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
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
|
// Copyright 2011 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build darwin || (openbsd && !mips64)
package syscall
import (
"internal/abi"
"runtime"
"unsafe"
)
type SysProcAttr struct {
Chroot string // Chroot.
Credential *Credential // Credential.
Ptrace bool // Enable tracing.
Setsid bool // Create session.
// Setpgid sets the process group ID of the child to Pgid,
// or, if Pgid == 0, to the new child's process ID.
Setpgid bool
// Setctty sets the controlling terminal of the child to
// file descriptor Ctty. Ctty must be a descriptor number
// in the child process: an index into ProcAttr.Files.
// This is only meaningful if Setsid is true.
Setctty bool
Noctty bool // Detach fd 0 from controlling terminal
Ctty int // Controlling TTY fd
// Foreground places the child process group in the foreground.
// This implies Setpgid. The Ctty field must be set to
// the descriptor of the controlling TTY.
// Unlike Setctty, in this case Ctty must be a descriptor
// number in the parent process.
Foreground bool
Pgid int // Child's process group ID if Setpgid.
}
// Implemented in runtime package.
func runtime_BeforeFork()
func runtime_AfterFork()
func runtime_AfterForkInChild()
// Fork, dup fd onto 0..len(fd), and exec(argv0, argvv, envv) in child.
// If a dup or exec fails, write the errno error to pipe.
// (Pipe is close-on-exec so if exec succeeds, it will be closed.)
// In the child, this function must not acquire any locks, because
// they might have been locked at the time of the fork. This means
// no rescheduling, no malloc calls, and no new stack segments.
// For the same reason compiler does not race instrument it.
// The calls to rawSyscall are okay because they are assembly
// functions that do not grow the stack.
//
//go:norace
func forkAndExecInChild(argv0 *byte, argv, envv []*byte, chroot, dir *byte, attr *ProcAttr, sys *SysProcAttr, pipe int) (pid int, err1 Errno) {
// Declare all variables at top in case any
// declarations require heap allocation (e.g., err1).
var (
r1 uintptr
nextfd int
i int
err error
pgrp _C_int
cred *Credential
ngroups, groups uintptr
)
rlim, rlimOK := origRlimitNofile.Load().(Rlimit)
// guard against side effects of shuffling fds below.
// Make sure that nextfd is beyond any currently open files so
// that we can't run the risk of overwriting any of them.
fd := make([]int, len(attr.Files))
nextfd = len(attr.Files)
for i, ufd := range attr.Files {
if nextfd < int(ufd) {
nextfd = int(ufd)
}
fd[i] = int(ufd)
}
nextfd++
// About to call fork.
// No more allocation or calls of non-assembly functions.
runtime_BeforeFork()
r1, _, err1 = rawSyscall(abi.FuncPCABI0(libc_fork_trampoline), 0, 0, 0)
if err1 != 0 {
runtime_AfterFork()
return 0, err1
}
if r1 != 0 {
// parent; return PID
runtime_AfterFork()
return int(r1), 0
}
// Fork succeeded, now in child.
// Enable tracing if requested.
if sys.Ptrace {
if err = ptrace(PTRACE_TRACEME, 0, 0, 0); err != nil {
err1 = err.(Errno)
goto childerror
}
}
// Session ID
if sys.Setsid {
_, _, err1 = rawSyscall(abi.FuncPCABI0(libc_setsid_trampoline), 0, 0, 0)
if err1 != 0 {
goto childerror
}
}
// Set process group
if sys.Setpgid || sys.Foreground {
// Place child in process group.
_, _, err1 = rawSyscall(abi.FuncPCABI0(libc_setpgid_trampoline), 0, uintptr(sys.Pgid), 0)
if err1 != 0 {
goto childerror
}
}
if sys.Foreground {
// This should really be pid_t, however _C_int (aka int32) is
// generally equivalent.
pgrp = _C_int(sys.Pgid)
if pgrp == 0 {
r1, _, err1 = rawSyscall(abi.FuncPCABI0(libc_getpid_trampoline), 0, 0, 0)
if err1 != 0 {
goto childerror
}
pgrp = _C_int(r1)
}
// Place process group in foreground.
_, _, err1 = rawSyscall(abi.FuncPCABI0(libc_ioctl_trampoline), uintptr(sys.Ctty), uintptr(TIOCSPGRP), uintptr(unsafe.Pointer(&pgrp)))
if err1 != 0 {
goto childerror
}
}
// Restore the signal mask. We do this after TIOCSPGRP to avoid
// having the kernel send a SIGTTOU signal to the process group.
runtime_AfterForkInChild()
// Chroot
if chroot != nil {
_, _, err1 = rawSyscall(abi.FuncPCABI0(libc_chroot_trampoline), uintptr(unsafe.Pointer(chroot)), 0, 0)
if err1 != 0 {
goto childerror
}
}
// User and groups
if cred = sys.Credential; cred != nil {
ngroups = uintptr(len(cred.Groups))
groups = uintptr(0)
if ngroups > 0 {
groups = uintptr(unsafe.Pointer(&cred.Groups[0]))
}
if !cred.NoSetGroups {
_, _, err1 = rawSyscall(abi.FuncPCABI0(libc_setgroups_trampoline), ngroups, groups, 0)
if err1 != 0 {
goto childerror
}
}
_, _, err1 = rawSyscall(abi.FuncPCABI0(libc_setgid_trampoline), uintptr(cred.Gid), 0, 0)
if err1 != 0 {
goto childerror
}
_, _, err1 = rawSyscall(abi.FuncPCABI0(libc_setuid_trampoline), uintptr(cred.Uid), 0, 0)
if err1 != 0 {
goto childerror
}
}
// Chdir
if dir != nil {
_, _, err1 = rawSyscall(abi.FuncPCABI0(libc_chdir_trampoline), uintptr(unsafe.Pointer(dir)), 0, 0)
if err1 != 0 {
goto childerror
}
}
// Pass 1: look for fd[i] < i and move those up above len(fd)
// so that pass 2 won't stomp on an fd it needs later.
if pipe < nextfd {
if runtime.GOOS == "openbsd" {
_, _, err1 = rawSyscall(dupTrampoline, uintptr(pipe), uintptr(nextfd), O_CLOEXEC)
} else {
_, _, err1 = rawSyscall(dupTrampoline, uintptr(pipe), uintptr(nextfd), 0)
if err1 != 0 {
goto childerror
}
_, _, err1 = rawSyscall(abi.FuncPCABI0(libc_fcntl_trampoline), uintptr(nextfd), F_SETFD, FD_CLOEXEC)
}
if err1 != 0 {
goto childerror
}
pipe = nextfd
nextfd++
}
for i = 0; i < len(fd); i++ {
if fd[i] >= 0 && fd[i] < i {
if nextfd == pipe { // don't stomp on pipe
nextfd++
}
if runtime.GOOS == "openbsd" {
_, _, err1 = rawSyscall(dupTrampoline, uintptr(fd[i]), uintptr(nextfd), O_CLOEXEC)
} else {
_, _, err1 = rawSyscall(dupTrampoline, uintptr(fd[i]), uintptr(nextfd), 0)
if err1 != 0 {
goto childerror
}
_, _, err1 = rawSyscall(abi.FuncPCABI0(libc_fcntl_trampoline), uintptr(nextfd), F_SETFD, FD_CLOEXEC)
}
if err1 != 0 {
goto childerror
}
fd[i] = nextfd
nextfd++
}
}
// Pass 2: dup fd[i] down onto i.
for i = 0; i < len(fd); i++ {
if fd[i] == -1 {
rawSyscall(abi.FuncPCABI0(libc_close_trampoline), uintptr(i), 0, 0)
continue
}
if fd[i] == i {
// dup2(i, i) won't clear close-on-exec flag on Linux,
// probably not elsewhere either.
_, _, err1 = rawSyscall(abi.FuncPCABI0(libc_fcntl_trampoline), uintptr(fd[i]), F_SETFD, 0)
if err1 != 0 {
goto childerror
}
continue
}
// The new fd is created NOT close-on-exec,
// which is exactly what we want.
_, _, err1 = rawSyscall(abi.FuncPCABI0(libc_dup2_trampoline), uintptr(fd[i]), uintptr(i), 0)
if err1 != 0 {
goto childerror
}
}
// By convention, we don't close-on-exec the fds we are
// started with, so if len(fd) < 3, close 0, 1, 2 as needed.
// Programs that know they inherit fds >= 3 will need
// to set them close-on-exec.
for i = len(fd); i < 3; i++ {
rawSyscall(abi.FuncPCABI0(libc_close_trampoline), uintptr(i), 0, 0)
}
// Detach fd 0 from tty
if sys.Noctty {
_, _, err1 = rawSyscall(abi.FuncPCABI0(libc_ioctl_trampoline), 0, uintptr(TIOCNOTTY), 0)
if err1 != 0 {
goto childerror
}
}
// Set the controlling TTY to Ctty
if sys.Setctty {
_, _, err1 = rawSyscall(abi.FuncPCABI0(libc_ioctl_trampoline), uintptr(sys.Ctty), uintptr(TIOCSCTTY), 0)
if err1 != 0 {
goto childerror
}
}
// Restore original rlimit.
if rlimOK && rlim.Cur != 0 {
rawSyscall(abi.FuncPCABI0(libc_setrlimit_trampoline), uintptr(RLIMIT_NOFILE), uintptr(unsafe.Pointer(&rlim)), 0)
}
// Time to exec.
_, _, err1 = rawSyscall(abi.FuncPCABI0(libc_execve_trampoline),
uintptr(unsafe.Pointer(argv0)),
uintptr(unsafe.Pointer(&argv[0])),
uintptr(unsafe.Pointer(&envv[0])))
childerror:
// send error code on pipe
rawSyscall(abi.FuncPCABI0(libc_write_trampoline), uintptr(pipe), uintptr(unsafe.Pointer(&err1)), unsafe.Sizeof(err1))
for {
rawSyscall(abi.FuncPCABI0(libc_exit_trampoline), 253, 0, 0)
}
}
|
