diff options
| author | Craig Topper <craig.topper@intel.com> | 2019-10-28 10:11:20 -0700 |
|---|---|---|
| committer | Craig Topper <craig.topper@intel.com> | 2019-10-28 10:58:22 -0700 |
| commit | 18824d25d8aa8727d9f64f8002f2533d57627bd5 (patch) | |
| tree | b53d65ebfcd2a4e0b52401e9a685ab183452ad99 /llvm/test/Transforms/LoopVectorize | |
| parent | 3f345732b4f88e8d0d302470929cbe33d65a7435 (diff) | |
| download | bcm5719-llvm-18824d25d8aa8727d9f64f8002f2533d57627bd5.tar.gz bcm5719-llvm-18824d25d8aa8727d9f64f8002f2533d57627bd5.zip | |
[LV] Interleaving should not exceed estimated loop trip count.
Currently we may do iterleaving by more than estimated trip count
coming from the profile or computed maximum trip count. The solution is to
use "best known" trip count instead of exact one in interleaving analysis.
Patch by Evgeniy Brevnov.
Differential Revision: https://reviews.llvm.org/D67948
Diffstat (limited to 'llvm/test/Transforms/LoopVectorize')
| -rw-r--r-- | llvm/test/Transforms/LoopVectorize/X86/no_fpmath_with_hotness.ll | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/llvm/test/Transforms/LoopVectorize/X86/no_fpmath_with_hotness.ll b/llvm/test/Transforms/LoopVectorize/X86/no_fpmath_with_hotness.ll index fbc4ab2aa86..f593b9423e3 100644 --- a/llvm/test/Transforms/LoopVectorize/X86/no_fpmath_with_hotness.ll +++ b/llvm/test/Transforms/LoopVectorize/X86/no_fpmath_with_hotness.ll @@ -3,7 +3,7 @@ ; CHECK: remark: no_fpmath.c:6:11: loop not vectorized: cannot prove it is safe to reorder floating-point operations (hotness: 300) ; CHECK: remark: no_fpmath.c:6:14: loop not vectorized -; CHECK: remark: no_fpmath.c:17:14: vectorized loop (vectorization width: 2, interleaved count: 2) (hotness: 300) +; CHECK: remark: no_fpmath.c:17:14: vectorized loop (vectorization width: 2, interleaved count: 1) (hotness: 300) target datalayout = "e-m:o-i64:64-f80:128-n8:16:32:64-S128" target triple = "x86_64-apple-macosx10.10.0" |
