Re: 2.6.14-rc1 wait()/SIG_CHILD bevahiour

[Edgar Toernig]

Badari Pulavarty wrote:
>
> I am looking at a problem where the parent process doesn't seem 
> to cleanup the exited children (with a webserver). We narrowed it
> down to a simple testcase. Seems more like a lost SIG_CHILD.

You are aware that the defunct processes are all grandchildren
and the sigchild handler of the children is the one inherited from
the main process?  That sighandler is pretty bogus for the children,
i.e. pid_exited never does anything useful.  Further, the children
terminate unconditionally after some time - all still present grand-
children are inherited by init (pid 1) who does the clean up.

But I think the major problem is that the main-loop of the children
and the signal handler both call usleep(rand()) - I wouldn't be
surprised if this combo isn't reentrant, especially from signal
handlers ...

Ciao, ET.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/