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
|
; RUN: llc -mtriple=arm-apple-darwin %s -o - | FileCheck %s
; RUN: llc -mtriple=arm-eabi -mattr=+vfp2 %s -o - \
; RUN: | FileCheck %s --check-prefix=CHECK-VFP
; RUN: llc -mtriple=thumbv7-apple-darwin -mattr=+neon,+thumb2 %s -o - \
; RUN: | FileCheck %s --check-prefix=CHECK-NEON
define i32 @f1(i32 %a.s) {
;CHECK-LABEL: f1:
;CHECK: moveq
entry:
%tmp = icmp eq i32 %a.s, 4
%tmp1.s = select i1 %tmp, i32 2, i32 3
ret i32 %tmp1.s
}
define i32 @f2(i32 %a.s) {
;CHECK-LABEL: f2:
;CHECK: movgt
entry:
%tmp = icmp sgt i32 %a.s, 4
%tmp1.s = select i1 %tmp, i32 2, i32 3
ret i32 %tmp1.s
}
define i32 @f3(i32 %a.s, i32 %b.s) {
;CHECK-LABEL: f3:
;CHECK: movlt
entry:
%tmp = icmp slt i32 %a.s, %b.s
%tmp1.s = select i1 %tmp, i32 2, i32 3
ret i32 %tmp1.s
}
define i32 @f4(i32 %a.s, i32 %b.s) {
;CHECK-LABEL: f4:
;CHECK: movle
entry:
%tmp = icmp sle i32 %a.s, %b.s
%tmp1.s = select i1 %tmp, i32 2, i32 3
ret i32 %tmp1.s
}
define i32 @f5(i32 %a.u, i32 %b.u) {
;CHECK-LABEL: f5:
;CHECK: movls
entry:
%tmp = icmp ule i32 %a.u, %b.u
%tmp1.s = select i1 %tmp, i32 2, i32 3
ret i32 %tmp1.s
}
define i32 @f6(i32 %a.u, i32 %b.u) {
;CHECK-LABEL: f6:
;CHECK: movhi
entry:
%tmp = icmp ugt i32 %a.u, %b.u
%tmp1.s = select i1 %tmp, i32 2, i32 3
ret i32 %tmp1.s
}
define double @f7(double %a, double %b) {
;CHECK-LABEL: f7:
;CHECK: movmi
;CHECK: movpl
;CHECK-VFP-LABEL: f7:
;CHECK-VFP: vmovmi
%tmp = fcmp olt double %a, 1.234e+00
%tmp1 = select i1 %tmp, double -1.000e+00, double %b
ret double %tmp1
}
; <rdar://problem/7260094>
;
; We used to generate really horrible code for this function. The main cause was
; a lack of a custom lowering routine for an ISD::SELECT. This would result in
; two "it" blocks in the code: one for the "icmp" and another to move the index
; into the constant pool based on the value of the "icmp". If we have one "it"
; block generated, odds are good that we have close to the ideal code for this:
;
; CHECK-NEON-LABEL: f8:
; CHECK-NEON: adr [[R2:r[0-9]+]], LCPI7_0
; CHECK-NEON: movw [[R3:r[0-9]+]], #1123
; CHECK-NEON-NEXT: cmp r0, [[R3]]
; CHECK-NEON-NEXT: it eq
; CHECK-NEON-NEXT: addeq{{.*}} [[R2]], #4
; CHECK-NEON-NEXT: ldr
; CHECK-NEON: bx
define arm_apcscc float @f8(i32 %a) nounwind {
%tmp = icmp eq i32 %a, 1123
%tmp1 = select i1 %tmp, float 0x3FF3BE76C0000000, float 0x40030E9A20000000
ret float %tmp1
}
; <rdar://problem/9049552>
; Glue values can only have a single use, but the following test exposed a
; case where a SELECT was lowered with 2 uses of a comparison, causing the
; scheduler to assert.
; CHECK-VFP-LABEL: f9:
declare ptr @objc_msgSend(ptr, ptr, ...)
define void @f9() optsize {
entry:
%cmp = icmp eq ptr undef, inttoptr (i32 4 to ptr)
%conv191 = select i1 %cmp, float -3.000000e+00, float 0.000000e+00
%conv195 = select i1 %cmp, double -1.000000e+00, double 0.000000e+00
%add = fadd double %conv195, 1.100000e+01
%conv196 = fptrunc double %add to float
%add201 = fadd float undef, %conv191
%tmp484 = bitcast float %conv196 to i32
%tmp478 = bitcast float %add201 to i32
%tmp490 = insertvalue [2 x i32] undef, i32 %tmp484, 0
%tmp493 = insertvalue [2 x i32] %tmp490, i32 %tmp478, 1
call void @objc_msgSend(ptr undef, ptr undef, [2 x i32] %tmp493, i32 0, float 1.000000e+00) optsize
ret void
}
; CHECK-LABEL: f10:
define float @f10(i32 %a, i32 %b) nounwind uwtable readnone ssp {
; CHECK-NOT: floatsisf
%1 = icmp eq i32 %a, %b
%2 = zext i1 %1 to i32
%3 = sitofp i32 %2 to float
ret float %3
}
; CHECK-LABEL: f11:
define float @f11(i32 %a, i32 %b) nounwind uwtable readnone ssp {
; CHECK-NOT: floatsisf
%1 = icmp eq i32 %a, %b
%2 = sitofp i1 %1 to float
ret float %2
}
; CHECK-LABEL: f12:
define float @f12(i32 %a, i32 %b) nounwind uwtable readnone ssp {
; CHECK-NOT: floatunsisf
%1 = icmp eq i32 %a, %b
%2 = uitofp i1 %1 to float
ret float %2
}
; CHECK-LABEL: test_overflow_recombine:
define i1 @test_overflow_recombine(i32 %in1, i32 %in2) {
; CHECK: smull [[LO:r[0-9]+]], [[HI:r[0-9]+]]
; CHECK: subs [[ZERO:r[0-9]+]], [[HI]], [[LO]], asr #31
; CHECK: movne [[ZERO]], #1
%prod = call { i32, i1 } @llvm.smul.with.overflow.i32(i32 %in1, i32 %in2)
%overflow = extractvalue { i32, i1 } %prod, 1
ret i1 %overflow
}
declare { i32, i1 } @llvm.smul.with.overflow.i32(i32, i32)
|
