diff options
| author | Chad Rosier <mcrosier@codeaurora.org> | 2018-03-09 16:43:46 +0000 |
|---|---|---|
| committer | Chad Rosier <mcrosier@codeaurora.org> | 2018-03-09 16:43:46 +0000 |
| commit | 95d9ccb2a0542d79ecf40f14784be20f96b494b1 (patch) | |
| tree | c51f472f4da6ea423d01362cea32aa957a550bcf /llvm | |
| parent | ebda41f812925cf1332b317ed725da2dd181182a (diff) | |
| download | bcm5719-llvm-95d9ccb2a0542d79ecf40f14784be20f96b494b1.tar.gz bcm5719-llvm-95d9ccb2a0542d79ecf40f14784be20f96b494b1.zip | |
[JumpThreading] Don't restrict cast-traversal to i1
In r263618, JumpThreading learned to look trough simple cast instructions, but
only if the source of those cast instructions was a phi/cmp i1 (in an effort to
limit compile time effects). I think this condition is too restrictive. For
switches with limited value range, InstCombine will readily introduce an extra
trunc instruction to a smaller integer type (e.g. from i8 to i2), leaving us in
the somewhat perverse situation that jump-threading would work before running
instcombine, but not after. Since instcombine produces this pattern, I think we
need to consider it canonical and support it in JumpThreading. In general,
for limiting recursion, I think the existing restriction to phi and cmp nodes
should be sufficient to avoid looking through unprofitable chains of
instructions.
Patch by Keno Fischer!
Differential Revision: https://reviews.llvm.org/D42262
llvm-svn: 327150
Diffstat (limited to 'llvm')
| -rw-r--r-- | llvm/lib/Transforms/Scalar/JumpThreading.cpp | 4 | ||||
| -rw-r--r-- | llvm/test/Transforms/JumpThreading/basic.ll | 28 |
2 files changed, 29 insertions, 3 deletions
diff --git a/llvm/lib/Transforms/Scalar/JumpThreading.cpp b/llvm/lib/Transforms/Scalar/JumpThreading.cpp index 15cc709451b..ee1de8edc1d 100644 --- a/llvm/lib/Transforms/Scalar/JumpThreading.cpp +++ b/llvm/lib/Transforms/Scalar/JumpThreading.cpp @@ -656,11 +656,9 @@ bool JumpThreadingPass::ComputeValueKnownInPredecessors( } // Handle Cast instructions. Only see through Cast when the source operand is - // PHI or Cmp and the source type is i1 to save the compilation time. + // PHI or Cmp to save the compilation time. if (CastInst *CI = dyn_cast<CastInst>(I)) { Value *Source = CI->getOperand(0); - if (!Source->getType()->isIntegerTy(1)) - return false; if (!isa<PHINode>(Source) && !isa<CmpInst>(Source)) return false; ComputeValueKnownInPredecessors(Source, BB, Result, Preference, CxtI); diff --git a/llvm/test/Transforms/JumpThreading/basic.ll b/llvm/test/Transforms/JumpThreading/basic.ll index ce86cba50c6..16e7549e3fc 100644 --- a/llvm/test/Transforms/JumpThreading/basic.ll +++ b/llvm/test/Transforms/JumpThreading/basic.ll @@ -547,6 +547,34 @@ l5: ; CHECK: } } +define i1 @trunc_switch(i1 %arg) { +; CHECK-LABEL: @trunc_switch +top: +; CHECK: br i1 %arg, label %exitA, label %exitB + br i1 %arg, label %common, label %B + +B: + br label %common + +common: + %phi = phi i8 [ 2, %B ], [ 1, %top ] + %trunc = trunc i8 %phi to i2 +; CHECK-NOT: switch + switch i2 %trunc, label %unreach [ + i2 1, label %exitA + i2 -2, label %exitB + ] + +unreach: + unreachable + +exitA: + ret i1 true + +exitB: + ret i1 false +} + ; CHECK-LABEL: define void @h_con(i32 %p) { define void @h_con(i32 %p) { %x = icmp ult i32 %p, 5 |

