Linux-USB Archive on lore.kernel.org
 help / color / Atom feed
From: Damien Le Moal <Damien.LeMoal@wdc.com>
To: Ming Lei <ming.lei@redhat.com>
Cc: Alan Stern <stern@rowland.harvard.edu>,
	Andrea Vai <andrea.vai@unipv.it>,
	Johannes Thumshirn <jthumshirn@suse.de>,
	Jens Axboe <axboe@kernel.dk>,
	USB list <linux-usb@vger.kernel.org>,
	SCSI development list <linux-scsi@vger.kernel.org>,
	Himanshu Madhani <himanshu.madhani@cavium.com>,
	Hannes Reinecke <hare@suse.com>, Omar Sandoval <osandov@fb.com>,
	"Martin K. Petersen" <martin.petersen@oracle.com>,
	Greg KH <gregkh@linuxfoundation.org>,
	Hans Holmberg <Hans.Holmberg@wdc.com>
Subject: Re: Slow I/O on USB media after commit f664a3cc17b7d0a2bc3b3ab96181e1029b0ec0e6
Date: Thu, 19 Sep 2019 09:09:33 +0000
Message-ID: <BYAPR04MB58163D3C66FEA00A754CF7C7E7890@BYAPR04MB5816.namprd04.prod.outlook.com> (raw)
In-Reply-To: <20190919085555.GA4541@ming.t460p>

On 2019/09/19 10:56, Ming Lei wrote:
> On Thu, Sep 19, 2019 at 08:26:32AM +0000, Damien Le Moal wrote:
>> On 2019/09/18 18:30, Alan Stern wrote:
>>> On Wed, 18 Sep 2019, Andrea Vai wrote:
>>>
>>>>> Also, I wonder if the changing the size of the data transfers would
>>>>> make any difference.  This is easy to try; just write "64" to
>>>>> /sys/block/sd?/queue/max_sectors_kb (where the ? is the appropriate
>>>>> drive letter) after the drive is plugged in but before the test
>>>>> starts.
>>>>
>>>> ok, so I duplicated the tests above for the "64" case (it was
>>>> initially set as "120", if it is relevant to know), leading to 40 tests named as
>>>>
>>>> bad.mon.out_50000000_64_TIMESTAMP
>>>> bad.mon.out_50000000_non64_TIMESTAMP
>>>> good.mon.out_50000000_64_TIMESTAMP
>>>> good.mon.out_50000000_non64_TIMESTAMP
>>>>
>>>> where "64" denotes the ones done with that value in max_sectors_kb,
>>>> and "not64" the ones without it (as far as I can tell, it has been
>>>> always "120").
>>>>
>>>> So, we have 40 traces total. Each set of 10 trials is identified by
>>>> a text file, which contains the output log of the test script (and the
>>>> timestamps), also available in the download zipfile.
>>>>
>>>> Just to summarize here the times, they are respectively (number
>>>> expressed  in seconds):
>>>>
>>>> BAD:
>>>>   Logs: log_10trials_50MB_BAD_NonCanc_64.txt,
>>>> log_10trials_50MB_BAD_NonCanc_non64.txt
>>>>   64: 34, 34, 35, 39, 37, 32, 42, 44, 43, 40
>>>>   not64: 61, 71, 59, 71, 62, 75, 62, 70, 62, 68
>>>> GOOD:
>>>>   Logs: log_10trials_50MB_GOOD_NonCanc_64.txt,
>>>> log_10trials_50MB_GOOD_NonCanc_non64.txt
>>>>   64: 34, 32, 35, 34, 35, 33, 34, 33, 33, 33
>>>>   not64: 32, 30, 32, 31, 31, 30, 32, 30, 32, 31
>>>
>>> The improvement from using "64" with the bad kernel is quite large.  
>>> That alone would be a big help for you.
>>>
>>> However, I did see what appears to be a very significant difference 
>>> between the bad and good kernel traces.  It has to do with the order in 
>>> which the blocks are accessed.
>>>
>>> Here is an extract from one of the bad traces.  I have erased all the 
>>> information except for the columns containing the block numbers to be 
>>> written:
>>>
>>> 00019628 00
>>> 00019667 00
>>> 00019628 80
>>> 00019667 80
>>> 00019629 00
>>> 00019668 00
>>> 00019629 80
>>> 00019668 80
>>>
>>> Here is the equivalent portion from one of the good traces:
>>>
>>> 00019628 00
>>> 00019628 80
>>> 00019629 00
>>> 00019629 80
>>> 0001962a 00
>>> 0001962a 80
>>> 0001962b 00
>>> 0001962b 80
>>>
>>> Notice that under the good kernel, the block numbers increase
>>> monotonically in a single sequence.  But under the bad kernel, the
>>> block numbers are not monotonic -- it looks like there are two separate
>>> threads each with its own strictly increasing sequence.
>>>
>>> This is exactly the sort of difference one might expect to see from
>>> the commit f664a3cc17b7 ("scsi: kill off the legacy IO path") you
>>> identified as the cause of the problem.  With multiqueue I/O, it's not 
>>> surprising to see multiple sequences of block numbers.
>>>
>>> Add it's not at all surprising that a consumer-grade USB storage device 
>>> might do a much worse job of handling non-sequential writes than 
>>> sequential ones.
>>>
>>> Which leads to a simple question for the SCSI or block-layer 
>>> maintainers:  Is there a sysfs setting Andrea can tweak which will 
>>> effectively restrict a particular disk device down to a single I/O
>>> queue, forcing sequential access?
>>
>> The scheduling inefficiency you are seeing may be coming from the fact that the
>> block layer does a direct issue of requests, bypassing the elevator, under some
>> conditions. One of these is sync requests on a multiqueue device. We hit this
>> problem on Zoned disks which can easily return an error for write requests
>> without the elevator throttling writes per zones (zone write locking). This
>> problem was discovered by Hans (on CC).
>>
>> I discussed this with Hannes yesterday and we think we have a fix, but we'll
>> need to do a lot of testing as all block devices are potentially impacted by the
>> change, including stacked drivers (DM). Performance regression is scary with any
>> change in that area (see blk_mq_make_request() and use of
>> blk_mq_try_issue_directly() vs blk_mq_sched_insert_request()).
> 
> Not sure this one is same with yours, for USB, mq-deadline is used at
> default, and direct issue won't be possible. Direct issue is only used
> in case of none or underlying queues of DM multipath.

For a multi-queue zoned disk, mq-deadline is also set, but we have observed
unaligned write IO errors for sync writes because of mq-deadline being bypassed
and as a result zones not being write-locked.

In blk_mq_make_request(), at the end, you have:

	} else if ((q->nr_hw_queues > 1 && is_sync) || (!q->elevator &&
			!data.hctx->dispatch_busy)) {
		blk_mq_try_issue_directly(data.hctx, rq, &cookie);
	} else {
		blk_mq_sched_insert_request(rq, false, true, true);
	}

Which I read as "for a sync req on a multi-queue device, direct issue",
regardless of the elevator being none or something else.

The correct test should probably be:

	} else if (!q->elevator &&
		   ((q->nr_hw_queues > 1 && is_sync) || 	
		     !data.hctx->dispatch_busy))) {
		blk_mq_try_issue_directly(data.hctx, rq, &cookie);
	} else {
		blk_mq_sched_insert_request(rq, false, true, true);
	}

That is, never bypass the elevator if one is set. Thoughts ?

-- 
Damien Le Moal
Western Digital Research

  reply index

Thread overview: 39+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <307581a490b610c3025ee80f79a465a89d68ed19.camel@unipv.it>
2019-08-20 17:13 ` Alan Stern
2019-08-23 10:39   ` Andrea Vai
2019-08-23 20:42     ` Alan Stern
2019-08-26  6:09       ` Andrea Vai
2019-08-26 16:33         ` Alan Stern
2019-09-18 15:25           ` Andrea Vai
2019-09-18 16:30             ` Alan Stern
2019-09-19  7:33               ` Andrea Vai
2019-09-19 17:54                 ` Alan Stern
2019-09-20  7:25                   ` Andrea Vai
2019-09-20  7:44                     ` Greg KH
2019-09-19  8:26               ` Damien Le Moal
2019-09-19  8:55                 ` Ming Lei
2019-09-19  9:09                   ` Damien Le Moal [this message]
2019-09-19  9:21                     ` Ming Lei
2019-09-19 14:01                 ` Alan Stern
2019-09-19 14:14                   ` Damien Le Moal
2019-09-20  7:03                     ` Andrea Vai
2019-09-25 19:30                       ` Alan Stern
2019-09-25 19:36                         ` Jens Axboe
2019-09-27 15:47                           ` Andrea Vai
     [not found] <e3f87757f7a0fdf551e911ad32fc8122eebe04c7.camel@unipv.it>
2019-08-13 19:52 ` Alan Stern
2019-07-02 10:46 Andrea Vai
2019-07-02 11:51 ` Johannes Thumshirn
2019-07-02 22:36   ` Andrea Vai
2019-07-03  7:29     ` Johannes Thumshirn
2019-07-03 14:23       ` Alan Stern
2019-07-06 22:06         ` Andrea Vai
2019-07-08 15:38           ` Alan Stern
2019-07-02 12:01 ` Ming Lei
2019-07-02 22:39   ` Andrea Vai
2019-07-03  2:01     ` Ming Lei
2019-07-03  5:11       ` Andrea Vai
2019-07-03  6:36         ` Ming Lei
2019-07-03 15:27           ` Chris Murphy
2019-07-06  9:33           ` Andrea Vai
2019-07-08  1:01             ` Ming Lei
2019-07-09 21:18               ` Andrea Vai
2019-07-10  2:44                 ` Ming Lei

Reply instructions:

You may reply publically 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=BYAPR04MB58163D3C66FEA00A754CF7C7E7890@BYAPR04MB5816.namprd04.prod.outlook.com \
    --to=damien.lemoal@wdc.com \
    --cc=Hans.Holmberg@wdc.com \
    --cc=andrea.vai@unipv.it \
    --cc=axboe@kernel.dk \
    --cc=gregkh@linuxfoundation.org \
    --cc=hare@suse.com \
    --cc=himanshu.madhani@cavium.com \
    --cc=jthumshirn@suse.de \
    --cc=linux-scsi@vger.kernel.org \
    --cc=linux-usb@vger.kernel.org \
    --cc=martin.petersen@oracle.com \
    --cc=ming.lei@redhat.com \
    --cc=osandov@fb.com \
    --cc=stern@rowland.harvard.edu \
    /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

Linux-USB Archive on lore.kernel.org

Archives are clonable:
	git clone --mirror https://lore.kernel.org/linux-usb/0 linux-usb/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 linux-usb linux-usb/ https://lore.kernel.org/linux-usb \
		linux-usb@vger.kernel.org linux-usb@archiver.kernel.org
	public-inbox-index linux-usb

Example config snippet for mirrors

Newsgroup available over NNTP:
	nntp://nntp.lore.kernel.org/org.kernel.vger.linux-usb


AGPL code for this site: git clone https://public-inbox.org/ public-inbox