2010-08-14 13:43:53

by Axel Lin

Subject: [PATCH] regulator: ab8500 - fix the logic to remove already registered regulators in error path

In current implementation, ab8500_regulator_info[0].regulator is not
unregistered if the error happen at i > 0.

This patch fixes the resource leak and also improves the readability.

Signed-off-by: Axel Lin <[email protected]>
---
I think this change is better ( in readability ) than simply change "while (i > 0)" to "while (i >= 0)".

drivers/regulator/ab8500.c | 4 +---
1 files changed, 1 insertions(+), 3 deletions(-)

diff --git a/drivers/regulator/ab8500.c b/drivers/regulator/ab8500.c
index dc3f1a4..775688e 100644
--- a/drivers/regulator/ab8500.c
+++ b/drivers/regulator/ab8500.c
@@ -368,11 +368,9 @@ static __devinit int ab8500_regulator_probe(struct platform_device *pdev)
dev_err(&pdev->dev, "failed to register regulator %s\n",
info->desc.name);
/* when we fail, un-register all earlier regulators */
- i--;
- while (i > 0) {
+ while (--i >= 0) {
info = &ab8500_regulator_info[i];
regulator_unregister(info->regulator);
- i--;
}
return err;
}
--
1.7.0.4

2010-08-14 14:35:25

by Sundar R IYER

Subject: RE: [PATCH] regulator: ab8500 - fix the logic to remove already registered regulators in error path

Thanx!

BTW, could you also include the AB3100 as well into this patch? I derived it from
the AB3100 implementation.

Linus, any comments?

Regards,
Sundar

>-----Original Message-----
>From: Axel Lin [mailto:[email protected]]
>Sent: Saturday, August 14, 2010 7:14 PM
>To: linux-kernel
>Cc: Liam Girdwood; Mark Brown; Sundar R IYER
>Subject: [PATCH] regulator: ab8500 - fix the logic to remove already registered
>regulators in error path
>
>In current implementation, ab8500_regulator_info[0].regulator is not
>unregistered if the error happen at i > 0.
>
>This patch fixes the resource leak and also improves the readability.
>
>Signed-off-by: Axel Lin <[email protected]>
>---
>I think this change is better ( in readability ) than simply change "while (i > 0)" to
>"while (i >= 0)".
>
> drivers/regulator/ab8500.c | 4 +---
> 1 files changed, 1 insertions(+), 3 deletions(-)
>
>diff --git a/drivers/regulator/ab8500.c b/drivers/regulator/ab8500.c
>index dc3f1a4..775688e 100644
>--- a/drivers/regulator/ab8500.c
>+++ b/drivers/regulator/ab8500.c
>@@ -368,11 +368,9 @@ static __devinit int ab8500_regulator_probe(struct
>platform_device *pdev)
> dev_err(&pdev->dev, "failed to register regulator %s\n",
> info->desc.name);
> /* when we fail, un-register all earlier regulators */
>- i--;
>- while (i > 0) {
>+ while (--i >= 0) {
> info = &ab8500_regulator_info[i];
> regulator_unregister(info->regulator);
>- i--;
> }
> return err;
> }
>--
>1.7.0.4
>
>

????{.n?+???????+%?????ݶ??w??{.n?+????{??G?????{ay?ʇڙ?,j??f???h?????????z_??(?階?ݢj"???m??????G????????????&???~???iO???z??v?^?m????????????I?

2010-08-14 14:39:13

by Sundar R IYER

Subject: RE: [PATCH] regulator: ab8500 - fix the logic to remove already registered regulators in error path

Oops, I didn’t see the below patch.

[PATCH] regulator: ab3100 - fix the logic to remove already registered regulators in error path

Sorry for the inconvenience!

Cheers!
Sundar

>-----Original Message-----
>From: Sundar R IYER
>Sent: Saturday, August 14, 2010 8:05 PM
>To: 'Axel Lin'; linux-kernel
>Cc: Liam Girdwood; Mark Brown; Linus WALLEIJ
>Subject: RE: [PATCH] regulator: ab8500 - fix the logic to remove already registered
>regulators in error path
>
>Thanx!
>
>BTW, could you also include the AB3100 as well into this patch? I derived it from
>the AB3100 implementation.
>
>Linus, any comments?
>
>Regards,
>Sundar
>
>>-----Original Message-----
>>From: Axel Lin [mailto:[email protected]]
>>Sent: Saturday, August 14, 2010 7:14 PM
>>To: linux-kernel
>>Cc: Liam Girdwood; Mark Brown; Sundar R IYER
>>Subject: [PATCH] regulator: ab8500 - fix the logic to remove already registered
>>regulators in error path
>>
>>In current implementation, ab8500_regulator_info[0].regulator is not
>>unregistered if the error happen at i > 0.
>>
>>This patch fixes the resource leak and also improves the readability.
>>
>>Signed-off-by: Axel Lin <[email protected]>
>>---
>>I think this change is better ( in readability ) than simply change "while (i > 0)" to
>>"while (i >= 0)".
>>
>> drivers/regulator/ab8500.c | 4 +---
>> 1 files changed, 1 insertions(+), 3 deletions(-)
>>
>>diff --git a/drivers/regulator/ab8500.c b/drivers/regulator/ab8500.c
>>index dc3f1a4..775688e 100644
>>--- a/drivers/regulator/ab8500.c
>>+++ b/drivers/regulator/ab8500.c
>>@@ -368,11 +368,9 @@ static __devinit int ab8500_regulator_probe(struct
>>platform_device *pdev)
>> dev_err(&pdev->dev, "failed to register regulator %s\n",
>> info->desc.name);
>> /* when we fail, un-register all earlier regulators */
>>- i--;
>>- while (i > 0) {
>>+ while (--i >= 0) {
>> info = &ab8500_regulator_info[i];
>> regulator_unregister(info->regulator);
>>- i--;
>> }
>> return err;
>> }
>>--
>>1.7.0.4
>>
>>

????{.n?+???????+%?????ݶ??w??{.n?+????{??G?????{ay?ʇڙ?,j??f???h?????????z_??(?階?ݢj"???m??????G????????????&???~???iO???z??v?^?m????????????I?

2010-08-15 15:42:51

by Mark Brown

Subject: Re: [PATCH] regulator: ab8500 - fix the logic to remove already registered regulators in error path

On Sat, Aug 14, 2010 at 09:44:04PM +0800, Axel Lin wrote:
> In current implementation, ab8500_regulator_info[0].regulator is not
> unregistered if the error happen at i > 0.
>
> This patch fixes the resource leak and also improves the readability.
>
> Signed-off-by: Axel Lin <[email protected]>

Acked-by: Mark Brown <[email protected]>

2010-08-15 15:55:44

by Linus Walleij

Subject: Re: [PATCH] regulator: ab8500 - fix the logic to remove already registered regulators in error path

2010/8/14 Sundar R IYER <[email protected]>:

> BTW, could you also include the AB3100 as well into this patch? I derived it from
> the AB3100 implementation.

Aha the same thing for AB8500, yeah I confess it is probably my coding
mistake replicated so
Acked-by: Linus Walleij <[email protected]>
on this one too then!

Thanks.
Linus Walleij

2010-08-18 13:29:42

by Liam Girdwood

Subject: Re: [PATCH] regulator: ab8500 - fix the logic to remove already registered regulators in error path

On Sat, 2010-08-14 at 21:44 +0800, Axel Lin wrote:
> In current implementation, ab8500_regulator_info[0].regulator is not
> unregistered if the error happen at i > 0.
>
> This patch fixes the resource leak and also improves the readability.
>
> Signed-off-by: Axel Lin <[email protected]>
> ---
> I think this change is better ( in readability ) than simply change "while (i > 0)" to "while (i >= 0)".
>
> drivers/regulator/ab8500.c | 4 +---
> 1 files changed, 1 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/regulator/ab8500.c b/drivers/regulator/ab8500.c
> index dc3f1a4..775688e 100644
> --- a/drivers/regulator/ab8500.c
> +++ b/drivers/regulator/ab8500.c
> @@ -368,11 +368,9 @@ static __devinit int ab8500_regulator_probe(struct platform_device *pdev)
> dev_err(&pdev->dev, "failed to register regulator %s\n",
> info->desc.name);
> /* when we fail, un-register all earlier regulators */
> - i--;
> - while (i > 0) {
> + while (--i >= 0) {
> info = &ab8500_regulator_info[i];
> regulator_unregister(info->regulator);
> - i--;
> }
> return err;
> }

Applied.

Thanks

Liam
--
Freelance Developer, SlimLogic Ltd
ASoC and Voltage Regulator Maintainer.
http://www.slimlogic.co.uk