[Mlir-commits] [mlir] [MLIR] Folding unpack and pack sequence in data layout propagation from padded domain (PR #138332)

[Zhuoran Yin]

================
@@ -298,20 +298,60 @@ getOrCreatePackedViewOfOperand(OpBuilder &b, Location loc, PackInfo packInfo,
   return std::make_tuple(packedOperand, indexingMap);
 }
 
+static bool isGenericOutsNotUsed(linalg::GenericOp genericOp) {
+  int numDpsOuts = genericOp.getNumDpsInits();
+  Block *block = genericOp.getBody();
+  int numBlockArgs = block->getNumArguments();
+  int initArgStartIndex = numBlockArgs - numDpsOuts;
+  for (int i = 0; i < numDpsOuts; ++i) {
+    int matchingInitArgIndex = initArgStartIndex + i;
+    return block->getArgument(matchingInitArgIndex).use_empty();
+  }
+  return true;
+}
+
 /// Pack a genericOp and return it.
 static GenericOp packGenericOp(RewriterBase &rewriter, GenericOp genericOp,
                                Value dest, AffineMap packedOutIndexingMap,
-                               const PackInfo &packInfo) {
+                               const PackInfo &packInfo,
+                               bool canUnpackPackFold) {
----------------
jerryyin wrote:

Yes, that's better name for sure. Also, I've added a comment for `packGenericOp` in the latest commit.

https://github.com/llvm/llvm-project/pull/138332