From f8407299f61681e9733f06c7214e81002cb459b3 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 24 Jul 2013 11:59:49 -0400 Subject: NFS Remove unused authflavour parameter from init_client Signed-off-by: Andy Adamson Signed-off-by: Trond Myklebust --- include/linux/nfs_xdr.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 8651574a305b..ddc3e32178c3 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -1442,7 +1442,7 @@ struct nfs_rpc_ops { struct nfs_client *(*alloc_client) (const struct nfs_client_initdata *); struct nfs_client * (*init_client) (struct nfs_client *, const struct rpc_timeout *, - const char *, rpc_authflavor_t); + const char *); void (*free_client) (struct nfs_client *); struct nfs_server *(*create_server)(struct nfs_mount_info *, struct nfs_subversion *); struct nfs_server *(*clone_server)(struct nfs_server *, struct nfs_fh *, -- cgit v1.2.3 From 1dada8e1f94e863a94f6622f0ddb5d1b01420150 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 26 Aug 2013 19:58:45 -0400 Subject: SUNRPC: Remove unused struct rpc_clnt field cl_protname Signed-off-by: Trond Myklebust --- include/linux/sunrpc/clnt.h | 1 - 1 file changed, 1 deletion(-) (limited to 'include') diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index bfe11be81f6f..481f9c07631e 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -41,7 +41,6 @@ struct rpc_clnt { cl_vers, /* RPC version number */ cl_maxproc; /* max procedure number */ - const char * cl_protname; /* protocol name */ struct rpc_auth * cl_auth; /* authenticator */ struct rpc_stat * cl_stats; /* per-program statistics */ struct rpc_iostats * cl_metrics; /* per-client statistics */ -- cgit v1.2.3 From c2190661039b3817b4cc1cbfea620b3f7dbe5cd8 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 26 Aug 2013 19:23:04 -0400 Subject: SUNRPC: Replace clnt->cl_principal The clnt->cl_principal is being used exclusively to store the service target name for RPCSEC_GSS/krb5 callbacks. Replace it with something that is stored only in the RPCSEC_GSS-specific code. Signed-off-by: Trond Myklebust --- include/linux/sunrpc/auth.h | 10 ++++++++-- include/linux/sunrpc/clnt.h | 1 - 2 files changed, 8 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/linux/sunrpc/auth.h b/include/linux/sunrpc/auth.h index 0dd00f4f6810..a9ab577b0c4a 100644 --- a/include/linux/sunrpc/auth.h +++ b/include/linux/sunrpc/auth.h @@ -87,6 +87,11 @@ struct rpc_auth { /* per-flavor data */ }; +struct rpc_auth_create_args { + rpc_authflavor_t pseudoflavor; + const char *target_name; +}; + /* Flags for rpcauth_lookupcred() */ #define RPCAUTH_LOOKUP_NEW 0x01 /* Accept an uninitialised cred */ @@ -97,7 +102,7 @@ struct rpc_authops { struct module *owner; rpc_authflavor_t au_flavor; /* flavor (RPC_AUTH_*) */ char * au_name; - struct rpc_auth * (*create)(struct rpc_clnt *, rpc_authflavor_t); + struct rpc_auth * (*create)(struct rpc_auth_create_args *, struct rpc_clnt *); void (*destroy)(struct rpc_auth *); struct rpc_cred * (*lookup_cred)(struct rpc_auth *, struct auth_cred *, int); @@ -140,7 +145,8 @@ struct rpc_cred * rpc_lookup_cred(void); struct rpc_cred * rpc_lookup_machine_cred(const char *service_name); int rpcauth_register(const struct rpc_authops *); int rpcauth_unregister(const struct rpc_authops *); -struct rpc_auth * rpcauth_create(rpc_authflavor_t, struct rpc_clnt *); +struct rpc_auth * rpcauth_create(struct rpc_auth_create_args *, + struct rpc_clnt *); void rpcauth_release(struct rpc_auth *); rpc_authflavor_t rpcauth_get_pseudoflavor(rpc_authflavor_t, struct rpcsec_gss_info *); diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index 481f9c07631e..b93b4aa49ffd 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -60,7 +60,6 @@ struct rpc_clnt { struct rpc_rtt cl_rtt_default; struct rpc_timeout cl_timeout_default; const struct rpc_program *cl_program; - char *cl_principal; /* target to authenticate to */ }; /* -- cgit v1.2.3 From 6739ffb754b47e6c0fa9d9e268bde828f6856528 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 26 Aug 2013 15:38:11 -0400 Subject: SUNRPC: Add a framework to clean up management of rpc_pipefs directories The current system requires everyone to set up notifiers, manage directory locking, etc. What we really want to do is have the rpc_client create its directory, and then create all the entries. This patch will allow the RPCSEC_GSS and NFS code to register all the objects that they want to have appear in the directory, and then have the sunrpc code call them back to actually create/destroy their pipefs dentries when the rpc_client creates/destroys the parent. Signed-off-by: Trond Myklebust --- include/linux/sunrpc/clnt.h | 2 ++ include/linux/sunrpc/rpc_pipe_fs.h | 33 ++++++++++++++++++++++++++++++++- 2 files changed, 34 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index b93b4aa49ffd..0dccd01ae2de 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -21,6 +21,7 @@ #include #include #include +#include #include #include #include @@ -55,6 +56,7 @@ struct rpc_clnt { int cl_nodelen; /* nodename length */ char cl_nodename[UNX_MAXNODENAME]; + struct rpc_pipe_dir_head cl_pipedir_objects; struct dentry * cl_dentry; struct rpc_clnt * cl_parent; /* Points to parent of clones */ struct rpc_rtt cl_rtt_default; diff --git a/include/linux/sunrpc/rpc_pipe_fs.h b/include/linux/sunrpc/rpc_pipe_fs.h index aa5b582cc471..188e7fca67e1 100644 --- a/include/linux/sunrpc/rpc_pipe_fs.h +++ b/include/linux/sunrpc/rpc_pipe_fs.h @@ -5,6 +5,26 @@ #include +struct rpc_pipe_dir_head { + struct list_head pdh_entries; + struct dentry *pdh_dentry; +}; + +struct rpc_pipe_dir_object_ops; +struct rpc_pipe_dir_object { + struct list_head pdo_head; + const struct rpc_pipe_dir_object_ops *pdo_ops; + + void *pdo_data; +}; + +struct rpc_pipe_dir_object_ops { + int (*create)(struct dentry *dir, + struct rpc_pipe_dir_object *pdo); + void (*destroy)(struct dentry *dir, + struct rpc_pipe_dir_object *pdo); +}; + struct rpc_pipe_msg { struct list_head list; void *data; @@ -74,7 +94,18 @@ extern int rpc_queue_upcall(struct rpc_pipe *, struct rpc_pipe_msg *); struct rpc_clnt; extern struct dentry *rpc_create_client_dir(struct dentry *, const char *, struct rpc_clnt *); -extern int rpc_remove_client_dir(struct dentry *); +extern int rpc_remove_client_dir(struct dentry *, struct rpc_clnt *); + +extern void rpc_init_pipe_dir_head(struct rpc_pipe_dir_head *pdh); +extern void rpc_init_pipe_dir_object(struct rpc_pipe_dir_object *pdo, + const struct rpc_pipe_dir_object_ops *pdo_ops, + void *pdo_data); +extern int rpc_add_pipe_dir_object(struct net *net, + struct rpc_pipe_dir_head *pdh, + struct rpc_pipe_dir_object *pdo); +extern void rpc_remove_pipe_dir_object(struct net *net, + struct rpc_pipe_dir_head *pdh, + struct rpc_pipe_dir_object *pdo); struct cache_detail; extern struct dentry *rpc_create_cache_dir(struct dentry *, -- cgit v1.2.3 From 5f42b016d7341871948a2b8cc8fb654691522d1a Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 26 Aug 2013 16:47:20 -0400 Subject: SUNRPC: Remove the obsolete auth-only interface for pipefs dentry management Signed-off-by: Trond Myklebust --- include/linux/sunrpc/auth.h | 2 -- 1 file changed, 2 deletions(-) (limited to 'include') diff --git a/include/linux/sunrpc/auth.h b/include/linux/sunrpc/auth.h index a9ab577b0c4a..6de26f2491e5 100644 --- a/include/linux/sunrpc/auth.h +++ b/include/linux/sunrpc/auth.h @@ -107,8 +107,6 @@ struct rpc_authops { struct rpc_cred * (*lookup_cred)(struct rpc_auth *, struct auth_cred *, int); struct rpc_cred * (*crcreate)(struct rpc_auth*, struct auth_cred *, int); - int (*pipes_create)(struct rpc_auth *); - void (*pipes_destroy)(struct rpc_auth *); int (*list_pseudoflavors)(rpc_authflavor_t *, int); rpc_authflavor_t (*info2flavor)(struct rpcsec_gss_info *); int (*flavor2info)(rpc_authflavor_t, -- cgit v1.2.3 From c36dcfe1f7712b7c12df2d80359e638b9d246ce6 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 26 Aug 2013 17:44:26 -0400 Subject: SUNRPC: Remove the rpc_client->cl_dentry It is now redundant. Signed-off-by: Trond Myklebust --- include/linux/sunrpc/clnt.h | 1 - include/linux/sunrpc/rpc_pipe_fs.h | 2 +- 2 files changed, 1 insertion(+), 2 deletions(-) (limited to 'include') diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index 0dccd01ae2de..76c0bf6ecd17 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -57,7 +57,6 @@ struct rpc_clnt { int cl_nodelen; /* nodename length */ char cl_nodename[UNX_MAXNODENAME]; struct rpc_pipe_dir_head cl_pipedir_objects; - struct dentry * cl_dentry; struct rpc_clnt * cl_parent; /* Points to parent of clones */ struct rpc_rtt cl_rtt_default; struct rpc_timeout cl_timeout_default; diff --git a/include/linux/sunrpc/rpc_pipe_fs.h b/include/linux/sunrpc/rpc_pipe_fs.h index 188e7fca67e1..b0cf1812be72 100644 --- a/include/linux/sunrpc/rpc_pipe_fs.h +++ b/include/linux/sunrpc/rpc_pipe_fs.h @@ -94,7 +94,7 @@ extern int rpc_queue_upcall(struct rpc_pipe *, struct rpc_pipe_msg *); struct rpc_clnt; extern struct dentry *rpc_create_client_dir(struct dentry *, const char *, struct rpc_clnt *); -extern int rpc_remove_client_dir(struct dentry *, struct rpc_clnt *); +extern int rpc_remove_client_dir(struct rpc_clnt *); extern void rpc_init_pipe_dir_head(struct rpc_pipe_dir_head *pdh); extern void rpc_init_pipe_dir_object(struct rpc_pipe_dir_object *pdo, -- cgit v1.2.3 From 298fc3558b9c1f5324c5ec6d5c587ca9ae6cc826 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 27 Aug 2013 16:27:04 -0400 Subject: SUNRPC: Add a helper to allow sharing of rpc_pipefs directory objects Add support for looking up existing objects and creating new ones if there is no match. Signed-off-by: Trond Myklebust --- include/linux/sunrpc/rpc_pipe_fs.h | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'include') diff --git a/include/linux/sunrpc/rpc_pipe_fs.h b/include/linux/sunrpc/rpc_pipe_fs.h index b0cf1812be72..a353e0300b54 100644 --- a/include/linux/sunrpc/rpc_pipe_fs.h +++ b/include/linux/sunrpc/rpc_pipe_fs.h @@ -106,6 +106,12 @@ extern int rpc_add_pipe_dir_object(struct net *net, extern void rpc_remove_pipe_dir_object(struct net *net, struct rpc_pipe_dir_head *pdh, struct rpc_pipe_dir_object *pdo); +extern struct rpc_pipe_dir_object *rpc_find_or_alloc_pipe_dir_object( + struct net *net, + struct rpc_pipe_dir_head *pdh, + int (*match)(struct rpc_pipe_dir_object *, void *), + struct rpc_pipe_dir_object *(*alloc)(void *), + void *data); struct cache_detail; extern struct dentry *rpc_create_cache_dir(struct dentry *, -- cgit v1.2.3 From 4de6caa270afaa381dd3373e9e6d148b1090e0ec Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 14 Aug 2013 11:59:15 -0400 Subject: SUNRPC new rpc_credops to test credential expiry This patch provides the RPC layer helper functions to allow NFS to manage data in the face of expired credentials - such as avoiding buffered WRITEs and COMMITs when the gss context will expire before the WRITEs are flushed and COMMITs are sent. These helper functions enable checking the expiration of an underlying credential key for a generic rpc credential, e.g. the gss_cred gss context gc_expiry which for Kerberos is set to the remaining TGT lifetime. A new rpc_authops key_timeout is only defined for the generic auth. A new rpc_credops crkey_to_expire is only defined for the generic cred. A new rpc_credops crkey_timeout is only defined for the gss cred. Set a credential key expiry watermark, RPC_KEY_EXPIRE_TIMEO set to 240 seconds as a default and can be set via a module parameter as we need to ensure there is time for any dirty data to be flushed. If key_timeout is called on a credential with an underlying credential key that will expire within watermark seconds, we set the RPC_CRED_KEY_EXPIRE_SOON flag in the generic_cred acred so that the NFS layer can clean up prior to key expiration. Checking a generic credential's underlying credential involves a cred lookup. To avoid this lookup in the normal case when the underlying credential has a key that is valid (before the watermark), a notify flag is set in the generic credential the first time the key_timeout is called. The generic credential then stops checking the underlying credential key expiry, and the underlying credential (gss_cred) match routine then checks the key expiration upon each normal use and sets a flag in the associated generic credential only when the key expiration is within the watermark. This in turn signals the generic credential key_timeout to perform the extra credential lookup thereafter. Signed-off-by: Andy Adamson Signed-off-by: Trond Myklebust --- include/linux/sunrpc/auth.h | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) (limited to 'include') diff --git a/include/linux/sunrpc/auth.h b/include/linux/sunrpc/auth.h index 6de26f2491e5..790be1472792 100644 --- a/include/linux/sunrpc/auth.h +++ b/include/linux/sunrpc/auth.h @@ -24,12 +24,21 @@ struct rpcsec_gss_info; +/* auth_cred ac_flags bits */ +enum { + RPC_CRED_NO_CRKEY_TIMEOUT = 0, /* underlying cred has no key timeout */ + RPC_CRED_KEY_EXPIRE_SOON = 1, /* underlying cred key will expire soon */ + RPC_CRED_NOTIFY_TIMEOUT = 2, /* nofity generic cred when underlying + key will expire soon */ +}; + /* Work around the lack of a VFS credential */ struct auth_cred { kuid_t uid; kgid_t gid; struct group_info *group_info; const char *principal; + unsigned long ac_flags; unsigned char machine_cred : 1; }; @@ -111,6 +120,8 @@ struct rpc_authops { rpc_authflavor_t (*info2flavor)(struct rpcsec_gss_info *); int (*flavor2info)(rpc_authflavor_t, struct rpcsec_gss_info *); + int (*key_timeout)(struct rpc_auth *, + struct rpc_cred *); }; struct rpc_credops { @@ -127,6 +138,8 @@ struct rpc_credops { void *, __be32 *, void *); int (*crunwrap_resp)(struct rpc_task *, kxdrdproc_t, void *, __be32 *, void *); + int (*crkey_timeout)(struct rpc_cred *); + bool (*crkey_to_expire)(struct rpc_cred *); }; extern const struct rpc_authops authunix_ops; @@ -166,6 +179,9 @@ int rpcauth_uptodatecred(struct rpc_task *); int rpcauth_init_credcache(struct rpc_auth *); void rpcauth_destroy_credcache(struct rpc_auth *); void rpcauth_clear_credcache(struct rpc_cred_cache *); +int rpcauth_key_timeout_notify(struct rpc_auth *, + struct rpc_cred *); +bool rpcauth_cred_key_to_expire(struct rpc_cred *); static inline struct rpc_cred * get_rpccred(struct rpc_cred *cred) -- cgit v1.2.3 From ba6c05928dcafc7e0a0c8e4ee6a293ba47190fd4 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 30 Aug 2013 12:24:25 -0400 Subject: NFS: Ensure that rmdir() waits for sillyrenames to complete If an NFS client does mkdir("dir"); fd = open("dir/file"); unlink("dir/file"); close(fd); rmdir("dir"); then the asynchronous nature of the sillyrename operation means that we can end up getting EBUSY for the rmdir() in the above test. Fix that by ensuring that we wait for any in-progress sillyrenames before sending the rmdir() to the server. Signed-off-by: Trond Myklebust --- include/linux/nfs_fs.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h index 7125cef74164..3ea4cde8701c 100644 --- a/include/linux/nfs_fs.h +++ b/include/linux/nfs_fs.h @@ -524,6 +524,7 @@ static inline void nfs4_label_free(void *label) {} * linux/fs/nfs/unlink.c */ extern void nfs_complete_unlink(struct dentry *dentry, struct inode *); +extern void nfs_wait_on_sillyrename(struct dentry *dentry); extern void nfs_block_sillyrename(struct dentry *dentry); extern void nfs_unblock_sillyrename(struct dentry *dentry); extern int nfs_sillyrename(struct inode *dir, struct dentry *dentry); -- cgit v1.2.3 From abf79bb341bf52f75f295b850abdf5f78f584311 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 9 Aug 2013 12:49:11 -0400 Subject: NFS: Add a slot table to struct nfs_client for NFSv4.0 transport blocking Anchor an nfs4_slot_table in the nfs_client for use with NFSv4.0 transport blocking. It is initialized only for NFSv4.0 nfs_client's. Introduce appropriate minor version ops to handle nfs_client initialization and shutdown requirements that differ for each minor version. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/nfs_fs_sb.h | 3 +++ 1 file changed, 3 insertions(+) (limited to 'include') diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h index d2212432c456..fc83d3d384d0 100644 --- a/include/linux/nfs_fs_sb.h +++ b/include/linux/nfs_fs_sb.h @@ -78,6 +78,9 @@ struct nfs_client { u32 cl_cb_ident; /* v4.0 callback identifier */ const struct nfs4_minor_version_ops *cl_mvops; + /* NFSv4.0 transport blocking */ + struct nfs4_slot_table *cl_slot_tbl; + /* The sequence id to use for the next CREATE_SESSION */ u32 cl_seqid; /* The flags used for obtaining the clientid during EXCHANGE_ID */ -- cgit v1.2.3 From 40b5ea0c25669cb99ba7f4836437a7ebaba91408 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 4 Sep 2013 12:16:23 -0400 Subject: SUNRPC: Add tracepoints to help debug socket connection issues Add client side debugging to help trace socket connection/disconnection and unexpected state change issues. Signed-off-by: Trond Myklebust --- include/trace/events/sunrpc.h | 131 ++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 131 insertions(+) (limited to 'include') diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index 43be87d5dd58..b74a8acd38fa 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -6,6 +6,8 @@ #include #include +#include +#include #include DECLARE_EVENT_CLASS(rpc_task_status, @@ -172,6 +174,135 @@ DEFINE_EVENT(rpc_task_queued, rpc_task_wakeup, ); +#define rpc_show_socket_state(state) \ + __print_symbolic(state, \ + { SS_FREE, "FREE" }, \ + { SS_UNCONNECTED, "UNCONNECTED" }, \ + { SS_CONNECTING, "CONNECTING," }, \ + { SS_CONNECTED, "CONNECTED," }, \ + { SS_DISCONNECTING, "DISCONNECTING" }) + +#define rpc_show_sock_state(state) \ + __print_symbolic(state, \ + { TCP_ESTABLISHED, "ESTABLISHED" }, \ + { TCP_SYN_SENT, "SYN_SENT" }, \ + { TCP_SYN_RECV, "SYN_RECV" }, \ + { TCP_FIN_WAIT1, "FIN_WAIT1" }, \ + { TCP_FIN_WAIT2, "FIN_WAIT2" }, \ + { TCP_TIME_WAIT, "TIME_WAIT" }, \ + { TCP_CLOSE, "CLOSE" }, \ + { TCP_CLOSE_WAIT, "CLOSE_WAIT" }, \ + { TCP_LAST_ACK, "LAST_ACK" }, \ + { TCP_LISTEN, "LISTEN" }, \ + { TCP_CLOSING, "CLOSING" }) + +DECLARE_EVENT_CLASS(xs_socket_event, + + TP_PROTO( + struct rpc_xprt *xprt, + struct socket *socket + ), + + TP_ARGS(xprt, socket), + + TP_STRUCT__entry( + __field(unsigned int, socket_state) + __field(unsigned int, sock_state) + __field(unsigned long long, ino) + __string(dstaddr, + xprt->address_strings[RPC_DISPLAY_ADDR]) + __string(dstport, + xprt->address_strings[RPC_DISPLAY_PORT]) + ), + + TP_fast_assign( + struct inode *inode = SOCK_INODE(socket); + __entry->socket_state = socket->state; + __entry->sock_state = socket->sk->sk_state; + __entry->ino = (unsigned long long)inode->i_ino; + __assign_str(dstaddr, + xprt->address_strings[RPC_DISPLAY_ADDR]); + __assign_str(dstport, + xprt->address_strings[RPC_DISPLAY_PORT]); + ), + + TP_printk( + "socket:[%llu] dstaddr=%s/%s " + "state=%u (%s) sk_state=%u (%s)", + __entry->ino, __get_str(dstaddr), __get_str(dstport), + __entry->socket_state, + rpc_show_socket_state(__entry->socket_state), + __entry->sock_state, + rpc_show_sock_state(__entry->sock_state) + ) +); +#define DEFINE_RPC_SOCKET_EVENT(name) \ + DEFINE_EVENT(xs_socket_event, name, \ + TP_PROTO( \ + struct rpc_xprt *xprt, \ + struct socket *socket \ + ), \ + TP_ARGS(xprt, socket)) + +DECLARE_EVENT_CLASS(xs_socket_event_done, + + TP_PROTO( + struct rpc_xprt *xprt, + struct socket *socket, + int error + ), + + TP_ARGS(xprt, socket, error), + + TP_STRUCT__entry( + __field(int, error) + __field(unsigned int, socket_state) + __field(unsigned int, sock_state) + __field(unsigned long long, ino) + __string(dstaddr, + xprt->address_strings[RPC_DISPLAY_ADDR]) + __string(dstport, + xprt->address_strings[RPC_DISPLAY_PORT]) + ), + + TP_fast_assign( + struct inode *inode = SOCK_INODE(socket); + __entry->socket_state = socket->state; + __entry->sock_state = socket->sk->sk_state; + __entry->ino = (unsigned long long)inode->i_ino; + __entry->error = error; + __assign_str(dstaddr, + xprt->address_strings[RPC_DISPLAY_ADDR]); + __assign_str(dstport, + xprt->address_strings[RPC_DISPLAY_PORT]); + ), + + TP_printk( + "error=%d socket:[%llu] dstaddr=%s/%s " + "state=%u (%s) sk_state=%u (%s)", + __entry->error, + __entry->ino, __get_str(dstaddr), __get_str(dstport), + __entry->socket_state, + rpc_show_socket_state(__entry->socket_state), + __entry->sock_state, + rpc_show_sock_state(__entry->sock_state) + ) +); +#define DEFINE_RPC_SOCKET_EVENT_DONE(name) \ + DEFINE_EVENT(xs_socket_event_done, name, \ + TP_PROTO( \ + struct rpc_xprt *xprt, \ + struct socket *socket, \ + int error \ + ), \ + TP_ARGS(xprt, socket, error)) + +DEFINE_RPC_SOCKET_EVENT(rpc_socket_state_change); +DEFINE_RPC_SOCKET_EVENT_DONE(rpc_socket_connect); +DEFINE_RPC_SOCKET_EVENT_DONE(rpc_socket_reset_connection); +DEFINE_RPC_SOCKET_EVENT(rpc_socket_close); +DEFINE_RPC_SOCKET_EVENT(rpc_socket_shutdown); + #endif /* _TRACE_SUNRPC_H */ #include -- cgit v1.2.3 From ef1820f9be27b6ad158f433ab38002ab8131db4d Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Wed, 4 Sep 2013 17:04:49 +1000 Subject: NFSv4: Don't try to recover NFSv4 locks when they are lost. When an NFSv4 client loses contact with the server it can lose any locks that it holds. Currently when it reconnects to the server it simply tries to reclaim those locks. This might succeed even though some other client has held and released a lock in the mean time. So the first client might think the file is unchanged, but it isn't. This isn't good. If, when recovery happens, the locks cannot be claimed because some other client still holds the lock, then we get a message in the kernel logs, but the client can still write. So two clients can both think they have a lock and can both write at the same time. This is equally not good. There was a patch a while ago http://comments.gmane.org/gmane.linux.nfs/41917 which tried to address some of this, but it didn't seem to go anywhere. That patch would also send a signal to the process. That might be useful but for now this patch just causes writes to fail. For NFSv4 (unlike v2/v3) there is a strong link between the lock and the write request so we can fairly easily fail any IO of the lock is gone. While some applications might not expect this, it is still safer than allowing the write to succeed. Because this is a fairly big change in behaviour a module parameter, "recover_locks", is introduced which defaults to true (the current behaviour) but can be set to "false" to tell the client not to try to recover things that were lost. Signed-off-by: NeilBrown Signed-off-by: Trond Myklebust --- include/linux/nfs_xdr.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index ddc3e32178c3..7c3956d65b3e 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -1419,12 +1419,12 @@ struct nfs_rpc_ops { void (*read_setup) (struct nfs_read_data *, struct rpc_message *); void (*read_pageio_init)(struct nfs_pageio_descriptor *, struct inode *, const struct nfs_pgio_completion_ops *); - void (*read_rpc_prepare)(struct rpc_task *, struct nfs_read_data *); + int (*read_rpc_prepare)(struct rpc_task *, struct nfs_read_data *); int (*read_done) (struct rpc_task *, struct nfs_read_data *); void (*write_setup) (struct nfs_write_data *, struct rpc_message *); void (*write_pageio_init)(struct nfs_pageio_descriptor *, struct inode *, int, const struct nfs_pgio_completion_ops *); - void (*write_rpc_prepare)(struct rpc_task *, struct nfs_write_data *); + int (*write_rpc_prepare)(struct rpc_task *, struct nfs_write_data *); int (*write_done) (struct rpc_task *, struct nfs_write_data *); void (*commit_setup) (struct nfs_commit_data *, struct rpc_message *); void (*commit_rpc_prepare)(struct rpc_task *, struct nfs_commit_data *); -- cgit v1.2.3 From 8d1018c7745e5ab861cd6cfd3dc4fa425b90575d Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 4 Sep 2013 14:45:13 -0400 Subject: SUNRPC: Ensure rpc_task->tk_pid is available for tracepoints Signed-off-by: Trond Myklebust --- include/linux/sunrpc/sched.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/sunrpc/sched.h b/include/linux/sunrpc/sched.h index 6d870353674a..8ebb7c043f05 100644 --- a/include/linux/sunrpc/sched.h +++ b/include/linux/sunrpc/sched.h @@ -79,7 +79,7 @@ struct rpc_task { unsigned short tk_flags; /* misc flags */ unsigned short tk_timeouts; /* maj timeouts */ -#ifdef RPC_DEBUG +#if defined(RPC_DEBUG) || defined(RPC_TRACEPOINTS) unsigned short tk_pid; /* debugging aid */ #endif unsigned char tk_priority : 2,/* Task priority */ -- cgit v1.2.3 From 2f048db4680ae19da13df15d352ac02748781ecb Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 4 Sep 2013 21:51:44 -0400 Subject: SUNRPC: Add an identifier for struct rpc_clnt Add an identifier in order to aid debugging. Signed-off-by: Trond Myklebust --- include/linux/sunrpc/clnt.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index 76c0bf6ecd17..6740801aa71a 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -33,6 +33,7 @@ struct rpc_inode; */ struct rpc_clnt { atomic_t cl_count; /* Number of references */ + unsigned int cl_clid; /* client id */ struct list_head cl_clients; /* Global list of clients */ struct list_head cl_tasks; /* List of tasks */ spinlock_t cl_lock; /* spinlock */ -- cgit v1.2.3 From 92cb6c5be8134db6f7c38f25f6afd13e444cebaf Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 4 Sep 2013 22:09:50 -0400 Subject: SUNRPC: Replace pointer values with task->tk_pid and rpc_clnt->cl_clid Instead of the pointer values, use the task and client identifier values for tracing purposes. Signed-off-by: Trond Myklebust --- include/trace/events/sunrpc.h | 50 ++++++++++++++++++++++--------------------- 1 file changed, 26 insertions(+), 24 deletions(-) (limited to 'include') diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index b74a8acd38fa..d51d16c7afd8 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -17,18 +17,20 @@ DECLARE_EVENT_CLASS(rpc_task_status, TP_ARGS(task), TP_STRUCT__entry( - __field(const struct rpc_task *, task) - __field(const struct rpc_clnt *, clnt) + __field(unsigned int, task_id) + __field(unsigned int, client_id) __field(int, status) ), TP_fast_assign( - __entry->task = task; - __entry->clnt = task->tk_client; + __entry->task_id = task->tk_pid; + __entry->client_id = task->tk_client->cl_clid; __entry->status = task->tk_status; ), - TP_printk("task:%p@%p, status %d",__entry->task, __entry->clnt, __entry->status) + TP_printk("task:%u@%u, status %d", + __entry->task_id, __entry->client_id, + __entry->status) ); DEFINE_EVENT(rpc_task_status, rpc_call_status, @@ -49,18 +51,20 @@ TRACE_EVENT(rpc_connect_status, TP_ARGS(task, status), TP_STRUCT__entry( - __field(const struct rpc_task *, task) - __field(const struct rpc_clnt *, clnt) + __field(unsigned int, task_id) + __field(unsigned int, client_id) __field(int, status) ), TP_fast_assign( - __entry->task = task; - __entry->clnt = task->tk_client; + __entry->task_id = task->tk_pid; + __entry->client_id = task->tk_client->cl_clid; __entry->status = status; ), - TP_printk("task:%p@%p, status %d",__entry->task, __entry->clnt, __entry->status) + TP_printk("task:%u@%u, status %d", + __entry->task_id, __entry->client_id, + __entry->status) ); DECLARE_EVENT_CLASS(rpc_task_running, @@ -70,8 +74,8 @@ DECLARE_EVENT_CLASS(rpc_task_running, TP_ARGS(clnt, task, action), TP_STRUCT__entry( - __field(const struct rpc_clnt *, clnt) - __field(const struct rpc_task *, task) + __field(unsigned int, task_id) + __field(unsigned int, client_id) __field(const void *, action) __field(unsigned long, runstate) __field(int, status) @@ -79,17 +83,16 @@ DECLARE_EVENT_CLASS(rpc_task_running, ), TP_fast_assign( - __entry->clnt = clnt; - __entry->task = task; + __entry->client_id = clnt->cl_clid; + __entry->task_id = task->tk_pid; __entry->action = action; __entry->runstate = task->tk_runstate; __entry->status = task->tk_status; __entry->flags = task->tk_flags; ), - TP_printk("task:%p@%p flags=%4.4x state=%4.4lx status=%d action=%pf", - __entry->task, - __entry->clnt, + TP_printk("task:%u@%u flags=%4.4x state=%4.4lx status=%d action=%pf", + __entry->task_id, __entry->client_id, __entry->flags, __entry->runstate, __entry->status, @@ -128,8 +131,8 @@ DECLARE_EVENT_CLASS(rpc_task_queued, TP_ARGS(clnt, task, q), TP_STRUCT__entry( - __field(const struct rpc_clnt *, clnt) - __field(const struct rpc_task *, task) + __field(unsigned int, task_id) + __field(unsigned int, client_id) __field(unsigned long, timeout) __field(unsigned long, runstate) __field(int, status) @@ -138,8 +141,8 @@ DECLARE_EVENT_CLASS(rpc_task_queued, ), TP_fast_assign( - __entry->clnt = clnt; - __entry->task = task; + __entry->client_id = clnt->cl_clid; + __entry->task_id = task->tk_pid; __entry->timeout = task->tk_timeout; __entry->runstate = task->tk_runstate; __entry->status = task->tk_status; @@ -147,9 +150,8 @@ DECLARE_EVENT_CLASS(rpc_task_queued, __assign_str(q_name, rpc_qname(q)); ), - TP_printk("task:%p@%p flags=%4.4x state=%4.4lx status=%d timeout=%lu queue=%s", - __entry->task, - __entry->clnt, + TP_printk("task:%u@%u flags=%4.4x state=%4.4lx status=%d timeout=%lu queue=%s", + __entry->task_id, __entry->client_id, __entry->flags, __entry->runstate, __entry->status, -- cgit v1.2.3 From 2031cd1af143f47dacacbb35efbef22f2fd079e6 Mon Sep 17 00:00:00 2001 From: Weston Andros Adamson Date: Tue, 13 Aug 2013 16:37:32 -0400 Subject: nfs4.1: Minimal SP4_MACH_CRED implementation This is a minimal client side implementation of SP4_MACH_CRED. It will attempt to negotiate SP4_MACH_CRED iff the EXCHANGE_ID is using krb5i or krb5p auth. SP4_MACH_CRED will be used if the server supports the minimal operations: BIND_CONN_TO_SESSION EXCHANGE_ID CREATE_SESSION DESTROY_SESSION DESTROY_CLIENTID This patch only includes the EXCHANGE_ID negotiation code because the client will already use the machine cred for these operations. If the server doesn't support SP4_MACH_CRED or doesn't support the minimal operations, the exchange id will be resent with SP4_NONE. Signed-off-by: Weston Andros Adamson Signed-off-by: Trond Myklebust --- include/linux/nfs_fs_sb.h | 4 ++++ include/linux/nfs_xdr.h | 19 +++++++++++++++++++ 2 files changed, 23 insertions(+) (limited to 'include') diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h index fc83d3d384d0..9f7b9192cf5e 100644 --- a/include/linux/nfs_fs_sb.h +++ b/include/linux/nfs_fs_sb.h @@ -90,6 +90,10 @@ struct nfs_client { struct nfs41_server_owner *cl_serverowner; struct nfs41_server_scope *cl_serverscope; struct nfs41_impl_id *cl_implid; + /* nfs 4.1+ state protection modes: */ + unsigned long cl_sp4_flags; +#define NFS_SP4_MACH_CRED_MINIMAL 1 /* Minimal sp4_mach_cred - state ops + * must use machine cred */ #endif /* CONFIG_NFS_V4 */ #ifdef CONFIG_NFS_FSCACHE diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 7c3956d65b3e..01fd84b566f7 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -1107,6 +1107,23 @@ struct pnfs_ds_commit_info { struct pnfs_commit_bucket *buckets; }; +#define NFS4_OP_MAP_NUM_LONGS \ + DIV_ROUND_UP(LAST_NFS4_OP, 8 * sizeof(unsigned long)) +#define NFS4_OP_MAP_NUM_WORDS \ + (NFS4_OP_MAP_NUM_LONGS * sizeof(unsigned long) / sizeof(u32)) +struct nfs4_op_map { + union { + unsigned long longs[NFS4_OP_MAP_NUM_LONGS]; + u32 words[NFS4_OP_MAP_NUM_WORDS]; + } u; +}; + +struct nfs41_state_protection { + u32 how; + struct nfs4_op_map enforce; + struct nfs4_op_map allow; +}; + #define NFS4_EXCHANGE_ID_LEN (48) struct nfs41_exchange_id_args { struct nfs_client *client; @@ -1114,6 +1131,7 @@ struct nfs41_exchange_id_args { unsigned int id_len; char id[NFS4_EXCHANGE_ID_LEN]; u32 flags; + struct nfs41_state_protection state_protect; }; struct nfs41_server_owner { @@ -1146,6 +1164,7 @@ struct nfs41_exchange_id_res { struct nfs41_server_owner *server_owner; struct nfs41_server_scope *server_scope; struct nfs41_impl_id *impl_id; + struct nfs41_state_protection state_protect; }; struct nfs41_create_session_args { -- cgit v1.2.3 From fa940720ceff2d7b24b4898c51a440ab59f39a4d Mon Sep 17 00:00:00 2001 From: Weston Andros Adamson Date: Tue, 13 Aug 2013 16:37:34 -0400 Subject: nfs4.1: Add SP4_MACH_CRED cleanup support CLOSE and LOCKU can use the machine credential. Signed-off-by: Weston Andros Adamson Signed-off-by: Trond Myklebust --- include/linux/nfs_fs_sb.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h index 9f7b9192cf5e..c0ba33874a85 100644 --- a/include/linux/nfs_fs_sb.h +++ b/include/linux/nfs_fs_sb.h @@ -94,6 +94,7 @@ struct nfs_client { unsigned long cl_sp4_flags; #define NFS_SP4_MACH_CRED_MINIMAL 1 /* Minimal sp4_mach_cred - state ops * must use machine cred */ +#define NFS_SP4_MACH_CRED_CLEANUP 2 /* CLOSE and LOCKU */ #endif /* CONFIG_NFS_V4 */ #ifdef CONFIG_NFS_FSCACHE -- cgit v1.2.3 From 8b5bee2e1b60a611b15b1862f55e60cc9f668324 Mon Sep 17 00:00:00 2001 From: Weston Andros Adamson Date: Tue, 13 Aug 2013 16:37:35 -0400 Subject: nfs4.1: Add SP4_MACH_CRED secinfo support SECINFO and SECINFO_NONAME can use the machine credential. Signed-off-by: Weston Andros Adamson Signed-off-by: Trond Myklebust --- include/linux/nfs_fs_sb.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h index c0ba33874a85..800da116950b 100644 --- a/include/linux/nfs_fs_sb.h +++ b/include/linux/nfs_fs_sb.h @@ -95,6 +95,7 @@ struct nfs_client { #define NFS_SP4_MACH_CRED_MINIMAL 1 /* Minimal sp4_mach_cred - state ops * must use machine cred */ #define NFS_SP4_MACH_CRED_CLEANUP 2 /* CLOSE and LOCKU */ +#define NFS_SP4_MACH_CRED_SECINFO 3 /* SECINFO and SECINFO_NO_NAME */ #endif /* CONFIG_NFS_V4 */ #ifdef CONFIG_NFS_FSCACHE -- cgit v1.2.3 From 3787d5063c52b0c38003e6293f24839508604070 Mon Sep 17 00:00:00 2001 From: Weston Andros Adamson Date: Tue, 13 Aug 2013 16:37:36 -0400 Subject: nfs4.1: Add SP4_MACH_CRED stateid support TEST_STATEID and FREE_STATEID can use the machine credential. Signed-off-by: Weston Andros Adamson Signed-off-by: Trond Myklebust --- include/linux/nfs_fs_sb.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h index 800da116950b..179586555bbe 100644 --- a/include/linux/nfs_fs_sb.h +++ b/include/linux/nfs_fs_sb.h @@ -96,6 +96,7 @@ struct nfs_client { * must use machine cred */ #define NFS_SP4_MACH_CRED_CLEANUP 2 /* CLOSE and LOCKU */ #define NFS_SP4_MACH_CRED_SECINFO 3 /* SECINFO and SECINFO_NO_NAME */ +#define NFS_SP4_MACH_CRED_STATEID 4 /* TEST_STATEID and FREE_STATEID */ #endif /* CONFIG_NFS_V4 */ #ifdef CONFIG_NFS_FSCACHE -- cgit v1.2.3 From 8c21c62c4452f4e66c3dac9b3f6b74474fad3e08 Mon Sep 17 00:00:00 2001 From: Weston Andros Adamson Date: Tue, 13 Aug 2013 16:37:37 -0400 Subject: nfs4.1: Add SP4_MACH_CRED write and commit support WRITE and COMMIT can use the machine credential. If WRITE is supported and COMMIT is not, make all (mach cred) writes FILE_SYNC4. Signed-off-by: Weston Andros Adamson Signed-off-by: Trond Myklebust --- include/linux/nfs_fs_sb.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include') diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h index 179586555bbe..e8ff178c3d33 100644 --- a/include/linux/nfs_fs_sb.h +++ b/include/linux/nfs_fs_sb.h @@ -97,6 +97,8 @@ struct nfs_client { #define NFS_SP4_MACH_CRED_CLEANUP 2 /* CLOSE and LOCKU */ #define NFS_SP4_MACH_CRED_SECINFO 3 /* SECINFO and SECINFO_NO_NAME */ #define NFS_SP4_MACH_CRED_STATEID 4 /* TEST_STATEID and FREE_STATEID */ +#define NFS_SP4_MACH_CRED_WRITE 5 /* WRITE */ +#define NFS_SP4_MACH_CRED_COMMIT 6 /* COMMIT */ #endif /* CONFIG_NFS_V4 */ #ifdef CONFIG_NFS_FSCACHE -- cgit v1.2.3 From 0e20162ed1e9a639fc61d62c71327169fb1a1970 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Fri, 6 Sep 2013 14:14:00 -0400 Subject: NFSv4.1 Use MDS auth flavor for data server connection Commit 4edaa308 "NFS: Use "krb5i" to establish NFSv4 state whenever possible" uses the nfs_client cl_rpcclient for all state management operations, and will use krb5i or auth_sys with no regard to the mount command authflavor choice. The MDS, as any NFSv4.1 mount point, uses the nfs_server rpc client for all non-state management operations with a different nfs_server for each fsid encountered traversing the mount point, each with a potentially different auth flavor. pNFS data servers are not mounted in the normal sense as there is no associated nfs_server structure. Data servers can also export multiple fsids, each with a potentially different auth flavor. Data servers need to use the same authflavor as the MDS server rpc client for non-state management operations. Populate a list of rpc clients with the MDS server rpc client auth flavor for the DS to use. Signed-off-by: Andy Adamson Signed-off-by: Trond Myklebust --- include/linux/nfs_fs_sb.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h index e8ff178c3d33..b8cedced50c9 100644 --- a/include/linux/nfs_fs_sb.h +++ b/include/linux/nfs_fs_sb.h @@ -56,6 +56,7 @@ struct nfs_client { struct rpc_cred *cl_machine_cred; #if IS_ENABLED(CONFIG_NFS_V4) + struct list_head cl_ds_clients; /* auth flavor data servers */ u64 cl_clientid; /* constant */ nfs4_verifier cl_confirm; /* Clientid verifier */ unsigned long cl_state; -- cgit v1.2.3