[llvm-branch-commits] [llvm] [LV][REVEC] Initial support for re-vectorisation (PR #208213)

Andrei Elovikov via llvm-branch-commits llvm-branch-commits at lists.llvm.org
Thu Aug 13 11:54:14 PDT 2026


=?utf-8?q?Gaëtan?= Bossu <gaetan.bossu at arm.com>,
=?utf-8?q?Gaëtan?= Bossu <gaetan.bossu at arm.com>,
=?utf-8?q?Gaëtan?= Bossu <gaetan.bossu at arm.com>,
=?utf-8?q?Gaëtan?= Bossu <gaetan.bossu at arm.com>,
=?utf-8?q?Gaëtan?= Bossu <gaetan.bossu at arm.com>,
=?utf-8?q?Gaëtan?= Bossu <gaetan.bossu at arm.com>
Message-ID:
In-Reply-To: <llvm.org/llvm/llvm-project/pull/208213 at github.com>


================
@@ -1001,7 +1026,8 @@ bool LoopVectorizationLegality::canVectorizeInstr(Instruction &I) {
     // supported on the target.
     if (ST->getMetadata(LLVMContext::MD_nontemporal)) {
       // Arbitrarily try a vector of 2 elements.
-      auto *VecTy = FixedVectorType::get(T, /*NumElts=*/2);
+      Type *VecTy =
----------------
eas wrote:

Would it be better to do

```c++
    // For nontemporal stores, check that a nontemporal vector version is
    // supported on the target. If this a vector store already (revectorization), 
    // skip the check, we know it's legal.
    if (ST->getMetadata(LLVMContext::MD_nontemporal) && !T->isVectorTy() {
```
?

https://github.com/llvm/llvm-project/pull/208213


More information about the llvm-branch-commits mailing list