[PATCH] D90554: [CostModel] remove cost-kind predicate for intrinsics in basic TTI implementation

Sanjay Patel via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Fri Nov 20 03:35:45 PST 2020


spatel added a comment.

In D90554#2407132 <https://reviews.llvm.org/D90554#2407132>, @MaskRay wrote:

> In D90554#2406927 <https://reviews.llvm.org/D90554#2406927>, @brooksmoses wrote:
>
>> As a heads-up: This is causing a lot of Clang segfaults in Google's builds with sanitizers enabled.  We're working work on a reduced testcase, but wanted to let you know while we do that.
>
> `opt -O2 < a.ll` on the following IR triggers an assertion failure:
>
>   616│ FixedVectorType *FixedVectorType::get(Type *ElementType, unsigned NumElts) {
>   617│   assert(NumElts > 0 && "#Elements of a VectorType must be greater than 0");
>   ////////////////// ElementType is MetadataTy, not Integer/FloatingPoint/Pointer
>   618│   assert(isValidElementType(ElementType) && "Element type of a VectorType must " 
>   619│                                             "be an integer, floating point, or "
>   620│                                             "pointer type.");

Thanks, all. Looks like we are blindly walking the intrinsic params (similar problem as 74ffc823ed21 <https://reviews.llvm.org/rG74ffc823ed2128dfce40181a13e3b2076a61126f>) without checking types. I'll work on a fix today.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D90554/new/

https://reviews.llvm.org/D90554



More information about the llvm-commits mailing list