maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity

[X86][SSE] Added additional fp/int tests. 

Demonstrates some shortfalls in subvector(cvt(x)) compared to cvt(subvector(x)) patterns - especially on AVX/AVX2 targets.

llvm-svn: 242614
This commit is contained in:
Simon Pilgrim 2015-07-18 17:05:39 +00:00
parent c77f72c4e4
commit 8c81678dfa
2 changed files with 951 additions and 60 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

289
llvm/test/CodeGen/X86/vec_fp_to_int.ll
View File

@ -62,6 +62,33 @@ define <4 x i32> @fptosi_2f64_to_2i32(<2 x double> %a) {
ret <4 x i32> %ext
}
define <4 x i32> @fptosi_4f64_to_2i32(<2 x double> %a) {
; SSE-LABEL: fptosi_4f64_to_2i32:
; SSE: # BB#0:
; SSE-NEXT: cvttsd2si %xmm0, %rax
; SSE-NEXT: movd %rax, %xmm1
; SSE-NEXT: shufpd {{.*#+}} xmm0 = xmm0[1,0]
; SSE-NEXT: cvttsd2si %xmm0, %rax
; SSE-NEXT: movd %rax, %xmm0
; SSE-NEXT: punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm1[0,2,2,3]
; SSE-NEXT: cvttsd2si %xmm0, %rax
; SSE-NEXT: movd %rax, %xmm1
; SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm1[0,1,0,1]
; SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm1[0,2,2,3]
; SSE-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE-NEXT: retq
;
; AVX-LABEL: fptosi_4f64_to_2i32:
; AVX: # BB#0:
; AVX-NEXT: vcvttpd2dqy %ymm0, %xmm0
; AVX-NEXT: vzeroupper
; AVX-NEXT: retq
%ext = shufflevector <2 x double> %a, <2 x double> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
%cvt = fptosi <4 x double> %ext to <4 x i32>
ret <4 x i32> %cvt
}
define <4 x i64> @fptosi_4f64_to_4i64(<4 x double> %a) {
; SSE-LABEL: fptosi_4f64_to_4i64:
; SSE: # BB#0:
@ -239,6 +266,56 @@ define <4 x i32> @fptoui_2f64_to_2i32(<2 x double> %a) {
ret <4 x i32> %ext
}
define <4 x i32> @fptoui_4f64_to_2i32(<2 x double> %a) {
; SSE-LABEL: fptoui_4f64_to_2i32:
; SSE: # BB#0:
; SSE-NEXT: movsd {{.*#+}} xmm1 = mem[0],zero
; SSE-NEXT: movapd %xmm0, %xmm2
; SSE-NEXT: subsd %xmm1, %xmm2
; SSE-NEXT: cvttsd2si %xmm2, %rax
; SSE-NEXT: movabsq $-9223372036854775808, %rcx # imm = 0x8000000000000000
; SSE-NEXT: xorq %rcx, %rax
; SSE-NEXT: cvttsd2si %xmm0, %rdx
; SSE-NEXT: ucomisd %xmm1, %xmm0
; SSE-NEXT: cmovaeq %rax, %rdx
; SSE-NEXT: movd %rdx, %xmm2
; SSE-NEXT: shufpd {{.*#+}} xmm0 = xmm0[1,0]
; SSE-NEXT: movapd %xmm0, %xmm3
; SSE-NEXT: subsd %xmm1, %xmm3
; SSE-NEXT: cvttsd2si %xmm3, %rax
; SSE-NEXT: xorq %rcx, %rax
; SSE-NEXT: cvttsd2si %xmm0, %rdx
; SSE-NEXT: ucomisd %xmm1, %xmm0
; SSE-NEXT: cmovaeq %rax, %rdx
; SSE-NEXT: movd %rdx, %xmm0
; SSE-NEXT: punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE-NEXT: pshufd {{.*#+}} xmm0 = xmm2[0,2,2,3]
; SSE-NEXT: cvttsd2si %xmm0, %rax
; SSE-NEXT: xorq %rax, %rcx
; SSE-NEXT: ucomisd %xmm1, %xmm0
; SSE-NEXT: cmovbq %rax, %rcx
; SSE-NEXT: movd %rcx, %xmm1
; SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm1[0,1,0,1]
; SSE-NEXT: pshufd {{.*#+}} xmm1 = xmm1[0,2,2,3]
; SSE-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE-NEXT: retq
;
; AVX-LABEL: fptoui_4f64_to_2i32:
; AVX: # BB#0:
; AVX-NEXT: vpermilpd {{.*#+}} xmm1 = xmm0[1,0]
; AVX-NEXT: vcvttsd2si %xmm1, %rax
; AVX-NEXT: vcvttsd2si %xmm0, %rcx
; AVX-NEXT: vmovd %ecx, %xmm0
; AVX-NEXT: vpinsrd $1, %eax, %xmm0, %xmm0
; AVX-NEXT: vcvttsd2si %xmm0, %rax
; AVX-NEXT: vpinsrd $2, %eax, %xmm0, %xmm0
; AVX-NEXT: vpinsrd $3, %eax, %xmm0, %xmm0
; AVX-NEXT: retq
%ext = shufflevector <2 x double> %a, <2 x double> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
%cvt = fptoui <4 x double> %ext to <4 x i32>
ret <4 x i32> %cvt
}
define <4 x i64> @fptoui_4f64_to_4i64(<4 x double> %a) {
; SSE-LABEL: fptoui_4f64_to_4i64:
; SSE: # BB#0:
@ -435,6 +512,32 @@ define <2 x i64> @fptosi_2f32_to_2i64(<4 x float> %a) {
ret <2 x i64> %cvt
}
define <2 x i64> @fptosi_4f32_to_2i64(<4 x float> %a) {
; SSE-LABEL: fptosi_4f32_to_2i64:
; SSE: # BB#0:
; SSE-NEXT: cvttss2si %xmm0, %rax
; SSE-NEXT: movd %rax, %xmm1
; SSE-NEXT: shufps {{.*#+}} xmm0 = xmm0[1,1,2,3]
; SSE-NEXT: cvttss2si %xmm0, %rax
; SSE-NEXT: movd %rax, %xmm0
; SSE-NEXT: punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE-NEXT: movdqa %xmm1, %xmm0
; SSE-NEXT: retq
;
; AVX-LABEL: fptosi_4f32_to_2i64:
; AVX: # BB#0:
; AVX-NEXT: vmovshdup {{.*#+}} xmm1 = xmm0[1,1,3,3]
; AVX-NEXT: vcvttss2si %xmm1, %rax
; AVX-NEXT: vcvttss2si %xmm0, %rcx
; AVX-NEXT: vmovq %rcx, %xmm0
; AVX-NEXT: vmovq %rax, %xmm1
; AVX-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX-NEXT: retq
%cvt = fptosi <4 x float> %a to <4 x i64>
%shuf = shufflevector <4 x i64> %cvt, <4 x i64> undef, <2 x i32> <i32 0, i32 1>
ret <2 x i64> %shuf
}
define <8 x i32> @fptosi_8f32_to_8i32(<8 x float> %a) {
; SSE-LABEL: fptosi_8f32_to_8i32:
; SSE: # BB#0:
@ -493,6 +596,49 @@ define <4 x i64> @fptosi_4f32_to_4i64(<8 x float> %a) {
ret <4 x i64> %cvt
}
define <4 x i64> @fptosi_8f32_to_4i64(<8 x float> %a) {
; SSE-LABEL: fptosi_8f32_to_4i64:
; SSE: # BB#0:
; SSE-NEXT: cvttss2si %xmm0, %rax
; SSE-NEXT: movd %rax, %xmm2
; SSE-NEXT: movaps %xmm0, %xmm1
; SSE-NEXT: shufps {{.*#+}} xmm1 = xmm1[1,1,2,3]
; SSE-NEXT: cvttss2si %xmm1, %rax
; SSE-NEXT: movd %rax, %xmm1
; SSE-NEXT: punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm1[0]
; SSE-NEXT: movaps %xmm0, %xmm1
; SSE-NEXT: shufps {{.*#+}} xmm1 = xmm1[3,1,2,3]
; SSE-NEXT: cvttss2si %xmm1, %rax
; SSE-NEXT: movd %rax, %xmm3
; SSE-NEXT: shufpd {{.*#+}} xmm0 = xmm0[1,0]
; SSE-NEXT: cvttss2si %xmm0, %rax
; SSE-NEXT: movd %rax, %xmm1
; SSE-NEXT: punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm3[0]
; SSE-NEXT: movdqa %xmm2, %xmm0
; SSE-NEXT: retq
;
; AVX-LABEL: fptosi_8f32_to_4i64:
; AVX: # BB#0:
; AVX-NEXT: vpermilps {{.*#+}} xmm1 = xmm0[3,1,2,3]
; AVX-NEXT: vcvttss2si %xmm1, %rax
; AVX-NEXT: vmovq %rax, %xmm1
; AVX-NEXT: vpermilpd {{.*#+}} xmm2 = xmm0[1,0]
; AVX-NEXT: vcvttss2si %xmm2, %rax
; AVX-NEXT: vmovq %rax, %xmm2
; AVX-NEXT: vpunpcklqdq {{.*#+}} xmm1 = xmm2[0],xmm1[0]
; AVX-NEXT: vcvttss2si %xmm0, %rax
; AVX-NEXT: vmovq %rax, %xmm2
; AVX-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX-NEXT: vcvttss2si %xmm0, %rax
; AVX-NEXT: vmovq %rax, %xmm0
; AVX-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm2[0],xmm0[0]
; AVX-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
; AVX-NEXT: retq
%cvt = fptosi <8 x float> %a to <8 x i64>
%shuf = shufflevector <8 x i64> %cvt, <8 x i64> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
ret <4 x i64> %shuf
}
;
; Float to Unsigned Integer
;
@ -589,6 +735,58 @@ define <2 x i64> @fptoui_2f32_to_2i64(<4 x float> %a) {
ret <2 x i64> %cvt
}
define <2 x i64> @fptoui_4f32_to_2i64(<4 x float> %a) {
; SSE-LABEL: fptoui_4f32_to_2i64:
; SSE: # BB#0:
; SSE-NEXT: movss {{.*#+}} xmm2 = mem[0],zero,zero,zero
; SSE-NEXT: movaps %xmm0, %xmm1
; SSE-NEXT: subss %xmm2, %xmm1
; SSE-NEXT: cvttss2si %xmm1, %rax
; SSE-NEXT: movabsq $-9223372036854775808, %rcx # imm = 0x8000000000000000
; SSE-NEXT: xorq %rcx, %rax
; SSE-NEXT: cvttss2si %xmm0, %rdx
; SSE-NEXT: ucomiss %xmm2, %xmm0
; SSE-NEXT: cmovaeq %rax, %rdx
; SSE-NEXT: movd %rdx, %xmm1
; SSE-NEXT: shufps {{.*#+}} xmm0 = xmm0[1,1,2,3]
; SSE-NEXT: movaps %xmm0, %xmm3
; SSE-NEXT: subss %xmm2, %xmm3
; SSE-NEXT: cvttss2si %xmm3, %rax
; SSE-NEXT: xorq %rcx, %rax
; SSE-NEXT: cvttss2si %xmm0, %rcx
; SSE-NEXT: ucomiss %xmm2, %xmm0
; SSE-NEXT: cmovaeq %rax, %rcx
; SSE-NEXT: movd %rcx, %xmm0
; SSE-NEXT: punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE-NEXT: movdqa %xmm1, %xmm0
; SSE-NEXT: retq
;
; AVX-LABEL: fptoui_4f32_to_2i64:
; AVX: # BB#0:
; AVX-NEXT: vmovshdup {{.*#+}} xmm1 = xmm0[1,1,3,3]
; AVX-NEXT: vmovss {{.*#+}} xmm2 = mem[0],zero,zero,zero
; AVX-NEXT: vsubss %xmm2, %xmm1, %xmm3
; AVX-NEXT: vcvttss2si %xmm3, %rax
; AVX-NEXT: movabsq $-9223372036854775808, %rcx # imm = 0x8000000000000000
; AVX-NEXT: xorq %rcx, %rax
; AVX-NEXT: vcvttss2si %xmm1, %rdx
; AVX-NEXT: vucomiss %xmm2, %xmm1
; AVX-NEXT: cmovaeq %rax, %rdx
; AVX-NEXT: vsubss %xmm2, %xmm0, %xmm1
; AVX-NEXT: vcvttss2si %xmm1, %rax
; AVX-NEXT: xorq %rcx, %rax
; AVX-NEXT: vcvttss2si %xmm0, %rcx
; AVX-NEXT: vucomiss %xmm2, %xmm0
; AVX-NEXT: cmovaeq %rax, %rcx
; AVX-NEXT: vmovq %rcx, %xmm0
; AVX-NEXT: vmovq %rdx, %xmm1
; AVX-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX-NEXT: retq
%cvt = fptoui <4 x float> %a to <4 x i64>
%shuf = shufflevector <4 x i64> %cvt, <4 x i64> undef, <2 x i32> <i32 0, i32 1>
ret <2 x i64> %shuf
}
define <8 x i32> @fptoui_8f32_to_8i32(<8 x float> %a) {
; SSE-LABEL: fptoui_8f32_to_8i32:
; SSE: # BB#0:
@ -749,6 +947,97 @@ define <4 x i64> @fptoui_4f32_to_4i64(<8 x float> %a) {
ret <4 x i64> %cvt
}
define <4 x i64> @fptoui_8f32_to_4i64(<8 x float> %a) {
; SSE-LABEL: fptoui_8f32_to_4i64:
; SSE: # BB#0:
; SSE-NEXT: movss {{.*#+}} xmm1 = mem[0],zero,zero,zero
; SSE-NEXT: movaps %xmm0, %xmm2
; SSE-NEXT: subss %xmm1, %xmm2
; SSE-NEXT: cvttss2si %xmm2, %rcx
; SSE-NEXT: movabsq $-9223372036854775808, %rax # imm = 0x8000000000000000
; SSE-NEXT: xorq %rax, %rcx
; SSE-NEXT: cvttss2si %xmm0, %rdx
; SSE-NEXT: ucomiss %xmm1, %xmm0
; SSE-NEXT: cmovaeq %rcx, %rdx
; SSE-NEXT: movd %rdx, %xmm2
; SSE-NEXT: movaps %xmm0, %xmm3
; SSE-NEXT: shufps {{.*#+}} xmm3 = xmm3[1,1,2,3]
; SSE-NEXT: movaps %xmm3, %xmm4
; SSE-NEXT: subss %xmm1, %xmm4
; SSE-NEXT: cvttss2si %xmm4, %rcx
; SSE-NEXT: xorq %rax, %rcx
; SSE-NEXT: cvttss2si %xmm3, %rdx
; SSE-NEXT: ucomiss %xmm1, %xmm3
; SSE-NEXT: cmovaeq %rcx, %rdx
; SSE-NEXT: movd %rdx, %xmm3
; SSE-NEXT: punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm3[0]
; SSE-NEXT: movaps %xmm0, %xmm3
; SSE-NEXT: shufps {{.*#+}} xmm3 = xmm3[3,1,2,3]
; SSE-NEXT: movaps %xmm3, %xmm4
; SSE-NEXT: subss %xmm1, %xmm4
; SSE-NEXT: cvttss2si %xmm4, %rcx
; SSE-NEXT: xorq %rax, %rcx
; SSE-NEXT: cvttss2si %xmm3, %rdx
; SSE-NEXT: ucomiss %xmm1, %xmm3
; SSE-NEXT: cmovaeq %rcx, %rdx
; SSE-NEXT: movd %rdx, %xmm3
; SSE-NEXT: shufpd {{.*#+}} xmm0 = xmm0[1,0]
; SSE-NEXT: movapd %xmm0, %xmm4
; SSE-NEXT: subss %xmm1, %xmm4
; SSE-NEXT: cvttss2si %xmm4, %rcx
; SSE-NEXT: xorq %rax, %rcx
; SSE-NEXT: cvttss2si %xmm0, %rax
; SSE-NEXT: ucomiss %xmm1, %xmm0
; SSE-NEXT: cmovaeq %rcx, %rax
; SSE-NEXT: movd %rax, %xmm1
; SSE-NEXT: punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm3[0]
; SSE-NEXT: movdqa %xmm2, %xmm0
; SSE-NEXT: retq
;
; AVX-LABEL: fptoui_8f32_to_4i64:
; AVX: # BB#0:
; AVX-NEXT: vpermilps {{.*#+}} xmm2 = xmm0[3,1,2,3]
; AVX-NEXT: vmovss {{.*#+}} xmm1 = mem[0],zero,zero,zero
; AVX-NEXT: vsubss %xmm1, %xmm2, %xmm3
; AVX-NEXT: vcvttss2si %xmm3, %rax
; AVX-NEXT: movabsq $-9223372036854775808, %rcx # imm = 0x8000000000000000
; AVX-NEXT: xorq %rcx, %rax
; AVX-NEXT: vcvttss2si %xmm2, %rdx
; AVX-NEXT: vucomiss %xmm1, %xmm2
; AVX-NEXT: cmovaeq %rax, %rdx
; AVX-NEXT: vmovq %rdx, %xmm2
; AVX-NEXT: vpermilpd {{.*#+}} xmm3 = xmm0[1,0]
; AVX-NEXT: vsubss %xmm1, %xmm3, %xmm4
; AVX-NEXT: vcvttss2si %xmm4, %rax
; AVX-NEXT: xorq %rcx, %rax
; AVX-NEXT: vcvttss2si %xmm3, %rdx
; AVX-NEXT: vucomiss %xmm1, %xmm3
; AVX-NEXT: cmovaeq %rax, %rdx
; AVX-NEXT: vmovq %rdx, %xmm3
; AVX-NEXT: vpunpcklqdq {{.*#+}} xmm2 = xmm3[0],xmm2[0]
; AVX-NEXT: vsubss %xmm1, %xmm0, %xmm3
; AVX-NEXT: vcvttss2si %xmm3, %rax
; AVX-NEXT: xorq %rcx, %rax
; AVX-NEXT: vcvttss2si %xmm0, %rdx
; AVX-NEXT: vucomiss %xmm1, %xmm0
; AVX-NEXT: cmovaeq %rax, %rdx
; AVX-NEXT: vmovq %rdx, %xmm3
; AVX-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX-NEXT: vsubss %xmm1, %xmm0, %xmm4
; AVX-NEXT: vcvttss2si %xmm4, %rax
; AVX-NEXT: xorq %rcx, %rax
; AVX-NEXT: vcvttss2si %xmm0, %rcx
; AVX-NEXT: vucomiss %xmm1, %xmm0
; AVX-NEXT: cmovaeq %rax, %rcx
; AVX-NEXT: vmovq %rcx, %xmm0
; AVX-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm3[0],xmm0[0]
; AVX-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
; AVX-NEXT: retq
%cvt = fptoui <8 x float> %a to <8 x i64>
%shuf = shufflevector <8 x i64> %cvt, <8 x i64> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
ret <4 x i64> %shuf
}
;
; Constant Folding
;

722
llvm/test/CodeGen/X86/vec_int_to_fp.ll
View File

File diff suppressed because it is too large Load Diff