mirror of
https://github.com/c64scene-ar/llvm-6502.git
synced 2024-07-21 02:29:22 +00:00
eb32048f80
This patch adds supports for the vector constant folding of TRUNCATE and FP_EXTEND instructions and tidies up the SINT_TO_FP and UINT_TO_FP instructions to match. It also moves the vector constant folding for the FNEG and FABS instructions to use the DAG.getNode() functionality like the other unary instructions. Differential Revision: http://reviews.llvm.org/D8593 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@233224 91177308-0d34-0410-b5e6-96231b3b80d8
57 lines
1.9 KiB
LLVM
57 lines
1.9 KiB
LLVM
; RUN: llc < %s -march=x86 -mattr=+sse4.1,-avx | FileCheck %s
|
|
; RUN: llc < %s -march=x86 -mcpu=corei7-avx | FileCheck --check-prefix=AVX %s
|
|
|
|
; PR11674
|
|
define void @fpext_frommem(<2 x float>* %in, <2 x double>* %out) {
|
|
; CHECK-LABEL: fpext_frommem:
|
|
; AVX-LABEL: fpext_frommem:
|
|
entry:
|
|
; CHECK: cvtps2pd (%{{.+}}), %xmm{{[0-9]+}}
|
|
; AVX: vcvtps2pd (%{{.+}}), %xmm{{[0-9]+}}
|
|
%0 = load <2 x float>, <2 x float>* %in, align 8
|
|
%1 = fpext <2 x float> %0 to <2 x double>
|
|
store <2 x double> %1, <2 x double>* %out, align 1
|
|
ret void
|
|
}
|
|
|
|
define void @fpext_frommem4(<4 x float>* %in, <4 x double>* %out) {
|
|
; CHECK-LABEL: fpext_frommem4:
|
|
; AVX-LABEL: fpext_frommem4:
|
|
entry:
|
|
; CHECK: cvtps2pd (%{{.+}}), %xmm{{[0-9]+}}
|
|
; CHECK: cvtps2pd 8(%{{.+}}), %xmm{{[0-9]+}}
|
|
; AVX: vcvtps2pd (%{{.+}}), %ymm{{[0-9]+}}
|
|
%0 = load <4 x float>, <4 x float>* %in
|
|
%1 = fpext <4 x float> %0 to <4 x double>
|
|
store <4 x double> %1, <4 x double>* %out, align 1
|
|
ret void
|
|
}
|
|
|
|
define void @fpext_frommem8(<8 x float>* %in, <8 x double>* %out) {
|
|
; CHECK-LABEL: fpext_frommem8:
|
|
; AVX-LABEL: fpext_frommem8:
|
|
entry:
|
|
; CHECK: cvtps2pd (%{{.+}}), %xmm{{[0-9]+}}
|
|
; CHECK: cvtps2pd 8(%{{.+}}), %xmm{{[0-9]+}}
|
|
; CHECK: cvtps2pd 16(%{{.+}}), %xmm{{[0-9]+}}
|
|
; CHECK: cvtps2pd 24(%{{.+}}), %xmm{{[0-9]+}}
|
|
; AVX: vcvtps2pd 16(%{{.+}}), %ymm{{[0-9]+}}
|
|
; AVX: vcvtps2pd (%{{.+}}), %ymm{{[0-9]+}}
|
|
%0 = load <8 x float>, <8 x float>* %in
|
|
%1 = fpext <8 x float> %0 to <8 x double>
|
|
store <8 x double> %1, <8 x double>* %out, align 1
|
|
ret void
|
|
}
|
|
|
|
define <2 x double> @fpext_fromconst() {
|
|
; CHECK-LABEL: fpext_fromconst:
|
|
; AVX-LABEL: fpext_fromconst:
|
|
entry:
|
|
; CHECK: movaps {{.*#+}} xmm0 = [1.000000e+00,-2.000000e+00]
|
|
; AVX: vmovaps {{.*#+}} xmm0 = [1.000000e+00,-2.000000e+00]
|
|
%0 = insertelement <2 x float> undef, float 1.0, i32 0
|
|
%1 = insertelement <2 x float> %0, float -2.0, i32 1
|
|
%2 = fpext <2 x float> %1 to <2 x double>
|
|
ret <2 x double> %2
|
|
}
|