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
|
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=i686-unknown -mattr=+mmx,+sse2 | FileCheck %s --check-prefix=X86
; RUN: llc < %s -mtriple=x86_64-unknown -mattr=+mmx,+sse2 | FileCheck %s --check-prefix=X64
define i32 @t0(i64 %x) nounwind {
; X86-LABEL: t0:
; X86: # %bb.0: # %entry
; X86-NEXT: pshufw $238, {{[0-9]+}}(%esp), %mm0 # mm0 = mem[2,3,2,3]
; X86-NEXT: movd %mm0, %eax
; X86-NEXT: retl
;
; X64-LABEL: t0:
; X64: # %bb.0: # %entry
; X64-NEXT: movq %rdi, %mm0
; X64-NEXT: pshufw $238, %mm0, %mm0 # mm0 = mm0[2,3,2,3]
; X64-NEXT: movd %mm0, %eax
; X64-NEXT: retq
entry:
%0 = bitcast i64 %x to <4 x i16>
%1 = bitcast <4 x i16> %0 to x86_mmx
%2 = tail call x86_mmx @llvm.x86.sse.pshuf.w(x86_mmx %1, i8 -18)
%3 = bitcast x86_mmx %2 to <4 x i16>
%4 = bitcast <4 x i16> %3 to <1 x i64>
%5 = extractelement <1 x i64> %4, i32 0
%6 = bitcast i64 %5 to <2 x i32>
%7 = extractelement <2 x i32> %6, i32 0
ret i32 %7
}
define i64 @t1(i64 %x, i32 %n) nounwind {
; X86-LABEL: t1:
; X86: # %bb.0: # %entry
; X86-NEXT: pushl %ebp
; X86-NEXT: movl %esp, %ebp
; X86-NEXT: andl $-8, %esp
; X86-NEXT: subl $8, %esp
; X86-NEXT: movq 8(%ebp), %mm0
; X86-NEXT: movd 16(%ebp), %mm1
; X86-NEXT: psllq %mm1, %mm0
; X86-NEXT: movq %mm0, (%esp)
; X86-NEXT: movl (%esp), %eax
; X86-NEXT: movl {{[0-9]+}}(%esp), %edx
; X86-NEXT: movl %ebp, %esp
; X86-NEXT: popl %ebp
; X86-NEXT: retl
;
; X64-LABEL: t1:
; X64: # %bb.0: # %entry
; X64-NEXT: movq %rdi, %mm0
; X64-NEXT: movd %esi, %mm1
; X64-NEXT: psllq %mm1, %mm0
; X64-NEXT: movq %mm0, %rax
; X64-NEXT: retq
entry:
%0 = bitcast i64 %x to x86_mmx
%1 = tail call x86_mmx @llvm.x86.mmx.pslli.q(x86_mmx %0, i32 %n)
%2 = bitcast x86_mmx %1 to i64
ret i64 %2
}
define i64 @t2(i64 %x, i32 %n, i32 %w) nounwind {
; X86-LABEL: t2:
; X86: # %bb.0: # %entry
; X86-NEXT: pushl %ebp
; X86-NEXT: movl %esp, %ebp
; X86-NEXT: andl $-8, %esp
; X86-NEXT: subl $8, %esp
; X86-NEXT: movd 20(%ebp), %mm0
; X86-NEXT: movd 16(%ebp), %mm1
; X86-NEXT: psllq %mm1, %mm0
; X86-NEXT: por 8(%ebp), %mm0
; X86-NEXT: movq %mm0, (%esp)
; X86-NEXT: movl (%esp), %eax
; X86-NEXT: movl {{[0-9]+}}(%esp), %edx
; X86-NEXT: movl %ebp, %esp
; X86-NEXT: popl %ebp
; X86-NEXT: retl
;
; X64-LABEL: t2:
; X64: # %bb.0: # %entry
; X64-NEXT: movd %edx, %mm0
; X64-NEXT: movd %esi, %mm1
; X64-NEXT: psllq %mm1, %mm0
; X64-NEXT: movq %rdi, %mm1
; X64-NEXT: por %mm0, %mm1
; X64-NEXT: movq %mm1, %rax
; X64-NEXT: retq
entry:
%0 = insertelement <2 x i32> undef, i32 %w, i32 0
%1 = insertelement <2 x i32> %0, i32 0, i32 1
%2 = bitcast <2 x i32> %1 to x86_mmx
%3 = tail call x86_mmx @llvm.x86.mmx.pslli.q(x86_mmx %2, i32 %n)
%4 = bitcast i64 %x to x86_mmx
%5 = tail call x86_mmx @llvm.x86.mmx.por(x86_mmx %4, x86_mmx %3)
%6 = bitcast x86_mmx %5 to i64
ret i64 %6
}
define i64 @t3(<1 x i64>* %y, i32* %n) nounwind {
; X86-LABEL: t3:
; X86: # %bb.0: # %entry
; X86-NEXT: pushl %ebp
; X86-NEXT: movl %esp, %ebp
; X86-NEXT: andl $-8, %esp
; X86-NEXT: subl $8, %esp
; X86-NEXT: movl 12(%ebp), %eax
; X86-NEXT: movl 8(%ebp), %ecx
; X86-NEXT: movq (%ecx), %mm0
; X86-NEXT: movd (%eax), %mm1
; X86-NEXT: psllq %mm1, %mm0
; X86-NEXT: movq %mm0, (%esp)
; X86-NEXT: movl (%esp), %eax
; X86-NEXT: movl {{[0-9]+}}(%esp), %edx
; X86-NEXT: movl %ebp, %esp
; X86-NEXT: popl %ebp
; X86-NEXT: retl
;
; X64-LABEL: t3:
; X64: # %bb.0: # %entry
; X64-NEXT: movq (%rdi), %mm0
; X64-NEXT: movd (%rsi), %mm1
; X64-NEXT: psllq %mm1, %mm0
; X64-NEXT: movq %mm0, %rax
; X64-NEXT: retq
entry:
%0 = bitcast <1 x i64>* %y to x86_mmx*
%1 = load x86_mmx, x86_mmx* %0, align 8
%2 = load i32, i32* %n, align 4
%3 = tail call x86_mmx @llvm.x86.mmx.pslli.q(x86_mmx %1, i32 %2)
%4 = bitcast x86_mmx %3 to i64
ret i64 %4
}
declare x86_mmx @llvm.x86.sse.pshuf.w(x86_mmx, i8)
declare x86_mmx @llvm.x86.mmx.pslli.q(x86_mmx, i32)
declare x86_mmx @llvm.x86.mmx.por(x86_mmx, x86_mmx)
|
