[vlc-devel] [vlc-commits] video_output: let the deinterlacing code access vout_thread_interlacing_t

Steve Lhomme robux4 at ycbcr.xyz
Mon Nov 16 08:23:37 CET 2020


On 2020-11-13 19:28, Rémi Denis-Courmont wrote:
> Le vendredi 13 novembre 2020, 17:52:52 EET Steve Lhomme a écrit :
>> vlc | branch: master | Steve Lhomme <robux4 at ycbcr.xyz> | Mon Nov  9 09:44:36
>> 2020 +0100| [05acaa1e459ed2e1b0f163518e3248ba19bb23a3] | committer: Steve
>> Lhomme
>>
>> video_output: let the deinterlacing code access vout_thread_interlacing_t
>>
>> This is still private compared to the rest of the core. deinterlacing.c
>> contains the video_output logic to switch the deinterlacing mode.
>>
>> We don't pass the vout_thread_interlacing_t anymore as it can be deduced
>> from the vout_thread_t pointer.
> 
> I'm pertty sure I rejected this.

Here is your "rejection:

 >> I don't think this patch is necessary. The previous one should be 
sufficient.
 >> And TBH, it still looks needlessly ugly to me.
 >
 >This is needed by the next patch(es). The DeinterlaceCallback didn't
 >know about the vout_thread_interlacing_t/whatever it is now. And so it
 >couldn't set the values (under lock) on its own. That's what it was
 >using a variable in the vout object to write the new state to use.

It doesn't look like a rejection at all. And the patchset, in the end, 
achieves something important: applying the deinterlacing before the 
interlaced picture is displayed, nor after.

> Please revert.
> 
> -- 
> Rémi Denis-Courmont
> 
> 
> _______________________________________________
> vlc-devel mailing list
> To unsubscribe or modify your subscription options:
> https://mailman.videolan.org/listinfo/vlc-devel
> 


More information about the vlc-devel mailing list