All of lore.kernel.org
 help / color / mirror / Atom feed
From: Christoph Hellwig <hch@infradead.org>
To: Chuck Lever <chuck.lever@oracle.com>
Cc: Christoph Hellwig <hch@infradead.org>,
	Anna Schumaker <anna.schumaker@netapp.com>,
	linux-rdma@vger.kernel.org,
	Linux NFS Mailing List <linux-nfs@vger.kernel.org>
Subject: Re: [PATCH v3 05/24] xprtrdma: Reduce max_frwr_depth
Date: Tue, 11 Dec 2018 23:18:42 -0800	[thread overview]
Message-ID: <20181212071842.GB23247@infradead.org> (raw)
In-Reply-To: <69E69F06-4B3C-4C91-9412-3E7FE14C2ADA@oracle.com>

On Tue, Dec 11, 2018 at 10:30:24AM -0500, Chuck Lever wrote:
> 
> 
> > On Dec 11, 2018, at 9:02 AM, Christoph Hellwig <hch@infradead.org> wrote:
> > 
> > On Mon, Dec 10, 2018 at 11:29:45AM -0500, Chuck Lever wrote:
> >> 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.
> > 
> > Can you list which devices for reference in the changelog?
> 
> I have only three devices here. I can't make an exhaustive list.

Just list the onces you've tested.

  reply	other threads:[~2018-12-12  7:18 UTC|newest]

Thread overview: 37+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-12-10 16:29 [PATCH v3 00/24] NFS/RDMA client for next Chuck Lever
2018-12-10 16:29 ` [PATCH v3 01/24] xprtrdma: Prevent leak of rpcrdma_rep objects Chuck Lever
2018-12-10 16:29 ` [PATCH v3 02/24] IB/rxe: IB_WR_REG_MR does not capture MR's iova field Chuck Lever
2018-12-11 14:00   ` Christoph Hellwig
2018-12-11 15:26     ` Chuck Lever
2018-12-10 16:29 ` [PATCH v3 03/24] xprtrdma: Remove support for FMR memory registration Chuck Lever
2018-12-11 14:02   ` Christoph Hellwig
2018-12-11 15:29     ` Chuck Lever
2018-12-12  7:18       ` Christoph Hellwig
2018-12-10 16:29 ` [PATCH v3 04/24] xprtrdma: Fix ri_max_segs and the result of ro_maxpages Chuck Lever
2018-12-10 16:29 ` [PATCH v3 05/24] xprtrdma: Reduce max_frwr_depth Chuck Lever
2018-12-11 14:02   ` Christoph Hellwig
2018-12-11 15:30     ` Chuck Lever
2018-12-12  7:18       ` Christoph Hellwig [this message]
2018-12-10 16:29 ` [PATCH v3 06/24] xprtrdma: Plant XID in on-the-wire RDMA offset (FRWR) Chuck Lever
2018-12-10 16:29 ` [PATCH v3 07/24] xprtrdma: Recognize XDRBUF_SPARSE_PAGES Chuck Lever
2018-12-10 16:30 ` [PATCH v3 08/24] xprtrdma: Remove request_module from backchannel Chuck Lever
2018-12-10 16:30 ` [PATCH v3 09/24] xprtrdma: Expose transport header errors Chuck Lever
2018-12-10 16:30 ` [PATCH v3 10/24] xprtrdma: Simplify locking that protects the rl_allreqs list Chuck Lever
2018-12-10 16:30 ` [PATCH v3 11/24] xprtrdma: Cull dprintk() call sites Chuck Lever
2018-12-10 16:30 ` [PATCH v3 12/24] xprtrdma: Clean up of xprtrdma chunk trace points Chuck Lever
2018-12-10 16:30 ` [PATCH v3 13/24] xprtrdma: Relocate the xprtrdma_mr_map " Chuck Lever
2018-12-10 16:30 ` [PATCH v3 14/24] xprtrdma: Add trace points for calls to transport switch methods Chuck Lever
2018-12-10 16:30 ` [PATCH v3 15/24] NFS: Make "port=" mount option optional for RDMA mounts Chuck Lever
2018-12-10 16:30 ` [PATCH v3 16/24] SUNRPC: Remove support for kerberos_v1 Chuck Lever
2018-12-12 21:20   ` Chuck Lever
2018-12-14 21:16     ` Chuck Lever
2018-12-10 16:30 ` [PATCH v3 17/24] SUNRPC: Fix some kernel doc complaints Chuck Lever
2018-12-10 16:30 ` [PATCH v3 18/24] NFS: Fix NFSv4 symbolic trace point output Chuck Lever
     [not found]   ` <632f5635-4c37-16ae-cdd0-65679d21c9ec@oracle.com>
2018-12-11 19:19     ` Calum Mackay
2018-12-10 16:31 ` [PATCH v3 19/24] SUNRPC: Simplify defining common RPC trace events Chuck Lever
2018-12-10 16:31 ` [PATCH v3 20/24] xprtrdma: Trace mapping, alloc, and dereg failures Chuck Lever
2018-12-10 16:31 ` [PATCH v3 21/24] xprtrdma: Update comments in frwr_op_send Chuck Lever
2018-12-10 16:31 ` [PATCH v3 22/24] xprtrdma: Replace outdated comment for rpcrdma_ep_post Chuck Lever
2018-12-10 16:31 ` [PATCH v3 23/24] xprtrdma: Add documenting comment for rpcrdma_buffer_destroy Chuck Lever
2018-12-10 16:31 ` [PATCH v3 24/24] xprtrdma: Clarify comments in rpcrdma_ia_remove Chuck Lever
2018-12-10 17:55 ` [PATCH v3 00/24] NFS/RDMA client for next Jason Gunthorpe

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=20181212071842.GB23247@infradead.org \
    --to=hch@infradead.org \
    --cc=anna.schumaker@netapp.com \
    --cc=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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.