[SCF][Transform] Add support for scf.for in LoopFuseSibling op #81495
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
|
Thank you for submitting a Pull Request (PR) to the LLVM Project! This PR will be automatically labeled and the relevant teams will be If you wish to, you can add reviewers by using the "Reviewers" section on this page. If this is not working for you, it is probably because you do not have write If you have received no comments on your PR for a week, you can request a review If you have further questions, they may be answered by the LLVM GitHub User Guide. You can also ask questions in a comment on this PR, on the LLVM Discord or on the forums. |
|
@llvm/pr-subscribers-mlir-scf @llvm/pr-subscribers-mlir Author: Rolf Morel (rolfmorel) ChangesAdds support for fusing two scf.for loops occurring in the same block. Implementation mirrors that of LoopFuseSibling's support for scf.forall, including only rudimentary checks, like the target loop's operands being dominated by the source loop. Fixes a bug in the dominance check whereby it was checked that values in the target loop themselves dominated the source loop rather than (the ops) where these values originate. Adds tests for using LoopFuseSibling on scf.for loops, including one which fails without the fix for the dominance check. Patch is 23.12 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/81495.diff 5 Files Affected:
diff --git a/mlir/include/mlir/Dialect/SCF/TransformOps/SCFTransformOps.td b/mlir/include/mlir/Dialect/SCF/TransformOps/SCFTransformOps.td
index cef73689c072b8..89d32ebcc24b10 100644
--- a/mlir/include/mlir/Dialect/SCF/TransformOps/SCFTransformOps.td
+++ b/mlir/include/mlir/Dialect/SCF/TransformOps/SCFTransformOps.td
@@ -342,11 +342,13 @@ def LoopFuseSibling : Op<Transform_Dialect, "loop.fuse_sibling",
Fuses the `target` loop into the `source` loop assuming they are
independent of each other. It is the responsibility of the user to ensure
that the given two loops are independent of each other, this operation will
- not performa any legality checks and will simply fuse the two given loops.
+ not perform any legality checks and will simply fuse the two given loops.
- Currently, the only fusion supported is when both `target` and `source`
- are `scf.forall` operations. For `scf.forall` fusion, the bounds and the
- mapping must match, otherwise a silencable failure is produced.
+ Currently, fusion is only supported in case both `target` and `source` are
+ `scf.for` operations or both are `scf.forall` operations. For `scf.for`
+ fusion the bounds and step size must match. For `scf.forall` fusion the
+ bounds and the mapping must match. Otherwise a silencable failure is
+ produced.
The input handles `target` and `source` must map to exactly one operation,
a definite failure is produced otherwise.
diff --git a/mlir/include/mlir/Dialect/SCF/Utils/Utils.h b/mlir/include/mlir/Dialect/SCF/Utils/Utils.h
index 9bdd6eb833876f..883d11bcc4df06 100644
--- a/mlir/include/mlir/Dialect/SCF/Utils/Utils.h
+++ b/mlir/include/mlir/Dialect/SCF/Utils/Utils.h
@@ -162,6 +162,16 @@ scf::ForallOp fuseIndependentSiblingForallLoops(scf::ForallOp target,
scf::ForallOp source,
RewriterBase &rewriter);
+/// Given two scf.for loops, `target` and `source`, fuses `target` into
+/// `source`. Assumes that the given loops are siblings and are independent of
+/// each other.
+///
+/// This function does not perform any legality checks and simply fuses the
+/// loops. The caller is responsible for ensuring that the loops are legal to
+/// fuse.
+scf::ForOp fuseIndependentSiblingForLoops(scf::ForOp target, scf::ForOp source,
+ RewriterBase &rewriter);
+
} // namespace mlir
#endif // MLIR_DIALECT_SCF_UTILS_UTILS_H_
diff --git a/mlir/lib/Dialect/SCF/TransformOps/SCFTransformOps.cpp b/mlir/lib/Dialect/SCF/TransformOps/SCFTransformOps.cpp
index bc2fe5772af9d6..7056185aeb456d 100644
--- a/mlir/lib/Dialect/SCF/TransformOps/SCFTransformOps.cpp
+++ b/mlir/lib/Dialect/SCF/TransformOps/SCFTransformOps.cpp
@@ -441,8 +441,11 @@ static DiagnosedSilenceableFailure isOpSibling(Operation *target,
bool failed = false;
OpOperand *failedValue = nullptr;
visitUsedValuesDefinedAbove(target->getRegions(), [&](OpOperand *operand) {
- if (!domInfo.properlyDominates(operand->getOwner(), source,
- /*enclosingOpOk=*/false)) {
+ Operation *operandOp = operand->get().getDefiningOp();
+ if (operandOp && !domInfo.properlyDominates(operandOp, source,
+ /*enclosingOpOk=*/false)) {
+ // `operand` is not a block argument and its defining op does not
+ // dominate `source`
failed = true;
failedValue = operand;
}
@@ -476,15 +479,34 @@ static bool isForallWithIdenticalConfiguration(Operation *target,
targetOp.getMapping() == sourceOp.getMapping();
}
-/// Fuse `target` into `source` assuming they are siblings and indepndent.
-/// TODO: Add fusion for more operations. Currently, we handle only scf.forall.
+static bool isForWithIdenticalConfiguration(Operation *target,
+ Operation *source) {
+ auto targetOp = dyn_cast<scf::ForOp>(target);
+ auto sourceOp = dyn_cast<scf::ForOp>(source);
+ if (!targetOp || !sourceOp)
+ return false;
+
+ return targetOp.getLowerBound() == sourceOp.getLowerBound() &&
+ targetOp.getUpperBound() == sourceOp.getUpperBound() &&
+ targetOp.getStep() == sourceOp.getStep();
+}
+
+/// Fuse `target` into `source` assuming they are siblings and independent.
+/// TODO: Support fusion for operations besides scf.for and scf.forall.
static Operation *fuseSiblings(Operation *target, Operation *source,
RewriterBase &rewriter) {
- auto targetOp = dyn_cast<scf::ForallOp>(target);
- auto sourceOp = dyn_cast<scf::ForallOp>(source);
- if (!targetOp || !sourceOp)
- return nullptr;
- return fuseIndependentSiblingForallLoops(targetOp, sourceOp, rewriter);
+ auto targetForOp = dyn_cast<scf::ForOp>(target);
+ auto sourceForOp = dyn_cast<scf::ForOp>(source);
+ if (targetForOp && sourceForOp)
+ return fuseIndependentSiblingForLoops(targetForOp, sourceForOp, rewriter);
+
+ auto targetForallOp = dyn_cast<scf::ForallOp>(target);
+ auto sourceForallOp = dyn_cast<scf::ForallOp>(source);
+ if (targetForallOp && sourceForallOp)
+ return fuseIndependentSiblingForallLoops(targetForallOp, sourceForallOp,
+ rewriter);
+
+ return nullptr;
}
DiagnosedSilenceableFailure
@@ -511,7 +533,8 @@ transform::LoopFuseSibling::apply(transform::TransformRewriter &rewriter,
return diag;
// Check if the target can be fused into source.
- if (!isForallWithIdenticalConfiguration(target, source)) {
+ if (!isForallWithIdenticalConfiguration(target, source) &&
+ !isForWithIdenticalConfiguration(target, source)) {
return emitSilenceableFailure(target->getLoc())
<< "operations cannot be fused";
}
diff --git a/mlir/lib/Dialect/SCF/Utils/Utils.cpp b/mlir/lib/Dialect/SCF/Utils/Utils.cpp
index cdd85ddeb93add..f5836edf5eeb59 100644
--- a/mlir/lib/Dialect/SCF/Utils/Utils.cpp
+++ b/mlir/lib/Dialect/SCF/Utils/Utils.cpp
@@ -970,3 +970,69 @@ scf::ForallOp mlir::fuseIndependentSiblingForallLoops(scf::ForallOp target,
return fusedLoop;
}
+
+scf::ForOp mlir::fuseIndependentSiblingForLoops(scf::ForOp target,
+ scf::ForOp source,
+ RewriterBase &rewriter) {
+ // Create fused init_args.
+ auto targetInitArgs = target.getInitArgs();
+ auto sourceInitArgs = source.getInitArgs();
+ SmallVector<Value> fusedInitArgs;
+ fusedInitArgs.reserve(targetInitArgs.size() + sourceInitArgs.size());
+ fusedInitArgs.append(sourceInitArgs.begin(), sourceInitArgs.end());
+ fusedInitArgs.append(targetInitArgs.begin(), targetInitArgs.end());
+
+ // Create a new scf::for op after the source loop.
+ rewriter.setInsertionPointAfter(source);
+ scf::ForOp fusedLoop = rewriter.create<scf::ForOp>(
+ source.getLoc(), source.getLowerBound(), source.getUpperBound(),
+ source.getStep(), fusedInitArgs);
+
+ SmallVector<Value> yieldResults;
+
+ // First merge source loop into the new (fused) for loop and then target loop.
+ rewriter.setInsertionPointToEnd(fusedLoop.getBody());
+ for (auto loopAndInitArgsBegin :
+ {std::pair(source, (unsigned int)0),
+ std::pair(target, source.getNumRegionIterArgs())}) {
+ auto origLoop = loopAndInitArgsBegin.first;
+ IRMapping mapping;
+
+ mapping.map(origLoop.getInductionVar(), fusedLoop.getInductionVar());
+ for (size_t i = 0; i < origLoop.getNumRegionIterArgs(); ++i) {
+ mapping.map(
+ origLoop.getRegionIterArgs()[i],
+ fusedLoop.getRegionIterArgs()[loopAndInitArgsBegin.second + i]);
+ }
+
+ for (Operation &op : origLoop.getBody()->getOperations()) {
+ rewriter.clone(op, mapping);
+ }
+
+ if (origLoop.getNumResults() > 0) {
+ scf::YieldOp yieldFromOrigLoop =
+ cast<scf::YieldOp>(fusedLoop.getBody()->getTerminator());
+ yieldResults.append(yieldFromOrigLoop.getOperands().begin(),
+ yieldFromOrigLoop.getOperands().end());
+ rewriter.eraseOp(yieldFromOrigLoop);
+ }
+ }
+
+ // Construct combined YieldOp
+ rewriter.setInsertionPointToEnd(fusedLoop.getBody());
+ rewriter.create<scf::YieldOp>(source.getLoc(), yieldResults);
+
+ // Replace all uses of the old loops with the fused loop.
+ unsigned numSourceOuts = source.getNumResults();
+ rewriter.replaceAllUsesWith(source.getResults(),
+ fusedLoop.getResults().slice(0, numSourceOuts));
+ rewriter.replaceAllUsesWith(
+ target.getResults(),
+ fusedLoop.getResults().slice(numSourceOuts, target.getNumResults()));
+
+ // Erase the old loops.
+ rewriter.eraseOp(target);
+ rewriter.eraseOp(source);
+
+ return fusedLoop;
+}
\ No newline at end of file
diff --git a/mlir/test/Dialect/SCF/transform-loop-fuse-sibling.mlir b/mlir/test/Dialect/SCF/transform-loop-fuse-sibling.mlir
index faaa2db3aa57de..332caf9cdf0516 100644
--- a/mlir/test/Dialect/SCF/transform-loop-fuse-sibling.mlir
+++ b/mlir/test/Dialect/SCF/transform-loop-fuse-sibling.mlir
@@ -1,6 +1,6 @@
// RUN: mlir-opt %s -transform-interpreter --cse --canonicalize -split-input-file -verify-diagnostics | FileCheck %s
-func.func @test(%A : tensor<128x128xf32>, %B1 : tensor<128x128xf32>, %B2 : tensor<128x128xf32>) -> (tensor<128x128xf32>, tensor<128x128xf32>) {
+func.func @matmul_fuse_1st_forall_into_2nd(%A : tensor<128x128xf32>, %B1 : tensor<128x128xf32>, %B2 : tensor<128x128xf32>) -> (tensor<128x128xf32>, tensor<128x128xf32>) {
%zero = arith.constant 0.0 : f32
%out_alloc = tensor.empty() : tensor<128x128xf32>
%out = linalg.fill ins(%zero : f32) outs(%out_alloc : tensor<128x128xf32>) -> tensor<128x128xf32>
@@ -38,7 +38,7 @@ module attributes {transform.with_named_sequence} {
// -----
-func.func @test(%A : tensor<128x128xf32>, %B1 : tensor<128x128xf32>, %B2 : tensor<128x128xf32>) -> (tensor<128x128xf32>, tensor<128x128xf32>) {
+func.func @matmul_fuse_1st_forall_into_2nd_err(%A : tensor<128x128xf32>, %B1 : tensor<128x128xf32>, %B2 : tensor<128x128xf32>) -> (tensor<128x128xf32>, tensor<128x128xf32>) {
%zero = arith.constant 0.0 : f32
%out_alloc = tensor.empty() : tensor<128x128xf32>
%out = linalg.fill ins(%zero : f32) outs(%out_alloc : tensor<128x128xf32>) -> tensor<128x128xf32>
@@ -66,7 +66,7 @@ module attributes {transform.with_named_sequence} {
// -----
-func.func @test(%A : tensor<128x128xf32>, %B1 : tensor<128x128xf32>, %B2 : tensor<128x128xf32>) -> (tensor<128x128xf32>, tensor<128x128xf32>) {
+func.func @matmul_fuse_2nd_forall_into_1st_err(%A : tensor<128x128xf32>, %B1 : tensor<128x128xf32>, %B2 : tensor<128x128xf32>) -> (tensor<128x128xf32>, tensor<128x128xf32>) {
%zero = arith.constant 0.0 : f32
%out_alloc = tensor.empty() : tensor<128x128xf32>
%out = linalg.fill ins(%zero : f32) outs(%out_alloc : tensor<128x128xf32>) -> tensor<128x128xf32>
@@ -94,7 +94,7 @@ module attributes {transform.with_named_sequence} {
// -----
-func.func @test(%A : tensor<128x128xf32>, %B1 : tensor<128x128xf32>, %B2 : tensor<128x128xf32>) -> (tensor<128x128xf32>, tensor<128x128xf32>) {
+func.func @matmul_fuse_2nd_forall_into_1st_err(%A : tensor<128x128xf32>, %B1 : tensor<128x128xf32>, %B2 : tensor<128x128xf32>) -> (tensor<128x128xf32>, tensor<128x128xf32>) {
%zero = arith.constant 0.0 : f32
%out_alloc = tensor.empty() : tensor<128x128xf32>
%out = linalg.fill ins(%zero : f32) outs(%out_alloc : tensor<128x128xf32>) -> tensor<128x128xf32>
@@ -119,3 +119,184 @@ module attributes {transform.with_named_sequence} {
transform.yield
}
}
+
+// -----
+
+// CHECK: func.func @test([[A:%.*]]: {{.*}}, [[B1:%.*]]: {{.*}}, [[B2:%.*]]: {{.*}} {{.*}}
+func.func @test(%A : tensor<128x128xf32>, %B1 : tensor<128x128xf32>, %B2 : tensor<128x128xf32>) -> (tensor<128x128xf32>, tensor<128x128xf32>) {
+ %zero = arith.constant 0.0 : f32
+ %out_alloc = tensor.empty() : tensor<128x128xf32>
+ %out = linalg.fill ins(%zero : f32) outs(%out_alloc : tensor<128x128xf32>) -> tensor<128x128xf32>
+
+ // CHECK-DAG: [[C0:%.*]] = arith.constant 0 : index
+ // CHECK-DAG: [[C32:%.*]] = arith.constant 32 : index
+ // CHECK-DAG: [[C128:%.*]] = arith.constant 128 : index
+ // CHECK-DAG: [[ZERO:%.*]] = arith.constant 0.000000e+00 : f32
+ // CHECK-DAG: [[EMPTY:%.*]] = tensor.empty() : tensor<128x128xf32>
+ // CHECK-DAG: [[BUF:%.*]] = linalg.fill ins([[ZERO]] : {{.*}}) outs([[EMPTY]] : {{.*}}) {{.*}}
+ // CHECK: [[RST:%.*]]:2 = scf.for [[IV:%.*]] = [[C0]] to [[C128]] step [[C32]] iter_args([[IA0:%.*]] = [[BUF]], [[IA1:%.*]] = [[BUF]]) {{.*}}
+ // CHECK-DAG: [[ASLICE:%.*]] = tensor.extract_slice [[A]][[[IV]], 0] [32, 128] [1, 1]
+ // CHECK-DAG: [[SLICE0:%.*]] = tensor.extract_slice [[IA0]][[[IV]], 0] [32, 128] [1, 1]
+ // CHECK: [[OUT1:%.*]] = linalg.matmul ins([[ASLICE]], [[B1]] : {{.*}}) outs([[SLICE0]]
+ // CHECK-NEXT: [[INS0:%.*]] = tensor.insert_slice [[OUT1]] into [[IA0]][[[IV]], 0] [32, 128] [1, 1]
+ // CHECK-DAG: [[SLICE1:%.*]] = tensor.extract_slice [[IA1]][[[IV]], 0] [32, 128] [1, 1]
+ // CHECK: [[OUT2:%.*]] = linalg.matmul ins([[ASLICE]], [[B2]] : {{.*}}) outs([[SLICE1]]
+ // CHECK-NEXT: [[INS1:%.*]] = tensor.insert_slice [[OUT2]] into [[IA1]][[[IV]], 0] [32, 128] [1, 1]
+ // CHECK: scf.yield [[INS0]], [[INS1]] : {{.*}}
+ %out1 = linalg.matmul ins(%A, %B1 : tensor<128x128xf32>, tensor<128x128xf32>) outs(%out : tensor<128x128xf32>) -> tensor<128x128xf32>
+ %out2 = linalg.matmul ins(%A, %B2 : tensor<128x128xf32>, tensor<128x128xf32>) outs(%out : tensor<128x128xf32>) -> tensor<128x128xf32>
+
+ // CHECK: return [[RST]]#0, [[RST]]#1 : {{.*}}
+ func.return %out1, %out2 : tensor<128x128xf32>, tensor<128x128xf32>
+}
+
+module attributes {transform.with_named_sequence} {
+ transform.named_sequence @__transform_main(%variant_op : !transform.any_op {transform.readonly}) {
+ %matched = transform.structured.match ops{["linalg.matmul"]} in %variant_op : (!transform.any_op) -> (!transform.any_op)
+
+ %mm1, %mm2 = transform.split_handle %matched : (!transform.any_op) -> (!transform.any_op, !transform.any_op)
+
+ %tiled_mm1, %loop1 = transform.structured.tile_using_for %mm1 [32] : (!transform.any_op) -> (!transform.any_op, !transform.any_op)
+ %tiled_mm2, %loop2 = transform.structured.tile_using_for %mm2 [32] : (!transform.any_op) -> (!transform.any_op, !transform.any_op)
+ //transform.print %variant_op : !transform.any_op
+
+ %func = transform.structured.match ops{["func.func"]} in %variant_op : (!transform.any_op) -> (!transform.any_op)
+ %cse_func = transform.apply_registered_pass "cse" to %func : (!transform.any_op) -> (!transform.any_op)
+ %for_loops = transform.structured.match ops{["scf.for"]} in %cse_func : (!transform.any_op) -> (!transform.any_op)
+ %for_loop1, %for_loop2 = transform.split_handle %for_loops : (!transform.any_op) -> (!transform.any_op, !transform.any_op)
+ %fused_loop = transform.loop.fuse_sibling %for_loop2 into %for_loop1 : (!transform.any_op, !transform.any_op) -> !transform.any_op
+ transform.yield
+ }
+}
+
+// -----
+
+// CHECK: func.func @test([[A:%.*]]: {{.*}}, [[B1:%.*]]: {{.*}}, [[B2:%.*]]: {{.*}} {{.*}}
+func.func @test(%A : tensor<128x128xf32>, %B1 : tensor<128x128xf32>, %B2 : tensor<128x128xf32>) -> (tensor<128x128xf32>, tensor<128x128xf32>) {
+ %zero = arith.constant 0.0 : f32
+ %out_alloc = tensor.empty() : tensor<128x128xf32>
+ %out = linalg.fill ins(%zero : f32) outs(%out_alloc : tensor<128x128xf32>) -> tensor<128x128xf32>
+
+ // CHECK-DAG: [[C0:%.*]] = arith.constant 0 : index
+ // CHECK-DAG: [[C32:%.*]] = arith.constant 32 : index
+ // CHECK-DAG: [[C128:%.*]] = arith.constant 128 : index
+ // CHECK-DAG: [[ZERO:%.*]] = arith.constant 0.000000e+00 : f32
+ // CHECK-DAG: [[EMPTY:%.*]] = tensor.empty() : tensor<128x128xf32>
+ // CHECK-DAG: [[BUF:%.*]] = linalg.fill ins([[ZERO]] : {{.*}}) outs([[EMPTY]] : {{.*}}) {{.*}}
+ // CHECK: [[RST:%.*]]:2 = scf.for [[IV:%.*]] = [[C0]] to [[C128]] step [[C32]] iter_args([[IA0:%.*]] = [[BUF]], [[IA1:%.*]] = [[BUF]]) {{.*}}
+ // CHECK-DAG: [[ASLICE:%.*]] = tensor.extract_slice [[A]][[[IV]], 0] [32, 128] [1, 1]
+ // CHECK-DAG: [[SLICE0:%.*]] = tensor.extract_slice [[IA0]][[[IV]], 0] [32, 128] [1, 1]
+ // CHECK: [[OUT1:%.*]] = linalg.matmul ins([[ASLICE]], [[B2]] : {{.*}}) outs([[SLICE0]]
+ // CHECK-NEXT: [[INS0:%.*]] = tensor.insert_slice [[OUT1]] into [[IA0]][[[IV]], 0] [32, 128] [1, 1]
+ // CHECK-DAG: [[SLICE1:%.*]] = tensor.extract_slice [[IA1]][[[IV]], 0] [32, 128] [1, 1]
+ // CHECK: [[OUT2:%.*]] = linalg.matmul ins([[ASLICE]], [[B1]] : {{.*}}) outs([[SLICE1]]
+ // CHECK-NEXT: [[INS1:%.*]] = tensor.insert_slice [[OUT2]] into [[IA1]][[[IV]], 0] [32, 128] [1, 1]
+ // CHECK: scf.yield [[INS0]], [[INS1]] : {{.*}}
+ %out1 = linalg.matmul ins(%A, %B1 : tensor<128x128xf32>, tensor<128x128xf32>) outs(%out : tensor<128x128xf32>) -> tensor<128x128xf32>
+ %out2 = linalg.matmul ins(%A, %B2 : tensor<128x128xf32>, tensor<128x128xf32>) outs(%out : tensor<128x128xf32>) -> tensor<128x128xf32>
+
+ // CHECK: return [[RST]]#1, [[RST]]#0 : {{.*}}
+ func.return %out1, %out2 : tensor<128x128xf32>, tensor<128x128xf32>
+}
+
+module attributes {transform.with_named_sequence} {
+ transform.named_sequence @__transform_main(%variant_op : !transform.any_op {transform.readonly}) {
+ %matched = transform.structured.match ops{["linalg.matmul"]} in %variant_op : (!transform.any_op) -> (!transform.any_op)
+
+ %mm1, %mm2 = transform.split_handle %matched : (!transform.any_op) -> (!transform.any_op, !transform.any_op)
+
+ %tiled_mm1, %loop1 = transform.structured.tile_using_for %mm1 [32] : (!transform.any_op) -> (!transform.any_op, !transform.any_op)
+ %tiled_mm2, %loop2 = transform.structured.tile_using_for %mm2 [32] : (!transform.any_op) -> (!transform.any_op, !transform.any_op)
+
+ %func = transform.structured.match ops{["func.func"]} in %variant_op : (!transform.any_op) -> (!transform.any_op)
+ %cse_func = transform.apply_registered_pass "cse" to %func : (!transform.any_op) -> (!transform.any_op)
+ %for_loops = transform.structured.match ops{["scf.for"]} in %cse_func : (!transform.any_op) -> (!transform.any_op)
+ %for_loop1, %for_loop2 = transform.split_handle %for_loops : (!transform.any_op) -> (!transform.any_op, !transform.any_op)
+ %fused_loop = transform.loop.fuse_sibling %for_loop1 into %for_loop2 : (!transform.any_op, !transform.any_op) -> !transform.any_op
+ transform.yield
+ }
+}
+
+// -----
+
+// transform.loop.fuse_sibling used to silently fail on the following due to a bug in the dominance check
+
+// CHECK: func.func @no_dominance_bug([[A:%.*]]: {{.*}}, [[B:%.*]]: {{.*}}
+func.func @no_dominance_bug(%arg1: tensor<128xf32>, %arg2: tensor<128xf32>) -> (tensor<128xf32>, tensor<128xf32>) {
+ // CHECK-DAG: [[C0:%.*]] = arith.constant 0 : index
+ // CHECK-DAG: [[C16:%.*]] = arith.constant 16 : index
+ // CHECK-DAG: [[C128:%.*]] = arith.constant 128 : index
+ // CHECK-DAG: [[ZERO:%.*]] = arith.constant 0.000000e+00 : f32
+ %c0 = arith.constant 0 : index
+ %c16 = arith.constant 16 : index
+ %c128 = arith.constant 128 : index
+ %cst = arith.constant 0.000000e+00 : f32
+ // CHECK: [[RST:%.*]]:2 = scf.for [[IV:%.*]] = [[C0]] to [[C128]] step [[C16]] iter_args([[IB0:%.*]] = [[B]], [[IB1:%.*]] = [[B]]) {{.*}}
+ %1 = scf.for %arg3 = %c0 to %c128 step %c16 iter_args(%arg4 = %arg2) -> (tensor<128xf32>) {
+ // CHECK-DAG: [[ASLICE:%.*]] = vector.transfer_read [[A]][[[IV]]], [[ZERO]]
+ // CHECK-DAG: [[SLICE0:%.*]] = vector.transfer_read [[IB0]][[[IV]]], [[ZERO]]
+ // CHECK: [[OUT1:%.*]] = arith.addf [[SLICE0]], [[ASLICE]]
+ // CHECK-NEXT: [[WRT0:%.*]] = vector.transfer_write [[OUT1]], [[IB0]][[[IV]]]
+ %2 = vector.transfer_read %arg1[%arg3], %cst {in_bounds = [true]} : tensor<128xf32>, vector<16xf32>
+ %3 = vector.transfer_read %arg4[%arg3], %cst {in_bounds = [true]} : tensor<128xf32>, vector<16xf32>
+ %5 = arith.addf %3, %2 : vector<16xf32>
+ %6 = vector.transfer_write %5, %arg4[%arg3] {in_bounds = [true]} : vector<16xf32>, tensor<128xf32>
+ scf.yield %6 : tensor<128xf32>
+ }
+ %dup1 = scf.for %arg3 = %c0 to %c128 step %c16 iter_args(%arg4 = %arg2) -> (tensor<128xf32>) {
+ // CHECK-DAG: [[SLICE1:%.*]] = vector.transfer_read [[IB1]][[[IV]]], [[ZERO]]
+ // CHECK...
[truncated]
|
|
@Groverkss, @ftynse, @banach-space, @matthias-springer, as you guys have worked on |
|
Self-nit: pretty much all of the Transform dialect operations have the class that implements a transform end in the suffix Should I change the name of the class from |
rolfmorel
force-pushed
the
fuseloopop
branch
from
70888b0 to
51d524a
Compare
rolfmorel
changed the title
Yes. |
Groverkss
reviewed
Feb 13, 2024
Groverkss
reviewed
Feb 13, 2024
Groverkss
reviewed
Feb 13, 2024
rolfmorel
force-pushed
the
fuseloopop
branch
3 times, most recently
from
4e04706 to
69e5620
Compare
|
✅ With the latest revision this PR passed the C/C++ code formatter. |
|
✅ With the latest revision this PR passed the Python code formatter. |
|
Hi @ftynse, @Groverkss (and @nicolasvasilache) Thank you for the comments and suggestions. After a long delay I have now done a thorough clean-up of the PR. I think it's now in a good state. If you could have another look, that would be much appreciated! |
Adds support for fusing two scf.for loops occurring in the same block. Uses the rudimentary checks already in place for scf.for_all (like the target loop's operands being dominated by the source loop). - Fixes a bug in the dominance check whereby it was checked that values in the target loop themselves dominated the source loop rather than the ops that define these operands. - Renames the LoopFuseSibling op to LoopFuseSiblingOp. - Updates the LoopFuseSiblingOp's description. - Adds tests for using LoopFuseSiblingOp on scf.for loops, including one which fails without the fix for the dominance check. - Adds tests checking the different failure modes of the dominance checker. - Adds test for case whereby scf.yield is automatically generated when there are no loop-carried variables.
|
Thanks @ftynse. Your latest comments are now also addressed. If somebody could help with merging, that would be much appreciated! |
|
@rolfmorel Congratulations on having your first Pull Request (PR) merged into the LLVM Project! Your changes will be combined with recent changes from other authors, then tested Please check whether problems have been caused by your change specifically, as How to do this, and the rest of the post-merge process, is covered in detail here. If your change does cause a problem, it may be reverted, or you can revert it yourself. If you don't get any reports, no action is required from you. Your changes are working as expected, well done! |
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.
Adds support for fusing two scf.for loops occurring in the same block. Uses the rudimentary checks already in place for scf.forall (like the target loop's operands being dominated by the source loop).