projects / oota-llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 012d959) | patch
[AArch64]Enable the narrow ld promotion only on profitable microarchitectures
authorJun Bum Lim <junbuml@codeaurora.org>
Fri, 6 Nov 2015 16:27:47 +0000 (16:27 +0000)
committerJun Bum Lim <junbuml@codeaurora.org>
Fri, 6 Nov 2015 16:27:47 +0000 (16:27 +0000)
commitde4969d37c17d57e87ce0365071922232dc323e0
tree1fd3016b8e287179692f00b7ead2c001cf37ab17tree | snapshot
parent012d959bc51efaba09bac21ada64315143726b77commit | diff
[AArch64]Enable the narrow ld promotion only on profitable microarchitectures

The benefit from converting narrow loads into a wider load (r251438) could be
micro-architecturally dependent, as it assumes that a single load with two bitfield
extracts is cheaper than two narrow loads. Currently, this conversion is
enabled only in cortex-a57 on which performance benefits were verified.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@252316 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Target/AArch64/AArch64LoadStoreOptimizer.cpp diff | blob | history
test/CodeGen/AArch64/arm64-ldp.ll diff | blob | history
test/CodeGen/AArch64/arm64-ldr-merge.ll [new file with mode: 0644] blob
C/C++ LLVM-based compilers that forbids OOTA behaviors