From ec56dc0b7f6c3fec20bbc2e98ff1a06edf2fc9b9 Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Mon, 28 May 2007 14:37:27 +0300 Subject: IPoIB/cm: Fix performance regression on Mellanox commit 518b1646 ("IPoIB/cm: Fix SRQ WR leak") introduced a severe performance regression on Mellanox cards, because keeping a QP in the error state for extended periods of time moves hardware to the slow path (until the QP is destroyed). For example, MPI latency goes from ~3 usecs to ~7 usecs. Fix this by posting a send WR on one of the QPs that are being flushed, instead of using a separate drain QP that is kept in the error state. This fixes bug , reported and bisected by Scott Weitzenkamp at Cisco and debugged by Sasha Mikheev at Voltaire. Signed-off-by: Michael S. Tsirkin Signed-off-by: Roland Dreier --- drivers/infiniband/ulp/ipoib/ipoib.h | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'drivers/infiniband/ulp/ipoib/ipoib.h') diff --git a/drivers/infiniband/ulp/ipoib/ipoib.h b/drivers/infiniband/ulp/ipoib/ipoib.h index 158759e28a5..285c143115c 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib.h +++ b/drivers/infiniband/ulp/ipoib/ipoib.h @@ -156,7 +156,7 @@ struct ipoib_cm_data { * - and then invoke a Destroy QP or Reset QP. * * We use the second option and wait for a completion on the - * rx_drain_qp before destroying QPs attached to our SRQ. + * same CQ before destroying QPs attached to our SRQ. */ enum ipoib_cm_state { @@ -199,7 +199,6 @@ struct ipoib_cm_dev_priv { struct ib_srq *srq; struct ipoib_cm_rx_buf *srq_ring; struct ib_cm_id *id; - struct ib_qp *rx_drain_qp; /* generates WR described in 10.3.1 */ struct list_head passive_ids; /* state: LIVE */ struct list_head rx_error_list; /* state: ERROR */ struct list_head rx_flush_list; /* state: FLUSH, drain not started */ -- cgit v1.2.3