1 # RUN: llc %s -start-after=shrink-wrap -march=mips64 -mcpu=mips64r6 -mattr=+fp64,+msa -o /dev/null
3 # Test that estimated size of the stack leads to the creation of an emergency
4 # spill when MSA is in use. Previously, this test case would fail during
5 # register scavenging due to the lack of a spill slot.
7 define inreg { i64, i64 } @test(i64 inreg %a.coerce0, i64 inreg %a.coerce1, i64 inreg %b.coerce0, i64 inreg %b.coerce1, i32 signext %c) #0 {
9 %retval = alloca <16 x i8>, align 16
10 %a = alloca <16 x i8>, align 16
11 %b = alloca <16 x i8>, align 16
12 %a.addr = alloca <16 x i8>, align 16
13 %b.addr = alloca <16 x i8>, align 16
14 %c.addr = alloca i32, align 4
15 %g = alloca <16 x i8>*, align 8
16 %d = alloca i8*, align 8
17 %0 = bitcast <16 x i8>* %a to { i64, i64 }*
18 %1 = getelementptr inbounds { i64, i64 }, { i64, i64 }* %0, i32 0, i32 0
19 store i64 %a.coerce0, i64* %1, align 16
20 %2 = getelementptr inbounds { i64, i64 }, { i64, i64 }* %0, i32 0, i32 1
21 store i64 %a.coerce1, i64* %2, align 8
22 %a1 = load <16 x i8>, <16 x i8>* %a, align 16
23 %3 = bitcast <16 x i8>* %b to { i64, i64 }*
24 %4 = getelementptr inbounds { i64, i64 }, { i64, i64 }* %3, i32 0, i32 0
25 store i64 %b.coerce0, i64* %4, align 16
26 %5 = getelementptr inbounds { i64, i64 }, { i64, i64 }* %3, i32 0, i32 1
27 store i64 %b.coerce1, i64* %5, align 8
28 %b2 = load <16 x i8>, <16 x i8>* %b, align 16
29 store <16 x i8> %a1, <16 x i8>* %a.addr, align 16
30 store <16 x i8> %b2, <16 x i8>* %b.addr, align 16
31 store i32 %c, i32* %c.addr, align 4
32 %6 = alloca i8, i64 6400, align 16
33 %7 = bitcast i8* %6 to <16 x i8>*
34 store <16 x i8>* %7, <16 x i8>** %g, align 8
35 %8 = load <16 x i8>*, <16 x i8>** %g, align 8
36 call void @h(<16 x i8>* %b.addr, <16 x i8>* %8)
37 %9 = load <16 x i8>*, <16 x i8>** %g, align 8
38 %10 = bitcast <16 x i8>* %9 to i8*
39 store i8* %10, i8** %d, align 8
40 %11 = load <16 x i8>, <16 x i8>* %a.addr, align 16
41 %12 = load i8*, i8** %d, align 8
42 %arrayidx = getelementptr inbounds i8, i8* %12, i64 0
43 %13 = load i8, i8* %arrayidx, align 1
44 %conv = sext i8 %13 to i32
45 %14 = call <16 x i8> @llvm.mips.fill.b(i32 %conv)
46 %add = add <16 x i8> %11, %14
47 %15 = load i8*, i8** %d, align 8
48 %arrayidx3 = getelementptr inbounds i8, i8* %15, i64 1
49 %16 = load i8, i8* %arrayidx3, align 1
50 %conv4 = sext i8 %16 to i32
51 %17 = call <16 x i8> @llvm.mips.fill.b(i32 %conv4)
52 %add5 = add <16 x i8> %add, %17
53 %18 = load <16 x i8>, <16 x i8>* %b.addr, align 16
54 %add6 = add <16 x i8> %18, %add5
55 store <16 x i8> %add6, <16 x i8>* %b.addr, align 16
56 %19 = load <16 x i8>, <16 x i8>* %b.addr, align 16
57 store <16 x i8> %19, <16 x i8>* %retval, align 16
58 %20 = bitcast <16 x i8>* %retval to { i64, i64 }*
59 %21 = load { i64, i64 }, { i64, i64 }* %20, align 16
63 declare void @h(<16 x i8>*, <16 x i8>*)
65 declare <16 x i8> @llvm.mips.fill.b(i32)
67 declare void @llvm.stackprotector(i8*, i8**)
73 exposesReturnsTwice: false
75 regBankSelected: false
77 tracksRegLiveness: true
80 - { reg: '$a0_64', virtual-reg: '' }
81 - { reg: '$a1_64', virtual-reg: '' }
82 - { reg: '$a2_64', virtual-reg: '' }
83 - { reg: '$a3_64', virtual-reg: '' }
84 - { reg: '$t0_64', virtual-reg: '' }
86 isFrameAddressTaken: false
87 isReturnAddressTaken: false
96 maxCallFrameSize: 4294967295
97 hasOpaqueSPAdjustment: false
99 hasMustTailInVarArgFunc: false
104 - { id: 0, name: retval, type: default, offset: 0, size: 16, alignment: 16,
105 callee-saved-register: '', debug-info-variable: '',
106 debug-info-expression: '', debug-info-location: '' }
107 - { id: 1, name: a, type: default, offset: 0, size: 16, alignment: 16,
108 callee-saved-register: '', debug-info-variable: '',
109 debug-info-expression: '', debug-info-location: '' }
110 - { id: 2, name: b, type: default, offset: 0, size: 16, alignment: 16,
111 callee-saved-register: '', debug-info-variable: '',
112 debug-info-expression: '', debug-info-location: '' }
113 - { id: 3, name: a.addr, type: default, offset: 0, size: 16, alignment: 16,
114 callee-saved-register: '', debug-info-variable: '',
115 debug-info-expression: '', debug-info-location: '' }
116 - { id: 4, name: b.addr, type: default, offset: 0, size: 16, alignment: 16,
117 callee-saved-register: '', debug-info-variable: '',
118 debug-info-expression: '', debug-info-location: '' }
119 - { id: 5, name: c.addr, type: default, offset: 0, size: 4, alignment: 4,
120 callee-saved-register: '', debug-info-variable: '',
121 debug-info-expression: '', debug-info-location: '' }
122 - { id: 6, name: g, type: default, offset: 0, size: 8, alignment: 8,
123 callee-saved-register: '', debug-info-variable: '',
124 debug-info-expression: '', debug-info-location: '' }
125 - { id: 7, name: d, type: default, offset: 0, size: 8, alignment: 8,
126 callee-saved-register: '', debug-info-variable: '',
127 debug-info-expression: '', debug-info-location: '' }
128 - { id: 8, name: '', type: default, offset: 0, size: 6400,
129 alignment: 16, callee-saved-register: '', debug-info-variable: '',
130 debug-info-expression: '', debug-info-location: '' }
134 liveins: $a0_64, $a1_64, $a2_64, $a3_64, $t0_64
136 SD killed $a0_64, %stack.1.a, 0 :: (store 8 into %ir.1, align 16)
137 SD killed $a1_64, %stack.1.a, 8 :: (store 8 into %ir.2)
138 $w0 = LD_B %stack.1.a, 0 :: (dereferenceable load 16 from %ir.a)
139 SD killed $a2_64, %stack.2.b, 0 :: (store 8 into %ir.4, align 16)
140 SD killed $a3_64, %stack.2.b, 8 :: (store 8 into %ir.5)
141 $w1 = LD_B %stack.2.b, 0 :: (dereferenceable load 16 from %ir.b)
142 ST_B killed $w0, %stack.3.a.addr, 0 :: (store 16 into %ir.a.addr)
143 ST_B killed $w1, %stack.4.b.addr, 0 :: (store 16 into %ir.b.addr)
144 SW $t0, %stack.5.c.addr, 0, implicit killed $t0_64 :: (store 4 into %ir.c.addr)
145 $at_64 = LEA_ADDiu64 %stack.8, 0
146 SD killed $at_64, %stack.6.g, 0 :: (store 8 into %ir.g)
147 $a1_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
148 ADJCALLSTACKDOWN 0, 0, implicit-def dead $sp, implicit $sp
149 $a0_64 = LEA_ADDiu64 %stack.4.b.addr, 0
150 JAL @h, csr_n64, implicit-def dead $ra, implicit $a0_64, implicit $a1_64, implicit-def $sp
151 ADJCALLSTACKUP 0, 0, implicit-def dead $sp, implicit $sp
152 $at_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
153 $v0_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
154 $v1_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
155 $a0_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
156 $a1_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
157 $a2_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
158 $a3_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
159 $t0_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
160 $t1_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
161 $t2_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
162 $t3_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
163 $t4_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
164 $t5_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
165 $t6_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
166 $t7_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
167 $s0_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
168 $s1_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
169 $s2_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
170 $s3_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
171 $s4_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
172 $s5_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
173 $s6_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
174 $s7_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
175 $t8_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
176 $t9_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
177 $ra_64 = LD %stack.6.g, 0 :: (dereferenceable load 8 from %ir.g)
178 $w0 = LD_B %stack.3.a.addr, 0 :: (dereferenceable load 16 from %ir.a.addr)
179 SD $at_64, %stack.7.d, 0 :: (store 8 into %ir.d)
180 SD $v0_64, %stack.7.d, 0 :: (store 8 into %ir.d)
181 SD $v1_64, %stack.7.d, 0 :: (store 8 into %ir.d)
182 SD $a0_64, %stack.7.d, 0 :: (store 8 into %ir.d)
183 SD $a1_64, %stack.7.d, 0 :: (store 8 into %ir.d)
184 SD $a2_64, %stack.7.d, 0 :: (store 8 into %ir.d)
185 SD $a3_64, %stack.7.d, 0 :: (store 8 into %ir.d)
186 SD $t0_64, %stack.7.d, 0 :: (store 8 into %ir.d)
187 SD $t1_64, %stack.7.d, 0 :: (store 8 into %ir.d)
188 SD $t2_64, %stack.7.d, 0 :: (store 8 into %ir.d)
189 SD $t3_64, %stack.7.d, 0 :: (store 8 into %ir.d)
190 SD $t4_64, %stack.7.d, 0 :: (store 8 into %ir.d)
191 SD $t5_64, %stack.7.d, 0 :: (store 8 into %ir.d)
192 SD $t6_64, %stack.7.d, 0 :: (store 8 into %ir.d)
193 SD $t7_64, %stack.7.d, 0 :: (store 8 into %ir.d)
194 SD $s0_64, %stack.7.d, 0 :: (store 8 into %ir.d)
195 SD $s1_64, %stack.7.d, 0 :: (store 8 into %ir.d)
196 SD $s2_64, %stack.7.d, 0 :: (store 8 into %ir.d)
197 SD $s3_64, %stack.7.d, 0 :: (store 8 into %ir.d)
198 SD $s4_64, %stack.7.d, 0 :: (store 8 into %ir.d)
199 SD $s5_64, %stack.7.d, 0 :: (store 8 into %ir.d)
200 SD $s6_64, %stack.7.d, 0 :: (store 8 into %ir.d)
201 SD $s7_64, %stack.7.d, 0 :: (store 8 into %ir.d)
202 SD $t8_64, %stack.7.d, 0 :: (store 8 into %ir.d)
203 SD $t9_64, %stack.7.d, 0 :: (store 8 into %ir.d)
204 SD $ra_64, %stack.7.d, 0 :: (store 8 into %ir.d)
205 $at_64 = LD %stack.7.d, 0 :: (dereferenceable load 8 from %ir.d)
206 $v0 = LB $at_64, 0 :: (load 1 from %ir.arrayidx)
207 $w1 = FILL_B killed $v0
208 $w0 = ADDV_B killed $w0, killed $w1
209 $at = LB killed $at_64, 1 :: (load 1 from %ir.arrayidx3)
210 $w1 = FILL_B killed $at
211 $w0 = ADDV_B killed $w0, killed $w1
212 $w1 = LD_B %stack.4.b.addr, 0 :: (dereferenceable load 16 from %ir.b.addr)
213 $w0 = ADDV_B killed $w1, killed $w0
214 ST_B killed $w0, %stack.4.b.addr, 0 :: (store 16 into %ir.b.addr)
215 $w0 = LD_B %stack.4.b.addr, 0 :: (dereferenceable load 16 from %ir.b.addr)
216 ST_B killed $w0, %stack.0.retval, 0 :: (store 16 into %ir.retval)
217 $v0_64 = LD %stack.0.retval, 0 :: (dereferenceable load 8 from %ir.20, align 16)
218 $v1_64 = LD %stack.0.retval, 8 :: (dereferenceable load 8 from %ir.20 + 8, align 16)
219 RetRA implicit $v0_64, implicit $v1_64