From: Tom Tucker <tom@opengridcomputing.com>
To: bfields@fieldses.org
Cc: linux-nfs@vger.kernel.org, Tom Tucker <tom@opengridcomputing.com>
Subject: [PATCH 00/11] svcrdma: WR context management bug fixes and cleanup
Date: Wed, 2 Jul 2008 21:27:32 -0500 [thread overview]
Message-ID: <1215052063-7330-1-git-send-email-tom@opengridcomputing.com> (raw)
Bruce:
This is version 2 of the WR context cleanup patchset. It includes modifications
per your review. It has been tested with iozone and Connectathon over IB
and iWARP on x86_64. The x86_32 platform has only been compile tested for
warnings.
This set of patches fixes bugs related to DMA mapping and unmapping and
in the process redesigns and cleans up the way that WR contexts are managed.
In addition to the bug fixes, there are three major changes:
- The data structure that maps client side memory to server side
memory has been separated from the WR context. This makes for cleaner
type checking and a smaller memory footprint for the RDMA transport.
- The mapping and unmapping of DMA is performed independently from the
allocation and deallocation of the WR context. This is to allow the
unmapping to occur early on recv where the unmapping of the DMA needs
to be done prior to looking at the data and therefore before
deallocating the WR context.
- The WR context cache now uses a kmem_cache and is shared across
all transport instances. This allows contexts to be shared between
mounts and idle mounts don't consume context memory.
This patchset is intended for 2.6.27 and is based on 2.6.26-rc8. These patches
are also available at the following git url:
git://git.linux-nfs.org/projects/tomtucker/xprt-switch-2.6.git
[PATCH 01/11] svcrdma: Add a type for keeping NFS RPC mapping
include/linux/sunrpc/svc_rdma.h | 27 +++++++++++++++++++++++++++
net/sunrpc/xprtrdma/svc_rdma.c | 19 +++++++++++++++++++
net/sunrpc/xprtrdma/svc_rdma_transport.c | 26 ++++++++++++++++++++++++++
3 files changed, 72 insertions(+), 0 deletions(-)
[PATCH 02/11] svcrdma: Use RPC reply map for RDMA_WRITE processing
net/sunrpc/xprtrdma/svc_rdma_sendto.c | 163 ++++++++++++++----------------
net/sunrpc/xprtrdma/svc_rdma_transport.c | 5 +-
2 files changed, 80 insertions(+), 88 deletions(-)
[PATCH 03/11] svcrdma: Use reply and chunk map for RDMA_READ processing
include/linux/sunrpc/svc_rdma.h | 1 +
net/sunrpc/xprtrdma/svc_rdma_recvfrom.c | 83 ++++++++++++++-----------------
2 files changed, 39 insertions(+), 45 deletions(-)
[PATCH 04/11] svcrdma: Move the DMA unmap logic to the CQ handler
net/sunrpc/xprtrdma/svc_rdma_transport.c | 20 ++++++++++++++------
1 files changed, 14 insertions(+), 6 deletions(-)
[PATCH 05/11] svcrdma: Add dma map count and WARN_ON
include/linux/sunrpc/svc_rdma.h | 1 +
net/sunrpc/xprtrdma/svc_rdma_recvfrom.c | 1 +
net/sunrpc/xprtrdma/svc_rdma_sendto.c | 3 +++
net/sunrpc/xprtrdma/svc_rdma_transport.c | 5 +++++
4 files changed, 10 insertions(+), 0 deletions(-)
[PATCH 06/11] svcrdma: Remove unneeded spin locks from __svc_rdma_free
net/sunrpc/xprtrdma/svc_rdma_transport.c | 4 ----
1 files changed, 0 insertions(+), 4 deletions(-)
[PATCH 07/11] svcrdma: Remove unused wait q from svcrdma_xprt structure
include/linux/sunrpc/svc_rdma.h | 1 -
1 files changed, 0 insertions(+), 1 deletions(-)
[PATCH 08/11] svcrdma: Limit ORD based on client's advertised IRD
net/sunrpc/xprtrdma/svc_rdma_transport.c | 16 ++++++++++++----
1 files changed, 12 insertions(+), 4 deletions(-)
[PATCH 09/11] svcrdma: Add flush_scheduled_work to module exit function
net/sunrpc/xprtrdma/svc_rdma.c | 1 +
1 files changed, 1 insertions(+), 0 deletions(-)
[PATCH 10/11] svcrdma: Create a kmem cache for the WR contexts
net/sunrpc/xprtrdma/svc_rdma.c | 23 +++++++++++++++++++----
1 files changed, 19 insertions(+), 4 deletions(-)
[PATCH 11/11] svcrdma: Change WR context get/put to use the kmem cache
include/linux/sunrpc/svc_rdma.h | 6 --
net/sunrpc/xprtrdma/svc_rdma_transport.c | 121 +++---------------------------
2 files changed, 12 insertions(+), 115 deletions(-)
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
next reply other threads:[~2008-07-03 2:27 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-07-03 2:27 Tom Tucker [this message]
2008-07-03 2:27 ` [PATCH 01/11] svcrdma: Add a type for keeping NFS RPC mapping Tom Tucker
2008-07-03 2:27 ` [PATCH 02/11] svcrdma: Use RPC reply map for RDMA_WRITE processing Tom Tucker
2008-07-03 2:27 ` [PATCH 03/11] svcrdma: Use reply and chunk map for RDMA_READ processing Tom Tucker
2008-07-03 2:27 ` [PATCH 04/11] svcrdma: Move the DMA unmap logic to the CQ handler Tom Tucker
2008-07-03 2:27 ` [PATCH 05/11] svcrdma: Add dma map count and WARN_ON Tom Tucker
2008-07-03 2:27 ` [PATCH 06/11] svcrdma: Remove unneeded spin locks from __svc_rdma_free Tom Tucker
2008-07-03 2:27 ` [PATCH 07/11] svcrdma: Remove unused wait q from svcrdma_xprt structure Tom Tucker
2008-07-03 2:27 ` [PATCH 08/11] svcrdma: Limit ORD based on client's advertised IRD Tom Tucker
2008-07-03 2:27 ` [PATCH 09/11] svcrdma: Add flush_scheduled_work to module exit function Tom Tucker
2008-07-03 2:27 ` [PATCH 10/11] svcrdma: Create a kmem cache for the WR contexts Tom Tucker
2008-07-03 2:27 ` [PATCH 11/11] svcrdma: Change WR context get/put to use the kmem cache Tom Tucker
2008-07-03 20:27 ` [PATCH 00/11] svcrdma: WR context management bug fixes and cleanup J. Bruce Fields
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=1215052063-7330-1-git-send-email-tom@opengridcomputing.com \
--to=tom@opengridcomputing.com \
--cc=bfields@fieldses.org \
--cc=linux-nfs@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