[kraibse]

• I have read and followed the Contribution Guidelines.
• I have read and followed the AI Policy.
• I have read and followed the Coding Guidelines to the best of my ability.
• I have put [HUMANMADE] in the title if the pr was human-made

---

Summary

Intel Arc dGPUs lost the device soon after boot in Mario Kart World, and also showed banding and crashed on shutdown. This PR fixes those.

• Task shaders: payload word 1 was never set, so a task that culls all its threads launched a garbage number of mesh workgroups. It is now zeroed at the start of the task program, followed by a barrier, as the mesh prologue already does.
• Banding: shaders that lost a conditional barrier raced on Intel's Windows driver. Those shaders now require the guest warp size (32). This is a workaround, not a real barrier. Shader::Info::UsesWarpOps() holds the warp-op condition so compute and graphics can't drift apart.
• Pipeline cache: Intel's Windows driver crashes in vkGetPipelineCacheData while pipelines build. The periodic flush is skipped, and queued builds are dropped and running ones drained before the final save, on that driver only.
• Shutdown: the state tracker kept a stale pointer to the channel's dirty flags. ClearChannel() resets it when the bound channel is released and when the rasterizer is destroyed.
• Mesh/task subgroup size: applies vulkan: Fix shaders that assume 32 wide warps when running on hardware with natively 16 wide warps like Intel #745's change (graphics mesh and task stages that use warp ops require the guest warp size). It uses the shared helper, so shaders that lost a conditional barrier are covered there too.

Vendor reach

Change	Applies to
Subgroup-32 for lost conditional barriers	Intel proprietary Windows only. The pass that sets the flag only runs there.
Pipeline cache flush skip and drain	Intel proprietary Windows only.
Task payload zero plus barrier	All vendors.
ClearChannel()	All vendors.
Mesh/task required subgroup size (#745)	Where requiredSubgroupSizeStages includes the stage: ANV, RADV on GFX10+, others per driver.

Checked against Mesa (ANV, RADV) and the Vulkan spec. ANV reports subgroup sizes 8 to 32 and lists compute, task and mesh as required-size stages. RADV on GFX10+ reports 32 to 64 and accepts 32 or 64.

#541 sets supports_conditional_barriers = true. On my A770 keeping those barriers on Windows causes device loss. I am not doing that here and will revisit it separately.

Assisted-by: Claude Opus 5.5 (debugging, code, testing, review)

Type of change

• Bug fix
• Game compatibility / game-support
• Performance
• New feature
• Refactor / cleanup
• CI / build
• Documentation

Testing

• Tested on Windows
• Tested on Linux
• Tested on macOS
• Tested on NVIDIA
• Tested on AMD (9600X iGPU)
• Tested on Intel (Arc A770 16gb)

Cross-testing on NVIDIA/Linux, Intel/Linux and AMD/Linux is in progress. Results will be posted here. The change most worth watching on other vendors is the task prologue.

Games tested:

Mario Kart World 1.8.0: no more crash on startup or banding
Donkey Kong Bananza 3.2.0: black screen with these changes and with the stock build

Checklist

• I built it locally and it compiles without new warnings
• I did not add any copyrighted game data, keys or firmware
• Docs updated if needed
• I added AI Disclosure if one was used

Related to #130, #706, #66, #745

Fix these before you post it:

• vulkan: Fix shaders that assume 32 wide warps when running on hardware with natively 16 wide warps like Intel #745 wording: the summary describes vulkan: Fix shaders that assume 32 wide warps when running on hardware with natively 16 wide warps like Intel #745's change as included.

Summary by CodeRabbit

• Bug Fixes
  • Improved Vulkan pipeline handling for task and mesh shaders that use warp-wide operations, helping ensure shader stages run with compatible execution settings.
  • Improved pipeline-cache handling on systems where the graphics driver cannot safely read the cache while builds are in progress.
  • Improved cleanup of queued work and graphics state when channels are released, reducing the risk of stale work or state being reused.
  • Added safer handling for shader barriers in conditional control flow.

[github-actions]

Retargeted to nx2-experimental for pre-build testing. It will be promoted to nx2 after validation.

[coderabbitai]

Note

Currently processing new changes in this PR. This may take a few minutes, please wait...

Review info

Run configuration

• Source mode: Auto
• Repository: Unavailable
• Branch: Unavailable
• File: No configuration file from this repository applied
• Commit: Unavailable
• Layers: Repository UI
• Review profile: CHILL
• Plan: Advanced
• Run ID: 5b690628-a4a0-4282-b571-4f22d7757d90

Commits

Reviewing files that changed from the base of the PR and between aff8b25 and 0e82056.

Files selected for processing: 11

• src/common/thread_worker.h
• src/shader_recompiler/backend/spirv/emit_spirv_mesh.cpp
• src/shader_recompiler/ir_opt/conditional_barrier_pass.cpp
• src/shader_recompiler/shader_info.h
• src/video_core/renderer_vulkan/vk_compute_pipeline.cpp
• src/video_core/renderer_vulkan/vk_graphics_pipeline.cpp
• src/video_core/renderer_vulkan/vk_pipeline_cache.cpp
• src/video_core/renderer_vulkan/vk_rasterizer.cpp
• src/video_core/renderer_vulkan/vk_rasterizer.h
• src/video_core/renderer_vulkan/vk_state_tracker.h
• src/video_core/vulkan_common/vulkan_device.h

[kraibse]

Test results

Intel GPU users

• Can now boot the game and don't crash from any of the previously occurring crashes described in vulkan: intel arc fixes for MKW #684
• Visual artifacts remain but the before and after comparison showed these issues already existing. Will be tackled in another PR.

Before

After

Remaining visual artifacts include:

• Blocky shadows under the player character
• Pink structures and landscape

AMD/NVIDIA users

• Report no regression as expected. Changes are guarded and solely limited to intel GPUs.