Re: [PATCH 9/9] modpost: Handle malformed WMI GUID strings

[Armin Wolf <W_Armin@gmx.de>]

Am 09.03.26 um 17:07 schrieb Mario Limonciello:

>
>
> On 3/7/2026 6:25 PM, Armin Wolf wrote:
>> Some WMI GUIDs found inside binary MOF files contain both
>> uppercase and lowercase characters. Blindly copying such
>> GUIDs will prevent the associated WMI driver from loading
>> automatically because the WMI GUID found inside WMI device ids
>> always contains uppercase characters.
>>
>> Avoid this issue by always converting WMI GUID strings to
>> uppercase. Also verify that the WMI GUID string actually looks
>> like a valid GUID.
>>
>> Signed-off-by: Armin Wolf <W_Armin@gmx.de>
>> ---
>>   .../wmi/driver-development-guide.rst          |  2 +-
>>   scripts/mod/file2alias.c                      | 28 ++++++++++++++++++-
>>   2 files changed, 28 insertions(+), 2 deletions(-)
>>
>> diff --git a/Documentation/wmi/driver-development-guide.rst 
>> b/Documentation/wmi/driver-development-guide.rst
>> index fbc2d9b12fe9..74bb156ad9cc 100644
>> --- a/Documentation/wmi/driver-development-guide.rst
>> +++ b/Documentation/wmi/driver-development-guide.rst
>> @@ -54,7 +54,7 @@ to matching WMI devices using a struct 
>> wmi_device_id table:
>>   ::
>>       static const struct wmi_device_id foo_id_table[] = {
>> -         /* Only use uppercase letters! */
>> +         /* Using only uppercase letters is recommended */
>>            { "936DA01F-9ABD-4D9D-80C7-02AF85C822A8", NULL },
>>            { }
>>     };
>> diff --git a/scripts/mod/file2alias.c b/scripts/mod/file2alias.c
>> index 4e99393a35f1..20e542a888c4 100644
>> --- a/scripts/mod/file2alias.c
>> +++ b/scripts/mod/file2alias.c
>> @@ -1253,6 +1253,8 @@ static void do_tee_entry(struct module *mod, 
>> void *symval)
>>   static void do_wmi_entry(struct module *mod, void *symval)
>>   {
>>       DEF_FIELD_ADDR(symval, wmi_device_id, guid_string);
>> +    char result[sizeof(*guid_string)];
>> +    int i;
>>         if (strlen(*guid_string) != UUID_STRING_LEN) {
>>           warn("Invalid WMI device id 'wmi:%s' in '%s'\n",
>> @@ -1260,7 +1262,31 @@ static void do_wmi_entry(struct module *mod, 
>> void *symval)
>>           return;
>>       }
>>   -    module_alias_printf(mod, false, WMI_MODULE_PREFIX "%s", 
>> *guid_string);
>> +    for (i = 0; i < UUID_STRING_LEN; i++) {
>> +        char value = (*guid_string)[i];
>> +        bool valid = false;
>> +
>> +        if (i == 8 || i == 13 || i == 18 || i == 23) {
>> +            if (value == '-')
>> +                valid = true;
>> +        } else {
>> +            if (isxdigit(value))
>> +                valid = true;
>> +        }
>> +
>> +        if (!valid) {
>> +            warn("Invalid character %c inside WMI GUID string '%s' 
>> in '%s'\n",
>> +                 value, *guid_string, mod->name);
>> +            return;
>> +        }
>> +
>> +        /* Some GUIDs from BMOF definitions contain lowercase 
>> characters */
>> +        result[i] = toupper(value);
>> +    }
>
> Minor logic change that could drop the boolean variable in the for loop:
>
> for (i = 0; i < UUID_STRING_LEN; i++) {
>     char value = (*guid_string)[i];
>
>     if (isxdigit(value)) {
>         result[i] = toupper(value);
>         continue;
>     }

This would not catch invalid GUID strings containing only digits.

Thanks,
Armin Wolf

>
>     if (value == '-' && (i == 8 || i == 13 || i == 18 || i == 23)) {
>         result[i] = value;
>         continue;
>     }
>
>     warn("Invalid character %c inside WMI GUID string '%s' in '%s'\n",
>          value, *guid_string, mod->name);
>     return;
> }
>
>> +
>> +    result[i] = '\0';
>> +
>> +    module_alias_printf(mod, false, WMI_MODULE_PREFIX "%s", result);
>>   }
>>     /* Looks like: mhi:S */
>
>