Re[2]: [Gc] [PATCH] define HANDLE_FORK on unix platforms with pthreads

[Ivan Maidanski]

Hi Andy,

Yes, you're right.
I'll try at least to check presence of pthread_atfork in various mentioned systems next week.

Regards.

19 02 2012, 00:59 Andy Wingo <wingo at pobox.com>:
> On Fri 17 Feb 2012 21:45, Ivan Maidanski <ivmai at mail.ru> writes:
> 
> > It seems you're proposing the right thing but I guess your patch could
> > break build for some targets (there are some many Unix targets supported
> > like xBSD, RTEMS, NaCl, Android, Darwin, DG/UX, HP/UX, Aix, Irix,
> > Solaris... - does at least pthread_atfork exist everywhere?
> 
> I don't know.  It's from the SUSV2 standard, released in 1997.  I would
> think it's fairly portable, but I really don't know.  The bigger problem
> is that that part of libgc probably isn't as tested.  It does seem to be
> the right thing, though -- as much as there can be a right thing, when
> you are combining threads, gc, and fork() ;-)
> 
> If we want to restrict this to e.g. glibc targets, that's fine; though,
> there is no inherent reason why this couldn't work on other POSIX
> platforms as well.
> 
> Regards,
> 
> Andy
> --
> http://wingolog.org/
> _______________________________________________
> Gc mailing list
> Gc at linux.hpl.hp.com
> http://www.hpl.hp.com/hosted/linux/mail-archives/gc/
>