diff options
author | Richard Lowe <richlowe@richlowe.net> | 2011-03-14 14:05:30 -0400 |
---|---|---|
committer | Richard Lowe <richlowe@richlowe.net> | 2011-03-14 14:05:30 -0400 |
commit | c10c16dec587a0662068f6e2991c29ed3a9db943 (patch) | |
tree | f414286f4bba41d75683ed4fbbaa6bfa4bf7fabd /usr/src/man/man2/vfork.2 | |
parent | 68caef18a23a498d9e3017b983562c0f4fd8ab23 (diff) | |
download | illumos-joyent-c10c16dec587a0662068f6e2991c29ed3a9db943.tar.gz |
243 system manual pages should live with the software
Reviewed by: garrett@nexenta.com
Reviewed by: gwr@nexenta.com
Reviewed by: trisk@opensolaris.org
Approved by: gwr@nexenta.com
--HG--
extra : rebase_source : 0c599d0bec0dc8865fbba67721a7a6cd6b1feefb
Diffstat (limited to 'usr/src/man/man2/vfork.2')
-rw-r--r-- | usr/src/man/man2/vfork.2 | 186 |
1 files changed, 186 insertions, 0 deletions
diff --git a/usr/src/man/man2/vfork.2 b/usr/src/man/man2/vfork.2 new file mode 100644 index 0000000000..c93b666fc5 --- /dev/null +++ b/usr/src/man/man2/vfork.2 @@ -0,0 +1,186 @@ +'\" te +.\" Copyright (c) 2004, Sun Microsystems, Inc. All Rights Reserved. +.\" Copyright 1989 AT&T. +.\" Copyright (c) 1980 Regents of the University of California. All rights reserved. The Berkeley software License Agreement specifies the terms and conditions for redistribution. +.TH vfork 2 "13 Dec 2006" "SunOS 5.11" "System Calls" +.SH NAME +vfork, vforkx \- spawn new process in a virtual memory efficient way +.SH SYNOPSIS +.LP +.nf +#include <unistd.h> + +\fBpid_t\fR \fBvfork\fR(\fBvoid\fR); +.fi + +.LP +.nf +#include <sys/fork.h> + +\fBpid_t\fR \fBvforkx\fR(\fBint\fR \fIflags\fR); +.fi + +.SH DESCRIPTION +.sp +.LP +The \fBvfork()\fR and \fBvforkx()\fR functions create a new process without +fully copying the address space of the old process. These functions are useful +in instances where the purpose of a \fBfork\fR(2) operation is to create a new +system context for an \fBexecve()\fR operation (see \fBexec\fR(2)). +.sp +.LP +Unlike with the \fBfork()\fR function, the child process borrows the parent's +memory and thread of control until a call to \fBexecve()\fR or an exit (either +abnormally or by a call to \fB_exit()\fR (see \fBexit\fR(2)). Any modification +made during this time to any part of memory in the child process is reflected +in the parent process on return from \fBvfork()\fR or \fBvforkx()\fR. The +parent process is suspended while the child is using its resources. +.sp +.LP +In a multithreaded application, \fBvfork()\fR and \fBvforkx()\fR borrow only +the thread of control that called \fBvfork()\fR or \fBvforkx()\fR in the +parent; that is, the child contains only one thread. The use of \fBvfork()\fR +or \fBvforkx()\fR in multithreaded applications, however, is unsafe due to race +conditions that can cause the child process to become deadlocked and +consequently block both the child and parent process from execution +indefinitely. +.sp +.LP +The \fBvfork()\fR and \fBvforkx()\fR functions can normally be used the same +way as \fBfork()\fR and \fBforkx()\fR, respectively. The calling procedure, +however, should not return while running in the child's context, since the +eventual return from \fBvfork()\fR or \fBvforkx()\fR in the parent would be to +a stack frame that no longer exists. The \fB_exit()\fR function should be used +in favor of \fBexit\fR(3C) if unable to perform an \fBexecve()\fR operation, +since \fBexit()\fR will invoke all functions registered by \fBatexit\fR(3C) and +will flush and close standard I/O channels, thereby corrupting the parent +process's standard I/O data structures. Care must be taken in the child process +not to modify any global or local data that affects the behavior of the parent +process on return from \fBvfork()\fR or \fBvforkx()\fR, unless such an effect +is intentional. +.sp +.LP +Unlike \fBfork()\fR and \fBforkx()\fR, fork handlers are not run when +\fBvfork()\fR and \fBvforkx()\fR are called. +.sp +.LP +The \fBvfork()\fR and \fBvforkx()\fR functions are deprecated. Their sole +legitimate use as a prelude to an immediate call to a function from the +\fBexec\fR family can be achieved safely by \fBposix_spawn\fR(3C) or +\fBposix_spawnp\fR(3C). +.SS "Fork Extensions" +.sp +.LP +The \fBvforkx()\fR function accepts a \fIflags\fR argument consisting of a +bitwise inclusive-OR of zero or more of the following flags, which are defined +in the header \fB<sys/fork.h>\fR: +.br +.in +2 +\fBFORK_NOSIGCHLD\fR +.in -2 +.br +.in +2 +\fBFORK_WAITPID\fR +.in -2 +.sp +.LP +See \fBfork\fR(2) for descriptions of these flags. If the \fIflags\fR argument +is 0, \fBvforkx()\fR is identical to \fBvfork()\fR. +.SH RETURN VALUES +.sp +.LP +Upon successful completion, \fBvfork()\fR and \fBvforkx()\fR return \fB0\fR to +the child process and returns the process ID of the child process to the parent +process. Otherwise, \fB\(mi1\fR is returned to the parent process, no child +process is created, and \fBerrno\fR is set to indicate the error. +.SH ERRORS +.sp +.LP +The \fBvfork()\fR and \fBvforkx()\fR functions will fail if: +.sp +.ne 2 +.mk +.na +\fB\fBEAGAIN\fR\fR +.ad +.RS 10n +.rt +The system-imposed limit on the total number of processes under execution +(either system-quality or by a single user) would be exceeded. This limit is +determined when the system is generated. +.RE + +.sp +.ne 2 +.mk +.na +\fB\fBENOMEM\fR\fR +.ad +.RS 10n +.rt +There is insufficient swap space for the new process. +.RE + +.sp +.LP +The \fBvforkx()\fR function will fail if: +.sp +.ne 2 +.mk +.na +\fB\fBEINVAL\fR\fR +.ad +.RS 10n +.rt +The \fIflags\fR argument is invalid. +.RE + +.SH ATTRIBUTES +.sp +.LP +See \fBattributes\fR(5) for descriptions of the following attributes: +.sp + +.sp +.TS +tab() box; +cw(2.75i) |cw(2.75i) +lw(2.75i) |lw(2.75i) +. +ATTRIBUTE TYPEATTRIBUTE VALUE +_ +Interface StabilityObsolete +_ +MT-LevelUnsafe +.TE + +.SH SEE ALSO +.sp +.LP +\fBexec\fR(2), \fBexit\fR(2), \fBfork\fR(2), \fBioctl\fR(2), \fBatexit\fR(3C), +\fBexit\fR(3C), \fBposix_spawn\fR(3C), \fBposix_spawnp\fR(3C), +\fBsignal.h\fR(3HEAD), \fBwait\fR(3C), \fBattributes\fR(5), \fBstandards\fR(5) +.SH NOTES +.sp +.LP +To avoid a possible deadlock situation, processes that are children in the +middle of a \fBvfork()\fR or \fBvforkx()\fR are never sent \fBSIGTTOU\fR or +\fBSIGTTIN\fR signals; rather, output or ioctls are allowed and input attempts +result in an \fBEOF\fR indication. +.sp +.LP +To forestall parent memory corruption due to race conditions with signal +handling, \fBvfork()\fR and \fBvforkx()\fR treat signal handlers in the child +process in the same manner as the \fBexec\fR(2) functions: signals set to be +caught by the parent process are set to the default action (\fBSIG_DFL\fR) in +the child process (see \fBsignal.h\fR(3HEAD)). Any attempt to set a signal +handler in the child before \fBexecve()\fR to anything other than \fBSIG_DFL\fR +or \fBSIG_IGN\fR is disallowed and results in setting the handler to +\fBSIG_DFL\fR. +.sp +.LP +On some systems, the implementation of \fBvfork()\fR and \fBvforkx()\fR cause +the parent to inherit register values from the child. This can create problems +for certain optimizing compilers if \fB<unistd.h>\fR is not included in the +source calling \fBvfork()\fR or if \fB<sys/fork.h>\fR is not included in the +source calling \fBvforkx()\fR. |