Re: Dettach problem

grape wrote: > Hello Richard, > > Richard Lynch skrev: > >> >> This seems to me like a cogent bug report... >> http://bugs.php.net/ >> >> > Yup, I'll report it if we cannot find a explaination in a couple of days. as a last resort you might as on internals@lists.php.net as to whether anyone there considers this a possible bug. > >> But what does the posix_setsid() bit do? Seems like you could take >> that out too, no?... Or does that "promote" the process to "be" the >> parent somehow?... Thereby confusing "fork" into thinking that it's >> not the child process in some twisted weird way? I read the docs ; >> But comprehension is not implied. :-) >> >> > The documentation on posix_setsid() is a bit sparse on php.net, but the > manpage for setsid on my FreeBSD system gives: > > DESCRIPTION > The setsid() system call creates a new session. The calling process is > the session leader of the new session, is the process group leader of a > new process group and has no controlling terminal. The calling process > is the only process in either the session or the process group. > > I found a post on the net describing how to run a PHP-script as a > daemon, and the author used posix_setsid() for it to work. Actually, the > posix_setsid() isn't really required to disconnect the childs I/O using > PHP 5.0.4, I can just do: > > <? > echo "Hello from parent\n"; > > if(pcntl_fork()) { > exit; > } > > fclose( STDIN ); > fclose( STDOUT ); > fclose( STDERR ); > > echo "This message should not go to stdout\n"; > ?> > > This is was I originally had in my code, but after moving to 5.1.2 I > started to experiment... :-) > > A while ago, I discovered that PHP 4.3.10-2 (cli) on a Debian Linux with > 2.4.27 kernel "misbehaves" as well, using posix_setsid() or not. I see > the output of the last echo command on stdout of the parent process > there also. > > The plot thickens! > > /Grape >