From 2116271a347d1181b5497602c2bfada1de8fd53b Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 20 May 2008 19:34:39 -0400 Subject: NFS: Add correct bounds checking to NFSv2 locks NFSv2 file locking currently fails the Connectathon tests, because the calls to the VFS locking code do not return an EINVAL error if the struct file_lock overflows the 32-bit boundaries. The problem is due to the fact that we occasionally call helpers from fs/locks.c in order to avoid RPC calls to the server when we know that a local process holds the lock. These helpers are, of course, always 64-bit enabled, so EINVAL is not returned in cases when it would if the call had gone to the NLM code. For consistency, we therefore add support for a bounds-checking helper. Signed-off-by: Trond Myklebust --- include/linux/nfs_xdr.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 24263bb8e0b..8d780de371f 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -832,6 +832,7 @@ struct nfs_rpc_ops { int (*file_open) (struct inode *, struct file *); int (*file_release) (struct inode *, struct file *); int (*lock)(struct file *, int, struct file_lock *); + int (*lock_check_bounds)(const struct file_lock *); void (*clear_acl_cache)(struct inode *); }; -- cgit v1.2.3 From b6b6152c46861dd914d0e6cea9c27df057d6e235 Mon Sep 17 00:00:00 2001 From: Olga Kornievskaia Date: Mon, 9 Jun 2008 16:51:31 -0400 Subject: rpc: bring back cl_chatty The cl_chatty flag alows us to control whether a given rpc client leaves "server X not responding, timed out" messages in the syslog. Such messages make sense for ordinary nfs clients (where an unresponsive server means applications on the mountpoint are probably hanging), but not for the callback client (which can fail more commonly, with the only result just of disabling some optimizations). Previously cl_chatty was removed, do to lack of users; reinstate it, and use it for the nfsd's callback client. Signed-off-by: J. Bruce Fields Signed-off-by: Trond Myklebust --- include/linux/sunrpc/clnt.h | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index 6fff7f82ef1..764fd4c286e 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -42,7 +42,8 @@ struct rpc_clnt { unsigned int cl_softrtry : 1,/* soft timeouts */ cl_discrtry : 1,/* disconnect before retry */ - cl_autobind : 1;/* use getport() */ + cl_autobind : 1,/* use getport() */ + cl_chatty : 1;/* be verbose */ struct rpc_rtt * cl_rtt; /* RTO estimator data */ const struct rpc_timeout *cl_timeout; /* Timeout strategy */ @@ -114,6 +115,7 @@ struct rpc_create_args { #define RPC_CLNT_CREATE_NONPRIVPORT (1UL << 3) #define RPC_CLNT_CREATE_NOPING (1UL << 4) #define RPC_CLNT_CREATE_DISCRTRY (1UL << 5) +#define RPC_CLNT_CREATE_QUIET (1UL << 6) struct rpc_clnt *rpc_create(struct rpc_create_args *args); struct rpc_clnt *rpc_bind_new_program(struct rpc_clnt *, -- cgit v1.2.3 From a486aeda9b2b0d944aecce7871b3186379b898de Mon Sep 17 00:00:00 2001 From: "\\\\\\\"J. Bruce Fields\\\\\\" Date: Mon, 9 Jun 2008 16:51:35 -0400 Subject: rpc: minor cleanup of scheduler callback code Try to make the comment here a little more clear and concise. Also, this macro definition seems unnecessary. Signed-off-by: J. Bruce Fields Signed-off-by: Trond Myklebust --- include/linux/sunrpc/sched.h | 1 - 1 file changed, 1 deletion(-) (limited to 'include') diff --git a/include/linux/sunrpc/sched.h b/include/linux/sunrpc/sched.h index d1a5c8c1a0f..64981a2f1ca 100644 --- a/include/linux/sunrpc/sched.h +++ b/include/linux/sunrpc/sched.h @@ -135,7 +135,6 @@ struct rpc_task_setup { #define RPC_IS_SWAPPER(t) ((t)->tk_flags & RPC_TASK_SWAPPER) #define RPC_DO_ROOTOVERRIDE(t) ((t)->tk_flags & RPC_TASK_ROOTCREDS) #define RPC_ASSASSINATED(t) ((t)->tk_flags & RPC_TASK_KILLED) -#define RPC_DO_CALLBACK(t) ((t)->tk_callback != NULL) #define RPC_IS_SOFT(t) ((t)->tk_flags & RPC_TASK_SOFT) #define RPC_TASK_RUNNING 0 -- cgit v1.2.3 From 46cb650c224bb8e64a749090105d74b9e8eda669 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 11 Jun 2008 16:32:46 -0400 Subject: NFS: Remove the redundant file_open entry from struct nfs_rpc_ops All instances are set to nfs_open(), so we should just remove the redundant indirection. Ditto for the file_release op Signed-off-by: Trond Myklebust --- include/linux/nfs_xdr.h | 2 -- 1 file changed, 2 deletions(-) (limited to 'include') diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 8d780de371f..8c77c11224d 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -829,8 +829,6 @@ struct nfs_rpc_ops { int (*write_done) (struct rpc_task *, struct nfs_write_data *); void (*commit_setup) (struct nfs_write_data *, struct rpc_message *); int (*commit_done) (struct rpc_task *, struct nfs_write_data *); - int (*file_open) (struct inode *, struct file *); - int (*file_release) (struct inode *, struct file *); int (*lock)(struct file *, int, struct file_lock *); int (*lock_check_bounds)(const struct file_lock *); void (*clear_acl_cache)(struct inode *); -- cgit v1.2.3 From 34e8f92831cb5c40b3137e47a3daf4c09016ef02 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 12 Jun 2008 12:32:25 -0400 Subject: NFS: Move fs/nfs/iostat.h to include/linux The fs/nfs/iostat.h header has definitions that were designed to be exposed to user space. Move these definitions under include/linux so user space can use the definitions in applications that read /proc/self/mountstats. Also address a handful of coding style issues called out by checkpatch.pl in fs/nfs/iostat.h. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/nfs_iostat.h | 119 +++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 119 insertions(+) create mode 100644 include/linux/nfs_iostat.h (limited to 'include') diff --git a/include/linux/nfs_iostat.h b/include/linux/nfs_iostat.h new file mode 100644 index 00000000000..1cb9a3fed2b --- /dev/null +++ b/include/linux/nfs_iostat.h @@ -0,0 +1,119 @@ +/* + * User-space visible declarations for NFS client per-mount + * point statistics + * + * Copyright (C) 2005, 2006 Chuck Lever + * + * NFS client per-mount statistics provide information about the + * health of the NFS client and the health of each NFS mount point. + * Generally these are not for detailed problem diagnosis, but + * simply to indicate that there is a problem. + * + * These counters are not meant to be human-readable, but are meant + * to be integrated into system monitoring tools such as "sar" and + * "iostat". As such, the counters are sampled by the tools over + * time, and are never zeroed after a file system is mounted. + * Moving averages can be computed by the tools by taking the + * difference between two instantaneous samples and dividing that + * by the time between the samples. + */ + +#ifndef _LINUX_NFS_IOSTAT +#define _LINUX_NFS_IOSTAT + +#define NFS_IOSTAT_VERS "1.0" + +/* + * NFS byte counters + * + * 1. SERVER - the number of payload bytes read from or written + * to the server by the NFS client via an NFS READ or WRITE + * request. + * + * 2. NORMAL - the number of bytes read or written by applications + * via the read(2) and write(2) system call interfaces. + * + * 3. DIRECT - the number of bytes read or written from files + * opened with the O_DIRECT flag. + * + * These counters give a view of the data throughput into and out + * of the NFS client. Comparing the number of bytes requested by + * an application with the number of bytes the client requests from + * the server can provide an indication of client efficiency + * (per-op, cache hits, etc). + * + * These counters can also help characterize which access methods + * are in use. DIRECT by itself shows whether there is any O_DIRECT + * traffic. NORMAL + DIRECT shows how much data is going through + * the system call interface. A large amount of SERVER traffic + * without much NORMAL or DIRECT traffic shows that applications + * are using mapped files. + * + * NFS page counters + * + * These count the number of pages read or written via nfs_readpage(), + * nfs_readpages(), or their write equivalents. + * + * NB: When adding new byte counters, please include the measured + * units in the name of each byte counter to help users of this + * interface determine what exactly is being counted. + */ +enum nfs_stat_bytecounters { + NFSIOS_NORMALREADBYTES = 0, + NFSIOS_NORMALWRITTENBYTES, + NFSIOS_DIRECTREADBYTES, + NFSIOS_DIRECTWRITTENBYTES, + NFSIOS_SERVERREADBYTES, + NFSIOS_SERVERWRITTENBYTES, + NFSIOS_READPAGES, + NFSIOS_WRITEPAGES, + __NFSIOS_BYTESMAX, +}; + +/* + * NFS event counters + * + * These counters provide a low-overhead way of monitoring client + * activity without enabling NFS trace debugging. The counters + * show the rate at which VFS requests are made, and how often the + * client invalidates its data and attribute caches. This allows + * system administrators to monitor such things as how close-to-open + * is working, and answer questions such as "why are there so many + * GETATTR requests on the wire?" + * + * They also count anamolous events such as short reads and writes, + * silly renames due to close-after-delete, and operations that + * change the size of a file (such operations can often be the + * source of data corruption if applications aren't using file + * locking properly). + */ +enum nfs_stat_eventcounters { + NFSIOS_INODEREVALIDATE = 0, + NFSIOS_DENTRYREVALIDATE, + NFSIOS_DATAINVALIDATE, + NFSIOS_ATTRINVALIDATE, + NFSIOS_VFSOPEN, + NFSIOS_VFSLOOKUP, + NFSIOS_VFSACCESS, + NFSIOS_VFSUPDATEPAGE, + NFSIOS_VFSREADPAGE, + NFSIOS_VFSREADPAGES, + NFSIOS_VFSWRITEPAGE, + NFSIOS_VFSWRITEPAGES, + NFSIOS_VFSGETDENTS, + NFSIOS_VFSSETATTR, + NFSIOS_VFSFLUSH, + NFSIOS_VFSFSYNC, + NFSIOS_VFSLOCK, + NFSIOS_VFSRELEASE, + NFSIOS_CONGESTIONWAIT, + NFSIOS_SETATTRTRUNC, + NFSIOS_EXTENDWRITE, + NFSIOS_SILLYRENAME, + NFSIOS_SHORTREAD, + NFSIOS_SHORTWRITE, + NFSIOS_DELAY, + __NFSIOS_COUNTSMAX, +}; + +#endif /* _LINUX_NFS_IOSTAT */ -- cgit v1.2.3 From e468bae97d243fe0e1515abaa1f7d0edf1476ad0 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 13 Jun 2008 13:25:22 -0400 Subject: NFS: Allow redirtying of a completed unstable write. Currently, if an unstable write completes, we cannot redirty the page in order to reflect a new change in the page data until after we've sent a COMMIT request. This patch allows a page rewrite to proceed without the unnecessary COMMIT step, putting it immediately back onto the dirty page list, undoing the VM unstable write accounting, and removing the NFS_PAGE_TAG_COMMIT tag from the NFS radix tree. Signed-off-by: Trond Myklebust --- include/linux/nfs_page.h | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/linux/nfs_page.h b/include/linux/nfs_page.h index a1676e19e49..3c60685d972 100644 --- a/include/linux/nfs_page.h +++ b/include/linux/nfs_page.h @@ -27,9 +27,12 @@ /* * Valid flags for a dirty buffer */ -#define PG_BUSY 0 -#define PG_NEED_COMMIT 1 -#define PG_NEED_RESCHED 2 +enum { + PG_BUSY = 0, + PG_CLEAN, + PG_NEED_COMMIT, + PG_NEED_RESCHED, +}; struct nfs_inode; struct nfs_page { -- cgit v1.2.3 From ce3b7e1906ebbe96753fe090b36de6ffb8e0e0e7 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 23 Jun 2008 12:36:53 -0400 Subject: NFS: Add string length argument to nfs_parse_server_address To make nfs_parse_server_address() more generally useful, allow it to accept input strings that are not terminated with '\0'. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/inet.h | 7 +++++++ 1 file changed, 7 insertions(+) (limited to 'include') diff --git a/include/linux/inet.h b/include/linux/inet.h index 1354080cf8c..4cca05c9678 100644 --- a/include/linux/inet.h +++ b/include/linux/inet.h @@ -44,6 +44,13 @@ #include +/* + * These mimic similar macros defined in user-space for inet_ntop(3). + * See /usr/include/netinet/in.h . + */ +#define INET_ADDRSTRLEN (16) +#define INET6_ADDRSTRLEN (48) + extern __be32 in_aton(const char *str); extern int in4_pton(const char *src, int srclen, u8 *dst, int delim, const char **end); extern int in6_pton(const char *src, int srclen, u8 *dst, int delim, const char **end); -- cgit v1.2.3 From 259875efed06d6936f54c9a264e868937f1bc217 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 2 Jul 2008 14:43:47 -0400 Subject: NFS: set transport defaults after mount option parsing is finished Move the UDP/TCP default timeo/retrans settings for text mounts to nfs_init_timeout_values(), which was were they were always being initialised (and sanity checked) for binary mounts. Document the default timeout values using appropriate #defines. Ensure that we initialise and sanity check the transport protocols that may have been specified by the user. Signed-off-by: Trond Myklebust --- include/linux/nfs_fs.h | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'include') diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h index 27d6a8d98ce..830d9cc8cdc 100644 --- a/include/linux/nfs_fs.h +++ b/include/linux/nfs_fs.h @@ -12,6 +12,11 @@ #include /* Default timeout values */ +#define NFS_DEF_UDP_TIMEO (11) +#define NFS_DEF_UDP_RETRANS (3) +#define NFS_DEF_TCP_TIMEO (600) +#define NFS_DEF_TCP_RETRANS (2) + #define NFS_MAX_UDP_TIMEOUT (60*HZ) #define NFS_MAX_TCP_TIMEOUT (600*HZ) -- cgit v1.2.3 From 0e0cab744b17a70ef0f08d818d66935feade7cad Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 26 Jun 2008 17:47:12 -0400 Subject: NFS: use documenting macro constants for initializing ac{reg, dir}{min, max} Clean up. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/nfs_fs.h | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'include') diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h index 830d9cc8cdc..29d26191873 100644 --- a/include/linux/nfs_fs.h +++ b/include/linux/nfs_fs.h @@ -20,6 +20,11 @@ #define NFS_MAX_UDP_TIMEOUT (60*HZ) #define NFS_MAX_TCP_TIMEOUT (600*HZ) +#define NFS_DEF_ACREGMIN (3) +#define NFS_DEF_ACREGMAX (60) +#define NFS_DEF_ACDIRMIN (30) +#define NFS_DEF_ACDIRMAX (60) + /* * When flushing a cluster of dirty pages, there can be different * strategies: -- cgit v1.2.3 From c2e1b09ff237c0a3687b9a804cc8bf489743cffc Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 14 Jul 2008 16:03:30 -0400 Subject: SUNRPC: Support registering IPv6 interfaces with local rpcbind daemon Introduce a new API to register RPC services on IPv6 interfaces to allow the NFS server and lockd to advertise on IPv6 networks. Unlike rpcb_register(), the new rpcb_v4_register() function uses rpcbind protocol version 4 to contact the local rpcbind daemon. The version 4 SET/UNSET procedures allow services to register address families besides AF_INET, register at specific network interfaces, and register transport protocols besides UDP and TCP. All of this functionality is exposed via the new rpcb_v4_register() kernel API. A user-space rpcbind daemon implementation that supports version 4 of the rpcbind protocol is required in order to make use of this new API. Note that rpcbind version 3 is sufficient to support the new rpcbind facilities listed above, but most extant implementations use version 4. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields Signed-off-by: Trond Myklebust --- include/linux/sunrpc/clnt.h | 3 +++ 1 file changed, 3 insertions(+) (limited to 'include') diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index 764fd4c286e..e5bfe01ee30 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -125,6 +125,9 @@ void rpc_shutdown_client(struct rpc_clnt *); void rpc_release_client(struct rpc_clnt *); int rpcb_register(u32, u32, int, unsigned short, int *); +int rpcb_v4_register(const u32 program, const u32 version, + const struct sockaddr *address, + const char *netid, int *result); int rpcb_getport_sync(struct sockaddr_in *, u32, u32, int); void rpcb_getport_async(struct rpc_task *); -- cgit v1.2.3