qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Avi Kivity <avi@redhat.com>
To: Paolo Bonzini <pbonzini@redhat.com>
Cc: Jan Kiszka <jan.kiszka@siemens.com>,
	Marcelo Tosatti <mtosatti@redhat.com>,
	liu ping fan <qemulist@gmail.com>,
	Anthony Liguori <anthony@codemonkey.ws>,
	qemu-devel@nongnu.org
Subject: Re: [Qemu-devel] [big lock] Discussion about the convention of device's DMA each other after breaking down biglock
Date: Wed, 19 Sep 2012 13:06:52 +0300	[thread overview]
Message-ID: <5059993C.1070506@redhat.com> (raw)
In-Reply-To: <505995B7.4010709@redhat.com>

On 09/19/2012 12:51 PM, Paolo Bonzini wrote:
> Il 19/09/2012 11:21, Avi Kivity ha scritto:
>>> > I don't know if the front-end (device) lock should come before or after
>>> > the back-end (e.g. netdev) lock in the hierarchy, but that's another story.

>> I would say device -> backend.  It's natural if the backend is the timer
>> subsystem, so extend it from there to the block and network layers.  Of
>> course callbacks want it to be the other way round.
> 
> Yes, that's what I wasn't sure about.  In many cases I believe callbacks
> can just release the backend lock.  It works for timers, for example:
> 
>     for(;;) {
>         ts = clock->active_timers;
>         if (!qemu_timer_expired_ns(ts, current_time)) {
>             break;
>         }
>         /* remove timer from the list before calling the callback */
>         clock->active_timers = ts->next;
>         ts->next = NULL;
> 
>         /* run the callback (the timer list can be modified) */
> -       ts->cb(ts->opaque);
> +       cb = ts->cb;
> +       opaque = ts->opaque;
> +       unlock();
> +       cb(opaque);
> +       lock();
>     }
> 
> (The hunch is that ts could be deleted exactly at the moment the
> callback is unlocked.  This can be solved with ref/unref on the opaque
> value, as you mention below).

Are you saying that this works as is or not?  It does seem broken wrt
deletion; after qemu_del_timer() completes the caller expects the
callback not to be called.

This isn't trivial to guarantee, we need something like

   dispatch_timer():
      pending += 1
      timer.ref()
      drop lock
      timer.cb()
      take lock
      timer.unref()
      pending -= 1
      notify

   del_timer():
      take lock
      timer.unlink()
      while pending:
          wait for notification
      drop lock

but, if del_timer is called with the device lock held, we deadlock. ugh.

-- 
error compiling committee.c: too many arguments to function

  reply	other threads:[~2012-09-19 10:07 UTC|newest]

Thread overview: 45+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-09-19  3:02 [Qemu-devel] [big lock] Discussion about the convention of device's DMA each other after breaking down biglock liu ping fan
2012-09-19  8:06 ` Avi Kivity
2012-09-19  9:00   ` liu ping fan
2012-09-19  9:07     ` Avi Kivity
2012-09-19  9:11       ` liu ping fan
2012-09-19  9:14         ` Paolo Bonzini
2012-09-19  9:19           ` liu ping fan
2012-09-19  9:23             ` Avi Kivity
2012-09-19  9:27               ` Jan Kiszka
2012-09-19  9:28                 ` Jan Kiszka
2012-09-20  7:51               ` liu ping fan
2012-09-20  7:54                 ` Paolo Bonzini
2012-09-20  8:09                   ` liu ping fan
2012-09-20  8:27                     ` Paolo Bonzini
2012-09-20  9:07                 ` Avi Kivity
2012-09-21  7:27                   ` liu ping fan
2012-09-21  8:21                     ` Paolo Bonzini
2012-09-19  9:21           ` Avi Kivity
2012-09-19  9:51             ` Paolo Bonzini
2012-09-19 10:06               ` Avi Kivity [this message]
2012-09-19 10:19                 ` Paolo Bonzini
2012-09-19 10:27                   ` Avi Kivity
2012-09-19  9:34     ` Jan Kiszka
2012-09-19  9:50       ` Avi Kivity
2012-09-19 10:18         ` Jan Kiszka
2012-09-24  6:33         ` liu ping fan
2012-09-24  7:44           ` Avi Kivity
2012-09-24  8:32             ` liu ping fan
2012-09-24  9:42               ` Avi Kivity
2012-09-27  3:13                 ` liu ping fan
2012-09-27  9:16                   ` Avi Kivity
2012-09-27  9:29                     ` Paolo Bonzini
2012-09-27  9:34                       ` Avi Kivity
2012-09-27  9:36                         ` Paolo Bonzini
2012-09-27 10:08                           ` Avi Kivity
2012-09-27 10:22                             ` Paolo Bonzini
2012-09-27 10:48                               ` Avi Kivity
2012-09-29  9:20                     ` liu ping fan
2012-09-30  8:13                       ` Avi Kivity
2012-09-30  8:48                         ` liu ping fan
2012-09-30 11:18                           ` Avi Kivity
2012-09-30 11:04                         ` Blue Swirl
2012-09-30 11:17                           ` Avi Kivity
2012-09-30 11:48                             ` Blue Swirl
2012-09-20  8:11       ` liu ping fan

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=5059993C.1070506@redhat.com \
    --to=avi@redhat.com \
    --cc=anthony@codemonkey.ws \
    --cc=jan.kiszka@siemens.com \
    --cc=mtosatti@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=qemulist@gmail.com \
    /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).