On 2/26/26 17:05, Tvrtko Ursulin wrote:
> On 26/02/2026 15:51, Sunil Khatri wrote:
>> update the type for num_syncobj_handles from __u64 to _u16 with
>> required padding.
>>
>> Signed-off-by: Sunil Khatri <[email protected]>
>> ---
>>   include/uapi/drm/amdgpu_drm.h | 4 +++-
>>   1 file changed, 3 insertions(+), 1 deletion(-)
>>
>> diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h
>> index 33e8738ce91f..42b4c0f6746b 100644
>> --- a/include/uapi/drm/amdgpu_drm.h
>> +++ b/include/uapi/drm/amdgpu_drm.h
>> @@ -479,7 +479,9 @@ struct drm_amdgpu_userq_signal {
>>        * @num_syncobj_handles: A count that represents the number of syncobj 
>> handles in
>>        * @syncobj_handles.
>>        */
>> -    __u64    num_syncobj_handles;
>> +    __u16    num_syncobj_handles;
> 
> Could probably cause build warnings/errors for something which builds fine 
> today. Hm on big endian it actually breaks completely.

Yeah this is a clear UAPI break, no go.


>> +    __u16    pad0;
>> +    __u32    pad1;
>>       /**
>>        * @bo_read_handles: The list of BO handles that the submitted user 
>> queue job
>>        * is using for read only. This will update BO fences in the kernel.
> 


-- 
Earthling Michel Dänzer       \        GNOME / Xwayland / Mesa developer
https://redhat.com             \               Libre software enthusiast

Reply via email to