linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Oleg Nesterov <oleg@redhat.com>
To: Tejun Heo <tj@kernel.org>
Cc: Denys Vlasenko <vda.linux@googlemail.com>,
	jan.kratochvil@redhat.com, linux-kernel@vger.kernel.org,
	torvalds@linux-foundation.org, akpm@linux-foundation.org,
	indan@nul.nu, bdonlan@gmail.com
Subject: Re: [PATCH 03/10] ptrace: implement PTRACE_SEIZE
Date: Mon, 30 May 2011 21:22:55 +0200	[thread overview]
Message-ID: <20110530192255.GB20616@redhat.com> (raw)
In-Reply-To: <20110527182121.GA3212@mtj.dyndns.org>

On 05/27, Tejun Heo wrote:
>
> On Thu, May 26, 2011 at 05:01:50PM +0200, Oleg Nesterov wrote:
> > On 05/26, Tejun Heo wrote:
> > > > Or we can change do_wait() to detach a zombie leader. In this case it
> > > > is not clear what should we do if the debugger is the real parent.
> > > > Perhaps do_wait() should do the same: detach a leader (but not reap).
> > > > When the last thread does, the real parent will be notified again.
> > > > IOW, wait(tgid) can succeed twice.
> > >
> > > Just letting PTRACE_DETACH work for zombies sounds much simpler to me.
> >
> > Probably, but please note we have to modify do_wait() anyway. Otherwise
> > in general the tracer simply can not know the tracee has exited. IOW,
> > waitpid(zombie_leader_pid, WEXITED) should succeed without reaping if
> > delay_group_leader(), then the tracer can do PTRACE_DETACH. But this is
> > not symmetrical with sub-thread zombies.
>
> Yes, complicated.  The task/process duality conflicts.  wait(2)
> already is different for group leader and succeeds twice for the
> ptracer and the real parent (when they're different).
>
> If we relocate ptrace group leader zombie wait, as suggested, such
> that it waits for the task itself rather than the whole group, we
> would be taking away a feature - ptracer waiting for the whole process

Partly yes. The parent will be notified again and it can do another
wait() later. The problem is, this is confusing.

> I think group leader wait becoming asymmetrical with sub-thread
> zombies is perfectly fine - it already is.

Well, probably yes... But why? We have to change do_wait() anyway to
report the death of the leader. If we do this, we can detach it as
well.

> But would it be okay to
> change ptrace wait(2) on group leader to wait for the task itself
> rather than the whole group?

Yes, this is iffy. Simply because this can confuse the userspace.
May be we can add W_PTRACED_THREAD_EXITED? (should be used instead
of WEXITED by ptracer). Or, perhaps WEXITED should succeed but put
something special into status?

Oleg.


  reply	other threads:[~2011-05-30 19:24 UTC|newest]

Thread overview: 88+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-05-16 18:17 [PATCHSET ptrace] ptrace: implement PTRACE_SEIZE/INTERRUPT and group stop notification, take#2 Tejun Heo
2011-05-16 18:17 ` [PATCH 01/10] signal: remove three noop tracehooks Tejun Heo
2011-05-17 16:22   ` Christoph Hellwig
2011-05-17 16:27     ` Tejun Heo
2011-05-18 18:45   ` Oleg Nesterov
2011-05-19 12:11     ` Tejun Heo
2011-05-19 16:10       ` Oleg Nesterov
2011-05-16 18:17 ` [PATCH 02/10] job control: introduce JOBCTL_TRAP_STOP and use it for group stop trap Tejun Heo
2011-05-18 16:48   ` Oleg Nesterov
2011-05-18 16:57     ` Oleg Nesterov
2011-05-19 10:19     ` Tejun Heo
2011-05-19 16:19       ` Oleg Nesterov
2011-05-16 18:17 ` [PATCH 03/10] ptrace: implement PTRACE_SEIZE Tejun Heo
2011-05-18  0:40   ` Denys Vlasenko
2011-05-18  9:55     ` Tejun Heo
2011-05-18 10:44       ` Denys Vlasenko
2011-05-18 11:14         ` Tejun Heo
2011-05-19 14:17       ` Tejun Heo
2011-05-19 15:02         ` Tejun Heo
2011-05-19 19:31         ` Pedro Alves
2011-05-19 22:42           ` Denys Vlasenko
2011-05-19 23:00             ` Pedro Alves
2011-05-20  1:44               ` Denys Vlasenko
2011-05-20  8:56                 ` Pedro Alves
2011-05-20  9:12                   ` Tejun Heo
2011-05-20  9:07               ` Tejun Heo
2011-05-20  9:27                 ` Pedro Alves
2011-05-20  9:31                   ` Tejun Heo
2011-05-24  9:49                     ` Pedro Alves
2011-05-24 12:00                       ` Tejun Heo
2011-05-24 12:36                         ` Pedro Alves
2011-05-24 14:02                           ` Tejun Heo
2011-05-24 14:55                             ` Pedro Alves
2011-05-25 18:18                             ` Oleg Nesterov
2011-05-26  9:10                               ` Tejun Heo
2011-05-26 10:01                                 ` Pedro Alves
2011-05-26 10:11                                   ` Tejun Heo
2011-05-26 14:55                                 ` Oleg Nesterov
2011-05-23 13:09         ` Oleg Nesterov
2011-05-23 12:43       ` Oleg Nesterov
2011-05-24 10:28         ` Tejun Heo
2011-05-25 18:29           ` Oleg Nesterov
2011-05-26  9:14             ` Tejun Heo
2011-05-26 15:01               ` Oleg Nesterov
2011-05-27 18:21                 ` Tejun Heo
2011-05-30 19:22                   ` Oleg Nesterov [this message]
     [not found]                     ` <BANLkTimupSd774N-VBoswOj+Dza=5ofvWQ@mail.gmail.com>
2011-05-31 19:08                       ` Oleg Nesterov
2011-05-31 21:32                         ` Linus Torvalds
2011-06-01 20:04                           ` Oleg Nesterov
2011-06-01  5:34                         ` Tejun Heo
2011-06-01 20:08                           ` Oleg Nesterov
2011-06-02  5:01                             ` Tejun Heo
2011-05-18 18:17   ` Oleg Nesterov
2011-05-19 10:34     ` Tejun Heo
2011-05-16 18:17 ` [PATCH 04/10] ptrace: implement PTRACE_INTERRUPT Tejun Heo
2011-05-18 18:38   ` Oleg Nesterov
2011-05-19 12:07     ` Tejun Heo
2011-05-19 16:21       ` Oleg Nesterov
2011-05-16 18:17 ` [PATCH 05/10] ptrace: restructure ptrace_getsiginfo() Tejun Heo
2011-05-16 18:17 ` [PATCH 06/10] ptrace: add siginfo.si_pt_flags Tejun Heo
2011-05-16 18:17 ` [PATCH 07/10] ptrace: make group stop state visible via PTRACE_GETSIGINFO Tejun Heo
2011-05-19 16:27   ` Oleg Nesterov
2011-05-19 16:40     ` Tejun Heo
2011-05-16 18:17 ` [PATCH 08/10] ptrace: don't let PTRACE_SETSIGINFO override __SI_TRAP siginfo Tejun Heo
2011-05-16 18:17 ` [PATCH 09/10] ptrace: add JOBCTL_BLOCK_NOTIFY Tejun Heo
2011-05-19 16:32   ` Oleg Nesterov
2011-05-19 16:44     ` Tejun Heo
2011-05-19 16:48       ` Oleg Nesterov
2011-05-19 16:58         ` Tejun Heo
2011-05-16 18:17 ` [PATCH 10/10] ptrace: implement group stop notification for ptracer Tejun Heo
2011-05-19 16:32   ` Oleg Nesterov
2011-05-19 16:57     ` Tejun Heo
2011-05-19 17:13       ` Oleg Nesterov
2011-05-19 22:48         ` Denys Vlasenko
2011-05-20  8:59           ` Tejun Heo
2011-05-23 13:34             ` Oleg Nesterov
2011-05-20  8:46         ` Tejun Heo
2011-05-19 16:58     ` Oleg Nesterov
2011-05-23 11:45       ` Oleg Nesterov
2011-05-24 13:44         ` Tejun Heo
2011-05-24 15:44           ` Tejun Heo
2011-05-26 14:44           ` Oleg Nesterov
2011-05-28  7:32             ` Tejun Heo
2011-05-18 18:50 ` [PATCHSET ptrace] ptrace: implement PTRACE_SEIZE/INTERRUPT and group stop notification, take#2 Oleg Nesterov
2011-05-19 12:08   ` Tejun Heo
2011-05-19 15:04 ` Linus Torvalds
2011-05-19 15:19   ` Tejun Heo
2011-05-19 22:45   ` Denys Vlasenko

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20110530192255.GB20616@redhat.com \
    --to=oleg@redhat.com \
    --cc=akpm@linux-foundation.org \
    --cc=bdonlan@gmail.com \
    --cc=indan@nul.nu \
    --cc=jan.kratochvil@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=tj@kernel.org \
    --cc=torvalds@linux-foundation.org \
    --cc=vda.linux@googlemail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).