public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Tejun Heo <tj@kernel.org>
To: Vivek Goyal <vgoyal@redhat.com>
Cc: axboe@kernel.dk, ctalbott@google.com, rni@google.com,
	linux-kernel@vger.kernel.org, cgroups@vger.kernel.org,
	containers@lists.linux-foundation.org
Subject: Re: [PATCH 07/11] blkcg: make request_queue bypassing on allocation
Date: Fri, 13 Apr 2012 14:38:52 -0700	[thread overview]
Message-ID: <20120413213852.GJ12233@google.com> (raw)
In-Reply-To: <20120413213344.GA1825@redhat.com>

On Fri, Apr 13, 2012 at 05:33:44PM -0400, Vivek Goyal wrote:
> On Fri, Apr 13, 2012 at 02:05:48PM -0700, Tejun Heo wrote:
> > On Fri, Apr 13, 2012 at 04:55:01PM -0400, Vivek Goyal wrote:
> > > But neither seems to be the case here. So to make sure that blkg_lookup()
> > > under rcu will see the updated value of queue flag (bypass), are we
> > > relying on the fact that caller should see the DEAD flag and not go
> > > ahead with blkg_lookup()?  If yes, atleast it is not obivious.
> > 
> > We're relying on the fact that it doesn't matter anymore because all
> > blkgs will be shoot down in queue cleanup path which goes through rcu
> > free, which is different from deactivating individual policies.  It
> > indeed is subtle.  Umm... this is starting to get ridiculous.  Why the
> > hell was megaraid messing with so many queues anyways?
> 
> Well, blkcg_deactivate_policy() frees the policy data in a non-rcu
> manner. So group is around but policy data is gone. So technically if some
> IO submitter does not see the queue bypass flag, he might still try to
> access blkg->pd[pol->plid] after being freed.

No, we always go through blkg_destroy_all() and each blkg along with
any attached policy_data will go through RCU grace period before
getting destroyed.  It is stupid subtle but nevertheless correct.

> Having said that, in this case we are probably fine as blk_release_queue()
> is executed after last reference to queue is dropped and no more IO can
> come. May be a 2 line comment will help.

Yeah, we're guaranteed that by the time blk_release_queue() executes
nobody is traversing the queue.  Hmmm... right, this is much easier to
wrap one's head around.  I'll use this explanation in the comment.

> BTW, looks like blkio_exit_group_fn() probably is not a good name anymore
> as it is not even called when policy is being deactivated. It should
> probably be now .blkio_exit_policy_data_fn() or something like that.

Heh, I'm brewing mass blkcg API rename patch as we speak.

Thanks.

-- 
tejun

  reply	other threads:[~2012-04-13 21:38 UTC|newest]

Thread overview: 32+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-04-13 20:11 [PATCHSET] block: per-queue policy activation, take#2 Tejun Heo
2012-04-13 20:11 ` [PATCH 01/11] cfq: fix build breakage & warnings Tejun Heo
2012-04-13 20:11 ` [PATCH 02/11] blkcg: kill blkio_list and replace blkio_list_lock with a mutex Tejun Heo
2012-04-13 20:11 ` [PATCH 03/11] blkcg: use @pol instead of @plid in update_root_blkg_pd() and blkcg_print_blkgs() Tejun Heo
2012-04-13 20:11 ` [PATCH 04/11] blkcg: remove static policy ID enums Tejun Heo
2012-04-13 20:11 ` [PATCH 05/11] blkcg: make blkg_conf_prep() take @pol and return with queue lock held Tejun Heo
2012-04-13 20:11 ` [PATCH 06/11] blkcg: make sure blkg_lookup() returns %NULL if @q is bypassing Tejun Heo
2012-04-13 21:50   ` [PATCH UPDATED " Tejun Heo
2012-04-13 20:11 ` [PATCH 07/11] blkcg: make request_queue bypassing on allocation Tejun Heo
2012-04-13 20:32   ` Vivek Goyal
2012-04-13 20:37     ` Tejun Heo
2012-04-13 20:44       ` Vivek Goyal
2012-04-13 20:47         ` Tejun Heo
2012-04-13 20:55           ` Vivek Goyal
2012-04-13 21:05             ` Tejun Heo
2012-04-13 21:16               ` Tejun Heo
2012-04-13 21:31                 ` Tejun Heo
2012-04-17 12:04                 ` James Bottomley
2012-04-18 21:42                   ` Tejun Heo
2012-04-13 21:33               ` Vivek Goyal
2012-04-13 21:38                 ` Tejun Heo [this message]
2012-04-16 12:39                   ` Vivek Goyal
2012-04-13 20:37     ` Vivek Goyal
2012-04-13 20:11 ` [PATCH 08/11] blkcg: add request_queue->root_blkg Tejun Heo
2012-04-13 20:11 ` [PATCH 09/11] blkcg: implement per-queue policy activation Tejun Heo
2012-04-13 20:11 ` [PATCH 10/11] blkcg: drop stuff unused after per-queue policy activation update Tejun Heo
2012-04-13 20:11 ` [PATCH 11/11] blkcg: shoot down blkgs if all policies are deactivated Tejun Heo
2012-04-20  8:09 ` [PATCHSET] block: per-queue policy activation, take#2 Jens Axboe
2012-04-20 12:02   ` Jens Axboe
2012-04-20 17:17     ` Tejun Heo
2012-04-20 19:08       ` Jens Axboe
2012-04-25 18:19         ` Tejun Heo

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=20120413213852.GJ12233@google.com \
    --to=tj@kernel.org \
    --cc=axboe@kernel.dk \
    --cc=cgroups@vger.kernel.org \
    --cc=containers@lists.linux-foundation.org \
    --cc=ctalbott@google.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=rni@google.com \
    --cc=vgoyal@redhat.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