[PATCH] D93745: [RISCV] Define the vfsqrt RVV intrinsics

[Craig Topper via Phabricator via llvm-commits]

craig.topper added inline comments.


================
Comment at: llvm/test/CodeGen/RISCV/rvv/vfsqrt-rv32.ll:512
+  ret <vscale x 16 x float> %a
+}
----------------
evandro wrote:
> craig.topper wrote:
> > We're missing tests for double in the rv32 file.
> Ain't it the case for virtually all other RVV intrinsics so far?
Yes. I didn’t catch it until after a bunch of patches had been committed. Some of the most recent patches have it right.

Would you rather fix this patch now or wait and fix all the tests together?


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D93745/new/

https://reviews.llvm.org/D93745