[MLIR] Fix unchecked use of memref memory space attr in affine data copy generate #116763
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
…opy generate Fix unchecked use of memref memory space attr in affine data copy generate. In the case of memory accesses without a memory space attribute or those other than integer attributes, the pass treats them as slow memory spaces. Fixes llvm#116536
Member
|
@llvm/pr-subscribers-mlir-affine Author: Uday Bondhugula (bondhugula) ChangesFix unchecked use of memref memory space attr in affine data copy Fixes #116536 Full diff: https://github.com/llvm/llvm-project/pull/116763.diff 2 Files Affected:
diff --git a/mlir/lib/Dialect/Affine/Utils/LoopUtils.cpp b/mlir/lib/Dialect/Affine/Utils/LoopUtils.cpp
index d6fc4ed07bfab3..0db71f866c0243 100644
--- a/mlir/lib/Dialect/Affine/Utils/LoopUtils.cpp
+++ b/mlir/lib/Dialect/Affine/Utils/LoopUtils.cpp
@@ -2298,21 +2298,26 @@ mlir::affine::affineDataCopyGenerate(Block::iterator begin, Block::iterator end,
// Walk this range of operations to gather all memory regions.
block->walk(begin, end, [&](Operation *opInst) {
+ Value memref;
+ MemRefType memrefType;
// Gather regions to allocate to buffers in faster memory space.
if (auto loadOp = dyn_cast<AffineLoadOp>(opInst)) {
- if ((filterMemRef.has_value() && filterMemRef != loadOp.getMemRef()) ||
- (loadOp.getMemRefType().getMemorySpaceAsInt() !=
- copyOptions.slowMemorySpace))
- return;
+ memref = loadOp.getMemRef();
+ memrefType = loadOp.getMemRefType();
} else if (auto storeOp = dyn_cast<AffineStoreOp>(opInst)) {
- if ((filterMemRef.has_value() && filterMemRef != storeOp.getMemRef()) ||
- storeOp.getMemRefType().getMemorySpaceAsInt() !=
- copyOptions.slowMemorySpace)
- return;
- } else {
- // Neither load nor a store op.
- return;
+ memref = storeOp.getMemRef();
+ memrefType = storeOp.getMemRefType();
}
+ // Neither load nor a store op.
+ if (!memref)
+ return;
+
+ auto memorySpaceAttr =
+ dyn_cast_or_null<IntegerAttr>(memrefType.getMemorySpace());
+ if ((filterMemRef.has_value() && filterMemRef != memref) ||
+ (memorySpaceAttr &&
+ memrefType.getMemorySpaceAsInt() != copyOptions.slowMemorySpace))
+ return;
// Compute the MemRefRegion accessed.
auto region = std::make_unique<MemRefRegion>(opInst->getLoc());
diff --git a/mlir/test/Dialect/Affine/affine-data-copy.mlir b/mlir/test/Dialect/Affine/affine-data-copy.mlir
index fe3b4a206e2b9d..330cf92bafba4a 100644
--- a/mlir/test/Dialect/Affine/affine-data-copy.mlir
+++ b/mlir/test/Dialect/Affine/affine-data-copy.mlir
@@ -333,3 +333,23 @@ func.func @index_elt_type(%arg0: memref<1x2x4x8xindex>) {
// CHECK-NEXT: affine.for %{{.*}} = 0 to 8
return
}
+
+#map = affine_map<(d0) -> (d0 + 1)>
+
+// CHECK-LABEL: func @arbitrary_memory_space
+func.func @arbitrary_memory_space() {
+ %alloc = memref.alloc() : memref<256x8xi8, #spirv.storage_class<StorageBuffer>>
+ affine.for %arg0 = 0 to 32 step 4 {
+ %0 = affine.apply #map(%arg0)
+ affine.for %arg1 = 0 to 8 step 2 {
+ %1 = affine.apply #map(%arg1)
+ affine.for %arg2 = 0 to 8 step 2 {
+ // CHECK: memref.alloc() : memref<1x7xi8>
+ %2 = affine.apply #map(%arg2)
+ %3 = affine.load %alloc[%0, %1] : memref<256x8xi8, #spirv.storage_class<StorageBuffer>>
+ affine.store %3, %alloc[%0, %2] : memref<256x8xi8, #spirv.storage_class<StorageBuffer>>
+ }
+ }
+ }
+ return
+}
|
Member
|
@llvm/pr-subscribers-mlir Author: Uday Bondhugula (bondhugula) ChangesFix unchecked use of memref memory space attr in affine data copy Fixes #116536 Full diff: https://github.com/llvm/llvm-project/pull/116763.diff 2 Files Affected:
diff --git a/mlir/lib/Dialect/Affine/Utils/LoopUtils.cpp b/mlir/lib/Dialect/Affine/Utils/LoopUtils.cpp
index d6fc4ed07bfab3..0db71f866c0243 100644
--- a/mlir/lib/Dialect/Affine/Utils/LoopUtils.cpp
+++ b/mlir/lib/Dialect/Affine/Utils/LoopUtils.cpp
@@ -2298,21 +2298,26 @@ mlir::affine::affineDataCopyGenerate(Block::iterator begin, Block::iterator end,
// Walk this range of operations to gather all memory regions.
block->walk(begin, end, [&](Operation *opInst) {
+ Value memref;
+ MemRefType memrefType;
// Gather regions to allocate to buffers in faster memory space.
if (auto loadOp = dyn_cast<AffineLoadOp>(opInst)) {
- if ((filterMemRef.has_value() && filterMemRef != loadOp.getMemRef()) ||
- (loadOp.getMemRefType().getMemorySpaceAsInt() !=
- copyOptions.slowMemorySpace))
- return;
+ memref = loadOp.getMemRef();
+ memrefType = loadOp.getMemRefType();
} else if (auto storeOp = dyn_cast<AffineStoreOp>(opInst)) {
- if ((filterMemRef.has_value() && filterMemRef != storeOp.getMemRef()) ||
- storeOp.getMemRefType().getMemorySpaceAsInt() !=
- copyOptions.slowMemorySpace)
- return;
- } else {
- // Neither load nor a store op.
- return;
+ memref = storeOp.getMemRef();
+ memrefType = storeOp.getMemRefType();
}
+ // Neither load nor a store op.
+ if (!memref)
+ return;
+
+ auto memorySpaceAttr =
+ dyn_cast_or_null<IntegerAttr>(memrefType.getMemorySpace());
+ if ((filterMemRef.has_value() && filterMemRef != memref) ||
+ (memorySpaceAttr &&
+ memrefType.getMemorySpaceAsInt() != copyOptions.slowMemorySpace))
+ return;
// Compute the MemRefRegion accessed.
auto region = std::make_unique<MemRefRegion>(opInst->getLoc());
diff --git a/mlir/test/Dialect/Affine/affine-data-copy.mlir b/mlir/test/Dialect/Affine/affine-data-copy.mlir
index fe3b4a206e2b9d..330cf92bafba4a 100644
--- a/mlir/test/Dialect/Affine/affine-data-copy.mlir
+++ b/mlir/test/Dialect/Affine/affine-data-copy.mlir
@@ -333,3 +333,23 @@ func.func @index_elt_type(%arg0: memref<1x2x4x8xindex>) {
// CHECK-NEXT: affine.for %{{.*}} = 0 to 8
return
}
+
+#map = affine_map<(d0) -> (d0 + 1)>
+
+// CHECK-LABEL: func @arbitrary_memory_space
+func.func @arbitrary_memory_space() {
+ %alloc = memref.alloc() : memref<256x8xi8, #spirv.storage_class<StorageBuffer>>
+ affine.for %arg0 = 0 to 32 step 4 {
+ %0 = affine.apply #map(%arg0)
+ affine.for %arg1 = 0 to 8 step 2 {
+ %1 = affine.apply #map(%arg1)
+ affine.for %arg2 = 0 to 8 step 2 {
+ // CHECK: memref.alloc() : memref<1x7xi8>
+ %2 = affine.apply #map(%arg2)
+ %3 = affine.load %alloc[%0, %1] : memref<256x8xi8, #spirv.storage_class<StorageBuffer>>
+ affine.store %3, %alloc[%0, %2] : memref<256x8xi8, #spirv.storage_class<StorageBuffer>>
+ }
+ }
+ }
+ return
+}
|
Contributor
Author
|
Straightforward fix for a crash. Merging. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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.
Fix unchecked use of memref memory space attr in affine data copy
generate. In the case of memory accesses without a memory space
attribute or those other than integer attributes, the pass treats them
as slow memory spaces.
Fixes #116536