mirror of
https://github.com/RPCS3/llvm-mirror.git
synced 2025-01-31 20:51:52 +01:00
21418c85a1
If the mask input to getV4X86ShuffleImm8 only refers to a single source element (+ undefs) then canonicalize to a full broadcast. getV4X86ShuffleImm8 defaults to inline values for undefs, which can be useful for shuffle widening/narrowing but does leave SimplifyDemanded* calls thinking the shuffle depends on unnecessary elements. I'm still investigating what we should do more generally to avoid these undemanded elements, but broadcast cases was a simpler win.
188 lines
6.1 KiB
LLVM
188 lines
6.1 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc < %s -mtriple=i386-apple-darwin10 -mattr=+avx | FileCheck %s
|
|
|
|
define <8 x float> @cvt_v8i8_v8f32(<8 x i8> %src) {
|
|
; CHECK-LABEL: cvt_v8i8_v8f32:
|
|
; CHECK: ## %bb.0:
|
|
; CHECK-NEXT: vpmovsxbd %xmm0, %xmm1
|
|
; CHECK-NEXT: vpshufd {{.*#+}} xmm0 = xmm0[1,1,1,1]
|
|
; CHECK-NEXT: vpmovsxbd %xmm0, %xmm0
|
|
; CHECK-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
|
|
; CHECK-NEXT: vcvtdq2ps %ymm0, %ymm0
|
|
; CHECK-NEXT: retl
|
|
%res = sitofp <8 x i8> %src to <8 x float>
|
|
ret <8 x float> %res
|
|
}
|
|
|
|
define <8 x float> @cvt_v8i16_v8f32(<8 x i16> %src) {
|
|
; CHECK-LABEL: cvt_v8i16_v8f32:
|
|
; CHECK: ## %bb.0:
|
|
; CHECK-NEXT: vpmovsxwd %xmm0, %xmm1
|
|
; CHECK-NEXT: vpshufd {{.*#+}} xmm0 = xmm0[2,3,2,3]
|
|
; CHECK-NEXT: vpmovsxwd %xmm0, %xmm0
|
|
; CHECK-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
|
|
; CHECK-NEXT: vcvtdq2ps %ymm0, %ymm0
|
|
; CHECK-NEXT: retl
|
|
%res = sitofp <8 x i16> %src to <8 x float>
|
|
ret <8 x float> %res
|
|
}
|
|
|
|
define <4 x float> @cvt_v4i8_v4f32(<4 x i8> %src) {
|
|
; CHECK-LABEL: cvt_v4i8_v4f32:
|
|
; CHECK: ## %bb.0:
|
|
; CHECK-NEXT: vpmovsxbd %xmm0, %xmm0
|
|
; CHECK-NEXT: vcvtdq2ps %xmm0, %xmm0
|
|
; CHECK-NEXT: retl
|
|
%res = sitofp <4 x i8> %src to <4 x float>
|
|
ret <4 x float> %res
|
|
}
|
|
|
|
define <4 x float> @cvt_v4i16_v4f32(<4 x i16> %src) {
|
|
; CHECK-LABEL: cvt_v4i16_v4f32:
|
|
; CHECK: ## %bb.0:
|
|
; CHECK-NEXT: vpmovsxwd %xmm0, %xmm0
|
|
; CHECK-NEXT: vcvtdq2ps %xmm0, %xmm0
|
|
; CHECK-NEXT: retl
|
|
%res = sitofp <4 x i16> %src to <4 x float>
|
|
ret <4 x float> %res
|
|
}
|
|
|
|
define <8 x float> @cvt_v8u8_v8f32(<8 x i8> %src) {
|
|
; CHECK-LABEL: cvt_v8u8_v8f32:
|
|
; CHECK: ## %bb.0:
|
|
; CHECK-NEXT: vpmovzxbd {{.*#+}} xmm1 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero
|
|
; CHECK-NEXT: vpshufd {{.*#+}} xmm0 = xmm0[1,1,1,1]
|
|
; CHECK-NEXT: vpmovzxbd {{.*#+}} xmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero
|
|
; CHECK-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
|
|
; CHECK-NEXT: vcvtdq2ps %ymm0, %ymm0
|
|
; CHECK-NEXT: retl
|
|
%res = uitofp <8 x i8> %src to <8 x float>
|
|
ret <8 x float> %res
|
|
}
|
|
|
|
define <8 x float> @cvt_v8u16_v8f32(<8 x i16> %src) {
|
|
; CHECK-LABEL: cvt_v8u16_v8f32:
|
|
; CHECK: ## %bb.0:
|
|
; CHECK-NEXT: vpxor %xmm1, %xmm1, %xmm1
|
|
; CHECK-NEXT: vpunpckhwd {{.*#+}} xmm1 = xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
|
|
; CHECK-NEXT: vpmovzxwd {{.*#+}} xmm0 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero
|
|
; CHECK-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
|
|
; CHECK-NEXT: vcvtdq2ps %ymm0, %ymm0
|
|
; CHECK-NEXT: retl
|
|
%res = uitofp <8 x i16> %src to <8 x float>
|
|
ret <8 x float> %res
|
|
}
|
|
|
|
define <4 x float> @cvt_v4u8_v4f32(<4 x i8> %src) {
|
|
; CHECK-LABEL: cvt_v4u8_v4f32:
|
|
; CHECK: ## %bb.0:
|
|
; CHECK-NEXT: vpmovzxbd {{.*#+}} xmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero
|
|
; CHECK-NEXT: vcvtdq2ps %xmm0, %xmm0
|
|
; CHECK-NEXT: retl
|
|
%res = uitofp <4 x i8> %src to <4 x float>
|
|
ret <4 x float> %res
|
|
}
|
|
|
|
define <4 x float> @cvt_v4u16_v4f32(<4 x i16> %src) {
|
|
; CHECK-LABEL: cvt_v4u16_v4f32:
|
|
; CHECK: ## %bb.0:
|
|
; CHECK-NEXT: vpmovzxwd {{.*#+}} xmm0 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero
|
|
; CHECK-NEXT: vcvtdq2ps %xmm0, %xmm0
|
|
; CHECK-NEXT: retl
|
|
%res = uitofp <4 x i16> %src to <4 x float>
|
|
ret <4 x float> %res
|
|
}
|
|
|
|
define <8 x i8> @cvt_v8f32_v8i8(<8 x float> %src) {
|
|
; CHECK-LABEL: cvt_v8f32_v8i8:
|
|
; CHECK: ## %bb.0:
|
|
; CHECK-NEXT: vcvttps2dq %ymm0, %ymm0
|
|
; CHECK-NEXT: vextractf128 $1, %ymm0, %xmm1
|
|
; CHECK-NEXT: vpackssdw %xmm1, %xmm0, %xmm0
|
|
; CHECK-NEXT: vpacksswb %xmm0, %xmm0, %xmm0
|
|
; CHECK-NEXT: vzeroupper
|
|
; CHECK-NEXT: retl
|
|
%res = fptosi <8 x float> %src to <8 x i8>
|
|
ret <8 x i8> %res
|
|
}
|
|
|
|
define <8 x i16> @cvt_v8f32_v8i16(<8 x float> %src) {
|
|
; CHECK-LABEL: cvt_v8f32_v8i16:
|
|
; CHECK: ## %bb.0:
|
|
; CHECK-NEXT: vcvttps2dq %ymm0, %ymm0
|
|
; CHECK-NEXT: vextractf128 $1, %ymm0, %xmm1
|
|
; CHECK-NEXT: vpackssdw %xmm1, %xmm0, %xmm0
|
|
; CHECK-NEXT: vzeroupper
|
|
; CHECK-NEXT: retl
|
|
%res = fptosi <8 x float> %src to <8 x i16>
|
|
ret <8 x i16> %res
|
|
}
|
|
|
|
define <4 x i8> @cvt_v4f32_v4i8(<4 x float> %src) {
|
|
; CHECK-LABEL: cvt_v4f32_v4i8:
|
|
; CHECK: ## %bb.0:
|
|
; CHECK-NEXT: vcvttps2dq %xmm0, %xmm0
|
|
; CHECK-NEXT: vpackssdw %xmm0, %xmm0, %xmm0
|
|
; CHECK-NEXT: vpacksswb %xmm0, %xmm0, %xmm0
|
|
; CHECK-NEXT: retl
|
|
%res = fptosi <4 x float> %src to <4 x i8>
|
|
ret <4 x i8> %res
|
|
}
|
|
|
|
define <4 x i16> @cvt_v4f32_v4i16(<4 x float> %src) {
|
|
; CHECK-LABEL: cvt_v4f32_v4i16:
|
|
; CHECK: ## %bb.0:
|
|
; CHECK-NEXT: vcvttps2dq %xmm0, %xmm0
|
|
; CHECK-NEXT: vpackssdw %xmm0, %xmm0, %xmm0
|
|
; CHECK-NEXT: retl
|
|
%res = fptosi <4 x float> %src to <4 x i16>
|
|
ret <4 x i16> %res
|
|
}
|
|
|
|
define <8 x i8> @cvt_v8f32_v8u8(<8 x float> %src) {
|
|
; CHECK-LABEL: cvt_v8f32_v8u8:
|
|
; CHECK: ## %bb.0:
|
|
; CHECK-NEXT: vcvttps2dq %ymm0, %ymm0
|
|
; CHECK-NEXT: vextractf128 $1, %ymm0, %xmm1
|
|
; CHECK-NEXT: vpackssdw %xmm1, %xmm0, %xmm0
|
|
; CHECK-NEXT: vpackuswb %xmm0, %xmm0, %xmm0
|
|
; CHECK-NEXT: vzeroupper
|
|
; CHECK-NEXT: retl
|
|
%res = fptoui <8 x float> %src to <8 x i8>
|
|
ret <8 x i8> %res
|
|
}
|
|
|
|
define <8 x i16> @cvt_v8f32_v8u16(<8 x float> %src) {
|
|
; CHECK-LABEL: cvt_v8f32_v8u16:
|
|
; CHECK: ## %bb.0:
|
|
; CHECK-NEXT: vcvttps2dq %ymm0, %ymm0
|
|
; CHECK-NEXT: vextractf128 $1, %ymm0, %xmm1
|
|
; CHECK-NEXT: vpackusdw %xmm1, %xmm0, %xmm0
|
|
; CHECK-NEXT: vzeroupper
|
|
; CHECK-NEXT: retl
|
|
%res = fptoui <8 x float> %src to <8 x i16>
|
|
ret <8 x i16> %res
|
|
}
|
|
|
|
define <4 x i8> @cvt_v4f32_v4u8(<4 x float> %src) {
|
|
; CHECK-LABEL: cvt_v4f32_v4u8:
|
|
; CHECK: ## %bb.0:
|
|
; CHECK-NEXT: vcvttps2dq %xmm0, %xmm0
|
|
; CHECK-NEXT: vpackusdw %xmm0, %xmm0, %xmm0
|
|
; CHECK-NEXT: vpackuswb %xmm0, %xmm0, %xmm0
|
|
; CHECK-NEXT: retl
|
|
%res = fptoui <4 x float> %src to <4 x i8>
|
|
ret <4 x i8> %res
|
|
}
|
|
|
|
define <4 x i16> @cvt_v4f32_v4u16(<4 x float> %src) {
|
|
; CHECK-LABEL: cvt_v4f32_v4u16:
|
|
; CHECK: ## %bb.0:
|
|
; CHECK-NEXT: vcvttps2dq %xmm0, %xmm0
|
|
; CHECK-NEXT: vpackusdw %xmm0, %xmm0, %xmm0
|
|
; CHECK-NEXT: retl
|
|
%res = fptoui <4 x float> %src to <4 x i16>
|
|
ret <4 x i16> %res
|
|
}
|
|
|