Jakub Kuderski bdb5c9c3fa

[mlir][spirv][gpu] Convert remaining wmma ops to KHR coop matrix (#66455) ...

These do not produce extension-specific ops and are handled via common
patterns for both the KHR and the NV coop matrix extension.

Also improve match failure reporting and error handling in type
conversion.

2023-09-19 13:04:34 -04:00

..

builtins-opencl.mlir

[mlir][spirv] Fix Physical32/Physical64 support for OpenCL

2023-02-27 06:22:59 +00:00

builtins-vulkan.mlir

[mlir][spirv] Fix Physical32/Physical64 support for OpenCL

2023-02-27 06:22:59 +00:00

entry-point.mlir

[mlir][spirv] Allow controlling subgroup size

2022-11-30 12:34:09 -05:00

gpu-to-spirv.mlir

[mlir][spirv] Allow controlling subgroup size

2022-11-30 12:34:09 -05:00

load-store.mlir

[mlir][spirv] Allow controlling subgroup size

2022-11-30 12:34:09 -05:00

module-opencl.mlir

[mlir][spirv] Fix Physical32/Physical64 support for OpenCL

2023-02-27 06:22:59 +00:00

reductions.mlir

[mlir][gpu][spirv] Gpu min/max reductions lowering to SPIR-V

2023-01-29 12:44:33 +01:00

shuffle.mlir

[mlir][spirv] Allow controlling subgroup size

2022-11-30 12:34:09 -05:00

wmma-ops-to-spirv-khr-coop-matrix.mlir

[mlir][spirv][gpu] Convert remaining wmma ops to KHR coop matrix (#66455)

2023-09-19 13:04:34 -04:00

wmma-ops-to-spirv-nv-coop-matrix.mlir

[mlir][spirv][gpu] Add conversion for load/store/mad coop matrix ops (#66311)

2023-09-14 22:16:27 -04:00