From: Oleg Nesterov <oleg@redhat.com>
To: Anton Vorontsov <anton.vorontsov@linaro.org>
Cc: Mike Frysinger <vapier@gentoo.org>,
Peter Zijlstra <a.p.zijlstra@chello.nl>,
user-mode-linux-devel@lists.sourceforge.net,
linux-sh@vger.kernel.org, Richard Weinberger <richard@nod.at>,
linux-kernel@vger.kernel.org,
uclinux-dist-devel@blackfin.uclinux.org, linux-mm@kvack.org,
Paul Mundt <lethal@linux-sh.org>,
John Stultz <john.stultz@linaro.org>,
KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>,
Russell King <linux@arm.linux.org.uk>,
Andrew Morton <akpm@linux-foundation.org>,
linuxppc-dev@lists.ozlabs.org,
linux-arm-kernel@lists.infradead.org
Subject: Re: [PATCH v2.1 01/10] cpu: Introduce clear_tasks_mm_cpumask() helper
Date: Sun, 25 Mar 2012 19:42:10 +0200 [thread overview]
Message-ID: <20120325174210.GA23605@redhat.com> (raw)
In-Reply-To: <20120324164316.GB3640@lizard>
On 03/24, Anton Vorontsov wrote:
>
> Many architctures clear tasks' mm_cpumask like this:
>
> read_lock(&tasklist_lock);
> for_each_process(p) {
> if (p->mm)
> cpumask_clear_cpu(cpu, mm_cpumask(p->mm));
> }
> read_unlock(&tasklist_lock);
Namely arm, powerpc, and sh.
> The code above has several problems, such as:
>
> 1. Working with task->mm w/o getting mm or grabing the task lock is
> dangerous as ->mm might disappear (exit_mm() assigns NULL under
> task_lock(), so tasklist lock is not enough).
This is not actually true for arm and sh, afaics. They do not even
need tasklist or rcu lock for for_each_process().
__cpu_disable() is called by __stop_machine(), we know that nobody
can preempt us and other CPUs can do nothing.
> 2. Checking for process->mm is not enough because process' main
> thread may exit or detach its mm via use_mm(), but other threads
> may still have a valid mm.
Yes,
> Also, Per Peter Zijlstra's idea, now we don't grab tasklist_lock in
> the new helper, instead we take the rcu read lock. We can do this
> because the function is called after the cpu is taken down and marked
> offline, so no new tasks will get this cpu set in their mm mask.
And only powerpc needs rcu_read_lock() and task_lock().
OTOH, I do not understand why powepc does this on CPU_DEAD...
And probably CPU_UP_CANCELED doesn't need to clear mm_cpumask().
That said, personally I think these patches are fine, the common
helper makes sense.
Oleg.
next prev parent reply other threads:[~2012-03-25 17:51 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-03-24 10:26 [PATCH v2 0/10] Fixes for common mistakes w/ for_each_process and task->mm Anton Vorontsov
2012-03-24 10:27 ` [PATCH 01/10] cpu: Introduce clear_tasks_mm_cpumask() helper Anton Vorontsov
2012-03-24 12:43 ` Peter Zijlstra
2012-03-24 16:43 ` [PATCH v2.1 " Anton Vorontsov
2012-03-25 17:42 ` Oleg Nesterov [this message]
2012-03-26 7:59 ` Peter Zijlstra
2012-03-26 17:04 ` Oleg Nesterov
2012-03-26 17:23 ` Peter Zijlstra
2012-03-28 0:01 ` Benjamin Herrenschmidt
2012-03-24 10:28 ` [PATCH 02/10] arm: Use clear_tasks_mm_cpumask() Anton Vorontsov
2012-03-24 10:28 ` [PATCH 03/10] powerpc: " Anton Vorontsov
2012-03-24 10:28 ` [PATCH 04/10] sh: " Anton Vorontsov
2012-03-24 10:29 ` [PATCH 05/10] blackfin: A couple of task->mm handling fixes Anton Vorontsov
2012-03-24 10:30 ` [PATCH 06/10] blackfin: Fix possible deadlock in decode_address() Anton Vorontsov
2012-03-24 10:30 ` [PATCH 07/10] um: Should hold tasklist_lock while traversing processes Anton Vorontsov
2012-03-24 11:12 ` Richard Weinberger
2012-03-24 12:48 ` Peter Zijlstra
2012-03-24 16:43 ` Anton Vorontsov
2012-03-24 10:30 ` [PATCH 08/10] um: Fix possible race on task->mm Anton Vorontsov
2012-03-24 11:12 ` Richard Weinberger
2012-03-24 10:31 ` [PATCH 09/10] um: Properly check all process' threads for a live mm Anton Vorontsov
2012-03-24 11:12 ` Richard Weinberger
2012-03-24 10:31 ` [PATCH 10/10] oom: Make find_lock_task_mm() sparse-aware Anton Vorontsov
2012-03-24 12:52 ` Peter Zijlstra
2012-03-24 16:21 ` Anton Vorontsov
2012-03-24 16:43 ` Peter Zijlstra
2012-03-28 7:20 ` David Rientjes
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=20120325174210.GA23605@redhat.com \
--to=oleg@redhat.com \
--cc=a.p.zijlstra@chello.nl \
--cc=akpm@linux-foundation.org \
--cc=anton.vorontsov@linaro.org \
--cc=john.stultz@linaro.org \
--cc=kosaki.motohiro@jp.fujitsu.com \
--cc=lethal@linux-sh.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linux-sh@vger.kernel.org \
--cc=linux@arm.linux.org.uk \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=richard@nod.at \
--cc=uclinux-dist-devel@blackfin.uclinux.org \
--cc=user-mode-linux-devel@lists.sourceforge.net \
--cc=vapier@gentoo.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).