X-Git-Url: https://git.whamcloud.com/?p=fs%2Flustre-release.git;a=blobdiff_plain;f=lnet%2Fklnds%2Fo2iblnd%2Fo2iblnd.h;h=f696eadebb2341152ecfc1f398006062ddd445d0;hp=ed5bacfd328ebc7185b623fe23010215835294e3;hb=ac54ab5516bc3c1fb995f8189234123cbbac61e7;hpb=50fa6533220487ab928404d14a95fe5503dc5ab7 diff --git a/lnet/klnds/o2iblnd/o2iblnd.h b/lnet/klnds/o2iblnd/o2iblnd.h index ed5bacf..f696ead 100644 --- a/lnet/klnds/o2iblnd/o2iblnd.h +++ b/lnet/klnds/o2iblnd/o2iblnd.h @@ -26,7 +26,7 @@ * GPL HEADER END */ /* - * Copyright 2008 Sun Microsystems, Inc. All rights reserved + * Copyright (c) 2007, 2010, Oracle and/or its affiliates. All rights reserved. * Use is subject to license terms. */ /* @@ -65,7 +65,6 @@ #include #include #include -#include #include #include @@ -89,7 +88,7 @@ typedef int gfp_t; /* tunables fixed at compile time */ #ifdef CONFIG_SMP -# define IBLND_N_SCHED num_online_cpus() /* # schedulers */ +# define IBLND_N_SCHED cfs_num_online_cpus() /* # schedulers */ #else # define IBLND_N_SCHED 1 /* # schedulers */ #endif @@ -107,7 +106,8 @@ typedef struct int *kib_keepalive; /* keepalive timeout (seconds) */ int *kib_ntx; /* # tx descs */ int *kib_credits; /* # concurrent sends */ - int *kib_peercredits; /* # concurrent sends to 1 peer */ + int *kib_peertxcredits; /* # concurrent sends to 1 peer */ + int *kib_peerrtrcredits; /* # per-peer router buffer credits */ int *kib_peercredits_hiw; /* # when eagerly to return credits */ int *kib_peertimeout; /* seconds to consider peer dead */ char **kib_default_ipif; /* default IPoIB interface */ @@ -132,11 +132,11 @@ extern kib_tunables_t kiblnd_tunables; #define IBLND_CREDIT_HIGHWATER_V1 7 /* V1 only : when eagerly to return credits */ #define IBLND_CREDITS_DEFAULT 8 /* default # of peer credits */ -#define IBLND_CREDITS_MAX 4096 /* Max # of peer credits */ +#define IBLND_CREDITS_MAX ((typeof(((kib_msg_t*) 0)->ibm_credits)) - 1) /* Max # of peer credits */ #define IBLND_MSG_QUEUE_SIZE(v) ((v) == IBLND_MSG_VERSION_1 ? \ IBLND_MSG_QUEUE_SIZE_V1 : \ - *kiblnd_tunables.kib_peercredits) /* # messages/RDMAs in-flight */ + *kiblnd_tunables.kib_peertxcredits) /* # messages/RDMAs in-flight */ #define IBLND_CREDITS_HIGHWATER(v) ((v) == IBLND_MSG_VERSION_1 ? \ IBLND_CREDIT_HIGHWATER_V1 : \ *kiblnd_tunables.kib_peercredits_hiw) /* when eagerly to return credits */ @@ -173,8 +173,6 @@ kiblnd_concurrent_sends_v1(void) /* TX messages (shared by all connections) */ #define IBLND_TX_MSGS() (*kiblnd_tunables.kib_ntx) -#define IBLND_TX_MSG_BYTES() (IBLND_TX_MSGS() * IBLND_MSG_SIZE) -#define IBLND_TX_MSG_PAGES() ((IBLND_TX_MSG_BYTES() + PAGE_SIZE - 1) / PAGE_SIZE) /* RX messages (per connection) */ #define IBLND_RX_MSGS(v) (IBLND_MSG_QUEUE_SIZE(v) * 2 + IBLND_OOB_MSGS(v)) @@ -188,27 +186,7 @@ kiblnd_concurrent_sends_v1(void) typedef struct { - struct ib_device *ibp_device; /* device for mapping */ - int ibp_npages; /* # pages */ - struct page *ibp_pages[0]; -} kib_pages_t; - -typedef struct { - spinlock_t ibmp_lock; /* serialize */ - int ibmp_allocated; /* MR in use */ - struct list_head ibmp_free_list; /* pre-allocated MR */ -} kib_phys_mr_pool_t; - -typedef struct { - struct list_head ibpm_link; /* link node */ - struct ib_mr *ibpm_mr; /* MR */ - __u64 ibpm_iova; /* Virtual I/O address */ - int ibpm_refcount; /* reference count */ -} kib_phys_mr_t; - -typedef struct -{ - struct list_head ibd_list; /* chain on kib_devs */ + cfs_list_t ibd_list; /* chain on kib_devs */ __u32 ibd_ifip; /* IPoIB interface IP */ char ibd_ifname[32]; /* IPoIB interface name */ int ibd_nnets; /* # nets extant */ @@ -225,49 +203,149 @@ typedef struct struct ib_mr **ibd_mrs; /* MR for non RDMA I/O */ } kib_dev_t; +#define IBLND_POOL_DEADLINE 300 /* # of seconds to keep pool alive */ + +typedef struct +{ + struct ib_device *ibp_device; /* device for mapping */ + int ibp_npages; /* # pages */ + struct page *ibp_pages[0]; /* page array */ +} kib_pages_t; + +struct kib_pmr_pool; + +typedef struct { + cfs_list_t pmr_list; /* chain node */ + struct ib_phys_buf *pmr_ipb; /* physical buffer */ + struct ib_mr *pmr_mr; /* IB MR */ + struct kib_pmr_pool *pmr_pool; /* owner of this MR */ + __u64 pmr_iova; /* Virtual I/O address */ + int pmr_refcount; /* reference count */ +} kib_phys_mr_t; + +struct kib_pool; +struct kib_poolset; + +typedef int (*kib_ps_pool_create_t)(struct kib_poolset *ps, int inc, struct kib_pool **pp_po); +typedef void (*kib_ps_pool_destroy_t)(struct kib_pool *po); +typedef void (*kib_ps_node_init_t)(struct kib_pool *po, + cfs_list_t *node); +typedef void (*kib_ps_node_fini_t)(struct kib_pool *po, + cfs_list_t *node); + +struct kib_net; + +#define IBLND_POOL_NAME_LEN 32 + +typedef struct kib_poolset +{ + cfs_spinlock_t ps_lock; /* serialize */ + struct kib_net *ps_net; /* network it belongs to */ + char ps_name[IBLND_POOL_NAME_LEN]; /* pool set name */ + cfs_list_t ps_pool_list; /* list of pools */ + cfs_time_t ps_next_retry; /* time stamp for retry if failed to allocate */ + int ps_increasing; /* is allocating new pool */ + int ps_pool_size; /* new pool size */ + + kib_ps_pool_create_t ps_pool_create; /* create a new pool */ + kib_ps_pool_destroy_t ps_pool_destroy; /* destroy a pool */ + kib_ps_node_init_t ps_node_init; /* initialize new allocated node */ + kib_ps_node_fini_t ps_node_fini; /* finalize node */ +} kib_poolset_t; + +typedef struct kib_pool +{ + cfs_list_t po_list; /* chain on pool list */ + cfs_list_t po_free_list; /* pre-allocated node */ + kib_poolset_t *po_owner; /* pool_set of this pool */ + cfs_time_t po_deadline; /* deadline of this pool */ + int po_allocated; /* # of elements in use */ + int po_size; /* # of pre-allocated elements */ +} kib_pool_t; + +typedef struct { + kib_poolset_t tps_poolset; /* pool-set */ + __u64 tps_next_tx_cookie; /* cookie of TX */ +} kib_tx_poolset_t; + +typedef struct { + kib_pool_t tpo_pool; /* pool */ + struct kib_tx *tpo_tx_descs; /* all the tx descriptors */ + kib_pages_t *tpo_tx_pages; /* premapped tx msg pages */ +} kib_tx_pool_t; + +typedef struct { + kib_poolset_t pps_poolset; /* pool-set */ +} kib_pmr_poolset_t; + +typedef struct kib_pmr_pool { + kib_pool_t ppo_pool; /* pool */ +} kib_pmr_pool_t; + +typedef struct +{ + cfs_spinlock_t fps_lock; /* serialize */ + struct kib_net *fps_net; /* IB network */ + cfs_list_t fps_pool_list; /* FMR pool list */ + __u64 fps_version; /* validity stamp */ + int fps_increasing; /* is allocating new pool */ + cfs_time_t fps_next_retry; /* time stamp for retry if failed to allocate */ +} kib_fmr_poolset_t; + typedef struct { + cfs_list_t fpo_list; /* chain on pool list */ + kib_fmr_poolset_t *fpo_owner; /* owner of this pool */ + struct ib_fmr_pool *fpo_fmr_pool; /* IB FMR pool */ + cfs_time_t fpo_deadline; /* deadline of this pool */ + int fpo_map_count; /* # of mapped FMR */ +} kib_fmr_pool_t; + +typedef struct { + struct ib_pool_fmr *fmr_pfmr; /* IB pool fmr */ + kib_fmr_pool_t *fmr_pool; /* pool of FMR */ +} kib_fmr_t; + +typedef struct kib_net +{ __u64 ibn_incarnation; /* my epoch */ int ibn_init; /* initialisation state */ int ibn_shutdown; /* shutting down? */ + unsigned int ibn_with_fmr:1; /* FMR? */ + unsigned int ibn_with_pmr:1; /* PMR? */ - atomic_t ibn_npeers; /* # peers extant */ - atomic_t ibn_nconns; /* # connections extant */ - - __u64 ibn_tx_next_cookie; /* RDMA completion cookie */ - struct kib_tx *ibn_tx_descs; /* all the tx descriptors */ - kib_pages_t *ibn_tx_pages; /* premapped tx msg pages */ - struct list_head ibn_idle_txs; /* idle tx descriptors */ - spinlock_t ibn_tx_lock; /* serialise */ + cfs_atomic_t ibn_npeers; /* # peers extant */ + cfs_atomic_t ibn_nconns; /* # connections extant */ - struct ib_fmr_pool *ibn_fmrpool; /* FMR pool for RDMA I/O */ - kib_phys_mr_pool_t *ibn_pmrpool; /* Physical MR pool for RDMA I/O */ + kib_tx_poolset_t ibn_tx_ps; /* tx pool-set */ + kib_fmr_poolset_t ibn_fmr_ps; /* fmr pool-set */ + kib_pmr_poolset_t ibn_pmr_ps; /* pmr pool-set */ kib_dev_t *ibn_dev; /* underlying IB device */ } kib_net_t; typedef struct { - int kib_init; /* initialisation state */ - int kib_shutdown; /* shut down? */ - struct list_head kib_devs; /* IB devices extant */ - atomic_t kib_nthreads; /* # live threads */ - rwlock_t kib_global_lock; /* stabilize net/dev/peer/conn ops */ + int kib_init; /* initialisation state */ + int kib_shutdown; /* shut down? */ + cfs_list_t kib_devs; /* IB devices extant */ + cfs_atomic_t kib_nthreads; /* # live threads */ + cfs_rwlock_t kib_global_lock; /* stabilize net/dev/peer/conn ops */ - struct list_head *kib_peers; /* hash table of all my known peers */ - int kib_peer_hash_size; /* size of kib_peers */ + cfs_list_t *kib_peers; /* hash table of all my known peers */ + int kib_peer_hash_size;/* size of kib_peers */ - void *kib_connd; /* the connd task (serialisation assertions) */ - struct list_head kib_connd_conns; /* connections to setup/teardown */ - struct list_head kib_connd_zombies; /* connections with zero refcount */ - wait_queue_head_t kib_connd_waitq; /* connection daemon sleeps here */ - spinlock_t kib_connd_lock; /* serialise */ + void *kib_connd; /* the connd task (serialisation assertions) */ + cfs_list_t kib_connd_conns; /* connections to setup/teardown */ + cfs_list_t kib_connd_zombies;/* connections with zero refcount */ + cfs_waitq_t kib_connd_waitq; /* connection daemon sleeps here */ + cfs_spinlock_t kib_connd_lock; /* serialise */ - wait_queue_head_t kib_sched_waitq; /* schedulers sleep here */ - struct list_head kib_sched_conns; /* conns to check for rx completions */ - spinlock_t kib_sched_lock; /* serialise */ + cfs_waitq_t kib_sched_waitq; /* schedulers sleep here */ + cfs_list_t kib_sched_conns; /* conns to check for rx completions */ + cfs_spinlock_t kib_sched_lock; /* serialise */ - struct ib_qp_attr kib_error_qpa; /* QP->ERROR */ + struct ib_qp_attr kib_error_qpa; /* QP->ERROR */ } kib_data_t; #define IBLND_INIT_NOTHING 0 @@ -397,7 +475,7 @@ typedef struct { typedef struct kib_rx /* receive message */ { - struct list_head rx_list; /* queue for attention */ + cfs_list_t rx_list; /* queue for attention */ struct kib_conn *rx_conn; /* owning conn */ int rx_nob; /* # bytes received (-1 while posted) */ enum ib_wc_status rx_status; /* completion status */ @@ -415,11 +493,12 @@ typedef struct kib_rx /* receive message */ typedef struct kib_tx /* transmit message */ { - struct list_head tx_list; /* queue on idle_txs ibc_tx_queue etc. */ + cfs_list_t tx_list; /* queue on idle_txs ibc_tx_queue etc. */ + kib_tx_pool_t *tx_pool; /* pool I'm from */ struct kib_conn *tx_conn; /* owning conn */ - int tx_sending; /* # tx callbacks outstanding */ - int tx_queued; /* queued for sending */ - int tx_waiting; /* waiting for peer */ + short tx_sending; /* # tx callbacks outstanding */ + short tx_queued; /* queued for sending */ + short tx_waiting; /* waiting for peer */ int tx_status; /* LNET completion status */ unsigned long tx_deadline; /* completion deadline */ __u64 tx_cookie; /* completion cookie */ @@ -433,11 +512,10 @@ typedef struct kib_tx /* transmit message */ kib_rdma_desc_t *tx_rd; /* rdma descriptor */ int tx_nfrags; /* # entries in... */ struct scatterlist *tx_frags; /* dma_map_sg descriptor */ - struct ib_phys_buf *tx_ipb; /* physical buffer (for iWARP) */ __u64 *tx_pages; /* rdma phys page addrs */ union { - kib_phys_mr_t *pmr; /* MR for physical buffer */ - struct ib_pool_fmr *fmr; /* rdma mapping (mapped if != NULL) */ + kib_phys_mr_t *pmr; /* MR for physical buffer */ + kib_fmr_t fmr; /* FMR */ } tx_u; int tx_dmadir; /* dma direction */ } kib_tx_t; @@ -450,36 +528,37 @@ typedef struct kib_connvars typedef struct kib_conn { - struct kib_peer *ibc_peer; /* owning peer */ - struct list_head ibc_list; /* stash on peer's conn list */ - struct list_head ibc_sched_list; /* schedule for attention */ - __u16 ibc_version; /* version of connection */ - __u64 ibc_incarnation; /* which instance of the peer */ - atomic_t ibc_refcount; /* # users */ - int ibc_state; /* what's happening */ - int ibc_nsends_posted; /* # uncompleted sends */ - int ibc_noops_posted; /* # uncompleted NOOPs */ - int ibc_credits; /* # credits I have */ - int ibc_outstanding_credits; /* # credits to return */ - int ibc_reserved_credits;/* # ACK/DONE msg credits */ - int ibc_comms_error; /* set on comms error */ - int ibc_nrx:16; /* receive buffers owned */ - int ibc_scheduled:1; /* scheduled for attention */ - int ibc_ready:1; /* CQ callback fired */ - unsigned long ibc_last_send; /* time of last send */ - struct list_head ibc_early_rxs; /* rxs completed before ESTABLISHED */ - struct list_head ibc_tx_queue; /* sends that need a credit */ - struct list_head ibc_tx_queue_nocred;/* sends that don't need a credit */ - struct list_head ibc_tx_queue_rsrvd; /* sends that need to reserve an ACK/DONE msg */ - struct list_head ibc_active_txs; /* active tx awaiting completion */ - spinlock_t ibc_lock; /* serialise */ - kib_rx_t *ibc_rxs; /* the rx descs */ - kib_pages_t *ibc_rx_pages; /* premapped rx msg pages */ - - struct rdma_cm_id *ibc_cmid; /* CM id */ - struct ib_cq *ibc_cq; /* completion queue */ - - kib_connvars_t *ibc_connvars; /* in-progress connection state */ + struct kib_peer *ibc_peer; /* owning peer */ + cfs_list_t ibc_list; /* stash on peer's conn list */ + cfs_list_t ibc_sched_list; /* schedule for attention */ + __u16 ibc_version; /* version of connection */ + __u64 ibc_incarnation; /* which instance of the peer */ + cfs_atomic_t ibc_refcount; /* # users */ + int ibc_state; /* what's happening */ + int ibc_nsends_posted; /* # uncompleted sends */ + int ibc_noops_posted; /* # uncompleted NOOPs */ + int ibc_credits; /* # credits I have */ + int ibc_outstanding_credits; /* # credits to return */ + int ibc_reserved_credits;/* # ACK/DONE msg credits */ + int ibc_retry_noop; /* need to retry returning credits */ + int ibc_comms_error; /* set on comms error */ + int ibc_nrx:16; /* receive buffers owned */ + int ibc_scheduled:1; /* scheduled for attention */ + int ibc_ready:1; /* CQ callback fired */ + unsigned long ibc_last_send; /* time of last send */ + cfs_list_t ibc_early_rxs; /* rxs completed before ESTABLISHED */ + cfs_list_t ibc_tx_queue; /* sends that need a credit */ + cfs_list_t ibc_tx_queue_nocred;/* sends that don't need a credit */ + cfs_list_t ibc_tx_queue_rsrvd; /* sends that need to reserve an ACK/DONE msg */ + cfs_list_t ibc_active_txs; /* active tx awaiting completion */ + cfs_spinlock_t ibc_lock; /* serialise */ + kib_rx_t *ibc_rxs; /* the rx descs */ + kib_pages_t *ibc_rx_pages; /* premapped rx msg pages */ + + struct rdma_cm_id *ibc_cmid; /* CM id */ + struct ib_cq *ibc_cq; /* completion queue */ + + kib_connvars_t *ibc_connvars; /* in-progress connection state */ } kib_conn_t; #define IBLND_CONN_INIT 0 /* being intialised */ @@ -491,18 +570,18 @@ typedef struct kib_conn typedef struct kib_peer { - struct list_head ibp_list; /* stash on global peer list */ - lnet_nid_t ibp_nid; /* who's on the other end(s) */ - lnet_ni_t *ibp_ni; /* LNet interface */ - atomic_t ibp_refcount; /* # users */ - struct list_head ibp_conns; /* all active connections */ - struct list_head ibp_tx_queue; /* msgs waiting for a conn */ - __u16 ibp_version; /* version of peer */ - __u64 ibp_incarnation; /* incarnation of peer */ - int ibp_connecting; /* current active connection attempts */ - int ibp_accepting; /* current passive connection attempts */ - int ibp_error; /* errno on closing this peer */ - cfs_time_t ibp_last_alive; /* when (in jiffies) I was last alive */ + cfs_list_t ibp_list; /* stash on global peer list */ + lnet_nid_t ibp_nid; /* who's on the other end(s) */ + lnet_ni_t *ibp_ni; /* LNet interface */ + cfs_atomic_t ibp_refcount; /* # users */ + cfs_list_t ibp_conns; /* all active connections */ + cfs_list_t ibp_tx_queue; /* msgs waiting for a conn */ + __u16 ibp_version; /* version of peer */ + __u64 ibp_incarnation; /* incarnation of peer */ + int ibp_connecting; /* current active connection attempts */ + int ibp_accepting; /* current passive connection attempts */ + int ibp_error; /* errno on closing this peer */ + cfs_time_t ibp_last_alive; /* when (in jiffies) I was last alive */ } kib_peer_t; extern kib_data_t kiblnd_data; @@ -510,50 +589,51 @@ extern kib_data_t kiblnd_data; #define kiblnd_conn_addref(conn) \ do { \ CDEBUG(D_NET, "conn[%p] (%d)++\n", \ - (conn), atomic_read(&(conn)->ibc_refcount)); \ - LASSERT(atomic_read(&(conn)->ibc_refcount) > 0); \ - atomic_inc(&(conn)->ibc_refcount); \ + (conn), cfs_atomic_read(&(conn)->ibc_refcount)); \ + LASSERT(cfs_atomic_read(&(conn)->ibc_refcount) > 0); \ + cfs_atomic_inc(&(conn)->ibc_refcount); \ } while (0) -#define kiblnd_conn_decref(conn) \ -do { \ - unsigned long flags; \ - \ - CDEBUG(D_NET, "conn[%p] (%d)--\n", \ - (conn), atomic_read(&(conn)->ibc_refcount)); \ - LASSERT(atomic_read(&(conn)->ibc_refcount) > 0); \ - if (atomic_dec_and_test(&(conn)->ibc_refcount)) { \ - spin_lock_irqsave(&kiblnd_data.kib_connd_lock, flags); \ - list_add_tail(&(conn)->ibc_list, \ - &kiblnd_data.kib_connd_zombies); \ - wake_up(&kiblnd_data.kib_connd_waitq); \ - spin_unlock_irqrestore(&kiblnd_data.kib_connd_lock, flags); \ - } \ +#define kiblnd_conn_decref(conn) \ +do { \ + unsigned long flags; \ + \ + CDEBUG(D_NET, "conn[%p] (%d)--\n", \ + (conn), cfs_atomic_read(&(conn)->ibc_refcount)); \ + LASSERT(cfs_atomic_read(&(conn)->ibc_refcount) > 0); \ + if (cfs_atomic_dec_and_test(&(conn)->ibc_refcount)) { \ + cfs_spin_lock_irqsave(&kiblnd_data.kib_connd_lock, flags); \ + cfs_list_add_tail(&(conn)->ibc_list, \ + &kiblnd_data.kib_connd_zombies); \ + cfs_waitq_signal(&kiblnd_data.kib_connd_waitq); \ + cfs_spin_unlock_irqrestore(&kiblnd_data.kib_connd_lock, flags);\ + } \ } while (0) #define kiblnd_peer_addref(peer) \ do { \ CDEBUG(D_NET, "peer[%p] -> %s (%d)++\n", \ (peer), libcfs_nid2str((peer)->ibp_nid), \ - atomic_read (&(peer)->ibp_refcount)); \ - LASSERT(atomic_read(&(peer)->ibp_refcount) > 0); \ - atomic_inc(&(peer)->ibp_refcount); \ + cfs_atomic_read (&(peer)->ibp_refcount)); \ + LASSERT(cfs_atomic_read(&(peer)->ibp_refcount) > 0); \ + cfs_atomic_inc(&(peer)->ibp_refcount); \ } while (0) #define kiblnd_peer_decref(peer) \ do { \ CDEBUG(D_NET, "peer[%p] -> %s (%d)--\n", \ (peer), libcfs_nid2str((peer)->ibp_nid), \ - atomic_read (&(peer)->ibp_refcount)); \ - LASSERT(atomic_read(&(peer)->ibp_refcount) > 0); \ - if (atomic_dec_and_test(&(peer)->ibp_refcount)) \ + cfs_atomic_read (&(peer)->ibp_refcount)); \ + LASSERT(cfs_atomic_read(&(peer)->ibp_refcount) > 0); \ + if (cfs_atomic_dec_and_test(&(peer)->ibp_refcount)) \ kiblnd_destroy_peer(peer); \ } while (0) -static inline struct list_head * +static inline cfs_list_t * kiblnd_nid2peerlist (lnet_nid_t nid) { - unsigned int hash = ((unsigned int)nid) % kiblnd_data.kib_peer_hash_size; + unsigned int hash = + ((unsigned int)nid) % kiblnd_data.kib_peer_hash_size; return (&kiblnd_data.kib_peers [hash]); } @@ -562,24 +642,24 @@ static inline int kiblnd_peer_active (kib_peer_t *peer) { /* Am I in the peer hash table? */ - return (!list_empty(&peer->ibp_list)); + return (!cfs_list_empty(&peer->ibp_list)); } static inline kib_conn_t * kiblnd_get_conn_locked (kib_peer_t *peer) { - LASSERT (!list_empty(&peer->ibp_conns)); + LASSERT (!cfs_list_empty(&peer->ibp_conns)); /* just return the first connection */ - return list_entry(peer->ibp_conns.next, kib_conn_t, ibc_list); + return cfs_list_entry(peer->ibp_conns.next, kib_conn_t, ibc_list); } static inline int kiblnd_send_keepalive(kib_conn_t *conn) { return (*kiblnd_tunables.kib_keepalive > 0) && - time_after(jiffies, conn->ibc_last_send + - *kiblnd_tunables.kib_keepalive*HZ); + cfs_time_after(jiffies, conn->ibc_last_send + + *kiblnd_tunables.kib_keepalive*CFS_HZ); } static inline int @@ -592,14 +672,15 @@ kiblnd_send_noop(kib_conn_t *conn) !kiblnd_send_keepalive(conn)) return 0; /* No need to send NOOP */ - if (!list_empty(&conn->ibc_tx_queue_nocred)) + if (!cfs_list_empty(&conn->ibc_tx_queue_nocred)) return 0; /* NOOP can be piggybacked */ if (!IBLND_OOB_CAPABLE(conn->ibc_version)) - return list_empty(&conn->ibc_tx_queue); /* can't piggyback? */ + /* can't piggyback? */ + return cfs_list_empty(&conn->ibc_tx_queue); /* No tx to piggyback NOOP onto or no credit to send a tx */ - return (list_empty(&conn->ibc_tx_queue) || conn->ibc_credits == 0); + return (cfs_list_empty(&conn->ibc_tx_queue) || conn->ibc_credits == 0); } static inline void @@ -610,7 +691,7 @@ kiblnd_abort_receives(kib_conn_t *conn) } static inline const char * -kiblnd_queue2str (kib_conn_t *conn, struct list_head *q) +kiblnd_queue2str (kib_conn_t *conn, cfs_list_t *q) { if (q == &conn->ibc_tx_queue) return "tx_queue"; @@ -662,7 +743,7 @@ static inline void kiblnd_set_conn_state (kib_conn_t *conn, int state) { conn->ibc_state = state; - mb(); + cfs_mb(); } static inline void @@ -848,19 +929,24 @@ struct ib_mr *kiblnd_find_dma_mr(kib_net_t *net, __u64 addr, __u64 size); void kiblnd_map_rx_descs(kib_conn_t *conn); void kiblnd_unmap_rx_descs(kib_conn_t *conn); -void kiblnd_map_tx_descs (lnet_ni_t *ni); -void kiblnd_unmap_tx_descs(lnet_ni_t *ni); int kiblnd_map_tx(lnet_ni_t *ni, kib_tx_t *tx, kib_rdma_desc_t *rd, int nfrags); void kiblnd_unmap_tx(lnet_ni_t *ni, kib_tx_t *tx); -kib_phys_mr_t *kiblnd_phys_mr_map(kib_net_t *net, kib_rdma_desc_t *rd, - struct ib_phys_buf *ipb, __u64 *iova); -void kiblnd_phys_mr_unmap(kib_net_t *net, kib_phys_mr_t *pmr); +void kiblnd_pool_free_node(kib_pool_t *pool, cfs_list_t *node); +cfs_list_t *kiblnd_pool_alloc_node(kib_poolset_t *ps); + +int kiblnd_fmr_pool_map(kib_fmr_poolset_t *fps, __u64 *pages, + int npages, __u64 iov, kib_fmr_t *fmr); +void kiblnd_fmr_pool_unmap(kib_fmr_t *fmr, int status); + +int kiblnd_pmr_pool_map(kib_pmr_poolset_t *pps, kib_rdma_desc_t *rd, + __u64 *iova, kib_phys_mr_t **pp_pmr); +void kiblnd_pmr_pool_unmap(kib_phys_mr_t *pmr); int kiblnd_startup (lnet_ni_t *ni); void kiblnd_shutdown (lnet_ni_t *ni); int kiblnd_ctl (lnet_ni_t *ni, unsigned int cmd, void *arg); -void kiblnd_query (struct lnet_ni *ni, lnet_nid_t nid, time_t *when); +void kiblnd_query (struct lnet_ni *ni, lnet_nid_t nid, cfs_time_t *when); int kiblnd_tunables_init(void); void kiblnd_tunables_fini(void); @@ -901,7 +987,8 @@ void kiblnd_launch_tx (lnet_ni_t *ni, kib_tx_t *tx, lnet_nid_t nid); void kiblnd_queue_tx_locked (kib_tx_t *tx, kib_conn_t *conn); void kiblnd_queue_tx (kib_tx_t *tx, kib_conn_t *conn); void kiblnd_init_tx_msg (lnet_ni_t *ni, kib_tx_t *tx, int type, int body_nob); -void kiblnd_txlist_done (lnet_ni_t *ni, struct list_head *txlist, int status); +void kiblnd_txlist_done (lnet_ni_t *ni, cfs_list_t *txlist, + int status); void kiblnd_check_sends (kib_conn_t *conn); void kiblnd_qp_event(struct ib_event *event, void *arg);