This is an archive of the discontinued LLVM Phabricator instance.

When optimizing for size, enable loop rerolling by default.
ClosedPublic

Authored by hfinkel on Oct 7 2016, 6:22 PM.

Details

Reviewers
jmolloy
Summary

We have a loop-rerolling optimization which can be enabled by using -freroll-loops. While sometimes loops are hand-unrolled for performance reasons, when optimizing for size, we should always undo this manual optimization to produce smaller code (our optimizer's unroller will still unroll the rerolled loops if it thinks that is a good idea).

Diff Detail

Event Timeline

hfinkel updated this revision to Diff 74007.Oct 7 2016, 6:22 PM
hfinkel retitled this revision from to When optimizing for size, enable loop rerolling by default..
hfinkel updated this object.
hfinkel added a reviewer: jmolloy.
hfinkel added a subscriber: cfe-commits.
jmolloy accepted this revision.Oct 8 2016, 3:29 AM
jmolloy edited edge metadata.

Good idea! LGTM!

This revision is now accepted and ready to land.Oct 8 2016, 3:29 AM
hfinkel closed this revision.Oct 8 2016, 8:16 PM

r283685