dri-devel Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Pekka Paalanen <ppaalanen@gmail.com>
To: Simon Ser <contact@emersion.fr>
Cc: dri-devel@lists.freedesktop.org, wayland-devel@lists.freedesktop.org
Subject: Re: [PATCH] drm/atomic: add quirks for blind save/restore
Date: Thu, 17 Nov 2022 10:35:14 +0200	[thread overview]
Message-ID: <20221117103514.47dedae9@eldfell> (raw)
In-Reply-To: <20221117075433.222968-1-contact@emersion.fr>

[-- Attachment #1: Type: text/plain, Size: 3053 bytes --]

On Thu, 17 Nov 2022 07:54:40 +0000
Simon Ser <contact@emersion.fr> wrote:

> Two quirks to make blind atomic save/restore [1] work correctly:
> 
> - Mark the DPMS property as immutable for atomic clients, since
>   atomic clients cannot change it.
> - Allow user-space to set content protection to "enabled", interpret
>   it as "desired".
> 
> [1]: https://gitlab.freedesktop.org/wlroots/wlroots/-/merge_requests/3794
> 
> Signed-off-by: Simon Ser <contact@emersion.fr>

That seems to be the sensible thing to do to these properties.

Acked-by: Pekka Paalanen <pekka.paalanen@collabora.com>

I'm really to happy see wlroots doing this. This has been in my mind to
do in Weston for years, but never had the chance to implement it yet. I
have just one little detail different: I would snapshot the "original"
KMS state at Weston start-up only once, and then use that snapshot as
the base every time Weston needs to sanitize the KMS state, like when
re-gaining DRM master. It's a bit more complicated, but it goes better
with the idea of resetting KMS state or starting from clean state, in
case KMS ever gains UAPI for that.


Thanks,
pq

> ---
> 
> I don't have the motivation to write IGT tests for this.
> 
>  drivers/gpu/drm/drm_atomic_uapi.c | 5 +++--
>  drivers/gpu/drm/drm_property.c    | 7 +++++++
>  2 files changed, 10 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/gpu/drm/drm_atomic_uapi.c b/drivers/gpu/drm/drm_atomic_uapi.c
> index c06d0639d552..95363aac7f69 100644
> --- a/drivers/gpu/drm/drm_atomic_uapi.c
> +++ b/drivers/gpu/drm/drm_atomic_uapi.c
> @@ -741,8 +741,9 @@ static int drm_atomic_connector_set_property(struct drm_connector *connector,
>  		state->scaling_mode = val;
>  	} else if (property == config->content_protection_property) {
>  		if (val == DRM_MODE_CONTENT_PROTECTION_ENABLED) {
> -			drm_dbg_kms(dev, "only drivers can set CP Enabled\n");
> -			return -EINVAL;
> +			/* Degrade ENABLED to DESIRED so that blind atomic
> +			 * save/restore works as intended. */
> +			val = DRM_MODE_CONTENT_PROTECTION_DESIRED;
>  		}
>  		state->content_protection = val;
>  	} else if (property == config->hdcp_content_type_property) {
> diff --git a/drivers/gpu/drm/drm_property.c b/drivers/gpu/drm/drm_property.c
> index dfec479830e4..dde42986f8cb 100644
> --- a/drivers/gpu/drm/drm_property.c
> +++ b/drivers/gpu/drm/drm_property.c
> @@ -474,7 +474,14 @@ int drm_mode_getproperty_ioctl(struct drm_device *dev,
>  		return -ENOENT;
>  
>  	strscpy_pad(out_resp->name, property->name, DRM_PROP_NAME_LEN);
> +
>  	out_resp->flags = property->flags;
> +	if (file_priv->atomic && property == dev->mode_config.dpms_property) {
> +		/* Quirk: indicate that the legacy DPMS property is not
> +		 * writable from atomic user-space, so that blind atomic
> +		 * save/restore works as intended. */
> +		out_resp->flags |= DRM_MODE_PROP_IMMUTABLE;
> +	}
>  
>  	value_count = property->num_values;
>  	values_ptr = u64_to_user_ptr(out_resp->values_ptr);


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

  reply	other threads:[~2022-11-17  8:35 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-11-17  7:54 [PATCH] drm/atomic: add quirks for blind save/restore Simon Ser
2022-11-17  8:35 ` Pekka Paalanen [this message]
2022-11-17 15:28 ` Daniel Vetter

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20221117103514.47dedae9@eldfell \
    --to=ppaalanen@gmail.com \
    --cc=contact@emersion.fr \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=wayland-devel@lists.freedesktop.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox