diff options
| author | Chandler Carruth <chandlerc@gmail.com> | 2012-08-07 09:45:24 +0000 |
|---|---|---|
| committer | Chandler Carruth <chandlerc@gmail.com> | 2012-08-07 09:45:24 +0000 |
| commit | e6450dc2afc18531bf9b70180a9f67376d9f00c7 (patch) | |
| tree | 88d92e7932ae7528c097316ea06db60180d92d94 /test/Transforms/LoopStrengthReduce | |
| parent | ba86b13ad9cd6a9707a954598863da1e2a9f773b (diff) | |
| download | llvm-e6450dc2afc18531bf9b70180a9f67376d9f00c7.tar.gz llvm-e6450dc2afc18531bf9b70180a9f67376d9f00c7.tar.bz2 llvm-e6450dc2afc18531bf9b70180a9f67376d9f00c7.tar.xz | |
Add a much more conservative strategy for aligning branch targets.
Previously, MBP essentially aligned every branch target it could. This
bloats code quite a bit, especially non-looping code which has no real
reason to prefer aligned branch targets so heavily.
As Andy said in review, it's still a bit odd to do this without a real
cost model, but this at least has much more plausible heuristics.
Fixes PR13265.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@161409 91177308-0d34-0410-b5e6-96231b3b80d8
Diffstat (limited to 'test/Transforms/LoopStrengthReduce')
| -rw-r--r-- | test/Transforms/LoopStrengthReduce/X86/2012-01-13-phielim.ll | 1 |
1 files changed, 0 insertions, 1 deletions
diff --git a/test/Transforms/LoopStrengthReduce/X86/2012-01-13-phielim.ll b/test/Transforms/LoopStrengthReduce/X86/2012-01-13-phielim.ll index a9b815fc45..c3b8b8910d 100644 --- a/test/Transforms/LoopStrengthReduce/X86/2012-01-13-phielim.ll +++ b/test/Transforms/LoopStrengthReduce/X86/2012-01-13-phielim.ll @@ -99,7 +99,6 @@ while.end: ; preds = %entry ; CHECK: %for.body3.lr.ph.us.i.loopexit ; CHECK-NEXT: in Loop: Header ; CHECK-NEXT: incq -; CHECK-NEXT: .align ; CHECK-NEXT: %for.body3.us.i ; CHECK-NEXT: Inner Loop ; CHECK: testb |