maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity

[InstCombine] Fix a vector splat handling bug in transformZExtICmp. 

We were using an i1 type and then zero extending to a vector. Instead just create the 0/1 directly as a ConstantInt with the correct type. No need to ask ConstantExpr to zero extend for us.

This bug is a bit tricky to hit because it requires us to visit a zext of an icmp that would normally be simplified to true/false, but that icmp hasnt' been visited yet. In the test case this zext and icmp were created by visiting a udiv and due to worklist ordering we got to the zext first.

Fixes PR34841.

llvm-svn: 314971
This commit is contained in:
Craig Topper 2017-10-05 07:59:11 +00:00
parent 902eef32eb
commit 17b0c78447
2 changed files with 10 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
llvm/lib/Transforms/InstCombine/InstCombineCasts.cpp
View File

@ -818,9 +818,7 @@ Instruction *InstCombiner::transformZExtICmp(ICmpInst *ICI, ZExtInst &CI,
if (!Op1CV->isNullValue() && (*Op1CV != KnownZeroMask)) {
// (X&4) == 2 --> false
// (X&4) != 2 --> true
Constant *Res = ConstantInt::get(Type::getInt1Ty(CI.getContext()),
isNE);
Res = ConstantExpr::getZExt(Res, CI.getType());
Constant *Res = ConstantInt::get(CI.getType(), isNE);
return replaceInstUsesWith(CI, Res);
}

9
llvm/test/Transforms/InstCombine/div.ll
View File

@ -553,3 +553,12 @@ define i32 @shrink_no3(i16 %x) {
ret i32 %div
}
; This previously crashed when trying to simplify the zext/icmp this becomes.
define <2 x i8> @PR34841(<2 x i8> %x) {
; CHECK-LABEL: @PR34841(
; CHECK-NEXT: ret <2 x i8> zeroinitializer
;
%neg = and <2 x i8> %x, <i8 2, i8 2>
%div = udiv <2 x i8> <i8 1, i8 1>, %neg
ret <2 x i8> %div
}