From: Denys Vlasenko <vda.linux@googlemail.com>
To: Tejun Heo <tj@kernel.org>
Cc: Oleg Nesterov <oleg@redhat.com>,
Roland McGrath <roland@redhat.com>,
jan.kratochvil@redhat.com, linux-kernel@vger.kernel.org,
torvalds@linux-foundation.org, akpm@linux-foundation.org
Subject: Re: [PATCH 1/1] ptrace: make sure do_wait() won't hang after PTRACE_ATTACH
Date: Sun, 13 Feb 2011 22:24:25 +0100 [thread overview]
Message-ID: <201102132224.25197.vda.linux@googlemail.com> (raw)
In-Reply-To: <20110204144858.GI12133@htj.dyndns.org>
On Friday 04 February 2011 15:48, Tejun Heo wrote:
> > But even this doesn't matter. We can not change ptrace API so that,
> > say, it does not reparent the tracee. Once we do this, we already
> > have the new API.
>
> I would argue that we can get by well enough by trimming and updating
> the curren ptrace API.
In the past Roland wasn't very enthusiastic about changes which
were fixing ptrace's bugs-turned-features.
If you want to do that, you need to convince him to change
his position a bit.
For example, PTRACE_DETACH requires tracee to be stopped to succeed.
If debugger tries to detach while the tracee is running, it will get
an error. This forces debugger to do stupid things like sending SIGSTOP,
then waiting for tracee to stop, then doing PTRACE_DETACH, then
sending SIGCONT. Of course, while this dance is performed,
any SIGSTOPs/SIGCONTs which may be sent to the tracee by other processes
are totally disrupted by this.
The natural (for me) fix is to make PTRACE_DETACH work even on running
tracee. It simply makes a lot of sense. Why on earth do we need tracee
to be stopped? There is no reason.
But this is a change in ptrace behavior, and therefore is not acceptable
for Roland.
Basically, we have slightly idea what are ptrace's development goals are.
>From my POV, we want to have strace tool which is *completely*, 100%
transparent for traced process and its parent (sans decrease in speed).
No "vfork turns into fork under strace" (we had this sometime ago),
no missed or misinterpreted SIGSTOPs or SIGTRAPs. Real parent
should still see its child stopping on ^Z even if strace -p PID
attached itself to the child. Etc etc etc.
If this can only be achieved by slightly changing (basically, fixing)
ptrace API, I'd go for it.
In many cases, Roland won't.
He is the maintainer. If you want changes which break some aspects of
current behavior, even quirky ones, you need to convince him.
> I could be wrong (with pretty high probability) but I don't really see
> the pressing need for a completely new API. ptrace sure is ugly and
> quirky but it's something people are already used to.
--
vda
next prev parent reply other threads:[~2011-02-13 21:24 UTC|newest]
Thread overview: 160+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-01-28 15:08 [PATCHSET] ptrace,signal: group stop / ptrace updates Tejun Heo
2011-01-28 15:08 ` [PATCH 01/10] signal: fix SIGCONT notification code Tejun Heo
2011-01-28 15:08 ` [PATCH 02/10] ptrace: remove the extra wake_up_process() from ptrace_detach() Tejun Heo
2011-01-28 18:46 ` Roland McGrath
2011-01-31 10:38 ` Tejun Heo
2011-02-01 10:26 ` [PATCH] ptrace: use safer wake up on ptrace_detach() Tejun Heo
2011-02-01 13:40 ` Oleg Nesterov
2011-02-01 15:07 ` Tejun Heo
2011-02-01 19:17 ` Oleg Nesterov
2011-02-02 5:31 ` Roland McGrath
2011-02-02 10:35 ` Tejun Heo
2011-02-02 0:27 ` Andrew Morton
2011-02-02 5:33 ` Roland McGrath
2011-02-02 5:38 ` Andrew Morton
2011-02-02 10:34 ` Tejun Heo
2011-02-02 19:33 ` Andrew Morton
2011-02-02 20:01 ` Tejun Heo
2011-02-02 21:40 ` Oleg Nesterov
2011-02-02 5:29 ` Roland McGrath
2011-02-02 5:28 ` [PATCH 02/10] ptrace: remove the extra wake_up_process() from ptrace_detach() Roland McGrath
2011-01-28 15:08 ` [PATCH 03/10] signal: remove superflous try_to_freeze() loop in do_signal_stop() Tejun Heo
2011-01-28 18:46 ` Roland McGrath
2011-01-28 15:08 ` [PATCH 04/10] ptrace: kill tracehook_notify_jctl() Tejun Heo
2011-01-28 21:09 ` Roland McGrath
2011-01-28 15:08 ` [PATCH 05/10] ptrace: add @why to ptrace_stop() Tejun Heo
2011-01-28 18:48 ` Roland McGrath
2011-01-28 15:08 ` [PATCH 06/10] signal: fix premature completion of group stop when interfered by ptrace Tejun Heo
2011-01-28 21:22 ` Roland McGrath
2011-01-31 11:00 ` Tejun Heo
2011-02-02 5:44 ` Roland McGrath
2011-02-02 10:56 ` Tejun Heo
2011-01-28 15:08 ` [PATCH 07/10] signal: use GROUP_STOP_PENDING to stop once for a single group stop Tejun Heo
2011-01-28 15:08 ` [PATCH 08/10] ptrace: participate in group stop from ptrace_stop() iff the task is trapping for " Tejun Heo
2011-01-28 21:30 ` Roland McGrath
2011-01-31 11:26 ` Tejun Heo
2011-02-02 5:57 ` Roland McGrath
2011-02-02 10:53 ` Tejun Heo
2011-02-03 10:02 ` Tejun Heo
2011-02-01 19:36 ` Oleg Nesterov
2011-01-28 15:08 ` [PATCH 09/10] ptrace: make do_signal_stop() use ptrace_stop() if the task is being ptraced Tejun Heo
2011-01-28 15:08 ` [PATCH 10/10] ptrace: clean transitions between TASK_STOPPED and TRACED Tejun Heo
2011-02-03 20:41 ` [PATCH 0/1] (Was: ptrace: clean transitions between TASK_STOPPED and TRACED) Oleg Nesterov
2011-02-03 20:41 ` [PATCH 1/1] ptrace: make sure do_wait() won't hang after PTRACE_ATTACH Oleg Nesterov
2011-02-03 21:36 ` Roland McGrath
2011-02-03 21:44 ` Oleg Nesterov
2011-02-04 10:53 ` Tejun Heo
2011-02-04 13:04 ` Oleg Nesterov
2011-02-04 14:48 ` Tejun Heo
2011-02-04 17:06 ` Oleg Nesterov
2011-02-05 13:39 ` Tejun Heo
2011-02-07 13:42 ` Oleg Nesterov
2011-02-07 14:11 ` Tejun Heo
2011-02-07 15:37 ` Oleg Nesterov
2011-02-07 16:31 ` Tejun Heo
2011-02-07 17:48 ` Oleg Nesterov
2011-02-09 14:18 ` Tejun Heo
2011-02-09 14:21 ` Tejun Heo
2011-02-09 21:25 ` Oleg Nesterov
2011-02-13 23:01 ` Denys Vlasenko
2011-02-14 9:03 ` Jan Kratochvil
2011-02-14 11:39 ` Denys Vlasenko
2011-02-14 17:32 ` Oleg Nesterov
2011-02-14 16:01 ` Oleg Nesterov
2011-02-26 3:59 ` Pavel Machek
2011-02-14 15:51 ` Oleg Nesterov
2011-02-14 14:50 ` Tejun Heo
2011-02-14 18:53 ` Oleg Nesterov
2011-02-13 22:25 ` Denys Vlasenko
2011-02-14 15:13 ` Tejun Heo
2011-02-14 16:15 ` Oleg Nesterov
2011-02-14 16:33 ` Tejun Heo
2011-02-14 17:23 ` Oleg Nesterov
2011-02-14 17:20 ` Denys Vlasenko
2011-02-14 17:30 ` Tejun Heo
2011-02-14 17:45 ` Oleg Nesterov
2011-02-14 17:54 ` Denys Vlasenko
2011-02-21 15:16 ` Tejun Heo
2011-02-21 15:28 ` Oleg Nesterov
2011-02-21 16:11 ` [pseudo patch] ptrace should respect the group stop Oleg Nesterov
2011-02-22 16:24 ` [PATCH 1/1] ptrace: make sure do_wait() won't hang after PTRACE_ATTACH Tejun Heo
2011-02-24 21:08 ` Oleg Nesterov
2011-02-25 15:45 ` Tejun Heo
2011-02-25 17:42 ` Roland McGrath
2011-02-28 15:23 ` Oleg Nesterov
2011-02-14 17:51 ` Oleg Nesterov
2011-02-14 18:55 ` Denys Vlasenko
2011-02-14 19:01 ` Oleg Nesterov
2011-02-14 19:42 ` Denys Vlasenko
2011-02-14 20:01 ` Oleg Nesterov
2011-02-15 15:24 ` Tejun Heo
2011-02-15 15:58 ` Oleg Nesterov
2011-02-15 17:31 ` Roland McGrath
2011-02-15 20:27 ` Oleg Nesterov
2011-02-18 17:02 ` Tejun Heo
2011-02-18 19:37 ` Oleg Nesterov
2011-02-21 16:22 ` Tejun Heo
2011-02-21 16:49 ` Oleg Nesterov
2011-02-21 16:59 ` Tejun Heo
2011-02-23 19:31 ` Oleg Nesterov
2011-02-25 15:10 ` Tejun Heo
2011-02-24 20:29 ` Oleg Nesterov
2011-02-25 15:51 ` Tejun Heo
2011-02-26 2:48 ` Denys Vlasenko
2011-02-28 12:56 ` Tejun Heo
2011-02-28 13:16 ` Denys Vlasenko
2011-02-28 13:29 ` Tejun Heo
2011-02-28 13:41 ` Denys Vlasenko
2011-02-28 13:53 ` Tejun Heo
2011-02-28 14:25 ` Denys Vlasenko
2011-02-28 14:39 ` Tejun Heo
2011-02-28 16:48 ` Oleg Nesterov
2011-02-28 14:36 ` Oleg Nesterov
2011-02-16 21:51 ` Jan Kratochvil
2011-02-17 3:37 ` Denys Vlasenko
2011-02-17 19:19 ` Oleg Nesterov
2011-02-18 21:11 ` Jan Kratochvil
2011-02-19 20:16 ` Oleg Nesterov
2011-02-17 16:49 ` Oleg Nesterov
2011-02-17 18:58 ` Roland McGrath
2011-02-17 19:33 ` Oleg Nesterov
2011-02-18 21:34 ` Jan Kratochvil
2011-02-19 20:06 ` Oleg Nesterov
2011-02-20 9:40 ` Jan Kratochvil
2011-02-20 17:06 ` Denys Vlasenko
2011-02-20 17:48 ` Oleg Nesterov
2011-02-20 19:10 ` Jan Kratochvil
2011-02-20 19:16 ` Oleg Nesterov
2011-02-20 17:16 ` Oleg Nesterov
2011-02-20 18:52 ` Jan Kratochvil
2011-02-20 20:38 ` Oleg Nesterov
2011-02-20 21:06 ` `(T) stopped' preservation after _exit() [Re: [PATCH 1/1] ptrace: make sure do_wait() won't hang after PTRACE_ATTACH] Jan Kratochvil
2011-02-20 21:19 ` Oleg Nesterov
2011-02-20 21:20 ` [PATCH 1/1] ptrace: make sure do_wait() won't hang after PTRACE_ATTACH Jan Kratochvil
2011-02-21 14:23 ` Oleg Nesterov
2011-02-23 16:44 ` Jan Kratochvil
2011-02-14 15:31 ` Oleg Nesterov
2011-02-14 17:24 ` Denys Vlasenko
2011-02-14 17:39 ` Oleg Nesterov
2011-02-14 17:57 ` Denys Vlasenko
2011-02-14 18:00 ` Oleg Nesterov
2011-02-14 18:06 ` Oleg Nesterov
2011-02-14 18:59 ` Denys Vlasenko
2011-02-13 21:24 ` Denys Vlasenko [this message]
2011-02-14 15:06 ` Oleg Nesterov
2011-02-14 15:19 ` Tejun Heo
2011-02-14 16:20 ` Oleg Nesterov
2011-02-14 17:05 ` Denys Vlasenko
2011-02-14 17:18 ` Oleg Nesterov
2011-01-28 16:54 ` [PATCHSET] ptrace,signal: group stop / ptrace updates Ingo Molnar
2011-01-28 17:41 ` Thomas Gleixner
2011-01-28 18:04 ` Anca Emanuel
2011-01-28 18:36 ` Mathieu Desnoyers
2011-01-28 17:55 ` Oleg Nesterov
2011-01-28 18:29 ` Bash not reacting to Ctrl-C Ingo Molnar
2011-02-05 20:34 ` Oleg Nesterov
2011-02-07 13:08 ` Oleg Nesterov
2011-02-09 6:17 ` Michael Witten
2011-02-09 14:53 ` Ingo Molnar
2011-02-09 19:37 ` Michael Witten
2011-02-11 14:41 ` Pavel Machek
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=201102132224.25197.vda.linux@googlemail.com \
--to=vda.linux@googlemail.com \
--cc=akpm@linux-foundation.org \
--cc=jan.kratochvil@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=oleg@redhat.com \
--cc=roland@redhat.com \
--cc=tj@kernel.org \
--cc=torvalds@linux-foundation.org \
/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).