[Mlir-commits] [mlir] 5ab1a8a - [mlir][sparse] fix crash when calling getTuple on non-sparse tensors.

[Peiming Liu]

Author: Peiming Liu
Date: 2022-11-11T21:15:44Z
New Revision: 5ab1a8aef454b82c409587aa849f4a1d2f1a86f9

URL: https://github.com/llvm/llvm-project/commit/5ab1a8aef454b82c409587aa849f4a1d2f1a86f9
DIFF: https://github.com/llvm/llvm-project/commit/5ab1a8aef454b82c409587aa849f4a1d2f1a86f9.diff

LOG: [mlir][sparse] fix crash when calling getTuple on non-sparse tensors.

This enables full sparse convolution codegen in D137298

Reviewed By: bixia

Differential Revision: https://reviews.llvm.org/D137853

Added: 
    

Modified: 
    mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorCodegen.cpp

Removed: 
    


################################################################################
diff  --git a/mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorCodegen.cpp b/mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorCodegen.cpp
index 499dc4c1ff33..d7dd8240a28c 100644
--- a/mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorCodegen.cpp
+++ b/mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorCodegen.cpp
@@ -59,15 +59,16 @@ static void flattenOperands(ValueRange operands,
   // ==>
   // memref ..., c, memref ...
   for (auto operand : operands) {
-    if (auto tuple = getTuple(operand);
-        tuple && getSparseTensorEncoding(tuple->getResultTypes()[0]))
+    if (getSparseTensorEncoding(operand.getType())) {
+      auto tuple = getTuple(operand);
       // An unrealized_conversion_cast will be inserted by type converter to
       // inter-mix the gap between 1:N conversion between sparse tensors and
       // fields. In this case, take the operands in the cast and replace the
       // sparse tensor output with the flattened type array.
       flattened.append(tuple.getOperands().begin(), tuple.getOperands().end());
-    else
+    } else {
       flattened.push_back(operand);
+    }
   }
 }