projects / oota-llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: cbc85a2) | patch
LoopVectorizer: Use abi alignment for accesses with no alignment
authorArnold Schwaighofer <aschwaighofer@apple.com>
Fri, 15 Nov 2013 23:09:33 +0000 (23:09 +0000)
committerArnold Schwaighofer <aschwaighofer@apple.com>
Fri, 15 Nov 2013 23:09:33 +0000 (23:09 +0000)
commit4634338655449a6e76b6a948c91fbe5bc736d24b
treef426ed0662f93ab45e2b5cb927f819dfd28c62b2tree | snapshot
parentcbc85a270d2546c49fc09f700687de2ecd46bd87commit | diff
LoopVectorizer: Use abi alignment for accesses with no alignment

When we vectorize a scalar access with no alignment specified, we have to set
the target's abi alignment of the scalar access on the vectorized access.
Using the same alignment of zero would be wrong because most targets will have a
bigger abi alignment for vector types.

This probably fixes PR17878.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@194876 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Transforms/Vectorize/LoopVectorize.cpp diff | blob | history
test/Transforms/LoopVectorize/align.ll [new file with mode: 0644] blob
C/C++ LLVM-based compilers that forbids OOTA behaviors