mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Trond Myklebust <trond.myklebust@fys.uio.no>
To: john cooper <john.cooper@timesys.com>
Cc: Oleg Nesterov <oleg@tv-sign.ru>,
	linux-kernel@vger.kernel.org, Ingo Molnar <mingo@elte.hu>,
	Olaf Kirch <okir@suse.de>
Subject: Re: RT and Cascade interrupts
Date: Wed, 01 Jun 2005 14:31:58 -0400	[thread overview]
Message-ID: <1117650718.10733.65.camel@lade.trondhjem.org> (raw)
In-Reply-To: <429DF8DE.7060008@timesys.com>

on den 01.06.2005 Klokka 14:05 (-0400) skreiv john cooper:
> I've done so and the second assert is generated
> when running the test.  So here we have a case
> of RPC_TASK_HAS_TIMER set but the associated
> timer->base == NULL.  It would seem this could
> easily be the scenario of executing in:
> 
> __run_timers()
>      timer->base = NULL;
>          rpc_run_timer()
>              task->tk_timeout_fn(task)
>              /* ksoftirqd preempted */
>                                               :
>                                           /* RPC client */
>                                           rpc_execute()
>                                              rpc_delete_timer()
>                                                  del_timer() returns 0
>                                                  BUG_ON(test_bit(RPC_TASK_HAS_TIMER,
>                                                    &task->tk_runstate));
>                                               :
>           /* rpc_run_timer() resumes */
>           clear_bit(RPC_TASK_HAS_TIMER,
>             &task->tk_runstate);
> 
> I don't see how this would imply a kernel/timer.c
> problem.  It also appears this wouldn't be causing
> the timer cascade corruption I've seen as the
> end result is deleting an already dequeued timer
> which is safe here.

If timer->base==NULL, then del_timer() returns 0 (as you correctly state
above). What you appear to be missing is that this will trigger a call
to del_timer_sync() inside del_singleshot_timer_sync().

In other words, your scenario above should, if the timer code is working
according to spec instead lead to

__run_timers()
     timer->base = NULL;
         rpc_run_timer()
             task->tk_timeout_fn(task)
             /* ksoftirqd preempted */
                                              :
                                       /* RPC client */
                                       rpc_execute()
                                          rpc_delete_timer()
					     del_singleshot_timer_sync()
                                                 del_timer() returns 0
						 del_timer_sync()
						/* wait until timer has
						   completed */
          /*rpc_run_timer() resumes */
          clear_bit(RPC_TASK_HAS_TIMER,
            &task->tk_runstate);
         ....
     set_running_timer(some_other_value);

				    BUG_ON(test_bit(RPC_TASK_HAS_TIMER,
                                                   &task->tk_runstate));
                                              :

Cheers,
  Trond


  reply	other threads:[~2005-06-01 18:50 UTC|newest]

Thread overview: 34+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2005-05-27 16:47 Oleg Nesterov
2005-05-27 23:37 ` john cooper
2005-05-28  8:52   ` Oleg Nesterov
2005-05-28 14:02     ` john cooper
2005-05-28 16:34       ` Oleg Nesterov
2005-05-28 17:48     ` john cooper
2005-05-28 20:35       ` Trond Myklebust
2005-05-29  3:12         ` john cooper
2005-05-29  7:40           ` Trond Myklebust
2005-05-30 21:32             ` john cooper
2005-05-31 23:09               ` john cooper
2005-06-01 14:22               ` Oleg Nesterov
2005-06-01 18:05                 ` john cooper
2005-06-01 18:31                   ` Trond Myklebust [this message]
2005-06-01 19:20                     ` john cooper
2005-06-01 19:46                       ` Trond Myklebust
2005-06-01 20:21                       ` Trond Myklebust
2005-06-01 20:59                         ` john cooper
2005-06-01 22:51                           ` Trond Myklebust
2005-06-01 23:09                             ` Trond Myklebust
2005-06-02  3:31                             ` john cooper
2005-06-02  4:26                               ` Trond Myklebust
2005-06-09 23:17                                 ` George Anzinger
2005-06-09 23:52                                   ` john cooper
2005-05-29 11:31         ` Oleg Nesterov
2005-05-29 13:58           ` Trond Myklebust
2005-05-30 14:50             ` Ingo Molnar
2005-05-28 22:17       ` Trond Myklebust
  -- strict thread matches above, loose matches on Subject: below --
2005-05-12 14:43 Daniel Walker
2005-05-13  7:44 ` Ingo Molnar
2005-05-13 13:12   ` john cooper
2005-05-24 16:32     ` john cooper
2005-05-27  7:25       ` Ingo Molnar
2005-05-27 13:53         ` john cooper

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=1117650718.10733.65.camel@lade.trondhjem.org \
    --to=trond.myklebust@fys.uio.no \
    --cc=john.cooper@timesys.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=okir@suse.de \
    --cc=oleg@tv-sign.ru \
    /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

all inboxes | Powered by JetHome®