This is an archive of the discontinued LLVM Phabricator instance.

[AMDGPU] Handle R_AMDGPU_ABS64 relocation
ClosedPublic

Authored by kzhuravl on Oct 19 2016, 11:21 PM.

Details

Summary

This is needed for the following case (OpenCL example):

__global int Var = 0; 
__global int* Ptr[] = {&Var}; 
...

Diff Detail

Repository
rL LLVM

Event Timeline

kzhuravl updated this revision to Diff 75268.Oct 19 2016, 11:21 PM
kzhuravl retitled this revision from to [AMDGPU] Handle R_AMDGPU_ABS64 relocation.
kzhuravl updated this object.
kzhuravl added a subscriber: llvm-commits.
ruiu accepted this revision.Oct 19 2016, 11:27 PM
ruiu edited edge metadata.

LGTM with a nit.

ELF/Target.cpp
1491 ↗(On Diff #75268)

Move to line 1487 to sort.

This revision is now accepted and ready to land.Oct 19 2016, 11:27 PM
This revision was automatically updated to reflect the committed changes.
kzhuravl marked an inline comment as done.
kzhuravl added inline comments.Oct 20 2016, 11:44 AM
ELF/Target.cpp
1491 ↗(On Diff #75268)

Thanks, fixed this before committing.