From: Bart Van Assche <bvanassche@acm.org>
To: "Asutosh Das (asd)" <quic_asutoshd@quicinc.com>,
John Garry <john.garry@huawei.com>,
Can Guo <quic_cang@quicinc.com>,
stanley.chu@mediatek.com, adrian.hunter@intel.com,
alim.akhtar@samsung.com, avri.altman@wdc.com, beanhuo@micron.com,
quic_nguyenb@quicinc.com, quic_ziqichen@quicinc.com,
linux-scsi@vger.kernel.org, kernel-team@android.com
Cc: "James E.J. Bottomley" <jejb@linux.ibm.com>,
"Martin K. Petersen" <martin.petersen@oracle.com>,
Daejun Park <daejun7.park@samsung.com>,
Jinyoung Choi <j-young.choi@samsung.com>,
Kiwoong Kim <kwmad.kim@samsung.com>,
open list <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 1/2] scsi: ufs: Add Multi-Circular Queue support
Date: Fri, 29 Jul 2022 11:59:08 -0700 [thread overview]
Message-ID: <afefc312-4996-e5f2-b33a-6a987b20961b@acm.org> (raw)
In-Reply-To: <59e835e6-b694-07c5-16a1-5f7774fc4177@quicinc.com>
On 7/29/22 09:43, Asutosh Das (asd) wrote:
> I went through the change and fwiu of your proposal is to use
> host_tagset = 1 if HC doesn't support EXT_IID capability.
> That way tags would be shared across 16 HWQs (4-bit IID encoding the
> queue-ids).
> That would also mean that the hba->nutrs would have to be adjusted such
> that the tags(8-bit) don't exceed 255.
>
> Summarily,
> if EXT_IID is not supported:
> host_tagset = 1, maximum configurable hba->nutrs = 16, maximum
> configurable nr_hw_queues = 16.
> maximum number of outstanding commands to host = 16 x 16 = 256.
>
> if EXT_IID is supported:
> host_tagset = 0, maximum confiugrable hba-nutrs = 255, maximum
> configurable nr_hw_queues = 255.
>
> Please let me know if I'm missing something.
Hi Asutosh,
I recommend to always set host_tagset = 1. The performance overhead of
host_tagset = 1 compared to host_tagset = 0 should be negligible for UFS
devices. This will simplify the UFS host controller driver and will
allow us to limit the total number of outstanding commands to the
maximum number of outstanding commands supported by the UFS device. I
expect that we will need to do this. If more commands are sent to the
controller than what the UFS device can queue, the host controller will
decide in which order commands are fetched and sent to the controller.
We want the host to decide in which order commands are queued and not
the host controller. As an example, adding support for HCTX_TYPE_READ
queues will only work as expected if the number of outstanding commands
is less than or equal to what the UFS device can queue.
For host_tagset = 1, we can combine the 4 EXT_IID bits with the 8 bit
task tag and create a single space of 4096 (2**12) tags. The block layer
sbitmap data structure has been designed to minimize contention even if
a single sbitmap data structure is shared across multiple CPU cores.
Thanks,
Bart.
next prev parent reply other threads:[~2022-07-29 18:59 UTC|newest]
Thread overview: 48+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <1658214120-22772-1-git-send-email-quic_cang@quicinc.com>
2022-07-19 7:01 ` [PATCH 1/2] scsi: ufs: Add Multi-Circular Queue support Can Guo
2022-07-19 23:01 ` Bart Van Assche
2022-07-19 23:07 ` Bart Van Assche
2022-07-20 17:34 ` Asutosh Das (asd)
2022-07-20 7:57 ` kernel test robot
2022-07-20 11:41 ` kernel test robot
2022-07-20 16:36 ` kernel test robot
2022-07-22 7:31 ` Avri Altman
2022-07-22 17:35 ` Asutosh Das (asd)
2022-07-22 19:37 ` Avri Altman
2022-07-22 20:14 ` Asutosh Das (asd)
2022-07-22 20:22 ` Avri Altman
2022-07-22 21:05 ` Asutosh Das (asd)
2022-07-22 17:58 ` Bart Van Assche
2022-07-26 6:48 ` Can Guo
2022-07-22 14:42 ` Avri Altman
2022-07-23 14:59 ` Avri Altman
2022-07-26 2:55 ` Can Guo
2022-07-23 15:26 ` Avri Altman
2022-07-24 3:14 ` Bart Van Assche
2022-07-25 16:24 ` Asutosh Das (asd)
2022-07-23 20:22 ` Avri Altman
2022-07-25 16:26 ` Asutosh Das (asd)
2022-07-25 19:50 ` Avri Altman
2022-07-25 20:24 ` Asutosh Das (asd)
2022-07-23 21:23 ` Avri Altman
2022-07-24 3:15 ` Bart Van Assche
2022-07-25 16:35 ` Asutosh Das (asd)
2022-07-26 22:47 ` Bart Van Assche
2022-07-24 4:07 ` Avri Altman
2022-07-25 16:38 ` Asutosh Das (asd)
2022-07-26 6:35 ` Can Guo
2022-07-26 9:46 ` Avri Altman
2022-07-24 4:32 ` Avri Altman
2022-07-26 6:21 ` Can Guo
2022-07-24 7:21 ` Avri Altman
2022-07-24 21:54 ` Bean Huo
2022-07-25 17:35 ` Asutosh Das (asd)
2022-07-25 9:16 ` Dan Carpenter
2022-07-28 19:10 ` John Garry
2022-07-28 19:15 ` Asutosh Das (asd)
2022-07-28 20:29 ` Bart Van Assche
2022-07-28 21:07 ` Asutosh Das (asd)
2022-07-29 16:43 ` Asutosh Das (asd)
2022-07-29 18:59 ` Bart Van Assche [this message]
2022-07-28 19:38 ` Bart Van Assche
2022-07-28 20:00 ` Asutosh Das (asd)
2022-07-19 7:01 ` [PATCH 2/2] scsi: ufs-qcom: Implement three CMQ related vops Can Guo
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=afefc312-4996-e5f2-b33a-6a987b20961b@acm.org \
--to=bvanassche@acm.org \
--cc=adrian.hunter@intel.com \
--cc=alim.akhtar@samsung.com \
--cc=avri.altman@wdc.com \
--cc=beanhuo@micron.com \
--cc=daejun7.park@samsung.com \
--cc=j-young.choi@samsung.com \
--cc=jejb@linux.ibm.com \
--cc=john.garry@huawei.com \
--cc=kernel-team@android.com \
--cc=kwmad.kim@samsung.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-scsi@vger.kernel.org \
--cc=martin.petersen@oracle.com \
--cc=quic_asutoshd@quicinc.com \
--cc=quic_cang@quicinc.com \
--cc=quic_nguyenb@quicinc.com \
--cc=quic_ziqichen@quicinc.com \
--cc=stanley.chu@mediatek.com \
/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;
as well as URLs for NNTP newsgroup(s).