projects / oota-llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 2cf5425) | patch
SLPVectorizer: Speedup isConsecutive by manually checking GEPs with multiple indices.
authorNadav Rotem <nrotem@apple.com>
Thu, 18 Jul 2013 18:20:45 +0000 (18:20 +0000)
committerNadav Rotem <nrotem@apple.com>
Thu, 18 Jul 2013 18:20:45 +0000 (18:20 +0000)
commitdfacdd04cd2dd3b474fcabc5497255548f5506d5
treee532adfe63257a671849e843fba07223a1a69001tree | snapshot
parent2cf5425d0a048fbf36fc15035487f427499b514acommit | diff
SLPVectorizer: Speedup isConsecutive by manually checking GEPs with multiple indices.
This brings the compile time of the SLP-Vectorizer to about 2.5% of OPT for my testcase.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@186592 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Transforms/Vectorize/SLPVectorizer.cpp diff | blob | history
C/C++ LLVM-based compilers that forbids OOTA behaviors