[RFC PATCH 03/17] drm/exynos: dsi: Use the drm_panel_bridge API

[Jagan Teki]

On Mon, Jul 5, 2021 at 5:43 PM Marek Szyprowski
<m.szyprowski at samsung.com> wrote:
>
> On 05.07.2021 14:00, Jagan Teki wrote:
> > On Mon, Jul 5, 2021 at 5:18 PM Marek Szyprowski
> > <m.szyprowski at samsung.com> wrote:
> >> On 04.07.2021 11:02, Jagan Teki wrote:
> >>> Use drm_panel_bridge to replace manual panel and
> >>> bridge_chain handling code.
> >>>
> >>> This makes the driver simpler to allow all components
> >>> in the display pipeline to be treated as bridges by
> >>> cleaning the way to generic connector handling.
> >>>
> >>> Signed-off-by: Jagan Teki <jagan at amarulasolutions.com>
> >> This breaks Exysos DSI driver operation (Trats board worked fine with
> >> only patches 1-2):
> >>
> >> [    2.540066] exynos4-fb 11c00000.fimd: Adding to iommu group 0
> >> [    2.554733] OF: graph: no port node found in /soc/fimd at 11c00000
> >> [    2.602819] [drm] Exynos DRM: using 11c00000.fimd device for DMA
> >> mapping operations
> >> [    2.609649] exynos-drm exynos-drm: bound 11c00000.fimd (ops
> >> fimd_component_ops)
> >> [    2.632558] exynos-drm exynos-drm: failed to bind 11c80000.dsi (ops
> >> exynos_dsi_component_ops): -22
> >> [    2.642263] exynos-drm exynos-drm: master bind failed: -22
> >> [    2.651017] exynos-drm: probe of exynos-drm failed with error -22
> > Thanks for testing it.
> >
> > Can you check Squash of 3,4 or 3,4,5 will work or not?
>
> I've check both sets: 1-4 and 1-5 and none of them works. The result is
> same as above. If I remember correctly, last time when I played with
> that code, there was a problem with DRM core calling bridge ops in
> different order than when they are used by the Exynos DSI driver.

Okay. Let me check with sun6i-mipi-dsi as it is component_ops based.

Jagan.