Re: Fix signalfd interaction with thread-private signals

[Oleg Nesterov <oleg@tv-sign.ru>]

On 06/22, Linus Torvalds wrote:
> 
> On Sat, 23 Jun 2007, Benjamin Herrenschmidt wrote:
> > 
> > > 
> > > It does exactly so, please note this chunk
> > > 
> > >  @@ -330,7 +339,7 @@ asmlinkage long sys_signalfd(int ufd, si
> > > 
> > >                  init_waitqueue_head(&ctx->wqh);
> > >                  ctx->sigmask = sigmask;
> > >  -               ctx->tsk = current;
> > >  +               ctx->tsk = current->group_leader;
> > > 
> > > > It might well be that signalfd's concept of context is wrong in the
> > > > first place and it should be attached to processes rather than threads
> > > > and that made more explicit in the first place...
> > 
> > Yup, looks like I was looking at a wrong patch... I think it's the right
> > thing to do indeed.
> 
> Quite frankly, it strikes me that if we want to do this, then we shouldn't 
> save the _process_ information at all, we should save the "sighand" 
> instead.

Yes, unless we pass signalfd to another process.

> So either we save the process info, or we save the sighand, but saving the 
> "group_leader" seems totally bogus. Especially as the group leader can 
> change (by execve()).

Note that exec() can change ->sighand as well, so in general this can't help.

Currently not a problem, exec() detaches process from signalfd anyway.


This reminds me, we have a similar problem with !CPUCLOCK_PERTHREAD()
cpu-timers, iirc. They can survive after exec(), but only if it was
->group_leader who does exec.

Oleg.

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