Re: [Intel-gfx] [drm-tip:drm-tip 4/8] drivers/gpu/drm/solomon/ssd130x.c:451:18: error: incomplete definition of type 'struct dma_buf_map'

[Javier Martinez Canillas <javierm@redhat.com>]

On 2/17/22 10:25, Lucas De Marchi wrote:
> On Thu, Feb 17, 2022 at 10:00:42AM +0100, Javier Martinez Canillas wrote:

[snip]

>>> this is now called iosys_map in drm-intel... drm-tip will need a fixup
>>> for the merge.
>>>
>>
>> I thought that the drm-intel tree was only for Intel DRM drivers changes and
>> subsystem wide changes should be merged through drm-mic ?
>>
>> Doing refactoring in that tree will likely lead to merge conflicts like this.
> 
> Yes, I know. My initial proposal was to split the rename and do it per
> branch to avoid this kind of situation, but it was requested to be done
> all in a single patch. Since I had other ~15 patches dependent on that
> one to be merged in drm-intel, it was agreed to do the rename via
> drm-intel. See 
> https://lore.kernel.org/lkml/e3813696-7b91-510c-987f-85ed2fd502d6@suse.de/
>

Got it. Thanks for the explanation.
 
> I guess the conflicts won't be that terrible and can be fixed as they
> show up.
>

Agreed.

>> Noticed your series in dri-devel but missed that already landed in drm-intel.
>>
>> The resolution should just be [0] right? If you confirm that then I can post
>> a proper patch to dri-devel.
>>
>>>>>> drivers/gpu/drm/solomon/ssd130x.c:451:18: error: incomplete definition of type 'struct dma_buf_map'
>>>>           void *vmap = map->vaddr; /* TODO: Use mapping abstraction properly */
>>>>                        ~~~^
>>>
>>> this shouldn't really be done.
>>>
>>
>> Yes, I know but asked what would be the proper way and didn't get an answer.
>> We have many drivers doing the same and I couldn't find one that was doing
>> it correctly to use as a reference:
>>
>> $ git grep "TODO: Use mapping abstraction properly" | wc -l
>> 15
>>
>> If you point me the proper way, I'll be happy to post a patch to change it.
> 
> It depends what you want to do with the address. There are APIs to copy
> from/to. I also added a few to read/write to an offset. It seems the
> problem here is that you need to pass that to a helper,
> drm_fb_xrgb8888_to_mono_reversed(). I think the proper solution would be
> to change the helper to accept an iosys_map* as argument rather than a
> void*.
>

That makes a lot of sense. Once the dust settles and your series land in
drm-misc-next, I can take a look at this and removing the TODO comment.

> Lucas De Marchi
>>>
>>> Lucas De Marchi
Best regards,
-- 
Javier Martinez Canillas
Linux Engineering
Red Hat