fix(shader): leave the workgroup memory set out of a compute module's bindings by BunnyFl · Pull Request #11951 · boykopovar/AnyPS5

BunnyFl · 2026-10-27T13:46:54Z

What

A compute shader whose LDS plus 4 bytes is larger than the device's maxWorkgroupSharedMemoryBytes keeps its LDS in a storage buffer declared in descriptor set 1, binding 0 (WorkgroupMemoryDescriptorSet). buildSpecializedModule lists the binding number of every descriptor variable that remains after specialization and left out only the pixel function LDS set (set 2), so it listed that buffer as binding 0. DescriptorBindingBuilder::Select then looks the module's binding numbers up in the prepared plan, which holds only set 0. When the shader has guest buffers, set 0's binding 0 (Buffers) is there and the lookup succeeds by number. When it has none, the recompile fails with specialized module bindings disagree with the prepared plan.

This leaves the variables of WorkgroupMemoryDescriptorSet out for a GLCompute entry point, as the pixel function LDS set already is. Set 1 is also FragmentDescriptorSet, so pixel shaders keep listing their set 1 bindings.

Tested

Windows 11 (10.0.26200), WinLibs GCC 15.2.0 r7 (MinGW-w64 UCRT), Release, BUILD_TESTING=ON, on main 75f3702aed:

New case in agc_lds_lock_budget: a compute program with one image load, 16 dwords of LDS stored with ds_write_b32 and no buffers, compiled for a 64-byte shared memory limit. It checks that the program compiles, that its LDS is in device memory (workgroupMemoryDwords == 18) and that it was specialized. It passes in three runs; with only Recompiler.cpp put back to main's version, it fails in all three runs with specialized module bindings disagree with the prepared plan. The 12 existing cases pass in both. The runs set ANYPS5_REQUIRE_VULKAN=1, ANYPS5_NO_SHADER_CACHE=1 and VK_INSTANCE_LAYERS=VK_LAYER_KHRONOS_validation; the test only recompiles and creates no Vulkan device, so neither log has a VUID- line or a missing-layer message.
tools/check_conventions.py --base upstream/main passes.

Planet Zoo (PPSA18026 01.009.002), AMD Radeon RX 9070 XT (maxComputeSharedMemorySize 32768, driver 32.0.32015.7029):

main aborts before any frame with compute shader 0xb401f08800: specialized module bindings disagree with the prepared plan: 55 and 46 s in on 3fea2a937b, 49 s in on 8ff4baf534. That shader uses 8192 dwords of LDS and no buffers. Its request, saved with APS5_DUMP_SHADERS, fails the same way in agc_shader_replay with main's Recompiler.cpp and recompiles with this change (on 62b3ba2522 too), as do the other nine requests saved in that run.
With this change, on 8ff4baf534, the title no longer aborts there. In one run it showed the Frontier logo and the EULA screen (288 presents before a timed APS5_HANG_DUMP_SECS dump at 60 s). In two other 300 s runs its log stopped after about 80 s and 30 s, with 3 and 0 presents; that stop is not investigated here.

Rebased onto main 08a10d2253 (head 40775bc85a). main had added MeshWindowDescriptorSet to the same condition (6f79055); the rebased line keeps it next to this change's compute-only WorkgroupMemoryDescriptorSet. On that head, same machine and compiler (these runs did not set the Vulkan variables above):

agc_lds_lock_budget passes in three runs; with only Recompiler.cpp put back to main's version it fails in all three with LDS lock budget: specialized image program without buffers: ShaderRecompiler::Recompile: specialized module bindings disagree with the prepared plan.
tools/check_conventions.py --base upstream/main passes.
Planet Zoo, freshly converted with this build (AMD driver 26.10.1, 3840x2160): one unattended 180 s run with Cross pressed every 3 s presents steadily from 40 s on (about 59 presents per 10 s) and its last saved frame shows the EULA screen, with no disagree with the prepared plan line; the run stopped at the time cap, not by itself. The same title on main 99b571f879 aborted 49 s in with that error.

Checklist

🤖 Generated with Claude Code

github-actions · 2026-10-27T14:30:38Z

Some comparisons could not be completed because a pull request conflicts with the base branch. Review these shared files manually before concluding that there is no overlap.

Pull request	Shared files not compared
PR 5021	`core/shader/recompiler/Recompiler.cpp`
PR 5609	`core/shader/recompiler/Recompiler.cpp`
PR 6412	`core/shader/recompiler/Recompiler.cpp`
PR 6488	`core/shader/recompiler/Recompiler.cpp`
PR 6858	`core/shader/recompiler/Recompiler.cpp`
PR 6924	`core/shader/recompiler/Recompiler.cpp`
PR 7223	`core/shader/recompiler/Recompiler.cpp`
PR 7410	`core/shader/recompiler/Recompiler.cpp`
PR 7642	`core/shader/recompiler/Recompiler.cpp`
PR 8004	`core/shader/recompiler/Recompiler.cpp`
PR 9020	`core/shader/recompiler/Recompiler.cpp`
PR 9335	`core/shader/recompiler/Recompiler.cpp`
PR 9356	`core/shader/recompiler/Recompiler.cpp`
PR 9387	`core/shader/recompiler/Recompiler.cpp`
PR 9508	`core/shader/recompiler/Recompiler.cpp`
PR 9748	`core/shader/recompiler/Recompiler.cpp`
PR 10193	`core/shader/recompiler/Recompiler.cpp`
PR 10359	`core/shader/recompiler/Recompiler.cpp`
PR 10757	`core/shader/recompiler/Recompiler.cpp`
PR 11351	`core/shader/recompiler/Recompiler.cpp`
PR 11483	`core/shader/recompiler/Recompiler.cpp`
PR 11551	`core/shader/recompiler/Recompiler.cpp`
PR 12324	`core/shader/recompiler/Recompiler.cpp`
PR 12548	`core/shader/recompiler/Recompiler.cpp`
PR 12662	`core/shader/recompiler/Recompiler.cpp`
PR 12790	`core/shader/recompiler/Recompiler.cpp`
PR 12828	`core/shader/recompiler/Recompiler.cpp`
PR 12840	`core/shader/recompiler/Recompiler.cpp`
PR 12944	`core/shader/recompiler/Recompiler.cpp`
PR 12980	`core/shader/recompiler/Recompiler.cpp`
PR 13928	`core/shader/recompiler/Recompiler.cpp`
PR 13929	`core/shader/recompiler/Recompiler.cpp`
PR 14032	`core/shader/recompiler/Recompiler.cpp`
PR 14217	`core/shader/recompiler/Recompiler.cpp`
PR 14853	`core/shader/recompiler/Recompiler.cpp`
PR 14871	`core/shader/recompiler/Recompiler.cpp`
PR 14944	`core/shader/recompiler/Recompiler.cpp`

maskym · 2026-10-27T22:36:56Z

Windows + NVIDIA result for head 22d08b63a4, merged into main f759e165d9 (the PR's GPUs tested line names no NVIDIA run on Windows).

This PR's own tests: agc_lds_lock_budget (0.42 s) and agc_driver_lds_lock_budget (0.50 s), all passed.

Host: Windows 11 26300, Intel i7-12700K, NVIDIA GeForce RTX 2070 SUPER (vendor 0x10de, device 0x1e84; Vulkan loader 1.4.357 from the NVIDIA driver), Khronos Validation Layers nightly 1.4.366 (b2b44c0d37) through VK_ADD_LAYER_PATH, WinLibs GCC 15.2.0 (x86_64-ucrt-posix-seh, r6), Release, BUILD_TESTING=ON, Ninja. Build clean.

Same focused run on main f759e165d9 and on this tree, same machine and GPU: VK_INSTANCE_LAYERS=VK_LAYER_KHRONOS_validation, ANYPS5_NO_SHADER_CACHE=1, ANYPS5_REQUIRE_VULKAN=1, VK_LOADER_DEBUG=error,warn,layer, ctest --test-dir build -R '^(agc|guest_video_out|video_out)' --parallel 4 -V.

main: 1127/1128 passed (9 skipped), the one failure being agc_driver_mesh_barycentric_emulation (AGC: agc_driver_mesh_barycentric_emulation fails since #9193 on devices with mesh shaders and VK_KHR_fragment_shader_barycentric #12121, pre-existing).
This tree: every pre-existing test gives the same result; the loader inserted VK_LAYER_KHRONOS_validation in every test (998 and 1012 insertions), neither log contains was not found but was requested by env var VK_INSTANCE_LAYERS, and both report exactly one VUID, VUID-VkPipelineColorBlendStateCreateInfo-pAttachments-00605 18 times (pre-existing on main, AGC: pipelines use differing per-attachment blend states while independentBlend is never enabled (VUID-VkPipelineColorBlendStateCreateInfo-pAttachments-00605) #11990 / fix(agc): enable independentBlend for color targets with different blend states #6782): this PR adds no new VUID- error.

The tree also carried the open PRs #11711, #11704, #11692, #11691, #11690, #11794, #11949, all merged cleanly, so the run is a joint one; of those, #11692's new agc_driver_depth_plane_cpu_write_after_page_unwatch fails on this host, reported on that PR. Not run on a PS5.

AI-assisted: yes (Claude Code), the runs and the numbers above are from this machine.

TechLevelZero · 2026-10-29T16:06:34Z

✅ NVIDIA GPU test: passed

PR head 22d08b6 merged onto main ba5b533 · tested 2026-10-29 16:06 UTC · 25 min


GPU	NVIDIA RTX 2000 Ada Generation, driver 595.99.02, Vulkan 1.4.329
Host	Ubuntu 26.04.1 LTS, kernel 7.0.0-38-generic, container ubuntu:24.04
Logs	build and test logs for this run
Build	✅ passed (Release, `-DANYPS5_ENABLE_SPIRV_TOOLS=ON`)
Tests	✅ 2,209 passed · 5 failed (0 new, 5 also fail on main) · 39 skipped (`ANYPS5_REQUIRE_VULKAN=1`)
Test setup	CPU-only tests without a GPU; GPU tests (and any CPU-only test that failed or skipped without one) once on the GPU with `VK_LAYER_KHRONOS_validation` and `ANYPS5_NO_SHADER_CACHE=1`; main ran the same way
Test selection	2,253 of 2,891 tests run: the ones this change can reach, chosen by `tools/select_tests.py` from main's build graph
Vulkan validation	✅ no new validation IDs vs main (855 tests run on the GPU)
Slow tests (>30 s, GPU tests timed with validation layers)	none

Failing on main too (5, not counted against this PR)

agc_compute_pipeline_inspection
agc_submit_progress
agc_submit_progress_no_timeline
guest_mmap_executable
guest_socket_options_multicast

Validation IDs already present on main (4 IDs in 404 tests)

VUID-VkDeviceCreateInfo-pNext-pNext (404 tests): agc_compute_pipeline_replay, agc_driver, agc_driver_address_passes, agc_driver_address_passes_draw_transitions, agc_driver_address_passes_targets_split, agc_driver_alpha_to_coverage, agc_driver_array_layer_target_cpu, agc_driver_bc_block_copy, and 396 more
VUID-vkCmdCopyImage-srcImage-01548 (11 tests): agc_driver_depth_clear_hle, agc_driver_depth_plane_buffer_memory, agc_driver_depth_plane_cpu_write_after_memo, agc_driver_depth_plane_cpu_write_after_page_unwatch, agc_driver_depth_plane_cpu_write_after_partial_unwatch, agc_driver_depth_plane_partial_store, agc_driver_depth_plane_partial_store_unimported, agc_driver_depth_storage_dispatch, and 3 more
VUID-vkCmdCopyImage-srcImage-01551 (11 tests): agc_driver_depth_clear_hle, agc_driver_depth_plane_buffer_memory, agc_driver_depth_plane_cpu_write_after_memo, agc_driver_depth_plane_cpu_write_after_page_unwatch, agc_driver_depth_plane_cpu_write_after_partial_unwatch, agc_driver_depth_plane_partial_store, agc_driver_depth_plane_partial_store_unimported, agc_driver_depth_storage_dispatch, and 3 more
VUID-VkPipelineColorBlendStateCreateInfo-pAttachments-00605 (2 tests): agc_driver_prepared_exports, agc_driver_srgb8_color_target

Skipped (39)

agc_driver_bda_reads_shadowed_results: skipped, the device reports watched host memory written when it imports or reads it
agc_driver_bda_reads_shadowed_results_per_build: skipped, the device reports watched host memory written when it imports or reads it
agc_driver_msaa_dcc_compared_import_2: skipped, the device reports host imports that keep write protection, so the write-watch probe compares none of them
agc_driver_msaa_dcc_compared_import_4: skipped, the device reports host imports that keep write protection, so the write-watch probe compares none of them
agc_driver_msaa_dcc_compared_import_8: skipped, the device reports host imports that keep write protection, so the write-watch probe compares none of them
agc_driver_sample_order_grid_reference: skipped, the device does not keep depth drawn at custom sample locations at those locations, even after a transition that names them: the composite of uniform renders holds the guest's depth in 0 of 3005 sloped samples, so no path can be compared with it bit for bit
agc_driver_sparse_dma_buf_import: skipped, host imports are unavailable: /dev/udmabuf is not accessible
agc_driver_stencil_ref_export: skipped, the device lacks VK_EXT_shader_stencil_export
agc_host_import_sparse: skipped, host imports are unavailable: /dev/udmabuf is not accessible
agc_import_settle_initial_dcc: skipped, the device has no deferred settle of host imports: imports under 1 MiB wait for one only on Windows
agc_unit_shadow_completion_store: skipped, the device reports watched host memory written when it imports or reads it
guest_net_ctl: skipped, the host has no IPv4 address
guest_video_out_color_space: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_hdr: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_high_frame_rate: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_output_mode: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_resolution_status: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_wqhd_detection: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
video_out_flip_controls: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_cpuflip: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_cpuflip_in_order: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_eop: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_eop_aftergpu: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_eop_main: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_fresh: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_labels: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SD

Reproduce

git fetch origin 22d08b63a4b379f880b428884f4158dce05601e0 ba5b5330b37724bca4b99f30faab504653635a04
git checkout ba5b5330b37724bca4b99f30faab504653635a04 && git merge --no-ff 22d08b63a4b379f880b428884f4158dce05601e0
git submodule update --init --recursive
cmake -S . -B build -G Ninja -DCMAKE_BUILD_TYPE=Release -DBUILD_TESTING=ON -DANYPS5_ENABLE_SPIRV_TOOLS=ON
cmake --build build && cmake --build build --target libs
VK_DRIVER_FILES=/nonexistent/icd.json ctest --test-dir build -E '^(agc|shader|recompiler|videoout|video_out|guest_video_out|vulkan)|gpu' --timeout 120 --parallel 32 --output-on-failure
ANYPS5_REQUIRE_VULKAN=1 ANYPS5_NO_SHADER_CACHE=1 VK_INSTANCE_LAYERS=VK_LAYER_KHRONOS_validation \
  ctest --test-dir build -R '^(agc|shader|recompiler|videoout|video_out|guest_video_out|vulkan)|gpu' --timeout 120 --parallel 4 --output-on-failure
# failures are then re-run alone: --timeout 360 --parallel 2; new validation IDs: the test alone, twice, --parallel 1

_{Automated build and test on a contributor's NVIDIA machine; it does not replace upstream CI. A new comment is posted for each new push. Full logs.}

… bindings A compute shader whose LDS does not fit the device limit keeps it in a storage buffer in descriptor set 1, binding 0. Specialization listed that variable with the module's set 0 bindings, so selecting the prepared plan failed for a shader with no guest buffers, because nothing in set 0 uses binding 0. Only the pixel function LDS set was left out before. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

BunnyFl · 2026-10-30T03:01:13Z

Rebased onto main 08a10d2253 (head 40775bc85a) to resolve the conflict with 6f79055, which added MeshWindowDescriptorSet to the same condition; both exclusions are kept. Retested on the new head: agc_lds_lock_budget passes 3 of 3 and fails 3 of 3 with main's Recompiler.cpp, conventions pass, and Planet Zoo now presents steadily up to the EULA screen in a 180 s run instead of aborting at 49 s. Details are in the Tested section.

AI-assisted: yes (Claude Code)

TechLevelZero · 2026-10-30T15:01:37Z

✅ NVIDIA GPU test: passed

PR head 40775bc merged onto main 93dbd0f · tested 2026-11-02 06:44 UTC · 18 min


GPU	NVIDIA RTX 2000 Ada Generation, driver 595.99.02, Vulkan 1.4.329
Host	Ubuntu 26.04.1 LTS, kernel 7.0.0-38-generic, container ubuntu:24.04
Logs	build and test logs for this run
Build	✅ passed (Release, `-DANYPS5_ENABLE_SPIRV_TOOLS=ON`)
Tests	✅ 2,442 passed · 5 failed (0 new, 3 also fail on main, 2 unstable on main) · 45 skipped (`ANYPS5_REQUIRE_VULKAN=1`)
Test setup	CPU-only tests without a GPU; GPU tests (and any CPU-only test that failed or skipped without one) once on the GPU with `VK_LAYER_KHRONOS_validation` and `ANYPS5_NO_SHADER_CACHE=1`; main ran the same way
Test selection	2,492 of 3,165 tests run: the ones this change can reach, chosen by `tools/select_tests.py` from main's build graph
Vulkan validation	✅ no new validation IDs vs main (930 tests run on the GPU)
Slow tests (>30 s, GPU tests timed with validation layers)	⚠️ `agc_driver_pipeline_journal` 120.3 s, ⚠️ `agc_driver_lookup_log_fits_build` 94.4 s, ⚠️ `agc_driver_graphics_maps_fallback` 32.5 s, ⚠️ `agc_driver_graphics` 31.6 s

Unstable on main (2, these also failed in recent main runs on this machine; not counted against this PR)

agc_driver_pipeline_journal: failed in 1 of the last 10 main runs
agc_submit_progress: failed in 6 of the last 10 main runs

Failing on main too (3, not counted against this PR)

agc_driver_image_table
guest_mmap_executable
guest_socket_options_multicast

Validation IDs already present on main (5 IDs in 484 tests)

VUID-VkDeviceCreateInfo-pNext-pNext (484 tests): agc_driver, agc_driver_address_passes, agc_driver_alias_copy_label, agc_driver_alpha_to_coverage, agc_driver_array_layer_target, agc_driver_bc_block_copy, agc_driver_bda_anchor_reads, agc_driver_bda_span_reads, and 476 more
VUID-vkCmdCopyImage-srcImage-01548 (7 tests): agc_driver_depth_clear_hle, agc_driver_depth_plane_buffer_memory, agc_driver_depth_plane_partial_store, agc_driver_depth_plane_sampling, agc_driver_depth_storage_dispatch, agc_driver_depth_surface_from_storage, agc_driver_dispatch_move_depth
VUID-vkCmdCopyImage-srcImage-01551 (7 tests): agc_driver_depth_clear_hle, agc_driver_depth_plane_buffer_memory, agc_driver_depth_plane_partial_store, agc_driver_depth_plane_sampling, agc_driver_depth_storage_dispatch, agc_driver_depth_surface_from_storage, agc_driver_dispatch_move_depth
VUID-VkPipelineColorBlendStateCreateInfo-pAttachments-00605 (1 test): agc_driver_prepared_exports
VUID-VkShaderModuleCreateInfo-pCode-08737 (1 test): agc_driver_image_sample_derivatives

Skipped (45)

agc_driver_bda_reads_shadowed_results: skipped, the device reports watched host memory written when it imports or reads it
agc_driver_bda_reads_shadowed_results_per_build: skipped, the device reports watched host memory written when it imports or reads it
agc_driver_sample_order_grid_reference: skipped, the device does not keep depth drawn at custom sample locations at those locations, even after a transition that names them: the composite of uniform renders holds the guest's depth in 0 of 3005 sloped samples, so no path can be compared with it bit for bit
agc_host_import_sparse: skipped, host imports are unavailable: /dev/udmabuf is not accessible
agc_import_settle_initial_dcc: skipped, the device has no deferred settle of host imports: imports under 1 MiB wait for one only on Windows
agc_unit_shadow_completion_store: skipped, the device reports watched host memory written when it imports or reads it
guest_http_uri_locale: 
guest_net_ctl: skipped, the host has no IPv4 address
guest_video_out_color_space: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_hdr: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_high_frame_rate: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_output_mode: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_resolution_status: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_wqhd_detection: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
video_out_flip_aftergpu: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_backtoback: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_controls: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_cpuflip: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_cpuflip_in_order: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_dcc: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_deferred: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_eop: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_eop_aftergpu: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_eop_main: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_fps_limit: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_o

Reproduce

git fetch origin 40775bc85aa664ca8dfaae5a99fad64e3c80ed9a 93dbd0f82aea684e95f15f06e7d42c0064af18cf
git checkout 93dbd0f82aea684e95f15f06e7d42c0064af18cf && git merge --no-ff 40775bc85aa664ca8dfaae5a99fad64e3c80ed9a
git submodule update --init --recursive
cmake -S . -B build -G Ninja -DCMAKE_BUILD_TYPE=Release -DBUILD_TESTING=ON -DANYPS5_ENABLE_SPIRV_TOOLS=ON
cmake --build build && cmake --build build --target libs
VK_DRIVER_FILES=/nonexistent/icd.json ctest --test-dir build -E '^(agc|shader|recompiler|videoout|video_out|guest_video_out|vulkan)|gpu' --timeout 120 --parallel 32 --output-on-failure
ANYPS5_REQUIRE_VULKAN=1 ANYPS5_NO_SHADER_CACHE=1 VK_INSTANCE_LAYERS=VK_LAYER_KHRONOS_validation \
  ctest --test-dir build -R '^(agc|shader|recompiler|videoout|video_out|guest_video_out|vulkan)|gpu' --timeout 120 --parallel 4 --output-on-failure
# failures are then re-run alone: --timeout 360 --parallel 2; new validation IDs: the test alone, twice, --parallel 1

_{Automated build and test on a contributor's NVIDIA machine; it does not replace upstream CI. A new comment is posted for each new push. Full logs.}

oneandonlydean · 2026-11-02T08:13:05Z

Hi BunnyFl, #6959 merged today, and this PR no longer merges cleanly with main. These conflicts overlap your changes, so they need your call:

core/shader/recompiler/Recompiler.cpp (main side from fix(agc): render layered mesh targets within device limits #6959)

Could you merge main into this branch and resolve them, keeping what that PR added? A merge commit is fine; no need to rebase or force-push. Once it's pushed we'll re-review the merge.

Keeps the specialization constants boykopovar#6959 records next to this branch's compute workgroup memory exclusion. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

BunnyFl · 2026-11-02T10:09:59Z

@oneandonlydean Merged main (88e804b0a7) into the branch as merge commit c258c310db, no rebase or force-push. The only conflict was in buildSpecializedModule in core/shader/recompiler/Recompiler.cpp. The resolution keeps #6959's DecorationSpecId collection unchanged and keeps this branch's ldsBuffers check, which still leaves out WorkgroupMemoryDescriptorSet for a GLCompute entry point. #6959 doesn't touch that set, so the two changes are independent.

On the merge, Windows 11, WinLibs GCC 15.2.0, Release, with ANYPS5_REQUIRE_VULKAN=1 ANYPS5_NO_SHADER_CACHE=1 VK_INSTANCE_LAYERS=VK_LAYER_KHRONOS_validation:

agc_lds_lock_budget and fix(agc): render layered mesh targets within device limits #6959's agc_shader_memory pass in three runs. No log has a VUID- or SYNC-HAZARD line or a missing-layer message.
With only Recompiler.cpp put back to main's version, agc_lds_lock_budget fails with specialized module bindings disagree with the prepared plan.
tools/check_conventions.py --base upstream/main passes.

TechLevelZero · 2026-11-02T10:59:22Z

✅ NVIDIA GPU test: passed

PR head c258c31 merged onto main 88e804b · tested 2026-11-02 10:59 UTC · 21 min


GPU	NVIDIA RTX 2000 Ada Generation, driver 595.99.02, Vulkan 1.4.329
Host	Ubuntu 26.04.1 LTS, kernel 7.0.0-38-generic, container ubuntu:24.04
Logs	build and test logs for this run
Build	✅ passed (Release, `-DANYPS5_ENABLE_SPIRV_TOOLS=ON`)
Tests	✅ 2,496 passed · 5 failed (0 new, 3 also fail on main, 2 unstable on main) · 45 skipped (`ANYPS5_REQUIRE_VULKAN=1`)
Test setup	CPU-only tests without a GPU; GPU tests (and any CPU-only test that failed or skipped without one) once on the GPU with `VK_LAYER_KHRONOS_validation` and `ANYPS5_NO_SHADER_CACHE=1`; main ran the same way
Test selection	2,546 of 3,234 tests run: the ones this change can reach, chosen by `tools/select_tests.py` from main's build graph
Vulkan validation	✅ no new validation IDs vs main (970 tests run on the GPU)
Slow tests (>30 s, GPU tests timed with validation layers)	⚠️ `agc_driver_pipeline_journal` 120.1 s, ⚠️ `agc_driver_lookup_log_fits_build` 36.3 s, ⚠️ `agc_driver_graphics_maps_fallback` 30.8 s, ⚠️ `agc_driver_graphics` 30.4 s

Unstable on main (2, these also failed in recent main runs on this machine; not counted against this PR)

agc_driver_pipeline_journal: failed in 2 of the last 10 main runs
agc_submit_progress: failed in 8 of the last 10 main runs

Failing on main too (3, not counted against this PR)

agc_driver_image_table
guest_mmap_executable
guest_socket_options_multicast

Validation IDs already present on main (4 IDs in 495 tests)

VUID-VkDeviceCreateInfo-pNext-pNext (495 tests): agc_driver, agc_driver_address_passes, agc_driver_alias_copy_label, agc_driver_alpha_to_coverage, agc_driver_array_layer_target, agc_driver_bc_block_copy, agc_driver_bda_anchor_reads, agc_driver_bda_span_reads, and 487 more
VUID-vkCmdCopyImage-srcImage-01548 (7 tests): agc_driver_depth_clear_hle, agc_driver_depth_plane_buffer_memory, agc_driver_depth_plane_partial_store, agc_driver_depth_plane_sampling, agc_driver_depth_storage_dispatch, agc_driver_depth_surface_from_storage, agc_driver_dispatch_move_depth
VUID-vkCmdCopyImage-srcImage-01551 (7 tests): agc_driver_depth_clear_hle, agc_driver_depth_plane_buffer_memory, agc_driver_depth_plane_partial_store, agc_driver_depth_plane_sampling, agc_driver_depth_storage_dispatch, agc_driver_depth_surface_from_storage, agc_driver_dispatch_move_depth
VUID-VkShaderModuleCreateInfo-pCode-08737 (1 test): agc_driver_image_sample_derivatives

Skipped (45)

agc_driver_bda_reads_shadowed_results: skipped, the device reports watched host memory written when it imports or reads it
agc_driver_bda_reads_shadowed_results_per_build: skipped, the device reports watched host memory written when it imports or reads it
agc_driver_sample_order_grid_reference: skipped, the device does not keep depth drawn at custom sample locations at those locations, even after a transition that names them: the composite of uniform renders holds the guest's depth in 0 of 3005 sloped samples, so no path can be compared with it bit for bit
agc_host_import_sparse: skipped, host imports are unavailable: /dev/udmabuf is not accessible
agc_import_settle_initial_dcc: skipped, the device has no deferred settle of host imports: imports under 1 MiB wait for one only on Windows
agc_unit_shadow_completion_store: skipped, the device reports watched host memory written when it imports or reads it
guest_http_uri_locale: 
guest_net_ctl: skipped, the host has no IPv4 address
guest_video_out_color_space: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_hdr: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_high_frame_rate: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_output_mode: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_resolution_status: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_wqhd_detection: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
video_out_flip_aftergpu: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_backtoback: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_controls: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_cpuflip: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_cpuflip_in_order: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_dcc: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_deferred: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_eop: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_eop_aftergpu: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_eop_main: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_fps_limit: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_o

Reproduce

git fetch origin c258c310db3055bb980b540dcbcdd97a7c457f09 88e804b0a7df1f1ac373a9e2f72fb378d925fdd4
git checkout 88e804b0a7df1f1ac373a9e2f72fb378d925fdd4 && git merge --no-ff c258c310db3055bb980b540dcbcdd97a7c457f09
git submodule update --init --recursive
cmake -S . -B build -G Ninja -DCMAKE_BUILD_TYPE=Release -DBUILD_TESTING=ON -DANYPS5_ENABLE_SPIRV_TOOLS=ON
cmake --build build && cmake --build build --target libs
VK_DRIVER_FILES=/nonexistent/icd.json ctest --test-dir build -E '^(agc|shader|recompiler|videoout|video_out|guest_video_out|vulkan)|gpu' --timeout 120 --parallel 32 --output-on-failure
ANYPS5_REQUIRE_VULKAN=1 ANYPS5_NO_SHADER_CACHE=1 VK_INSTANCE_LAYERS=VK_LAYER_KHRONOS_validation \
  ctest --test-dir build -R '^(agc|shader|recompiler|videoout|video_out|guest_video_out|vulkan)|gpu' --timeout 120 --parallel 4 --output-on-failure
# failures are then re-run alone: --timeout 360 --parallel 2; new validation IDs: the test alone, twice, --parallel 1

_{Automated build and test on a contributor's NVIDIA machine; it does not replace upstream CI. A new comment is posted for each new push. Full logs.}

github-actions Bot added the unblocks title Names a title ID on the Unblocks line label Oct 27, 2026

github-actions Bot added shader core/shader agc libSceAgc* labels Oct 27, 2026

maskym mentioned this pull request Oct 27, 2026

build(windows): quote the forced standard-stream include path #11949

Merged

12 tasks done

fogydes mentioned this pull request Oct 28, 2026

fix(shader): keep mesh register windows out of local memory on NVIDIA #12233

Merged

12 tasks done

oneandonlydean self-assigned this Oct 28, 2026

wrocketdev mentioned this pull request Oct 29, 2026

fix(libSceAmpr): declare the page-table pool occupancy threshold argument #13007

Merged

12 tasks done

github-actions Bot added the needs rebase Conflicts with main, rebase needed label Oct 30, 2026

BunnyFl force-pushed the fix/planet-zoo-binding-select branch from 22d08b6 to 40775bc Compare October 30, 2026 03:00

github-actions Bot removed the needs rebase Conflicts with main, rebase needed label Oct 30, 2026

github-actions Bot added the needs rebase Conflicts with main, rebase needed label Nov 2, 2026

Merge upstream/main into fix/planet-zoo-binding-select

c258c31

Keeps the specialization constants boykopovar#6959 records next to this branch's compute workgroup memory exclusion. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

github-actions Bot removed the needs rebase Conflicts with main, rebase needed label Nov 2, 2026

github-actions Bot added the review queue Next in the maintainer review queue: checks pass, no open change request, no conflict label Nov 2, 2026

oneandonlydean merged commit 26b0933 into boykopovar:main Nov 2, 2026
14 checks passed

github-actions Bot removed the review queue Next in the maintainer review queue: checks pass, no open change request, no conflict label Nov 2, 2026

Navigation Menu

Repository navigation

fix(shader): leave the workgroup memory set out of a compute module's bindings - #11951

fix(shader): leave the workgroup memory set out of a compute module's bindings#11951
oneandonlydean merged 2 commits into
boykopovar:mainfrom
BunnyFl:fix/planet-zoo-binding-select

BunnyFl commented Oct 27, 2026 •

edited

Loading

Uh oh!

github-actions Bot commented Oct 27, 2026 •

edited

Loading

Uh oh!

maskym commented Oct 27, 2026

Uh oh!

TechLevelZero commented Oct 29, 2026

Uh oh!

BunnyFl commented Oct 30, 2026

Uh oh!

TechLevelZero commented Oct 30, 2026 •

edited

Loading

Uh oh!

oneandonlydean commented Nov 2, 2026

Uh oh!

BunnyFl commented Nov 2, 2026

Uh oh!

TechLevelZero commented Nov 2, 2026

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Conversation

BunnyFl commented Oct 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What

Tested

Checklist

Uh oh!

github-actions Bot commented Oct 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

maskym commented Oct 27, 2026

Uh oh!

TechLevelZero commented Oct 29, 2026

✅ NVIDIA GPU test: passed

Uh oh!

BunnyFl commented Oct 30, 2026

Uh oh!

TechLevelZero commented Oct 30, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

✅ NVIDIA GPU test: passed

Uh oh!

oneandonlydean commented Nov 2, 2026

Uh oh!

BunnyFl commented Nov 2, 2026

Uh oh!

TechLevelZero commented Nov 2, 2026

✅ NVIDIA GPU test: passed

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

BunnyFl commented Oct 27, 2026 •

edited

Loading

github-actions Bot commented Oct 27, 2026 •

edited

Loading

TechLevelZero commented Oct 30, 2026 •

edited

Loading