summaryrefslogtreecommitdiffstats
AgeCommit message (Expand)AuthorFilesLines
2010-09-08rds: more FMRs are fasterChris Mason1-1/+1
2010-09-08rds: recycle FMRs through lockless listsChris Mason2-42/+282
2010-09-08rds: fix rds_send_xmit() serializationZach Brown5-58/+52
2010-09-08rds: block ints when acquiring c_lock in rds_conn_message_info()Zach Brown1-2/+3
2010-09-08rds: remove unused rds_send_acked_before()Zach Brown2-30/+0
2010-09-08RDS: use friendly gfp masks for prefillChris Mason1-9/+18
2010-09-08RDS/IB: Add caching of frags and incsChris Mason3-52/+297
2010-09-08RDS/IB: Remove ib_recv_unmap_page()Andy Grover1-20/+2
2010-09-08RDS: Assume recv->r_frag is always NULL in refill_one()Andy Grover1-13/+16
2010-09-08RDS: Use page_remainder_alloc() for recv bufsAndy Grover3-73/+29
2010-09-08RDS/IB: disconnect when IB devices are removedZach Brown1-1/+24
2010-09-08RDS: introduce rds_conn_connect_if_down()Zach Brown3-14/+15
2010-09-08RDS/IB: add refcount tracking to struct rds_ib_deviceZach Brown4-55/+125
2010-09-08RDS/IB: get the xmit max_sge from the RDS IB device on the connectionZach Brown1-7/+5
2010-09-08RDS/IB: rds_ib_cm_handle_connect() forgot to unlock c_cm_lockZach Brown1-9/+6
2010-09-08rds: Fix reference counting on the for xmit_atomic and xmit_rdmaChris Mason2-9/+11
2010-09-08rds: use RCU to protect the connection hashChris Mason1-22/+22
2010-09-08RDS: use locking on the connection hash listChris Mason1-0/+3
2010-09-08rds: Fix RDMA message reference countingChris Mason2-8/+14
2010-09-08rds: don't let RDS shutdown a connection while senders are presentChris Mason5-16/+26
2010-09-08rds: Use RCU for the bind lookup searchesChris Mason4-45/+57
2010-09-08RDS/IB: add _to_node() macros for numa and use {k,v}malloc_node()Andy Grover4-5/+14
2010-09-08RDS/IB: Remove unused variable in ib_remove_addr()Andy Grover1-1/+1
2010-09-08rds: rcu-ize rds_ib_get_device()Chris Mason2-8/+17
2010-09-08rds: per-rm flush_wait waitqChris Mason2-4/+5
2010-09-08rds: switch to rwlock on bind_lockChris Mason1-7/+7
2010-09-08RDS: Update comments in rds_send_xmit()Andy Grover1-3/+2
2010-09-08RDS: Use a generation counter to avoid rds_send_xmit loopChris Mason3-4/+9
2010-09-08RDS: Get pong working againAndy Grover1-1/+4
2010-09-08RDS: Do wait_event_interruptible instead of wait_eventAndy Grover1-2/+2
2010-09-08RDS: Remove send_quota from send_xmit()Andy Grover1-15/+15
2010-09-08RDS: Move atomic stats from general to ib-specific areaAndy Grover5-6/+6
2010-09-08RDS: rds_message_unmapped() doesn't need to check if queue activeAndy Grover1-2/+1
2010-09-08RDS: Fix locking in send on m_rs_lockAndy Grover1-2/+3
2010-09-08RDS: Use NOWAIT in message_map_pages()Andy Grover1-1/+1
2010-09-08RDS: Bypass workqueue when queueing cong updatesAndy Grover1-1/+1
2010-09-08RDS: Call rds_send_xmit() directly from sendmsg()Andy Grover1-1/+1
2010-09-08RDS: rds_send_xmit() locking/irq fixesAndy Grover1-9/+12
2010-09-08RDS: Change send lock from a mutex to a spinlockAndy Grover4-29/+18
2010-09-08RDS: Refill recv ring directly from taskletAndy Grover3-17/+10
2010-09-08RDS: Stop supporting old cong map sending methodAndy Grover8-106/+13
2010-09-08RDS/IB: Do not wait for send ring to be empty on conn shutdownAndy Grover1-1/+4
2010-09-08RDS: Perform unmapping ops in stagesAndy Grover5-126/+142
2010-09-08RDS: Make sure cmsgs aren't used in improper waysAndy Grover1-0/+9
2010-09-08RDS: Add flag for silent ops. Do atomic op before RDMAAndy Grover4-24/+36
2010-09-08RDS: Move some variables around for consistencyAndy Grover2-4/+7
2010-09-08RDS: queue failure notifications for dropped atomic opsAndy Grover1-2/+10
2010-09-08RDS: Add a warning if trying to allocate 0 sgsAndy Grover1-0/+1
2010-09-08RDS: Do not set op_active in r_m_copy_from_user().Andy Grover2-7/+8
2010-09-08RDS: Rewrite rds_send_xmitAndy Grover2-63/+73