Page MenuHomePhabricator

[OPENMP][NVPTX]Improve code by using parallel level counter.
ClosedPublic

Authored by ABataev on May 1 2019, 1:02 PM.

Details

Summary

Previously for the different purposes we need to get the active/common
parallel level and with full runtime we iterated over all the records to
calculate this level. Instead, we can used the warp-based parallel level
counters used in no-runtime mode.

Diff Detail

Repository
rL LLVM

Event Timeline

ABataev created this revision.May 1 2019, 1:02 PM
Herald added a project: Restricted Project. · View Herald TranscriptMay 1 2019, 1:02 PM
ABataev updated this revision to Diff 197776.May 2 2019, 7:15 AM

Fixed detection of non-active parallel regions.

ABataev updated this revision to Diff 197788.May 2 2019, 7:59 AM

Update to the latest version

grokos accepted this revision.May 2 2019, 12:51 PM

LGTM

This revision is now accepted and ready to land.May 2 2019, 12:51 PM
This revision was automatically updated to reflect the committed changes.
Herald added a project: Restricted Project. · View Herald TranscriptMay 2 2019, 1:03 PM