All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH 00/13] fbdev locking rework and deferred setup, take 2
@ 2017-06-27 14:59 Daniel Vetter
  2017-06-27 14:59 ` [PATCH 01/13] drm/fb-helper: Push down modeset lock into FB helpers Daniel Vetter
                   ` (16 more replies)
  0 siblings, 17 replies; 43+ messages in thread
From: Daniel Vetter @ 2017-06-27 14:59 UTC (permalink / raw)
  To: DRI Development; +Cc: Daniel Vetter, Intel Graphics Development

Hi all,

Thanks to Liviu's help I realized that I fumbled the locking rework completely.
This one here should be better, but somehow I'm having a real bad day today and
I spent all day typing shit code, and then making it worse.

This here seems to work at first glance, but please test and review carefully.

Thanks a lot.

Cheers, Daniel

Daniel Vetter (9):
  drm/i915: Drop FBDEV #ifdev in mst code
  drm/fb-helper: Push locking in fb_is_bound
  drm/fb-helper: Drop locking from the vsync wait ioctl code
  drm/fb-helper: Push locking into pan_display_atomic|legacy
  drm/fb-helper: Push locking into restore_fbdev_mode_atomic|legacy
  drm/fb-helper: Stop using mode_config.mutex for internals
  drm/fb-helper: Split dpms handling into legacy and atomic paths
  drm/fb-helper: Support deferred setup
  drm/atomic-helper: Realign function parameters

Thierry Reding (4):
  drm/fb-helper: Push down modeset lock into FB helpers
  drm/fb-helper: Add top-level lock
  drm/exynos: Remove custom FB helper deferred setup
  drm/hisilicon: Remove custom FB helper deferred setup

 drivers/gpu/drm/drm_atomic_helper.c             |  24 +-
 drivers/gpu/drm/drm_fb_helper.c                 | 362 +++++++++++++++++-------
 drivers/gpu/drm/drm_vblank.c                    |   2 +-
 drivers/gpu/drm/exynos/exynos_drm_drv.c         |   6 +-
 drivers/gpu/drm/exynos/exynos_drm_fbdev.c       |  26 +-
 drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c |  21 +-
 drivers/gpu/drm/i915/intel_dp_mst.c             |  43 +--
 drivers/gpu/drm/i915/intel_fbdev.c              |  16 +-
 drivers/gpu/drm/radeon/radeon_dp_mst.c          |   7 -
 include/drm/drm_fb_helper.h                     |  42 ++-
 10 files changed, 345 insertions(+), 204 deletions(-)

-- 
2.11.0

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

^ permalink raw reply	[flat|nested] 43+ messages in thread
* [PATCH 10/12] drm/fb-helper: Support deferred setup
@ 2017-06-21 18:28 Daniel Vetter
  2017-06-23 13:31 ` [PATCH] " Daniel Vetter
  0 siblings, 1 reply; 43+ messages in thread
From: Daniel Vetter @ 2017-06-21 18:28 UTC (permalink / raw)
  To: DRI Development
  Cc: Daniel Vetter, Intel Graphics Development, Thierry Reding, John Stultz

From: Thierry Reding <treding@nvidia.com>

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.

v2: Rebase onto my entirely reworked fbdev helper locking. One big
difference is that this version again drops&reacquires the fbdev lock
(which is now fb_helper->lock, but before this patch series it was
mode_config->mutex), because register_framebuffer must be able to
recurse back into fbdev helper code for the initial screen setup.

Cc: John Stultz <john.stultz@linaro.org>
Cc: Thierry Reding <treding@nvidia.com> (v1)
Tested-by: John Stultz <john.stultz@linaro.org>
Signed-off-by: Thierry Reding <treding@nvidia.com> (v1)
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/drm_fb_helper.c | 111 ++++++++++++++++++++++++++++++----------
 include/drm/drm_fb_helper.h     |  23 +++++++++
 2 files changed, 107 insertions(+), 27 deletions(-)

diff --git a/drivers/gpu/drm/drm_fb_helper.c b/drivers/gpu/drm/drm_fb_helper.c
index e3d033df068f..a4cfef9bcfa7 100644
--- a/drivers/gpu/drm/drm_fb_helper.c
+++ b/drivers/gpu/drm/drm_fb_helper.c
@@ -521,6 +521,9 @@ int drm_fb_helper_restore_fbdev_mode_unlocked(struct drm_fb_helper *fb_helper)
 	if (!drm_fbdev_emulation)
 		return -ENODEV;
 
+	if (READ_ONCE(fb_helper->deferred_setup))
+		return 0;
+
 	mutex_lock(&fb_helper->lock);
 	ret = restore_fbdev_mode(fb_helper);
 
@@ -1693,6 +1696,23 @@ 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
@@ -2352,8 +2372,6 @@ 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->lock);
@@ -2431,6 +2449,61 @@ static void drm_setup_crtcs(struct drm_fb_helper *fb_helper,
 	kfree(enabled);
 }
 
+/* Note: Drops&reacquires fb_helper->lock */
+static int __drm_fb_helper_initial_config(struct drm_fb_helper *fb_helper,
+					  int bpp_sel)
+{
+	struct drm_device *dev = fb_helper->dev;
+	struct fb_info *info;
+	unsigned int width, height;
+	int ret;
+
+	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);
+	if (ret) {
+		mutex_unlock(&fb_helper->lock);
+		return ret;
+	}
+
+	fb_helper->deferred_setup = false;
+	mutex_unlock(&fb_helper->lock);
+
+	info = fb_helper->fbdev;
+	info->var.pixclock = 0;
+	ret = register_framebuffer(info);
+	if (ret < 0)
+		return ret;
+
+	dev_info(dev->dev, "fb%d: %s frame buffer device\n",
+		 info->node, info->fix.id);
+
+	mutex_lock(&kernel_fb_helper_lock);
+	if (list_empty(&kernel_fb_helper_list))
+		register_sysrq_key('v', &sysrq_drm_fb_helper_restore_op);
+
+	list_add(&fb_helper->kernel_fb_list, &kernel_fb_helper_list);
+
+	return 0;
+}
+
 /**
  * drm_fb_helper_initial_config - setup a sane initial connector configuration
  * @fb_helper: fb_helper device struct
@@ -2475,39 +2548,16 @@ 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;
-	struct fb_info *info;
 	int ret;
 
 	if (!drm_fbdev_emulation)
 		return 0;
 
 	mutex_lock(&fb_helper->lock);
-	drm_setup_crtcs(fb_helper,
-			dev->mode_config.max_width,
-			dev->mode_config.max_height);
-	ret = drm_fb_helper_single_fb_probe(fb_helper, bpp_sel);
-	mutex_unlock(&fb_helper->lock);
-	if (ret)
-		return ret;
-
-	info = fb_helper->fbdev;
-	info->var.pixclock = 0;
-	ret = register_framebuffer(info);
-	if (ret < 0)
-		return ret;
-
-	dev_info(dev->dev, "fb%d: %s frame buffer device\n",
-		 info->node, info->fix.id);
-
-	mutex_lock(&kernel_fb_helper_lock);
-	if (list_empty(&kernel_fb_helper_list))
-		register_sysrq_key('v', &sysrq_drm_fb_helper_restore_op);
-
-	list_add(&fb_helper->kernel_fb_list, &kernel_fb_helper_list);
+	ret = __drm_fb_helper_initial_config(fb_helper, bpp_sel);
 	mutex_unlock(&kernel_fb_helper_lock);
 
-	return 0;
+	return ret;
 }
 EXPORT_SYMBOL(drm_fb_helper_initial_config);
 
@@ -2540,6 +2590,13 @@ int drm_fb_helper_hotplug_event(struct drm_fb_helper *fb_helper)
 		return 0;
 
 	mutex_lock(&fb_helper->lock);
+	if (fb_helper->deferred_setup) {
+		err = __drm_fb_helper_initial_config(fb_helper,
+						     fb_helper->preferred_bpp);
+		mutex_unlock(&fb_helper->lock);
+		return err;
+	}
+
 	if (!fb_helper->fb || !drm_fb_helper_is_bound(fb_helper)) {
 		fb_helper->delayed_hotplug = true;
 		mutex_unlock(&fb_helper->lock);
diff --git a/include/drm/drm_fb_helper.h b/include/drm/drm_fb_helper.h
index ea170b96e88d..a5ea6ffdfecc 100644
--- a/include/drm/drm_fb_helper.h
+++ b/include/drm/drm_fb_helper.h
@@ -232,6 +232,29 @@ 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.
+	 *
+	 * Protected by @lock.
+	 */
+	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;
 };
 
 /**
-- 
2.11.0

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

^ permalink raw reply related	[flat|nested] 43+ messages in thread
end of thread, other threads:[~2017-07-04 15:16 UTC | newest]

Thread overview: 43+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-06-27 14:59 [PATCH 00/13] fbdev locking rework and deferred setup, take 2 Daniel Vetter
2017-06-27 14:59 ` [PATCH 01/13] drm/fb-helper: Push down modeset lock into FB helpers Daniel Vetter
2017-06-29  9:10   ` Maarten Lankhorst
2017-06-29  9:23     ` Daniel Vetter
2017-06-29  9:33       ` Maarten Lankhorst
2017-06-29  9:44         ` Daniel Vetter
2017-06-29 11:13           ` Maarten Lankhorst
2017-06-29 12:38             ` Daniel Vetter
2017-06-27 14:59 ` [PATCH 02/13] drm/i915: Drop FBDEV #ifdev in mst code Daniel Vetter
2017-06-27 14:59 ` [PATCH 03/13] drm/fb-helper: Add top-level lock Daniel Vetter
2017-06-27 14:59 ` [PATCH 04/13] drm/fb-helper: Push locking in fb_is_bound Daniel Vetter
2017-06-27 14:59 ` [PATCH 05/13] drm/fb-helper: Drop locking from the vsync wait ioctl code Daniel Vetter
2017-06-27 14:59 ` [PATCH 06/13] drm/fb-helper: Push locking into pan_display_atomic|legacy Daniel Vetter
2017-06-27 14:59 ` [PATCH 07/13] drm/fb-helper: Push locking into restore_fbdev_mode_atomic|legacy Daniel Vetter
2017-06-27 14:59 ` [PATCH 08/13] drm/fb-helper: Stop using mode_config.mutex for internals Daniel Vetter
2017-06-27 14:59 ` [PATCH 09/13] drm/fb-helper: Split dpms handling into legacy and atomic paths Daniel Vetter
2017-06-29 10:22   ` Maarten Lankhorst
2017-06-29 10:31     ` Daniel Vetter
2017-06-29 10:58       ` Maarten Lankhorst
2017-06-29 11:00         ` Daniel Vetter
2017-06-29 11:23           ` Maarten Lankhorst
2017-06-27 14:59 ` [PATCH 10/13] drm/fb-helper: Support deferred setup Daniel Vetter
2017-06-28 11:32   ` [PATCH] " Daniel Vetter
2017-06-28 16:24     ` Liviu Dudau
2017-06-29 10:59     ` Maarten Lankhorst
2017-06-29 12:36       ` Daniel Vetter
2017-06-30 16:51     ` [PATCH] drm/fb-helper: Restore first connection behaviour on " Liviu Dudau
2017-06-30 18:13       ` Daniel Vetter
2017-07-03  8:44         ` Liviu Dudau
2017-07-03 16:33           ` Daniel Vetter
2017-06-27 14:59 ` [PATCH 11/13] drm/exynos: Remove custom FB helper " Daniel Vetter
2017-06-27 14:59 ` [PATCH 12/13] drm/hisilicon: " Daniel Vetter
2017-06-28  9:08   ` [PATCH] " Daniel Vetter
2017-06-27 14:59 ` [PATCH 13/13] drm/atomic-helper: Realign function parameters Daniel Vetter
2017-06-27 15:01   ` Deucher, Alexander
2017-06-27 15:42   ` Harry Wentland
2017-07-04 15:16     ` Daniel Vetter
2017-06-27 15:30 ` ✓ Fi.CI.BAT: success for fbdev locking rework and deferred setup, take 2 Patchwork
2017-06-27 23:02 ` [PATCH 00/13] " John Stultz
2017-06-28  7:36   ` Daniel Vetter
2017-06-28  9:28 ` ✓ Fi.CI.BAT: success for fbdev locking rework and deferred setup, take 2 (rev2) Patchwork
2017-06-28 12:28 ` ✓ Fi.CI.BAT: success for fbdev locking rework and deferred setup, take 2 (rev3) Patchwork
  -- strict thread matches above, loose matches on Subject: below --
2017-06-21 18:28 [PATCH 10/12] drm/fb-helper: Support deferred setup Daniel Vetter
2017-06-23 13:31 ` [PATCH] " Daniel Vetter

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.