Annotation of cci/usr/src/man/man2/vfork.2, revision 1.1.1.1

1.1       root        1: .TH VFORK 2 "2 July 1983"
                      2: .UC 4
                      3: .SH NAME
                      4: vfork \- spawn new process in a virtual memory efficient way
                      5: .SH SYNOPSIS
                      6: .B pid = vfork()
                      7: .br
                      8: .B int pid;
                      9: .SH DESCRIPTION
                     10: .I Vfork
                     11: can be used to create new processes without fully copying the address
                     12: space of the old process, which is horrendously inefficient in a paged
                     13: environment.  It is useful when the purpose of
                     14: .IR fork (2)
                     15: would have been to create a new system context for an
                     16: .IR execve .
                     17: .I Vfork
                     18: differs from
                     19: .I fork
                     20: in that the child borrows the parent's memory and thread of
                     21: control until a call to
                     22: .IR execve (2)
                     23: or an exit (either by a call to
                     24: .IR exit (2)
                     25: or abnormally.)
                     26: The parent process is suspended while the child is using its resources.
                     27: .PP
                     28: .I Vfork
                     29: returns 0 in the child's context and (later) the pid of the child in
                     30: the parent's context.
                     31: .PP
                     32: .I Vfork
                     33: can normally be used just like
                     34: .I fork.
                     35: It does not work, however, to return while running in the childs context
                     36: from the procedure which called
                     37: .I vfork
                     38: since the eventual return from
                     39: .I vfork
                     40: would then return to a no longer existent stack frame.
                     41: Be careful, also, to call
                     42: .I _exit
                     43: rather than
                     44: .I exit
                     45: if you can't
                     46: .IR execve ,
                     47: since
                     48: .I exit
                     49: will flush and close standard I/O channels, and thereby mess up the
                     50: parent processes standard I/O data structures.
                     51: (Even with
                     52: .I fork
                     53: it is wrong to call
                     54: .I exit
                     55: since buffered data would then be flushed twice.)
                     56: .SH SEE ALSO
                     57: fork(2), execve(2), sigvec(2), wait(2),
                     58: .SH DIAGNOSTICS
                     59: Same as for
                     60: .IR fork .
                     61: .SH BUGS
                     62: This system call will be eliminated when proper system sharing
                     63: mechanisms are implemented. 
                     64: Users should not depend on the memory
                     65: sharing semantics of
                     66: .I vfork
                     67: as it will, in that case, be made synonymous to
                     68: .IR fork .
                     69: .PP
                     70: To avoid a possible deadlock situation,
                     71: processes which are children in the middle
                     72: of a
                     73: .I vfork
                     74: are never sent SIGTTOU or SIGTTIN signals; rather,
                     75: output or
                     76: .IR ioctl s
                     77: are allowed
                     78: and input attempts result in an end-of-file indication.

unix.superglobalmegacorp.com

This archive runs on limited infrastructure. Preserving old code on modern bandwidth. Automated agents are requested to crawl responsibly.