[PATCH] D137280: [RISCV] Prevent autovectorization using vscale with Zvl32b.

Wed Nov 2 13:06:24 PDT 2022

craig.topper added a comment.

In D137280#3903292 <https://reviews.llvm.org/D137280#3903292>, @reames wrote:

> LGTM
>
> I'm fine with this, but I thought we didn't support zve32f during compilation at all right now?  Is this the only issue which needs fixed.

I think zve32f works fine with Zvl64b or larger. The only issues I know of are related to use of llvm.vscale with Zvl32b. The type system limps along as long as we don't try to figure out what the value of vscale is.

Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D137280/new/

https://reviews.llvm.org/D137280