linux-kernel - Re: [PATCH 2/3] asus-wmi: Add dgpu disable method

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

Message-ID: <e933f74e-50d3-8de9-258a-a4000f3b6403@redhat.com>
Date:   Tue, 6 Jul 2021 12:17:13 +0200
From:   Hans de Goede <hdegoede@...hat.com>
To:     Barnabás Pőcze <pobrn@...tonmail.com>,
        "Luke D. Jones" <luke@...nes.dev>
Cc:     corentin.chary@...il.com, mgross@...ux.intel.com,
        jdelvare@...e.com, linux@...ck-us.net,
        platform-driver-x86@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 2/3] asus-wmi: Add dgpu disable method

Hi,

Barnabás made some good points which I missed.

See me reply inline.

On 7/5/21 2:47 AM, Barnabás Pőcze wrote:
> Hi
> 
> I have added a couple comments inline.
> 
> 
> 2021. július 5., hétfő 0:21 keltezéssel, Luke D. Jones írta:
> 

<snip>

>> +static ssize_t dgpu_disable_store(struct device *dev,
>> +				    struct device_attribute *attr,
>> +				    const char *buf, size_t count)
>> +{
>> +	int result;
>> +	u8 disable;
>> +	struct asus_wmi *asus = dev_get_drvdata(dev);
>> +
>> +	result = kstrtou8(buf, 10, &disable);
> 
> You could use `kstrtobool()`. I think that would be better since it accepts
> 'y', 'n', etc. in addition to 0 and 1.

Good point and the same applies to patch 1/3.

>> +	if (result < 0)
>> +		return result;
>> +
>> +	if (disable > 1 || disable < 0)
>> +		return -EINVAL;
>> +
>> +	asus->dgpu_disable_mode = disable;
>> +	/*
>> +	 * The ACPI call used does not save the mode unless the call is run twice.
>> +	 * Once to disable, then once to check status and save - this is two code
>> +	 * paths in the method in the ACPI dumps.
>> +	*/
>> +	dgpu_disable_write(asus);
>> +	dgpu_disable_write(asus);
> 
> Is there any reason the potential error codes are not returned?

Good question.

<snip>

>> @@ -2699,6 +2792,10 @@ static int asus_wmi_add(struct platform_device *pdev)
>>  	if (err)
>>  		goto fail_platform;
>>
>> +	err = dgpu_disable_check_present(asus);
>> +	if (err)
>> +		goto fail_dgpu_disable;
>> +
> 
> Should this really be considered a "fatal" error?

Well dgpu_disable_check_present() does already contain:

		if (err == -ENODEV)
			return 0;

IOW it only returns an error on unexpected errors and asus_wmi_add()
already contains a couple of other foo_present() checks which are
dealt with in the same way, so this is consistent with that and
being consistent is good, so I think this is fine.

Regards,

Hans