public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
To: linux-kernel@vger.kernel.org, linux-s390@vger.kernel.org
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>,
	Ursula Braun <ursula.braun@de.ibm.com>,
	Martin Schwidefsky <schwidefsky@de.ibm.com>
Subject: [patch 01/21] qdio: speed up multicast traffic on full HiperSocket queue
Date: Wed, 01 Oct 2008 10:33:18 +0200	[thread overview]
Message-ID: <20081001084125.504618165@de.ibm.com> (raw)
In-Reply-To: 20081001083317.782967473@de.ibm.com

[-- Attachment #1: 100-qdio-perf.diff --]
[-- Type: text/plain, Size: 2118 bytes --]

From: Ursula Braun <ursula.braun@de.ibm.com>

If an asynchronous HiperSockets queue runs full, no further packet
can be sent. In this case the next initiative to give transmitted
skbs back to the stack is triggered only by a 10-seconds qdio timer.
This timer has been introduced for low multicast traffic scenarios
to guarantee freeing of skbs in a limited amount of time. For high
HiperSocket multicast traffic scenarios progress checking on the
outbound queue should be enforced by tasklet rescheduling.

Signed-off-by: Ursula Braun <ursula.braun@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---

 drivers/s390/cio/qdio.h      |    8 ++++++++
 drivers/s390/cio/qdio_main.c |    6 ++++++
 2 files changed, 14 insertions(+)

Index: quilt-2.6/drivers/s390/cio/qdio.h
===================================================================
--- quilt-2.6.orig/drivers/s390/cio/qdio.h
+++ quilt-2.6/drivers/s390/cio/qdio.h
@@ -16,6 +16,14 @@
 #define QDIO_BUSY_BIT_GIVE_UP		2000000	/* 2 seconds = eternity */
 #define QDIO_INPUT_THRESHOLD		500	/* 500 microseconds */
 
+/*
+ * if an asynchronous HiperSockets queue runs full, the 10 seconds timer wait
+ * till next initiative to give transmitted skbs back to the stack is too long.
+ * Therefore polling is started in case of multicast queue is filled more
+ * than 50 percent.
+ */
+#define QDIO_IQDIO_POLL_LVL		65	/* HS multicast queue */
+
 enum qdio_irq_states {
 	QDIO_IRQ_STATE_INACTIVE,
 	QDIO_IRQ_STATE_ESTABLISHED,
Index: quilt-2.6/drivers/s390/cio/qdio_main.c
===================================================================
--- quilt-2.6.orig/drivers/s390/cio/qdio_main.c
+++ quilt-2.6/drivers/s390/cio/qdio_main.c
@@ -851,6 +851,12 @@ static void __qdio_outbound_processing(s
 	if (queue_type(q) == QDIO_IQDIO_QFMT && !multicast_outbound(q))
 		return;
 
+	if ((queue_type(q) == QDIO_IQDIO_QFMT) &&
+	    (atomic_read(&q->nr_buf_used)) > QDIO_IQDIO_POLL_LVL) {
+		tasklet_schedule(&q->tasklet);
+		return;
+	}
+
 	if (q->u.out.pci_out_enabled)
 		return;
 

-- 
blue skies,
   Martin.

"Reality continues to ruin my life." - Calvin.


  reply	other threads:[~2008-10-01  9:02 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-10-01  8:33 [patch 00/21] s390 patches for the 2.6.28 merge window Martin Schwidefsky
2008-10-01  8:33 ` Martin Schwidefsky [this message]
2008-10-01  8:33 ` [patch 02/21] cio: move device unregistration to dedicated work queue Martin Schwidefsky
2008-10-01  8:33 ` [patch 03/21] cio: introduce purge function for /proc/cio_ignore Martin Schwidefsky
2008-10-01  8:33 ` [patch 04/21] cio: Update cio_ignore documentation Martin Schwidefsky
2008-10-01  8:33 ` [patch 05/21] cio: Exorcise cio_msg= from documentation Martin Schwidefsky
2008-10-01  8:33 ` [patch 06/21] bus_id -> dev_name conversions Martin Schwidefsky
2008-10-01  8:33 ` [patch 07/21] bus_id -> dev_set_name() changes Martin Schwidefsky
2008-10-01  8:33 ` [patch 08/21] more bus_id -> dev_name conversions Martin Schwidefsky
2008-10-01  8:33 ` [patch 09/21] Use s390_root_dev_* in kvm_virtio Martin Schwidefsky
2008-10-01  8:33 ` [patch 10/21] bus_id ->dev_name() conversions in qdio Martin Schwidefsky
2008-10-01  8:33 ` [patch 11/21] bus_id -> dev_set_name() for css and ccw busses Martin Schwidefsky
2008-10-01  8:33 ` [patch 12/21] cio: inline assembly cleanup Martin Schwidefsky
2008-10-01  8:33 ` [patch 13/21] qdio enhanced SIGA (iqdio) support Martin Schwidefsky
2008-10-01  8:33 ` [patch 14/21] s390: use sys_pause for 31bit pause entry point Martin Schwidefsky
2008-10-01  8:33 ` [patch 15/21] ptrace changes Martin Schwidefsky
2008-11-03 17:14   ` David Smith
2008-11-05 11:41     ` Martin Schwidefsky
2008-11-06 18:24       ` David Smith
2008-11-07  9:14         ` Martin Schwidefsky
2008-11-07 15:32         ` Martin Schwidefsky
2008-10-01  8:33 ` [patch 16/21] dcssblk: add >2G DCSSs support and stacked contiguous DCSSs support Martin Schwidefsky
2008-10-01  8:33 ` [patch 17/21] nohz: Fix __udelay Martin Schwidefsky
2008-10-01  8:33 ` [patch 18/21] Move private simple udelay function to arch/s390/lib/delay.c Martin Schwidefsky
2008-10-01  8:33 ` [patch 19/21] dasd: fix message flood for unsolicited interrupts Martin Schwidefsky
2008-10-01  8:33 ` [patch 20/21] xpram: per device block request queues Martin Schwidefsky
2008-10-01  8:33 ` [patch 21/21] Add ioctl support for EMC Symmetrix Subsystem Control I/O Martin Schwidefsky
2008-10-01 11:03   ` Christoph Hellwig
2008-10-01 11:17     ` Martin Schwidefsky

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=20081001084125.504618165@de.ibm.com \
    --to=schwidefsky@de.ibm.com \
    --cc=heiko.carstens@de.ibm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-s390@vger.kernel.org \
    --cc=ursula.braun@de.ibm.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