OSDN Git Service

Reapply "[LV] Extend trunc optimization to all IVs with constant integer steps"
authorMatthew Simpson <mssimpso@codeaurora.org>
Tue, 14 Feb 2017 16:28:32 +0000 (16:28 +0000)
committerMatthew Simpson <mssimpso@codeaurora.org>
Tue, 14 Feb 2017 16:28:32 +0000 (16:28 +0000)
commit06f1a4b52beb58edd122765fa23e65808d814ef5
tree7f21e8347c0873a8a85636eb2e949fc0ecf5e720
parent2fce16a04e36b219362b0022a944d829abeaafc2
Reapply "[LV] Extend trunc optimization to all IVs with constant integer steps"

This reapplies commit r294967 with a fix for the execution time regressions
caught by the clang-cmake-aarch64-quick bot. We now extend the truncate
optimization to non-primary induction variables only if the truncate isn't
already free.

Differential Revision: https://reviews.llvm.org/D29847

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@295063 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Transforms/Vectorize/LoopVectorize.cpp
test/Transforms/LoopVectorize/AArch64/induction-trunc.ll [new file with mode: 0644]
test/Transforms/LoopVectorize/induction.ll
test/Transforms/LoopVectorize/reverse_iter.ll