[PATCH 3/6] xinput.sys: Introduce new xinput.sys driver.

[Rémi Bernon]

On 8/27/21 7:56 PM, Zebediah Figura (she/her) wrote:
> On 8/26/21 12:59 AM, Rémi Bernon wrote:
>> +static WCHAR *query_instance_id(DEVICE_OBJECT *device)
>> +{
>> +    struct device_extension *ext = ext_from_DEVICE_OBJECT(device);
>> +    struct device_state *state = ext->state;
>> +    DWORD len = wcslen(state->instance_id);
>> +    WCHAR *dst;
>> +
>> +    if ((dst = ExAllocatePool(PagedPool, len * sizeof(WCHAR))))
>> +        wcscpy(dst, state->instance_id);
>> +
>> +    return dst;
>> +}
>> +
>> +static WCHAR *query_device_id(DEVICE_OBJECT *device)
>> +{
>> +    struct device_extension *ext = ext_from_DEVICE_OBJECT(device);
>> +    DWORD len = wcslen(ext->device_id);
>> +    WCHAR *dst;
>> +
>> +    if ((dst = ExAllocatePool(PagedPool, len * sizeof(WCHAR))))
>> +        wcscpy(dst, ext->device_id);
>> +
>> +    return dst;
>> +}
>> +
> 
> Missing space for a null terminator in these two.
> 
> In general these could also be memcpy(), although that's a bit of a 
> nitpick.
> 

Indeed.

>> +static WCHAR *query_hardware_ids(DEVICE_OBJECT *device)
>> +{
>> +    struct device_extension *ext = (struct device_extension 
>> *)device->DeviceExtension;
>> +    DWORD len = wcslen(ext->device_id);
>> +    WCHAR *dst;
>> +
>> +    if ((dst = ExAllocatePool(PagedPool, (len + 2) * sizeof(WCHAR))))
>> +    {
>> +        wcscpy(dst, ext->device_id);
>> +        dst[len + 1] = 0;
>> +    }
>> +
>> +    return dst;
>> +}
>> +
>> +static WCHAR *query_compatible_ids(DEVICE_OBJECT *device)
>> +{
>> +    DWORD len = wcslen(L"WINEBUS\\WINE_COMP_HID");
>> +    WCHAR *dst;
>> +
>> +    if ((dst = ExAllocatePool(PagedPool, (len + 2) * sizeof(WCHAR))))
>> +    {
>> +        wcscpy(dst, L"WINEBUS\\WINE_COMP_HID");
>> +        dst[len + 1] = 0;
>> +    }
>> +
>> +    return dst;
>> +}
>> +
> 
> Same comment from last time (on a different patch?) about defining a 
> static const WCHAR[] variable for that string.
> 
> Actually, it seems like quite a lot of feedback from 
> <https://www.winehq.org/pipermail/wine-devel/2021-August/193365.html> 
> has not been addressed, or responded to.
> 

Indeed, I missed them somehow... sorry.
-- 
Rémi Bernon <rbernon at codeweavers.com>