From: Jens Axboe <axboe@kernel.dk>
To: "Michael Kelley (LINUX)" <mikelley@microsoft.com>,
Sasha Levin <sashal@kernel.org>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
"stable@vger.kernel.org" <stable@vger.kernel.org>
Cc: Dexuan Cui <decui@microsoft.com>, Ming Lei <ming.lei@redhat.com>,
"linux-block@vger.kernel.org" <linux-block@vger.kernel.org>
Subject: Re: [PATCH AUTOSEL 5.15 20/29] block: reduce kblockd_mod_delayed_work_on() CPU consumption
Date: Tue, 21 Dec 2021 08:36:33 -0700 [thread overview]
Message-ID: <ad76826e-73b2-b2f0-3cd4-8481645a6568@kernel.dk> (raw)
In-Reply-To: <MWHPR21MB1593141494C76CF5A0BDDB11D77C9@MWHPR21MB1593.namprd21.prod.outlook.com>
On 12/21/21 8:35 AM, Michael Kelley (LINUX) wrote:
> From: Sasha Levin <sashal@kernel.org> Sent: Monday, December 20, 2021 5:58 PM
>>
>> From: Jens Axboe <axboe@kernel.dk>
>>
>> [ Upstream commit cb2ac2912a9ca7d3d26291c511939a41361d2d83 ]
>>
>> Dexuan reports that he's seeing spikes of very heavy CPU utilization when
>> running 24 disks and using the 'none' scheduler. This happens off the
>> sched restart path, because SCSI requires the queue to be restarted async,
>> and hence we're hammering on mod_delayed_work_on() to ensure that the work
>> item gets run appropriately.
>>
>> Avoid hammering on the timer and just use queue_work_on() if no delay
>> has been specified.
>>
>> Reported-and-tested-by: Dexuan Cui <decui@microsoft.com>
>> Link: https://lore.kernel.org/linux-block/BYAPR21MB1270C598ED214C0490F47400BF719@BYAPR21MB1270.namprd21.prod.outlook.com/
>> Reviewed-by: Ming Lei <ming.lei@redhat.com>
>> Signed-off-by: Jens Axboe <axboe@kernel.dk>
>> Signed-off-by: Sasha Levin <sashal@kernel.org>
>> ---
>> block/blk-core.c | 2 ++
>> 1 file changed, 2 insertions(+)
>>
>> diff --git a/block/blk-core.c b/block/blk-core.c
>> index c2d912d0c976c..a728434fcff87 100644
>> --- a/block/blk-core.c
>> +++ b/block/blk-core.c
>> @@ -1625,6 +1625,8 @@ EXPORT_SYMBOL(kblockd_schedule_work);
>> int kblockd_mod_delayed_work_on(int cpu, struct delayed_work *dwork,
>> unsigned long delay)
>> {
>> + if (!delay)
>> + return queue_work_on(cpu, kblockd_workqueue, &dwork->work);
>> return mod_delayed_work_on(cpu, kblockd_workqueue, dwork, delay);
>> }
>> EXPORT_SYMBOL(kblockd_mod_delayed_work_on);
>> --
>> 2.34.1
>
> Sasha -- there are reports of this patch causing performance problems.
> See
> https://lore.kernel.org/lkml/1639853092.524jxfaem2.none@localhost/. I
> would suggest *not* backporting it to any of the stable branches until
> the issues are fully sorted out.
Both this and the revert were backported. Which arguably doesn't make a
lot of sense, but at least it's consistent and won't cause any issues...
--
Jens Axboe
next prev parent reply other threads:[~2021-12-21 15:37 UTC|newest]
Thread overview: 34+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-12-21 1:57 [PATCH AUTOSEL 5.15 01/29] ARM: dts: imx6qp-prtwd3: update RGMII delays for sja1105 switch Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 02/29] ARM: dts: ls1021a-tsn: " Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 03/29] ARM: rockchip: Use memcpy_toio instead of memcpy on smp bring-up Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 04/29] mac80211: fix TCP performance on mesh interface Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 05/29] mac80211: set up the fwd_skb->dev for mesh forwarding Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 06/29] mac80211: fix a memory leak where sta_info is not freed Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 07/29] nl80211: reset regdom when reloading regdb Sasha Levin
2021-12-25 18:23 ` Sedat Dilek
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 08/29] iwlwifi: fix LED dependencies Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 09/29] phonet: refcount leak in pep_sock_accep Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 10/29] net: usb: qmi_wwan: add Telit 0x1070 composition Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 11/29] dmaengine: ti: k3-udma: Fix smatch warnings Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 12/29] net: bcmgenet: Fix NULL vs IS_ERR() checking Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 13/29] net: ethernet: ti: add missing of_node_put before return Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 14/29] scsi: pm8001: Fix phys_to_virt() usage on dma_addr_t Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 15/29] mac80211: Fix the size used for building probe request Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 16/29] mac80211: update channel context before station state Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 17/29] mac80211: do drv_reconfig_complete() before restarting all Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 18/29] drm/amd/display: Reset DMCUB before HW init Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 19/29] drm/amdgpu: correct the wrong cached state for GMC on PICASSO Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 20/29] block: reduce kblockd_mod_delayed_work_on() CPU consumption Sasha Levin
2021-12-21 15:35 ` Michael Kelley (LINUX)
2021-12-21 15:36 ` Jens Axboe [this message]
2021-12-21 17:58 ` Sasha Levin
2021-12-21 18:11 ` Jens Axboe
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 21/29] net: usb: lan78xx: add Allied Telesis AT29M2-AF Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 22/29] virtio_net: fix rx_drops stat for small pkts Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 23/29] bpf: Make 32->64 bounds propagation slightly more robust Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 24/29] riscv: dts: sifive unmatched: Name gpio lines Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 25/29] riscv: dts: sifive unmatched: Expose the board ID eeprom Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 26/29] riscv: dts: sifive unmatched: Expose the PMIC sub-functions Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 27/29] riscv: dts: sifive unmatched: Fix regulator for board rev3 Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 28/29] riscv: dts: sifive unmatched: Link the tmp451 with its power supply Sasha Levin
2021-12-21 1:57 ` [PATCH AUTOSEL 5.15 29/29] Revert "block: reduce kblockd_mod_delayed_work_on() CPU consumption" Sasha Levin
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=ad76826e-73b2-b2f0-3cd4-8481645a6568@kernel.dk \
--to=axboe@kernel.dk \
--cc=decui@microsoft.com \
--cc=linux-block@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mikelley@microsoft.com \
--cc=ming.lei@redhat.com \
--cc=sashal@kernel.org \
--cc=stable@vger.kernel.org \
/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