From: Karsten Graul <kgraul@linux.ibm.com>
To: Tony Lu <tonylu@linux.alibaba.com>,
davem@davemloft.net, kuba@kernel.org, ubraun@linux.ibm.com
Cc: netdev@vger.kernel.org, linux-s390@vger.kernel.org,
linux-rdma@vger.kernel.org, jacob.qi@linux.alibaba.com,
xuanzhuo@linux.alibaba.com, guwen@linux.alibaba.com,
dust.li@linux.alibaba.com
Subject: Re: [PATCH net 4/4] net/smc: Fix wq mismatch issue caused by smc fallback
Date: Fri, 29 Oct 2021 11:40:17 +0200 [thread overview]
Message-ID: <acaf3d5a-219b-3eec-3a65-91d3fdfb21e9@linux.ibm.com> (raw)
In-Reply-To: <20211027085208.16048-5-tonylu@linux.alibaba.com>
On 27/10/2021 10:52, Tony Lu wrote:
> From: Wen Gu <guwen@linux.alibaba.com>
>
> A socket_wq mismatch issue may occur because of fallback.
>
> When use SMC to replace TCP, applications add an epoll entry into SMC
> socket's wq, but kernel uses clcsock's wq instead of SMC socket's wq
> once fallback occurs, which means the application's epoll fd dosen't
> work anymore.
I am not sure if I understand this fix completely, please explain your intentions
for the changes in more detail.
What I see so far:
- smc_create() swaps the sk->sk_wq of the clcsocket and the new SMC socket
- sets clcsocket sk->sk_wq to smcsocket->wq (why?)
- sets smcsocket sk->sk_wq to clcsocket->wq (why?)
- smc_switch_to_fallback() resets the clcsock sk->sk_wq to clcsocket->wq
- smc_accept() sets smcsocket sk->sk_wq to clcsocket->wq when it is NOT fallback
- but this was already done before in smc_create() ??
- smc_poll() now always uses clcsocket->wq for the call to sock_poll_wait()
In smc_poll() the comment says that now clcsocket->wq is used for poll, whats
the relation between socket->wq and socket->sk->sk_wq here?
next prev parent reply other threads:[~2021-10-29 9:40 UTC|newest]
Thread overview: 28+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-10-27 8:52 [PATCH net 0/4] Fixes for SMC Tony Lu
2021-10-27 8:52 ` [PATCH net 1/4] Revert "net/smc: don't wait for send buffer space when data was already sent" Tony Lu
2021-10-27 10:21 ` Karsten Graul
2021-10-27 15:08 ` Jakub Kicinski
2021-10-27 15:38 ` Karsten Graul
2021-10-27 15:47 ` Jakub Kicinski
2021-10-28 6:48 ` Tony Lu
2021-10-28 11:57 ` Karsten Graul
2021-10-28 14:38 ` Jakub Kicinski
2021-11-01 7:04 ` Tony Lu
2021-11-02 9:17 ` Karsten Graul
2021-11-03 3:06 ` Tony Lu
2021-11-06 12:46 ` Karsten Graul
2021-10-27 8:52 ` [PATCH net 2/4] net/smc: Fix smc_link->llc_testlink_time overflow Tony Lu
2021-10-27 10:24 ` Karsten Graul
2021-10-28 6:52 ` Tony Lu
2021-10-27 8:52 ` [PATCH net 3/4] net/smc: Correct spelling mistake to TCPF_SYN_RECV Tony Lu
2021-10-27 10:23 ` Karsten Graul
2021-10-28 6:53 ` Tony Lu
2021-10-27 8:52 ` [PATCH net 4/4] net/smc: Fix wq mismatch issue caused by smc fallback Tony Lu
2021-10-28 14:16 ` Karsten Graul
2021-10-29 9:40 ` Karsten Graul [this message]
2021-11-01 6:15 ` Wen Gu
2021-11-02 9:25 ` Karsten Graul
2021-11-03 8:56 ` Wen Gu
2021-11-04 4:39 ` Wen Gu
2021-11-06 12:51 ` Karsten Graul
2021-11-10 12:50 ` Wen Gu
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=acaf3d5a-219b-3eec-3a65-91d3fdfb21e9@linux.ibm.com \
--to=kgraul@linux.ibm.com \
--cc=davem@davemloft.net \
--cc=dust.li@linux.alibaba.com \
--cc=guwen@linux.alibaba.com \
--cc=jacob.qi@linux.alibaba.com \
--cc=kuba@kernel.org \
--cc=linux-rdma@vger.kernel.org \
--cc=linux-s390@vger.kernel.org \
--cc=netdev@vger.kernel.org \
--cc=tonylu@linux.alibaba.com \
--cc=ubraun@linux.ibm.com \
--cc=xuanzhuo@linux.alibaba.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