linux-kernel - Re: [PATCH v1] watchdog: starfive: Fix the probe return error if PM and early

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <dc4190cd-5803-5d58-5421-a1df9c015ffc@starfivetech.com>
Date:   Wed, 26 Apr 2023 10:44:15 +0800
From:   Xingyu Wu <xingyu.wu@...rfivetech.com>
To:     Guenter Roeck <linux@...ck-us.net>,
        <linux-watchdog@...r.kernel.org>,
        "Wim Van Sebroeck" <wim@...ux-watchdog.org>
CC:     Samin Guo <samin.guo@...rfivetech.com>,
        <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v1] watchdog: starfive: Fix the probe return error if PM
 and early_enable are both disabled

On 2023/4/26 10:01, Guenter Roeck wrote:
> On 4/25/23 18:49, Xingyu Wu wrote:
>> On 2023/4/25 22:06, Guenter Roeck wrote:
>>> On 4/25/23 03:04, Xingyu Wu wrote:
>>>> When the starfive watchdog driver uses 'pm_runtime_put_sync()' as probe
>>>> return value at last and 'early_enable' is disabled, it could return the
>>>> error '-ENOSYS' if the CONFIG_PM is disabled, but the driver should works
>>>> normally.
>>>>
>>>> Add a check to make sure the PM is enabled and then use
>>>> 'pm_runtime_put_sync()' as return value when 'early_enable' is disabled.
>>>>
>>>> Fixes: db728ea9c7be ("drivers: watchdog: Add StarFive Watchdog driver")
>>>> Signed-off-by: Xingyu Wu <xingyu.wu@...rfivetech.com>
>>>> ---
>>>>
>>>> Hi, Guenter and Wim,
>>>>
>>>> This patch fixes the issue of StarFive watchdog driver and rebases on
>>>> the master branch of linux-next.
>>>>
>>>> Thanks.
>>>>    ---
>>>>    drivers/watchdog/starfive-wdt.c | 3 ++-
>>>>    1 file changed, 2 insertions(+), 1 deletion(-)
>>>>
>>>> diff --git a/drivers/watchdog/starfive-wdt.c b/drivers/watchdog/starfive-wdt.c
>>>> index 1995cceca51e..51e487e09960 100644
>>>> --- a/drivers/watchdog/starfive-wdt.c
>>>> +++ b/drivers/watchdog/starfive-wdt.c
>>>> @@ -492,7 +492,8 @@ static int starfive_wdt_probe(struct platform_device *pdev)
>>>>            goto err_exit;
>>>>          if (!early_enable)
>>>> -        return pm_runtime_put_sync(&pdev->dev);
>>>> +        if (pm_runtime_enabled(&pdev->dev))
>>>> +            return pm_runtime_put_sync(&pdev->dev);
>>>>    
>>>
>>> Why not just
>>>
>>>      if (!early_enable)
>>>          pm_runtime_put_sync(&pdev->dev)
>>>
>>> like almost every other caller of pm_runtime_put_sync() ?
>>>
>>
>> The function of pm_runtime_put_sync() is that:
>>
>> static inline int pm_runtime_put_sync(struct device *dev)
>> {
>>     return __pm_runtime_idle(dev, RPM_GET_PUT);
>> }
>>
>> and when do not enable CONFIG_PM, the __pm_runtime_idle() is that:
>>
>> static inline int __pm_runtime_idle(struct device *dev, int rpmflags)
>> {
>>     return -ENOSYS;
>> }
>>
>> If I do not open PM, it will return error saying probe failed
>> but it works fine without PM. I had tested that and probe really
>> was failed. So I should add this check before using pm_runtime_put_sync().
>>
> 
> You did not answer my question.
> 

Oh sorry, I did not notice that you dropped the 'return'. It looks simpler.

Best regards,
Xingyu Wu