From: Oleg Nesterov <oleg@redhat.com>
To: Alan Stern <stern@rowland.harvard.edu>
Cc: Peter Zijlstra <peterz@infradead.org>,
Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
linux-kernel@vger.kernel.org, Ingo Molnar <mingo@kernel.org>,
Thomas Gleixner <tglx@linutronix.de>,
Steven Rostedt <rostedt@goodmis.org>,
Paul McKenney <paulmck@linux.vnet.ibm.com>,
Linus Torvalds <torvalds@linux-foundation.org>
Subject: Re: check && lockdep_no_validate (Was: lockdep: Introduce wait-type checks)
Date: Fri, 17 Jan 2014 17:31:11 +0100 [thread overview]
Message-ID: <20140117163111.GA5764@redhat.com> (raw)
In-Reply-To: <Pine.LNX.4.44L0.1401161515190.1109-100000@iolanthe.rowland.org>
On 01/16, Alan Stern wrote:
>
> On Thu, 16 Jan 2014, Peter Zijlstra wrote:
>
> > On Thu, Jan 16, 2014 at 06:43:48PM +0100, Oleg Nesterov wrote:
>
> > > Perhaps we should change the meaning of lockdep_set_novalidate_class?
> > > (perhaps with rename). What do you think about the patch below?
> > >
> > > With this patch __lockdep_no_validate__ means "automatically nested",
> >
> > Yes, I suppose that might work, it would allow some validation.
>
> I haven't seen the patch, but I'm not so sure it will work. Suppose we
> have two devices, D1 and D2, and some other mutex, M. Then the locking
> pattern:
>
> lock(D1);
> lock(M);
> unlock(M);
> unlock(D1);
>
> generally should not conflict with:
>
> lock(M);
> lock(D2);
> unlock(D2);
> unlock(M);
Yes, sure. This change assumes that the only problem in drivers/base is
dev->parent->mutex / dev->mutex dependency. If the locking is even more
"broken" (wrt lockdep), we can't replace lockdep_set_novalidate_class()
with lockdep_set_auto_nested().
And, otoh, with this change lockdep can miss the real problems too, for
example:
func1(dev)
{
device_lock(dev->parent);
mutex_lock(MUTEX);
device_lock(dev);
...
}
func2(dev)
{
device_lock(dev);
mutex_lock(MUTEX);
...
}
lockdep will only notice dev -> MUTEX dependency.
I booted the kernel (under kvm) with this change and there is nothing
in dmesg, but of course this is not the real testing.
So do you think that dev->mutex should not be validated at all ?
Just in case... Of course, if we actually add auto_nested we should not
use a single class unless dev->mutex will be the only user.
Oleg.
next prev parent reply other threads:[~2014-01-17 16:31 UTC|newest]
Thread overview: 50+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-01-09 11:15 [RFC][PATCH] lockdep: Introduce wait-type checks Peter Zijlstra
2014-01-09 11:49 ` Peter Zijlstra
2014-01-09 16:31 ` Oleg Nesterov
2014-01-09 17:08 ` Peter Zijlstra
2014-01-09 17:54 ` check && lockdep_no_validate (Was: lockdep: Introduce wait-type checks) Oleg Nesterov
2014-01-12 20:58 ` Peter Zijlstra
2014-01-13 16:07 ` Oleg Nesterov
2014-01-16 17:43 ` Oleg Nesterov
2014-01-16 18:09 ` Peter Zijlstra
2014-01-16 20:26 ` Alan Stern
2014-01-17 16:31 ` Oleg Nesterov [this message]
2014-01-17 18:01 ` Alan Stern
2014-01-20 18:19 ` [PATCH 0/5] lockdep: (Was: check && lockdep_no_validate) Oleg Nesterov
2014-01-20 18:20 ` [PATCH 1/5] lockdep: make held_lock->check and "int check" argument bool Oleg Nesterov
2014-02-10 13:32 ` [tip:core/locking] lockdep: Make " tip-bot for Oleg Nesterov
2014-01-20 18:20 ` [PATCH 2/5] lockdep: don't create the wrong dependency on hlock->check == 0 Oleg Nesterov
2014-02-10 13:33 ` [tip:core/locking] lockdep: Don' t " tip-bot for Oleg Nesterov
2014-01-20 18:20 ` [PATCH 3/5] lockdep: change mark_held_locks() to check hlock->check instead of lockdep_no_validate Oleg Nesterov
2014-02-10 13:33 ` [tip:core/locking] lockdep: Change " tip-bot for Oleg Nesterov
2014-01-20 18:20 ` [PATCH 4/5] lockdep: change lockdep_set_novalidate_class() to use _and_name Oleg Nesterov
2014-02-10 13:33 ` [tip:core/locking] lockdep: Change " tip-bot for Oleg Nesterov
2014-01-20 18:20 ` [PATCH 5/5] lockdep: pack subclass/trylock/read/check into a single argument Oleg Nesterov
2014-01-21 14:10 ` Peter Zijlstra
2014-01-21 17:27 ` Oleg Nesterov
2014-01-21 17:35 ` Dave Jones
2014-01-21 18:43 ` Oleg Nesterov
2014-01-21 18:53 ` Steven Rostedt
2014-01-21 20:06 ` Oleg Nesterov
2014-01-21 19:39 ` uninline rcu_lock_acquire/etc ? Oleg Nesterov
2014-01-22 3:54 ` Paul E. McKenney
2014-01-22 18:31 ` Oleg Nesterov
2014-01-22 19:34 ` Paul E. McKenney
2014-01-22 19:39 ` Oleg Nesterov
2014-01-20 18:37 ` [PATCH 0/5] lockdep: (Was: check && lockdep_no_validate) Alan Stern
2014-01-20 18:54 ` Oleg Nesterov
2014-01-20 21:42 ` Alan Stern
2014-01-12 9:40 ` [RFC][PATCH] lockdep: Introduce wait-type checks Ingo Molnar
2014-01-12 17:45 ` [PATCH 0/1] lockdep: Kill held_lock->check and "int check" arg of __lock_acquire() Oleg Nesterov
2014-01-12 17:45 ` [PATCH 1/1] " Oleg Nesterov
2014-01-13 0:28 ` Dave Jones
2014-01-13 16:20 ` Oleg Nesterov
2014-01-13 17:06 ` Oleg Nesterov
2014-01-13 17:28 ` Peter Zijlstra
2014-01-13 18:52 ` Oleg Nesterov
2014-01-13 22:34 ` Paul E. McKenney
2014-01-12 20:00 ` [PATCH 0/1] " Peter Zijlstra
2014-01-13 18:35 ` Oleg Nesterov
2014-01-09 17:33 ` [RFC][PATCH] lockdep: Introduce wait-type checks Dave Jones
2014-01-09 22:12 ` Peter Zijlstra
2014-01-10 12:11 ` Peter Zijlstra
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=20140117163111.GA5764@redhat.com \
--to=oleg@redhat.com \
--cc=gregkh@linuxfoundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@kernel.org \
--cc=paulmck@linux.vnet.ibm.com \
--cc=peterz@infradead.org \
--cc=rostedt@goodmis.org \
--cc=stern@rowland.harvard.edu \
--cc=tglx@linutronix.de \
--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).