[NFS] bk update ready for 2.5.69

Trond Myklebust (trond.myklebust@fys.uio.no)
Wed, 7 May 2003 15:21:04 +0200


Linus, please do a

bk pull http://nfsclient.bkbits.net/linux-2.5

This will update the following files:

fs/fs-writeback.c | 3
fs/lockd/host.c | 14 --
fs/lockd/mon.c | 1
fs/nfs/dir.c | 2
fs/nfs/file.c | 2
fs/nfs/write.c | 6 -
include/linux/sunrpc/xdr.h | 5
include/linux/sunrpc/xprt.h | 2
mm/page-writeback.c | 2
net/sunrpc/clnt.c | 58 ++++++-----
net/sunrpc/sched.c | 5
net/sunrpc/xdr.c | 111 ++++++++++++++++++++-
net/sunrpc/xprt.c | 229 ++++++++++++++++++++++----------------------
13 files changed, 281 insertions(+), 159 deletions(-)

through these ChangeSets:

<trond.myklebust@fys.uio.no> (03/05/07 1.1098)
UDP and TCP zero copy code for the NFS client. The main interest
of this patch is to eliminate the use of xdr_kmap() and xdr_unmap()
by replacing them with MSG_MORE. xdr_kmap() is deadlock-prone
due to the fact that it has to kmap() several pages at the same time.

<trond.myklebust@fys.uio.no> (03/05/07 1.1097)
Ensure that Lockd and the NSM (statd) clients always use privileged
ports. Remove the existing code to temporarily raise privileges in
fs/lockd/host.c, and use the new code in net/sunrpc/xprt.c

There should no longer be a need to temporarily change the fsuid.
Remove this feature.

<cel@citi.umich.edu> (03/05/07 1.1096)
the recently-applied patch to fix the rpc_show_tasks() Oops is incomplete.
this applies over 2.5.68 and should address all of the issues in
rpc_show_tasks().

<trond.myklebust@fys.uio.no> (03/05/07 1.1095)
Add the sk->callback_lock spinlocks to the RPC socket callbacks
in order to protect the socket from being released by one
CPU while the other is in a soft interrupt.

<trond.myklebust@fys.uio.no> (03/05/07 1.1094)
Ensure that if we need to reconnect the socket, we also resend
the entire RPC message

Assorted TCP reconnection fixes.

Temporarily raise the necessary CAP_NET_BIND_SERVICE capability
if we need to bind the socket to a reserved port during a TCP
reconnection. Check for CAP_NET_BIND_SERVICE at mount time.

<trond.myklebust@fys.uio.no> (03/05/07 1.1093)
Don't use an RPC child process when reconnecting to a TCP server.
This is more efficient, and also fixes an existing deadlock
situation in which the child could be waiting for an xprt_write_lock
that was being held by the parent.

<trond.myklebust@fys.uio.no> (03/05/07 1.1092)
Fix a TCP race: check whether or not the socket has been disconnected
before we allow an RPC request to wait on a reply.

<trond.myklebust@fys.uio.no> (03/05/07 1.1091)
Fix typos in close-to-open cache consistency checking.

<trond.myklebust@fys.uio.no> (03/05/07 1.1090)
Decrement the nr_unstable page state after the COMMIT RPC call
completes instead of before. This ensures that writeback
WB_SYNC_ALL does wait on completion.

Don't overreport the number of pages we wrote out. It is safer to
underreport.

Fix missing NFSv3 unstable write accounting in fs/fs-writeback.c
and mm/page-writeback.c

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/