Re: [PATCH 03/12] drm/msm/dpu: use format-related definitions from mdp_common.xml.h

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Thu, 11 Apr 2024 at 04:20, Dmitry Baryshkov
<dmitry.baryshkov@xxxxxxxxxx> wrote:
>
> On Thu, 11 Apr 2024 at 02:54, Abhinav Kumar <quic_abhinavk@xxxxxxxxxxx> wrote:
> >
> >
> >
> > On 4/10/2024 2:12 PM, Dmitry Baryshkov wrote:
> > > On Wed, Apr 10, 2024 at 01:18:42PM -0700, Abhinav Kumar wrote:
> > >>
> > >>
> > >> On 4/10/2024 1:16 PM, Dmitry Baryshkov wrote:
> > >>> On Wed, 10 Apr 2024 at 23:00, Abhinav Kumar <quic_abhinavk@xxxxxxxxxxx> wrote:
> > >>>>
> > >>>>
> > >>>>
> > >>>> On 12/2/2023 1:40 PM, Dmitry Baryshkov wrote:
> > >>>>> Instead of having DPU-specific defines, switch to the definitions from
> > >>>>> the mdp_common.xml.h file. This is the preparation for merged of DPU and
> > >>>>> MDP format tables.
> > >>>>>
> > >>>>
> > >>>> Adding MDP_***__ usages in DPU driver is quite confusing.
> > >>>>
> > >>>> Can we align to a common naming scheme such as DISP_***?
> > >>>
> > >>> No, it's not something display-generic. It is specific to MDP
> > >>> platforms. In the end DPU is a continuation of the MDP lineup, isn't
> > >>> it?
> > >>>
> > >>
> > >> No some aspects of the hw are completely different as you already know
> > >> between MDP4/MDP5 and DPU. Bringing back MDP usages into DPU does not seem
> > >> right.
> > >
> > > MDP4 is different, it's true. But there is a lot of common between MDP5
> > > and DPU. Frakly speaking, I don't see an issue with using the constant
> > > that was defined for MDP5 for DPU layer. Especially since we are also
> > > going to use mdp_ functions for format handling.
> > >
> >
> > All the HW naming etc in the doc has migrated to DPU and in fact it only
> > makes sense to start using DPU for MDP5 as we plan to move mdp5 targets
> > to DPU anyway. Not the other way around.
> >
> > MDP4 remains different.
> >
> > How about MSM_DISP then? I dont get why this is MDP platform specific.
>
> I expect MSM_DISP to be applicable to all MSM displays, even if e.g.
> at some point DPU2 switches colour component encoding.
>
> > Because the term MDP no longer holds true for DPU.
>
> The XML is still called mdp_common. And the functions are in the mdp_
> namespace. I don't think we should be changing them just because the
> name has changed.
> Likewise if MDP3 is not compatible with these definitions (to be
> honest, I didn't check) I still don't think we should change these
> names.
>
> > I am even looking for future chipsets. We cannot live with MDP5 names.
> > Have to think of generic names for formats.
>
> Ok, I'm open for suggestions from your side.

My proposal: MDPU? Neither MDP nor DPU.

-- 
With best wishes
Dmitry




[Index of Archives]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [Linux for Sparc]     [IETF Annouce]     [Security]     [Bugtraq]     [Linux MIPS]     [ECOS]     [Asterisk Internet PBX]     [Linux API]

  Powered by Linux