diff options
| author | Jeroen Ketema <j.ketema@imperial.ac.uk> | 2015-09-21 20:28:04 +0000 |
|---|---|---|
| committer | Jeroen Ketema <j.ketema@imperial.ac.uk> | 2015-09-21 20:28:04 +0000 |
| commit | 41681a5329ea53adc2e77f9fbaa6ecd2e908decc (patch) | |
| tree | c97c285b4fe3a35270eac79907a25b8f067446cf /llvm/test/CodeGen/ARM | |
| parent | 8abf7c87cdb5c4f6ef9a50c4f90372aa12b32f66 (diff) | |
| download | bcm5719-llvm-41681a5329ea53adc2e77f9fbaa6ecd2e908decc.tar.gz bcm5719-llvm-41681a5329ea53adc2e77f9fbaa6ecd2e908decc.zip | |
[ARM] Do not scale vext with a factor
The vext pseudo-instruction takes the number of elements that need to be
extracted, not the number of bytes. Hence, use the number of elements
directly instead of scaling them with a factor.
Reviewers: Silviu Baranga, James Molloy
(not reflected in the differential revision)
Differential Revision: http://reviews.llvm.org/D12974
llvm-svn: 248208
Diffstat (limited to 'llvm/test/CodeGen/ARM')
| -rw-r--r-- | llvm/test/CodeGen/ARM/vzip.ll | 11 |
1 files changed, 11 insertions, 0 deletions
diff --git a/llvm/test/CodeGen/ARM/vzip.ll b/llvm/test/CodeGen/ARM/vzip.ll index 24386a236f4..29385d0cc9d 100644 --- a/llvm/test/CodeGen/ARM/vzip.ll +++ b/llvm/test/CodeGen/ARM/vzip.ll @@ -305,3 +305,14 @@ entry: store <4 x i32> %0, <4 x i32>* %B ret void } + +define void @vzip_vext_factor(<8 x i16>* %A, <4 x i16>* %B) { +entry: + ; CHECK-LABEL: vzip_vext_factor + ; CHECK: vext.16 d16, d16, d17, #3 + ; CHECK: vzip + %tmp1 = load <8 x i16>, <8 x i16>* %A + %0 = shufflevector <8 x i16> %tmp1, <8 x i16> undef, <4 x i32> <i32 4, i32 4, i32 5, i32 3> + store <4 x i16> %0, <4 x i16>* %B + ret void +} |
