public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: sukadev@us.ibm.com
To: Oleg Nesterov <oleg@tv-sign.ru>, Andrew Morton <akpm@osdl.org>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>,
	Cedric Le Goater <clg@fr.ibm.com>,
	Dave Hansen <haveblue@us.ibm.com>,
	Serge Hallyn <serue@us.ibm.com>,
	Herbert Poetzel <herbert@13thfloor.at>,
	Pavel Emelianov <xemul@openvz.org>,
	Linux Containers <containers@lists.osdl.org>,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH 3/5] Use task_pid() to find leader's pid
Date: Mon, 16 Jul 2007 12:59:52 -0700	[thread overview]
Message-ID: <20070716195952.GG25053@us.ibm.com> (raw)
In-Reply-To: <20070715121750.GA173@tv-sign.ru>

Oleg Nesterov [oleg@tv-sign.ru] wrote:
| Sukadev Bhattiprolu wrote:
| >
| > Use task_pid() to get leader's pid since find_pid() cannot be used
| > after detach_pid(). See comments in the code below for more details.
| >
| > ...
| >
| > +		 * Note: With multiple pid namespaces, active pid namespace of
| > +		 * 	 a process is stored in its struct pid. The detach_pid
| > +		 * 	 below frees the struct pid, so we will have no notion
| > +		 * 	 of an active pid namespace until we complete the
| > +		 * 	 subsequent attach_pid(). Which means - calls like
| > +		 * 	 find_pid()/pid_to_nr() return NULL and cannot be used
| > +		 * 	 between the detach_pid() and attach_pid() calls.
| 
| I think both the changelog and the comment are confusing,
| 
| >  		detach_pid(tsk, PIDTYPE_PID);
| >  		tsk->pid = leader->pid;
| > -		attach_pid(tsk, PIDTYPE_PID,  find_pid(tsk->pid));
| > +		attach_pid(tsk, PIDTYPE_PID,  task_pid(leader));
| 
| because the change itself looks like an obvious performance fix, even
| we don't use multiple pid namespaces. I don't think it is good idea to
| add a fat comment which doesn't match the current reality, and find_pid()
| should be avoided anyway.

Its a performance fix but also a correctness issue with multiple pid
namespaces. Here is the modified patch with the simplified changelog and
comment removed.

| 
| Stupid question: why do we need to put the pid namespace into the struct
| pid? Isn't it better if the user of the struct pid should know its ns?
| For example, if /proc does put_pid(), that pid should be from the active
| namespace.

Not sure I fully understand this. A process, and by extension its 'struct
pid' is visible in multiple namespaces and we maintain this list of
namespaces in each 'struct pid'.

Are you suggesting having a pid_namespace with a list of all 'struct pids'
that are visible in it ?

| 
| Sukadev, could you cc me if you do that kind of changes?

Sure - I will.
---


Subject: [PATCH 3/5] Use task_pid() to find leader's pid

From: Sukadev Bhattiprolu <sukadev@us.ibm.com>

Use task_pid() to get leader's 'struct pid' and avoid the find_pid().

Signed-off-by: Sukadev Bhattiprolu <sukadev@us.ibm.com>
Acked-by: Pavel Emelianov <xemul@openvz.org>

Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Cedric Le Goater <clg@fr.ibm.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: Serge Hallyn <serue@us.ibm.com>
Cc: Herbert Poetzel <herbert@13thfloor.at>
---
 fs/exec.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Index: lx26-22-rc6-mm1a/fs/exec.c
===================================================================
--- lx26-22-rc6-mm1a.orig/fs/exec.c	2007-07-13 18:23:55.000000000 -0700
+++ lx26-22-rc6-mm1a/fs/exec.c	2007-07-16 12:56:22.000000000 -0700
@@ -908,7 +908,7 @@ static int de_thread(struct task_struct 
 		 */
 		detach_pid(tsk, PIDTYPE_PID);
 		tsk->pid = leader->pid;
-		attach_pid(tsk, PIDTYPE_PID,  find_pid(tsk->pid));
+		attach_pid(tsk, PIDTYPE_PID,  task_pid(leader));
 		transfer_pid(leader, tsk, PIDTYPE_PGID);
 		transfer_pid(leader, tsk, PIDTYPE_SID);
 		list_replace_rcu(&leader->tasks, &tsk->tasks);

  reply	other threads:[~2007-07-16 20:00 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-07-15 12:17 [PATCH 3/5] Use task_pid() to find leader's pid Oleg Nesterov
2007-07-16 19:59 ` sukadev [this message]
2007-07-17 13:44   ` Oleg Nesterov
  -- strict thread matches above, loose matches on Subject: below --
2007-07-15  4:57 sukadev

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=20070716195952.GG25053@us.ibm.com \
    --to=sukadev@us.ibm.com \
    --cc=akpm@osdl.org \
    --cc=clg@fr.ibm.com \
    --cc=containers@lists.osdl.org \
    --cc=ebiederm@xmission.com \
    --cc=haveblue@us.ibm.com \
    --cc=herbert@13thfloor.at \
    --cc=linux-kernel@vger.kernel.org \
    --cc=oleg@tv-sign.ru \
    --cc=serue@us.ibm.com \
    --cc=xemul@openvz.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