This is an archive of the discontinued LLVM Phabricator instance.

[CUDA] Make sure we emit all templated __global__ functions on device side.
ClosedPublic

Authored by tra on Jul 30 2015, 3:48 PM.

Diff Detail

Repository
rL LLVM

Event Timeline

tra updated this revision to Diff 31082.Jul 30 2015, 3:48 PM
tra retitled this revision from to [CUDA] Make sure we emit all templated __global__ functions on device side..
tra updated this object.
tra added reviewers: echristo, eliben.
tra added a subscriber: cfe-commits.

Couldn't you just add an implicit UsedAttr when processing the CUDAGlobalAttr
and LangOpts.CUDAIsDevice was set to true?

tra updated this revision to Diff 31128.Jul 31 2015, 10:11 AM
tra updated this object.

Changed to use implicit 'used' attribute.

This revision was automatically updated to reflect the committed changes.