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
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
|
; REQUIRES: asserts
; RUN: llc < %s -mtriple=armv8r-eabi -mcpu=cortex-a57 -enable-misched -verify-misched -debug-only=machine-scheduler -o - 2>&1 > /dev/null | FileCheck %s --check-prefix=CHECK --check-prefix=CHECK-DEFAULT
; RUN: llc < %s -mtriple=armv8r-eabi -mcpu=cortex-a57 -enable-misched -verify-misched -debug-only=machine-scheduler -o - 2>&1 > /dev/null -fp-contract=fast | FileCheck %s --check-prefix=CHECK --check-prefix=CHECK-FAST
; Check latencies of vmul/vfma accumulate chains.
define arm_aapcs_vfpcc float @Test1(float %f1, float %f2, float %f3, float %f4, float %f5, float %f6) {
; CHECK: ********** MI Scheduling **********
; CHECK: Test1:%bb.0
; CHECK: VMULS
; > VMULS common latency = 5
; CHECK: Latency : 5
; CHECK: Successors:
; CHECK: Data
; > VMULS read-advanced latency to VMLAS = 0
; CHECK-SAME: Latency=0
; CHECK-DEFAULT: VMLAS
; CHECK-FAST: VFMAS
; > VMLAS common latency = 9
; CHECK: Latency : 9
; CHECK: Successors:
; CHECK: Data
; > VMLAS read-advanced latency to the next VMLAS = 4
; CHECK-SAME: Latency=4
; CHECK-DEFAULT: VMLAS
; CHECK-FAST: VFMAS
; CHECK: Latency : 9
; CHECK: Successors:
; CHECK: Data
; > VMLAS not-optimized latency to VMOVRS = 9
; CHECK-SAME: Latency=9
; f1 * f2 + f3 * f4 + f5 * f6 ==> VMULS, VMLAS, VMLAS
%mul1 = fmul float %f1, %f2
%mul2 = fmul float %f3, %f4
%mul3 = fmul float %f5, %f6
%add1 = fadd float %mul1, %mul2
%add2 = fadd float %add1, %mul3
ret float %add2
}
; ASIMD form
define arm_aapcs_vfpcc <2 x float> @Test2(<2 x float> %f1, <2 x float> %f2, <2 x float> %f3, <2 x float> %f4, <2 x float> %f5, <2 x float> %f6) {
; CHECK: ********** MI Scheduling **********
; CHECK: Test2:%bb.0
; CHECK: VMULfd
; > VMULfd common latency = 5
; CHECK: Latency : 5
; CHECK: Successors:
; CHECK: Data
; VMULfd read-advanced latency to VMLAfd = 0
; CHECK-SAME: Latency=0
; CHECK-DEFAULT: VMLAfd
; CHECK-FAST: VFMAfd
; > VMLAfd common latency = 9
; CHECK: Latency : 9
; CHECK: Successors:
; CHECK: Data
; > VMLAfd read-advanced latency to the next VMLAfd = 4
; CHECK-SAME: Latency=4
; CHECK-DEFAULT: VMLAfd
; CHECK-FAST: VFMAfd
; CHECK: Latency : 9
; CHECK: Successors:
; CHECK: Data
; > VMLAfd not-optimized latency to VMOVRRD = 9
; CHECK-SAME: Latency=9
; f1 * f2 + f3 * f4 + f5 * f6 ==> VMULS, VMLAS, VMLAS
%mul1 = fmul <2 x float> %f1, %f2
%mul2 = fmul <2 x float> %f3, %f4
%mul3 = fmul <2 x float> %f5, %f6
%add1 = fadd <2 x float> %mul1, %mul2
%add2 = fadd <2 x float> %add1, %mul3
ret <2 x float> %add2
}
define arm_aapcs_vfpcc float @Test3(float %f1, float %f2, float %f3, float %f4, float %f5, float %f6) {
; CHECK: ********** MI Scheduling **********
; CHECK: Test3:%bb.0
; CHECK: VMULS
; > VMULS common latency = 5
; CHECK: Latency : 5
; CHECK: Successors:
; CHECK: Data
; > VMULS read-advanced latency to VMLSS = 0
; CHECK-SAME: Latency=0
; CHECK-DEFAULT: VMLSS
; CHECK-FAST: VFNMSS
; > VFNMSS common latency = 9
; CHECK: Latency : 9
; CHECK: Successors:
; CHECK: Data
; > VFNMSS read-advanced latency to the next VMLSS = 4
; CHECK-SAME: Latency=4
; CHECK-DEFAULT: VMLSS
; CHECK-FAST: VFMSS
; CHECK: Latency : 9
; CHECK: Successors:
; CHECK: Data
; > VMLSS not-optimized latency to VMOVRS = 9
; CHECK-SAME: Latency=9
; f1 * f2 + f3 * f4 + f5 * f6 ==> VMULS, VMLSS, VMLSS
%mul1 = fmul float %f1, %f2
%mul2 = fmul float %f3, %f4
%mul3 = fmul float %f5, %f6
%sub1 = fsub float %mul1, %mul2
%sub2 = fsub float %sub1, %mul3
ret float %sub2
}
; ASIMD form
define arm_aapcs_vfpcc <2 x float> @Test4(<2 x float> %f1, <2 x float> %f2, <2 x float> %f3, <2 x float> %f4, <2 x float> %f5, <2 x float> %f6) {
; CHECK: ********** MI Scheduling **********
; CHECK: Test4:%bb.0
; CHECK: VMULfd
; > VMULfd common latency = 5
; CHECK: Latency : 5
; CHECK: Successors:
; CHECK: Data
; VMULfd read-advanced latency to VMLSfd = 0
; CHECK-SAME: Latency=0
; CHECK-DEFAULT: VMLSfd
; CHECK-FAST: VFMSfd
; > VMLSfd common latency = 9
; CHECK: Latency : 9
; CHECK: Successors:
; CHECK: Data
; > VMLSfd read-advanced latency to the next VMLSfd = 4
; CHECK-SAME: Latency=4
; CHECK-DEFAULT: VMLSfd
; CHECK-FAST: VFMSfd
; CHECK: Latency : 9
; CHECK: Successors:
; CHECK: Data
; > VMLSfd not-optimized latency to VMOVRRD = 9
; CHECK-SAME: Latency=9
; f1 * f2 + f3 * f4 + f5 * f6 ==> VMULS, VMLSS, VMLSS
%mul1 = fmul <2 x float> %f1, %f2
%mul2 = fmul <2 x float> %f3, %f4
%mul3 = fmul <2 x float> %f5, %f6
%sub1 = fsub <2 x float> %mul1, %mul2
%sub2 = fsub <2 x float> %sub1, %mul3
ret <2 x float> %sub2
}
define arm_aapcs_vfpcc float @Test5(float %f1, float %f2, float %f3) {
; CHECK: ********** MI Scheduling **********
; CHECK: Test5:%bb.0
; CHECK-DEFAULT: VNMLS
; CHECK-FAST: VFNMS
; CHECK: Latency : 9
; CHECK: Successors:
; CHECK: Data
; > VMLAS not-optimized latency to VMOVRS = 9
; CHECK-SAME: Latency=9
; f1 * f2 - f3 ==> VNMLS/VFNMS
%mul = fmul float %f1, %f2
%sub = fsub float %mul, %f3
ret float %sub
}
define arm_aapcs_vfpcc float @Test6(float %f1, float %f2, float %f3) {
; CHECK: ********** MI Scheduling **********
; CHECK: Test6:%bb.0
; CHECK-DEFAULT: VNMLA
; CHECK-FAST: VFNMA
; CHECK: Latency : 9
; CHECK: Successors:
; CHECK: Data
; > VMLAS not-optimized latency to VMOVRS = 9
; CHECK-SAME: Latency=9
; f1 * f2 - f3 ==> VNMLA/VFNMA
%mul = fmul float %f1, %f2
%sub1 = fsub float -0.0, %mul
%sub2 = fsub float %sub1, %f2
ret float %sub2
}
|
