Re: permissions (and other) problems

On Fri, Sep 13, 2002 at 01:13:57AM +0200, Branko Cibej wrote:
> Jon Watte wrote:
>...
> >While that is probably true, the benefit of using the signal handler
> >approach is that you can also catch SIGSEGV to do the same clean-up :-)
>
> There's also the small problem that longjmp'ing out of a signal handler
> loses context, which means that we can't clean up the database locks
> anyway -- defeating the purpose of this exercice.
>
> And trying to clean up after a seg faulr simply doesn't work in general.

Right. At one point, the Apache httpd grew a signal handler for SIGSEGV. One
of the hardcore geeks ripped a couple people a new one for that. There is
simply *no* effective response to a SIGSEGV. There is no way to clean up,
and there is even no way to issue a message to the user (no I/O at all). You
can't be sure data is around (the SEGV may have been an invalid access to
data), and all your memory and I/O structures could be toast.

Nope... just avoid trying to catch SIGSEGV. Die, and let your parent process
do something about it. (of course, that is usually the shell in terms of the
cmdline client :-)

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org