From: Alexey Lyashkov <alexey.lyashkov@seagate.com>
Date: Mon, 27 Nov 2017 12:44:56 +0000 (+0300)
Subject: LU-9810 lnd: use less CQ entries for each connection
X-Git-Tag: 2.10.57~81
X-Git-Url: https://git.whamcloud.com/?a=commitdiff_plain;h=refs%2Fchanges%2F79%2F28279%2F10;p=fs%2Flustre-release.git

LU-9810 lnd: use less CQ entries for each connection

Currently we have a 2 work requests chains per transfer.
It mean OFED stack will generate only 2 events if transfer will
faild. Reduce number CQ entries to avoid extra resource consumption.

Test-Parameters: trivial
Seagate-bug-id: MRP-4508
Signed-off-by: Alexey Lyashkov <alexey.lyashkov@seagate.com>
Change-Id: I0c06fef9589478f40ef7e1eeacff2aec7013e562
Reviewed-on: https://review.whamcloud.com/28279
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Doug Oucharek <dougso@me.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
---

diff --git a/lnet/klnds/o2iblnd/o2iblnd.h b/lnet/klnds/o2iblnd/o2iblnd.h
index 0c6bddf..6b265c2 100644
--- a/lnet/klnds/o2iblnd/o2iblnd.h
+++ b/lnet/klnds/o2iblnd/o2iblnd.h
@@ -167,7 +167,11 @@ extern kib_tunables_t  kiblnd_tunables;
 #define IBLND_SEND_WRS(c)	\
 	((c->ibc_max_frags + 1) * kiblnd_concurrent_sends(c->ibc_version, \
 							  c->ibc_peer->ibp_ni))
-#define IBLND_CQ_ENTRIES(c)         (IBLND_RECV_WRS(c) + IBLND_SEND_WRS(c))
+
+/* 2 = LNet msg + Transfer chain */
+#define IBLND_CQ_ENTRIES(c)	\
+	(IBLND_RECV_WRS(c) + 2 * kiblnd_concurrent_sends(c->ibc_version, \
+							 c->ibc_peer->ibp_ni))
 
 struct kib_hca_dev;