From: Chuck Lever <chuck.lever@oracle.com>
To: linux-rdma@vger.kernel.org
Cc: linux-nfs@vger.kernel.org
Subject: [PATCH v1 3/4] xprtrdma: Reduce max_frwr_depth
Date: Mon, 19 Nov 2018 10:46:01 -0500 [thread overview]
Message-ID: <20181119154601.10832.97570.stgit@manet.1015granger.net> (raw)
In-Reply-To: <20181119153707.10832.42881.stgit@manet.1015granger.net>
Some devices advertise a large max_fast_reg_page_list_len
capability, but perform optimally when MRs are significantly smaller
than that depth -- probably when the MR itself is no larger than a
page.
By default, the RDMA R/W core API uses max_sge_rd as the maximum
page depth for MRs. For some devices, the value of max_sge_rd is
1, which is also not optimal. Thus, when max_sge_rd is larger than
1, use that value. Otherwise use the value of the
max_fast_reg_page_list_len attribute.
I've tested this with a couple of devices, and it reproducibly
improves the throughput of large I/Os by several percent.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
---
net/sunrpc/xprtrdma/frwr_ops.c | 15 +++++++++++----
1 file changed, 11 insertions(+), 4 deletions(-)
diff --git a/net/sunrpc/xprtrdma/frwr_ops.c b/net/sunrpc/xprtrdma/frwr_ops.c
index 448e063..49b314d 100644
--- a/net/sunrpc/xprtrdma/frwr_ops.c
+++ b/net/sunrpc/xprtrdma/frwr_ops.c
@@ -191,10 +191,17 @@
if (attrs->device_cap_flags & IB_DEVICE_SG_GAPS_REG)
ia->ri_mrtype = IB_MR_TYPE_SG_GAPS;
- ia->ri_max_frwr_depth =
- min_t(unsigned int, RPCRDMA_MAX_DATA_SEGS,
- attrs->max_fast_reg_page_list_len);
- dprintk("RPC: %s: device's max FR page list len = %u\n",
+ /* Quirk: Some devices advertise a large max_fast_reg_page_list_len
+ * capability, but perform optimally when the MRs are not larger
+ * than a page.
+ */
+ if (attrs->max_sge_rd > 1)
+ ia->ri_max_frwr_depth = attrs->max_sge_rd;
+ else
+ ia->ri_max_frwr_depth = attrs->max_fast_reg_page_list_len;
+ if (ia->ri_max_frwr_depth > RPCRDMA_MAX_DATA_SEGS)
+ ia->ri_max_frwr_depth = RPCRDMA_MAX_DATA_SEGS;
+ dprintk("RPC: %s: max FR page list depth = %u\n",
__func__, ia->ri_max_frwr_depth);
/* Add room for frwr register and invalidate WRs.
next prev parent reply other threads:[~2018-11-19 15:46 UTC|newest]
Thread overview: 31+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-11-19 15:45 [PATCH v1 0/4] NFS/RDMA client for v4.21 (part 1) Chuck Lever
2018-11-19 15:45 ` [PATCH v1 1/4] xprtrdma: Remove support for FMR memory registration Chuck Lever
2018-11-19 16:16 ` Bart Van Assche
2018-11-19 19:09 ` Leon Romanovsky
2018-11-19 20:52 ` Bart Van Assche
2018-11-20 5:37 ` Leon Romanovsky
2018-11-19 22:41 ` Jason Gunthorpe
2018-11-19 22:56 ` Chuck Lever
2018-11-19 23:10 ` Jason Gunthorpe
2018-11-20 15:22 ` Dennis Dalessandro
2018-11-19 15:45 ` [PATCH v1 2/4] xprtrdma: mrs_create off-by-one Chuck Lever
2018-11-19 15:46 ` Chuck Lever [this message]
2018-11-19 15:46 ` [PATCH v1 4/4] xprtrdma: Plant XID in on-the-wire RDMA offset (FRWR) Chuck Lever
2018-11-19 17:47 ` Olga Kornievskaia
2018-11-19 17:58 ` Chuck Lever
2018-11-19 18:08 ` Olga Kornievskaia
2018-11-19 18:18 ` Chuck Lever
2018-11-19 18:47 ` Olga Kornievskaia
2018-11-19 18:58 ` Chuck Lever
2018-11-19 21:22 ` Olga Kornievskaia
2018-11-19 21:32 ` Chuck Lever
2018-11-19 21:42 ` Mora, Jorge
2018-11-19 22:46 ` Jason Gunthorpe
2018-11-20 2:45 ` Tom Talpey
2018-11-20 3:09 ` Jason Gunthorpe
2018-11-20 3:25 ` Tom Talpey
2018-11-20 3:32 ` Jason Gunthorpe
2018-11-20 3:38 ` Tom Talpey
2018-11-20 18:02 ` Anna Schumaker
2018-11-20 18:07 ` Chuck Lever
[not found] ` <94ff7ec712e086bfdd9c217a5f97c293a07151b9.camel@gmail.com>
2018-11-20 21:31 ` Chuck Lever
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=20181119154601.10832.97570.stgit@manet.1015granger.net \
--to=chuck.lever@oracle.com \
--cc=linux-nfs@vger.kernel.org \
--cc=linux-rdma@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