Message-ID: <cacc6eee-9b7b-00d2-d573-a303b3bb57b8@huawei.com>
Date:   Sat, 25 Jun 2022 16:10:36 +0800
From:   Yu Kuai <yukuai3@...wei.com>
To:     Paolo Valente <paolo.valente@...aro.org>
CC:     Jan Kara <jack@...e.cz>, <cgroups@...r.kernel.org>,
        linux-block <linux-block@...r.kernel.org>,
        Tejun Heo <tj@...nel.org>, Jens Axboe <axboe@...nel.dk>,
        <linux-kernel@...r.kernel.org>, <yi.zhang@...wei.com>
Subject: Re: [PATCH -next v10 3/4] block, bfq: refactor the counting of
 'num_groups_with_pending_reqs'

在 2022/06/24 9:26, Yu Kuai 写道:
> 在 2022/06/23 23:32, Paolo Valente 写道:
>> Sorry for the delay.
>>
>>> Il giorno 10 giu 2022, alle ore 04:17, Yu Kuai <yukuai3@...wei.com> 
>>> ha scritto:
>>>
>>> Currently, bfq can't handle sync io concurrently as long as they
>>> are not issued from root group. This is because
>>> 'bfqd->num_groups_with_pending_reqs > 0' is always true in
>>> bfq_asymmetric_scenario().
>>>
>>> The way that bfqg is counted into 'num_groups_with_pending_reqs':
>>>
>>> Before this patch:
>>> 1) root group will never be counted.
>>> 2) Count if bfqg or it's child bfqgs have pending requests.
>>> 3) Don't count if bfqg and it's child bfqgs complete all the requests.
>>>
>>> After this patch:
>>> 1) root group is counted.
>>> 2) Count if bfqg have pending requests.
>>> 3) Don't count if bfqg complete all the requests.
>>>
>>> With this change, the occasion that only one group is activated can be
>>> detected, and next patch will support concurrent sync io in the
>>> occasion.
>>>
>>> Signed-off-by: Yu Kuai <yukuai3@...wei.com>
>>> Reviewed-by: Jan Kara <jack@...e.cz>
>>> ---
>>> block/bfq-iosched.c | 42 ------------------------------------------
>>> block/bfq-iosched.h | 18 +++++++++---------
>>> block/bfq-wf2q.c    | 19 ++++---------------
>>> 3 files changed, 13 insertions(+), 66 deletions(-)
>>>
>>> diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c
>>> index 0ec21018daba..03b04892440c 100644
>>> --- a/block/bfq-iosched.c
>>> +++ b/block/bfq-iosched.c
>>> @@ -970,48 +970,6 @@ void __bfq_weights_tree_remove(struct bfq_data 
>>> *bfqd,
>>> void bfq_weights_tree_remove(struct bfq_data *bfqd,
>>>                  struct bfq_queue *bfqq)
>>> {
>>> -    struct bfq_entity *entity = bfqq->entity.parent;
>>> -
>>> -    for_each_entity(entity) {
>>> -        struct bfq_sched_data *sd = entity->my_sched_data;
>>> -
>>> -        if (sd->next_in_service || sd->in_service_entity) {
>>> -            /*
>>> -             * entity is still active, because either
>>> -             * next_in_service or in_service_entity is not
>>> -             * NULL (see the comments on the definition of
>>> -             * next_in_service for details on why
>>> -             * in_service_entity must be checked too).
>>> -             *
>>> -             * As a consequence, its parent entities are
>>> -             * active as well, and thus this loop must
>>> -             * stop here.
>>> -             */
>>> -            break;
>>> -        }
>>> -
>>> -        /*
>>> -         * The decrement of num_groups_with_pending_reqs is
>>> -         * not performed immediately upon the deactivation of
>>> -         * entity, but it is delayed to when it also happens
>>> -         * that the first leaf descendant bfqq of entity gets
>>> -         * all its pending requests completed. The following
>>> -         * instructions perform this delayed decrement, if
>>> -         * needed. See the comments on
>>> -         * num_groups_with_pending_reqs for details.
>>> -         */
>>> -        if (entity->in_groups_with_pending_reqs) {
>>> -            entity->in_groups_with_pending_reqs = false;
>>> -            bfqd->num_groups_with_pending_reqs--;
>>> -        }
>>> -    }
>>
>> With this part removed, I'm missing how you handle the following
>> sequence of events:
>> 1.  a queue Q becomes non busy but still has dispatched requests, so
>> it must not be removed from the counter of queues with pending reqs
>> yet
>> 2.  the last request of Q is completed with Q being still idle (non
>> busy).  At this point Q must be removed from the counter.  It seems to
>> me that this case is not handled any longer
>>
> Hi, Paolo
> 
> 1) At first, patch 1 support to track if bfqq has pending requests, it's
> done by setting the flag 'entity->in_groups_with_pending_reqs' when the
> first request is inserted to bfqq, and it's cleared when the last
> request is completed.
> 
> 2) Then, patch 2 add a counter in bfqg: how many bfqqs have pending
> requests, which is updated while tracking if bfqq has pending requests.
> 
> 3) Finally, patch 3 tracks 'num_groups_with_pending_reqs' based on the
> new counter in patch 2:
>   - if the counter(how many bfqqs have pending requests) increased from 0
>     to 0, increase 'num_groups_with_pending_reqs'.
Hi, Paolo

Sorry that I made a mistake here:
increased from 0 to 0 -> increased from 0 to 1.

look forward to your reply
Kuai
>   - if the counter is decreased from 1 to 0, decrease
>     'num_groups_with_pending_reqs'
> 
>> Additional comment: if your changes do not cpus the problem above,
>> then this function only invokes __bfq_weights_tree_remove.  So what's
>> the point in keeping this function)
> 
> If this patchset is applied, there are following cleanup patches to
> remove this function.
> 
> multiple cleanup patches for bfq:
> https://lore.kernel.org/all/20220528095958.270455-1-yukuai3@huawei.com/
>>
>>> -
>>> -    /*
>>> -     * Next function is invoked last, because it causes bfqq to be
>>> -     * freed if the following holds: bfqq is not in service and
>>> -     * has no dispatched request. DO NOT use bfqq after the next
>>> -     * function invocation.
>>> -     */
>>
>> I would really love it if you leave this comment.  I added it after
>> suffering a lot for a nasty UAF.  Of course the first sentence may
>> need to be adjusted if the code that precedes it is to be removed.
>>
> 
> Same as above, if this patch is applied, this function will be gone.
> 
> Thanks,
> Kuai
>> Thanks,
>> Paolo
>>
>>
>>>     __bfq_weights_tree_remove(bfqd, bfqq,
>>>                   &bfqd->queue_weights_tree);
>>> }
>>> diff --git a/block/bfq-iosched.h b/block/bfq-iosched.h
>>> index de2446a9b7ab..f0fce94583e4 100644
>>> --- a/block/bfq-iosched.h
>>> +++ b/block/bfq-iosched.h
>>> @@ -496,27 +496,27 @@ struct bfq_data {
>>>     struct rb_root_cached queue_weights_tree;
>>>
>>>     /*
>>> -     * Number of groups with at least one descendant process that
>>> +     * Number of groups with at least one process that
>>>      * has at least one request waiting for completion. Note that
>>>      * this accounts for also requests already dispatched, but not
>>>      * yet completed. Therefore this number of groups may differ
>>>      * (be larger) than the number of active groups, as a group is
>>>      * considered active only if its corresponding entity has
>>> -     * descendant queues with at least one request queued. This
>>> +     * queues with at least one request queued. This
>>>      * number is used to decide whether a scenario is symmetric.
>>>      * For a detailed explanation see comments on the computation
>>>      * of the variable asymmetric_scenario in the function
>>>      * bfq_better_to_idle().
>>>      *
>>>      * However, it is hard to compute this number exactly, for
>>> -     * groups with multiple descendant processes. Consider a group
>>> -     * that is inactive, i.e., that has no descendant process with
>>> +     * groups with multiple processes. Consider a group
>>> +     * that is inactive, i.e., that has no process with
>>>      * pending I/O inside BFQ queues. Then suppose that
>>>      * num_groups_with_pending_reqs is still accounting for this
>>> -     * group, because the group has descendant processes with some
>>> +     * group, because the group has processes with some
>>>      * I/O request still in flight. num_groups_with_pending_reqs
>>>      * should be decremented when the in-flight request of the
>>> -     * last descendant process is finally completed (assuming that
>>> +     * last process is finally completed (assuming that
>>>      * nothing else has changed for the group in the meantime, in
>>>      * terms of composition of the group and active/inactive state of 
>>> child
>>>      * groups and processes). To accomplish this, an additional
>>> @@ -525,7 +525,7 @@ struct bfq_data {
>>>      * we resort to the following tradeoff between simplicity and
>>>      * accuracy: for an inactive group that is still counted in
>>>      * num_groups_with_pending_reqs, we decrement
>>> -     * num_groups_with_pending_reqs when the first descendant
>>> +     * num_groups_with_pending_reqs when the first
>>>      * process of the group remains with no request waiting for
>>>      * completion.
>>>      *
>>> @@ -533,12 +533,12 @@ struct bfq_data {
>>>      * carefulness: to avoid multiple decrements, we flag a group,
>>>      * more precisely an entity representing a group, as still
>>>      * counted in num_groups_with_pending_reqs when it becomes
>>> -     * inactive. Then, when the first descendant queue of the
>>> +     * inactive. Then, when the first queue of the
>>>      * entity remains with no request waiting for completion,
>>>      * num_groups_with_pending_reqs is decremented, and this flag
>>>      * is reset. After this flag is reset for the entity,
>>>      * num_groups_with_pending_reqs won't be decremented any
>>> -     * longer in case a new descendant queue of the entity remains
>>> +     * longer in case a new queue of the entity remains
>>>      * with no request waiting for completion.
>>>      */
>>>     unsigned int num_groups_with_pending_reqs;
>>> diff --git a/block/bfq-wf2q.c b/block/bfq-wf2q.c
>>> index 6f36f3fe5cc8..9c2842bedf97 100644
>>> --- a/block/bfq-wf2q.c
>>> +++ b/block/bfq-wf2q.c
>>> @@ -984,19 +984,6 @@ static void __bfq_activate_entity(struct 
>>> bfq_entity *entity,
>>>         entity->on_st_or_in_serv = true;
>>>     }
>>>
>>> -#ifdef CONFIG_BFQ_GROUP_IOSCHED
>>> -    if (!bfq_entity_to_bfqq(entity)) { /* bfq_group */
>>> -        struct bfq_group *bfqg =
>>> -            container_of(entity, struct bfq_group, entity);
>>> -        struct bfq_data *bfqd = bfqg->bfqd;
>>> -
>>> -        if (!entity->in_groups_with_pending_reqs) {
>>> -            entity->in_groups_with_pending_reqs = true;
>>> -            bfqd->num_groups_with_pending_reqs++;
>>> -        }
>>> -    }
>>> -#endif
>>> -
>>>     bfq_update_fin_time_enqueue(entity, st, backshifted);
>>> }
>>>
>>> @@ -1654,7 +1641,8 @@ void 
>>> bfq_add_bfqq_in_groups_with_pending_reqs(struct bfq_queue *bfqq)
>>>     if (!entity->in_groups_with_pending_reqs) {
>>>         entity->in_groups_with_pending_reqs = true;
>>> #ifdef CONFIG_BFQ_GROUP_IOSCHED
>>> -        bfqq_group(bfqq)->num_queues_with_pending_reqs++;
>>> +        if (!(bfqq_group(bfqq)->num_queues_with_pending_reqs++))
>>> +            bfqq->bfqd->num_groups_with_pending_reqs++;
>>> #endif
>>>     }
>>> }
>>> @@ -1666,7 +1654,8 @@ void 
>>> bfq_del_bfqq_in_groups_with_pending_reqs(struct bfq_queue *bfqq)
>>>     if (entity->in_groups_with_pending_reqs) {
>>>         entity->in_groups_with_pending_reqs = false;
>>> #ifdef CONFIG_BFQ_GROUP_IOSCHED
>>> -        bfqq_group(bfqq)->num_queues_with_pending_reqs--;
>>> +        if (!(--bfqq_group(bfqq)->num_queues_with_pending_reqs))
>>> +            bfqq->bfqd->num_groups_with_pending_reqs--;
>>> #endif
>>>     }
>>> }
>>> -- 
>>> 2.31.1
>>>
>>
>> .
>>

Powered by blists - more mailing lists