projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: c1679a9) | patch
Fix a bug in the vectorization of wide load/store operations.
authorNadav Rotem <nrotem@apple.com>
Sun, 21 Oct 2012 06:49:10 +0000 (06:49 +0000)
committerNadav Rotem <nrotem@apple.com>
Sun, 21 Oct 2012 06:49:10 +0000 (06:49 +0000)
commitfe88c6716143eace05fd49b8291e128768c33891
tree78e5d081993ab80ac5f975a6f196e42b68be0803tree | snapshot
parentc1679a95b6b46a48bf5951cda1373a123b19c1d3commit | diff
Fix a bug in the vectorization of wide load/store operations.
We used a SCEV to detect that A[X] is consecutive. We assumed that X was
the induction variable. But X can be any expression that uses the induction
for example: X = i + 2;

llvm-svn: 166388
llvm/lib/Transforms/Vectorize/LoopVectorize.cpp diff | blob | history
llvm/test/Transforms/LoopVectorize/induction_plus.ll [new file with mode: 0644] blob
Domain: System / Toolchain;