NFS: introduce mechanism for tracking NFS client metrics

Add a per-superblock performance counter facility to the NFS client.  This
facility mimics the counters available for block devices and for
networking.  Expose these new counters via the new /proc/self/mountstats
interface.

Thanks to Andrew Morton and Trond Myklebust for their review and comments.

Test plan:
fsx and iozone on UP and SMP systems, with and without pre-emption.  Watch
for memory overwrite bugs, and performance loss (significantly more CPU
required per op).

Signed-off-by: Chuck Lever <cel@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>

1 files changed, 3 insertions, 0 deletions

diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h
index a522ab97358d..d65e69a06b72 100644
--- a/include/linux/nfs_fs_sb.h
+++ b/include/linux/nfs_fs_sb.h
@@ -4,6 +4,8 @@
 #include <linux/list.h>
 #include <linux/backing-dev.h>
 
+struct nfs_iostats;
+
 /*
  * NFS client parameters stored in the superblock.
  */
@@ -12,6 +14,7 @@ struct nfs_server {
         struct rpc_clnt *       client_sys;     /* 2nd handle for FSINFO */
         struct rpc_clnt *       client_acl;     /* ACL RPC client handle */
         struct nfs_rpc_ops *    rpc_ops;        /* NFS protocol vector */
+        struct nfs_iostats *    io_stats;       /* I/O statistics */
         struct backing_dev_info backing_dev_info;
         int                     flags;          /* various flags */
         unsigned int            caps;           /* server capabilities */