linux-arm-kernel.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
From: u.kleine-koenig@pengutronix.de (Uwe Kleine-König)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH 1/2] arm/at91: Don't disable irqs in set_next_event and set_mode callbacks
Date: Thu, 17 Dec 2009 14:33:40 +0100	[thread overview]
Message-ID: <1261056821-28521-1-git-send-email-u.kleine-koenig@pengutronix.de> (raw)
In-Reply-To: <1259231164-21242-1-git-send-email-u.kleine-koenig@pengutronix.de>

on AT91 the timer irq is shared, so the handler might be entered without
irqs being disabled.  Though this should not happen as the timer irq is
registered early, there have been some reports on the mailing list.

To make debugging that problem easier next time it pops up a WARN_ON is
added in the handler if irqs are not off.

Signed-off-by: Uwe Kleine-K?nig <u.kleine-koenig@pengutronix.de>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Magnus Damm <damm@opensource.se>
Cc: Hugh Dickins <hugh@veritas.com>
Cc: John Stultz <johnstul@us.ibm.com>
---
 arch/arm/mach-at91/at91rm9200_time.c  |   20 ++++++--------------
 arch/arm/mach-at91/at91sam926x_time.c |   11 ++++++-----
 2 files changed, 12 insertions(+), 19 deletions(-)

diff --git a/arch/arm/mach-at91/at91rm9200_time.c b/arch/arm/mach-at91/at91rm9200_time.c
index 309f351..f494ce9 100644
--- a/arch/arm/mach-at91/at91rm9200_time.c
+++ b/arch/arm/mach-at91/at91rm9200_time.c
@@ -58,6 +58,12 @@ static irqreturn_t at91rm9200_timer_interrupt(int irq, void *dev_id)
 {
 	u32	sr = at91_sys_read(AT91_ST_SR) & irqmask;
 
+	/*
+	 * irqs should be disabled here, but as the irq is shared they are only
+	 * guaranteed to be off if the timer irq is registered first.
+	 */
+	WARN_ON(!irqs_disabled());
+
 	/* simulate "oneshot" timer with alarm */
 	if (sr & AT91_ST_ALMS) {
 		clkevt.event_handler(&clkevt);
@@ -132,24 +138,11 @@ clkevt32k_mode(enum clock_event_mode mode, struct clock_event_device *dev)
 static int
 clkevt32k_next_event(unsigned long delta, struct clock_event_device *dev)
 {
-	unsigned long	flags;
 	u32		alm;
 	int		status = 0;
 
 	BUG_ON(delta < 2);
 
-	/* Use "raw" primitives so we behave correctly on RT kernels. */
-	raw_local_irq_save(flags);
-
-	/*
-	 * According to Thomas Gleixner irqs are already disabled here.  Simply
-	 * removing raw_local_irq_save above (and the matching
-	 * raw_local_irq_restore) was not accepted.  See
-	 * http://thread.gmane.org/gmane.linux.ports.arm.kernel/41174
-	 * So for now (2008-11-20) just warn once if irqs were not disabled ...
-	 */
-	WARN_ON_ONCE(!raw_irqs_disabled_flags(flags));
-
 	/* The alarm IRQ uses absolute time (now+delta), not the relative
 	 * time (delta) in our calling convention.  Like all clockevents
 	 * using such "match" hardware, we have a race to defend against.
@@ -169,7 +162,6 @@ clkevt32k_next_event(unsigned long delta, struct clock_event_device *dev)
 	alm += delta;
 	at91_sys_write(AT91_ST_RTAR, alm);
 
-	raw_local_irq_restore(flags);
 	return status;
 }
 
diff --git a/arch/arm/mach-at91/at91sam926x_time.c b/arch/arm/mach-at91/at91sam926x_time.c
index 4bd56ae..786f172 100644
--- a/arch/arm/mach-at91/at91sam926x_time.c
+++ b/arch/arm/mach-at91/at91sam926x_time.c
@@ -62,16 +62,12 @@ static struct clocksource pit_clk = {
 static void
 pit_clkevt_mode(enum clock_event_mode mode, struct clock_event_device *dev)
 {
-	unsigned long	flags;
-
 	switch (mode) {
 	case CLOCK_EVT_MODE_PERIODIC:
-		/* update clocksource counter, then enable the IRQ */
-		raw_local_irq_save(flags);
+		/* update clocksource counter */
 		pit_cnt += pit_cycle * PIT_PICNT(at91_sys_read(AT91_PIT_PIVR));
 		at91_sys_write(AT91_PIT_MR, (pit_cycle - 1) | AT91_PIT_PITEN
 				| AT91_PIT_PITIEN);
-		raw_local_irq_restore(flags);
 		break;
 	case CLOCK_EVT_MODE_ONESHOT:
 		BUG();
@@ -100,6 +96,11 @@ static struct clock_event_device pit_clkevt = {
  */
 static irqreturn_t at91sam926x_pit_interrupt(int irq, void *dev_id)
 {
+	/*
+	 * irqs should be disabled here, but as the irq is shared they are only
+	 * guaranteed to be off if the timer irq is registered first.
+	 */
+	WARN_ON(!irqs_disabled());
 
 	/* The PIT interrupt may be disabled, and is shared */
 	if ((pit_clkevt.mode == CLOCK_EVT_MODE_PERIODIC)
-- 
1.6.5.2

  parent reply	other threads:[~2009-12-17 13:33 UTC|newest]

Thread overview: 68+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-09-21  7:39 [PATCH] Don't disable irqs in set_next_event and set_mode callbacks Uwe Kleine-König
2009-09-21  9:04 ` Russell King - ARM Linux
2009-09-21  9:16   ` Uwe Kleine-König
2009-09-21  9:30     ` Russell King - ARM Linux
2009-09-21 10:48     ` Alessandro Rubini
2009-09-21 12:32 ` Kristoffer Ericson
2009-09-23 19:01   ` Eric Miao
2009-09-23 21:04 ` Remy Bohmer
2009-11-26 10:26 ` [RESENT PATCH] " Uwe Kleine-König
2009-11-26 10:50   ` Russell King - ARM Linux
2009-11-26 11:31     ` Russell King - ARM Linux
2009-11-27 10:44       ` Uwe Kleine-König
2009-11-27 19:08         ` Thomas Gleixner
2009-11-27 19:58         ` Russell King - ARM Linux
2009-11-27 20:38           ` Uwe Kleine-König
2009-11-27 20:44             ` Russell King - ARM Linux
2009-11-27 21:31               ` Thomas Gleixner
2009-11-27 21:59                 ` Russell King - ARM Linux
2009-11-27 22:20                   ` Thomas Gleixner
2009-11-27 21:10           ` [PATCH] warn about shared irqs requesting IRQF_DISABLED registered with setup_irq Uwe Kleine-König
2009-11-27 22:18             ` Thomas Gleixner
2009-11-28 20:03               ` Uwe Kleine-König
2009-11-28 21:50                 ` Thomas Gleixner
2009-11-28 22:13                   ` David Brownell
2009-11-29  2:31                   ` Jamie Lokier
2009-11-29 10:26                     ` Uwe Kleine-König
2009-11-29 15:18                       ` Jamie Lokier
2009-11-29 15:27                         ` Russell King - ARM Linux
2009-11-30 20:39                           ` Jamie Lokier
2009-11-30  9:28                         ` Uwe Kleine-König
2009-11-30  9:54                         ` Thomas Gleixner
2009-11-28 22:09                 ` David Brownell
2009-11-30 10:47             ` [PATCH] genirq: warn about IRQF_SHARED|IRQF_DISABLED at the right place Uwe Kleine-König
2009-11-30 13:54               ` Get rid of IRQF_DISABLED - (was [PATCH] genirq: warn about IRQF_SHARED|IRQF_DISABLED) Thomas Gleixner
2009-11-30 14:03                 ` Peter Zijlstra
2009-11-30 14:24                   ` Thomas Gleixner
2009-11-30 14:47                     ` Alan Cox
2009-11-30 15:01                       ` Russell King - ARM Linux
2009-11-30 15:32                         ` Alan Cox
2009-11-30 15:43                           ` Russell King - ARM Linux
2009-11-30 20:15                         ` Andrew Victor
2009-11-30 20:53                         ` David Brownell
2009-11-30 20:38                       ` David Brownell
2009-12-01  1:42                         ` Andy Walls
2009-11-30 19:59                     ` Benjamin Herrenschmidt
2009-11-30 21:31                       ` Thomas Gleixner
2009-11-30 21:42                         ` Benjamin Herrenschmidt
2009-11-30 21:54                           ` Thomas Gleixner
2009-11-30 14:37                 ` Russell King - ARM Linux
2009-11-30 14:39                   ` Russell King - ARM Linux
2009-11-30 17:48                     ` Thomas Gleixner
2009-11-30 14:51                   ` Alan Cox
2009-11-30 21:59                     ` Thomas Gleixner
2009-11-30 23:30                       ` Alan Cox
2009-11-30 15:38                   ` Nicolas Pitre
2009-11-30 17:46                   ` Thomas Gleixner
2009-11-30 19:51                 ` Uwe Kleine-König
2009-11-30 21:23                   ` Thomas Gleixner
2009-11-30 20:21               ` [PATCH] genirq: warn about IRQF_SHARED|IRQF_DISABLED at the right place David Brownell
2009-11-30 20:27                 ` Uwe Kleine-König
2010-01-12 15:42               ` [RESEND PATCH] " Uwe Kleine-König
2009-11-26 10:51   ` [RESENT PATCH] Don't disable irqs in set_next_event and set_mode callbacks Eric Miao
2009-12-17 13:33   ` Uwe Kleine-König [this message]
2009-12-17 13:33     ` [PATCH 2/2] arm/{pxa, sa1100, nomadik}: Don't disable irqs in set_next_event and set_mode Uwe Kleine-König
2010-01-22 16:08     ` [PATCH 1/2] arm/at91: Don't disable irqs in set_next_event and set_mode callbacks Uwe Kleine-König
2010-01-22 16:36       ` Russell King - ARM Linux
2010-01-22 16:52         ` Uwe Kleine-König
2010-02-12 10:35           ` Uwe Kleine-König

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=1261056821-28521-1-git-send-email-u.kleine-koenig@pengutronix.de \
    --to=u.kleine-koenig@pengutronix.de \
    --cc=linux-arm-kernel@lists.infradead.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;
as well as URLs for NNTP newsgroup(s).