Richard Braun, le Mon 18 Nov 2013 00:36:14 +0100, a écrit : > On Sun, Nov 17, 2013 at 09:41:03PM +0100, Samuel Thibault wrote: > > Richard Braun, le Sun 17 Nov 2013 17:13:23 +0100, a écrit : > > > This patch makes use of a new GNU Mach specific call > > > (thread_terminate_release [1]) so that threads do terminate themselves > > > and release their stack, and in addition their last self reference, > > > and their reply port. > > > > I was first wondering whether I liked that. After thinking more about > > it, I do. Thread termination has always been a burden in thread > > libraries, and it's really way simpler when it's the kernel which > > releases the last bits. Your patch lines count shows that. So I'd say > > go ahead. > > What do you think about error handling in that kernel call though ? > For now, they're ignored on the account that the thread ought to pass > correct values, but I'm not really happy about it.
I'd be OK with it. There is not much to do here. > On the other hand, I wonder what returning on first error would bring > besides help with debugging, which can already be done from userspace > right before calling the RPC. Yes. And at best userland would, in the non-debugging case, try to cope with the issue the same way the kernel would with the behavior proposed above. Samuel
