summaryrefslogtreecommitdiffstats
path: root/net/sunrpc/xprtrdma
AgeCommit message (Expand)AuthorFilesLines
2008-10-10RPC/RDMA: optionally emit useful transport info upon connect/disconnect.Tom Talpey2-1/+22
2008-10-10RPC/RDMA: reformat a debug printk to keep lines together.Tom Talpey1-2/+2
2008-10-10RPC/RDMA: harden connection logic against missing/late rdma_cm upcalls.Tom Talpey2-4/+10
2008-10-10RPC/RDMA: fix connect/reconnect resource leak.Tom Talpey1-5/+4
2008-10-10RPC/RDMA: return a consistent error, when connect fails.Tom Talpey1-1/+1
2008-10-10RPC/RDMA: adhere to protocol for unpadded client trailing write chunks.Tom Talpey3-2/+33
2008-10-10RPC/RDMA: avoid an oops due to disconnect racing with async upcalls.Tom Talpey1-11/+9
2008-10-10RPC/RDMA: maintain the RPC task bytes-sent statistic.Tom Talpey1-0/+1
2008-10-10RPC/RDMA: suppress retransmit on RPC/RDMA clients.Tom Talpey3-4/+15
2008-10-10RPC/RDMA: fix connection IRD/ORD settingTom Tucker1-37/+14
2008-10-10RPC/RDMA: support FRMR client memory registration.Tom Talpey2-6/+167
2008-10-10RPC/RDMA: check selected memory registration mode at runtime.Tom Talpey1-15/+80
2008-10-10RPC/RDMA: add data types and new FRMR memory registration enum.Tom Talpey1-1/+7
2008-10-10RPC/RDMA: refactor the inline memory registration code.Tom Talpey1-158/+207
2008-08-13svcrdma: Fix race between svc_rdma_recvfrom thread and the dto_taskletTom Tucker2-7/+6
2008-07-26dma-mapping: add the device argument to dma_mapping_error()FUJITA Tomonori1-1/+2
2008-07-02svcrdma: Change WR context get/put to use the kmem cacheTom Tucker1-109/+12
2008-07-02svcrdma: Create a kmem cache for the WR contextsTom Tucker1-4/+19
2008-07-02svcrdma: Add flush_scheduled_work to module exit functionTom Tucker1-0/+1
2008-07-02svcrdma: Limit ORD based on client's advertised IRDTom Tucker1-4/+12
2008-07-02svcrdma: Remove unneeded spin locks from __svc_rdma_freeTom Tucker1-4/+0
2008-07-02svcrdma: Add dma map count and WARN_ONTom Tucker3-0/+9
2008-07-02svcrdma: Move the DMA unmap logic to the CQ handlerTom Tucker1-6/+14
2008-07-02svcrdma: Use reply and chunk map for RDMA_READ processingTom Tucker1-45/+38
2008-07-02svcrdma: Use RPC reply map for RDMA_WRITE processingTom Tucker2-88/+80
2008-07-02svcrdma: Add a type for keeping NFS RPC mappingTom Tucker2-0/+45
2008-05-19svcrdma: Verify read-list fits within RPCSVC_MAXPAGESTom Tucker1-0/+2
2008-05-19svcrdma: Change svc_rdma_send_error return type to voidTom Tucker2-6/+5
2008-05-19svcrdma: Copy transport address and arm CQ before calling rdma_acceptTom Tucker1-9/+14
2008-05-19svcrdma: Set rqstp transport address in rdma_read_complete functionTom Tucker1-0/+1
2008-05-19svcrdma: Use ib verbs version of dma_unmapTom Tucker1-4/+4
2008-05-19svcrdma: Cleanup queued, but unprocessed I/O in svc_rdma_freeTom Tucker1-3/+35
2008-05-19svcrdma: Move the QP and cm_id destruction to svc_rdma_freeTom Tucker1-20/+20
2008-05-19svcrdma: Add reference for each SQ/RQ WRTom Tucker1-2/+16
2008-05-19svcrdma: Move destroy to kernel threadTom Tucker1-3/+14
2008-05-19svcrdma: Shrink scope of spinlock on RQ CQTom Tucker1-2/+2
2008-05-19svcrdma: Use standard Linux lists for context cacheTom Tucker1-21/+26
2008-05-19svcrdma: Simplify RDMA_READ deferral buffer managementTom Tucker2-48/+15
2008-05-19svcrdma: Remove unused READ_DONE context flags bitTom Tucker2-2/+0
2008-05-19svcrdma: Return error from rdma_read_xdr so caller knows to free contextTom Tucker1-7/+16
2008-05-19svcrdma: Fix error handling during listening endpoint creationTom Tucker1-11/+13
2008-05-19svcrdma: Free context on post_recv error in send_replyTom Tucker1-1/+2
2008-05-19svcrdma: Free context on ib_post_recv errorTom Tucker1-0/+2
2008-05-19svcrdma: Add put of connection ESTABLISHED reference in rdma_cma_handlerTom Tucker1-0/+1
2008-05-19svcrdma: Fix return value in svc_rdma_sendTom Tucker1-1/+1
2008-05-19svcrdma: Fix race with dto_tasklet in svc_rdma_sendTom Tucker1-18/+22
2008-05-19svcrdma: Simplify receive buffer postingTom Tucker3-35/+11
2008-04-23SVCRDMA: Add check for XPT_CLOSE in svc_rdma_sendTom Tucker1-0/+2
2008-04-16IB/core: Add support for "send with invalidate" work requestsRoland Dreier1-1/+0
2008-03-26SVCRDMA: Check num_sge when setting LAST_CTXT bitTom Tucker1-10/+11