Skip to content

[mlir][XeGPU][VectorToXeGPU] Use 'xegpu.load' to lower 1D 'vector.transfer_read' for PVC & BMG #168910

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Nov 24, 2025
Merged

[mlir][XeGPU][VectorToXeGPU] Use 'xegpu.load' to lower 1D 'vector.transfer_read' for PVC & BMG #168910

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 6 additions & 1 deletion mlir/lib/Conversion/VectorToXeGPU/VectorToXeGPU.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -519,8 +519,13 @@ struct TransferReadLowering : public OpRewritePattern<vector::TransferReadOp> {
return lowerToScatteredLoadOp(readOp, rewriter);
}

// Perform common data transfer checks.
VectorType vecTy = readOp.getVectorType();

// Lower using load.gather in 1D case
if (vecTy.getRank() == 1 && !readOp.hasOutOfBoundsDim())
return lowerToScatteredLoadOp(readOp, rewriter);

// Perform common data transfer checks.
if (failed(storeLoadPreconditions(rewriter, readOp, vecTy)))
return failure();

Expand Down
82 changes: 63 additions & 19 deletions mlir/test/Conversion/VectorToXeGPU/transfer-read-to-xegpu.mlir
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,14 +11,15 @@ gpu.func @load_1D_vector(%source: memref<8x16x32xf32>, %offset: index) -> vector

// LOAD-ND-LABEL: @load_1D_vector(
// LOAD-ND-SAME: %[[SRC:.+]]: memref<8x16x32xf32>,
// LOAD-ND-SAME: %[[OFFSET:.+]]: index
// LOAD-ND: %[[COLLAPSED:.+]] = memref.subview %[[SRC]][%[[OFFSET]], %[[OFFSET]], 0]
// LOAD-ND: %[[DESC:.+]] = xegpu.create_nd_tdesc
// LOAD-ND-SAME: %[[COLLAPSED]]
// LOAD-ND-SAME: memref<32xf32, strided<[1], offset: ?>> -> !xegpu.tensor_desc<8xf32,
// LOAD-ND-SAME: boundary_check = false
// LOAD-ND: %[[VEC:.+]] = xegpu.load_nd %[[DESC]][%[[OFFSET]]]{{.*}}-> vector<8xf32>
// LOAD-ND: return %[[VEC]]
// LOAD-ND: %[[CST:.+]] = arith.constant dense<true> : vector<8xi1>
// LOAD-ND: %[[STEP:.+]] = vector.step : vector<8xindex>
// LOAD-ND-COUNT2: arith.muli {{.*}} : index
// LOAD-ND-COUNT2: arith.addi {{.*}} : index
// LOAD-ND: %[[SPLAT:.+]] = vector.broadcast {{.*}}: index to vector<8xindex>
// LOAD-ND: %[[IDX:.+]] = arith.addi %[[SPLAT]], %[[STEP]] : vector<8xindex>
// LOAD-ND: %[[COLLAPSE:.+]] = memref.extract_aligned_pointer_as_index %[[SRC]] : memref<8x16x32xf32> -> index
// LOAD-ND: %[[COLLAPSE_I:.+]] = arith.index_cast %[[COLLAPSE]] : index to i64
// LOAD-ND: %[[VEC:.+]] = xegpu.load %[[COLLAPSE_I]]{{\[}}%[[IDX]]{{\]}}, %[[CST]] : i64, vector<8xindex>, vector<8xi1> -> vector<8xf32>

// LOAD-GATHER-LABEL: @load_1D_vector(
// LOAD-GATHER-SAME: %[[SRC:.+]]: memref<8x16x32xf32>,
Expand Down Expand Up @@ -404,27 +405,31 @@ gpu.func @no_load_unsupported_map(%source: memref<16x32x64xf32>,

// -----
gpu.module @xevm_module {
gpu.func @load_from_subview(%source: memref<4096x4096xf16>, %off1: index, %off2: index) -> vector<8xf16> {
gpu.func @load_from_subview_1D(%source: memref<4096x4096xf16>, %off1: index, %off2: index) -> vector<8xf16> {
%c0 = arith.constant 0.0 : f16
%subview = memref.subview %source[%off1, %off2] [256, 256] [1, 1] : memref<4096x4096xf16> to memref<256x256xf16, strided<[4096, 1], offset: ?>>
%0 = vector.transfer_read %subview[%off2, %off2], %c0
{in_bounds = [true]} : memref<256x256xf16, strided<[4096, 1], offset: ?>>, vector<8xf16>
gpu.return %0 : vector<8xf16>
}

// LOAD-ND-LABEL: @load_from_subview(
// LOAD-ND-LABEL: @load_from_subview_1D(
// LOAD-ND-SAME: %[[SRC:.+]]: memref<4096x4096xf16>,
// LOAD-ND-SAME: %[[OFF1:.+]]: index, %[[OFF2:.+]]: index
// LOAD-ND: %[[CST:.+]] = arith.constant dense<true> : vector<8xi1>
// LOAD-ND: %[[SUBVIEW:.+]] = memref.subview %[[SRC]][%[[OFF1]], %[[OFF2]]] [256, 256] [1, 1] : memref<4096x4096xf16> to memref<256x256xf16, strided<[4096, 1], offset: ?>>
// LOAD-ND: %[[COLLAPSED:.+]] = memref.subview %[[SUBVIEW]][%[[OFF2]], 0]
// LOAD-ND: %[[DESC:.+]] = xegpu.create_nd_tdesc
// LOAD-ND-SAME: %[[COLLAPSED]]
// LOAD-ND-SAME: memref<256xf16, strided<[1], offset: ?>> -> !xegpu.tensor_desc<8xf16,
// LOAD-ND-SAME: boundary_check = false
// LOAD-ND: %[[VEC:.+]] = xegpu.load_nd %[[DESC]][%[[OFF2]]]{{.*}}-> vector<8xf16>
// LOAD-ND: return %[[VEC]]

// LOAD-GATHER-LABEL: @load_from_subview(
// LOAD-ND: %[[BB:.+]], %[[OFFSET:.+]],{{.*}},{{.*}} = memref.extract_strided_metadata %[[SUBVIEW]] : memref<256x256xf16, strided<[4096, 1], offset: ?>> -> memref<f16>, index, index, index, index, index
// LOAD-ND: %[[STEP:.+]] = vector.step : vector<8xindex>
// LOAD-ND: arith.muli {{.*}} : index
// LOAD-ND: arith.addi %[[OFFSET]]{{.*}} : index
// LOAD-ND: arith.addi {{.*}} : index
// LOAD-ND: %[[SPLAT:.+]] = vector.broadcast {{.*}}: index to vector<8xindex>
// LOAD-ND: %[[IDX:.+]] = arith.addi %[[SPLAT]], %[[STEP]] : vector<8xindex>
// LOAD-ND: %[[COLLAPSE:.+]] = memref.extract_aligned_pointer_as_index %[[SUBVIEW]] : memref<256x256xf16, strided<[4096, 1], offset: ?>> -> index
// LOAD-ND: %[[COLLAPSE_I:.+]] = arith.index_cast %[[COLLAPSE]] : index to i64
// LOAD-ND: %[[VEC:.+]] = xegpu.load %[[COLLAPSE_I]]{{\[}}%[[IDX]]{{\]}}, %[[CST]] : i64, vector<8xindex>, vector<8xi1> -> vector<8xf16>

// LOAD-GATHER-LABEL: @load_from_subview_1D(
// LOAD-GATHER-SAME: %[[SRC:.+]]: memref<4096x4096xf16>,
// LOAD-GATHER-SAME: %[[OFF1:.+]]: index, %[[OFF2:.+]]: index
// LOAD-GATHER: %[[CST:.+]] = arith.constant dense<true> : vector<8xi1>
Expand All @@ -440,3 +445,42 @@ gpu.func @load_from_subview(%source: memref<4096x4096xf16>, %off1: index, %off2:
// LOAD-GATHER: %[[COLLAPSE_I:.+]] = arith.index_cast %[[COLLAPSE]] : index to i64
// LOAD-GATHER: %[[VEC:.+]] = xegpu.load %[[COLLAPSE_I]]{{\[}}%[[IDX]]{{\]}}, %[[CST]] : i64, vector<8xindex>, vector<8xi1> -> vector<8xf16>
}

// -----
gpu.module @xevm_module {
gpu.func @load_from_subview_2D(%source: memref<4096x4096xf16>, %off1: index, %off2: index) -> vector<8x16xf16> {
%c0 = arith.constant 0.0 : f16
%subview = memref.subview %source[%off1, %off2] [256, 256] [1, 1] : memref<4096x4096xf16> to memref<256x256xf16, strided<[4096, 1], offset: ?>>
%0 = vector.transfer_read %subview[%off2, %off2], %c0
{in_bounds = [true, true]} : memref<256x256xf16, strided<[4096, 1], offset: ?>>, vector<8x16xf16>
gpu.return %0 : vector<8x16xf16>
}

// LOAD-ND-LABEL: @load_from_subview_2D(
// LOAD-ND-SAME: %[[SRC:.+]]: memref<4096x4096xf16>,
// LOAD-ND-SAME: %[[OFF1:.+]]: index, %[[OFF2:.+]]: index
// LOAD-ND: %[[SUBVIEW:.+]] = memref.subview %[[SRC]][%[[OFF1]], %[[OFF2]]] [256, 256] [1, 1] : memref<4096x4096xf16> to memref<256x256xf16, strided<[4096, 1], offset: ?>>
// LOAD-ND: %[[DESC:.+]] = xegpu.create_nd_tdesc
// LOAD-ND-SAME: %[[SUBVIEW]]
// LOAD-ND-SAME: memref<256x256xf16, strided<[4096, 1], offset: ?>> -> !xegpu.tensor_desc<8x16xf16,
// LOAD-ND-SAME: boundary_check = false
// LOAD-ND: %[[VEC:.+]] = xegpu.load_nd %[[DESC]][%[[OFF2]], %[[OFF2]]]{{.*}}-> vector<8x16xf16>
// LOAD-ND: return %[[VEC]]

// LOAD-GATHER-LABEL: @load_from_subview_2D(
// LOAD-GATHER-SAME: %[[SRC:.+]]: memref<4096x4096xf16>,
// LOAD-GATHER-SAME: %[[OFF1:.+]]: index, %[[OFF2:.+]]: index
// LOAD-GATHER: %[[CST:.+]] = arith.constant dense<true> : vector<8x16xi1>
// LOAD-GATHER: %[[SUBVIEW:.+]] = memref.subview %[[SRC]][%[[OFF1]], %[[OFF2]]] [256, 256] [1, 1] : memref<4096x4096xf16> to memref<256x256xf16, strided<[4096, 1], offset: ?>>
// LOAD-GATHER: %[[BB:.+]], %[[OFFSET:.+]],{{.*}},{{.*}} = memref.extract_strided_metadata %[[SUBVIEW]] : memref<256x256xf16, strided<[4096, 1], offset: ?>> -> memref<f16>, index, index, index, index, index
// LOAD-GATHER-COUNT2: vector.step
// LOAD-GATHER-COUNT2: vector.shape_cast
// LOAD-GATHER-COUNT2: vector.broadcast
// LOAD-GATHER-COUNT2: arith.muli {{.*}} : index
// LOAD-GATHER-COUNT2: arith.addi {{.*}} : index
// LOAD-GATHER: %[[SPLAT:.+]] = vector.broadcast {{.*}}: index to vector<8x16xindex>
// LOAD-GATHER: %[[IDX:.+]] = arith.addi %[[SPLAT]], {{.*}} : vector<8x16xindex>
// LOAD-GATHER: %[[COLLAPSE:.+]] = memref.extract_aligned_pointer_as_index %[[SUBVIEW]] : memref<256x256xf16, strided<[4096, 1], offset: ?>> -> index
// LOAD-GATHER: %[[COLLAPSE_I:.+]] = arith.index_cast %[[COLLAPSE]] : index to i64
// LOAD-GATHER: %[[VEC:.+]] = xegpu.load %[[COLLAPSE_I]]{{\[}}%[[IDX]]{{\]}}, %[[CST]] : i64, vector<8x16xindex>, vector<8x16xi1> -> vector<8x16xf16>
}