This is an archive of the discontinued LLVM Phabricator instance.

Differential D36965

[RelNotes,AArch64] Mention improved instruction fusion and fun alignment.
ClosedPublic

Authored by fhahn on Aug 21 2017, 8:25 AM.

Details

Reviewers
t.p.northover
hans
kristof.beyls
Commits
rGa2c40bdbaf4b: [RelNotes,AArch64] Mention improved instruction fusion and fun alignment.
rL311539: [RelNotes,AArch64] Mention improved instruction fusion and fun alignment.

Diff Detail

Event Timeline

fhahn created this revision.Aug 21 2017, 8:25 AM
Herald added subscribers: rengolin, aemerson. · View Herald TranscriptAug 21 2017, 8:25 AM
fhahn added a comment.Aug 21 2017, 8:32 AM

This is for the 5.0 release notes

hans accepted this revision.Aug 21 2017, 1:01 PM

Looks great, thanks!

This revision is now accepted and ready to land.Aug 21 2017, 1:01 PM
fhahn updated this revision to Diff 112314.Aug 23 2017, 2:10 AM

rebased

fhahn closed this revision.Aug 23 2017, 2:13 AM

Revision Contents

PathSize
docs/
8 lines

Diff 112314

docs/ReleaseNotes.rst

Context not available.
Changes to the Arm Targets Changes to the Arm Targets
-------------------------- --------------------------
During this release the AArch64 target has:
* Made instruction fusion more aggressive, resulting in speedups
for code making use of AArch64 AES instructions. AES fusion has been
enabled for most Cortex-A cores and the AArch64MacroFusion pass was moved
to the generic MacroFusion pass.
* Added preferred function alignments for most Cortex-A cores.
During this release the ARM target has: During this release the ARM target has:
* Improved mixed ARM/Thumb code generation. Some cases in which wrong * Improved mixed ARM/Thumb code generation. Some cases in which wrong
Context not available.