mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: "Wolfgang Mües" <wolfgang.mues@auerswald.de>
To: Pierre Ossman <drzeus@drzeus.cx>
Cc: "Andrew Morton" <akpm@linux-foundation.org>,
	"Matt Fleming" <matt@console-pimps.org>,
	"David Brownell" <dbrownell@users.sourceforge.net>,
	"Mike Frysinger" <vapier.adi@gmail.com>,
	linux-kernel@vger.kernel.org
Subject: [PATCH 6/7 revised] mmc_spi: convert timeout handling to jiffies and avoid busy waiting
Date: Tue, 7 Apr 2009 15:26:30 +0100	[thread overview]
Message-ID: <200904071626.30917.wolfgang.mues@auerswald.de> (raw)

From: Wolfgang Muees <wolfgang.mues@auerswald.de>

o SD/MMC card timeouts can be very high. So avoid busy-waiting,
  using the scheduler. Calculate all timeouts in jiffies units,
  because this will give us the correct sign when to involve
  the scheduler.

Signed-off-by: Wolfgang Muees <wolfgang.mues@auerswald.de>

---

This patch was discussed on lkml. Some where concerned about
performance loss, but no performance loss was visible in real-
life tests, and nobody came up with a better solution. And it
was clear that a busy waiting of several 100 of ms has to be
ended.

diff -uprN 2_6_29_rc7_patch12534/drivers/mmc/host/mmc_spi.c 2_6_29_rc7_patch125346/drivers/mmc/host/mmc_spi.c
--- 2_6_29_rc7_patch12534/drivers/mmc/host/mmc_spi.c	2009-03-16 12:15:36.000000000 +0100
+++ 2_6_29_rc7_patch125346/drivers/mmc/host/mmc_spi.c	2009-04-07 15:59:54.000000000 +0200
@@ -24,7 +24,7 @@
  * along with this program; if not, write to the Free Software
  * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  */
-#include <linux/hrtimer.h>
+#include <linux/sched.h>
 #include <linux/delay.h>
 #include <linux/bio.h>
 #include <linux/dma-mapping.h>
@@ -95,7 +95,7 @@
  * reads which takes nowhere near that long.  Older cards may be able to use
  * shorter timeouts ... but why bother?
  */
-#define r1b_timeout		ktime_set(3, 0)
+#define r1b_timeout		(HZ * 3)
 
 
 /****************************************************************************/
@@ -183,12 +183,11 @@ mmc_spi_readbytes(struct mmc_spi_host *h
 	return status;
 }
 
-static int
-mmc_spi_skip(struct mmc_spi_host *host, ktime_t timeout, unsigned n, u8 byte)
+static int mmc_spi_skip(struct mmc_spi_host *host, unsigned long timeout,
+			unsigned n, u8 byte)
 {
 	u8		*cp = host->data->status;
-
-	timeout = ktime_add(timeout, ktime_get());
+	unsigned long start = jiffies;
 
 	while (1) {
 		int		status;
@@ -203,22 +202,26 @@ mmc_spi_skip(struct mmc_spi_host *host, 
 				return cp[i];
 		}
 
-		/* REVISIT investigate msleep() to avoid busy-wait I/O
-		 * in at least some cases.
-		 */
-		if (ktime_to_ns(ktime_sub(ktime_get(), timeout)) > 0)
+		if (time_is_before_jiffies(start + timeout))
 			break;
+
+		/* If we need long timeouts, we may release the CPU.
+		 * We use jiffies here because we want to have a relation
+		 * between elapsed time and the blocking of the scheduler.
+		 */
+		if (time_is_before_jiffies(start+1))
+			schedule();
 	}
 	return -ETIMEDOUT;
 }
 
 static inline int
-mmc_spi_wait_unbusy(struct mmc_spi_host *host, ktime_t timeout)
+mmc_spi_wait_unbusy(struct mmc_spi_host *host, unsigned long timeout)
 {
 	return mmc_spi_skip(host, timeout, sizeof(host->data->status), 0);
 }
 
-static int mmc_spi_readtoken(struct mmc_spi_host *host, ktime_t timeout)
+static int mmc_spi_readtoken(struct mmc_spi_host *host, unsigned long timeout)
 {
 	return mmc_spi_skip(host, timeout, 1, 0xff);
 }
@@ -607,7 +610,7 @@ mmc_spi_setup_data_message(
  */
 static int
 mmc_spi_writeblock(struct mmc_spi_host *host, struct spi_transfer *t,
-	ktime_t timeout)
+	unsigned long timeout)
 {
 	struct spi_device	*spi = host->spi;
 	int			status, i;
@@ -717,7 +720,7 @@ mmc_spi_writeblock(struct mmc_spi_host *
  */
 static int
 mmc_spi_readblock(struct mmc_spi_host *host, struct spi_transfer *t,
-	ktime_t timeout)
+	unsigned long timeout)
 {
 	struct spi_device	*spi = host->spi;
 	int			status;
@@ -803,7 +806,7 @@ mmc_spi_data_do(struct mmc_spi_host *hos
 	unsigned		n_sg;
 	int			multiple = (data->blocks > 1);
 	u32			clock_rate;
-	ktime_t			timeout;
+	unsigned long		timeout;
 
 	if (data->flags & MMC_DATA_READ)
 		direction = DMA_FROM_DEVICE;
@@ -817,8 +820,9 @@ mmc_spi_data_do(struct mmc_spi_host *hos
 	else
 		clock_rate = spi->max_speed_hz;
 
-	timeout = ktime_add_ns(ktime_set(0, 0), data->timeout_ns +
-			data->timeout_clks * 1000000 / clock_rate);
+	timeout = data->timeout_ns +
+		  data->timeout_clks * 1000000 / clock_rate;
+	timeout = usecs_to_jiffies((unsigned int)(timeout / 1000)) + 1;
 
 	/* Handle scatterlist segments one at a time, with synch for
 	 * each 512-byte block

---
regards

i. A. Wolfgang Mües
-- 
Auerswald GmbH & Co. KG
Hardware Development
Telefon: +49 (0)5306 9219 0
Telefax: +49 (0)5306 9219 94 
E-Mail: Wolfgang.Mues@Auerswald.de
Web: http://www.auerswald.de
 
--------------------------------------------------------------
Auerswald GmbH & Co. KG, Vor den Grashöfen 1, 38162 Cremlingen
Registriert beim AG Braunschweig HRA 13289
p.h.G Auerswald Geschäftsführungsges. mbH
Registriert beim AG Braunschweig HRB 7463
Geschäftsführer: Dipl-Ing. Gerhard Auerswald

             reply	other threads:[~2009-04-07 14:27 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-04-07 14:26 Wolfgang Mües [this message]
2009-04-08 18:39 ` Pierre Ossman

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=200904071626.30917.wolfgang.mues@auerswald.de \
    --to=wolfgang.mues@auerswald.de \
    --cc=akpm@linux-foundation.org \
    --cc=dbrownell@users.sourceforge.net \
    --cc=drzeus@drzeus.cx \
    --cc=linux-kernel@vger.kernel.org \
    --cc=matt@console-pimps.org \
    --cc=vapier.adi@gmail.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

all inboxes | Powered by JetHome®