Lines Matching +full:fossilize +full:- +full:runner
1 Mesa 20.1.0 Release Notes / 2020-05-27
21 ---------------
25 2109055d7660514fc4c1bcd861bcba9db00c026119ae222720111732dba27c83 mesa-20.1.0.tar.xz
28 ------------
30 - GL_ARB_compute_variable_group_size on i965.
31 - GL_EXT_depth_bounds_test on Iris.
32 - GL_EXT_texture_shadow_lod on radeonsi, nvc0.
33 - GL_NV_alpha_to_coverage_dither_control on radeonsi
34 - GL_NV_copy_image on all gallium drivers.
35 - GL_NV_pixel_buffer_object on all gallium drivers, i915, i965, swrast.
36 - GL_NV_viewport_array2 on nvc0 (GM200+).
37 - GL_NV_viewport_swizzle on nvc0 (GM200+).
38 - VK_AMD_memory_overallocation_behavior on RADV.
39 - VK_KHR_shader_non_semantic_info on Intel, RADV.
40 - GL_EXT_draw_instanced on gles2
41 - VK_KHR_8bit_storage for ACO on GFX8+
42 - VK_KHR_16bit_storage for ACO on GFX8+ (storageInputOutput16 is still
44 - shaderInt16 for ACO on GFX9+
45 - VK_KHR_shader_float16_int8 for ACO on GFX8+ (shaderFloat16 is still
47 - VK_EXT_robustness2 on Intel, RADV.
48 - Add Rocket Lake (RKL) support on anvil and iris.
51 ---------
53 - Reproduceable i915 gpu hang Intel Iris Plus Graphics (Ice Lake 8x8
55 - glsl: regression affecting shader compilation time
56 - freedreno: glamor issue with x11 desktops
57 - [gles3] supertuxkart: some textures are incorrect
58 - Double lock in fbobject.c
59 - [bisected] Steam crashes when newest Iris built with LTO
60 - i965/vec4: opt_cse_local cause the out of bound array access
61 - NIR: Regression on shader using 8/16-bit integers
62 - lp_bld_intr.c:70:16: error: use of undeclared identifier
64 - Deadlock in anv_timelines_wait()
65 - post_version.py does not work with release candidates
66 - post_version.py does not work with release candidates
67 - radv regression on android
68 - src\util\meson.build:294:4: ERROR: Program or command 'winepath' not
70 - debug builds are massively broken on Windows
71 - heavy glitches on amd ryzen 5 since version 20.x
72 - zink asserts with 32-bit boolean
73 - Dirt: Showdown bad performance and broken rendering with enabled
75 - gravit & Firefox WebGL broken since
78 - mesa 20.0.5 causing kitty to crash
79 - radeonsi: "Torchlight II" trace showing regression on mesa-20.0.6
81 - [RADV/LLVM/ACO/Regression] After mesa commit
83 - Android building error after commit 2ab45f41
84 - iris: Crash when trying to capture window in OBS Studio
85 - Properly annotate control flow convergence points
86 - intel/compiler: Register coalesce doesn't move conditional modifiers
87 - [bisected] [iris] mpv under wayland: failed to import supplied
89 - [Bisected][Iris] piglit.spec.!opengl 1_1.max-texture-size crashes on
91 - anv : android deqp assert
92 dEQP-VK.api.external.memory.android_hardware_buffer.dedicated.image#export_import_bind_bind
93 - GL cts gtf30.GL3Tests.sgis_texture_lod.sgis_texture_lod_basic_getter
95 - freedreno/a6xx: texture cache vs realloc_bo()
96 - [Bisected]
97 dEQP-VK.subgroups.ballot_mask.ext_shader_subgroup_ballot.\* failures
98 - dEQP-VK.subgroups.size_control.compute.\* crashes on HSW and TGL
99 - zink: framebuffer and pipeline caches accumulate due to
101 - FTBFS due to LLVM commit 2dea3f129878 (LLVMVectorTypeKind is gone)
102 - [r600/Turks] 20.0.2: modesetting/radeon driver SIGABRT at loading X
104 - piglit spec.!opengl 1.0.gl-1.0-fpexceptions crash on Iris
105 - ci: Update the Wine version
106 - SPIR-V: Failure in dEQP-VK.graphicsfuzz.control-flow-switch
107 - SPIR-V: OpConvertUToPtr from spec constant fails to compile
108 - ACO: Regression: Texture corruption
109 - radv: Reading ViewportIndex in fragment shader returns garbage
110 - piglit
111 spec.arb_gpu_shader_fp64.execution.arb_gpu_shader_fp64-vs-non-uniform-control-flow-ssbo
113 - piglit
114 spec/arb_gpu_shader_fp64/execution/built-in-functions/vs-sign-neg-abs.shader_test
116 - [ANV] gfxbench Aztec Ruins misrenders on gen11+
117 - glxinfo cmd crashed
118 - radeonsi: GL_LINES rendering is affected by GL_POINT_SPRITE
119 - nir: nir_lower_returns can't handle nested loops
120 - Graphic artifacts with Mesa 20.0.4 on intel HD 510 GPU
121 - [Iris] [Bisected] Some KHR-GL46.arrays_of_arrays_gl. tests are
123 - Mesa 20 regression makes Lightsprint demos crash
124 - metro redux games crash upon loading certain levels on amdgpu
125 - dri_common.h:58:8: error: unknown type name '__GLXDRIdrawable'
126 - Graphical glitches on Intel Graphics when Xorg started on Iris driver
127 - GL/GLES test crashes on G33/i915 platforms
128 - GL/GLES test crashes on G33/i915 platforms
129 - GL/GLES test crashes on G33/i915 platforms
130 - SIGSEGV src/compiler/glsl/ast_function.cpp:53
131 - manywin aborts with "i965: Failed to submit batchbuffer: Invalid
133 - manywin aborts with "i965: Failed to submit batchbuffer: Invalid
135 - manywin aborts with "i965: Failed to submit batchbuffer: Invalid
137 - manywin aborts with "i965: Failed to submit batchbuffer: Invalid
139 - v3d: transform feedback issue
140 - radv: Enable TC-compat HTILE in VK_IMAGE_LAYOUT_GENERAL.
141 - radv:
142 …dEQP-VK.binding_model.descriptorset_random.sets4.noarray.ubolimitlow.sbolimitlow.imglimitlow.noiub…
144 - radv: RAVEN fails
145 dEQP-VK.pipeline.timestamp.misc_tests.reset_query_before_copy
146 - buffer overflow in nouveau driver on mesa 20.0.2
147 - xmlconfig sha1 code has overflow and possible bug
148 - enable storageBuffer16BitAccess feature in radv for SI and CIK
149 - Build Fails with Clang Shared Library
150 - Thousands of 32 bit regressions in VulkanCTS and GL test suites due
151 to handling of cross-invocation
152 - anv: isl assert when running dEQP-VK.geometry.layered.3d.*.readback
153 - Weston drm-backend.so seems to fail with Mesa master and
155 - freedreno/turnip: Don't request pixlodenable when we don't use it
156 - VulkanCTS uniform_buffer_block_geom spins forever
157 - freedreno: dEQP-GLES3.functional.fbo.msaa.4_samples.r16f flakiness in
159 - src\util\meson.build:291:4: ERROR: Program or command 'winepath' not
161 - RADV: flickering textures in Q.U.B.E. 2 through Proton
162 - Missing ENDBR in entry_x86-64_tls.h, entry_x86_tls.h and
164 - [regression][bisected] Android build test fails:
166 - Missing ENDBR in rtasm_x86sse.c
167 - src/intel/tools/aubinator_viewer.cpp:383:52: error: format ‘%lx’
169 ‘uint64_t {aka long long unsigned int}’ [-Werror=format=]
170 - src/compiler/glsl/ast_to_hir.cpp:2134: ir_rvalue\*
173 - process_test fails on macOS
174 - Vulkan Overlay is blinking
175 - Regression: 9d64ad2fe79 broke Rocket League
176 - GameMaker games (Memoranda and Undertale) + amdgpu — Segmentation
178 - Civilization VI - Animated leader characters small black squares
180 - [ACO] Reliable crash with RPCS3 that is not present with LLVM
181 - [RADV] vkCmdBindTransformFeedbackBuffersEXT pSizes optional parameter
183 - [RadeonSI] - Curse of the Dead Gods (1123770) - Lighting is not
185 - soft-fp64: \__fsat64 incorrectly returns NaN for a NaN input. It
187 - Hang when using glWaitSync with multithreaded shared GL contexts
188 - RPCS3 / Persona 5 - Performance regression [RADV / Navi]
189 - [ANV] Rendering corruption in Shadow of the Tomb Raider
190 - src/compiler/glsl/glcpp/glcpp-parse.y:1297: \_token_print: Assertion
192 - [CTS] dEQP-VK.descriptor_indexing.\* fails on RADV/LLVM
193 - Unigine Valley failure / assert
194 - [Gen9/icl] [Bisected] [Regression]
195 dEQP-GLES3.functional.shaders.loops.short_circuit.do_while_fragment
197 - [RadeonSI][gfx10/navi] Kerbal Space Program crash: si_draw_vbo:
199 - Budget Cuts hits VK_AMD_shader_fragment_mask assert
200 - Follow-up from "i965/blorp: Don't resolve HiZ unless we're
202 - crash in vc4_write_uniforms with shaders involving YUV textures
203 - Corrupted output with vaapi 10 bit -> 8 bit transcoding on AMD RAVEN
204 - tessellator.cpp:78:7: error: 'fmin' is missing exception
206 - Please add Raspberry Pi 4 to features.txt
207 - Build failure with bison 2.3.
208 - Mesa build fails on 32 bit architecture
209 - Mesa build fails on 32 bit architecture
210 - Incorrect rendering with vaapi + uyvy422
211 - V3D/Broadcom (Raspberry Pi 4) - GLES 3.1 - GL_EXT_texture_norm16
213 - mesa-20.0.0/src/amd/compiler/aco_instruction_selection.cpp:7221:55:
215 - i965 assertion failure in fallback_rgbx_to_rgba
216 - vaapi bob deinterlacer produces wrong output height on AMD
217 - Compute copies do not handle SUBSAMPLED formats
218 - Please document RADV_TEX_ANISO variable in envvars.html
219 - unexpected CI failure
220 - Multiple glapi_mapi_tmp.h
221 - drisw crashes on calling NULL putImage on EGL surfaceless platform
223 - VRAM leak with vuilkan external memory + opengl memory objects
224 - [radeonsi][vaapi][bisected] invalid VASurfaceID when playing
226 - [RADV] GPU hangs while the cutscene plays in the game Assassin's
228 - ACO: The Elder Scrolls Online crashes on startup (Navi)
229 - Broken rendering of glxgears on S/390 architecture (64bit, BigEndian)
230 - aco: sun flickering with Assassins Creeds Origins
231 - !1896 broke ext_image_dma_buf_import piglit tests with radeonsi
232 - aco: wrong geometry with Assassins Creed Origins on GFX6
233 - valgrind errors since commit a8ec4082a41
234 - src/broadcom/qpu/qpu_pack.c:962:25: error: implicit declaration of
236 [-Werror,-Wimplicit-function-declaration] mux_b =
237 ffs(desc->mux_b_mask) - 1;
238 - X fails to start with amdgpu and Mesa 20.1 on Fedora
239 - GPU hangs in Factorio on Radeon RX 5700 XT (MSI GAMING X)
240 - OSMesa osmesa_choose_format returns a format not supported by
242 - Build error with VS on WIN
243 - Using EGL_KHR_surfaceless_context causes spurious "libEGL warning:
245 - !3460 broke texsubimage test with piglit on zink+anv
246 - VERSION needs to be bumped for trunk master
247 - The screen is black when using ACO
250 -------
254 - anv/android: fix assert in anv_import_ahw_memory
258 - gallium: enable EGL_EXT_image_dma_buf_import_modifiers
263 - cube_face_coord: Use fabsf instead of fabs since we know it's floats
264 - cube_face_index: Use fabsf instead of fabs since we know it's floats
265 - aco: Minor optimization in spill_ctx constructor
266 - aco: pass vars by const &
267 - Fix promotion of floats to doubles
271 - docs/features: add v3d driver
272 - nir/linker: remove reference to just SPIR-V linking
273 - v3d/tex: don't configure tmu config 1 if not needed
274 - v3d/tex: Configuration Parameter 1 can be only skipped if P2 can be
276 - v3d/packet: fixing TMU_Config_Parameter_2 definition
277 - nir: add nir_tex_instr_need_sampler helper
278 - v3d: support for textureQueryLOD
282 - gitlab-ci: Automated testing with OpenGL traces
283 - gitlab-ci: Fix traces caching in tracie
284 - gitlab-ci: Check the Mesa version used for tracie tests
288 - pan/midgard: Break out one-src read_components
289 - pan/midgard: Implement mixed-type constant packing
290 - panfrost: Avoid overlapping copy
291 - pan/midgard: Check for null consts
292 - pan/midgard: Remove unused variable
293 - panfrost: Use size0 when calculating the offset to a depth level
294 - pan/midgard: Fix scheduling issue with csel + render target reference
295 - panfrost: Simplify swizzle translation
296 - panfrost: Update comment about magic number relating to barriers
297 - panfrost: Ensure compute shader_meta is zeroed
298 - panfrost: Identify mali_shared_memory structure
299 - panfrost: Unify bifrost_scratchpad with mali_shared_memory
300 - panfrost: Rename bifrost_framebuffer->mali_framebuffer
301 - panfrost: Rename unknown2_8 to padding
302 - panfrost: Allocate RAM backing of shared memory
303 - pan/midgard: Track pressure when scheduling ld/st
304 - pan/midgard: Fix missing prefixes
305 - pan/midgard: Fix swizzles harder
306 - pan/midgard: Implement barriers
307 - pan/midgard: Allow jumping out of a shader
308 - pan/midgard: Fix 32/64 mixed swizzle packing
309 - pan/midgard: Use dummy tag for empty shaders
310 - pan/midgard: Improve barrier disassembly
311 - pan/midgard: Overhaul tag handling
312 - pan/midgard: Imply next tags
313 - pan/midgard: Infer tags entirely
314 - pan/midgard: Set xyzx swizzle for load_compute_arg
315 - pan/midgard: Identify stack barrier flag
316 - pan/midgard: Don't crash with constants on unknown ops
317 - pan/midgard: Use fprintf instead of printf for constants
318 - pan/decode: Remove extraneous newline
319 - pan/decode: Add \`minimal\` mode
320 - pan/decode: Cleanup pandecode_jc
321 - panfrost: Implement PAN_DBG_SYNC with pandecode/minimal
322 - panfrost: Print synced traces to stderr
323 - panfrost: Rewrite scoreboarding routines
324 - panfrost: Update scoreboarding notes
325 - panfrost: Cleanup transfer_map
326 - panfrost: Avoid reading GPU memory when packing vertices
327 - panfrost: Debitfieldize mali_uniform_buffer_meta
328 - panfrost: Remove enum panfrost_memory_layout
329 - panfrost: Remove dirty tracking
330 - panfrost: Remove old comment
331 - panfrost: Remove old hack
332 - panfrost: Remove flush_frontbuffer
333 - pan/midgard: Identify clamp(x, -1.0, 1.0) flag
334 - panfrost: Move checksum routines to root panfrost
335 - panfrost: Move pan_afbc.c to root
336 - panfrost: Move format translation to root
337 - panfrost: Rewrite texture descriptor creation logic
338 - nir: Add SSBO->global lowering pass
339 - pan/midgard: Lower SSBOs in NIR
340 - pan/midgard: Implement nir_intrinsic_get_buffer_size
341 - pan/midgard: Implement load/store_shared
342 - panfrost: Combine get_index_buffer with bound computation
343 - panfrost: Implement index buffer cache
344 - pan/decode: Dump scratchpad size if present
345 - pan/midgard: Don't spill near a branch
346 - panfrost: Fix gl_VertexID/InstanceID
347 - panfrost: Fix padded_vertex_count generation
348 - panfrost: Update spilling comment framebuffer->shared
349 - panfrost: Don't set shared->unk0
350 - panfrost: Fix param getting
351 - panfrost: Default to 256 threads for TLS
352 - panfrost: Reserve an extra page for spilling
353 - panfrost: Simplify stack shift calculation
354 - panfrost: Expose PIPE_CAP_PRIMITIVE_RESTART
355 - panfrost: Add PAN_MESA_DEBUG=gles3 option
356 - panfrost: Increase SSBO/image limit from 4->8
357 - pan/midgard: Allow inverted inverted ops
358 - pan/midgard: Allow fusing inverted sources for inverted ops
359 - pan/midgard: Partially fix 64-bit swizzle alignment
360 - pan/midgard: Extract nir_ssa_index helper
361 - pan/midgard: Add LDST_ADDRESS property
362 - pan/midgard: Fix load/store argument sizing
363 - pan/midgard: Round up bytemasks when promoting uniforms
364 - pan/midgard: Force address alignment
365 - pan/midgard: Add address analysis framework
366 - pan/midgard: Use address analysis for globals, etc
367 - pan/decode: Calm an assert to a pandecode error
368 - pan/decode: Restore bifrost sample_locations
369 - pan/decode: Fix tiler weights printing
370 - pan/decode: Skip analysis for Bifrost tiler structures
371 - pan/bi: Add discard ops
372 - pan/bi: Add ICMP.GL.NEQ op
373 - pan/bi: Move notes on FMA opcodes from disassembler
374 - pan/bi: Introduce CSEL4 class
375 - pan/bi: Move notes on ADD ops to notes file
376 - pan/bi: Decode FMA_SHIFT properly
377 - pan/bi: Add v4i8 mode to FMA_SHIFT
378 - pan/bi: Identify extended FMA opcodes
379 - pan/bi: Decode ADD_SHIFT properly
380 - pan/bi: Combine LOAD_VARYING_ADDRESS instructions by type
381 - pan/bi: Squash LD_ATTR ops together
382 - pan/bi: Structify FMA_FADD
383 - pan/bi: Move some definitions from disasm to bifrost.h
384 - panfrost: Add note about preloaded varyings
385 - pan/bi: Gut old compiler
386 - pan/bi: Stub out new compiler
387 - pan/bi: Add the control flow graph
388 - pan/bi: Add src/dest fields to bifrost_instruction
389 - pan/bi: Add class properties
390 - pan/bi: Add modifiers to bi_instruction
391 - pan/bi: Add BI_GENERIC property
392 - pan/bi: Factor out enum bifrost_minmax_mode
393 - pan/bi: Add a bifrost_roundmode field
394 - pan/bi: Add bifrost_minmax_mode field
395 - pan/bi: Add bi_load structure
396 - pan/bi: Pull out bifrost_load_var
397 - pan/bi: Add bi_load_vary structure
398 - pan/bi: Add PAN_SCHED\_\* flags
399 - pan/bi: Add bi_clause, bi_bundle abstractions
400 - pan/bi: Add dest_type field to bifrost_instruction
401 - pan/bi: Add special indices
402 - pan/bi: Add constant field to bi_instruction
403 - pan/bi: Add class-specific ops
404 - pan/bi: Add clause header fields to bi_clause
405 - pan/bi: Clarify special op scheduling
406 - pan/bi: Add swizzles
407 - pan/bi: Add source type for conversions
408 - pan/bi: Add EXTRACT, MAKE_VEC synthetic ops
409 - pan/bi: Add constants to bi_clause
410 - pan/bi: Add pred/successors to build CFG
411 - pan/bi: Extract bifrost_branch structure
412 - pan/bi: Add bi_branch data
413 - pan/bi: Add CSEL condition
414 - pan/bi: Add high-latency property for classes
415 - pan/bi: Add quirks system
416 - pan/bi: Add IR iteration macros
417 - pan/bi: Move some print routines out of the disasm
418 - pan/bi: Add BIR manipulation routines to bir.c
419 - pan/bi: Move bi_interp_mode_name to bi_print
420 - pan/bi: Add bi_instruction printing
421 - pan/bi: Add bi_print_bundle for printing bi_bundle
422 - pan/bi: Add bi_print_clause
423 - pan/bi: Add bi_print_block
424 - pan/bi: Add bi_print_shader
425 - pan/bi: Lower and optimize NIR
426 - pan/bi: Walk through the NIR control flow graph
427 - pan/bi: Improve block printing
428 - pan/bi: Don't print types for unconditional branches
429 - pan/bi: Print branch target
430 - pan/bi: Add instruction emit/remove helpers
431 - pan/bi: Call nir_lower_io_to_temporaries in cmdline
432 - pan/bi: Add support for if-else blocks
433 - pan/bi: Handle loops when ingesting CFG
434 - pan/bi: Handle jumps (breaks, continues)
435 - pan/bi: Fix destination printing
436 - pan/bi: Implement nir_intrsinic_load_interpolated_input
437 - pan/bi: Add blend_location to IR for BI_BLEND
438 - pan/bi: Add bi_schedule_barrier helper
439 - pan/bi: Implement store_output for fragment shaders
440 - pan/bi: Implement load_input for vertex shaders
441 - pan/bi: Add helpers for creating temporaries
442 - pan/bi: Implement store_vary for vertex shaders
443 - pan/bi: Add preliminary LOAD_UNIFORM implementation
444 - pan/bi: Implement load_const
445 - pan/bi: Add dummy scheduler
446 - pan/bi: Rename next-wait to simply 'wait'
447 - pan/bi: Fix Android.mk
448 - panfrost: Move mir_to_bytemask to common code
449 - pan/bi: Generalize swizzles to avoid extracts
450 - pan/bi: Introduce writemasks
451 - pan/bi: Remove bi_load
452 - pan/bi: Lower vec\* to writemasks in NIR
453 - pan/bi: Add initial handling of ALU ops
454 - pan/bi: Allow inlining constants
455 - pan/bi: Implement fsat as mov.sat
456 - pan/bi: Add a bunch of ALU ops
457 - pan/bi: Add BI_SPECIAL\_\* enum
458 - pan/bi: Handle special ops in NIR->BIR
459 - pan/bi: Implement fabs, fneg as fmov with mods
460 - pan/bi: Disable lower_sub
461 - pan/bi: Add isub op
462 - pan/bi: Import algebraic pass from midgard
463 - pan/bi: Implement nir_op_bcsel
464 - pan/bi: Lower b2f to bcsel
465 - pan/bi: Specify comparison op for BI_CMP
466 - pan/bi: Print source types unconditionally
467 - pan/bi: Implement comparison opcodes via BI_CMP
468 - panfrost: Promote midgard_program to panfrost/util
469 - pan/midgard: Remove unused iterators
470 - pan/midgard: Adjust sysval-related prototypes
471 - pan/midgard: Remove indexing dependency of sysvals
472 - pan/midgard: Decontextualize midgard_nir_assign_sysval_body
473 - pan/midgard: Remove dest_override sysval argument
474 - panfrost: Move Midgard sysval code to common Panfrost
475 - pan/bi: Switch to panfrost_program
476 - pan/bi: Implement sysvals
477 - pan/midgard: Localize \`visited\` tracking
478 - pan/midgard: Decontextualize liveness analysis core
479 - pan/midgard: Sync midgard_block field names with Bifrost
480 - pan/midgard: Subclass midgard_block from pan_block
481 - panfrost: Move liveness analysis to root panfrost/
482 - panfrost: Sync Midgard/Bifrost control flow
483 - pan/bi: Paste over bi_has_arg
484 - pan/bi: Add bi_bytemask_of_read_components helpers
485 - pan/bi: Add bi_next/prev_op helpers
486 - pan/bi: Add bi_max_temp helper
487 - pan/bi: Add liveness analysis pass
488 - pan/bi: Add dead code elimination pass
489 - pan/bi: Implement nir_op_ffma
490 - pan/bi: Fix swizzle for second argument to ST_VARY
491 - panfrost: Move lcra to panfrost/util
492 - pan/midgard: Remove incorrect comment in RA
493 - pan/bi: Minor fixes in iteration macros
494 - pan/bi: Fix vector handling of readmasks
495 - pan/bi: Fix missing src_types
496 - pan/bi: Add register allocator
497 - pan/bi: Interpret register allocation results
498 - pan/bi: Setup initial clause packing
499 - pan/bi: Sketch out instruction word packing
500 - pan/bi: Add packing for register control field
501 - pan/bi: Pack register fields
502 - pan/bi: Add missing \__attribute__((packed))
503 - pan/bi: Assign registers to ports
504 - pan/bi: Route through first_instruction field
505 - pan/bi: Model 3-bit Bifrost srcs in IR
506 - pan/bi: Add struct bifrost_fma_fma
507 - pan/bi: Pack BI_FMA ops
508 - pan/bi: Pack fadd32
509 - pan/bi: List ADD classes in bi_pack_add
510 - pan/bi: Generalize bi_get_src a bit
511 - pan/bi: Pass second src for load_vary ops
512 - pan/bi: Emit load_vary ops
513 - pan/bi: Skip over data registers in port assignment
514 - pan/bi: Route through clause header
515 - pan/bi: Pretty-print clause types in disassembler
516 - pan/bi: Don't hide SCHED_ADD inside HI_LATENCY
517 - pan/bi: Track clause types during scheduling
518 - pan/bi: Flesh out ATEST in IR
519 - pan/bi: Add ATEST packing
520 - pan/bi: Flesh out BI_BLEND
521 - pan/bi: Pack BI_BLEND
522 - pan/bi: Implement FMA/MOV without modifiers
523 - pan/bi: Add bi_emit_before helper
524 - pan/bi: Add move lowering pass
525 - pan/bi: Pack a constant quadword
526 - pan/bi: Document constant related errata(?)
527 - pan/bi: Index out constants in instructions
528 - pan/bi: Include UBO index for sysval reads
529 - pan/bi: Add bi_load32_components helper
530 - pan/bi: Pack ld_ubo ops
531 - pan/bi: Pack ld_var_addr
532 - pan/bi: Flesh out st_vary IR
533 - pan/bi: Generalize data register setting
534 - pan/bi: Add store_channels property
535 - pan/bi: Pack st_vary
536 - pan/bi: Pack LD_ATTR
537 - pan/bi: Lower bool to ints
538 - pan/bi: Remove hacks for 1-bit booleans in IR
539 - pan/bi: Add \`soft\` NIR->BIR condition translation
540 - pan/bi: Implement csel fusing
541 - pan/bi: Respect shift when printing immediates
542 - pan/bi: Use bi_lookup_immediate when packing
543 - pan/bi: Default csel to "!= 0" mode
544 - pan/bi: Pack csel4 opcodes
545 - pan/bi: Ingest vecN directly (again)
546 - pan/bi: Lower combines to rewrites for scalars
547 - pan/bi: Rewrite aligned vectors as well
548 - panfrost: Split panfrost_device from panfrost_screen
549 - panfrost: Isolate panfrost_bo_access_for_stage to pan_cmdstream.c
550 - panfrost: Inline reference counting routines
551 - panfrost: Move pan_bo to root panfrost
552 - pan/bit: Link standalone compiler with en/decoder
553 - panfrost: Move device open/close to root panfrost
554 - pan/bit: Open up the device
555 - panfrost: Stub out G31/G52 quirks
556 - pan/bit: Submit a WRITE_VALUE job as a sanity check
557 - pan/bit: Begin generating a vertex job
558 - pan/bi: Fix overzealous write barriers
559 - pan/bi: Fix off-by-one in scoreboarding packing
560 - pan/bi: Enable precision lowering in standalone compiler
561 - panfrost: Enable PIPE_SHADER_CAP_FP16 on Bifrost
562 - pan/bi: Handle f2f\* opcodes
563 - pan/bi: Ignore swizzle in unwritten component
564 - pan/bi: Finish FMA structures
565 - pan/bi: Fix missing type for fmul
566 - pan/bi: Add FMA16 packing
567 - pan/bi: Pack outmod and roundmode with FMA
568 - pan/bi: Expand out FMA conversion opcodes
569 - pan/bi: Enumerate conversions
570 - pan/bi: Handle standard FMA conversions
571 - pan/bi: Add bifrost_fma_2src generic
572 - pan/bi: Add one-source f32->f16 op
573 - pan/bi: Assert out i16 related converts for now
574 - pan/bi: Handle round opcodes in frontend
575 - pan/bi: Add v2f16 versions of rounding ops
576 - pan/bi: Structify fadd/min/max16
577 - pan/bi: Handle core faddminmax16 packing
578 - pan/bi: Handle abs packing for fp16/FMA add/min
579 - pan/bi: Handle fp16/abs scheduling restriction
580 - pan/bi: Fix handling of constants with COMBINE
581 - pan/bit: Add \`run\` mode to the cmdline
582 - pan/bit: Wire through I/O
583 - pan/bi: Fix writes_component for VECTOR
584 - pan/bi: Use STAGE srcs for scheduler nops
585 - pan/bi: Don't set the back-to-back bit yet
586 - pan/bi: Add cmdline option for verbose disassembly
587 - pan/bi: Fix unused port swapping
588 - pan/bi: Handle fmov class ops
589 - pan/bi: Fix outmod/roundmode flip
590 - pan/bi: Export bi_class_name
591 - pan/bi: Fix duplicated source in ADD.v2f16
592 - pan/bi: Fix negation in ADD.v2f16
593 - pan/bi: Don't gobble zero ports
594 - pan/bi: Allow BI_FMA to take mods
595 - pan/bi: Handle BIFROST_FIRST_WRITE_FMA_P2_READ_P3
596 - pan/bi: Add helper to debug port assignment
597 - pan/bi: Match CSEL argument order with hw
598 - pan/bit: Stub out BIR interpreter
599 - pan/bit: Handle read/write
600 - pan/bit: Add preliminary FMA/ADD/MOV implementations
601 - pan/bit: Implement outmods
602 - pan/bit: Implement floating source mods
603 - pan/bit: Add packing test framework
604 - pan/bit: Add helper for generating floating mod tests
605 - pan/bit: Add verbose printing for tests
606 - pan/bit: Add 16-bit fmod tests
607 - pan/bit: Add FMA tests
608 - pan/bit: Add CSEL to interpreter
609 - pan/bit: Add csel tests
610 - pan/bit: Make run more useful
611 - pan/bit: Add mode to run unit tests
612 - pan/bi: Remove nontrivial SPECIAL ops
613 - pan/bi: Add 32-bit \_FAST packing
614 - pan/bi: Add fp16 support for frcp/frsq
615 - pan/bit: Add special op interpreting
616 - pan/bit: Add special unit test
617 - pan/bi: Implement min/max on FMA
618 - pan/bi: Structify ADD unit add/min/max
619 - pan/bi: Add ADD add/min/max fp32 packing
620 - pan/bi: Set BI_MODS for MINMAX
621 - pan/bi: Fix incorrect abs flip in fma/fadd16
622 - pan/bi: Force ADD scheduling for MINMAX
623 - pan/bit: Unify test frontends
624 - pan/bit: Add min/max support to interpreter
625 - pan/bit: Enable more debug for \`run\`
626 - pan/bit: Add fmin/max16 tests
627 - pan/bit: Wire up add/add op+test
628 - panfrost: Add IS_BIFROST quirk
629 - panfrost: Populate bifrost-specific structs within mali_shader_meta
630 - panfrost: Staticize a few cmdstream functions
631 - panfrost: Unify vertex/tiler structures
632 - panfrost: Set mfbd.msaa.sample_locations on Bifrost
633 - panfrost: Call the Bifrost compiler on bi devices
634 - pan/bi: Fix nondeterministic register packing
635 - pan/midgard: Remove unused max_varying variable
636 - panfrost: Move varying linking to cmdstream
637 - panfrost: Move uniform_count to pan_assemble
638 - panfrost: Pass compiler-appropriate options
639 - pan/bi: Fix backwards registers ports
640 - panfrost: Fix BI_BLEND packing
641 - pan/bi: Let !b2b imply branch_cond
642 - pan/decode: Print Bifrost blend descriptor
643 - panfrost: Drop dependency on nonexistant write_value
644 - pan/bi: Lower fsqrt
645 - pan/midgard: Fix f2u naming confusion
646 - pan/bi: Set BI_ROUNDMODE for BI_CONVERT
647 - pan/bi: Fix incorrect swizzle packing assert
648 - pan/bi: Rewrite conversion packing
649 - pan/bi: ADD packing for CONVERT
650 - pan/bit: Add BI_CONVERT interpretation
651 - pan/bit: Add BI_CONVERT tests
652 - pan/bi: Add disasm for ADD.i8
653 - pan/bi: Disable FMA scheduling for CONVERT
654 - pan/bi: Add BI_TABLE for fast table accesses
655 - pan/bi: Add special op for exp2
656 - pan/bi: Add op for ADD_FREXPM
657 - pan/bi: Add FLOG2_U op to disassembler
658 - pan/bi: Add log_frexpe op to IR
659 - pan/bi: Add frexp_log packing
660 - pan/bi: Add bi_pack_fma_2src helper
661 - pan/bi: Pack ADD_FREXPM
662 - pan/bi: Add log2_help packing
663 - pan/bi: Add \_MSCALE flag for FMA/ADD
664 - pan/bi: Structify FMA_MSCALE
665 - pan/bi: Pack FMA_MSCALE
666 - pan/bi: Add fexp2_fast packing
667 - pan/bi: Split src/dest index printing
668 - pan/bi: Ensure CONSTANT srcs have types
669 - pan/bi: Fix bi_get_immediate with multiple imms
670 - pan/bi: Fix packing with multiple constants
671 - pan/bi: Fix packing with low-nibble-set on hi constant
672 - pan/bi: Fix lower_combine swizzle rewrite
673 - pan/bi: Add fexp2 implementation
674 - pan/bi: Implement flog2
675 - pan/bi: Fix vec2/3 handling
676 - pan/bi: Handle st_vary with <4 components
677 - pan/bi: Try to reuse constants in ALU
678 - pan/bi: Workaround constant packing errata
679 - pan/bi: Structify add and min/max fp16 ADD
680 - pan/bi: Pack ADD.v2f16
681 - pan/bi: Pack MAX.v2f16
682 - pan/bi: Dump extra bits for disasm
683 - pan/bi: Round constants to 32-bit
684 - pan/bi: Lower special ops to 32-bit
685 - pan/bit: Add FREXP interp support
686 - pan/bit: Add frexp_log test
687 - pan/bit: Add BI_REDUCE_FMA interp
688 - pan/bit: Add FMA_REDUCE test
689 - pan/bit: Add log2 helper interp
690 - pan/bit: Add BI_TABLE test
691 - pan/bit: \_MSCALE interp
692 - pan/bit: Add FMA_MSCALE test
693 - pan/bit: Add fexp2_fast interp
694 - pan/bit: Add fexp2_fast test
695 - pan/bit: Add constants test
696 - pan/bit: Add fp16 min/max tests
697 - pan/bi: Print tex_compact coordinates
698 - pan/bi: Document when dual-tex is triggered
699 - pan/bi: Disassemble f16 dual tex
700 - pan/bi: Structify TEX compact
701 - pan/bi: Include TEX_COMPACT f16 opcode
702 - pan/bi: Feed data register to BI_TEX
703 - pan/bi: Add normal/compact/dual switch to IR
704 - pan/bi: Stub out tex_compact logic
705 - pan/bi: Generate TEX_COMPACT instruction
706 - pan/bi: Pack TEX compact instructions
707 - pan/bi: Assert out multiple textures
708 - panfrost: Fix crashes with small BOs
709 - panfrost: Assert on unimplemented fragcoord etc
710 - panfrost: Set clear_color_[12] in the extra fb desc
711 - panfrost: Add tentative bifrost_texture_descriptor
712 - panfrost: decode textures and samplers on bifrost
713 - pan/decode: Remove is_zs weirdness
714 - panfrost: Identify texture layout field
715 - panfrost: The texture descriptor has a pointer to a trampoline
716 - pan/bi: Pack fp16 ATEST
717 - pan/bi: Passthrough type for ATEST
718 - pan/bi: Passthrough blend types
719 - pan/bi: Assign blend descriptor for BLEND op
720 - pan/bi: Add missing BI_VECTOR
721 - pan/bi: Fix ADD.v4i8 opcode
722 - pan/bi: Eliminate writemasks in the IR
723 - pan/bi: Rename BI_SWIZZLE to BI_SELECT
724 - pan/bi: Pack FMA SEL16
725 - pan/bi: Pack FMA SEL8
726 - pan/bi: Pack ADD SEL16
727 - pan/bi: Force BI_SELECT arguments scalar
728 - pan/bit: Interpret BI_SELECT
729 - pan/bit: Add SELECT tests
730 - pan/bi: Fix RA wrt 16-bit swizzles
731 - pan/bi: Implement 16-bit COMBINE lowering
732 - nir: Move nir_lower_mediump_outputs from ir3
733 - ir3: Use shared mediump output lowering
734 - pan/bi: Add bool->float opcodes
735 - pan/bi: Add CSEL.64 opcode
736 - pan/bi: Add some 8-bit compares
737 - pan/bi: Add 64-bit int compares
738 - pan/bi: Add FCMP.GL.v2f16 on ADD opcode
739 - pan/bi: Add CSEL.8 opcode
740 - pan/bi(t): Fix SELECT tests
741 - pan/bi: Deduplicate csel/cmp cond
742 - pan/bi: Remove bi_round_op
743 - pan/bi: Structify FMA FCMP
744 - pan/bi Strucitfy ADD FCMP 32
745 - pan/bi: Structify FMA FCMP16
746 - pan/bi: Structify ADD FCMP16
747 - pan/bi: Structify FMA ICMP 32
748 - pan/bi: Structify FMA ICMP 16
749 - pan/bi: Structify ADD ICMP 32
750 - pan/bi: Fix source mod testing for CMP
751 - pan/bi: Pack FMA 32 FCMP
752 - pan/bi: Factor out fp16 abs logic
753 - pan/bi: Pack fma.fcmp16
754 - pan/bi: Relax double-abs condition
755 - pan/bit: Prepare condition evaluation for vectors
756 - pan/bit: Interpret CMP
757 - pan/bi: Add initial fcmp test
758 - pan/bi: Add bitwise modifiers
759 - pan/bi: Pack BI_BITWISE
760 - pan/bi: Handle iand/ior/ixor in NIR->BIR
761 - pan/bit: Interpret BI_BITWISE
762 - pan/bit: Add BITWISE test
763 - panfrost: Fix BO reference counting
764 - panfrost: Move Bifrost IR indexing to common
765 - pan/bi: Use common IR indices
766 - pan/mdg: Remove nir_alu_src_index
767 - pan/mdg: Use PAN_IS_REG
768 - pan/mdg: SSA_FIXED_MINIMUM already covered by PAN_IS_REG
769 - pan/mdg: Don't break SSA
770 - pan/mdg: Remove goofy 16-bit comment
771 - pan/mdg: Remove old hack
772 - pan/mdg: Set lower_flrp16
773 - pan/bi: Share ALU type printing
774 - pan/mdg: Add type fields to IR
775 - pan/mdg: Track ALU src types
776 - pan/mdg: Track ALU dest type
777 - pan/mdg: Another goofy comment gone
778 - pan/mdg: Track a primary type for I/O
779 - pan/mdg: Denoise prints
780 - pan/mdg: Track v_mov type (force uint32 for now?)
781 - pan/mdg: Track texture types
782 - pan/mdg: Set texture full fields at pack time
783 - pan/mdg: Move sampler_type emission to pack time
784 - pan/mdg: Lower specials to 32-bit
785 - pan/mdg: Specialize swizzle to type
786 - pan/mdg: Always print the mask
787 - pan/mdg: Make some branch targets more explicit
788 - pan/mdg: Don't crash on unknown branch target
789 - pan/mdg: Pass through some types from scheduling
790 - pan/mdg: Move condense_writemask to disasm
791 - pan/mdg: Ensure fdot is scalar out in disasm
792 - pan/mdg: Replicate 16-bit swizzles
796 - lima/parser: Fix RSW depth test parsing
797 - lima/parser: Extend AUX0 findings
798 - lima/parser: Change value name in RSW parser
799 - lima/parser: Extend rsw parsing showing strings instead of numbers
800 - gitlab-ci: lima: Add flaky tests to the skips list
801 - gitlab-ci: Enable the lima job again
802 - gitlab-ci: Add add a set of lima flakes
803 - lima: Add etc1 support
807 - tracie: correct typo
808 - gitlab-ci: add missing popd to the build-deqp-vk.sh script
809 - gitlab-ci: build gfxreconstruct into the Vulkan testing container
810 - gitlab-ci: build VulkanTools into the Vulkan testing container
811 - gitlab-ci: Change devices format to <api-vendor-deviceId>
812 - gitlab-ci: Add gfxreconstruct traces support
813 - gitlab-ci: Add jobs to be able to test Vulkan
814 - gitlab-ci: Fix indentation and dangerous "\" in the last multiline
816 - gitlab-ci: Remove unneeded python3-pilkit dependency
817 - gitlab-ci: Sort packages to install alphabetically
818 - gitlab-ci: add python3-requests to the test-vk container
819 - gitlab-ci/traces: Add Vulkan sample entries for POLARIS10
820 - gitlab-ci: Don't use buster-backports packages by default for
821 x86_test-vk
822 - gitlab-ci: add Wine, win64's apitrace and DXVK to the Vulkan testing
824 - gitlab-ci: add apitrace's DXGI traces support
825 - gitlab-ci: replay apitrace traces in headless mode
826 - gitlab-ci: add Wine and DXVK env variables to Vulkan's tracie runner
827 - gitlab-ci/traces: Add D3D11 sample entry for POLARIS10
828 - gitlab-ci: Vulkan tracie runner to return last command exit code
829 - gitlab-ci: protect usage of shell variables with double quotes
830 - gitlab-ci: make explicit tracie is gitlab specific
831 - gitlab-ci: adapt query_traces_yaml to gitlab specific changes
832 - gitlab-ci: install winehq-stable to get 5.0 instead of 4.0
833 - Revert "meson,ci: Disable sparse_array tests on windows"
834 - gitlab-ci: update tracie README after changes in main script
835 - gitlab-ci: create always the "results" directory with tracie
836 - gitlab-ci: correct tracie behavior with replay errors
840 - Revert "glx: convert glx_config_create_list to one big calloc"
841 - i965/vec4: Ignore swizzle of VGRF for use by var_range_end()
845 - intel/gen12+: Reserve 4KB of URB space per bank for Compute Engine
846 - intel/gen12+: Set way_size_per_bank to 4
848 Arcady Goldmints-Orlov (7):
850 - compiler/nir: Add support for variable initialization from a pointer
851 - compiler/spirv: Add support for non-constant initializers
852 - Rename nir_lower_constant_initializers to
854 - spirv: Remove outdated SPIR-V decoration warnings
855 - nir: Lower returns correctly inside nested loops
856 - anv: increase minUniformBufferOffsetAlignment to 64
857 - intel/compiler: fix alignment assert in nir_emit_intrinsic
861 - gallium/util: Fix leak in the live shader cache
865 - radv: Allow non-dedicated linear images and buffer.
866 - radv: Do not set SX DISABLE bits for RB+ with unused surfaces.
867 - radv: Optimize emitting index buffer changes.
868 - radv: Do not redundantly set the RB+ regs on pipeline switch.
869 - radeonsi: Fix compute copies for subsampled formats.
870 - amd/llvm: Fix divergent descriptor indexing. (v3)
871 - amd/llvm: Fix divergent descriptor regressions with radeonsi.
872 - radv: Store 64-bit availability bools if requested.
873 - radv: Consider maximum sample distances for entire grid.
874 - radv: Whitespace fixup.
875 - radv: Use correct buffer count with variable descriptor set sizes.
876 - winsys/amdgpu: Retrieve WC flags from imported buffers.
877 - drm-uapi,radv,radeonsi: Add amdgpu_drm.h header.
878 - vulkan/wsi: Add callback to set ownership of buffer.
879 - radv: Add WSI buffers to BO list only if they can be used.
880 - st/dri: Set next in template instead of after creation. (v2)
881 - radeonsi: Count planes for imported textures.
882 - radv: Use actual memory type count for setting app-visible bitset.
883 - radv: Stop using memory type indices.
884 - radv/winsys: Add function to get domains/flags from fd.
885 - radv: Determine memory type for import based on fd.
886 - radv: Expose 4G element texel buffers.
887 - radv: Fix implicit sync with recent allocation changes.
888 - radv: Extend tiling flags to 64-bit.
889 - radv: Provide a better error for permission issues with priorities.
890 - radv/winsys: Remove extra sizeof multiply.
891 - radv: Handle failing to create .cache dir.
892 - radv: Do not close fd -1 when NULL-winsys creation fails.
893 - radv: Implement vkGetSwapchainGrallocUsage2ANDROID.
897 - util/os_socket: Include unistd.h to fix build error
901 - radeonsi: Fix omitted flush when moving suballocated texture
905 - pan/midgard: Add an enum to describe the render targets
906 - pan/midgard: Make sure we pass the right RT id to
908 - pan/midgard: Lower bitfield extract to shifts
909 - pan/midgard: Don't check 'branch && branch->writeout' twice in
911 - pan/midgard: Stop leaking instruction objects in mir_schedule_alu()
912 - panfrost: Fix the damage box clamping logic
913 - pan/midgard: Turn Z/S stores into zs_output_pan intrinsics
914 - pan/midgard: Add nir_intrinsic_store_zs_output_pan support
915 - panfrost: Z24 variants should be sampled as R32UI
916 - panfrost: Add the MALI_WRITES_{Z,S} flags
917 - panfrost: Set the MALI_WRITES_{Z,S} flags when needed
918 - Revert "panfrost: Z24 variants should be sampled as R32UI"
919 - panfrost: Pass the sampler view format when creating a tex descriptor
920 - panfrost: Assign primitive_size.pointer only if writes_point_size()
922 - panfrost: Add an helper to retrieve the currently active shader state
923 - panfrost: Move the batch stack size adjustment out of
925 - panfrost: Move viewport desc emission out of panfrost_emit_for_draw()
926 - panfrost: Move the const buf emission logic out of
928 - panfrost: Move shared mem desc emission out of panfrost_launch_grid()
929 - panfrost: Dissociate shader meta patching from the desc emission
930 - panfrost: Move panfrost_attach_vt_framebuffer() to pan_cmdstream.c
931 - panfrost: Stop using panfrost_emit_for_draw() for compute jobs
932 - panfrost: Simplify panfrost_emit_for_draw() and make it private
933 - panfrost: Add an helper to update the occclusion query part of a
935 - panfrost: Add an helper to update the rasterizer part of a tiler job
937 - panfrost: Prepare things to get rid of panfrost_shader_state.tripipe
938 - panfrost: Prepare shader_meta descriptors at emission time
939 - panfrost: Add a panfrost_sampler_desc_init() helper
940 - panfrost: Move sampler/tex descs emission helpers to pan_cmdstream.c
941 - panfrost: Add an helper to emit a pair of vertex/tiler jobs
942 - panfrost: Drop initial mali_attr_meta.src_offset assignment
943 - panfrost: Ignore BO start addr when adjusting src_offset
944 - panfrost: Prepare attribute for builtins at state creation time
945 - panfrost: Emit attribute descriptors after patching the templates
946 - panfrost: Move the mali_attr.src_offset adjustment to a sub-function
947 - panfrost: Rename panfrost_stage_attributes()
948 - panfrost: Move streamout offset update out of panfrost_draw_vbo()
949 - panfrost: Move vertex/tiler payload initialization out of
951 - panfrost: Inline panfrost_queue_draw() and panfrost_emit_for_draw()
952 - panfrost: Move panfrost_emit_vertex_data() to pan_cmdstream.c
953 - panfrost: Move panfrost_emit_varying_descriptor() to pan_cmdstream.c
954 - panfrost: Re-init the VT payloads at draw/launch_grid() time
955 - panfrost: Use ctx->active_prim in panfrost_writes_point_size()
956 - panfrost: Get rid of ctx->payloads[]
957 - vtn/opencl: add rint-support
961 - turnip: Promote tu_cs_get_size/is_empty to header
962 - turnip: Execute main cs for secondary command buffers
963 - turnip: Advertise 8 bit subpixel precision
964 - ir3: Disable copy prop for immediate ldlw offsets
965 - turnip: Set has_gs in ir3_shader_key
966 - turnip: Emit geometry shader obj and related consts
967 - turnip: Configure VPC for geometry shaders
968 - turnip: Configure VFD_CONTROL with gsheader and primitiveid
969 - turnip: Set up REG_A6XX_SP_GS_CONFIG
970 - turnip: Selectively configure GRAS_LAYER_CNTL
971 - turnip: Update maxGeometryShaderInvocations to match blob
972 - turnip: Populate tu_pipeline.active_stages
973 - turnip: Enable geometry shaders for CP_DRAWs
974 - turnip: Enable geometryShader device feature
975 - turnip: Correctly set layer stride for 3D images
976 - turnip: Emit geometry shader descriptor consts
977 - freedreno/turnip: Update GRAS_LAYER_CNTL to GRAS_MAX_LAYER_INDEX
981 - anv: Advertise VK_KHR_shader_non_semantic_info
982 - radv: Advertise VK_KHR_shader_non_semantic_info
983 - intel/gen12: Take into account opcode when decoding SWSB
984 - spirv: Be consistent when checking for Shader/Kernel
985 - anv: Use intel_debug_flag_for_shader_stage()
986 - anv: Add pipe_state_for_stage() helper
987 - nir/builder: Add nir_scoped_memory_barrier()
988 - nir: Add the alias NIR_MEMORY_ACQ_REL
989 - nir/tests: Use nir_scoped_memory_barrier() helper
990 - nir, intel: Move use_scoped_memory_barrier to nir_options
991 - anv: Remove unused field xfb_used from anv_pipeline
992 - anv: Remove unused field \`urb.total_size\`
993 - nir: Don't skip a bit in nir_memory_semantics
994 - nir: Reorder nir_scopes so wider scope has larger numeric value
995 - nir: Add pass to combine adjacent scoped memory barriers
996 - intel/fs: Combine adjacent memory barriers
997 - anv: Add a new enum to identify the pipeline type
998 - anv: Use pipeline type to decide whether or not lower multiview
999 - anv: Use a dynamic array for storing executables in pipeline
1000 - anv: Keep the shader stage in anv_shader_bin
1001 - anv: Pass the right pipe_state to flush_descriptor_sets()
1002 - anv: Remove redundant check in flush_descriptor_sets() helpers
1003 - anv: Decouple flush_descriptor_sets() helpers from pipeline struct
1004 - anv: Decouple flush_descriptor_sets() from pipeline struct
1005 - anv: Use a separate field in the pipeline for compute shader
1006 - anv: Split graphics and compute bits from anv_pipeline
1007 - anv: Reduce compute pipeline batch_data size
1008 - anv: Remove duplicate code in anv_cmd_buffer_bind_descriptor_set
1009 - intel/blorp: Plumb the stage through blorp upload_shader
1010 - mesa/main: Fix overflow in validation of DispatchComputeGroupSizeARB
1011 - nir: Add per_view attribute to nir_variable
1012 - intel/gen12: Add XML description for 3DSTATE_PRIMITIVE_REPLICATION
1013 - intel/fs: Allow multiple slots for position
1014 - anv/gen12: Lower VK_KHR_multiview using Primitive Replication
1015 - intel/compiler: Replace cs_prog_data->push.total with a helper
1016 - anv: Stop using cs_prog_data->threads
1017 - iris: Stop using cs_prog_data->threads
1018 - intel/compiler: Remove cs_prog_data->threads
1019 - intel/fs,vec4: Properly account SENDs in IVB memory fence
1020 - spirv: Fix propagation of OpVariable access flags
1021 - spirv: Handle instruction aliases in vtn_gather_types
1022 - spirv: Update the headers from latest Khronos master
1023 - intel/fs: Allow FS_OPCODE_SCHEDULING_FENCE stall on registers
1024 - intel/fs,vec4: Pull stall logic for memory fences up into the IR
1025 - intel/fs: Only stall after sending all memory fence messages
1026 - i965: Use correct constant for max_variable_local_size
1030 - anv: Drop unused anv_image_get_surface_for_aspect_mask()
1031 - anv: Rename param make_surface::dev to device
1032 - anv: Delete anv_image::ccs_e_compatible
1033 - anv: Clarify behavior of anv_image_aspect_to_plane()
1034 - anv: Respect ISL_SURF_USAGE_DISABLE_AUX_BIT in make_surface()
1035 - turnip: Add magic register values to tu_physical_device
1036 - turnip: Add a618 support
1037 - anv: Drop anv_image.c:get_surface()
1038 - anv: Add anv_image_plane_needs_shadow_surface() (v2)
1039 - anv: Refactor creation of aux surfaces (v2)
1040 - anv: Flatten the logic add_aux_surface_if_supported (v3)
1041 - anv: Use isl_drm_modifier_get_default_aux_state()
1043 Chia-I Wu (2):
1045 - egl/android: require ANDROID_native_fence_sync for buffer age
1046 - egl/android: enable/disable KHR_partial_update correctly
1050 - vc4: fix vc4_yuv_blit overwriting fragment constant buffer slot 0
1051 - vc4: Fix query_dmabuf_modifiers mis-reporting external_only property
1055 - iris: Fix import sync-file into syncobj
1059 - etnaviv: enable texture upload memory throttling
1060 - etnaviv: update headers from rnndb
1061 - etnaviv: fix alpha test on GC3000
1062 - etnaviv: add etna_constbuf_state object
1063 - etnaviv: ask kernel for max number of supported varyings
1064 - etnaviv: update headers from rnndb
1065 - etnaviv: increase number of supported varyings to 16
1066 - etnaviv: implement emit_string_marker
1067 - etnaviv: get rid of etna_spec in etna_context
1068 - etnaviv: enable shareable shaders
1069 - freedreno: calculate modified bit mask only once
1070 - freedreno: simplify fd_set_shader_buffers(..)
1071 - freedreno: ssbo: keep track if a buffer gets written
1072 - freedreno: ssbo: mark resource read or written depending on usage
1073 - etnaviv: get rid of SE_CLIP\_\*
1074 - etnaviv: rework clippling calculation to be a derived state
1075 - etnaviv: do the left shift by 16 at emit time
1076 - etnaviv: get rid of struct compiled_scissor_state
1077 - etnaviv: s/scissor_s/scissor
1078 - etnaviv: compiled_framebuffer_state: get rid of SE_SCISSOR\_\*
1079 - etnaviv: rename hw queries to acc queries
1080 - etnaviv: rework etna_acc_sample_provider
1081 - etnaviv: explicitly call resource_written(..)
1082 - etnaviv: reset no_wait_cnt after triggered flush
1083 - etnaviv: rework wait/flush logic
1084 - etnaviv: extend acc query provider with supports(..) function
1085 - etnaviv: make use of a fixed size array to track of all acc query
1087 - etnaviv: extend result(..) to return if data is ready
1088 - etnaviv: extend acc sample provide with an allocate(..)
1089 - etnaviv: move generic perfmon functionality into own file
1090 - etnaviv: convert perfmon queries to acc queries
1091 - etnaviv: drop redundant calls to etna_acc_query_suspend(..)
1092 - etnaviv: change begin_query(..) to a void function
1093 - etnaviv: remove the "active" member of queries
1094 - etnaviv: anisotropic filtering is supported starting with HALTI0
1095 - etnaviv: update headers from rnndb
1096 - etnaviv: add anisotropic filter support
1097 - docs/features: mark GL_ARB_texture_filter_anisotropic as done for
1099 - etnaviv: drop default state for FE_HALTI5_ID_CONFIG
1100 - etnaviv: call util_blitter_save_fragment_constant_buffer_slot(..)
1101 - etnaviv: support for using generic blit path
1102 - ci: bare-metal: power down device after tests
1103 - etnaviv: fix SAMP_ANISOTROPY register value
1104 - etnaviv: do not use int filter when anisotropic filtering is used
1108 - radv: use util_float_to_half_rtz
1112 - egl/wayland: Fix zwp_linux_dmabuf usage
1116 - freedreno: Fix CP_COND_REG_EXEC bit positions
1117 - freedreno: Add CP_REG_WRITE documentation
1118 - freedreno: Fix CP_COND_EXEC
1119 - tu: Move vsc_data and vsc_data2 allocation into the device
1120 - tu: Don't emit initial render target state in tile_load_ib
1121 - tu: Properly set UBWC flags in RB_RENDER_CNTL
1122 - tu/blit: Support blits in secondary cmdstreams
1123 - tu: Support multisample image clears
1124 - tu: Disable linear depth attachments
1125 - tu: Sysmem rendering
1126 - tu: Add helper for CP_COND_REG_EXEC
1127 - tu: Handle vkCmdClearAttachments() with sysmem
1128 - tu: Support resolve ops with sysmem rendering
1129 - tu: Support input attachments with sysmem
1130 - tu: Force sysmem with mipmapped non-aligned linear stores
1131 - tu: Rewrite border color handling
1132 - lima/gpir: Make lima_gpir_node_insert_child() useful
1133 - lima/gpir: Optimize conditional break/continue
1134 - lima/gpir: Optimize nots created from branch lowering
1135 - tu: Fix border color with compute shaders
1136 - freedreno/fdl: Add base_align
1137 - tu: Return the correct alignment for images
1138 - freedreno: Cleanup event names
1139 - freedreno: Rename RB_DONE_TS
1140 - tu: Dump out shader assembly when requested
1141 - tu: ir3: Emit push constants directly
1142 - freedreno/a6xx: Add UBO size field
1143 - freedreno/a6xx: Add registers for the bindless model
1144 - ir3: Add bindless instruction encoding
1145 - ir3: Plumb through support for a1.x
1146 - ir3: Also don't propagate immediate offset with LDC
1147 - ir3: LDC also has a destination
1148 - ir3: Plumb through bindless support
1149 - ir3: Rewrite UBO push analysis to support bindless
1150 - tu: Switch to the bindless descriptor model
1151 - tu: Emit CP_LOAD_STATE6 for descriptors
1152 - tu: Add missing code for immutable samplers
1153 - tu: Implement descriptor set update templates
1154 - ir3: Fix txs with bindless
1155 - ir3: Fix LDC offset units
1156 - ir3: Handle load_ubo_ir3 when promoting to constants
1157 - tu: Align GMEM resolve blit scissor
1158 - tu: Use tu_cs_add_entries() with non-render-pass secondaries
1159 - ir3/ra: Fix off-by-one issues with live-range extension
1160 - freedreno/a6xx: Expand various varying-count bitfields
1161 - tu: Fix the advertised maxFragmentInputComponents
1162 - ir3: Don't double-insert the first block
1163 - ir3: Fix bug with shaders that only exit via discard
1164 - freedreno/a6xx: Document PrimID passthrough registers
1165 - ir3: Skip missing VS outputs in VS out map when linking
1166 - tu: Implement PrimID passthrough
1167 - freedreno/a6xx: Implement PrimID passthrough
1168 - st/nir: Fix assigning PointCoord location with !PIPE_CAP_TEXCOORD
1169 - ir3: Remove VARYING_SLOT_PNTC remapping hack
1170 - tu: Don't invert point coords
1174 - intel/tools/aubinator_error_decode: read HW Context before other
1176 - intel/tools/aubinator_error_decode: Decode ring buffers from HEAD to
1178 - util/sparse_array: don't stomp head's counter on pop operations
1179 - intel/fs: Update location of Render Target Array Index for gen12
1180 - anv,iris: Fix input vertex max for tcs on gen12
1181 - anv/gen11+: Disable object level preemption
1185 - aco: fix image_atomic_cmp_swap
1186 - nir: gather info whether a shader uses demote_to_helper
1187 - nir: add pass to lower discard() to demote()
1188 - amd/llvm: implement nir_intrinsic_demote(_if) and
1190 - radeonsi: lower discard to demote when FS_CORRECT_DERIVS_AFTER_KILL
1192 - radv: use nir_lower_discard_to_demote to work around game bugs
1193 - amd: join emit_kill() from radv and radeonsi in ac_nir_to_llvm
1194 - nir: fix unpack_64_4x16 in lower_alu_to_scalar()
1195 - aco: add comparison operators for PhysReg
1196 - aco: add sub-dword regclasses
1197 - aco: refactor regClass setup for subdword VGPRs
1198 - aco: validate p_create_vector with subdword elements properly
1199 - aco: validate register alignment of subdword operands and definitions
1200 - aco: validate uninitialized operands
1201 - aco: validate RA of subdword assignments
1202 - aco: print subdword registers
1203 - aco: fix Temp and assignment of renamed operands during RA
1204 - aco: remove unnecessary reg_file.fill() operation in
1206 - aco: add notion of subdword registers to register allocator
1207 - aco: create helper function to collect variables from register area
1208 - aco: adapt register allocation for subdword registers
1209 - aco: align subdword registers during RA when necessary
1210 - aco: small refactoring of shuffle code lowering
1211 - aco: add builder function for subdword copy()
1212 - aco: lower subdword shuffles correctly.
1213 - aco: don't propagate SGPRs into subdword PSEUDO instructions
1214 - aco: don't assume split_vector(create_vector) has the same number of
1216 - aco: don't vectorize 8/16bit load/store_ssbo
1217 - aco: add missing conversion operations for small bitsizes
1218 - aco: add byte_align_scalar() & trim_subdword_vector() helper
1220 - aco: prepare helper functions for subdword handling
1221 - aco: implement vec2/3/4 with subdword operands
1222 - aco: implement storagePushConstant8 & storagePushConstant16
1223 - aco: implement 8bit/16bit load_buffer
1224 - aco: implement 8bit/16bit store_ssbo
1225 - aco: use MUBUF to load subdword SSBO
1226 - aco: guarantee that Temp fits in 4 bytes
1227 - aco: add explicit padding for all Instruction sub-structs
1228 - aco: improve hashing for value numbering
1229 - aco: improve register assignment when live-range splits are necessary
1230 - aco: replace assignment hashmap by std::vector in register allocation
1231 - aco: during RA only insert into renames table if a variable got
1233 - aco: improve speed of live_var_analysis
1234 - aco: refactor try_remove_trivial_phi() in RA
1235 - aco: change some std::map to std::unordered_map in
1237 - aco: change live_out variables to std::unordered_set
1238 - aco: move all needed helper containers to ra_ctx
1239 - aco: RA - move all std::function objects into proper functions
1240 - aco: setup subdword regclasses for ssa_undef & load_const
1241 - aco: ensure correct bit representation of subdword constants
1242 - aco: don't constant-propagate into subdword PSEUDO instructions
1243 - aco: lower subdword phis with SGPR operands
1244 - aco: rename aco_lower_bool_phis() -> aco_lower_phis()
1245 - aco: make some reg_file helpers private and fix their uses
1246 - aco: fix p_extract_vector optimization in presence of unequally sized
1248 - aco: use v_subrev_f32 for fsub with an sgpr operand in src1
1249 - aco: fix 64bit fsub
1250 - aco: move src1 to vgpr instead of using VOP3 for VOP2 instructions
1252 - aco: simplify operand handling in RA
1253 - aco: refactor get_reg() to take Temp instead of RegClass
1254 - aco: refactor get_reg() to also handle affinities
1255 - aco: create pseudo dummy instruction in RA to be used for live-range
1257 - aco: create and use DefInfo struct in RA
1258 - aco: use DefInfo in more places to simplify RA
1259 - aco: move attempt to find strided register into get_reg_simple()
1260 - aco: allocate full register for subdword definitions if HW doesn't
1262 - aco: don't create vector affinities for operands which are not killed
1264 - aco: refactor get_reg_simple() to return early on exact matches
1265 - aco: stop get_reg_simple after reaching max_used_gpr
1266 - aco: try to always find a register with stride for even sizes
1267 - aco: use upper part of gap in register file if it is beneficial for
1269 - aco: coalesce v_mad's accumulator with definition's affinities
1270 - aco: either copy-propagate or inline create_vector operands
1274 - Revert "gitlab-ci: disable panfrost runners"
1275 - egl/wayland: Don't invalidate buffers on no-op resize
1276 - util/test: Use MAX_PATH on Windows
1277 - CI: Add native Windows VS2019 build
1278 - CI: Windows: Fix Docker tag argument inversion
1279 - CI: Disable Panfrost Mali-T820 jobs
1280 - CI: Avoid htz4 runner for VS2019
1281 - meson: Add VS 4624 warning exclusion to remove piles of LLVM warnings
1282 - CI: Re-enable Windows VS2019 builds
1283 - EGL: Add eglSetDamageRegionKHR to GLVND dispatch list
1284 - meson: Make shared-llvm into a tri-state boolean
1285 - CI: Disable Windows/VS2019 builds
1286 - Revert "CI: Disable Windows/VS2019 builds"
1287 - ci/windows: Make Chocolatey installs more reliable
1288 - CI: Disable Lima jobs due to lab unhealthiness
1292 - i965: Do not set front_buffer_dirty if there is no front buffer
1293 - st/mesa: Handle the rest renderbuffer formats from OSMesa
1294 - osmesa/tests: Cover OSMESA_RGB GL_UNSIGNED_BYTE case
1295 - st/nir: Unify inputs_read/outputs_written before serializing NIR
1296 - brw_nir: Cast bitshift to unsigned
1297 - brw_fs: Avoid zero size vla
1298 - intel/compiler: Do not qsort zero sized array
1299 - intel/bufmgr: Cast bitshift to unsigned
1300 - glsl/blob: Do not call memcpy if there is nothing to copy
1301 - iris: Do not dereference nullptr with pipe_reference
1302 - i965: Do not generate D16 B5G6R5_UNORM configs on gen < 8
1303 - intel/tools: Fix compilation with UBSan
1304 - glsl: do not crash if string literal is used outside of
1306 - st/mesa: Fix signed integer overflow when using
1308 - intel/aub_viewer: Fix format specifier for uint64_t
1309 - nir: Fix breakage of foreach_list_typed_safe assumptions in loop
1311 - anv: Do not sample from 3d depth image with HiZ
1312 - glsl/list: Fix undefined behaviour of foreach\_\* macros
1313 - st/mesa: Update shader info of ffvp/ARB_vp after translation to NIR
1314 - st/mesa: Re-assign vs in locations after updating nir info for
1316 - spirv: Expand workaround for OpControlBarrier on old GLSLang
1317 - st/mesa: Treat vertex inputs absent in inputMapping as zero in
1319 - iris/bufmgr: Check if iris_bo_gem_mmap failed
1320 - i965: Fix out-of-bounds access to brw_stage_state::surf_offset
1321 - anv: Translate relative timeout to absolute when calling
1323 - anv: Fix deadlock in anv_timelines_wait
1324 - meson: Disable GCC's dead store elimination for memory zeroing custom
1326 - mesa: Fix double-lock of Shared->FrameBuffers and usage of wrong
1328 - intel/fs: Work around dual-source blending hangs in combination with
1333 - llvmpipe/query: add support for indexed queries
1334 - gallivm/swr: add stream_id to geom epilogue emit
1335 - gallivm/nir: add support for multiple vertex streams
1336 - draw: change geom shader output to an array of outputs.
1337 - draw/gs: track emitted prims + verts per stream.
1338 - draw: emit multiple streams to streamout.
1339 - draw: don't emit vertex to streams with no outputs
1340 - llvmpipe: advertise 4 vertex streams
1341 - gallivm/s390: fix pass init order on s390 with llvm 8 (v2)
1342 - ci: bump debian image and change llvm deps to 8
1343 - dri: add another get shm variant.
1344 - glx/drisw: add getImageShm2 path
1345 - glx/drisw: return false if shmid == -1
1346 - glx/drisw: fix shm put image fallback
1347 - gallivm/tgsi: fix stream id regression
1348 - gallivm/nir: fix integer divide SIGFPE
1349 - gallivm/nir: handle mod 0 better.
1350 - gallium/auxiliary: add the microsoft tessellator and a pipe wrapper.
1351 - gallivm/nir: split out 64-bit splitting code
1352 - gallivm/nir: add support for tess system values
1353 - gallivm/nir: align store_var param order with load_var
1354 - gallivm/tgsi/swr: add mask vec to the tcs store
1355 - gallivm/nir: add tessellation i/o support.
1356 - draw: add JIT context/functions for tess stages.
1357 - draw: add main tessellation code
1358 - draw: hook up final bits of tessellation
1359 - gallium/nir/tgsi: only scan fragment shader inputs for usage_mask
1360 - llvmpipe: add support for tessellation shaders
1361 - gallivm/tessellator: use private functions for min/max to avoid
1363 - gallium: fix build with latest meson and gcc10
1364 - gallivm/s3tc: split out dxt5 alpha code
1365 - gallivm: add support for rgtc/latc fetches.
1366 - gallium/llvmpipe: add an optimised 32-bit memset
1367 - gallivm/rgtc: fix the truncation to 8-bit
1368 - gallivm/rgtc: enable fast path for snorm types.
1369 - Revert "gallivm: disable rgtc/latc SNORM accellerated fetches"
1370 - llvmpipe: fixup context leaks.
1371 - draw: collect tessellation invocations statistics
1372 - llvmpipe: report tessellation shader statistics.
1373 - llvmpipe/query: fix transform feedback overflow any queries.
1374 - gallivm: fix left over shader vote debug
1375 - gallivm/nir: lower implicit lod to tex.
1376 - gallivm/draw: calloc prim id toavoid undef
1377 - llvmpipe: fix no tokens detections.
1378 - draw: fix tessellation stats query
1379 - llvmpipe/setup: move line stats collection earlier.
1380 - draw/cull: run pipeline for culled points.
1381 - draw: fix user culling pipeline order. (v2)
1382 - u_blitter: fix stencil blitting
1383 - draw: free the NIR IR.
1384 - draw/tess: free the NIR
1385 - llvmpipe/nir: free the nir shader
1386 - nir/linking: fix issue with two compact variables in a row. (v2)
1387 - gallivm/nir: fix image store conversions
1388 - gallivm/nir: add helper invocation support
1389 - util/indirect: handle stride less than number of parameters.
1390 - llvmpipe: bump max images to 16
1391 - llvmpipe: fix ssbo alignment
1392 - draw/tess: fix TES patch vertices in.
1393 - llvmpipe: fix d32 unorm depth conversions.
1394 - llvmpipe/setup: add point size clamping
1395 - llvmpipe: enable stencil only formats. (v2)
1396 - llvmpipe: clamp color storage for integer types.
1397 - gallivm: fix stencil border
1398 - vulkan: add initial device selection layer. (v6.1)
1399 - ci: add llvmpipe paths to virgl rules
1400 - draw/tess: free tessellation control shader i/o memory.
1401 - llvmpipo/nir: free compute shader NIR
1402 - llvmpipe: compute shaders work better with all the threads.
1406 - egl/android: set window usage flags
1410 - gitlab: add bug report template
1414 - meson: fix debug build on Android
1418 - radv: Filter extensions not whitelisted for Android
1422 - zink. Added storage CISto descriptor pool. Added storage in
1426 - zink: zero out zink_render_pass_state
1430 - docs/release-calendar: 20.0.0-rc1 has been released
1431 - docs: Mark 20.0-rc2 as done
1432 - docs: Add release notes for 19.3.4
1433 - docs: Add SHA256 sum for 19.3.4
1434 - docs: Mark 19.3.4 as done
1435 - docs: Mark 20.0.0-rc3 as done
1436 - Docs: Add 20.0.0 release notes
1437 - docs: Update index, relnotes, and release-calendar for 20.0
1438 - docs: Update stable process around using fixes: and gitlab
1439 - docs/submittingpatches: Fix confusing typo + missing pronoun
1440 - docs: Update release notes with current process
1441 - bin/post_version.py: Update the release calendar as well
1442 - bin/post_version.py: Pretty print the html
1443 - bin/post_version.py: Make the git commit as well.
1444 - docs: update releasing to cover updated post_version.py
1445 - docs: add relnotes for 20.0.1
1446 - docs: Add sha256sums for 20.0.1
1447 - docs: update news, calendar, and link release notes for 20.0.1
1448 - Docs: Add release notes for 20.0.2
1449 - docs/relnotes: Add sha256 sums for 20.0.2
1450 - docs: update calendar, add news item, and link releases notes for
1452 - docs/release-calendar: Add calendar for 20.1 Release candidates
1453 - bin/gen_release_notes.py: Fix version detection for .0 release
1454 - bin/pick-ui: Add a new maintainer script for picking patches
1455 - replace \_mesa_is_pow_two with util_is_power_of_two\_\*
1456 - replace \_mesa_next_pow_two\_\* with util_next_power_of_two\_\*
1457 - replace \_mesa_logbase2 with util_logbase2
1458 - replace LOG2 with util_fast_log2
1459 - u_math: add x86 optimized version of ifloor
1460 - replace IFLOOR with util_ifloor
1461 - Replace IROUND_POS with \_mesa_roundevenf
1462 - mesa/main: remove unused IROUNDD
1463 - replace IROUND with util functions
1464 - move windows strtok_r define to u_string
1465 - Replace IS_INF_OR_NAN with util_is_inf_or_nan
1466 - replace malloc macros in imports.h with u_memory.h versions
1467 - util: Add an aligned realloc function
1468 - replace imports memory functions with utils memory functions
1469 - mesa|mapi: replace \_mesa_[v]snprintf with [v]snprintf
1470 - mesa: move ADD_POINTERS to macros.h
1471 - dri/nouveau: replace assert with unreachable
1472 - remove final imports.h and imports.c bits
1473 - meson: update llvm dependency logic for meson 0.54.0
1474 - docs: Add relnotes for 20.0.5
1475 - docs: Add sha256 sums for 20.0.5
1476 - docs: update calendar, add news item, and link releases notes for
1478 - mesa: Follow OpenGL conversion rules for values that exceed storage
1480 - tests: Make tests aware of meson test wrapper
1484 - radv/sqtt: fix RADV_THREAD_TRACE_BUFFER_SIZE spelling
1488 - turnip/pipeline: Don't assume tu_shader is a valid object
1489 - turnip: Instance can be NULL resolving 'GetInstanceProcAddr' entry
1491 - anv/radv: Resolving 'GetInstanceProcAddr' should not require a valid
1496 - docs: fix typo in v20 release notes
1500 - spirv2nir: print nir shader if translation succed
1501 - spirv2nir: Add kernel spirv support
1502 - docs/features: Update virgl OpenGL 4.5 features GL_ARB_clip_control
1507 - meson: glx: drop with_glx == dri check
1508 - glx: set the loader_logger early and for everyone
1509 - egl/drm: reinstate (kms\_)swrast support
1510 - Revert "egl/dri2: Don't dlclose() the driver on
1512 - loader: use a maximum of 64 drmDevices
1513 - loader: simplify loader_get_user_preferred_fd()
1514 - loader: simplify codeflow in drm_get_pci_id_for_fd
1515 - loader: move "using driver..." message to
1517 - loader: fallback to kernel name, if PCI fails
1518 - glx: omit loader_loader() for macOS
1519 - egl: simplify client/platform extension handling
1523 - Expose EGL_KHR_platform\_\* when EXT is supported
1527 - gallium/osmesa: Fix a typo in the unit test's test names.
1528 - gallium/osmesa: Fix MakeCurrent of non-8888 contexts.
1529 - gallium/osmesa: Fill out other format tests.
1530 - gallium/osmesa: Try to fix the test for big-endian.
1531 - util: Make helper functions for pack/unpacking pixel rows.
1532 - mesa/st: Use direct util_format_pack/unpack instead of u_tile.
1533 - gallium/util: Remove pipe_get_tile_z/put_tile_z.
1534 - softpipe: Drop the raw_to\* part of the tile cache interface.
1535 - softpipe: Refactor pipe_get/put_tile_rgba\_\* paths.
1536 - gallium: Add and use a helper for packing uc from a color_union.
1537 - gallium: Refactor some single-pixel util_format_read/writes.
1538 - util: Drop unpacking from int signed to unsigned and vice versa.
1539 - freedreno: Move the layout debug under FD_MESA_DEBUG=layout.
1540 - freedreno: Include the layer size in layout debug.
1541 - freedreno: Rename the UBWC layer size field and store it as bytes.
1542 - freedreno/a6xx: Disable the core layer-size setup.
1543 - freedreno: Swap the whole resource layout in shadowing.
1544 - freedreno: Blit all array levels when uncompressing UBWC.
1545 - freedreno: Disable UBWC on Z24S8 if not TEXTURE_2D.
1546 - freedreno: Allow UBWC on textures with multiple mipmap levels.
1547 - mesa: Clean up some endianness adapters for shader image formats.
1548 - intel/isl: Move iris's pipe-to-isl format function to isl.
1549 - glsl,nir: Switch the enum representing shader image formats to
1551 - mesa/st: Move the SYSTEM_VALUE -> TGSI_SEMANTIC map to
1553 - nouveau: Reuse tgsi_get_sysval_semantic().
1554 - nouveau: reuse tgsi_get_gl_frag_result_semantic().
1555 - nouveau: Reuse tgsi_get_gl_varying_semantic().
1556 - u_tile: Skip the packed temporary and just store tiles directly.
1557 - ci: Disable a bunch of tests on freedreno a630.
1558 - ci: Bump the GLES CTS version to 3.2.6.1.
1559 - Revert "gallium: Fix big-endian addressing of non-bitmask array
1561 - ci: Extend the a630 flake list to reduce spurious failures.
1562 - radv: Squelch possibly-undefined warning
1563 - llvmpipe: Fix real uninitialized use of "atype" for SEMANTIC_FACE
1564 - llvmpipe: Silence "possibly uninitialized value" warning for
1566 - llvmpipe: Silence uninitialized variable warning about "chan"
1567 - llvmpipe: Fix warning about uninitialized "op" in the NIR path.
1568 - llvmpipe: Silence uninitialized variable warning about "vals"
1569 - llvmpipe: Silence uninitialized variable warning about "scissor"
1570 - llvmpipe: Fix another uninitialized value warning, on init_val.
1571 - gallium: Only define PIPE_ALIGNSTACK on x86.
1572 - ci: prepare-artifacts: Make the indent here match previously in the
1574 - ci: Make sure that we have a proper shell prompt for LAVA.
1575 - ci: Make LAVA job fails emit the full list of unexpected test
1577 - ci: Document how LAVA runners work.
1578 - ci: Don't bother generating deqp junit results since we don't present
1580 - ci: Remove a useless filtering of the lava logs.
1581 - nir: Rename gl_nir_lower_bindless_images.c in preparation for
1583 - nir: Make image lowering optionally handle the !bindless case as
1585 - gallium: Add a cap for enabling lowering of image load/store
1587 - v3d: Ask the state tracker to lower image accesses off of derefs.
1588 - glsl: Factor out the sampler dim coordinate components switch
1590 - spirv_to_nir: Reuse glsl_sampler_dim_coordinate_components().
1591 - freedreno/ir3: Reuse glsl_get_sampler_dim_coordinate_components() in
1593 - tgsi_to_nir: Reuse glsl_get_sampler_dim_coordinate_components().
1594 - prog_to_nir: Reuse glsl_get_sampler_dim_coordinate_components().
1595 - freedreno/ir3: Fix the arg to
1597 - nir: Move intel's intrinsic_image_coordinate_components() to core
1599 - freedreno: Switch to using lowered image intrinsics.
1600 - ci: Blacklist another freedreno flaky test.
1601 - meson: Disable bison's -Wdeprecated since we still support old bison.
1602 - turnip: Fix compiler warning about casting a nondispatchable handle.
1603 - freedreno/computerator: Fix defined-but-not-used warnings from
1605 - ci: Remove LLVM from ARM test drivers.
1606 - ci: Stop disabling ACPI in the LAVA arm64 kernel build.
1607 - ci: Shrink the arm64 kernel build a bit.
1608 - ci: Include db410c support in the ARM container.
1609 - aco: Fix signed-vs-unsigned warning.
1610 - ci: Enable -Werror on meson-vulkan and meson-testing.
1611 - ci: Switch testing on db410c over to LAVA.
1612 - ci: Add a disabled-by-default job for GLES3 testing on db410c.
1613 - ci: Flip db410c back to docker mode.
1614 - ci: Print the renderer/version that our dEQP invocation is using.
1615 - ci: Fix installation of firmware for db410c's nic.
1616 - ci: Make a simple little bare-metal fastboot mode for db410c.
1617 - glsl/tests: Catch mkdir errors to help explain when they happen.
1618 - glsl/tests: Fix waiting for disk_cache_put() to finish.
1619 - ci: Update the ci-templates commit.
1620 - ci: Enable ccache in the container builds.
1621 - ci: Enable ccaching of CMake builds as well.
1622 - ci: Enable testing GLES2-3 on a530 (Dragonboard 820c).
1623 - freedreno/a5xx: Fix min-vs-mag filtering decisions on non-mipmap tex.
1624 - gallium/util: Switch util_float_to_half to \_mesa_float_to_half()'s
1626 - ci: Ban the recent popular freedreno a630 flakes.
1627 - ci: Disable tests that showed intermittent fails on a530 in day 1.
1628 - ci: Only run the freedreno baremetal tests when freedreno/core
1630 - freedreno: Switch to exposing only half-integer pixel centers.
1631 - ci: Move db820c and db410c's gles3 tests to manual, like radv did.
1632 - glsl: Restore the IsES flag on the shader when reading from cache.
1633 - ci: Ban the recent popular freedreno a630 intermittent failure.
1634 - freedreno: Remove always-true return from per-gen begin_query.
1635 - freedreno: Remove the "active" member of queries.
1636 - freedreno: Fix acc query handling in the presence of batch
1638 - freedreno: Associate the acc query bo with the batch.
1639 - freedreno: Count blits in GL_TIME_ELAPSED and perf counter queries.
1640 - freedreno/a6xx: Fix timestamp queries.
1641 - freedreno: Rename "is_blit" to "is_discard_blit"
1642 - freedreno: Fix detection of being in a blit for acc queries.
1643 - freedreno: Work around UBWC flakiness.
1644 - freedreno: Drop an unnecessary include marked "this should go away"
1645 - freedreno/turnip: Use the NIR info to decide if we need helper
1647 - loader: Warn when we fail to open a device node due to permissions.
1648 - ci: Consistently use -j4 across x86 build jobs and -j8 on ARM.
1649 - freedreno/a6xx: Sink the per-level size temps inside the loop.
1650 - freedreno/a6xx: Remove the "aligned_height" temporary.
1651 - freedreno/a6xx: Drop the "alignment" layout temporary.
1652 - freedreno: Add the outline of a test for a6xx texture layout.
1653 - freedreno/a6xx: Set a level's pitch based on minified level0 pitch,
1655 - freedreno: Fix leak of binning shader variants.
1656 - freedreno/ir3: Stop doing b2n on the SEL condition.
1657 - freedreno/ir3: CSE the up/downconversion of SEL's cond's size.
1658 - freedreno/a5xx+: Skip compiling the old gmem blit programs.
1659 - freedreno/drm-shim: Add support for faking other adreno chips.
1660 - freedreno/ir3: Drop handling FRAG_RESULT_DEPTH writing to .z
1661 - freedreno: Introduce a "cpp_shift" value for cpp divs/muls.
1662 - freedreno: Make the slice pitch be bytes, not pixels.
1663 - drm-shim: Let the driver choose to overwrite the first render node.
1664 - nir/lower_two_sided_color: Fix picking of new driver location.
1665 - nir/lower_clip: Fix picking of unused driver locations.
1666 - gallium: Fix setup of pstipple frag coord var.
1667 - freedreno/ir3: Fix driver_location of the added vertex_flags varying.
1668 - freedreno/ir3: Fix sizing of the inputs/outputs array.
1669 - vc4: Use NIR shader's num_outputs for generating our new output.
1670 - ci: Drop redundant freedreno stage specification.
1671 - ci: Enable GLES3 testing on db410c/db820c (freedreno a306 and a530).
1672 - freedreno: Fix derivatives without texturing on a3xx-a5xx.
1673 - ci: Enable GLES 3.1 testing on db820c (a530).
1674 - freedreno/ir3: Fix the disasm of half-float STG dests.
1675 - freedreno/ir3: Print a space after nop counts, like qcom's disasm.
1676 - freedreno/ir3: Add a unit test for our disassembler.
1677 - freedreno/ir3: Convert remaining disasm src prints to reginfo.
1678 - freedreno/ir3: Refactor out print_reg_src().
1679 - freedreno/ir3: Add support for disasm of cat2 float32 immediates.
1680 - ci: Enable --compact-display false on all dEQP runs.
1681 - ci: Add sanity checking that dEQP gets the expected GL_RENDERER.
1682 - freedreno: Fix calculation of the const buffer cmdstream size.
1683 - ci: Allow namespacing of dEQP run results files.
1684 - ci: Clean up some excessive use of pipes in dEQP results processing.
1685 - ci/freedreno: Add a test run of a few driver options.
1686 - util/ra: Sanity check that the driver selected a valid reg.
1687 - util/ra: Sanity check that we're adding a valid reg to a class.
1688 - util/ra: Use util_dynarray for the adjacency list.
1689 - util/ra: Use util_dynarray for handling the conflict lists.
1690 - util/ra: Improve ra_set_finalize() performance.
1694 - VERSION: bump after 20.0 branch point
1695 - egl: put full path to libEGL_mesa.so in GLVND json
1696 - gitlab-ci: disable a630 tests as mesa-cheza is down
1697 - util/os_socket: fix header unavailable on windows
1698 - freedreno/perfcntrs: fix fd leak
1699 - dri: delete gen-symbol-redefs.py
1700 - util/disk_cache: check for write() failure in the zstd path
1701 - meson: don't bother trying \`python2\`
1702 - Revert "egl: put full path to libEGL_mesa.so in GLVND json"
1703 - egl: directly access static members instead of using
1705 - meson: explicitly disallow unsupported build directory layout
1706 - docs: fix typos in the release docs
1707 - bin/gen_release_notes.py: fix commit list command
1708 - gen_release_notes: fix vulkan version reported
1709 - docs/relnotes/19.3: fix vulkan version reported
1710 - docs/relnotes/20.0: fix vulkan version reported
1711 - Revert "docs/relnotes/19.3: fix vulkan version reported"
1712 - docs: trivial fix for html structure
1713 - docs/releasing: add missing </li> tags
1714 - docs: add release notes for 19.3.5
1715 - docs: update calendar, add news item, and link releases notes for
1717 - vulkan/wsi: fix cleanup when dup() fails
1718 - gen_release_notes: fix version in "you should wait" message
1719 - gen_release_notes: resolve ambiguity by renaming \`version\` to
1721 - meson: use existing variables in inc_common
1722 - meson: inline \`inc_common\`
1723 - vulkan: drop unused include directories
1724 - intel: drop unused include directories
1725 - scons: prune unused Makefile.sources
1726 - docs: add release notes for 20.0.3
1727 - docs/relnotes: add sha256sum for 20.0.3
1728 - docs: update calendar, add news item, and link releases notes for
1730 - docs: add release notes for 20.0.4
1731 - docs/relnotes: add sha256sum for 20.0.4
1732 - docs: update calendar, add news item, and link releases notes for
1734 - glx: fix 630 times -Wlto-type-mismatch when building with LTO enabled
1735 - glx: use anonymous namespace to avoid -Wodr issues when building with
1737 - pick-ui: auto-scroll the feedback window
1738 - pick-ui: compute .pick_status.json path only once
1739 - pick-ui: make .pick_status.json path relative to the git root instead
1741 - pick-ui: show commit sha in the pick list
1742 - VERSION: bump to 20.1.0-rc1
1743 - .pick_status.json: Update to af55bdd05d94eda59ee1c9331a50045000da5db5
1744 - .pick_status.json: Update to 57796946985de60204189426ca8eb7bbfa97c396
1745 - .pick_status.json: Mark 3fac55ce0d066d767d6c6c8308f79d0c3e566ec0 as
1747 - .pick_status.json: Update to 29da52128090a1ef8ef782188c0f67c7f5ec8d19
1748 - VERSION: bump to 20.1.0-rc2
1749 - .pick_status.json: Update to 772b15ad3227e08bb4e18932ac9ecf4c29271160
1750 - .pick_status.json: Update to 56f955e4850035d915a2a87e2ebea7fa66ab5e19
1751 - .pick_status.json: Update to c1c0cf7a66905e8d7ad506842a41b0ad0c5b10da
1752 - VERSION: bump to 20.1.0-rc3
1753 - .pick_status.json: Update to 5a6beb6a24aa084adfd6c57edd0a64f0a044611a
1754 - post_version.py: fix branch name construction for release candidates
1755 - post_version.py: invert \`is_point\` into \`is_first_release\` to
1757 - post_version.py: stop adding release candidates to the index and
1759 - VERSION: bump to 20.1.0-rc4
1760 - .pick_status.json: Update to a91306677c613ba7511b764b3decc9db42b24de1
1761 - tree-wide: fix deprecated GitLab URLs
1763 Erik Faye-Lund (154):
1765 - zink: enable texture-buffer objects
1766 - zink: implement load_instance_id
1767 - zink: implement support for derivative-control
1768 - zink: be more careful about the mask-check
1769 - zink: disallow depth-stencil blits with format-change
1770 - st/mesa: use uint-result for sampling stencil buffers
1771 - zink: lower away fdph
1772 - zink: fixup sampler-usage
1773 - zink: replace unset buffer with a dummy-buffer
1774 - zink: emit blend-target index
1775 - zink: only inspect dual-src limit if feature enabled
1776 - Revert "nir: Add a couple trivial abs optimizations"
1777 - zink: do not use SpvDimRect
1778 - zink: fix binding-usage
1779 - zink: do not report texture-samplers for unsupported stages
1780 - zink/spirv: do not reinvent store_dest
1781 - zink/spirv: prefer store_dest over store_dest_uint
1782 - zink/spirv: rename functions a bit
1783 - zink/spirv: unit_value -> raw_value
1784 - zink/spirv: uint -> raw
1785 - zink: do not convert bools to/from uint
1786 - util: promote u_debug_memory.c to src/util
1787 - util: move debug_memory_{begin,end} to os_memory_debug.h
1788 - gallium/util: do not use debug_print_format
1789 - gallium/util: remove unused debug_print_foo helpers
1790 - zink/spirv: do not use bitwise operations on booleans
1791 - pipebuffer: clean up cast-warnings
1792 - rbug: clean up cast-warnings
1793 - rbug: do not return void-value
1794 - vtn/opencl: fully enable OpenCLstd_Clz
1795 - compiler/nir: move build_exp helper into builtin-builder
1796 - compiler/nir: move build_log helper into builtin-builder
1797 - vtn/opencl: add native exp/log-support
1798 - vtn/opencl: add native exp10/log10-support
1799 - vtn/opencl: add native exp2/log2-support
1800 - nv50: remove unused variable
1801 - meson: disable some more warnings on msvc
1802 - mesa/main: correct extension-checks for GL_BLACKHOLE_RENDER_INTEL
1803 - mesa/main: clean-up extension-checks for point-sprites
1804 - mesa/main: clean up extension-check for GL_VERTEX_PROGRAM
1805 - mesa/main: clean up extension-check for GL_VERTEX_PROGRAM_TWO_SIDE
1806 - mesa/main: clean up extension-check for GL_VERTEX_PROGRAM_POINT_SIZE
1807 - mesa/main: clean up extension-check for GL_TEXTURE_RECTANGLE
1808 - mesa/main: clean up extension-check for GL_STENCIL_TEST_TWO_SIDE
1809 - mesa/main: clean up extension-check for GL_DEPTH_BOUNDS_TEST
1810 - mesa/main: clean up extension-check for AMD_depth_clamp_separate
1811 - mesa/main: clean up extension-check for GL_FRAGMENT_SHADER_ATI
1812 - mesa/main: clean up extension-check for GL_TEXTURE_CUBE_MAP_SEAMLESS
1813 - mesa/main: clean up extension-check for GL_RASTERIZER_DISCARD
1814 - mesa/main: clean up extension-check for GL_TEXTURE_EXTERNAL
1815 - mesa/main: remove unused macro
1816 - wgl: drop pointless debug_printf
1817 - wgl: drop unused member
1818 - wgl: move screen-init to a helper
1819 - wgl: do not create screen from DllMain
1820 - st/dri: make sure software color-buffers are linear
1821 - zink: be less picky about tiled resources
1822 - .mailmap: add an alias for Alan Swanson
1823 - .mailmap: add an alias for Alyssa Rosenzweig
1824 - .mailmap: add an alias for Andrii Simiklit
1825 - .mailmap: add an alias for Anuj Phogat
1826 - .mailmap: add an alias for Axel Davy
1827 - .mailmap: add an alias for Boris Brezillon
1828 - .mailmap: add an alias for Bruce Cherniak
1829 - .mailmap: update aliases for Carl-Philip Hänsch
1830 - .mailmap: add an alias for Chad Versace
1831 - .mailmap: add a couple of aliases for Chandu Babu Namburu
1832 - .mailmap: add alias for Chenglei Ren
1833 - .mailmap: add an alias for Christian Gmeiner
1834 - .mailmap: add an alias for Christian Inci
1835 - .mailmap: add a few aliases for Christoph Haag
1836 - .mailmap: add an alias for Colin McDonald
1837 - .mailmap: specify spelling for Constantine Kharlamov
1838 - .mailmap: add an alias for Craig Stout
1839 - .mailmap: add an alias for Daniel Schürmann
1840 - .mailmap: add an alias for Danylo Piliaiev
1841 - .mailmap: add an alias for Dave Airlie
1842 - .mailmap: add an alias for Dylan Baker
1843 - .mailmap: add a couple of aliases for Dylan Noblesmith
1844 - .mailmap: add an alias for Emmanuel Gil Peyrot
1845 - .mailmap: add an alias for Erik Faye-Lund
1846 - .mailmap: specify spelling for Francesco Ansanelli
1847 - .mailmap: specify spelling for Gurchetan Singh
1848 - .mailmap: add an alias for Haihao Xiang
1849 - .mailmap: add an alias for Harish Krupo
1850 - .mailmap: specify spelling for Heinrich Fink
1851 - .mailmap: specify spelling for Henri Verbeet
1852 - .mailmap: add an alias for Igor Gnatenko
1853 - .mailmap: add an alias for Illia Iorin
1854 - .mailmap: specify spelling for James Zhu
1855 - .mailmap: add an alias for Jan Beich
1856 - .mailmap: clean up aliases for Jeremy Huddleston
1857 - .mailmap: add an alias for Julien Isorce
1858 - .mailmap: add a few aliases for Karol Herbst
1859 - .mailmap: add a few aliases for Kevin Rogovin
1860 - .mailmap: add a few aliases for Kristian Høgsberg
1861 - .mailmap: add an alias for Lionel Landwerlin
1862 - .mailmap: specify spelling for Liviu Prodea
1863 - .mailmap: update aliases for Marc-André Lureau
1864 - .mailmap: add alias for Matthias Groß
1865 - .mailmap: add an alias for Neha Bhende
1866 - .mailmap: add an alias for Neil Roberts
1867 - .mailmap: specify spelling for Nian Wu
1868 - .mailmap: add an alias for Nicholas Bishop
1869 - .mailmap: update aliases for Nicolai Hähnle
1870 - .mailmap: add an alias for Philipp Zabel
1871 - .mailmap: update aliases for Pierre-Eric Pelloux-Prayer
1872 - .mailmap: add an alias for Plamena Manolova
1873 - .mailmap: add an alias for Qiang Yu
1874 - .mailmap: specify spelling for Randy Xu
1875 - .mailmap: add an alias for Renato Caldas
1876 - .mailmap: add an alias for Rob Clark
1877 - .mailmap: add an alias for Rodrigo Vivi
1878 - .mailmap: add an alias for Samuel Li
1879 - .mailmap: add an alias for Sergii Romantsov
1880 - .mailmap: specify spelling for Sonny Jiang
1881 - .mailmap: add a couple of aliases for Steinar H. Gunderson
1882 - .mailmap: add a couple of aliases for Suresh Guttula
1883 - .mailmap: add an alias for Thierry Reding
1884 - .mailmap: add an alias for Timo Aaltonen
1885 - .mailmap: add a couple of aliases for Timothy Arceri
1886 - .mailmap: add an alias for Tim Wiederhake
1887 - .mailmap: add an alias for Tom Stellard
1888 - .mailmap: add an alias for Tomasz Figa
1889 - .mailmap: add an alias for Topi Pohjolainen
1890 - .mailmap: add an alias for Vadym Shovkoplias
1891 - .mailmap: add an alias for Varad Gautam
1892 - .mailmap: specify spelling for Vivek Kasireddy
1893 - .mailmap: specify spelling for Wladimir J. van der Laan
1894 - .mailmap: add an alias for Xavier Bouchoux
1895 - .mailmap: add an alias for Yaakov Selkowitz
1896 - .mailmap: add alias for Zhaowei Yuan
1897 - .mailmap: add an alias for Zhongmin Wu
1898 - meson: use override_options to change warning-level
1899 - wgl: silence some cast-warnings
1900 - util/tests: initialize variable
1901 - mesa: fixup cast expression
1902 - vbo: avoid including wingdi.h on win32
1903 - meson: tell flex that we support c99
1904 - gtest: Update to 1.10.0
1905 - meson: do not disable incremental linking for debug-builds
1906 - docs: remove outdated sentence
1907 - mesa/gallium: do not use enum for bit-allocated member
1908 - meson: correct windows-version define
1909 - mesa/main: do not store unrecognized extensions in context
1910 - mesa/main: do not pass context to one-time extension init
1911 - mesa/main: do not init remap-table per api
1912 - mesa/main: Do not pass context to one_time_init
1913 - mesa/main: one_time_init() -> \_mesa_initialize()
1914 - mesa/st: call \_mesa_initialize() early
1915 - zink: lower b2b to b2i
1916 - util/os_memory: never use os_memory_debug.h
1917 - zink: implement i2b1
1918 - zink: use general-layout when blitting to/from same resource
1922 - intel/fs/cse: Make HALT instruction act as CSE barrier.
1923 - intel/fs/gen7: Fix fs_inst::flags_written() for
1925 - intel/fs: Add virtual instruction to load mask of live channels into
1927 - intel/fs/gen12: Workaround unwanted SEND execution due to broken
1929 - intel/fs/gen12: Fixup/simplify SWSB annotations of SIMD32 scratch
1931 - intel/fs/gen12: Workaround data coherency issues due to broken NoMask
1933 - intel/fs: Set src0 alpha present bit in header when provided in
1935 - intel/fs/gen11: Work around dual-source blending hangs in combination
1937 - intel/fs: Make sample_mask_reg() local to brw_fs.cpp and use it in
1939 - intel/fs: Use helper for discard sample mask flag subregister number.
1940 - intel/fs/gen7+: Swap sample mask flag register and FIND_LIVE_CHANNEL
1942 - intel/fs: Refactor predication on sample mask into helper function.
1943 - intel/fs: Return consistent UW types from sample_mask_reg() in
1945 - intel/fs/gen7+: Implement discard/demote for SIMD32 programs.
1946 - intel/compiler: Move base IR definitions into a separate header file
1947 - intel/compiler: Reverse inclusion dependency between brw_cfg.h and
1949 - intel/compiler: Nest definition of live variables block_data
1951 - intel/compiler: Reverse inclusion dependency between
1953 - intel/compiler: Reverse inclusion dependency between
1955 - intel/compiler: Introduce simple IR analysis pass framework
1956 - intel/compiler: Introduce backend_shader method to propagate IR
1958 - intel/compiler: Define more detailed analysis dependency classes
1959 - intel/compiler: Pass detailed dependency classes to
1961 - intel/compiler: Mark virtual_grf_interferes and vars_interfere as
1963 - intel/compiler: Move all live interval analysis results into
1965 - intel/compiler: Move all live interval analysis results into
1967 - intel/compiler: Restructure live intervals computation code
1968 - intel/compiler: Pass single backend_shader argument to the
1970 - intel/compiler: Pass single backend_shader argument to the
1972 - intel/compiler/fs: Add live interval validation pass
1973 - intel/compiler/vec4: Add live interval validation pass
1974 - intel/compiler/fs: Switch liveness analysis to IR analysis framework
1975 - intel/compiler/vec4: Switch liveness analysis to IR analysis
1977 - intel/compiler: Drop invalidate_live_intervals()
1978 - intel/compiler: Move idom tree calculation and related logic into
1980 - intel/compiler: Move dominance tree data structure into idom_tree
1982 - entel/compiler: Simplify new_idom reduction in dominance tree
1984 - intel/compiler: Move register pressure calculation into IR analysis
1986 - intel/compiler: Calculate num_instructions in O(1) during register
1988 - intel/fs: Fix workaround for VxH indirect addressing bug under
1990 - intel/fs/gen12: Fix interaction of SWSB dependency combination with
1992 - intel/fs/gen12: Fix hangs with per-sample SIMD32 fragment shader
1994 - intel/fs/gen12: Work around dual-source blending hangs in combination
1996 - intel/fs/gen12: Fix Render Target Read header setup for new thread
1998 - intel/ir: Add missing initialization of backend_reg::offset during
2000 - intel/fs: Rename half() helpers to quarter(), allow index up to 3.
2001 - intel/fs: Fix constness of argument of
2003 - intel/fs: Replace fs_visitor::bank_conflict_cycles() with stand-alone
2005 - intel/vec4: Fix constness of vec4_instruction::reads_flag() and
2007 - intel/ir: Import shader performance analysis pass.
2008 - intel/fs: Heap-allocate fs_visitors in brw_compile_fs().
2009 - intel/fs: Implement performance analysis-based SIMD32 heuristic for
2011 - intel/fs: Add INTEL_DEBUG=no32 debugging flag.
2012 - intel/ir: Use brw::performance object instead of CFG cycle counts for
2014 - intel/ir: Pass block cycle count information explicitly to
2016 - intel/ir: Remove scheduling-based cycle count estimates.
2017 - intel/ir: Update performance analysis parameters for memory fence
2022 - Revert "gitlab-ci: disable a630 tests as mesa-cheza is down"
2023 - Revert "gitlab-ci: disable a630 tests as mesa-cheza is down (again)"
2024 - freedreno: allow FMT6_8_UNORM as a UBWC format
2028 - Correctly wait in the fragment stage until all semaphores are
2030 - Vulkan Overlay: Don't try to change the image layout to present twice
2031 - Vulkan overlay: use the corresponding image index for each swapchain
2035 - r600: force new CF with TEX only if any texture value is written
2036 - r600: Increase space for IO values to agree with
2038 - r600: Add NIR compiler options
2039 - r600: Update state code to accept NIR shaders
2040 - r600/sfn: Add a basic nir shader backend
2041 - r600: enable NIR backend DEBUG flag for supported architectures
2042 - r600/sfn: Add the VS in and FS out vectorization
2043 - r600/sfn: Add the WaitAck instruction
2044 - r600/sfn: add live range evaluation for the GPR
2045 - r600/sfn: add register remapping
2046 - r600/sfn: Add lowering arrays to scratch and according instructions
2047 - r600/sfn: Add a load GDS result instruction
2048 - r600/sfn: Add MemRingOut instructions
2049 - r600/sfn: add emitVertex instructions
2050 - r600/sfn: Add support for geometry shader
2051 - r600/sfn: Add VS for TCS shader skeleton
2052 - r600/sfn: Add compute shader skeleton
2053 - r600/sfn: Add GDS instructions
2054 - r600/sfn: Add lowering UBO access to r600 specific codes
2055 - r600: Make sure LLVM is not used for DRAW
2056 - r600/sfn: Add support for atomic instructions
2057 - r600/sfn: Add support for SSBO load and store
2058 - r600/sfn: Add .editorconfig file
2059 - r600/sfn: Add some documentation
2060 - r600/sfn: Avoid using dynamic_cast to identify type
2061 - r600/sfn: Use static_cast when type is already known
2062 - r600/sfn: Don't try to catch exceptions, the driver doesn't throw any
2063 - gallium/tgsi_to_nir: Set nir_intrinsic_align_mul to 16 and offset to
2065 - r600: Dump a few more variables when requested
2066 - r600/sfn: Reduce array limit for scratch usage
2067 - r600/sfn: Fix setting alignments when lowering UBOs
2068 - r600/sfn: Implementing instructions blocks
2069 - r600/nir: Pin interpolation results to channel
2070 - r600/sfn: Fix null pointer deref in live range evalation
2071 - r600/sfn: Handle b2b1 like it was a mov
2072 - r600/sfn: Fix handling of GS inputs
2073 - r600/sfn: Fix using the result of a fetch instruction in next fetch
2074 - r600/sfn: Count only literals that are not inline to split
2076 - r600/sfn: use new temp register allocation when loading single value
2078 - nir: Add r600 specific intrinsics for tesselation shader IO
2079 - nir: Add umad24 and umul24 opcodes
2080 - r600: Handle texcoord semantics in LDS index evaluation
2081 - r600/sfn: simplify UBO lowering pass
2082 - r600/sfn: Don't emit inline constants in the r600 IR
2083 - r600/sfn: Add LDS IO instructions to r600 IR
2084 - r600/sfn: Add LDS instruction to assembly conversion
2085 - r600/sfn: Add TF write instruction
2086 - r600/sfn: Add IR instruction to fetch the TESS parameters
2087 - r600/sfn: Handle umul24 and umad24
2088 - r600/sfn: Emit some LDS instructions
2089 - r600/sfn: Move emission of barrier from compute shader to shader base
2090 - r600/sfn: Add methods to valuepool to get a vector of values
2091 - r600/sfn: Move some shader base methods to the public interface
2092 - r600/sfn: extract class to handle the VS export to different stages
2093 - r600/sfn: derive the GS from the vertex stage for a common interface
2094 - r600/sfn: Handle LDS output in VS
2095 - r600/sfn: Move removing of unused variables
2096 - r600/sfn: Add lowering passes for Tesselation IO
2097 - r600/sfn: Add tesselation shaders
2098 - r600: Enable tesselation for NIR
2099 - r600: Fix nir compiler options, i.e. don't lower IO to temps for TESS
2100 - r600/sfn: Fix printing vertex fetch instruction flags
2101 - r600: Fix duplicated subexpression in r600_asm.c
2105 - amd/addrlib: fix build on non-x86 platforms
2106 - r600: add missing <array> include
2107 - svga: fix build on FreeBSD
2111 - x86_init_func_common: Add ENDBR at function entry
2112 - x86: Add ENDBR at function entries
2116 - Properly check mmap return value
2120 - freedreno/ir3: fix printing half constant registers.
2121 - freedreno/ir3: Add cat4 mediump opcodes
2122 - freedreno/ir3: put the conversion back for half const to the right
2124 - freedreno/ir3: Fold const only when the type is float
2125 - freedreno/ir3: Add new ir3 pass to fold out fp16 conversions
2126 - nir: Add optimization for doing removing f16/f32 conversions
2127 - freedreno/ir3: handle half registers for arrays during register
2129 - turnip: support indirect draw
2130 - glsl: Handle fp16 unary operations when lowering matrix operations
2131 - glsl/lower_instructions: Handle fp16 for MOD_TO_FLOOR
2132 - turnip: Gather information for transform feedback
2133 - turnip: Define structs for transform feedback
2134 - turnip: Setup stream-output when linking program
2135 - turnip: Implement stream-out emit and vkApis for transform feedback
2136 - turnip: Implement an empty function vkCmdDrawIndirectByteCountEXT
2137 - turnip: Enable VK_EXT_transform_feedback
2138 - turnip: Add tu6_control struct.
2139 - turnip: Fix wrong assignment of xfb output's offset.
2140 - turnip: Do gathering xfb info after nir_remove_dead_variables
2141 - freedreno: Enable mediump lowering
2142 - freedreno/ir3: enable nir_opt_loop_unroll on a6xx
2143 - nir: fix wrong assignment to buffer in xfb_varyings_info
2144 - turnip: make the struct slot_value of queries get 2 values
2145 - turnip: Implement and enable
2147 - turnip : Fix wrong offset calculation for xfb buffer.
2148 - turnip: Skip unused regs when setting up streamout buffers
2149 - turnip: Fix crashes when geometry shader constants aren't used
2153 - nir: add a bool bitsize lowering pass
2157 - intel/fs: Don't count integer instructions as being possibly coissue
2158 - nir: Mark fmin and fmax as commutative and associative
2159 - mesa/draw: Make sure all the unused fields are initialized to zero
2160 - nir/search: Use larger type to hold linearized index
2161 - intel/fs: Correctly handle multiply of fsign with a source modifier
2162 - intel/fs: Do cmod prop again after scheduling
2163 - intel/fs: Allow NOT instructions in conditional discard optimization
2164 - intel/fs: Fix NULL destinations on 3-source instructions again after
2166 - nir/algebraic: Simplify logic to detect sign of an integer
2167 - nir/algebraic: optimize ior(ine(a, 0), ine(b, 0)) to ine(ior(a, b),
2169 - nir/algebraic: Generalize some and-of-shift-right patterns [v2]
2170 - nir/algebraic: Constant reassociation for bitwise operations too
2171 - nir/algebraic: Simplify a contradiction that can occur in
2173 - soft-fp64/b2f: Reimplement using bitwise logic ops
2174 - soft-fp64: Don't open-code umulExtended
2175 - soft-fp64: Simplify \__countLeadingZeros32 function
2176 - soft-fp64: Pick a single idiom for treating sign value as a Boolean
2177 - soft-fp64: Store sign value as 0 or 0x80000000
2178 - soft-fp64/fneg: Don't treat NaN specially
2179 - soft-fp64/flt: Perform checks in a different order
2180 - soft-fp64/fsat: Correctly handle NaN
2181 - soft-fp64/fsat: Micro-optimize x < 0 test
2182 - soft-fp64/fsat: Micro-optimize x >= 1 test
2183 - soft-fp64: Relax the way NaN is propagated
2184 - soft-fp64/ffloor: Simplify the >= 0 comparison
2185 - soft-fp64: Optimize \__fmin64 and \__fmax64 by using different
2187 - soft-fp64/fadd: Instead of tracking "b < a", track sign of the
2189 - soft-fp64/fadd: Massively split the live range of zFrac0 and zFrac1
2190 - soft-fp64/fadd: Pick zero or non-zero result based on subtraction
2192 - soft-fp64/fadd: Just let the subtraction happen when the result will
2194 - soft-fp64/fadd: Delete a redundant condition check
2195 - soft-fp64/fadd: Reformat after previous commit
2196 - soft-fp64/fadd: Combine an if-statement into the preceeding
2197 else-clause
2198 - soft-fp64/fadd: Rename aFrac and bFrac variables
2199 - soft-fp64/fadd: Use absolute value of expDiff
2200 - soft-fp64/fadd: Move common code out of both branches of an
2201 if-statement
2202 - soft-fp64/fadd: Common code optimization for differing sign case
2203 - soft-fp64: Split a block that was missing a cast on a comparison
2204 - intel/vec4: Allow late copy propagation on vec4
2205 - nir/algebraic: Change the default cursor location when replacing a
2207 - nir/algebraic: Distribute source modifiers into instructions
2208 - nir/algebraic: Use value range analysis to convert fmax to fsat
2209 - nir/algebraic: Remove a redundant fabs pattern
2210 - tnl: Don't dereference NULL obj pointer in bind_indices
2211 - tnl: Don't dereference NULL obj pointer in replay_init
2212 - tnl: Don't dereference NULL obj pointer in t_rebase_prims
2213 - tnl: Silence unused parameter 'attrib' warning in
2215 - tnl: Silence unused parameter warnings in \_tnl_draw_prims
2216 - tnl: Silence unused parameter warnings in dump_draw_info
2217 - tnl: Silence unused parameter warnings in \_tnl_split_inplace
2218 - tnl: Code formatting in t_draw.c
2219 - tnl: Code formatting in t_rebase.c
2220 - intel/compiler: Silence unused parameter warnings in vec4_tcs_visitor
2221 - intel/compiler: Silence unused parameter warning in
2223 - intel/compiler: Silence unused parameter warning in
2225 - intel/compiler: Only GE and L modifiers are commutative for SEL
2226 - intel/compiler: CSEL can do saturate
2227 - intel/compiler: Fixup operands in fs_builder::emit() that takes array
2228 - nir/algebraic: Detect some kinds of malformed variable names
2229 - nir/algebraic: Require operands to iand be 32-bit
2230 - nir/algebraic: Optimize ushr of pack_half, not ishr
2231 - anv/tests: Don't rely on assert or changing NDEBUG in tests
2235 - panfrost: Fix non-debug builds
2236 - panfrost: Inline panfrost_get_default_swizzle
2237 - panfrost: LogicOp support
2238 - nir: Allow nir_format conversions to work on 32-bit values
2239 - panfrost: LogicOp fixes and non 8-bit format support
2240 - mesa/format_utils: Add a fast-path for RGBA to BGRA
2241 - panfrost: Extend the tiled store fast-path to loads
2242 - panfrost: Mark 64-bit formats as unsupported
2243 - panfrost: Add support for B5G5R5X1
2244 - st/mesa: Fall back on R3G3B2 for R3_G3_B2
2245 - panfrost: Add support for R3G3B2
2246 - panfrost: Correctly identify format 0x4c
2247 - pan/midgard: Fix a divide by zero in emit_alu_bundle
2248 - panfrost: Fix GL_EXT_vertex_array_bgra
2249 - panfrost: Enable PIPE_CAP_VERTEX_COLOR_UNCLAMPED
2250 - panfrost: Fix background showing when using discard
2254 - lima: remove its hash table entry when invalidating a resource
2255 - lima: expose fragment shader derivatives capability
2256 - lima: implement zsbuf reload
2260 - nv50: report max lod bias of 15.0
2261 - gitlab-ci: disable panfrost runners
2262 - mesa: fix \_mesa_draw_nonzero_divisor_bits to return nonzero divisors
2263 - nv50,nvc0: add newly added PIPE_CAP's to list
2264 - st/mesa: allow TXB2/TXL2 to work with cube array shadow textures
2265 - nvc0: enable EXT_texture_shadow_lod
2266 - st/vdpau: avoid asserting on new VDP_YCBCR\_\* formats
2267 - st/vdpau: make query test for 2D support
2268 - nv50: don't try to upload MSAA settings for BUFFER textures
2269 - gallium: add viewport swizzling state and cap
2270 - mesa: add GL_NV_viewport_swizzle support
2271 - st/mesa: add NV_viewport_swizzle support
2272 - nvc0: add NV_viewport_swizzle support for GM200+
2273 - compiler: add VARYING_SLOT_VIEWPORT_MASK
2274 - glsl: add NV_viewport_array2 support
2275 - mesa: add NV_viewport_array2 enable, attach to glsl
2276 - gallium: add TGSI_SEMANTIC_VIEWPORT_MASK
2277 - gallium: add TGSI_PROPERTY_LAYER_VIEWPORT_RELATIVE
2278 - gallium: add PIPE_CAP_VIEWPORT_MASK
2279 - st/mesa: add support for GL_NV_viewport_array2
2280 - nvc0: enable GL_NV_viewport_array2
2281 - nv50,nvc0: update with latest caps
2282 - docs: update for recently-added nvc0 features
2283 - mesa: add interaction between compute derivatives and variable local
2288 - glapi/copyimage: Implement CopyImageSubDataNV
2289 - gallium: prepare framework for supporting
2291 - mesa: add support for AlphaToCoverageDitherControlNV
2292 - radeonsi: enable support for AlphaToCoverageDitherControlNV
2296 - egl: allow INVALID format for linux_dmabuf
2300 - iris: handle the failure of converting unsupported yuv formats to isl
2301 - gallium: let the pipe drivers decide the supported modifiers
2305 - radeonsi: fix Segmentation fault during vaapi enc test
2309 - targets/opencl: fix build against LLVM>=10 with Polly support
2313 - clover: Use explicit conversion from llvm::StringRef to std::string
2314 - clover: Check if the detected clang libraries are usable
2318 - gallium/swr: Fix various asserts and security issues
2319 - gallium/swr: fix corruptions in Unigine Heaven
2320 - gallium/swr: use ElementCount type arguments for getSplat()
2321 - gallium/gallivm: Remove workaround disabling AVX code for newer CPUs
2322 - gallium/gallivm: fix compilation issues with llvm 11
2323 - gallium/gallivm: remove unused header include for newer LLVM
2324 - gallium/swr: Fix LLVM 11 compilation issues
2325 - gallium/swr: Fix crashes and failures in vertex fetch
2329 - genxml: Add a new 3DSTATE_SF field on gen12
2330 - anv,iris: Set 3DSTATE_SF::DerefBlockSize to per-poly on Gen12+
2331 - intel/genxml: Drop SLMEnable from L3CNTLREG on Gen11
2332 - iris: Set SLMEnable based on the L3$ config
2333 - iris: Store the L3$ configs in the screen
2334 - iris: Use the URB size from the L3$ config
2335 - i965: Re-emit l3 state before BLORP executes
2336 - intel: Take a gen_l3_config in gen_get_urb_config
2337 - intel/blorp: Always emit URB config on Gen7+
2338 - iris: Consolodate URB emit
2339 - anv: Emit URB setup earlier
2340 - intel/common: Return the block size from get_urb_config
2341 - intel/blorp: Plumb deref block size through to 3DSTATE_SF
2342 - anv: Plumb deref block size through to 3DSTATE_SF
2343 - iris: Plumb deref block size through to 3DSTATE_SF
2344 - anv: Always fill out the AUX table even if CCS is disabled
2345 - intel/eu/validate: Don't validate regions of sends
2346 - intel/disasm: SEND has two sources on Gen12+
2347 - intel/tools: Handle strides better when dumping buffers
2348 - intel/fs: Write the address register with NoMask for MOV_INDIRECT
2349 - anv/blorp: Use the correct size for vkCmdCopyBufferToImage
2350 - anv: No-op submit and wait calls when no_hw is set
2351 - anv: Reject modifiers on depth/stencil formats
2352 - vulkan: Update the XML and headers to 1.2.133
2353 - nir: Fix the nir_builder include path for nir_builtin_builder
2354 - nir/builder: Return an integer from nir_get_texture_size
2355 - intel/isl: Add isl_aux_info.c to Makefile.sources
2356 - anv: Always enable the data cache
2357 - nir: Drop nir_tex_instr::texture_array_size
2358 - anv: Use the PIPE_CONTROL instead of bits for the CS stall W/A
2359 - anv: Use a proper end-of-pipe sync instead of just CS stall
2360 - anv: Do end-of-pipe sync around MCS/CCS ops instead of CS stall
2361 - nir: Flush to zero with OOB low exponents in ldexp
2362 - isl: Set 3DSTATE_DEPTH_BUFFER::Depth correctly for 3D surfaces
2363 - iris: Allow HiZ on blit sources
2364 - blorp: Write to depth/stencil images as depth/stencil when possible
2365 - anv: Enable HiZ for VK_IMAGE_LAYOUT_TRANSFER_DST_OPTIMAL
2366 - iris: Enable CCS for copies from HiZ+CCS depth buffers
2367 - iris: Enable HiZ and stencil CCS for blorp blit destinations
2368 - iris: Don't skip fast depth clears if the color changed
2369 - anv: Parse VkPhysicalDeviceFeatures2 in CreateDevice
2370 - anv: Mark max_push_range UNUSED and simplify the code
2371 - anv: Pass buffer addresses into emit_push_constant\*
2372 - anv: Delete some pointless break statements
2373 - anv: Align UBO sizes to 32B
2374 - anv: Add an align_down_u32 helper
2375 - anv: Bounds-check pushed UBOs when robustBufferAccess = true
2376 - vulkan/wsi: Don't leak the FD when
2378 - vulkan/wsi: Return an error if dup() fails
2379 - intel/isl: Clean up some aux surface logic
2380 - intel/isl: Add a separate ISL_AUX_USAGE_HIZ_CCS_WT
2381 - intel/blorp: Allow HIZ_CCS_WT in copy sources
2382 - iris: Use ISL_AUX_USAGE_HIZ_CCS_WT to indicate write-through HiZ
2383 - intel/isl: Require ISL_AUX_USAGE_HIZ_CCS_WT for HZ+CCS WT mode
2384 - intel/isl: Add a separate ISL_AUX_USAGE_STC_CCS
2385 - intel/blorp: Allow STC_CCS in blit sources
2386 - iris: Use ISL_AUX_USAGE_STC_CCS for stencil CCS
2387 - intel: Require ISL_AUX_USAGE_STC_CCS for stencil CCS
2388 - intel/isl: Set DepthStencilResource based on aux usage
2389 - anv: Dump push ranges via VK_KHR_pipeline_executable_properties
2390 - anv: Fix the comparison in an assert
2391 - anv: Push UBO ranges relative to the start of the binding
2392 - anv: Do an end-of-pipe sync before updating AUX table entries
2393 - intel/isl: Don't align linear images to 64K on Gen12+
2394 - intel/blorp: Add support for swizzling fast-clear colors
2395 - anv: Swizzle fast-clear values
2396 - intel/iris: Always initialize CCS to 0
2397 - anv: Only add END_OF_PIPE_SYNC if we actually have AUX_INVAL
2398 - util/sparse_array: Finish the sparse_array in the tests
2399 - util/sparse_array: Add a node_size_log2 temporary
2400 - meson,ci: Disable sparse_array tests on windows
2401 - util/sparse_array: Stash the node level in the node pointer
2402 - anv: Stop fetching the timestamp frequency ourselves
2403 - intel/dump_gpu: Add an ensure_device_info helper
2404 - intel/dump_gpu: Handle a bunch of getparam in the no-HW case
2405 - intel/nir: Run copy-prop and DCE after lower_bool_to_int32
2406 - nir: Add b2b opcodes
2407 - aco: Implement b2b32 and b2b1
2408 - nir: Use b2b opcodes for shared and constant memory
2409 - nir: Insert b2b1s around booleans in nir_lower_to
2410 - anv: Set alignments on descriptor and constant loads
2411 - nir: Validate that memory load/store ops work on whole bytes
2412 - nir: Set UBO alignments in lower_uniforms_to_ubo
2413 - nir/opt_loop_unroll: Fix has_nested_loop handling
2414 - nir/lower_int64: Lower 8 and 16-bit downcasts with nir_lower_mov64
2415 - nir/algebraic: Add downcast-of-pack opts
2416 - nir: Add a nir_op_is_vec helper
2417 - nir: Copy propagate through vec8s and vec16s
2418 - nir: Handle vec8/16 in bool_to_bitsize
2419 - nir: Handle vec8/16 in gather_ssa_types
2420 - nir: Handle vec8/16 in lower_phis_to_scalar
2421 - nir: Handle vec8/16 in lower_regs_to_ssa
2422 - nir: Handle vec8/16 in opt_split_alu_of_phi
2423 - nir: Treat vec8/16 as select in opt_peephole_select
2424 - nir: Handle vec8/16 in opt_undef_vecN
2425 - nir: Handle vec8/16 in nir_shrink_array_vars
2426 - anv: Account for the header in anv_state_stream_alloc
2427 - anv/allocator: Use util_dynarray for blocks in anv_state_stream
2428 - spirv: Implement OpCopyObject and OpCopyLogical as blind copies
2429 - Revert "spirv: Implement OpCopyObject and OpCopyLogical as blind
2431 - anv/image: Use align_u64 for image offsets
2432 - nir/from_ssa: Only chain movs when a src is also a dest
2433 - intel/fs: Choose memory message type based on bit size
2434 - anv: Improve brw_nir_lower_mem_access_bit_sizes
2435 - iris: Set alignments on cbuf0 and constant reads
2436 - intel/nir: Lower memory access bit sizes later
2437 - nir/load_store_vectorize: Fix shared atomic info
2438 - nir/load_store_vectorize: Use nir_iadd_imm for offsets
2439 - nir/load_store_vectorize: Add support for nir_var_mem_global
2440 - intel/nir: Enable load/store vectorization
2441 - spirv: Add a vtn_block() helper
2442 - spirv: Add cast and loop helpers for vtn_cf_node
2443 - spirv: Make vtn_case a vtn_cf_node
2444 - spirv: Make vtn_function a vtn_cf_node
2445 - spirv: Add a parent field to vtn_cf_node
2446 - spirv: Rewrite CFG construction
2447 - Revert "spirv: Rewrite CFG construction"
2448 - nir: Assert memory loads are aligned
2449 - anv: Advertise SEND count through
2451 - anv: Fix UBO range detection in anv_nir_compute_push_layout
2452 - nir: Add an alignment to nir_intrinsic_load_constant
2453 - nir: Add some sanity assertions in opt_large_constants
2454 - intel: Add \_const versions of prog_data cast helpers
2455 - anv: Report correct SLM size
2456 - intel/batch_decoder: Stop printing to stdout
2457 - intel/cfg: Add first/last_block helpers
2458 - anv: Emit pushed UBO bounds checking code in the back-end compiler
2459 - intel/blorp: Delete an unused enum
2460 - spirv: Handle OOB vector extract operations
2461 - spirv,nir: Add a better vector_insert
2462 - spirv: Error if OpCompositeInsert/Extract has OOB indices
2463 - nir/builder: Handle any bit-size selector in nir_extract
2464 - spirv: Call nir_builder directly for vector_extract
2465 - spirv,nir: Move the SPIR-V vector insert code to NIR
2466 - anv: Move vb_emit setup closer to where it's used in flush_state
2467 - anv: Apply any needed PIPE_CONTROLs before emitting state
2468 - nir/dominance: Better handle unreachable blocks
2469 - nir/gcm: Loop over blocks in pin_instructions
2470 - nir/gcm: Use an array for storing the early block
2471 - nir/gcm: Move block choosing into a helper function
2472 - nir/gcm: Add a real concept of "progress"
2473 - nir/gcm: Delete dead instructions
2474 - nir/gcm: Prefer the instruction's original block
2475 - intel/fs: Rename block to scan_block in can_coalesce_vars
2476 - intel/fs: Coalesce when the src live range is contained in the dst
2477 - glsl: Hard-code noise to zero in builtin_functions.cpp
2478 - nir: Delete the fnoise opcodes
2479 - meta,i965: Rip GL_EXT_texture_multisample_blit_scaled support out of
2481 - spirv: Allow constants and NULLs in SpvOpConvertUToPtr
2482 - anv: Properly handle all sizes of specialization constants
2483 - radv: Properly handle all sizes of specialization constants
2484 - turnip: Properly handle all sizes of specialization constants
2485 - spirv: Use nir_const_value for spec constants
2486 - nir/opt_deref: Remove certain sampler type casts
2487 - spirv: Fix passing combined image/samplers through function calls
2488 - anv: Drop an assert
2489 - nir/lower_subgroups: Mask off unused bits in ballot ops
2490 - anv: Add a vk_image_layout_to_usage_flags helper
2491 - anv: Move vk_image_layout_is_read_only higher
2492 - anv: Be more conservative about image view usage
2493 - anv: Rework anv_layout_to_aux_state
2494 - anv/blorp: Do less hard-coding of aux usages
2495 - anv: Generalize some aux usage checks
2496 - intel/blorp: Allow more HiZ usages in hiz_clear_depth_stencil
2497 - anv: Simplify a case in layout_to_aux_usage
2498 - anv/cmd_buffer: Move anv_image_init_aux_tt higher
2499 - intel/isl: Delete a misleading comment
2500 - intel/isl: Refactor isl_surf_get_ccs_surf
2501 - anv: Add support for HiZ+CCS
2502 - spirv: Rewrite CFG construction
2503 - intel/devinfo: Compute the correct L3$ size for Gen12
2504 - anv: Expose CS workgroup sizes based on a maximum of 64 threads
2505 - anv: Return an error if allocating attachment memory fails
2506 - anv: Add TRANSFER_SRC to pass usage not subpass usage
2507 - anv: Stop filling out the clear color in compute_aux_usage
2508 - anv: Assert surface states are valid
2509 - anv: Use ANV_FROM_HANDLE for pInheritanceInfo fields
2510 - anv: Mark images written in end_subpass
2511 - anv: Split command buffer attachment setup in three
2512 - anv: Allocate surface states per-subpass
2513 - intel: Move swizzle_color_value from blorp to ISL
2514 - anv: Disallow fast-clears which require format-reinterpretation
2515 - anv: Stop allowing non-zero clear colors in input attachments
2516 - anv: Refactor cmd_buffer_setup_attachments
2517 - anv: Rework depth_stencil_attachment_compute_aux_usage
2518 - anv: Split color_attachment_compute_aux_usage in two
2519 - anv: Use anv_layout_to_aux_usage for color during render passes
2520 - anv: Allow all clear colors for texturing on Gen11+
2521 - vulkan: Update Vulkan XML and headers to 1.2.139
2522 - nir/copy_prop_vars: Handle volatile better
2523 - nir/copy_prop_vars: Report progress when deleting self-copies
2524 - nir/dead_write_vars: Handle volatile
2525 - nir/combine_stores: Handle volatile
2526 - anv: Handle NULL descriptors
2527 - anv: Handle null vertex buffer bindings
2528 - anv: Claim VK_EXT_robustness2 support
2529 - intel/fs: Don't delete coalesced MOVs if they have a cmod
2530 - vulkan: Allow destroying NULL debug report callbacks
2531 - anv:gpu_memcpy: Emit 3DSTATE_VF_INDEXING on Gen8+
2532 - nir/lower_double_ops: Rework the if (progress) tree
2533 - nir/opt_deref: Report progress if we remove a deref
2534 - nir/copy_prop_vars: Record progress in more places
2538 - wgl: add official gldrv.h header-file
2539 - wgl: use gldrv.h instead of stw_icd.h
2540 - util/ralloc: fix ralloc alignment on Win64
2544 - freedreno: Add ir3_cf.c and ir3_delay.c to Makefile.sources
2545 - panfrost: Move pan_afbc.c file to the the right Makefile.source file
2546 - gallium: hud_context: Fix scalar initializer warning.
2547 - Android.mk: Tweak MESA_ENABLE_LLVM checks
2548 - etnaviv: Avoid shift overflow
2549 - vc4_bufmgr: Remove duplicative VC definition
2550 - r600: Fix build error in sfn_nir_lower_fs_out_to_vector.cpp
2554 - Fix util/process test on Cygwin
2558 - freedreno/a6xx: use single format enum
2559 - freedreno/a6xx: fix Z24_UNORM_S8_UINT_AS_R8G8B8A8
2560 - freedreno: name sysmem color/depth flush events
2561 - freedreno/a6xx: document some unknown bits
2562 - turnip: add option to force use of hw binning
2563 - turnip: fix COND_EXEC reserved size in tu_query
2564 - turnip: add tu_device pointer to tu_cs
2565 - turnip: automatically reserve cmdstream space in emit_pkt4/emit_pkt7
2566 - turnip: remove marker seqno
2567 - turnip: make cond_exec helper easier to use
2568 - turnip: move tile_load_ib/sysmem_clear_ib into draw_cs
2569 - hud: add GALLIUM_HUD_SCALE
2570 - turnip: enable sampleRateShading feature
2571 - turnip: enable
2573 - etnaviv: disable INT_FILTER for ASTC
2574 - util/format: add missing BC4/BC5 vulkan formats
2575 - turnip: rework format table to support r5g5b5a1_unorm/b5g5r5a1_unorm
2576 - turnip: add r5g5b5a1_unorm/b5g5r5a1_unorm formats
2577 - turnip: check the right alignment requirement on shader iova
2578 - turnip: move some constant state to tu6_init_hw
2579 - turnip: remove unecessary MRT_CONTROL fill
2580 - turnip: minify image_view extent
2581 - turnip: fix hw binning + render_area offset interaction
2582 - turnip: fix srgb MRT
2583 - turnip: don't hardcode gmem base for input attachment
2584 - turnip: remove unnecessary fb size check
2585 - turnip: fall back to sysmem when attachments don't fit into gmem
2586 - turnip: increase array sizes in tu_descriptor_map
2587 - turnip: improve binning pipe layout config
2588 - turnip: fix tile->slot calculation
2589 - etnaviv: nir: add compile_check_limits
2590 - freedreno/registers: more GRAS_CL_CNTL bits, Z_CLAMP
2591 - turnip: fix znear clipping
2592 - turnip: implement depth clamp
2593 - turnip: implement timestamp query
2594 - turnip: fix compute shaders crashing after geometry shader change
2595 - turnip: improve vertex input handling
2596 - turnip: use buffer size instead of bo size for VFD_FETCH_SIZE
2597 - freedreno/registers: add RB_CCU_CNTL bitfields
2598 - freedreno/a6xx: set bypass RB_CCU_CNTL value for blitter
2599 - turnip: RB_CCU_CNTL fixes
2600 - turnip: split up gmem/tile alignment
2601 - turnip: fix nir validate failure from push constant lowering
2602 - turnip: disable 8x msaa
2603 - turnip: save attachment samples in renderpass state
2604 - turnip: use dirty bits for dynamic viewport/scissor state
2605 - turnip: rework format helpers
2606 - turnip: add vk_format_is_snorm/is_float
2607 - turnip: new clear/blit implementation with shader path fallback
2608 - freedreno/computerator: support nop prefix
2609 - freedreno/computerator: support bindless sampler instructions
2610 - freedreno/ir3: fix emit_tex_info split_dest
2611 - freedreno/ir3: don't overwrite wrmask in ir3_SAM
2612 - turnip: compute render_components/srgb_cntl at renderpass creation
2614 - turnip: don't limit framebuffer size to image size
2615 - turnip: image_view rework
2616 - nir: add common convert_ycbcr for vulkan csc
2617 - nir: convert_ycbcr: preserve alpha channel
2618 - anv: use common nir_convert_ycbcr
2619 - radv: use common nir_convert_ycbcr
2620 - turnip: fix GMEM resolve in CmdNextSubpass
2621 - turnip: disable depth test for S8_UINT attachment
2622 - turnip: improve GMEM load/store logic
2623 - turnip: enable VK_FORMAT_S8_UINT as stencil format
2624 - turnip: set shader key msaa field
2625 - turnip: implement VK_EXT_sample_locations
2626 - turnip: implement VK_EXT_filter_cubic
2627 - turnip: enable cube arrays
2628 - turnip: implement VK_EXT_sampler_filter_minmax
2629 - turnip: divide cube map depth by 6
2630 - freedreno/ir3: fix 16-bit ssbo access
2631 - freedreno/ir3: set even bit for f2f16_rtne
2632 - freedreno/ir3: fix incorrect conversion folding
2633 - turnip: remove unused RB_UNKNOWN_8E04_blit
2634 - turnip: use RESOLVE_TS event
2635 - turnip: add adreno 650
2636 - nir: add pack_32_2x16_split/unpack_32_2x16_split lowering
2637 - freedreno/ir3: run nir_lower_pack
2638 - turnip: fix wrong substream size in parse_multisample_and_color_blend
2642 - intel/compiler: Restrict cs_threads to 64
2643 - intel: Update TGL PCI strings
2644 - intel: Add TGL PCI ID
2645 - intel/dev: Split .num_subslices out of GEN12_FEATURES macro
2646 - intel/dev: Add device info for RKL
2647 - docs/relnotes/new_features.txt: Add RKL to 20.1 release notes
2651 - broadcom: Fix implicit declaration of ffs for Android build
2652 - v3d: Sync on last CS when non-compute stage uses resource written by
2654 - v3d: Primitive Counts Feedback needs an extra 32-bit padding.
2655 - v3d: Fix swizzle in DXT3 and DXT5 formats
2656 - v3d: Include supported DXT formats to enable s3tc/dxt extensions
2660 - radv: Use TRUNC_COORD on samplers
2661 - radv: Pass logical device to si_emit_graphics
2662 - radeonsi: Use TRUNC_COORD on samplers
2666 - meson: Avoid duplicate symbols.
2667 - scons: Prune out unnecessary targets.
2668 - gitlab-ci: Prune all SCons jobs except scons-win64, and allows
2670 - appveyor: Remove Meson job.
2674 - nir/lower_double_ops: add note for lowering mod
2675 - nir/lower_double_ops: relax lower mod()
2676 - nir/algebraic: coalesce fmod lowering
2677 - anv: use urb_setup_attribs in SBE
2678 - intel/compiler: store the FS inputs in WM prog data
2679 - anv/pipeline: allow more than 16 FS inputs
2683 - clover: add trivial clCreateCommandQueueWithProperties implementation
2684 - nir/lower_ssbo: handle atomics
2685 - gallium: make handles of set_global_binding 64 bit
2686 - Revert "gallium: make handles of set_global_binding 64 bit"
2687 - nv50, nvc0: fix must_check warning of util_dynarray_resize_bytes
2688 - clover: fix build with single library clang build
2689 - gallium: add PIPE_CAP_SYSTEM_SVM
2690 - clover: add stubs for SVM
2691 - clover: implement CL_DEVICE_SVM_CAPABILITIES
2692 - clover: implement clSetKernelArgSVMPointer
2693 - clover: implement SVM functions for devices with fine grained system
2695 - clover: implement cl_arm_shared_virtual_memory
2696 - clover: expose cl_arm_shared_virtual_memory for devices with SVM
2698 - nvc0: enable ASTC and ETC on GM20B
2699 - mesa: fix enum value of VIEWPORT_SWIZZLE_POSITIVE_W_NV
2700 - gallium: initialize viewport swizzle in cso_set_viewport_dims
2701 - Revert "nvc0: fix line width on GM20x+"
2702 - st/mesa: properly guard fallback_copy_texsubimage aginst failed maps
2706 - intel/genxml: Drop "reserved" enum
2707 - isl: Fix the android build.
2708 - iris: Dump frame markers with INTEL_DEBUG=submit
2709 - iris: Trim "../../src/gallium/drivers/iris/" out of debug dump
2711 - iris: Make mocs an inline helper in iris_resource.h
2712 - iris: Fix BLORP vertex buffers to respect ISL MOCS settings
2713 - iris: Set MOCS for constant packets on Gen12+
2714 - intel/compiler: Drop nir_lower_to_source_mods() and related handling.
2715 - intel/compiler: Put back saturate on [iu]add_sat opcodes
2716 - intel/compiler: Don't copy prop source mods into PICK_HIGH_32BIT
2717 - intel/compiler: Delete abs/neg handling in fsign code
2718 - intel/compiler: Don't create 64-bit src1 immediates in
2720 - nir: Actually do load/store vectorization beyond vec2
2721 - iris: Fix downcast of bound_vertex_buffers from uint64_t to int
2725 - freedreno/a4xx: enable A405
2729 - nir: Delete unused is_var_constant() helper
2730 - nir: Make unroll pragma work on clang
2731 - freedreno/fdperf: Cast away some ignored return values
2732 - spirv/opencl: Cast opcode up front to avoid warnings
2733 - glsl: Use 'using' to be explicit about visitor overloads
2734 - nir: Remove always-true assert
2735 - turnip: Be explicit about converting vk compare func to a6xx
2736 - freedreno/a6xx: Add fd6_resource_screen_init()
2737 - freedreno: Set up supported modifiers in fd*_resource_screen_init()
2738 - freedreno: Add layout_resource_for_modifier screen vfunc
2739 - freedreno/a6xx: Implement layout for DRM_FORMAT_MOD_QCOM_COMPRESSED
2740 - turnip: Drop explicit configure opt-in for turnip
2741 - ci: Drop turnip opt-in option
2742 - freedreno/ir3: Set IR3_REG_HALF flag on src as well in immediate MOV
2743 - Mark a few static inline helpers with ASSERTED
2744 - main/get: Converted type conversion macros to inline functions
2745 - nir/types: Add glsl_float16_type() helper
2746 - freedreno/ir3: Lower output precision
2747 - Revert "glsl: Use a simpler formula for tanh"
2748 - Revert "spirv: Use a simpler and more correct implementaiton of
2750 - freedreno/ir3: Don't fold conversions into sign
2751 - glsl: Add ir_constant constructor for fp16
2752 - glsl: Add fp16 case for ir_triop_lrp optimization
2753 - glsl: Implement constant propagation for fp16
2754 - glsl: Expand fp16 to float before constant expression evaluation
2755 - glsl: Add type queries for fp16+float and fp16+float+double
2756 - glsl/lower_instructions: Handle fp16 for FDIV_TO_MUL_RCP
2757 - radeonsi: Stop exposing PIPE_SHADER_CAP_FP16
2758 - turnip: Add missing VKAPI_ATTR annotations
2759 - turnip: Stub out VK_KHR_external_{fence,semaphore}_fd
2760 - turnip: Make Android platform build
2761 - turnip: Drop dep_llvm from dependencies
2762 - freedreno/ir3: Fix sz vs class confusion
2763 - freedreno/computerator: Decouple ir3 assembler
2764 - freedreno/ir3: Move ir3 assembler to backend compiler
2765 - freedreno/ir3: Parse, but ignore @in, @out and @tex headers
2766 - freedreno/ir3: Reset lex line number when we start parsing
2767 - freedreno/ir3: Print @tex write mask using 0x%x
2768 - freedreno: Use the right amount of &'s
2772 - gallium/swr: fix gcc warnings
2773 - gallium/swr: Fix gcc 4.8.5 compile error
2774 - gallium/swr: Fix llvm11 compilation issues
2775 - gallium/swr: simplify environmental variabled expansion code
2776 - gallium/swr: fix rdtsc debug statistics mechanism
2777 - gallium/swr: Fix min/max range index draw
2778 - Revert "gallium/swr: Fix min/max range index draw"
2779 - gallium/swr: Fix vcvtph2ps llvm intrinsic compile error
2780 - gallium/swr: Fix array stride problem.
2781 - gallium/swr: Re-enable scratch space for client-memory buffers
2785 - i965: remove duplicated comment
2789 - radeon/jpeg: fix the jpeg dt_pitch with YUYV format
2793 - virgl: Use ETC2 formats directly when possible.
2797 - iris: implement gen12 post sync pipe control workaround
2798 - anv: implement gen9 post sync pipe control workaround
2799 - anv: implement gen12 post sync pipe control workaround
2800 - anv: set MOCS on push constants
2801 - mesa: add INTEL_blackhole_render
2802 - i965: enable INTEL_blackhole_render
2803 - st: add support for INTEL_blackhole_render
2804 - iris: add support INTEL_blackhole_render
2805 - intel/tools/aub_dump: move aub file initialization to maybe_init()
2806 - intel/tools/aub_dump: fix crash when using the default legacy context
2807 - intel/aub_dump: stub the waits when overriding the device
2808 - intel/tools/dump_gpu: fix getparam values
2809 - anv: stop storing prog param data into shader blobs
2810 - intel/decoder: don't consider header fields past dword0
2811 - isl: implement linear tiling row pitch requirement for display
2812 - isl: properly filter supported display modifiers on Gen9+
2813 - isl: only apply main surface ccs pitch constraint with CCS
2814 - isl: drop min row pitch alignment when set by the driver
2815 - intel: add new TGL pci ids
2816 - i965/iris: fix crash when calling GetPerfQueryDataINTEL
2817 - vulkan/overlay: Add a workaround semaphore for application presenting
2819 - intel/perf: move register definition to special file
2820 - intel/perf: break GL query stuff away
2821 - intel/perf: move mdapi query definitions to their own file
2822 - intel/perf: document meaning of query field
2823 - intel/perf: store the probed i915-perf version
2824 - isl: set bpb for Y8_UNORM
2825 - isl: don't warn in physical extent calculation for yuv formats
2826 - intel/aub_viewer: fix access to freed memory
2827 - drm-shim: return device platform as specified
2828 - drm-shim: stub libdrm's use of realpath()
2829 - iris: properly free resources on BO allocation failure
2830 - iris: share buffer managers accross screens
2831 - iris: make resources take a ref on the screen object
2832 - i965: store DRM fd on intel_screen
2833 - i965: share buffer managers across screens
2834 - iris: drop cache coherent cpu mapping for external BO
2835 - intel/perf: Enable MDAPI queries for Gen12
2836 - anv: skip writing perfcntr in results on Gen12+
2837 - util/sparse_free_list: manipulate node pointers using atomic
2839 - iris: fail screen creation when kernel support is not there
2840 - include/drm-uapi: bump headers
2841 - intel/perf: store default sseu configuration
2842 - intel/perf: specify sseu configuration when supported
2843 - anv: force whole EU array to be powered for perf queries
2844 - drm-shim: provide a valid fake syncobj handle at creation
2845 - drm-shim: stub syncobj wait ioctl
2846 - iris: don't assert on unfinished aux import in copy paths
2847 - anv: don't expose VK_INTEL_performance_query without kernel support
2851 - scons/windows: Support build with LLVM 10.
2852 - util: Make process_test path compatible with mingw native toolchains
2854 Louis-Francis Ratté-Boulianne (7):
2856 - glsl/linker: add DisableTransformFeedbackPacking workaround
2857 - glsl/linker: handle array/struct members for DisableXfbPacking
2858 - glsl/linker: add xfb workaround for modified built-in variables
2859 - gallium: add PIPE_CAP_PACKED_STREAM_OUTPUT
2860 - gallium: add PIPE_CAP_VIEWPORT_TRANSFORM_LOWERED
2861 - gallium: add PIPE_CAP_PSIZ_CLAMPED
2862 - panfrost: fix transform feedback
2866 - etnaviv: retarget transfer to render resource when necessary
2870 - vbo: move GLvertexformat initialization into a template header file
2872 - vbo: use the template for noop GLvertexformat initialization
2873 - vbo: use the template for save GLvertexformat initialization
2874 - vbo: move reusable code from vbo_attrib_tmp.h into vbo_util.h
2875 - mesa: implement missing display list functions while switching to the
2877 - radeonsi: don't report that multi-plane formats are supported
2878 - radeonsi: fix the DCC MSAA bug workaround
2879 - radeonsi: don't update states for the DCC MSAA bug on GFX6-7
2880 - glx: print FPS with 2 decimal places
2881 - mesa: fix incorrect uses of FLUSH_CURRENT
2882 - mesa: remove FLUSH_CURRENT calls that have no effect
2883 - mesa: import PIPE_CAP_SIGNED_VERTEX_BUFFER_OFFSET handling
2884 - vbo: create the immediate mode buffer only in vbo_exec_vtx_map
2885 - vbo: skip FlushMappedBufferRange for glBegin/End by using a
2887 - vbo: don't unmap persistent buffer mappings for glBegin/End
2888 - vbo: remove immediate mode code that doesn't do anything and simplify
2890 - vbo: interleave attrsz, attrtype, and active_sz in memory
2891 - vbo: remove a funky recursive call in glBegin
2892 - vbo: don't check ctx->NewState twice in glBegin
2893 - vbo: keep the immediate mode buffer always mapped for simplicity
2894 - vbo: don't set FLUSH_UPDATE_CURRENT for glVertex
2895 - vbo: pass only either uint32_t or uint64_t into ATTR_UNION
2896 - vbo: don't store glVertex values temporarily into exec
2897 - vbo: optimize resizing vertex attributes during immediate mode
2898 - vbo: fix resizing 64-bit vertex attributes
2899 - vbo: use FlushVertices flags properly and clear NeedFlush correctly
2900 - vbo: increase the size of the immediate mode buffer to decrease draw
2902 - vbo: add/update unlikely statements in ATTR_UNION
2903 - vbo: delay flagging FLUSH_STORED_VERTICES until glEnd
2904 - vbo: also map the immediate mode buffer for read
2905 - vbo: clean up resetting vertex attribs
2906 - vbo: merge use_buffer_objects into vbo_CreateContext to skip the big
2908 - í965: don't use \_mesa_prim::is_indirect
2909 - mesa: remove unused \_mesa_prim::is_indirect
2910 - mesa: don't use bitfields in \_mesa_prim
2911 - st/mesa: optimize st_update_array with ALWAYSINLINE
2912 - radeonsi: don't wait for shader compilation to finish when destroying
2914 - mesa: translate into gallium vertex formats in mesa/main
2915 - mesa: remove unused \_mesa_draw_indirect
2916 - st/mesa: always inline the code setting non-64bit vertex elements
2917 - st/mesa: simplify determination whether a draw has user vertex
2919 - st/mesa: simplify determination whether a draw needs min/max index
2920 - st/mesa: change some loops from while to do..while in st_atom_array.c
2921 - st/mesa: make st_setup_current static
2922 - st/mesa: simplify releasing the current attrib buffer
2923 - gallium/u_upload_mgr: reduce dereferences by adding buffer_size
2924 - gallium/u_upload_mgr: don't do align twice in the u_upload_alloc fast
2926 - gallium/u_vbuf: adjust the heuristic for unrolling indices
2927 - gallium/cso_hash: inline a bunch of functions
2928 - gallium/cso_hash: make cso_hash declared within structures instead of
2930 - gallium/cso_hash: remove always constant variable nodeSize
2931 - gallium/cso_hash: cosmetic changes, no behavior changes
2932 - gallium/cso_hash: remove another layer of pointer indirection
2933 - st/mesa: try to fix MSVC build failure due to ALWAYS_INLINE
2934 - vbo: remove dead code in vbo_can_merge_prims
2935 - vbo: remove redundant code in vbo_exec_fixup_vertex
2936 - mesa: document \_mesa_prim::begin/end
2937 - mesa: don't use memset in glDrawArrays
2938 - mesa: fix immediate mode with tessellation and varying patch vertices
2939 - gallium/util: remove unused u_surfaces.c/h
2940 - util: remove the dependency on kcmp.h
2941 - nir: fix gl_nir_lower_images for bindless images
2942 - tgsi_to_nir: set num_images and num_samplers with holes correctly
2943 - gallium/hash_table: consolidate hash tables with pointer keys
2944 - gallium/hash_table: consolidate hash tables with FD keys
2945 - gallium/hash_table: use the same callback signatures as
2947 - gallium/hash_table: turn it into a wrapper around util/hash_table
2948 - gallium/hash_table: remove some function wrappers
2949 - mesa: remove leftovers from ARB_shadow_ambient
2950 - mesa: call FLUSH_VERTICES before updating CoordReplace
2951 - i965: stop using "indirect" parameter from Driver.Draw (non-indirect)
2952 - mesa: remove unused "indirect" parameter from Driver.Draw
2953 - gallium/cso_hash: pack cso_node better
2954 - gallium/cso_hash: inline struct cso_hash_data
2955 - gallium: pass cso_velems_state into cso_context instead of
2957 - gallium/u_threaded: fix uploading user indices with start != 0
2958 - gallium/u_threaded: convert dividing by index_size to a bit shift
2959 - mesa/i965: remove \_mesa_prim::indirect_offset
2960 - mesa: remove redundant \_mesa_prim::is_indexed
2961 - mesa: move num_instances and base_instance out of \_mesa_prim
2962 - mesa: clean up glMultiDrawElements code, use alloca for small draw
2964 - mesa: don't unroll glMultiDrawElements if one count is 0
2965 - mesa: optimize glMultiDrawArrays, call Draw only once (v2)
2966 - mesa: fix incorrect prim.begin/end for glMultiDrawElements
2967 - nir: replace GCC unroll with an option that works on GCC < 8.0
2968 - gallivm: fix 5 warnings
2969 - nir: fix 5 warnings
2970 - mesa: fix 11 warnings
2971 - gallium/u_vbuf: silence a warning by using unreachable
2972 - mesa: add index_size_shift = log2(index_size) into
2974 - mesa: replace some index_size multiplications and divisions with
2976 - vbo: don't look at the second draw's count when merging 2 glBegin/End
2978 - vbo: deduplicate copy_vertices functions
2979 - vbo: clean up vbo_copy_vertices
2980 - vbo: handle GS and tess primitive types when splitting Begin/End
2981 - vbo: clean up conditional blocks in ATTR_UNION
2982 - vbo: fold code from vbo_exec_fixup_vertex to
2984 - Revert "mesa: check for z=0 in \_mesa_Vertex3dv()"
2985 - mesa: remove \_mesa_index_buffer::index_size in favor of
2987 - mesa: optimize get_index_size
2988 - mesa: deduplicate draw indirect functions
2989 - vbo: merge more primitive types for glBegin/End (v2)
2990 - vbo: merge draws even when begin==0 or end==0
2991 - glthread: don't generate the sync fallback if the call size is not
2993 - glthread: don't prefix variable_data with const
2994 - glthread: inline \_mesa_unmarshal_dispatch_cmd and convert the switch
2996 - glthread: reduce pointer dereferences in glthread_unmarshal_batch
2997 - glthread: use int instead of size_t where it's OK
2998 - glthread: simplify repeated function sequences in marshal_generated.c
2999 - glthread: don't insert \_mesa_post_marshal_hook into every function
3000 - glthread: don't increment variable_data if it's the last
3001 variable-size param
3002 - glthread: add GL_DRAW_INDIRECT_BUFFER tracking and generator support
3003 - glthread: add/update count and marshal fields for many GL functions
3004 - glthread: handle complex pointer parameters and support GL functions
3006 - glthread: check the size of all variable params and clean up the code
3007 - glthread: replace custom ClearBuffer marshalling with generated one
3008 - glthread: add support for TexParameteri and SamplerParameteri
3010 - glthread: add support for glFog, glLight, glLightModel, glTexEnv,
3012 - glthread: add support for glClearNamedFramebuffer, glMaterial,
3014 - glthread: add support for glCallLists, glPatchParameterfv
3015 - glthread: add support for glMemoryObjectParameteriv,
3017 - glthread: don't insert an empty line after (void) cmd;
3018 - glthread: add marshal_call_after and remove custom glFlush and
3020 - glthread: track for each VAO whether the user has set a user pointer
3021 - glthread: sync instead of disabling glthread for non-VBO pointers
3022 - glthread: replace custom glBindBuffer marshalling with generated one
3023 - glthread: merge glBufferData and glNamedBufferData into 1 set of
3025 - glthread: merge glBufferSubData and glNamedBufferSubData into 1 set
3027 - glthread: add custom marshalling for glNamedBuffer(Sub)DataEXT
3028 - glthread: fix a crash with incorrect glShaderSource parameters
3029 - glthread: fall back if a param size is non-zero and a pointer param
3031 - radeonsi: add a bug workaround for NGG - LATE_ALLOC_GS
3032 - ac: add a bug workaround for the 100% NGG culling case
3033 - radeonsi: determine uses_bindless_samplers correctly
3034 - st/mesa: flush the bitmap cache before st/dri and vbo flushes
3035 - st/mesa: fix a possible crash with selection and feedback modes
3036 - gallium/cso_context: remove cso_delete_xxx_shader helpers to fix the
3038 - st/mesa: keep serialized NIR instead of nir_shader in st_program
3039 - vbo: use vbo_exec_wrap_upgrade_vertex for glVertex in ATTR_UNION
3040 - vbo: fix transitions from glVertexN to glVertexM where M < N
3041 - vbo: fix vbo_copy_vertices for GL_PATCHES and adjacency primitive
3043 - gallium: add PIPE_CAP_DRAW_INFO_START_WITH_USER_INDICES
3044 - mesa: don't unroll glMultiDrawElements with user indices for gallium
3045 - radeonsi/gfx10: cache metadata in L2 on small chips
3046 - radeonsi: set better tessellation tunables on gfx9 and gfx10
3047 - radeonsi: tune primitive binning for small chips
3048 - ac: add radeon_info::use_late_alloc to control LATE_ALLOC globally
3049 - ac: disable late alloc on small gfx10 chips
3050 - gallium/u_threaded: don't sync the thread for all unsychronized
3052 - gallium/u_vbuf: simplify the first if statement in
3054 - ac: unify denorm setting enforcement
3055 - ac: set new LLVM denormal flags
3056 - ac: don't set old denormals flags with LLVM >= 11
3057 - nir: fix clip/cull_distance_array_size in
3059 - mesa: use vbo_attrib_tmp.h to generate display list vertex attrib
3061 - mesa: remove redundant api_loopback functions
3062 - glthread: align the batch buffer to 8 bytes for pointers and doubles
3064 - glthread: enable display lists
3065 - glthread: track VAOs created by CreateVertexArrays
3066 - glthread: don't execute any custom VAO and BindBuffer code in the
3068 - glthread: remove debug_print_marshal function
3069 - glthread: clean up debug_print_sync code
3070 - glthread: don't declare unmarshal functions as inline
3071 - winsys/radeon: change to 3-space indentation
3072 - driconf: enable glthread for "From The Depths"
3073 - glthread: remove \_mesa_post_marshal_hook, because it's not very
3075 - glthread: simplify printing safe_mul in gl_marshal.py
3076 - glthread: autogenerate prototypes for custom-marshalled functions
3077 - glthread: move buffer functions into glthread_bufferobj.c
3078 - glthread: rename marshal.h/c to glthread_marshal.h and
3080 - mesa: put gl_thread_state inside gl_context to remove pointer
3082 - glthread: handle buffer unbinding via glDeleteBuffers
3083 - glthread: rename non_vbo helper functions
3084 - glthread: track which vertex array attribs are enabled
3085 - glthread: ignore vertex arrays with user pointers if they're disabled
3086 - glthread: remove the marshal_fail XML attribute
3087 - vbo,gallium: make glBegin/End buffer size configurable by drivers
3088 - ac: fix fast division
3089 - st/mesa: fix use of uninitialized memory due to st_nir_lower_builtin
3090 - glthread: inline SET_func and add -O1 to build
3092 - glthread: declare marshal and unmarshal functions as non-static
3093 - glthread: compile marshal_generated.c faster by breaking it up into 8
3095 - nir: add and gather shader_info::writes_memory
3096 - glsl_to_tgsi: set shader_info::writes_memory
3097 - mesa: allow out-of-order drawing to optimize immediate mode if it's
3099 - radeonsi: enable full out-of-order drawing when
3101 - mesa: try to fix the android build
3102 - Move compiler.h and imports.h/c from src/mesa/main into src/util
3103 - mesa: don't use <> for including internal headers
3104 - util: stop including files from mesa/main
3105 - radv: stop including files from mesa/main
3106 - util: don't include p_defines.h and u_pointer.h from gallium
3107 - util: remove duplicated MALLOC_STRUCT and CALLOC_STRUCT
3108 - radeonsi: remove obsolete TODO comment related to compute-based
3110 - radeonsi: fix incorrect ordered_wave_id initilization for
3111 compute-based culling
3112 - radeonsi: set amdgpu-gds-size for mode == 2 of compute-based culling
3113 - radeonsi: always create wait_mem_scratch for compute-based culling
3114 - radeonsi: add num_vbos_in_user_sgprs into the shader cache key
3115 - radeonsi/gfx10: don't use NGG culling if compute-based culling is
3117 - radeonsi/gfx10: fix ds.ordered.add intrinsic for compute-based
3119 - radeonsi/gfx10: user correct ACQUIRE_MEM packet for compute-based
3121 - radeonsi/gfx10: fix the wave size for compute-based culling
3122 - radeonsi/gfx10: fix descriptors and compute registers for
3123 compute-based culling
3124 - gallium/u_threaded: call the driver to pin threads to L3 immediately
3125 - st/mesa: add environment variable pin_app_thread for faster glthread
3127 - driconf: whilelist more games for glthread
3128 - mesa: optimize initialization of new VAOs
3129 - mesa: don't ever set NullBufferObj in gl_vertex_array_binding
3130 - mesa: don't ever bind NullBufferObj for glBindBuffer targets
3131 - mesa: don't ever bind NullBufferObj to glBindBuffer(Base,Range) slots
3132 - mesa: remove NullBufferObj
3133 - mesa: remove no longer needed \_mesa_is_bufferobj function
3134 - mesa: precompute \_mesa_primitive_restart_index during state changes
3135 - mesa: split \_mesa_primitive_restart_index into a function without
3137 - vbo: expose helper function vbo_get_minmax_index_mapped for glthread
3138 - util: move and adjust the vertex upload heuristic equation from
3140 - st/mesa: fix a crash due to passing a draw vertex shader into the
3142 - ac: out-of-order rasterization is not supported on gfx10
3143 - ac,radeonsi: simplify checking for Navi1x chips
3144 - radeonsi: use pipe_blend_state::max_rt to update fewer blend
3146 - ac: force enable -structurizecfg-skip-uniform-regions for LLVM 11
3147 - ac: update and document fast math flags used by radeonsi
3148 - ac: generate FMA for inexact instructions for radeonsi
3149 - ac: reassociate FP expressions for inexact instructions for radeonsi
3150 - mesa: replace \_NEW_EVAL with vbo_exec_update_eval_maps
3151 - mesa: reset primitive restart state in glClientAttribDefaultEXT
3152 - mesa: remove exec="dynamic" from Draw functions that are not really
3154 - glthread: use 32-bit align instead of 64-bit ALIGN
3155 - glthread: reduce dereferences of the next batch
3156 - glthread: use GLenum16 in batch buffers to save space
3157 - glthread: sort variables in marshal structures to pack them optimally
3158 - gallium: add PIPE_CAP_MAP_UNSYNCHRONIZED_THREAD_SAFE for glthread
3159 - mesa: add Const.BufferCreateMapUnsynchronizedThreadSafe &
3161 - mesa: add offset_is_int32 param into \_mesa_bind_vertex_buffer for
3163 - mesa: extend \_mesa_bind_vertex_buffer to take ownership of the
3165 - mesa: replace GLenum target with gl_shader_stage in NewProgram
3166 - ac/surface: rename micro tile mode enums like gfx10 uses them
3167 - ac/surface: remove RADEON_SURF_TC_COMPATIBLE_HTILE and assume it's
3169 - ac/surface: replace RADEON_SURF_OPTIMIZE_FOR_SPACE with
3171 - ac/surface: match get_display_flag() with expectations for
3173 - ac/surface: don't compute DCC if it's unsupported by DCN on gfx9+
3174 - ac/surface: move non-displayable DCC to the end of the buffer
3175 - ac/surface: add code for gfx10 displayable DCC
3176 - ac/surface: validate that DCC is enabled correctly on gfx9+
3177 - ac: enable displayable DCC on Navi12 & Navi14
3178 - mesa: report GL_INVALID_OPERATION for invalid glTextureBuffer target
3179 - st/mesa: expose more SPIR-V capabilities
3180 - radeonsi: unify and align down the max SSBO/TBO/UBO buffer binding
3182 - radeonsi: revert an accidental change in si_clear_buffer
3183 - Revert "ac/surface: remove RADEON_SURF_TC_COMPATIBLE_HTILE and assume
3185 - Revert "ac: reassociate FP expressions for inexact instructions for
3187 - ac/surface: fix MSAA crash with FORCE_SWIZZLE_MODE on gfx9
3188 - radeonsi: fix compilation of monolithic PS
3189 - radeonsi: don't expose 16xAA on chips with 1 RB due to an occlusion
3194 - etnaviv: Destroy rsc->pending_ctx set in etna_resource_destroy()
3195 - etnaviv: Emit PE.ALPHA_COLOR_EXT\* on GPUs with half-float support
3196 - etnaviv: Fix depth stencil ops on GC880/GC2000
3197 - etnaviv: Disable seamless cube map on GC880
3201 - nir: check shader type before writing to shaderinfo.tess union
3202 - nir: place aligned members after bitfields in shader_info.tess
3206 - util/blob: Add overwrite function for uint8
3207 - tgsi/util: Change boolean for bool
3211 - freedreno: android: fix build failure on android due to python
3213 - freedreno: android: add a6xx-pack.xml.h generation to android build
3214 - freedreno: android: fix build of perfcounters.
3218 - egl: Implement getImage/putImage on pbuffer swrast.
3219 - mesa: Fix FLUSH_VERTICES in SubpixelPrecisionBiasNV.
3220 - egl: Fix A2RGB10 platform_{device,surfaceless} PBuffer configs.
3221 - egl: Factor out dri2_add_pbuffer_configs_for_visuals
3223 - mesa: Check for OpenGL state change before flushing vertices.
3224 - mesa: Flush vertices before changing the OpenGL state.
3225 - i965: Move down genX_upload_sbe in profiles.
3226 - iris: Move down iris_emit_sbe_swiz in profiles.
3227 - i965: Use 32 bit u_bit_scan for vertex attribute setup.
3228 - i965: Use the VAOs binding information in array setup.
3229 - i965: Test original vertex array pointer to skip array upload.
3230 - i965: Split merge_inputs and clear_buffers.
3231 - i965: Reorder workaround flags computation.
3232 - i965: Remove glbinding from brw_vertex_element.
3233 - mesa: Remove now unused \_mesa_draw_attrib_and_binding.
3234 - mesa: Remove now unused \_mesa_draw_attrib.
3235 - mesa: Provide gl_vertex_format accessors.
3236 - i965: Make use of the vertex format functions in i965.
3237 - i965: Use gl_vertex_format in brw_vertex_element.
3241 - intel/tools: Do not print type/qualifiers/name for c_literal
3242 - intel/vec4: Make implied_mrf_writes() a vec4_instruction method
3243 - intel/compiler: Remove unnecessary local variables
3244 - intel/compiler: Make instructions_to_schedule a local variable
3245 - intel/compiler: Mark some methods and parameters const
3246 - intel/compiler: Mark visitor parameters to scheduler const
3247 - intel/compiler: Pass backend_shader \* to cfg_t()
3248 - intel/compiler: Pass shader_stats for each SIMD mode
3249 - intel/compiler: Discount NOPs from instruction counts
3250 - isl: Avoid EXPECT_DEATH in unit tests
3251 - meson: Specify the maximum required libdrm in dri.pc
3255 - android: gallium/auxiliary: fix "Unused source files" in tesselator
3256 - android: aco: fix PIPE_FORMAT related building errors
3257 - android: r600/sfn: fix includes and libmesa_nir dependency
3258 - android: r600/sfn: Add GDS instructions
3259 - android: aco: add various compiler statistics
3263 - gitlab-ci: Update to latest ci-templates HEAD
3264 - gitlab-ci: Pass -j4 to make
3265 - gitlab-ci: Merge ccache and libxml2-utils into main apt-get install
3266 - gitlab-ci: Add ppc64el and s390x cross-build jobs
3267 - gitlab-ci: Build radeonsi & RADV in the ppc64el job
3268 - llvmpipe: Bump test timeout to 180 seconds
3269 - gitlab-ci: Only use gstreamer runners for the s390x job for now
3270 - gitlab-ci: Sort random failure softpipe skips
3271 - gitlab-ci: Add three more dEQP-GLES31 tests to softpipe skips
3272 - st/vdpau: Only call is_video_format_supported hook if needed
3273 - winsys/amdgpu: Make local variable r signed
3274 - util: Change os_same_file_description return type from bool to int
3275 - gitlab-ci: Drop "test-" prefix from llvmpipe/softpipe job names
3276 - gitlab-ci: Distribute jobs across more stages
3277 - gitlab-ci: Always name artifacts archive after the job producing it
3278 - gitlab-ci: Don't restrict ppc64el/s390x build jobs to gstreamer
3280 - gitlab-ci: Don't use buster-backports packages by default for
3282 - gitlab-ci: Fold scons-swr job into scons job
3283 - gitlab-ci: Move classic driver testing to a new meson-classic job
3284 - llvmpipe: Use uintptr_t for pointer values
3285 - gitlab-ci: Enable more Gallium drivers in meson-i386 job
3286 - gitlab-ci: Restrict s390x/ppc64el jobs to packet runners
3287 - gitlab-ci: Update to current templates
3288 - gitlab-ci: Rename "paths" YAML anchor to "all_paths"
3289 - gitlab-ci/lava: Add needs: for container image to test jobs (again)
3290 - gitlab-ci: Don't require triggering build/test jobs manually
3291 - gitlab-ci: Run merge request pipelines automatically only for Marge
3293 - gitlab-ci: Use all_paths in .test-manual rules
3294 - gbm/dri: Propagate queryDmaBufModifiers return value
3295 - amd/addrlib: Use enum instead of sparse chars to identify dimensions
3296 - mesa: Skip 3-byte array formats in \_mesa_array_format_flip_channels
3297 - Revert "ac,radeonsi: fix compilations issues with LLVM 11"
3298 - Revert "gallium/gallivm: fix compilation issues with llvm 11"
3302 - zink: set UBO alignments in nir_intrinsic_load_uniform lowering
3303 - zink: remove framebuffer cache
3304 - zink: explicitly unref old fb object when setting new one
3305 - iris: move iris_vtable to iris_screen
3306 - gallium: add pipe cap for scissored clears and pass scissor state to
3308 - iris: handle PIPE_CAP_CLEAR_SCISSORED
3312 - isl: Add a module which manages aux resolves
3313 - iris: Use isl_aux_usage_has_fast_clear()
3314 - iris: Use ISL's access preparation functions
3315 - iris: Use isl_aux_state_transition_write()
3316 - i965: Use ISL's access preparation functions
3317 - i965: Use isl_aux_state_transition_write()
3321 - dri_util: Update internal_format to GL_RGB8 for
3326 - svga: fix size of format_conversion_table[]
3327 - svga: Use pipe_shader_state_from_tgsi to set shader state
3331 - gitlab-ci/lava: fix handling of lava tags
3332 - Revert "ci: Remove T820 from CI temporarily"
3333 - gitlab-ci: add FILES_HOST_URL and move FILES_HOST_NAME into jobs
3334 - gitlab-ci: re-enable mali400/450 and t820 jobs
3338 - nir/opcodes: Add nir_op_f2fmp
3339 - glsl: Add support for float16 types in the IR tree
3340 - glsl: Add IR conversion ops for 16-bit float types
3341 - glsl: Add b2f16 and f162b conversion operations
3342 - glsl: Add ir_unop_f2fmp
3343 - glsl/validate: Allow float16 in the expression tree
3344 - glsl/lower_instructions: Use float16 constants when appropriate
3345 - glsl/opt_minmax: Add support for float16
3346 - glsl: Add a method to get precision from a deref instruction
3347 - glsl/hierarchical_visitor: Call leave_callback on leaf nodes
3348 - glsl: Add an IR lowering pass to convert mediump operations to 16-bit
3349 - glsl/standalone: Add an option to lower the precision
3350 - glsl: Add unit tests for the lower_precision pass
3351 - freedreno/ir3: Lower bools to bitsize
3352 - glsl: Inline builtins in a separate pass
3353 - glsl/lower_precision: Lower builtins depending on arguments
3354 - glsl/lower_precision: Use vector.back() instead of vector.end()[-1]
3358 - intel: fix the gen 11 compute shader scratch IDs
3359 - intel: fix the gen 12 compute shader scratch IDs
3360 - intel/device: bdw_gt1 actually has 6 eus per subslice
3361 - anv: multiply the scratch space by 4 on gen9-10 like iris and i965
3362 - iris: remove hole from struct iris_bo
3363 - iris: remove unnecessary forward declaration
3364 - iris: remove useless bo->gtt_offset assignment
3365 - iris: make BATCH_SZ smaller by BATCH_RESERVED bytes
3369 - radeonsi: make si_fence_server_signal flush pipe without work
3373 - clover/nir: Check the result of spirv_to_nir
3375 Pierre-Eric Pelloux-Prayer (44):
3377 - radeonsi/ngg: add VGT_FLUSH when enabling fast launch
3378 - radeonsi: test subsampled format in testdma
3379 - format: add format_to_chroma_format
3380 - gallium/video: remove pipe_video_buffer.chroma_format
3381 - gallium/vl: add 4:2:2 support
3382 - radeonsi: fix surf_pitch for subsampled surface
3383 - st/va: enable 4:2:2 chroma format
3384 - st/va: add support YUY2
3385 - radeonsi: remove AMD_DEBUG=sisched option
3386 - omx: fix build with gcc 10
3387 - meson: enable -fno-common by default
3388 - gitlab-ci: rules:changes to test on tested drivers changes
3389 - vdpau: remove bogus assert
3390 - st/mesa: disallow deferred flush if there are multiple contexts
3391 - radeonsi: enable glsl_zero_init for Curse of the Dead Gods
3392 - radeonsi: clarify the conditions when FLUSH_AND_INV_DB is needed
3393 - util/os_file: extend os_read_file to return the file size
3394 - util/u_process: add util_get_process_exec_path
3395 - util/xmlconfig: add new sha1 application attribute
3396 - radeonsi: enable workarounds for YoYo engine based games
3397 - util/u_process: fix Windows build
3398 - nir: update uses_demote flag in discard_to_demote pass
3399 - ac: fix ac_build_is_helper_invocation when postponed_kill is null
3400 - util: fix process_test path
3401 - ddebug: add missing forward declaration
3402 - radeon: fix includes
3403 - radeonsi: switch to 3-spaces style
3404 - radeon: switch to 3-spaces style
3405 - gallium/util: let shader live cache users know if a hit occured
3406 - radeonsi: dump shader stats when hitting the live cache
3407 - util/xmlconfig: fix sha1 comparison code
3408 - mesa: update pipeline when re-linking a program in use
3409 - gallium/u_threaded: flush batch when hitting mapping limit
3410 - radeonsi: use thread_context::bytes_mapped_limit
3411 - radeonsi: don't assume ctx is always a threaded_context
3412 - radeonsi: skip vs output optimizations for some outputs
3413 - mesa: fix crash in find_value
3414 - gallium/utils: silence strncpy warning
3415 - st/omx: fix gcc warnings
3416 - radeonsi: fix export count
3417 - mesa: add gl_coontext::ForceIntegerTexNearest
3418 - driconf: add force_integer_tex_nearest option
3419 - radeonsi: don't print gs_copy_shader stats for shaderdb
3420 - amd/addrlib: fix forgotten char -> enum conversions
3424 - intel/compiler: Add support for variable workgroup size
3425 - i965: Implement ARB_compute_variable_group_size
3429 - lima: remove definition of lima_is_scanout
3430 - lima: use util_copy_framebuffer_state
3431 - lima: always add texture bo to submit
3432 - lima: remove lima_ctx_buff_va submit flags (v2)
3433 - lima: pass array as parameter to PLBU and VS command macros
3434 - lima: delay add plb buffer to submit when flush
3435 - lima: delay plbu head command generation to flush stage (v2)
3436 - lima: add render target to submit by dirty buffer flags
3437 - lima: add missing resolve check for damage and reload
3438 - lima: move syncobj from lima_submit to lima_context
3439 - lima: merge gp/pp submit
3440 - lima: put hardware related info to lima_gpu.h
3441 - lima: move flush code to lima_submit.c
3442 - lima: pass submit parameter for functions in lima_submic.c (v2)
3443 - lima: add lima_submit_create_stream_bo
3444 - lima: adjust pp_stream to use lima_submit_create_stream_bo
3445 - lima: use lima_submit_create_stream_bo for plbu/vs_cmd and pp_stack
3446 - lima: add lima_submit_get
3447 - lima: make lima_submit one time use drop data (v3)
3448 - lima: track write submits of context (v3)
3449 - lima: move plbu/vs_cmd_array into lima_submit
3450 - lima: move resolve into lima_submit
3451 - lima: move pp_max_stack_size to lima_submit
3452 - lima: move damage_rect into lima_submit
3453 - lima: move clear into submit (v2)
3454 - lima: move framebuffer info to lima_submit
3455 - lima: use per submit dump file
3456 - lima: optinal flush submit in lima_clear
3457 - lima: enable multi submit optimization
3458 - lima: move dump check to macro for lima_dump_command_stream_print
3459 - lima: rename lima_submit to lima_job
3460 - lima: fix buffer import with offset
3461 - lima: also check tiled and depth case when import
3462 - lima: set offset when export resource
3463 - panfrost: don't always build bifrost_compiler
3467 - meson: Use dependency.partial_dependency()
3471 - intel: Load the driver even if I915_PARAM_REVISION is not found.
3472 - intel/tools: Update aubinator_error_decode.
3473 - intel/blorp: Implement GEN:BUG:1605967699.
3474 - iris: Apply the flushes when switching pipelines.
3475 - anv: Wait for the GPU to be idle before invalidating the aux table.
3476 - iris: Split aux map initialization from invalidation.
3477 - iris: Wait for the GPU to be idle before invalidating the aux table.
3478 - intel/isl: Implement D16_UNORM workarounds.
3479 - intel/gen12+: Disable mid thread preemption.
3480 - iris: Enable EXT_depth_bounds_test extension.
3481 - drm-uapi: Update headers from Linux 5.7-rc1.
3482 - i965/bufmgr: Factor out GEM_MMAP ioctl from mmap_cpu and mmap_wc.
3483 - iris/bufmgr: Factor out GEM_MMAP ioctl from mmap_cpu and mmap_wc.
3484 - i965/bufmgr: Add support for MMAP_OFFSET ioctl.
3485 - iris/bufmgr: Add support for MMAP_OFFSET ioctl.
3486 - anv: Add anv_device parameter to anv_gem_munmap.
3487 - anv: Add support for new MMAP_OFFSET ioctl.
3488 - anv: Enable HiZ on multi-layer depth buffers.
3492 - aco: fix gfx10_wave64_bpermute
3493 - aco: gfx10_wave64_bpermute reduce op to print_ir
3494 - aco: disable some instruction combining if it could change an exec
3496 - aco: improve SCC handling in some SALU combines
3497 - nir: fix nir_const_value_as_uint bit size in load/store vectorizer
3499 - gitlab-ci: remove load_store_vectorizer from expected s390x test
3501 - aco: add RegisterFile
3502 - aco: add some helpers for filling/testing register ranges
3503 - aco: improve GFX9 1D ddx/ddy assertion
3504 - spirv: improve creation of memory_barrier
3505 - spirv: fix memory_barrier_tcs_patch emission
3506 - aco: keep track of which events are used in a barrier
3507 - aco: fix carry-out size for wave32 v_add_co_u32_e64
3508 - aco: handle v_add_co_u32_e64 in parse_base_offset()
3509 - aco: add new NOP insertion pass for GFX6-9
3510 - aco: improve get_wait_states()
3511 - aco: consider non-hazard writes in handle_raw_hazard_internal
3512 - aco: improve control flow handling in GFX6-9 NOP pass
3513 - aco: only reserve sgprs for vcc if it's used
3514 - aco: fix uninitialized data error in waitcnt pass
3515 - glsl/list: use uintptr_t for exec_node_data()'s subtraction
3516 - aco: add helpers for moving instructions for scheduling
3517 - aco: add helpers for ensuring correct ordering while scheduling
3518 - aco: allow barriers to be skipped during scheduling
3519 - aco: don't stop scheduling at exports
3520 - aco: move some register demand helpers into aco_live_var_analysis.cpp
3521 - aco: add a late kill flag
3522 - aco: set late kill for v_interp_p1_f32 for some APUs
3523 - aco: fix instruction encoding for LS VGPR init bug workaround
3524 - aco: fix operand order for LS VGPR init bug workaround
3525 - nir/gather_info: handle emit_vertex_with_counter
3526 - radv: call nir_shader_gather_info again
3527 - radv/winsys: set has_syncobj_wait_for_submit in the null winsys
3528 - aco: set has_divergent_branch for discards in loops
3529 - aco: handle missing second predecessors at merge block phis
3530 - aco: handle when ACO adds new continue edges
3531 - aco: skip NIR in unreachable merge blocks
3532 - aco: improve check for unreachable loop continue blocks
3533 - aco: emit IR in IF's merge block instead if the other side ends in a
3535 - aco: fix boolean undef regclass
3536 - nir/gather_info: fix per-vertex handling in try_mask_partial_io
3537 - aco: remove dead code in handle_operands()
3538 - aco: implement 64-bit VGPR constant copies in handle_operands()
3539 - aco: look at p_{extract,split}_vector's definitions in
3541 - glsl: fix race in instance getters
3542 - util/u_queue: fix race in total_jobs_size access
3543 - radv: add code for exposing compiler statistics
3544 - aco: add various compiler statistics
3545 - aco: add vmem/smem score statistic
3546 - radv, aco: collect statistics if requested but executables are not
3547 - radv: fix null winsys gpu_info array
3548 - aco: make PhysReg in units of bytes
3549 - aco: add SDWA_instruction
3550 - aco: print and validate opsel
3551 - aco: add emission support for register-allocated sdwa sels
3552 - aco: remove divergence check in sanitize_if()
3553 - aco: zero-initialize Temp
3554 - aco: improve vector optimization with sub-dword vectors
3555 - aco: fix p_extract_vector validation
3556 - aco: improve p_create_vector RA for sub-dword operands
3557 - aco: clear moved operands in get_reg_create_vector()
3558 - aco: fix 1D textureGrad() on GFX9
3559 - aco: implement various 8/16-bit conversions
3560 - aco: add missing scc clobber to nir_op_unpack_32_2x16_split_y
3561 - aco: fix copy statistic for 64-bit vgpr constant copy
3562 - aco: add VOP3P_instruction
3563 - aco: implement sub-dword swaps
3564 - aco: implement 64-bit sgpr swaps
3565 - nir/lower_bit_size: fix lowering of shifts
3566 - nir/lower_bit_size: fix lowering of {imul,umul}_high
3567 - nir/algebraic: don't undo lowering of 8/16-bit comparisons to 32-bit
3568 - aco: decrease the uses of other copy operations after
3570 - aco: copy-propagate p_create_vector copies of vectors
3571 - aco: remove copy in load_input_from_temps()
3572 - aco: move call to store_output_to_temps in store_ls_or_es_output
3574 - aco: combine VALU and SALU into various VOP3 instructions
3575 - aco: improve code for 32-bit isign
3576 - aco: fix v_or(s_lshl) and v_add(s_lshl) optimizations
3577 - aco: fix outdated label_vec from p_create_vector labelling
3578 - radv: align buffer descriptor sizes to dword
3579 - radv: allocate larger shader memory slabs if needed
3580 - aco: be more careful about using SMEM for load_global
3581 - aco: add and use RegClass::get() helper
3582 - aco: add emit_load helper
3583 - aco: refactor load_lds to use new helpers
3584 - aco: use emit_load helper for VMEM/SMEM loads
3585 - aco: add helpers for splitting stores
3586 - aco: refactor store_lds() to use new helpers
3587 - aco: refactor store_vmem_mubuf() to use new helpers
3588 - aco: refactor visit_store_ssbo() to use new helpers
3589 - aco: refactor visit_store_global() to use new helpers
3590 - aco: refactor visit_store_scratch() to use new helpers
3591 - aco: add and use get_buffer_store_op() helper
3592 - aco: allow 8/16-bit shared loads
3593 - aco: vectorize global loads/stores
3594 - aco: handle undef p_create_vector operands in the optimizer
3595 - aco: clobber scc in s_bfe_u32 in get_alu_src()
3596 - aco: improve sub-dword emit_split_vector() with sgprs
3597 - aco: lower 8/16-bit integer arithmetic
3598 - radv/aco: enable 8/16-bit storage and int8/int16 on GFX8+
3599 - aco: make RegisterFile::block() take a regclass
3600 - aco: check alignment of non-subdword registers in get_reg_specified()
3601 - aco: fix neighboring register check in get_reg_simple()
3602 - aco: split self-intersecting copies instead of swapping
3603 - aco: don't recurse in sub-dword get_reg_simple()
3604 - aco: improve RA for uneven p_split_vector
3605 - aco: add missing adjust_max_used_regs()
3606 - aco: fix sub-dword out-of-bounds check in RA validator
3607 - aco: fix sub-dword overwrite check in RA validator
3608 - aco: add various GFX10 int16 opcodes
3609 - aco: improve clamped integer addition disassembly workaround
3610 - aco: fix vgpr nir_op_vecn with sgpr operands
3611 - aco: consider blocks unreachable if they are in the logical cfg
3612 - aco: remove use of f-strings
3613 - aco: add message to static_assert
3614 - nir: add missing group_memory_barrier handling
3615 - nir/opt_if: run opt_peel_loop_initial_if after all other
3617 - nir: fix lowering to scratch with boolean access
3621 - freedreno/drm: readonly cmdstream
3622 - freedreno/ir3: shuffle a few ir3_register fields
3623 - freedreno/ir3: cleanup after lower_locals_to_regs
3624 - freedreno/ir3: fix crash when no non-input instructions
3625 - freedreno/ir3: split out delay helpers
3626 - freedreno/ir3: move nop padding to legalize
3627 - freedreno/ir3: move block-scheduling into legalize
3628 - freedreno/ir3: move atomic fixup after RA
3629 - freedreno/ir3: a bit more optmsgs debug
3630 - freedreno/ir3/ra: make use()/def() functions instead of macros
3631 - freedreno/ir3: fix kill scheduling
3632 - freedreno/ir3: post-RA sched pass
3633 - freedreno/ir3: number instructions from one
3634 - freedreno/ir3: add is_tex_or_prefetch()
3635 - freedreno/ir3: don't precolor unused inputs
3636 - freedreno/ir3: two pass register allocation
3637 - freedreno/a6xx: fix lrz overflow
3638 - freedreno/ir3: add RA sanity check
3639 - freedreno/ir3: remove unused tex arg harder
3640 - freedreno/ir3: create fragcoord instructions in input block
3641 - freedreno/ir3: simplify split from collect
3642 - freedreno/ir3: fix a dirty lie
3643 - freedreno: allow ctx->batch to be NULL
3644 - freedreno/ir3: fold const conversion into consumer
3645 - freedreno: allow INVALID modifier
3646 - freedreno/registers: teach gen_header.py about a3xx_regid
3647 - freedreno/a6xx: few register updates
3648 - freedreno: quiet INFO_MSG
3649 - freedreno/registers: cleanup CP_SET_MARKER
3650 - freedreno/computerator: import parser/lexer from fdre-a3xx
3651 - freedreno/computerator: polish out some of the rust
3652 - freedreno/computerator: rename prefix asm->ir3
3653 - freedreno/ir3: allow block->predecessors to be null
3654 - freedreno/computerator: add computerator
3655 - freedreno/computerator: fix build dependency
3656 - freedreno/ir3: remove from_tgsi
3657 - freedreno/a6xx: remove unused param
3658 - freedreno/a6xx: emit LRZ clear in sysmem too
3659 - freedreno/a6xx: whitespace fix
3660 - freedreno/a6xx: don't emit YIELD packet
3661 - freedreno/a6xx: enable SKIP_IB2_ENABLE properly
3662 - freedreno: honor FD_MESA_DEBUG=nogrow
3663 - freedreno/ir3: remove regmask_set_if_not()
3664 - freedreno/ir3: rewrite regmask to better support a6xx+
3665 - freedreno/ir3: don't hide latency when there is none to hide
3666 - freedreno/ir3: track half-precision live values
3667 - freedreno/ir3: update SFU delay
3668 - freedreno/ir3: fix crash with samgq workaround
3669 - freedreno/ir3: don't precolor unassigned inputs
3670 - freedreno/ir3: fix assert with getinfo
3671 - freedreno/ir3: add assert
3672 - nir/print: show variable precision
3673 - freedreno/ir3: also lower lowp frag outputs
3674 - freedreno/computerator: add hrsq/hlog2/hexp2
3675 - freedreno/ir3: remove extra nops inserted in scheduler
3676 - freedreno/ir3: add simplified stall estimation
3677 - freedreno: fix FD_MESA_DEBUG=inorder
3678 - util/ra: spiff out select_reg_callback
3679 - util/ra: move NO_REG to header
3680 - freedreno/ir3: split out has_latency_to_hide()
3681 - freedreno/ir3: fix has_latency_to_hide
3682 - freedreno/ir3: track register usage in first RA pass
3683 - freedreno/ir3: round-robin RA
3684 - freedreno/ir3: try to avoid syncs
3685 - freedreno/computerator: add performance counter support
3686 - freedreno/fdperf: set locale
3687 - freedreno/a6xx: register update
3688 - freedreno/ir3: small cleanup and comments
3689 - freedreno/ir3: add bary_ij as src for meta:tex_prefetch
3690 - freedreno/ir3: remove unused helper
3691 - freedreno/ir3: fix bogus register footprint with tess/gs
3692 - freedreno/ir3: reformat disasm output
3693 - freedreno/ir3: convert debug bitfield to BITFIELD_BIT()
3694 - freedreno/ir3/ra: add debug option for RA debug msgs
3695 - freedreno/ir3/ra: split-up
3696 - freedreno/ir3/ra: add helper to map name to instruction
3697 - freedreno/ir3/ra: fix target register calculation
3698 - freedreno/ir3/ra: add helper to map name to array
3699 - freedreno/ir3/ra: drop extending output live-ranges
3700 - freedreno/ir3/ra: add def/use iterators
3701 - freedreno/ir3/ra: fix array liveranges
3702 - freedreno/ir3/ra: compute register target from liveranges
3703 - freedreno/ir3/ra: pick higher numbered scalars in first pass
3704 - freedreno/ir3/ra: split building regs/classes and conflicts
3705 - freedreno/ir3/ra: re-work a6xx merged register file conflicts
3706 - gitlab-ci: disable vs2019 build
3707 - freedreno: remove some obsolete debug options
3708 - util: fix u_fifo_pop()
3709 - freedreno: add logging infrastructure
3710 - freedreno/a6xx: timestamp logging support
3711 - freedreno: add some initial fd_log tracepoints
3712 - freedreno/a6xx: add some more tracepoints
3713 - freedreno/log: avoid duplicate ts's
3714 - util: move ALIGN/ROUND_DOWN_TO to u_math.h
3715 - freedreno/ir3: fix android build
3716 - freedreno/log: fix build error
3717 - nir: fix definition of imadsh_mix16 for vectors
3718 - freedreno/ir3/cf: handle widening too
3719 - freedreno/ir3: fixup cat3 32b vs 16b
3720 - freedreno/ir3/cf: skip array load/store
3721 - freedreno/ir3: add a pass to collect SSA uses
3722 - freedreno/ir3/cf: use ssa-uses
3723 - freedreno/a6xx: add some compute logging
3724 - freedreno: fix missing locking
3725 - freedreno/ir3: also precompile compute shaders for shaderdb
3726 - freedreno: limit fp16 to frag and compute
3727 - glsl: don't limit fp16 lowering to frag
3728 - nir: add some swizzle helpers
3729 - nir/lower_amul: fix slot calculation
3730 - freedreno/log: android support
3731 - freedreno/log: spiff out parser some more
3732 - freedreno/log: better decoding for multiple chunks per batch
3733 - freedreno/ir3: spiff out disasm a bit
3734 - freedreno/ir3: make falsedep use's optional
3735 - freedreno/ir3: simplify grouping pass
3736 - freedreno/ir3: fix location of inserted mov's
3737 - freedreno/ir3: new pre-RA scheduler
3738 - freedreno/ir3/sched: awareness of partial liveness
3739 - freedreno/ir3/postsched: remove some leftovers
3740 - freedreno/ir3/postsched: avoid moving tex ahead of kill
3741 - freedreno/ir3: add mov/cov stats
3742 - freedreno/ir3/ra: handle array case for SFU select_reg opt
3743 - freedreno/ir3: better cleanup when removing unused instructions
3744 - freedreno/ir3: rename depth->dce
3745 - freedreno/ir3/ra: cleanup some leftovers
3746 - mesa: avoid redundant VBO updates
3747 - mesa/st: avoid u_vbuf for GLES
3748 - gallium: add # of MRT to blend state
3749 - freedreno/computer: add script to test widening/narrowing
3750 - freedreno/ir3/ra: remove unused variable
3751 - freedreno/ir3/ra: use ir3_debug_print helper
3752 - freedreno/ir3/ra: split out helper for array assignment
3753 - freedreno/ir3/ra: only assign array base in first pass
3754 - freedreno/a6xx+tu: rename VSC_DATA/VSC_DATA2
3755 - freedreno: add helper to estimate # of bins per pipe
3756 - freedreno/a6xx: pre-calculate expected vsc stream sizes
3757 - freedreno/log-parser: support to read gzip'd logs
3758 - freedreno: small whitespace fix
3759 - freedreno: don't realloc idle bo's
3760 - freedreno: mark more state dirty when rebinding resources
3761 - freedreno: optimize rebind_resource()
3762 - freedreno: rebind resource in all contexts
3763 - freedreno: rebind_resource() \*before\* bo changes
3764 - freedreno/a6xx: invalidate tex state cache entries on rebind
3765 - freedreno: fix buffer import
3766 - freedreno/ir3: fix indirect cb0 load_ubo lowering
3767 - freedreno: clear last_fence after resource tracking
3771 - ci: Split out radv build-testing on arm64
3772 - ci: Drop the git dependency in tracie
3773 - tracie: Switch to using shutil.move for cross filesystem moves
3774 - tracie: Print results in a machine readable format
3775 - tracie: Reformat code to fix indentation
3779 - gallivm: fix crash with bptc border color sampling
3780 - gallivm: fix crash in emit_get_buffer_size
3781 - gallivm: disable rgtc/latc SNORM accellerated fetches
3782 - gallium/util: Add back (and rename) util_float_to_half implementation
3783 - gallivm: fix rgtc2 format
3784 - gallivm: switch the mask6/mask7 cases for signed rgtc formats
3785 - gallivm: fix stream id fetch
3789 - panfrost: Align Android makefiles with recent changes
3790 - lima: Add missing source file to Android.mk
3791 - panfrost: Align Android makefiles with recent changes
3795 - intel/isl: Move get_format_encoding function to isl
3796 - intel/isl: Switch to R8_UNORM format for compatiblity
3797 - intel/tools: Handle illegal instruction
3798 - intel/tools: Handle STATE_REG in typed source operand
3799 - intel/tools: Set correct address register file and number in i965_asm
3800 - intel/tools: Add test for address register as source
3801 - intel/tools: Add test for state register as source
3802 - intel/tools: Print c_literals 4 byte wide
3803 - intel/tools: Allow i965_disasm to disassemble c_literal input type
3804 - intel/genxml: Add patch count threshold field on gen12
3805 - intel/compiler: Track patch count threshold
3806 - anv: Set patch count threshold in 3DSTATE_HS
3807 - iris: Set patch count threshold in 3DSTATE_HS
3811 - radv: check buffer size in vkCreateBuffer()
3812 - radv: set sparseAddressSpaceSize to RADV_MAX_MEMORY_ALLOCATION_SIZE
3816 - aco: fix MUBUF VS input loads when expanding vec3 to vec4 on GFX6
3817 - aco: do not use ds_{read,write}2 on GFX6
3818 - gitlab-ci: disable a630 tests as mesa-cheza is down (again)
3819 - aco: fix waiting for scalar stores before "writing back" data on
3820 GFX8-GFX9
3821 - radv: make sure to not submit any IBs when RADV_FORCE_FAMILY is set
3822 - radv: set the chip name to GCN-NOOP when RADV_FORCE_FAMILY is set
3823 - aco: fix creating v_madak if v_mad_f32 has two sgpr literals
3824 - nir: do not use De Morgan's Law rules for flt and fge
3825 - radv: fix line width range and granularity
3826 - radv: implement VK_EXT_line_rasterization
3827 - radv: remove LLVM sicheduler enable for The Talos Principle
3828 - radv: remove RADV_DEBUG=nosisched and RADV_PERFTEST=sisched
3829 - radv: remove unused RADV_HASH_SHADER_IS_GEOM_COPY_SHADER
3830 - radv: remove unnecessary RADV_DEBUG=nobatchchain option
3831 - docs/new_features: empty the feature list for the 20.1 cycle
3832 - radv: enable shaderStorageImageMultisample on GFX6-GFX7
3833 - radv: enable VK_EXT_sampler_filter_minmax on GFX6
3834 - radv: enable VK_NV_compute_shader_derivatives on GFX6-GFX7
3835 - radv: add a comment about VK_AMD_mixed_attachment_samples on
3836 GFX6-GFX7
3837 - docs/envvars: document RADV_TEX_ANISO
3838 - radv/winsys: add a new flag that requests zerovram allocations
3839 - radv: use RADEON_FLAG_ZERO_VRAM when creating the trace BO
3840 - radv: add the trace BO to the BO list at submit time
3841 - radv: implement a dummy winsys for creating devices without AMDGPU
3842 - ac,radeonsi: add ac_gpu_info::lds_size_per_cu
3843 - ac: add more ac_gpu_info related shader fields
3844 - radv/gfx10: adjust the number of simd per compute unit
3845 - radv/gfx10: adjust SGPRs/VGPRs related info
3846 - radv/gfx10: adjust the LDS size used to compute waves
3847 - radv/gfx10: adjust the number of VGPRs used to compute waves
3848 - radv: make use of ac_gpu_info::max_wave64_per_simd
3849 - radv: fix creating null devices if KHR_display is enabled
3850 - ac/llvm: fix 64-bit fmed3
3851 - ac/llvm: fix 16-bit fmed3 on GFX8 and older gens
3852 - ac/llvm: flush denorms for nir_op_fmed3 on GFX8 and older gens
3853 - ac: add more fields to ac_gpu_info
3854 - ac/registers: add definitions for thread trace
3855 - radv: add a small helper that allows to submit internal CS
3856 - radv: add initial SQ Thread Trace support for GFX9
3857 - radv: emit thread trace markers after every draw/dispatch call
3858 - radv: add initial SQTT files generation support
3859 - radv: allow to capture SQTT traces with
3861 - radv: fix 32-bit build failure in radv_queue_internal_submit()
3862 - radv: fix size of sqtt_file_chunk_asic_info on 32-bit system
3863 - radv/rgp: adjust trace memory/shader clocks to fix frame duration
3864 - radv/sqtt: do not assume that the number of shader engines is 4
3865 - radv/sqtt: update SPI_CONFIG_CNTL.EXP_PRIORITY_ORDER value
3866 - ac/registers: add definitions for thread trace on GFX10
3867 - radv/sqtt: add support for GFX10
3868 - radv: update entrypoints generation from ANV
3869 - ac: rename lds_size_per_cu to lds_size_per_workgroup
3870 - ac: rename vgpr_alloc_granularity to wave64_vgpr_alloc_granularity
3871 - ac: rename min_vgpr_alloc to min_wave64_vgpr_alloc
3872 - aco: fix image load/store with lod and 1D images
3873 - gitlab-ci: build Fossilize in the test image for VK
3874 - gitlab-ci: add Fossilize support to detect compiler regressions
3875 - gitlab-ci: enable building the test image for VK unconditionally
3876 - gitlab-ci: add a job that runs Fossilize on RADV/Polaris10
3877 - radv/winsys: fix missing initializations of shader info in the null
3879 - radv/sqtt: fix wrong check in radv_is_thread_trace_complete()
3880 - radv/sqtt: tidy up radv_emit_thread_trace_{start,stop}
3881 - radv/sqtt: add radv_copy_thread_trace_info_regs() helper
3882 - ac/registers: adjust some definitions for thread trace on GFX8
3883 - radv/sqtt: add support for GFX8
3884 - radv/sqtt: abort if SQTT is used on GFX6-GFX7
3885 - ac: add ac_gpu_info::cu_mask to store bitmask of compute units
3886 - radv/rgp: report correct cu_mask info
3887 - radv/rgp: report correct system ram size
3888 - nir/lower_input_attachments: remove bogus assert in
3890 - radv/entrypoints: declare a driver internal layer for SQTT
3891 - radv: use device entrypoints from the SQTT layer if enabled
3892 - radv/sqtt: add a helper that emits thread trace userdata markers
3893 - radv: initial implementation of the driver internal layer SQTT
3894 - radv/sqtt: describe begin/end command buffers with user markers
3895 - radv/sqtt: describe draw/dispatch and emit event markers
3896 - radv/sqtt: describe render pass color/depthstencil clears
3897 - radv/rgp: bump the instrumentation spec version to 1
3898 - radv/sqtt: describe pipeline and wait events barriers
3899 - gitlab-ci: add rules:changes for RADV
3900 - radv: do not recursively begin/end render pass for meta operations
3901 - radv: fix 32-bits build (again)
3902 - gitlab-ci: build RADV in meson-i386 to avoid 32-bit build failures
3903 - ac/llvm: add missing optimization barrier for 64-bit readlanes
3904 - radv/sqtt: describe begin/end subpass barriers with user markers
3905 - radv/sqtt: describe layout transitions with user markers
3906 - radv/gfx10: cache metadata in L2 on small chips
3907 - radv: use better tessellation tunables on GFX9+
3908 - radv: tune primitive binning for small chips
3909 - radv: rewrite late alloc computation
3910 - radv: use ac_gpu_info::use_late_alloc
3911 - radv: cleanup occurences of use_aco everywhere
3912 - radv: remove radv_shader_variant::aco_used
3913 - radv: remove unnecessary LLVM includes
3914 - radv: add llvm_compiler_shader() helper
3915 - gitlab-ci: remove useless 'patch' package in the VK test image
3916 - gitlab-ci: allow deqp-runner to use the maximum number of jobs
3917 - gitlab-ci: do not set the number of deqp-parallel jobs for RADV CTS
3918 - gitlab-ci: bump Vulkan CTS to 1.2.1.0
3919 - radv/sqtt: handle thread trace capture in sqtt_QueuePresentKHR()
3920 - radv: only inject implicit subpass dependencies if necessary
3921 - radv/gfx10: fix required subgroup size with
3923 - radv/gfx10: fix required ballot size with
3925 - radv: fix random depth range unrestricted failures due to a cache
3927 - radv: remove wrong assert that checks compute subgroup size
3928 - radv: fix optional pSizes parameter when binding streamout buffers
3929 - radv/winsys: fix wrong PCI ID for Vega10 in the null winsys
3930 - radv/winsys: spoof some values for num_render_backends in the null
3932 - gitlab-ci: compile fossils with both RADV compiler backends
3934 - gitlab-ci: compile fossils with more ASICs
3935 - gitlab-ci: add a new stage for RADV CI
3936 - gitlab-ci: add a bunch of new fossils from the Sascha Vulkan demos
3937 - radv/llvm: fix subgroup shuffle for chips without bpermute
3938 - radv: enable VK_KHR_8bit_storage on GFX6-GFX7
3939 - ac/nir: use llvm.amdgcn.rcp for nir_op_frcp
3940 - ac/nir: use llvm.amdgcn.rsq for nir_op_frsq
3941 - ac/nir: use llvm.amdgcn.rcp in ac_build_fdiv()
3942 - nir/algebraic: add fexp2(fmul(flog2(a), 0.5) -> fsqrt(a) optimization
3943 - aco: only break SMEM clauses if XNACK is enabled (mostly APUs)
3944 - aco: always optimize v_mad to v_madak in presence of literals
3945 - ac/nir: split 8-bit load/store to global memory on GFX6
3946 - ac/nir: split 8-bit SSBO stores on GFX6
3947 - radv/llvm: enable 8-bit storage features on GFX6-GFX7
3948 - ac/nir: split 16-bit load/store to global memory on GFX6
3949 - ac/nir: split 16-bit SSBO stores on GFX6
3950 - radv/llvm: enable 16-bit storage features on GFX6-GFX7
3951 - radv: rename decompress/resummarize depth/stencil functions
3952 - radv: rename extra graphics pipeline decompress/resummarize fields
3953 - radv: cleanup creating the decompress/resummarize pipelines
3954 - radv: remove radv_layout_has_htile() helper
3955 - radv: enable lowering of GS intrinsics for the LLVM backend
3956 - ac,radv: add ac_gpu_info::has_double_rate_fp16
3957 - radv: only expose shaderFloat16 for chips with double rate fp16
3958 - radv: only expose storageInputOutput16 for chips with double rate
3960 - radv: only expose fp16 control features for chips with double rate
3962 - radv: only enable TC-compat HTILE for images readable by a shader
3963 - radv: allow TC-compat HTILE with GENERAL outside of render loops
3964 - aco: implement 16-bit nir_op_frexp_sig/nir_op_frexp_exp
3965 - aco: implement 16-bit nir_op_ffract
3966 - aco: implement 16-bit nir_op_fexp2/nir_op_flog2
3967 - aco: implement 16-bit nir_op_ftrunc/nir_op_fround_even
3968 - aco: implement 16-bit nir_op_fsqrt/nir_op_frcp/nir_op_frsq
3969 - aco: implement 16-bit nir_op_ffloor/nir_op_fceil
3970 - aco: implement 16-bit nir_op_fmax/nir_op_fmin
3971 - aco: implement 16-bit nir_op_fabs/nir_op_fneg
3972 - aco: implement 16-bit nir_op_fsub/nir_op_fadd
3973 - aco: implement 16-bit nir_op_fcos/nir_op_fsin
3974 - aco: implement 16-bit nir_op_fmul
3975 - aco: implement 16-bit nir_op_fsat
3976 - aco: implement 16-bit nir_op_fsign
3977 - aco: implement 16-bit nir_op_bcsel
3978 - aco: implement 16-bit nir_op_f2i32/nir_op_f2u32
3979 - aco: implement 16-bit nir_op_ldexp
3980 - aco: implement 16-bit nir_op_fmax3/nir_op_fmin3/nir_op_fmed3
3981 - aco: implement 16-bit comparisons
3982 - aco: implement nir_op_b2f16/nir_op_i2f16/nir_op_u2f16
3983 - aco: fix f2i64/f2u64 with sgprs if the exponent computation overflow
3984 - aco: implement 16-bit nir_op_f2i64/nir_op_f2u64
3985 - aco: fix nir_op_pack_32_2x16_split if one operand is a constant
3986 - radv: add radeon_set_context_reg_rmw() helper
3987 - radv: use RMW packets for updating the maximum sample distance
3988 - aco: fix nir_op_frexp_exp with 16-bit floats and negative exponents
3989 - radv/aco: do not advertise VK_KHR_shader_subgroup_extended_types
3990 - aco: implement nir_op_f2i8/nir_op_f2u8
3991 - aco: fix emitting stream output with tess eval shaders
3992 - radv: do not abort with unknown/unimplemented descriptor types
3993 - radv: fix geometry shader primitives query with ACO on GFX10
3994 - radv: set missing SHARED_VGPR_CNT for NGG VS and ACO
3995 - radv/llvm: fix exporting the viewport index if the fragment shader
3997 - aco: fix exporting the viewport index if the fragment shader needs it
3998 - nir/lower_int64: lower imin3/imax3/umin3/umax3/imed3/umed3
3999 - nir/opt_algebraic: lower 64-bit fmin3/fmax3/fmed3
4000 - gitlab-ci: add a list of excluded tests for RADV
4001 - radv: make sure to export the viewport index if FS needs it
4002 - radv: simplify checking for Navi1x chips
4003 - radv: adjust the supported subgroup stages
4004 - radv: fix robust_buffer_access if enabled via
4006 - gitlab-ci: add lists of expected failures for RADV CI
4007 - ac,radeonsi: fix compilations issues with LLVM 11
4008 - radv: do not expose GTT as device local memory mostly for APUs
4009 - radv: enable FMASK for color attachments only
4010 - radv: remove unused radv_device_memory::map_size field
4011 - radv: track memory heaps usage if overallocation is explicitly
4013 - radv: advertise VK_AMD_memory_overallocation_behavior
4014 - ac/llvm: fix nir_texop_texture_samples with NULL descriptors
4015 - aco: fix nir_texop_texture_samples with NULL descriptors
4016 - aco: fix adjusting the sample index with FMASK if value is negative
4017 - radv: handle NULL descriptors
4018 - radv: handle NULL vertex bindings
4019 - radv: advertise VK_EXT_robustness2
4020 - gitlab-ci: add a list of expected failures for FIJI with ACO
4021 - ci: fix reporting the number of unexpected/flakes
4022 - radv: report INITIALIZATION_FAILED when the amdgpu winsys init failed
4023 - radv: don't report error with other vendor DRM devices
4024 - aco: fix 64-bit trunc with negative exponents on GFX6
4025 - radv: limit the Vulkan version to 1.1 for Android
4026 - radv: handle different Vulkan API versions correctly
4027 - radv: update the list of allowed Android extensions
4031 - st/va: GetConfigAttributes: check profile and entrypoint combination
4035 - mesa: add support for NV_pixel_buffer_object
4039 - mesa: enable GL_EXT_draw_instanced for gles2
4043 - radeonsi: enable EXT_texture_shadow_lod
4047 - virgl: Use align_free for align_malloc allocated buffer
4051 - intel/vec4: fix valgrind errors with vf_values array
4052 - glsl: fix a memory leak with resource_set
4053 - iris: fix aux buf map failure in 32bits app on Android
4054 - mesa: introduce boolean toggle for EXT_texture_norm16
4055 - i965: toggle on EXT_texture_norm16
4056 - mesa/st: toggle EXT_texture_norm16 based on format support
4057 - mesa/st: fix formats required for EXT_texture_norm16
4058 - nir: fix compilation warning on glsl_get_internal_ifc_packing
4059 - iris: toggle on PIPE_CAP_MIXED_COLOR_DEPTH_BITS
4060 - nir/glsl: gather bitmask of images used by program
4061 - iris: use the images_used mask in resolve pass
4062 - intel/compiler: detect if atomic load store operations are used
4063 - iris: provide dummy iris_image_view_aux_usage
4064 - iris: move existing image format fallback as a helper function
4065 - iris: determine aux usage during predraw and state setup
4066 - isl: allow compression for storage images on gen12+
4067 - iris: allow compression conditionally for images on gen12
4068 - glsl: set error_emitted true if type not ok for assignment
4069 - mesa/st: unbind shader state before deleting it
4070 - mesa/st: release variants for active programs before unref
4071 - mesa: remove redudant check
4072 - mesa: remove redudant assignment
4073 - glsl: remove redudant assignment
4074 - glsl: stop processing function parameters if error happened
4075 - mesa/st: initialize all winsys_handle fields for memory objects
4076 - anv: remove assert from GetImageMemoryRequirements[2]
4077 - st/mesa: destroy only own program variants when program is released
4081 - svga: Fix banded DMA upload
4082 - svga, winsys/svga: Fix persistent memory discard maps
4083 - svga: Treat forced coherent maps as maps of persistent memory
4084 - gallium/pipebuffer: Use persistent maps for slabs
4085 - winsys/svga: Optionally avoid caching buffer maps
4089 - Revert "st/va: Convert interlaced NV12 to progressive"
4090 - gallium/auxiliary/vl: fix bob compute shaders for deint yuv
4091 - st/va: remove unneeded code
4092 - st/va/postproc: reallocate interlaced destination buffer
4093 - radeonsi: add 10-bit HEVC encode support for VCN2.0 devices
4094 - radeon: add support for 10-bit HEVC encoding to VCN 2.0
4095 - st/va: add check for P010 and P016 encode/decode support
4099 - glsl: fix gl_nir_set_uniform_initializers() for image arrays
4100 - glsl: fix possible memory leak in nir uniform linker
4101 - glsl: set the correct number of samplers in a shader
4102 - glsl: set the correct number of images in a shader
4103 - glsl: fix resizing of the uniform remap table
4104 - glsl: reset next_image_index count for each shader stage
4105 - glsl: fix sampler index calculation in nir linker
4106 - glsl: add some error checks to the nir uniform linker
4107 - glsl: move nir link uniforms struct defs earlier
4108 - glsl: move add_parameter() earlier in nir link uniforms
4109 - glsl: move get_next_index() earlier in nir link uniforms
4110 - glsl: add name support to nir uniform linker
4111 - glsl: correctly find block index when linking glsl with nir linker
4112 - nir: add glsl_get_internal_ifc_packing() helper
4113 - nir: add glsl_get_std140_base_alignment() helper
4114 - nir: add glsl_get_std140_size() helper
4115 - nir: add glsl_get_std430_base_alignment() helper
4116 - nir: add glsl_get_std430_size() helper
4117 - glsl: add std140 and std430 layouts to nir uniform linker
4118 - glsl: correctly set explicit offsets for struct members
4119 - glsl: find the base offset for block members from unnamed blocks
4120 - glsl: nir linker fix setting of ssbo top level array
4121 - glsl: set ShaderStorageBlocksWriteAccess in the nir linker
4122 - glsl: add support for builtins to the nir uniform linker
4123 - glsl: dont try to assign uniform storage for uniform blocks
4124 - glsl: add subroutine support to nir linker
4125 - glsl: fix varying packing for 64bit integers
4126 - nir: fix packing of TCS varyings not read by the TES
4127 - nir: fix crash in varying packing on interface mismatch
4128 - glsl_to_nir: remove dead code
4129 - radeonsi: don't lower constant arrays to uniforms in GLSL IR
4130 - nir: make opt_if_loop_terminator() less strict
4131 - nir: add matrix_layout to nir_variable data
4132 - glsl: fix struct offsets in the nir uniform linker
4133 - glsl: tidy up uniform storage value count code in NIR linker
4134 - Revert "glsl: fix resizing of the uniform remap table"
4135 - glsl: fix explicit locations for the glsl linker
4136 - glsl: error check max user assignable uniform locations
4137 - glsl: fix block index in NIR uniform linker
4138 - glsl: pull mark_array_elements_referenced() out into common helper
4139 - glsl: only set stage ref when uniforms referenced in stage
4140 - nir/gcm: allow derivative dependent intrinisics to be moved earlier
4141 - nir/gcm: be more conservative about moving instructions from loops
4142 - nir/gcm: dont move movs unless we can replace them later with their
4144 - glsl: add bindless support to nir uniform linker
4145 - glsl: fix gl_nir_set_uniform_initializers() for bindless textures
4146 - st/glsl_to_nir: make use of nir linker for linking uniforms
4147 - glsl: some nir uniform linker fixes
4148 - glsl: remove some duplicate code from the nir uniform linker
4149 - glsl: stop cascading errors if process_parameters() fails
4150 - glsl: fix slow linking of uniforms in the nir linker
4154 - aco/optimizer: Don't combine uniform bool s_and to s_andn2.
4155 - radv: Move some helper functions to the radv_shader.h header file.
4156 - aco: Extract setup_gs_variables into a separate function.
4157 - aco: Setup tessellation control shader variables.
4158 - aco: Implement load_tess_coord.
4159 - aco: Implement load_primitive_id for tessellation shaders.
4160 - aco: Implement load_patch_vertices_in.
4161 - aco: Implement load_invocation_id for tessellation control shaders.
4162 - aco: Implement control_barrier for tessellation control shaders.
4163 - aco: Implement memory_barrier_tcs_patch.
4164 - aco: Implement load_view_index for TCS and TES.
4165 - aco: Setup correct HW stages when tessellation is used.
4166 - aco: Use mesa shader stage when loading inputs.
4167 - aco: Remove vertex_geometry_gs assertion from merged shaders.
4168 - aco: Extract LDS alignment calculation to a separate function.
4169 - aco: Remove esgs_itemsize from LDS alignment calculation.
4170 - aco: Introduce new VMEM load/store helpers.
4171 - aco: Introduce new helpers for calculating address offsets.
4172 - aco: Refactor load_per_vertex_input in preparation for tessellation.
4173 - aco: Refactor VS output stores in preparation for tessellation.
4174 - aco: Slight fix to lds_store and lds_load.
4175 - aco: Fix combining DS additions in the optimizer.
4176 - aco: Implement tessellation control shader input/output.
4177 - aco: Store VS outputs correctly when tessellation is used.
4178 - aco: Fix LS VGPR init bug on affected hardware.
4179 - radv: Enable ACO for tessellation control shaders.
4180 - aco: Setup tessellation evaluation shader variables.
4181 - aco: Use TES output info when TES runs on the VS stage.
4182 - aco: Store TES outputs when TES runs on the HW VS stage.
4183 - aco: Enable streamout when TES runs on the HW VS stage.
4184 - aco: Implement loading TES inputs.
4185 - radv: Enable ACO for TES when there is no GS.
4186 - aco: Enable running TES as ES, including merged TES+GS.
4187 - radv: Enable ACO on all stages.
4188 - aco: Don't generate an if when the first part of a merged HS or GS is
4190 - aco: Store tess factors in VMEM only at the end of the shader.
4191 - aco: Only write TCS outputs to LDS when they are read by the TCS.
4192 - aco: Don't store TCS outputs to LDS when we're sure that none are
4194 - nir: Add ability to lower non-const quad broadcasts to const ones.
4195 - radv: Enable lowering dynamic quad broadcasts.
4196 - radv: Enable subgroup shuffle on GFX10 when ACO is used.
4197 - aco: Create null exports in instruction selection instead of
4199 - aco: Extract tcs_driver_location_matches_api_mask to separate
4201 - aco: Fix handling of tess factors.
4202 - aco: Allow combining TCS output VMEM stores.
4203 - aco: Allow combining LDS loads when loading tess factors.
4204 - aco: Skip 2nd read of merged wave info when TCS in/out vertices are
4206 - aco: Use more optimal sequence at the beginning of merged shaders.
4207 - nir: Collect if shader uses cross-invocation or indirect I/O.
4208 - aco: Treat outputs of the previous stage as inputs of the next stage.
4209 - aco: Change isel inputs/outputs to a flat array.
4210 - aco: Zero-fill undefined elements in create_vec_from_array.
4211 - aco: Extract setup_tcs_info to a separate function.
4212 - aco: Fix workgroup size calculation.
4213 - aco: Extract store_output_to_temps into a separate function.
4214 - aco: When LS and HS invocations are the same, pass LS outputs in
4216 - aco: Don't store LS VS outputs to LDS when TCS doesn't need them.
4217 - aco: Fix crash in insert_wait_states.
4218 - aco: Extract uniform if handling to separate functions.
4219 - aco: Print block_kind_export_end.
4220 - aco: Extract merged_wave_info_to_mask to its own function.
4221 - aco: Treat s_setprio as a scheduling barrier.
4222 - aco/ngg: Add new stage for hw_ngg_gs.
4223 - aco/ngg: Initialize exec mask for NGG VS and TES.
4224 - aco/ngg: Fix exports for NGG VS and TES.
4225 - aco/ngg: Setup NGG VS and TES stages.
4226 - aco/ngg: Implement NGG VS and TES.
4227 - aco/ngg: Schedule position exports of NGG VS/TES.
4228 - aco/ngg: Run GS_ALLOC_REQ on priority 3 for NGG VS and TES.
4229 - radv: Enable ACO for NGG VS/TES, but disable NGG for ACO GS.
4230 - aco: Print shader stage in aco_print_program.
4231 - radv: Print shader stage before disassembly.
4232 - radv: Add inputs read by TES to radv_shader_info.
4233 - aco: Only store TCS outputs to VMEM when they are read by TES.
4234 - aco: Increase barrier_count to 7 to include barrier_barrier.
4235 - aco: Abort when RA can't find a register.
4236 - aco: Const correctness for get_barrier_interaction.
4237 - aco: Const correctness for aco_print_ir.
4238 - aco: Use 24-bit multiplication in TCS I/O
4239 - aco: Use 24-bit multiplication for NGG wave id and thread id.
4240 - aco: Move s_setprio to correct place after the gs_alloc_req.
4241 - radv: Refactor calculate_tess_lds_size and get_tcs_num_patches.
4242 - aco: Use context variables instead of calculating TCS inputs/outputs.
4243 - aco: Remember VS/TCS output driver locations.
4244 - aco: Calculate workgroup size of legacy GS.
4245 - aco: Set config->lds_size when TES or VS is running on HW ESGS.
4246 - nir: Add new linking helper to set linked driver locations.
4247 - radv: Use new linking helper to set default driver locations.
4248 - aco: Use new default driver locations.
4249 - radv: Use smaller esgs_itemsize for ACO.
4253 - meson: Link Gallium Nine with ld_args_build_id
4257 - gallium/swr: spin-lock performance improvement
4261 - panfrost: Print intended field when decoding
4262 - panfrost: Add more info to some assertions
4263 - pan/midgard: Handle nir_intrinsic_load_barycentric_centroid
4264 - panfrost: Use DBG macro to avoid noise in the console
4265 - panfrost: Fix decoding of tiled 3D textures
4266 - panfrost: Only clamp the LOD to disable mipmapping when needed
4267 - gitlab-ci: Switch kernel for LAVA jobs to 5.5
4268 - gitlab-ci: Disable the lima job for now
4269 - gitlab-ci: Run GLES3 tests in dEQP on Panfrost
4270 - panfrost: Remove some more prints to stdout
4271 - gitlab-ci: Move to 5.5 kernel plus fixes for Panfrost
4272 - gitlab-ci: Use PAN_MESA_DEBUG=gles3 for Panfrost
4273 - gitlab-ci: Remove GLES3 test from Panfrost fails list
4274 - gitlab-ci: Skip dEQP-GLES3.functional.shaders.derivate.\*
4275 - gallium: Add forgotten docs for new CAPs related to transform
4277 - gitlab-ci: Update renderdoc
4278 - gitlab-ci: Use surfaceless platform also for apitrace
4279 - gitlab-ci: Place files from the Mesa repo into the build tarball
4280 - gitlab-ci: Serve files for LAVA via separate service
4281 - gitlab-ci: Disable jobs for Collabora's LAVA lab
4282 - Revert "gitlab-ci: Disable jobs for Collabora's LAVA lab"
4283 - panfrost: Remove most usage of midgard_payload_vertex_tiler
4284 - panfrost: Pass IS_BIFROST to pandecode_jc
4285 - panfrost: Don't emit write_value jobs on Bifrost
4286 - panfrost: On Bifrost, set the right tiler descriptor
4287 - gitlab-ci: Test virgl driver
4288 - panfrost: Clean up a bit the tiler structs for Bifrost
4289 - panfrost: Emit sampler descriptor on bifrost
4290 - panfrost: Emit texture descriptor on bifrost
4291 - gitlab-ci: Update virglrenderer in the x86_test-gl image
4292 - gitlab-ci: Allow test jobs to add options to the dEQP invocation
4293 - gitlab-ci: Test OpenGL ES 3.1 on virgl
4294 - gitlab-ci: Test Virgl with traces
4295 - panfrost: Add Bifrost texture trampoline BO to batch
4299 - doc: Update features.txt for r600 with misc supported features
4303 - lima: handle early-z and pixel kill better
4304 - lima: implement PLB PP stream cache
4305 - lima: add RGBA5551 and RGBA4444 formats
4306 - lima: don't disable tiling if there's linear modifier in list
4307 - lima: gpir: enforce instruction limit earlier
4308 - panfrost: split index cache into shared part
4309 - lima: enable minmax cache for index buffers
4310 - lima: print gp uniforms if gp debug is enabled
4311 - lima/gpir: improve disassembler output
4312 - lima/gpir: print acc ops even if we have only one source
4313 - lima/gpir: kill dead writes to regs in DCE
4314 - lima/gpir: add better lowering for ftrunc
4315 - lima/gpir: fix crash in schedule_insert_ready_list()
4316 - lima: disable Z16 format
4317 - lima: decode depth/stencil write bits in RSW
4318 - lima: split pixel and texel format tables
4319 - lima: add support for R and RG formats
4320 - lima: Implement lima_texture_subdata
4321 - lima: avoid situations when scissor minx > maxx or miny > maxy
4325 - radeon/vce: Move global function pointer si_get_pic_param to local
4331 - zink. Don't set incorrect sType in VkImportMemoryFdInfoKHR struct
4335 - swr: Fix build with GCC 10.
4336 - lima: Fix build with GCC 10.
4337 - swr: Fix GCC 4.9 checks.
4338 - panfrost: Remove unused anonymous enum variables.
4339 - meson: Enable -Wno-deprecated only for bison > 2.3.
4340 - swr: Fix non-pod-varargs error.
4341 - st/nine: Fix incompatible-pointer-types-discards-qualifiers errors.
4342 - panfrost: Fix gnu-empty-initializer error.
4343 - util/u_process: Add util_get_process_exec_path for macOS.
4344 - mesa: Change \_mesa_exec_malloc argument type.
4345 - gallivm: Add missing header for powf.
4346 - swr/rasterizer: Use private functions for min/max to avoid namespace
4348 - swr: Remove Byte Order Mark.
4349 - r600/sfn: Initialize VertexStageExportForGS m_num_clip_dist member
4351 - r600/sfn: Use correct setter method.
4352 - freedreno: Add missing va_end.
4356 - intel/compiler: fix cmod propagation optimisations
4360 - radeonsi: Add support for midstream bitrate change in encoder
4364 - zink: confused compilation macro usage for zink in target helpers.