caio/clang-p2996
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Fix invalid op result access in SparseTensorCodegen

Browse Source 
This trigges an assertion introduced in https://reviews.llvm.org/D147883

Differential Revision: https://reviews.llvm.org/D147950
This commit is contained in:
Frederik Gossen
2023-04-10 12:30:27 -04:00
parent f34a1953ce
commit 4a33c67995
1 changed files with 1 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorCodegen.cpp
View File

@@ -1418,8 +1418,7 @@ struct SparseNewOpConverter : public OpConversionPattern<NewOp> {
// instead of just accessing the reader's memory directly.
Value dimSizes = genAlloca(rewriter, loc, dimRank, indexTp);
createFuncCall(rewriter, loc, "copySparseTensorReaderDimSizes", {},
{reader, dimSizes}, EmitCInterface::On)
.getResult(0);
{reader, dimSizes}, EmitCInterface::On);
for (const auto &d : llvm::enumerate(dstTp.getDimShape()))
if (ShapedType::isDynamic(d.value()))
dynSizes.push_back(rewriter.create<memref::LoadOp>(