From: James Smart <jsmart2021@gmail.com>
To: Max Gurtovoy <maxg@mellanox.com>, Christoph Hellwig <hch@lst.de>
Cc: Nigel Kirkland <nigel.kirkland@broadcom.com>,
mark.wunderlich@intel.com, sagi@grimberg.me,
linux-nvme@lists.infradead.org
Subject: Re: [PATCH 2/2] nvme-rdma/nvme-tcp: Move err_work to nvme_reset_wq
Date: Tue, 11 Feb 2020 09:56:06 -0800 [thread overview]
Message-ID: <35b81d7a-3783-c81b-a45b-f22260671444@gmail.com> (raw)
In-Reply-To: <ec482da7-1794-eea1-555c-31c76e792e4d@mellanox.com>
On 2/11/2020 2:12 AM, Max Gurtovoy wrote:
>
> On 2/10/2020 7:04 PM, Christoph Hellwig wrote:
>> On Thu, Feb 06, 2020 at 11:13:42AM -0800, James Smart wrote:
>>> With keep alive processing being moved to nvme_wq, it potentially
>>> creates
>>> a conflicting position with err_work also processing on nvme_wq and
>>> needing
>>> to flush/stop keep alives.
>>>
>>> To avoid issues, schedule err_work on nvme_reset_wq. It looks like
>>> this is
>>> not only a good thing for keep alives, but also brings the transports in
>>> line with the RESETTING state and processing work relative to RESETTING
>>> on nvme_reset_wq.
>>>
>>> This change is made to both nvme-rdma and nvme-tcp which have like code.
>> Shouldn't we move both in one patch to create a regression in one
>> cycle while fixing the other one?
>
> Yup, seems like these should be squashed.
note: see the v2 posted yesterday.
>
> James,
>
> did you run some testing for RDMA/tcp that will trigger this code ?
The testing of the keepalive movement was all in FC. FC already has it's
calls from either nvme_reset_wq or nvme_delete_wq.
For RDMA/TCP - it was not tested. It was by code inspection and looking
at who calls the keepalive cancellation calls.
-- james
_______________________________________________
linux-nvme mailing list
linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme
prev parent reply other threads:[~2020-02-11 18:20 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-02-06 19:13 [PATCH 0/2] nvme: prevent warning triggered by nvme_stop_keep_alive James Smart
2020-02-06 19:13 ` [PATCH 1/2] nvme: move keep alive processing to nvme-wq James Smart
2020-02-07 0:05 ` Sagi Grimberg
2020-02-10 17:04 ` Christoph Hellwig
2020-02-11 3:00 ` Keith Busch
2020-02-11 17:50 ` James Smart
2020-02-06 19:13 ` [PATCH 2/2] nvme-rdma/nvme-tcp: Move err_work to nvme_reset_wq James Smart
2020-02-07 0:06 ` Sagi Grimberg
2020-02-10 17:04 ` Christoph Hellwig
2020-02-11 10:12 ` Max Gurtovoy
2020-02-11 17:56 ` James Smart [this message]
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=35b81d7a-3783-c81b-a45b-f22260671444@gmail.com \
--to=jsmart2021@gmail.com \
--cc=hch@lst.de \
--cc=linux-nvme@lists.infradead.org \
--cc=mark.wunderlich@intel.com \
--cc=maxg@mellanox.com \
--cc=nigel.kirkland@broadcom.com \
--cc=sagi@grimberg.me \
/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