This is an archive of the discontinued LLVM Phabricator instance.

AMDGPU/R600: Expand unaligned writes to local and global AS
ClosedPublic

Authored by jvesely on Aug 28 2016, 12:36 AM.

Diff Detail

Repository
rL LLVM

Event Timeline

jvesely updated this revision to Diff 69506.Aug 28 2016, 12:36 AM
jvesely retitled this revision from to AMDGPU/R600: Expand unaligned writes to local and global AS.
jvesely updated this object.
jvesely set the repository for this revision to rL LLVM.
jvesely updated this revision to Diff 69525.Aug 28 2016, 4:12 PM
jvesely edited edge metadata.

fix store_v4i16 test

arsenm added inline comments.Aug 29 2016, 12:07 PM
lib/Target/AMDGPU/R600ISelLowering.cpp
1131 ↗(On Diff #69525)

C++ style comments

1142–1143 ↗(On Diff #69525)

Ditto

1151–1152 ↗(On Diff #69525)

These can be in one && assert

jvesely updated this revision to Diff 69614.Aug 29 2016, 2:10 PM

fix c style comments

jvesely marked 3 inline comments as done.Aug 29 2016, 2:12 PM
jvesely added inline comments.
lib/Target/AMDGPU/R600ISelLowering.cpp
1151–1153 ↗(On Diff #69614)

I'd prefer to keep them separate. the assertion message clearly indicates what part failed

arsenm accepted this revision.Aug 29 2016, 4:57 PM
arsenm added a reviewer: arsenm.

LGTM

This revision is now accepted and ready to land.Aug 29 2016, 4:57 PM
This revision was automatically updated to reflect the committed changes.