nvir/nir: implement nir_op_extract_u16
949 - nvir/nir: implement nir_op_extract_i16
950 - nvir/nir: implement nir_op_urol
951 - nvir/nir: implement nir_op_uror
952 - nvir/nir: nir expects the shift amount to wrap, rather than clamp
953 - nvir/nir: use nir_lower_idiv
954 - nvir/gm107: implement OP_PERMT
955 - nvir/gm107: replace SHR+AND+AND with PRMT+PRMT in PFETCH lowering
956 - nvir/gm107: separate out header for sched data calculator
957 - nvir/nir/gm107: split nir shader compiler options from gf100
958 - nvir/nir/gm107: turn on nir_lower_extract64
959 - nvir/nir/gm107: switch off lower_extract_byte
960 - nvir/nir/gm107: switch off lower_extract_word
961 - nvir/gv100: initial support
962 - nvir/gv100: enable support for tu1xx
963 - nvc0: use NVIDIA headers for GK104->GM2xx compute QMD
964 - nvc0: use NVIDIA headers for GP100- compute QMD
965 - nvc0: move setting of entrypoint for a shader stage to a function
966 - nvc0: remove hardcoded blitter vertprog
967 - nvc0: initial support for gv100
968 - nvc0: initial support for tu1xx
972 - drirc: Add picom to adaptive_sync exclusion list
976 - CI: reduce bandwidth for git pull
977 - gitlab-ci: update ci-fairy minio to latest upstream
978 - gitlab-ci: do not run full CI on scheduled pipelines
982 - radeonsi: Fix omitted flush when moving suballocated texture
986 - spirv: Split the vtn_emit_scoped_memory_barrier() logic
987 - nir: Replace the scoped_memory barrier by a scoped_barrier
988 - intel/compiler: Extract control barriers from scoped barriers
989 - spirv: Use scoped barriers for SpvOpControlBarrier
990 - nir: Add new rules to optimize NOOP pack/unpack pairs
991 - nir: Use a switch in build_deref_offset()/deref_instr_get_const_offset()
992 - nir: Allow casts in nir_deref_instr_get[_const]_offset()
993 - freedreno: Initialize lower_int64_options to a proper value
994 - nir: Stop passing an options arg to nir_lower_int64()
995 - nir: Extend nir_lower_int64() to support i2f/f2i lowering
996 - intel: Set int64_options to ~0 when lowering 64b ops
997 - nir: Get rid of __[u]int64_to_fp32() and __fp32_to_[u]int64()
998 - nir: Fix i64tof32 lowering
999 - spirv: Add a vtn_get_mem_operands() helper
1003 - radeon/vcn/enc: Re-write PPS encoding for HEVC
1004 - radeon/vcn: bump vcn3.0 encode major version to 1
1008 - turnip: Execute ir3_nir_lower_gs pass again
1009 - turnip: Fill out VkPhysicalDeviceSubgroupProperties
1010 - nir: Support sysval tess levels in SPIR-V to NIR
1011 - nir: Add an option for lowering TessLevelInner/Outer to vecs
1012 - turnip: Lower shaders for tessellation
1013 - turnip: Offset by component when lowering gl_TessLevel*
1014 - turnip: Parse tess state and support PATCH primtype
1015 - turnip: Allocate tess BOs as a function of draw size
1016 - turnip: Update VFD_CONTROL with tess system values
1017 - turnip: Emit HS/DS user consts as draw states
1018 - turnip: Support tess for draws
1019 - turnip: Force sysmem for tessellation
1020 - ir3: Unconditionally enable MERGEDREGS on a6xx
1021 - turnip: Enable tessellationShader physical device feature
1025 - intel/dev: Bail when INTEL_DEVID_OVERRIDE is not valid
1026 - intel/fs: Clean up variable group size handling in backend
1027 - intel/fs: Add an option to lower variable group size in backend
1028 - intel/fs: Add and use a new load_simd_width_intel intrinsic
1029 - intel: Let drivers call brw_nir_lower_cs_intrinsics()
1030 - iris: Implement ARB_compute_variable_group_size
1031 - util/list: Add list_foreach_entry_from_safe
1032 - nir: Use deref intrinsics to set writes_memory when gathering info
1033 - intel/fs: Use writes_memory from shader_info
1034 - nir: Consider atomic counter intrinsics when setting writes_memory
1035 - intel/fs: Remove unused emission of load_simd_with_intel
1036 - intel/fs: Remove unused state from brw_nir_lower_cs_intrinsics
1037 - intel/fs: Early return when can't satisfy explicit group size
1038 - intel/fs: Remove redundant assert()
1039 - intel/fs: Remove min_dispatch_width spilling decision from RA
1040 - intel/fs: Support INTEL_DEBUG=no8,no32 in compute shaders
1041 - intel/fs: Add helper to get prog_offset and simd_size
1042 - i965: Use new helper functions to pick SIMD variant for CS
1043 - iris: Set CS KernelStatePointer at dispatch
1044 - iris: Use new helper functions to pick SIMD variant for CS
1045 - anv: Use new helper functions to pick SIMD variant for CS
1046 - intel/fs: Generate multiple CS SIMD variants for variable group size
1047 - iris, i965: Drop max_variable_local_size
1048 - iris, i965: Update limits for ARB_compute_variable_group_size
1049 - intel: Add helper to calculate GPGPU_WALKER::RightExecutionMask
1050 - nir: Fix printing execution scope of a scoped barrier
1051 - spirv: Memory semantics is optional for OpControlBarrier
1052 - intel/fs: Add Fall-through comment
1053 - nir: Fix logic that ends combine barrier sequence
1054 - spirv: Handle most execution modes earlier
1055 - nir: Filter modes of scoped memory barrier in nir_opt_load_store_vectorize
1056 - spirv: Propagate explicit layout only in types that need it
1060 - llvmpipe: do not enable tessellation shader without llvm coroutines support
1064 - bifrost: Set RTZ rounding mode for f2i conversion
1065 - bifrost: Lower x->bool conversions to != 0
1066 - bifrost: Emit "d3d" variant of comparison instructions
1067 - bifrost: Document d3d/gl comparison control bit
1068 - bifrost: Add lowering for b2i32
1069 - bifrost: Add support for nir_op_inot
1070 - bifrost: Add support for nir_op_ishl
1071 - bifrost: Add support for nir_op_uge
1072 - bifrost: Add support for nir_op_imul
1073 - bifrost: Add support for nir_op_iabs
1074 - bifrost: Honor src swizzle in special math ops
1075 - bifrost: Fix packing of ADD_FEXP2_FAST
1079 - iris: Place a seqno at the end of every batch
1080 - iris: Convert fences to using lightweight seqno
1081 - iris: Store a seqno for each batch in the fence
1082 - iris: Initialise stub iris_seqno to 0
1083 - iris: Rename iris_seqno to iris_fine_fence
1084 - iris: Fixup copy'n'paste mistake in Makefile.sources
1088 - etnaviv: fix SAMP_ANISOTROPY register value
1089 - etnaviv: do not use int filter when anisotropic filtering is used
1090 - ci: bare-metal: make it possible to use a script for serial
1091 - ci: extend expect-output.sh
1092 - ci: add U-Boot specific fetch strings
1093 - etnaviv: drop translate_blend(..)
1094 - ci: add arm_test-base docker image
1095 - ci: use separate docker images for baremetal builds
1096 - ci: fix possible spuriously run of jobs
1097 - etnaviv: delete not used struct
1098 - etnaviv: convert enums
1099 - etnaviv: move etna_lower_io(..) to etnaviv_nir.c
1100 - etnaviv: get rid of etna_compile dependency
1101 - etnaviv: move etna_lower_alu(..) to etnaviv_nir.c
1102 - etnaviv: drop OPT_V define
1103 - etnaviv: make more use of compile_error(..)
1104 - etnaviv: move liveness related stuff into own file
1105 - etnaviv: merge struct etna_compile and etna_state
1106 - etnaviv: drop emit macro
1107 - etnaviv: move functions that generate asm to own file
1108 - etnaviv: move nir compiler related stuff into .c file
1109 - etnaviv: move ra into own file
1110 - etnaviv: replace prims-emitted query
1111 - ci: bare-metal: use nginx to get results from DUT
1112 - etnaviv: explicitly set nir_variable_mode
1113 - etnaviv: introduce struct etna_compiler
1114 - etnaviv: move shader_count to etna_compiler
1115 - etnaviv: do register setup only once
1116 - etnaviv: fix nir validation problem
1117 - etnaviv: call nir_lower_bool_to_bitsize
1118 - etnaviv: completely turn off MSAA
1122 - radv: use util_float_to_half_rtz
1123 - r600: Use TRUNC_COORD on samplers
1127 - radv: Always expose non-visible local memory type on dedicated GPUs
1131 - Linux: Change minimum priority threads from SCHED_IDLE to nice 19 SCHED_BATCH.
1135 - tu: Support pipelines without a fragment shader
1136 - tu: Add a "scratch bo" allocation mechanism
1137 - tu: Add noubwc debug flag to disable UBWC
1138 - tu: Implement fallback linear staging blit for CopyImage
1139 - freedreno/a6xx: Document dual-src blending enable bits
1140 - ir3: Fixup dual-source blending slot
1141 - tu: Move RENDER_COMPONENTS setting to pipeline state
1142 - tu: Implement dual-src blending
1143 - tu: Advertise COLOR_ATTACHMENT_BLEND_BIT for blendable formats
1144 - tu: Always initialize image_view fields for blit sources
1145 - tu: Fall back to 3d blit path for BC1_RGB_* formats
1146 - tu: Fix buffer compressed pitch calculation with unaligned sizes
1147 - tu: Support VK_FORMAT_FEATURE_BLIT_SRC_BIT for texture-only formats
1148 - tu: Fix IBO descriptor for cubes
1149 - tu: Respect VK_IMAGE_CREATE_MUTABLE_FORMAT_BIT
1150 - tu: Add missing storage image/texel buffer bits
1151 - tu: Remove useless post-binning flushes
1152 - tu: Don't actually track seqno's for events
1153 - tu: Remove useless event_write helpers
1154 - tu: Rewrite flushing to use barriers
1155 - tu: Fix context faults loading unused descriptor sets
1156 - ir3: Pass reserved_user_consts to ir3_shader_from_nir()
1157 - tu: Remove num_samp hack
1158 - tu: Use the ir3 shader API
1159 - tu: Remove tu_shader_compile_options
1160 - tu: Set num_components to 0 when building bindless intrinsics
1161 - ir3: Don't calculate num_samp ourselves
1162 - tu: Actually remove dead variables after io lowering
1163 - ir3: Split out variant-specific lowering and optimizations
1164 - ir3, freedreno: Round up constlen earlier
1165 - ir3: Include ir3_compiler from ir3_shader
1166 - ir3: Support variants with different constlen's
1167 - ir3: Add ir3_trim_constlen()
1168 - tu: Share constlen between different stages properly
1169 - freedreno: Refactor ir3_cache shader compilation
1170 - freedreno: Share constlen between different stages properly
1171 - freedreno: On a5xx+ INDX_SIZE is MAX_INDICES
1172 - freedreno/registers: Label firstIndex field in CP_DRAW_INDX_OFFSET
1173 - tu: Pass firstIndex directly to CP_DRAW_INDX_OFFSET
1174 - freedreno/a6xx: use firstIndex field
1175 - nir: Refactor load/store intrinsic helper
1176 - nir: add vec2_index_32bit_offset address format
1177 - tu: Rewrite variable lowering
1178 - tu: Enable KHR_variable_pointers
1179 - ir3: Add layer_zero variant bit
1180 - tu: Force gl_Layer to 0 when necessary
1181 - freedreno/a6xx: Force gl_Layer to 0 when necessary
1182 - freedreno: Include adreno_pm4.xml.h before adreno_a6xx.xml.h
1183 - freedreno: Sync registers with envytools
1184 - freedreno/a6xx: Rename and document HLSQ_UPDATE_CNTL
1185 - freedreno/a6xx: Add some documentation for shared consts
1186 - tu: Don't invalidate irrelevant state when changing pipeline
1187 - freedreno/a6xx: Add stencilref register info
1188 - ir3: Handle gl_FragStencilRefARB
1189 - tu: Enable VK_EXT_shader_stencil_export
1190 - freedreno: Add a helper for computing guardband sizes
1191 - tu: Use common guardband helper
1192 - freedreno: Use common guardband helper
1193 - freedreno/ir3: Fix SSBO size for bindless SSBO's
1194 - tu: Enable VK_EXT_depth_clip_enable
1195 - freedreno: Clean up CP_DRAW_MULTI_INDIRECT definition
1196 - freedreno: Add INDIRECT_COUNT CP_DRAW_INDIRECT_MULTI variants
1197 - tu: Integrate WFI/WAIT_FOR_ME/WAIT_MEM_WRITES with cache tracking
1198 - tu: Add missing wfi to tu6_emit_hw()
1199 - tu: Implement VK_KHR_draw_indirect_count
1200 - tu: Fix empty blit scissor case
1201 - tu: Fix hangs for DS with no output
1202 - tu: Detect invalid-for-binning renderpass dependencies
1203 - tu: Enable vertex & fragment stores & atomics
1204 - tu: Fix descriptor update templates with input attachments
1205 - ir3: Validate bindless samp_tex correctly
1206 - ir3: Remove redundant samp_tex validation
1207 - ir3: Fix incorrect src flags for samp_tex
1208 - tu: Enable resource dynamic indexing
1209 - freedreno/rnn: Return success when parsing addvariant
1210 - tu: Dump CP_DRAW_INDIRECT_MULTI draw BO's
1211 - freedreno/rnn: Support stripes in rnndec_decodereg
1212 - freedreno/cffdec: Handle CP_DRAW_INDIRECT_MULTI like other draws
1213 - freedreno: Add trace for CP_DRAW_INDIRECT_MULTI
1214 - freedreno/a6xx: Fix CP_BIN_SIZE_ADDRESS name
1215 - freedreno/rnn: Make rnn_decode_enum() respect variants
1216 - freedreno/cffdec: Stop open-coding enum parsing
1217 - freedreno/afuc: Add missing rnn_prepdb()
1218 - freedreno/afuc: Fix PM4 enum parsing
1219 - tu: Fix DST_INCOHERENT_FLUSH copy/paste error
1220 - freedreno: Document draw predication packets
1221 - tu: Reset has_tess after renderpass
1222 - tu: Implement VK_EXT_conditional_rendering
1226 - intel/fs: Update location of Render Target Array Index for gen12
1227 - anv,iris: Fix input vertex max for tcs on gen12
1228 - intel/dump_gpu: Fix name of LD_PRELOAD in env append logic
1229 - anv/gen11+: Disable object level preemption
1233 - aco: either copy-propagate or inline create_vector operands
1234 - aco: coalesce parallelcopies during register allocation
1235 - nir: add nir_intrinsic_elect to divergence analysis
1236 - nir: refactor divergence analysis state
1237 - nir: rework phi handling in divergence analysis
1238 - nir: simplify phi handling in divergence analysis
1239 - nir: reset ssa-defs as non-divergent during divergence analysis instead of upfront
1240 - aco: fix WQM coalescing
1241 - aco: restrict copying of create_vector operands to GFX9+
1242 - aco: don't move create_vector subdword operands to unsupported register offsets
1243 - aco: fix corner case in register allocation
1244 - aco: don't allow unaligned subdword accesses on GFX6/7
1245 - aco: fix register assignment for p_create_vector on GFX6/7
1246 - aco: simplify statistics collection for copies
1247 - aco: use full-register instructions to implement subdword packing on GFX6/7
1248 - aco: Workarounds subdword lowering on GFX6/7
1249 - aco: adjust GFX6 subdword lowering workarounds for 8bit
1250 - aco: add and use scratch SGPR to lower subdword p_create_vector on GFX6/7
1251 - aco: coalesce copies more aggressively when lowering to hw
1252 - aco: skip partial copies on first iteration when lowering to hw
1253 - aco: optimize packing of 16bit subdword registers on GFX6/7
1254 - aco: remove unnecessary split- and create_vector instructions for subdword loads
1255 - aco: fix shared subdword loads
1256 - aco: reorder calls to aco_validate() and cleanup aco_compile_shader()
1257 - aco: don't allow SGPRs on logical phis
1258 - aco: fix WQM handling in nested loops
1259 - radv/aco: implement logic64 instead of lowering
1260 - aco: align swap operations to 4 bytes on GFX6/7
1261 - aco: don't allow partial copies on GFX6/7
1262 - radv: introduce RADV_DEBUG=llvm option
1263 - radv: change use_aco -> use_llvm
1264 - radv: enable ACO by default
1265 - aco: fix partial copies on GFX6/7
1266 - aco: remove superflous (bool & exec) if the result comes from VOPC
1267 - nir: also move vecN in case of nir_move_copies
1268 - nir: refactor nir_can_move_instr
1269 - nir/algebraic: optimize bcsel(a, 0, 1) to b2i
1270 - nir: also move b2i in case of nir_move_copies
1271 - nir/algebraic: optimize iand/ior of (n)eq zero
1272 - nir/algebraic: add optimizations for fsign/isign
1273 - nir/algebraic: add some more unop + bcsel optimizations
1274 - nir/algebraic: optimize fmul(x, bcsel(c, -1.0, 1.0)) -> bcsel(c, -x, x)
1275 - nir/algebraic: optimize (a < 0.0) ? -a : a -> fabs(a)
1276 - nir/algebraic: add distributive rules for ior/iand
1277 - nir/algebraic: propagate b2i out of ior/iand
1278 - nir/algebraic: fold some nested bcsel
1279 - aco: fix scratch loads which cross element_size boundaries
1280 - aco: ensure to not extract more components than have been fetched
1281 - aco: don't split store data if it was already split into more elements
1282 - aco: prevent infinite recursion in RA for subdword variables
1283 - aco: ensure readfirstlane subdword operands are always dword aligned
1284 - radv: call radv_nir_lower_ycbcr_textures after first optimizations
1285 - aco: add GFX6/7 subdword lowering tests
1286 - aco: execute branch instructions in WQM if necessary
1290 - CI: Disable Panfrost T7x0 jobs
1291 - CI: Re-enable Panfrost T7x0 jobs
1292 - llvmpipe: Expect increased exp precision on Windows
1293 - CI: Windows: Build LLVM and llvmpipe
1294 - CI: Disable Panfrost T720/T760
1295 - Revert "CI: Disable Panfrost T720/T760"
1296 - CI: Enable assertions on Windows
1297 - CI: Try shared libraries on Windows
1298 - CI: Correct build-directory path on Windows, and keep it
1299 - CI: Re-enable the Windows VS2019 build job
1300 - CI: Temporarily disable Panfrost T860 jobs
1301 - CI: Re-enable Panfrost T860 jobs
1302 - CI: Disable Windows build due to unstable infrastructure
1306 - glsl: rename has_implicit_uint_to_int_conversion to *_int_to_uint_*
1307 - i965: Fix out-of-bounds access to brw_stage_state::surf_offset
1308 - anv: Translate relative timeout to absolute when calling anv_timelines_wait
1309 - anv: Fix deadlock in anv_timelines_wait
1310 - meson: Disable GCC's dead store elimination for memory zeroing custom new
1311 - mesa: Fix double-lock of Shared->FrameBuffers and usage of wrong mutex
1312 - st/mesa: Clear texture's views when texture is removed from Shared->TexObjects
1313 - intel/fs: Work around dual-source blending hangs in combination with SIMD16
1314 - glsl: Don't replace lrp pattern with lrp if arguments are not floats
1315 - glsl: inline functions with unsupported return type before converting to nir
1316 - i965: Work around incorrect usage of glDrawRangeElements in UE4
1317 - st/mesa: account for "loose", per-mipmap level textures in CopyImageSubData
1318 - iris: Honor scanout requirement from DRI
1319 - iris: Fix fast-clearing of depth via glClearTex(Sub)Image
1320 - nir/opt_if: Fix opt_if_simplification when else branch has jump
1321 - nir/tests: Add tests for opt_if_simplification
1322 - st/mesa: Treat vertex outputs absent in outputMapping as zero in mesa_to_tgsi
1323 - anv/nir: Unify inputs_read/outputs_written between geometry stages
1324 - spirv: Only require bare types to match when copying variables
1325 - glsl: Eliminate out-of-bounds triop_vector_insert
1326 - intel/compiler: Fix pointer arithmetic when reading shader assembly
1327 - glsl: Eliminate assigments to out-of-bounds elements of vector
1328 - nir/lower_io: Eliminate oob writes and return zero for oob reads
1329 - nir/large_constants: Eliminate out-of-bounds writes to large constants
1330 - nir/lower_samplers: Clamp out-of-bounds access to array of samplers
1334 - prevent multiply defined symbols
1338 - i965: add support for gen 5 pipelined pointers to dump
1339 - i965: disable shadow batches when batch debugging.
1340 - draw/tess: free tessellation control shader i/o memory.
1341 - llvmpipo/nir: free compute shader NIR
1342 - llvmpipe: simple texture barrier implementation.
1343 - gallivm/sample: add multisample support for texel fetch
1344 - gallivm/sample: add multisample image operation support
1345 - gallivm/nir/tgsi: add multisample texture sampling.
1346 - gallivm/nir: add multisample support to image size
1347 - gallivm/nir: add multisample image operations
1348 - draw: introduce sampler num samples + stride members
1349 - draw: add support for num_samples + sample_stride to the image paths
1350 - llvmpipe: add num_samples/sample_stride support to jit textures
1351 - llvmpipe: add samples support to image jit
1352 - util: add a resource wrapper to get resource samples
1353 - llvmpipe: add multisample support to texture allocator.
1354 - llvmpipe: add a max samples define set to 4.
1355 - gallium/util: split out zstencil clearing code.
1356 - llvmpipe: fix race between draw and setting fragment shader.
1357 - llvmpipe: add get_sample_position support (v2)
1358 - llvmpipe/jit: pass fragment sample mask via jit context.
1359 - llvmpipe: pass incoming sample_mask into fragment shader context.
1360 - llvmpipe: add internal multisample texture mapping path.
1361 - llvmpipe: add multisample resource copy region support.
1362 - llvmpipe: add clear texture support for multisample textures.
1363 - llvmpipe: handle multisample render target clears
1364 - draw: disable point/line smoothing for multisample (v2)
1365 - llvmpipe: pass color and depth sample strides into fragment shader.
1366 - llvmpipe: record sample info for color/depth buffers in scene
1367 - llvmpipe/rast: fix tile clearing for multisample color and depth tiles
1368 - llvmpipe: plumb multisample state bit into setup code.
1369 - llvmpipe: add multisample bit to fragment shader key.
1370 - llvmpipe: change mask input to fragment shader to 64-bit.
1371 - llvmpipe: add cbuf/zsbuf + coverage samples to the fragment shader key.
1372 - gallivm: add sample id/pos intrinsic support
1373 - gallivm: add mask api to force mask
1374 - nir/tgsi: translate the interp location
1375 - llvmpipe: pass interp location into interpolation code.
1376 - llvmpipe: add centroid interpolation support.
1377 - llvmpipe: add per-sample interpolation.
1378 - llvmpipe: move getting mask value out of depth code. (v2)
1379 - llvmpipe: add per-sample depth/stencil test
1380 - llvmpipe: move some fs code around
1381 - llvmpipe: multisample sample mask + early/late depth pass
1382 - llvmpipe: handle multisample early depth test/late depth write
1383 - llvmpipe: interpolate Z at sample points for early depth test.
1384 - llvmpipe: handle multisample color stores.
1385 - llvmpipe: hook up sample position system value
1386 - llvmpipe: add multisample alpha to coverage support.
1387 - llvmpipe: add multisample alpha to one support
1388 - llvmpipe: handle gl_SampleMask writing.
1389 - llvmpipe: don't allow branch to end for early Z with multisample
1390 - llvmpipe: pass mask store into interp for centroid interpolation
1391 - llvmpipe: move color storing earlier in frag shader
1392 - llvmpipe: fix multisample occlusion queries.
1393 - llvmpipe: disable opaque variant for multisample
1394 - llvmpipe: add new rast api to pass full 64-bit mask.
1395 - llvmpipe: add fixed point sample positions to scene.
1396 - llvmpipe: build 64-bit coverage mask in rasterizer
1397 - llvmpipe: fixup multisample coverage masks for covered tiles
1398 - llvmpipe: generate multisample triangle rasterizer functions (v2)
1399 - llvmpipe: choose multisample rasterizer functions per triangle (v2)
1400 - llvmpipe: choose correct position for multisample
1401 - llvmpipe: don't choose pixel centers for multisample
1402 - drisw: add multisample support to sw dri layer.
1403 - llvmpipe: enable 4x sample MSAA + texture multisample
1404 - gallivm/sample: add num samples query for txqs (v2)
1405 - gallivm/nir: hooks up texture samples queries
1406 - llvmpipe: enable GL_ARB_shader_texture_image_samples
1407 - llvmpipe: add min samples support to the fragment shader.
1408 - llvmpipe: enable ARB_sample_shading
1409 - llvmpipe: make sample position a global array.
1410 - zink: enable conditional rendering if available
1411 - r600: enable TEXCOORD semantic for TGSI.
1412 - r600/sfn: plumb the chip class into the instruction emission
1413 - r600/sfn: fix cayman float instruction emission.
1414 - r600/sfn: cayman fix int trans op2
1415 - r600/sfn: add callstack non-evergreen support
1416 - r600/sfn: add emit if start cayman support
1417 - llvmpipe: don't use sample mask with 0 samples
1418 - llvmpipe: use per-sample position not sample id for interp
1419 - llvmpipe/interp: fix interpolating frag pos for sample shading
1420 - llvmpipe: remove non-simple interpolation paths.
1421 - gallivm/nir: add an interpolation interface.
1422 - llvmpipe/interp: refactor out use of pixel center offset
1423 - llvmpipe/interp: refactor out centroid calculations
1424 - llvmpipe: add interp instruction support
1425 - llvmpipe/fs: hook up the interpolation APIs.
1426 - gallivm/nir: add sample_mask_in support
1427 - llvmpipe: add gl_SampleMaskIn support.
1428 - r600/sfn: fix nop channel assignment.
1429 - llvmpipe: compute shaders work better with all the threads.
1430 - llvmpipe: move coroutines out of noopt case
1431 - ci: bump virglrenderer to latest version
1432 - util/disk_cache: add fallback for disk_cache_get_function_identifier
1433 - llvmpipe/cs: overhaul cs variant key state.
1434 - llvmpipe/draw: drop variant number from function names.
1435 - gallivm: rework coroutine malloc/free callouts.
1436 - gallivm: rework debug printf hook to use global mapping.
1437 - gallivm: add support for a cache object
1438 - gallivm: skip operations if we have a cached object.
1439 - gallivm: add cache interface to mcjit
1440 - llvmpipe: add infrastructure for disk cache support
1441 - gallivm: don't cache shaders that use fetch functions.
1442 - llvmpipe/fs: add caching support
1443 - llvmpipe/cs: add shader caching
1444 - draw: add disk cache callbacks for draw shaders
1445 - llvmpipe: hook draw disk cache up
1446 - draw: add disk caching for draw shaders
1447 - draw/gs: fix emitting inactive primitives crash
1448 - draw/gs: add more info to debugging.
1449 - gallivm/nir: add group barrier support
1450 - llvmpipe: fix subpixel bits reporting.
1451 - gallivm/format: convert unsigned values to float properly.
1452 - gallivm/conv: enable conversion min code. (v2)
1453 - gallivm/sample: fix texel type for stencil 8-bit
1454 - llvmpipe/setup: add planes for draw regions if no scissor.
1455 - gallivm/cache: don't require a null terminator for cache data.
1456 - mesa/gles3: add support for GL_EXT_shader_group_vote
1457 - virgl: change vendor id to reflect reality more.
1458 - llvmpipe: change vendor to be more generic.
1459 - softpipe: change vendor name to something more generic.
1460 - gallivm/nir: fix const loading on big endian systems
1461 - glsl: fix constant packing for 64-bit big endian.
1462 - gallivm/nir: fix big-endian 64-bit splitting/merging.
1463 - llvmpipe: fix occlusion queries on big-endian.
1464 - mesa/get: fix enum16 big-endian getting.
1465 - draw/llvm: fix big-endian mask adjusting
1466 - draw: pass nr_samplers into llvm sample state creation.
1467 - llvmpipe: pass number of samplers into llvm sampler code.
1468 - gallivm/sample: change texture function generator api
1469 - gallivm: add indirect texture switch statement builder.
1470 - draw: add support for indirect texture access
1471 - llvmpipe: add support for indirect texture access.
1472 - gallivm/nir: add texture unit indexing
1473 - gallivm/nir: handle non-uniform texture offsets
1474 - gallivm/sample: pass indirect offset into texture/image units
1475 - llvmpipe/draw: wire up indirect offset
1476 - gallivm/sample: handle size unit offset
1477 - llvmpipe: enable ARB_gpu_shader5
1478 - draw: pass number of images to image soa create
1479 - llvmpipe: pass number of images into image soa create
1480 - gallivm/nir: support passing image index into image code.
1481 - gallivm/nir: refactor image operations for indirect support.
1482 - gallivm/img: refactor out the texel return type (v2)
1483 - gallivm/nir: add support for indirect image loading
1484 - draw/sample: add support for indirect images
1485 - llvmpipe: handle indirect images properly
1486 - ci: fixup tests after all indirect images fixes.
1487 - docs: update llvmpipe GL 4.0 status
1488 - draw/clip: cleanup viewport index handling code.
1489 - draw/clip: fix viewport index for geometry shaders
1490 - mesa/version: only enable GL4.1 with correct limits.
1491 - llvmpipe: bump texture/scene limits to enable GL 4.1
1492 - llvmpipe: bump to GL support to GL 4.1
1493 - llvmpipe: enable GL 4.2
1494 - gallivm/nir: call end prim at end on all GS streams.
1495 - draw: emit so primitives before ending empty pipeline.
1496 - draw/gs: fix up current verts in output fetching.
1497 - gallivm/draw/gs: pass vertex stream count into shader build
1498 - draw/gs: only allocate memory for streams needed.
1499 - gallivm/gs_iface: pass stream into end primitive interface.
1500 - gallivm/nir: don't access stream var outside bounds
1501 - gallivm/nir: end primitive for all streams.
1502 - draw: account primitive lengths for all streams.
1503 - draw/gs: reverse the polarity of the invocation/prims execution
1504 - draw: use common exit path in pipeline finish.
1505 - draw: free vertex info from geometry streams.
1506 - draw/gs: use mask to limit vertex emission.
1507 - ci/virgl: update results after streams fixes.
1508 - llvmpipe: add ARB_post_depth_coverage support.
1509 - llvmpipe: denote NEW fs when images change.
1510 - llvmpipe: flush resources on sampler view binding
1511 - llvmpipe/cs: fix image/sampler binding for compute
1512 - nouveau: avoid LTO ODR warning (v2)
1513 - gallivm/sample: always square rho before fast log2
1514 - llvmpipe/format: fix snorm conversion
1515 - mesa: change dsa texture error codes for GL 4.6
1516 - ci: bump piglit checkout for dsa tests
1517 - llvmpipe: fix stencil only formats.
1518 - llvmpipe: fix position offset interpolation
1519 - llvmpipe/cs: respect render condition
1520 - llvmpipe: add framebuffer fetching support (v1.1)
1521 - ci/llvmpipe: reenable gpu shader5 tests
1522 - llvmpipe: enable EXT_texture_shadow_lod
1523 - llvmpipe/draw: handle constant buffer limits and robustness (v1.1)
1524 - drisw: add robustness extension support.
1525 - glx/drisw: add robustness support
1526 - llvmpipe: add device reset query context hook.
1527 - llvmpipe: enable robust buffer access + GL 4.3, GLES 3.2 and robust buffer access behaviour
1528 - llvmpipe/ms: fix sign extension bug in rasterizer.
1529 - Revert "llvmpipe: Use the default behavior of ALLOW_MAPPED_BUFFERS."
1530 - radv: cleanup locking around timeline waiting.
1531 - llvmpipe: only read 0 for channels being read
1532 - llvmpipe/blit: for 32-bit unorm depth blits just copy 32-bit
1533 - llvmpipe: enable GL 4.5
1534 - llvmpipe/cs: update compute counters not fragment shader.
1535 - llvmpipe: include gallivm perf flags in shader cache.
1536 - gallivm: disable brilinear for lod bias and explicit lod.
1540 - radv: link with ld_args_build_id
1544 - nir: Add colorspace support to YUV lowering pass
1545 - i965/i915: Add colorspace support to YUV sampling
1549 - gitlab: Ask about reproduction rate in the issue template
1553 - mesa: check draw buffer completeness on glClearBufferfv/glClearBufferuiv
1554 - nir: replace fnv1a hash function with xxhash
1555 - freedreno: replace fnv1a hash function with xxhash
1556 - i965: replace fnv1a hash function with xxhash
1557 - util/hash_table: replace fnv1a hash function with xxhash
1558 - r600: replace fnv1a hash function with xxhash
1559 - zink: replace fnv1a hash function with xxhash
1560 - util: delete fnv1a hash function
1564 - zink. Changed sampler default name.
1568 - docs: Add release notes for 20.0.6
1569 - docs: Add SHA256 sums for 20.0.6
1570 - docs: update calendar, add news item, and link releases notes for 20.0.6
1571 - docs: Add release notes for 20.0.7
1572 - docs/relnotes Add sha256 sums to 20.0.7
1573 - docs: update calendar, add news item, and link releases notes for 20.0.7
1574 - tests: Make tests aware of meson test wrapper
1575 - meson: Bump required version to 0.52.0
1576 - meson: Use the check_header function
1577 - meson: Use build_always_stale instead of build_always
1578 - meson: Use builtins for checking gnu __attributes__
1579 - drm-shim/meson: The name of the target is a string not a list
1580 - drm-shim/meson: Use portable override_options for setting C standard
1581 - meson: use gnu_symbol_visibility argument
1582 - meson: use 2 space not 3 space indent
1583 - meson: deprecated 'true' and 'false' in combo options for 'enabled' and 'disabled'
1584 - vulkan-overlay/meson: use install_data instead of configure_file
1585 - docs: Add release notes for 20.0.8
1586 - docs: Add sha256sums for 20.0.8
1587 - docs: update calendar, add news item, and link releases notes for 20.0.8
1588 - mesa/swrast: use logf2 instead of util_fast_log2
1589 - VERSION: bump for 20.2.0-rc1
1590 - .pick_status.json: Update to 9333a8570d2174b73da63c3ee6f1a740ae487ab8
1591 - .pick_status.json: Update to 1e28745bc0d3528c1dfc25459456849feb58d407
1592 - meson/freedreno: Fix lua requirement
1593 - .pick_status.json: Update to fdb97d3d2914c8f887a7968432db4fdbd35d8376
1594 - bump version for 20.2.0-rc2
1595 - .pick_status.json: Update to 61042b1bdb199f98dd34085ed29a8c492ed9b2a3
1596 - .pick_status.json: Update to 6d28270968e0728bf8bdf48a6abd261c50d9ef07
1597 - .pick_status.json: Update to ca7d66e847d08914cec0a5e003b400da9c0a2695
1598 - VERSION: bump for 20.2.0-rc3
1599 - .pick_status.json: Update to 7fbded8b5821a47c26245b181446f972f920a96e
1600 - .pick_status.json: Mark e93979ba599355c42df01a89073362b970489a3a as denominated
1601 - .pick_status.json: Update to b9927c8c8d0c105699306a68773c015930ff9509
1602 - VERSION: bump for 20.2.0-rc4
1603 - .pick_status.json: Update to ef980ac0c1cd65993ba0c1d20e1c09b45bfef99d
1604 - fix: gallivm: disable brilenear for lod bias and explicit lod.
1605 - .pick_status.json: Update to a1f46d7b6943699e5efb60fbcfdd1450db85adb1
1606 - amd/ac_surface: convert tabs to 3 spaces
1607 - .pick_status.json: Update to 90b98c06493f8a9759e5496d5ec91fb60edf7b92
1608 - .pick_status.json: Update to 472a20c5fc0feda0f074b4ff95fd7c7a6305c8cd
1612 - freedreno: Centralize UUID generation into new files freedreno_uuid.c/h
1613 - freedreno/uuid: Generate meaningful device and driver UUID
1617 - virgl: implement ARB_clear_texture
1618 - virgl: Enable CAP_CLEAR_TEXTURE if host supports it
1619 - docs/features: Add ARB_clear_texture to virgl
1620 - gallium: add TGSI_PROPERTY_FS_BLEND_EQUATION_ADVANCED
1621 - glsl_to_tgsi: Set TGSI_PROPERTY_FS_BLEND_EQUATION_ADVANCED
1622 - virgl: Reserved last caps of capability_bits
1623 - gallium: Add PIPE_CAP_BLEND_EQUATION_ADVANCED
1624 - st: expose KHR_blend_equation_advanced if PIPE_CAP_BLEND_EQUATION_ADVANCED
1625 - glsl_to_ir: do lower_blend_equation if PIPE_CAP_FBFETCH
1626 - virgl: Use alpha_src_factor to store blend_equation_advenced value
1627 - virgl: Encode barrier for blend_equation_advanced
1628 - virgl: set PIPE_CAP_BLEND_EQUATION_ADVANCED
1632 - meson: Do not enable USE_ELF_TLS for FreeBSD
1633 - iris: Explicitly cast value to uint64_t
1634 - i965: Explicitly cast value to uint64_t
1638 - util/rand_xor: use getrandom() when available
1639 - Expose EGL_KHR_platform_* when EXT is supported
1643 - meson: Add versioning for xvmc tracker
1647 - freedreno/ir3: Initialize the unused dwords of the immediates consts.
1648 - freedreno/ir3: Drop redundant IR3_REG_HALF setup in ALU ops.
1649 - freedreno/ir3: Leave bools as 1-bit, storing them in full regs.
1650 - freedreno/ir3: Set up the block predecessors for a3xx TF
1651 - freedreno/ir3: Fix the a3xx TF outputs stores.
1652 - freedreno/ir3: Fix register allocation assertion failures.
1653 - freedreno: Stop doing binning shaders other than the VS in shader-db.
1654 - freedreno/ir3: Skip tess epilogue if the program is missing stores.
1655 - freedreno: Fix assertion failures on GS/tess shaders with shader-db enabled.
1656 - freedreno/ir3: Remove unused half precision shader key flag.
1657 - freedreno: Emit debug messages when doing draw-time recompiles of shaders.
1658 - freedreno/ir3: Improve shader key normalization.
1659 - freedreno/ir3: Stop initializing regid of so->outputs during setup.
1660 - freedreno/ir3: Set up outputs for multi-slot varyings.
1661 - freedreno: Immediately compile a default variant of shaders.
1662 - freedreno/ir3: Set the FS .msaa flag to true during precompiles.
1663 - freedreno/ir3: Add some more tests of cat6 disasm.
1664 - freedreno/ir3: Sync some new changes from envytools.
1665 - freedreno/ir3: Define the bindful uniform/nonuniform desc modes for cat6 a6xx.
1666 - freedreno/ir3: Disable sin/cos range reduction for mediump.
1667 - ci: Clean up setup of the job-specific env vars in baremetal testing.
1668 - ci: Enable IRC flake reporting on freedreno baremetal boards.
1669 - ci: Improve the flakes reports on IRC.
1670 - ci: Fix the nick used in IRC reporting.
1671 - freedreno: Deduplicate ringbuffer macros with computerator/fdperf
1672 - freedreno: Clean up tests around ORing in the reloc flags.
1673 - freedreno: Rename append_bo() in case it doesn't get inlined.
1674 - freedreno: Initialize the bo's iova at creation time.
1675 - freedreno: Start moving relocs flags into the BOs.
1676 - freedreno: Replace OUT_RELOCD with permanently flagging shader BOs for it.
1677 - freedreno: Mark all ringbuffer BOs as to be dumped on crash.
1678 - freedreno: Tell the kernel that all BOs are for writing.
1679 - freedreno: Replace OUT_RELOCW with OUT_RELOC.
1680 - freedreno: Drop the "write" arg to emit_const_bo now relocs don't care.
1681 - nir: Fix count when we didn't lower load_uniforms but did shift load_ubos.
1682 - freedreno: Fix non-constbuf-upload UBO block indices and count.
1683 - freedreno: Add a nohw flag to skip submitting to the kernel.
1684 - freedreno: Split the fd_batch_resource_used by read vs write.
1685 - freedreno: Add an early out for preparing to read a resource.
1686 - freedreno: Move the resource_read early out to an inline.
1687 - freedreno: Skip taking the lock for resource usage if it's already flagged.
1688 - freedreno/a4xx+: Increase max texture size to 16384.
1689 - freedreno/a6xx: Improve layout testcase logging for UBWC fails.
1690 - freedreno/a6xx: Add a testcase for UBWC buffer sharing.
1691 - freedreno: Pull the tile_alignment lookup for a layout to a helper.
1692 - freedreno/a6xx: Fix UBWC blockheight for RG8.
1693 - freedreno/a6xx: Fix UBWC mipmap sizing.
1694 - freedreno/a6xx: Fix UBWC mipmapping height alignment.
1695 - nir: Include num_ubos in the printed shader (if nonzero).
1696 - freedreno/ir3: Clean up a silly nir_src_for_ssa(src.ssa).
1697 - freedreno/ir3: Leave the cursor alone during ir3_nir_try_propagate_bit_shift.
1698 - freedreno/ir3: Move i/o offset lowering after analyze_ubo_ranges.
1699 - freedreno: Trim num_ubos to just the ones we haven't lowered to constbuf.
1700 - freedreno/a6xx: Use LDC for UBO loads.
1701 - freedreno: Drop the noubo fails list for CI, since there aren't any now.
1702 - freedreno: Fix attempts to push UBO contents past the constlen on pre-a6xx.
1703 - freedreno: Fix resource layout dump loop.
1704 - freedreno: Avoid duplicate BO relocs in FD_RINGBUFFER_OBJECTs.
1705 - ci: Move cross file generation to a shared script.
1706 - ci: Autodetect whether we need cross setup in lava_arm builds.
1707 - ci: Make cmake toolchain file for deqp cross build setup.
1708 - ci: Make the create-rootfs more resilient.
1709 - ci: Update versions of packages to remove from rootfses.
1710 - ci: Switch the baremetal runner to be an x86 docker image.
1711 - ci: Disable SMP on the a5xx boards.
1712 - ci: Make a530's GLES3/31 fractional runs much more complete.
1713 - freedreno/a5xx: Move resource layout to fdl.
1714 - freedreno/fdl: Separate the list of a6xx testcases from the the test code.
1715 - freedreno/a5xx: Add the outline of a unit test for a5xx layout.
1716 - freedreno/a5xx: Set MIN_LAYERSZ on 3D textures like we do on a6xx.
1717 - freedreno/a5xx: Define the 2D blit UBWC pitch fields
1718 - ci: Fix DEQP_CASELIST_FILTER (used by a630 noubo run)
1719 - ci: Do an explicit NIR validation-enabled pass on freedreno a630.
1720 - ci: Don't forget to set NIR_VALIDATE in baremetal runs.
1721 - ci: Enable a fractional run with UBO-to-constbuf disabled on a3xx.
1722 - ci: Improve baremetal's logging of the job env var passthrough.
1723 - freedreno/a6xx: Fix the size of buffer image views.
1724 - freedreno: Fix printing of unused src in disasm of cat6 RESINFO.
1725 - freedreno: Add more resinfo/ldgb testcases.
1726 - freedreno: Fix resinfo asm, which doesn't have srcs besides IBO number.
1727 - freedreno: Set the immediate flag in a4/a5xx resinfos.
1728 - freedreno/ir3: Refactor out IBO source references.
1729 - freedreno/ir3: Move handle_bindless_cat6 to compiler_nir and reuse.
1730 - freedreno/ir3: Use RESINFO for a6xx image size queries.
1731 - ci: Drop double ".txt" suffix on the unexpected results file.
1732 - ci: Drop old comment about enabling --deqp-watchdog.
1733 - ci: Auto-detect the architecture for VK ICD filenames.
1734 - ci: Add DEQP_EXPECTED_RENDERER support for VK tests.
1735 - ci: Move baremetal DEQP_NO_SAVE_RESULTS setup to the yml.
1736 - ci: Quick exit qpa extraction for non-matching qpas.
1737 - ci: Disable the firmware loader user helper option in arm64 kernels.
1738 - ci: Build a cheza kernel.
1739 - ci: Add scripts for controlling bare-metal chezas.
1740 - ci: Switch cheza (freedreno a630) testing to baremetal.
1741 - ci: Don't build an arm_test container now that the last user is gone.
1742 - ci: Rename x86_cross_arm_test to just arm_test.
1743 - turnip: Move vertex buffer bindings to SET_DRAW_STATE.
1744 - turnip: Don't bother clamping VB size.
1745 - turnip: Simplify vertex buffer bindings.
1746 - turnip: Use tu_cs_emit_regs() for BLEND_CONTROL.
1747 - turnip: Add support for alphaToOne.
1748 - freedreno/a6xx: Add support for ALPHA_TO_ONE.
1749 - freedreno: Upload gallium constbufs as needed when referenced as a UBO.
1750 - freedreno/ir3: Refactor ir3_cp's lower_immed().
1751 - freedreno/ir3: Stop pushing immediates once we've filled the constbuf.
1752 - freedreno/ir3: Drop unnecessary alignment of pushed UBO size.
1753 - freedreno/ir3: Stop shifting UBO 1 down to be UBO 0.
1754 - freedreno/ir3: Account for driver params in UBO max const upload.
1755 - freedreno/ir3: Drop the max_const on a6xx to 512.
1756 - freedreno/ir3: Handle cases where we decide not to lower UBO 0 loads.
1757 - turnip: Fix crashes in compute with no descriptors to load.
1758 - ci: Bump up to the current version of the VK CTS.
1759 - ci: Disable shader cache on vulkan CI runs.
1760 - ci: Build the full VK CTS for baremetal testing.
1761 - ci: Enable pre-merge fractional vulkan CTS runs on the turnip driver.
1762 - ci: Use rsync for initial nfsroot population on cheza.
1763 - turnip: Expose robustBufferAccess.
1764 - freedreno/a6xx: Fix clip_halfz support.
1765 - ci: Leave a note as to what might be going on with a test.
1766 - ci: Fix weird filesystem globs appearing in failed test .qpa files.
1767 - ci: Disable some flaky tests on turnip.
1768 - ci/bare-metal: Reword the final output of the init script on the board.
1769 - ci/bare-metal: Make which test to run configurable.
1770 - ci/bare-metal: Use the deqp-runner bits straight out of the artifacts.
1771 - ci/bare-metal: Stop fetching the git tree.
1772 - ci/bare-metal: Terminate the job with an error on kernel panic.
1773 - docs: Replace ancient swrast conformance docs with more current information.
1774 - docs: Add dri-devel to the mailing lists and drop the DRI wiki link.
1775 - ci: disable the windows tests until the runner can be stabilized again
1776 - ci: Bump vulkan CTS to 1.2.3.0.
1777 - ci: Enable NIR validation on a630 GLES2 and VK tests.
1778 - ci/bare-metal: Skip setting of unset variables at startup.
1779 - ci/bare-metal: Don't include dev packages in arm*test.
1780 - ci/tracie: Print the path if the trace isn't found.
1781 - ci/tracie: Fix apitrace dump using "less" which isn't in the ARM rootfs.
1782 - ci: Add a freedreno a630 tracie run.
1783 - freedreno/a6xx: Define the register fields for polygon fill mode.
1784 - turnip: Add support for polygon fill modes.
1785 - freedreno/a6xx: Add support for polygon fill mode (as long as front==back).
1786 - ci: Remove a stray "always" on the freedreno traces job.
1787 - ci/bare-metal: Fail early when we get stuck powering on a cheza.
1788 - ci/baremetal: Bump the kernel to a recent drm-msm-fixes for msm semaphores.
1789 - turnip: Do better TU_DEBUG=startup logging of drmGetDevices2() failure.
1790 - turnip: Fix error handling of DRM_MSM_GEM_INFO ioctls.
1791 - turnip: Properly return VK_DEVICE_LOST on queuesubmit failures.
1792 - gallium/util: Add a helper function for point sprite handling.
1793 - vc4: Enable PIPE_CAP_TGSI_TEXCOORD.
1794 - v3d: Enable PIPE_CAP_TGSI_TEXCOORD.
1795 - v3d: Fix -Wmaybe-uninitialized compiler warning in the v33 code.
1796 - ci: Disable pixmark-piano trace on a630 due to GPU hangs.
1797 - util: Avoid strict aliasing bugs in xxhash.
1798 - util: Mark util_format_description() as a const function.
1799 - softpipe: Clean up softpipe's SSBO load/store interpreting instructions.
1800 - util: Remove unused util_format_planar_is_supported().
1801 - etnaviv: Use the util_pack_color_union() helper.
1802 - gallium/util: Fix location of the comment about S8_UINT handling.
1803 - gallium/util: Clean up the Z/S tile write path.
1804 - gallium/util: Move the Z/S handling to the outside of get_tile().
1805 - svga: Reuse util_format_unpack_rgba().
1806 - util: Merge util_format_write_4* functions.
1807 - util: Merge util_format_read_4* functions.
1808 - util: Use designated initializers to clean up the format tables' pack/unpack.
1809 - llvmpipe: Generalize "could llvmpipe fetch this format" check in unit testing.
1810 - util: Remove the stub pack/unpack functions for YUV formats.
1811 - util: Share a single function pointer for the 4-byte rgba unpack function.
1812 - docs: Move the current CI .rst doc to docs/ci/ and link to it from .gitlab-ci.
1813 - docs: Move the conformance and the CI docs to a top level Testing section.
1814 - docs: Move the gitlab-ci docs to RST.
1815 - docs: Relax the expectations of HW CI farms.
1816 - docs: Document how to interact with docker containers.
1817 - freedreno/ir3_cmdline: Fix an uninit var warning.
1818 - freedreno/ir3: Fix uninit var warning.
1819 - intel: Fix release-build warnings about sf_entry_size.
1820 - intel/perf: Fix unused var warning in release builds.
1821 - intel/perf: Move perf query register programming to static tables.
1822 - freedreno/a2xx: Fix compiler warning in disasm.
1823 - meson: Enable GCing of functions and data from compilation units by default.
1824 - freedreno/ir3: Fix duplicated fine derivatives instructions.
1825 - freedreno/ir3: Add unit tests for derivatives disasm.
1826 - ci: Use FDO_CI_CONCURRENT as our -j flags when present in the runner env.
1827 - freedreno/ir3: Add a note about the instructions in the disasm test.
1828 - freedreno/ir3: Add a bunch more tests for cat6 opcodes.
1829 - freedreno/ir3: Refactor cat6 general dst printing.
1830 - freedreno/ir3: Fix disasm of register offsets in ldp/stp.
1831 - freedreno/ir3: Add missing ld_args_build_id to the ir3_delay unit test.
1832 - ci: Set XDG_CACHE_HOME to tmpfs for bare-metal runners to avoid NFS.
1833 - ci: Update checksums for freedreno traces.
1834 - llvmpipe: Remove a bunch of default handling of pipe caps.
1835 - llvmpipe: Use the default behavior of ALLOW_MAPPED_BUFFERS.
1836 - softpipe: Remove a bunch of default handling of pipe caps.
1837 - softpipe: Use the default behavior of ALLOW_MAPPED_BUFFERS.
1838 - virgl: Remove a bunch of default handling of pipe caps.
1839 - swr: Remove a bunch of default handling of pipe caps.
1840 - swr: Use the default behavior of ALLOW_MAPPED_BUFFERS.
1841 - svga: Remove a bunch of default handling of pipe caps.
1842 - i915: Remove a bunch of default handling of pipe caps.
1843 - softpipe: Refactor pipe_shader_state setup.
1844 - softpipe: Convert to comma-separated SOFTPIPE_DEBUG for debug options.
1845 - softpipe: Add support for reporting shader-db output.
1846 - softpipe: Enable PIPE_CAP_TGSI_TEXCOORD.
1847 - softpipe: Enable PIPE_CAP_TGSI_ANY_REG_AS_ADDRESS;
1848 - ci/bare-metal: Capture the first devcoredump a job produces.
1849 - drm-shim: Return -EINVAL instead of abort()ing on unknown ioctls.
1850 - docs: Explain how to set up a personal gitlab runner.
1851 - nir: Add a pass to cut the trailing ends of vectors.
1852 - i965: Enable vector shrinking in the vec4 backend.
1853 - amd: Swap from nir_opt_shrink_load() to nir_opt_shrink_vectors().
1854 - nir: Remove the old nir_opt_shrink_load.
1855 - freedreno: Fix "Offset of packed bitfield changed" warnings:
1856 - nir/lower_amul: Use num_ubos/ssbos instead of recomputing it.
1857 - nir: Add a little more docs about NIR's constant_data.
1858 - nir: Print the constant data size associated with a shader.
1859 - freedreno/ir3: Fix the type of half-float indirect uniform loads.
1860 - freedreno/a6xx: Document the bit for the magic 32bit-uniforms-as-16b mode.
1861 - freedreno/computerator: Set SP_MODE_CONTROL to the same value as vulkan/GL
1862 - freedreno/ir3: Merge the redundant immediate_idx/immediates_count fields
1863 - freedreno/ir3: Simpify the immediates from an array of vec4 to array of dwords.
1864 - freedreno: Rename emit_const_bo() to emit_const_ptrs().
1865 - freedreno: Split ir3_const's user buffer and indirect upload APIs.
1866 - freedreno/ir3: Clean up instrlen setup.
1867 - freedreno: Increase the NUM_UNIT on compute's consts in indirect dispatch.
1868 - freedreno: Add more asserts for DST_OFF/NUM_UNIT in indirect const uploads.
1869 - freedreno/ir3: Fix assertion failures dumping CS high full regs.
1870 - turnip: Make sure we include the build id.
1871 - gallium/tgsi_exec: Fix up NumOutputs counting
1872 - freedreno: Make the pack struct have a .qword for wide addresses.
1873 - turnip: Fix truncation of CS shader iovas to 32 bits.
1874 - turnip: Fix truncation of iovas to 32 bits in queries.
1878 - cut 20.1 branch
1879 - docs: update calendar for 20.1.0-rc2
1880 - post_version.py: fix branch name construction for release candidates
1881 - post_version.py: invert `is_point` into `is_first_release` to make its purpose clearer
1882 - post_version.py: stop adding release candidates to the index and relnotes
1883 - docs: update calendar for 20.1.0-rc3
1884 - gitlab-ci: exclude scripts that don't affect the build
1885 - util/rand_xor: make it clear that {,s_}rand_xorshift128plus take *exactly 2* uint64_t
1886 - util/rand_xor: drop unused header
1887 - util/rand_xor: fallback Linux to time-based instead of fixed seed
1888 - util/rand_xor: extend the urandom path to all non-Windows platforms
1889 - docs: update calendar for 20.1.0-rc4
1890 - anv: pass the fd directly to anv_gem_reg_read()
1891 - anv: replace magic `| 1` with already #define'd name
1892 - anv: disable VK_EXT_calibrated_timestamps when the timestamp register is unreadable
1893 - git_sha1_gen.py: fix out-of-date comment
1894 - git_sha1_gen.py: fix code style
1895 - git_sha1_gen.py: fix whitespace
1896 - compiler: delete leftover autotools test wrapper
1897 - no_extern_c.h: fix typo in comment
1898 - tree-wide: fix deprecated GitLab URLs
1899 - docs: drop no-longer-relevant comment about bugzilla
1900 - docs: Add release notes for 20.1.0
1901 - docs: update calendar, add news item, and link releases notes for 20.1.0
1902 - meson: remove "empty array"/"array of an empty string" confusion
1903 - glapi: remove deprecated .getchildren() that has been replace with an iterator
1904 - intel/genxml: drop sort_xml.sh and move the loop directly in gen_sort_tags.py
1905 - intel: fix gen_sort_tags.py
1906 - docs: Add release notes for 20.1.1
1907 - docs: update calendar, add news item, and link releases notes for 20.1.1
1908 - v3d: add missing unlock() in error path
1909 - intel/genxml: drop python 2 support for gen_sort_tags.py
1910 - intel/genxml: replace gen_sort_tags.py MIT licence with SPDX equivalent
1911 - docs: update the blocks of unused EGL enums assigned to us
1912 - i965: drop dead #include "config.h"
1913 - iris: drop dead #include "config.h"
1914 - gen_release_notes.py: update script to the new rST way of things
1915 - post_version.py: update script to the new rST way of things
1916 - intel/tools: rewrite run-test.sh in python
1917 - intel/tools: make test aware of the meson test wrapper
1918 - khronos-update.py: add script to simplify update of Khronos headers & xml files
1919 - docs: remove plain-text copy of versions.rst
1920 - util/os_file: replace broken windows-detection code with detect_os.h
1921 - util: introduce os_dupfd_cloexec() helper
1922 - replace all F_DUPFD_CLOEXEC with os_dupfd_cloexec()
1923 - vulkan/wsi: replace all dup() with os_dupfd_cloexec()
1924 - radv: replace all dup() with os_dupfd_cloexec()
1925 - anv: replace all dup() with os_dupfd_cloexec()
1926 - iris: replace all dup() with os_dupfd_cloexec()
1927 - i965: replace all dup() with os_dupfd_cloexec()
1928 - egl: replace all dup() with os_dupfd_cloexec()
1929 - etnaviv: replace all dup() with os_dupfd_cloexec()
1930 - freedreno: replace all dup() with os_dupfd_cloexec()
1931 - svga: replace all dup() with os_dupfd_cloexec()
1932 - virgl: replace all dup() with os_dupfd_cloexec()
1933 - docs: publish our release maintainers' keys
1934 - docs: remind release maintainers to sign the tarballs and publish their key
1935 - docs: suggest alternative installation methods for meson
1936 - docs: stop considering `Cc: mesa-stable` as an email address
1937 - docs: reword "sending a patch revision" to "updating a merge request"
1938 - docs: drop `git sendemail` instructions
1939 - docs: prefer `Fixes:` over `Cc: mesa-stable`
1940 - docs: add some formatting to the "backport merge request" option
1941 - docs: reword a sentence a bit
1942 - docs: make it clear that the tags needs to be in the commit message
1943 - docs: move `Fixes:` tag explanation to its own section
1944 - docs: move "stable" tag explanation next to `Fixes:`
1945 - driconf: drop 28% catalan translation
1946 - driconf: drop 15% german translation
1947 - driconf: drop 26% spanish translation
1948 - driconf: drop 6% french translation
1949 - driconf: drop 8% dutch translation
1950 - driconf: drop 9% swedish translation
1951 - driconf: drop now unused translation facility
1952 - util: rename xmlpool.h to driconf.h
1953 - gitlab-ci: drop gettext from the build images
1954 - docs: drop deleted file from extra sphinx files
1955 - docs: cat maintainer keys to a single file
1956 - docs: add some padding to the release calendar
1957 - docs: add planning for 20.2
1958 - bin/symbols-check: explain C++ symbols workaround
1959 - docs: Add release notes for 20.1.2
1960 - docs: update calendar and link releases notes for 20.1.2
1961 - docs: fix 20.1.2 relnotes
1962 - docs: add a page explaining the GitLab CI and the Intel CI
1963 - mesa/glformats: make _mesa_gles_error_check_format_and_type() more consistent
1964 - docs: add release notes for 20.1.3
1965 - docs: update calendar and link releases notes for 20.1.3
1966 - docs: fix a bunch of typos
1967 - egl: always compile surfaceless
1968 - vulkan: automatically compile the `display` platform when available
1969 - meson: move xlib-lease block further down
1970 - egl: automatically compile the `drm` platform when available
1971 - introduce `commit_in_branch.py` script to help devs figure this out
1972 - bin/gen_release_notes.py: drop new_features.txt when we release XX.Y.0
1973 - egl/wayland: add missing newline between functions
1974 - glx: drop always-true #ifdef
1975 - docs/submittingpatches: add more than one `Cc: mesa-stable` example to the examples list
1976 - meson/intel: add missing dep on git_sha1.h
1977 - meson: fix android vulkan build
1978 - egl: inline fallback for create_pixmap_surface
1979 - egl: inline fallback for create_pbuffer_surface
1980 - egl: drop unused fallback function
1981 - egl: inline fallback for swap_buffers_with_damage
1982 - egl: inline fallback for swap_buffers_region
1983 - egl: inline fallback for post_sub_buffer
1984 - egl: inline fallback for copy_buffers
1985 - egl: inline fallback for query_buffer_age
1986 - egl: inline fallback for create_wayland_buffer_from_image
1987 - egl: inline fallback for get_sync_values
1988 - egl: drop now empty egl_dri2_fallbacks.h
1989 - egl: mark the rest of the callbacks as mandatory or optional
1990 - egl: inline _EGLAPI into _EGLDriver
1991 - docs: add release notes for 20.1.4
1992 - docs: update calendar and link releases notes for 20.1.4
1993 - post_version.py: don't generate relnotes twice
1994 - post_version.py: drop incorrect conf.py changes
1995 - post_version.py: stop using non-existent functions and fix commit message
1996 - post_version.py: update the files in the current worktree, not the one with the script that we run
1997 - post_version.py: fix relnotes links
1998 - bin/gen_release_notes: automatically commit release notes
1999 - docs/releasing: improve wording
2000 - bin/khronos-update: having a folder in include/ is not a requirement
2001 - bin/khronos-update: add support for the SPIRV files
2002 - bin/khronos-update: add workaround for python bug 9625
2003 - egl: replace _eglInitDriver() with a simple variable
2004 - egl: drop unnecessary _eglGetDriver()
2005 - egl: fix _eglMatchDriver() return type
2006 - egl: inline _eglMatchAndInitialize() and refactor _eglMatchDriver()
2007 - egl: rename _eglMatchDriver() to _eglInitializeDisplay()
2008 - egl: drop left-over function prototype
2009 - egl: const _eglDriver
2010 - egl/haiku: drop overwritten preset of EGL version
2011 - egl: consistently use dri2_egl_display() helper macro
2012 - meson: fix `-D xlib-lease=auto` detection
2013 - docs: add release notes for 20.1.5
2014 - docs: update calendar and link releases notes for 20.1.5
2015 - pick-ui: specify git commands in "resolve cherry pick" message
2016 - egl/entrypoint-check: split sort-check into a function
2017 - egl/entrypoint-check: add check that GLVND and plain EGL have the same entrypoints
2018 - driconf: fix force_gl_vendor description
2019 - meson: bump required glvnd version
2020 - egl/x11_dri3: enable & require xfixes 2.0
2021 - egl/x11_dri3: implement EGL_KHR_swap_buffers_with_damage
2022 - meson: don't advertise TLS support if glx wasn't build with it
2023 - meson: drop leftover PTHREAD_SETAFFINITY_IN_NP_HEADER
2027 - lima/ppir: introduce liveness internal live set
2028 - lima/ppir: fix lod bias register codegen
2029 - lima/ppir: do not assume single src for pipeline outputs
2030 - lima/ppir: combine varying loads in node_to_instr
2031 - lima/ppir: duplicate intrinsics in nir
2032 - lima/ppir: duplicate consts in nir
2033 - lima/ppir: remove unused clone functions
2034 - lima/ppir: rework emit nir to ppir
2035 - lima/ppir: rework store output
2036 - lima/ppir: add fallback mov option for const scheduler
2037 - lima/ppir: rework select conditions
2038 - lima/ppir: handle failures on all ppir_emit_cf_list paths
2039 - lima/ppir: improve handling for successors in other blocks
2040 - lima/ppir: rework tex lowering
2041 - lima/ppir: optimize tex loads with single successor
2042 - lima/ppir: use a ready list in node_to_instr
2044 Erik Faye-Lund (124):
2046 - compiler/nir: move tan-calculation to helper
2047 - vtn/opencl: add native_tan-support
2048 - vtn/opencl: native variants of sin/cos
2049 - vtn/opencl: native divide support
2050 - vtn/opencl: native powr support
2051 - vtn/opencl: native recip support
2052 - vtn/opencl: native rsqrt support
2053 - vtn/opencl: native sqrt support
2054 - compiler/glsl: explicitly store NumUniformBlocks
2055 - mesa/st: consider NumUniformBlocks instead of num_ubos when binding
2056 - zink: use nir_lower_uniforms_to_ubo
2057 - zink: lower b2b to b2i
2058 - util/os_memory: never use os_memory_debug.h
2059 - st/wgl: pass st_context_iface into stw_st_framebuffer_present_locked
2060 - st/wgl: allocate and resolve msaa-textures
2061 - docs/features: add zink features
2062 - zink: load vk_GetMemoryFdKHR while creating screen
2063 - zink: add a GET_PROC_ADDR macro to simplify load_device_extensions
2064 - docs/features: mark GL_NV_conditional_render as done for zink
2065 - zink: disable vkCmdResolveImage when respecting render-condition
2066 - zink: do not expose real value for PIPE_CAP_MAX_VIEWPORTS
2067 - zink: correct PIPE_SHADER_CAP_MAX_SHADER_IMAGES
2068 - zink: mark depth-component cube-maps as done
2069 - zink: implement i2b1
2070 - docs: fix broken release-calendar
2071 - zink: hammer in an explicit wait when retrieving buffer contents for reading
2072 - zink: use samples from state
2073 - zink: do not dig into resource for nr_samples
2074 - zink: pass batch instead of context for queries
2075 - zink: implement nir_texop_txf_ms
2076 - zink: expose PIPE_CAP_TEXTURE_MULTISAMPLE
2077 - docs/features: mark GL_ARB_texture_multisample as done for zink
2078 - zink: use general-layout when blitting to/from same resource
2079 - zink: Use store_dest_raw instead of storing an uint
2080 - nir: reuse existing psiz-variable
2081 - zink: emulate B8G8R8X8_SRGB with B8G8R8A8_SRGB
2082 - zink: assert that image-view format isn't undefined
2083 - zink: only report device-local memory as video-memory
2084 - gallium/hud: do not specify potentially invalid depth-range
2085 - TEMP: add rst-conversion scripts
2086 - docs: convert articles to reructuredtext
2087 - TEMP: remove rst-conversion scripts
2088 - docs: delete no longer needed file
2089 - docs: fixup botched table
2090 - docs: escape double colons
2091 - docs: escape asterisks
2092 - docs: escape trailing underscores properly
2093 - docs: fixup broken rst
2094 - docs: fixup heading-levels
2095 - docs: use sphinx
2096 - docs: disable syntax-highlighting by default
2097 - docs: use code-block with caption instead of table
2098 - docs: format notes as rst-notes
2099 - docs: use code-blocks
2100 - docs: drop open-coded toc for articles
2101 - docs: add xlibdriver to table-of-contents
2102 - docs: do not copy source-files to site
2103 - docs: use rst footnotes instead of manual ones
2104 - docs: reformat license table as rst table
2105 - docs: use rst-note for highlighted text
2106 - docs: bundle extra files
2107 - docs: include specs into the generated docs
2108 - gitlab-ci: build and deploy docs
2109 - docs: drop news in favour of the introduction as index-page
2110 - README: update references to internal docs
2111 - docs: update internal references
2112 - docs/relnotes: update internal references
2113 - radv: update internal reference
2114 - bin/perf-annotate-jit.py: update internal reference
2115 - docs/release-calendar: restore missing id
2116 - nir: do not try to merge xfb-outputs
2117 - Revert "gallium/hud: don't use user vertex buffers"
2118 - gallium/hud: don't use user vertex buffers
2119 - zink: enable cull-distance if supported
2120 - zink: expose GLSL 1.30
2121 - docs: update internal references
2122 - docs/relnotes: update internal references
2123 - docs: fixup relnotes after rst-conversion
2124 - docs/features: mark GL3 as complete for zink
2125 - docs/features: update ARB_texture_buffer_object line
2126 - docs/features: remove driver-list for forward-compatible context
2127 - mesa/main: fix inverted condition
2128 - gallium/os: call "ANSI" version of GetCommandLine
2129 - graw/gdi: do not depend on UNICODE macro
2130 - gallium/util: limit STACK_LEN on Windows
2131 - gallium/util: add missing include
2132 - docs: update favicon
2133 - docs: remove non-existent reference
2134 - docs: restore accidentally dropped labels
2135 - docs: fix internal references
2136 - docs: use ref-links for internal references
2137 - gallium/docs: update to recent sphinx
2138 - gallium/docs: fixup formatting of numbered lists
2139 - gallium/docs: remove reference to non-existent label
2140 - gallium/docs: use none for highlight_language
2141 - gallium/docs: prefix exts dir with underscore
2142 - gallium/docs: remove non-existent static dir
2143 - gallium/docs: remove unused imgmath extension
2144 - ci: only build docs in the upstream-repo
2145 - ci: only build docs if any docs changed
2146 - ci: test docs for non-master builds
2147 - ci: move deploy-stage later in the pipeline
2148 - ci: move test-docs to container stage
2149 - ci: add graphviz to the .docs-base template
2150 - merge gallium docs into main docs
2151 - docs: clean up gallium index-file
2152 - docs: add an extension to generate redirects
2153 - docs: move gallium specific docs into gallium folder
2154 - docs: use svg for graphviz output
2155 - docs: fixup envvar output
2156 - zink: expose depth-clip if supported
2157 - mesa/main: factor out one-time-init into a helper
2158 - mesa/main: use call_once instead of open-coding
2159 - gallium/util: do not use _MTX_INITIALIZER_NP on Windows
2160 - mesa/main: use p_atomic_inc_return instead of locking
2161 - mesa: do not use bitfields for advanced-blend state
2162 - mesa: treat Color._AdvancedBlendMode as enum
2163 - zink: use ralloc in nir-to-spirv
2164 - zink: use ralloc for plain malloc-calls
2165 - zink: pass mem_ctx to ralloc_size-call
2166 - zink: use ralloc for spirv_builder as well
2167 - mesa/program: fix shadow property for samplers
2168 - docs: add some very basic documentation about zink
2169 - mesa: handle GL_FRONT after translating to it
2173 - intel/ir: Update performance analysis parameters for memory fence codegen changes.
2174 - iris: Simplify iris_batch_prepare_noop().
2175 - iris: Extend iris_context dirty state flags to 128 bits.
2176 - iris: Add batch-local synchronization book-keeping to iris_bo.
2177 - iris: Add infrastructure to partition batch into sync boundaries.
2178 - iris: Bracket batch operations which access memory within sync regions.
2179 - iris: Annotate all BO uses with domain and sequence number information.
2180 - iris: Drop redundant iris_address::write flag.
2181 - iris: Report use of any in-flight buffers on first draw call after sync boundary.
2182 - iris: Introduce cache coherency matrix for batch-local memory ordering.
2183 - iris: Update cache coherency matrix on PIPE_CONTROL.
2184 - iris: Implement buffer-local memory barrier based on cache coherency matrix.
2185 - iris: Insert buffer barrier in existing cache flush helpers.
2186 - iris: Remove batch argument of iris_resource_prepare_access() and friends.
2187 - iris: Perform compute predraw flushes from compute batch.
2188 - iris: Remove depth cache set tracking and synchronization.
2189 - iris: Remove render cache hash table-based synchronization.
2190 - iris: Open-code iris_cache_flush_for_read() and iris_cache_flush_for_depth().
2191 - iris: Emit single render target flush PIPE_CONTROL on format mismatch.
2192 - iris: Remove iris_flush_depth_and_render_caches().
2193 - OPTIONAL: iris: Perform BLORP buffer barriers outside of iris_blorp_exec() hook.
2194 - iris/icl+: Report same caching domain as main surface for clear color BO.
2195 - intel/ir/gen12+: Work around FS performance regressions due to SIMD32 discard divergence.
2199 - docs: change "Fixes:" tag example to match git fixes output
2200 - egl/dri2: only take a dri2_dpy reference when binding a new context/surfaces
2204 - clover: Fix types collision between c++ and altivec
2205 - meson: Revert commit overriding C++ standard with gnu++11 on ppc64el
2209 - r600: Annotate some case fallthroughs
2210 - r600: remove unused static functions
2211 - r600/sb: replace memset by using member initialization/assignment
2212 - r600: remove some unused variables to silence warnings
2213 - r600: Fix warning regarding mixing enums and unsigned in ?: expression
2214 - r600: Fix nir compiler options, i.e. don't lower IO to temps for TESS
2215 - r600/sfn: Unify semantic name and index query and use TEXCOORD semantic
2216 - r600/sfn: Fix printing vertex fetch instruction flags
2217 - r600: Lower int64 ops from TGSI-to-NIR shaders too
2218 - r600: Lower lerp after tgsi_to_nir
2219 - r600: Add support for loading index register from other than chan X
2220 - r600/sfn: Handle CF index loading from non-X channel
2221 - r600/sfn: rework getting a vector and uniforms from the value pool
2222 - r600/sfn: Skip move instructions if they are only ssa and without modifiers
2223 - r600/sfn: re-use an allocated register in lookup
2224 - r600/sfn: skip copying LOD if the target register is is the same
2225 - r600/sfn: Fix memring print output
2226 - r600/sfn: Fix RING instruction assembly emission
2227 - r600/sfn: Fix GDS assembly emission
2228 - r600/sfn: Fix RAT instruction assembly emission
2229 - r600/sfn: Make allocate_reserved_registers forward to a virtual function
2230 - r600/sfn: Fix handling of output register index
2231 - r600/sfn: Make 3vec loads skip possible moves
2232 - r600/sfn: Add support for viewport index output
2233 - r600/sfn: Take FOGC, and backcolors into account im GS outputs
2234 - r600/sfn: Handle loading sample_pos
2235 - r600/sfn: Add FS output sample_mask
2236 - r600/sfn: Don't reject VARYING_SLOT_PCNT
2237 - r600/sfn: remove pointless check
2238 - r600/sfn: assert when alu dest is missing
2239 - r600/sfn: support indirect sampler buffer reads.
2240 - r600/sfn: Add support for texture_samples
2241 - r600/sfn: use the per shader atomic base
2242 - r600/sfn: SSBO: Fix query of dest components
2243 - r600/sfn: Fix clip vertex output as possible stream variable
2244 - r600/sfn: Fix splitting constants that come from different kcache banks.
2245 - r600/sfn: Don't reorder outputs by location
2246 - r600/sfn: Fix printing ALU op without dest
2247 - r600: Fix duplicated subexpression in r600_asm.c
2248 - r600/sfn: Fix mapping for f32tof64 and f64tof32
2249 - r600/sfn: use modern c++ in printing LDS read instruction
2250 - r600/sfn: Correctly update the number of literals when forcing a new group
2251 - r600/sfn: remove debug output leftover
2252 - nir: lower_tex: Don't normalize coordinates for TXF with RECT
2253 - r600/sfn: lower image derefs
2254 - r600/sfn: Add imageio support
2255 - r600/sfn: Add support for image_size
2256 - r600/sfn: Add support for reading cube image array dim.
2257 - r600/sfn: Take SSBO buffer ID offset into account
2258 - r600/sfn: Handle memory_barrier
2259 - r600/sfn: Add lowering pass for shared IO
2260 - r600/sfn: Add support for shared atomics
2261 - r600/sfn: Don't set num_components on TESS sysvalue intrinsics
2262 - r600/sfn: lower rotate ALU ops
2263 - r600/sfn: Pipe through requesting a register at a given channel
2264 - r600/sfn: emit texture instructions in one block
2265 - r600/sfn: Add option to get a temp value for a specific channel
2266 - r600/sfn: correct handling of loading vec4 with fetching constants
2267 - r600/sfn: Add a forced output swizzle for depth write
2268 - r600/sfn: Fix Ring output swizzle masks
2269 - r600/sfn: Fix default z swizzle for GDS instructions
2270 - r600: Add shader key item to identify when the sample mask should be used
2271 - r600/sfn: Only use sample mask if the according shader key is set
2272 - r600/sfn: Make the pin_to_channel generic
2273 - d600/sfn: write stream outputs to correct mem ring
2274 - gallivm/nir: Lower uniforms to UBOs in llvm draw if the driver didn't request this already
2278 - gallium,util: undef ALIGN on FreeBSD to prevent name clash
2282 - etnaviv: drm: Use NSEC_PER_SEC
2283 - etnaviv: drm: Normalize nano seconds
2287 - virgl: apply bgra dest swizzle and add Portal 2
2291 - Properly check mmap return value
2295 - freedreno,tu: Don't request fragcoord components not being read.
2296 - tu,radv: fix potentially wrong offset of flexible array.
2297 - vulkan: Adds helpers for vk_object (de)alloation and (de)initialization.
2298 - tu: Fix wrong copies of sampler descriptor.
2299 - turnip: Use the common base object type and struct.
2300 - turnip: implement VK_EXT_private_data
2304 - v3d/compiler: don't rewrite unused temporaries to point to NOP register
2305 - v3d/compiler: fix spill offset
2306 - v3d/compiler: fix image size for 1D arrays
2307 - nir/lower_clip: make the pass compatible with Vulkan semantics
2308 - v3d/compiler: handle compact varyings
2309 - v3d/compiler: request fragment shader clip lowering to be vulkan compatible.
2310 - nir/lower_tex: skip lower_tex_packing for the texture samples query
2314 - nir/algebraic: Recognize open-coded byte or word extract from bfe
2315 - nir/algebraic: Split ibfe and ubfe with two constant sources
2316 - nir/algebraic: Optimize some bfe patterns
2317 - nir/algebraic: Optimize ushr of pack_half, not ishr
2318 - nir/algebraic: Add some half packing optimizations for pack_half_2x16_split
2319 - nir/algebraic: Eliminate useless extract before unpack
2320 - i965: Assert that blorp always handles color blits
2321 - meta: Make _mesa_meta_texture_object_from_renderbuffer static
2322 - meta: Make _mesa_meta_setup_sampler static
2323 - meta: Remove support for clearing integer buffers
2324 - mesa: Add matrix utility functions to load matrices
2325 - mesa: Add function to calculate an orthographic projection
2326 - meta: Stop frobbing MatrixMode
2327 - meta: Use same vertex coordinates for GLSL and FF clears
2328 - meta: Coalesce the GLSL and FF paths in meta_clear
2329 - meta: Remove support for multisample blits
2330 - anv/tests: Don't rely on assert or changing NDEBUG in tests
2331 - anv/tests: Silence unused parameter warnings in main
2332 - anv: Silence unused parameter warning in anv_image_get_clear_color_addr
2333 - intel: Silence unused parameter warning in __intel_log_use_args
2334 - intel/drm-shim: Add noop ioctl handler for set_tiling
2335 - intel/drm-shim: Return correct values for I915_PARAM_HAS_ALIASING_PPGTT
2336 - glsl: Remove integer matrix support from ir_dereference_array::constant_expression_value
2337 - nir/algebraic: Don't distrubte absolute-value into dot-products
2341 - pan/midgard: Fix old style shadows
2342 - panfrost: Fix background showing when using discard
2343 - panfrost: Enable PIPE_CAP_VERTEX_COLOR_UNCLAMPED
2344 - panfrost: Decode AFBC flag bits
2345 - panfrost: Only use AFBC YTR with RGB and RGBA
2346 - pan/midgard: Use a signed value for checking inline constants
2347 - Revert "panfrost: Keep cached BOs mmap'd"
2348 - panfrost: Mark PIPE_BUFFER BOs as not renderable
2349 - pan/mdg: Add a macro for printing instruction source information
2350 - pan/mdg: Move r1.w writeout to branch->dest
2351 - pan/mdg: Remove old zs store lowering
2352 - pan/mdg: Remove old depth writeout code
2353 - pan/mdg: Remove writeout case from bytemask_of_read_components
2354 - nir: Replace the zs_output_pan intrinsic with combined_output_pan
2355 - pan/mdg: Replace writeout booleans with a single value
2356 - pan/mdg: Add new depth writeout code
2357 - pan/mdg: Move search_var to earlier in midgard_compile.c
2358 - pan/mdg: Add depth/stencil support to emit_fragment_store
2359 - pan/mdg: Add new depth store lowering
2360 - pan/mdg: Print writeout sources in mir_print_instruction
2361 - panfrost: Add writes_stencil to the EARLY_Z disable list
2362 - panfrost: Move sampler view bo creation to a separate function
2363 - panfrost: Create a new sampler view bo when the layout changes
2364 - panfrost: Tiled to linear layout conversion
2365 - panfrost: Clean up panfrost_frag_meta_rasterizer_update
2366 - panfrost: Implement ARB_depth_clamp
2367 - pan/decode: Fix helper invocations when tracing
2368 - pan/decode: Add missing wrap modes
2369 - pan/mdg: Fix max_comp calculation for constant printing
2370 - panfrost: RGBA4 and RGB5_A1 framebuffer support
2371 - panfrost: Update sampler views when the texture bo changes
2372 - panfrost: Copy resources when mapping to avoid waiting for readers
2373 - panfrost: Only copy resources when they are in a pending batch
2374 - panfrost: Add PAN_MESA_DEBUG=gl3 flag
2375 - panfrost: Do fine-grained flushing for occlusion query results
2376 - pan/mdg: Vectorize vlut operations
2377 - pan/decode: Make mapped memory read-only while decoding
2378 - nir: Add a base value to load_raw_output_pan
2379 - panfrost: Fix MALI_READS_TILEBUFFER
2380 - pan/mdg: Handle tilebuffer wait loops
2381 - pan/mdg: Use the writeout tag for tilebuffer wait loops
2382 - panfrost: Add rt formats to shader state
2383 - panfrost: Add a bitset of render targets read by shaders
2384 - pan/mdg: Do the pan_lower_framebuffer pass later
2385 - pan/mdg: Emit a tilebuffer wait loop when needed
2386 - pan/mdg: Handle non-blend framebuffer lowering
2387 - pan/mdg: Support MRT in output load lowering
2388 - pan/mdg: Set the z/s store intrinsic base correctly
2389 - pan/mdg: Use a 32-bit ld_color_buffer op when needed
2390 - panfrost: Implement texture_barrier
2391 - panfrost: Stop keying on rt format when using native loads
2392 - panfrost: Use f2fmp for framebuffer lowering conversions
2393 - panfrost: Enable framebuffer fetch
2394 - pan/mdg: Fix non-debug compiliation
2395 - compiler: Add dual-source factors to blend_factor
2396 - gallium: Dual source support in blend_factor_to_shader
2397 - pan/mdg: Add a nir pass to reorder store_output intrinsics
2398 - pan/mdg: Dual source blend input/writeout support
2399 - pan/mdg: Skip z/s combining for dual-source writes
2400 - panfrost: Dual source blend support
2401 - pan/decode: Open the dump file later
2402 - pan/mdg: Don't disassemble blit shaders
2403 - panfrost: Rename lower_store to is_blend in pan_lower_framebuffer
2404 - pan/mdg: Do per-sample framebuffer loads
2405 - panfrost: Do per-sample shading when outputs are read
2406 - nir: Add a face_sysval argument to nir_lower_two_sided_color
2407 - nir: Fix lower_two_sided_color when the face is an input
2408 - panfrost: Report TEXTURE_BUFFER_OBJECTS cap when gl3 flag set
2409 - panfrost: Set depth_enabled when stencil is enabled
2410 - nir: Set the alignment for SSBO lowering
2411 - panfrost: Make panfrost_bo_wait take a wait_readers bool
2412 - panfrost: Fix calls to panfrost_flush_batches_accessing_bo
2413 - panfrost: Fake RGTC support
2414 - panfrost: Use more tilebuffer sizes
2415 - panfrost: 8x MRT support
2416 - pan/mdg: Use the blend RT for blend shader framebuffer fetches
2417 - panfrost: Allow PIPE_TEXTURE_1D_ARRAY textures
2418 - pan/mdg: Fix spilling of non-32-bit types
2422 - panfrost: signal syncobj if nothing is going to be flushed
2426 - freedreno/a3xx: there's no r8i/ui rb format, only rg8i/rg8ui
2427 - freedreno/a3xx: reinstate rgb10_a2ui texture format
2428 - freedreno/ir3: avoid applying (sat) on bary.f
2429 - freedreno/a3xx: fix const footprint
2430 - freedreno: fix off-by-one in assertions checking for const sizes
2431 - freedreno/a3xx: parameterize ubo optimization
2432 - freedreno/a3xx: fix rasterizer discard
2433 - nouveau: allow invalidating coherent/persistent buffer backings
2434 - st/mesa: allow R8 to not be exposed as renderable by driver
2435 - a4xx: add noperspective interpolation support
2436 - a4xx: add polygon offset clamp, fix units
2437 - ir3: mark ucp_enables as allowed values on all keys
2438 - a4xx: hook up centroid ij coords
2439 - ir3: use empirical size for params as used by the shader
2443 - st/mesa: use fragment shader to copy stencil buffer
2444 - st/mesa: optimize DEPTH_STENCIL copies using fragment shader
2448 - panfrost: Fix outmods on int to float conversions
2449 - pan/mdg: fix src_type in instructions that need a implicit zero
2450 - pan/mdg: prepare effective_writemask()
2451 - pan/mdg: eliminate references to ins->alu.op
2452 - pan/mdg: eliminate references to ins->alu.reg_mode
2453 - pan/mdg: fix comment
2454 - pan/mdg: eliminate references to ins->alu.outmod
2455 - pan/mdg: apply float outmods to textures
2456 - pan/mdg: eliminate references to ins->texture.op
2457 - pan/mdg: eliminate references to ins->load_store.op
2458 - pan/mdg: defer register packing
2459 - pan/mdg: externalize mir_pack_mod
2460 - pan/mdg: remove ins->alu
2461 - pan/mdg: refactor emit_alu_bundle
2462 - pan/mdg: defer branch packing
2463 - pan/mdg: remove ins->br_compact and ins->branch_extended
2464 - pan/mdg: emit REGISTER_UNUSED on unused ALU src2
2468 - anv: use the correct format on Android
2469 - anv: Disable B5G6R5_UNORM_PACK16
2470 - anv: Add a way to reserve states from a pool
2471 - anv: Implement VK_EXT_custom_border_color
2472 - anv: support externally synchronized pipeline caches
2473 - anv: implement VK_PIPELINE_CREATE_FAIL_ON_PIPELINE_COMPILE_REQUIRED_BIT_EXT
2474 - anv: enable VK_EXT_pipeline_creation_cache_control
2475 - anv: Add VK_EXT_custom_border_color to relnotes
2476 - anv: fix allocation of custom border color pool
2480 - amd/llvm: Reorder LLVM headers
2484 - ac/gpu_info: Correct Acturus cu bitmap
2488 - drm-uapi: Add sync_file.h
2489 - anv,iris: unbreak on BSDs after 812cf5f522ab,abf8aed68047
2490 - util: enable futex usage on BSDs after 7dc2f4788288
2491 - meson: unbreak sysctl.h detection on BSDs
2492 - anv: disable i915_perf warning on non-Linux
2496 - targets/opencl: fix build against LLVM>=10 with Polly support
2500 - gallium/swr: Fix crashes in sampling code
2504 - intel/eu: Use non-coherent mode (BTI=253) for stateless A64 messages
2505 - Revert "anv/gen12: Temporarily disable VK_KHR_buffer_device_address (and EXT)"
2506 - vulkan: Allow destroying NULL debug report callbacks
2507 - vulkan,anv: Add a common base object type for VkDevice
2508 - anv: Stop clflushing events
2509 - anv: Allocate CPU-side memory for events
2510 - vulkan,anv: Add a base object struct type
2511 - vulkan,anv: Move the DEFINE_HANDLE_CASTS macros to vk_object.h
2512 - anv: Refactor setting descriptors with immutable sampler
2513 - vulkan: Add run-time object type asserts in handle casts
2514 - vulkan/wsi: Make wsi_swapchain inherit from vk_object_base
2515 - anv/allocator: Add a start_offset to anv_state_pool
2516 - vulkan/object: Always include the type
2517 - anv,vulkan: Implement VK_EXT_private_data
2518 - vulkan: Handle vkGet/SetPrivateDataEXT on Android swapchains
2519 - nir: Make "divergent" a property of an SSA value
2520 - util/list: Add a list pair iterator
2521 - util/vma: Add an option to configure high/low preference
2522 - util/vma: Add a debug print helper
2523 - util/ra: Add [de]serialization support
2524 - anv: Set 3DSTATE_VF_INSTANCING on the SVGS element
2525 - anv: Set MOCS in 3DSTATE_CONSTANT_* on Gen9+
2526 - nir: Add some docs to the metadata types
2527 - anv: Call vk_object_base_finish for image views
2528 - anv: Fix descriptor set clean-up on BO allocation failure
2529 - nir: Use 8-bit types for most info fields
2530 - anv:gpu_memcpy: Emit 3DSTATE_VF_INDEXING on Gen8+
2531 - nir: Validate jump instructions as an instruction type
2532 - nir: Use a switch statement in nir_handle_add_jump
2533 - nir: Add documentation for each jump instruction type
2534 - nir/clone: Re-use clone_alu for nir_alu_instr_clone
2535 - nir: Add a new helper for iterating phi sources leaving a block
2536 - nir: Add a store_reg helper and use the builder in phis_to_regs
2537 - nir: Add const to nir_intrinsic_src_components
2538 - nir/lower_double_ops: Rework the if (progress) tree
2539 - nir/opt_deref: Report progress if we remove a deref
2540 - nir/copy_prop_vars: Record progress in more places
2541 - nir: Fix sources for image atomic fadd
2542 - intel/vec4: Stomp the return type of RESINFO to UINT32
2543 - intel/fs: Fix unused texture coordinate zeroing on Gen4-5
2544 - intel/fs: Emit HALT for discard on Gen4-5
2545 - anv/allocator: Compare to start_offset in state_pool_free_no_vg
2546 - nir: Add a nir_metadata_all enum value
2547 - nir: Add a nir_shader_preserve_all_metadata helper
2548 - nir: Call nir_metadata_preserve on !progress
2549 - nir: Properly preserve metadata in more cases
2550 - intel/nir: Call nir_metadata_preserve on !progress
2551 - iris: Better handle metadata in NIR passes
2552 - anv: Add an anv_batch_set_storage helper
2553 - anv: Add anv_pipeline_init/finish helpers
2554 - nir/intrinsics: Put the _intel intrinsics together at the end
2555 - anv: Use resolve_device_entrypoint for dispatch init
2556 - vulkan: Update Vulkan XML and headers to 1.2.145
2557 - anv: Bump the advertised patch version to 145
2558 - intel/fs: Expose a couple of NIR lowering helpers
2559 - intel/fs: Break wm_prog_data setup into a helper
2560 - intel/fs: Move more prog_data setup into populate_wm_prog_data
2561 - intel/compiler: Expose brw_texture_offset to C
2562 - intel/eu: Add a brw_urb_dest_msg_type helper
2563 - intel/eu: Set the right subnr for ALIGN16 destinations
2564 - intel/eu: Add the RNDU opcode
2565 - vulkan/wsi: Don't consider VK_SUBOPTIMAL_KHR to be an error condition
2566 - wsi/x11: Log swapchain status changes
2567 - freedreno: Only call nir_lower_io on shader_in/out
2568 - lima: Only call nir_lower_io on shader_in/out
2569 - nouveau: Only call nir_lower_io on shader_in/out
2570 - vc4: Only call nir_lower_io on shader_in/out
2571 - v3d: Only call nir_lower_io on shader_in/out
2572 - panfrost: Only call nir_lower_io on shader_in/out
2573 - nir: Assert that nir_lower_io is only called with allowed modes
2574 - nir: Remove shared support from lower_io
2575 - nir: Add docs to nir_lower[_explicit]_io
2576 - anv: Handle clamping of inverted depth ranges
2577 - nir/validate: Don't abort() until after the shader has printed
2578 - spirv: Skip phis in unreachable blocks in the second phi pass
2579 - spirv: Allow block-decorated struct types for constants
2580 - vulkan: Update Vulkan XML and headers to 1.2.148
2581 - anv: Advertise VK_EXT_image_robustness
2582 - spirv: Update headers and grammar json
2583 - spirv: Add support for SPV_EXT_shader_atomic_float
2584 - intel/fs: Use the correct logical op for global float atomics
2585 - anv: Advertise support for VK_EXT_shader_atomic_float
2586 - nir: Allow for system values with variable numbers of destination components
2587 - nir/lower_io: Choose to set access based on intrinsic metadata
2588 - nir/lower_io: Use b2b for shader and function temporaries
2589 - nir/lower_io: Add support for global scratch addressing
2590 - spirv: Simplify our handling of NonUniform
2591 - spirv: Drop the void \*ptr from vtn_value
2592 - spirv: Fix indentation in vtn_handle_ptr
2593 - spirv: Clean up OpSignBitSet
2594 - spirv: Use nir_bany/ball for OpAny/All
2595 - spirv: Add a helpers for getting types of values
2596 - spirv: Rename push_value_pointer to push_pointer
2597 - spirv: Add a vtn_push_nir_ssa helper
2598 - spirv/amd: Use vtn_push_nir_ssa
2599 - spirv: Add a vtn_get_nir_ssa helper
2600 - spirv: Use the new helpers in OpConvertUToPtr/PtrToU
2601 - spirv: Refactor vtn_push_ssa
2602 - spirv/alu: Use vtn_push_ssa_value
2603 - spirv/glsl450: Use vtn_push_ssa_value
2604 - spirv/subgroups: Stop incrementing w
2605 - spirv/subgroups: Refactor to use vtn_push_ssa
2606 - spirv: Simplify vtn_ssa_value creation
2607 - spirv: Hand-roll fewer vtn_ssa_value creations
2608 - spirv: Add better checks for SSA value types
2609 - spirv: Drop the sampled boolean from vtn_type
2610 - spirv: Give atomic counters their own variable mode
2611 - spirv: Add a helper for getting the NIR type of a vtn_type
2612 - spirv: Remove a dead case in function parameter handling
2613 - spirv: More heavily use vtn_ssa_value in function parameter handling
2614 - anv,turnip,radv,clover,glspirv: Run nir_copy_prop before nir_opt_deref
2615 - spirv: Rework our handling of images and samplers
2616 - spirv: Also copy over binding information for atomic counters
2617 - nir: Take a mode in remove_unused_io_vars
2618 - nir/dead_variables: Respect the modes passed to remove_dead_vars
2619 - nir: Add nir_foreach_shader_in/out_variable helpers
2620 - nir: Add a nir_foreach_function_temp_variable helper
2621 - nir: Add a nir_foreach_uniform_variable helper
2622 - nir: Add a nir_foreach_gl_uniform_variable helper for GL linking
2623 - nir: Add and use a nir_variable_list_for_mode helper
2624 - nir: Take a nir_shader and variable mode in assign_var_locations
2625 - nir: Take a shader and variable mode in nir_assign_io_var_locations
2626 - nir/linking: Rework some internal helpers
2627 - st/nir: Rework fixup_varying_slots
2628 - nir/split_vars: Add mode checks to list walks
2629 - nir: Split nir_index_vars into two functions
2630 - nir/lower_amul: Add a variable mode check
2631 - nir: Use a nir_shader and mode in lower_clip_cull_distance_arrays
2632 - nir/lower_io_to_temporaries: Use a separate list for new inputs
2633 - nir/io_to_vector: Use nir_foreach_variable_with_modes
2634 - nir/lower_two_sided_color: Use nir_variable_create
2635 - nir/lower_uniforms_to_ubo: Use nir_foreach_variable_with_modes
2636 - nir/split_per_member_structs: Use nir_variable_with_modes_safe
2637 - nir/lower_variable_initializers: Restrict the modes we lower
2638 - nir/gl_nir_linker: Use nir_foreach_variable_with_modes
2639 - freedreno/ir3_lower_tess: Rework var list helpers
2640 - lima/standalone: Rework i/o variable fixup
2641 - freedreno/ir3_cmdline: Rework i/o variable fixup
2642 - r600/sfn/lower_tess_io: Rework get_tcs_varying_offset
2643 - r600/sfn/lower_tex: Get rid of the lower_sampler vector
2644 - r600/sfn: Use nir_foreach_variable_with_modes in IO vectorization
2645 - panfrost/midgard: Make search_var take a nir_shader and mode
2646 - panfrost: Use nir_foreach_variable_with_modes in pan_compile
2647 - aco: Use nir_foreach_variable_with_modes to walk SSBOs
2648 - mesa/ptn: Use nir_variable_create
2649 - gallium/ttn: Use variable create/add helpers
2650 - nir: Use a single list for all shader variables
2651 - nir/split_per_member_structs: Inline split_variables_in_list
2652 - nir/gl_nir_linker: Call add_vars_with_modes once for GL_PROGRAM_INPUT
2653 - nir: Add a find_variable_with\_[driver\_]location helper
2654 - vulkan: Update Vulkan XML and headers to 1.2.149
2655 - anv: Implement VK_EXT_4444_formats
2656 - nir/deref: Don't try to compare derefs containing casts
2657 - compiler/types: Add a struct_type_is_packed wrapper
2658 - spirv: Do more complex unwrapping in get_nir_type
2659 - anv: Advertise shaderIntegerFunctions2
2660 - spirv: Don't emit RMW for vector indexing in shared or global
2661 - clover/spirv: Don't call llvm::regularizeLlvmForSpirv
2662 - intel/nir: Pass the nir_builder by reference in lower_alpha_to_coverage
2663 - intel/nir: Rewrite the guts of lower_alpha_to_coverage
2664 - intel/fs: Fix MOV_INDIRECT and BROADCAST of Q types on Gen11+
2665 - intel/fs: Don't copy-propagate stride=0 sources into ddx/ddy
2666 - iris: Re-emit push constants if we have a varying workgroup size
2667 - spirv: Run repair_ssa if there are discard instructions
2668 - nir: More NIR_MAX_VEC_COMPONENTS fixes
2669 - intel/fs/swsb: SCHEDULING_FENCE only emits SYNC_NOP
2670 - radeonsi: Only call nir_lower_var_copies at the end of the opt loop
2674 - nir_lower_io: Add addr_format_is_offset helper
2675 - nir: When nir_lower_vars_to_explicit_types is run on temps, update scratch_size
2676 - nir: Support load/store of temps as scratch in nir_lower_explicit_io
2677 - nir: Support vec8/vec16 in nir_lower_bit_size
2678 - nir: Support algebraic opts on vectors larger than 4
2679 - nir: Support 8 and 16 component vectors for reduceable intrinsics
2680 - nir/vtn: Add support for 8 and 16 vector ball/bany
2681 - u_debug_stack_test: Fix MSVC compiling by using ATTRIBUTE_NOINLINE
2682 - nir: More NIR_MAX_VEC_COMPONENTS fixes
2683 - glsl_type: Add packed to structure type comparison for hash map
2687 - drirc: Enable glthread for PCSX2
2691 - glthread: Fix use of alloca() without #include "c99_alloca.h"
2695 - util: unbreak endian detection on OpenBSD
2696 - util/anon_file: add OpenBSD shm_mkstemp() path
2697 - meson: build with _ISOC11_SOURCE on OpenBSD
2698 - meson: don't build with USE_ELF_TLS on OpenBSD
2699 - meson: conditionally include -ldl in gbm pkg-config file
2700 - util: futex fixes for OpenBSD
2701 - util/u_thread: include pthread_np.h if found
2702 - anv: use os_get_total_physical_memory()
2703 - util/os_misc: add os_get_available_system_memory()
2704 - anv: use os_get_available_system_memory()
2705 - util/os_misc: os_get_available_system_memory() for OpenBSD
2706 - radv: remove seccomp includes
2707 - vulkan: make VK_TIME_DOMAIN_CLOCK_MONOTONIC_RAW_EXT conditional
2711 - turnip: update "fetchsize" value to match fdl6_layout changes
2712 - turnip: enable tiling for compressed formats
2713 - util/format: translate 422_UNORM and 420_UNORM vulkan formats
2714 - freedreno/registers: document 422_UNORM and 420_UNORM formats
2715 - turnip: implement VK_KHR_sampler_ycbcr_conversion
2716 - turnip: enable 422_UNORM formats
2717 - freedreno: move a4xx specific layout code to a4xx code
2718 - freedreno/a5xx: remove unused reference to gmem_alignw in layout code
2719 - freedreno/a6xx: don't use gmem_alignw for imported buffers
2720 - freedreno/a6xx: split up gmem/tile alignment requirements
2721 - freedreno: reduce extra height alignment in a6xx layout
2722 - freedreno/a6xx: use RESOLVE_TS event
2723 - freedreno: add adreno 650
2724 - freedreno/layout: add explicit offset/pitch argument to fdl6_layout
2725 - turnip: support VkImageDrmFormatModifierExplicitCreateInfoEXT
2726 - turnip: fix RENDER_COMPONENTS value
2727 - turnip: move HLSQ_UPDATE_CNTL write to before xs config writes
2728 - turnip: update some properties based on blob driver
2729 - turnip: clamp sampler minLod/maxLod
2730 - freedreno/a6xx: use nonbinning VS when GS is used
2731 - turnip: correctly emit non-binning vs in transform feedback case
2732 - turnip: fix HW binning with geometry shader
2733 - turnip: use common emit_xs_cntl to fill a6xx_sp_xs_ctrl_reg0
2734 - turnip: fix VFD_CONTROL for binning pass
2735 - turnip: pipeline program state refactor
2736 - turnip: share code between 3D blit/clear path and tu_pipeline
2737 - turnip: add layered 3D path clear for CmdClearAttachments
2738 - turnip: add emit renderpass cache flushes for sysmem 3D CmdClearAttachments
2739 - turnip: remove some dead/redundant code
2740 - freedreno/ir3: fix ir3_nir_move_varying_inputs
2741 - turnip: remove duplicated stage2opcode and stage2shaderdb
2742 - turnip: simplify stage2 helpers
2743 - turnip: set VFD_INDEX_OFFSET in 3D clear/blit path
2744 - turnip: fix 3D path always being used for CmdBlitImage
2745 - turnip: fix cubic filtering with CmdBlitImage
2746 - turnip: compute and graphics have completely separate state
2747 - turnip: move descriptor set BO tracking to CmdBindDescriptorSets
2748 - turnip: improve dirty bit handling a bit
2749 - turnip: delete dead dynamic state code
2750 - turnip: refactor draw states and dynamic states
2751 - turnip: input attachment descriptor set rework
2752 - turnip: use draw states for input attachments
2753 - turnip: use u_format for packing gmem clear values
2754 - freedreno/a6xx: FETCHSIZE is PITCHALIGN
2755 - freedreno/fdl6: rework layout code a bit (reduce linear align to 64 bytes)
2756 - turnip: fix a crash when rasterizerDiscardEnable is set
2757 - turnip: fix a sample shading case
2758 - turnip: fix renderpass gmem configs when there are too many attachments
2759 - turnip: set the API version
2760 - turnip: move enum translation functions to a common header
2761 - freedreno/a6xx: VSC "STRM_ARRAY_PITCH" is "STRM_LIMIT"
2762 - freedreno/a6xx: remove unnecessary OVERFLOW_FLAG_REG check
2763 - turnip: remove unnecessary OVERFLOW_FLAG_REG check
2764 - freedreno/a4xx: restore pitch to bytes change to layout code
2765 - freedreno/a4xx: simplify setup_slices
2766 - turnip: rework streamout state and add missing counter buffer read/writes
2767 - turnip: refactor CmdDraw* functions (and a few fixes)
2768 - turnip: enable VK_EXT_index_type_uint8
2769 - turnip: implement CmdDrawIndirectByteCountEXT
2770 - turnip: fix ts_cs_memory typo
2771 - turnip: use pipeline cs for shader programs instead of separate bo
2772 - freedreno/registers: a6xx depth bounds test registers
2773 - turnip: implement depthBounds
2774 - turnip: translate CreateRenderPass to CreateRenderPass2
2775 - turnip: replace a memset(0) with zalloc in CreateRenderPass
2776 - turnip: use RenderPassCreateInfo for render_pass_add_implicit_deps
2777 - turnip: move some logic out of create_render_pass_common
2778 - turnip: implement VK_EXT_vertex_attribute_divisor
2779 - turnip: fix empty scissor case
2780 - turnip: fix update_stencil_mask
2781 - turnip: disable early_z for VK_FORMAT_S8_UINT
2782 - freedreno/registers: add CP_DRAW_INDIRECT_MULTI
2783 - freedreno/ir3: add support for load_draw_id
2784 - turnip: implement VK_KHR_shader_draw_parameters
2785 - turnip: fix VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VULKAN_1_1_FEATURES
2786 - turnip: fix huge scissor min/max case
2787 - freedreno/ir3: fix resinfo wrmask
2788 - freedreno/regs: add extra bits for UBWC array pitch
2789 - turnip: enable largePoints
2790 - turnip: enable depthBiasClamp
2791 - freedreno/registers: update varying-related registers
2792 - freedreno/a3xx: support LINEAR_PIXEL/PERSP_CENTROID/LINEAR_CENTROID sysvals
2793 - freedreno/a4xx: fake LINEAR_PIXEL varying support for u_blitter
2794 - freedreno/ir3: add generic get_barycentric()
2795 - freedreno/a5xx: set missing bary sysvals
2796 - freedreno/a6xx: set missing bary sysvals
2797 - turnip: set missing bary sysvals
2798 - freedreno/ir3: add support for INTERP_MODE_NOPERSPECTIVE
2799 - turnip: make tiling config part of framebuffer state
2800 - turnip: rework render_tiles loop
2801 - turnip: vsc improvements
2802 - turnip: fix tess param bo size calculation
2803 - turnip: clear_blit: pass aspect mask to setup function
2804 - turnip: support multi-image layouts
2805 - turnip: enable 420_UNORM formats
2806 - freedreno/layout: fix explicit layout offset not added to slice offset
2807 - freedreno/ir3: fix/rework tess levels
2808 - Revert "nir: Add an option for lowering TessLevelInner/Outer to vecs"
2809 - Revert "nir: Support sysval tess levels in SPIR-V to NIR"
2810 - freedreno/regs: document SS6_UBO state src
2811 - turnip: use global bo for clear blit shaders
2812 - freedreno/ir3: add support for a650 tess shared storage
2813 - freedreno/regs: document CS shared storage size bit
2814 - freedreno/a2xx: fix compressed textures
2815 - freedreno: add a fd_resource_pitch helper
2816 - freedreno/layout: layout simplifications and pitch from level 0 pitch
2817 - turnip: fix active_desc_sets not being set for compute pipeline
2818 - freedreno/ir3: fix setup_input for sparse vertex inputs
2819 - freedreno/ir3: run nir_opt_loop_unroll in optimization loop
2820 - freedreno: fix layout pitchalign field not being set for imported buffers
2821 - freedreno/regs: update primitive output related registers
2822 - turnip: clean up primitive output state
2823 - turnip: drop GS clear path
2824 - turnip: use DIRTY SDS bit to avoid making copies of pipeline load state ib
2825 - turnip: emit compute pipeline directly in CmdBindPipeline
2826 - turnip: fix inconsistencies with tu6_load_state_size
2827 - turnip: remove use of tu_cs_entry for draw states
2828 - gitlab-ci: re-enable arm64_a630_vk
2829 - freedreno/regs: update a6xx GRAS registers
2830 - freedreno/regs: update a6xx RB regs
2831 - freedreno/regs: update a6xx VPC regs
2832 - freedreno/regs: update a6xx PC regs
2833 - turnip: disable tiling for NV12/IYUV formats
2834 - turnip: remove extra gmem alignment
2835 - freedreno/ir3: fix wrong local_primitive_id_start type
2836 - turnip: move WFI out of draw state to fix a650 hangs
2837 - turnip: use patchControlPoints for HS_INPUT_SIZE value
2838 - turnip: fix SP_HS_UNKNOWN_A831 value for A650
2839 - turnip: workaround for a630 d24_unorm_s8_uint fails
2840 - turnip: fix sysmem CmdClearAttachments 3D fallback breaking GMEM path flush
2841 - turnip: delete tu_clear_sysmem_attachments_2d
2842 - turnip: add support for D32_SFLOAT_S8_UINT
2843 - turnip: rework extended formats to allow more extended formats
2844 - util/format: translate A4R4G4B4_UNORM and A4B4G4R4_UNORM vulkan formats
2845 - turnip: implement VK_EXT_4444_formats
2849 - intel/dev: Split .num_subslices out of GEN12_FEATURES macro
2850 - intel/dev: Add device info for RKL
2851 - intel/l3: Don't rely on cfg entry URB size being 0 as a sentinal
2852 - intel/l3: Allow platforms to have no l3 configurations
2853 - iris/l3: Enable L3 full way allocation when L3 config is NULL
2854 - anv: Set L3 full way allocation at context init if L3 cfg is NULL
2855 - intel/dev: Add device info for DG1
2856 - iris: Make use of devinfo has_aux_map field
2857 - anv: Make use of devinfo has_aux_map field
2858 - anv/pipeline: Split VFE/INTERFACE_DESCRIPTOR out to emit_media_cs_state
2859 - anv/cmd_buffer: Split GPGPU_WALKER out to emit_gpgpu_walker
2860 - iris: Split walker and state update into iris_upload_gpgpu_walker
2861 - iris/compute: Split out iris_load_indirect_location
2862 - intel/compiler/cs: Allow simd32 in some more cases with no8 and/or no16
2863 - intel/compiler/fs: Still attempt simd32 when INTEL_DEBUG=no16 is used
2864 - iris: Add missing break in switch in modifier_is_supported
2865 - anv, iris: Set MediaSamplerDOPClockGateEnable for gen12+
2869 - v3d: Fix swizzle in DXT3 and DXT5 formats
2870 - v3d: Include supported DXT formats to enable s3tc/dxt extensions
2871 - vc4: don't relay on intr->num_components for non-vectorized intrinsics
2872 - nir: only uniforms with dynamically_uniform offset are dynamically_uniform
2876 - anv: Remove RANGE_SIZE usage
2877 - radv: Remove RANGE_SIZE usage
2878 - turnip: Remove RANGE_SIZE usage
2879 - vulkan: Update Vulkan XML and headers to 1.2.140
2880 - radv: Implement VK_EXT_custom_border_color
2881 - radeonsi: Use TRUNC_COORD on samplers
2882 - radv: Implement VK_EXT_4444_formats
2886 - glthread: Add GLAPIENTRY to _mesa_marshal_MultiDrawArrays.
2887 - appveyor: Upgrade pip.
2888 - appveyor: Use Python3.
2892 - nir/deref: copy ptr_stride when rematerializing
2893 - nir/validate: validate the stride for deref_ptr_as_array
2894 - Revert "nir/validate: validate the stride for deref_ptr_as_array"
2895 - nvir/nir: use component helpers instead of insn->num_components
2896 - st/mesa: lower images when needed
2897 - nir/lower_images: fix for array of arrays
2898 - nir/lower_images: handle dec and inc
2899 - nv50/ir/nir: move away from image_deref intrinsics
2900 - nv50/ir/nir: handle image atomic inc and dec
2901 - nv50/ir/nir: remove image uniform hack
2902 - gv100/ir: fix atom cas
2903 - gv100/ir: fix shift lowering
2904 - gv100/ir: fix OP_TXG for shadow textures
2905 - nv50/ir/nir: add workaround for double vertex attribs
2906 - nv50/ir/print: add missing VIEWPORT_MASK handling
2907 - nv50/ir/nir: fix ext_demote_to_helper_invocation
2908 - nv50/ir/nir: fix nv_viewport_array2
2909 - nvc0: enable spirv caps with nir
2910 - nv50/ir/nir: don't emit a restart with set a stream_id
2911 - nv50/ir/nir: handle clip vertex for tess eval shaders
2912 - nv50/ir/nir: rework input output handling
2913 - nv50/ir/nir: rework CFG handling
2914 - nv50/ir/ra: convert some for loops to Range-based for loops
2915 - nv50/ir/ra: fix memory corruption when spilling
2916 - nv50/ir/nir: fix interpolation on explicit operations
2917 - gv100/ir: implement sample shading
2918 - gv100/ir: fix coherent and volatile memory access
2919 - nv50/ir/nir: fix cache mode conversion
2920 - nv50/ir: fix memset on non trivial types warning
2921 - nv50/ir/tgsi: move call to tgsi_scan_shader inside Source constructor
2922 - nvc0: set local mem size for compute on gv100
2923 - nvc0: set sampler index mode to independently on gv100 compute
2924 - gv100/ir: set ftz bit on floating point operations
2925 - ci: bump libdrm to 2.4.102
2926 - nouveau: enable HMM
2927 - gallium: add PIPE_CAP_RESOURCE_FROM_USER_MEMORY_COMPUTE_ONLY
2928 - nvc0: support PIPE_CAP_RESOURCE_FROM_USER_MEMORY_COMPUTE_ONLY
2929 - nouveau: expose HMM
2930 - ci: need to install wget in order to download libdrm
2931 - ci: bump libdrm to 2.4.102
2932 - nouveau: enable HMM
2933 - gallium: add PIPE_CAP_RESOURCE_FROM_USER_MEMORY_COMPUTE_ONLY
2934 - nvc0: support PIPE_CAP_RESOURCE_FROM_USER_MEMORY_COMPUTE_ONLY
2935 - nouveau: expose HMM
2936 - st/mesa: fix st_CopyPixels without support for stencil exports
2937 - nv50/ir/tgsi: silence warning about unhandled GS_INPUT_PRIM property
2938 - nv50/ir: initialize persampleInvocation to false
2939 - nir/lower_io: assert that offsets are used for shader_in
2940 - nv50/ir/nir: fix global_atomic_comp_swap
2941 - spirv: extract switch parsing into its own function
2945 - iris: Include linux/sync_file.h instead of cut and pasting contents
2946 - anv: Include linux/sync_file.h instead of cut and pasting contents
2947 - iris: Rename iris_syncpt to iris_syncobj for clarity.
2948 - iris: Give up on not passing ice to iris_init_batch
2949 - iris: Destroy transfer slab after batches
2950 - iris: Flush any current work in iris_fence_await before adding deps
2951 - intel: Move anv_gem_supports_syncobj_wait to common code.
2952 - iris: Detect DRM_SYNCOBJ_WAIT_FLAGS_WAIT_FOR_SUBMIT kernel support
2953 - iris: Implement PIPE_FLUSH_DEFERRED support.
2954 - intel: Delete hardcoded devinfo->urb.size values for Gen7+ (sans DG1).
2955 - iris: Delete useless #define
2956 - intel/eu: Add a brw_urb_desc helper
2957 - CI: Disable Panfrost Mali-T820, Lima Mali-400 and Lima Mali-450 jobs
2958 - intel: Disable loading drivers on DG1 devices for now
2959 - nir: Fix divergence analysis for tessellation input/outputs
2960 - iris: Implement pipe->texture_subdata directly
2961 - iris: Fix CCS check in iris_texture_subdata().
2962 - iris: Delete shader variants when deleting the API-facing shader
2963 - iris: Reorder the loops in iris_fence_await() for clarity.
2964 - iris: Drop stale syncobj references in fence_server_sync
2968 - freedreno/ir3: Pass stream output info to ir3_shader_from_nir
2969 - freedreno/ir3: Rename ir3_nir_lower_to_explicit_io
2970 - freedreno/ir3: Add ir3_nir_lower_to_explicit_input() pass
2971 - freedreno/ir3: Lower GS builtins before lowering IO
2972 - freedreno/ir3: Drop hack to clean up split vars
2973 - freedreno/fdl: Align after dividing by block size
2974 - freedreno/a6xx: Set tfetch correctly for compressed formats
2975 - freedreno/ir3: Drop wrmask for ir3 local and global store intrinsics
2976 - freedreno/a6xx: Create shader dependent streamout state at compile time
2977 - freedreno/a6xx: Map inputs to VFD entries up front
2978 - freedreno/a6xx: Allocate ringbuffer based on VFD count
2979 - freedreno/a6xx: Emit VFD setup as array writes
2980 - freedreno/a6xx: Avoid stalling for occlusion queries
2981 - freedreno: Use the right amount of &'s
2982 - freedreno: Use explicit \*_NONE enum for undefined formats
2983 - turnip: Use hw enum when emitting A6XX_RB_STENCIL_CONTROL
2984 - turnip: Use tu6_reduction_mode() to avoid warning
2985 - turnip: Use {} initializer to silence warning
2986 - freedreno/ir3: Avoid {0} initializer for struct reginfo
2987 - src/util: Remove out-of-range comparison
2988 - mapi: Fix a couple of warning in generated code
2989 - mesa/st: Use memset to zero out struct
2990 - egl/android: Move get_format under HAVE_DRM_GRALLOC guard where it's used
2991 - egl/android: Drop unused variable
2992 - freedreno/a6xx: Move per element offset to VFD_DECODE
2993 - freedreno/a6xx: Decouple VFD_FETCH and VFD_DECODE
2994 - freedreno/a6xx: Create stateobj for VFD_DECODE
2995 - freedreno/a6xx: Program VFD_DEST_CNTL from program stateobj
2996 - freedreno/a6xx: Turn on robustness extensions
2997 - docs/features.txt: Update for freedreno
2998 - freedreno/a6xx: Fix VFD_CONTROL emit
2999 - freedreno/a6xx: Don't write REG_A6XX_RB_SRGB_CNTL in restore
3000 - freedreno/a6xx: Set index buffer size to bo size
3001 - freedreno: Handle DRM_FORMAT_MOD_INVALID in shared code
3002 - turnip: Put VK_KHR_external_fence_fd stubs back
3003 - freedreno/a6xx: Don't blit with R2D_RAW
3004 - freedreno/a6xx: Move fd6_ifmt into fd6_blitter.c
3005 - freedreno/a6xx: Split out src and dst setup helpers for blit
3006 - freedreno/a6xx: Don't set unknown bit when tiling differs
3007 - freedreno/a6xx: Set src and dst rects outside blit loop
3008 - freedreno/a6xx: Program SP_2D_SRC_FORMAT outside blit loop
3009 - freedreno/a6xx: Consolidate computing blit_cntl
3010 - freedreno/a6xx: Don't emit src state when clearing
3011 - freedreno/a6xx: Separate stencil sysmem clear fix
3012 - freedreno/a6xx: Enable FMT6_10_10_10_2_UNORM blitting
3013 - freedreno/a6xx: Make blit_control helper a little more helpful
3014 - freedreno/a6xx: Program A6XX_SP_2D_SRC_FORMAT_COLOR_FORMAT based on dst format
3015 - freedreno/a6xx: Move REG_A6XX_SP_2D_SRC_FORMAT programming to helper
3016 - freedreno/a6xx: Move CP_SET_MARKER to setup helper
3017 - freedreno/a6xx: Program RB_UNKNOWN_8C01 in setup helper
3018 - freedreno/a6xx: Don't take pipe_blit_info in emit_blit_dst
3019 - freedreno/a6xx: Split clear and blit texture into different functions
3020 - freedreno/registers: Rename SP_2D_SRC_FORMAT
3021 - turnip: Move device enumeration and feature discovery to tu_drm.c
3022 - turnip: Move tu_bo functions to tu_drm.c
3023 - turnip: Collapse some tu_drm wrappers
3024 - turnip: Move remaining drm code to tu_drm.c
3025 - turnip: Only include msm_drm in tu_drm.c
3026 - egl/android: Remove unused variable
3027 - mapi/test: Change type to unsigned for offset
3028 - gallium: Switch u_debug_stack/symbol.c to util/hash_table.h
3029 - util: Move stack debug functions to src/util
3030 - util: Add unit test for stack backtrace caputure
3031 - gallium/android: Rewrite backtrace helper for android
3032 - ci: Include enough Android headers to let us compile test EGL
3033 - mapi: Mark TLS symbols as optional in glapi-symbols.txt
3034 - turnip: Make tu_android.c compile again
3035 - meson: Define ANDROID and ANDROID_API_LEVEL when compiling for Android
3036 - anv: Pass device to setup_gralloc0_usage for error reporting
3037 - anv: Add stub for anv_gem_get_tiling() for Android
3038 - vulkan: Allow global symbol HMI for Android
3039 - radv/android: Remove unused variable
3040 - ci: Add a build test for the Android platform
3044 - gallium/swr: Fix building swr with MSVC
3048 - docs: include meson in the toctree
3049 - docs: Remove version.
3050 - docs: Add the favicon to the new page.
3054 - radeon/vcn: reset the decode flags from message buffer
3055 - radeon/vcn: add Sienna to use internal register offset
3056 - radeon/vcn/dec: add db_aligned_height to message buffer
3060 - mapi: x86: Fix dynamic entries in x86 tsd stubs.
3061 - mapi: Return NULL function pointers for GL_EXT_debug_marker
3062 - egl: Allow software rendering for vgem/virtio_gpu in platform_device
3066 - drm-shim: move handle lock to shim_fd
3067 - drm-shim: don't create a memfd per BO
3068 - drm-shim: silence warnings
3069 - intel/dev: print out error when platform is not found by name
3070 - intel: add stub_gpu tool
3071 - ci: Add intel to shaderdb runs
3072 - iris: don't assert on unfinished aux import in copy paths
3073 - anv: don't expose VK_INTEL_performance_query without kernel support
3074 - anv: fix alignments for uniform buffers
3075 - genxml: run sorting script
3076 - genxml: fix invalid end value for video fields
3077 - genxml: factor out utility functions
3078 - genxml: pack: deal with default field not being simple integers
3079 - intel/genxml: fix bits generation for MI_LOAD_REGISTER_IMM
3080 - intel/mi-builder: add framework for self modifying batches
3081 - anv: don't reserve a particular register for draw count
3082 - anv: add a new execution mode for secondary command buffers
3083 - intel/genxml: add PIPE_CONTROL command cache invalidate bit
3084 - intel/perf: make pipeline statistic query loading optional
3085 - intel/perf: store the appropriate OA formats in queries
3086 - intel/perf: update generated code to ralloc all data
3087 - intel/perf: create a unique list of counters
3088 - intel/perf: compute number of passes for a set of counters
3089 - intel/perf: emit counter units in generated code
3090 - intel/perf: add helper to compute metrics from counters
3091 - intel/perf: add counter category to generated code
3092 - intel/perf: report whether the platform supported
3093 - anv: use a query filled by the perf code
3094 - intel/perf: reuse offset specified in the query
3095 - anv: Implement VK_KHR_performance_query
3096 - intel/perf: repurpose INTEL_DEBUG=no-oaconfig
3097 - anv: fixup unwinding of device create failure
3098 - blorp: rename workaround address function
3099 - anv: store the workaround address
3100 - iris: store workaround address
3101 - i965: store workaround_bo offset
3102 - intel: add identifier for debug purposes
3103 - iris: add identifier BO
3104 - i965: add identifier BO
3105 - anv: add identifier BO
3106 - intel/aub_error_decoder: print driver identifier if found
3107 - iris: fix BO destruction in error path
3108 - i965: don't forget to set screen on duped image
3109 - iris: fix export of GEM handles
3110 - i965: fix export of GEM handles
3111 - anv: add an option to disable secondary command buffer calls
3112 - anv: garbage collect timeline semaphore when querying value
3113 - iris: fix fallback to swrast driver
3114 - anv: fix uninitialized variable access
3115 - anv: properly handle fence import of sync_fd = -1
3116 - anv: fix descriptor set free
3117 - anv: fix incorrect realloc failure handling
3118 - anv: centralize vk to gen arrays
3119 - anv: fix up dynamic clip emission
3120 - anv: don't fail userspace relocation with perf queries
3121 - anv: fix transform feedback surface size
3122 - anv: VK_INTEL_performance_query interaction with VK_EXT_private_data
3123 - intel/perf: store query symbol name
3124 - intel/perf: fix raw query kernel metric selection
3125 - intel/compiler: fixup Gen12 workaround for array sizes
3129 - util: Make process_test path compatible with mingw native toolchains
3131 Louis-Francis Ratté-Boulianne (1):
3133 - nir: Always create UBO variable when lowering uniforms to ubo
3137 - etnaviv: generalize FE stall before loading shader and sampler states
3138 - etnaviv: retarget transfer to render resource when necessary
3139 - etnaviv: don't expose timer queries
3143 - dri2: dri2_make_current() fold multiple if blocks
3144 - dri2: do not conflate unbind and bindContext() failure
3145 - egl/dri2: try to bind old context if bindContext failed
3149 - i965: remove unused variable
3150 - glsl_to_tgsi: add fallthrough comments
3151 - glsl: cleanup vertex shader input checks
3152 - iris: remove unused iris_bo->swizzle_mode
3153 - intel/compiler: fix Android build
3154 - st/mesa: fix reporting of float perf counters max value
3155 - iris: return max counter value for AMD_performance_monitor
3156 - iris: remove iris_monitor_config
3157 - intel/perf: move query_mask and location out of gen_perf_query_counter
3158 - iris: propagate error from gen_perf_begin_query to glBeginPerfQueryINTEL
3159 - i965: propagate error from gen_perf_begin_query to glBeginPerfQueryINTEL
3160 - util: fix possible fd leaks in os_socket_listen_abstract
3161 - glsl: catch out of bounds access in the debug version
3162 - util: fix possible buffer overflow in util_get_process_exec_path
3163 - util/format: initialize non-important components to 0
3164 - mesa: fix out of bounds access in glGetFramebufferParameterivEXT
3165 - mesa: quiet down static analyzers
3166 - iris: quiet down static analyzers
3167 - intel/vec4: fix out of bounds read
3168 - intel/perf: fix performance counters availability after glFinish
3169 - anv: refresh cached current batch bo after emitting some commands
3170 - anv: fix minor gen_ioctl(I915_PERF_IOCTL_CONFIG) error handling issue
3171 - intel/perf: split load_oa_metrics
3172 - intel/perf: export performance counters sorted by [group|set] and name
3176 - mesa: optimize glPush/PopClientAttrib by removing malloc overhead
3177 - mesa: don't call _mesa_update_state for _mesa_get_clamp_fragment_color
3178 - mesa: don't set unnecessary program flags in _mesa_update_state
3179 - mesa: don't update shaders on fixed-func state changes if user shaders are bound
3180 - mesa,st/mesa: add a fast path for non-static VAOs
3181 - mesa: inline vbo_context inside gl_context to remove vbo_context dereferences
3182 - mesa: add glInternalBufferSubDataCopyMESA for glthread
3183 - mesa: add _mesa_InternalBind{ElementBuffer,VertexBuffers} for glthread
3184 - glthread: do glBufferSubData as unsynchronized upload + GPU copy
3185 - glthread: don't use atomics for refcounting to decrease overhead on AMD Zen
3186 - glthread: track pointers and strides for Pointer & EXT_dsa attrib functions
3187 - glthread: track instance divisor changes
3188 - glthread: track primitive restart state
3189 - glthread: initialize VAOs properly
3190 - glthread: handle POS vs GENERIC0 aliasing
3191 - glthread: handle gl{Push,Pop}ClientAttrib{DefaultEXT} for glthread states
3192 - glthread: upload non-VBO vertices and indices for non-Indirect non-IBM draws
3193 - tgsi_to_nir: handle TGSI_SEMANTIC_BLOCK_SIZE
3194 - tgsi_to_nir: handle TGSI_OPCODE_BARRIER
3195 - radeonsi: unify and align down the max SSBO/TBO/UBO buffer binding size
3196 - radeonsi: clean up and deduplicate code around internal compute dispatches
3197 - radeonsi: bind shader images after DCC is disabled for image stores
3198 - radeonsi: add SI_IMAGE_ACCESS_DCC_OFF to ignore DCC for shader images
3199 - radeonsi: implement and use compute-based DCC decompression on gfx9-10
3200 - radeonsi: add a workaround to fix KHR-GL45.texture_view.view_classes on gfx9
3201 - radeonsi: fix si_compute_clear_render_target with render condition enabled
3202 - radeonsi: revert an accidental change in si_clear_buffer
3203 - Revert "ac/surface: remove RADEON_SURF_TC_COMPATIBLE_HTILE and assume it's always set"
3204 - Revert "ac: reassociate FP expressions for inexact instructions for radeonsi"
3205 - ac/surface: fix MSAA crash with FORCE_SWIZZLE_MODE on gfx9
3206 - radeonsi: don't wait for idle at the end of gfx IBs
3207 - ac/surface: unset RADEON_SURF_TC_COMPATIBLE_HTILE if HTILE hasn't been computed
3208 - radeonsi/gfx9: always use IMG_DATA_FORMAT_S8_32 for 8-bit stencil
3209 - radeonsi: allow tc_compatible_htile to be mutable
3210 - radeonsi: enable TC-compatible HTILE on demand for best Z/S performance
3211 - tgsi_to_nir: translate non-vec4 image stores correctly
3212 - radeonsi: fix compilation of monolithic PS
3213 - amd: update amdgpu_drm.h
3214 - amd: remove duplicated definitions from amdgpu_drm.h
3215 - amd: assume CMASK is always rb/pipe_aligned, remove ac_surface.u.gfx9.cmask
3216 - amd: assume HTILE is always rb/pipe_aligned, remove ac_surface.u.gfx9.htile
3217 - ac/surface,radeonsi: move the set/get_bo_metadata code to ac_surface.c
3218 - ac/surface,radeonsi: move the set/get_umd_metadata code into ac_surface.c
3219 - amd: unify code for overriding offset and stride for imported buffers
3220 - ac/surface: override all offsets including metadata offsets
3221 - ac/surface: fix broken pitch override on gfx8
3222 - gallium: rename 'state tracker' to 'frontend'
3223 - gallium: change comments to remove 'state tracker'
3224 - gallium: rename PIPE_RESOURCE_FLAG_ST_PRIV to FRONTEND_PRIV
3225 - gallium: remove more "state tracker" occurences
3226 - radeonsi: also enable tgsi_to_nir caching for compute shaders
3227 - glthread: stop using GLenum16 to get correct GL errors for out-of-bounds enums
3228 - radeonsi: don't expose 16xAA on chips with 1 RB due to an occlusion query issue
3229 - ac/nir: honor ACCESS_STREAM_CACHE_POLICY for L1 and L0 caches too
3230 - radeonsi: use correct clear value size for EQAA in expand_fmask
3231 - radeonsi: optimize access pattern for compute blits with linear textures
3232 - radeonsi: tweak clear/copy_buffer limits when to use compute
3233 - radeonsi: simplify setting resource usage for si_init_temp_resource_from_box
3234 - radeonsi: rename SI_RESOURCE_FLAG_TRANSFER to FORCE_LINEAR
3235 - radeonsi: use vi_dcc_enabled instead of using tex->surface.dcc_offset directly
3236 - radeonsi: use display_dcc_offset for setting displayable_dcc_cb_mask
3237 - winsys/amdgpu: add RADEON_FLAG_UNCACHED for faster blits over PCIe
3238 - radeonsi: disable the L2 cache for most CPU mappings of textures
3239 - radeonsi: disable the L2 cache for CPU read mappings of buffers
3240 - radeonsi: compute perf tests - don't test 1 wave/SA limit, test no limit first
3241 - radeonsi: test uncached clear/copy buffer performance with compute shaders
3242 - gallium/u_threaded: execute transfer_unmap with THREAD_SAFE directly
3243 - ac/gpu_info: compute the best safe IB alignment
3244 - ac/surface: don't compute single-sample CMASK if it's unaligned
3245 - radeonsi: don't use INDIRECT_BUFFER within IBs
3246 - radeonsi: decrease the max GS invocation count to 32
3247 - Revert "radeonsi: don't wait for idle at the end of gfx IBs"
3248 - ac: update register and packet definitions for preemption
3249 - radeonsi: move resetting tracked registers into a new function
3250 - radeonsi: split si_all_descriptors_begin_new_cs and rename functions
3251 - radeonsi: don't enable TC-compatible HTILE for stencil if stencil doesn't use it
3252 - radeonsi/gfx8: enable TC-compatible HTILE from the beginning as before
3253 - radeonsi: don't hardcode most perf counter block counts
3254 - ac/gpu_info: replace num_good_cu_per_sh with min/max_good_cu_per_sa
3255 - amd: replace SH -> SA (shader array) in comments
3256 - radeonsi/gfx10: implement most performance counters
3257 - glthread: don't upload for glDraw inside a display list and always sync
3258 - nir: add i2imp and u2ump opcodes for conversions to mediump
3259 - nir: add int16 and uint16 type helpers
3260 - nir: lower int16 and uint16 in nir_lower_mediump_outputs
3261 - nir: fix lower_wpos for 16-bit fddy
3262 - nir: add options::vectorize_vec2_16bit to limit vectorization to vec2 16
3263 - glsl: treat lowp as mediump when lowering builtins
3264 - glsl: handle int16 and uint16 types and add instructions for mediump
3265 - glsl: lower mediump integer types to int16 and uint16
3266 - glsl: lower mediump partial derivatives
3267 - glsl: lower the precision of imageLoad
3268 - glsl: lower samplers with highp coordinates correctly
3269 - gallium: add shader caps INT16 and FP16_DERIVATIVES
3270 - ac: rename has_double_rate_fp16 -> has_packed_math_16bit
3271 - ac/nir: use more types from ac_llvm_context
3272 - ac/nir: support vector types in the type suffix of overloaded intrinsics
3273 - ac/nir: remove type and num_channels args from ac_build_buffer_store_common
3274 - ac/nir: support 16-bit data in buffer_load_format opcodes
3275 - ac/nir: support 16-bit data in image opcodes
3276 - ac/nir: handle nir_op_[fiu]2[fiu]mp opcodes
3277 - ac/nir: select v_cvt_pkrtz for all conversions from f32 to f16 for radeonsi
3278 - ac/nir: set the second v_cvt_pkrtz argument to undef if it's unused
3279 - ac/nir: support v2f16 derivatives
3280 - nir: don't count samplers and images in interface blocks
3281 - nir: gather which images are buffers
3282 - nir: gather which images are MSAA
3283 - radeonsi: remove unused leftover code for INDIRECT_BUFFER inside IBs
3284 - radeonsi: remove const_buffers_declared hacks
3285 - radeonsi: pass at most 3 images and/or shader buffers via user SGPRs for compute
3286 - radeonsi: add a hack to disable TRUNC_COORD for shadow samplers
3287 - gallium/u_vbuf: get rid of some pointer dereferences
3288 - gallium/u_vbuf: add a faster path for uploading non-interleaved attribs
3289 - glthread: sync in glFlush for multiple contexts
3290 - radeonsi: enable ARB_sparse_buffer
3291 - ac,radeonsi: replace == GFX10 with >= GFX10 where it's needed
3292 - ac,radeonsi: start adding support for gfx10.3
3293 - ac/surface: add displayable DCC code for gfx10.3
3294 - radeonsi: honor a user-specified pitch on gfx10.3
3295 - radeonsi: enable larger SDMA clears and copies on gfx10.3
3296 - radeonsi: implement R9G9B9E5 render target and image store support on gfx10.3
3297 - radeonsi: move L2_CACHE_CONTROL registers into si_emit_framebuffer_state
3298 - radeonsi: set BIG_PAGE fields on gfx10.3
3299 - radeonsi: don't set any XNACK options on gfx10.3
3300 - ac: align num_vgprs for gfx10.3
3301 - radeonsi: add support for Sienna Cichlid
3302 - radeonsi: require LLVM 11 for gfx10.3
3303 - ac/surface: don't recompute the DCC retile map for imported textures
3304 - amd/addrlib: don't recompute DCC info for every ComputeDccAddrFromCoord call
3305 - amd/addrlib: remove unused members of ADDR2_COMPUTE_DCC_ADDRFROMCOORD_INPUT
3306 - ac/surface: add a wrapper structure to hold ADDR_HANDLE
3307 - ac/surface: cache DCC retile maps (v2)
3308 - amd/addrlib: fix the C++ one definition rule violation
3309 - ac/surface: don't set is_displayable if displayable DCC is missing
3310 - ac/surface: require that gfx8 doesn't have DCC in order to be displayable
3311 - ac/surface: enable DCC for the first level in the mip tail on gfx10
3312 - ac/surface: don't free dcc_retile_map on failure
3313 - radeonsi: compact MRTs to save PS export memory space
3314 - ac/nir: fix 64-bit division for GL CTS
3315 - glapi: fix incorrect param names in ARB_vertex_attrib_binding functions
3316 - glthread: rename non_vbo_attrib_mask -> user_buffer_mask, attribs -> buffers
3317 - glthread: handle ARB_vertex_attrib_binding
3318 - radeonsi: don't wait for idle at the end of gfx IBs
3319 - radeonsi: replace ctx->screen with sscreen in si_flush_gfx_cs
3320 - glsl,driconf: add allow_glsl_120_subset_in_110 for SPECviewperf13
3321 - driconf: add workarounds for SPECviewperf13
3322 - amd: add proper definitions for NOP packets
3323 - ac,winsys/amdgpu: align IBs the same as the kernel
3324 - radeonsi: don't add the border color buffer into the init_config state
3325 - radeonsi: rename init_config states to cs_preamble states
3326 - radeonsi: don't add the tess ring buffers into the cs_preamble state
3327 - radeonsi: make wait_mem_scratch unmappable
3328 - radeonsi: disallow adding BOs into si_pm4_state except 1 shader BO per state
3329 - radeonsi: make si_pm4_cmd_begin/end static and simplify all usages
3330 - radeonsi: clear per-context buffers at the end of si_create_context
3331 - radeonsi: remove tabs
3332 - radeonsi: don't flush in fence_server_sync
3333 - ac/gpu_info: fix num_physical_sgprs_per_simd for gfx10
3334 - radeonsi: fix NGG culling for Wave64
3335 - radeonsi: always use Wave32 for GS fast launch, because Wave64 hangs
3336 - radeonsi: always use Wave64 for HS/GS/VS shader stages (except GS fast launch)
3337 - radeonsi: don't try to enable NGG culling for GS
3338 - radeonsi: add a debug option to enable NGG culling for tessellation
3339 - glsl: make print_type non-static for debugging
3340 - glsl: print precision qualifiers in IR dumps
3341 - glsl: print constant initializers
3342 - glsl: fix the type of ir_constant_data::u16
3343 - glsl: fix evaluating float16 constant expression matrices
3344 - glsl: run validate_ir_tree if GLSL_VALIDATE=1 regardless of the build config
3345 - glsl: validate more stuff
3346 - glsl: convert reusable lower_precision util code into helper functions
3347 - glsl: remove the return type from lower_precision
3348 - glsl: cleanups in lower_precision
3349 - glsl: flatten a tautological conditional in lower_precision
3350 - glsl: don't lower precision of textureSize
3351 - glsl: don't lower builtins to mediump that don't allow it
3352 - glsl: lower builtins to mediump that ignore precision of certain parameters
3353 - glsl: lower builtins to mediump that always return mediump or lowp
3354 - glsl: add capability to lower mediump array types
3355 - glsl: lower mediump temporaries to 16 bits except structures (v2)
3356 - gallium: add PIPE_SHADER_CAP_GLSL_16BIT_TEMPS for LowerPrecisionTemporaries
3357 - Revert "ac/surface: require that gfx8 doesn't have DCC in order to be displayable"
3358 - glsl: don't validate array types in ir_dereference_variable
3359 - radeonsi: prevent a gfx10_ngg_calculate_subgroup_info failure for TES+NGG GS
3360 - radeonsi: add missing initialization of registers
3361 - radeonsi/gfx10: set the correct value for OFFCHIP_BUFFERING
3362 - radeonsi: sort registers in si_emit_initial_compute_regs according to GPU gen
3363 - radeonsi: sort registers in si_init_cs_preamble_state according to GPU gen
3364 - ac: add helper ac_get_register_name
3365 - ac: add tables for CP register shadowing
3366 - winsys/amdgpu: make amdgpu_bo_unmap non-static
3367 - radeonsi: make cs_preamble_state optional
3368 - radeonsi: reorder code in update_gs_ring_buffers and init_tess_factor_ring
3369 - radeonsi: implement CP register shadowing
3370 - radeonsi: add reg shadowing codepaths to GS and tess ring setup
3371 - radeonsi: add debug code for register shadowing
3372 - radeonsi: don't restore states at the beginning of IBs if they're shadowed
3373 - radeonsi: set up IBs for preemption
3374 - radeonsi: enable preemption if the kernel enabled it
3375 - amd: rename SIENNA -> SIENNA_CICHLID
3376 - amd: add support for Navy Flounder
3377 - amd: enable displayable DCC for everything newer than Navi1x
3378 - radeonsi: disable SDMA on gfx9
3379 - radeonsi: reorder NIR optimizations
3380 - radeonsi: call nir_split_array_vars/shrink_vec_array_vars/opt_find_array_copies
3381 - glsl: lower_precision - fix assertion failure with dereferences of constants
3382 - glsl: fix constant expression evaluation for 16-bit types
3383 - glsl: don't lower atomic functions to mediump
3384 - glsl: don't create conversion opcodes for array types
3385 - glsl: don't lower to mediump for desktop OpenGL
3386 - glsl: improve precision determination for calls
3387 - Revert "radeonsi: honor a user-specified pitch on gfx10.3"
3388 - radeonsi: use correct wave size in gfx10_ngg_calculate_subgroup_info
3389 - radeonsi: use the same units for esgs_ring_size and ngg_emit_size
3390 - radeonsi: increase minimum NGG vertex count requirement per workgroup on gfx 10.3
3391 - radeonsi: fix applying the NGG minimum vertex count requirement
3392 - radeonsi: don't count unusable vertices to the NGG LDS size
3393 - radeonsi: add a common function for getting the size of gs_ngg_scratch
3394 - radeonsi: remove the NGG hack decreasing LDS usage to deal with overflows
3395 - radeonsi: various fixes for gfx10.3
3396 - radeonsi: disable NGG culling on gfx10.3 because of hangs
3397 - st/mesa: don't generate NIR for ARB_vp/fp if NIR is not preferred
3398 - radeonsi: fix tess levels coming as scalar arrays from SPIR-V
3399 - gallivm: fix build on LLVM 12 due to LLVMAddConstantPropagationPass removal
3400 - ac/llvm: fix unaligned VS input loads on gfx10.3
3401 - Revert "ac: generate FMA for inexact instructions for radeonsi"
3405 - etnaviv: Disable seamless cube map on GC880
3406 - etnaviv: Remove etna_resource_get_status()
3407 - etnaviv: Add lock around pending_ctx
3411 - vulkan/wsi: Really terminate DRM lease in wsi_release_display().
3415 - st/mesa: Move _NEW_FRAG_CLAMP to NewFragClamp driver flag.
3416 - mesa: set _NEW_FRAG_CLAMP only when needed
3420 - intel/compiler: Drop opt_sampler_eot()
3421 - intel/tools: Remove unnecessary reg number checking
3422 - intel/tools: Drop srctype from ipreg
3423 - intel/tools: Require explicit regions/types for special regs
3424 - intel/tools: Disallow control subregisters > 3
3425 - intel/tools: Add assembler tests for the cr0 register
3426 - intel/compiler: Add assert that set bits are within mask
3427 - intel/compiler: Don't emit no-op cr0 changes
3428 - intel/tools: Fix typos
3429 - intel/tools: Remove stray newline
3430 - intel/tools: Don't allow empty type specifier
3431 - intel/tools: Simplify register type handling
3432 - intel/tools: Make swizzle an integer
3433 - intel/tools: Make writemask an integer
3434 - intel/tools: Simplify immediate handling
3435 - intel/tools: Simplify dstregion
3436 - intel/compiler: Relax SENDS regioning assertions
3437 - intel/tools: Pass integers, not enums, to stride()
3438 - intel/tools: Manually set ARF register file/nr/subnr
3439 - intel/tools: Don't hardcode notification register
3440 - intel/tools: Simplify notification register handling
3441 - intel/tools: Test notification subregisters
3445 - android: iris: add iris_seqno.{c,h} to Makefile.sources
3446 - freedreno/drm: android: add libfreedreno_registers static dependency
3447 - freedreno: android: add adreno-pm4-pack.xml.h generation to android build
3448 - android: util: fix build for GL4.1 support
3449 - android: svga: fix build for GL4.1 support
3450 - android: aco: add aco_ir.cpp to Makefile.sources
3451 - android: nvir/gv100: update sources in Makefile.sources
3452 - android: freedreno: add fd5_layout.c to Makefile.sources
3453 - android: freedreno/ir3: add missing generated sources and rules
3454 - android: freedreno/ir3: simplify generated sources rules
3455 - android: panfrost/encoder: add libmesa_nir static dependency
3456 - radv: fix build on Android 7 (v2)
3457 - android: freedreno/registers: fix generated headers rules
3458 - android: freedreno/ir3: fix include paths
3459 - android: freedreno/common: add support for libfreedreno_common static
3460 - android: freedreno: move a2xx disasm out of gallium
3461 - android: freedreno/common: add libmesa_git_sha1 static dependency
3465 - gitlab-ci: Use YAML anchor for llvmpipe paths in virgl rules
3466 - gitlab-ci: Update to current templates
3467 - gitlab-ci: Move down container_pre_build.sh invocation in x86_build.sh
3468 - gitlab-ci: Add Debian testing repository for x86_build image
3469 - gitlab-ci: Install WINE from Debian testing
3470 - gitlab-ci: Move lib{drm,pciaccess}-dev cross packages out of loop
3471 - gitlab-ci: Install g++-mingw-w64-x86-64-win32 instead of mingw-w64
3472 - Revert "ac,radeonsi: fix compilations issues with LLVM 11"
3473 - Revert "gallium/gallivm: fix compilation issues with llvm 11"
3474 - gitlab-ci: Enable -Werror in `meson-s390x` job
3475 - gitlab-ci: Also list arm/x86_build in needs: of test jobs
3476 - gitlab-ci: x86_test-base image as common base for x86_test-gl/vk
3477 - gitlab-ci: Pull in GCC 9 from Debian testing in x86_test-gl/vk images
3478 - gitlab-ci: Move LLVM/clang 6/7 packages to the x86_build_old image
3479 - gitlab-ci: Use Debian 10 wine-development packages
3480 - gitlab-ci: Stop using packages from Debian testing
3481 - gitlab-ci: Move meson back to x86_test-gl/vk ephemeral packages lists
3482 - gitlab-ci: Add x86_build-base docker image
3483 - gitlab-ci: Use separate docker images for cross builds
3484 - loader/dri3: Add dri3_wait_for_event_locked full_sequence out parameter
3485 - loader/dri3: Use dri3_wait_for_event_locked in loader_dri3_wait_for_msc
3486 - loader/dri3: Check for window destruction in dri3_wait_for_event_locked
3487 - gitlab-ci: Automatically run pipelines for Marge Bot pre-merge only
3488 - gitlab-ci: Use rules: instead of except:/only: for test-docs job
3489 - gitlab-ci: Extend .ci-run-policy template for docs jobs
3490 - gitlab-ci: Do not create the "success" job when the test-docs job exists
3491 - ci: Use "when: always" for pages job
3492 - ci: Move deploy stage between container & build stages
3493 - Revert "loader/dri3: Check for window destruction in dri3_wait_for_event_locked"
3494 - gitlab-ci: Remove indirect dependencies from needs:
3495 - gitlab-ci: Drop dependencies:
3496 - Revert https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4580
3497 - gitlab-ci: Fix "triggered by Marge for a merge request" rule
3498 - gitlab-ci: Only trigger test-docs job automatically for MRs
3499 - ci: Use FDO_CI_CONCURRENT in run-shader-db.sh as well
3500 - ci: Do not mark container / pages jobs as interruptible
3501 - ci: Use half as many parallel softpipe / virgl test jobs
3502 - ci: Use ignore_scheduled_pipelines anchor in .radeonsi-rules
3506 - swr: fix build with mingw
3510 - zink: explicitly zero some arrays in ntv
3511 - zink: add SpvId returns to a couple ntv functions
3512 - zink: flush active queries on destroy and free query object
3513 - zink: fix vkCmdResetQueryPool usage
3514 - zink: reset query on-demand when beginning a new query from resume
3515 - zink: always use logical eq ops in ntv with 1bit inputs
3516 - zink: track program usages for each shader
3517 - zink: emit interpolation decorations for ntv outputs
3518 - zink: handle more glsl->spirv builtin translation
3519 - zink: rework input/output location emission
3520 - zink: use '2' variants for device props/feats, check features for ext enabling
3521 - zink: add spirv builder util functions for emitting xfb decorations
3522 - zink: add spirv_builder methods for OpVectorExtractDynamic and OpVectorInsertDynamic
3523 - zink: implement streamout and xfb handling in ntv
3524 - zink: implement transform feedback support to finish off opengl 3.0
3525 - zink: set PIPE_CAP_VIEWPORT_TRANSFORM_LOWERED and remove POS special casing
3526 - zink: switch to passing VkPhysicalDeviceFeatures2 in VkDeviceCreateInfo
3527 - zink: enable xfb extension in screen creation
3528 - zink: use int assignment for vk int type
3529 - zink: use correct define value for reserved slot count in ntv
3530 - zink: clamp VkImageCreateInfo.arrayLayers to 1 for image resource creation
3531 - zink: unify code for setting resource barriers
3532 - zink: handle signed and unsigned min/max ops in ntv
3533 - zink: add ult handling for ntv
3534 - zink: add bitfield_reverse handling to ntv
3535 - zink: lower byte/word extract ops in nir
3536 - zink: handle ixor in ntv
3537 - zink: handle isign alu in ntv
3538 - zink: set lower_mul_high and lower_rotate in ntv compiler options
3539 - zink: use OpFUnordNotEqual for nir_op_fne
3540 - zink: set lower_uadd_carry in nir options
3541 - zink: implement Vk_EXT_index_type_uint8
3542 - nir: add lowering pass for clip plane enabling
3543 - st/program: use nir_lower_clip_disable instead of nir_lower_clip_vs conditionally
3544 - nir: add lowering pass for fragcolor -> fragdata
3545 - zink: translate gl_FragColor to gl_FragData before ntv to fix multi-rt output
3546 - u_prim_restart: handle user buffers in util_translate_prim_restart_ib()
3547 - nir: allow nir_lower_point_size_mov to run in geometry shader
3548 - nir: allow nir_lower_clip_halfz to run in geometry shaders
3549 - zink: rework query handling
3550 - zink: use #define for number of queries per-pool
3551 - zink: only stall during query destroy for xfb queries
3552 - zink: properly handle query pool overflows
3553 - zink: only reset query pool on query end if current batch isn't in renderpass
3554 - zink: use right vulkan type for GL_PRIMITIVES_GENERATED queries
3555 - zink: handle ntv case of nested loop instructions more permissively
3556 - zink: add lengthy comment and remove assert from discard_if ntv pass
3557 - zink: use type of src[0] for ntv store and load ops
3558 - zink: try copy_region hook for blits where we can't do a regular blit or resolve
3559 - zink: block vkCmdBlitImage usage for multi sampled blits
3560 - zink: block resolve blits for depth/stencil buffers
3561 - zink: handle empty attachments
3562 - zink: try to handle multisampled null buffers
3563 - zink: enable tgsi texcoord pipe cap
3564 - zink: destroy gfx program when a shader is freed
3565 - zink: destroy descriptor pools on context destroy
3566 - zink: free pipeline cache during program destroy
3567 - zink: free all ntv allocations after creating shader module
3568 - zink: use helper function to handle uvec/bvec types
3569 - zink: handle texelFetchOffset with offsets
3570 - zink: add some asserts for building access chains in ntv
3571 - zink: omit Lod image operand in ntv when not using an image texture dim
3572 - nir: allow lower_psiz_mov to run in tessellation stages
3573 - nir\_ allow nir_lower_clip_halfz to run in tess eval shader
3574 - u_prim_restart: handle indirect draws
3575 - zink: add extension loading framework for spirv builder
3576 - zink: implement VK_EXT_robustness2
3577 - zink: clamp PIPE_SHADER_CAP_MAX_SHADER_BUFFERS to PIPE_MAX_SHADER_BUFFERS
3578 - zink: handle VK_EXT_vertex_attribute_divisor setup
3579 - zink: store valid timestamp bits onto zink_screen
3580 - zink: implement handling for VK_EXT_calibrated_timestamps
3581 - u_prim_restart: add inline function for getting restart index based on index size
3582 - zink: reorder create_stream_output_target to fix failure case leak
3586 - docs: add some missing stuff to sourcetree.rst
3590 - iris: Drop can_fast_clear_color's format parameter
3591 - iris: Remove the CCS_D fallback
3592 - iris: Avoid fast-clear with incompatible view
3593 - iris: Disable sRGB fast-clears for non-0/1 values
3594 - intel: Add ISL_AUX_USAGE_GEN12_CCS_E
3595 - iris: Don't support sRGB + Y_TILED_CCS on gen9
3596 - iris: Use ISL_AUX_USAGE_GEN12_CCS_E on gen12
3597 - isl/drm: Support I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS
3598 - gallium/dri2: Support I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS
3599 - iris: Handle importing aux-enabled surfaces on TGL
3600 - iris: Refactor modifier_is_supported for gen12
3601 - iris: Support I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS
3602 - iris: Zero the add-on clear color BO on import
3603 - dri_util: Update internal_format to GL_RGB8 for MESA_FORMAT_B8G8R8X8_UNORM
3604 - iris: Don't call SET_TILING for dmabuf imports
3605 - gallium/dri2: Report correct YUYV and UYVY plane count
3606 - iris: Fix aux assertion in resource_get_handle
3607 - blorp: Fix alignment test for HIZ_CCS_WT fast-clears
3611 - anv: Limit vulkan version to 1.1 for Android
3612 - anv: Disable extensions based on Android versions
3613 - dri_util: Update internal_format to GL_RGB8 for MESA_FORMAT_R8G8B8X8_UNORM
3617 - util: Initialize pipe_shader_state for passthrough and transform shaders
3618 - util: Add util functionality for GL4.1 support
3619 - winsys/drm: Add GL4.1 support in drm winsys
3620 - svga/include: Headers for GL4.1 support
3621 - svga: Add GL4.1(compatibility profile) support in svga driver
3622 - svga: Performance fixes
3626 - Revert "CI: Disable Lima jobs due to lab unhealthiness"
3627 - Revert "CI: Disable Panfrost Mali-T820 jobs"
3631 - nir/scheduler: Handle nir_intrinsic_load_per_vertex_input
3632 - v3d: Remove unused member of v3d_compile
3633 - nir/schedule: Store a pointer to the scoreboard in nir_deps_state
3634 - nir/scheduler: Add an option to specify what stages share memory for I/O
3635 - v3d: Let scheduler know GS doesn’t have shared I/O memory
3636 - gallium: Add pipe cap for primitive restart with fixed index
3637 - mesa: Add PrimitiveRestartFixedIndex to gl_constants
3638 - v3d: Disable PIPE_CAP_PRIMITIVE_RESTART
3639 - v3d: Add missing macro for stvpmd instruction
3640 - v3d: Use stvpmd for non-uniform offsets in GS
3641 - compiler: Add a system value for the line coord
3642 - v3d: Implement the line coord intrinsic
3643 - nir: Add intrinsics for the line width
3644 - v3d: Handle the line width intrinsics
3645 - v3d: Add a lowering pass for line smoothing
3646 - v3d: Enable perpendicular line caps when line smoothing
3647 - broadcom/qpu: set VC5_QPU_RADDR_A out of the switch at _pack_branch
3648 - v3d/compiler: Fix sorting the gs and fs inputs
3649 - v3d/compiler: Lower geometry output store base into offset src
3650 - nir/scheduler: Move nir_scheduler to its own header
3651 - nir/schedule: Store a pointer to the options struct in scoreboard
3652 - nir/schedule: Add a callback for backend-specific dependencies
3653 - v3d: Mark scheduling dependency for prim id and first output
3654 - nir/schedule: Add an option for a fallback scheduling algorithm
3655 - v3d: Changed v3d_compile:failed to an enum
3656 - v3d: Retry with the fallback scheduler when RA fails
3660 - radv: Don't take absolute value of unsigned type.
3661 - aco: Don't declare 'Block' as class, but define as struct.
3662 - aco: Don't std::move temporary object.
3663 - aco: Use correct reference type in for-range-loop.
3664 - radv: Explicitly cast TIMESTAMP_NOT_READY value to uin32_t where needed.
3668 - ci: TRACES_DB_PATH and RESULTS_PATH defined as relative paths
3669 - ci: ArgumentParser receives the args from the main parameters
3670 - ci: Migrate tracie tests done in shell script to pytest
3671 - ci: Split test_tracie_skips_traces_without_checksum in separate cases
3672 - ci: Fix TypoError error when traces in traces.yml is an empty list
3676 - vulkan/overlay: fix crash on destroying NULL swapchain
3680 - vc4_bufmgr: fix time_t printf
3681 - pan_bo.h: add time.h include for time_t
3682 - v3d_bufmgr: fix time_t printf
3686 - clover/nir: Check the result of spirv_to_nir
3687 - clover/api: Address missing braces for subobj init
3688 - clover: Address unnecessary copy warnings
3689 - clover/spirv: Remove unused tuple header
3691 Pierre-Eric Pelloux-Prayer (62):
3693 - radeonsi: fix export count
3694 - mesa: add gl_coontext::ForceIntegerTexNearest
3695 - driconf: add force_integer_tex_nearest option
3696 - radeonsi: add workaround for issue 2647
3697 - radeonsi: don't print gs_copy_shader stats for shaderdb
3698 - glsl: init gl_FragColor if zero_init=true
3699 - glsl: rework zero initialization
3700 - glsl: add a is_implicit_initializer flag
3701 - mesa: extend GLSLZeroInit semantics
3702 - gallium: add a new cap PIPE_CAP_GLSL_ZERO_INIT
3703 - ac/nir: export some undef as zero
3704 - ac/surface: remove shadowing declaration
3705 - amdgpu/radeon: add secure api
3706 - radeonsi: add AMD_DEBUG=tmz option
3707 - radeon: add RADEON_CREATE_ENCRYPTED flag
3708 - radeonsi: allocate framebuffer texture as secure when using tmz
3709 - amdgpu: add encrypted slabs support
3710 - radeonsi: force using staging texture when uploading to secure texture
3711 - radeonsi/sdma: implement tmz support
3712 - gallium: PIPE_RESOURCE_FLAG_ENCRYPTED
3713 - radeonsi: add support for PIPE_RESOURCE_FLAG_ENCRYPTED
3714 - amdgpu: use AMDGPU_IB_FLAGS_SECURE when requested
3715 - radeonsi: determine secure flag must be set for gfx IB
3716 - radeonsi: do not use cmask with encrypted texture
3717 - amd/addrlib: fix forgotten char -> enum conversions
3718 - radeonsi: fix inversed arguments in si_test_gds_memory_management
3719 - amdgpu: fix unitialized variable
3720 - radeonsi/sdma: remove useless compare
3721 - radeonsi/drirc: enable zerovram option for 7 Days to Die
3722 - winsys/radeon: do not cast bo->va as void*
3723 - radeonsi: add return value to gfx10_ngg_calculate_subgroup_info
3724 - radeonsi/ngg: try GS multi-cycling mode if default mode failed
3725 - ac/surface: set SCANOUT if surf->is_displayable
3726 - ac/surface: fix epitch when modifying surf_pitch
3727 - ac/llvm: load 1 byte at a time if unaligned on gfx10
3728 - st/mesa: make texture views inherit compressed_data storage
3729 - radeonsi: bump SI_NUM_SHADER_BUFFERS to 32
3730 - st/mesa: do not clear NewDriverState for inactive states
3731 - glsl: reject size1x8 for image variable with floating-point data types
3732 - ac/llvm: remove the -1 hack from ac_atomic_inc_wrap
3733 - glsl: don't expose imageAtomicIncWrap for signed image
3734 - glsl: only allow 32 bits atomic operations on images
3735 - glsl: declare gl_Layer/gl_ViewportIndex/gl_ViewportMask as vs builtins
3736 - st/mesa: set compressed_data to NULL when freed
3737 - bin/symbols-check.py: add --ignore-symbol argument
3738 - ac/llvm: export ac_init_llvm_once in targets
3739 - mesa: rename _mesa_free_errors_data
3740 - mesa: add bool param to _mesa_free_context_data
3741 - mesa/st: release debug_output after destroying the context
3742 - ac/surface: adapt surf_size when modifying surf_pitch
3743 - radeonsi: adjust epitch for PIPE_FORMAT_R8G8_R8B8_UNORM
3744 - radeonsi: extend workaround for KHR-GL45.texture_view.view_classes on gfx9
3745 - ac/llvm: handle static/shared llvm init separately
3746 - mesa/st: introduce PIPE_CAP_NO_CLIP_ON_COPY_TEX
3747 - radeonsi: enable PIPE_CAP_NO_CLIP_ON_COPY_TEX
3748 - ac/llvm: add option to clamp division by zero
3749 - radeonsi,driconf: add clamp_div_by_zero option
3750 - radeonsi: use radeonsi_clamp_div_by_zero for SPECviewperf13, Road Redemption
3751 - glsl: fix per_vertex_accumulator::fields size
3752 - r600/uvd: set dec->bs_ptr = NULL on unmap
3753 - radeon/vcn: set dec->bs_ptr = NULL on unmap
3754 - mesa: fix glUniform* when a struct contains a bindless sampler
3756 Pierre-Loup A. Griffais (2):
3758 - radv: fix null descriptor for dynamic buffers
3759 - radv: fix vertex buffer null descriptors
3763 - radeonsi: remove emacs style config file
3764 - panfrost: don't always build bifrost_compiler
3765 - radeonsi: fix syncobj wait timeout
3766 - radeonsi: fix user fence space when MCBP is enabled
3767 - radeonsi: fix max syncobj wait timeout
3768 - radeonsi: fix user fence GPU address
3772 - intel: Store the aperture size in devinfo.
3773 - intel/isl: Update mocs for DG1
3774 - intel/l3: Return the URB size from devinfo for DG1
3775 - intel/devinfo: Add function to check for DRM_I915_GEM_GET_TILING.
3776 - iris/bufmgr: Do not use map_gtt or use set/get_tiling on DG1
3777 - anv/dg1: Don't use SET_TILING kernel uapi.
3778 - iris: Align last_seqnos to 64 bits.
3779 - anv: Align "used" attribute to 64 bits.
3783 - nv50_2d: regenerate envytools-based rnndb headers
3784 - nv50_2d,nvc0_2d: Document SET_PIXELS_FROM_MEMORY_SAFE_OVERLAP from rnndb
3785 - nvc0_2d: Document SET_PIXELS_FROM_MEMORY_CORRAL_SIZE from rnndb
3786 - nvc0: fix macro define for NVE4_COPY()
3787 - nvc0: add documentation for nve4+ (Kepler) COPY class
3791 - aco: remove use of f-strings
3792 - aco: add message to static_assert
3793 - nir: add missing group_memory_barrier handling
3794 - compiler/spirv: flag nclamp/nmin/nmax as exact
3795 - nir: make fsat return 0.0 with NaN instead of passing it through
3796 - docs: add src/amd/ to sourcetree.html
3797 - docs/envvars: document ACO_DEBUG
3798 - docs/envvars: update RADV_FORCE_FAMILY
3799 - aco: simplify consecutive ordered vmem/lds writes optimization
3800 - aco: fix consecutively written vgprs from vmem instructions
3801 - aco: mark phi definitions as last-seen phi operands
3802 - aco: consider affinities when creating v_mac_f32
3803 - aco: improve phi affinities with p_split_vector
3804 - aco: split operations that use a swap's definition
3805 - aco: fix disassembly with LLVM 11
3806 - nir/opt_if: run opt_peel_loop_initial_if after all other optimizations
3807 - nir/opt_if: use nir_src_as_bool in opt_peel_loop_initial_if helper
3808 - aco: fix typo in insert_waitcnt's kill()
3809 - nir: fix lowering to scratch with boolean access
3810 - aco: fix interaction with 3f branch workaround and p_constaddr
3811 - aco: consider SDWA during value numbering
3812 - aco: check instruction format before waiting for a previous SMEM store
3813 - aco: preserve more fields when combining additions into SMEM
3814 - aco: don't reorder barriers in the scheduler
3815 - aco: fix 64-bit shared_atomic_exchange
3816 - docs: add missing "shader\_" in VK_KHR_shader_subgroup_extended_types
3817 - radv: set keep_statistic_info with RADV_DEBUG=shaderstats
3818 - ac/gpu_info, radv: set max_wave64_per_simd to 20 on GFX10
3819 - aco: use v_xor3_b32
3820 - aco: validate instructions reading/writing upper halves/bytes
3821 - aco: p_extract_vector in 64-bit u2f16/i2f16
3822 - aco: allow reading/writing upper halves/bytes when possible
3823 - aco: prefer 4-byte aligned definitions
3824 - aco: add Info::{operand_size,definition_size}
3825 - aco: use Info::definition_size instead of definition's regclass
3826 - aco: fix moving sub-dword values out of a register for a fixed definition
3827 - aco: use num_opcodes instead of last_opcode
3828 - aco: improve code for f2{i,u}{8,16}
3829 - aco: use p_as_uniform in emit_vop1_instruction
3830 - aco: add and set precise flag
3831 - aco: create mads when signed zeros should be preserved
3832 - aco: try to use fma instead of mad when denormals are enabled
3833 - aco: create 16-bit mad/fma
3834 - aco: update comment about preserving fp16/fp64 denormals
3835 - aco: create 16-bit input and output modifiers
3836 - aco: improve sub-dword check for sgpr/constant propagation
3837 - aco: fix half_pi constant for 16-bit fsin/fcos
3838 - aco: use 32-bit inline constants for 16-bit integer instructions
3839 - aco: improve 8/16-bit constants
3840 - aco: copy-propagate constants through p_extract_vector/p_split_vector
3841 - aco: optimize 16-bit and 64-bit float comparisons
3842 - aco: validate sub-dword pseudo instructions
3843 - aco: add more opcodes to can_swap_operands
3844 - aco: allow GFX9 partial writes with instructions which use opsel
3845 - aco: improve check for moving temporaries out of fixed definitions
3846 - aco: fix encoding of certain s_setreg_imm32_b32 instructions
3847 - aco: fix validation error from vgpr spill/restore code
3848 - aco: fix sub-dword opsel/sdwa checks
3849 - aco: fix validation of opsel when set for the definition
3850 - aco: shrink ssa_info
3851 - aco: make ssa_info::label 64-bit
3852 - aco: shrink mad_info
3853 - aco: fix edge check with sub-dword temporaries
3854 - aco: use the same regclass as the definition for undef phi operands
3855 - radv: add new drirc option radv_no_dynamic_bounds
3856 - radv: enable radv_no_dynamic_bounds for Path of Exile
3857 - radv: enable radv_no_dynamic_bounds for more Path of Exile executables
3858 - nir: slight correction to cube_face_coord constant folding
3859 - spirv: set variables to restrict by default
3860 - radv: fix image variable types in meta shaders
3861 - aco: only use SMEM if we can prove it's safe
3862 - aco: allow SMEM for some sub-dword accesses
3863 - radv/aco,aco: allow SMEM SSBO loads on GFX6/7
3864 - aco: fix copy+paste error in split_buffer_store
3865 - aco: don't store byte-aligned short stores
3866 - aco: add missing bld.scc() in byte_align_scalar()
3867 - aco: don't create byte-aligned short loads
3868 - aco: fix when sub-dword create_vector operand cannot be placed perfectly
3869 - aco: improve vectorization of 8/16-bit loads/stores
3870 - aco: ignore blocked registers when checking edges in get_reg_impl()
3871 - aco: remove outdated assert in handle_operands()
3872 - radv: enable zerovram for Quantic Dream games
3873 - aco: use VOP2 version of v_mbcnt_hi_u32_b32 on GFX6/7
3874 - aco: rework boolean phi pass
3875 - aco: create better code for boolean phis with constant operands
3876 - aco: optimize boolean phis with uniform selections
3877 - aco: don't create phis with undef operands in the boolean phi pass
3878 - aco: read 0 from inactive lanes when using dpp
3879 - aco: optimize some masked swizzles to DPP
3880 - aco: implement <32-bit masked_swizzle_amd
3881 - nir/lower_subgroups: pass options struct to lower_shuffle
3882 - nir/lower_subgroups: add lower_shuffle_to_swizzle_amd
3883 - radv: use lower_shuffle_to_swizzle_amd
3884 - aco: add 32-bit integer addition to can_swap_operands
3885 - aco: fix underestimated pressure in spiller when a phi has a killed def
3886 - aco: rewrite graph coloring in spiller
3887 - aco: use unordered_set for spill id interferences
3888 - aco: add add_interference() helper
3889 - aco: use s_round_mode/s_denorm_mode
3890 - aco: flush denormals before fp16 fabs/fneg if needed
3891 - aco: fix nir_op_f2f16_rtne with non-default rounding modes
3892 - aco: set tcs_in_out_eq=false if float controls of VS and TCS stages differ
3893 - radv: enable more float_controls features
3894 - aco: properly recognize that s_waitcnt mitigates VMEMtoScalarWriteHazard
3895 - aco: use s_waitcnt_depctr to mitigate VMEMtoScalarWriteHazard
3896 - spirv: don't split memory barriers
3897 - nir/lower_int64: lower 64-bit amul
3898 - aco: always set FI on GFX10
3899 - radv: replace discard with demote for Quantic Dream games
3900 - aco: implement b2i8/b2i16
3901 - aco: be more careful combining additions that could wrap into loads/stores
3902 - aco: allow overflow for some SMEM instructions
3903 - aco: add NUW flag
3904 - nir: add nir_unsigned_upper_bound and nir_addition_might_overflow
3905 - aco: use nir_addition_might_overflow to combine additions into SMEM
3906 - aco: move some setup code into helpers
3907 - aco: make validate() usable in tests
3908 - aco: print ACO IR before scheduling instead of after
3909 - radv: fix invalid conversion warnings in vk_format.h
3910 - aco: fix copy of uninitialized boolean
3911 - aco: fix includes in aco_ir.cpp
3912 - aco: add missing add_to_hazard_query
3913 - aco: rework barriers and replace can_reorder
3914 - radv/aco,aco: use scoped barriers
3915 - aco: consider intrinsic access in visit_{load,store}_image
3916 - nir,radv/aco: add and use pass to lower make available/visible barriers
3917 - aco: enable value numbering of s_buffer_load_*
3918 - aco: use storage_scratch
3919 - aco: improve sync_info for TCS output stores
3920 - aco: improve workgroup-scope and lower vmem/smem barriers
3921 - aco: create acq+rel barriers instead of acq/rel
3922 - nir/load_store_vectorize: fix indentation
3923 - ac/nir: implement scoped_barrier
3924 - radv: use scoped barriers
3925 - aco: remove isel for GLSL-style barriers
3926 - aco: add framework for unit testing
3927 - aco: add a few tests for the assembler and optimizer
3928 - aco: add framework for testing isel and integration tests
3929 - ci: enable ACO tests
3930 - aco/tests: add tests for sub-dword swaps
3931 - aco: optimize swizzled SALU 8/16-bit conversions
3932 - aco: fix waitcnt insertion on GFX10.3
3933 - aco: don't create v_mad_f32 on GFX10.3
3934 - aco: update bug workarounds for GFX10_3
3935 - aco: fix max_waves_per_simd on Polaris, VegaM and GFX10.3
3936 - aco: update vgpr_alloc_granule for GFX10.3
3937 - aco: implement subgroup shader_clock on GFX10.3
3938 - aco: update aco_opcodes.py for GFX10.3
3939 - aco: disable SMEM stores on GFX10.3
3940 - aco: replace MADs in isel with FMA on GFX10.3
3941 - spirv: set ACCESS_COHERENT for ssbo/global/image atomic load/store
3942 - radv/aco: enable VK_KHR_memory_model
3943 - ac/nir: consider an image load/store intrinsic's access
3944 - ac/nir: fix coherent global loads/stores
3945 - radv/llvm: enable VK_KHR_memory_model
3946 - aco: fix C++11/C++14 compilation
3947 - aco: set constant_data_offset correctly in the case of merged shaders
3948 - aco: don't move memory accesses to before control barriers
3949 - aco: fix non-rtz pack_half_2x16
3950 - aco: consider branch definitions in spiller
3951 - aco: don't consider the first partial spill if it's the wrong type
3952 - aco: don't fix break condition for break+discard to exec
3953 - aco: fix regclass checks when fixing to vcc/exec with Builder
3954 - aco: fix spills_entry heuristic for branch blocks in init_live_in_vars()
3955 - aco: keep loop live-through variables spilled
3956 - aco: reserve 2 sgprs for each branch
3957 - aco: create long jumps
3958 - aco: fix byte_align_scalar for 3 dword vectors
3959 - aco: fix one-off error in Operand(uint16_t)
3960 - nir/opt_if: fix opt_if_merge when destination branch has a jump
3961 - aco: fix v_writelane_b32 with two sgprs
3962 - aco: don't apply constant to SDWA on GFX8
3963 - radv: initialize with expanded cmask if the destination layout needs it
3964 - radv,aco: fix reading primitive ID in FS after TES
3968 - util/simple_mtx: add assert_locked()
3969 - freedreno: add screen lock wrappers
3970 - freedreno: switch to simple_mtx
3971 - freedreno: fix buffer import
3972 - gallium: extract out logicop helper
3973 - freedreno/drm: drop atomic refcnts
3974 - freedreno/drm: inline the things
3975 - freedreno/a6xx: small query cleanup
3976 - freedreno/a6xx: avoid unnecessary clearing VS DP state
3977 - freedreno/a6xx: move const state to single stateobj
3978 - freedreno/a6xx: move scissor state to stateobj
3979 - freedreno/a6xx: limit PROG_FB_RAST state emit
3980 - freedreno/a6xx: limit LRZ state emit
3981 - freedreno/a6xx: move blend-color to stateobj
3982 - freedreno/a6xx: combine sample mask into blend state
3983 - freedreno/a6xx: skip unnecessary MRT blend state
3984 - freedreno/a6xx: add OUT_PKT()
3985 - freedreno/a6xx: convert draw packet to OUT_PKT()
3986 - freedreno/a6xx: split out const emit
3987 - freedreno/ir3: inline const emit
3988 - freedreno/a6xx: convert const emit to OUT_PKT()
3989 - freedreno: scissor vs disabled scissor micro-opt
3990 - freedreno/a6xx: more OUT_REG()
3991 - freedreno: sync registers with envytools
3992 - freedreno/a6xx: don't set SP_FS_CTRL_REG0.VARYING for fragcoord
3993 - freedreno/a6xx: fix LRZ hang
3994 - freedreno/a6xx: add some more formats
3995 - freedreno: we don't need aligned vbo's
3996 - freedreno/a6xx: compressed blit fixes
3997 - freedreno/a6xx: enable tiled compressed textures
3998 - freedreno/gmem: don't assume scissor opt when estimating # of bins
3999 - freedreno: initialize max_scissor
4000 - freedreno/gmem: add div_align() helper
4001 - freedreno/gmem: add helper to dump GMEM layout
4002 - freedreno: add gmemtool
4003 - freedreno/gmem: relax alignment on a6xx
4004 - freedreno/gmem: rework gmem layout algo
4005 - freedreno/ir3: don't allow negative const_offset
4006 - freedreno/ir3: fix indirect cb0 load_ubo lowering
4007 - freedreno/ir3: limit # of tex prefetch by shader size
4008 - freedreno/ir3/postsched: reset sfu_delay on sync
4009 - freedreno/ir3/postsched: try to avoid (sy) syncs
4010 - freedreno/ir3/sched: avoid scheduling outputs
4011 - freedreno/ir3/sched: try to avoid syncs
4012 - freedreno/a6xx: fix max-scissor opt
4013 - freedreno/ir3: use const_index accessors
4014 - nir: fix indices for ir3 ssbo_atomic intrinsics
4015 - nir: add helper to copy const_index[]
4016 - nir: add pass to lower disjoint wrmask's
4017 - freedreno/ir3: use lower_wrmasks pass
4018 - freedreno/fdperf: add dependency on generated headers
4019 - freedreno/drm: don't pass thru 'DUMP' flag on older kernels
4020 - freedreno/drm: handle ancient kernels
4021 - freedreno/ir3: remove Sethi-Ullman numbering pass
4022 - freedreno/ir3: juggle around ir3_debug_print()
4023 - freedreno/ir3/dce: report progress
4024 - freedreno/cf: report progress
4025 - freedreno/ir3/cp: report progress
4026 - freedreno/ir3/deps: report progress
4027 - freedreno/ir3/group: report progress
4028 - freedreno/ir3/legalize: report progress
4029 - freedreno/ir3/postsched: report progress
4030 - freedreno/ir3: add IR3_PASS() macro
4031 - freedreno/ir3: move where we preserve binning pass inputs
4032 - freedreno/ir3: be iterative
4033 - freedreno/ir3: make foreach_src declare cursor ptr
4034 - freedreno/ir3: make foreach_ssa_src declar cursor ptr
4035 - freedreno/ir3: make input/output iterators declare cursor ptr
4036 - freedreno/ir3/group: fix for half-regs
4037 - freedreno/ir3: fix mismatched flags on split
4038 - freedreno/ir3/cf: handle multiple cov's properly
4039 - freedreno/ir3: fix immed type in create_addr0()
4040 - freedreno/ir3/print: print cat2 condition
4041 - freedreno/ir3/cp: fix cmps folding
4042 - freedreno/ir3: fix mismatched wrmask for overlapping VS inputs
4043 - freedreno/ir3: add simple validate pass
4044 - freedreno/ir3: add helpers to deal with src/dst types
4045 - freedreno/ir3/validate: add checking for types and opcodes
4046 - freedreno/drm: disallow exported buffers in bo cache
4047 - freedreno: add batch debugging
4048 - freedreno: clear last_fence after resource tracking
4049 - freedreno: handle PIPE_TRANSFER_MAP_DIRECTLY
4050 - freedreno/gmem: make noscis debug actually do something on a6xx
4051 - freedreno/gmemtool: make GMEM alignment per-gen
4052 - freedreno/gmemtool: add a405
4053 - freedreno/gmemtool: add verbose mode
4054 - freedreno/gmem: add some asserts
4055 - freedreno/gmem: fix nbins_x/y mismatch
4056 - freedreno/gmem: split out helper to calc # of bins
4057 - freedreno/a6xx: LRZ fix for alpha-test
4058 - freedreno/a6xx: document LRZ flag buffer
4059 - freedreno/a6xx: fix vsc assert
4060 - nir: get_base_type() should return enum type
4061 - nir: extract out convert_to_bitsize() helper
4062 - nir/builder: add bitsize conversion helpers
4063 - nir/lower_tex: fixes for fp16 yuv lowering
4064 - freedreno/ir3: split kill from no_earlyz
4065 - freedreno/a6xx: sync registers from envytools
4066 - freedreno/a6xx: update depth-plane control regs
4067 - freedreno/a6xx: re-work LRZ state tracking
4068 - freedreno/a6xx: add early-lrz-late-z mode
4069 - freedreno/a6xx: also consider alpha-test for ztest-mode
4070 - freedreno/a6xx: more early-z
4071 - freedreno/computerator: fix missing dependency on generated header
4072 - nir/print: print tex dest type
4073 - freedreno/ir3: add debug code to print conflicting half-regs
4074 - freedreno/ir3: respect tex prefetch limits
4075 - freedreno/ir3: remove RA "q-values" optimization
4076 - freedreno/ir3: limit pre-fetched tex dest
4077 - freedreno/ir3: unify shader create/delete paths
4078 - freedreno/ir3: move the libdrm dependency out of shared code
4079 - turnip: drop linking libfreedreno_drm
4080 - freedreno/ir3: don't rely on intr->num_components
4081 - radv: don't set num_components for non-vectorized intrinsics
4082 - nir/builder: don't set intr->num_components
4083 - nir/lower-atomics-to-ssbo: don't set num_components
4084 - spriv: don't set num_components for non-vectorised intrinsics
4085 - v3d: don't use intr->num_components for non-vectorized intrinsics
4086 - nir/validate: validate intr->num_components
4087 - freedreno/log-parser: fix compute times
4088 - freedreno/sched: reset delay counters at start of block
4089 - freedreno/ir3/validate: also check instr->address
4090 - freedreno/ir3/cp: properly handle already-folded RELATIV
4091 - freedreno: splitup emit_string_marker
4092 - freedreno/a6xx: emit shader names in debug builds
4093 - freedreno/ir3/legalize: don't allow (nopN) if (rptN)
4094 - freedreno/ir3/print: print (r) flag
4095 - freedreno/ir3: add test for delay slot calculation
4096 - freedreno/ir3/delay: calculate delay properly for (rptN)'d instructions
4097 - freedreno/ir3: add helpers to move instructions
4098 - freedreno/ir3: delay test support for vectorish instructions
4099 - freedreno/ir3/cp: extract valid_flags
4100 - freedreno/ir3: add post-scheduler cp pass
4101 - freedreno/ir3: convert regmask_t to struct
4102 - freedreno/ir3: move mergedreg state out of reg
4103 - freedreno/ir3: decouple regset from gpu gen
4104 - freedreno/ir3: pass variant to postsched
4105 - freedreno/ir3: re-work assembler API
4106 - freedreno/ir3: make mergedregs a property of the variant
4107 - freedreno/a6xx: set .MERGEREGS based on variant
4108 - turnip: set .MERGEDREGS based on variant
4109 - freedreno/computerator: MERGEDREGS update
4110 - freedreno/ir3: update obsolete comment
4111 - spirv: atomic_counter_read_deref is not vectorized
4112 - spirv: drop some dead code
4113 - glsl_to_nir: fix is_helper_invocation
4114 - glsl_to_nir: fix shader_clock
4115 - glsl_to_nir: fix vote_any/vote_all
4116 - freedreno/ir3: refactor out helper to compile shader from asm
4117 - freedreno/ir3: add accessor for const_state
4118 - freedreno/a6xx: defer userconst cmdstream size calculation
4119 - freedreno/ir3: move ubo_state into const_state
4120 - freedreno/ir3: drop shader->num_ubos
4121 - freedreno/ir3: constify shader key
4122 - freedreno/ir3: pass variant to ir3_create()
4123 - freedreno/ir3: convert over to ralloc
4124 - freedreno/ir3: move num_reserved_user_consts out of const_state
4125 - freedreno/ir3: un-embed const_state
4126 - freedreno/ir3: move const_state back to variant
4127 - freedreno/ir3: move output_loc to variant
4128 - freedreno/ir3: split out ubo info from range
4129 - freedreno/ir3: splitup get_existing_range()
4130 - freedreno/ir3: split ubo analysis/lowering passes
4131 - ci: remove some freedreno a6xx skips
4132 - freedreno/ir3: add helper to determine point-coord inputs
4133 - freedreno/a6xx: de-duplicate vinterp/vpsrepl state building
4134 - freedreno/a6xx: use point-coord helper
4135 - freedreno/a5xx: use point-coord helper
4136 - freedreno/a4xx: use point-coord helper
4137 - freedreno/a3xx: use point-coord helper
4138 - freedreno: convert builtin blit VS prog to ureg builder
4139 - freedreno/ir3: switch PIPE_CAP_TGSI_TEXCOORD
4140 - freedreno: make foreach_bit() declare it's cursor
4141 - freedreno: split out batch draw tracking helper
4142 - freedreno: split out batch clear tracking helper
4143 - freedreno: handle batch flush in resource tracking
4144 - freedreno/ir3/ra: fix pre-color edge case
4145 - freedreno/ir3: add ir3_finalize_nir()
4146 - freedreno/ir3: move finalize_nir to pscreen hook
4147 - freedreno/ir3: add ir3_compiler_destroy()
4148 - freedreno/ir3: shuffle some variant fields
4149 - freedreno/a6xx+ir3: stop generating pointless binning shaders
4150 - freedreno/ir3: build binning variant at same time as draw variant
4151 - freedreno/ir3: disk-cache support
4152 - freedreno/ir3: move nir finalization to after cache miss
4153 - freedreno/fdperf: fix print of base address
4154 - freedreno/fdperf: better compatible string matching
4155 - freedreno/fdperf: prefer render node
4156 - gitlab-ci: reduce a630 runner load
4157 - freedreno/ir3: add missing VS driver params
4158 - freedreno/ir3: make compile fails more visible
4159 - freedreno/a6xx: bail instead of crash for compile fails
4160 - freedreno/ir3/ra: be better at failing
4161 - freedreno/a6xx: don't enable early-z/lrz if no z-test
4162 - freedreno/ir3: DCE unused arrays
4163 - driconf: allowlist/denylist
4164 - gitlab-ci: re-enable all a630 jobs
4165 - freedreno: small comment re-word
4166 - freedreno: whitespace fix
4167 - freedreno/ir3/parser: half-precision relative regs
4168 - freedreno/ir3: set array precision on creation
4169 - freedreno/ir3: fix half-reg array stores
4170 - freedreno/ir3/ra: debug msgs tweak
4171 - freedreno/ir3/ra: assign vreg names to all array elements
4172 - freedreno/ir3/ra: fix array conflicts for split/merged
4173 - freedreno: sync registers from envytools
4174 - freedreno: make gen_header.py check parent directory
4175 - freedreno: slurp in rnndb
4176 - freedreno: slurp in rnn
4177 - freedreno: slurp in decode tools
4178 - freedreno: slurp in afuc
4179 - freedreno/rnn: warnings cleanup
4180 - freedreno/decode: warnings cleanup
4181 - freedreno/afuc: warnings cleanup
4182 - freedreno: add CI for envytools tools
4183 - freedreno/ir3: split out regmask
4184 - freedreno: drop shader_t
4185 - freedreno: deduplicate a3xx+ disasm
4186 - freedreno: move a2xx disasm out of gallium
4187 - freedreno: deduplicate a2xx disasm
4188 - freedreno/ci: add a2xx trace to CI job
4189 - freedreno/tools: check rnn parse status
4190 - freedreno/rnn: split out helper to find files
4191 - freedreno/rnn: add error helper
4192 - freedreno/rnn: rename schema file
4193 - freedreno/rnn: update schema for 'pos'
4194 - freedreno/rnn: add relaxed boolean type
4195 - freedreno/rnn: add high/low/pos to registers
4196 - freedreno/rnn: add radix/align
4197 - freedreno/rnn: relax Hexadecimal to HexOrNumber
4198 - freedreno/rnn: add variants/varset to domain
4199 - freedreno/registers/a2xx: fix validation error
4200 - freedreno/registers/a4xx: fix validation error
4201 - freedreno/registers/adreno_pm4: fix validation errors
4202 - freedreno/rnn: describe copyright element in schema
4203 - freedreno/rnn: add "addvariant" to schema
4204 - freedreno/rnn: allow name to be optional in arrays
4205 - freedreno/rnn: fix use-group
4206 - freedreno/registers/mdp5: fix validation error
4207 - freedreno/rnn: schema updates for dynamic/irregular offsets
4208 - freedreno/rnn: add schema validation
4209 - freedreno/rnn: headergen2 warnings cleanup
4210 - freedreno/decode: cffdec warnings cleanup
4211 - freedreno/ir3: add missing track_ubo_use()
4212 - freedreno/a6xx: don't emit a bogus size for empty cb slots
4213 - freedreno/a6xx: fixup draw state earlier
4214 - freedreno/rnn: also look for .xml.gz
4215 - freedreno/rnn: rework RNN_DEF_PATH construction
4216 - freedreno/registers: add .gitignore
4217 - freedreno/registers: split header build into subdirs
4218 - freedreno/registers: install gzip'd register database
4219 - freedreno/decode: move dependencies up a level
4220 - freedreno: allow fence_fd fences to be recycled
4221 - freedreno/ir3: ir3_cmdline updates
4222 - freedreno/ir3: lower local_index using local_id
4223 - glsl/lower_precision: split out const lowering
4224 - gallium: replace 16BIT_TEMPS cap with 16BIT_CONSTS
4225 - glsl: remove LowerPrecisionTemporaries
4226 - glsl: don't inline intrinsics for mediump
4227 - glsl_to_nir: fix bitfield_extract with 16-bit operands
4228 - freedreno/registers: add some missing regs to build
4229 - freedreno/crashdec: handle section name typos
4230 - freedreno/a6xx: fix occlusion query with more than one tile
4231 - freedreno: handle case of shadowing current render target
4232 - freedreno/gmemtool: add tile_alignw/h and a650
4236 - iris: Fix documentation for _iris_batch_flush
4237 - ci: Include trace replay support in ARM rootfses.
4238 - gitlab-ci: Replay traces on lava devices
4242 - gallivm: fix half to float conversions with llvm 11
4246 - vulkan/wsi/x11: add sent image counter
4247 - vulkan/wsi/x11: wait for acquirable images in FIFO mode
4251 - egl: Build surfaceless platform on Android
4252 - Android: Fixes for Q and R
4253 - panfrost: Android build fixes 2020 week 31
4254 - lima: Fix lima_screen_query_dmabuf_modifiers()
4255 - android: freedreno: Another build fix
4259 - iris: Use modfiy disables for 3DSTATE_WM_DEPTH_STENCIL command
4260 - intel/compiler: Optimize integer add with 0 into mov
4261 - intel/compiler: Remove unnecessary optimization for MUL
4265 - ci: fix reporting the number of unexpected/flakes
4266 - ci: add lists of expected failures & skipped tests for RAVEN with ACO
4267 - aco: remove unecessary p_split_vector with v2b reg class
4268 - radv: enable shaderInt16 unconditionally with LLVM and only GFX8+ with ACO
4269 - radv: cleanup radv_CreateInstance()
4270 - radv: rename radv_devices() to radv_enumerate_physical_devices()
4271 - radv: fix a memleak if the physical device initialization failed
4272 - radv: report INITIALIZATION_FAILED when the amdgpu winsys init failed
4273 - radv: don't report error with other vendor DRM devices
4274 - radv: use a linked list for physical devices
4275 - radv: display an error message if the winsys init failed
4276 - radv/winsys: do not count visible VRAM buffers twice in the budget
4277 - ci: remove unused .test-radv-fossilize rule
4278 - ci: set ACO_DEBUG=validateir,validatera global for RADV testing
4279 - ci: run radv-fossils with Pitcairn (GFX6) and Bonaire (GFX7) too
4280 - radv: remove the LLVM version string when ACO is used
4281 - radv: do not print the LLVM version string twice in hang reports
4282 - radv: report correct backend IR in hang reports when ACO is used
4283 - aco: fix 64-bit trunc with negative exponents on GFX6
4284 - nir: do not vectorize load/store if offset can overflow and robustness enabled
4285 - aco: prevent invalid loads/stores vectorization if robustness is enabled
4286 - radv: limit the Vulkan version to 1.1 for Android
4287 - radv: handle different Vulkan API versions correctly
4288 - radv: update the list of allowed Android extensions
4289 - aco: optimize add/sub(a, cndmask(b, 0, 1, cond)) -> addc/subbrev_co(0, a, b)
4290 - radv: use the common base object type for VkDevice
4291 - radv: use the base object struct types
4292 - radv: implement VK_EXT_private_data
4293 - vulkan: import common code for generating extensions
4294 - radv: use the common code for generating extensions and dispatch tables
4295 - anv: use the common code for generating extensions and dispatch tables
4296 - turnip: use the common code for generating extensions and dispatch tables
4297 - radv: add a LLVM version string workaround for SotTR and ACO
4298 - aco: remove useless check for nir_tex_src_bias
4299 - aco: add support for texturing with clamped LOD
4300 - ac/llvm: add support for texturing with clamped LOD
4301 - radv: enable shaderResourceMinLod
4302 - spirv: handle OpCopyObject correctly with any types
4303 - radv: fix missing break in radv_GetPhysicalDeviceProperties2()
4304 - aco: store 16-bit temporary outputs as v2b
4305 - aco: convert 16-bit values before exporting MRTs
4306 - aco: allow to load/store 16-bit values in VMEM for tess and geom
4307 - aco: implement 8-bit/16-bit mov's with p_create_vector
4308 - aco: implement 16-bit vertex fetches with tbuffer_load_format_d16_*
4309 - aco: validate v_interp_*_f16 as VOP3 instructions instead of VINTRP
4310 - aco: emit v_interp_*_f16 instructions as VOP3 instead of VINTRP
4311 - aco: implement 16-bit interp
4312 - aco: fix off-by-one error with 16-bit MTBUF opcodes on GFX10
4313 - radv/aco: enable storageInputOutput16 on GFX9+
4314 - aco: fix missing break in label_instruction()
4315 - radv: fix missing break in radv_GetPhysicalDeviceFeatures2()
4316 - radv: fix duplicated expression in ac_setup_rings()
4317 - radv/winsys: remove useless free in radv_amdgpu_create_bo_list()
4318 - aco: declare 8-bit/16-bit reduce operations
4319 - aco: implement 8-bit/16-bit reductions
4320 - aco: validate 8-bit/16-bit VGPR operands for readfirstlane/readlane/writelane
4321 - aco: implement 8-bit/16-bit nir_intrinsic_read_first_invocation
4322 - aco: implement 8-bit/16-bit nir_intrinsic_{shuffle,_read_invocation}
4323 - aco: implement 8-bit/16-bit nir_intrinsic_quad_*
4324 - aco: use a temporary SGPR for 8-bit/16-bit literal reduction identities
4325 - aco: sign-extend the input and identity for 8-bit subgroup operations
4326 - radv: do not return from radv_GetPhysicalDeviceFeatures2()
4327 - radv: cleanup physical device features
4328 - radv: remove useless assignment in build_streamout_vertex()
4329 - spirv: add ReadClockKHR support with device scope
4330 - aco: implement nir_intrinsic_shader_clock with device scope
4331 - ac/nir: fix shader clock with subgroup scope
4332 - ac/nir: implement nir_intrinsic_shader_clock with device scope
4333 - radv: advertise shaderDeviceClock on GFX8+
4334 - spirv: add SpvCapabilityImageGatherBiasLodAMD
4335 - spirv: add support for bias/lod with OpImageGather
4336 - ac/nir: add support for bias/lod with texture gather
4337 - aco: add support for bias/lod with texture gather
4338 - radv: add support for querying which formats support texture gather LOD
4339 - radv: advertise VK_AMD_texture_gather_bias_lod
4340 - spirv,radv,anv: implement no-op VK_GOOGLE_user_type
4341 - radv/aco: enable VK_EXT_subgroup_size_control
4342 - aco: fix register allocation for subdword instructions on GFX10
4343 - aco: implement 8-bit/16-bit reductions on GFX10
4344 - aco: allocate a temp VGPR for some 8-bit/16-bit reduction ops on GFX10
4345 - aco: allow gfx10_wave64_bpermute with 8-bit/16-bit input
4346 - aco: sign-extend input/indentity for 32-bit reduce ops on GFX10
4347 - radv/aco: enable VK_KHR_subgroup_extended_types on GFX8+
4348 - radv: enable zero VRAM for Doom Eternal
4349 - radv: enable zero VRAM for all VKD3D (DX12->VK) games
4350 - aco: implement 16-bit reduce operations on GFX6-GFX7
4351 - aco: implement 16-bit nir_intrinsic_quad_* on GFX6-GFX7
4352 - aco: fix subdword copies on GFX6-GFX7
4353 - aco: sign-extend input/identity for 16-bit subgroup ops on GFX6-GFX7
4354 - radv/aco: enable 64-bit atomic features if RADV is linked with LLVM 8
4355 - aco: use v_bfe_u32 for unsigned reductions sign-extension on GFX6-GFX7
4356 - aco: fix sign-extend 8-bit subgroup operations on GFX6-GFX7
4357 - aco: fix nir_intrinsic_quad_* with 8-bit in GFX6-GFX7
4358 - radv/aco: enable VK_KHR_shader_subgroup_extended_types on GFX6-GFX7
4359 - ac/nir: adjust an assertion for D16 on GFX6-GFX7
4360 - nir/lower_explicit_io: fix NON_UNIFORM access for UBO loads
4361 - radv/llvm: expose VK_EXT_shader_demote_to_helper_invocation with LLVM 9+
4362 - aco: implement 8-bit/16-bit conversions on GFX6-GFX7
4363 - aco: fix alignment of vectors with 4 elements
4364 - radv/aco: enable 8-bit/16-bit storage on GFX6-GFX7
4365 - radv/aco: enable shaderInt16 on GFX6-GFX7
4366 - radv/aco: enable shaderInt8 and VK_KHR_shader_float16_int8 on GFX6-GFX7
4367 - ac/nir: fix integer comparisons with pointers
4368 - radv: set DB_SHADER_CONTROL.CONSERVATIVE_Z_EXPORT correctly
4369 - radv: add new drirc option radv_enable_mrt_output_nan_fixup
4370 - aco: implement radv_enable_mrt_output_nan_fixup workaround
4371 - radv/llvm: implement radv_enable_mrt_output_nan_fixup workaround
4372 - radv: enable radv_enable_mrt_output_nan_fixup for RAGE 2
4373 - ac: add ac_choose_spi_color_formats() to common code
4374 - spirv: fix using OpSampledImage with OpUndef instead of OpType{Image,Sampler}
4375 - aco: allow to swap operands for some 16-bit float instructions
4376 - spirv: do not set num_components for non-vectorized mbcnt_amd intrinsic
4377 - radv/aco: enable FP16 features/extensions on GFX9+
4378 - radv: lower discards to demote to workaround a RDR2 game bug
4379 - radv: make sure to set CB_SHADER_MASK correctly for internal CB operations
4380 - radv: compute CB_SHADER_MASK from the fragment shader outputs
4381 - radv: only requires LLVM 9 for GFX10 if not using ACO
4382 - radv: replace == GFX10 with >= GFX10 where it's needed
4383 - aco: replace == GFX10 with >= GFX10 where it's needed
4384 - radv: add support for Sienna Cichlid
4385 - radv: require LLVM 11+ for GFX 10.3 if not using ACO
4386 - aco: fix printing ASM on GFX6-7 if clrxdisasm is not found
4387 - aco: improve validation checks for readlane/writelane
4388 - aco: fix printing ASM on GFX6-7 again
4389 - gitlab-ci: stop testing RADV with LLVM
4390 - gitlab-ci: update the list of expected CTS failures for RADV/ACO
4391 - gitlab-ci: update the list of expected failures for Pitcairn
4392 - radv: fix checking the return value of cs_finalize()
4393 - gitlab-ci: add parallel-rdp fossils
4394 - radv: lower 64-bit drcp/dsqrt/drsq for fixing precision issues
4395 - radv: lower 64-bit dfloor on GFX6 for fixing precision issues
4396 - gitlab-ci: add a list of expected failures for RADV/ACO on NAVI14
4397 - gitlab-ci: set the number of Fossilize threads to 4
4398 - gitlab-ci: append Fossilize stdout/stderr to a file to reduce spam
4399 - gitlab-ci: attach the Fossilize log file as artifact on failure
4400 - radv: remove the shader ballot workaround for Youngblood with LLVM
4401 - radv: remove the load/store workaround for Monster Hunter World with LLVM
4402 - radv: enable VK_AMD_shader_ballot on GFX6-7 with both compiler backends
4403 - radv: adjust CB_SHADER_MASK for dual-source blending in the shader info pass
4404 - radv: rework 8/16-bit color attachment formats detection
4405 - radv: use SPI_SHADER_ZERO for non-written color attachments
4406 - radv: add support for MRTs compaction to avoid holes
4407 - radv: fix wide points and lines
4408 - radv: fix wide lines with multisample enabled
4409 - Revert "vulkan/wsi/x11: Ensure we create at least minImageCount images."
4410 - radv,vulkan: add a new x11 wsi drirc workaround for DOOM Eternal
4411 - radv: disable FMASK compression when drawing with GENERAL layout
4412 - radv: set depth/stencil enable values correctly for the meta clear path
4413 - radv: implement missing VK_ACCESS_MEMORY_{READ,WRITE}_BIT
4414 - radv: store the primitive topology hardware value in the pipeline
4415 - radv: adjust IA_MULTI_VGT_PARAM.WD_SWITCH_ON_EOP at draw time
4416 - radv: adjust IA_MULTI_VGT_PARAM.PARTIAL_VS_WAVE at draw time
4417 - radv: compute prim_vertex_count at draw time
4418 - aco: fix more validation errors from vgpr spill/restore code
4419 - radv: return VK_ERROR_DEVICE_LOST if wait-for-idle failed or expired
4420 - radv: remove the secure compile support feature
4421 - radv: rework dynamic viewports/scissors support
4422 - radv: add VK_EXT_extended_dynamic_state but leave it disabled
4423 - radv: declare new extended dynamic states
4424 - radv: add support for dynamic cull mode and front face
4425 - radv: add support for dynamic primitive topology
4426 - radv: add support for dynamic and scissor count
4427 - radv: add support for dynamic depth/stencil states
4428 - radv: add support for dynamic vertex input binding stride
4429 - radv: advertise VK_EXT_extended_dynamic_state
4430 - radv: add the custom border color BO to the list of buffers
4431 - radv: destroy the base object if VkCreateQueryPool() failed
4432 - radv: destroy the base object if VkCreateRenderPass*() failed
4433 - radv: destroy the base object if VkCreateImage() failed
4434 - radv: destroy the base object if VkCreateBuffer() failed
4435 - radv: destroy the base object if VkCreateEvent() failed
4436 - radv: destroy the base object if VkCreateSemaphore() failed
4437 - radv: destroy the base object if VkCreateFence() failed
4438 - radv: destroy the base object if VkAllocateCommandBuffers() failed
4439 - radv: destroy the base object if VkCreateInstance() failed
4440 - radv/winsys: replace alloca() by malloc() everywhere
4441 - radv/winsys: pass the buffer list via the CS ioctl for less CPU overhead
4442 - radv: fix destroying the syncobj when exporting a fence FD
4443 - radv: fix the error code when exporting a semaphore/fence fails
4444 - radv: fix the error code when allocating a fresh imported syncobj fails
4445 - radv: optimize creating signaled syncobj with amdgpu_cs_create_syncobj2()
4446 - radv: split fence into two parts as enum+union.
4447 - radv: remove one useless goto in radv_queue_submit_deferred()
4448 - radv: improve the error messages when a CS submission failed
4449 - radv: return better Vulkan error codes when VkQueueSubmit() fails
4450 - radv: disable CPU caching for IBS to reduce fetch latency
4451 - radv/winsys: always allow GTT placements on APUs
4452 - radv: advertise VK_EXT_image_robustness
4453 - radv: do not perform read-modify-write with the upload BO
4454 - radv: disable CPU caching for the upload BO to reduce fetch latency
4455 - aco: add support for nir_intrinsic_shared_atomic_fadd
4456 - ac/nir: add support for nir_intrinsic_shared_atomic_fadd
4457 - radv: advertise VK_EXT_shader_atomic_float
4458 - radv: add missing return values check for some winsys calls
4459 - radv/winsys: check more allocation failures
4460 - radv/winsys: remove useless check when binding virtual buffers/images
4461 - radv/winsys: return a Vulkan error code when binding virtual buffers/images
4462 - radv/winsys: be more robust when a CS failed during recording
4463 - radv: remove declared but unused radv_pipeline::is_dual_src
4464 - radv: remove set but unused radv_pipeline::vertex_elements
4465 - radv: remove outdated TODO related to PA_SU_VTX_CNTL.PIX_CENTER
4466 - radv: emit more invariant registers as part of the initial gfx state
4467 - radv: emit PA_SC_LINE_CNTL as part of the rasterization state
4468 - radv: clean up VGT_SHADER_STAGES_EN emission
4469 - radv: clean up PA_SC_CLIPRECT_RULE emission
4470 - radv: reduce the number of allocated dwords for compute CS
4471 - radv: clean up radv_compute_generate_pm4()
4472 - radv: remove unnecessary radv_tessellation_state::num_patches
4473 - radv: remove no-op si_multiwave_lds_size_workaround()
4474 - radv: remove one unnecessary param to radv_generate_graphics_pipeline_key()
4475 - radv: align the LDS size in calculate_tess_lds_size()
4476 - radv: set LDS TCS size at shaders creation for GFX9+
4477 - radv: remove unnecessary radv_tessellation_state::lds_size
4478 - radv: clean up tessellation state emission
4479 - radv: add radv_pipeline_init_input_assembly_state()
4480 - radv: add radv_pipeline_generate_vgt_gs_out()
4481 - radv: clean up adjusting MSAA state if conservative rast is enabled
4482 - radv: clean up binning state initialization
4483 - radv: assign pipeline gfx fields before PM4 emission
4484 - radv: constify all radv_pipeline_generate_*() helpers
4485 - radv: add radv_pipeline_init_shader_stages_state()
4486 - radv: remove useless return value to radv_pipeline_scratch_init()
4487 - radv: clean up remaining pipeline init functions
4488 - radv: print warnings for famous RADV_PERFTEST options that no longer exist
4489 - radv: do not honor a user-specified pitch on GFX 10.3
4490 - radv: increase minimum NGG vertex count requirement per workgroup on GFX 10.3
4491 - radv: fix sample shading on GFX 10.3
4492 - radv: set BYPASS_VTX_RATE_COMBINER_GFX103 on GFX 10.3
4493 - radv/gfx10: add missing initialization of registers
4494 - radv: limit LATE_ALLOC_GS to prevent a GPU hang on GFX10
4495 - radv: fix emitting the border color pointer on the compute queue
4496 - nir/algebraic: mark some optimizations with fsat(NaN) as inexact
4497 - aco: handle unaligned loads on GFX10.3
4498 - spirv: fix emitting switch cases that directly jump to the merge block
4499 - radv: fix transform feedback crashes if pCounterBufferOffsets is NULL
4503 - frontends/va: Handle dynamic resolution/SVC for VP9
4507 - intel/dev: Don't consider all TGL SKUs as GT1 only
4511 - amd/common: Fix incorrect use of asprintf instead of vasprintf
4512 - clover: add more cl_mem_object_type to pipe_texture_target mapping
4513 - clover: implements clEnqueueFillBuffer
4517 - freedreno/a4xx: fix \*_NONE enum conversion
4521 - EGL: sync headers with Khronos
4522 - gbm: document that gbm_bo_map exposes a linear view
4523 - radv: use bitshifts for debug enum values
4527 - radeon/vcn: Corrected vp9 ref associated data incase of target->codec is NULL
4531 - st/mesa: destroy only own program variants when program is released
4532 - anv: call base finish only if pass given in DestroyRenderPass
4533 - anv: add VK_EXT_extended_dynamic_state but leave it disabled
4534 - anv: add new dynamic states
4535 - anv: consider dynamic state when creating pipeline
4536 - anv: handle dynamic viewport count
4537 - anv: add support for dynamic cull mode and winding order
4538 - anv: add support for dynamic viewport and scissor with count
4539 - anv: add support for dynamic primitive topology change
4540 - anv: depth/stencil dynamic state support
4541 - anv: dynamic vertex input binding stride and size support
4542 - anv: toggle on VK_EXT_extended_dynamic_state
4543 - anv: add a check for depthStencilState before using it
4544 - anv: null check for buffer before reading size
4548 - radeon: Fix whitespaces
4549 - gallium/auxiliary/vl: Fix compute shader scaling for non-square pixels
4550 - gallium/auxiliary/vl: Fix compute shader scale_y for interlaced videos
4551 - frontends/va: Fix deinterlace bottom field first flag
4552 - frontends/vdpau: Default destination rect to source rect
4553 - radeon/vcn: add vcn 3.0 encode support
4554 - radeonsi: use PIPE_FORMAT_P010 for 10-bit VP9 decoding
4555 - radeon/vcn: increase render_pic_list size
4559 - glsl: stop cascading errors if process_parameters() fails
4560 - glsl: fix slow linking of uniforms in the nir linker
4561 - radv: fix regression with builtin cache
4562 - nir: add glsl_get_ifc_packing() helper
4563 - nir: add callback to nir_remove_dead_variables()
4564 - glsl: add can_remove_uniform() helper to the NIR linker
4565 - glsl: remove dead uniforms in the nir linker
4566 - glsl/spirv: remove dead uniforms in spirv nir linker
4567 - gitlab-ci: bump piglit checkout commit
4568 - i965: call brw_nir_lower_uniforms() after uniform linking is complete
4569 - util: add BITSET_LAST_BIT() helper
4570 - glsl: add struct to gather more info about uniform array access
4571 - glsl: add update_array_sizes() helper to the NIR uniform linker
4572 - glsl: gather uniform dereference info before main linking loop
4573 - glsl: when NIR linker enable use it to resize uniform arrays
4574 - glsl: fix potential slow compile times for GLSLOptimizeConservatively
4575 - glsl: fix incorrect optimisation in opt_constant_variable()
4576 - glsl: fix uniform array resizing in the nir linker
4577 - glsl: small optimisation fix for uniform array resizing
4578 - st_glsl_to_nir: fix potential use after free
4579 - mesa: remove _mesa prefix from static function
4580 - mesa: add _mesa_program_state_value_size() helper
4581 - glsl: define gl_LightSource members in ARB_vertex_program order
4582 - st/glsl_to_nir: disable st_nir_lower_builtin() when packing supported
4583 - glsl: remove stale FIXME
4584 - i965: add and fix fallthrough comments
4585 - llvmpipe: add missing fallthrough comments
4586 - gallivm: add missing break
4587 - anv: update fallthrough comment so gcc sees it
4588 - intel/compiler: add and fix up fallthrough comments for gcc warnings
4589 - iris: add missing fallthrough comment
4590 - egl: move fallthrough comment so gcc can see it
4591 - nir: add missing break to nir_opt_access()
4592 - mesa: fix fallthrough in glformats
4593 - mesa: add fallthrough comments to glformats.c
4594 - mesa: add fallthrough comments to get.c
4595 - nir: fix implicit fallthrough warnings
4596 - mesa: add fallthrough comments to COPY_SZ_4V()
4597 - radeonsi: add missing fallthrough comment
4598 - glx: add missing fallthrough comment
4599 - glsl: move fallthrough comment to where gcc can see it
4600 - radeon: add missing fallthrough comments
4601 - spirv: add missing fallthrough comments
4602 - mesa/vbo: add some missing fallthrough comments
4603 - mesa: add missing fallthrough comment to teximage.c
4604 - mesa: fix unintended fallthrough in glIsEnabled()
4605 - r300: add and fix up fallthrough comments
4606 - svga: add missing fallthrough comments
4607 - mesa: update fallthrough comment so gcc can see it
4608 - nv30: add missing fallthrough comment
4609 - meson: turn on Wimplicit-fallthrough project wide
4610 - nouveau: fix pointer-sign warning
4611 - gitlab-ci: Enable -Werror in `meson-classic` job
4612 - r600/radeonsi: silence zero-length-bounds gcc warnings
4613 - radeonsi: fix SI_NUM_ATOMS
4614 - iris: fix maybe-uninitialized warning for initial_state variable
4615 - iris: silence maybe-uninitialized for stc_dst_aux_usage variable
4616 - nouveau/nvc0: silence maybe-uninitialized warning
4617 - panfrost: add some missing fallthrough comments
4618 - panfrost: hide more unused code in bi_lower_combine.c
4619 - panfrost: add some missing fallthrough comments to bi_pack.c
4620 - freedreno: fix missing fallthrough comments
4621 - v3d: remove redefine of VG(x)
4622 - zink: fix missing fallthrough comment
4623 - nine: remove unused var
4624 - etnaviv: add missing fallthrough comments
4625 - lima: add missing fallthrough comments
4626 - lima: add missing break
4627 - gitlab-ci: Enable -Werror in `meson-gallium` job
4631 - aco/gfx10: Refactor of GFX10 wave64 bpermute.
4632 - aco: Implement subgroup shuffle on GFX6-7.
4633 - radv/aco: Always enable subgroup shuffle.
4634 - aco: Fix emit_boolean_exclusive_scan in wave32 mode.
4638 - panfrost: Emit blend descriptors on Bifrost
4639 - panfrost: Don't leak temporary descriptors array
4640 - pan/decode: Check for correct unknown field
4641 - pan/decode: Use correct printf modifier for long int
4642 - panfrost: Split bit out of format.unk3
4643 - panfrost: Create additional BO for the checksum of imported BOs (Bifrost)
4644 - panfrost: Add a bit more info about some tiler fields
4645 - pan/bi: Print shaders only if BIFROST_MESA_DEBUG=shaders
4646 - pan/decode: Trace to stderr with PANDECODE_DUMP_FILE=stderr
4647 - panfrost: GPUs newer than G-71 don't have swizzles...
4648 - panfrost: mali_attr_meta.unknown1 is zero on Bifrost
4649 - panfrost: Add Bifrost texture trampoline BO to batch
4650 - pan/decode: Properly print tripped zeroes
4651 - virgl: Properly check for encode_stride when encoding transfers
4652 - panfrost: Add checksum BOs to batch
4653 - panfrost: Don't trample on top of Bifrost-specific unions
4654 - panfrost: Handle MALI_RGB8_UNORM in panfrost_format_to_bifrost_blend
4655 - gitlab-ci: Run more dEQP tests for virgl
4656 - gitlab-ci: Add manual tests for Virgl using GLES on the host
4657 - gitlab-ci: Test virgl with Khronos' OpenGL CTS
4658 - gitlab-ci: Update CTS runner
4659 - ci: Don't call renderdoc's ReplayController.Shutdown()
4660 - ci: Move ARM rootfses to stable
4661 - gitlab-ci: Build kernel drivers for a few ethernet USB dongles
4662 - gitlab-ci: More stable URL for kernel and ramdisk artifacts, for LAVA
4663 - gitlab-ci: Remove left-behind rules:
4664 - gitlab-ci: Don't rebuild kernels and rootfs if they have been already built in mainline
4665 - gitlab-ci: Run all of GLES3 tests for Panfrost
4666 - gitlab-ci: Re-add kernels for bare-metal
4667 - gitlab-ci: Download traces from MinIO
4668 - gitlab-ci: Upload tracie artifacts to MinIO
4669 - gitlab-ci: Fix needs: of the arm64 LAVA test jobs
4670 - ci: Upload images of failed replays to MinIO
4671 - ci: Use smaller glxgears trace
4672 - ci: Prefix tracie artifacts with the device name
4673 - ci: Test with more traces
4674 - ci: Disable trace testing on Mali T760
4675 - ci: Fix the overwriting of traces.yml for baremetal
4676 - ci: Namespace trace artifacts to the job number
4677 - ci: Always print status code of HTTP uploads in tracie
4678 - ci: Print load stats after running dEQP
4679 - ci: Fix URL for glslang
4680 - ci: Don't ship vk-build-programs after building dEQP
4681 - ci: Split building of libdrm to its own script
4682 - ci: Build kernels and rootfs for x86 devices
4683 - ci: Upload reference images for traces
4684 - ci: Print URL to image diff when a trace replay fails
4685 - ci: Generate MinIO credentials within LAVA jobs
4686 - ci: Set date in LAVA DUTs from NTP servers
4687 - ci: Build-test Panfrost tools
4688 - ci: Upload traces' reference and actual images to MinIO
4689 - ci: Download traces from MinIO in baremetal runs
4690 - ci: Remove kernel module build that slipped in
4691 - ci: Actually upload trace artifacts to MinIO for baremetal
4692 - ci: Use a rootfs tarball for NFS root, instead of a ramdisk (for LAVA)
4696 - nir/lower_idiv: Port recent LLVM fixes to emit_udiv
4697 - radv: Fix various non-critical integer overflows
4698 - aco: Fix integer overflows when emitting parallel copies during RA
4699 - amd/common: Fix various non-critical integer overflows
4703 - freedreno: Add missing break statement.
4704 - llvmpipe: Fix variable name.
4705 - r600/sfn: Initialize VertexStageExportForGS m_num_clip_dist member variable.
4706 - panfrost: Ensure final.no_colour is initialized.
4707 - r600/sfn: Use correct setter method.
4708 - freedreno: Add missing va_end.
4709 - pan/bi: Initialize struct fma_op_info member extended.
4710 - zink: Check fopen result.
4711 - etnaviv: Fix memory leak on error path.
4712 - panfrost: Fix printf format specifier.
4713 - r300g: Remove extra printf format specifiers.
4714 - vdpau: Fix wrong calloc sizeof argument.
4715 - mesa: Fix NetBSD compiler macro.
4716 - Switch from cElementTree to ElementTree.
4717 - intel/genxml: Migrate from deprecated xml.etree.ElementTree getchildren.
4718 - rbug: Fix rbug_delete_vs_state lock acquisition.
4719 - nir: Add nir_lower_clip_disable.c to SCons build.
4720 - util: Fix SCons build.
4721 - util: Fix memory leaks in unit test.
4722 - meson: Fix lmsensors warning message.
4723 - vulkan: Fix memory leaks.
4724 - freedreno: Fix file descriptor leak.
4725 - svga: Fix unused printf argument.
4726 - freedreno: Check file descriptor before write.
4727 - panfrost: Delete debug allocated syncobj.
4731 - st/mesa: fix corrupted texture levels, when adding more levels than expected
4735 - glsl: subroutine signatures must match exactly
4736 - nvir: don't use designated initialisers in C++ code
4737 - intel/compiler: don't propagate cmp to add if add is saturated
4738 - mesa: change error code of \*TextureSubImage\* for incorreect target
4739 - nine: fix incorrect calculation of layer count for 3D textures
4743 - gallium/swr: Fix compilation warnings
4744 - swr: Bump maximum 2D texture size to 16kx16k
4748 - nv50: Clear nv50_ir_prog_info of dead and codegen specific variables