linux-rdma.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Halil Pasic <pasic@linux.ibm.com>
To: Guangguan Wang <guangguan.wang@linux.alibaba.com>
Cc: Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
	Simon Horman <horms@kernel.org>,
	"D. Wythe" <alibuda@linux.alibaba.com>,
	Dust Li <dust.li@linux.alibaba.com>,
	Sidraya Jayagond <sidraya@linux.ibm.com>,
	Wenjia Zhang <wenjia@linux.ibm.com>,
	Mahanta Jambigi <mjambigi@linux.ibm.com>,
	Tony Lu <tonylu@linux.alibaba.com>,
	Wen Gu <guwen@linux.alibaba.com>,
	netdev@vger.kernel.org, linux-doc@vger.kernel.org,
	linux-kernel@vger.kernel.org, linux-rdma@vger.kernel.org,
	linux-s390@vger.kernel.org, Halil Pasic <pasic@linux.ibm.com>
Subject: Re: [PATCH net-next v3 1/2] net/smc: make wr buffer count configurable
Date: Fri, 26 Sep 2025 12:12:49 +0200	[thread overview]
Message-ID: <20250926121249.687b519d.pasic@linux.ibm.com> (raw)
In-Reply-To: <1aa764d0-0613-499e-bc44-52e70602b661@linux.alibaba.com>

On Fri, 26 Sep 2025 10:44:00 +0800
Guangguan Wang <guangguan.wang@linux.alibaba.com> wrote:

> > +
> > +smcr_max_send_wr - INTEGER
> > +	So called work request buffers are SMCR link (and RDMA queue pair) level
> > +	resources necessary for performing RDMA operations. Since up to 255
> > +	connections can share a link group and thus also a link and the number
> > +	of the work request buffers is decided when the link is allocated,
> > +	depending on the workload it can a bottleneck in a sense that threads
> > +	have to wait for work request buffers to become available. Before the
> > +	introduction of this control the maximal number of work request buffers
> > +	available on the send path used to be hard coded to 16. With this control
> > +	it becomes configurable. The acceptable range is between 2 and 2048.
> > +
> > +	Please be aware that all the buffers need to be allocated as a physically
> > +	continuous array in which each element is a single buffer and has the size
> > +	of SMC_WR_BUF_SIZE (48) bytes. If the allocation fails we give up much
> > +	like before having this control.
> > +
> > +	Default: 16
> > +
> > +smcr_max_recv_wr - INTEGER
> > +	So called work request buffers are SMCR link (and RDMA queue pair) level
> > +	resources necessary for performing RDMA operations. Since up to 255
> > +	connections can share a link group and thus also a link and the number
> > +	of the work request buffers is decided when the link is allocated,
> > +	depending on the workload it can a bottleneck in a sense that threads
> > +	have to wait for work request buffers to become available. Before the
> > +	introduction of this control the maximal number of work request buffers
> > +	available on the receive path used to be hard coded to 16. With this control
> > +	it becomes configurable. The acceptable range is between 2 and 2048.
> > +
> > +	Please be aware that all the buffers need to be allocated as a physically
> > +	continuous array in which each element is a single buffer and has the size
> > +	of SMC_WR_BUF_SIZE (48) bytes. If the allocation fails we give up much
> > +	like before having this control.
> > +
> > +	Default: 48  
> 
> Notice that the ratio of smcr_max_recv_wr to smcr_max_send_wr is set to 3:1, with the
> intention of ensuring that the peer QP's smcr_max_recv_wr is three times the local QP's
> smcr_max_send_wr and the local QP's smcr_max_recv_wr is three times the peer QP's
> smcr_max_send_wr, rather than making the local QP's smcr_max_recv_wr three times its own
> smcr_max_send_wr. The purpose of this design is to guarantee sufficient receive WRs on
> the side to receive incoming data when peer QP doing RDMA sends. Otherwise, RNR (Receiver
> Not Ready) may occur, leading to poor performance(RNR will drop the packet and retransmit
> happens in the transport layer of the RDMA).

Thank you Guangguan! I think we already had that discussion. 
> 
> Let us guess a scenario that have multiple hosts, and the multiple hosts have different
> smcr_max_send_wr and smcr_max_recv_wr configurations, mesh connections between these hosts.
> It is difficult to ensure that the smcr_max_recv_wr/smcr_max_send_wr is 3:1 on the connected
> QPs between these hosts, and it may even be hard to guarantee the smcr_max_recv_wr > smcr_max_send_wr
> on the connected QPs between these hosts.


It is not difficult IMHO. You just leave the knobs alone and you have
3:1 per default. If tuning is attempted that needs to be done carefully.
At least with SMC-R V2 there is this whole EID business, as well so it
is reasonable to assume that the environment can be tuned in a coherent
fashion. E.g. whoever is calling the EID could call use smcr_max_recv_wr:=32 and smcr_max_send_wr:=

> 
> Therefore, I believe that if these values are made configurable, additional mechanisms must be
> in place to prevent RNR from occurring. Otherwise we need to carefully configure smcr_max_recv_wr
> and smcr_max_send_wr, or ensure that all hosts capable of establishing SMC-R connections are configured
> smcr_max_recv_wr and smcr_max_send_wr with the same values.

Thank you for 

  reply	other threads:[~2025-09-26 10:13 UTC|newest]

Thread overview: 21+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-09-21 21:44 [PATCH net-next v3 0/2] net/smc: make wr buffer count configurable Halil Pasic
2025-09-21 21:44 ` [PATCH net-next v3 1/2] " Halil Pasic
2025-09-24 17:27   ` Sidraya Jayagond
2025-09-25  9:27   ` Paolo Abeni
2025-09-25 11:25     ` Halil Pasic
2025-09-27 22:55       ` Halil Pasic
2025-09-28  2:02         ` Dust Li
2025-09-28  2:12           ` Dust Li
2025-09-28  8:39           ` Halil Pasic
2025-09-28 11:42             ` Dust Li
2025-09-28 18:32               ` Halil Pasic
2025-09-26  2:44   ` Guangguan Wang
2025-09-26 10:12     ` Halil Pasic [this message]
2025-09-26 10:30       ` Halil Pasic
2025-09-28  3:05         ` Guangguan Wang
2025-09-21 21:44 ` [PATCH net-next v3 2/2] net/smc: handle -ENOMEM from smc_wr_alloc_link_mem gracefully Halil Pasic
2025-09-24 17:28   ` Sidraya Jayagond
2025-09-25  9:40   ` Paolo Abeni
2025-09-25 15:05     ` Halil Pasic
2025-09-25 15:41       ` Paolo Abeni
2025-09-25 21:46         ` Halil Pasic

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=20250926121249.687b519d.pasic@linux.ibm.com \
    --to=pasic@linux.ibm.com \
    --cc=alibuda@linux.alibaba.com \
    --cc=dust.li@linux.alibaba.com \
    --cc=guangguan.wang@linux.alibaba.com \
    --cc=guwen@linux.alibaba.com \
    --cc=horms@kernel.org \
    --cc=kuba@kernel.org \
    --cc=linux-doc@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-rdma@vger.kernel.org \
    --cc=linux-s390@vger.kernel.org \
    --cc=mjambigi@linux.ibm.com \
    --cc=netdev@vger.kernel.org \
    --cc=pabeni@redhat.com \
    --cc=sidraya@linux.ibm.com \
    --cc=tonylu@linux.alibaba.com \
    --cc=wenjia@linux.ibm.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).