linux-kernel - Re: [PATCH 3/7] platform/x86: think-lmi: Use WMI bus API when accessing BIOS settings

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: <62e5b2f6-23e8-44e2-80f8-2509dd8e5a17@gmx.de>
Date: Fri, 14 Feb 2025 04:07:07 +0100
From: Armin Wolf <W_Armin@....de>
To: Ilpo Järvinen <ilpo.jarvinen@...ux.intel.com>
Cc: james@...iv.tech, markpearson@...ovo.com, jorge.lopez2@...com,
 jdelvare@...e.com, linux@...ck-us.net, linux-hwmon@...r.kernel.org,
 LKML <linux-kernel@...r.kernel.org>, Hans de Goede <hdegoede@...hat.com>,
 platform-driver-x86@...r.kernel.org, corbet@....net,
 linux-doc@...r.kernel.org
Subject: Re: [PATCH 3/7] platform/x86: think-lmi: Use WMI bus API when
 accessing BIOS settings

Am 13.02.25 um 14:17 schrieb Ilpo Järvinen:

> On Mon, 3 Feb 2025, Armin Wolf wrote:
>
>> Since the driver already binds to LENOVO_BIOS_SETTING_GUID, using
>> wmidev_block_query() inside tlmi_setting() allows for faster
>> access to BIOS settings.
>>
>> Signed-off-by: Armin Wolf <W_Armin@....de>
>> ---
>>   drivers/platform/x86/think-lmi.c | 23 +++++++++--------------
>>   drivers/platform/x86/think-lmi.h |  2 ++
>>   2 files changed, 11 insertions(+), 14 deletions(-)
>>
>> diff --git a/drivers/platform/x86/think-lmi.c b/drivers/platform/x86/think-lmi.c
>> index 2c94a4af9a1d..0fc275e461be 100644
>> --- a/drivers/platform/x86/think-lmi.c
>> +++ b/drivers/platform/x86/think-lmi.c
>> @@ -344,20 +344,14 @@ static int tlmi_opcode_setting(char *setting, const char *value)
>>   	return ret;
>>   }
>>
>> -static int tlmi_setting(int item, char **value, const char *guid_string)
>> +static int tlmi_setting(struct wmi_device *wdev, int item, char **value)
>>   {
>> -	struct acpi_buffer output = { ACPI_ALLOCATE_BUFFER, NULL };
>>   	union acpi_object *obj;
>> -	acpi_status status;
>>   	int ret;
>>
>> -	status = wmi_query_block(guid_string, item, &output);
>> -	if (ACPI_FAILURE(status))
>> -		return -EIO;
>> -
>> -	obj = output.pointer;
>> +	obj = wmidev_block_query(wdev, item);
>>   	if (!obj)
>> -		return -ENODATA;
>> +		return -EIO;
> Hi Armin,
>
> I'm trying to understand why there are these back and forth changes in the
> error code.
>
> It almost looks to me like wmidev_block_query() would want to return the
> error code itself because after you abstracted this code using
> wmidev_block_query(), you had to change the error code because you no
> longer have access to the key detail to decide which error code should be
> returned. That is, use ERR_PTR() inside wmidev_block_query() and the
> callers should just pass that error code on with IS_ERR & friends?
>
Hi,

the reason why wmidev_block_query() only returns NULL in case of an error is that
according to the WMI-ACPI specification, querying a WMI data block should return data.

So we have two error scenarios:

- ACPI error => firmware error => EIO
- no data returned => violation of firmware spec => EIO

Because of this always returning EIO is the correct approach in my opinion.

Thanks,
Armin Wolf