diff options
author | Evgeniy Stepanov <eugeni.stepanov@gmail.com> | 2014-06-25 07:54:58 +0000 |
---|---|---|
committer | Evgeniy Stepanov <eugeni.stepanov@gmail.com> | 2014-06-25 07:54:58 +0000 |
commit | 98726c311b0dc19076bf242973d92ca1858febe7 (patch) | |
tree | 4f05d831dd12a24b620b5d87772468891b8d8e99 /test | |
parent | aa2e057bc108c32529815afb8be4ef090acddd94 (diff) | |
download | llvm-98726c311b0dc19076bf242973d92ca1858febe7.tar.gz llvm-98726c311b0dc19076bf242973d92ca1858febe7.tar.bz2 llvm-98726c311b0dc19076bf242973d92ca1858febe7.tar.xz |
[LICM] Don't create more than one copy of an instruction per loop exit block when sinking.
Fixes exponential compilation complexity in PR19835, caused by
LICM::sink not handling the following pattern well:
f = op g
e = op f, g
d = op e
c = op d, e
b = op c
a = op b, c
When an instruction with N uses is sunk, each of its operands gets N
new uses (all of them - phi nodes). In the example above, if a had 1
use, c would have 2, e would have 4, and g would have 8.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@211673 91177308-0d34-0410-b5e6-96231b3b80d8
Diffstat (limited to 'test')
-rw-r--r-- | test/Transforms/LICM/extra-copies.ll | 29 |
1 files changed, 29 insertions, 0 deletions
diff --git a/test/Transforms/LICM/extra-copies.ll b/test/Transforms/LICM/extra-copies.ll new file mode 100644 index 0000000000..ef52f9f404 --- /dev/null +++ b/test/Transforms/LICM/extra-copies.ll @@ -0,0 +1,29 @@ +; RUN: opt < %s -licm -S | FileCheck %s +; PR19835 +target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128" +target triple = "x86_64-unknown-linux-gnu" + +define i32 @f(i32 %x) { +entry: + br label %for.body + +for.body: ; preds = %entry, %for.body + %storemerge4 = phi i32 [ 0, %entry ], [ %inc, %for.body ] + %mul = mul nsw i32 %x, %x + %add2 = add nsw i32 %mul, %x + %mul3 = add nsw i32 %add2, %mul + %inc = add nsw i32 %storemerge4, 1 + %cmp = icmp slt i32 %inc, 100 + br i1 %cmp, label %for.body, label %for.end + +for.end: ; preds = %for.body + %a9.0.lcssa = phi i32 [ %mul3, %for.body ] + ret i32 %a9.0.lcssa +} + +; Test that there is exactly one copy of mul nsw i32 %x, %x in the exit block. +; CHECK: define i32 @f(i32 [[X:%.*]]) +; CHECK: for.end: +; CHECK-NOT: mul nsw i32 [[X]], [[X]] +; CHECK: mul nsw i32 [[X]], [[X]] +; CHECK-NOT: mul nsw i32 [[X]], [[X]] |