diff options
| author | Cameron Zwarich <zwarich@apple.com> | 2011-04-02 02:40:43 +0000 |
|---|---|---|
| committer | Cameron Zwarich <zwarich@apple.com> | 2011-04-02 02:40:43 +0000 |
| commit | 6fe5c294302615875d9975d6f3d38f3b9bb294a2 (patch) | |
| tree | baf172921f60b9dd82a2ca15ce204267b51cb174 /llvm/test/CodeGen/ARM/fp-arg-shuffle.ll | |
| parent | 8c7bbc09e2287533d48e932b4c833be3cdc37409 (diff) | |
| download | bcm5719-llvm-6fe5c294302615875d9975d6f3d38f3b9bb294a2.tar.gz bcm5719-llvm-6fe5c294302615875d9975d6f3d38f3b9bb294a2.zip | |
Do some peephole optimizations to remove pointless VMOVs from Neon to integer
registers that arise from argument shuffling with the soft float ABI. These
instructions are particularly slow on Cortex A8. This fixes one half of
<rdar://problem/8674845>.
llvm-svn: 128759
Diffstat (limited to 'llvm/test/CodeGen/ARM/fp-arg-shuffle.ll')
| -rw-r--r-- | llvm/test/CodeGen/ARM/fp-arg-shuffle.ll | 11 |
1 files changed, 11 insertions, 0 deletions
diff --git a/llvm/test/CodeGen/ARM/fp-arg-shuffle.ll b/llvm/test/CodeGen/ARM/fp-arg-shuffle.ll new file mode 100644 index 00000000000..59303ac06da --- /dev/null +++ b/llvm/test/CodeGen/ARM/fp-arg-shuffle.ll @@ -0,0 +1,11 @@ +; RUN: llc < %s -march=arm -mattr=+neon -float-abi=soft | FileCheck %s + +; CHECK: function1 +; CHECK-NOT: vmov r +define double @function1(double %a, double %b, double %c, double %d, double %e, double %f) nounwind noinline ssp { +entry: + %call = tail call double @function2(double %f, double %e, double %d, double %c, double %b, double %a) nounwind + ret double %call +} + +declare double @function2(double, double, double, double, double, double) |
