Lines Matching full:radv
33 - VK_EXT_attachment_feedback_loop_layout on RADV, lavapipe
34 - VK_KHR_global_priority on RADV
36 - VK_EXT_load_store_op_none on RADV
37 - VK_EXT_mutable_descriptor_type on RADV
43 - VK_EXT_extended_dynamic_state3 on RADV
47 - support for R8G8B8, B8G8R8, R16G16B16 and 64-bit vertex buffer formats on RADV
48 - initial GFX11/RDNA3 support on RADV
49 - various ray tracing optimizations on RADV
50 - extendedDynamicState2PatchControlPoints on RADV (VK_EXT_extended_dynamic_state2 feature)
73 - radv, dxvk: Rendering errors in World of Tanks after "Switch to dynamic rendering only"
80 - radv: Artifacts in multiple games with yuzu (regression, bisected)
82 - radv: Flickering in Spider-Man Remastered (Regression) (Bisected)
83 - radv: Hitman 2 using Direct3D 12 has discolored squares on RDNA2 with DCC enabled
89 - [radv/navi21] No sprites being rendered in Quake 2 RTX
99 - radv: incorrect query result for VK_QUERY_TYPE_PRIMITIVES_GENERATED_EXT
110 - radv: Crash in dEQP-VK.ray_query.misc.dynamic_indexing
117 - radv: dynamic patch control points broken
131 - RADV: Mafia & Mafia III :Definitive Edition graphical issues with certain effects
132 - Cyberpunk 2077 ground reflectiveness issues using RADV
133 - [RADV] Grand Theft Auto V: Shadow rendering bug on Steam Deck
141 - radeonsi/radv: Discrepancy in handling wd_switch_on_eop
143 - RADV: Regression with RPCS3 where nothing renders
144 - [RADV][bisected] commit 3714c89d0ebf0c541d8f7a7a807ce7c1bfe3cfea at least lead to GPU hang when I…
151 - RADV: VK_EXT_load_store_op_none support?
152 - radv: extendedDynamicState2PatchControlPoints
159 - radv: Enable instruction timing with RGP for RT
169 - ci: radeonsi-raven-skqp also tests RADV but only runs on radeonsi changes
175 - radv: missing support for some 3-component vertex formats in RDNA2
176 - radv: radv_translate_buffer_dataformat overhead
185 - RADV: vulkaninfo fails to report extended VkShaderStageFlagBits for VK_KHR_ray_tracing_pipeline
189 - [RADV] (Vega/Navi) Halo: The Master Chief Collection: Texture issue on Halo 3 and Reach
212 - radv regression: Hitman 2 rendering issue
216 - [Bisected][RADV] Rendering issues and performance drop with GFX10 on RPCS3
735 - radv: Remove redundant radv_QueuePresentKHR.
736 - radv: Add binding code for 3d sparse images.
737 - radv: Add 3d tile shapes for sparse binding.
738 - radv: Expose 3d sparse images.
741 - radv: Don't flatten bottom AS exit if statement.
742 - radv: Store top of stack in a register.
743 - Revert "radv/rt: use derefs for the traversal stack"
744 - radv: Use constant for ray traversal exit condition.
745 - radv: Use nested ifs for pushing child nodes in traversal loop.
746 - radv: Add scratch stack to reduce LDS stack in RT traversal.
748 - radv: Hardcode root node id.
749 - radv: Translate the BVH copy shader to glsl from nir_builder.
750 - radv: Remove aabb bounds from instance nodes.
751 - radv: Use proper matrices for instance nodes.
752 - radv: Use GLSL matrices for instance transforms in BVH.
753 - radv: Use deterministic order for dumping acceleration stuctures.
754 - radv: Properly initialize all memory in RRA dumps.
755 - radv: Do an early check of distance in triangle culling.
757 - radv: Clean up unused fields in BVH IR.
758 - radv: Add a field for the offset of the bvh in the blas.
759 - radv: Simplify buffer copy address generation.
760 - radv: Start the BVH after the header.
761 - radv: Generate parent links in BVH.
762 - radv: Only emit parents from parents that actually end up in the tree.
763 - radv: Split global & local bvh node variable.
764 - radv: Add traversal backtracking with a short stack.
765 - radv: Properly annotate all the invalid node usage.
766 - radv: Implement pipeline-wide skiptriangles/skipaabbs.
767 - radv: Consider inactive internal nodes never in the tree.
768 - radv: Constify entry_size.
769 - radv: Use a VkPipelineCache handle for meta shaders.
770 - radv: Rework modification detection of the meta cache.
771 - radv: Mark dEQP-VK.ray_query.misc.dynamic_indexing as crashing in CI.
773 - radv: Improve gfx11 clear colors for DCC.
774 - radv: Use correct primgrp size for gfx11.
775 - radv: Add BVH IR header.
776 - radv: Make the number of internal nodes be written on the GPU.
777 - radv: Add PLOC shader
778 - radv: Use PLOC for BVH building
779 - radv: Speculatively tune RT pipelines for GFX11.
780 - radv: Use correct types for loading the rings with LLVM.
781 - radv: Handle GSVS ring intrinsic correctly with LLVM.
782 - radv: Handle attribute ring intrinsic correctly with LLVM.
783 - radv: Fix compute scratch buffer emission.
784 - radv: Make the compute scratch waves per SE as well.
785 - radv: Fix custom build id with C90 compilation.
786 - radv: Use correct init order for ETC2 image views
787 - radv: Fix sampler types in ETC2 decode.
788 - vulkan: Remove asserts that weren't valid for RADV ETC2 emulation.
789 - radv: reserve space for the scissor in vkCmdBeginRendering.
1184 - radv: use nir_opt_uniform_access
1212 - radv/rt: create separate radv_rt_pipeline struct
1213 - radv/rt: inline radv_rt_pipeline_create_() helper into radv_rt_pipeline_create()
1223 - radv/rt: create traversal shader independent from main shader
1224 - radv/rt: overwrite hit args with undef in case of a miss
1301 - radv: move to using common record_result.
1650 - docs/features: add missing extensions supported by radv
1982 - radv/winsys: Return VK_ERROR_INVALID_EXTERNAL_HANDLE if buffer imports fail
1983 - radv: Generalize instance count in BVH states to leaf node count
1984 - radv: Add metadata to acceleration structures
1985 - radv: Add Radeon Raytracing Analyzer trace dumping utilities
1986 - radv: Add Radeon Raytracing Analyzer capturing layer
1987 - radv: Enable Radeon Raytracing Analyzer traces
1988 - docs: Add documentation about RADV RRA tracing
1989 - radv: Use radv_CmdUpdateBuffer instead of radv_update_buffer_cp for geometry info writes
1990 - radv: Free geometry infos in radv_CmdBuildAccelerationStructuresKHR
1991 - radv: Correct accel struct header size
1992 - radv: Rename internal node shader to lbvh_internal
1993 - radv: Split CmdBuildAccelerationStructuresKHR into different functions
1994 - radv: Add BVH IR types
1995 - radv: Add conversion shader for internal nodes
1996 - radv: Add conversion shader for leaf nodes
1997 - radv: Create pipelines for conversion meta shaders
1998 - radv: Build acceleration structures using BVH IR
1999 - radv: Rename emulated float helpers
2000 - radv: Use a struct for AABBs
2001 - radv: Add radv_indirect_unaligned_dispatch
2002 - radv/rt: Fix internal converter synchronization
2003 - radv/rt: Dispatch internal converter indirectly
2004 - radv/rt: Track number of inactive leaf nodes
2005 - radv: Add global sync utilities
2006 - radv: Add REF as a typename macro to .clang-format
2007 - radv: Use spirv1.5 instead of vulkan1.2
2021 - radv: Fold 16bit image sources.
2029 - radv: Enable VK_EXT_load_store_op_none.
2032 - radv: Fix GLSL BDA struct alignment and use pointer arithmetic SIZEOF.
2040 - radv,aco: Lower uclz in NIR.
2047 - radv: Use available dot product instructions on gfx11.
2196 - radv: Implement VK_EXT_mutable_descriptor_type.
2303 - radv: anv: Use the new vk_clock_gettime and vk_time_max_deviation functions
2378 - radv: Use both aspects for depth/stencil blit destinations
2379 - radv: Use VK_IMAGE_LAYOUT_ATTACHMENT_FEEDBACK_LOOP_OPTIMAL_EXT for render loops
2380 - radv/meta: Respect render area for MSAA resolves
2381 - radv/meta: Fix a stack use-after-free with sample locations
2382 - radv/meta: Drop subpass_att from radv_fast_clear_color
2383 - radv/meta: Use Begin/EndRendering for resolves
2384 - radv/meta: s/dest/dst/ in a few functions
2385 - radv/meta: Use Begin/EndRendering for clears
2386 - radv: Drop set/restore_subpass
2387 - radv/meta: Drop redundant depth_view_can_fast_clear
2388 - radv: Add a image_transition helper for separate depth/stencil layouts
2389 - radv: Don't assume layout == stencilLayout for stencil-only
2411 - radv: Move to the common command pool framework
2416 - radv: Set the window scissor to the render area, not framebuffer
2417 - radv: Only copy the render area from VRS to HTILE
2418 - radv: Leave image layouts alone when doing HW MSAA resolves
2419 - radv: Switch to dynamic rendering only
2586 - radv: Add get_addrlib function to radv_radeon_winsys
2588 - radv: Only apply mipmap view adjustments to block compressed images
2589 - radv: Fix corrupted mipmap copies on GFX9+
2590 - radv: Fix mipmap views on GFX10+
2630 - radv: Remove workarounds for D3D9 feedback loops
3013 - radv/nir_lower_abi: Use instructions_pass
3014 - radv: Move accel struct structs to bvh/bvh.h
3015 - radv: Move radv_acceleration_structure
3016 - radv: Remove acceleration structure host builds
3017 - radv: Always emulate fmin/fmax
3018 - radv: Remove accel_struct_build
3019 - radv: Remove unused push constant structs
3020 - radv: Add create_build_pipeline_spv helper
3021 - radv: Add the basics for GLSL bvh kerrnels
3022 - radv: Add a GLSL morton kernel implementation
3023 - radv: Switch to the GLSL morton implementation
3024 - radv: Add a GLSL internal kernel implementation
3025 - radv: Switch to the GLSL internal implementation
3026 - radv: Add a GLSL leaf kernel implementation
3027 - radv: Switch to the GLSL leaf implementation
3029 - radv: Fix stack size calculation with stage ids
3030 - radv: Inline bvh build headers
3031 - radv/ci: Remove host build related fails
3033 - radv: Advertise subgroup ops for rt stages
3035 - radv: Use the common physical device enumeration
3041 - radv: Proper handling for inactive instance nodes
3042 - radv: Deduplicate push constant structs
3043 - radv/rra: Map accel struct VAs to handles
3044 - radv/rra: Remove redundant bounds validation
3045 - radv: Make the radv_buffer_get_va parameter const
3046 - radv/rra: Replace aliasing assert with a warning
3047 - radv: Explicitly store the VA of accel structs
3048 - radv: Cleanup radv_GetInstanceProcAddr
3049 - radv: Add a common traversal build helper
3050 - radv/rq: Use the common traversal helper
3051 - radv/rtpso: Use the common traversal helper
3052 - radv: Use scalar layout for BDA references
3053 - radv: Add and use AS and scratch layout structs
3054 - radv: Fix incorrect build info indexing
3055 - radv: Use cache_uuid for accel struct compatibility
3056 - radv: Remove main_loop_case_visited
3057 - radv/rra: Transcode nodes recursively
3058 - radv: Rename node_internal to node_box32
3059 - radv: Add radv_bvh_node_box16
3060 - radv: Use half floats for box16 coords
3061 - radv/rra: Handle box16 nodes
3062 - radv/rra: Calculate bvh size requirements recursively
3063 - radv: Remove create_accel_build_shader
3064 - radv/rra: Fix dumps in the case of aliasing
3065 - radv/rra: Validate before gathering bvh info
3066 - radv/rra: Add basic header validation
3067 - radv/rra: Continue dumping accel structs if validation fails
3068 - radv/rra: Use the accel struct type for header validation
3069 - radv/rt: Fix setting tmax for opaque AABBs
3070 - radv/rt: Load instance id and custom index on demand
3071 - radv/rt: Set vars.arg in the traversal shader
3072 - radv/rt: Restore prev barycentrics when rejecting hits
3073 - radv/rra: Fix copying accel structs that were not built yet
3074 - radv/ray_queries: Fix AABB handling
3075 - radv/rt: Check space before emitting descriptors
3229 - radv: tweak lower_shader_calls parameters
3356 - nir, anv, hasvk, radv: pull uses_wide_subgroup_intrinsics into shader_info
3364 - radv: use nir_shader_instructions_pass in radv_nir_lower_ycbcr_textures
3404 - radeonsi,radv: run ac_nir_lower_resinfo
3566 - radv/ci: document a recent regression
3567 - radv/ci: move some tests from the vega10 fail to its flake list
3568 - radv/ci: add more tests to the flake list of vega10
3569 - radv/ci: skip host_write_transfer_src.1048576 on renoir
3570 - radv/ci: bump the console activity timeout of VanGogh to 3 minutes
3571 - radv/ci: uprev boot2container
3572 - radv/ci: disable the POLARIS10 jobs
3573 - radv/ci: disable kabini's vkd3d job
3574 - radv/ci: run vkcts on the two steam decks in parallel
3575 - radv/ci: document an unstable test
3576 - radv/ci: move some tests from the renoir fail to its flake list
3578 - radv/ci: bump the job count on navi21
3579 - radv/ci: run two jobs in parallel to reduce the execution time
3580 - zink: update the RADV expectations
3581 - zink: mark a test as a flake in RADV expectations
3583 - zink/ci: add another subtest to the list of known failures for radv
3694 - radv: fix return type for meta resolve shaders
3697 - radv: remove stupid conditional from draw path
3698 - radv: force inlining on another draw function
3868 - radv: avoid bottlenecking on sequential sparse buffer binds
3924 - radv: ALWAYS_INLINE radv_after_draw
3925 - radv: ALWAYS_INLINE radv_flush_descriptors
3926 - radv: ALWAYS_INLINE radv_is_streamout_enabled()
3927 - radv: store pointer to sgprs for last vertex stage
3928 - radv: use direct access to last_vgt_api_stage_locs for sgpr emission
4218 - radv: deal with ac_nir_translate failures
4246 - radv/llvm: use ac_build_gep0_type to get args types
4252 - radv/llvm: replace LLVMBuildLoad by LLVMBuildLoad2
4283 - ac/nir/ngg,radv: use nir_load_viewport_xy_scale_and_offset
4315 - nir,ac/llvm,radv: add nir_intrinsic_load_provoking_vtx_in_prim_amd
4317 - radv: count gen_prims_queries_enabled
4318 - radv: split active_pipeline_gds_queries
4319 - nir,ac/nir/ngg,radv: split shader_query_enabled_amd
4320 - ac/nir/ngg,radv: use different counters for shader queries
4321 - nir,ac,radv: add primitive count add intrinsics
4331 - ac/nir/ngg,radv: move LDS layout calculation out of nir ngg lowering
4403 - docs/radv: advertise VK_KHR_global_priority
4413 - radv: remove claimed support for sRGB vertex buffer formats
4414 - radv: fix 16-bit support in radv_lower_vs_input
4416 - radv/llvm: fix packed VS inputs on GFX6/GFX10+
4422 - radv: enable ac_nir_lower_resinfo for ACO
4432 - radv: add radv_is_vertex_buffer_format_supported
4433 - radv,aco: use pipe_format for static vertex input state
4434 - radv,aco: use pipe_format for dynamic vertex input state
4435 - radv: remove radv_translate_vertex_format
4436 - radv: enable R8G8B8, B8G8R8 and R16G16B16 vertex formats
4437 - radv,aco: implement 64-bit vertex inputs
4438 - radv: enable 64-bit vertex formats
4444 - radv: disable EXT_vertex_input_dynamic_state when using DGC
4445 - radv: shrink zero-initialization in vkCmdSetVertexInputEXT
4446 - radv: use nir_ubfe_imm
4447 - radv: remove unnecessary .align_mul=4
4448 - radv: fix dynamic RT stack size with VGPR spilling
4489 - radv,nir: add intrinsics for streamout and GS copy shaders
4495 - nir,ac/nir,aco,radv: replace has_input_*_amd with more general intrinsics
4498 - radv/gfx11: don't create fmask copy/expand pipelines
4499 - radv/gfx11: don't create texop_samples_identical
4500 - radv,aco: don't use lower_to_fragment_fetch_amd on GFX11+
4504 - radv/llvm: fix GS shaders on GFX8/9
4505 - docs: update new_features.txt for non-extension additions to RADV
4506 - radv: increase gfx1100/gfx1101 physical vgprs
4511 - radv: lower 8/16-bit uadd_carry/usub_borrow
4512 - radv: disable MRT compaction with dual-source blending
4712 - radv: implement VK_EXT_attachment_feedback_loop_layout
4713 - radv: ignore out-of-order rasterization if stencil write mask is dynamic
4714 - radv: rework shaders ref counting
4715 - radv: use ref counting for VS prologs and PS epilogs
4716 - radv: remove RADV_META_SAVE_SAMPLE_LOCATIONS
4717 - radv: simplify saving/restoring all dynamic states
4718 - radv: remove unused states parameter from some radv_emit_XXX() helpers
4719 - radv: simplify radv_bind_dynamic_state() slightly
4720 - radv: fix gathering XFB info if there is dead outputs
4721 - radv: respect the render area for depth/stencil resolves
4722 - radv: fix cleaning the meta query state if an error occured
4723 - radv: properly ignore PSO states that are dynamic
4724 - radv: remove one useless check about line width and dynamic state
4725 - radv: simplify dynamic states uses when determining NGG culling settings
4726 - radv: do not set registers for PSO states that are dynamic
4727 - radv: remove radv_invariant_geom workarounds for DX games
4728 - radv: explain why radv_invariant_geom is still needed for SotTR DX
4729 - radv: remove unecessary radv_pipeline::uses_dynamic_stride
4730 - radv: copy line stipple factor/pattern only if the state isn't dynamic
4731 - radv: copy the sample locations only if the state isn't dynamic
4736 - radv: store the provoking vertex mode in the graphics pipeline info
4737 - radv: simplify the depth clip enable/disable logic
4738 - radv: do not translate polygon mode during gfx info initialization
4739 - radv: do not translate logic op during gfx info initialization
4740 - radv: do not translate blend op/factor during gfx info initialization
4741 - radv: do not translate primitive topology during gfx info initialization
4742 - radv: slightly change the color write enable mask
4743 - radv: call vk_graphics_pipeline_state_fill() when initializing gfx pipelines
4744 - radv: convert radv_tessellation_info to vk_tessellation_state
4745 - radv: convert radv_viewport_info to vk_viewpoert_state
4746 - radv: convert radv_discard_rectangle_info to vk_discard_rectangles_state
4747 - radv: convert radv_rasterization_info to vk_rasterization_state
4748 - radv: convert radv_depth_stencil_info to vk_depth_stencil_state
4749 - radv: convert radv_fragment_shading_rate_info to vk_fragment_shading_rate_state
4750 - radv: convert radv_input_assembly_info to vk_input_assembly_state
4751 - radv: convert radv_multisample_info to vk_multisample_state
4752 - radv: convert radv_color_blend_info to vk_color_blend_state
4753 - radv: convert radv_rendering_info to vk_render_pass_state
4754 - radv: convert radv_vertex_input_info to vk_vertex_input_state
4755 - radv: remove radv_graphics_pipeline_info completely
4756 - radv: advertise VK_KHR_global_priority
4757 - radv: remove unused pCreateInfo from radv_pipeline_init_blend_state()
4758 - radv: add radv_graphics_pipeline_import_info() helper
4759 - radv: introduce a new pipeline type for graphics libs
4760 - radv: store vk_graphics_pipeline_state to radv_graphics_pipeline
4761 - radv: prevent accessing rasterization state when it's NULL
4762 - radv: add a very initial implementation of VK_EXT_graphics_pipeline_library
4763 - radv: do not compress DCC in presence of render loops on GFX10+
4764 - radv: remove useless check when creating the noop FS
4765 - radv: copy linked I/O info for merged stages in radv_fill_shader_info
4766 - radv: copy the whole radv_shader_info struct for merged stages
4767 - radv: add a radv_postprocess_nir() helper
4768 - radv: move lowering FS intrinsics to radv_postprocess_nir()
4769 - radv: move dumping NIR shaders to the previous similar loop
4770 - radv: remove unused parameter in radv_lower_io_to_mem()
4771 - radv: use NIR_PASS with radv_export_implicit_primitive_id
4772 - radv: remove useless parameter from radv_link_shaders()
4773 - radv: add radv_remove_point_size() linking helper
4774 - radv: rework linking shaders in NIR
4775 - radv: move assigning driver locations to the new linking helpers
4776 - radv: move lowering patch vertices to radv_pipeline_link_tcs()
4777 - radv: move lowering the view index to radv_pipeline_link_shaders()
4778 - radv: remove redundant assignment of tcs.tcs_vertices_out
4779 - radv: remove unused tcs_vertices_out assignment for VS
4780 - radv: remove unused num_tess_patches assignment for VS
4781 - radv: merge gather_tess_info() with radv_fill_shader_info()
4782 - radv: stop setting CB_COLOR_CONTROL.ROP3 from the pipeline
4783 - radv: stop clearing bitfields for registers that are emitted dynamically
4784 - radv: cleanup dynamic states in radv_emit_graphics_pipeline()
4785 - radv: stop emitting RMW context registers for updating sample locations
4786 - radv: re-emit viewports if negative one to one or depth clamp mode changed
4787 - radv: remove bogus assertion about independent set layouts with GPL
4788 - radv: fix missing initialization of the pipeline layout when creating a lib
4789 - radv: destroy the pipeline layout if creating a library failed
4792 - radv: stop duplicating radv_es_output_info
4793 - radv: use esgs_itemsize when calling ac_nir_lower_es_outputs_to_mem
4794 - radv: compute the ESGS itemsize outside of radv_nir_shader_info_pass()
4795 - radv: remove dead code about task ring when binding a compute pipeline
4796 - radv: replace cs.uses_task_rings by ms.has_task
4797 - radv/llvm: remove unused parameter in handle_vs_outputs_post()
4798 - radv: stop duplicating radv_vs_output_info
4799 - radv: fill radv_vs_output_info unconditionally for vertex related stages
4800 - radv: remove redundant VS output parameter assignments
4801 - radv: add a helper that links shader info between stages
4802 - radv: link primitive ID/clip distance shader info from the new helper
4803 - radv: stop gathering info for FS before other stages
4804 - radv: move filling cs.block_size
4805 - radv: add gather_shader_info_fs() helper
4806 - radv: add gather_shader_info_mesh() helper
4807 - radv: add gather_shader_info_gs() helper
4808 - radv: add gather_shader_info_tes() helper
4809 - radv: add gather_shader_info_tcs() helper
4810 - radv: add gather_shader_info_vs() helper
4811 - radv: add gather_shader_info_cs() helper
4812 - radv: add gather_shader_info_task() helper
4813 - radv: move more CS info to gather_shader_info_cs()
4814 - radv: move more MS info to gather_shader_info_ms()
4815 - radv: move determining NGG shader info to radv_fill_shader_info()
4816 - radv: store binning settings into the physical device
4817 - radv: make sure to emit BREAK_BATCH when color write enable is dynamic
4818 - radv: remove useless parameter in si_write_scissors()
4819 - radv: split emitting guardband into a separate helper
4820 - radv: rework dirtying scissors when the line width changes
4821 - radv: emit the guardband state separately from the scissor state
4822 - radv: allow null descriptor set layouts in pipeline layouts
4823 - radv: determine if the topology uses adjacency from the shaders on GFX9+
4824 - radv: fix bogus assertion with RADV_FORCE_VRS
4826 - radv: stop setting as_ls for the TCS stage
4827 - radv: stop setting as_es for the GS stage
4828 - radv: stop setting num_tess_patches for the GS stage
4829 - radv: stop setting esgs_itemsize to the GS stage
4830 - radv: cleanup helpers that compute NGG info and GS info on GFX9+
4831 - radv: move computing wave_size/ballot_bit_size to the shader info pass
4832 - radv: move computing NGG info and GS info to radv_nir_shader_info_link()
4833 - radv: refactor radv_nir_shader_info_link()
4834 - radv: move determining other NGG settings in radv_link_shaders_info()
4835 - radv: simplify determining if NGG passthrough can be enabled
4836 - radv: cleanup computing the workgroup size for all stages
4837 - radv: add and use a helper that merges shader info for merged stages on GFX9+
4838 - radv: remove unused parameters in radv_fill_shader_info()
4839 - radv: determine as_ls/as_es in radv_link_shaders_info()
4840 - radv: constify radv_shader_info where it's missing
4841 - radeonsi/ci: trigger radeonsi-raven-skqp for any RADV changes
4842 - radv: gather MRTs that are written by the fragment shader
4843 - radv: compact MRTs to save PS export memory space
4844 - radv: only expose sparseResidencyImage3D on GFX9+
4845 - radv: emit SQTT markers for RT related commands
4846 - radv: capture RT pipelines from the SQTT layer
4847 - radv: fix reporting RT shaders in RGP
4848 - radv: do not lower color exports for FS that need an epilog
4849 - radv: add radv_remove_color_exports() helper
4850 - radv: do not try to remove color exports for FS that need an epilog
4851 - radv: keep track of the code size for VS prologs and PS epilogs
4852 - radv: create a PS epilog from a library without the main FS
4853 - radv: add support for emitting and prefetching PS epilogs
4854 - radv: import PS epilog from libraries if present
4855 - radv: fix hw remapping of MRT holes with color attachments without export
4856 - radv: fix pipelineStageCreationFeedbackCount when it's 0
4857 - radv: add a pointer to radv_shader_binary in radv_shader
4858 - radv/ci: re-enable dEQP-VK.renderpass2.depth_stencil_resolve.*_samplemask
4859 - radv: store the VS prologs/PS epilogs VA at upload time
4860 - radv: split upload_shader_part() in two parts
4861 - radv: store the binary to radv_shader_part
4862 - radv: upload the PS epilog in the existing pipeline BO
4863 - radv: reword a comment about dynamic states when rasterization is disabled
4864 - radv: emit the LDS size for TCS directly from the pipeline on GFX9+
4865 - radv: add radv_pipeline_key::dynamic_patch_control_points
4866 - radv: skip computing some tess info when patch control points is dynamic
4867 - radv: set workgroup_size to 256 when patch control points is dynamic
4869 - radv: add shader arguments for dynamic patch control points
4870 - radv: add ABI lowering support for dynamic patch control points
4871 - radv: pass the number of patch control points to si_get_ia_multi_vgt_param()
4872 - radv: move emitting PRIMGROUP_SIZE for <= GFX9 from the cmdbuf
4873 - radv: move emitting GE_CNTL for non-NGG pipelines from the cmdbuf
4874 - radv: implement dynamic patch control points
4875 - radv: advertise extendedDynamicState2PatchControlPoints
4876 - radv: stop setting redundant viewport/scissor for internal operations
4877 - radv: do not bind NULL graphics pipeline when restoring the meta state
4878 - radv: reset the compute pipeline when the saved one was NULL
4879 - radv: stop dirtying the graphics pipeline when restoring it
4880 - radv: stop checking for NULL pipelines in radv_CmdBindPipeline()
4881 - radv/ci: cleanup lists of failures/flakes
4882 - radv: disable VK_EXT_graphics_pipeline_library with LLVM
4883 - radv: rename radv_pipeline_key::vs::dynamic_vs_input to has_prolog
4884 - radv: prepare the VS input state for prologs created with GPL
4885 - radv: bind the VS input state for prologs created with GPL
4886 - radv: enable the VS prologs cache if graphicsPipelineLibrary is enabled
4887 - radv: do not remove PSIZ for VS when the topology is unknown
4888 - Revert "radv: upload the PS epilog in the existing pipeline BO"
4890 - radv: add nir_intrinsic_load_sample_positions_amd in the ABI
4891 - radv: run nir_opt_cse before lowering FS intrinsics
4892 - radv,aco: lower barycentric_at_sample in NIR
4893 - radv: do not link shaders when the next stage is unknown
4894 - radv: allow to build the main FS in a graphics pipeline library
4895 - radv,aco: do not compact MRTs if the pipeline uses a PS epilog
4896 - radv: rename radv_thread_trace_set_pstate() to radv_device_set_pstate()
4897 - radv: acquire pstate on-demand when capturing with RGP
4898 - radv: constify radv_lookup_user_sgpr()
4899 - radv: add radv_pipeline_key::dynamic_rasterization_samples
4900 - radv: declare shader arguments for the number of samples for FS
4902 - radv: lower nir_load_rasterization_samples_amd in ABI
4903 - nir,radv: pass the number of samples to load_sample_positions_amd
4904 - radv: add barycentric_at_sample lowering when the number of samples is dynamic
4905 - radv: emit the rasterization samples through an user SGPR if needed
4906 - radv: make sure to initialize wd_switch_on_eop before checking its value
4907 - radv: pass a VkSampler to write_sampler_descriptor()
4908 - radv: remove unnecessary radv_buffer_view::vk_format
4909 - radv: rework and rename radv_make_buffer_descriptor()
4910 - zink/ci: skip ext_external_objects from Piglit with RADV
4911 - zink/ci: skip arb_shader_clock with RADV
4912 - zink/ci: skip GLX tests with RADV
4913 - zink/ci: update list of expected failures with RADV on NAVI10
4914 - zink/ci: update list of flakes for Piglit with RADV on NAVI10
4915 - radv/ci: add piglit testing with Zink on NAVI10
4916 - radv: do not free the GS copy shader binary if created from a library
4917 - radv: import the GS copy shader from a library if present
4918 - radv: determine the last VGT api stage from the active_stages bitfield
4919 - radv: allow to build the pre-rasterization stages in a library
4920 - radv: remove useless gfx10_ngg_info::enable_vertex_grouping
4921 - radv: determine the last VGT api stage also for GPL
4922 - radv: use the maximum number of vertices per primitives for NGG with GPL
4923 - radv: enable NGG culling unconditionally for GPL but disable it dynamically
4924 - radv: rework multiview lowering in NIR slightly
4925 - radv: fix handling ViewportIndex with GPL
4926 - radv: fix handling primitive ID and clip/cull distances with GPL
4927 - radv: fix handling multiview with GPL
4928 - radv: fix emitting RBPLUS state when MRTs are compacted
4929 - radv: remove NULL checks when binding a graphics pipeline
4930 - radv: fix initializing the attribute mask for VS input state and GPL
4931 - radv: fix initializing the blend state for NULL render pass with GPL
4932 - radv: ignore shader stages that don't need to be imported with GPL
4933 - radv: add a new drirc option to enable a unified heap on APUs
4934 - radv: enable the unified heap on APUs for Red Dead Redemption 2
4935 - radv: fix and rework shaders upload with GPL
4936 - radv: always re-emit patch control points when a pipeline with tess is bound
4937 - radv: move the disk cache to the vk_physical_device struct
4938 - radv: use SHA1_DIGEST_LENGTH in the pipeline cache
4939 - radv: do not create a noop FS when the FS is imported from a library
4940 - radv: discard the PS epilog when the pipeline doesn't use a fragment shader
4942 - radv: set SAMPLE_MASK_TRACKER_WATERMARK to 15 on GFX11
4943 - radv: use correct VGT_TESS_DISTRIBUTION settings on GFX11
4944 - radv: do not use memory for waiting for cache flushes on GFX11
4945 - radv: drop the ES vertex count requirement on GFX11
4946 - radv: add RADEON_FLAG_DISCARDABLE
4947 - radv: create and configure the ring BO for attributes
4948 - radv: declare shader arguments for attributes via memory
4950 - radv: add lowering for nir_intrinsic_load_ring_attr_{offset}_amd
4951 - radv: allocate more space for pipeline statistics query on GFX11
4953 - radv: advertise fragmentShadingRateWithShaderDepthStencilWrites on NAVI23+
4955 - radv: declare new dynamic states
4956 - radv: add support for dynamic polygon mode
4957 - radv: add support for dynamic tessellation domain origin
4958 - radv: add support for dynamic logic op enable
4959 - radv: add support for dynamic line stipple enable
4960 - radv: add support for dynamic alpha to coverage enable
4961 - radv: add support for dynamic sample mask
4962 - radv: add support for dynamic depth clip enable
4963 - radv: add support for dynamic conservative rasterization mode
4964 - radv: add support for dynamic depth clip negative one to one
4965 - radv: add support for dynamic provoking vertex mode
4966 - radv: add support for dynamic depth clamp enable
4967 - radv: advertise VK_EXT_extended_dynamic_state3
4968 - radv: disable VK_AMD_texture_gather_bias_lod on GFX11
4969 - radv: disable AMD_shader_explicit_vertex_parameter on GFX11
4970 - radv/llvm: do not set COMPR for exports on GFX11
4971 - ac/nir,radv/llvm: add support for the attribute ring
4972 - aco,radv/llvm: do not export parameters on GFX11
4974 - radv: disable {shader,sparse}ImageFloat32AtomicMinMax on GFX11
4975 - radv: disable shaderBufferFloat64AtomicMinMax on GFX11
4976 - radv: disable dual source blending in more situations
4982 - radv: fix suspending/resuming pipeline statistics queries with GDS
4983 - radv: fix VRS limit when attachmentFragmentShadingRate is disabled
4984 - radv/llvm: fix dual source blending on GFX11
4985 - radv: move nir_opt_idiv_const/nir_lower_idiv after NGG lowering
4987 - radv: implement transform feedback queries with NGG streamout
4988 - Revert "radv: add a pointer to radv_shader_binary in radv_shader"
4992 - radv/llvm: prevent emitting streamout outputs for NGG
4993 - radv: call nir_io_add_intrinsic_xfb_info() after IO lowering
4994 - radv: lower NIR intrinsics for NGG streamout
4995 - radv: do not unconditionally disable NGG streamout lowering in NIR
4996 - radv: reword an incorrect comment about primitives generated query
4997 - radv: use 32-bit GDS counters for primitives generated query with NGG
4998 - radv: suspend/resume primitives generated query separately for legacy/NGG
4999 - radv: fix primitives generated query with NGG only
5000 - radv: set the correct buffer size for NGG streamout
5001 - radv: enable all possible features with NGG streamout
5002 - radv: allow to enable NGG streamout with RADV_PERFTEST=ngg_streamout
5003 - radv: enable VK_EXT_transform_feedback with NGG streamout on GFX11
5005 - radv: assign IO var locations for VS/TES with NGG streamout
5006 - radv/amdgpu: add a kernel GDS management workaround for -ENOMEM
5007 - radv: use defines instead of magic values for GDS counters offset
5008 - radv: move GDS counters after reserved GDS offsets for streamout
5009 - radv: fix fallback for extreme geometry with tessellation on GFX11
5010 - radv: fix setting MIN_LOD for texture descriptors on GFX11
5011 - radv: make the GDS/GDS OA buffer objects resident
5012 - radv: invalidate L2 instead of only writeback L2 when using DCC stores
5013 - radv: re-emit the guardband state when restoring meta operations
5014 - ac/nir,radv: rework and fix NGG queries enables for VS/TES
5016 - radv: re-emit NGG culling settings when conservative rast mode is dynamic
5017 - radv: enable lowering of subgroup shuffle in NIR on GFX11+
5023 - radv: suspend/resume XFB queries with NGG for meta operations
5025 - radv: stop overallocating LDS for VS/TES when NGG streamout is enabled
5026 - radv: fix computing the pervertex LDS size with NGG streamout
5027 - radv: enable NGG XFB queries only if streamout is enabled
5028 - radv: re-mit streamout buffers to unbind them when NGG streamout is disabled
5029 - radv: fix NGG streamout when it's never enabled in runtime
5030 - radv: fix possible hangs with NGG streamout and secondary cmdbuf
5032 - radv: re-emit dynamic depth clamp enable if depth clip enable changed
5033 - radv: make sure to mark DCC as compressed on GFX11
5034 - radv: introduce RADV_DEBUG=nofmask
5035 - radv: set missing SPI_SHADER_PGM_xxx registers on GFX11
5036 - radv: set INTERPOLATE_COMP_Z to 0 on GFX11
5037 - radv: do not enable DCC for MSAA images without FMASK
5038 - radv: do not enable NGG culling on GFX11
5039 - radv: disable VRS entirely on GFX11
5054 - radv: fix setting results for initialization failures in thread trace and trap handler
5244 - radv: Implement radv_flush_before_query_copy to workaround UE Vulkan bugs.
5245 - radv: Remove unused functions.
5246 - radv: Remove unused radv_deferred_queue_submission.
5247 - radv,lvp,anv: Delete pre-common dispatch leftover.
5290 - radv: add radv_zero_vram workarounds for OpenGL games
5333 - radv: Implement EXT_mesh_shader draw calls.
5334 - radv: Enable EXT_mesh_shader on RDNA2 with RADV_PERFTEST=ext_ms
5337 - radv: Replace NV_mesh_shader macros with EXT.
5338 - radv: Enable subgroup feature bits for mesh and task shaders.
5339 - radv: Change max preferred task workgroup invocations to 64.
5340 - radv: Remove dead shader temps after linking.
5352 - radv: Use a fallback for marketing name when libdrm doesn't know it.
5374 - radv/ci: Put one more board to run the CTS on Stoney Ridge
5388 - radeonsi,radv/ci: Increase coverage
5392 - radv: refactor descriptor set layout tracking
5400 - radv: Use count_tes_user_sgprs return value.
5402 - radv: Fix file descriptor leak.
5580 - radv: Remove the redundant #include <gelf.h> and #include <libelf.h> in ac_binary.c
5585 - radv/r600/clover: Getting libelf to be optional
5586 - radv: Getting radeon_icd to be generated properly on win32
5681 - radv: Fixes prototypes