Re: [PATCH 3/7] drm: Add bmc_attached flag to connector

[Jani Nikula <jani.nikula@linux.intel.com>]

On Fri, 11 Oct 2024, Thomas Zimmermann <tzimmermann@suse.de> wrote:
> Add the bmc_attached flag to struct drm_connector to signal the
> presence of a virtual BMC output. The connector reports to be in
> status connected even without a physically connected display. Fbcon
> or userspace compositors would otherwise stop displaying to the
> BMC.
>
> Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
> ---
>  drivers/gpu/drm/drm_probe_helper.c | 6 +++++-
>  include/drm/drm_connector.h        | 8 ++++++++
>  2 files changed, 13 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/drm_probe_helper.c b/drivers/gpu/drm/drm_probe_helper.c
> index df44be128e72..83c3f2d42d49 100644
> --- a/drivers/gpu/drm/drm_probe_helper.c
> +++ b/drivers/gpu/drm/drm_probe_helper.c
> @@ -605,7 +605,11 @@ int drm_helper_probe_single_connector_modes(struct drm_connector *connector,
>  			ret = connector_status_unknown;
>  
>  		connector->physical_status = ret;
> -		connector->status = connector->physical_status;
> +
> +		if (connector->bmc_attached)
> +			connector->status = connector_status_connected;
> +		else
> +			connector->status = connector->physical_status;

Perhaps all this would make more sense squashed into the previous patch?

>  	}
>  
>  	/*
> diff --git a/include/drm/drm_connector.h b/include/drm/drm_connector.h
> index 37e951f04ae8..ed360ae35f21 100644
> --- a/include/drm/drm_connector.h
> +++ b/include/drm/drm_connector.h
> @@ -1802,6 +1802,14 @@ struct drm_connector {
>  	 */
>  	bool ycbcr_420_allowed;
>  
> +	/**
> +	 * @ bmc_attached:
> +	 * The connector has a BMC transparently attached to it. It has to
> +	 * report a connected status, even without a physically connected
> +	 * display.
> +	 */
> +	bool bmc_attached;

Who is supposed to set this and when? Can it be changed during connector
life cycle? (I find myself wishing for C++ public/private members,
*gasp*!)

Again, there's bound to be people for whom "BMC" means nothing, and
therefore this member means nothing. I'm wondering if the name could be
something with a higher level of abstraction. In some ways, this is
related to or a special case of connector->force, but without the
uapi. You know, "logical force", with the use case being BMC.

BR,
Jani.


> +
>  	/**
>  	 * @registration_state: Is this connector initializing, exposed
>  	 * (registered) with userspace, or unregistered?

-- 
Jani Nikula, Intel