Re: [patch 1/4] signalfd v1 - signalfd core ...

[Davide Libenzi <davidel@xmailserver.org>]

On Wed, 7 Mar 2007, Ulrich Drepper wrote:

> On 3/7/07, Linus Torvalds <torvalds@linux-foundation.org> wrote:
> > > 1) You want standard delivery only:
> > >
> > >    - Just dont use signalfd
> > >
> > > 2) you want signalfd only:
> > >
> > >    - Do a sigprocmask(SIG_BLOCK) of the same mask you pass to signalfd
> > >
> > > If you want both, you can have it. Race free.
> > 
> > .. but maybe with more code and lots of confusion. I'm still unclear on
> > any upsides here.
> > 
> > Choice is good, but only if it's *useful* choice.
> 
> Not only that.  If you don't force that the signal is blocked when
> using signalfd() you are bound to run into problems.  For the same
> reason is it required to have signals blocked when you use sigwait()
> etc.  Don't try to innovate here, I guarantee you it's going to break
> something somewhere.

Let's do this. How about you throw this way one of the case that would 
possibly break, and I test it?
Queues are separeted, so if you have a signal hitting you while in 
signalfd_dequeue, either the signal is not for your set (and you'd get an 
EINTR ATM), or it is for your set too, and signalfd_dequeue will return 
with it.
But again, I'd love to have some insides about possible breakages so that 
I can make test cases for them.



- Davide