18710 commits

Author	SHA1	Message	Date
Yonggang Luo	7db518cfe4	aco: Fixes warning: function get_branch_target/to_clrx_device_name defined but not used ... ../../src/amd/compiler/aco_print_asm.cpp:156:1: warning: 'bool aco::{anonymous}::get_branch_target(char**, aco::Program*, const std::vector<bool>&, char**)' defined but not used [-Wunused-function] 156 | get_branch_target(char** output, Program* program, const std::vector<bool>& referenced_blocks, | ^~~~~~~~~~~~~~~~~ ../../src/amd/compiler/aco_print_asm.cpp:105:1: warning: 'const char* aco::{anonymous}::to_clrx_device_name(amd_gfx_level, radeon_family)' defined but not used [-Wunused-function] 105 | to_clrx_device_name(amd_gfx_level gfx_level, radeon_family family) | ^~~~~~~~~~~~~~~~~~~ Signed-off-by: Yonggang Luo <luoyonggang@gmail.com> Reviewed-by: Georg Lehmann <dadschoorse@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37289>	2025-09-13 08:23:07 +00:00
Timur Kristóf	c183eb5bc8	radv: Flush L2 before CP DMA copy/fill when CP DMA doesn't use L2 ... In case the source or destination were previously written through L2, we need to writeback L2 to avoid the CP DMA accessing stale data. However, as the CP DMA doesn't write L2 either, an invalidation is also needed to make sure other clients don't access stale data when they read it through L2 after the CP DMA is complete. Doing an invalidation before the CP DMA operation should take care of both. Additionally, radv_src_access_flush also invalidates L2 before the copied data can be read. Cc: mesa-stable Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36820>	2025-09-13 05:18:28 +00:00
Timur Kristóf	8ef2f492e2	radv: Add comment to document CP DMA prefetch ... Explain which caches (MALL, L2) the prefetch works with. Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36820>	2025-09-13 05:18:28 +00:00
Val Packett	7f556805c1	radv: detect platform:virtio-mmio devices for virtgpu native context ... VirtIO devices can be configured as platform devices instead of PCI, which is especially common in microVM projects like libkrun. Let's allow RADV to probe MMIO virtgpu devices. Signed-off-by: Val Packett <val@invisiblethingslab.com> Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37281>	2025-09-12 12:56:46 +00:00
Karol Herbst	4f94ca6c96	ac/llvm: fix get_global_address for global atomics ... They do not have an ACCESS index. Fixes: 9a33c03654 ("ac/llvm: port load_smem_amd behavior to load_global_amd") Reviewed-by: Konstantin Seurer <konstantin.seurer@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37292>	2025-09-12 08:55:39 +00:00
Samuel Pitoiset	a52483d9e7	radv: fix capture/replay with sampler border color ... The border color index must be captured and returned to the app, otherwise it's broken if samplers aren't replayed in-order. Cc: mesa-stable Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37291>	2025-09-12 06:51:51 +00:00
Samuel Pitoiset	a658be8a24	radv: get NIR options after initializing the physical device cache key ... Otherwise, radv_split_fma isn't considered. Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13878 Fixes: 7304423b5c ("radv: mark RADV_DEBUG=splitfma as deprecated") Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37293>	2025-09-12 06:18:26 +00:00
Eric Engestrom	6ecb7f11b7	radv/ci: document recent flakes ... Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37302>	2025-09-11 16:02:38 +00:00
Samuel Pitoiset	b00b8c763b	radv: disable radv_disable_hiz_his_gfx12 for Mafia Definition Edition ... This should be properly fixed now. Also remove this drirc option which should no longer be needed. Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36174>	2025-09-11 15:21:50 +00:00
Samuel Pitoiset	ec09ac1501	radv: switch to the full HiZ workaround by default on GFX12 ... The full HiZ workaround is the only one that fixes the issue reliably. Sadly, the performance results are mixed and sometimes it hurts. To maintain performance, RADV will opt-in by selecting which workarounds to apply from drirc. As we can't know the full list of games that will be affected by a potential performance regression, users are encouraged to try with RADV_GFX12_HIZ_WA (see Mesa documentation for more explanations). Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36174>	2025-09-11 15:21:50 +00:00
Samuel Pitoiset	5f8f4686bf	radv: replace RADV_GFX12_HIZ_WA by a drirc option ... Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36174>	2025-09-11 15:21:50 +00:00
Samuel Pitoiset	0a2ef363a8	radv: report an message when RADV_GFX12_HIZ_WA value is invalid ... Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36174>	2025-09-11 15:21:50 +00:00
Samuel Pitoiset	ec87f1338f	radv: emit more push shader registers on GFX12 ... They are supposed to be slightly faster. Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37256>	2025-09-11 06:47:40 +00:00
Samuel Pitoiset	9039f33a8d	Revert "radv: handle fbfetch output after binding graphics shaders" ... This is actually wrong because if radv_handle_fbfetch_output() triggers a decompression pass and graphics shaders (ESO) are saved/restored they won't be updated because radv_bind_graphics_shaders() was called before. This fixes a very recent regression that I noticed while implementing a new extension. This reverts commit 9b912f00c7. Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37194>	2025-09-11 06:22:20 +00:00
Samuel Pitoiset	b69b953973	radv: add RADV_DEBUG=bo_history ... This dumps the BO history to /tmp/radv_bo_history.log after each BO operations. Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37062>	2025-09-11 06:03:15 +00:00
Samuel Pitoiset	1da270fb35	radv/amdgpu: add more helpers for managing virtual BOs ... All these new helpers will make the SMEM PRT workaround better organized. Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37193>	2025-09-10 14:50:25 +00:00
Samuel Pitoiset	3c4168a3cc	radv/amdgpu: return OOM device when BO mapping fails ... It's more appropriate than VK_ERROR_UNKNOWN. Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37193>	2025-09-10 14:50:24 +00:00
Konstantin Seurer	7c9e945460	radv,vulkan: Avoid a useless barrier in radv_update_bind_pipeline ... Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36982>	2025-09-10 08:35:50 +00:00
Konstantin Seurer	a35dfab281	radv: Use vk_barrier_compute_w_to_compute_r more ... vk_barrier_compute_w_to_compute_r shows up in rgp captures and is less code. Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36982>	2025-09-10 08:35:50 +00:00
Samuel Pitoiset	c739d836f7	radv: exclude dynamic vertex input stride for the late scissor workaround ... RADV_DYNAMIC_VERTEX_INPUT_BINDING_STRIDE doesn't emit any context registers, so it can be excluded for the late scissor workaround to avoid re-emitting scissors all the time it's dirty. This fixes a performance regression noticed with Cyberpunk on Vega10, but other games are likely affected too. The late scissor workaround is only applied on Raven/Vega10. Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13828 Fixes: d7f401c2bb ("radv: bind the vertex binding strides like a normal dynamic state") Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37252>	2025-09-10 07:09:48 +00:00
abdelhadi	3a41644165	aco, radv: remove line duplicate ... Signed-off-by: abdelhadi <abdelhadims@icloud.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37243>	2025-09-10 06:34:43 +00:00
Rhys Perry	e2181744c2	aco/tests: add barrier-to-waitcnt tests ... Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Georg Lehmann <dadschoorse@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36491>	2025-09-09 12:34:40 +00:00
Rhys Perry	0f32b573a4	aco/gfx10: skip waitcnts or use vm_vsrc(0) for workgroup lds barriers ... fossil-db (navi21): Totals from 36594 (45.84% of 79825) affected shaders: Instrs: 19922581 -> 19922563 (-0.00%) CodeSize: 103616980 -> 103616956 (-0.00%) Latency: 69862064 -> 69053273 (-1.16%) InvThroughput: 14607708 -> 14606308 (-0.01%); split: -0.01%, +0.00% fossil-db (navi31): Totals from 1641 (2.06% of 79825) affected shaders: Instrs: 1247591 -> 1247875 (+0.02%); split: -0.00%, +0.03% CodeSize: 6259516 -> 6260612 (+0.02%); split: -0.00%, +0.02% Latency: 7657224 -> 7577299 (-1.04%); split: -1.05%, +0.00% InvThroughput: 1150669 -> 1148171 (-0.22%); split: -0.22%, +0.00% Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Georg Lehmann <dadschoorse@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36491>	2025-09-09 12:34:40 +00:00
Rhys Perry	ac882985c0	aco/gfx10: skip waitcnts or use vm_vsrc(0) for workgroup vmem barriers ... Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Georg Lehmann <dadschoorse@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36491>	2025-09-09 12:34:40 +00:00
Rhys Perry	145b178de2	aco: fix workgroup-scope barrier between vmem and lds ... A barrier between two lds/vmem instructions needs to ensure that the second starts after the first finishes, which means that we can't just skip workgroup-scope vmem barriers if there is a lds instruction later. Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Georg Lehmann <dadschoorse@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36491>	2025-09-09 12:34:40 +00:00
Rhys Perry	02718fd4c5	aco: use a separate event for sendmsg_rtn ... Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Georg Lehmann <dadschoorse@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36491>	2025-09-09 12:34:40 +00:00
Rhys Perry	5812c2ea89	aco: update waitcnt events for exports ... Include primitive, dual source blend and POS4 exports. Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Georg Lehmann <dadschoorse@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36491>	2025-09-09 12:34:40 +00:00
Rhys Perry	711c023b55	aco: remove waitcnt code for POPS ... We now insert barriers around these instead. Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Georg Lehmann <dadschoorse@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36491>	2025-09-09 12:34:40 +00:00
Rhys Perry	005694fe1f	aco: remove waitcnt code for SMEM stores ... These were removed in GFX10.3 and we haven't used them in a while. Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Georg Lehmann <dadschoorse@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36491>	2025-09-09 12:34:40 +00:00
Rhys Perry	20cd5cf5f7	aco: delay barrier waitcnt until they are needed ... fossil-db (navi21): Totals from 44 (0.06% of 79825) affected shaders: Instrs: 16001 -> 15932 (-0.43%); split: -0.46%, +0.02% CodeSize: 85800 -> 85548 (-0.29%); split: -0.30%, +0.01% Latency: 190124 -> 173458 (-8.77%) InvThroughput: 23605 -> 22756 (-3.60%) Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Georg Lehmann <dadschoorse@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36491>	2025-09-09 12:34:40 +00:00
Rhys Perry	843acfa50b	aco: add a separate barrier_info for release/acquire barriers ... These can wait for different sets of accesses. Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Georg Lehmann <dadschoorse@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36491>	2025-09-09 12:34:40 +00:00
Rhys Perry	6c446c2f83	aco: refactor waitcnt pass to use barrier_info ... Currently there's just barrier_info_all, but more will be added later. Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Georg Lehmann <dadschoorse@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36491>	2025-09-09 12:34:40 +00:00
Rhys Perry	21332609b9	aco: don't move acquire barriers before interlock begin ... Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Georg Lehmann <dadschoorse@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36491>	2025-09-09 12:34:40 +00:00
Rhys Perry	0ee1c137f9	aco: don't move release barriers after interlock end ... Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Georg Lehmann <dadschoorse@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36491>	2025-09-09 12:34:40 +00:00
Rhys Perry	7c056dd473	aco: add is_atomic_or_control_instr helper ... Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Georg Lehmann <dadschoorse@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36491>	2025-09-09 12:34:40 +00:00
Rhys Perry	df6a3b7619	aco: reduce cost of using values defined in predecessors ... For code like: if (cond) { val = load() } use(val) The "use(val)" now has a similar cost to a use inside the IF. Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Georg Lehmann <dadschoorse@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36491>	2025-09-09 12:34:40 +00:00
Yonggang Luo	773a7f347a	clang-format: Update the .clang-format files to conformance clang-format json-schema ... The document is at https://clang.llvm.org/docs/ClangFormatStyleOptions.html The json-schema at https://www.schemastore.org/clang-format.json Signed-off-by: Yonggang Luo <luoyonggang@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37235>	2025-09-09 07:04:55 +00:00
Georg Lehmann	4143f0725a	radv/nir/lower_cmat: clean up GFX11 ACC->B convert ... Reviewed-by: Rhys Perry <pendingchaos02@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37213>	2025-09-09 06:08:55 +00:00
Georg Lehmann	5c0ebcdaef	radv/nir/lower_cmat: clean up gfx12 transpose ... Reviewed-by: Rhys Perry <pendingchaos02@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37213>	2025-09-09 06:08:55 +00:00
Georg Lehmann	2da7b4bd0a	radv/nir/lower_cmat: add shuffle_xor_imm helper ... Reviewed-by: Rhys Perry <pendingchaos02@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37213>	2025-09-09 06:08:54 +00:00
Christian Gmeiner	1492de1bc3	radv: re-format using clang-format ... No manual changes here, this is simply running $ ninja -C build/ clang-format Signed-off-by: Christian Gmeiner <cgmeiner@igalia.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37226>	2025-09-09 05:48:56 +00:00
Lars-Ivar Hesselberg Simonsen	ae9edb6fc0	radv/ci: Add uncovered CTS issue to gfx1201 fails ... gfx1201 has started hitting some failures in dEQP-VK.glsl.bfloat16.constant* that seem to be caused by a CTS issue. Suggested-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36985>	2025-09-08 16:54:06 +00:00
Konstantin Seurer	9a93f794cd	radv/bvh: Do not write pointer flag related data on GFX103 ... On GFX103, those fields are unused. VK_BUILD_FLAG_PROPAGATE_CULL_FLAGS is set if the fields are used so it can be used to skip writing them. Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37021>	2025-09-08 12:47:18 +00:00
David Rosca	c9b2c00de6	radv: Enable VK_EXT_ycbcr_2plane_444_formats ... Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/9486 Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37174>	2025-09-08 11:59:37 +00:00
David Rosca	5e7b6df860	radv/video: Support quantization map on VCN5 ... VCN5 needs to convert the input map image to internal format. Reviewed-by: Autumn Ashton <misyl@froggi.es> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37156>	2025-09-08 10:52:06 +00:00
David Rosca	a03e055cd3	radv/video: Set rate control to default on reset ... If we never get rate control info, we would treat it as rate control disabled and use QP value from slice info which is always 0 for default rate control. Cc: mesa-stable Reviewed-by: Autumn Ashton <misyl@froggi.es> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37156>	2025-09-08 10:52:05 +00:00
David Rosca	070c3d2b89	ac/vcn: Add RADEON_VCN_IB_COMMON_OP_RESOLVEINPUTPARAMLAYOUT ... Reviewed-by: Autumn Ashton <misyl@froggi.es> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37156>	2025-09-08 10:52:05 +00:00
Faith Ekstrand	446d5ef103	vulkan: Drop the driver_internal from vk_image_view_init/create() ... It alwways comes in through the create flags now. Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com> Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36957>	2025-09-05 23:34:14 +00:00
Faith Ekstrand	1897d5d9c9	radv: Use VK_IMAGE_VIEW_CREATE_DRIVER_INTERNAL_BIT_MESA ... This does mean having to set the flag everywhere, which is a bit annoying, but I don't think I missed any. Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36957>	2025-09-05 23:34:12 +00:00
Timur Kristóf	78efa4157a	radv: Don't use V_370_PFP or V_028A90_PS_DONE on compute queues ... The compute queue doesn't support these things. This change doesn't fix any known issues, but better to be safe. Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37121>	2025-09-05 13:42:56 +00:00