[Mlir-commits] [mlir] [mlir] Use range-based for loops (NFC) (PR #146946)

[Kazu Hirata]

================
@@ -602,8 +602,8 @@ void TosaReduceTransposes::runOnOperation() {
         !llvm::isa<RankedTensorType>(output.getType()))
       return;
 
-    llvm::for_each(transposeOp.getPerms(),
-                   [&perms](const auto i) { perms.emplace_back(i); });
+    for (const auto i : transposeOp.getPerms())
+      perms.emplace_back(i);
----------------
kazutakahirata wrote:

Fixed in the latest revision.  Thanks!

https://github.com/llvm/llvm-project/pull/146946