1 ; PR26374: Check no stack slots are allocated for vregs which have no real reference.
2 ; RUN: llc < %s | FileCheck %s
3 target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
4 target triple = "x86_64-unknown-linux-gnu"
6 %struct.ImageParameters = type { i32, i32, [0 x [16 x i16]] }
7 %struct.InputParameters = type { i32, i32 }
9 @c = common global %struct.ImageParameters* null, align 8
10 @a = common global i16** null, align 8
11 @d = common global [6 x i32] zeroinitializer, align 16
12 @b = common global %struct.InputParameters* null, align 8
13 @e = common global [4 x i32] zeroinitializer, align 16
15 ; It is not easy to check there is no unused holes in stack allocated for spills,
16 ; so simply check the size of stack allocated cannot exceed 350.
17 ; (408 is used before the fix for PR26374. 344 is used after the fix).
20 ; CHECK: subq {{\$3[0-4][0-9]}}, %rsp
22 ; Function Attrs: nounwind uwtable
25 %n = alloca [8 x [8 x i32]], align 16
26 %tmp = bitcast [8 x [8 x i32]]* %n to i8*
27 call void @llvm.lifetime.start.p0i8(i64 256, i8* %tmp) #3
28 %tmp1 = bitcast [8 x [8 x i32]]* %n to i8*
29 %arraydecay.1 = getelementptr inbounds [8 x [8 x i32]], [8 x [8 x i32]]* %n, i64 0, i64 1, i64 0
30 %tmp2 = bitcast i32* %arraydecay.1 to i8*
31 %arraydecay.2 = getelementptr inbounds [8 x [8 x i32]], [8 x [8 x i32]]* %n, i64 0, i64 2, i64 0
32 %tmp3 = bitcast i32* %arraydecay.2 to i8*
33 %arraydecay.3 = getelementptr inbounds [8 x [8 x i32]], [8 x [8 x i32]]* %n, i64 0, i64 3, i64 0
34 %tmp4 = bitcast i32* %arraydecay.3 to i8*
35 %arraydecay.4 = getelementptr inbounds [8 x [8 x i32]], [8 x [8 x i32]]* %n, i64 0, i64 4, i64 0
36 %tmp5 = bitcast i32* %arraydecay.4 to i8*
37 %arraydecay.5 = getelementptr inbounds [8 x [8 x i32]], [8 x [8 x i32]]* %n, i64 0, i64 5, i64 0
38 %tmp6 = bitcast i32* %arraydecay.5 to i8*
39 %arraydecay.6 = getelementptr inbounds [8 x [8 x i32]], [8 x [8 x i32]]* %n, i64 0, i64 6, i64 0
40 %tmp7 = bitcast i32* %arraydecay.6 to i8*
41 %arraydecay.7 = getelementptr inbounds [8 x [8 x i32]], [8 x [8 x i32]]* %n, i64 0, i64 7, i64 0
42 %tmp8 = bitcast i32* %arraydecay.7 to i8*
45 for.body: ; preds = %for.inc73, %entry
46 %q.0131 = phi i32 [ 0, %entry ], [ %inc74, %for.inc73 ]
47 %m.0130 = phi i32 [ 0, %entry ], [ %m.4, %for.inc73 ]
48 %div = sdiv i32 %q.0131, 2
49 %shl = shl i32 %div, 3
50 %rem = srem i32 %q.0131, 2
51 %shl1 = shl nsw i32 %rem, 3
52 %tmp9 = sext i32 %shl1 to i64
53 %tmp10 = sext i32 %shl to i64
54 %tmp11 = or i32 %shl1, 4
55 %tmp12 = sext i32 %tmp11 to i64
56 %tmp13 = or i32 %shl, 4
57 %tmp14 = sext i32 %tmp13 to i64
60 for.body4: ; preds = %for.inc48, %for.body
61 %indvars.iv148 = phi i64 [ %tmp10, %for.body ], [ %indvars.iv.next149, %for.inc48 ]
62 %m.1126 = phi i32 [ %m.0130, %for.body ], [ %m.3.lcssa, %for.inc48 ]
63 %tmp15 = load %struct.ImageParameters*, %struct.ImageParameters** @c, align 8
64 %opix_y = getelementptr inbounds %struct.ImageParameters, %struct.ImageParameters* %tmp15, i64 0, i32 1
65 %tmp16 = load i32, i32* %opix_y, align 4
66 %tmp17 = trunc i64 %indvars.iv148 to i32
67 %add5 = add nsw i32 %tmp16, %tmp17
68 %tmp18 = sub nuw nsw i64 %indvars.iv148, %tmp10
69 %tmp19 = sext i32 %add5 to i64
70 %tmp20 = add nsw i64 %tmp19, 1
71 %tmp21 = or i64 %indvars.iv148, 1
72 %tmp22 = or i64 %tmp18, 1
73 %tmp23 = add nsw i64 %tmp19, 2
74 %tmp24 = or i64 %indvars.iv148, 2
75 %tmp25 = or i64 %tmp18, 2
76 %tmp26 = add nsw i64 %tmp19, 3
77 %tmp27 = or i64 %indvars.iv148, 3
78 %tmp28 = or i64 %tmp18, 3
81 for.body9: ; preds = %for.inc45.for.body9_crit_edge, %for.body4
82 %tmp29 = phi %struct.ImageParameters* [ %tmp15, %for.body4 ], [ %.pre, %for.inc45.for.body9_crit_edge ]
83 %indvars.iv145 = phi i64 [ %tmp9, %for.body4 ], [ %indvars.iv.next146, %for.inc45.for.body9_crit_edge ]
84 %m.2124 = phi i32 [ %m.1126, %for.body4 ], [ %m.3, %for.inc45.for.body9_crit_edge ]
85 %opix_x = getelementptr inbounds %struct.ImageParameters, %struct.ImageParameters* %tmp29, i64 0, i32 0
86 %tmp30 = load i32, i32* %opix_x, align 4
87 %tmp31 = trunc i64 %indvars.iv145 to i32
88 %add10 = add nsw i32 %tmp30, %tmp31
89 tail call void @LumaPrediction4x4(i32 %tmp31, i32 %tmp17, i32 0, i32 0, i32 0, i16 signext 0, i16 signext 0) #3
90 %tmp32 = load i16**, i16*** @a, align 8
91 %tmp33 = load %struct.ImageParameters*, %struct.ImageParameters** @c, align 8
92 %tmp34 = sub nuw nsw i64 %indvars.iv145, %tmp9
93 %tmp35 = sext i32 %add10 to i64
94 br label %for.cond14.preheader
96 for.cond14.preheader: ; preds = %for.body9
97 %arrayidx = getelementptr inbounds i16*, i16** %tmp32, i64 %tmp19
98 %tmp36 = load i16*, i16** %arrayidx, align 8
99 %arrayidx20 = getelementptr inbounds i16, i16* %tmp36, i64 %tmp35
100 %arrayidx26 = getelementptr inbounds %struct.ImageParameters, %struct.ImageParameters* %tmp33, i64 0, i32 2, i64 %indvars.iv148, i64 %indvars.iv145
101 %arrayidx35 = getelementptr inbounds [8 x [8 x i32]], [8 x [8 x i32]]* %n, i64 0, i64 %tmp18, i64 %tmp34
102 %tmp37 = bitcast i16* %arrayidx20 to <4 x i16>*
103 %tmp38 = load <4 x i16>, <4 x i16>* %tmp37, align 2
104 %tmp39 = zext <4 x i16> %tmp38 to <4 x i32>
105 %tmp40 = bitcast i16* %arrayidx26 to <4 x i16>*
106 %tmp41 = load <4 x i16>, <4 x i16>* %tmp40, align 2
107 %tmp42 = zext <4 x i16> %tmp41 to <4 x i32>
108 %tmp43 = sub nsw <4 x i32> %tmp39, %tmp42
109 %tmp44 = bitcast i32* %arrayidx35 to <4 x i32>*
110 store <4 x i32> %tmp43, <4 x i32>* %tmp44, align 16
111 store <4 x i32> %tmp43, <4 x i32>* bitcast ([6 x i32]* @d to <4 x i32>*), align 16
112 %arrayidx.1 = getelementptr inbounds i16*, i16** %tmp32, i64 %tmp20
113 %tmp45 = load i16*, i16** %arrayidx.1, align 8
114 %arrayidx20.1 = getelementptr inbounds i16, i16* %tmp45, i64 %tmp35
115 %arrayidx26.1 = getelementptr inbounds %struct.ImageParameters, %struct.ImageParameters* %tmp33, i64 0, i32 2, i64 %tmp21, i64 %indvars.iv145
116 %arrayidx35.1 = getelementptr inbounds [8 x [8 x i32]], [8 x [8 x i32]]* %n, i64 0, i64 %tmp22, i64 %tmp34
117 %tmp46 = bitcast i16* %arrayidx20.1 to <4 x i16>*
118 %tmp47 = load <4 x i16>, <4 x i16>* %tmp46, align 2
119 %tmp48 = zext <4 x i16> %tmp47 to <4 x i32>
120 %tmp49 = bitcast i16* %arrayidx26.1 to <4 x i16>*
121 %tmp50 = load <4 x i16>, <4 x i16>* %tmp49, align 2
122 %tmp51 = zext <4 x i16> %tmp50 to <4 x i32>
123 %tmp52 = sub nsw <4 x i32> %tmp48, %tmp51
124 %tmp53 = bitcast i32* %arrayidx35.1 to <4 x i32>*
125 store <4 x i32> %tmp52, <4 x i32>* %tmp53, align 16
126 store <4 x i32> %tmp52, <4 x i32>* bitcast (i32* getelementptr inbounds ([6 x i32], [6 x i32]* @d, i64 0, i64 4) to <4 x i32>*), align 16
127 %arrayidx.2 = getelementptr inbounds i16*, i16** %tmp32, i64 %tmp23
128 %tmp54 = load i16*, i16** %arrayidx.2, align 8
129 %arrayidx20.2 = getelementptr inbounds i16, i16* %tmp54, i64 %tmp35
130 %arrayidx26.2 = getelementptr inbounds %struct.ImageParameters, %struct.ImageParameters* %tmp33, i64 0, i32 2, i64 %tmp24, i64 %indvars.iv145
131 %arrayidx35.2 = getelementptr inbounds [8 x [8 x i32]], [8 x [8 x i32]]* %n, i64 0, i64 %tmp25, i64 %tmp34
132 %tmp55 = bitcast i16* %arrayidx20.2 to <4 x i16>*
133 %tmp56 = load <4 x i16>, <4 x i16>* %tmp55, align 2
134 %tmp57 = zext <4 x i16> %tmp56 to <4 x i32>
135 %tmp58 = bitcast i16* %arrayidx26.2 to <4 x i16>*
136 %tmp59 = load <4 x i16>, <4 x i16>* %tmp58, align 2
137 %tmp60 = zext <4 x i16> %tmp59 to <4 x i32>
138 %tmp61 = sub nsw <4 x i32> %tmp57, %tmp60
139 %tmp62 = bitcast i32* %arrayidx35.2 to <4 x i32>*
140 store <4 x i32> %tmp61, <4 x i32>* %tmp62, align 16
141 store <4 x i32> %tmp61, <4 x i32>* bitcast (i32* getelementptr ([6 x i32], [6 x i32]* @d, i64 1, i64 2) to <4 x i32>*), align 16
142 %arrayidx.3 = getelementptr inbounds i16*, i16** %tmp32, i64 %tmp26
143 %tmp63 = load i16*, i16** %arrayidx.3, align 8
144 %arrayidx20.3 = getelementptr inbounds i16, i16* %tmp63, i64 %tmp35
145 %arrayidx26.3 = getelementptr inbounds %struct.ImageParameters, %struct.ImageParameters* %tmp33, i64 0, i32 2, i64 %tmp27, i64 %indvars.iv145
146 %arrayidx35.3 = getelementptr inbounds [8 x [8 x i32]], [8 x [8 x i32]]* %n, i64 0, i64 %tmp28, i64 %tmp34
147 %tmp64 = bitcast i16* %arrayidx20.3 to <4 x i16>*
148 %tmp65 = load <4 x i16>, <4 x i16>* %tmp64, align 2
149 %tmp66 = zext <4 x i16> %tmp65 to <4 x i32>
150 %tmp67 = bitcast i16* %arrayidx26.3 to <4 x i16>*
151 %tmp68 = load <4 x i16>, <4 x i16>* %tmp67, align 2
152 %tmp69 = zext <4 x i16> %tmp68 to <4 x i32>
153 %tmp70 = sub nsw <4 x i32> %tmp66, %tmp69
154 %tmp71 = bitcast i32* %arrayidx35.3 to <4 x i32>*
155 store <4 x i32> %tmp70, <4 x i32>* %tmp71, align 16
156 store <4 x i32> %tmp70, <4 x i32>* bitcast (i32* getelementptr ([6 x i32], [6 x i32]* @d, i64 2, i64 0) to <4 x i32>*), align 16
157 %tmp72 = load %struct.InputParameters*, %struct.InputParameters** @b, align 8
158 %rdopt = getelementptr inbounds %struct.InputParameters, %struct.InputParameters* %tmp72, i64 0, i32 0
159 %tmp73 = load i32, i32* %rdopt, align 4
160 %cmp42 = icmp eq i32 %tmp73, 0
161 br i1 %cmp42, label %land.lhs.true, label %if.then
163 land.lhs.true: ; preds = %for.cond14.preheader
164 %Transform8x8Mode = getelementptr inbounds %struct.InputParameters, %struct.InputParameters* %tmp72, i64 0, i32 1
165 %tmp74 = load i32, i32* %Transform8x8Mode, align 4
166 %tobool = icmp eq i32 %tmp74, 0
167 br i1 %tobool, label %if.then, label %for.inc45
169 if.then: ; preds = %land.lhs.true, %for.cond14.preheader
170 %call = tail call i32 @distortion4x4(i32* nonnull getelementptr inbounds ([6 x i32], [6 x i32]* @d, i64 0, i64 0)) #3
171 %add44 = add nsw i32 %call, %m.2124
174 for.inc45: ; preds = %if.then, %land.lhs.true
175 %m.3 = phi i32 [ %m.2124, %land.lhs.true ], [ %add44, %if.then ]
176 %cmp8 = icmp slt i64 %indvars.iv145, %tmp12
177 br i1 %cmp8, label %for.inc45.for.body9_crit_edge, label %for.inc48
179 for.inc45.for.body9_crit_edge: ; preds = %for.inc45
180 %indvars.iv.next146 = add nsw i64 %indvars.iv145, 4
181 %.pre = load %struct.ImageParameters*, %struct.ImageParameters** @c, align 8
184 for.inc48: ; preds = %for.inc45
185 %m.3.lcssa = phi i32 [ %m.3, %for.inc45 ]
186 %indvars.iv.next149 = add nsw i64 %indvars.iv148, 4
187 %cmp3 = icmp slt i64 %indvars.iv148, %tmp14
188 br i1 %cmp3, label %for.body4, label %for.end50
190 for.end50: ; preds = %for.inc48
191 %m.3.lcssa.lcssa = phi i32 [ %m.3.lcssa, %for.inc48 ]
192 %tmp75 = load %struct.InputParameters*, %struct.InputParameters** @b, align 8
193 %rdopt51 = getelementptr inbounds %struct.InputParameters, %struct.InputParameters* %tmp75, i64 0, i32 0
194 %tmp76 = load i32, i32* %rdopt51, align 4
195 %cmp52 = icmp eq i32 %tmp76, 0
196 br i1 %cmp52, label %land.lhs.true54, label %for.inc73
198 land.lhs.true54: ; preds = %for.end50
199 %Transform8x8Mode55 = getelementptr inbounds %struct.InputParameters, %struct.InputParameters* %tmp75, i64 0, i32 1
200 %tmp77 = load i32, i32* %Transform8x8Mode55, align 4
201 %tobool56 = icmp eq i32 %tmp77, 0
202 br i1 %tobool56, label %for.inc73, label %for.body61.preheader
204 for.body61.preheader: ; preds = %land.lhs.true54
205 call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 nonnull bitcast (i32* getelementptr ([4 x i32], [4 x i32]* @e, i64 4, i64 0) to i8*), i8* align 16 %tmp1, i64 32, i1 false)
206 call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 nonnull bitcast (i32* getelementptr ([4 x i32], [4 x i32]* @e, i64 6, i64 0) to i8*), i8* align 16 %tmp2, i64 32, i1 false)
207 call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 nonnull bitcast (i32* getelementptr ([4 x i32], [4 x i32]* @e, i64 8, i64 0) to i8*), i8* align 16 %tmp3, i64 32, i1 false)
208 call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 nonnull bitcast (i32* getelementptr ([4 x i32], [4 x i32]* @e, i64 10, i64 0) to i8*), i8* align 16 %tmp4, i64 32, i1 false)
209 call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 nonnull bitcast (i32* getelementptr ([4 x i32], [4 x i32]* @e, i64 12, i64 0) to i8*), i8* align 16 %tmp5, i64 32, i1 false)
210 call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 nonnull bitcast (i32* getelementptr ([4 x i32], [4 x i32]* @e, i64 14, i64 0) to i8*), i8* align 16 %tmp6, i64 32, i1 false)
211 call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 nonnull bitcast (i32* getelementptr ([4 x i32], [4 x i32]* @e, i64 16, i64 0) to i8*), i8* align 16 %tmp7, i64 32, i1 false)
212 call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 nonnull bitcast (i32* getelementptr ([4 x i32], [4 x i32]* @e, i64 18, i64 0) to i8*), i8* align 16 %tmp8, i64 32, i1 false)
213 %call70 = tail call i32 @distortion4x4(i32* nonnull getelementptr inbounds ([4 x i32], [4 x i32]* @e, i64 0, i64 0)) #3
214 %add71 = add nsw i32 %call70, %m.3.lcssa.lcssa
217 for.inc73: ; preds = %for.body61.preheader, %land.lhs.true54, %for.end50
218 %m.4 = phi i32 [ %add71, %for.body61.preheader ], [ %m.3.lcssa.lcssa, %land.lhs.true54 ], [ %m.3.lcssa.lcssa, %for.end50 ]
219 %inc74 = add nuw nsw i32 %q.0131, 1
220 %exitcond156 = icmp eq i32 %inc74, 4
221 br i1 %exitcond156, label %for.end75, label %for.body
223 for.end75: ; preds = %for.inc73
224 %m.4.lcssa = phi i32 [ %m.4, %for.inc73 ]
225 call void @llvm.lifetime.end.p0i8(i64 256, i8* %tmp) #3
229 ; Function Attrs: argmemonly nounwind
230 declare void @llvm.lifetime.start.p0i8(i64, i8* nocapture) #1
232 declare void @LumaPrediction4x4(i32, i32, i32, i32, i32, i16 signext, i16 signext) #2
234 declare i32 @distortion4x4(i32*) #2
236 ; Function Attrs: argmemonly nounwind
237 declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1) #1
239 ; Function Attrs: argmemonly nounwind
240 declare void @llvm.lifetime.end.p0i8(i64, i8* nocapture) #1
242 attributes #0 = { nounwind uwtable "disable-tail-calls"="false" "less-precise-fpmad"="false" "no-frame-pointer-elim"="false" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="x86-64" "target-features"="+fxsr,+mmx,+sse,+sse2" "unsafe-fp-math"="false" "use-soft-float"="false" }
243 attributes #1 = { argmemonly nounwind }
244 attributes #2 = { "disable-tail-calls"="false" "less-precise-fpmad"="false" "no-frame-pointer-elim"="false" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="x86-64" "target-features"="+fxsr,+mmx,+sse,+sse2" "unsafe-fp-math"="false" "use-soft-float"="false" }
245 attributes #3 = { nounwind }