llvm-project

History

[MLIR] Vector to XeGPU conversion: Use proper source variant for create_nd_tdesc op creation. (#171216 )

If source strided memref is not fully static - at least one of shape,
strides, offset is kDynamic - use i64 source variant.
With this change, xegpu.create_nd_tdesc created by lowering from vector
dialect, can rely on getMixedOffsets, getMixedSize and getMixedStrides
to get relevant values.

2025-12-18 11:03:51 -08:00

contract-to-xegpu.mlir

[mlir][xegpu] Remove vector contract to dpas size restriction (#147470 )

2025-07-09 22:37:06 +02:00

gather-to-xegpu.mlir

[MLIR][XeGPU][VectorToXeGPU] Add lowering from vector.gather/scatter to xegpu.load/store (#158024 )

2025-09-19 11:12:14 +02:00

load-to-xegpu.mlir

[MLIR] Vector to XeGPU conversion: Use proper source variant for create_nd_tdesc op creation. (#171216 )

2025-12-18 11:03:51 -08:00

scatter-to-xegpu.mlir

[MLIR][XeGPU][VectorToXeGPU] Add lowering from vector.gather/scatter to xegpu.load/store (#158024 )