[llvm] [llvm][RISCV] Do not assume V extension on seeing vector type. (PR #166994)
via llvm-commits
llvm-commits at lists.llvm.org
Fri Nov 7 10:53:50 PST 2025
llvmbot wrote:
<!--LLVM PR SUMMARY COMMENT-->
@llvm/pr-subscribers-backend-risc-v
Author: Chenguang Wang (wecing)
<details>
<summary>Changes</summary>
We have a private extension which also uses the vector type in the frontend. Our platform does not have the V extension, so it triggered assertion failures from within getLMULCost().
I am not sure what is the best way to handle this, or if there are more such assertions within the codebase. But it feels reasonable to check for V extension before assuming LMUL exists.
---
Full diff: https://github.com/llvm/llvm-project/pull/166994.diff
1 Files Affected:
- (modified) llvm/lib/Target/RISCV/RISCVTargetTransformInfo.cpp (+3-2)
``````````diff
diff --git a/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.cpp b/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.cpp
index 7bc0b5b394828..ababe42604ba5 100644
--- a/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.cpp
+++ b/llvm/lib/Target/RISCV/RISCVTargetTransformInfo.cpp
@@ -2140,8 +2140,9 @@ InstructionCost RISCVTTIImpl::getMemoryOpCost(unsigned Opcode, Type *Src,
// Assume memory ops cost scale with the number of vector registers
// possible accessed by the instruction. Note that BasicTTI already
// handles the LT.first term for us.
- if (LT.second.isVector() && CostKind != TTI::TCK_CodeSize)
- BaseCost *= TLI->getLMULCost(LT.second);
+ if (TLI->getSubtarget().hasVInstructions())
+ if (LT.second.isVector() && CostKind != TTI::TCK_CodeSize)
+ BaseCost *= TLI->getLMULCost(LT.second);
return Cost + BaseCost;
}
``````````
</details>
https://github.com/llvm/llvm-project/pull/166994
More information about the llvm-commits
mailing list