File: x32-va_start.ll

package info (click to toggle)
llvm-toolchain-14 1%3A14.0.6-12
  • links: PTS, VCS
  • area: main
  • in suites: bookworm
  • size: 1,496,180 kB
  • sloc: cpp: 5,593,972; ansic: 986,872; asm: 585,869; python: 184,223; objc: 72,530; lisp: 31,119; f90: 27,793; javascript: 9,780; pascal: 9,762; sh: 9,482; perl: 7,468; ml: 5,432; awk: 3,523; makefile: 2,538; xml: 953; cs: 573; fortran: 567
file content (165 lines) | stat: -rw-r--r-- 6,542 bytes parent folder | download | duplicates (6)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=x86_64-linux-gnux32 | FileCheck %s -check-prefix=SSE
; RUN: llc < %s -mtriple=x86_64-linux-gnux32 -mattr=-sse | FileCheck %s -check-prefix=NOSSE
; RUN: llc < %s -mtriple=i386-linux-gnux32 | FileCheck %s -check-prefix=32BITABI
; RUN: llc < %s -mtriple=i686-linux-gnux32 | FileCheck %s -check-prefix=32BITABI
;
; Verifies that x32 va_start lowering is sane. To regenerate this test, use
; cat <<EOF |
; #include <stdarg.h>
;
; int foo(float a, const char* fmt, ...) {
;   va_list ap;
;   va_start(ap, fmt);
;   int value = va_arg(ap, int);
;   va_end(ap);
;   return value;
; }
; EOF
; build/bin/clang -mx32 -O3 -o- -S -emit-llvm -xc -
;
target datalayout = "e-m:e-p:32:32-i64:64-f80:128-n8:16:32:64-S128"
target triple = "x86_64-unknown-linux-gnux32"

%struct.__va_list_tag = type { i32, i32, i8*, i8* }

define i32 @foo(float %a, i8* nocapture readnone %fmt, ...) nounwind {
; SSE-LABEL: foo:
; SSE:       # %bb.0: # %entry
; SSE-NEXT:    subl $72, %esp
; SSE-NEXT:    movq %rsi, -{{[0-9]+}}(%esp)
; SSE-NEXT:    movq %rdx, -{{[0-9]+}}(%esp)
; SSE-NEXT:    movq %rcx, -{{[0-9]+}}(%esp)
; SSE-NEXT:    movq %r8, -{{[0-9]+}}(%esp)
; SSE-NEXT:    movq %r9, -{{[0-9]+}}(%esp)
; SSE-NEXT:    testb %al, %al
; SSE-NEXT:    je .LBB0_5
; SSE-NEXT:  # %bb.4: # %entry
; SSE-NEXT:    movaps %xmm1, -{{[0-9]+}}(%esp)
; SSE-NEXT:    movaps %xmm2, -{{[0-9]+}}(%esp)
; SSE-NEXT:    movaps %xmm3, -{{[0-9]+}}(%esp)
; SSE-NEXT:    movaps %xmm4, (%esp)
; SSE-NEXT:    movaps %xmm5, {{[0-9]+}}(%esp)
; SSE-NEXT:    movaps %xmm6, {{[0-9]+}}(%esp)
; SSE-NEXT:    movaps %xmm7, {{[0-9]+}}(%esp)
; SSE-NEXT:  .LBB0_5: # %entry
; SSE-NEXT:    leal -{{[0-9]+}}(%rsp), %eax
; SSE-NEXT:    movl %eax, -{{[0-9]+}}(%esp)
; SSE-NEXT:    leal {{[0-9]+}}(%rsp), %eax
; SSE-NEXT:    movl %eax, -{{[0-9]+}}(%esp)
; SSE-NEXT:    movabsq $274877906952, %rax # imm = 0x4000000008
; SSE-NEXT:    movq %rax, -{{[0-9]+}}(%esp)
; SSE-NEXT:    movl $8, %ecx
; SSE-NEXT:    cmpl $40, %ecx
; SSE-NEXT:    ja .LBB0_2
; SSE-NEXT:  # %bb.1: # %vaarg.in_reg
; SSE-NEXT:    movl -{{[0-9]+}}(%esp), %eax
; SSE-NEXT:    addl %ecx, %eax
; SSE-NEXT:    addl $8, %ecx
; SSE-NEXT:    movl %ecx, -{{[0-9]+}}(%esp)
; SSE-NEXT:    jmp .LBB0_3
; SSE-NEXT:  .LBB0_2: # %vaarg.in_mem
; SSE-NEXT:    movl -{{[0-9]+}}(%esp), %eax
; SSE-NEXT:    leal 8(%rax), %ecx
; SSE-NEXT:    movl %ecx, -{{[0-9]+}}(%esp)
; SSE-NEXT:  .LBB0_3: # %vaarg.end
; SSE-NEXT:    movl (%eax), %eax
; SSE-NEXT:    addl $72, %esp
; SSE-NEXT:    retq
;
; NOSSE-LABEL: foo:
; NOSSE:       # %bb.0: # %entry
; NOSSE-NEXT:    movq %rsi, -{{[0-9]+}}(%esp)
; NOSSE-NEXT:    movq %rdx, -{{[0-9]+}}(%esp)
; NOSSE-NEXT:    movq %rcx, -{{[0-9]+}}(%esp)
; NOSSE-NEXT:    movq %r8, -{{[0-9]+}}(%esp)
; NOSSE-NEXT:    movq %r9, -{{[0-9]+}}(%esp)
; NOSSE-NEXT:    leal -{{[0-9]+}}(%rsp), %eax
; NOSSE-NEXT:    movl %eax, -{{[0-9]+}}(%esp)
; NOSSE-NEXT:    leal {{[0-9]+}}(%rsp), %eax
; NOSSE-NEXT:    movl %eax, -{{[0-9]+}}(%esp)
; NOSSE-NEXT:    movabsq $206158430216, %rax # imm = 0x3000000008
; NOSSE-NEXT:    movq %rax, -{{[0-9]+}}(%esp)
; NOSSE-NEXT:    movl $8, %ecx
; NOSSE-NEXT:    cmpl $40, %ecx
; NOSSE-NEXT:    ja .LBB0_2
; NOSSE-NEXT:  # %bb.1: # %vaarg.in_reg
; NOSSE-NEXT:    movl -{{[0-9]+}}(%esp), %eax
; NOSSE-NEXT:    addl %ecx, %eax
; NOSSE-NEXT:    addl $8, %ecx
; NOSSE-NEXT:    movl %ecx, -{{[0-9]+}}(%esp)
; NOSSE-NEXT:    movl (%eax), %eax
; NOSSE-NEXT:    retq
; NOSSE-NEXT:  .LBB0_2: # %vaarg.in_mem
; NOSSE-NEXT:    movl -{{[0-9]+}}(%esp), %eax
; NOSSE-NEXT:    leal 8(%rax), %ecx
; NOSSE-NEXT:    movl %ecx, -{{[0-9]+}}(%esp)
; NOSSE-NEXT:    movl (%eax), %eax
; NOSSE-NEXT:    retq
;
; 32BITABI-LABEL: foo:
; 32BITABI:       # %bb.0: # %entry
; 32BITABI-NEXT:    subl $28, %esp
; 32BITABI-NEXT:    leal {{[0-9]+}}(%esp), %ecx
; 32BITABI-NEXT:    movl %ecx, (%esp)
; 32BITABI-NEXT:    cmpl $40, %ecx
; 32BITABI-NEXT:    ja .LBB0_2
; 32BITABI-NEXT:  # %bb.1: # %vaarg.in_reg
; 32BITABI-NEXT:    movl {{[0-9]+}}(%esp), %eax
; 32BITABI-NEXT:    addl %ecx, %eax
; 32BITABI-NEXT:    addl $8, %ecx
; 32BITABI-NEXT:    movl %ecx, (%esp)
; 32BITABI-NEXT:    jmp .LBB0_3
; 32BITABI-NEXT:  .LBB0_2: # %vaarg.in_mem
; 32BITABI-NEXT:    movl {{[0-9]+}}(%esp), %eax
; 32BITABI-NEXT:    leal 8(%eax), %ecx
; 32BITABI-NEXT:    movl %ecx, {{[0-9]+}}(%esp)
; 32BITABI-NEXT:  .LBB0_3: # %vaarg.end
; 32BITABI-NEXT:    movl (%eax), %eax
; 32BITABI-NEXT:    addl $28, %esp
; 32BITABI-NEXT:    retl
entry:
  %ap = alloca [1 x %struct.__va_list_tag], align 16
  %0 = bitcast [1 x %struct.__va_list_tag]* %ap to i8*
  call void @llvm.lifetime.start.p0i8(i64 16, i8* %0) #2
  call void @llvm.va_start(i8* %0)
  %gp_offset_p = getelementptr inbounds [1 x %struct.__va_list_tag], [1 x %struct.__va_list_tag]* %ap, i32 0, i32 0, i32 0
  %gp_offset = load i32, i32* %gp_offset_p, align 16
  %fits_in_gp = icmp ult i32 %gp_offset, 41
  br i1 %fits_in_gp, label %vaarg.in_reg, label %vaarg.in_mem

vaarg.in_reg:                                     ; preds = %entry
  %1 = getelementptr inbounds [1 x %struct.__va_list_tag], [1 x %struct.__va_list_tag]* %ap, i32 0, i32 0, i32 3
  %reg_save_area = load i8*, i8** %1, align 4
  %2 = getelementptr i8, i8* %reg_save_area, i32 %gp_offset
  %3 = add i32 %gp_offset, 8
  store i32 %3, i32* %gp_offset_p, align 16
  br label %vaarg.end
vaarg.in_mem:                                     ; preds = %entry
  %overflow_arg_area_p = getelementptr inbounds [1 x %struct.__va_list_tag], [1 x %struct.__va_list_tag]* %ap, i32 0, i32 0, i32 2
  %overflow_arg_area = load i8*, i8** %overflow_arg_area_p, align 8
  %overflow_arg_area.next = getelementptr i8, i8* %overflow_arg_area, i32 8
  store i8* %overflow_arg_area.next, i8** %overflow_arg_area_p, align 8
  br label %vaarg.end

vaarg.end:                                        ; preds = %vaarg.in_mem, %vaarg.in_reg
  %vaarg.addr.in = phi i8* [ %2, %vaarg.in_reg ], [ %overflow_arg_area, %vaarg.in_mem ]
  %vaarg.addr = bitcast i8* %vaarg.addr.in to i32*
  %4 = load i32, i32* %vaarg.addr, align 4
  call void @llvm.va_end(i8* %0)
  call void @llvm.lifetime.end.p0i8(i64 16, i8* %0) #2
  ret i32 %4
}

; Function Attrs: nounwind argmemonly
declare void @llvm.lifetime.start.p0i8(i64, i8* nocapture) nounwind

; Function Attrs: nounwind
declare void @llvm.va_start(i8*) nounwind

; Function Attrs: nounwind
declare void @llvm.va_end(i8*) nounwind

; Function Attrs: nounwind argmemonly
declare void @llvm.lifetime.end.p0i8(i64, i8* nocapture) nounwind