lkml.org 
[lkml]   [2014]   [Nov]   [25]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH 1/2] exit: reparent: fix/update the comments in zap_pid_ns_processes()
forgot to mention...

On 11/25, Oleg Nesterov wrote:
>
> On 11/24, Eric W. Biederman wrote:
> >
> > - We won't wait for an injected process in a pid namespace,
> > or a processes debugged with gdb to be reaped before the pid
> > init process exits if we don't wait.
>
> Yes, and I do not see why this is bad, but this is off-topic.
>
> Again, lets discuss this in another thread. This patch doesn't try to
> document the desired semantics, it only tries to explain why zap_pid_ns
> _has_ to wait until EXIT_DEAD (and in fact EXIT_ZOMBIE) tasks go away.

Just in case, as for "processes debugged with gdb" I obviously agree,
but this has nothing to do with EXIT_DEAD threads. zap_pid_ns_processes()
will sleep in sys_wait4() until debugger detaches the tracee and reaps
it due to ignored SIGCHLD.

But again, this is off-topic now.

> Eric. I'd really ask you to take another look. But in any case: thanks for
> looking at this.

Yes, please.

Oleg.



\
 
 \ /
  Last update: 2014-11-25 18:42    [W:0.161 / U:0.660 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site