diff options
Diffstat (limited to 'net')
-rw-r--r-- | net/core/skbuff.c | 8 | ||||
-rw-r--r-- | net/core/sock.c | 2 | ||||
-rw-r--r-- | net/ipv4/inet_timewait_sock.c | 3 | ||||
-rw-r--r-- | net/iucv/af_iucv.c | 147 | ||||
-rw-r--r-- | net/iucv/iucv.c | 268 | ||||
-rw-r--r-- | net/rxrpc/ar-connection.c | 10 | ||||
-rw-r--r-- | net/rxrpc/ar-connevent.c | 7 | ||||
-rw-r--r-- | net/sunrpc/Makefile | 1 | ||||
-rw-r--r-- | net/sunrpc/backchannel_rqst.c | 281 | ||||
-rw-r--r-- | net/sunrpc/bc_svc.c | 81 | ||||
-rw-r--r-- | net/sunrpc/cache.c | 2 | ||||
-rw-r--r-- | net/sunrpc/clnt.c | 143 | ||||
-rw-r--r-- | net/sunrpc/sched.c | 2 | ||||
-rw-r--r-- | net/sunrpc/stats.c | 8 | ||||
-rw-r--r-- | net/sunrpc/sunrpc.h | 37 | ||||
-rw-r--r-- | net/sunrpc/svc.c | 136 | ||||
-rw-r--r-- | net/sunrpc/svc_xprt.c | 57 | ||||
-rw-r--r-- | net/sunrpc/svcsock.c | 161 | ||||
-rw-r--r-- | net/sunrpc/xprt.c | 60 | ||||
-rw-r--r-- | net/sunrpc/xprtrdma/svc_rdma_recvfrom.c | 8 | ||||
-rw-r--r-- | net/sunrpc/xprtsock.c | 217 |
21 files changed, 1497 insertions, 142 deletions
diff --git a/net/core/skbuff.c b/net/core/skbuff.c index a4c01f5c658..9e0597d189b 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -39,6 +39,7 @@ #include <linux/module.h> #include <linux/types.h> #include <linux/kernel.h> +#include <linux/kmemcheck.h> #include <linux/mm.h> #include <linux/interrupt.h> #include <linux/in.h> @@ -201,6 +202,8 @@ struct sk_buff *__alloc_skb(unsigned int size, gfp_t gfp_mask, skb->data = data; skb_reset_tail_pointer(skb); skb->end = skb->tail + size; + kmemcheck_annotate_bitfield(skb, flags1); + kmemcheck_annotate_bitfield(skb, flags2); #ifdef NET_SKBUFF_DATA_USES_OFFSET skb->mac_header = ~0U; #endif @@ -221,6 +224,8 @@ struct sk_buff *__alloc_skb(unsigned int size, gfp_t gfp_mask, struct sk_buff *child = skb + 1; atomic_t *fclone_ref = (atomic_t *) (child + 1); + kmemcheck_annotate_bitfield(child, flags1); + kmemcheck_annotate_bitfield(child, flags2); skb->fclone = SKB_FCLONE_ORIG; atomic_set(fclone_ref, 1); @@ -639,6 +644,9 @@ struct sk_buff *skb_clone(struct sk_buff *skb, gfp_t gfp_mask) n = kmem_cache_alloc(skbuff_head_cache, gfp_mask); if (!n) return NULL; + + kmemcheck_annotate_bitfield(n, flags1); + kmemcheck_annotate_bitfield(n, flags2); n->fclone = SKB_FCLONE_UNAVAILABLE; } diff --git a/net/core/sock.c b/net/core/sock.c index 06e26b77ad9..b0ba569bc97 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -945,6 +945,8 @@ static struct sock *sk_prot_alloc(struct proto *prot, gfp_t priority, sk = kmalloc(prot->obj_size, priority); if (sk != NULL) { + kmemcheck_annotate_bitfield(sk, flags); + if (security_sk_alloc(sk, family, priority)) goto out_free; diff --git a/net/ipv4/inet_timewait_sock.c b/net/ipv4/inet_timewait_sock.c index 68a8d892c71..61283f92882 100644 --- a/net/ipv4/inet_timewait_sock.c +++ b/net/ipv4/inet_timewait_sock.c @@ -9,6 +9,7 @@ */ #include <linux/kernel.h> +#include <linux/kmemcheck.h> #include <net/inet_hashtables.h> #include <net/inet_timewait_sock.h> #include <net/ip.h> @@ -120,6 +121,8 @@ struct inet_timewait_sock *inet_twsk_alloc(const struct sock *sk, const int stat if (tw != NULL) { const struct inet_sock *inet = inet_sk(sk); + kmemcheck_annotate_bitfield(tw, flags); + /* Give us an identity. */ tw->tw_daddr = inet->daddr; tw->tw_rcv_saddr = inet->rcv_saddr; diff --git a/net/iucv/af_iucv.c b/net/iucv/af_iucv.c index abadb4a846c..6be5f92d109 100644 --- a/net/iucv/af_iucv.c +++ b/net/iucv/af_iucv.c @@ -1,11 +1,12 @@ /* - * linux/net/iucv/af_iucv.c - * * IUCV protocol stack for Linux on zSeries * - * Copyright 2006 IBM Corporation + * Copyright IBM Corp. 2006, 2009 * * Author(s): Jennifer Hunt <jenhunt@us.ibm.com> + * Hendrik Brueckner <brueckner@linux.vnet.ibm.com> + * PM functions: + * Ursula Braun <ursula.braun@de.ibm.com> */ #define KMSG_COMPONENT "af_iucv" @@ -122,6 +123,122 @@ static inline void low_nmcpy(unsigned char *dst, char *src) memcpy(&dst[8], src, 8); } +static int afiucv_pm_prepare(struct device *dev) +{ +#ifdef CONFIG_PM_DEBUG + printk(KERN_WARNING "afiucv_pm_prepare\n"); +#endif + return 0; +} + +static void afiucv_pm_complete(struct device *dev) +{ +#ifdef CONFIG_PM_DEBUG + printk(KERN_WARNING "afiucv_pm_complete\n"); +#endif + return; +} + +/** + * afiucv_pm_freeze() - Freeze PM callback + * @dev: AFIUCV dummy device + * + * Sever all established IUCV communication pathes + */ +static int afiucv_pm_freeze(struct device *dev) +{ + struct iucv_sock *iucv; + struct sock *sk; + struct hlist_node *node; + int err = 0; + +#ifdef CONFIG_PM_DEBUG + printk(KERN_WARNING "afiucv_pm_freeze\n"); +#endif + read_lock(&iucv_sk_list.lock); + sk_for_each(sk, node, &iucv_sk_list.head) { + iucv = iucv_sk(sk); + skb_queue_purge(&iucv->send_skb_q); + skb_queue_purge(&iucv->backlog_skb_q); + switch (sk->sk_state) { + case IUCV_SEVERED: + case IUCV_DISCONN: + case IUCV_CLOSING: + case IUCV_CONNECTED: + if (iucv->path) { + err = iucv_path_sever(iucv->path, NULL); + iucv_path_free(iucv->path); + iucv->path = NULL; + } + break; + case IUCV_OPEN: + case IUCV_BOUND: + case IUCV_LISTEN: + case IUCV_CLOSED: + default: + break; + } + } + read_unlock(&iucv_sk_list.lock); + return err; +} + +/** + * afiucv_pm_restore_thaw() - Thaw and restore PM callback + * @dev: AFIUCV dummy device + * + * socket clean up after freeze + */ +static int afiucv_pm_restore_thaw(struct device *dev) +{ + struct iucv_sock *iucv; + struct sock *sk; + struct hlist_node *node; + +#ifdef CONFIG_PM_DEBUG + printk(KERN_WARNING "afiucv_pm_restore_thaw\n"); +#endif + read_lock(&iucv_sk_list.lock); + sk_for_each(sk, node, &iucv_sk_list.head) { + iucv = iucv_sk(sk); + switch (sk->sk_state) { + case IUCV_CONNECTED: + sk->sk_err = EPIPE; + sk->sk_state = IUCV_DISCONN; + sk->sk_state_change(sk); + break; + case IUCV_DISCONN: + case IUCV_SEVERED: + case IUCV_CLOSING: + case IUCV_LISTEN: + case IUCV_BOUND: + case IUCV_OPEN: + default: + break; + } + } + read_unlock(&iucv_sk_list.lock); + return 0; +} + +static struct dev_pm_ops afiucv_pm_ops = { + .prepare = afiucv_pm_prepare, + .complete = afiucv_pm_complete, + .freeze = afiucv_pm_freeze, + .thaw = afiucv_pm_restore_thaw, + .restore = afiucv_pm_restore_thaw, +}; + +static struct device_driver af_iucv_driver = { + .owner = THIS_MODULE, + .name = "afiucv", + .bus = &iucv_bus, + .pm = &afiucv_pm_ops, +}; + +/* dummy device used as trigger for PM functions */ +static struct device *af_iucv_dev; + /** * iucv_msg_length() - Returns the length of an iucv message. * @msg: Pointer to struct iucv_message, MUST NOT be NULL @@ -1617,8 +1734,30 @@ static int __init afiucv_init(void) err = sock_register(&iucv_sock_family_ops); if (err) goto out_proto; + /* establish dummy device */ + err = driver_register(&af_iucv_driver); + if (err) + goto out_sock; + af_iucv_dev = kzalloc(sizeof(struct device), GFP_KERNEL); + if (!af_iucv_dev) { + err = -ENOMEM; + goto out_driver; + } + dev_set_name(af_iucv_dev, "af_iucv"); + af_iucv_dev->bus = &iucv_bus; + af_iucv_dev->parent = iucv_root; + af_iucv_dev->release = (void (*)(struct device *))kfree; + af_iucv_dev->driver = &af_iucv_driver; + err = device_register(af_iucv_dev); + if (err) + goto out_driver; + return 0; +out_driver: + driver_unregister(&af_iucv_driver); +out_sock: + sock_unregister(PF_IUCV); out_proto: proto_unregister(&iucv_proto); out_iucv: @@ -1629,6 +1768,8 @@ out: static void __exit afiucv_exit(void) { + device_unregister(af_iucv_dev); + driver_unregister(&af_iucv_driver); sock_unregister(PF_IUCV); proto_unregister(&iucv_proto); iucv_unregister(&af_iucv_handler, 0); diff --git a/net/iucv/iucv.c b/net/iucv/iucv.c index 61e8038a55e..c833481d32e 100644 --- a/net/iucv/iucv.c +++ b/net/iucv/iucv.c @@ -1,7 +1,8 @@ /* * IUCV base infrastructure. * - * Copyright 2001, 2006 IBM Deutschland Entwicklung GmbH, IBM Corporation + * Copyright IBM Corp. 2001, 2009 + * * Author(s): * Original source: * Alan Altmark (Alan_Altmark@us.ibm.com) Sept. 2000 @@ -10,6 +11,8 @@ * Fritz Elfert (elfert@de.ibm.com, felfert@millenux.com) * Rewritten for af_iucv: * Martin Schwidefsky <schwidefsky@de.ibm.com> + * PM functions: + * Ursula Braun (ursula.braun@de.ibm.com) * * Documentation used: * The original source @@ -45,6 +48,7 @@ #include <linux/err.h> #include <linux/device.h> #include <linux/cpu.h> +#include <linux/reboot.h> #include <net/iucv/iucv.h> #include <asm/atomic.h> #include <asm/ebcdic.h> @@ -75,9 +79,24 @@ static int iucv_bus_match(struct device *dev, struct device_driver *drv) return 0; } +static int iucv_pm_prepare(struct device *); +static void iucv_pm_complete(struct device *); +static int iucv_pm_freeze(struct device *); +static int iucv_pm_thaw(struct device *); +static int iucv_pm_restore(struct device *); + +static struct dev_pm_ops iucv_pm_ops = { + .prepare = iucv_pm_prepare, + .complete = iucv_pm_complete, + .freeze = iucv_pm_freeze, + .thaw = iucv_pm_thaw, + .restore = iucv_pm_restore, +}; + struct bus_type iucv_bus = { .name = "iucv", .match = iucv_bus_match, + .pm = &iucv_pm_ops, }; EXPORT_SYMBOL(iucv_bus); @@ -147,6 +166,7 @@ enum iucv_command_codes { IUCV_RESUME = 14, IUCV_SEVER = 15, IUCV_SETMASK = 16, + IUCV_SETCONTROLMASK = 17, }; /* @@ -364,6 +384,18 @@ static void iucv_allow_cpu(void *data) parm->set_mask.ipmask = 0xf8; iucv_call_b2f0(IUCV_SETMASK, parm); + /* + * Enable all iucv control interrupts. + * ipmask contains bits for the different interrupts + * 0x80 - Flag to allow pending connections interrupts + * 0x40 - Flag to allow connection complete interrupts + * 0x20 - Flag to allow connection severed interrupts + * 0x10 - Flag to allow connection quiesced interrupts + * 0x08 - Flag to allow connection resumed interrupts + */ + memset(parm, 0, sizeof(union iucv_param)); + parm->set_mask.ipmask = 0xf8; + iucv_call_b2f0(IUCV_SETCONTROLMASK, parm); /* Set indication that iucv interrupts are allowed for this cpu. */ cpu_set(cpu, iucv_irq_cpumask); } @@ -389,6 +421,31 @@ static void iucv_block_cpu(void *data) } /** + * iucv_block_cpu_almost + * @data: unused + * + * Allow connection-severed interrupts only on this cpu. + */ +static void iucv_block_cpu_almost(void *data) +{ + int cpu = smp_processor_id(); + union iucv_param *parm; + + /* Allow iucv control interrupts only */ + parm = iucv_param_irq[cpu]; + memset(parm, 0, sizeof(union iucv_param)); + parm->set_mask.ipmask = 0x08; + iucv_call_b2f0(IUCV_SETMASK, parm); + /* Allow iucv-severed interrupt only */ + memset(parm, 0, sizeof(union iucv_param)); + parm->set_mask.ipmask = 0x20; + iucv_call_b2f0(IUCV_SETCONTROLMASK, parm); + + /* Clear indication that iucv interrupts are allowed for this cpu. */ + cpu_clear(cpu, iucv_irq_cpumask); +} + +/** * iucv_declare_cpu * @data: unused * @@ -758,6 +815,28 @@ void iucv_unregister(struct iucv_handler *handler, int smp) } EXPORT_SYMBOL(iucv_unregister); +static int iucv_reboot_event(struct notifier_block *this, + unsigned long event, void *ptr) +{ + int i, rc; + + get_online_cpus(); + on_each_cpu(iucv_block_cpu, NULL, 1); + preempt_disable(); + for (i = 0; i < iucv_max_pathid; i++) { + if (iucv_path_table[i]) + rc = iucv_sever_pathid(i, NULL); + } + preempt_enable(); + put_online_cpus(); + iucv_disable(); + return NOTIFY_DONE; +} + +static struct notifier_block iucv_reboot_notifier = { + .notifier_call = iucv_reboot_event, +}; + /** * iucv_path_accept * @path: address of iucv path structure @@ -777,6 +856,10 @@ int iucv_path_accept(struct iucv_path *path, struct iucv_handler *handler, int rc; local_bh_disable(); + if (!cpu_isset(smp_processor_id(), iucv_buffer_cpumask)) { + rc = -EIO; + goto out; + } /* Prepare parameter block. */ parm = iucv_param[smp_processor_id()]; memset(parm, 0, sizeof(union iucv_param)); @@ -792,6 +875,7 @@ int iucv_path_accept(struct iucv_path *path, struct iucv_handler *handler, path->msglim = parm->ctrl.ipmsglim; path->flags = parm->ctrl.ipflags1; } +out: local_bh_enable(); return rc; } @@ -821,6 +905,10 @@ int iucv_path_connect(struct iucv_path *path, struct iucv_handler *handler, spin_lock_bh(&iucv_table_lock); iucv_cleanup_queue(); + if (!cpu_isset(smp_processor_id(), iucv_buffer_cpumask)) { + rc = -EIO; + goto out; + } parm = iucv_param[smp_processor_id()]; memset(parm, 0, sizeof(union iucv_param)); parm->ctrl.ipmsglim = path->msglim; @@ -855,6 +943,7 @@ int iucv_path_connect(struct iucv_path *path, struct iucv_handler *handler, rc = -EIO; } } +out: spin_unlock_bh(&iucv_table_lock); return rc; } @@ -876,12 +965,17 @@ int iucv_path_quiesce(struct iucv_path *path, u8 userdata[16]) int rc; local_bh_disable(); + if (!cpu_isset(smp_processor_id(), iucv_buffer_cpumask)) { + rc = -EIO; + goto out; + } parm = iucv_param[smp_processor_id()]; memset(parm, 0, sizeof(union iucv_param)); if (userdata) memcpy(parm->ctrl.ipuser, userdata, sizeof(parm->ctrl.ipuser)); parm->ctrl.ippathid = path->pathid; rc = iucv_call_b2f0(IUCV_QUIESCE, parm); +out: local_bh_enable(); return rc; } @@ -903,12 +997,17 @@ int iucv_path_resume(struct iucv_path *path, u8 userdata[16]) int rc; local_bh_disable(); + if (!cpu_isset(smp_processor_id(), iucv_buffer_cpumask)) { + rc = -EIO; + goto out; + } parm = iucv_param[smp_processor_id()]; memset(parm, 0, sizeof(union iucv_param)); if (userdata) memcpy(parm->ctrl.ipuser, userdata, sizeof(parm->ctrl.ipuser)); parm->ctrl.ippathid = path->pathid; rc = iucv_call_b2f0(IUCV_RESUME, parm); +out: local_bh_enable(); return rc; } @@ -927,6 +1026,10 @@ int iucv_path_sever(struct iucv_path *path, u8 userdata[16]) int rc; preempt_disable(); + if (!cpu_isset(smp_processor_id(), iucv_buffer_cpumask)) { + rc = -EIO; + goto out; + } if (iucv_active_cpu != smp_processor_id()) spin_lock_bh(&iucv_table_lock); rc = iucv_sever_pathid(path->pathid, userdata); @@ -934,6 +1037,7 @@ int iucv_path_sever(struct iucv_path *path, u8 userdata[16]) list_del_init(&path->list); if (iucv_active_cpu != smp_processor_id()) spin_unlock_bh(&iucv_table_lock); +out: preempt_enable(); return rc; } @@ -956,6 +1060,10 @@ int iucv_message_purge(struct iucv_path *path, struct iucv_message *msg, int rc; local_bh_disable(); + if (!cpu_isset(smp_processor_id(), iucv_buffer_cpumask)) { + rc = -EIO; + goto out; + } parm = iucv_param[smp_processor_id()]; memset(parm, 0, sizeof(union iucv_param)); parm->purge.ippathid = path->pathid; @@ -967,6 +1075,7 @@ int iucv_message_purge(struct iucv_path *path, struct iucv_message *msg, msg->audit = (*(u32 *) &parm->purge.ipaudit) >> 8; msg->tag = parm->purge.ipmsgtag; } +out: local_bh_enable(); return rc; } @@ -1043,6 +1152,10 @@ int __iucv_message_receive(struct iucv_path *path, struct iucv_message *msg, if (msg->flags & IUCV_IPRMDATA) return iucv_message_receive_iprmdata(path, msg, flags, buffer, size, residual); + if (!cpu_isset(smp_processor_id(), iucv_buffer_cpumask)) { + rc = -EIO; + goto out; + } parm = iucv_param[smp_processor_id()]; memset(parm, 0, sizeof(union iucv_param)); parm->db.ipbfadr1 = (u32)(addr_t) buffer; @@ -1058,6 +1171,7 @@ int __iucv_message_receive(struct iucv_path *path, struct iucv_message *msg, if (residual) *residual = parm->db.ipbfln1f; } +out: return rc; } EXPORT_SYMBOL(__iucv_message_receive); @@ -1111,6 +1225,10 @@ int iucv_message_reject(struct iucv_path *path, struct iucv_message *msg) int rc; local_bh_disable(); + if (!cpu_isset(smp_processor_id(), iucv_buffer_cpumask)) { + rc = -EIO; + goto out; + } parm = iucv_param[smp_processor_id()]; memset(parm, 0, sizeof(union iucv_param)); parm->db.ippathid = path->pathid; @@ -1118,6 +1236,7 @@ int iucv_message_reject(struct iucv_path *path, struct iucv_message *msg) parm->db.iptrgcls = msg->class; parm->db.ipflags1 = (IUCV_IPTRGCLS | IUCV_IPFGMID | IUCV_IPFGPID); rc = iucv_call_b2f0(IUCV_REJECT, parm); +out: local_bh_enable(); return rc; } @@ -1145,6 +1264,10 @@ int iucv_message_reply(struct iucv_path *path, struct iucv_message *msg, int rc; local_bh_disable(); + if (!cpu_isset(smp_processor_id(), iucv_buffer_cpumask)) { + rc = -EIO; + goto out; + } parm = iucv_param[smp_processor_id()]; memset(parm, 0, sizeof(union iucv_param)); if (flags & IUCV_IPRMDATA) { @@ -1162,6 +1285,7 @@ int iucv_message_reply(struct iucv_path *path, struct iucv_message *msg, parm->db.iptrgcls = msg->class; } rc = iucv_call_b2f0(IUCV_REPLY, parm); +out: local_bh_enable(); return rc; } @@ -1190,6 +1314,10 @@ int __iucv_message_send(struct iucv_path *path, struct iucv_message *msg, union iucv_param *parm; int rc; + if (!cpu_isset(smp_processor_id(), iucv_buffer_cpumask)) { + rc = -EIO; + goto out; + } parm = iucv_param[smp_processor_id()]; memset(parm, 0, sizeof(union iucv_param)); if (flags & IUCV_IPRMDATA) { @@ -1212,6 +1340,7 @@ int __iucv_message_send(struct iucv_path *path, struct iucv_message *msg, rc = iucv_call_b2f0(IUCV_SEND, parm); if (!rc) msg->id = parm->db.ipmsgid; +out: return rc; } EXPORT_SYMBOL(__iucv_message_send); @@ -1272,6 +1401,10 @@ int iucv_message_send2way(struct iucv_path *path, struct iucv_message *msg, int rc; local_bh_disable(); + if (!cpu_isset(smp_processor_id(), iucv_buffer_cpumask)) { + rc = -EIO; + goto out; + } parm = iucv_param[smp_processor_id()]; memset(parm, 0, sizeof(union iucv_param)); if (flags & IUCV_IPRMDATA) { @@ -1297,6 +1430,7 @@ int iucv_message_send2way(struct iucv_path *path, struct iucv_message *msg, rc = iucv_call_b2f0(IUCV_SEND, parm); if (!rc) msg->id = parm->db.ipmsgid; +out: local_bh_enable(); return rc; } @@ -1687,6 +1821,130 @@ static void iucv_external_interrupt(u16 code) spin_unlock(&iucv_queue_lock); } +static int iucv_pm_prepare(struct device *dev) +{ + int rc = 0; + +#ifdef CONFIG_PM_DEBUG + printk(KERN_INFO "iucv_pm_prepare\n"); +#endif + if (dev->driver && dev->driver->pm && dev->driver->pm->prepare) + rc = dev->driver->pm->prepare(dev); + return rc; +} + +static void iucv_pm_complete(struct device *dev) +{ +#ifdef CONFIG_PM_DEBUG + printk(KERN_INFO "iucv_pm_complete\n"); +#endif + if (dev->driver && dev->driver->pm && dev->driver->pm->complete) + dev->driver->pm->complete(dev); +} + +/** + * iucv_path_table_empty() - determine if iucv path table is empty + * + * Returns 0 if there are still iucv pathes defined + * 1 if there are no iucv pathes defined + */ +int iucv_path_table_empty(void) +{ + int i; + + for (i = 0; i < iucv_max_pathid; i++) { + if (iucv_path_table[i]) + return 0; + } + return 1; +} + +/** + * iucv_pm_freeze() - Freeze PM callback + * @dev: iucv-based device + * + * disable iucv interrupts + * invoke callback function of the iucv-based driver + * shut down iucv, if no iucv-pathes are established anymore + */ +static int iucv_pm_freeze(struct device *dev) +{ + int cpu; + int rc = 0; + +#ifdef CONFIG_PM_DEBUG + printk(KERN_WARNING "iucv_pm_freeze\n"); +#endif + for_each_cpu_mask_nr(cpu, iucv_irq_cpumask) + smp_call_function_single(cpu, iucv_block_cpu_almost, NULL, 1); + if (dev->driver && dev->driver->pm && dev->driver->pm->freeze) + rc = dev->driver->pm->freeze(dev); + if (iucv_path_table_empty()) + iucv_disable(); + return rc; +} + +/** + * iucv_pm_thaw() - Thaw PM callback + * @dev: iucv-based device + * + * make iucv ready for use again: allocate path table, declare interrupt buffers + * and enable iucv interrupts + * invoke callback function of the iucv-based driver + */ +static int iucv_pm_thaw(struct device *dev) +{ + int rc = 0; + +#ifdef CONFIG_PM_DEBUG + printk(KERN_WARNING "iucv_pm_thaw\n"); +#endif + if (!iucv_path_table) { + rc = iucv_enable(); + if (rc) + goto out; + } + if (cpus_empty(iucv_irq_cpumask)) { + if (iucv_nonsmp_handler) + /* enable interrupts on one cpu */ + iucv_allow_cpu(NULL); + else + /* enable interrupts on all cpus */ + iucv_setmask_mp(); + } + if (dev->driver && dev->driver->pm && dev->driver->pm->thaw) + rc = dev->driver->pm->thaw(dev); +out: + return rc; +} + +/** + * iucv_pm_restore() - Restore PM callback + * @dev: iucv-based device + * + * make iucv ready for use again: allocate path table, declare interrupt buffers + * and enable iucv interrupts + * invoke callback function of the iucv-based driver + */ +static int iucv_pm_restore(struct device *dev) +{ + int rc = 0; + +#ifdef CONFIG_PM_DEBUG + printk(KERN_WARNING "iucv_pm_restore %p\n", iucv_path_table); +#endif + if (cpus_empty(iucv_irq_cpumask)) { + rc = iucv_query_maxconn(); + rc = iucv_enable(); + if (rc) + goto out; + } + if (dev->driver && dev->driver->pm && dev->driver->pm->restore) + rc = dev->driver->pm->restore(dev); +out: + return rc; +} + /** * iucv_init * @@ -1740,15 +1998,20 @@ static int __init iucv_init(void) rc = register_hotcpu_notifier(&iucv_cpu_notifier); if (rc) goto out_free; + rc = register_reboot_notifier(&iucv_reboot_notifier); + if (rc) + goto out_cpu; ASCEBC(iucv_error_no_listener, 16); ASCEBC(iucv_error_no_memory, 16); ASCEBC(iucv_error_pathid, 16); iucv_available = 1; rc = bus_register(&iucv_bus); if (rc) - goto out_cpu; + goto out_reboot; return 0; +out_reboot: + unregister_reboot_notifier(&iucv_reboot_notifier); out_cpu: unregister_hotcpu_notifier(&iucv_cpu_notifier); out_free: @@ -1783,6 +2046,7 @@ static void __exit iucv_exit(void) list_for_each_entry_safe(p, n, &iucv_work_queue, list) kfree(p); spin_unlock_irq(&iucv_queue_lock); + unregister_reboot_notifier(&iucv_reboot_notifier); unregister_hotcpu_notifier(&iucv_cpu_notifier); for_each_possible_cpu(cpu) { kfree(iucv_param_irq[cpu]); diff --git a/net/rxrpc/ar-connection.c b/net/rxrpc/ar-connection.c index 67e38a05624..9f1ce841a0b 100644 --- a/net/rxrpc/ar-connection.c +++ b/net/rxrpc/ar-connection.c @@ -444,6 +444,11 @@ int rxrpc_connect_call(struct rxrpc_sock *rx, conn = list_entry(bundle->avail_conns.next, struct rxrpc_connection, bundle_link); + if (conn->state >= RXRPC_CONN_REMOTELY_ABORTED) { + list_del_init(&conn->bundle_link); + bundle->num_conns--; + continue; + } if (--conn->avail_calls == 0) list_move(&conn->bundle_link, &bundle->busy_conns); @@ -461,6 +466,11 @@ int rxrpc_connect_call(struct rxrpc_sock *rx, conn = list_entry(bundle->unused_conns.next, struct rxrpc_connection, bundle_link); + if (conn->state >= RXRPC_CONN_REMOTELY_ABORTED) { + list_del_init(&conn->bundle_link); + bundle->num_conns--; + continue; + } ASSERTCMP(conn->avail_calls, ==, RXRPC_MAXCALLS); conn->avail_calls = RXRPC_MAXCALLS - 1; ASSERT(conn->channels[0] == NULL && diff --git a/net/rxrpc/ar-connevent.c b/net/rxrpc/ar-connevent.c index dc5cb1e1950..0505cdc4d6d 100644 --- a/net/rxrpc/ar-connevent.c +++ b/net/rxrpc/ar-connevent.c @@ -150,11 +150,15 @@ static int rxrpc_process_event(struct rxrpc_connection *conn, u32 serial; int loop, ret; - if (conn->state >= RXRPC_CONN_REMOTELY_ABORTED) + if (conn->state >= RXRPC_CONN_REMOTELY_ABORTED) { + kleave(" = -ECONNABORTED [%u]", conn->state); return -ECONNABORTED; + } serial = ntohl(sp->hdr.serial); + _enter("{%d},{%u,%%%u},", conn->debug_id, sp->hdr.type, serial); + switch (sp->hdr.type) { case RXRPC_PACKET_TYPE_ABORT: if (skb_copy_bits(skb, 0, &tmp, sizeof(tmp)) < 0) @@ -199,6 +203,7 @@ static int rxrpc_process_event(struct rxrpc_connection *conn, return 0; default: + _leave(" = -EPROTO [%u]", sp->hdr.type); return -EPROTO; } } diff --git a/net/sunrpc/Makefile b/net/sunrpc/Makefile index 5369aa369b3..db73fd2a3f0 100644 --- a/net/sunrpc/Makefile +++ b/net/sunrpc/Makefile @@ -13,5 +13,6 @@ sunrpc-y := clnt.o xprt.o socklib.o xprtsock.o sched.o \ rpcb_clnt.o timer.o xdr.o \ sunrpc_syms.o cache.o rpc_pipe.o \ svc_xprt.o +sunrpc-$(CONFIG_NFS_V4_1) += backchannel_rqst.o bc_svc.o sunrpc-$(CONFIG_PROC_FS) += stats.o sunrpc-$(CONFIG_SYSCTL) += sysctl.o diff --git a/net/sunrpc/backchannel_rqst.c b/net/sunrpc/backchannel_rqst.c new file mode 100644 index 00000000000..553621fb2c4 --- /dev/null +++ b/net/sunrpc/backchannel_rqst.c @@ -0,0 +1,281 @@ +/****************************************************************************** + +(c) 2007 Network Appliance, Inc. All Rights Reserved. +(c) 2009 NetApp. All Rights Reserved. + +NetApp provides this source code under the GPL v2 License. +The GPL v2 license is available at +http://opensource.org/licenses/gpl-license.php. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR +CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, +EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, +PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR +PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF +LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING +NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS +SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + +******************************************************************************/ + +#include <linux/tcp.h> +#include <linux/sunrpc/xprt.h> + +#ifdef RPC_DEBUG +#define RPCDBG_FACILITY RPCDBG_TRANS +#endif + +#if defined(CONFIG_NFS_V4_1) + +/* + * Helper routines that track the number of preallocation elements + * on the transport. + */ +static inline int xprt_need_to_requeue(struct rpc_xprt *xprt) +{ + return xprt->bc_alloc_count > 0; +} + +static inline void xprt_inc_alloc_count(struct rpc_xprt *xprt, unsigned int n) +{ + xprt->bc_alloc_count += n; +} + +static inline int xprt_dec_alloc_count(struct rpc_xprt *xprt, unsigned int n) +{ + return xprt->bc_alloc_count -= n; +} + +/* + * Free the preallocated rpc_rqst structure and the memory + * buffers hanging off of it. + */ +static void xprt_free_allocation(struct rpc_rqst *req) +{ + struct xdr_buf *xbufp; + + dprintk("RPC: free allocations for req= %p\n", req); + BUG_ON(test_bit(RPC_BC_PA_IN_USE, &req->rq_bc_pa_state)); + xbufp = &req->rq_private_buf; + free_page((unsigned long)xbufp->head[0].iov_base); + xbufp = &req->rq_snd_buf; + free_page((unsigned long)xbufp->head[0].iov_base); + list_del(&req->rq_bc_pa_list); + kfree(req); +} + +/* + * Preallocate up to min_reqs structures and related buffers for use + * by the backchannel. This function can be called multiple times + * when creating new sessions that use the same rpc_xprt. The + * preallocated buffers are added to the pool of resources used by + * the rpc_xprt. Anyone of these resources may be used used by an + * incoming callback request. It's up to the higher levels in the + * stack to enforce that the maximum number of session slots is not + * being exceeded. + * + * Some callback arguments can be large. For example, a pNFS server + * using multiple deviceids. The list can be unbound, but the client + * has the ability to tell the server the maximum size of the callback + * requests. Each deviceID is 16 bytes, so allocate one page + * for the arguments to have enough room to receive a number of these + * deviceIDs. The NFS client indicates to the pNFS server that its + * callback requests can be up to 4096 bytes in size. + */ +int xprt_setup_backchannel(struct rpc_xprt *xprt, unsigned int min_reqs) +{ + struct page *page_rcv = NULL, *page_snd = NULL; + struct xdr_buf *xbufp = NULL; + struct rpc_rqst *req, *tmp; + struct list_head tmp_list; + int i; + + dprintk("RPC: setup backchannel transport\n"); + + /* + * We use a temporary list to keep track of the preallocated + * buffers. Once we're done building the list we splice it + * into the backchannel preallocation list off of the rpc_xprt + * struct. This helps minimize the amount of time the list + * lock is held on the rpc_xprt struct. It also makes cleanup + * easier in case of memory allocation errors. + */ + INIT_LIST_HEAD(&tmp_list); + for (i = 0; i < min_reqs; i++) { + /* Pre-allocate one backchannel rpc_rqst */ + req = kzalloc(sizeof(struct rpc_rqst), GFP_KERNEL); + if (req == NULL) { + printk(KERN_ERR "Failed to create bc rpc_rqst\n"); + goto out_free; + } + + /* Add the allocated buffer to the tmp list */ + dprintk("RPC: adding req= %p\n", req); + list_add(&req->rq_bc_pa_list, &tmp_list); + + req->rq_xprt = xprt; + INIT_LIST_HEAD(&req->rq_list); + INIT_LIST_HEAD(&req->rq_bc_list); + + /* Preallocate one XDR receive buffer */ + page_rcv = alloc_page(GFP_KERNEL); + if (page_rcv == NULL) { + printk(KERN_ERR "Failed to create bc receive xbuf\n"); + goto out_free; + } + xbufp = &req->rq_rcv_buf; + xbufp->head[0].iov_base = page_address(page_rcv); + xbufp->head[0].iov_len = PAGE_SIZE; + xbufp->tail[0].iov_base = NULL; + xbufp->tail[0].iov_len = 0; + xbufp->page_len = 0; + xbufp->len = PAGE_SIZE; + xbufp->buflen = PAGE_SIZE; + + /* Preallocate one XDR send buffer */ + page_snd = alloc_page(GFP_KERNEL); + if (page_snd == NULL) { + printk(KERN_ERR "Failed to create bc snd xbuf\n"); + goto out_free; + } + + xbufp = &req->rq_snd_buf; + xbufp->head[0].iov_base = page_address(page_snd); + xbufp->head[0].iov_len = 0; + xbufp->tail[0].iov_base = NULL; + xbufp->tail[0].iov_len = 0; + xbufp->page_len = 0; + xbufp->len = 0; + xbufp->buflen = PAGE_SIZE; + } + + /* + * Add the temporary list to the backchannel preallocation list + */ + spin_lock_bh(&xprt->bc_pa_lock); + list_splice(&tmp_list, &xprt->bc_pa_list); + xprt_inc_alloc_count(xprt, min_reqs); + spin_unlock_bh(&xprt->bc_pa_lock); + + dprintk("RPC: setup backchannel transport done\n"); + return 0; + +out_free: + /* + * Memory allocation failed, free the temporary list + */ + list_for_each_entry_safe(req, tmp, &tmp_list, rq_bc_pa_list) + xprt_free_allocation(req); + + dprintk("RPC: setup backchannel transport failed\n"); + return -1; +} +EXPORT_SYMBOL(xprt_setup_backchannel); + +/* + * Destroys the backchannel preallocated structures. + * Since these structures may have been allocated by multiple calls + * to xprt_setup_backchannel, we only destroy up to the maximum number + * of reqs specified by the caller. + * @xprt: the transport holding the preallocated strucures + * @max_reqs the maximum number of preallocated structures to destroy + */ +void xprt_destroy_backchannel(struct rpc_xprt *xprt, unsigned int max_reqs) +{ + struct rpc_rqst *req = NULL, *tmp = NULL; + + dprintk("RPC: destroy backchannel transport\n"); + + BUG_ON(max_reqs == 0); + spin_lock_bh(&xprt->bc_pa_lock); + xprt_dec_alloc_count(xprt, max_reqs); + list_for_each_entry_safe(req, tmp, &xprt->bc_pa_list, rq_bc_pa_list) { + dprintk("RPC: req=%p\n", req); + xprt_free_allocation(req); + if (--max_reqs == 0) + break; + } + spin_unlock_bh(&xprt->bc_pa_lock); + + dprintk("RPC: backchannel list empty= %s\n", + list_empty(&xprt->bc_pa_list) ? "true" : "false"); +} +EXPORT_SYMBOL(xprt_destroy_backchannel); + +/* + * One or more rpc_rqst structure have been preallocated during the + * backchannel setup. Buffer space for the send and private XDR buffers + * has been preallocated as well. Use xprt_alloc_bc_request to allocate + * to this request. Use xprt_free_bc_request to return it. + * + * We know that we're called in soft interrupt context, grab the spin_lock + * since there is no need to grab the bottom half spin_lock. + * + * Return an available rpc_rqst, otherwise NULL if non are available. + */ +struct rpc_rqst *xprt_alloc_bc_request(struct rpc_xprt *xprt) +{ + struct rpc_rqst *req; + + dprintk("RPC: allocate a backchannel request\n"); + spin_lock(&xprt->bc_pa_lock); + if (!list_empty(&xprt->bc_pa_list)) { + req = list_first_entry(&xprt->bc_pa_list, struct rpc_rqst, + rq_bc_pa_list); + list_del(&req->rq_bc_pa_list); + } else { + req = NULL; + } + spin_unlock(&xprt->bc_pa_lock); + + if (req != NULL) { + set_bit(RPC_BC_PA_IN_USE, &req->rq_bc_pa_state); + req->rq_reply_bytes_recvd = 0; + req->rq_bytes_sent = 0; + memcpy(&req->rq_private_buf, &req->rq_rcv_buf, + sizeof(req->rq_private_buf)); + } + dprintk("RPC: backchannel req=%p\n", req); + return req; +} + +/* + * Return the preallocated rpc_rqst structure and XDR buffers + * associated with this rpc_task. + */ +void xprt_free_bc_request(struct rpc_rqst *req) +{ + struct rpc_xprt *xprt = req->rq_xprt; + + dprintk("RPC: free backchannel req=%p\n", req); + + smp_mb__before_clear_bit(); + BUG_ON(!test_bit(RPC_BC_PA_IN_USE, &req->rq_bc_pa_state)); + clear_bit(RPC_BC_PA_IN_USE, &req->rq_bc_pa_state); + smp_mb__after_clear_bit(); + + if (!xprt_need_to_requeue(xprt)) { + /* + * The last remaining session was destroyed while this + * entry was in use. Free the entry and don't attempt + * to add back to the list because there is no need to + * have anymore preallocated entries. + */ + dprintk("RPC: Last session removed req=%p\n", req); + xprt_free_allocation(req); + return; + } + + /* + * Return it to the list of preallocations so that it + * may be reused by a new callback request. + */ + spin_lock_bh(&xprt->bc_pa_lock); + list_add(&req->rq_bc_pa_list, &xprt->bc_pa_list); + spin_unlock_bh(&xprt->bc_pa_lock); +} + +#endif /* CONFIG_NFS_V4_1 */ diff --git a/net/sunrpc/bc_svc.c b/net/sunrpc/bc_svc.c new file mode 100644 index 00000000000..13f214f5312 --- /dev/null +++ b/net/sunrpc/bc_svc.c @@ -0,0 +1,81 @@ +/****************************************************************************** + +(c) 2007 Network Appliance, Inc. All Rights Reserved. +(c) 2009 NetApp. All Rights Reserved. + +NetApp provides this source code under the GPL v2 License. +The GPL v2 license is available at +http://opensource.org/licenses/gpl-license.php. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR +CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, +EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, +PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR +PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF +LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING +NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS +SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + +******************************************************************************/ + +/* + * The NFSv4.1 callback service helper routines. + * They implement the transport level processing required to send the + * reply over an existing open connection previously established by the client. + */ + +#if defined(CONFIG_NFS_V4_1) + +#include <linux/module.h> + +#include <linux/sunrpc/xprt.h> +#include <linux/sunrpc/sched.h> +#include <linux/sunrpc/bc_xprt.h> + +#define RPCDBG_FACILITY RPCDBG_SVCDSP + +void bc_release_request(struct rpc_task *task) +{ + struct rpc_rqst *req = task->tk_rqstp; + + dprintk("RPC: bc_release_request: task= %p\n", task); + + /* + * Release this request only if it's a backchannel + * preallocated request + */ + if (!bc_prealloc(req)) + return; + xprt_free_bc_request(req); +} + +/* Empty callback ops */ +static const struct rpc_call_ops nfs41_callback_ops = { +}; + + +/* + * Send the callback reply + */ +int bc_send(struct rpc_rqst *req) +{ + struct rpc_task *task; + int ret; + + dprintk("RPC: bc_send req= %p\n", req); + task = rpc_run_bc_task(req, &nfs41_callback_ops); + if (IS_ERR(task)) + ret = PTR_ERR(task); + else { + BUG_ON(atomic_read(&task->tk_count) != 1); + ret = task->tk_status; + rpc_put_task(task); + } + return ret; + dprintk("RPC: bc_send ret= %d \n", ret); +} + +#endif /* CONFIG_NFS_V4_1 */ diff --git a/net/sunrpc/cache.c b/net/sunrpc/cache.c index 20029a79a5d..ff0c23053d2 100644 --- a/net/sunrpc/cache.c +++ b/net/sunrpc/cache.c @@ -488,7 +488,7 @@ static void do_cache_clean(struct work_struct *work) { int delay = 5; if (cache_clean() == -1) - delay = 30*HZ; + delay = round_jiffies_relative(30*HZ); if (list_empty(&cache_list)) delay = 0; diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 5abab094441..5bc2f45bddf 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -36,7 +36,9 @@ #include <linux/sunrpc/clnt.h> #include <linux/sunrpc/rpc_pipe_fs.h> #include <linux/sunrpc/metrics.h> +#include <linux/sunrpc/bc_xprt.h> +#include "sunrpc.h" #ifdef RPC_DEBUG # define RPCDBG_FACILITY RPCDBG_CALL @@ -63,6 +65,9 @@ static void call_decode(struct rpc_task *task); static void call_bind(struct rpc_task *task); static void call_bind_status(struct rpc_task *task); static void call_transmit(struct rpc_task *task); +#if defined(CONFIG_NFS_V4_1) +static void call_bc_transmit(struct rpc_task *task); +#endif /* CONFIG_NFS_V4_1 */ static void call_status(struct rpc_task *task); static void call_transmit_status(struct rpc_task *task); static void call_refresh(struct rpc_task *task); @@ -613,6 +618,50 @@ rpc_call_async(struct rpc_clnt *clnt, const struct rpc_message *msg, int flags, } EXPORT_SYMBOL_GPL(rpc_call_async); +#if defined(CONFIG_NFS_V4_1) +/** + * rpc_run_bc_task - Allocate a new RPC task for backchannel use, then run + * rpc_execute against it + * @ops: RPC call ops + */ +struct rpc_task *rpc_run_bc_task(struct rpc_rqst *req, + const struct rpc_call_ops *tk_ops) +{ + struct rpc_task *task; + struct xdr_buf *xbufp = &req->rq_snd_buf; + struct rpc_task_setup task_setup_data = { + .callback_ops = tk_ops, + }; + + dprintk("RPC: rpc_run_bc_task req= %p\n", req); + /* + * Create an rpc_task to send the data + */ + task = rpc_new_task(&task_setup_data); + if (!task) { + xprt_free_bc_request(req); + goto out; + } + task->tk_rqstp = req; + + /* + * Set up the xdr_buf length. + * This also indicates that the buffer is XDR encoded already. + */ + xbufp->len = xbufp->head[0].iov_len + xbufp->page_len + + xbufp->tail[0].iov_len; + + task->tk_action = call_bc_transmit; + atomic_inc(&task->tk_count); + BUG_ON(atomic_read(&task->tk_count) != 2); + rpc_execute(task); + +out: + dprintk("RPC: rpc_run_bc_task: task= %p\n", task); + return task; +} +#endif /* CONFIG_NFS_V4_1 */ + void rpc_call_start(struct rpc_task *task) { @@ -695,6 +744,19 @@ void rpc_force_rebind(struct rpc_clnt *clnt) EXPORT_SYMBOL_GPL(rpc_force_rebind); /* + * Restart an (async) RPC call from the call_prepare state. + * Usually called from within the exit handler. + */ +void +rpc_restart_call_prepare(struct rpc_task *task) +{ + if (RPC_ASSASSINATED(task)) + return; + task->tk_action = rpc_prepare_task; +} +EXPORT_SYMBOL_GPL(rpc_restart_call_prepare); + +/* * Restart an (async) RPC call. Usually called from within the * exit handler. */ @@ -1085,7 +1147,7 @@ call_transmit(struct rpc_task *task) * in order to allow access to the socket to other RPC requests. */ call_transmit_status(task); - if (task->tk_msg.rpc_proc->p_decode != NULL) + if (rpc_reply_expected(task)) return; task->tk_action = rpc_exit_task; rpc_wake_up_queued_task(&task->tk_xprt->pending, task); @@ -1120,6 +1182,72 @@ call_transmit_status(struct rpc_task *task) } } +#if defined(CONFIG_NFS_V4_1) +/* + * 5b. Send the backchannel RPC reply. On error, drop the reply. In + * addition, disconnect on connectivity errors. + */ +static void +call_bc_transmit(struct rpc_task *task) +{ + struct rpc_rqst *req = task->tk_rqstp; + + BUG_ON(task->tk_status != 0); + task->tk_status = xprt_prepare_transmit(task); + if (task->tk_status == -EAGAIN) { + /* + * Could not reserve the transport. Try again after the + * transport is released. + */ + task->tk_status = 0; + task->tk_action = call_bc_transmit; + return; + } + + task->tk_action = rpc_exit_task; + if (task->tk_status < 0) { + printk(KERN_NOTICE "RPC: Could not send backchannel reply " + "error: %d\n", task->tk_status); + return; + } + + xprt_transmit(task); + xprt_end_transmit(task); + dprint_status(task); + switch (task->tk_status) { + case 0: + /* Success */ + break; + case -EHOSTDOWN: + case -EHOSTUNREACH: + case -ENETUNREACH: + case -ETIMEDOUT: + /* + * Problem reaching the server. Disconnect and let the + * forechannel reestablish the connection. The server will + * have to retransmit the backchannel request and we'll + * reprocess it. Since these ops are idempotent, there's no + * need to cache our reply at this time. + */ + printk(KERN_NOTICE "RPC: Could not send backchannel reply " + "error: %d\n", task->tk_status); + xprt_conditional_disconnect(task->tk_xprt, + req->rq_connect_cookie); + break; + default: + /* + * We were unable to reply and will have to drop the + * request. The server should reconnect and retransmit. + */ + BUG_ON(task->tk_status == -EAGAIN); + printk(KERN_NOTICE "RPC: Could not send backchannel reply " + "error: %d\n", task->tk_status); + break; + } + rpc_wake_up_queued_task(&req->rq_xprt->pending, task); +} +#endif /* CONFIG_NFS_V4_1 */ + /* * 6. Sort out the RPC call status */ @@ -1130,8 +1258,8 @@ call_status(struct rpc_task *task) struct rpc_rqst *req = task->tk_rqstp; int status; - if (req->rq_received > 0 && !req->rq_bytes_sent) - task->tk_status = req->rq_received; + if (req->rq_reply_bytes_recvd > 0 && !req->rq_bytes_sent) + task->tk_status = req->rq_reply_bytes_recvd; dprint_status(task); @@ -1248,7 +1376,7 @@ call_decode(struct rpc_task *task) /* * Ensure that we see all writes made by xprt_complete_rqst() - * before it changed req->rq_received. + * before it changed req->rq_reply_bytes_recvd. */ smp_rmb(); req->rq_rcv_buf.len = req->rq_private_buf.len; @@ -1289,7 +1417,7 @@ out_retry: task->tk_status = 0; /* Note: rpc_verify_header() may have freed the RPC slot */ if (task->tk_rqstp == req) { - req->rq_received = req->rq_rcv_buf.len = 0; + req->rq_reply_bytes_recvd = req->rq_rcv_buf.len = 0; if (task->tk_client->cl_discrtry) xprt_conditional_disconnect(task->tk_xprt, req->rq_connect_cookie); @@ -1377,13 +1505,14 @@ rpc_verify_header(struct rpc_task *task) } if ((len -= 3) < 0) goto out_overflow; - p += 1; /* skip XID */ + p += 1; /* skip XID */ if ((n = ntohl(*p++)) != RPC_REPLY) { dprintk("RPC: %5u %s: not an RPC reply: %x\n", - task->tk_pid, __func__, n); + task->tk_pid, __func__, n); goto out_garbage; } + if ((n = ntohl(*p++)) != RPC_MSG_ACCEPTED) { if (--len < 0) goto out_overflow; diff --git a/net/sunrpc/sched.c b/net/sunrpc/sched.c index ff50a054686..1102ce1251f 100644 --- a/net/sunrpc/sched.c +++ b/net/sunrpc/sched.c @@ -569,7 +569,7 @@ EXPORT_SYMBOL_GPL(rpc_delay); /* * Helper to call task->tk_ops->rpc_call_prepare */ -static void rpc_prepare_task(struct rpc_task *task) +void rpc_prepare_task(struct rpc_task *task) { task->tk_ops->rpc_call_prepare(task, task->tk_calldata); } diff --git a/net/sunrpc/stats.c b/net/sunrpc/stats.c index 1ef6e46d9da..1b4e6791ecf 100644 --- a/net/sunrpc/stats.c +++ b/net/sunrpc/stats.c @@ -141,12 +141,14 @@ EXPORT_SYMBOL_GPL(rpc_free_iostats); void rpc_count_iostats(struct rpc_task *task) { struct rpc_rqst *req = task->tk_rqstp; - struct rpc_iostats *stats = task->tk_client->cl_metrics; + struct rpc_iostats *stats; struct rpc_iostats *op_metrics; long rtt, execute, queue; - if (!stats || !req) + if (!task->tk_client || !task->tk_client->cl_metrics || !req) return; + + stats = task->tk_client->cl_metrics; op_metrics = &stats[task->tk_msg.rpc_proc->p_statidx]; op_metrics->om_ops++; @@ -154,7 +156,7 @@ void rpc_count_iostats(struct rpc_task *task) op_metrics->om_timeouts += task->tk_timeouts; op_metrics->om_bytes_sent += task->tk_bytes_sent; - op_metrics->om_bytes_recv += req->rq_received; + op_metrics->om_bytes_recv += req->rq_reply_bytes_recvd; queue = (long)req->rq_xtime - task->tk_start; if (queue < 0) diff --git a/net/sunrpc/sunrpc.h b/net/sunrpc/sunrpc.h new file mode 100644 index 00000000000..5d9dd742264 --- /dev/null +++ b/net/sunrpc/sunrpc.h @@ -0,0 +1,37 @@ +/****************************************************************************** + +(c) 2008 NetApp. All Rights Reserved. + +NetApp provides this source code under the GPL v2 License. +The GPL v2 license is available at +http://opensource.org/licenses/gpl-license.php. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR +CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, +EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, +PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR +PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF +LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING +NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS +SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + +******************************************************************************/ + +/* + * Functions and macros used internally by RPC + */ + +#ifndef _NET_SUNRPC_SUNRPC_H +#define _NET_SUNRPC_SUNRPC_H + +static inline int rpc_reply_expected(struct rpc_task *task) +{ + return (task->tk_msg.rpc_proc != NULL) && + (task->tk_msg.rpc_proc->p_decode != NULL); +} + +#endif /* _NET_SUNRPC_SUNRPC_H */ + diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c index 8847add6ca1..952f206ff30 100644 --- a/net/sunrpc/svc.c +++ b/net/sunrpc/svc.c @@ -25,6 +25,7 @@ #include <linux/sunrpc/stats.h> #include <linux/sunrpc/svcsock.h> #include <linux/sunrpc/clnt.h> +#include <linux/sunrpc/bc_xprt.h> #define RPCDBG_FACILITY RPCDBG_SVCDSP @@ -124,7 +125,7 @@ svc_pool_map_choose_mode(void) { unsigned int node; - if (num_online_nodes() > 1) { + if (nr_online_nodes > 1) { /* * Actually have multiple NUMA nodes, * so split pools on NUMA node boundaries @@ -486,6 +487,10 @@ svc_destroy(struct svc_serv *serv) if (svc_serv_is_pooled(serv)) svc_pool_map_put(); +#if defined(CONFIG_NFS_V4_1) + svc_sock_destroy(serv->bc_xprt); +#endif /* CONFIG_NFS_V4_1 */ + svc_unregister(serv); kfree(serv->sv_pools); kfree(serv); @@ -970,20 +975,18 @@ svc_printk(struct svc_rqst *rqstp, const char *fmt, ...) } /* - * Process the RPC request. + * Common routine for processing the RPC request. */ -int -svc_process(struct svc_rqst *rqstp) +static int +svc_process_common(struct svc_rqst *rqstp, struct kvec *argv, struct kvec *resv) { struct svc_program *progp; struct svc_version *versp = NULL; /* compiler food */ struct svc_procedure *procp = NULL; - struct kvec * argv = &rqstp->rq_arg.head[0]; - struct kvec * resv = &rqstp->rq_res.head[0]; struct svc_serv *serv = rqstp->rq_server; kxdrproc_t xdr; __be32 *statp; - u32 dir, prog, vers, proc; + u32 prog, vers, proc; __be32 auth_stat, rpc_stat; int auth_res; __be32 *reply_statp; @@ -993,19 +996,6 @@ svc_process(struct svc_rqst *rqstp) if (argv->iov_len < 6*4) goto err_short_len; - /* setup response xdr_buf. - * Initially it has just one page - */ - rqstp->rq_resused = 1; - resv->iov_base = page_address(rqstp->rq_respages[0]); - resv->iov_len = 0; - rqstp->rq_res.pages = rqstp->rq_respages + 1; - rqstp->rq_res.len = 0; - rqstp->rq_res.page_base = 0; - rqstp->rq_res.page_len = 0; - rqstp->rq_res.buflen = PAGE_SIZE; - rqstp->rq_res.tail[0].iov_base = NULL; - rqstp->rq_res.tail[0].iov_len = 0; /* Will be turned off only in gss privacy case: */ rqstp->rq_splice_ok = 1; /* Will be turned off only when NFSv4 Sessions are used */ @@ -1014,17 +1004,13 @@ svc_process(struct svc_rqst *rqstp) /* Setup reply header */ rqstp->rq_xprt->xpt_ops->xpo_prep_reply_hdr(rqstp); - rqstp->rq_xid = svc_getu32(argv); svc_putu32(resv, rqstp->rq_xid); - dir = svc_getnl(argv); vers = svc_getnl(argv); /* First words of reply: */ svc_putnl(resv, 1); /* REPLY */ - if (dir != 0) /* direction != CALL */ - goto err_bad_dir; if (vers != 2) /* RPC version number */ goto err_bad_rpc; @@ -1147,7 +1133,7 @@ svc_process(struct svc_rqst *rqstp) sendit: if (svc_authorise(rqstp)) goto dropit; - return svc_send(rqstp); + return 1; /* Caller can now send it */ dropit: svc_authorise(rqstp); /* doesn't hurt to call this twice */ @@ -1161,12 +1147,6 @@ err_short_len: goto dropit; /* drop request */ -err_bad_dir: - svc_printk(rqstp, "bad direction %d, dropping request\n", dir); - - serv->sv_stats->rpcbadfmt++; - goto dropit; /* drop request */ - err_bad_rpc: serv->sv_stats->rpcbadfmt++; svc_putnl(resv, 1); /* REJECT */ @@ -1220,6 +1200,100 @@ err_bad: EXPORT_SYMBOL_GPL(svc_process); /* + * Process the RPC request. + */ +int +svc_process(struct svc_rqst *rqstp) +{ + struct kvec *argv = &rqstp->rq_arg.head[0]; + struct kvec *resv = &rqstp->rq_res.head[0]; + struct svc_serv *serv = rqstp->rq_server; + u32 dir; + int error; + + /* + * Setup response xdr_buf. + * Initially it has just one page + */ + rqstp->rq_resused = 1; + resv->iov_base = page_address(rqstp->rq_respages[0]); + resv->iov_len = 0; + rqstp->rq_res.pages = rqstp->rq_respages + 1; + rqstp->rq_res.len = 0; + rqstp->rq_res.page_base = 0; + rqstp->rq_res.page_len = 0; + rqstp->rq_res.buflen = PAGE_SIZE; + rqstp->rq_res.tail[0].iov_base = NULL; + rqstp->rq_res.tail[0].iov_len = 0; + + rqstp->rq_xid = svc_getu32(argv); + + dir = svc_getnl(argv); + if (dir != 0) { + /* direction != CALL */ + svc_printk(rqstp, "bad direction %d, dropping request\n", dir); + serv->sv_stats->rpcbadfmt++; + svc_drop(rqstp); + return 0; + } + + error = svc_process_common(rqstp, argv, resv); + if (error <= 0) + return error; + + return svc_send(rqstp); +} + +#if defined(CONFIG_NFS_V4_1) +/* + * Process a backchannel RPC request that arrived over an existing + * outbound connection + */ +int +bc_svc_process(struct svc_serv *serv, struct rpc_rqst *req, + struct svc_rqst *rqstp) +{ + struct kvec *argv = &rqstp->rq_arg.head[0]; + struct kvec *resv = &rqstp->rq_res.head[0]; + int error; + + /* Build the svc_rqst used by the common processing routine */ + rqstp->rq_xprt = serv->bc_xprt; + rqstp->rq_xid = req->rq_xid; + rqstp->rq_prot = req->rq_xprt->prot; + rqstp->rq_server = serv; + + rqstp->rq_addrlen = sizeof(req->rq_xprt->addr); + memcpy(&rqstp->rq_addr, &req->rq_xprt->addr, rqstp->rq_addrlen); + memcpy(&rqstp->rq_arg, &req->rq_rcv_buf, sizeof(rqstp->rq_arg)); + memcpy(&rqstp->rq_res, &req->rq_snd_buf, sizeof(rqstp->rq_res)); + + /* reset result send buffer "put" position */ + resv->iov_len = 0; + + if (rqstp->rq_prot != IPPROTO_TCP) { + printk(KERN_ERR "No support for Non-TCP transports!\n"); + BUG(); + } + + /* + * Skip the next two words because they've already been + * processed in the trasport + */ + svc_getu32(argv); /* XID */ + svc_getnl(argv); /* CALLDIR */ + + error = svc_process_common(rqstp, argv, resv); + if (error <= 0) + return error; + + memcpy(&req->rq_snd_buf, &rqstp->rq_res, sizeof(req->rq_snd_buf)); + return bc_send(req); +} +EXPORT_SYMBOL(bc_svc_process); +#endif /* CONFIG_NFS_V4_1 */ + +/* * Return (transport-specific) limit on the rpc payload. */ u32 svc_max_payload(const struct svc_rqst *rqstp) diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index c200d92e57e..6f33d33cc06 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -11,6 +11,7 @@ #include <net/sock.h> #include <linux/sunrpc/stats.h> #include <linux/sunrpc/svc_xprt.h> +#include <linux/sunrpc/svcsock.h> #define RPCDBG_FACILITY RPCDBG_SVCXPRT @@ -1097,36 +1098,58 @@ struct svc_xprt *svc_find_xprt(struct svc_serv *serv, const char *xcl_name, } EXPORT_SYMBOL_GPL(svc_find_xprt); -/* - * Format a buffer with a list of the active transports. A zero for - * the buflen parameter disables target buffer overflow checking. +static int svc_one_xprt_name(const struct svc_xprt *xprt, + char *pos, int remaining) +{ + int len; + + len = snprintf(pos, remaining, "%s %u\n", + xprt->xpt_class->xcl_name, + svc_xprt_local_port(xprt)); + if (len >= remaining) + return -ENAMETOOLONG; + return len; +} + +/** + * svc_xprt_names - format a buffer with a list of transport names + * @serv: pointer to an RPC service + * @buf: pointer to a buffer to be filled in + * @buflen: length of buffer to be filled in + * + * Fills in @buf with a string containing a list of transport names, + * each name terminated with '\n'. + * + * Returns positive length of the filled-in string on success; otherwise + * a negative errno value is returned if an error occurs. */ -int svc_xprt_names(struct svc_serv *serv, char *buf, int buflen) +int svc_xprt_names(struct svc_serv *serv, char *buf, const int buflen) { struct svc_xprt *xprt; - char xprt_str[64]; - int totlen = 0; - int len; + int len, totlen; + char *pos; /* Sanity check args */ if (!serv) return 0; spin_lock_bh(&serv->sv_lock); + + pos = buf; + totlen = 0; list_for_each_entry(xprt, &serv->sv_permsocks, xpt_list) { - len = snprintf(xprt_str, sizeof(xprt_str), - "%s %d\n", xprt->xpt_class->xcl_name, - svc_xprt_local_port(xprt)); - /* If the string was truncated, replace with error string */ - if (len >= sizeof(xprt_str)) - strcpy(xprt_str, "name-too-long\n"); - /* Don't overflow buffer */ - len = strlen(xprt_str); - if (buflen && (len + totlen >= buflen)) + len = svc_one_xprt_name(xprt, pos, buflen - totlen); + if (len < 0) { + *buf = '\0'; + totlen = len; + } + if (len <= 0) break; - strcpy(buf+totlen, xprt_str); + + pos += len; totlen += len; } + spin_unlock_bh(&serv->sv_lock); return totlen; } diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c index 9d504234af4..23128ee191a 100644 --- a/net/sunrpc/svcsock.c +++ b/net/sunrpc/svcsock.c @@ -240,42 +240,76 @@ out: /* * Report socket names for nfsdfs */ -static int one_sock_name(char *buf, struct svc_sock *svsk) +static int svc_one_sock_name(struct svc_sock *svsk, char *buf, int remaining) { + const struct sock *sk = svsk->sk_sk; + const char *proto_name = sk->sk_protocol == IPPROTO_UDP ? + "udp" : "tcp"; int len; - switch(svsk->sk_sk->sk_family) { - case AF_INET: - len = sprintf(buf, "ipv4 %s %pI4 %d\n", - svsk->sk_sk->sk_protocol == IPPROTO_UDP ? - "udp" : "tcp", - &inet_sk(svsk->sk_sk)->rcv_saddr, - inet_sk(svsk->sk_sk)->num); + switch (sk->sk_family) { + case PF_INET: + len = snprintf(buf, remaining, "ipv4 %s %pI4 %d\n", + proto_name, + &inet_sk(sk)->rcv_saddr, + inet_sk(sk)->num); + break; + case PF_INET6: + len = snprintf(buf, remaining, "ipv6 %s %pI6 %d\n", + proto_name, + &inet6_sk(sk)->rcv_saddr, + inet_sk(sk)->num); break; default: - len = sprintf(buf, "*unknown-%d*\n", - svsk->sk_sk->sk_family); + len = snprintf(buf, remaining, "*unknown-%d*\n", + sk->sk_family); + } + + if (len >= remaining) { + *buf = '\0'; + return -ENAMETOOLONG; } return len; } -int -svc_sock_names(char *buf, struct svc_serv *serv, char *toclose) +/** + * svc_sock_names - construct a list of listener names in a string + * @serv: pointer to RPC service + * @buf: pointer to a buffer to fill in with socket names + * @buflen: size of the buffer to be filled + * @toclose: pointer to '\0'-terminated C string containing the name + * of a listener to be closed + * + * Fills in @buf with a '\n'-separated list of names of listener + * sockets. If @toclose is not NULL, the socket named by @toclose + * is closed, and is not included in the output list. + * + * Returns positive length of the socket name string, or a negative + * errno value on error. + */ +int svc_sock_names(struct svc_serv *serv, char *buf, const size_t buflen, + const char *toclose) { struct svc_sock *svsk, *closesk = NULL; int len = 0; if (!serv) return 0; + spin_lock_bh(&serv->sv_lock); list_for_each_entry(svsk, &serv->sv_permsocks, sk_xprt.xpt_list) { - int onelen = one_sock_name(buf+len, svsk); - if (toclose && strcmp(toclose, buf+len) == 0) + int onelen = svc_one_sock_name(svsk, buf + len, buflen - len); + if (onelen < 0) { + len = onelen; + break; + } + if (toclose && strcmp(toclose, buf + len) == 0) closesk = svsk; else len += onelen; } spin_unlock_bh(&serv->sv_lock); + if (closesk) /* Should unregister with portmap, but you cannot * unregister just one protocol... @@ -346,6 +380,7 @@ static void svc_sock_setbufsize(struct socket *sock, unsigned int snd, sock->sk->sk_sndbuf = snd * 2; sock->sk->sk_rcvbuf = rcv * 2; sock->sk->sk_userlocks |= SOCK_SNDBUF_LOCK|SOCK_RCVBUF_LOCK; + sock->sk->sk_write_space(sock->sk); release_sock(sock->sk); #endif } @@ -387,6 +422,15 @@ static void svc_write_space(struct sock *sk) } } +static void svc_tcp_write_space(struct sock *sk) +{ + struct socket *sock = sk->sk_socket; + + if (sk_stream_wspace(sk) >= sk_stream_min_wspace(sk) && sock) + clear_bit(SOCK_NOSPACE, &sock->flags); + svc_write_space(sk); +} + /* * Copy the UDP datagram's destination address to the rqstp structure. * The 'destination' address in this case is the address to which the @@ -427,13 +471,14 @@ static int svc_udp_recvfrom(struct svc_rqst *rqstp) long all[SVC_PKTINFO_SPACE / sizeof(long)]; } buffer; struct cmsghdr *cmh = &buffer.hdr; - int err, len; struct msghdr msg = { .msg_name = svc_addr(rqstp), .msg_control = cmh, .msg_controllen = sizeof(buffer), .msg_flags = MSG_DONTWAIT, }; + size_t len; + int err; if (test_and_clear_bit(XPT_CHNGBUF, &svsk->sk_xprt.xpt_flags)) /* udp sockets need large rcvbuf as all pending @@ -465,8 +510,8 @@ static int svc_udp_recvfrom(struct svc_rqst *rqstp) return -EAGAIN; } len = svc_addr_len(svc_addr(rqstp)); - if (len < 0) - return len; + if (len == 0) + return -EAFNOSUPPORT; rqstp->rq_addrlen = len; if (skb->tstamp.tv64 == 0) { skb->tstamp = ktime_get_real(); @@ -980,25 +1025,16 @@ static void svc_tcp_prep_reply_hdr(struct svc_rqst *rqstp) static int svc_tcp_has_wspace(struct svc_xprt *xprt) { struct svc_sock *svsk = container_of(xprt, struct svc_sock, sk_xprt); - struct svc_serv *serv = svsk->sk_xprt.xpt_server; + struct svc_serv *serv = svsk->sk_xprt.xpt_server; int required; - int wspace; - /* - * Set the SOCK_NOSPACE flag before checking the available - * sock space. - */ + if (test_bit(XPT_LISTENER, &xprt->xpt_flags)) + return 1; + required = atomic_read(&xprt->xpt_reserved) + serv->sv_max_mesg; + if (sk_stream_wspace(svsk->sk_sk) >= required) + return 1; set_bit(SOCK_NOSPACE, &svsk->sk_sock->flags); - required = atomic_read(&svsk->sk_xprt.xpt_reserved) + serv->sv_max_mesg; - wspace = sk_stream_wspace(svsk->sk_sk); - - if (wspace < sk_stream_min_wspace(svsk->sk_sk)) - return 0; - if (required * 2 > wspace) - return 0; - - clear_bit(SOCK_NOSPACE, &svsk->sk_sock->flags); - return 1; + return 0; } static struct svc_xprt *svc_tcp_create(struct svc_serv *serv, @@ -1054,7 +1090,7 @@ static void svc_tcp_init(struct svc_sock *svsk, struct svc_serv *serv) dprintk("setting up TCP socket for reading\n"); sk->sk_state_change = svc_tcp_state_change; sk->sk_data_ready = svc_tcp_data_ready; - sk->sk_write_space = svc_write_space; + sk->sk_write_space = svc_tcp_write_space; svsk->sk_reclen = 0; svsk->sk_tcplen = 0; @@ -1148,9 +1184,19 @@ static struct svc_sock *svc_setup_socket(struct svc_serv *serv, return svsk; } -int svc_addsock(struct svc_serv *serv, - int fd, - char *name_return) +/** + * svc_addsock - add a listener socket to an RPC service + * @serv: pointer to RPC service to which to add a new listener + * @fd: file descriptor of the new listener + * @name_return: pointer to buffer to fill in with name of listener + * @len: size of the buffer + * + * Fills in socket name and returns positive length of name if successful. + * Name is terminated with '\n'. On error, returns a negative errno + * value. + */ +int svc_addsock(struct svc_serv *serv, const int fd, char *name_return, + const size_t len) { int err = 0; struct socket *so = sockfd_lookup(fd, &err); @@ -1190,7 +1236,7 @@ int svc_addsock(struct svc_serv *serv, sockfd_put(so); return err; } - return one_sock_name(name_return, svsk); + return svc_one_sock_name(svsk, name_return, len); } EXPORT_SYMBOL_GPL(svc_addsock); @@ -1327,3 +1373,42 @@ static void svc_sock_free(struct svc_xprt *xprt) sock_release(svsk->sk_sock); kfree(svsk); } + +/* + * Create a svc_xprt. + * + * For internal use only (e.g. nfsv4.1 backchannel). + * Callers should typically use the xpo_create() method. + */ +struct svc_xprt *svc_sock_create(struct svc_serv *serv, int prot) +{ + struct svc_sock *svsk; + struct svc_xprt *xprt = NULL; + + dprintk("svc: %s\n", __func__); + svsk = kzalloc(sizeof(*svsk), GFP_KERNEL); + if (!svsk) + goto out; + + xprt = &svsk->sk_xprt; + if (prot == IPPROTO_TCP) + svc_xprt_init(&svc_tcp_class, xprt, serv); + else if (prot == IPPROTO_UDP) + svc_xprt_init(&svc_udp_class, xprt, serv); + else + BUG(); +out: + dprintk("svc: %s return %p\n", __func__, xprt); + return xprt; +} +EXPORT_SYMBOL_GPL(svc_sock_create); + +/* + * Destroy a svc_sock. + */ +void svc_sock_destroy(struct svc_xprt *xprt) +{ + if (xprt) + kfree(container_of(xprt, struct svc_sock, sk_xprt)); +} +EXPORT_SYMBOL_GPL(svc_sock_destroy); diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 06ca058572f..f412a852bc7 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -12,8 +12,9 @@ * - Next, the caller puts together the RPC message, stuffs it into * the request struct, and calls xprt_transmit(). * - xprt_transmit sends the message and installs the caller on the - * transport's wait list. At the same time, it installs a timer that - * is run after the packet's timeout has expired. + * transport's wait list. At the same time, if a reply is expected, + * it installs a timer that is run after the packet's timeout has + * expired. * - When a packet arrives, the data_ready handler walks the list of * pending requests for that transport. If a matching XID is found, the * caller is woken up, and the timer removed. @@ -46,6 +47,8 @@ #include <linux/sunrpc/clnt.h> #include <linux/sunrpc/metrics.h> +#include "sunrpc.h" + /* * Local variables */ @@ -192,8 +195,8 @@ EXPORT_SYMBOL_GPL(xprt_load_transport); */ int xprt_reserve_xprt(struct rpc_task *task) { - struct rpc_xprt *xprt = task->tk_xprt; struct rpc_rqst *req = task->tk_rqstp; + struct rpc_xprt *xprt = req->rq_xprt; if (test_and_set_bit(XPRT_LOCKED, &xprt->state)) { if (task == xprt->snd_task) @@ -803,9 +806,10 @@ void xprt_complete_rqst(struct rpc_task *task, int copied) list_del_init(&req->rq_list); req->rq_private_buf.len = copied; - /* Ensure all writes are done before we update req->rq_received */ + /* Ensure all writes are done before we update */ + /* req->rq_reply_bytes_recvd */ smp_wmb(); - req->rq_received = copied; + req->rq_reply_bytes_recvd = copied; rpc_wake_up_queued_task(&xprt->pending, task); } EXPORT_SYMBOL_GPL(xprt_complete_rqst); @@ -820,7 +824,7 @@ static void xprt_timer(struct rpc_task *task) dprintk("RPC: %5u xprt_timer\n", task->tk_pid); spin_lock_bh(&xprt->transport_lock); - if (!req->rq_received) { + if (!req->rq_reply_bytes_recvd) { if (xprt->ops->timer) xprt->ops->timer(task); } else @@ -842,8 +846,8 @@ int xprt_prepare_transmit(struct rpc_task *task) dprintk("RPC: %5u xprt_prepare_transmit\n", task->tk_pid); spin_lock_bh(&xprt->transport_lock); - if (req->rq_received && !req->rq_bytes_sent) { - err = req->rq_received; + if (req->rq_reply_bytes_recvd && !req->rq_bytes_sent) { + err = req->rq_reply_bytes_recvd; goto out_unlock; } if (!xprt->ops->reserve_xprt(task)) @@ -855,7 +859,7 @@ out_unlock: void xprt_end_transmit(struct rpc_task *task) { - xprt_release_write(task->tk_xprt, task); + xprt_release_write(task->tk_rqstp->rq_xprt, task); } /** @@ -872,8 +876,11 @@ void xprt_transmit(struct rpc_task *task) dprintk("RPC: %5u xprt_transmit(%u)\n", task->tk_pid, req->rq_slen); - if (!req->rq_received) { - if (list_empty(&req->rq_list)) { + if (!req->rq_reply_bytes_recvd) { + if (list_empty(&req->rq_list) && rpc_reply_expected(task)) { + /* + * Add to the list only if we're expecting a reply + */ spin_lock_bh(&xprt->transport_lock); /* Update the softirq receive buffer */ memcpy(&req->rq_private_buf, &req->rq_rcv_buf, @@ -908,8 +915,13 @@ void xprt_transmit(struct rpc_task *task) /* Don't race with disconnect */ if (!xprt_connected(xprt)) task->tk_status = -ENOTCONN; - else if (!req->rq_received) + else if (!req->rq_reply_bytes_recvd && rpc_reply_expected(task)) { + /* + * Sleep on the pending queue since + * we're expecting a reply. + */ rpc_sleep_on(&xprt->pending, task, xprt_timer); + } spin_unlock_bh(&xprt->transport_lock); } @@ -982,11 +994,17 @@ static void xprt_request_init(struct rpc_task *task, struct rpc_xprt *xprt) */ void xprt_release(struct rpc_task *task) { - struct rpc_xprt *xprt = task->tk_xprt; + struct rpc_xprt *xprt; struct rpc_rqst *req; + int is_bc_request; if (!(req = task->tk_rqstp)) return; + + /* Preallocated backchannel request? */ + is_bc_request = bc_prealloc(req); + + xprt = req->rq_xprt; rpc_count_iostats(task); spin_lock_bh(&xprt->transport_lock); xprt->ops->release_xprt(xprt, task); @@ -999,10 +1017,19 @@ void xprt_release(struct rpc_task *task) mod_timer(&xprt->timer, xprt->last_used + xprt->idle_timeout); spin_unlock_bh(&xprt->transport_lock); - xprt->ops->buf_free(req->rq_buffer); + if (!bc_prealloc(req)) + xprt->ops->buf_free(req->rq_buffer); task->tk_rqstp = NULL; if (req->rq_release_snd_buf) req->rq_release_snd_buf(req); + + /* + * Early exit if this is a backchannel preallocated request. + * There is no need to have it added to the RPC slot list. + */ + if (is_bc_request) + return; + memset(req, 0, sizeof(*req)); /* mark unused */ dprintk("RPC: %5u release request %p\n", task->tk_pid, req); @@ -1049,6 +1076,11 @@ found: INIT_LIST_HEAD(&xprt->free); INIT_LIST_HEAD(&xprt->recv); +#if defined(CONFIG_NFS_V4_1) + spin_lock_init(&xprt->bc_pa_lock); + INIT_LIST_HEAD(&xprt->bc_pa_list); +#endif /* CONFIG_NFS_V4_1 */ + INIT_WORK(&xprt->task_cleanup, xprt_autoclose); setup_timer(&xprt->timer, xprt_init_autodisconnect, (unsigned long)xprt); diff --git a/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c b/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c index 42a6f9f2028..9e884383134 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c +++ b/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c @@ -397,14 +397,14 @@ static int rdma_read_xdr(struct svcxprt_rdma *xprt, if (!ch) return 0; - /* Allocate temporary reply and chunk maps */ - rpl_map = svc_rdma_get_req_map(); - chl_map = svc_rdma_get_req_map(); - svc_rdma_rcl_chunk_counts(ch, &ch_count, &byte_count); if (ch_count > RPCSVC_MAXPAGES) return -EINVAL; + /* Allocate temporary reply and chunk maps */ + rpl_map = svc_rdma_get_req_map(); + chl_map = svc_rdma_get_req_map(); + if (!xprt->sc_frmr_pg_list_len) sge_count = map_read_chunks(xprt, rqstp, hdr_ctxt, rmsgp, rpl_map, chl_map, ch_count, diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 6c2d6158655..83c73c4d017 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -34,6 +34,9 @@ #include <linux/sunrpc/sched.h> #include <linux/sunrpc/xprtsock.h> #include <linux/file.h> +#ifdef CONFIG_NFS_V4_1 +#include <linux/sunrpc/bc_xprt.h> +#endif #include <net/sock.h> #include <net/checksum.h> @@ -270,6 +273,13 @@ struct sock_xprt { #define TCP_RCV_COPY_FRAGHDR (1UL << 1) #define TCP_RCV_COPY_XID (1UL << 2) #define TCP_RCV_COPY_DATA (1UL << 3) +#define TCP_RCV_READ_CALLDIR (1UL << 4) +#define TCP_RCV_COPY_CALLDIR (1UL << 5) + +/* + * TCP RPC flags + */ +#define TCP_RPC_REPLY (1UL << 6) static inline struct sockaddr *xs_addr(struct rpc_xprt *xprt) { @@ -956,7 +966,7 @@ static inline void xs_tcp_read_fraghdr(struct rpc_xprt *xprt, struct xdr_skb_rea transport->tcp_offset = 0; /* Sanity check of the record length */ - if (unlikely(transport->tcp_reclen < 4)) { + if (unlikely(transport->tcp_reclen < 8)) { dprintk("RPC: invalid TCP record fragment length\n"); xprt_force_disconnect(xprt); return; @@ -991,33 +1001,77 @@ static inline void xs_tcp_read_xid(struct sock_xprt *transport, struct xdr_skb_r if (used != len) return; transport->tcp_flags &= ~TCP_RCV_COPY_XID; - transport->tcp_flags |= TCP_RCV_COPY_DATA; + transport->tcp_flags |= TCP_RCV_READ_CALLDIR; transport->tcp_copied = 4; - dprintk("RPC: reading reply for XID %08x\n", + dprintk("RPC: reading %s XID %08x\n", + (transport->tcp_flags & TCP_RPC_REPLY) ? "reply for" + : "request with", ntohl(transport->tcp_xid)); xs_tcp_check_fraghdr(transport); } -static inline void xs_tcp_read_request(struct rpc_xprt *xprt, struct xdr_skb_reader *desc) +static inline void xs_tcp_read_calldir(struct sock_xprt *transport, + struct xdr_skb_reader *desc) { - struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt); - struct rpc_rqst *req; + size_t len, used; + u32 offset; + __be32 calldir; + + /* + * We want transport->tcp_offset to be 8 at the end of this routine + * (4 bytes for the xid and 4 bytes for the call/reply flag). + * When this function is called for the first time, + * transport->tcp_offset is 4 (after having already read the xid). + */ + offset = transport->tcp_offset - sizeof(transport->tcp_xid); + len = sizeof(calldir) - offset; + dprintk("RPC: reading CALL/REPLY flag (%Zu bytes)\n", len); + used = xdr_skb_read_bits(desc, &calldir, len); + transport->tcp_offset += used; + if (used != len) + return; + transport->tcp_flags &= ~TCP_RCV_READ_CALLDIR; + transport->tcp_flags |= TCP_RCV_COPY_CALLDIR; + transport->tcp_flags |= TCP_RCV_COPY_DATA; + /* + * We don't yet have the XDR buffer, so we will write the calldir + * out after we get the buffer from the 'struct rpc_rqst' + */ + if (ntohl(calldir) == RPC_REPLY) + transport->tcp_flags |= TCP_RPC_REPLY; + else + transport->tcp_flags &= ~TCP_RPC_REPLY; + dprintk("RPC: reading %s CALL/REPLY flag %08x\n", + (transport->tcp_flags & TCP_RPC_REPLY) ? + "reply for" : "request with", calldir); + xs_tcp_check_fraghdr(transport); +} + +static inline void xs_tcp_read_common(struct rpc_xprt *xprt, + struct xdr_skb_reader *desc, + struct rpc_rqst *req) +{ + struct sock_xprt *transport = + container_of(xprt, struct sock_xprt, xprt); struct xdr_buf *rcvbuf; size_t len; ssize_t r; - /* Find and lock the request corresponding to this xid */ - spin_lock(&xprt->transport_lock); - req = xprt_lookup_rqst(xprt, transport->tcp_xid); - if (!req) { - transport->tcp_flags &= ~TCP_RCV_COPY_DATA; - dprintk("RPC: XID %08x request not found!\n", - ntohl(transport->tcp_xid)); - spin_unlock(&xprt->transport_lock); - return; + rcvbuf = &req->rq_private_buf; + + if (transport->tcp_flags & TCP_RCV_COPY_CALLDIR) { + /* + * Save the RPC direction in the XDR buffer + */ + __be32 calldir = transport->tcp_flags & TCP_RPC_REPLY ? + htonl(RPC_REPLY) : 0; + + memcpy(rcvbuf->head[0].iov_base + transport->tcp_copied, + &calldir, sizeof(calldir)); + transport->tcp_copied += sizeof(calldir); + transport->tcp_flags &= ~TCP_RCV_COPY_CALLDIR; } - rcvbuf = &req->rq_private_buf; len = desc->count; if (len > transport->tcp_reclen - transport->tcp_offset) { struct xdr_skb_reader my_desc; @@ -1054,7 +1108,7 @@ static inline void xs_tcp_read_request(struct rpc_xprt *xprt, struct xdr_skb_rea "tcp_offset = %u, tcp_reclen = %u\n", xprt, transport->tcp_copied, transport->tcp_offset, transport->tcp_reclen); - goto out; + return; } dprintk("RPC: XID %08x read %Zd bytes\n", @@ -1070,11 +1124,125 @@ static inline void xs_tcp_read_request(struct rpc_xprt *xprt, struct xdr_skb_rea transport->tcp_flags &= ~TCP_RCV_COPY_DATA; } -out: + return; +} + +/* + * Finds the request corresponding to the RPC xid and invokes the common + * tcp read code to read the data. + */ +static inline int xs_tcp_read_reply(struct rpc_xprt *xprt, + struct xdr_skb_reader *desc) +{ + struct sock_xprt *transport = + container_of(xprt, struct sock_xprt, xprt); + struct rpc_rqst *req; + + dprintk("RPC: read reply XID %08x\n", ntohl(transport->tcp_xid)); + + /* Find and lock the request corresponding to this xid */ + spin_lock(&xprt->transport_lock); + req = xprt_lookup_rqst(xprt, transport->tcp_xid); + if (!req) { + dprintk("RPC: XID %08x request not found!\n", + ntohl(transport->tcp_xid)); + spin_unlock(&xprt->transport_lock); + return -1; + } + + xs_tcp_read_common(xprt, desc, req); + if (!(transport->tcp_flags & TCP_RCV_COPY_DATA)) xprt_complete_rqst(req->rq_task, transport->tcp_copied); + spin_unlock(&xprt->transport_lock); - xs_tcp_check_fraghdr(transport); + return 0; +} + +#if defined(CONFIG_NFS_V4_1) +/* + * Obtains an rpc_rqst previously allocated and invokes the common + * tcp read code to read the data. The result is placed in the callback + * queue. + * If we're unable to obtain the rpc_rqst we schedule the closing of the + * connection and return -1. + */ +static inline int xs_tcp_read_callback(struct rpc_xprt *xprt, + struct xdr_skb_reader *desc) +{ + struct sock_xprt *transport = + container_of(xprt, struct sock_xprt, xprt); + struct rpc_rqst *req; + + req = xprt_alloc_bc_request(xprt); + if (req == NULL) { + printk(KERN_WARNING "Callback slot table overflowed\n"); + xprt_force_disconnect(xprt); + return -1; + } + + req->rq_xid = transport->tcp_xid; + dprintk("RPC: read callback XID %08x\n", ntohl(req->rq_xid)); + xs_tcp_read_common(xprt, desc, req); + + if (!(transport->tcp_flags & TCP_RCV_COPY_DATA)) { + struct svc_serv *bc_serv = xprt->bc_serv; + + /* + * Add callback request to callback list. The callback + * service sleeps on the sv_cb_waitq waiting for new + * requests. Wake it up after adding enqueing the + * request. + */ + dprintk("RPC: add callback request to list\n"); + spin_lock(&bc_serv->sv_cb_lock); + list_add(&req->rq_bc_list, &bc_serv->sv_cb_list); + spin_unlock(&bc_serv->sv_cb_lock); + wake_up(&bc_serv->sv_cb_waitq); + } + + req->rq_private_buf.len = transport->tcp_copied; + + return 0; +} + +static inline int _xs_tcp_read_data(struct rpc_xprt *xprt, + struct xdr_skb_reader *desc) +{ + struct sock_xprt *transport = + container_of(xprt, struct sock_xprt, xprt); + + return (transport->tcp_flags & TCP_RPC_REPLY) ? + xs_tcp_read_reply(xprt, desc) : + xs_tcp_read_callback(xprt, desc); +} +#else +static inline int _xs_tcp_read_data(struct rpc_xprt *xprt, + struct xdr_skb_reader *desc) +{ + return xs_tcp_read_reply(xprt, desc); +} +#endif /* CONFIG_NFS_V4_1 */ + +/* + * Read data off the transport. This can be either an RPC_CALL or an + * RPC_REPLY. Relay the processing to helper functions. + */ +static void xs_tcp_read_data(struct rpc_xprt *xprt, + struct xdr_skb_reader *desc) +{ + struct sock_xprt *transport = + container_of(xprt, struct sock_xprt, xprt); + + if (_xs_tcp_read_data(xprt, desc) == 0) + xs_tcp_check_fraghdr(transport); + else { + /* + * The transport_lock protects the request handling. + * There's no need to hold it to update the tcp_flags. + */ + transport->tcp_flags &= ~TCP_RCV_COPY_DATA; + } } static inline void xs_tcp_read_discard(struct sock_xprt *transport, struct xdr_skb_reader *desc) @@ -1114,9 +1282,14 @@ static int xs_tcp_data_recv(read_descriptor_t *rd_desc, struct sk_buff *skb, uns xs_tcp_read_xid(transport, &desc); continue; } + /* Read in the call/reply flag */ + if (transport->tcp_flags & TCP_RCV_READ_CALLDIR) { + xs_tcp_read_calldir(transport, &desc); + continue; + } /* Read in the request data */ if (transport->tcp_flags & TCP_RCV_COPY_DATA) { - xs_tcp_read_request(xprt, &desc); + xs_tcp_read_data(xprt, &desc); continue; } /* Skip over any trailing bytes on short reads */ @@ -1792,6 +1965,7 @@ static void xs_tcp_setup_socket(struct rpc_xprt *xprt, */ set_bit(XPRT_CONNECTION_CLOSE, &xprt->state); xprt_force_disconnect(xprt); + break; case -ECONNREFUSED: case -ECONNRESET: case -ENETUNREACH: @@ -2010,6 +2184,9 @@ static struct rpc_xprt_ops xs_tcp_ops = { .buf_free = rpc_free, .send_request = xs_tcp_send_request, .set_retrans_timeout = xprt_set_retrans_timeout_def, +#if defined(CONFIG_NFS_V4_1) + .release_request = bc_release_request, +#endif /* CONFIG_NFS_V4_1 */ .close = xs_tcp_close, .destroy = xs_destroy, .print_stats = xs_tcp_print_stats, |