[mlir][vector] proper masking support for contract lowering #67145
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Support all known permutations when lowering masked vector.contract to vector.outerproduct, and not just the canonical permutation.
|
@llvm/pr-subscribers-mlir @llvm/pr-subscribers-mlir-vector ChangesSupport all known permutations when lowering masked vector.contract to vector.outerproduct, and not just the canonical permutation. Full diff: https://github.com/llvm/llvm-project/pull/67145.diff 2 Files Affected:
diff --git a/mlir/lib/Dialect/Vector/Transforms/LowerVectorContract.cpp b/mlir/lib/Dialect/Vector/Transforms/LowerVectorContract.cpp
index 64ab0abda26e640..de7aba6e84b748b 100644
--- a/mlir/lib/Dialect/Vector/Transforms/LowerVectorContract.cpp
+++ b/mlir/lib/Dialect/Vector/Transforms/LowerVectorContract.cpp
@@ -457,38 +457,44 @@ struct UnrolledOuterProductGenerator
// Set up the parallel/reduction structure in the right form.
AffineExpr m, n, k;
bindDims(rewriter.getContext(), m, n, k);
+ Value transposedMask = t(mask, {2, 0, 1});
// Classical row-major matmul: Just permute the lhs.
if (layout({{m, k}, {k, n}, {m, n}}))
- return outerProd(t(lhs), rhs, res, lhsType.getDimSize(1),
- t(mask, {2, 0, 1}));
+ return outerProd(t(lhs), rhs, res, lhsType.getDimSize(1), transposedMask);
// TODO: may be better to fail and use some vector<k> -> scalar reduction.
if (layout({{m, k}, {n, k}, {m, n}})) {
Value tlhs = t(lhs);
- return outerProd(tlhs, t(rhs), res, lhsType.getDimSize(1));
+ return outerProd(tlhs, t(rhs), res, lhsType.getDimSize(1),
+ transposedMask);
}
// No need to permute anything.
if (layout({{k, m}, {k, n}, {m, n}}))
- return outerProd(lhs, rhs, res, lhsType.getDimSize(0));
+ return outerProd(lhs, rhs, res, lhsType.getDimSize(0), transposedMask);
// Just permute the rhs.
if (layout({{k, m}, {n, k}, {m, n}}))
- return outerProd(lhs, t(rhs), res, lhsType.getDimSize(0));
+ return outerProd(lhs, t(rhs), res, lhsType.getDimSize(0), transposedMask);
// Transposed output: swap RHS and LHS.
// Classical row-major matmul: permute the lhs.
if (layout({{m, k}, {k, n}, {n, m}}))
- return outerProd(rhs, t(lhs), res, lhsType.getDimSize(1));
+ return outerProd(rhs, t(lhs), res, lhsType.getDimSize(1), transposedMask);
// TODO: may be better to fail and use some vector<k> -> scalar reduction.
if (layout({{m, k}, {n, k}, {n, m}})) {
Value trhs = t(rhs);
- return outerProd(trhs, t(lhs), res, lhsType.getDimSize(1));
+ return outerProd(trhs, t(lhs), res, lhsType.getDimSize(1),
+ transposedMask);
}
if (layout({{k, m}, {k, n}, {n, m}}))
- return outerProd(rhs, lhs, res, lhsType.getDimSize(0));
+ return outerProd(rhs, lhs, res, lhsType.getDimSize(0), transposedMask);
if (layout({{k, m}, {n, k}, {n, m}}))
- return outerProd(t(rhs), lhs, res, lhsType.getDimSize(0));
+ return outerProd(t(rhs), lhs, res, lhsType.getDimSize(0), transposedMask);
return failure();
}
- /// One outer parallel, one inner reduction (matvec flavor)
+ //
+ // One outer parallel, one inner reduction (matvec flavor).
+ // Mask needs to be transposed everywhere to turn the reduction dimension
+ // outermost as required by outerproduct.
+ //
FailureOr<Value> matvec() {
if (!iters({Par(), Red()}))
return failure();
@@ -500,18 +506,19 @@ struct UnrolledOuterProductGenerator
return outerProd(t(lhs), rhs, res, lhsType.getDimSize(1), t(mask));
// Case mat-trans-vec: ready to go.
if (layout({{k, m}, {k}, {m}}))
- return outerProd(lhs, rhs, res, lhsType.getDimSize(0));
+ return outerProd(lhs, rhs, res, lhsType.getDimSize(0), t(mask));
// Case vec-mat: swap and transpose.
if (layout({{k}, {m, k}, {m}}))
- return outerProd(t(rhs), lhs, res, lhsType.getDimSize(0));
+ return outerProd(t(rhs), lhs, res, lhsType.getDimSize(0), t(mask));
// Case vec-mat-trans: swap and ready to go.
if (layout({{k}, {k, m}, {m}}))
- return outerProd(rhs, lhs, res, lhsType.getDimSize(0));
+ return outerProd(rhs, lhs, res, lhsType.getDimSize(0), t(mask));
return failure();
}
//
- // One outer reduction, one inner parallel (tmatvec flavor)
+ // One outer reduction, one inner parallel (tmatvec flavor).
+ // Mask already has the shape of the outer product.
//
FailureOr<Value> tmatvec() {
if (!iters({Red(), Par()}))
@@ -521,16 +528,16 @@ struct UnrolledOuterProductGenerator
// Case mat-vec: transpose.
if (layout({{m, k}, {k}, {m}}))
- return outerProd(t(lhs), rhs, res, lhsType.getDimSize(1));
+ return outerProd(t(lhs), rhs, res, lhsType.getDimSize(1), mask);
// Case mat-trans-vec: ready to go.
if (layout({{k, m}, {k}, {m}}))
- return outerProd(lhs, rhs, res, lhsType.getDimSize(0));
+ return outerProd(lhs, rhs, res, lhsType.getDimSize(0), mask);
// Case vec-mat: swap and transpose.
if (layout({{k}, {m, k}, {m}}))
- return outerProd(t(rhs), lhs, res, lhsType.getDimSize(0));
+ return outerProd(t(rhs), lhs, res, lhsType.getDimSize(0), mask);
// Case vec-mat-trans: swap and ready to go.
if (layout({{k}, {k, m}, {m}}))
- return outerProd(rhs, lhs, res, lhsType.getDimSize(0));
+ return outerProd(rhs, lhs, res, lhsType.getDimSize(0), mask);
return failure();
}
diff --git a/mlir/test/Dialect/Vector/vector-contract-to-outerproduct-transforms.mlir b/mlir/test/Dialect/Vector/vector-contract-to-outerproduct-transforms.mlir
index aee4cf6b1379bd0..1b4428431a92ba3 100644
--- a/mlir/test/Dialect/Vector/vector-contract-to-outerproduct-transforms.mlir
+++ b/mlir/test/Dialect/Vector/vector-contract-to-outerproduct-transforms.mlir
@@ -341,6 +341,174 @@ func.func @matmul_7(%arg0: vector<2x1xf32>, %arg1: vector<1x3xf32>, %arg2: vecto
return %0 : vector<3x2xf32>
}
+// CHECK-LABEL: @masked_matvec_mk_k_m
+// CHECK-SAME: %[[MAT:.+]]: vector<4x2xf32>
+// CHECK-SAME: %[[VEC:.+]]: vector<2xf32>
+// CHECK-SAME: %[[INIT:.+]]: vector<4xf32>
+// CHECK-SAME: %[[MASK:.+]]: vector<4x2xi1>
+func.func @masked_matvec_mk_k_m(%arg0: vector<4x2xf32>, %arg1: vector<2xf32>, %arg2: vector<4xf32>, %mask: vector<4x2xi1>) -> vector<4xf32> {
+ // CHECK: vector.transpose %[[MAT]]
+ // CHECK: vector.transpose %[[MASK]]
+ // CHECK-COUNT-2: vector.mask %{{.*}} { vector.outerproduct %{{.*}}, %{{.*}}, %{{.*}} {kind = #vector.kind<add>} : vector<4xf32>, f32 }
+ %res = vector.mask %mask {
+ vector.contract {
+ indexing_maps = [affine_map<(m, k) -> (m, k)>,
+ affine_map<(m, k) -> (k)>,
+ affine_map<(m, k) -> (m)>],
+ iterator_types = ["parallel", "reduction"],
+ kind = #vector.kind<add>
+ } %arg0, %arg1, %arg2 : vector<4x2xf32>, vector<2xf32>, vector<4xf32> into vector<4xf32>
+ } : vector<4x2xi1> -> vector<4xf32>
+ return %res : vector<4xf32>
+}
+
+// CHECK-LABEL: @masked_matvec_km_k_m
+// CHECK-SAME: %[[MAT:.+]]: vector<2x4xf32>
+// CHECK-SAME: %[[VEC:.+]]: vector<2xf32>
+// CHECK-SAME: %[[INIT:.+]]: vector<4xf32>
+// CHECK-SAME: %[[MASK:.+]]: vector<4x2xi1>
+func.func @masked_matvec_km_k_m(%arg0: vector<2x4xf32>, %arg1: vector<2xf32>, %arg2: vector<4xf32>, %mask: vector<4x2xi1>) -> vector<4xf32> {
+ // CHECK-NOT: vector.transpose %[[MAT]]
+ // CHECK: vector.transpose %[[MASK]]
+ // CHECK-COUNT-2: vector.mask %{{.*}} { vector.outerproduct %{{.*}}, %{{.*}}, %{{.*}} {kind = #vector.kind<add>} : vector<4xf32>, f32 }
+ %res = vector.mask %mask {
+ vector.contract {
+ indexing_maps = [affine_map<(m, k) -> (k, m)>,
+ affine_map<(m, k) -> (k)>,
+ affine_map<(m, k) -> (m)>],
+ iterator_types = ["parallel", "reduction"],
+ kind = #vector.kind<add>
+ } %arg0, %arg1, %arg2 : vector<2x4xf32>, vector<2xf32>, vector<4xf32> into vector<4xf32>
+ } : vector<4x2xi1> -> vector<4xf32>
+ return %res : vector<4xf32>
+}
+
+// CHECK-LABEL: @masked_matvec_k_mk_m
+// CHECK-SAME: %[[MAT:.+]]: vector<4x2xf32>
+// CHECK-SAME: %[[VEC:.+]]: vector<2xf32>
+// CHECK-SAME: %[[INIT:.+]]: vector<4xf32>
+// CHECK-SAME: %[[MASK:.+]]: vector<4x2xi1>
+func.func @masked_matvec_k_mk_m(%arg0: vector<4x2xf32>, %arg1: vector<2xf32>, %arg2: vector<4xf32>, %mask: vector<4x2xi1>) -> vector<4xf32> {
+ // CHECK: vector.transpose %[[MAT]]
+ // CHECK: vector.transpose %[[MASK]]
+ // CHECK-COUNT-2: vector.mask %{{.*}} { vector.outerproduct %{{.*}}, %{{.*}}, %{{.*}} {kind = #vector.kind<add>} : vector<4xf32>, f32 }
+ %res = vector.mask %mask {
+ vector.contract {
+ indexing_maps = [affine_map<(m, k) -> (k)>,
+ affine_map<(m, k) -> (m, k)>,
+ affine_map<(m, k) -> (m)>],
+ iterator_types = ["parallel", "reduction"],
+ kind = #vector.kind<add>
+ } %arg1, %arg0, %arg2 : vector<2xf32>, vector<4x2xf32>, vector<4xf32> into vector<4xf32>
+ } : vector<4x2xi1> -> vector<4xf32>
+ return %res : vector<4xf32>
+}
+
+// CHECK-LABEL: @masked_matvec_k_km_m
+// CHECK-SAME: %[[MAT:.+]]: vector<2x4xf32>
+// CHECK-SAME: %[[VEC:.+]]: vector<2xf32>
+// CHECK-SAME: %[[INIT:.+]]: vector<4xf32>
+// CHECK-SAME: %[[MASK:.+]]: vector<4x2xi1>
+func.func @masked_matvec_k_km_m(%arg0: vector<2x4xf32>, %arg1: vector<2xf32>, %arg2: vector<4xf32>, %mask: vector<4x2xi1>) -> vector<4xf32> {
+ // CHECK-NOT: vector.transpose %[[MAT]]
+ // CHECK: vector.transpose %[[MASK]]
+ // CHECK-COUNT-2: vector.mask %{{.*}} { vector.outerproduct %{{.*}}, %{{.*}}, %{{.*}} {kind = #vector.kind<add>} : vector<4xf32>, f32 }
+ %res = vector.mask %mask {
+ vector.contract {
+ indexing_maps = [affine_map<(m, k) -> (k)>,
+ affine_map<(m, k) -> (k, m)>,
+ affine_map<(m, k) -> (m)>],
+ iterator_types = ["parallel", "reduction"],
+ kind = #vector.kind<add>
+ } %arg1, %arg0, %arg2 : vector<2xf32>, vector<2x4xf32>, vector<4xf32> into vector<4xf32>
+ } : vector<4x2xi1> -> vector<4xf32>
+ return %res : vector<4xf32>
+}
+
+// CHECK-LABEL: @masked_tmatvec_mk_k_m
+// CHECK-SAME: %[[MAT:.+]]: vector<4x2xf32>
+// CHECK-SAME: %[[VEC:.+]]: vector<2xf32>
+// CHECK-SAME: %[[INIT:.+]]: vector<4xf32>
+// CHECK-SAME: %[[MASK:.+]]: vector<2x4xi1>
+func.func @masked_tmatvec_mk_k_m(%arg0: vector<4x2xf32>, %arg1: vector<2xf32>, %arg2: vector<4xf32>, %mask: vector<2x4xi1>) -> vector<4xf32> {
+ // CHECK: vector.transpose %[[MAT]]
+ // CHECK-NOT: vector.transpose %[[MASK]]
+ // CHECK-COUNT-2: vector.mask %{{.*}} { vector.outerproduct %{{.*}}, %{{.*}}, %{{.*}} {kind = #vector.kind<add>} : vector<4xf32>, f32 }
+ %res = vector.mask %mask {
+ vector.contract {
+ indexing_maps = [affine_map<(k, m) -> (m, k)>,
+ affine_map<(k, m) -> (k)>,
+ affine_map<(k, m) -> (m)>],
+ iterator_types = ["reduction", "parallel"],
+ kind = #vector.kind<add>
+ } %arg0, %arg1, %arg2 : vector<4x2xf32>, vector<2xf32>, vector<4xf32> into vector<4xf32>
+ } : vector<2x4xi1> -> vector<4xf32>
+ return %res : vector<4xf32>
+}
+
+// CHECK-LABEL: @masked_tmatvec_km_k_m
+// CHECK-SAME: %[[MAT:.+]]: vector<2x4xf32>
+// CHECK-SAME: %[[VEC:.+]]: vector<2xf32>
+// CHECK-SAME: %[[INIT:.+]]: vector<4xf32>
+// CHECK-SAME: %[[MASK:.+]]: vector<2x4xi1>
+func.func @masked_tmatvec_km_k_m(%arg0: vector<2x4xf32>, %arg1: vector<2xf32>, %arg2: vector<4xf32>, %mask: vector<2x4xi1>) -> vector<4xf32> {
+ // CHECK-NOT: vector.transpose %[[MAT]]
+ // CHECK-NOT: vector.transpose %[[MASK]]
+ // CHECK-COUNT-2: vector.mask %{{.*}} { vector.outerproduct %{{.*}}, %{{.*}}, %{{.*}} {kind = #vector.kind<add>} : vector<4xf32>, f32 }
+ %res = vector.mask %mask {
+ vector.contract {
+ indexing_maps = [affine_map<(k, m) -> (k, m)>,
+ affine_map<(k, m) -> (k)>,
+ affine_map<(k, m) -> (m)>],
+ iterator_types = ["reduction", "parallel"],
+ kind = #vector.kind<add>
+ } %arg0, %arg1, %arg2 : vector<2x4xf32>, vector<2xf32>, vector<4xf32> into vector<4xf32>
+ } : vector<2x4xi1> -> vector<4xf32>
+ return %res : vector<4xf32>
+}
+
+// CHECK-LABEL: @masked_tmatvec_k_mk_m
+// CHECK-SAME: %[[MAT:.+]]: vector<4x2xf32>
+// CHECK-SAME: %[[VEC:.+]]: vector<2xf32>
+// CHECK-SAME: %[[INIT:.+]]: vector<4xf32>
+// CHECK-SAME: %[[MASK:.+]]: vector<2x4xi1>
+func.func @masked_tmatvec_k_mk_m(%arg0: vector<4x2xf32>, %arg1: vector<2xf32>, %arg2: vector<4xf32>, %mask: vector<2x4xi1>) -> vector<4xf32> {
+ // CHECK: vector.transpose %[[MAT]]
+ // CHECK-NOT: vector.transpose %[[MASK]]
+ // CHECK-COUNT-2: vector.mask %{{.*}} { vector.outerproduct %{{.*}}, %{{.*}}, %{{.*}} {kind = #vector.kind<add>} : vector<4xf32>, f32 }
+ %res = vector.mask %mask {
+ vector.contract {
+ indexing_maps = [affine_map<(k, m) -> (k)>,
+ affine_map<(k, m) -> (m, k)>,
+ affine_map<(k, m) -> (m)>],
+ iterator_types = ["reduction", "parallel"],
+ kind = #vector.kind<add>
+ } %arg1, %arg0, %arg2 : vector<2xf32>, vector<4x2xf32>, vector<4xf32> into vector<4xf32>
+ } : vector<2x4xi1> -> vector<4xf32>
+ return %res : vector<4xf32>
+}
+
+// CHECK-LABEL: @masked_tmatvec_k_km_m
+// CHECK-SAME: %[[MAT:.+]]: vector<2x4xf32>
+// CHECK-SAME: %[[VEC:.+]]: vector<2xf32>
+// CHECK-SAME: %[[INIT:.+]]: vector<4xf32>
+// CHECK-SAME: %[[MASK:.+]]: vector<2x4xi1>
+func.func @masked_tmatvec_k_km_m(%arg0: vector<2x4xf32>, %arg1: vector<2xf32>, %arg2: vector<4xf32>, %mask: vector<2x4xi1>) -> vector<4xf32> {
+ // CHECK-NOT: vector.transpose %[[MAT]]
+ // CHECK-NOT: vector.transpose %[[MASK]]
+ // CHECK-COUNT-2: vector.mask %{{.*}} { vector.outerproduct %{{.*}}, %{{.*}}, %{{.*}} {kind = #vector.kind<add>} : vector<4xf32>, f32 }
+ %res = vector.mask %mask {
+ vector.contract {
+ indexing_maps = [affine_map<(k, m) -> (k)>,
+ affine_map<(k, m) -> (k, m)>,
+ affine_map<(k, m) -> (m)>],
+ iterator_types = ["reduction", "parallel"],
+ kind = #vector.kind<add>
+ } %arg1, %arg0, %arg2 : vector<2xf32>, vector<2x4xf32>, vector<4xf32> into vector<4xf32>
+ } : vector<2x4xi1> -> vector<4xf32>
+ return %res : vector<4xf32>
+}
+
transform.sequence failures(propagate) {
^bb1(%module_op: !transform.any_op):
|
nicolasvasilache
approved these changes
Sep 22, 2023
Calling t() constructs IR, so it should not happen from multiple function arguments that will be executed in different order on Windows and Linux.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Support all known permutations when lowering masked vector.contract to vector.outerproduct, and not just the canonical permutation.