linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
To: Vojtech Pavlik <vojtech@suse.cz>
Cc: Jiri Slaby <jslaby@suse.cz>, Tejun Heo <tj@kernel.org>,
	Jiri Kosina <jkosina@suse.cz>,
	linux-kernel@vger.kernel.org, jirislaby@gmail.com,
	Michael Matz <matz@suse.de>, Steven Rostedt <rostedt@goodmis.org>,
	Frederic Weisbecker <fweisbec@gmail.com>,
	Ingo Molnar <mingo@redhat.com>,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	"Theodore Ts'o" <tytso@mit.edu>,
	Dipankar Sarma <dipankar@in.ibm.com>
Subject: Re: [RFC 09/16] kgr: mark task_safe in some kthreads
Date: Wed, 14 May 2014 08:30:08 -0700	[thread overview]
Message-ID: <20140514153008.GD4570@linux.vnet.ibm.com> (raw)
In-Reply-To: <20140514151501.GA24142@suse.cz>

On Wed, May 14, 2014 at 05:15:01PM +0200, Vojtech Pavlik wrote:
> On Wed, May 14, 2014 at 04:59:05PM +0200, Jiri Slaby wrote:
> 
> > I see the worst case scenario. (For curious readers, it is for example
> > this kthread body:
> > while (1) {
> >   some_paired_call(); /* invokes pre-patched code */
> >   if (kthread_should_stop()) { /* kgraft switches to the new code */
> >     its_paired_function(); /* invokes patched code (wrong) */
> >     break;
> >   }
> >   its_paired_function(); /* the same (wrong) */
> > })
> > 
> > What to do with that now? We have come up with a couple possibilities.
> > Would you consider try_to_freeze() a good state-defining function? As it
> > is called when a kthread expects weird things can happen, it should be
> > safe to switch to the patched version in our opinion.
> > 
> > The other possibility is to patch every kthread loop (~300) and insert
> > kgr_task_safe() semi-manually at some proper place.
> > 
> > Or if you have any other suggestions we would appreciate that?
> 
> A heretic idea would be to convert all kernel threads into functions
> that do not sleep and exit after a single iteration and are called from
> a central kthread main loop function. That would get all of
> kthread_should_stop() and try_to_freeze() and kgr_task_safe() nicely
> into one place and at the same time put enough constraint on what the
> thread function can do to prevent it from breaking the assumptions of
> each of these calls. 

Some substantial restructuring would be required for several of
the kthreads I am aware of, which contain kthread_should_stop()
inside loop bodies as well as on their conditions.  Also, a number
of them do things like wait_event() and the like, which would mean
that the central kthread main loop function would need to know
about the wait queues and wait conditions and handle them properly.
See for example rcu_torture_barrier_cbs() and rcu_torture_barrier()
in kernel/rcu/rcutorture.c [*], which wait on each other in order to
test RCU's rcu_barrier() primitives.

							Thanx, Paul

* In older kernels, this is kernel/rcu/torture.c or kernel/rcutorture.c.


  reply	other threads:[~2014-05-14 15:30 UTC|newest]

Thread overview: 59+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-04-30 14:30 [RFC 00/16] kGraft Jiri Slaby
2014-04-30 14:30 ` [RFC 01/16] ftrace: Add function to find fentry of function Jiri Slaby
2014-04-30 14:48   ` Steven Rostedt
2014-04-30 14:58     ` Jiri Slaby
2014-04-30 14:30 ` [RFC 02/16] ftrace: Make ftrace_is_dead available globally Jiri Slaby
2014-04-30 14:30 ` [RFC 03/16] kgr: initial code Jiri Slaby
2014-04-30 14:56   ` Steven Rostedt
2014-04-30 14:57     ` Jiri Slaby
2014-05-01 20:20   ` Andi Kleen
2014-05-01 20:37     ` Jiri Kosina
2014-05-14  9:28   ` Aravinda Prasad
2014-05-14 10:12     ` Jiri Slaby
2014-05-14 10:41       ` Aravinda Prasad
2014-05-14 10:44         ` Jiri Slaby
2014-05-14 11:19           ` Aravinda Prasad
2014-05-20 11:36     ` Jiri Slaby
2014-05-21 18:28       ` Aravinda Prasad
2014-05-26  8:50       ` Jiri Kosina
2014-04-30 14:30 ` [RFC 04/16] kgr: add testing kgraft patch Jiri Slaby
2014-05-06 11:03   ` Pavel Machek
2014-05-12 12:50     ` Jiri Slaby
2014-04-30 14:30 ` [RFC 05/16] kgr: update Kconfig documentation Jiri Slaby
2014-05-03 14:32   ` Randy Dunlap
2014-04-30 14:30 ` [RFC 06/16] kgr: add Documentation Jiri Slaby
2014-05-06 11:03   ` Pavel Machek
2014-05-09  9:31     ` kgr: dealing with optimalizations? (was Re: [RFC 06/16] kgr: add Documentat)ion Pavel Machek
2014-05-09 12:22       ` Michael Matz
2014-04-30 14:30 ` [RFC 07/16] kgr: trigger the first check earlier Jiri Slaby
2014-04-30 14:30 ` [RFC 08/16] kgr: sched.h, introduce kgr_task_safe helper Jiri Slaby
2014-04-30 14:30 ` [RFC 09/16] kgr: mark task_safe in some kthreads Jiri Slaby
2014-04-30 15:49   ` Greg Kroah-Hartman
2014-04-30 16:55   ` Paul E. McKenney
2014-04-30 18:33     ` Vojtech Pavlik
2014-04-30 19:07       ` Paul E. McKenney
2014-05-01 14:24   ` Tejun Heo
2014-05-01 20:17     ` Jiri Kosina
2014-05-01 21:02       ` Tejun Heo
2014-05-01 21:09         ` Tejun Heo
2014-05-14 14:59           ` Jiri Slaby
2014-05-14 15:15             ` Vojtech Pavlik
2014-05-14 15:30               ` Paul E. McKenney [this message]
2014-05-14 16:32               ` Tejun Heo
2014-05-15  3:53                 ` Mike Galbraith
2014-05-15  4:06                   ` Tejun Heo
2014-05-15  4:46                     ` Mike Galbraith
2014-05-15  4:50                       ` Tejun Heo
2014-05-15  5:04                         ` Mike Galbraith
2014-05-15  5:09                           ` Tejun Heo
2014-05-15  5:32                             ` Mike Galbraith
2014-05-15  6:05                               ` Tejun Heo
2014-05-15  6:32                                 ` Mike Galbraith
2014-04-30 14:30 ` [RFC 10/16] kgr: kthreads support Jiri Slaby
2014-04-30 14:30 ` [RFC 11/16] kgr: handle irqs Jiri Slaby
2014-04-30 14:30 ` [RFC 12/16] kgr: add tools Jiri Slaby
2014-05-06 11:03   ` Pavel Machek
2014-04-30 14:30 ` [RFC 13/16] kgr: add MAINTAINERS entry Jiri Slaby
2014-04-30 14:30 ` [RFC 14/16] kgr: x86: refuse to build without fentry support Jiri Slaby
2014-04-30 14:30 ` [RFC 15/16] kgr: add procfs interface for per-process 'kgr_in_progress' Jiri Slaby
2014-04-30 14:30 ` [RFC 16/16] kgr: make a per-process 'in progress' flag a single bit Jiri Slaby

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=20140514153008.GD4570@linux.vnet.ibm.com \
    --to=paulmck@linux.vnet.ibm.com \
    --cc=dipankar@in.ibm.com \
    --cc=fweisbec@gmail.com \
    --cc=gregkh@linuxfoundation.org \
    --cc=jirislaby@gmail.com \
    --cc=jkosina@suse.cz \
    --cc=jslaby@suse.cz \
    --cc=linux-kernel@vger.kernel.org \
    --cc=matz@suse.de \
    --cc=mingo@redhat.com \
    --cc=rostedt@goodmis.org \
    --cc=tj@kernel.org \
    --cc=tytso@mit.edu \
    --cc=vojtech@suse.cz \
    /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).