Re: [PATCH v8 16/23] drm/mediatek: mtk_hdmi: Split driver and add common probe function

[CK Hu (胡俊光)]

On Mon, 2025-04-14 at 10:52 +0200, AngeloGioacchino Del Regno wrote:
> External email : Please do not click links or open attachments until you have verified the sender or the content.
> 
> 
> Il 11/04/25 10:18, CK Hu (胡俊光) ha scritto:
> > On Wed, 2025-04-09 at 15:13 +0200, AngeloGioacchino Del Regno wrote:
> > > External email : Please do not click links or open attachments until you have verified the sender or the content.
> > > 
> > > 
> > > In preparation for adding a new driver for the HDMI TX v2 IP,
> > > split out the functions that will be common between the already
> > > present mtk_hdmi (v1) driver and the new one.
> > > 
> > > Since the probe flow for both drivers is 90% similar, add a common
> > > probe function that will be called from each driver's .probe()
> > > callback, avoiding lots of code duplication.
> > > 
> > > Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@xxxxxxxxxxxxx>
> > > ---
> > >   drivers/gpu/drm/mediatek/Kconfig           |  11 +-
> > >   drivers/gpu/drm/mediatek/Makefile          |   1 +
> > >   drivers/gpu/drm/mediatek/mtk_hdmi.c        | 542 +--------------------
> > >   drivers/gpu/drm/mediatek/mtk_hdmi_common.c | 426 ++++++++++++++++
> > >   drivers/gpu/drm/mediatek/mtk_hdmi_common.h | 188 +++++++
> > >   5 files changed, 633 insertions(+), 535 deletions(-)
> > >   create mode 100644 drivers/gpu/drm/mediatek/mtk_hdmi_common.c
> > >   create mode 100644 drivers/gpu/drm/mediatek/mtk_hdmi_common.h
> > > 
> > > diff --git a/drivers/gpu/drm/mediatek/Kconfig b/drivers/gpu/drm/mediatek/Kconfig
> > > index e47debd60619..994b48b82d44 100644
> > > --- a/drivers/gpu/drm/mediatek/Kconfig
> > > +++ b/drivers/gpu/drm/mediatek/Kconfig
> > > @@ -30,9 +30,18 @@ config DRM_MEDIATEK_DP
> > >          help
> > >            DRM/KMS Display Port driver for MediaTek SoCs.
> > > 
> > > +config DRM_MEDIATEK_HDMI_COMMON
> > > +       tristate
> > > +       depends on DRM_MEDIATEK
> > > +       select DRM_DISPLAY_HDMI_HELPER
> > > +       select DRM_DISPLAY_HELPER
> > > +       select SND_SOC_HDMI_CODEC if SND_SOC
> > > +       help
> > > +         MediaTek SoC HDMI common library
> > > +
> > >   config DRM_MEDIATEK_HDMI
> > >          tristate "DRM HDMI Support for Mediatek SoCs"
> > >          depends on DRM_MEDIATEK
> > > -       select SND_SOC_HDMI_CODEC if SND_SOC
> > > +       select DRM_MEDIATEK_HDMI_COMMON
> > >          help
> > >            DRM/KMS HDMI driver for Mediatek SoCs
> > > diff --git a/drivers/gpu/drm/mediatek/Makefile b/drivers/gpu/drm/mediatek/Makefile
> > > index 43afd0a26d14..78cf2d4fc85f 100644
> > > --- a/drivers/gpu/drm/mediatek/Makefile
> > > +++ b/drivers/gpu/drm/mediatek/Makefile
> > > @@ -21,6 +21,7 @@ mediatek-drm-y := mtk_crtc.o \
> > > 
> > >   obj-$(CONFIG_DRM_MEDIATEK) += mediatek-drm.o
> > > 
> > > +obj-$(CONFIG_DRM_MEDIATEK_HDMI_COMMON) += mtk_hdmi_common.o
> > 
> > The whole patch looks good to me.
> > But make common part to a module looks a little bit more modularized.
> > Let things to be simple, I would like the whole MediaTek hdmi driver be a single module.
> > For MediaTek drm driver, it is already broken to mediatek-drm, hdmi, dp modules.
> > Maybe someday dsi or dpi would be broken to modules.
> > So I would like hdmi to be a single module which include v1, v2 and common part.
> > If someday we need to optimize code size, then send patch to break hdmi module.
> > 
> 
> Hello CK,
> 
> The HDMIv1 and HDMIv2 drivers are handling two *very* different IPs that can never
> be *both* present at the same time on one SoC: having one Kconfig for both would
> not only add unnecessary bloat (especially on older SoCs!!), but would actually be
> wrong.
> 
> Also, please be aware that old LK bootloaders (smartphones/tablets) have a very
> limited RAM carveout for kernel + ramdisk that should be around 8MB if I recall
> correctly.
> 
> Remember as well that these devices *cannot* run different bootloaders as those
> are only able to run *OEM signed* binaries in the early bootchain, so replacing
> the bootloader is not a viable solution.
> 
> It's true, people can (and most probably want to and will) build this as a module
> but, in my opinion, freedom should be given to have the HDMI driver built-in even
> in these cases - and that's another good reason to keep them split.
> 
> This is done to avoid being hostile to old platforms, in full upstream spirit.

OK, it seems there is some case need to shrink code size, so

Reviewed-by: CK Hu <ck.hu@xxxxxxxxxxxx>

> 
> Regards,
> Angelo
> 
> > Regards,
> > CK
> > 
> > >   obj-$(CONFIG_DRM_MEDIATEK_HDMI) += mtk_cec.o
> > >   obj-$(CONFIG_DRM_MEDIATEK_HDMI) += mtk_hdmi.o
> > >   obj-$(CONFIG_DRM_MEDIATEK_HDMI) += mtk_hdmi_ddc.o
> > > diff --git a/drivers/gpu/drm/mediatek/mtk_hdmi.c b/drivers/gpu/drm/mediatek/mtk_hdmi.c
> > 
>