On 4/6/06, Michael S. Tsirkin <[EMAIL PROTECTED]> wrote:
> Quoting r. Muli Ben-Yehuda <[EMAIL PROTECTED]>:
> > Subject: Re: [openib-general] Re: RE: Re: [PATCH] ipoib_flush_paths
> >
> > On Thu, Apr 06, 2006 at 04:38:33PM +0300, Michael S. Tsirkin wrote:
> >
> > > No, since we are keeping a callback pointer into that module.
> >
> > Sorry if I'm being dense but I don't see it in this patch. Point me at
> > it?
>
> You don't see it in the patch because SA already kept a callback pointer -
> that's the race I'm solving. Look in sa_query.c
>
>
> If I have
>
> struct query {
> void (*callback)();
> struct module *owner;
> }
>
> Then it is always safe to do
>
> __get_module(query->owner);
> query->callback();
> put_module(query->owner);
>
> since it is the called module's responsibility to invalidate
> all such query objects before its unloaded.
Wait, why are you doing __get_module just before the callback? This
leaves the possibility of crashing - sure, you'll detect that things
went wrong, but you haven't solved the issue. The whole point of the
reference is to prevent the crash.
You need to call __get_module from the context of teh caller making the request.
- Fab
_______________________________________________
openib-general mailing list
[email protected]
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
