linux-kernel - Re: [PATCH 1/7] Input: mtk-pmic-keys - check for NULL on of_match

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]

Date:   Thu, 29 Apr 2021 16:48:55 +0200
From:   Mattijs Korpershoek <mkorpershoek@...libre.com>
To:     Dmitry Torokhov <dmitry.torokhov@...il.com>
Cc:     Matthias Brugger <matthias.bgg@...il.com>, fparent@...libre.com,
        linux-input@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
        linux-mediatek@...ts.infradead.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/7] Input: mtk-pmic-keys - check for NULL on
 of_match_device()

Hi Dmitry,

Mattijs Korpershoek <mkorpershoek@...libre.com> writes:

> Hi Dmitry,
>
> Dmitry Torokhov <dmitry.torokhov@...il.com> writes:
>
>> Hi Mattijs,
>>
>> On Wed, Apr 28, 2021 at 06:42:13PM +0200, Mattijs Korpershoek wrote:
>>> mtk-pmic-keys being a child device of mt6397, it will always get probed
>>> when mt6397_probe() is called.
>>> 
>>> This also happens when we have no device tree node matching
>>> mediatek,mt6397-keys.
>>
>> It sounds for me that creating a platform device instance in case where
>> we know need OF node, but do not have one, is wasteful. Can
>> mt6397-core.c and/or MFD core be adjusted to not do that.
>
> You are right. Maybe I can fix MFD core instead. I will look into it.
>
> Thanks for your review.
>>
>>> 
>>> In that case, the mfd core warns us:
>>> 
>>>   [    0.352175] mtk-pmic-keys: Failed to locate of_node [id: -1]
>>> 
>>> Check return value from call to of_match_device()
>>> in order to prevent a NULL pointer dereference.
>>> 
>>> In case of NULL print error message and return -ENODEV
>>> 
>>> Signed-off-by: Mattijs Korpershoek <mkorpershoek@...libre.com>
>>> ---
>>>  drivers/input/keyboard/mtk-pmic-keys.c | 3 +++
>>>  1 file changed, 3 insertions(+)
>>> 
>>> diff --git a/drivers/input/keyboard/mtk-pmic-keys.c b/drivers/input/keyboard/mtk-pmic-keys.c
>>> index 62391d6c7da6..12c449eed026 100644
>>> --- a/drivers/input/keyboard/mtk-pmic-keys.c
>>> +++ b/drivers/input/keyboard/mtk-pmic-keys.c
>>> @@ -247,6 +247,9 @@ static int mtk_pmic_keys_probe(struct platform_device *pdev)
>>>  	const struct of_device_id *of_id =
>>>  		of_match_device(of_mtk_pmic_keys_match_tbl, &pdev->dev);
>>>  
>>> +	if (!of_id)
>>> +		return -ENODEV;
>>> +
>>
>> So if we make MFD/6396 core smarter we would not be needing this. I
>> guess there is still a possibility of someone stuffing "mtk-pmic-keys"
>> into "driver_override" attribute of a random platform device but I
>> wonder if we really need to take care of such scenarios...
It turns out it was possible to make 6397-core smarter.
I've submitted [1] to replace this patch.

Thanks again for your suggestion. Please let me know if I should add
your Suggested-by: in [1].

[1] https://patchwork.kernel.org/project/linux-mediatek/patch/20210429143811.2030717-1-mkorpershoek@baylibre.com/ instead
>>
>>>  	keys = devm_kzalloc(&pdev->dev, sizeof(*keys), GFP_KERNEL);
>>>  	if (!keys)
>>>  		return -ENOMEM;
>>> -- 
>>> 2.27.0
>>> 
>>
>> Thanks.
>>
>> -- 
>> Dmitry