[mlir][sparsifier] fix isAdmissibleBSR

[makslevental]

Fixes #72194.

[llvmbot]

@llvm/pr-subscribers-mlir-sparse

@llvm/pr-subscribers-mlir

Author: Maksim Levental (makslevental)

Changes

Fixes #72194.

---

Full diff: https://github.com/llvm/llvm-project/pull/72195.diff

1 Files Affected:

• (modified) mlir/lib/Dialect/SparseTensor/Transforms/SparseGPUCodegen.cpp (+1-1)

diff --git a/mlir/lib/Dialect/SparseTensor/Transforms/SparseGPUCodegen.cpp b/mlir/lib/Dialect/SparseTensor/Transforms/SparseGPUCodegen.cpp
index fb090890a46dfd5..927fc71f84a85a8 100644
--- a/mlir/lib/Dialect/SparseTensor/Transforms/SparseGPUCodegen.cpp
+++ b/mlir/lib/Dialect/SparseTensor/Transforms/SparseGPUCodegen.cpp
@@ -436,7 +436,7 @@ static bool isAdmissibleBSR(SparseTensorType &aTp) {
     // CuSparse only supports "square" blocks currently.
     SmallVector<unsigned> dims = getBlockSize(aTp.getDimToLvl());
     assert(dims.size() == 2);
-    return dims[0] = dims[1] && dims[0] > 1;
+    return dims[0] == dims[1] && dims[0] > 1;
   }
   return false;
 }

[aartbik]

Thanks for fixing this! Not sure how this slipped through ;-)