[PATCH] D97299: [IR][SVE] Add new llvm.experimental.stepvector intrinsic

[Florian Hahn via Phabricator via llvm-commits]

fhahn added inline comments.


================
Comment at: llvm/docs/LangRef.rst:16613
+with a step of 1.  It is recommended that this experimental intrinsic only be
+used for scalable vectors with integer elements. The intrinsic also works for
+fixed-width vectors, but the user will probably get better code quality
----------------
Would it make sense to limit the intrinsic to integer vectors only? Otherwise, why is it not recommended to use it with different element types? I think the behaviour with other element types should also be spelled out, if other element types are supported.


================
Comment at: llvm/docs/LangRef.rst:16614
+used for scalable vectors with integer elements. The intrinsic also works for
+fixed-width vectors, but the user will probably get better code quality
+generating constant vectors instead. If the sequence value exceeds the allowed
----------------
Can we make the wording here consistent with the other scalable intrinsics?


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D97299/new/

https://reviews.llvm.org/D97299