misc: atmel_pwm: fix probe dependencies
cover misc: atmel_pwm: fix probe dependencies
Two drivers (atmel-pwm-bl and leds-atmel-pwm) currently depend on the atmel_pwm driver to have bound to any pwm-device before their devices are probed.
The first patch adds deferred-probing support to the atmel_pwm driver to handle such dependencies. Note that the atmel-pwm-bl driver in linux-next supports deferred probing since commit 9d3fde86b ("backlight:
atmel-pwm-bl: fix deferred probe from __init").
Although deferred probing is sufficient to deal with the dependency problem, relying on deferred probing for the backlight driver is not desirable as it may cause unnecessary delays before enabling the backlight at boot. The second patch fixes this by making sure pwm-devices are probed before any backlight devices.
Johan
Johan Hovold (2):
misc: atmel_pwm: add deferred-probing support
misc: atmel_pwm: set initcall level to subsys
drivers/misc/atmel_pwm.c | 18 +++++++++++++++--- 1 file changed, 15 insertions(+), 3 deletions(-)
Signature · 1 line
1.8.4
PATCH 1/2 misc: atmel_pwm: add deferred-probing support
Two drivers (atmel-pwm-bl and leds-atmel-pwm) currently depend on the atmel_pwm driver to have bound to any pwm-device before their devices are probed.
Support deferred probing of such devices by making sure to return -EPROBE_DEFER from pwm_channel_alloc when no pwm-device has yet been bound.
--- drivers/misc/atmel_pwm.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/misc/atmel_pwm.c +4 -2
@@ -90,8 +90,10 @@ int pwm_channel_alloc(int index, struct pwm_channel *ch)unsigned long flags;int status =0 ; -/* insist on PWM init, with this signal pinned out */ -if (!pwm || !(pwm->mask &1 << index)) +if (!pwm) +return -EPROBE_DEFER ; + +if (!(pwm->mask &1 << index))return -ENODEV ;if (index <0 || index >=PWM_NCHAN || !ch)
Signature · 1 line
1.8.4
Re: [PATCH 1/2] misc: atmel_pwm: add deferred-probing support
Quoted 9 lines · Johan Hovold
Two drivers (atmel-pwm-bl and leds-atmel-pwm) currently depend on the
atmel_pwmdriver to have bound to any pwm-device before their devices are probed.Support deferred probing of such devices by making sure to return -EPROBE_DEFER from
pwm_channel_allocwhen no pwm-device has yet been bound.Signed-off-by Johan Hovold <[email protected]>
This patch, 5c6d6fd1 ("misc: atmel_pwm: add deferred-probing support"), and 9d3fde86 ("backlight: atmel-pwm-bl: fix deferred probe from __init") in linux-next should probably be considered for inclusion in stable.
Without them the atmel-pwm-bl driver may always fail to probe depending on link order.
Thanks,
Johan
Quoted 24 lines
--- drivers/misc/atmel_pwm.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)drivers/misc/atmel_pwm.c +4
-2@@ -90,8 +90,10 @@ int pwm_channel_alloc(int index, struct pwm_channel *ch)unsigned long flags;int status =0 ; -/* insist on PWM init, with this signal pinned out */ -if (!pwm || !(pwm->mask &1 << index)) +if (!pwm) +return -EPROBE_DEFER ; + +if (!(pwm->mask &1 << index))return -ENODEV ;if (index <0 || index >=PWM_NCHAN || !ch)Signature · 1 line
1.8.4
Greg Kroah-Hartman
maintainerRe: [PATCH 1/2] misc: atmel_pwm: add deferred-probing support
Quoted 9 lines · Johan Hovold
Quoted 8 lines · Johan Hovold
Two drivers (atmel-pwm-bl and leds-atmel-pwm) currently depend on the
atmel_pwmdriver to have bound to any pwm-device before their devices are probed.Support deferred probing of such devices by making sure to return -EPROBE_DEFER from
pwm_channel_allocwhen no pwm-device has yet been bound.
Quoted 8 lines
Quoted 1 line
Signed-off-by Johan Hovold <[email protected]>This patch,
5c6d6fd1("misc:atmel_pwm: add deferred-probing support"), and9d3fde86("backlight: atmel-pwm-bl: fix deferred probe from__init") in linux-next should probably be considered for inclusion in stable.Without them the atmel-pwm-bl driver may always fail to probe depending on link order.
Ok, I'll watch this for when it hits Linus's tree, thanks for letting me know.
greg k-h
PATCH 2/2 misc: atmel_pwm: set initcall level to subsys
In reply to Johan HovoldEven with the atmel_pwm driver and the atmel-pwm-bl backlight driver supporting deferred probing, we still want to make sure that any pwm-device is available when the backlight devices are probed to avoid any unnecessary delays before enabling the backlight.
--- drivers/misc/atmel_pwm.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-)
drivers/misc/atmel_pwm.c +11 -1
@@ -395,7 +395,17 @@ static struct platform_driver atmel_pwm_driver = {*/ }; -module_platform_driver_probe (atmel_pwm_driver, pwm_probe); +static int __init pwm_init (void ) +{ +return platform_driver_probe (&atmel_pwm_driver, pwm_probe); +} +subsys_initcall (pwm_init); + +static void __exit pwm_exit(void) +{ +platform_driver_unregister (&atmel_pwm_driver); +} +module_exit (pwm_exit);MODULE_DESCRIPTION ("Driver for AT32/AT91 PWM module" );MODULE_LICENSE ("GPL" );
Signature · 1 line
1.8.4
Greg Kroah-Hartman
maintainerRe: [PATCH 2/2] misc: atmel_pwm: set initcall level to subsys
Quoted 15 lines · Johan Hovold
Even with the
atmel_pwmdriver and the atmel-pwm-bl backlight driver supporting deferred probing, we still want to make sure that any pwm-device is available when the backlight devices are probed to avoid any unnecessary delays before enabling the backlight.Signed-off-by Johan Hovold <[email protected]>--- drivers/misc/atmel_pwm.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-)drivers/misc/atmel_pwm.c +0
-0
drivers/misc/atmel_pwm.c static struct platform_driver atmel_pwm_driver = { · line 402
Quoted 9 lines
drivers/misc/atmel_pwm.c +5
-1@@ -395,7 +395,17 @@ static struct platform_driver atmel_pwm_driver = {*/ }; -module_platform_driver_probe (atmel_pwm_driver, pwm_probe); +static int __init pwm_init (void ) +{ +return platform_driver_probe (&atmel_pwm_driver, pwm_probe); +} +subsys_initcall (pwm_init);
I really hate this type of patch, as it's papering over the real problem. What happens when someone else moves their driver to this level? Then you are back to the original problem.
This is what deferred probing was supposed to fix. If it doesn't, then something else needs to be done, or fix the deferred probing mess...
Sorry, I can't take this.
greg k-h
Re: [PATCH 2/2] misc: atmel_pwm: set initcall level to subsys
Quoted 19 lines · Greg KH
Quoted 18 lines · Johan Hovold
Even with the
atmel_pwmdriver and the atmel-pwm-bl backlight driver supporting deferred probing, we still want to make sure that any pwm-device is available when the backlight devices are probed to avoid any unnecessary delays before enabling the backlight.Signed-off-by Johan Hovold <[email protected]>--- drivers/misc/atmel_pwm.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-)drivers/misc/atmel_pwm.c +0
-0@@ -395,7 +395,17 @@ static struct platform_driver atmel_pwm_driver = {*/ };
Quoted 10 lines
drivers/misc/atmel_pwm.c static struct platform_driver atmel_pwm_driver = { · line 402
Quoted 6 lines
drivers/misc/atmel_pwm.c +5
-1-module_platform_driver_probe (atmel_pwm_driver, pwm_probe); +static int __init pwm_init (void ) +{ +return platform_driver_probe (&atmel_pwm_driver, pwm_probe); +} +subsys_initcall (pwm_init);I really hate this type of patch, as it's papering over the real problem. What happens when someone else moves their driver to this level? Then you are back to the original problem.
Yes, it's crude, but it's currently the only way to express a preferred probe order.
Quoted 2 lines
This is what deferred probing was supposed to fix. If it doesn't, then something else needs to be done, or fix the deferred probing mess...
Deferred probing (the first patch) fixes the dependency problem, but may introduce delays. I can live with that.
Quoted 1 line
Sorry, I can't take this.
Fair enough.
Thanks,
Johan
Arnd Bergmann
maintainerRe: [PATCH 2/2] misc: atmel_pwm: set initcall level to subsys
Quoted 6 lines · Johan Hovold
Quoted 3 lines
I really hate this type of patch, as it's papering over the real problem. What happens when someone else moves their driver to this level? Then you are back to the original problem.
Yes, it's crude, but it's currently the only way to express a preferred probe order.
Does the same problem occur with the generic PWM subsystem?
The atmel_pwm driver apparently is the last PWM driver that has not been converted to the generic subsystem yet. Doing that would also let us drop the atmel-pwm-bl driver in favor of the generic backlight.
Arnd
Re: [PATCH 2/2] misc: atmel_pwm: set initcall level to subsys
Hi Arnd,
Quoted 8 lines · Arnd Bergmann
Quoted 6 lines · Johan Hovold
Quoted 3 lines
I really hate this type of patch, as it's papering over the real problem. What happens when someone else moves their driver to this level? Then you are back to the original problem.
Yes, it's crude, but it's currently the only way to express a preferred probe order.
Quoted 6 lines
Does the same problem occur with the generic PWM subsystem?
The
atmel_pwmdriver apparently is the last PWM driver that has not been converted to the generic subsystem yet. Doing that would also let us drop the atmel-pwm-bl driver in favor of the generic backlight.
The atmel pwm driver is on the way to generic subsystem. more information at:
--->8---
PWM: atmel-pwm: add PWM controller driver:
https://patchwork.kernel.org/patch/2963081/
---<8---
Best Regards,
Bo Shen
Re: [PATCH 2/2] misc: atmel_pwm: set initcall level to subsys
Quoted 18 lines · Bo Shen
Hi Arnd,
Quoted 8 lines · Arnd Bergmann
Quoted 6 lines · Johan Hovold
Quoted 3 lines
I really hate this type of patch, as it's papering over the real problem. What happens when someone else moves their driver to this level? Then you are back to the original problem.
Yes, it's crude, but it's currently the only way to express a preferred probe order.
Quoted 6 lines
Does the same problem occur with the generic PWM subsystem?
The
atmel_pwmdriver apparently is the last PWM driver that has not been converted to the generic subsystem yet. Doing that would also let us drop the atmel-pwm-bl driver in favor of the generic backlight.
Quoted 5 lines
The atmel pwm driver is on the way to generic subsystem. more information at:
--->8---
PWM: atmel-pwm: add PWM controller driver:https://patchwork.kernel.org/patch/2963081/
so the fix can be drop in favor of the conversion
Best Regards,
J.
Quoted 10 lines
---<8---
Best Regards,
Bo Shen_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel at lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Re: [PATCH 2/2] misc: atmel_pwm: set initcall level to subsys
Quoted 19 lines · Jean-Christophe PLAGNIOL-VILLARD
Quoted 18 lines · Bo Shen
Hi Arnd,
Quoted 14 lines · Arnd Bergmann
Quoted 6 lines · Johan Hovold
Quoted 3 lines
I really hate this type of patch, as it's papering over the real problem. What happens when someone else moves their driver to this level? Then you are back to the original problem.
Yes, it's crude, but it's currently the only way to express a preferred probe order.
Does the same problem occur with the generic PWM subsystem?
The
atmel_pwmdriver apparently is the last PWM driver that has not been converted to the generic subsystem yet. Doing that would also let us drop the atmel-pwm-bl driver in favor of the generic backlight.
Quoted 7 lines
Quoted 5 lines
The atmel pwm driver is on the way to generic subsystem. more information at:
--->8---
PWM: atmel-pwm: add PWM controller driver:https://patchwork.kernel.org/patch/2963081/so the fix can be drop in favor of the conversion
This patch (2/2) has been dropped, but the dependency issue still has to be addressed (by adding deferred-probing support). The atmel-pwm-bl driver has not been removed yet and the fix is also needed in stable.
Johan
Re: [PATCH 2/2] misc: atmel_pwm: set initcall level to subsys
In reply to Arnd BergmannQuoted 9 lines · Arnd Bergmann
Quoted 6 lines · Johan Hovold
Quoted 3 lines
I really hate this type of patch, as it's papering over the real problem. What happens when someone else moves their driver to this level? Then you are back to the original problem.
Yes, it's crude, but it's currently the only way to express a preferred probe order.
Does the same problem occur with the generic PWM subsystem?
It does not suffer from the problem with failed probe as it supports deferred probing, but there could be added delays depending on probe order.
Quoted 4 lines
The
atmel_pwmdriver apparently is the last PWM driver that has not been converted to the generic subsystem yet. Doing that would also let us drop the atmel-pwm-bl driver in favor of the generic backlight.
As Bo mentioned, it's work in progress that hasn't been merged yet.
Johan