mirror of
https://github.com/RPCS3/llvm-mirror.git
synced 2024-11-24 03:33:20 +01:00
7e501cf4c3
This update was done with the following bash script:
find test/CodeGen -name "*.ll" | \
while read NAME; do
echo "$NAME"
if ! grep -q "^; *RUN: *llc.*debug" $NAME; then
TEMP=`mktemp -t temp`
cp $NAME $TEMP
sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \
while read FUNC; do
sed -i '' "s/;\(.*\)\([A-Za-z0-9_-]*\):\( *\)$FUNC: *\$/;\1\2-LABEL:\3$FUNC:/g" $TEMP
done
sed -i '' "s/;\(.*\)-LABEL-LABEL:/;\1-LABEL:/" $TEMP
sed -i '' "s/;\(.*\)-NEXT-LABEL:/;\1-NEXT:/" $TEMP
sed -i '' "s/;\(.*\)-NOT-LABEL:/;\1-NOT:/" $TEMP
sed -i '' "s/;\(.*\)-DAG-LABEL:/;\1-DAG:/" $TEMP
mv $TEMP $NAME
fi
done
llvm-svn: 186280
76 lines
1.7 KiB
LLVM
76 lines
1.7 KiB
LLVM
; RUN: llc -mtriple armv7 %s -o - | FileCheck %s
|
|
|
|
; CHECK-LABEL: f:
|
|
define float @f(<4 x i16>* nocapture %in) {
|
|
; CHECK: vldr
|
|
; CHECK: vmovl.u16
|
|
%1 = load <4 x i16>* %in
|
|
; CHECK: vcvt.f32.u32
|
|
%2 = uitofp <4 x i16> %1 to <4 x float>
|
|
%3 = extractelement <4 x float> %2, i32 0
|
|
%4 = extractelement <4 x float> %2, i32 1
|
|
%5 = extractelement <4 x float> %2, i32 2
|
|
|
|
; CHECK: vadd.f32
|
|
%6 = fadd float %3, %4
|
|
%7 = fadd float %6, %5
|
|
|
|
ret float %7
|
|
}
|
|
|
|
; CHECK-LABEL: g:
|
|
define float @g(<4 x i8>* nocapture %in) {
|
|
; Note: vld1 here is reasonably important. Mixing VFP and NEON
|
|
; instructions is bad on some cores
|
|
; CHECK: vld1
|
|
; CHECK: vmovl.u8
|
|
; CHECK: vmovl.u16
|
|
%1 = load <4 x i8>* %in
|
|
; CHECK: vcvt.f32.u32
|
|
%2 = uitofp <4 x i8> %1 to <4 x float>
|
|
%3 = extractelement <4 x float> %2, i32 0
|
|
%4 = extractelement <4 x float> %2, i32 1
|
|
%5 = extractelement <4 x float> %2, i32 2
|
|
|
|
; CHECK: vadd.f32
|
|
%6 = fadd float %3, %4
|
|
%7 = fadd float %6, %5
|
|
|
|
ret float %7
|
|
}
|
|
|
|
; CHECK-LABEL: h:
|
|
define <4 x i8> @h(<4 x float> %v) {
|
|
; CHECK: vcvt.{{[us]}}32.f32
|
|
; CHECK: vmovn.i32
|
|
%1 = fptoui <4 x float> %v to <4 x i8>
|
|
ret <4 x i8> %1
|
|
}
|
|
|
|
; CHECK-LABEL: i:
|
|
define <4 x i8> @i(<4 x i8>* %x) {
|
|
; Note: vld1 here is reasonably important. Mixing VFP and NEON
|
|
; instructions is bad on some cores
|
|
; CHECK: vld1
|
|
; CHECK: vmovl.s8
|
|
; CHECK: vmovl.s16
|
|
; CHECK: vrecpe
|
|
; CHECK: vrecps
|
|
; CHECK: vmul
|
|
; CHECK: vmovn
|
|
%1 = load <4 x i8>* %x, align 4
|
|
%2 = sdiv <4 x i8> zeroinitializer, %1
|
|
ret <4 x i8> %2
|
|
}
|
|
; CHECK-LABEL: j:
|
|
define <4 x i32> @j(<4 x i8>* %in) nounwind {
|
|
; CHECK: vld1
|
|
; CHECK: vmovl.u8
|
|
; CHECK: vmovl.u16
|
|
; CHECK-NOT: vand
|
|
%1 = load <4 x i8>* %in, align 4
|
|
%2 = zext <4 x i8> %1 to <4 x i32>
|
|
ret <4 x i32> %2
|
|
}
|
|
|