On Tue, Mar 21, 2017 at 09:13:54AM +0100, Thierry Reding wrote:
> From: Thierry Reding <treding@xxxxxxxxxx>
>
> FB helper code falls back to a 1024x768 mode if no outputs are connected
> or don't report back any modes upon initialization. This can be annoying
> because outputs that are added to FB helper later on can't be used with
> FB helper if they don't support a matching mode.
>
> The fallback is in place because VGA connectors can happen to report an
> unknown connection status even when they are in fact connected.
>
> Some drivers have custom solutions in place to defer FB helper setup
> until at least one output is connected. But the logic behind these
> solutions is always the same and there is nothing driver-specific about
> it, so a better alterative is to fix the FB helper core and add support
> for all drivers automatically.
>
> This patch adds support for deferred FB helper setup. It checks all the
> connectors for their connection status, and if all of them report to be
> disconnected marks the FB helper as needing deferred setup. Whet setup
> is deferred, the FB helper core will automatically retry setup after a
> hotplug event, and it will keep trying until it succeeds.
>
> Tested-by: John Stultz <john.stultz@xxxxxxxxxx>
> Signed-off-by: Thierry Reding <treding@xxxxxxxxxx>
Ok 2nd attempt at making this work, probably easier to go back to v2.
> ---
> drivers/gpu/drm/drm_fb_helper.c | 60 +++++++++++++++++++++++++++++++++++++----
> include/drm/drm_fb_helper.h | 21 +++++++++++++++
> 2 files changed, 76 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/gpu/drm/drm_fb_helper.c b/drivers/gpu/drm/drm_fb_helper.c
> index 9060adcf7cf8..d4a2c97d8b02 100644
> --- a/drivers/gpu/drm/drm_fb_helper.c
> +++ b/drivers/gpu/drm/drm_fb_helper.c
> @@ -511,6 +511,9 @@ int drm_fb_helper_restore_fbdev_mode_unlocked(struct drm_fb_helper *fb_helper)
> if (!drm_fbdev_emulation)
> return -ENODEV;
>
> + if (fb_helper->deferred_setup)
> + return 0;
Please wrap in READ_ONCE to make it clear we're doing lockless checking
here.
> +
> mutex_lock(&fb_helper->lock);
> drm_modeset_lock_all(dev);
>
> @@ -1597,6 +1600,23 @@ int drm_fb_helper_pan_display(struct fb_var_screeninfo *var,
> }
> EXPORT_SYMBOL(drm_fb_helper_pan_display);
>
> +static bool drm_fb_helper_maybe_connected(struct drm_fb_helper *helper)
> +{
> + bool connected = false;
> + unsigned int i;
> +
> + for (i = 0; i < helper->connector_count; i++) {
> + struct drm_fb_helper_connector *fb = helper->connector_info[i];
> +
> + if (fb->connector->status != connector_status_disconnected) {
> + connected = true;
> + break;
> + }
> + }
> +
> + return connected;
> +}
> +
> /*
> * Allocates the backing storage and sets up the fbdev info structure through
> * the ->fb_probe callback and then registers the fbdev and sets up the panic
> @@ -2254,8 +2274,6 @@ static void drm_setup_crtcs(struct drm_fb_helper *fb_helper,
> int i;
>
> DRM_DEBUG_KMS("\n");
> - if (drm_fb_helper_probe_connector_modes(fb_helper, width, height) == 0)
> - DRM_DEBUG_KMS("No connectors reported connected with modes\n");
>
> /* prevent concurrent modification of connector_count by hotplug */
> lockdep_assert_held(&fb_helper->dev->mode_config.mutex);
> @@ -2378,6 +2396,7 @@ static void drm_setup_crtcs(struct drm_fb_helper *fb_helper,
> int drm_fb_helper_initial_config(struct drm_fb_helper *fb_helper, int bpp_sel)
> {
> struct drm_device *dev = fb_helper->dev;
> + unsigned int width, height;
> struct fb_info *info;
> int ret;
>
> @@ -2385,14 +2404,34 @@ int drm_fb_helper_initial_config(struct drm_fb_helper *fb_helper, int bpp_sel)
> return 0;
>
>From here ...
> mutex_lock(&dev->mode_config.mutex);
> - drm_setup_crtcs(fb_helper,
> - dev->mode_config.max_width,
> - dev->mode_config.max_height);
> +
> + width = dev->mode_config.max_width;
> + height = dev->mode_config.max_height;
> +
> + if (drm_fb_helper_probe_connector_modes(fb_helper, width, height) == 0)
> + DRM_DEBUG_KMS("No connectors reported connected with modes\n");
> +
> + /*
> + * If everything's disconnected, there's no use in attempting to set
> + * up fbdev.
> + */
> + if (!drm_fb_helper_maybe_connected(fb_helper)) {
> + DRM_INFO("No outputs connected, deferring setup\n");
> + fb_helper->preferred_bpp = bpp_sel;
> + fb_helper->deferred_setup = true;
> + mutex_unlock(&dev->mode_config.mutex);
> + return 0;
> + }
> +
> + drm_setup_crtcs(fb_helper, width, height);
> +
> ret = drm_fb_helper_single_fb_probe(fb_helper, bpp_sel);
> mutex_unlock(&dev->mode_config.mutex);
> if (ret)
> return ret;
>
> + fb_helper->deferred_setup = false;
To here you need to hold the new fb_helper->mutex. Plus at the top, after
you've grabbed the lock, you need to re-check ->deferred_setup. This way
we guaranteed that only 1 thread can ever do the deferred setup, while not
holding the lock while we call register_framebuffer. If you hold any fbdev
or kms lock while calling register_framebuffer you will deadlock.
> +
> info = fb_helper->fbdev;
> info->var.pixclock = 0;
> ret = register_framebuffer(info);
> @@ -2437,11 +2476,16 @@ EXPORT_SYMBOL(drm_fb_helper_initial_config);
> int drm_fb_helper_hotplug_event(struct drm_fb_helper *fb_helper)
> {
> struct drm_device *dev = fb_helper->dev;
> + unsigned int width, height;
> int err = 0;
>
> if (!drm_fbdev_emulation)
> return 0;
>
> + if (fb_helper->deferred_setup)
Same here, please wrap in READ_ONCE, this is just a fastpath check.
With those changes your prep patch to roll out fb_helper->mutex more
shouldn't be needed at all.
Cheers, Daniel
> + return drm_fb_helper_initial_config(fb_helper,
> + fb_helper->preferred_bpp);
> +
> mutex_lock(&fb_helper->lock);
> mutex_lock(&dev->mode_config.mutex);
>
> @@ -2453,6 +2497,12 @@ int drm_fb_helper_hotplug_event(struct drm_fb_helper *fb_helper)
>
> DRM_DEBUG_KMS("\n");
>
> + width = dev->mode_config.max_width;
> + height = dev->mode_config.max_height;
> +
> + if (drm_fb_helper_probe_connector_modes(fb_helper, width, height) == 0)
> + DRM_DEBUG_KMS("No connectors reported connected with modes\n");
> +
> drm_setup_crtcs(fb_helper, fb_helper->fb->width, fb_helper->fb->height);
>
> mutex_unlock(&dev->mode_config.mutex);
> diff --git a/include/drm/drm_fb_helper.h b/include/drm/drm_fb_helper.h
> index 33771d0c44e3..41c3587751f8 100644
> --- a/include/drm/drm_fb_helper.h
> +++ b/include/drm/drm_fb_helper.h
> @@ -222,6 +222,27 @@ struct drm_fb_helper {
> * needs to be reprobe when fbdev is in control again.
> */
> bool delayed_hotplug;
> +
> + /**
> + * @deferred_setup:
> + *
> + * If no outputs are connected (disconnected or unknown) the FB helper
> + * code will defer setup until at least one of the outputs shows up.
> + * This field keeps track of the status so that setup can be retried
> + * at every hotplug event until it succeeds eventually.
> + */
> + bool deferred_setup;
> +
> + /**
> + * @preferred_bpp:
> + *
> + * Temporary storage for the driver's preferred BPP setting passed to
> + * FB helper initialization. This needs to be tracked so that deferred
> + * FB helper setup can pass this on.
> + *
> + * See also: @deferred_setup
> + */
> + int preferred_bpp;
> };
>
> /**
> --
> 2.12.0
>
> _______________________________________________
> dri-devel mailing list
> dri-devel@xxxxxxxxxxxxxxxxxxxxx
> https://lists.freedesktop.org/mailman/listinfo/dri-devel
--
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
dri-devel mailing list
dri-devel@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/dri-devel
