From 52782c92ac85c4e393eb4a903a62e6c24afa633f Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Wed, 6 May 2020 12:09:34 -0400 Subject: kthread: save thread function It's handy to keep the kthread_fn just as a unique cookie to identify classes of kthreads. E.g. if you can verify that a given task is running your thread_fn, then you may know what sort of type kthread_data points to. We'll use this in nfsd to pass some information into the vfs. Note it will need kthread_data() exported too. Original-patch-by: Tejun Heo Signed-off-by: J. Bruce Fields --- include/linux/kthread.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux') diff --git a/include/linux/kthread.h b/include/linux/kthread.h index 8bbcaad7ef0f..c2a274b79c42 100644 --- a/include/linux/kthread.h +++ b/include/linux/kthread.h @@ -57,6 +57,7 @@ bool kthread_should_stop(void); bool kthread_should_park(void); bool __kthread_should_park(struct task_struct *k); bool kthread_freezable_should_stop(bool *was_frozen); +void *kthread_func(struct task_struct *k); void *kthread_data(struct task_struct *k); void *kthread_probe_data(struct task_struct *k); int kthread_park(struct task_struct *k); -- cgit v1.2.3 From 28df3d1539de5090f7916f6fff03891b67f366f4 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Fri, 28 Jul 2017 16:35:15 -0400 Subject: nfsd: clients don't need to break their own delegations We currently revoke read delegations on any write open or any operation that modifies file data or metadata (including rename, link, and unlink). But if the delegation in question is the only read delegation and is held by the client performing the operation, that's not really necessary. It's not always possible to prevent this in the NFSv4.0 case, because there's not always a way to determine which client an NFSv4.0 delegation came from. (In theory we could try to guess this from the transport layer, e.g., by assuming all traffic on a given TCP connection comes from the same client. But that's not really correct.) In the NFSv4.1 case the session layer always tells us the client. This patch should remove such self-conflicts in all cases where we can reliably determine the client from the compound. To do that we need to track "who" is performing a given (possibly lease-breaking) file operation. We're doing that by storing the information in the svc_rqst and using kthread_data() to map the current task back to a svc_rqst. Signed-off-by: J. Bruce Fields --- include/linux/fs.h | 1 + include/linux/sunrpc/svc.h | 1 + 2 files changed, 2 insertions(+) (limited to 'include/linux') diff --git a/include/linux/fs.h b/include/linux/fs.h index 4f6f59b4f22a..4b784560ffaf 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1045,6 +1045,7 @@ struct lock_manager_operations { bool (*lm_break)(struct file_lock *); int (*lm_change)(struct file_lock *, int, struct list_head *); void (*lm_setup)(struct file_lock *, void **); + bool (*lm_breaker_owns_lease)(struct file_lock *); }; struct lock_manager { diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h index fd390894a584..abf4a57ce4a7 100644 --- a/include/linux/sunrpc/svc.h +++ b/include/linux/sunrpc/svc.h @@ -299,6 +299,7 @@ struct svc_rqst { struct net *rq_bc_net; /* pointer to backchannel's * net namespace */ + void ** rq_lease_breaker; /* The v4 client breaking a lease */ }; #define SVC_NET(rqst) (rqst->rq_xprt ? rqst->rq_xprt->xpt_net : rqst->rq_bc_net) -- cgit v1.2.3 From ca4faf543a33373bed3650812d5f0cd0bd295b1a Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Sat, 2 May 2020 10:37:44 -0400 Subject: SUNRPC: Move xpt_mutex into socket xpo_sendto methods It appears that the RPC/RDMA transport does not need serialization of calls to its xpo_sendto method. Move the mutex into the socket methods that still need that serialization. Tail latencies are unambiguously better with this patch applied. fio randrw 8KB 70/30 on NFSv3, smaller numbers are better: clat percentiles (usec): With xpt_mutex: r | 99.99th=[ 8848] w | 99.99th=[ 9634] Without xpt_mutex: r | 99.99th=[ 8586] w | 99.99th=[ 8979] Serializing the construction of RPC/RDMA transport headers is not really necessary at this point, because the Linux NFS server implementation never changes its credit grant on a connection. If that should change, then svc_rdma_sendto will need to serialize access to the transport's credit grant fields. Reported-by: kbuild test robot [ cel: fix uninitialized variable warning ] Signed-off-by: Chuck Lever --- include/linux/sunrpc/svc_xprt.h | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'include/linux') diff --git a/include/linux/sunrpc/svc_xprt.h b/include/linux/sunrpc/svc_xprt.h index 9e1e046de176..aca35ab5cff2 100644 --- a/include/linux/sunrpc/svc_xprt.h +++ b/include/linux/sunrpc/svc_xprt.h @@ -117,6 +117,12 @@ static inline int register_xpt_user(struct svc_xprt *xpt, struct svc_xpt_user *u return 0; } +static inline bool svc_xprt_is_dead(const struct svc_xprt *xprt) +{ + return (test_bit(XPT_DEAD, &xprt->xpt_flags) != 0) || + (test_bit(XPT_CLOSE, &xprt->xpt_flags) != 0); +} + int svc_reg_xprt_class(struct svc_xprt_class *); void svc_unreg_xprt_class(struct svc_xprt_class *); void svc_xprt_init(struct net *, struct svc_xprt_class *, struct svc_xprt *, -- cgit v1.2.3 From ea740bd5f58e2912e74f401fd01a9d6aa985ca05 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 20 Mar 2020 17:32:41 -0400 Subject: svcrdma: Fix backchannel return code Way back when I was writing the RPC/RDMA server-side backchannel code, I misread the TCP backchannel reply handler logic. When svc_tcp_recvfrom() successfully receives a backchannel reply, it does not return -EAGAIN. It sets XPT_DATA and returns zero. Update svc_rdma_recvfrom() to return zero. Here, XPT_DATA doesn't need to be set again: it is set whenever a new message is received, behind a spin lock in a single threaded context. Also, if handling the cb reply is not successful, the message is simply dropped. There's no special message framing to deal with as there is in the TCP case. Now that the handle_bc_reply() return value is ignored, I've removed the dprintk call sites in the error exit of handle_bc_reply() in favor of trace points in other areas that already report the error cases. Signed-off-by: Chuck Lever --- include/linux/sunrpc/svc_rdma.h | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/svc_rdma.h b/include/linux/sunrpc/svc_rdma.h index cbcfbd0521e3..8518c3f37e56 100644 --- a/include/linux/sunrpc/svc_rdma.h +++ b/include/linux/sunrpc/svc_rdma.h @@ -160,9 +160,8 @@ struct svc_rdma_send_ctxt { }; /* svc_rdma_backchannel.c */ -extern int svc_rdma_handle_bc_reply(struct rpc_xprt *xprt, - __be32 *rdma_resp, - struct xdr_buf *rcvbuf); +extern void svc_rdma_handle_bc_reply(struct svc_rqst *rqstp, + struct svc_rdma_recv_ctxt *rctxt); /* svc_rdma_recvfrom.c */ extern void svc_rdma_recv_ctxts_destroy(struct svcxprt_rdma *rdma); -- cgit v1.2.3 From 08e3c9f181bfb78d842c4f432888116d66e07c2f Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 29 Apr 2020 16:00:12 -0400 Subject: svcrdma: Remove the SVCRDMA_DEBUG macro Clean up: Commit d21b05f101ae ("rdma: SVCRMDA Header File") introduced the SVCRDMA_DEBUG macro, but it doesn't seem to have been used. Signed-off-by: Chuck Lever --- include/linux/sunrpc/svc_rdma.h | 1 - 1 file changed, 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/svc_rdma.h b/include/linux/sunrpc/svc_rdma.h index 8518c3f37e56..7ed82625dc0b 100644 --- a/include/linux/sunrpc/svc_rdma.h +++ b/include/linux/sunrpc/svc_rdma.h @@ -48,7 +48,6 @@ #include #include #include -#define SVCRDMA_DEBUG /* Default and maximum inline threshold sizes */ enum { -- cgit v1.2.3 From 02648908d19a99532b0839959e38ae53d95d2798 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Tue, 17 Mar 2020 14:12:15 -0400 Subject: SUNRPC: Rename svc_sock::sk_reclen Clean up. I find the name of the svc_sock::sk_reclen field confusing, so I've changed it to better reflect its function. This field is not read directly to get the record length. Rather, it is a buffer containing a record marker that needs to be decoded. Signed-off-by: Chuck Lever --- include/linux/sunrpc/svcsock.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/svcsock.h b/include/linux/sunrpc/svcsock.h index 771baadaee9d..b7ac7fe68306 100644 --- a/include/linux/sunrpc/svcsock.h +++ b/include/linux/sunrpc/svcsock.h @@ -28,7 +28,7 @@ struct svc_sock { /* private TCP part */ /* On-the-wire fragment header: */ - __be32 sk_reclen; + __be32 sk_marker; /* As we receive a record, this includes the length received so * far (including the fragment header): */ u32 sk_tcplen; @@ -41,12 +41,12 @@ struct svc_sock { static inline u32 svc_sock_reclen(struct svc_sock *svsk) { - return ntohl(svsk->sk_reclen) & RPC_FRAGMENT_SIZE_MASK; + return be32_to_cpu(svsk->sk_marker) & RPC_FRAGMENT_SIZE_MASK; } static inline u32 svc_sock_final_rec(struct svc_sock *svsk) { - return ntohl(svsk->sk_reclen) & RPC_LAST_STREAM_FRAGMENT; + return be32_to_cpu(svsk->sk_marker) & RPC_LAST_STREAM_FRAGMENT; } /* -- cgit v1.2.3 From ca07eda33e01eafa7a26ec06974f7eacee6a89c8 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 20 May 2020 17:30:12 -0400 Subject: SUNRPC: Refactor svc_recvfrom() This function is not currently "generic" so remove the documenting comment and rename it appropriately. Its internals are converted to use bio_vecs for reading from the transport socket. In existing typical sunrpc uses of bio_vecs, the bio_vec array is allocated dynamically. Here, instead, an array of bio_vecs is added to svc_rqst. The lifetime of this array can be greater than one call to xpo_recvfrom(): - Multiple calls to xpo_recvfrom() might be needed to read an RPC message completely. - At some later point, rq_arg.bvecs will point to this array and it will carry the received message into svc_process(). I also expect that a future optimization will remove either the rq_vec or rq_pages array in favor of rq_bvec, thus conserving the size of struct svc_rqst. Signed-off-by: Chuck Lever --- include/linux/sunrpc/svc.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux') diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h index fd390894a584..05da19a0516d 100644 --- a/include/linux/sunrpc/svc.h +++ b/include/linux/sunrpc/svc.h @@ -254,6 +254,7 @@ struct svc_rqst { struct page * *rq_page_end; /* one past the last page */ struct kvec rq_vec[RPCSVC_MAXPAGES]; /* generally useful.. */ + struct bio_vec rq_bvec[RPCSVC_MAXPAGES]; __be32 rq_xid; /* transmission id */ u32 rq_prog; /* program number */ -- cgit v1.2.3 From 24c5efe41c29ee3e55bcf5a1c9f61ca8709622e8 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Fri, 22 May 2020 12:01:33 +1000 Subject: sunrpc: clean up properly in gss_mech_unregister() gss_mech_register() calls svcauth_gss_register_pseudoflavor() for each flavour, but gss_mech_unregister() does not call auth_domain_put(). This is unbalanced and makes it impossible to reload the module. Change svcauth_gss_register_pseudoflavor() to return the registered auth_domain, and save it for later release. Cc: stable@vger.kernel.org (v2.6.12+) Link: https://bugzilla.kernel.org/show_bug.cgi?id=206651 Signed-off-by: NeilBrown Signed-off-by: J. Bruce Fields --- include/linux/sunrpc/gss_api.h | 1 + include/linux/sunrpc/svcauth_gss.h | 3 ++- 2 files changed, 3 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/gss_api.h b/include/linux/sunrpc/gss_api.h index bc07e51f20d1..bf4ac8a0268c 100644 --- a/include/linux/sunrpc/gss_api.h +++ b/include/linux/sunrpc/gss_api.h @@ -84,6 +84,7 @@ struct pf_desc { u32 service; char *name; char *auth_domain_name; + struct auth_domain *domain; bool datatouch; }; diff --git a/include/linux/sunrpc/svcauth_gss.h b/include/linux/sunrpc/svcauth_gss.h index ca39a388dc22..f09c82b0a7ae 100644 --- a/include/linux/sunrpc/svcauth_gss.h +++ b/include/linux/sunrpc/svcauth_gss.h @@ -20,7 +20,8 @@ int gss_svc_init(void); void gss_svc_shutdown(void); int gss_svc_init_net(struct net *net); void gss_svc_shutdown_net(struct net *net); -int svcauth_gss_register_pseudoflavor(u32 pseudoflavor, char * name); +struct auth_domain *svcauth_gss_register_pseudoflavor(u32 pseudoflavor, + char *name); u32 svcauth_gss_flavor(struct auth_domain *dom); #endif /* _LINUX_SUNRPC_SVCAUTH_GSS_H */ -- cgit v1.2.3