1. "Upgrage" comments.
authorWojciech Matyjewicz <wmatyjewicz@fastmail.fm>
Wed, 12 Dec 2007 15:21:32 +0000 (15:21 +0000)
committerWojciech Matyjewicz <wmatyjewicz@fastmail.fm>
Wed, 12 Dec 2007 15:21:32 +0000 (15:21 +0000)
commited2232502eac83d649c61c3f9bbfbd642b864319
tree256e0615d6af3f7ca5776dfc29fdda31013d7da8
parentbf3348d944c9b090ba9a162b170ee74bf95d6337
1. "Upgrage" comments.
2. Using zero-extended value of Scale and unsigned division is safe provided
   that Scale doesn't have the sign bit set.
   Previously these 2 instructions:
        %p = bitcast [100 x {i8,i8,i8}]* %x to i8*
        %q = getelementptr i8* %p, i32 -4
   were combined into:
        %q = getelementptr [100 x { i8, i8, i8 }]* %x, i32 0,
               i32 1431655764, i32 0
   what was incorrect.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@44936 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Transforms/Scalar/InstructionCombining.cpp
test/Transforms/InstCombine/2007-12-12-GEPScale.ll [new file with mode: 0644]