| blob | 5aeaca7e84c3b0d23727094b1e2183e3c9d694a7 |
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 ptr null, align 8
10 @a = common global ptr null, align 8
11 @d = common global [6 x i32] zeroinitializer, align 16
12 @b = common global ptr 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).
18 ;
19 ; CHECK-LABEL: @fn
20 ; CHECK: subq {{\$3[0-4][0-9]}}, %rsp
22 ; Function Attrs: nounwind uwtable
23 define i32 @fn() #0 {
24 entry:
25 %n = alloca [8 x [8 x i32]], align 16
26 call void @llvm.lifetime.start.p0(i64 256, ptr %n) #3
27 %arraydecay.1 = getelementptr inbounds [8 x [8 x i32]], ptr %n, i64 0, i64 1, i64 0
28 %arraydecay.2 = getelementptr inbounds [8 x [8 x i32]], ptr %n, i64 0, i64 2, i64 0
29 %arraydecay.3 = getelementptr inbounds [8 x [8 x i32]], ptr %n, i64 0, i64 3, i64 0
30 %arraydecay.4 = getelementptr inbounds [8 x [8 x i32]], ptr %n, i64 0, i64 4, i64 0
31 %arraydecay.5 = getelementptr inbounds [8 x [8 x i32]], ptr %n, i64 0, i64 5, i64 0
32 %arraydecay.6 = getelementptr inbounds [8 x [8 x i32]], ptr %n, i64 0, i64 6, i64 0
33 %arraydecay.7 = getelementptr inbounds [8 x [8 x i32]], ptr %n, i64 0, i64 7, i64 0
34 br label %for.body
36 for.body: ; preds = %for.inc73, %entry
37 %q.0131 = phi i32 [ 0, %entry ], [ %inc74, %for.inc73 ]
38 %m.0130 = phi i32 [ 0, %entry ], [ %m.4, %for.inc73 ]
39 %div = sdiv i32 %q.0131, 2
40 %shl = shl i32 %div, 3
41 %rem = srem i32 %q.0131, 2
42 %shl1 = shl nsw i32 %rem, 3
43 %tmp9 = sext i32 %shl1 to i64
44 %tmp10 = sext i32 %shl to i64
45 %tmp11 = or i32 %shl1, 4
46 %tmp12 = sext i32 %tmp11 to i64
47 %tmp13 = or i32 %shl, 4
48 %tmp14 = sext i32 %tmp13 to i64
49 br label %for.body4
51 for.body4: ; preds = %for.inc48, %for.body
52 %indvars.iv148 = phi i64 [ %tmp10, %for.body ], [ %indvars.iv.next149, %for.inc48 ]
53 %m.1126 = phi i32 [ %m.0130, %for.body ], [ %m.3.lcssa, %for.inc48 ]
54 %tmp15 = load ptr, ptr @c, align 8
55 %opix_y = getelementptr inbounds %struct.ImageParameters, ptr %tmp15, i64 0, i32 1
56 %tmp16 = load i32, ptr %opix_y, align 4
57 %tmp17 = trunc i64 %indvars.iv148 to i32
58 %add5 = add nsw i32 %tmp16, %tmp17
59 %tmp18 = sub nuw nsw i64 %indvars.iv148, %tmp10
60 %tmp19 = sext i32 %add5 to i64
61 %tmp20 = add nsw i64 %tmp19, 1
62 %tmp21 = or i64 %indvars.iv148, 1
63 %tmp22 = or i64 %tmp18, 1
64 %tmp23 = add nsw i64 %tmp19, 2
65 %tmp24 = or i64 %indvars.iv148, 2
66 %tmp25 = or i64 %tmp18, 2
67 %tmp26 = add nsw i64 %tmp19, 3
68 %tmp27 = or i64 %indvars.iv148, 3
69 %tmp28 = or i64 %tmp18, 3
70 br label %for.body9
72 for.body9: ; preds = %for.inc45.for.body9_crit_edge, %for.body4
73 %tmp29 = phi ptr [ %tmp15, %for.body4 ], [ %.pre, %for.inc45.for.body9_crit_edge ]
74 %indvars.iv145 = phi i64 [ %tmp9, %for.body4 ], [ %indvars.iv.next146, %for.inc45.for.body9_crit_edge ]
75 %m.2124 = phi i32 [ %m.1126, %for.body4 ], [ %m.3, %for.inc45.for.body9_crit_edge ]
76 %tmp30 = load i32, ptr %tmp29, align 4
77 %tmp31 = trunc i64 %indvars.iv145 to i32
78 %add10 = add nsw i32 %tmp30, %tmp31
79 tail call void @LumaPrediction4x4(i32 %tmp31, i32 %tmp17, i32 0, i32 0, i32 0, i16 signext 0, i16 signext 0) #3
80 %tmp32 = load ptr, ptr @a, align 8
81 %tmp33 = load ptr, ptr @c, align 8
82 %tmp34 = sub nuw nsw i64 %indvars.iv145, %tmp9
83 %tmp35 = sext i32 %add10 to i64
84 br label %for.cond14.preheader
86 for.cond14.preheader: ; preds = %for.body9
87 %arrayidx = getelementptr inbounds ptr, ptr %tmp32, i64 %tmp19
88 %tmp36 = load ptr, ptr %arrayidx, align 8
89 %arrayidx20 = getelementptr inbounds i16, ptr %tmp36, i64 %tmp35
90 %arrayidx26 = getelementptr inbounds %struct.ImageParameters, ptr %tmp33, i64 0, i32 2, i64 %indvars.iv148, i64 %indvars.iv145
91 %arrayidx35 = getelementptr inbounds [8 x [8 x i32]], ptr %n, i64 0, i64 %tmp18, i64 %tmp34
92 %tmp38 = load <4 x i16>, ptr %arrayidx20, align 2
93 %tmp39 = zext <4 x i16> %tmp38 to <4 x i32>
94 %tmp41 = load <4 x i16>, ptr %arrayidx26, align 2
95 %tmp42 = zext <4 x i16> %tmp41 to <4 x i32>
96 %tmp43 = sub nsw <4 x i32> %tmp39, %tmp42
97 store <4 x i32> %tmp43, ptr %arrayidx35, align 16
98 store <4 x i32> %tmp43, ptr @d, align 16
99 %arrayidx.1 = getelementptr inbounds ptr, ptr %tmp32, i64 %tmp20
100 %tmp45 = load ptr, ptr %arrayidx.1, align 8
101 %arrayidx20.1 = getelementptr inbounds i16, ptr %tmp45, i64 %tmp35
102 %arrayidx26.1 = getelementptr inbounds %struct.ImageParameters, ptr %tmp33, i64 0, i32 2, i64 %tmp21, i64 %indvars.iv145
103 %arrayidx35.1 = getelementptr inbounds [8 x [8 x i32]], ptr %n, i64 0, i64 %tmp22, i64 %tmp34
104 %tmp47 = load <4 x i16>, ptr %arrayidx20.1, align 2
105 %tmp48 = zext <4 x i16> %tmp47 to <4 x i32>
106 %tmp50 = load <4 x i16>, ptr %arrayidx26.1, align 2
107 %tmp51 = zext <4 x i16> %tmp50 to <4 x i32>
108 %tmp52 = sub nsw <4 x i32> %tmp48, %tmp51
109 store <4 x i32> %tmp52, ptr %arrayidx35.1, align 16
110 store <4 x i32> %tmp52, ptr getelementptr inbounds ([6 x i32], ptr @d, i64 0, i64 4), align 16
111 %arrayidx.2 = getelementptr inbounds ptr, ptr %tmp32, i64 %tmp23
112 %tmp54 = load ptr, ptr %arrayidx.2, align 8
113 %arrayidx20.2 = getelementptr inbounds i16, ptr %tmp54, i64 %tmp35
114 %arrayidx26.2 = getelementptr inbounds %struct.ImageParameters, ptr %tmp33, i64 0, i32 2, i64 %tmp24, i64 %indvars.iv145
115 %arrayidx35.2 = getelementptr inbounds [8 x [8 x i32]], ptr %n, i64 0, i64 %tmp25, i64 %tmp34
116 %tmp56 = load <4 x i16>, ptr %arrayidx20.2, align 2
117 %tmp57 = zext <4 x i16> %tmp56 to <4 x i32>
118 %tmp59 = load <4 x i16>, ptr %arrayidx26.2, align 2
119 %tmp60 = zext <4 x i16> %tmp59 to <4 x i32>
120 %tmp61 = sub nsw <4 x i32> %tmp57, %tmp60
121 store <4 x i32> %tmp61, ptr %arrayidx35.2, align 16
122 store <4 x i32> %tmp61, ptr getelementptr ([6 x i32], ptr @d, i64 1, i64 2), align 16
123 %arrayidx.3 = getelementptr inbounds ptr, ptr %tmp32, i64 %tmp26
124 %tmp63 = load ptr, ptr %arrayidx.3, align 8
125 %arrayidx20.3 = getelementptr inbounds i16, ptr %tmp63, i64 %tmp35
126 %arrayidx26.3 = getelementptr inbounds %struct.ImageParameters, ptr %tmp33, i64 0, i32 2, i64 %tmp27, i64 %indvars.iv145
127 %arrayidx35.3 = getelementptr inbounds [8 x [8 x i32]], ptr %n, i64 0, i64 %tmp28, i64 %tmp34
128 %tmp65 = load <4 x i16>, ptr %arrayidx20.3, align 2
129 %tmp66 = zext <4 x i16> %tmp65 to <4 x i32>
130 %tmp68 = load <4 x i16>, ptr %arrayidx26.3, align 2
131 %tmp69 = zext <4 x i16> %tmp68 to <4 x i32>
132 %tmp70 = sub nsw <4 x i32> %tmp66, %tmp69
133 store <4 x i32> %tmp70, ptr %arrayidx35.3, align 16
134 store <4 x i32> %tmp70, ptr getelementptr ([6 x i32], ptr @d, i64 2, i64 0), align 16
135 %tmp72 = load ptr, ptr @b, align 8
136 %tmp73 = load i32, ptr %tmp72, align 4
137 %cmp42 = icmp eq i32 %tmp73, 0
138 br i1 %cmp42, label %land.lhs.true, label %if.then
140 land.lhs.true: ; preds = %for.cond14.preheader
141 %Transform8x8Mode = getelementptr inbounds %struct.InputParameters, ptr %tmp72, i64 0, i32 1
142 %tmp74 = load i32, ptr %Transform8x8Mode, align 4
143 %tobool = icmp eq i32 %tmp74, 0
144 br i1 %tobool, label %if.then, label %for.inc45
146 if.then: ; preds = %land.lhs.true, %for.cond14.preheader
147 %call = tail call i32 @distortion4x4(ptr nonnull @d) #3
148 %add44 = add nsw i32 %call, %m.2124
149 br label %for.inc45
151 for.inc45: ; preds = %if.then, %land.lhs.true
152 %m.3 = phi i32 [ %m.2124, %land.lhs.true ], [ %add44, %if.then ]
153 %cmp8 = icmp slt i64 %indvars.iv145, %tmp12
154 br i1 %cmp8, label %for.inc45.for.body9_crit_edge, label %for.inc48
156 for.inc45.for.body9_crit_edge: ; preds = %for.inc45
157 %indvars.iv.next146 = add nsw i64 %indvars.iv145, 4
158 %.pre = load ptr, ptr @c, align 8
159 br label %for.body9
161 for.inc48: ; preds = %for.inc45
162 %m.3.lcssa = phi i32 [ %m.3, %for.inc45 ]
163 %indvars.iv.next149 = add nsw i64 %indvars.iv148, 4
164 %cmp3 = icmp slt i64 %indvars.iv148, %tmp14
165 br i1 %cmp3, label %for.body4, label %for.end50
167 for.end50: ; preds = %for.inc48
168 %m.3.lcssa.lcssa = phi i32 [ %m.3.lcssa, %for.inc48 ]
169 %tmp75 = load ptr, ptr @b, align 8
170 %tmp76 = load i32, ptr %tmp75, align 4
171 %cmp52 = icmp eq i32 %tmp76, 0
172 br i1 %cmp52, label %land.lhs.true54, label %for.inc73
174 land.lhs.true54: ; preds = %for.end50
175 %Transform8x8Mode55 = getelementptr inbounds %struct.InputParameters, ptr %tmp75, i64 0, i32 1
176 %tmp77 = load i32, ptr %Transform8x8Mode55, align 4
177 %tobool56 = icmp eq i32 %tmp77, 0
178 br i1 %tobool56, label %for.inc73, label %for.body61.preheader
180 for.body61.preheader: ; preds = %land.lhs.true54
181 call void @llvm.memcpy.p0.p0.i64(ptr align 16 nonnull getelementptr ([4 x i32], ptr @e, i64 4, i64 0), ptr align 16 %n, i64 32, i1 false)
182 call void @llvm.memcpy.p0.p0.i64(ptr align 16 nonnull getelementptr ([4 x i32], ptr @e, i64 6, i64 0), ptr align 16 %arraydecay.1, i64 32, i1 false)
183 call void @llvm.memcpy.p0.p0.i64(ptr align 16 nonnull getelementptr ([4 x i32], ptr @e, i64 8, i64 0), ptr align 16 %arraydecay.2, i64 32, i1 false)
184 call void @llvm.memcpy.p0.p0.i64(ptr align 16 nonnull getelementptr ([4 x i32], ptr @e, i64 10, i64 0), ptr align 16 %arraydecay.3, i64 32, i1 false)
185 call void @llvm.memcpy.p0.p0.i64(ptr align 16 nonnull getelementptr ([4 x i32], ptr @e, i64 12, i64 0), ptr align 16 %arraydecay.4, i64 32, i1 false)
186 call void @llvm.memcpy.p0.p0.i64(ptr align 16 nonnull getelementptr ([4 x i32], ptr @e, i64 14, i64 0), ptr align 16 %arraydecay.5, i64 32, i1 false)
187 call void @llvm.memcpy.p0.p0.i64(ptr align 16 nonnull getelementptr ([4 x i32], ptr @e, i64 16, i64 0), ptr align 16 %arraydecay.6, i64 32, i1 false)
188 call void @llvm.memcpy.p0.p0.i64(ptr align 16 nonnull getelementptr ([4 x i32], ptr @e, i64 18, i64 0), ptr align 16 %arraydecay.7, i64 32, i1 false)
189 %call70 = tail call i32 @distortion4x4(ptr nonnull @e) #3
190 %add71 = add nsw i32 %call70, %m.3.lcssa.lcssa
191 br label %for.inc73
193 for.inc73: ; preds = %for.body61.preheader, %land.lhs.true54, %for.end50
194 %m.4 = phi i32 [ %add71, %for.body61.preheader ], [ %m.3.lcssa.lcssa, %land.lhs.true54 ], [ %m.3.lcssa.lcssa, %for.end50 ]
195 %inc74 = add nuw nsw i32 %q.0131, 1
196 %exitcond156 = icmp eq i32 %inc74, 4
197 br i1 %exitcond156, label %for.end75, label %for.body
199 for.end75: ; preds = %for.inc73
200 %m.4.lcssa = phi i32 [ %m.4, %for.inc73 ]
201 call void @llvm.lifetime.end.p0(i64 256, ptr %n) #3
202 ret i32 %m.4.lcssa
203 }
205 ; Function Attrs: argmemonly nounwind
206 declare void @llvm.lifetime.start.p0(i64, ptr nocapture) #1
208 declare void @LumaPrediction4x4(i32, i32, i32, i32, i32, i16 signext, i16 signext) #2
210 declare i32 @distortion4x4(ptr) #2
212 ; Function Attrs: argmemonly nounwind
213 declare void @llvm.memcpy.p0.p0.i64(ptr nocapture, ptr nocapture readonly, i64, i1) #1
215 ; Function Attrs: argmemonly nounwind
216 declare void @llvm.lifetime.end.p0(i64, ptr nocapture) #1
218 attributes #0 = { nounwind uwtable "disable-tail-calls"="false" "less-precise-fpmad"="false" "frame-pointer"="none" "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" }
219 attributes #1 = { argmemonly nounwind }
220 attributes #2 = { "disable-tail-calls"="false" "less-precise-fpmad"="false" "frame-pointer"="none" "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" }
221 attributes #3 = { nounwind }