From: Bart Van Assche <bvanassche-HInyCGIudOg@public.gmane.org>
To: Roland Dreier <roland-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
Cc: David Dillow <dillowda-1Heg1YXhbW8@public.gmane.org>,
Vu Pham <vuhuong-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>,
Sebastian Riemer
<sebastian.riemer-EIkl63zCoXaH+58JC4qpiA@public.gmane.org>,
linux-rdma <linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org>
Subject: [PATCH 12/14] IB/srp: Make HCA completion vector configurable
Date: Wed, 12 Jun 2013 15:35:00 +0200 [thread overview]
Message-ID: <51B87904.1070803@acm.org> (raw)
In-Reply-To: <51B87501.4070005-HInyCGIudOg@public.gmane.org>
Several InfiniBand HCA's allow to configure the completion vector
per queue pair. This allows to spread the workload created by IB
completion interrupts over multiple MSI-X vectors and hence over
multiple CPU cores. In other words, configuring the completion
vector properly not only allows to reduce latency on an initiator
connected to multiple SRP targets but also allows to improve
throughput.
Signed-off-by: Bart Van Assche <bvanassche-HInyCGIudOg@public.gmane.org>
Cc: Roland Dreier <roland-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
Cc: David Dillow <dillowda-1Heg1YXhbW8@public.gmane.org>
Cc: Vu Pham <vu-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
Cc: Sebastian Riemer <sebastian.riemer-EIkl63zCoXaH+58JC4qpiA@public.gmane.org>
---
drivers/infiniband/ulp/srp/ib_srp.c | 26 ++++++++++++++++++++++++--
drivers/infiniband/ulp/srp/ib_srp.h | 1 +
2 files changed, 25 insertions(+), 2 deletions(-)
diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c
index dbedac9..49d8464 100644
--- a/drivers/infiniband/ulp/srp/ib_srp.c
+++ b/drivers/infiniband/ulp/srp/ib_srp.c
@@ -294,14 +294,16 @@ static int srp_create_target_ib(struct srp_target_port *target)
return -ENOMEM;
recv_cq = ib_create_cq(target->srp_host->srp_dev->dev,
- srp_recv_completion, NULL, target, SRP_RQ_SIZE, 0);
+ srp_recv_completion, NULL, target, SRP_RQ_SIZE,
+ target->comp_vector);
if (IS_ERR(recv_cq)) {
ret = PTR_ERR(recv_cq);
goto err;
}
send_cq = ib_create_cq(target->srp_host->srp_dev->dev,
- srp_send_completion, NULL, target, SRP_SQ_SIZE, 0);
+ srp_send_completion, NULL, target, SRP_SQ_SIZE,
+ target->comp_vector);
if (IS_ERR(send_cq)) {
ret = PTR_ERR(send_cq);
goto err_recv_cq;
@@ -2006,6 +2008,14 @@ static ssize_t show_local_ib_device(struct device *dev,
return sprintf(buf, "%s\n", target->srp_host->srp_dev->dev->name);
}
+static ssize_t show_comp_vector(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ struct srp_target_port *target = host_to_target(class_to_shost(dev));
+
+ return sprintf(buf, "%d\n", target->comp_vector);
+}
+
static ssize_t show_cmd_sg_entries(struct device *dev,
struct device_attribute *attr, char *buf)
{
@@ -2032,6 +2042,7 @@ static DEVICE_ATTR(req_lim, S_IRUGO, show_req_lim, NULL);
static DEVICE_ATTR(zero_req_lim, S_IRUGO, show_zero_req_lim, NULL);
static DEVICE_ATTR(local_ib_port, S_IRUGO, show_local_ib_port, NULL);
static DEVICE_ATTR(local_ib_device, S_IRUGO, show_local_ib_device, NULL);
+static DEVICE_ATTR(comp_vector, S_IRUGO, show_comp_vector, NULL);
static DEVICE_ATTR(cmd_sg_entries, S_IRUGO, show_cmd_sg_entries, NULL);
static DEVICE_ATTR(allow_ext_sg, S_IRUGO, show_allow_ext_sg, NULL);
@@ -2046,6 +2057,7 @@ static struct device_attribute *srp_host_attrs[] = {
&dev_attr_zero_req_lim,
&dev_attr_local_ib_port,
&dev_attr_local_ib_device,
+ &dev_attr_comp_vector,
&dev_attr_cmd_sg_entries,
&dev_attr_allow_ext_sg,
NULL
@@ -2140,6 +2152,7 @@ enum {
SRP_OPT_CMD_SG_ENTRIES = 1 << 9,
SRP_OPT_ALLOW_EXT_SG = 1 << 10,
SRP_OPT_SG_TABLESIZE = 1 << 11,
+ SRP_OPT_COMP_VECTOR = 1 << 12,
SRP_OPT_ALL = (SRP_OPT_ID_EXT |
SRP_OPT_IOC_GUID |
SRP_OPT_DGID |
@@ -2160,6 +2173,7 @@ static const match_table_t srp_opt_tokens = {
{ SRP_OPT_CMD_SG_ENTRIES, "cmd_sg_entries=%u" },
{ SRP_OPT_ALLOW_EXT_SG, "allow_ext_sg=%u" },
{ SRP_OPT_SG_TABLESIZE, "sg_tablesize=%u" },
+ { SRP_OPT_COMP_VECTOR, "comp_vector=%u" },
{ SRP_OPT_ERR, NULL }
};
@@ -2315,6 +2329,14 @@ static int srp_parse_options(const char *buf, struct srp_target_port *target)
target->sg_tablesize = token;
break;
+ case SRP_OPT_COMP_VECTOR:
+ if (match_int(args, &token) || token < 0) {
+ pr_warn("bad comp_vector parameter '%s'\n", p);
+ goto out;
+ }
+ target->comp_vector = token;
+ break;
+
default:
pr_warn("unknown parameter or missing value '%s' in target creation request\n",
p);
diff --git a/drivers/infiniband/ulp/srp/ib_srp.h b/drivers/infiniband/ulp/srp/ib_srp.h
index e45d9d0..cbc0b14 100644
--- a/drivers/infiniband/ulp/srp/ib_srp.h
+++ b/drivers/infiniband/ulp/srp/ib_srp.h
@@ -156,6 +156,7 @@ struct srp_target_port {
char target_name[32];
unsigned int scsi_id;
unsigned int sg_tablesize;
+ int comp_vector;
struct ib_sa_path_rec path;
__be16 orig_dgid[8];
--
1.7.10.4
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
next prev parent reply other threads:[~2013-06-12 13:35 UTC|newest]
Thread overview: 63+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-06-12 13:17 [PATCH 0/14] IB SRP initiator patches for kernel 3.11 Bart Van Assche
2013-06-12 13:28 ` [PATCH 07/14] scsi_transport_srp: Add transport layer error handling Bart Van Assche
[not found] ` <51B8777B.5050201-HInyCGIudOg@public.gmane.org>
2013-06-13 19:43 ` Vu Pham
2013-06-14 13:19 ` Bart Van Assche
[not found] ` <51BB1857.7040802-HInyCGIudOg@public.gmane.org>
2013-06-14 17:59 ` Vu Pham
[not found] ` <51BB5A04.3080901-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
2013-06-15 9:52 ` Bart Van Assche
[not found] ` <51BC3945.9030900-HInyCGIudOg@public.gmane.org>
2013-06-17 6:18 ` Hannes Reinecke
2013-06-17 7:04 ` Bart Van Assche
2013-06-17 7:14 ` Hannes Reinecke
2013-06-17 7:29 ` Bart Van Assche
[not found] ` <51BEBAEA.4080202-HInyCGIudOg@public.gmane.org>
2013-06-17 8:10 ` Hannes Reinecke
2013-06-17 10:13 ` Sebastian Riemer
2013-06-18 16:59 ` Vu Pham
[not found] ` <51C09202.2040503-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
2013-06-19 13:00 ` Bart Van Assche
2013-06-23 21:13 ` Mike Christie
[not found] ` <51C764FB.6070207-hcNo3dDEHLuVc3sceRu5cw@public.gmane.org>
2013-06-24 7:37 ` Bart Van Assche
[not found] ` <51B87501.4070005-HInyCGIudOg@public.gmane.org>
2013-06-12 13:20 ` [PATCH 01/14] IB/srp: Fix remove_one crash due to resource exhaustion Bart Van Assche
[not found] ` <51B875A4.7040903-HInyCGIudOg@public.gmane.org>
2013-06-12 13:38 ` Bart Van Assche
[not found] ` <51B879CF.1080802-HInyCGIudOg@public.gmane.org>
2013-06-12 14:24 ` Sebastian Riemer
2013-06-27 21:01 ` David Dillow
[not found] ` <1372366870.32164.30.camel-zHLflQxYYDO4Hhoo1DtQwJ9G+ZOsUmrO@public.gmane.org>
2013-06-27 23:45 ` Roland Dreier
[not found] ` <CAL1RGDWVgAKSL-GNZCkP1FEt9r_y5QWp+74NzDcga6+tcvWpXw-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2013-06-28 7:41 ` Sebastian Riemer
2013-06-12 13:21 ` [PATCH 02/14] IB/srp: Fix race between srp_queuecommand() and srp_claim_req() Bart Van Assche
[not found] ` <51B875EE.3030702-HInyCGIudOg@public.gmane.org>
2013-06-12 14:58 ` Sebastian Riemer
[not found] ` <51B88C7C.4030209-EIkl63zCoXaH+58JC4qpiA@public.gmane.org>
2013-06-12 15:14 ` Bart Van Assche
[not found] ` <51B8903E.3000609-HInyCGIudOg@public.gmane.org>
2013-06-27 21:02 ` David Dillow
[not found] ` <1372366945.32164.32.camel-zHLflQxYYDO4Hhoo1DtQwJ9G+ZOsUmrO@public.gmane.org>
2013-06-28 7:36 ` Bart Van Assche
2013-06-12 13:23 ` [PATCH 03/14] IB/srp: Avoid that srp_reset_host() is skipped after a TL error Bart Van Assche
[not found] ` <51B87638.50102-HInyCGIudOg@public.gmane.org>
2013-06-13 9:30 ` Sebastian Riemer
[not found] ` <51B99120.9000503-EIkl63zCoXaH+58JC4qpiA@public.gmane.org>
2013-06-13 9:57 ` Bart Van Assche
2013-06-27 21:03 ` David Dillow
2013-06-12 13:24 ` [PATCH 04/14] IB/srp: Skip host settle delay Bart Van Assche
[not found] ` <51B87689.8030806-HInyCGIudOg@public.gmane.org>
2013-06-13 9:53 ` Sebastian Riemer
[not found] ` <51B996A1.6080604-EIkl63zCoXaH+58JC4qpiA@public.gmane.org>
2013-06-13 13:06 ` Or Gerlitz
2013-06-27 21:04 ` David Dillow
2013-06-12 13:25 ` [PATCH 05/14] IB/srp: Maintain a single connection per I_T nexus Bart Van Assche
[not found] ` <51B876BF.4070400-HInyCGIudOg@public.gmane.org>
2013-06-13 13:57 ` Sebastian Riemer
[not found] ` <51B9CFC3.8080008-EIkl63zCoXaH+58JC4qpiA@public.gmane.org>
2013-06-13 15:07 ` Bart Van Assche
[not found] ` <51B9E046.3030008-HInyCGIudOg@public.gmane.org>
2013-06-13 15:35 ` Sebastian Riemer
2013-06-13 17:50 ` Vu Pham
[not found] ` <51BA0655.6090707-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
2013-06-13 18:25 ` Bart Van Assche
[not found] ` <51BA0E8F.3030104-HInyCGIudOg@public.gmane.org>
2013-06-13 23:27 ` Vu Pham
[not found] ` <51BA555F.9060807-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
2013-06-14 9:38 ` Sebastian Riemer
[not found] ` <51BAE482.1050304-EIkl63zCoXaH+58JC4qpiA@public.gmane.org>
2013-06-14 17:07 ` Vu Pham
[not found] ` <51BB4DBB.4070800-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
2013-06-17 9:41 ` Sebastian Riemer
2013-06-27 21:10 ` David Dillow
[not found] ` <1372367432.32164.36.camel-zHLflQxYYDO4Hhoo1DtQwJ9G+ZOsUmrO@public.gmane.org>
2013-06-28 7:40 ` Bart Van Assche
2013-06-12 13:26 ` [PATCH 06/14] IB/srp: Keep rport as long as the IB transport layer Bart Van Assche
2013-06-12 13:29 ` [PATCH 08/14] IB/srp: Add srp_terminate_io() Bart Van Assche
2013-06-12 13:30 ` [PATCH 09/14] IB/srp: Use SRP transport layer error recovery Bart Van Assche
2013-06-12 13:31 ` [PATCH 10/14] IB/srp: Start timers if a transport layer error occurs Bart Van Assche
2013-06-12 13:33 ` [PATCH 11/14] IB/srp: Fail SCSI commands silently Bart Van Assche
2013-06-12 13:35 ` Bart Van Assche [this message]
[not found] ` <51B87904.1070803-HInyCGIudOg@public.gmane.org>
2013-06-27 21:24 ` [PATCH 12/14] IB/srp: Make HCA completion vector configurable David Dillow
[not found] ` <1372368256.32164.41.camel-zHLflQxYYDO4Hhoo1DtQwJ9G+ZOsUmrO@public.gmane.org>
2013-06-28 8:18 ` Bart Van Assche
[not found] ` <51CD46F0.60301-HInyCGIudOg@public.gmane.org>
2013-06-28 12:04 ` David Dillow
[not found] ` <1372421041.28740.14.camel-a7a0dvSY7KqLUyTwlgNVppKKF0rrzTr+@public.gmane.org>
2013-06-28 12:29 ` Bart Van Assche
2013-06-12 13:36 ` [PATCH 13/14] IB/srp: Make transport layer retry count configurable Bart Van Assche
[not found] ` <51B8794F.6050003-HInyCGIudOg@public.gmane.org>
2013-06-27 21:22 ` David Dillow
[not found] ` <1372368138.32164.40.camel-zHLflQxYYDO4Hhoo1DtQwJ9G+ZOsUmrO@public.gmane.org>
2013-06-28 8:28 ` Bart Van Assche
[not found] ` <51CD4933.5080709-HInyCGIudOg@public.gmane.org>
2013-06-28 12:07 ` David Dillow
[not found] ` <1372421227.28740.17.camel-a7a0dvSY7KqLUyTwlgNVppKKF0rrzTr+@public.gmane.org>
2013-06-28 12:30 ` Bart Van Assche
2013-06-12 13:37 ` [PATCH 14/14] IB/srp: Bump driver version and release date Bart Van Assche
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=51B87904.1070803@acm.org \
--to=bvanassche-hinycgiudog@public.gmane.org \
--cc=dillowda-1Heg1YXhbW8@public.gmane.org \
--cc=linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=roland-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org \
--cc=sebastian.riemer-EIkl63zCoXaH+58JC4qpiA@public.gmane.org \
--cc=vuhuong-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org \
/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