extract-concat.ll
7.21 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=x86_64-- -mattr=sse4.2 | FileCheck %s --check-prefixes=SSE42
; RUN: llc < %s -mtriple=x86_64-- -mattr=avx | FileCheck %s --check-prefixes=AVX,AVX1
; RUN: llc < %s -mtriple=x86_64-- -mattr=avx2 | FileCheck %s --check-prefixes=AVX,AVX2
; RUN: llc < %s -mtriple=x86_64-- -mattr=avx512f | FileCheck %s --check-prefixes=AVX,AVX512F
define void @foo(<4 x float> %in, <4 x i8>* %out) {
; SSE42-LABEL: foo:
; SSE42: # %bb.0:
; SSE42-NEXT: cvttps2dq %xmm0, %xmm0
; SSE42-NEXT: pextrb $8, %xmm0, %eax
; SSE42-NEXT: pextrb $4, %xmm0, %ecx
; SSE42-NEXT: pextrb $0, %xmm0, %edx
; SSE42-NEXT: movd %edx, %xmm0
; SSE42-NEXT: pinsrb $1, %ecx, %xmm0
; SSE42-NEXT: pinsrb $2, %eax, %xmm0
; SSE42-NEXT: movl $255, %eax
; SSE42-NEXT: pinsrb $3, %eax, %xmm0
; SSE42-NEXT: movd %xmm0, (%rdi)
; SSE42-NEXT: retq
;
; AVX-LABEL: foo:
; AVX: # %bb.0:
; AVX-NEXT: vcvttps2dq %xmm0, %xmm0
; AVX-NEXT: vpextrb $8, %xmm0, %eax
; AVX-NEXT: vpextrb $4, %xmm0, %ecx
; AVX-NEXT: vpextrb $0, %xmm0, %edx
; AVX-NEXT: vmovd %edx, %xmm0
; AVX-NEXT: vpinsrb $1, %ecx, %xmm0, %xmm0
; AVX-NEXT: vpinsrb $2, %eax, %xmm0, %xmm0
; AVX-NEXT: movl $255, %eax
; AVX-NEXT: vpinsrb $3, %eax, %xmm0, %xmm0
; AVX-NEXT: vmovd %xmm0, (%rdi)
; AVX-NEXT: retq
%t0 = fptosi <4 x float> %in to <4 x i32>
%t1 = trunc <4 x i32> %t0 to <4 x i16>
%t2 = shufflevector <4 x i16> %t1, <4 x i16> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
%t3 = trunc <8 x i16> %t2 to <8 x i8>
%t4 = shufflevector <8 x i8> %t3, <8 x i8> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
%t5 = insertelement <4 x i8> %t4, i8 -1, i32 3
store <4 x i8> %t5, <4 x i8>* %out
ret void
}
define <16 x i64> @catcat(<4 x i64> %x) {
; SSE42-LABEL: catcat:
; SSE42: # %bb.0:
; SSE42-NEXT: movq %rdi, %rax
; SSE42-NEXT: pshufd {{.*#+}} xmm2 = xmm0[0,1,0,1]
; SSE42-NEXT: pshufd {{.*#+}} xmm0 = xmm0[2,3,2,3]
; SSE42-NEXT: pshufd {{.*#+}} xmm3 = xmm1[0,1,0,1]
; SSE42-NEXT: pshufd {{.*#+}} xmm1 = xmm1[2,3,2,3]
; SSE42-NEXT: movdqa %xmm1, 112(%rdi)
; SSE42-NEXT: movdqa %xmm1, 96(%rdi)
; SSE42-NEXT: movdqa %xmm3, 80(%rdi)
; SSE42-NEXT: movdqa %xmm3, 64(%rdi)
; SSE42-NEXT: movdqa %xmm0, 48(%rdi)
; SSE42-NEXT: movdqa %xmm0, 32(%rdi)
; SSE42-NEXT: movdqa %xmm2, 16(%rdi)
; SSE42-NEXT: movdqa %xmm2, (%rdi)
; SSE42-NEXT: retq
;
; AVX1-LABEL: catcat:
; AVX1: # %bb.0:
; AVX1-NEXT: vmovddup {{.*#+}} ymm1 = ymm0[0,0,2,2]
; AVX1-NEXT: vperm2f128 {{.*#+}} ymm2 = ymm1[2,3,2,3]
; AVX1-NEXT: vpermilpd {{.*#+}} ymm1 = ymm0[1,1,3,3]
; AVX1-NEXT: vperm2f128 {{.*#+}} ymm3 = ymm1[2,3,2,3]
; AVX1-NEXT: vpermilps {{.*#+}} xmm1 = xmm0[0,1,0,1]
; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm1, %ymm4
; AVX1-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[2,3,2,3]
; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm1
; AVX1-NEXT: vmovaps %ymm4, %ymm0
; AVX1-NEXT: retq
;
; AVX2-LABEL: catcat:
; AVX2: # %bb.0:
; AVX2-NEXT: vpermpd {{.*#+}} ymm1 = ymm0[1,1,1,1]
; AVX2-NEXT: vpermpd {{.*#+}} ymm2 = ymm0[2,2,2,2]
; AVX2-NEXT: vpermpd {{.*#+}} ymm3 = ymm0[3,3,3,3]
; AVX2-NEXT: vbroadcastsd %xmm0, %ymm0
; AVX2-NEXT: retq
;
; AVX512F-LABEL: catcat:
; AVX512F: # %bb.0:
; AVX512F-NEXT: # kill: def $ymm0 killed $ymm0 def $zmm0
; AVX512F-NEXT: vmovaps {{.*#+}} zmm1 = [0,0,0,0,1,1,1,1]
; AVX512F-NEXT: vpermpd %zmm0, %zmm1, %zmm2
; AVX512F-NEXT: vmovaps {{.*#+}} zmm1 = [2,2,2,2,3,3,3,3]
; AVX512F-NEXT: vpermpd %zmm0, %zmm1, %zmm1
; AVX512F-NEXT: vmovaps %zmm2, %zmm0
; AVX512F-NEXT: retq
%cat1 = shufflevector <4 x i64> %x, <4 x i64> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3>
%cat2 = shufflevector <8 x i64> %cat1, <8 x i64> undef, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
%r = shufflevector <16 x i64> %cat2, <16 x i64> undef, <16 x i32> <i32 0, i32 4, i32 8, i32 12, i32 1, i32 5, i32 9, i32 13, i32 2, i32 6, i32 10, i32 14, i32 3, i32 7, i32 11, i32 15>
ret <16 x i64> %r
}
define <16 x i64> @load_catcat(<4 x i64>* %p) {
; SSE42-LABEL: load_catcat:
; SSE42: # %bb.0:
; SSE42-NEXT: movq %rdi, %rax
; SSE42-NEXT: movdqa (%rsi), %xmm0
; SSE42-NEXT: movdqa 16(%rsi), %xmm1
; SSE42-NEXT: pshufd {{.*#+}} xmm2 = xmm0[0,1,0,1]
; SSE42-NEXT: pshufd {{.*#+}} xmm0 = xmm0[2,3,2,3]
; SSE42-NEXT: pshufd {{.*#+}} xmm3 = xmm1[0,1,0,1]
; SSE42-NEXT: pshufd {{.*#+}} xmm1 = xmm1[2,3,2,3]
; SSE42-NEXT: movdqa %xmm1, 112(%rdi)
; SSE42-NEXT: movdqa %xmm1, 96(%rdi)
; SSE42-NEXT: movdqa %xmm3, 80(%rdi)
; SSE42-NEXT: movdqa %xmm3, 64(%rdi)
; SSE42-NEXT: movdqa %xmm0, 48(%rdi)
; SSE42-NEXT: movdqa %xmm0, 32(%rdi)
; SSE42-NEXT: movdqa %xmm2, 16(%rdi)
; SSE42-NEXT: movdqa %xmm2, (%rdi)
; SSE42-NEXT: retq
;
; AVX1-LABEL: load_catcat:
; AVX1: # %bb.0:
; AVX1-NEXT: vbroadcastsd (%rdi), %ymm0
; AVX1-NEXT: vbroadcastsd 8(%rdi), %ymm1
; AVX1-NEXT: vbroadcastsd 16(%rdi), %ymm2
; AVX1-NEXT: vbroadcastsd 24(%rdi), %ymm3
; AVX1-NEXT: retq
;
; AVX2-LABEL: load_catcat:
; AVX2: # %bb.0:
; AVX2-NEXT: vmovaps (%rdi), %ymm3
; AVX2-NEXT: vbroadcastsd (%rdi), %ymm0
; AVX2-NEXT: vpermpd {{.*#+}} ymm1 = ymm3[1,1,1,1]
; AVX2-NEXT: vpermpd {{.*#+}} ymm2 = ymm3[2,2,2,2]
; AVX2-NEXT: vpermpd {{.*#+}} ymm3 = ymm3[3,3,3,3]
; AVX2-NEXT: retq
;
; AVX512F-LABEL: load_catcat:
; AVX512F: # %bb.0:
; AVX512F-NEXT: vbroadcasti64x4 {{.*#+}} zmm1 = mem[0,1,2,3,0,1,2,3]
; AVX512F-NEXT: vmovdqa64 {{.*#+}} zmm0 = [0,4,0,4,1,5,1,5]
; AVX512F-NEXT: vpermq %zmm1, %zmm0, %zmm0
; AVX512F-NEXT: vmovdqa64 {{.*#+}} zmm2 = [2,6,2,6,3,7,3,7]
; AVX512F-NEXT: vpermq %zmm1, %zmm2, %zmm1
; AVX512F-NEXT: retq
%x = load <4 x i64>, <4 x i64>* %p
%cat1 = shufflevector <4 x i64> %x, <4 x i64> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3>
%cat2 = shufflevector <8 x i64> %cat1, <8 x i64> undef, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
%r = shufflevector <16 x i64> %cat2, <16 x i64> undef, <16 x i32> <i32 0, i32 4, i32 8, i32 12, i32 1, i32 5, i32 9, i32 13, i32 2, i32 6, i32 10, i32 14, i32 3, i32 7, i32 11, i32 15>
ret <16 x i64> %r
}
; Use weird types to make sure we do not miscompile a case where
; the source ops are not an even multiple size of the result.
define <4 x i32> @cat_ext_straddle(<6 x i32>* %px, <6 x i32>* %py) {
; SSE42-LABEL: cat_ext_straddle:
; SSE42: # %bb.0:
; SSE42-NEXT: movaps 16(%rdi), %xmm0
; SSE42-NEXT: unpcklpd {{.*#+}} xmm0 = xmm0[0],mem[0]
; SSE42-NEXT: retq
;
; AVX-LABEL: cat_ext_straddle:
; AVX: # %bb.0:
; AVX-NEXT: vmovaps 16(%rdi), %xmm0
; AVX-NEXT: vunpcklpd {{.*#+}} xmm0 = xmm0[0],mem[0]
; AVX-NEXT: retq
%x = load <6 x i32>, <6 x i32>* %px
%y = load <6 x i32>, <6 x i32>* %py
%cat = shufflevector <6 x i32> %x, <6 x i32> %y, <12 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11>
%ext = shufflevector <12 x i32> %cat, <12 x i32> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
ret <4 x i32> %ext
}