linux-kernel - Re: [PATCH BUGFIX 3/6] block, bfq: get extra ref to prevent a queue from being freed during a group move

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Fri, 31 Jan 2020 11:20:59 +0100
From:   Oleksandr Natalenko <oleksandr@...alenko.name>
To:     Paolo Valente <paolo.valente@...aro.org>
Cc:     Jens Axboe <axboe@...nel.dk>, linux-block@...r.kernel.org,
        linux-kernel@...r.kernel.org, bfq-iosched@...glegroups.com,
        patdung100@...il.com, cevich@...hat.com
Subject: Re: [PATCH BUGFIX 3/6] block, bfq: get extra ref to prevent a queue
 from being freed during a group move

Hello.

On 31.01.2020 10:24, Paolo Valente wrote:
> In bfq_bfqq_move(), the bfq_queue, say Q, to be moved to a new group
> may happen to be deactivated in the scheduling data structures of the
> source group (and then activated in the destination group). If Q is
> referred only by the data structures in the source group when the
> deactivation happens, then Q is freed upon the deactivation.
> 
> This commit addresses this issue by getting an extra reference before
> the possible deactivation, and releasing this extra reference after Q
> has been moved.
> 
> Tested-by: Chris Evich <cevich@...hat.com>
> Signed-off-by: Paolo Valente <paolo.valente@...aro.org>
> ---
>  block/bfq-cgroup.c | 8 ++++++++
>  1 file changed, 8 insertions(+)
> 
> diff --git a/block/bfq-cgroup.c b/block/bfq-cgroup.c
> index e1419edde2ec..8ab7f18ff8cb 100644
> --- a/block/bfq-cgroup.c
> +++ b/block/bfq-cgroup.c
> @@ -651,6 +651,12 @@ void bfq_bfqq_move(struct bfq_data *bfqd, struct
> bfq_queue *bfqq,
>  		bfq_bfqq_expire(bfqd, bfqd->in_service_queue,
>  				false, BFQQE_PREEMPTED);
> 
> +	/*
> +	 * get extra reference to prevent bfqq from being freed in
> +	 * next possible deactivate
> +	 */
> +	bfqq->ref++;

Shouldn't this be hidden under some macro (bfq_get_queue_ref(), for 
instance) and also converted from int into refcount_t?

> +
>  	if (bfq_bfqq_busy(bfqq))
>  		bfq_deactivate_bfqq(bfqd, bfqq, false, false);
>  	else if (entity->on_st)
> @@ -670,6 +676,8 @@ void bfq_bfqq_move(struct bfq_data *bfqd, struct
> bfq_queue *bfqq,
> 
>  	if (!bfqd->in_service_queue && !bfqd->rq_in_driver)
>  		bfq_schedule_dispatch(bfqd);
> +	/* release extra ref taken above */
> +	bfq_put_queue(bfqq);
>  }
> 
>  /**

-- 
   Oleksandr Natalenko (post-factum)