Skip to contentSkip to conversation

clockevents: a couple of fixes for tip/timers/urgent

GIT PULL New Daniel Lezcano 4 participants 6 messages Aug 6, 2015 – Aug 9, 2015 Clocksource, clockevent drivers

lore
Share & tools
lore
#0 lore raw

Hi Ingo,

this pull request contains the following changes:

- Prevented to suspend or resume the sh_cmt clocksource when this one
is not enabled (Geert Uytterhoeven)
- Improved build coverage with COMPILE_TEST for the stm32 timer
(Maxime Coquelin)

Thanks !

-- Daniel

The following changes since commit 0f44705175347ec96935d60b765b5d14ecc763bb:

tick: Move the export of tick_broadcast_oneshot_control to the proper
place (2015-07-14 12:01:04 +0200)

are available in the git repository at:

http://git.linaro.org/people/daniel.lezcano/linux.git
clockevents/4.2-fixes

for you to fetch changes up to 1822a48be12f7ce2799d3423e49b967896794c4d:

clockevents/drivers/sh_cmt: Only perform clocksource suspend/resume
if enabled (2015-08-06 17:17:20 +0200)

----------------------------------------------------------------
Geert Uytterhoeven (1):
clockevents/drivers/sh_cmt: Only perform clocksource
suspend/resume if enabled

Maxime Coquelin (1):
clockevents/drivers/timer-stm32: Improve dependencies of timer-stm32

 drivers/clocksource/Kconfig  | 4 ++--
 drivers/clocksource/sh_cmt.c | 6 ++++++
 2 files changed, 8 insertions(+), 2 deletions(-)
Signature · 5 lines
 <http://www.linaro.org/> Linaro.org │ Open source software for ARM SoCs

Follow Linaro:  <http://www.facebook.com/pages/Linaro> Facebook |
<http://twitter.com/#!/linaroorg> Twitter |
<http://www.linaro.org/linaro-blog/> Blog
#1 lore raw

PATCH 1/2 clockevents/drivers/timer-stm32: Improve dependencies of timer-stm32

From: Maxime Coquelin <[email protected]>

This patch improves the build coverage with COMPILE_TEST so it is not only limited to ARM builds.

The fix consists in making the STM32 timer depend on GENERIC_CLOCKEVENTS.

---
 drivers/clocksource/Kconfig | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/clocksource/Kconfig +2 -2
index 4e57730..e110ec3 100644
@@ -112,8 +112,8 @@ config CLKSRC_LPC32XX
 	select CLKSRC_OF
 
 config CLKSRC_STM32
-	bool "Clocksource for STM32 SoCs" if !ARCH_STM32
-	depends on OF && ARM && (ARCH_STM32 || COMPILE_TEST)
+	bool "Clocksource for STM32 SoCs" if COMPILE_TEST
+	depends on OF && GENERIC_CLOCKEVENTS
 	select CLKSRC_MMIO
 
 config ARM_ARCH_TIMER
Signature · 1 line
1.9.1
#2 lore raw

PATCH 2/2 clockevents/drivers/sh_cmt: Only perform clocksource suspend/resume if enabled

From: Geert Uytterhoeven <[email protected]>

Currently the sh_cmt clocksource timer is disabled or enabled unconditionally on clocksource suspend resp. resume, even if a better clocksource is present (e.g. arch_sys_counter) and the sh_cmt clocksource is not enabled.

As sh_cmt is a syscore device when its timer is enabled, this may lead to a genpd.prepared_count imbalance in the presence of PM Domains, which may cause a lock-up during reboot after s2ram.

During suspend:
- pm_genpd_prepare() is called for all non-syscore devices (incl. sh_cmt), increasing genpd.prepared_count for each device,
- clocksource.suspend() is called for all clocksource devices,
- sh_cmt_clocksource_suspend() calls sh_cmt_stop(), which is a no-op as the clocksource was not enabled.

During resume:
- clocksource.resume() is called for all clocksource devices,
- sh_cmt_clocksource_resume() calls sh_cmt_start(), which enables the clocksource timer, and turns sh_cmt into a syscore device,
- pm_genpd_complete() is called for all non-syscore devices (excl. sh_cmt now!), decreasing genpd.prepared_count for each device but sh_cmt.

Now genpd.prepared_count of the PM Domain containing sh_cmt is still 1 instead of zero. On subsequent suspend/resume cycles, sh_cmt is still a syscore device, hence it's skipped for pm_genpd_{prepare,complete}(), keeping the imbalance of genpd.prepared_count at 1.

During reboot:
- platform_drv_shutdown() is called for any platform device that has a driver with a .shutdown() method (only rcar-dmac on R-Car Gen2),
- platform_drv_shutdown() calls dev_pm_domain_detach(), which calls genpd_dev_pm_detach(),
- genpd_dev_pm_detach() keeps calling pm_genpd_remove_device() until it doesn't return -EAGAIN[*],
- If the device is part of the same PM Domain as sh_cmt, pm_genpd_remove_device() always fails with -EAGAIN due to genpd.prepared_count > 0.
- Infinite loop in genpd_dev_pm_detach()[*].

[*] Commit 93af5e9354432828 ("PM / Domains: Avoid infinite loops in
attach/detach code") already limited the number of loop iterations, avoiding the lock-up.

To fix this, only disable or enable the clocksource timer on clocksource suspend resp. resume if the clocksource was enabled.

This was tested on r8a7791/koelsch with the CPG Clock Domain:
- using arch_sys_counter as the clocksource, which is the default, and which showed the problem,
- using sh_cmt as a clocksource ("echo ffca0000.timer > \
/sys/devices/system/clocksource/clocksource0/current_clocksource"), which behaves the same as before.

---
 drivers/clocksource/sh_cmt.c | 6 ++++++
 1 file changed, 6 insertions(+)
drivers/clocksource/sh_cmt.c +6 -0
index b8ff3c6..c96de14 100644
@@ -661,6 +661,9 @@ static void sh_cmt_clocksource_suspend(struct clocksource *cs)
 {
 	struct sh_cmt_channel *ch = cs_to_sh_cmt(cs);
 
+	if (!ch->cs_enabled)
+		return;
+
 	sh_cmt_stop(ch, FLAG_CLOCKSOURCE);
 	pm_genpd_syscore_poweroff(&ch->cmt->pdev->dev);
 }
@@ -669,6 +672,9 @@ static void sh_cmt_clocksource_resume(struct clocksource *cs)
 {
 	struct sh_cmt_channel *ch = cs_to_sh_cmt(cs);
 
+	if (!ch->cs_enabled)
+		return;
+
 	pm_genpd_syscore_poweron(&ch->cmt->pdev->dev);
 	sh_cmt_start(ch, FLAG_CLOCKSOURCE);
 }
Signature · 1 line
1.9.1
#3 lore raw

[tip:timers/urgent] clockevents/drivers/sh_cmt: Only perform clocksource suspend/resume if enabled

clockevents/drivers/sh_cmt: Only perform clocksource suspend/resume if enabled
Commit-ID:  54d46b7fbcbd00fe4b20a27208e5909facc714e3
Gitweb:     http://git.kernel.org/tip/54d46b7fbcbd00fe4b20a27208e5909facc714e3
Author:     Geert Uytterhoeven <[email protected]>
AuthorDate: Thu, 6 Aug 2015 17:32:06 +0200
Committer:  Ingo Molnar <[email protected]>
CommitDate: Sat, 8 Aug 2015 10:50:08 +0200

clockevents/drivers/sh_cmt: Only perform clocksource suspend/resume if enabled

Currently the sh_cmt clocksource timer is disabled or enabled unconditionally on clocksource suspend resp. resume, even if a better clocksource is present (e.g. arch_sys_counter) and the sh_cmt clocksource is not enabled.

As sh_cmt is a syscore device when its timer is enabled, this may lead to a genpd.prepared_count imbalance in the presence of PM Domains, which may cause a lock-up during reboot after s2ram.

During suspend:
- pm_genpd_prepare() is called for all non-syscore devices (incl. sh_cmt), increasing genpd.prepared_count for each device,
- clocksource.suspend() is called for all clocksource devices,
- sh_cmt_clocksource_suspend() calls sh_cmt_stop(), which is a no-op as the clocksource was not enabled.

During resume:
- clocksource.resume() is called for all clocksource devices,
- sh_cmt_clocksource_resume() calls sh_cmt_start(), which enables the clocksource timer, and turns sh_cmt into a syscore device,
- pm_genpd_complete() is called for all non-syscore devices (excl. sh_cmt now!), decreasing genpd.prepared_count for each device but sh_cmt.

Now genpd.prepared_count of the PM Domain containing sh_cmt is still 1 instead of zero. On subsequent suspend/resume cycles, sh_cmt is still a syscore device, hence it's skipped for pm_genpd_{prepare,complete}(), keeping the imbalance of genpd.prepared_count at 1.

During reboot:

- platform_drv_shutdown() is called for any platform device that has a driver with a .shutdown() method (only rcar-dmac on R-Car Gen2),

- platform_drv_shutdown() calls dev_pm_domain_detach(), which calls genpd_dev_pm_detach(),

- genpd_dev_pm_detach() keeps calling pm_genpd_remove_device() until it doesn't return -EAGAIN[*],

- If the device is part of the same PM Domain as sh_cmt, pm_genpd_remove_device() always fails with -EAGAIN due to genpd.prepared_count > 0.

- Infinite loop in genpd_dev_pm_detach()[*].

[*] Commit 93af5e9354432828 ("PM / Domains: Avoid infinite loops in
attach/detach code") already limited the number of loop iterations, avoiding the lock-up.

To fix this, only disable or enable the clocksource timer on clocksource suspend resp. resume if the clocksource was enabled.

This was tested on r8a7791/koelsch with the CPG Clock Domain:

- using arch_sys_counter as the clocksource, which is the default, and which showed the problem,

- using sh_cmt as a clocksource ("echo ffca0000.timer > \
/sys/devices/system/clocksource/clocksource0/current_clocksource"), which behaves the same as before.

---
 drivers/clocksource/sh_cmt.c | 6 ++++++
 1 file changed, 6 insertions(+)
drivers/clocksource/sh_cmt.c +6 -0
index b8ff3c6..c96de14 100644
@@ -661,6 +661,9 @@ static void sh_cmt_clocksource_suspend(struct clocksource *cs)
 {
 	struct sh_cmt_channel *ch = cs_to_sh_cmt(cs);
 
+	if (!ch->cs_enabled)
+		return;
+
 	sh_cmt_stop(ch, FLAG_CLOCKSOURCE);
 	pm_genpd_syscore_poweroff(&ch->cmt->pdev->dev);
 }
@@ -669,6 +672,9 @@ static void sh_cmt_clocksource_resume(struct clocksource *cs)
 {
 	struct sh_cmt_channel *ch = cs_to_sh_cmt(cs);
 
+	if (!ch->cs_enabled)
+		return;
+
 	pm_genpd_syscore_poweron(&ch->cmt->pdev->dev);
 	sh_cmt_start(ch, FLAG_CLOCKSOURCE);
 }
#4 lore raw
In reply to Daniel Lezcano
Quoted 34 lines · Daniel Lezcano

Hi Ingo,

this pull request contains the following changes:

- Prevented to suspend or resume the sh_cmt clocksource when this one is
not enabled (Geert Uytterhoeven)
- Improved build coverage with COMPILE_TEST for the stm32 timer (Maxime
Coquelin)

Thanks !

-- Daniel

The following changes since commit 0f44705175347ec96935d60b765b5d14ecc763bb:

tick: Move the export of tick_broadcast_oneshot_control to the proper
place (2015-07-14 12:01:04 +0200)

are available in the git repository at:

http://git.linaro.org/people/daniel.lezcano/linux.git
clockevents/4.2-fixes

for you to fetch changes up to 1822a48be12f7ce2799d3423e49b967896794c4d:

clockevents/drivers/sh_cmt: Only perform clocksource suspend/resume if
enabled (2015-08-06 17:17:20 +0200)

----------------------------------------------------------------
Geert Uytterhoeven (1):
clockevents/drivers/sh_cmt: Only perform clocksource suspend/resume if
enabled

Quoted 4 lines

Maxime Coquelin (1):
clockevents/drivers/timer-stm32: Improve dependencies of timer-stm32

 drivers/clocksource/Kconfig  | 4 ++--

Hm, so why is this an urgent fix, i.e. a regression fix? It appears to widen build coverage:

config CLKSRC_STM32

-       bool "Clocksource for STM32 SoCs" if !ARCH_STM32
-       depends on OF && ARM && (ARCH_STM32 || COMPILE_TEST)
+       bool "Clocksource for STM32 SoCs" if COMPILE_TEST
+       depends on OF && GENERIC_CLOCKEVENTS

that can at most introduce build regressions - not fix anything.

If there's any other purpose of this change it's not mentioned in the changelog.

Also, I think the change is broken to begin with:

+ bool "Clocksource for STM32 SoCs" if COMPILE_TEST

this will turn off the driver even in its target platform!

The right way to increase testing this way is via:

	depends on ... || COMPILE_TEST

Thanks,

Ingo

#5 lore raw
Quoted 16 lines · Ingo Molnar
Quoted 4 lines

Maxime Coquelin (1):
clockevents/drivers/timer-stm32: Improve dependencies of timer-stm32

 drivers/clocksource/Kconfig  | 4 ++--

Hm, so why is this an urgent fix, i.e. a regression fix? It appears to widen build coverage:

config CLKSRC_STM32

(quoted diff) +2 -2
-       bool "Clocksource for STM32 SoCs" if !ARCH_STM32
-       depends on OF && ARM && (ARCH_STM32 || COMPILE_TEST)
+       bool "Clocksource for STM32 SoCs" if COMPILE_TEST
+       depends on OF && GENERIC_CLOCKEVENTS

that can at most introduce build regressions - not fix anything.

Quoted 7 lines

If there's any other purpose of this change it's not mentioned in the changelog.

Also, I think the change is broken to begin with:

+ bool "Clocksource for STM32 SoCs" if COMPILE_TEST

this will turn off the driver even in its target platform!

FWIW, ARCH_STM32 in arch/arm/Kconfig selects CLKSRC_STM32

Quoted 3 lines

The right way to increase testing this way is via:

        depends on ... || COMPILE_TEST

Indeed.

Gr{oetje,eeting}s,

Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds