[PATCH] NFS: use a constant value for TCP retransmit timeouts
authorChuck Lever <cel@citi.umich.edu>
Thu, 11 Aug 2005 20:25:14 +0000 (16:25 -0400)
committerTrond Myklebust <Trond.Myklebust@netapp.com>
Fri, 23 Sep 2005 16:38:06 +0000 (12:38 -0400)
commiteab5c084b858fd95a873fc2b97de9a9ad937b4ed
tree6ba8287570ecc83fed1512bd4901df979221c2ab
parentda35187801732397a7e05fb9e77f3700cc35f5db
[PATCH] NFS: use a constant value for TCP retransmit timeouts

 Implement a best practice: don't use exponential backoff when computing
 retransmit timeout values on TCP connections, but simply retransmit
 at regular intervals.

 This also fixes a bug introduced when xprt_reset_majortimeo() was added.

 Test-plan:
 Enable RPC debugging and watch timeout behavior on a NFS/TCP mount.

 Version: Thu, 11 Aug 2005 16:02:19 -0400

Signed-off-by: Chuck Lever <cel@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
fs/nfs/inode.c
net/sunrpc/xprt.c