1 # NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
2 # RUN: llc -mtriple=thumbv8.1m.main-none-eabi -mcpu=cortex-m85 -mattr=+use-mipipeliner -run-pass=pipeliner --pipeliner-force-issue-width=10 -o - %s | FileCheck %s --check-prefix=CHECK
5 define hidden float @dot(ptr nocapture noundef readonly %a, ptr nocapture noundef readonly %b, i32 noundef %sz) local_unnamed_addr #0 {
7 %cmp8 = icmp sgt i32 %sz, 0
8 br i1 %cmp8, label %for.body.preheader, label %for.end
10 for.body.preheader: ; preds = %entry
11 %scevgep = getelementptr float, ptr %b, i32 -1
12 %scevgep4 = getelementptr float, ptr %a, i32 -1
15 for.body: ; preds = %for.body.preheader, %for.body
16 %lsr.iv5 = phi ptr [ %scevgep4, %for.body.preheader ], [ %scevgep6, %for.body ]
17 %lsr.iv1 = phi ptr [ %scevgep, %for.body.preheader ], [ %scevgep2, %for.body ]
18 %lsr.iv = phi i32 [ %sz, %for.body.preheader ], [ %lsr.iv.next, %for.body ]
19 %sum.010 = phi float [ %add, %for.body ], [ 0.000000e+00, %for.body.preheader ]
20 %scevgep7 = getelementptr float, ptr %lsr.iv5, i32 1
21 %0 = load float, ptr %scevgep7, align 4
22 %scevgep3 = getelementptr float, ptr %lsr.iv1, i32 1
23 %1 = load float, ptr %scevgep3, align 4
24 %mul = fmul fast float %1, %0
25 %add = fadd fast float %mul, %sum.010
26 %lsr.iv.next = add i32 %lsr.iv, -1
27 %scevgep2 = getelementptr float, ptr %lsr.iv1, i32 1
28 %scevgep6 = getelementptr float, ptr %lsr.iv5, i32 1
29 %exitcond.not = icmp eq i32 %lsr.iv.next, 0
30 br i1 %exitcond.not, label %for.end, label %for.body, !llvm.loop !0
32 for.end: ; preds = %for.body, %entry
33 %sum.0.lcssa = phi float [ 0.000000e+00, %entry ], [ %add, %for.body ]
34 ret float %sum.0.lcssa
37 !0 = distinct !{!0, !1, !2, !3}
38 !1 = !{!"llvm.loop.mustprogress"}
39 !2 = !{!"llvm.loop.unroll.disable"}
40 !3 = !{!"llvm.loop.pipeline.initiationinterval", i32 3}
46 tracksRegLiveness: true
49 value: 'float 0.000000e+00'
51 isTargetSpecific: false
53 ; CHECK-LABEL: name: dot
55 ; CHECK-NEXT: successors: %bb.2(0x50000000), %bb.1(0x30000000)
56 ; CHECK-NEXT: liveins: $r0, $r1, $r2
58 ; CHECK-NEXT: [[COPY:%[0-9]+]]:gprnopc = COPY $r2
59 ; CHECK-NEXT: [[COPY1:%[0-9]+]]:gprnopc = COPY $r1
60 ; CHECK-NEXT: [[COPY2:%[0-9]+]]:gprnopc = COPY $r0
61 ; CHECK-NEXT: t2CMPri [[COPY]], 1, 14 /* CC::al */, $noreg, implicit-def $cpsr
62 ; CHECK-NEXT: t2Bcc %bb.2, 10 /* CC::ge */, $cpsr
65 ; CHECK-NEXT: successors: %bb.4(0x80000000)
67 ; CHECK-NEXT: [[VLDRS:%[0-9]+]]:spr = VLDRS %const.0, 0, 14 /* CC::al */, $noreg :: (load (s32) from constant-pool)
68 ; CHECK-NEXT: t2B %bb.4, 14 /* CC::al */, $noreg
70 ; CHECK-NEXT: bb.2.for.body.preheader:
71 ; CHECK-NEXT: successors: %bb.5(0x80000000)
73 ; CHECK-NEXT: [[t2SUBri:%[0-9]+]]:rgpr = t2SUBri [[COPY1]], 4, 14 /* CC::al */, $noreg, $noreg
74 ; CHECK-NEXT: [[COPY3:%[0-9]+]]:gprnopc = COPY [[t2SUBri]]
75 ; CHECK-NEXT: [[t2SUBri1:%[0-9]+]]:rgpr = t2SUBri [[COPY2]], 4, 14 /* CC::al */, $noreg, $noreg
76 ; CHECK-NEXT: [[VLDRS1:%[0-9]+]]:spr = VLDRS %const.0, 0, 14 /* CC::al */, $noreg :: (load (s32) from constant-pool)
77 ; CHECK-NEXT: [[COPY4:%[0-9]+]]:gprnopc = COPY [[t2SUBri1]]
79 ; CHECK-NEXT: bb.5.for.body:
80 ; CHECK-NEXT: successors: %bb.6(0x40000000), %bb.9(0x40000000)
82 ; CHECK-NEXT: [[t2ADDri:%[0-9]+]]:rgpr = t2ADDri [[COPY4]], 4, 14 /* CC::al */, $noreg, $noreg
83 ; CHECK-NEXT: [[VLDRS2:%[0-9]+]]:spr = VLDRS [[COPY4]], 1, 14 /* CC::al */, $noreg :: (load (s32) from %ir.scevgep7)
84 ; CHECK-NEXT: [[t2ADDri1:%[0-9]+]]:rgpr = t2ADDri [[COPY3]], 4, 14 /* CC::al */, $noreg, $noreg
85 ; CHECK-NEXT: [[VLDRS3:%[0-9]+]]:spr = VLDRS [[COPY3]], 1, 14 /* CC::al */, $noreg :: (load (s32) from %ir.scevgep3)
86 ; CHECK-NEXT: [[VMULS:%[0-9]+]]:spr = nnan ninf nsz arcp contract afn reassoc VMULS [[VLDRS3]], [[VLDRS2]], 14 /* CC::al */, $noreg
87 ; CHECK-NEXT: [[t2SUBri2:%[0-9]+]]:rgpr = t2SUBri [[COPY]], 1, 14 /* CC::al */, $noreg, def $cpsr
88 ; CHECK-NEXT: [[COPY5:%[0-9]+]]:gprnopc = COPY [[t2SUBri2]]
89 ; CHECK-NEXT: [[COPY6:%[0-9]+]]:gprnopc = COPY [[t2ADDri1]]
90 ; CHECK-NEXT: [[COPY7:%[0-9]+]]:gprnopc = COPY [[t2ADDri]]
91 ; CHECK-NEXT: t2Bcc %bb.9, 0 /* CC::eq */, $cpsr
92 ; CHECK-NEXT: t2B %bb.6, 14 /* CC::al */, $noreg
94 ; CHECK-NEXT: bb.6.for.body:
95 ; CHECK-NEXT: successors: %bb.7(0x80000000), %bb.8(0x00000000)
97 ; CHECK-NEXT: [[t2ADDri2:%[0-9]+]]:rgpr = t2ADDri [[COPY7]], 4, 14 /* CC::al */, $noreg, $noreg
98 ; CHECK-NEXT: [[VLDRS4:%[0-9]+]]:spr = VLDRS [[COPY7]], 1, 14 /* CC::al */, $noreg :: (load unknown-size from %ir.scevgep7, align 4)
99 ; CHECK-NEXT: [[t2ADDri3:%[0-9]+]]:rgpr = t2ADDri [[COPY6]], 4, 14 /* CC::al */, $noreg, $noreg
100 ; CHECK-NEXT: [[VLDRS5:%[0-9]+]]:spr = VLDRS [[COPY6]], 1, 14 /* CC::al */, $noreg :: (load unknown-size from %ir.scevgep3, align 4)
101 ; CHECK-NEXT: [[VMULS1:%[0-9]+]]:spr = nnan ninf nsz arcp contract afn reassoc VMULS [[VLDRS5]], [[VLDRS4]], 14 /* CC::al */, $noreg
102 ; CHECK-NEXT: [[t2SUBri3:%[0-9]+]]:rgpr = t2SUBri [[COPY5]], 1, 14 /* CC::al */, $noreg, def $cpsr
103 ; CHECK-NEXT: [[COPY8:%[0-9]+]]:gpr = COPY [[t2SUBri3]]
104 ; CHECK-NEXT: [[COPY9:%[0-9]+]]:gpr = COPY [[t2ADDri3]]
105 ; CHECK-NEXT: [[COPY10:%[0-9]+]]:gpr = COPY [[t2ADDri2]]
106 ; CHECK-NEXT: t2Bcc %bb.8, 0 /* CC::eq */, $cpsr
107 ; CHECK-NEXT: t2B %bb.7, 14 /* CC::al */, $noreg
109 ; CHECK-NEXT: bb.7.for.body:
110 ; CHECK-NEXT: successors: %bb.8(0x04000000), %bb.7(0x7c000000)
112 ; CHECK-NEXT: [[PHI:%[0-9]+]]:gprnopc = PHI [[COPY10]], %bb.6, %49, %bb.7
113 ; CHECK-NEXT: [[PHI1:%[0-9]+]]:gprnopc = PHI [[COPY9]], %bb.6, %50, %bb.7
114 ; CHECK-NEXT: [[PHI2:%[0-9]+]]:gprnopc = PHI [[COPY8]], %bb.6, %51, %bb.7
115 ; CHECK-NEXT: [[PHI3:%[0-9]+]]:spr = PHI [[VLDRS1]], %bb.6, %43, %bb.7
116 ; CHECK-NEXT: [[PHI4:%[0-9]+]]:spr = PHI [[VMULS1]], %bb.6, %52, %bb.7
117 ; CHECK-NEXT: [[PHI5:%[0-9]+]]:spr = PHI [[VMULS]], %bb.6, [[PHI4]], %bb.7
118 ; CHECK-NEXT: [[VADDS:%[0-9]+]]:spr = nnan ninf nsz arcp contract afn reassoc VADDS [[PHI5]], [[PHI3]], 14 /* CC::al */, $noreg
119 ; CHECK-NEXT: [[t2SUBri4:%[0-9]+]]:rgpr = t2SUBri [[PHI2]], 1, 14 /* CC::al */, $noreg, def $cpsr
120 ; CHECK-NEXT: [[VLDRS6:%[0-9]+]]:spr = VLDRS [[PHI1]], 1, 14 /* CC::al */, $noreg :: (load unknown-size from %ir.scevgep3, align 4)
121 ; CHECK-NEXT: [[VLDRS7:%[0-9]+]]:spr = VLDRS [[PHI]], 1, 14 /* CC::al */, $noreg :: (load unknown-size from %ir.scevgep7, align 4)
122 ; CHECK-NEXT: [[t2ADDri4:%[0-9]+]]:rgpr = t2ADDri [[PHI]], 4, 14 /* CC::al */, $noreg, $noreg
123 ; CHECK-NEXT: [[t2ADDri5:%[0-9]+]]:rgpr = t2ADDri [[PHI1]], 4, 14 /* CC::al */, $noreg, $noreg
124 ; CHECK-NEXT: [[COPY11:%[0-9]+]]:gpr = COPY [[t2ADDri4]]
125 ; CHECK-NEXT: [[COPY12:%[0-9]+]]:gpr = COPY [[t2ADDri5]]
126 ; CHECK-NEXT: [[COPY13:%[0-9]+]]:gpr = COPY [[t2SUBri4]]
127 ; CHECK-NEXT: [[VMULS2:%[0-9]+]]:spr = nnan ninf nsz arcp contract afn reassoc VMULS [[VLDRS6]], [[VLDRS7]], 14 /* CC::al */, $noreg
128 ; CHECK-NEXT: t2Bcc %bb.7, 1 /* CC::ne */, $cpsr
129 ; CHECK-NEXT: t2B %bb.8, 14 /* CC::al */, $noreg
132 ; CHECK-NEXT: successors: %bb.9(0x80000000)
134 ; CHECK-NEXT: [[PHI6:%[0-9]+]]:spr = PHI [[VLDRS1]], %bb.6, [[VADDS]], %bb.7
135 ; CHECK-NEXT: [[PHI7:%[0-9]+]]:spr = PHI [[VMULS1]], %bb.6, [[VMULS2]], %bb.7
136 ; CHECK-NEXT: [[PHI8:%[0-9]+]]:spr = PHI [[VMULS]], %bb.6, [[PHI4]], %bb.7
137 ; CHECK-NEXT: [[VADDS1:%[0-9]+]]:spr = nnan ninf nsz arcp contract afn reassoc VADDS [[PHI8]], [[PHI6]], 14 /* CC::al */, $noreg
140 ; CHECK-NEXT: successors: %bb.4(0x80000000)
142 ; CHECK-NEXT: [[PHI9:%[0-9]+]]:spr = PHI [[VLDRS1]], %bb.5, [[VADDS1]], %bb.8
143 ; CHECK-NEXT: [[PHI10:%[0-9]+]]:spr = PHI [[VMULS]], %bb.5, [[PHI7]], %bb.8
144 ; CHECK-NEXT: [[VADDS2:%[0-9]+]]:spr = nnan ninf nsz arcp contract afn reassoc VADDS [[PHI10]], [[PHI9]], 14 /* CC::al */, $noreg
145 ; CHECK-NEXT: t2B %bb.4, 14 /* CC::al */, $noreg
147 ; CHECK-NEXT: bb.4.for.end:
148 ; CHECK-NEXT: [[PHI11:%[0-9]+]]:spr = PHI [[VLDRS]], %bb.1, [[VADDS2]], %bb.9
149 ; CHECK-NEXT: [[VMOVRS:%[0-9]+]]:gpr = VMOVRS [[PHI11]], 14 /* CC::al */, $noreg
150 ; CHECK-NEXT: $r0 = COPY [[VMOVRS]]
151 ; CHECK-NEXT: tBX_RET 14 /* CC::al */, $noreg, implicit $r0
153 successors: %bb.1(0x50000000), %bb.4(0x30000000)
154 liveins: $r0, $r1, $r2
156 %13:gprnopc = COPY $r2
157 %12:gprnopc = COPY $r1
158 %11:gprnopc = COPY $r0
159 t2CMPri %13, 1, 14 /* CC::al */, $noreg, implicit-def $cpsr
160 t2Bcc %bb.1, 10 /* CC::ge */, $cpsr
163 successors: %bb.3(0x80000000)
165 %14:spr = VLDRS %const.0, 0, 14 /* CC::al */, $noreg :: (load (s32) from constant-pool)
166 t2B %bb.3, 14 /* CC::al */, $noreg
168 bb.1.for.body.preheader:
169 successors: %bb.2(0x80000000)
171 %16:rgpr = t2SUBri %12, 4, 14 /* CC::al */, $noreg, $noreg
173 %17:rgpr = t2SUBri %11, 4, 14 /* CC::al */, $noreg, $noreg
174 %15:spr = VLDRS %const.0, 0, 14 /* CC::al */, $noreg :: (load (s32) from constant-pool)
178 successors: %bb.3(0x04000000), %bb.2(0x7c000000)
180 %2:gprnopc = PHI %1, %bb.1, %9, %bb.2
181 %3:gprnopc = PHI %0, %bb.1, %8, %bb.2
182 %4:gprnopc = PHI %13, %bb.1, %7, %bb.2
183 %5:spr = PHI %15, %bb.1, %6, %bb.2
184 %18:rgpr = t2ADDri %2, 4, 14 /* CC::al */, $noreg, $noreg
185 %19:spr = VLDRS %2, 1, 14 /* CC::al */, $noreg :: (load (s32) from %ir.scevgep7)
186 %20:rgpr = t2ADDri %3, 4, 14 /* CC::al */, $noreg, $noreg
187 %21:spr = VLDRS %3, 1, 14 /* CC::al */, $noreg :: (load (s32) from %ir.scevgep3)
188 %22:spr = nnan ninf nsz arcp contract afn reassoc VMULS killed %21, killed %19, 14 /* CC::al */, $noreg
189 %6:spr = nnan ninf nsz arcp contract afn reassoc VADDS killed %22, %5, 14 /* CC::al */, $noreg
190 %23:rgpr = t2SUBri %4, 1, 14 /* CC::al */, $noreg, def $cpsr
194 t2Bcc %bb.2, 1 /* CC::ne */, $cpsr
195 t2B %bb.3, 14 /* CC::al */, $noreg
198 %10:spr = PHI %14, %bb.4, %6, %bb.2
199 %24:gpr = VMOVRS %10, 14 /* CC::al */, $noreg
201 tBX_RET 14 /* CC::al */, $noreg, implicit $r0