X-Git-Url: https://git.whamcloud.com/?a=blobdiff_plain;f=lnet%2Finclude%2Flnet%2Flib-lnet.h;h=d245c1180b55e3c4418506b4e5e778e2e373a147;hb=a2e61838f8de89e0f7c80c3bf288cbeb1b358baa;hp=6d577e1e7df69efa2a497f2494f1c9ffeaaa17fc;hpb=17802b329cdea7057cb1db75a5b803f1c4e44a2b;p=fs%2Flustre-release.git

diff --git a/lnet/include/lnet/lib-lnet.h b/lnet/include/lnet/lib-lnet.h
index 6d577e1..d245c11 100644
--- a/lnet/include/lnet/lib-lnet.h
+++ b/lnet/include/lnet/lib-lnet.h
@@ -74,8 +74,10 @@ extern struct lnet the_lnet;			/* THE network */
 /** exclusive lock */
 #define LNET_LOCK_EX		CFS_PERCPT_LOCK_EX
 
-/* default timeout */
+/* default timeout and credits */
 #define DEFAULT_PEER_TIMEOUT    180
+#define DEFAULT_PEER_CREDITS    8
+#define DEFAULT_CREDITS         256
 
 #ifdef HAVE_KERN_SOCK_GETNAME_2ARGS
 #define lnet_kernel_getpeername(sock, addr, addrlen) \
@@ -91,11 +93,12 @@ extern struct lnet the_lnet;			/* THE network */
 
 /*
  * kernel 5.3: commit ef11db3310e272d3d8dbe8739e0770820dd20e52
+ * kernel 4.18.0-193.el8:
  * added in_dev_for_each_ifa_rtnl and in_dev_for_each_ifa_rcu
  * and removed for_ifa and endfor_ifa.
  * Use the _rntl variant as the current locking is rtnl.
  */
-#ifdef in_dev_for_each_ifa_rtnl
+#ifdef HAVE_IN_DEV_FOR_EACH_IFA_RTNL
 #define DECLARE_CONST_IN_IFADDR(ifa)		const struct in_ifaddr *ifa
 #define endfor_ifa(in_dev)
 #else
@@ -212,6 +215,62 @@ extern struct kmem_cache *lnet_small_mds_cachep; /* <= LNET_SMALL_MD_SIZE bytes
 extern struct kmem_cache *lnet_rspt_cachep;
 extern struct kmem_cache *lnet_msg_cachep;
 
+static inline bool
+lnet_ni_set_status_locked(struct lnet_ni *ni, __u32 status)
+__must_hold(&ni->ni_lock)
+{
+	bool update = false;
+
+	if (ni->ni_status && ni->ni_status->ns_status != status) {
+		CDEBUG(D_NET, "ni %s status changed from %#x to %#x\n",
+		       libcfs_nid2str(ni->ni_nid),
+		       ni->ni_status->ns_status, status);
+		ni->ni_status->ns_status = status;
+		update = true;
+	}
+
+	return update;
+}
+
+static inline bool
+lnet_ni_set_status(struct lnet_ni *ni, __u32 status)
+{
+	bool update;
+
+	lnet_ni_lock(ni);
+	update = lnet_ni_set_status_locked(ni, status);
+	lnet_ni_unlock(ni);
+
+	return update;
+}
+
+static inline void lnet_md_wait_handling(struct lnet_libmd *md, int cpt)
+{
+	wait_queue_head_t *wq = __var_waitqueue(md);
+#ifdef HAVE_WAIT_QUEUE_ENTRY
+	struct wait_bit_queue_entry entry;
+	wait_queue_entry_t *wqe = &entry.wq_entry;
+#else
+	struct wait_bit_queue entry;
+	wait_queue_entry_t *wqe = &entry.wait;
+#endif
+	init_wait_var_entry(&entry, md, 0);
+	prepare_to_wait_event(wq, wqe, TASK_IDLE);
+	if (md->md_flags & LNET_MD_FLAG_HANDLING) {
+		/* Race with unlocked call to ->md_handler.
+		 * It is safe to drop the res_lock here as the
+		 * caller has only just claimed it.
+		 */
+		lnet_res_unlock(cpt);
+		schedule();
+		/* Cannot check md now, it might be freed.  Caller
+		 * must reclaim reference and check.
+		 */
+		lnet_res_lock(cpt);
+	}
+	finish_wait(wq, wqe);
+}
+
 static inline void
 lnet_md_free(struct lnet_libmd *md)
 {
@@ -454,12 +513,14 @@ struct lnet_net *lnet_get_net_locked(__u32 net_id);
 int lnet_lib_init(void);
 void lnet_lib_exit(void);
 
+extern unsigned int lnet_response_tracking;
 extern unsigned lnet_transaction_timeout;
 extern unsigned lnet_retry_count;
 extern unsigned int lnet_lnd_timeout;
 extern unsigned int lnet_numa_range;
 extern unsigned int lnet_health_sensitivity;
 extern unsigned int lnet_recovery_interval;
+extern unsigned int lnet_recovery_limit;
 extern unsigned int lnet_peer_discovery_disabled;
 extern unsigned int lnet_drop_asym_route;
 extern unsigned int router_sensitivity_percentage;
@@ -609,6 +670,7 @@ void lnet_set_reply_msg_len(struct lnet_ni *ni, struct lnet_msg *msg,
 void lnet_detach_rsp_tracker(struct lnet_libmd *md, int cpt);
 void lnet_clean_zombie_rstqs(void);
 
+bool lnet_md_discarded(struct lnet_libmd *md);
 void lnet_finalize(struct lnet_msg *msg, int rc);
 bool lnet_send_error_simulation(struct lnet_msg *msg,
 				enum lnet_msg_hstatus *hstatus);
@@ -626,7 +688,6 @@ int lnet_msg_containers_create(void);
 
 char *lnet_health_error2str(enum lnet_msg_hstatus hstatus);
 char *lnet_msgtyp2str(int type);
-void lnet_print_hdr(struct lnet_hdr *hdr);
 int lnet_fail_nid(lnet_nid_t nid, unsigned int threshold);
 
 /** \addtogroup lnet_fault_simulation @{ */
@@ -649,7 +710,7 @@ bool lnet_delay_rule_match_locked(struct lnet_hdr *hdr, struct lnet_msg *msg);
 /** @} lnet_fault_simulation */
 
 void lnet_counters_get_common(struct lnet_counters_common *common);
-void lnet_counters_get(struct lnet_counters *counters);
+int lnet_counters_get(struct lnet_counters *counters);
 void lnet_counters_reset(void);
 
 unsigned int lnet_iov_nob(unsigned int niov, struct kvec *iov);
@@ -888,8 +949,8 @@ lnet_peer_needs_push(struct lnet_peer *lp)
 }
 
 /*
- * A peer is alive if it satisfies the following two conditions:
- *  1. peer health >= LNET_MAX_HEALTH_VALUE * router_sensitivity_percentage
+ * A peer NI is alive if it satisfies the following two conditions:
+ *  1. peer NI health >= LNET_MAX_HEALTH_VALUE * router_sensitivity_percentage
  *  2. the cached NI status received when we discover the peer is UP
  */
 static inline bool