linux-kernel - Re: [PATCH] gnss: Use devm_regulator_get_enable

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: <458e18b0-cdf7-1a27-ed18-00dd6858615e@wanadoo.fr>
Date:   Tue, 20 Jun 2023 19:16:59 +0200
From:   Christophe JAILLET <christophe.jaillet@...adoo.fr>
To:     Johan Hovold <johan@...nel.org>
Cc:     Matthias Brugger <matthias.bgg@...il.com>,
        AngeloGioacchino Del Regno 
        <angelogioacchino.delregno@...labora.com>,
        Liam Girdwood <lgirdwood@...il.com>,
        Mark Brown <broonie@...nel.org>, linux-kernel@...r.kernel.org,
        kernel-janitors@...r.kernel.org,
        linux-arm-kernel@...ts.infradead.org,
        linux-mediatek@...ts.infradead.org
Subject: Re: [PATCH] gnss: Use devm_regulator_get_enable_optional()

Le 20/06/2023 à 10:57, Johan Hovold a écrit :
> On Sat, Jun 17, 2023 at 11:57:09AM +0200, Christophe JAILLET wrote:
>> Use devm_regulator_get_enable_optional() instead of hand writing it. It
>> saves some line of code.
>>
>> Signed-off-by: Christophe JAILLET <christophe.jaillet@...adoo.fr>
>> ---
>> Note that regulator_disable() is now called after gnss_serial_free() in
>> the error handling path of the probe and in the remove function, but it
>> looks harmless to me.
> 
> Yeah, that bit should be fine.
> 
>> ---
>>   drivers/gnss/mtk.c | 26 ++++----------------------
>>   1 file changed, 4 insertions(+), 22 deletions(-)
>>
>> diff --git a/drivers/gnss/mtk.c b/drivers/gnss/mtk.c
>> index c62b1211f4fe..d3d31295d4e0 100644
>> --- a/drivers/gnss/mtk.c
>> +++ b/drivers/gnss/mtk.c
>> @@ -17,7 +17,6 @@
>>   #include "serial.h"
>>   
>>   struct mtk_data {
>> -	struct regulator *vbackup;
>>   	struct regulator *vcc;
>>   };
>>   
>> @@ -87,30 +86,16 @@ static int mtk_probe(struct serdev_device *serdev)
>>   		goto err_free_gserial;
>>   	}
>>   
>> -	data->vbackup = devm_regulator_get_optional(&serdev->dev, "vbackup");
>> -	if (IS_ERR(data->vbackup)) {
>> -		ret = PTR_ERR(data->vbackup);
>> -		if (ret == -ENODEV)
>> -			data->vbackup = NULL;
>> -		else
>> -			goto err_free_gserial;
>> -	}
>> -
>> -	if (data->vbackup) {
>> -		ret = regulator_enable(data->vbackup);
>> -		if (ret)
>> -			goto err_free_gserial;
>> -	}
>> +	ret = devm_regulator_get_enable_optional(&serdev->dev, "vbackup");
>> +	if (ret)
>> +		goto err_free_gserial;
> 
> But this breaks the driver as the new helper still returns -ENODEV when
> the optional is resource is not present.

Ouch!

Thanks for the feedback. I urgently need to double check some other 
similar proposals I've made.

Sorry for the broken patch :(

CJ

> 
> Wolfram already suggested using this new helper here:
> 
> 	https://lore.kernel.org/lkml/20230523064310.3005-3-wsa+renesas@sang-engineering.com
> 
> and also got the error handling right even if that patch will require a
> respin for other reasons.
> 
> As I mentioned in my reply to Wolfram, I'm generally sceptical of
> helpers like this one, but in this case where there are no dependencies
> on other resources I guess it's ok.
> 
> Johan
>