[PATCH] D60457: [CodeGen] Fixed de-optimization of legalize subvector extract

[Tim Renouf via Phabricator via llvm-commits]

tpr marked 2 inline comments as done.
tpr added a comment.

Hi Samuel. Sorry for the delay; I kind of lost track of this change.

Question for Nicolai below.



================
Comment at: test/CodeGen/AMDGPU/extract_subvector_vec4_vec3.ll:4
+; GCN-LABEL: body:
+; GCN-NOT: %stack
+
----------------
nhaehnle wrote:
> tpr wrote:
> > arsenm wrote:
> > > I wouldn’t trust this to check this, a generated check would be better
> > Not really sure what you're suggesting, but I hope this is better.
> Maybe you can use `util/update_llc_test_checks.py`?
You mean have a check line for each line of IR output in the function? Do you think that would be better than the negative check for storing to stack?


Repository:
  rL LLVM

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D60457/new/

https://reviews.llvm.org/D60457