From: Max Gurtovoy <[email protected]>
To: Jens Axboe <[email protected]>, Christoph Hellwig <[email protected]>
Cc: <[email protected]>, <[email protected]>,
"Hannes Reinecke" <[email protected]>
Subject: Re: [PATCH 4/4] nvme: add support for mq_ops->queue_rqs()
Date: Thu, 16 Dec 2021 18:34:40 +0200 [thread overview]
Message-ID: <[email protected]> (raw)
In-Reply-To: <[email protected]>
On 12/16/2021 6:25 PM, Jens Axboe wrote:
> On 12/16/21 9:19 AM, Max Gurtovoy wrote:
>> On 12/16/2021 6:05 PM, Jens Axboe wrote:
>>> On 12/16/21 9:00 AM, Max Gurtovoy wrote:
>>>> On 12/16/2021 5:48 PM, Jens Axboe wrote:
>>>>> On 12/16/21 6:06 AM, Max Gurtovoy wrote:
>>>>>> On 12/16/2021 11:08 AM, Christoph Hellwig wrote:
>>>>>>> On Wed, Dec 15, 2021 at 09:24:21AM -0700, Jens Axboe wrote:
>>>>>>>> + spin_lock(&nvmeq->sq_lock);
>>>>>>>> + while (!rq_list_empty(*rqlist)) {
>>>>>>>> + struct request *req = rq_list_pop(rqlist);
>>>>>>>> + struct nvme_iod *iod = blk_mq_rq_to_pdu(req);
>>>>>>>> +
>>>>>>>> + memcpy(nvmeq->sq_cmds + (nvmeq->sq_tail << nvmeq->sqes),
>>>>>>>> + absolute_pointer(&iod->cmd), sizeof(iod->cmd));
>>>>>>>> + if (++nvmeq->sq_tail == nvmeq->q_depth)
>>>>>>>> + nvmeq->sq_tail = 0;
>>>>>>> So this doesn't even use the new helper added in patch 2? I think this
>>>>>>> should call nvme_sq_copy_cmd().
>>>>>> I also noticed that.
>>>>>>
>>>>>> So need to decide if to open code it or use the helper function.
>>>>>>
>>>>>> Inline helper sounds reasonable if you have 3 places that will use it.
>>>>> Yes agree, that's been my stance too :-)
>>>>>
>>>>>>> The rest looks identical to the incremental patch I posted, so I guess
>>>>>>> the performance degration measured on the first try was a measurement
>>>>>>> error?
>>>>>> giving 1 dbr for a batch of N commands sounds good idea. Also for RDMA host.
>>>>>>
>>>>>> But how do you moderate it ? what is the batch_sz <--> time_to_wait
>>>>>> algorithm ?
>>>>> The batching is naturally limited at BLK_MAX_REQUEST_COUNT, which is 32
>>>>> in total. I do agree that if we ever made it much larger, then we might
>>>>> want to cap it differently. But 32 seems like a pretty reasonable number
>>>>> to get enough gain from the batching done in various areas, while still
>>>>> not making it so large that we have a potential latency issue. That
>>>>> batch count is already used consistently for other items too (like tag
>>>>> allocation), so it's not specific to just this one case.
>>>> I'm saying that the you can wait to the batch_max_count too long and it
>>>> won't be efficient from latency POV.
>>>>
>>>> So it's better to limit the block layar to wait for the first to come: x
>>>> usecs or batch_max_count before issue queue_rqs.
>>> There's no waiting specifically for this, it's just based on the plug.
>>> We just won't do more than 32 in that plug. This is really just an
>>> artifact of the plugging, and if that should be limited based on "max of
>>> 32 or xx time", then that should be done there.
>>>
>>> But in general I think it's saner and enough to just limit the total
>>> size. If we spend more than xx usec building up the plug list, we're
>>> doing something horribly wrong. That really should not happen with 32
>>> requests, and we'll never eg wait on requests if we're out of tags. That
>>> will result in a plug flush to begin with.
>> I'm not aware of the plug. I hope to get to it soon.
>>
>> My concern is if the user application submitted only 28 requests and
>> then you'll wait forever ? or for very long time.
>>
>> I guess not, but I'm asking how do you know how to batch and when to
>> stop in case 32 commands won't arrive anytime soon.
> The plug is in the stack of the task, so that condition can never
> happen. If the application originally asks for 32 but then only submits
> 28, then once that last one is submitted the plug is flushed and
> requests are issued.
So if I'm running fio with --iodepth=28 what will plug do ? send batches
of 28 ? or 1 by 1 ?
>>>> Also, This batch is per HW queue or SW queue or the entire request queue ?
>>> It's per submitter, so whatever the submitter ends up queueing IO
>>> against. In general it'll be per-queue.
>> struct request_queue ?
>>
>> I think the best is to batch per struct blk_mq_hw_ctx.
>>
>> I see that you check this in the nvme_pci driver but shouldn't it go to
>> the block layer ?
> That's not how plugging works. In general, unless your task bounces
> around, then it'll be a single queue and a single hw queue as well.
> Adding code to specifically check the mappings and flush at that point
> would be a net loss, rather than just deal with it if it happens for
> some cases.
>
next prev parent reply other threads:[~2021-12-16 16:35 UTC|newest]
Thread overview: 50+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-12-15 16:24 [PATCHSET v3 0/4] Add support for list issue Jens Axboe
2021-12-15 16:24 ` [PATCH 1/4] block: add mq_ops->queue_rqs hook Jens Axboe
2021-12-16 9:01 ` Christoph Hellwig
2021-12-20 20:36 ` Keith Busch
2021-12-20 20:47 ` Jens Axboe
2021-12-15 16:24 ` [PATCH 2/4] nvme: split command copy into a helper Jens Axboe
2021-12-16 9:01 ` Christoph Hellwig
2021-12-16 12:17 ` Max Gurtovoy
2021-12-15 16:24 ` [PATCH 3/4] nvme: separate command prep and issue Jens Axboe
2021-12-16 9:02 ` Christoph Hellwig
2021-12-15 16:24 ` [PATCH 4/4] nvme: add support for mq_ops->queue_rqs() Jens Axboe
2021-12-15 17:29 ` Keith Busch
2021-12-15 20:27 ` Jens Axboe
2021-12-16 9:08 ` Christoph Hellwig
2021-12-16 13:06 ` Max Gurtovoy
2021-12-16 15:48 ` Jens Axboe
2021-12-16 16:00 ` Max Gurtovoy
2021-12-16 16:05 ` Jens Axboe
2021-12-16 16:19 ` Max Gurtovoy
2021-12-16 16:25 ` Jens Axboe
2021-12-16 16:34 ` Max Gurtovoy [this message]
2021-12-16 16:36 ` Jens Axboe
2021-12-16 16:57 ` Max Gurtovoy
2021-12-16 17:16 ` Jens Axboe
2021-12-19 12:14 ` Max Gurtovoy
2021-12-19 14:48 ` Jens Axboe
2021-12-20 10:11 ` Max Gurtovoy
2021-12-20 14:19 ` Jens Axboe
2021-12-20 14:25 ` Jens Axboe
2021-12-20 15:29 ` Max Gurtovoy
2021-12-20 16:34 ` Jens Axboe
2021-12-20 18:48 ` Max Gurtovoy
2021-12-20 18:58 ` Jens Axboe
2021-12-21 10:20 ` Max Gurtovoy
2021-12-21 15:23 ` Jens Axboe
2021-12-21 15:29 ` Max Gurtovoy
2021-12-21 15:33 ` Jens Axboe
2021-12-21 16:08 ` Max Gurtovoy
2021-12-16 15:45 ` Jens Axboe
2021-12-16 16:15 ` Christoph Hellwig
2021-12-16 16:27 ` Jens Axboe
2021-12-16 16:30 ` Christoph Hellwig
2021-12-16 16:36 ` Jens Axboe
2021-12-16 13:02 ` Max Gurtovoy
2021-12-16 15:59 ` Jens Axboe
2021-12-16 16:06 ` Max Gurtovoy
2021-12-16 16:09 ` Jens Axboe
-- strict thread matches above, loose matches on Subject: below --
2021-12-16 16:05 [PATCHSET v4 0/4] Add support for list issue Jens Axboe
2021-12-16 16:05 ` [PATCH 4/4] nvme: add support for mq_ops->queue_rqs() Jens Axboe
2021-12-16 16:38 [PATCHSET v5 0/4] Add support for list issue Jens Axboe
2021-12-16 16:39 ` [PATCH 4/4] nvme: add support for mq_ops->queue_rqs() Jens Axboe
2021-12-16 17:53 ` Christoph Hellwig
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 \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
/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