linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH] regulator: core: Reduce busy-wait looping
@ 2013-09-20 11:51 Thierry Reding
  2013-09-20 16:42 ` Mark Brown
  0 siblings, 1 reply; 4+ messages in thread
From: Thierry Reding @ 2013-09-20 11:51 UTC (permalink / raw)
  To: Mark Brown, Liam Girdwood; +Cc: linux-kernel

Keep busy-wait looping to a minimum while waiting for a regulator to
ramp-up to the target voltage. This follows the guidelines set forth
in Documentation/timers/timers-howto.txt and assumes that regulators
are never enabled in atomic context.

Signed-off-by: Thierry Reding <treding@nvidia.com>
---
 drivers/regulator/core.c | 38 +++++++++++++++++++++++++++++++++-----
 1 file changed, 33 insertions(+), 5 deletions(-)

diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c
index 4eefcc3..4e18884 100644
--- a/drivers/regulator/core.c
+++ b/drivers/regulator/core.c
@@ -1566,11 +1566,39 @@ static int _regulator_do_enable(struct regulator_dev *rdev)
 	 * together.  */
 	trace_regulator_enable_delay(rdev_get_name(rdev));
 
-	if (delay >= 1000) {
-		mdelay(delay / 1000);
-		udelay(delay % 1000);
-	} else if (delay) {
-		udelay(delay);
+	/*
+	 * Delay for the requested amount of time as per the guidelines in:
+	 *
+	 *     Documentation/timers/timers-howto.txt
+	 *
+	 * The assumption here is that regulators will never be enabled in
+	 * atomic context and therefore sleeping functions can be used.
+	 */
+	if (delay) {
+		unsigned int ms = delay / 1000;
+		unsigned int us = delay % 1000;
+
+		if (ms > 0) {
+			/*
+			 * For small enough values, handle super-millisecond
+			 * delays in the usleep_range() call below.
+			 */
+			if (ms < 20)
+				us += ms * 1000;
+			else
+				msleep(ms);
+		}
+
+		/*
+		 * Give the scheduler some room to coalesce with any other
+		 * wakeup sources. For delays shorter than 10 us, don't even
+		 * bother setting up high-resolution timers and just busy-
+		 * loop.
+		 */
+		if (us >= 10)
+			usleep_range(us, us + 100);
+		else
+			udelay(us);
 	}
 
 	trace_regulator_enable_complete(rdev_get_name(rdev));
-- 
1.8.4


^ permalink raw reply related	[flat|nested] 4+ messages in thread
* [PATCH] regulator: core: Reduce busy-wait looping
@ 2014-03-21  2:31 Jonghwan Choi
  2014-03-21 11:50 ` Mark Brown
  0 siblings, 1 reply; 4+ messages in thread
From: Jonghwan Choi @ 2014-03-21  2:31 UTC (permalink / raw)
  To: 'open list'; +Cc: 'Liam Girdwood', 'Mark Brown'

Commit 5df529d440("regulator: core: Reduce busy-wait looping")
can also be used in regulator_do_set_voltage.

Signed-off-by: Jonghwan Choi <jhbird.choi@samsung.com>
---
 drivers/regulator/core.c |   39 +++++++++++++++++++++++++++++++++------
 1 file changed, 33 insertions(+), 6 deletions(-)

diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c
index afca1bc..c375bfb 100644
--- a/drivers/regulator/core.c
+++ b/d728099474rivers/regulator/core.c
@@ -2355,12 +2355,39 @@ static int _regulator_do_set_voltage(struct
regulator_dev *rdev,
 			delay = 0;
 		}
 
-		/* Insert any necessary delays */
-		if (delay >= 1000) {
-			mdelay(delay / 1000);
-			udelay(delay % 1000);
-		} else if (delay) {
-			udelay(delay);
+		/*
+		 * Delay for the requested amount of time as per the
guidelines in:
+		 *
+		 *     Documentation/timers/timers-howto.txt
+		 *
+		 * The assumption here is that regulators will never set the
voltage in
+		 * atomic context and therefore sleeping functions can be
used.
+		 */
+		if (delay) {
+			unsigned int ms = delay / 1000;
+			unsigned int us = delay % 1000;
+
+			if (ms > 0) {
+				/*
+				 * For small enough values, handle
super-millisecond
+				 * delays in the usleep_range() call below.
+				 */
+				if (ms < 20)
+					us += ms * 1000;
+				else
+					msleep(ms);
+			}
+
+			/*
+			 * Give the scheduler some room to coalesce with any
other
+			 * wakeup sources. For delays shorter than 10 us,
don't even
+			 * bother setting up high-resolution timers and just
busy-
+			 * loop.
+			 */
+			if (us >= 10)
+				usleep_range(us, us + 100);
+			else
+				udelay(us);
 		}
 	}
 
-- 
1.7.10.4


^ permalink raw reply related	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2014-03-21 11:50 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2013-09-20 11:51 [PATCH] regulator: core: Reduce busy-wait looping Thierry Reding
2013-09-20 16:42 ` Mark Brown
  -- strict thread matches above, loose matches on Subject: below --
2014-03-21  2:31 Jonghwan Choi
2014-03-21 11:50 ` Mark Brown

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).