1# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
2# RUN: llc -mtriple=thumbv8.1m.main -mattr=+mve -run-pass=arm-low-overhead-loops %s -o - --verify-machineinstrs | FileCheck %s
3
4# A decent sized test to handle a matrix, with scalar and vector low-overhead loops.
5
6--- |
7  define dso_local arm_aapcs_vfpcc signext i16 @matrix_test(i32 %d, i32* nocapture %e, i16* nocapture readonly %k, i16* nocapture readonly %l) {
8  entry:
9    %cmp19.i = icmp sgt i32 %d, 0
10    br i1 %cmp19.i, label %for.body.i.preheader, label %c.exit.thread
11
12  for.body.i.preheader:                             ; preds = %entry
13    %start1 = call i32 @llvm.start.loop.iterations.i32(i32 %d)
14    br label %for.body.i
15
16  c.exit.thread:                                    ; preds = %entry
17    %call169 = tail call arm_aapcs_vfpcc signext i16 bitcast (i16 (...)* @crc16 to i16 (i32)*)(i32 0)
18    %conv270 = sext i16 %call169 to i32
19    br label %c.exit59
20
21  for.body.i:                                       ; preds = %for.body.i, %for.body.i.preheader
22    %lsr.iv15 = phi i32* [ %e, %for.body.i.preheader ], [ %scevgep16, %for.body.i ]
23    %h.022.i = phi i16 [ %h.1.i, %for.body.i ], [ 0, %for.body.i.preheader ]
24    %f.020.i = phi i32 [ %f.1.i, %for.body.i ], [ undef, %for.body.i.preheader ]
25    %0 = phi i32 [ %start1, %for.body.i.preheader ], [ %2, %for.body.i ]
26    %1 = load i32, i32* %lsr.iv15, align 4
27    %add.i = add nsw i32 %1, %f.020.i
28    %cmp1.i = icmp sgt i32 %add.i, 0
29    %cmp3.i = icmp sgt i32 %1, 0
30    %f.1.i = select i1 %cmp1.i, i32 0, i32 %add.i
31    %narrow.i = and i1 %cmp3.i, %cmp1.i
32    %add6.i = zext i1 %narrow.i to i16
33    %h.1.i = add i16 %h.022.i, %add6.i
34    %scevgep16 = getelementptr i32, i32* %lsr.iv15, i32 1
35    %2 = call i32 @llvm.loop.decrement.reg.i32.i32.i32(i32 %0, i32 1)
36    %3 = icmp ne i32 %2, 0
37    br i1 %3, label %for.body.i, label %c.exit
38
39  c.exit:                                           ; preds = %for.body.i
40    %4 = icmp sgt i32 %d, 0
41    %phitmp = sext i16 %h.1.i to i32
42    %call1 = tail call arm_aapcs_vfpcc signext i16 bitcast (i16 (...)* @crc16 to i16 (i32)*)(i32 %phitmp)
43    %conv2 = sext i16 %call1 to i32
44    br i1 %4, label %for.cond4.preheader.us.preheader, label %c.exit59
45
46  for.cond4.preheader.us.preheader:                 ; preds = %c.exit
47    %n.rnd.up = add i32 %d, 3
48    %n.vec = and i32 %n.rnd.up, -4
49    %5 = shl i32 %d, 1
50    %6 = add i32 %n.vec, -4
51    %7 = lshr i32 %6, 2
52    %8 = add nuw nsw i32 %7, 1
53    %9 = shl i32 %7, 2
54    %10 = sub i32 %d, %9
55    br label %for.cond4.preheader.us
56
57  for.cond4.preheader.us:                           ; preds = %middle.block, %for.cond4.preheader.us.preheader
58    %lsr.iv7 = phi i16* [ %28, %middle.block ], [ %k, %for.cond4.preheader.us.preheader ]
59    %i.064.us = phi i32 [ %inc15.us, %middle.block ], [ 0, %for.cond4.preheader.us.preheader ]
60    %arrayidx12.us = getelementptr inbounds i32, i32* %e, i32 %i.064.us
61    %arrayidx12.promoted.us = load i32, i32* %arrayidx12.us, align 4
62    %11 = insertelement <4 x i32> <i32 undef, i32 0, i32 0, i32 0>, i32 %arrayidx12.promoted.us, i32 0
63    %start2 = call i32 @llvm.start.loop.iterations.i32(i32 %8)
64    br label %vector.body
65
66  vector.body:                                      ; preds = %vector.body, %for.cond4.preheader.us
67    %lsr.iv10 = phi i16* [ %scevgep11, %vector.body ], [ %lsr.iv7, %for.cond4.preheader.us ]
68    %lsr.iv4 = phi i16* [ %scevgep5, %vector.body ], [ %l, %for.cond4.preheader.us ]
69    %vec.phi = phi <4 x i32> [ %11, %for.cond4.preheader.us ], [ %19, %vector.body ]
70    %12 = phi i32 [ %start2, %for.cond4.preheader.us ], [ %20, %vector.body ]
71    %13 = phi i32 [ %d, %for.cond4.preheader.us ], [ %15, %vector.body ]
72    %lsr.iv1012 = bitcast i16* %lsr.iv10 to <4 x i16>*
73    %lsr.iv46 = bitcast i16* %lsr.iv4 to <4 x i16>*
74    %14 = call <4 x i1> @llvm.arm.mve.vctp32(i32 %13)
75    %15 = sub i32 %13, 4
76    %wide.masked.load = call <4 x i16> @llvm.masked.load.v4i16.p0v4i16(<4 x i16>* %lsr.iv1012, i32 2, <4 x i1> %14, <4 x i16> undef)
77    %16 = sext <4 x i16> %wide.masked.load to <4 x i32>
78    %wide.masked.load76 = call <4 x i16> @llvm.masked.load.v4i16.p0v4i16(<4 x i16>* %lsr.iv46, i32 2, <4 x i1> %14, <4 x i16> undef)
79    %17 = sext <4 x i16> %wide.masked.load76 to <4 x i32>
80    %18 = mul nsw <4 x i32> %17, %16
81    %19 = add <4 x i32> %18, %vec.phi
82    %scevgep5 = getelementptr i16, i16* %lsr.iv4, i32 4
83    %scevgep11 = getelementptr i16, i16* %lsr.iv10, i32 4
84    %20 = call i32 @llvm.loop.decrement.reg.i32.i32.i32(i32 %12, i32 1)
85    %21 = icmp ne i32 %20, 0
86    br i1 %21, label %vector.body, label %middle.block
87
88  middle.block:                                     ; preds = %vector.body
89    %vec.phi.lcssa = phi <4 x i32> [ %vec.phi, %vector.body ]
90    %.lcssa = phi <4 x i32> [ %19, %vector.body ]
91    %22 = call <4 x i1> @llvm.arm.mve.vctp32(i32 %10)
92    %23 = bitcast i16* %lsr.iv7 to i1*
93    %24 = select <4 x i1> %22, <4 x i32> %.lcssa, <4 x i32> %vec.phi.lcssa
94    %25 = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> %24)
95    %sunkaddr = mul i32 %i.064.us, 4
96    %26 = bitcast i32* %e to i8*
97    %sunkaddr17 = getelementptr inbounds i8, i8* %26, i32 %sunkaddr
98    %27 = bitcast i8* %sunkaddr17 to i32*
99    store i32 %25, i32* %27, align 4
100    %inc15.us = add nuw nsw i32 %i.064.us, 1
101    %scevgep9 = getelementptr i1, i1* %23, i32 %5
102    %28 = bitcast i1* %scevgep9 to i16*
103    %exitcond66 = icmp eq i32 %inc15.us, %d
104    br i1 %exitcond66, label %for.end16, label %for.cond4.preheader.us
105
106  for.end16:                                        ; preds = %middle.block
107    %29 = icmp sgt i32 %d, 0
108    br i1 %29, label %for.body.i57.preheader, label %c.exit59
109
110  for.body.i57.preheader:                           ; preds = %for.end16
111    %start3 = call i32 @llvm.start.loop.iterations.i32(i32 %d)
112    br label %for.body.i57
113
114  for.body.i57:                                     ; preds = %for.body.i57, %for.body.i57.preheader
115    %lsr.iv1 = phi i32* [ %e, %for.body.i57.preheader ], [ %scevgep, %for.body.i57 ]
116    %h.022.i44 = phi i16 [ %h.1.i54, %for.body.i57 ], [ 0, %for.body.i57.preheader ]
117    %f.020.i46 = phi i32 [ %f.1.i51, %for.body.i57 ], [ undef, %for.body.i57.preheader ]
118    %30 = phi i32 [ %start3, %for.body.i57.preheader ], [ %32, %for.body.i57 ]
119    %31 = load i32, i32* %lsr.iv1, align 4
120    %add.i48 = add nsw i32 %31, %f.020.i46
121    %cmp1.i49 = icmp sgt i32 %add.i48, 0
122    %cmp3.i50 = icmp sgt i32 %31, 0
123    %f.1.i51 = select i1 %cmp1.i49, i32 0, i32 %add.i48
124    %narrow.i52 = and i1 %cmp3.i50, %cmp1.i49
125    %add6.i53 = zext i1 %narrow.i52 to i16
126    %h.1.i54 = add i16 %h.022.i44, %add6.i53
127    %scevgep = getelementptr i32, i32* %lsr.iv1, i32 1
128    %32 = call i32 @llvm.loop.decrement.reg.i32.i32.i32(i32 %30, i32 1)
129    %33 = icmp ne i32 %32, 0
130    br i1 %33, label %for.body.i57, label %c.exit59.loopexit
131
132  c.exit59.loopexit:                                ; preds = %for.body.i57
133    %phitmp67 = sext i16 %h.1.i54 to i32
134    br label %c.exit59
135
136  c.exit59:                                         ; preds = %c.exit59.loopexit, %for.end16, %c.exit, %c.exit.thread
137    %conv27173 = phi i32 [ %conv2, %for.end16 ], [ %conv2, %c.exit59.loopexit ], [ %conv2, %c.exit ], [ %conv270, %c.exit.thread ]
138    %h.0.lcssa.i58 = phi i32 [ 0, %for.end16 ], [ %phitmp67, %c.exit59.loopexit ], [ 0, %c.exit ], [ 0, %c.exit.thread ]
139    %call19 = tail call arm_aapcs_vfpcc signext i16 bitcast (i16 (...)* @crc16 to i16 (i32, i32)*)(i32 %h.0.lcssa.i58, i32 %conv27173)
140    ret i16 %call19
141  }
142  declare dso_local arm_aapcs_vfpcc signext i16 @crc16(...) local_unnamed_addr #0
143  declare <4 x i16> @llvm.masked.load.v4i16.p0v4i16(<4 x i16>*, i32 immarg, <4 x i1>, <4 x i16>) #1
144  declare i32 @llvm.vector.reduce.add.v4i32(<4 x i32>) #2
145  declare i32 @llvm.start.loop.iterations.i32(i32) #3
146  declare i32 @llvm.loop.decrement.reg.i32.i32.i32(i32, i32) #3
147  declare <4 x i1> @llvm.arm.mve.vctp32(i32) #4
148
149...
150---
151name:            matrix_test
152alignment:       2
153exposesReturnsTwice: false
154legalized:       false
155regBankSelected: false
156selected:        false
157failedISel:      false
158tracksRegLiveness: true
159hasWinCFI:       false
160registers:       []
161liveins:
162  - { reg: '$r0', virtual-reg: '' }
163  - { reg: '$r1', virtual-reg: '' }
164  - { reg: '$r2', virtual-reg: '' }
165  - { reg: '$r3', virtual-reg: '' }
166frameInfo:
167  isFrameAddressTaken: false
168  isReturnAddressTaken: false
169  hasStackMap:     false
170  hasPatchPoint:   false
171  stackSize:       32
172  offsetAdjustment: 0
173  maxAlignment:    4
174  adjustsStack:    true
175  hasCalls:        true
176  stackProtector:  ''
177  maxCallFrameSize: 0
178  cvBytesOfCalleeSavedRegisters: 0
179  hasOpaqueSPAdjustment: false
180  hasVAStart:      false
181  hasMustTailInVarArgFunc: false
182  localFrameSize:  0
183  savePoint:       ''
184  restorePoint:    ''
185fixedStack:      []
186stack:
187  - { id: 0, name: '', type: spill-slot, offset: -4, size: 4, alignment: 4,
188      stack-id: default, callee-saved-register: '$lr', callee-saved-restored: true,
189      debug-info-variable: '', debug-info-expression: '', debug-info-location: '' }
190  - { id: 1, name: '', type: spill-slot, offset: -8, size: 4, alignment: 4,
191      stack-id: default, callee-saved-register: '$r10', callee-saved-restored: true,
192      debug-info-variable: '', debug-info-expression: '', debug-info-location: '' }
193  - { id: 2, name: '', type: spill-slot, offset: -12, size: 4, alignment: 4,
194      stack-id: default, callee-saved-register: '$r9', callee-saved-restored: true,
195      debug-info-variable: '', debug-info-expression: '', debug-info-location: '' }
196  - { id: 3, name: '', type: spill-slot, offset: -16, size: 4, alignment: 4,
197      stack-id: default, callee-saved-register: '$r8', callee-saved-restored: true,
198      debug-info-variable: '', debug-info-expression: '', debug-info-location: '' }
199  - { id: 4, name: '', type: spill-slot, offset: -20, size: 4, alignment: 4,
200      stack-id: default, callee-saved-register: '$r7', callee-saved-restored: true,
201      debug-info-variable: '', debug-info-expression: '', debug-info-location: '' }
202  - { id: 5, name: '', type: spill-slot, offset: -24, size: 4, alignment: 4,
203      stack-id: default, callee-saved-register: '$r6', callee-saved-restored: true,
204      debug-info-variable: '', debug-info-expression: '', debug-info-location: '' }
205  - { id: 6, name: '', type: spill-slot, offset: -28, size: 4, alignment: 4,
206      stack-id: default, callee-saved-register: '$r5', callee-saved-restored: true,
207      debug-info-variable: '', debug-info-expression: '', debug-info-location: '' }
208  - { id: 7, name: '', type: spill-slot, offset: -32, size: 4, alignment: 4,
209      stack-id: default, callee-saved-register: '$r4', callee-saved-restored: true,
210      debug-info-variable: '', debug-info-expression: '', debug-info-location: '' }
211callSites:       []
212constants:       []
213machineFunctionInfo: {}
214body:             |
215  ; CHECK-LABEL: name: matrix_test
216  ; CHECK: bb.0.entry:
217  ; CHECK:   successors: %bb.1(0x50000000), %bb.12(0x30000000)
218  ; CHECK:   liveins: $lr, $r0, $r1, $r2, $r3, $r4, $r5, $r6, $r7, $r8, $r9, $r10
219  ; CHECK:   $sp = frame-setup t2STMDB_UPD $sp, 14 /* CC::al */, $noreg, killed $r4, killed $r5, killed $r6, killed $r7, killed $r8, killed $r9, killed $r10, killed $lr
220  ; CHECK:   frame-setup CFI_INSTRUCTION def_cfa_offset 32
221  ; CHECK:   frame-setup CFI_INSTRUCTION offset $lr, -4
222  ; CHECK:   frame-setup CFI_INSTRUCTION offset $r10, -8
223  ; CHECK:   frame-setup CFI_INSTRUCTION offset $r9, -12
224  ; CHECK:   frame-setup CFI_INSTRUCTION offset $r8, -16
225  ; CHECK:   frame-setup CFI_INSTRUCTION offset $r7, -20
226  ; CHECK:   frame-setup CFI_INSTRUCTION offset $r6, -24
227  ; CHECK:   frame-setup CFI_INSTRUCTION offset $r5, -28
228  ; CHECK:   frame-setup CFI_INSTRUCTION offset $r4, -32
229  ; CHECK:   tCMPi8 renamable $r0, 1, 14 /* CC::al */, $noreg, implicit-def $cpsr
230  ; CHECK:   t2Bcc %bb.12, 11 /* CC::lt */, killed $cpsr
231  ; CHECK: bb.1.for.body.i.preheader:
232  ; CHECK:   successors: %bb.2(0x80000000)
233  ; CHECK:   liveins: $r0, $r1, $r2, $r3
234  ; CHECK:   $r5 = tMOVr killed $r2, 14 /* CC::al */, $noreg
235  ; CHECK:   $r8 = tMOVr killed $r3, 14 /* CC::al */, $noreg
236  ; CHECK:   $r4 = tMOVr $r1, 14 /* CC::al */, $noreg
237  ; CHECK:   renamable $r6, dead $cpsr = tMOVi8 0, 14 /* CC::al */, $noreg
238  ; CHECK:   renamable $r2 = IMPLICIT_DEF
239  ; CHECK:   $r10 = tMOVr $r0, 14 /* CC::al */, $noreg
240  ; CHECK:   $lr = t2DLS killed renamable $r0
241  ; CHECK: bb.2.for.body.i:
242  ; CHECK:   successors: %bb.2(0x7c000000), %bb.3(0x04000000)
243  ; CHECK:   liveins: $lr, $r1, $r2, $r4, $r5, $r6, $r8, $r10
244  ; CHECK:   renamable $r3, renamable $r1 = t2LDR_POST killed renamable $r1, 4, 14 /* CC::al */, $noreg :: (load 4 from %ir.lsr.iv15)
245  ; CHECK:   renamable $r2 = nsw tADDhirr killed renamable $r2, renamable $r3, 14 /* CC::al */, $noreg
246  ; CHECK:   tCMPi8 renamable $r2, 0, 14 /* CC::al */, $noreg, implicit-def $cpsr
247  ; CHECK:   renamable $r7 = t2CSINC $zr, $zr, 13, implicit killed $cpsr
248  ; CHECK:   tCMPi8 killed renamable $r3, 0, 14 /* CC::al */, $noreg, implicit-def $cpsr
249  ; CHECK:   renamable $r3 = t2CSINC $zr, $zr, 13, implicit killed $cpsr
250  ; CHECK:   tCMPi8 renamable $r2, 0, 14 /* CC::al */, $noreg, implicit-def $cpsr
251  ; CHECK:   renamable $r3 = t2ANDrr killed renamable $r3, killed renamable $r7, 14 /* CC::al */, $noreg, $noreg
252  ; CHECK:   t2IT 12, 8, implicit-def $itstate
253  ; CHECK:   $r2 = tMOVi8 $noreg, 0, 12 /* CC::gt */, killed $cpsr, implicit killed renamable $r2, implicit killed $itstate
254  ; CHECK:   renamable $r6 = tADDhirr killed renamable $r6, killed renamable $r3, 14 /* CC::al */, $noreg
255  ; CHECK:   $lr = t2LEUpdate killed renamable $lr, %bb.2
256  ; CHECK: bb.3.c.exit:
257  ; CHECK:   successors: %bb.4(0x50000000), %bb.14(0x30000000)
258  ; CHECK:   liveins: $r4, $r5, $r6, $r8, $r10
259  ; CHECK:   renamable $r0 = tSXTH killed renamable $r6, 14 /* CC::al */, $noreg
260  ; CHECK:   tBL 14 /* CC::al */, $noreg, @crc16, csr_aapcs, implicit-def dead $lr, implicit $sp, implicit killed $r0, implicit-def $sp, implicit-def $r0
261  ; CHECK:   $r12 = tMOVr killed $r0, 14 /* CC::al */, $noreg
262  ; CHECK:   renamable $r7, dead $cpsr = tMOVi8 0, 14 /* CC::al */, $noreg
263  ; CHECK:   t2CMPri $r10, 1, 14 /* CC::al */, $noreg, implicit-def $cpsr
264  ; CHECK:   tBcc %bb.14, 11 /* CC::lt */, killed $cpsr
265  ; CHECK: bb.4.for.cond4.preheader.us.preheader:
266  ; CHECK:   successors: %bb.5(0x80000000)
267  ; CHECK:   liveins: $r4, $r5, $r7, $r8, $r10, $r12
268  ; CHECK:   renamable $r0 = t2ADDri $r10, 3, 14 /* CC::al */, $noreg, $noreg
269  ; CHECK:   $lr = tMOVr $r10, 14 /* CC::al */, $noreg
270  ; CHECK:   renamable $r0 = t2BICri killed renamable $r0, 3, 14 /* CC::al */, $noreg, $noreg
271  ; CHECK:   renamable $r3 = t2LSLri $r10, 1, 14 /* CC::al */, $noreg, $noreg
272  ; CHECK:   renamable $r1, dead $cpsr = tSUBi3 killed renamable $r0, 4, 14 /* CC::al */, $noreg
273  ; CHECK:   renamable $r0, dead $cpsr = tMOVi8 1, 14 /* CC::al */, $noreg
274  ; CHECK:   renamable $q0 = MVE_VDUP32 renamable $r7, 0, $noreg, undef renamable $q0
275  ; CHECK:   renamable $r0 = nuw nsw t2ADDrs killed renamable $r0, renamable $r1, 19, 14 /* CC::al */, $noreg, $noreg
276  ; CHECK:   renamable $r1, dead $cpsr = tLSRri killed renamable $r1, 2, 14 /* CC::al */, $noreg
277  ; CHECK:   renamable $r9 = t2SUBrs $r10, killed renamable $r1, 18, 14 /* CC::al */, $noreg, $noreg
278  ; CHECK: bb.5.for.cond4.preheader.us:
279  ; CHECK:   successors: %bb.6(0x80000000)
280  ; CHECK:   liveins: $lr, $q0, $r0, $r3, $r4, $r5, $r7, $r8, $r9, $r10, $r12
281  ; CHECK:   renamable $r1 = t2LDRs renamable $r4, renamable $r7, 2, 14 /* CC::al */, $noreg :: (load 4 from %ir.arrayidx12.us)
282  ; CHECK:   $q1 = MVE_VORR $q0, $q0, 0, $noreg, undef $q1
283  ; CHECK:   $r2 = tMOVr killed $lr, 14 /* CC::al */, $noreg
284  ; CHECK:   renamable $q1 = MVE_VMOV_to_lane_32 killed renamable $q1, killed renamable $r1, 0, 14 /* CC::al */, $noreg
285  ; CHECK:   $r6 = tMOVr $r5, 14 /* CC::al */, $noreg
286  ; CHECK:   $r1 = tMOVr $r8, 14 /* CC::al */, $noreg
287  ; CHECK:   $lr = t2DLS renamable $r0
288  ; CHECK: bb.6.vector.body:
289  ; CHECK:   successors: %bb.6(0x7c000000), %bb.7(0x04000000)
290  ; CHECK:   liveins: $lr, $q0, $q1, $r0, $r1, $r2, $r3, $r4, $r5, $r6, $r7, $r8, $r9, $r10, $r12
291  ; CHECK:   renamable $vpr = MVE_VCTP32 renamable $r2, 0, $noreg
292  ; CHECK:   $q2 = MVE_VORR killed $q1, killed $q1, 0, $noreg, undef $q2
293  ; CHECK:   MVE_VPST 4, implicit $vpr
294  ; CHECK:   renamable $r6, renamable $q1 = MVE_VLDRHS32_post killed renamable $r6, 8, 1, renamable $vpr :: (load 8 from %ir.lsr.iv1012, align 2)
295  ; CHECK:   renamable $r1, renamable $q3 = MVE_VLDRHS32_post killed renamable $r1, 8, 1, killed renamable $vpr :: (load 8 from %ir.lsr.iv46, align 2)
296  ; CHECK:   renamable $r2, dead $cpsr = tSUBi8 killed renamable $r2, 4, 14 /* CC::al */, $noreg
297  ; CHECK:   renamable $q1 = nsw MVE_VMULi32 killed renamable $q3, killed renamable $q1, 0, $noreg, undef renamable $q1
298  ; CHECK:   renamable $q1 = MVE_VADDi32 killed renamable $q1, renamable $q2, 0, $noreg, undef renamable $q1
299  ; CHECK:   $lr = t2LEUpdate killed renamable $lr, %bb.6
300  ; CHECK: bb.7.middle.block:
301  ; CHECK:   successors: %bb.8(0x04000000), %bb.5(0x7c000000)
302  ; CHECK:   liveins: $q0, $q1, $q2, $r0, $r3, $r4, $r5, $r7, $r8, $r9, $r10, $r12
303  ; CHECK:   renamable $vpr = MVE_VCTP32 renamable $r9, 0, $noreg
304  ; CHECK:   renamable $r5 = tADDhirr killed renamable $r5, renamable $r3, 14 /* CC::al */, $noreg
305  ; CHECK:   renamable $q1 = MVE_VPSEL killed renamable $q1, killed renamable $q2, 0, killed renamable $vpr
306  ; CHECK:   $lr = tMOVr $r10, 14 /* CC::al */, $noreg
307  ; CHECK:   renamable $r2 = MVE_VADDVu32no_acc killed renamable $q1, 0, $noreg
308  ; CHECK:   t2STRs killed renamable $r2, renamable $r4, renamable $r7, 2, 14 /* CC::al */, $noreg :: (store 4 into %ir.27)
309  ; CHECK:   renamable $r7, dead $cpsr = nuw nsw tADDi8 killed renamable $r7, 1, 14 /* CC::al */, $noreg
310  ; CHECK:   tCMPhir renamable $r7, $r10, 14 /* CC::al */, $noreg, implicit-def $cpsr
311  ; CHECK:   tBcc %bb.5, 1 /* CC::ne */, killed $cpsr
312  ; CHECK: bb.8.for.end16:
313  ; CHECK:   successors: %bb.9(0x50000000), %bb.13(0x30000000)
314  ; CHECK:   liveins: $lr, $r4, $r12
315  ; CHECK:   t2CMPri renamable $lr, 1, 14 /* CC::al */, $noreg, implicit-def $cpsr
316  ; CHECK:   tBcc %bb.13, 11 /* CC::lt */, killed $cpsr
317  ; CHECK: bb.9.for.body.i57.preheader:
318  ; CHECK:   successors: %bb.10(0x80000000)
319  ; CHECK:   liveins: $lr, $r4, $r12
320  ; CHECK:   renamable $r0, dead $cpsr = tMOVi8 0, 14 /* CC::al */, $noreg
321  ; CHECK:   renamable $r1 = IMPLICIT_DEF
322  ; CHECK:   $lr = t2DLS killed renamable $lr
323  ; CHECK: bb.10.for.body.i57:
324  ; CHECK:   successors: %bb.10(0x7c000000), %bb.11(0x04000000)
325  ; CHECK:   liveins: $lr, $r0, $r1, $r4, $r12
326  ; CHECK:   renamable $r2, renamable $r4 = t2LDR_POST killed renamable $r4, 4, 14 /* CC::al */, $noreg :: (load 4 from %ir.lsr.iv1)
327  ; CHECK:   renamable $r1 = nsw tADDhirr killed renamable $r1, renamable $r2, 14 /* CC::al */, $noreg
328  ; CHECK:   tCMPi8 renamable $r1, 0, 14 /* CC::al */, $noreg, implicit-def $cpsr
329  ; CHECK:   renamable $r3 = t2CSINC $zr, $zr, 13, implicit killed $cpsr
330  ; CHECK:   tCMPi8 killed renamable $r2, 0, 14 /* CC::al */, $noreg, implicit-def $cpsr
331  ; CHECK:   renamable $r2 = t2CSINC $zr, $zr, 13, implicit killed $cpsr
332  ; CHECK:   tCMPi8 renamable $r1, 0, 14 /* CC::al */, $noreg, implicit-def $cpsr
333  ; CHECK:   renamable $r2 = t2ANDrr killed renamable $r2, killed renamable $r3, 14 /* CC::al */, $noreg, $noreg
334  ; CHECK:   t2IT 12, 8, implicit-def $itstate
335  ; CHECK:   $r1 = tMOVi8 $noreg, 0, 12 /* CC::gt */, killed $cpsr, implicit killed renamable $r1, implicit killed $itstate
336  ; CHECK:   renamable $r0 = tADDhirr killed renamable $r0, killed renamable $r2, 14 /* CC::al */, $noreg
337  ; CHECK:   $lr = t2LEUpdate killed renamable $lr, %bb.10
338  ; CHECK: bb.11.c.exit59.loopexit:
339  ; CHECK:   successors: %bb.14(0x80000000)
340  ; CHECK:   liveins: $r0, $r12
341  ; CHECK:   renamable $r7 = tSXTH killed renamable $r0, 14 /* CC::al */, $noreg
342  ; CHECK:   tB %bb.14, 14 /* CC::al */, $noreg
343  ; CHECK: bb.12.c.exit.thread:
344  ; CHECK:   successors: %bb.14(0x80000000)
345  ; CHECK:   $r0, dead $cpsr = tMOVi8 0, 14 /* CC::al */, $noreg
346  ; CHECK:   renamable $r7, dead $cpsr = tMOVi8 0, 14 /* CC::al */, $noreg
347  ; CHECK:   tBL 14 /* CC::al */, $noreg, @crc16, csr_aapcs, implicit-def dead $lr, implicit $sp, implicit killed $r0, implicit-def $sp, implicit-def $r0
348  ; CHECK:   $r12 = tMOVr killed $r0, 14 /* CC::al */, $noreg
349  ; CHECK:   tB %bb.14, 14 /* CC::al */, $noreg
350  ; CHECK: bb.13:
351  ; CHECK:   successors: %bb.14(0x80000000)
352  ; CHECK:   liveins: $r12
353  ; CHECK:   renamable $r7, dead $cpsr = tMOVi8 0, 14 /* CC::al */, $noreg
354  ; CHECK: bb.14.c.exit59:
355  ; CHECK:   liveins: $r7, $r12
356  ; CHECK:   $r0 = tMOVr killed $r7, 14 /* CC::al */, $noreg
357  ; CHECK:   $r1 = tMOVr killed $r12, 14 /* CC::al */, $noreg
358  ; CHECK:   $sp = t2LDMIA_UPD $sp, 14 /* CC::al */, $noreg, def $r4, def $r5, def $r6, def $r7, def $r8, def $r9, def $r10, def $lr
359  ; CHECK:   tTAILJMPdND @crc16, 14 /* CC::al */, $noreg, implicit $sp, implicit $sp, implicit killed $r0, implicit killed $r1
360  bb.0.entry:
361    successors: %bb.1(0x50000000), %bb.12(0x30000000)
362    liveins: $r0, $r1, $r2, $r3, $r4, $r5, $r6, $r7, $r8, $r9, $r10, $lr
363
364    $sp = frame-setup t2STMDB_UPD $sp, 14, $noreg, killed $r4, killed $r5, killed $r6, killed $r7, killed $r8, killed $r9, killed $r10, killed $lr
365    frame-setup CFI_INSTRUCTION def_cfa_offset 32
366    frame-setup CFI_INSTRUCTION offset $lr, -4
367    frame-setup CFI_INSTRUCTION offset $r10, -8
368    frame-setup CFI_INSTRUCTION offset $r9, -12
369    frame-setup CFI_INSTRUCTION offset $r8, -16
370    frame-setup CFI_INSTRUCTION offset $r7, -20
371    frame-setup CFI_INSTRUCTION offset $r6, -24
372    frame-setup CFI_INSTRUCTION offset $r5, -28
373    frame-setup CFI_INSTRUCTION offset $r4, -32
374    tCMPi8 renamable $r0, 1, 14, $noreg, implicit-def $cpsr
375    t2Bcc %bb.12, 11, killed $cpsr
376
377  bb.1.for.body.i.preheader:
378    successors: %bb.2(0x80000000)
379    liveins: $r0, $r1, $r2, $r3
380
381    $r5 = tMOVr killed $r2, 14, $noreg
382    $r8 = tMOVr killed $r3, 14, $noreg
383    $r4 = tMOVr $r1, 14, $noreg
384    renamable $r6, dead $cpsr = tMOVi8 0, 14, $noreg
385    renamable $r2 = IMPLICIT_DEF
386    $r10 = tMOVr $r0, 14, $noreg
387    $lr = tMOVr $r0, 14, $noreg
388    $lr = t2DoLoopStart killed renamable $r0
389
390  bb.2.for.body.i:
391    successors: %bb.2(0x7c000000), %bb.3(0x04000000)
392    liveins: $lr, $r1, $r2, $r4, $r5, $r6, $r8, $r10
393
394    renamable $r3, renamable $r1 = t2LDR_POST killed renamable $r1, 4, 14, $noreg :: (load 4 from %ir.lsr.iv15)
395    renamable $lr = t2LoopDec killed renamable $lr, 1
396    renamable $r2 = nsw tADDhirr killed renamable $r2, renamable $r3, 14, $noreg
397    tCMPi8 renamable $r2, 0, 14, $noreg, implicit-def $cpsr
398    renamable $r7 = t2CSINC $zr, $zr, 13, implicit killed $cpsr
399    tCMPi8 killed renamable $r3, 0, 14, $noreg, implicit-def $cpsr
400    renamable $r3 = t2CSINC $zr, $zr, 13, implicit killed $cpsr
401    tCMPi8 renamable $r2, 0, 14, $noreg, implicit-def $cpsr
402    renamable $r3 = t2ANDrr killed renamable $r3, killed renamable $r7, 14, $noreg, $noreg
403    t2IT 12, 8, implicit-def $itstate
404    $r2 = tMOVi8 $noreg, 0, 12, killed $cpsr, implicit killed renamable $r2, implicit killed $itstate
405    renamable $r6 = tADDhirr killed renamable $r6, killed renamable $r3, 14, $noreg
406    t2LoopEnd renamable $lr, %bb.2, implicit-def dead $cpsr
407    tB %bb.3, 14, $noreg
408
409  bb.3.c.exit:
410    successors: %bb.4(0x50000000), %bb.14(0x30000000)
411    liveins: $r4, $r5, $r6, $r8, $r10
412
413    renamable $r0 = tSXTH killed renamable $r6, 14, $noreg
414    tBL 14, $noreg, @crc16, csr_aapcs, implicit-def dead $lr, implicit $sp, implicit killed $r0, implicit-def $sp, implicit-def $r0
415    $r12 = tMOVr killed $r0, 14, $noreg
416    renamable $r7, dead $cpsr = tMOVi8 0, 14, $noreg
417    t2CMPri $r10, 1, 14, $noreg, implicit-def $cpsr
418    tBcc %bb.14, 11, killed $cpsr
419
420  bb.4.for.cond4.preheader.us.preheader:
421    successors: %bb.5(0x80000000)
422    liveins: $r4, $r5, $r7, $r8, $r10, $r12
423
424    renamable $r0 = t2ADDri $r10, 3, 14, $noreg, $noreg
425    $lr = tMOVr $r10, 14, $noreg
426    renamable $r0 = t2BICri killed renamable $r0, 3, 14, $noreg, $noreg
427    renamable $r3 = t2LSLri $r10, 1, 14, $noreg, $noreg
428    renamable $r1, dead $cpsr = tSUBi3 killed renamable $r0, 4, 14, $noreg
429    renamable $r0, dead $cpsr = tMOVi8 1, 14, $noreg
430    renamable $q0 = MVE_VDUP32 renamable $r7, 0, $noreg, undef renamable $q0
431    renamable $r0 = nuw nsw t2ADDrs killed renamable $r0, renamable $r1, 19, 14, $noreg, $noreg
432    renamable $r1, dead $cpsr = tLSRri killed renamable $r1, 2, 14, $noreg
433    renamable $r9 = t2SUBrs $r10, killed renamable $r1, 18, 14, $noreg, $noreg
434
435  bb.5.for.cond4.preheader.us:
436    successors: %bb.6(0x80000000)
437    liveins: $lr, $q0, $r0, $r3, $r4, $r5, $r7, $r8, $r9, $r10, $r12
438
439    renamable $r1 = t2LDRs renamable $r4, renamable $r7, 2, 14, $noreg :: (load 4 from %ir.arrayidx12.us)
440    $q1 = MVE_VORR $q0, $q0, 0, $noreg, undef $q1
441    $r2 = tMOVr killed $lr, 14, $noreg
442    renamable $q1 = MVE_VMOV_to_lane_32 killed renamable $q1, killed renamable $r1, 0, 14, $noreg
443    $r6 = tMOVr $r5, 14, $noreg
444    $r1 = tMOVr $r8, 14, $noreg
445    $lr = tMOVr $r0, 14, $noreg
446    $lr = t2DoLoopStart renamable $r0
447
448  bb.6.vector.body:
449    successors: %bb.6(0x7c000000), %bb.7(0x04000000)
450    liveins: $lr, $q0, $q1, $r0, $r1, $r2, $r3, $r4, $r5, $r6, $r7, $r8, $r9, $r10, $r12
451
452    renamable $vpr = MVE_VCTP32 renamable $r2, 0, $noreg
453    $q2 = MVE_VORR killed $q1, $q1, 0, $noreg, undef $q2
454    MVE_VPST 4, implicit $vpr
455    renamable $r6, renamable $q1 = MVE_VLDRHS32_post killed renamable $r6, 8, 1, renamable $vpr :: (load 8 from %ir.lsr.iv1012, align 2)
456    renamable $r1, renamable $q3 = MVE_VLDRHS32_post killed renamable $r1, 8, 1, killed renamable $vpr :: (load 8 from %ir.lsr.iv46, align 2)
457    renamable $r2, dead $cpsr = tSUBi8 killed renamable $r2, 4, 14, $noreg
458    renamable $q1 = nsw MVE_VMULi32 killed renamable $q3, killed renamable $q1, 0, $noreg, undef renamable $q1
459    renamable $lr = t2LoopDec killed renamable $lr, 1
460    renamable $q1 = MVE_VADDi32 killed renamable $q1, renamable $q2, 0, $noreg, undef renamable $q1
461    t2LoopEnd renamable $lr, %bb.6, implicit-def dead $cpsr
462    tB %bb.7, 14, $noreg
463
464  bb.7.middle.block:
465    successors: %bb.8(0x04000000), %bb.5(0x7c000000)
466    liveins: $q0, $q1, $q2, $r0, $r3, $r4, $r5, $r7, $r8, $r9, $r10, $r12
467
468    renamable $vpr = MVE_VCTP32 renamable $r9, 0, $noreg
469    renamable $r5 = tADDhirr killed renamable $r5, renamable $r3, 14, $noreg
470    renamable $q1 = MVE_VPSEL killed renamable $q1, killed renamable $q2, 0, killed renamable $vpr
471    $lr = tMOVr $r10, 14, $noreg
472    renamable $r2 = MVE_VADDVu32no_acc killed renamable $q1, 0, $noreg
473    t2STRs killed renamable $r2, renamable $r4, renamable $r7, 2, 14, $noreg :: (store 4 into %ir.27)
474    renamable $r7, dead $cpsr = nuw nsw tADDi8 killed renamable $r7, 1, 14, $noreg
475    tCMPhir renamable $r7, $r10, 14, $noreg, implicit-def $cpsr
476    tBcc %bb.5, 1, killed $cpsr
477
478  bb.8.for.end16:
479    successors: %bb.9(0x50000000), %bb.13(0x30000000)
480    liveins: $lr, $r4, $r12
481
482    t2CMPri renamable $lr, 1, 14, $noreg, implicit-def $cpsr
483    tBcc %bb.13, 11, killed $cpsr
484
485  bb.9.for.body.i57.preheader:
486    successors: %bb.10(0x80000000)
487    liveins: $lr, $r4, $r12
488
489    renamable $r0, dead $cpsr = tMOVi8 0, 14, $noreg
490    renamable $r1 = IMPLICIT_DEF
491    $lr = t2DoLoopStart renamable $lr
492
493  bb.10.for.body.i57:
494    successors: %bb.10(0x7c000000), %bb.11(0x04000000)
495    liveins: $lr, $r0, $r1, $r4, $r12
496
497    renamable $r2, renamable $r4 = t2LDR_POST killed renamable $r4, 4, 14, $noreg :: (load 4 from %ir.lsr.iv1)
498    renamable $lr = t2LoopDec killed renamable $lr, 1
499    renamable $r1 = nsw tADDhirr killed renamable $r1, renamable $r2, 14, $noreg
500    tCMPi8 renamable $r1, 0, 14, $noreg, implicit-def $cpsr
501    renamable $r3 = t2CSINC $zr, $zr, 13, implicit killed $cpsr
502    tCMPi8 killed renamable $r2, 0, 14, $noreg, implicit-def $cpsr
503    renamable $r2 = t2CSINC $zr, $zr, 13, implicit killed $cpsr
504    tCMPi8 renamable $r1, 0, 14, $noreg, implicit-def $cpsr
505    renamable $r2 = t2ANDrr killed renamable $r2, killed renamable $r3, 14, $noreg, $noreg
506    t2IT 12, 8, implicit-def $itstate
507    $r1 = tMOVi8 $noreg, 0, 12, killed $cpsr, implicit killed renamable $r1, implicit killed $itstate
508    renamable $r0 = tADDhirr killed renamable $r0, killed renamable $r2, 14, $noreg
509    t2LoopEnd renamable $lr, %bb.10, implicit-def dead $cpsr
510    tB %bb.11, 14, $noreg
511
512  bb.11.c.exit59.loopexit:
513    successors: %bb.14(0x80000000)
514    liveins: $r0, $r12
515
516    renamable $r7 = tSXTH killed renamable $r0, 14, $noreg
517    tB %bb.14, 14, $noreg
518
519  bb.12.c.exit.thread:
520    successors: %bb.14(0x80000000)
521
522    $r0, dead $cpsr = tMOVi8 0, 14, $noreg
523    renamable $r7, dead $cpsr = tMOVi8 0, 14, $noreg
524    tBL 14, $noreg, @crc16, csr_aapcs, implicit-def dead $lr, implicit $sp, implicit killed $r0, implicit-def $sp, implicit-def $r0
525    $r12 = tMOVr killed $r0, 14, $noreg
526    tB %bb.14, 14, $noreg
527
528  bb.13:
529    successors: %bb.14(0x80000000)
530    liveins: $r12
531
532    renamable $r7, dead $cpsr = tMOVi8 0, 14, $noreg
533
534  bb.14.c.exit59:
535    liveins: $r7, $r12
536
537    $r0 = tMOVr killed $r7, 14, $noreg
538    $r1 = tMOVr killed $r12, 14, $noreg
539    $sp = t2LDMIA_UPD $sp, 14, $noreg, def $r4, def $r5, def $r6, def $r7, def $r8, def $r9, def $r10, def $lr
540    tTAILJMPdND @crc16, 14, $noreg, implicit $sp, implicit $sp, implicit killed $r0, implicit killed $r1
541
542...
543