From 7e0a0e38fcfea47e74b0ff6da6266f00bcd2af43 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 1 May 2019 10:49:27 -0400 Subject: SUNRPC: Replace the queue timer with a delayed work function The queue timer function, which walks the RPC queue in order to locate candidates for waking up is one of the current constraints against removing the bh-safe queue spin locks. Replace it with a delayed work queue, so that we can do the actual rpc task wake ups from an ordinary process context. Signed-off-by: Trond Myklebust --- include/linux/sunrpc/sched.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/sched.h b/include/linux/sunrpc/sched.h index d0e451868f02..7d8db5dcac04 100644 --- a/include/linux/sunrpc/sched.h +++ b/include/linux/sunrpc/sched.h @@ -183,8 +183,9 @@ struct rpc_task_setup { #define RPC_NR_PRIORITY (1 + RPC_PRIORITY_PRIVILEGED - RPC_PRIORITY_LOW) struct rpc_timer { - struct timer_list timer; struct list_head list; + unsigned long expires; + struct delayed_work dwork; }; /* -- cgit v1.2.3 From 4f8943f8088348ec01456b075d44ad19dce3d698 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 1 May 2019 16:28:29 -0400 Subject: SUNRPC: Replace direct task wakeups from softirq context Replace the direct task wakeups from inside a softirq context with wakeups from a process context. Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprtsock.h | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprtsock.h b/include/linux/sunrpc/xprtsock.h index b81d0b3e0799..7638dbe7bc50 100644 --- a/include/linux/sunrpc/xprtsock.h +++ b/include/linux/sunrpc/xprtsock.h @@ -56,6 +56,7 @@ struct sock_xprt { */ unsigned long sock_state; struct delayed_work connect_worker; + struct work_struct error_worker; struct work_struct recv_worker; struct mutex recv_mutex; struct sockaddr_storage srcaddr; @@ -84,6 +85,10 @@ struct sock_xprt { #define XPRT_SOCK_CONNECTING 1U #define XPRT_SOCK_DATA_READY (2) #define XPRT_SOCK_UPD_TIMEOUT (3) +#define XPRT_SOCK_WAKE_ERROR (4) +#define XPRT_SOCK_WAKE_WRITE (5) +#define XPRT_SOCK_WAKE_PENDING (6) +#define XPRT_SOCK_WAKE_DISCONNECT (7) #endif /* __KERNEL__ */ -- cgit v1.2.3 From 21f0ffaff510b0530bfdf77da7133c0b99dee2fe Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 28 Apr 2017 10:52:42 -0400 Subject: SUNRPC: Add basic load balancing to the transport switch For now, just count the queue length. It is less accurate than counting number of bytes queued, but easier to implement. Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprt.h | 1 + include/linux/sunrpc/xprtmultipath.h | 2 ++ 2 files changed, 3 insertions(+) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index a6d9fce7f20e..15322c1d9c8c 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -238,6 +238,7 @@ struct rpc_xprt { /* * Send stuff */ + atomic_long_t queuelen; spinlock_t transport_lock; /* lock transport info */ spinlock_t reserve_lock; /* lock slot table */ spinlock_t queue_lock; /* send/receive queue lock */ diff --git a/include/linux/sunrpc/xprtmultipath.h b/include/linux/sunrpc/xprtmultipath.h index af1257c030d2..c6cce3fbf29d 100644 --- a/include/linux/sunrpc/xprtmultipath.h +++ b/include/linux/sunrpc/xprtmultipath.h @@ -15,6 +15,8 @@ struct rpc_xprt_switch { struct kref xps_kref; unsigned int xps_nxprts; + unsigned int xps_nactive; + atomic_long_t xps_queuelen; struct list_head xps_xprt_list; struct net * xps_net; -- cgit v1.2.3 From 1c341b777501613aad83f9c233a3fe5701cff083 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 22 May 2019 08:38:57 -0400 Subject: NFS: Add deferred cache invalidation for close-to-open consistency violations If the client detects that close-to-open cache consistency has been violated, and that the file or directory has been changed on the server, then do a cache invalidation when we're done working with the file. The reason we don't do an immediate cache invalidation is that we want to avoid performance problems due to false positives. Also, note that we cannot guarantee cache consistency in this situation even if we do invalidate the cache. Signed-off-by: Trond Myklebust --- include/linux/nfs_fs.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include/linux') diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h index d363d5765cdf..0a11712a80e3 100644 --- a/include/linux/nfs_fs.h +++ b/include/linux/nfs_fs.h @@ -223,6 +223,8 @@ struct nfs4_copy_state { #define NFS_INO_INVALID_MTIME BIT(10) /* cached mtime is invalid */ #define NFS_INO_INVALID_SIZE BIT(11) /* cached size is invalid */ #define NFS_INO_INVALID_OTHER BIT(12) /* other attrs are invalid */ +#define NFS_INO_DATA_INVAL_DEFER \ + BIT(13) /* Deferred cache invalidation */ #define NFS_INO_INVALID_ATTR (NFS_INO_INVALID_CHANGE \ | NFS_INO_INVALID_CTIME \ -- cgit v1.2.3 From 612b41f808a98a124b23d72229693c3181733291 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 27 Apr 2017 08:50:51 -0400 Subject: SUNRPC: Allow creation of RPC clients with multiple connections Add an argument to struct rpc_create_args that allows the specification of how many transport connections you want to set up to the server. Signed-off-by: Trond Myklebust --- include/linux/sunrpc/clnt.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux') diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index 6e8073140a5d..4619098affa3 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -124,6 +124,7 @@ struct rpc_create_args { u32 prognumber; /* overrides program->number */ u32 version; rpc_authflavor_t authflavor; + u32 nconnect; unsigned long flags; char *client_name; struct svc_xprt *bc_xprt; /* NFSv4.1 backchannel */ -- cgit v1.2.3 From 6619079d05404cb32be29af329b87ac3b0ad4f96 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 27 Apr 2017 11:13:40 -0400 Subject: NFSv4: Allow multiple connections to NFSv4.x (x>0) servers If the user specifies the -onconn= mount option, and the transport protocol is TCP, then set up connections to the server. The connections will all go to the same IP address. Signed-off-by: Trond Myklebust --- include/linux/nfs_fs_sb.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux') diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h index 1e78032a174b..a87fe854f008 100644 --- a/include/linux/nfs_fs_sb.h +++ b/include/linux/nfs_fs_sb.h @@ -58,6 +58,7 @@ struct nfs_client { struct nfs_subversion * cl_nfs_mod; /* pointer to nfs version module */ u32 cl_minorversion;/* NFSv4 minorversion */ + unsigned int cl_nconnect; /* Number of connections */ const char * cl_principal; /* used for machine cred */ #if IS_ENABLED(CONFIG_NFS_V4) -- cgit v1.2.3 From 5a0c257f8e0f4c4b3c33dff545317c21a921303e Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Thu, 30 May 2019 10:41:28 +1000 Subject: NFS: send state management on a single connection. With NFSv4.1, different network connections need to be explicitly bound to a session. During session startup, this is not possible so only a single connection must be used for session startup. So add a task flag to disable the default round-robin choice of connections (when nconnect > 1) and force the use of a single connection. Then use that flag on all requests for session management - for consistence, include NFSv4.0 management (SETCLIENTID) and session destruction Reported-by: Chuck Lever Signed-off-by: NeilBrown Signed-off-by: Trond Myklebust --- include/linux/sunrpc/sched.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux') diff --git a/include/linux/sunrpc/sched.h b/include/linux/sunrpc/sched.h index d0e451868f02..11424bdf09e6 100644 --- a/include/linux/sunrpc/sched.h +++ b/include/linux/sunrpc/sched.h @@ -126,6 +126,7 @@ struct rpc_task_setup { #define RPC_CALL_MAJORSEEN 0x0020 /* major timeout seen */ #define RPC_TASK_ROOTCREDS 0x0040 /* force root creds */ #define RPC_TASK_DYNAMIC 0x0080 /* task was kmalloc'ed */ +#define RPC_TASK_NO_ROUND_ROBIN 0x0100 /* send requests on "main" xprt */ #define RPC_TASK_SOFT 0x0200 /* Use soft timeouts */ #define RPC_TASK_SOFTCONN 0x0400 /* Fail if can't connect */ #define RPC_TASK_SENT 0x0800 /* message was sent */ -- cgit v1.2.3 From a332518fda4731c07394164b3edcbb6efaf4c4d7 Mon Sep 17 00:00:00 2001 From: Dave Wysochanski Date: Thu, 23 May 2019 16:13:50 -0400 Subject: SUNRPC: Count ops completing with tk_status < 0 We often see various error conditions with NFS4.x that show up with a very high operation count all completing with tk_status < 0 in a short period of time. Add a count to rpc_iostats to record on a per-op basis the ops that complete in this manner, which will enable lower overhead diagnostics. Signed-off-by: Dave Wysochanski Signed-off-by: Trond Myklebust --- include/linux/sunrpc/metrics.h | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/metrics.h b/include/linux/sunrpc/metrics.h index 1b3751327575..0ee3f7052846 100644 --- a/include/linux/sunrpc/metrics.h +++ b/include/linux/sunrpc/metrics.h @@ -30,7 +30,7 @@ #include #include -#define RPC_IOSTATS_VERS "1.0" +#define RPC_IOSTATS_VERS "1.1" struct rpc_iostats { spinlock_t om_lock; @@ -66,6 +66,11 @@ struct rpc_iostats { ktime_t om_queue, /* queued for xmit */ om_rtt, /* RPC RTT */ om_execute; /* RPC execution */ + /* + * The count of operations that complete with tk_status < 0. + * These statuses usually indicate error conditions. + */ + unsigned long om_error_status; } ____cacheline_aligned; struct rpc_task; -- cgit v1.2.3 From 675dd90ad0932f2c03912a5252458d792bd7033a Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 19 Jun 2019 10:33:42 -0400 Subject: xprtrdma: Modernize ops->connect Adapt and apply changes that were made to the TCP socket connect code. See the following commits for details on the purpose of these changes: Commit 7196dbb02ea0 ("SUNRPC: Allow changing of the TCP timeout parameters on the fly") Commit 3851f1cdb2b8 ("SUNRPC: Limit the reconnect backoff timer to the max RPC message timeout") Commit 02910177aede ("SUNRPC: Fix reconnection timeouts") Some common transport code is moved to xprt.c to satisfy the code duplication police. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- include/linux/sunrpc/xprt.h | 3 +++ 1 file changed, 3 insertions(+) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index a6d9fce7f20e..cc78fd38ea7d 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -334,6 +334,9 @@ struct xprt_class { */ struct rpc_xprt *xprt_create_transport(struct xprt_create *args); void xprt_connect(struct rpc_task *task); +unsigned long xprt_reconnect_delay(const struct rpc_xprt *xprt); +void xprt_reconnect_backoff(struct rpc_xprt *xprt, + unsigned long init_to); void xprt_reserve(struct rpc_task *task); void xprt_retry_reserve(struct rpc_task *task); int xprt_reserve_xprt(struct rpc_xprt *xprt, struct rpc_task *task); -- cgit v1.2.3 From a101b043c44dfcb63bed7f29a675e9fa0259005e Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 11 Jul 2019 16:33:12 -0400 Subject: SUNRPC: Fix transport accounting when caller specifies an rpc_xprt Ensure that we do the required accounting for the round robin queue when the caller to rpc_init_task() has passed in a transport to be used. Reported-by: Olga Kornievskaia Reported-by: Neil Brown Signed-off-by: Trond Myklebust --- include/linux/sunrpc/clnt.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include/linux') diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index 4619098affa3..4e070e00c143 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -164,6 +164,8 @@ void rpc_shutdown_client(struct rpc_clnt *); void rpc_release_client(struct rpc_clnt *); void rpc_task_release_transport(struct rpc_task *); void rpc_task_release_client(struct rpc_task *); +struct rpc_xprt *rpc_task_get_xprt(struct rpc_clnt *clnt, + struct rpc_xprt *xprt); int rpcb_create_local(struct net *); void rpcb_put_local(struct net *); -- cgit v1.2.3 From 7402a4fedc2bc448100c2d086406c708451b16dc Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 16 Jul 2019 13:51:29 -0400 Subject: SUNRPC: Fix up backchannel slot table accounting Add a per-transport maximum limit in the socket case, and add helpers to allow the NFSv4 code to discover that limit. Signed-off-by: Trond Myklebust --- include/linux/sunrpc/bc_xprt.h | 1 + include/linux/sunrpc/clnt.h | 1 + include/linux/sunrpc/xprt.h | 6 ++++-- 3 files changed, 6 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/bc_xprt.h b/include/linux/sunrpc/bc_xprt.h index d4229a78524a..87d27e13d885 100644 --- a/include/linux/sunrpc/bc_xprt.h +++ b/include/linux/sunrpc/bc_xprt.h @@ -43,6 +43,7 @@ void xprt_destroy_backchannel(struct rpc_xprt *, unsigned int max_reqs); int xprt_setup_bc(struct rpc_xprt *xprt, unsigned int min_reqs); void xprt_destroy_bc(struct rpc_xprt *xprt, unsigned int max_reqs); void xprt_free_bc_rqst(struct rpc_rqst *req); +unsigned int xprt_bc_max_slots(struct rpc_xprt *xprt); /* * Determine if a shared backchannel is in use diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index 4e070e00c143..abc63bd1be2b 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -194,6 +194,7 @@ void rpc_setbufsize(struct rpc_clnt *, unsigned int, unsigned int); struct net * rpc_net_ns(struct rpc_clnt *); size_t rpc_max_payload(struct rpc_clnt *); size_t rpc_max_bc_payload(struct rpc_clnt *); +unsigned int rpc_num_bc_slots(struct rpc_clnt *); void rpc_force_rebind(struct rpc_clnt *); size_t rpc_peeraddr(struct rpc_clnt *, struct sockaddr *, size_t); const char *rpc_peeraddr2str(struct rpc_clnt *, enum rpc_display_format_t); diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index ed76e5fb36c1..13e108bcc9eb 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -158,6 +158,7 @@ struct rpc_xprt_ops { int (*bc_setup)(struct rpc_xprt *xprt, unsigned int min_reqs); size_t (*bc_maxpayload)(struct rpc_xprt *xprt); + unsigned int (*bc_num_slots)(struct rpc_xprt *xprt); void (*bc_free_rqst)(struct rpc_rqst *rqst); void (*bc_destroy)(struct rpc_xprt *xprt, unsigned int max_reqs); @@ -251,8 +252,9 @@ struct rpc_xprt { #if defined(CONFIG_SUNRPC_BACKCHANNEL) struct svc_serv *bc_serv; /* The RPC service which will */ /* process the callback */ - int bc_alloc_count; /* Total number of preallocs */ - atomic_t bc_free_slots; + unsigned int bc_alloc_max; + unsigned int bc_alloc_count; /* Total number of preallocs */ + atomic_t bc_slot_count; /* Number of allocated slots */ spinlock_t bc_pa_lock; /* Protects the preallocated * items */ struct list_head bc_pa_list; /* List of preallocated -- cgit v1.2.3 From d5b9216fd5114be4ed98ca9c1ecc5f164cd8cf5e Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 18 Jul 2019 09:32:17 -0400 Subject: pnfs/flexfiles: Add tracepoints for detecting pnfs fallback to MDS Add tracepoints to allow debugging of the event chain leading to a pnfs fallback to doing I/O through the MDS. Signed-off-by: Trond Myklebust --- include/linux/nfs4.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux') diff --git a/include/linux/nfs4.h b/include/linux/nfs4.h index 22494d170619..fd59904a282c 100644 --- a/include/linux/nfs4.h +++ b/include/linux/nfs4.h @@ -660,6 +660,7 @@ enum pnfs_update_layout_reason { PNFS_UPDATE_LAYOUT_BLOCKED, PNFS_UPDATE_LAYOUT_INVALID_OPEN, PNFS_UPDATE_LAYOUT_SEND_LAYOUTGET, + PNFS_UPDATE_LAYOUT_EXIT, }; #define NFS4_OP_MAP_NUM_LONGS \ -- cgit v1.2.3