Chuck Lever [Fri, 11 Jan 2008 22:09:59 +0000 (17:09 -0500)]
NLM/NFS: Use cached nlm_host when calling nlmclnt_proc()
Now that each NFS mount point caches its own nlm_host structure, it can be
passed to nlmclnt_proc() for each lock request. By pinning an nlm_host for
each mount point, we trade the overhead of looking up or creating a fresh
nlm_host struct during every NLM procedure call for a little extra memory.
We also restrict the nlmclnt_proc symbol to limit the use of this call to
in-tree modules.
Note that nlm_lookup_host() (just removed from the client's per-request
NLM processing) could also trigger an nlm_host garbage collection. Now
client-side nlm_host garbage collection occurs only during NFS mount
processing. Since the NFS client now holds a reference on these nlm_host
structures, they wouldn't have been affected by garbage collection
anyway.
Given that nlm_lookup_host() reorders the global nlm_host chain after
every successful lookup, and that a garbage collection could be triggered
during the call, we've removed a significant amount of per-NLM-request
CPU processing overhead.
Sidebar: there are only a few remaining references to the internals of
NFS inodes in the client-side NLM code. The only references I found are
related to extracting or comparing the inode's file handle via NFS_FH().
One is in nlmclnt_grant(); the other is in nlmclnt_setlockargs().
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Fri, 11 Jan 2008 22:09:52 +0000 (17:09 -0500)]
NFS: Invoke nlmclnt_init during NFS mount processing
Cache an appropriate nlm_host structure in the NFS client's mount point
metadata for later use.
Note that there is no need to set NFS_MOUNT_NONLM in the error case -- if
nfs_start_lockd() returns a non-zero value, its callers ensure that the
mount request fails outright.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Fri, 11 Jan 2008 22:09:44 +0000 (17:09 -0500)]
NLM: Introduce external nlm_host set-up and tear-down functions
We would like to remove the per-lock-operation nlm_lookup_host() call from
nlmclnt_proc().
The new architecture pins an nlm_host structure to each NFS client
superblock that has the "lock" mount option set. The NFS client passes
in the pinned nlm_host structure during each call to nlmclnt_proc(). NFS
client unmount processing "puts" the nlm_host so it can be garbage-
collected later.
This patch introduces externally callable NLM functions that handle
mount-time nlm_host set up and tear-down.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Mon, 7 Jan 2008 23:34:48 +0000 (18:34 -0500)]
SUNRPC: fewer conditionals in the format_ip_address routines
Clean up: have the set up routines explicitly pass the strings to be used
for the transport name and NETID. This removes a number of conditionals
and dependencies on rpc_xprt.prot, which is overloaded.
Tighten up type checking on the address_strings array while we're at it.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Thu, 20 Dec 2007 19:55:04 +0000 (14:55 -0500)]
NFS: nfs_write_end clean up
Clean up: commit 4899f9c8 added nfs_write_end(), which introduces a
conditional expression that returns an unsigned integer in one arm and
a signed integer in the other.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Thu, 20 Dec 2007 19:54:42 +0000 (14:54 -0500)]
NFS: Fix use of copy_to_user() in idmap_pipe_upcall
The idmap_pipe_upcall() function expects the copy_to_user() function to
return a negative error value if the call fails, but copy_to_user()
returns an unsigned long number of bytes that couldn't be copied.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Thu, 20 Dec 2007 19:54:27 +0000 (14:54 -0500)]
SUNRPC: Fix use of copy_to_user() in gss_pipe_upcall()
The gss_pipe_upcall() function expects the copy_to_user() function to
return a negative error value if the call fails, but copy_to_user()
returns an unsigned long number of bytes that couldn't be copied.
Can rpc_pipefs actually retry a partially completed upcall read? If
not, then gss_pipe_upcall() should punt any partial read, just like the
upcall logic in net/sunrpc/cache.c.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Trond Myklebust [Thu, 3 Jan 2008 21:29:06 +0000 (16:29 -0500)]
NFS: Fix the 'proto=' mount option
Currently, if you have a server mounted using networking protocol, you
cannot specify a different value using the 'proto=' option on another
mountpoint.
Trond Myklebust [Thu, 20 Dec 2007 21:03:55 +0000 (16:03 -0500)]
SUNRPC: Add support for per-client timeout values
In order to be able to support setting the timeo and retrans parameters on
a per-mountpoint basis, we move the rpc_timeout structure into the
rpc_clnt.
Chuck Lever [Mon, 10 Dec 2007 19:59:28 +0000 (14:59 -0500)]
NFS: Support non-IPv4 addresses in nfs_parsed_mount_data
Replace the nfs_server and mount_server address fields in the
nfs_parsed_mount_data structure with a "struct sockaddr_storage"
instead of a "struct sockaddr_in".
Chuck Lever [Mon, 10 Dec 2007 19:59:21 +0000 (14:59 -0500)]
NFS: Refactor mount option address parsing into separate function
Refactor the logic to parse incoming text-based IP addresses. Use the
in4_pton() function instead of the older in_aton(), following the lead
of the in-kernel CIFS client.
Later we'll add IPv6 address parsing using the matching in6_pton()
function. For now we can't allow IPv6 address parsing: we must expand
the size of the address storage fields in the nfs_parsed_mount_options
struct before we can parse and store IPv6 addresses.
Chuck Lever [Mon, 10 Dec 2007 19:59:13 +0000 (14:59 -0500)]
NFS: Remove the NIPQUAD from nfs_try_mount
In the name of address family compatibility, we can't have the NIP_FMT and
NIPQUAD macros in nfs_try_mount(). Instead, we can make use of an unused
mount option to display the mount server's hostname.
Chuck Lever [Mon, 10 Dec 2007 19:59:06 +0000 (14:59 -0500)]
NFS: Adjust nfs_clone_mount structure to store "struct sockaddr *"
Change the addr field in the nfs_clone_mount structure to store a "struct
sockaddr *" to support non-IPv4 addresses in the NFS client.
Note this is mostly a cosmetic change, and does not actually allow
referrals using IPv6 addresses. The existing referral code assumes that
the server returns a string that represents an IPv4 address. This code
needs to support hostnames and IPv6 addresses as well as IPv4 addresses,
thus it will need to be reorganized completely (to handle DNS resolution
in user space).
Chuck Lever [Mon, 10 Dec 2007 19:58:59 +0000 (14:58 -0500)]
NFS: Change nfs4_set_client() to accept struct sockaddr *
Adjust the arguments and callers of nfs4_set_client() to pass a "struct
sockaddr *" instead of a "struct sockaddr_in *" to support non-IPv4
addresses in the NFS client.
Chuck Lever [Mon, 10 Dec 2007 19:58:44 +0000 (14:58 -0500)]
NFS: Change nfs_find_client() to take "struct sockaddr *"
Adjust arguments and callers of nfs_find_client() to pass a
"struct sockaddr *" instead of "struct sockaddr_in *" to support non-IPv4
addresses.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Cc: Aurelien Charbon <aurelien.charbon@ext.bull.net>
Trond: Also fix up protocol version number argument in nfs_find_client() to
use the correct u32 type.
Chuck Lever [Mon, 10 Dec 2007 19:58:15 +0000 (14:58 -0500)]
NFS: Expand server address storage in nfs_client struct
Prepare for managing larger addresses in the NFS client by widening the
nfs_client struct's cl_addr field.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Cc: Aurelien Charbon <aurelien.charbon@ext.bull.net>
(Modified to work with the new parameters for nfs_alloc_client) Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Mon, 10 Dec 2007 19:57:53 +0000 (14:57 -0500)]
NFS: Make setting a port number agostic
We'll need to set the port number of an AF_INET or AF_INET6 address in
several places in fs/nfs/super.c, so introduce a helper that can manage
this for us. We put this helper to immediate use.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Mon, 10 Dec 2007 19:57:38 +0000 (14:57 -0500)]
NFS: Add support for AF_INET6 addresses in nfs_compare_super()
Refactor nfs_compare_super() and add AF_INET6 support.
Replace the generic memcmp() to document explicitly what parts of the
addresses must match in this check, and make the comparison independent
of the lengths of both addresses.
A side benefit is both tests are more computationally efficient than a
memcmp().
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Mon, 10 Dec 2007 19:57:16 +0000 (14:57 -0500)]
NFS: eliminate NIPQUAD(clp->cl_addr.sin_addr)
To ensure the NFS client displays IPv6 addresses properly, replace
address family-specific NIPQUAD() invocations with a call to the RPC
client to get a formatted string representing the remote peer's
address.
Chuck Lever [Mon, 10 Dec 2007 19:57:09 +0000 (14:57 -0500)]
NFS: Enable NFS client to generate CLIENTID strings with IPv6 addresses
We recently added methods to RPC transports that provide string versions of
the remote peer address information. Convert the NFSv4 SETCLIENTID
procedure to use those methods instead of building the client ID out of
whole cloth.
Chuck Lever [Mon, 10 Dec 2007 19:56:54 +0000 (14:56 -0500)]
NFS: Ensure NFSv4 SETCLIENTID send buffer is large enough
Ensure that the RPC buffer size specified for NFSv4 SETCLIENTID procedures
matches what we are encoding into the buffer. See the definition of
struct nfs4_setclientid {} and the encode_setclientid() function.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Mon, 10 Dec 2007 19:56:46 +0000 (14:56 -0500)]
SUNRPC: Move universal address definitions to global header
Universal addresses are defined in RFC 1833 and clarified in RFC 3530. We
need to use them in several places in the NFS and RPC clients, so move the
relevant definition and block comment to an appropriate global include
file.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Mon, 10 Dec 2007 19:56:24 +0000 (14:56 -0500)]
SUNRPC: rpc_create() default hostname should support AF_INET6 addresses
If the ULP doesn't pass a hostname string to rpc_create(), it manufactures
one based on the passed-in address. Be smart enough to handle an AF_INET6
address properly in this case.
Move the default servername logic before the xprt_create_transport() call
to simplify error handling in rpc_create().
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Fri, 26 Oct 2007 17:32:45 +0000 (13:32 -0400)]
NFS: Clean up address comparison in __nfs_find_client()
The address comparison in the __nfs_find_client() function is deceptive.
It uses a memcmp() to check a pair of u32 fields for equality. Not only is
this inefficient, but usually memcmp() is used for comparing two *whole*
sockaddr_in's (which includes comparisons of the address family and port
number), so it's easy to mistake the comparison here for a whole sockaddr
comparison, which it isn't.
So for clarity and efficiency, we replace the memcmp() with a simple test
for equality between the two s_addr fields. This should have no
behavioral effect.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Fri, 26 Oct 2007 17:32:40 +0000 (13:32 -0400)]
NFS: Clean up: copy hostname with kstrndup during mount processing
Clean up: mount option parsing uses kstrndup in several places, rather than
using kzalloc. Replace the few remaining uses of kzalloc with kstrndup,
for consistency.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Fri, 26 Oct 2007 17:32:29 +0000 (13:32 -0400)]
NFS: Remove support for the 'mountprog' option
Remove the mount option that allows users to specify an alternate mountd
program number. The client hasn't support setting an alternate mountd
program number for a very long time.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Fri, 26 Oct 2007 17:32:24 +0000 (13:32 -0400)]
NFS: Remove support for the 'nfsprog' option
Remove the mount option that allows users to specify an alternate NFS
program number. The client hasn't support setting an alternate NFS
program number for a very long time.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Fri, 26 Oct 2007 17:32:19 +0000 (13:32 -0400)]
NFS: Ensure that NFS version 4 mounts use NFS_PORT if nfsport wasn't set
Text-based mount option parsing introduced a minor regression in the
behavior of NFS version 4 mounts. NFS version 4 is not supposed to require
a running rpcbind service on the server in order for a mount to succeed.
In other words, if the mount options don't specify a port number, the port
number is supposed to default to 2049. For earlier versions of NFS, the
default port number was zero in order to cause the RPC client to autobind
to the server's NFS service.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Fri, 26 Oct 2007 17:32:13 +0000 (13:32 -0400)]
NFS: Prevent nfs_getattr() hang during heavy write workloads
POSIX requires that ctime and mtime, as reported by the stat(2) call,
reflect the activity of the most recent write(2). To that end, nfs_getattr()
flushes pending dirty writes to a file before doing a GETATTR to allow the
NFS server to set the file's size, ctime, and mtime properly.
However, nfs_getattr() can be starved when a constant stream of application
writes to a file prevents nfs_wb_nocommit() from completing. This usually
results in hangs of programs doing a stat against an NFS file that is being
written. "ls -l" is a common victim of this behavior.
To prevent starvation, hold the file's i_mutex in nfs_getattr() to
freeze applications writes temporarily so the client can more quickly obtain
clean values for a file's size, mtime, and ctime.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Fri, 26 Oct 2007 17:31:47 +0000 (13:31 -0400)]
NFS: Ensure nfs_wcc_update_inode always converts file size to loff_t
The nfs_wcc_update_inode() function omits logic to convert the type of
the NFS on-the-wire value of a file's size (__u64) to the type of file
size value stored in struct inode (loff_t, which is signed).
Everywhere else in the NFS client I checked already correctly converts the
file size type.
This effects only very large files.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Fri, 26 Oct 2007 17:30:49 +0000 (13:30 -0400)]
SUNRPC: Fix an unnecessary implicit type cast in rpcrdma_count_chunks()
Nit: rl_nchunks is an unsigned integer, so pass it into
rpcrdma_count_chunks() via an unsigned integer argument. This eliminates
a harmless mixed sign comparison in rpcrdma_count_chunks()
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Cc: Thomas Talpey <Thomas.Talpey@netapp.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Fri, 26 Oct 2007 17:30:43 +0000 (13:30 -0400)]
SUNRPC: Prevent mixed sign comparisons in rpcrdma_convert_iovs()
Keep the type of the buffer position the same during iovec conversion to
reduce the likelihood of unexpected results from comparisons and length
computations.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Cc: Thomas Talpey <Thomas.Talpey@netapp.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
The sunrpc client exports are not meant to be part of any official kernel
API: they can change at the drop of a hat. Mark them as internal functions
using EXPORT_SYMBOL_GPL.
Trond Myklebust [Tue, 6 Nov 2007 23:44:20 +0000 (18:44 -0500)]
SUNRPC: Rename xprt_disconnect()
xprt_disconnect() should really only be called when the transport shutdown
is completed, and it is time to wake up any pending tasks. Rename it to
xprt_disconnect_done() in order to reflect the semantical change.
Trond Myklebust [Tue, 6 Nov 2007 23:40:12 +0000 (18:40 -0500)]
SUNRPC: Make call_status()/call_decode() call xprt_force_disconnect()
Move the calls to xprt_disconnect() over to xprt_force_disconnect() in
order to enable the transport layer to manage the state of the
XPRT_CONNECTED flag.
Ditto in xs_tcp_read_fraghdr().
Trond Myklebust [Tue, 6 Nov 2007 23:28:53 +0000 (18:28 -0500)]
SUNRPC: xprt_autoclose() should not call xprt_disconnect()
The transport layer should do that itself whenever appropriate.
Note that the RDMA transport already assumes that it needs to call
xprt_disconnect in xprt_rdma_close().
For TCP sockets, we want to call xprt_disconnect() only after the
connection has been closed by both ends.
Trond Myklebust [Mon, 5 Nov 2007 20:44:12 +0000 (15:44 -0500)]
SUNRPC: Use shutdown() instead of close() when disconnecting a TCP socket
By using shutdown() rather than close() we allow the RPC client to wait
for the TCP close handshake to complete before we start trying to reconnect
using the same port.
We use shutdown(SHUT_WR) only instead of shutting down both directions,
however we wait until the server has closed the connection on its side.
Trond Myklebust [Mon, 5 Nov 2007 22:42:39 +0000 (17:42 -0500)]
SUNRPC: Allow the client to detect if the TCP connection is closed
Add an xprt->state bit to enable the TCP ->state_change() method to signal
whether or not the TCP connection is in the process of closing down.
This will to be used by the reconnection logic in a separate patch.
Trond Myklebust [Mon, 5 Nov 2007 22:40:58 +0000 (17:40 -0500)]
SUNRPC: Fix TCP rebinding logic
Currently the TCP rebinding logic assumes that if we're not using a
reserved port, then we don't need to reconnect on the same port if a
disconnection event occurs. This breaks most RPC duplicate reply cache
implementations.
Also take into account the fact that xprt_min_resvport and
xprt_max_resvport may change while we're reconnecting, since the user may
change them at any time via the sysctls. Ensure that we check the port
boundaries every time we loop in xs_bind4/xs_bind6. Also ensure that if the
boundaries change, we only scan the ports a maximum of 2 times.
Trond Myklebust [Tue, 8 Jan 2008 22:56:07 +0000 (17:56 -0500)]
NFSv4: Clean up the OPEN/CLOSE serialisation code
Reduce the time spent locking the rpc_sequence structure by queuing the
nfs_seqid only when we are ready to take the lock (when calling
nfs_wait_on_sequence).
projects / linux-2.6-omap-h63xx.git / log