From 71d3d0ebc894294ef9454e45a3ac2e9ba60b3351 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 26 Jul 2021 08:01:27 -0400 Subject: SUNRPC: Convert rpc_client refcount to use refcount_t There are now tools in the refcount library that allow us to convert the client shutdown code. Reported-by: Xiyu Yang Signed-off-by: Trond Myklebust Signed-off-by: Anna Schumaker --- include/linux/sunrpc/clnt.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'include/linux/sunrpc') diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index 8b5d5c97553e..b2edd5fc2f0c 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -14,6 +14,7 @@ #include #include #include +#include #include #include @@ -35,7 +36,7 @@ struct rpc_sysfs_client; * The high-level client handle */ struct rpc_clnt { - atomic_t cl_count; /* Number of references */ + refcount_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 */ -- cgit v1.2.3 From c2dc3e5fad13aca5d7bdf4bcb52b1a1d707c8555 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 26 Jul 2021 07:59:23 -0400 Subject: SUNRPC: Fix potential memory corruption We really should not call rpc_wake_up_queued_task_set_status() with xprt->snd_task as an argument unless we are certain that is actually an rpc_task. Fixes: 0445f92c5d53 ("SUNRPC: Fix disconnection races") Signed-off-by: Trond Myklebust Signed-off-by: Anna Schumaker --- include/linux/sunrpc/xprt.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux/sunrpc') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index c8c39f22d3b1..59cd97da895b 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -432,6 +432,7 @@ void xprt_release_write(struct rpc_xprt *, struct rpc_task *); #define XPRT_CONGESTED (9) #define XPRT_CWND_WAIT (10) #define XPRT_WRITE_SPACE (11) +#define XPRT_SND_IS_COOKIE (12) static inline void xprt_set_connected(struct rpc_xprt *xprt) { -- cgit v1.2.3 From 438623a06bacd69c40c4af633bb09a3bbb9dfc78 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 15 Jul 2021 15:52:06 -0400 Subject: SUNRPC: Add svc_rqst::rq_auth_stat I'd like to take commit 4532608d71c8 ("SUNRPC: Clean up generic dispatcher code") even further by using only private local SVC dispatchers for all kernel RPC services. This change would enable the removal of the logic that switches between svc_generic_dispatch() and a service's private dispatcher, and simplify the invocation of the service's pc_release method so that humans can visually verify that it is always invoked properly. All that will come later. First, let's provide a better way to return authentication errors from SVC dispatcher functions. Instead of overloading the dispatch method's *statp argument, add a field to struct svc_rqst that can hold an error value. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- include/linux/sunrpc/svc.h | 1 + include/linux/sunrpc/svcauth.h | 4 ++-- 2 files changed, 3 insertions(+), 2 deletions(-) (limited to 'include/linux/sunrpc') diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h index e91d51ea028b..35f12963e1ff 100644 --- a/include/linux/sunrpc/svc.h +++ b/include/linux/sunrpc/svc.h @@ -282,6 +282,7 @@ struct svc_rqst { void * rq_argp; /* decoded arguments */ void * rq_resp; /* xdr'd results */ void * rq_auth_data; /* flavor-specific data */ + __be32 rq_auth_stat; /* authentication status */ int rq_auth_slack; /* extra space xdr code * should leave in head * for krb5i, krb5p. diff --git a/include/linux/sunrpc/svcauth.h b/include/linux/sunrpc/svcauth.h index b0003866a249..6d9cc9080aca 100644 --- a/include/linux/sunrpc/svcauth.h +++ b/include/linux/sunrpc/svcauth.h @@ -127,7 +127,7 @@ struct auth_ops { char * name; struct module *owner; int flavour; - int (*accept)(struct svc_rqst *rq, __be32 *authp); + int (*accept)(struct svc_rqst *rq); int (*release)(struct svc_rqst *rq); void (*domain_release)(struct auth_domain *); int (*set_client)(struct svc_rqst *rq); @@ -149,7 +149,7 @@ struct auth_ops { struct svc_xprt; -extern int svc_authenticate(struct svc_rqst *rqstp, __be32 *authp); +extern int svc_authenticate(struct svc_rqst *rqstp); extern int svc_authorise(struct svc_rqst *rqstp); extern int svc_set_client(struct svc_rqst *rqstp); extern int svc_auth_register(rpc_authflavor_t flavor, struct auth_ops *aops); -- cgit v1.2.3 From 9082e1d914f8b27114352b1940bbcc7522f682e7 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 15 Jul 2021 15:52:19 -0400 Subject: SUNRPC: Eliminate the RQ_AUTHERR flag Now that there is an alternate method for returning an auth_stat value, replace the RQ_AUTHERR flag with use of that new method. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- include/linux/sunrpc/svc.h | 2 -- 1 file changed, 2 deletions(-) (limited to 'include/linux/sunrpc') diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h index 35f12963e1ff..63c9210cae06 100644 --- a/include/linux/sunrpc/svc.h +++ b/include/linux/sunrpc/svc.h @@ -275,7 +275,6 @@ struct svc_rqst { #define RQ_VICTIM (5) /* about to be shut down */ #define RQ_BUSY (6) /* request is busy */ #define RQ_DATA (7) /* request has data */ -#define RQ_AUTHERR (8) /* Request status is auth error */ unsigned long rq_flags; /* flags field */ ktime_t rq_qtime; /* enqueue time */ @@ -533,7 +532,6 @@ unsigned int svc_fill_write_vector(struct svc_rqst *rqstp, char *svc_fill_symlink_pathname(struct svc_rqst *rqstp, struct kvec *first, void *p, size_t total); -__be32 svc_return_autherr(struct svc_rqst *rqstp, __be32 auth_err); __be32 svc_generic_init_request(struct svc_rqst *rqstp, const struct svc_program *progp, struct svc_process_info *procinfo); -- cgit v1.2.3 From 2f0f88f42f2eab0421ed37d7494de9124fdf0d34 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 1 Jul 2021 10:03:10 -0400 Subject: SUNRPC: Add svc_rqst_replace_page() API Replacing a page in rq_pages[] requires a get_page(), which is a bus-locked operation, and a put_page(), which can be even more costly. To reduce the cost of replacing a page in rq_pages[], batch the put_page() operations by collecting "freed" pages in a pagevec, and then release those pages when the pagevec is full. This pagevec is also emptied when each RPC completes. Signed-off-by: Chuck Lever --- include/linux/sunrpc/svc.h | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'include/linux/sunrpc') diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h index e91d51ea028b..ab9afbf0a0d8 100644 --- a/include/linux/sunrpc/svc.h +++ b/include/linux/sunrpc/svc.h @@ -19,6 +19,7 @@ #include #include #include +#include /* statistics for svc_pool structures */ struct svc_pool_stats { @@ -256,6 +257,7 @@ struct svc_rqst { struct page * *rq_next_page; /* next reply page to use */ struct page * *rq_page_end; /* one past the last page */ + struct pagevec rq_pvec; struct kvec rq_vec[RPCSVC_MAXPAGES]; /* generally useful.. */ struct bio_vec rq_bvec[RPCSVC_MAXPAGES]; @@ -502,6 +504,8 @@ struct svc_rqst *svc_rqst_alloc(struct svc_serv *serv, struct svc_pool *pool, int node); struct svc_rqst *svc_prepare_thread(struct svc_serv *serv, struct svc_pool *pool, int node); +void svc_rqst_replace_page(struct svc_rqst *rqstp, + struct page *page); void svc_rqst_free(struct svc_rqst *); void svc_exit_thread(struct svc_rqst *); unsigned int svc_pool_map_get(void); -- cgit v1.2.3 From 6c8c84f525100a1cade5698320b4abe43062e159 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 7 Jul 2021 14:57:28 -0400 Subject: svcrdma: Fewer calls to wake_up() in Send completion handler Because wake_up() takes an IRQ-safe lock, it can be expensive, especially to call inside of a single-threaded completion handler. What's more, the Send wait queue almost never has waiters, so most of the time, this is an expensive no-op. As always, the goal is to reduce the average overhead of each completion, because a transport's completion handlers are single- threaded on one CPU core. This change reduces CPU utilization of the Send completion thread by 2-3% on my server. Signed-off-by: Chuck Lever Reviewed-By: Tom Talpey --- include/linux/sunrpc/svc_rdma.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux/sunrpc') diff --git a/include/linux/sunrpc/svc_rdma.h b/include/linux/sunrpc/svc_rdma.h index 3184465de3a0..57c60ffe76dd 100644 --- a/include/linux/sunrpc/svc_rdma.h +++ b/include/linux/sunrpc/svc_rdma.h @@ -207,6 +207,7 @@ extern void svc_rdma_send_error_msg(struct svcxprt_rdma *rdma, struct svc_rdma_send_ctxt *sctxt, struct svc_rdma_recv_ctxt *rctxt, int status); +extern void svc_rdma_wake_send_waiters(struct svcxprt_rdma *rdma, int avail); extern int svc_rdma_sendto(struct svc_rqst *); extern int svc_rdma_result_payload(struct svc_rqst *rqstp, unsigned int offset, unsigned int length); -- cgit v1.2.3 From b6c2bfea096ba22583f1071c10ce0745804b9b95 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Tue, 9 Feb 2021 10:32:20 -0500 Subject: svcrdma: Relieve contention on sc_send_lock. /proc/lock_stat indicates the the sc_send_lock is heavily contended when the server is under load from a single client. To address this, convert the send_ctxt free list to an llist. Returning an item to the send_ctxt cache is now waitless, which reduces the instruction path length in the single-threaded Send handler (svc_rdma_wc_send). The goal is to enable the ib_comp_wq worker to handle a higher RPC/RDMA Send completion rate given the same CPU resources. This change reduces CPU utilization of Send completion by 2-3% on my server. Signed-off-by: Chuck Lever Reviewed-By: Tom Talpey --- include/linux/sunrpc/svc_rdma.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include/linux/sunrpc') diff --git a/include/linux/sunrpc/svc_rdma.h b/include/linux/sunrpc/svc_rdma.h index 57c60ffe76dd..5f8d5af6556c 100644 --- a/include/linux/sunrpc/svc_rdma.h +++ b/include/linux/sunrpc/svc_rdma.h @@ -90,7 +90,7 @@ struct svcxprt_rdma { struct ib_pd *sc_pd; spinlock_t sc_send_lock; - struct list_head sc_send_ctxts; + struct llist_head sc_send_ctxts; spinlock_t sc_rw_ctxt_lock; struct list_head sc_rw_ctxts; @@ -150,7 +150,7 @@ struct svc_rdma_recv_ctxt { }; struct svc_rdma_send_ctxt { - struct list_head sc_list; + struct llist_node sc_node; struct rpc_rdma_cid sc_cid; struct ib_send_wr sc_send_wr; -- cgit v1.2.3 From 07a92d009f0b1557d3d58905ce18821a483be2e1 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 8 Feb 2021 15:33:16 -0500 Subject: svcrdma: Convert rdma->sc_rw_ctxts to llist Relieve contention on sc_rw_ctxt_lock by converting rdma->sc_rw_ctxts to an llist. The goal is to reduce the average overhead of Send completions, because a transport's completion handlers are single-threaded on one CPU core. This change reduces CPU utilization of each Send completion by 2-3% on my server. Signed-off-by: Chuck Lever Reviewed-By: Tom Talpey --- include/linux/sunrpc/svc_rdma.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/linux/sunrpc') diff --git a/include/linux/sunrpc/svc_rdma.h b/include/linux/sunrpc/svc_rdma.h index 5f8d5af6556c..24aa159d29a7 100644 --- a/include/linux/sunrpc/svc_rdma.h +++ b/include/linux/sunrpc/svc_rdma.h @@ -92,7 +92,7 @@ struct svcxprt_rdma { spinlock_t sc_send_lock; struct llist_head sc_send_ctxts; spinlock_t sc_rw_ctxt_lock; - struct list_head sc_rw_ctxts; + struct llist_head sc_rw_ctxts; u32 sc_pending_recvs; u32 sc_recv_batch; -- cgit v1.2.3 From 5c11720767f70d34357d00a15ba5a0ad052c40fe Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 5 Aug 2021 15:11:24 -0400 Subject: SUNRPC: Fix a NULL pointer deref in trace_svc_stats_latency() Some paths through svc_process() leave rqst->rq_procinfo set to NULL, which triggers a crash if tracing happens to be enabled. Fixes: 89ff87494c6e ("SUNRPC: Display RPC procedure names instead of proc numbers") Signed-off-by: Chuck Lever --- include/linux/sunrpc/svc.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux/sunrpc') diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h index ab9afbf0a0d8..f0f846fa396e 100644 --- a/include/linux/sunrpc/svc.h +++ b/include/linux/sunrpc/svc.h @@ -527,6 +527,7 @@ void svc_wake_up(struct svc_serv *); void svc_reserve(struct svc_rqst *rqstp, int space); struct svc_pool * svc_pool_for_cpu(struct svc_serv *serv, int cpu); char * svc_print_addr(struct svc_rqst *, char *, size_t); +const char * svc_proc_name(const struct svc_rqst *rqstp); int svc_encode_result_payload(struct svc_rqst *rqstp, unsigned int offset, unsigned int length); -- cgit v1.2.3 From b4ab2fea7c797b0b8b92332c7e315703c12d37d8 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 30 Jul 2021 16:07:36 -0400 Subject: SUNRPC: Add RPC_AUTH_TLS protocol numbers Shared by client and server. See: https://www.iana.org/assignments/rpc-authentication-numbers/rpc-authentication-numbers.xhtml Signed-off-by: Chuck Lever --- include/linux/sunrpc/msg_prot.h | 1 + include/linux/sunrpc/xdr.h | 1 + 2 files changed, 2 insertions(+) (limited to 'include/linux/sunrpc') diff --git a/include/linux/sunrpc/msg_prot.h b/include/linux/sunrpc/msg_prot.h index 938c2bf29db8..02117ed0fa2e 100644 --- a/include/linux/sunrpc/msg_prot.h +++ b/include/linux/sunrpc/msg_prot.h @@ -20,6 +20,7 @@ enum rpc_auth_flavors { RPC_AUTH_DES = 3, RPC_AUTH_KRB = 4, RPC_AUTH_GSS = 6, + RPC_AUTH_TLS = 7, RPC_AUTH_MAXFLAVOR = 8, /* pseudoflavors: */ RPC_AUTH_GSS_KRB5 = 390003, diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h index a965cbc136ad..b519609af1d0 100644 --- a/include/linux/sunrpc/xdr.h +++ b/include/linux/sunrpc/xdr.h @@ -95,6 +95,7 @@ xdr_buf_init(struct xdr_buf *buf, void *start, size_t len) #define rpc_auth_unix cpu_to_be32(RPC_AUTH_UNIX) #define rpc_auth_short cpu_to_be32(RPC_AUTH_SHORT) #define rpc_auth_gss cpu_to_be32(RPC_AUTH_GSS) +#define rpc_auth_tls cpu_to_be32(RPC_AUTH_TLS) #define rpc_call cpu_to_be32(RPC_CALL) #define rpc_reply cpu_to_be32(RPC_REPLY) -- cgit v1.2.3 From a4ae308143961bf688e1c8a62f6604e62b491120 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 5 Aug 2021 10:25:49 -0400 Subject: SUNRPC: Move client-side disconnect injection Disconnect injection stress-tests the ability for both client and server implementations to behave resiliently in the face of network instability. Convert the existing client-side disconnect injection infrastructure to use the kernel's generic error injection facility. The generic facility has a richer set of injection criteria. Signed-off-by: Chuck Lever --- include/linux/sunrpc/xprt.h | 18 ------------------ 1 file changed, 18 deletions(-) (limited to 'include/linux/sunrpc') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index c8c39f22d3b1..b15c1f07162d 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -288,7 +288,6 @@ struct rpc_xprt { const char *address_strings[RPC_DISPLAY_MAX]; #if IS_ENABLED(CONFIG_SUNRPC_DEBUG) struct dentry *debugfs; /* debugfs directory */ - atomic_t inject_disconnect; #endif struct rcu_head rcu; const struct xprt_class *xprt_class; @@ -502,21 +501,4 @@ static inline int xprt_test_and_set_binding(struct rpc_xprt *xprt) return test_and_set_bit(XPRT_BINDING, &xprt->state); } -#if IS_ENABLED(CONFIG_SUNRPC_DEBUG) -extern unsigned int rpc_inject_disconnect; -static inline void xprt_inject_disconnect(struct rpc_xprt *xprt) -{ - if (!rpc_inject_disconnect) - return; - if (atomic_dec_return(&xprt->inject_disconnect)) - return; - atomic_set(&xprt->inject_disconnect, rpc_inject_disconnect); - xprt->ops->inject_disconnect(xprt); -} -#else -static inline void xprt_inject_disconnect(struct rpc_xprt *xprt) -{ -} -#endif - #endif /* _LINUX_SUNRPC_XPRT_H */ -- cgit v1.2.3 From 3a3f976639f267823e443fdd8bffa03848fa1c3f Mon Sep 17 00:00:00 2001 From: Olga Kornievskaia Date: Fri, 27 Aug 2021 14:37:14 -0400 Subject: SUNRPC keep track of number of transports to unique addresses Currently, xprt_switch keeps a number of all xprts (xps_nxprts) that were added to the switch regardless of whethere it's an nconnect transport or a transport to a trunkable address. Introduce a new counter to keep track of transports to unique destination addresses per xprt_switch. Signed-off-by: Olga Kornievskaia Signed-off-by: Anna Schumaker --- include/linux/sunrpc/xprtmultipath.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux/sunrpc') diff --git a/include/linux/sunrpc/xprtmultipath.h b/include/linux/sunrpc/xprtmultipath.h index b19addc8b715..bbb8a5fa0816 100644 --- a/include/linux/sunrpc/xprtmultipath.h +++ b/include/linux/sunrpc/xprtmultipath.h @@ -18,6 +18,7 @@ struct rpc_xprt_switch { unsigned int xps_id; unsigned int xps_nxprts; unsigned int xps_nactive; + unsigned int xps_nunique_destaddr_xprts; atomic_long_t xps_queuelen; struct list_head xps_xprt_list; -- cgit v1.2.3 From dc48e0abee245e2f0361bd8d4e3b00f70450fab2 Mon Sep 17 00:00:00 2001 From: Olga Kornievskaia Date: Fri, 27 Aug 2021 14:37:18 -0400 Subject: SUNRPC enforce creation of no more than max_connect xprts If we are adding new transports via rpc_clnt_test_and_add_xprt() then check if we've reached the limit. Currently only pnfs path adds transports via that function but this is done in preparation when the client would add new transports when session trunking is detected. A warning is logged if the limit is reached. Signed-off-by: Olga Kornievskaia Signed-off-by: Anna Schumaker --- include/linux/sunrpc/clnt.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include/linux/sunrpc') diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index b2edd5fc2f0c..a4661646adc9 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -82,6 +82,7 @@ struct rpc_clnt { struct work_struct cl_work; }; const struct cred *cl_cred; + unsigned int cl_max_connect; /* max number of transports not to the same IP */ }; /* @@ -136,6 +137,7 @@ struct rpc_create_args { char *client_name; struct svc_xprt *bc_xprt; /* NFSv4.1 backchannel */ const struct cred *cred; + unsigned int max_connect; }; struct rpc_add_xprt_test { -- cgit v1.2.3