From: Vipul Pandya <vipul@chelsio.com>
To: netdev@vger.kernel.org, linux-rdma@vger.kernel.org,
linux-scsi@vger.kernel.org
Cc: davem@davemloft.net, roland@purestorage.com,
JBottomley@parallels.com, dm@chelsio.com,
swise@opengridcomputing.com, leedom@chelsio.com,
naresh@chelsio.com, divy@chelsio.com, santosh@chelsio.com,
arvindb@chelsio.com, abhishek@chelsio.com,
Vipul Pandya <vipul@chelsio.com>
Subject: [PATCH net-next 13/22] RDMA/cxgb4: Turn off db coalescing when RDMA QPs are in use.
Date: Tue, 12 Mar 2013 17:16:25 +0530 [thread overview]
Message-ID: <1363088794-31453-14-git-send-email-vipul@chelsio.com> (raw)
In-Reply-To: <1363088794-31453-1-git-send-email-vipul@chelsio.com>
Signed-off-by: Vipul Pandya <vipul@chelsio.com>
---
drivers/infiniband/hw/cxgb4/qp.c | 20 ++++++++++++++++----
drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c | 19 +++++++++++++++++++
drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h | 3 +++
drivers/net/ethernet/chelsio/cxgb4/t4_regs.h | 4 ++++
4 files changed, 42 insertions(+), 4 deletions(-)
diff --git a/drivers/infiniband/hw/cxgb4/qp.c b/drivers/infiniband/hw/cxgb4/qp.c
index c460244..da4869f 100644
--- a/drivers/infiniband/hw/cxgb4/qp.c
+++ b/drivers/infiniband/hw/cxgb4/qp.c
@@ -42,10 +42,17 @@ static int ocqp_support = 1;
module_param(ocqp_support, int, 0644);
MODULE_PARM_DESC(ocqp_support, "Support on-chip SQs (default=1)");
-int db_fc_threshold = 2000;
+int db_fc_threshold = 1000;
module_param(db_fc_threshold, int, 0644);
-MODULE_PARM_DESC(db_fc_threshold, "QP count/threshold that triggers automatic "
- "db flow control mode (default = 2000)");
+MODULE_PARM_DESC(db_fc_threshold,
+ "QP count/threshold that triggers"
+ " automatic db flow control mode (default = 1000)");
+
+int db_coalescing_threshold;
+module_param(db_coalescing_threshold, int, 0644);
+MODULE_PARM_DESC(db_coalescing_threshold,
+ "QP count/threshold that triggers"
+ " disabling db coalescing (default = 0)");
static void set_state(struct c4iw_qp *qhp, enum c4iw_qp_state state)
{
@@ -1448,6 +1455,8 @@ int c4iw_destroy_qp(struct ib_qp *ib_qp)
rhp->db_state = NORMAL;
idr_for_each(&rhp->qpidr, enable_qp_db, NULL);
}
+ if (rhp->qpcnt <= db_coalescing_threshold)
+ cxgb4_enable_db_coalescing(rhp->rdev.lldi.ports[0]);
spin_unlock_irq(&rhp->lock);
atomic_dec(&qhp->refcnt);
wait_event(qhp->wait, !atomic_read(&qhp->refcnt));
@@ -1559,11 +1568,14 @@ struct ib_qp *c4iw_create_qp(struct ib_pd *pd, struct ib_qp_init_attr *attrs,
spin_lock_irq(&rhp->lock);
if (rhp->db_state != NORMAL)
t4_disable_wq_db(&qhp->wq);
- if (++rhp->qpcnt > db_fc_threshold && rhp->db_state == NORMAL) {
+ rhp->qpcnt++;
+ if (rhp->qpcnt > db_fc_threshold && rhp->db_state == NORMAL) {
rhp->rdev.stats.db_state_transitions++;
rhp->db_state = FLOW_CONTROL;
idr_for_each(&rhp->qpidr, disable_qp_db, NULL);
}
+ if (rhp->qpcnt > db_coalescing_threshold)
+ cxgb4_disable_db_coalescing(rhp->rdev.lldi.ports[0]);
ret = insert_handle_nolock(rhp, &rhp->qpidr, qhp, qhp->wq.sq.qid);
spin_unlock_irq(&rhp->lock);
if (ret)
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
index a59bb23..e76cf03 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
@@ -3397,6 +3397,25 @@ out:
}
EXPORT_SYMBOL(cxgb4_sync_txq_pidx);
+void cxgb4_disable_db_coalescing(struct net_device *dev)
+{
+ struct adapter *adap;
+
+ adap = netdev2adap(dev);
+ t4_set_reg_field(adap, A_SGE_DOORBELL_CONTROL, F_NOCOALESCE,
+ F_NOCOALESCE);
+}
+EXPORT_SYMBOL(cxgb4_disable_db_coalescing);
+
+void cxgb4_enable_db_coalescing(struct net_device *dev)
+{
+ struct adapter *adap;
+
+ adap = netdev2adap(dev);
+ t4_set_reg_field(adap, A_SGE_DOORBELL_CONTROL, F_NOCOALESCE, 0);
+}
+EXPORT_SYMBOL(cxgb4_enable_db_coalescing);
+
static struct pci_driver cxgb4_driver;
static void check_neigh_update(struct neighbour *neigh)
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h
index e2bbc7f..4faf4d0 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h
@@ -269,4 +269,7 @@ struct sk_buff *cxgb4_pktgl_to_skb(const struct pkt_gl *gl,
unsigned int skb_len, unsigned int pull_len);
int cxgb4_sync_txq_pidx(struct net_device *dev, u16 qid, u16 pidx, u16 size);
int cxgb4_flush_eq_cache(struct net_device *dev);
+void cxgb4_disable_db_coalescing(struct net_device *dev);
+void cxgb4_enable_db_coalescing(struct net_device *dev);
+
#endif /* !__CXGB4_OFLD_H */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_regs.h b/drivers/net/ethernet/chelsio/cxgb4/t4_regs.h
index 22cbcb3..ef146c0 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/t4_regs.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4_regs.h
@@ -241,6 +241,10 @@
#define SGE_DOORBELL_CONTROL 0x10a8
#define ENABLE_DROP (1 << 13)
+#define S_NOCOALESCE 26
+#define V_NOCOALESCE(x) ((x) << S_NOCOALESCE)
+#define F_NOCOALESCE V_NOCOALESCE(1U)
+
#define SGE_TIMER_VALUE_0_AND_1 0x10b8
#define TIMERVALUE0_MASK 0xffff0000U
#define TIMERVALUE0_SHIFT 16
--
1.7.1
next prev parent reply other threads:[~2013-03-12 11:53 UTC|newest]
Thread overview: 28+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-03-12 11:46 [PATCH net-next 00/22] Add support for Chelsio T5 adapter Vipul Pandya
2013-03-12 11:46 ` [PATCH net-next 01/22] cxgb4: Add register definations for T5 Vipul Pandya
2013-03-12 11:46 ` [PATCH net-next 02/22] cxgb4: Add macros, structures and inline functions " Vipul Pandya
2013-03-12 11:46 ` [PATCH net-next 03/22] cxgb4: Initialize T5 Vipul Pandya
2013-03-12 11:46 ` [PATCH net-next 06/22] cxgb4: Enable doorbell drop recovery only for T4 adapter Vipul Pandya
[not found] ` <1363088794-31453-1-git-send-email-vipul-ut6Up61K2wZBDgjK7y7TUQ@public.gmane.org>
2013-03-12 11:46 ` [PATCH net-next 04/22] cxgb4: Dump T5 registers Vipul Pandya
2013-03-12 11:46 ` [PATCH net-next 05/22] cxgb4: Add T5 write combining support Vipul Pandya
2013-03-12 12:19 ` David Miller
[not found] ` <20130312.081927.1036246728528667686.davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>
2013-03-12 14:42 ` Steve Wise
[not found] ` <513F3EBD.5020504-7bPotxP6k4+P2YhJcF5u+vpXobYPEAuW@public.gmane.org>
2013-03-13 14:36 ` Vipul Pandya
2013-03-13 15:43 ` David Laight
2013-03-13 22:54 ` Casey Leedom
2013-03-12 11:46 ` [PATCH net-next 07/22] cxgb4: Add T5 debugfs support Vipul Pandya
2013-03-12 11:46 ` [PATCH net-next 08/22] cxgb4: Add T5 PCI ids Vipul Pandya
2013-03-12 11:46 ` [PATCH net-next 09/22] cxgb4: Update driver version and description Vipul Pandya
2013-03-12 11:46 ` [PATCH net-next 10/22] cxgb4: Disable SR-IOV support for PF4-7 for T5 Vipul Pandya
2013-03-12 11:46 ` [PATCH net-next 11/22] cxgb4vf: Add support for Chelsio T5 adapter Vipul Pandya
2013-03-12 11:46 ` [PATCH net-next 12/22] RDMA/cxgb4: Add Support " Vipul Pandya
2013-03-12 11:46 ` Vipul Pandya [this message]
2013-03-12 11:46 ` [PATCH net-next 14/22] RDMA/cxgb4: Add module_params to enable DB FC & Coalescing on T5 Vipul Pandya
2013-03-12 11:46 ` [PATCH net-next 15/22] RDMA/cxgb4: Use DSGLs for fastreg and adapter memory writes for T5 Vipul Pandya
2013-03-12 11:46 ` [PATCH net-next 16/22] RDMA/cxgb4: Map pbl buffers for dma if using DSGL Vipul Pandya
2013-03-12 11:46 ` [PATCH net-next 17/22] RDMA/cxgb4: Bump tcam_full stat and WR reply timeout Vipul Pandya
2013-03-12 11:46 ` [PATCH net-next 18/22] RDMA/cxgb4: Fix onchip queue support for T5 Vipul Pandya
2013-03-12 11:46 ` [PATCH net-next 19/22] csiostor: Segregate T4 adapter operations Vipul Pandya
2013-03-12 11:46 ` [PATCH net-next 20/22] csiostor: Add T5 " Vipul Pandya
2013-03-12 11:46 ` [PATCH net-next 21/22] csiostor: Header file modifications for chip support and bug fixes Vipul Pandya
2013-03-12 11:46 ` [PATCH net-next 22/22] csiostor: Cleanup chip specific operations Vipul Pandya
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1363088794-31453-14-git-send-email-vipul@chelsio.com \
--to=vipul@chelsio.com \
--cc=JBottomley@parallels.com \
--cc=abhishek@chelsio.com \
--cc=arvindb@chelsio.com \
--cc=davem@davemloft.net \
--cc=divy@chelsio.com \
--cc=dm@chelsio.com \
--cc=leedom@chelsio.com \
--cc=linux-rdma@vger.kernel.org \
--cc=linux-scsi@vger.kernel.org \
--cc=naresh@chelsio.com \
--cc=netdev@vger.kernel.org \
--cc=roland@purestorage.com \
--cc=santosh@chelsio.com \
--cc=swise@opengridcomputing.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox