All of lore.kernel.org
 help / color / mirror / Atom feed
From: Daniel Vetter <daniel.vetter@ffwll.ch>
To: Intel Graphics Development <intel-gfx@lists.freedesktop.org>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Subject: [PATCH] drm/i915: compute masks of crtcs affected in set_mode
Date: Wed, 29 Aug 2012 12:34:04 +0200	[thread overview]
Message-ID: <1346236444-31640-1-git-send-email-daniel.vetter@ffwll.ch> (raw)
In-Reply-To: <1345403595-9678-51-git-send-email-daniel.vetter@ffwll.ch>

This is definetely a bit more generic than currently required, but
if we keep track of all crtcs that need to be disabled/enable (because
they loose an encoder or something similar), crtcs that get completely
disabled and those that we need to do an actual mode change on nicely
prepares us for global modeset operations on multiple crtcs.

The only big thing missing here would be a global resource allocation
step (for e.g. pch plls), which would equally frob these bitmasks if
e.g. a crtc only needs a new pll.

These masks aren't yet put to use in this patch, this will follow in the
next one.

v2-v5: Fix up the computations for good (hopefully).

v6: Fixup a confusion reported by Damien Lespiau: I've conserved the
(imo braindead) behaviour of the crtc helper to disable _any_
disconnected outputs if we do a modeset, even when that newly disabled
connector isn't connected to the crtc being changed by the modeset.

The effect of that is that we could disable an arbitrary number of
unrelated crtcs, which I haven't taken into account when writing this
code. Fix this up.

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
---
 drivers/gpu/drm/i915/intel_display.c | 96 ++++++++++++++++++++++++++++++++++++
 1 file changed, 96 insertions(+)

diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index c59569e..8f71957 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -6706,6 +6706,98 @@ fail:
 	return ERR_PTR(-EINVAL);
 }
 
+/* Computes which crtcs are affected and sets the relevant bits in the mask. For
+ * simplicity we use the crtc's pipe number (because it's easier to obtain). */
+static void
+intel_modeset_affected_pipes(struct drm_crtc *crtc, unsigned *modeset_pipes,
+			     unsigned *prepare_pipes, unsigned *disable_pipes)
+{
+	struct intel_crtc *intel_crtc;
+	struct drm_device *dev = crtc->dev;
+	struct intel_encoder *encoder;
+	struct intel_connector *connector;
+	struct drm_crtc *tmp_crtc;
+
+	*disable_pipes = *modeset_pipes = *prepare_pipes = 0;
+
+	/* Check which crtcs have changed outputs connected to them, these need
+	 * to be part of the prepare_pipes mask. We don't (yet) support global
+	 * modeset across multiple crtcs, so modeset_pipes will only have one
+	 * bit set at most. */
+	list_for_each_entry(connector, &dev->mode_config.connector_list,
+			    base.head) {
+		if (connector->base.encoder == &connector->new_encoder->base)
+			continue;
+
+		if (connector->base.encoder) {
+			tmp_crtc = connector->base.encoder->crtc;
+
+			*prepare_pipes |= 1 << to_intel_crtc(tmp_crtc)->pipe;
+		}
+
+		if (connector->new_encoder)
+			*prepare_pipes |=
+				1 << connector->new_encoder->new_crtc->pipe;
+	}
+
+	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
+			    base.head) {
+		if (encoder->base.crtc == &encoder->new_crtc->base)
+			continue;
+
+		if (encoder->base.crtc) {
+			tmp_crtc = encoder->base.crtc;
+
+			*prepare_pipes |= 1 << to_intel_crtc(tmp_crtc)->pipe;
+		}
+
+		if (encoder->new_crtc)
+			*prepare_pipes |= 1 << encoder->new_crtc->pipe;
+	}
+
+	/* Check for any pipes that will be fully disabled ... */
+	list_for_each_entry(intel_crtc, &dev->mode_config.crtc_list,
+			    base.head) {
+		bool used = false;
+
+		/* Don't try to disable disabled crtcs. */
+		if (!intel_crtc->base.enabled)
+			continue;
+
+		list_for_each_entry(encoder, &dev->mode_config.encoder_list,
+				    base.head) {
+			if (encoder->new_crtc == intel_crtc)
+				used = true;
+		}
+
+		if (!used)
+			*disable_pipes |= 1 << intel_crtc->pipe;
+	}
+
+
+	/* set_mode is also used to update properties on life display pipes. */
+	intel_crtc = to_intel_crtc(crtc);
+	if (crtc->enabled)
+		*prepare_pipes |= 1 << intel_crtc->pipe;
+
+	/* We only support modeset on one single crtc, hence we need to do that
+	 * only for the passed in crtc iff we change anything else than just
+	 * disable crtcs.
+	 *
+	 * This is actually not true, to be fully compatible with the old crtc
+	 * helper we automatically disable _any_ output (i.e. doesn't need to be
+	 * connected to the crtc we're modesetting on) if it's disconnected.
+	 * Which is a rather nutty api (since changed the output configuration
+	 * without userspace's explicit request can lead to confusion), but
+	 * alas. Hence we currently need to modeset on all pipes we prepare. */
+	if (*prepare_pipes)
+		*modeset_pipes = *prepare_pipes;
+
+	/* ... and mask these out. */
+	*modeset_pipes &= ~(*disable_pipes);
+	*prepare_pipes &= ~(*disable_pipes);
+}
+
 bool intel_set_mode(struct drm_crtc *crtc,
 		    struct drm_display_mode *mode,
 		    int x, int y, struct drm_framebuffer *fb)
@@ -6715,8 +6807,12 @@ bool intel_set_mode(struct drm_crtc *crtc,
 	struct drm_display_mode *adjusted_mode, saved_mode, saved_hwmode;
 	struct drm_encoder_helper_funcs *encoder_funcs;
 	struct drm_encoder *encoder;
+	unsigned disable_pipe, prepare_pipes, modeset_pipes;
 	bool ret = true;
 
+	intel_modeset_affected_pipes(crtc, &modeset_pipes,
+				     &prepare_pipes, &disable_pipe);
+
 	intel_modeset_commit_output_state(dev);
 
 	crtc->enabled = drm_helper_crtc_in_use(crtc);
-- 
1.7.11.2

  reply	other threads:[~2012-08-29 11:41 UTC|newest]

Thread overview: 173+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-08-19 19:12 [PATCH 00/58] modeset-rework, the basic conversion Daniel Vetter
2012-08-19 19:12 ` [PATCH 01/58] drm/i915: add crtc->enable/disable vfuncs insted of dpms Daniel Vetter
2012-08-29 17:51   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 02/58] drm/i915: rip out crtc prepare/commit indirection Daniel Vetter
2012-08-29 17:52   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 03/58] drm/i915: add direct encoder disable/enable infrastructure Daniel Vetter
2012-08-29 18:01   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 04/58] drm/i915/hdmi: convert to encoder->disable/enable Daniel Vetter
2012-09-04 19:24   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 05/58] drm/i915/tv: convert to encoder enable/disable Daniel Vetter
2012-09-04 19:25   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 06/58] drm/i915/lvds: convert to encoder disable/enable Daniel Vetter
2012-09-04 19:26   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 07/58] drm/i915/dp: " Daniel Vetter
2012-09-04 19:33   ` Jesse Barnes
2012-09-04 19:42     ` Daniel Vetter
2012-08-19 19:12 ` [PATCH 08/58] drm/i915/crt: " Daniel Vetter
2012-09-04 19:50   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 09/58] drm/i915/sdvo: " Daniel Vetter
2012-09-04 19:52   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 10/58] drm/i915/dvo: " Daniel Vetter
2012-09-04 19:53   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 11/58] drm/i915: convert dpms functions of dvo/sdvo/crt Daniel Vetter
2012-08-29  7:12   ` [PATCH] " Daniel Vetter
2012-09-04 20:10     ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 12/58] drm/i915: rip out encoder->disable/enable checks Daniel Vetter
2012-09-04 20:11   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 13/58] drm/i915: clean up encoder_prepare/commit Daniel Vetter
2012-09-04 20:12   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 14/58] drm/i915: copy&paste drm_crtc_helper_set_config Daniel Vetter
2012-09-04 20:13   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 15/58] drm/i915: call set_base directly Daniel Vetter
2012-09-04 20:15   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 16/58] drm/i915: inline intel_best_encoder Daniel Vetter
2012-09-04 20:18   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 17/58] drm/i915: copy&paste drm_crtc_helper_set_mode Daniel Vetter
2012-09-04 20:19   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 18/58] drm/i915: simplify intel_crtc_prepare_encoders Daniel Vetter
2012-09-04 20:20   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 19/58] drm/i915: rip out encoder->prepare/commit Daniel Vetter
2012-09-04 20:21   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 20/58] drm/i915: call crtc functions directly Daniel Vetter
2012-09-04 20:22   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 21/58] drm/i915: WARN when trying to enabled an unused crtc Daniel Vetter
2012-09-04 20:23   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 22/58] drm/i915: Add interfaces to read out encoder/connector hw state Daniel Vetter
2012-09-04 20:25   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 23/58] drm/i915/dp: implement get_hw_state Daniel Vetter
2012-09-04 20:26   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 24/58] drm/i915/hdmi: " Daniel Vetter
2012-09-04 20:28   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 25/58] drm/i915/tv: " Daniel Vetter
2012-09-04 20:28   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 26/58] drm/i915/lvds: " Daniel Vetter
2012-09-04 20:28   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 27/58] drm/i915/crt: " Daniel Vetter
2012-09-04 20:29   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 28/58] drm/i915/sdvo: " Daniel Vetter
2012-09-04 20:31   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 29/58] drm/i915/dvo: " Daniel Vetter
2012-09-04 20:32   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 30/58] drm/i915: read out the modeset hw state at load and resume time Daniel Vetter
2012-09-05 16:14   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 31/58] drm/i915: check connector hw/sw state Daniel Vetter
2012-09-05 16:26   ` Jesse Barnes
2012-09-05 19:10     ` Daniel Vetter
2012-08-19 19:12 ` [PATCH 32/58] drm/i915: rip out intel_crtc->dpms_mode Daniel Vetter
2012-09-05 16:27   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 33/58] drm/i915: rip out intel_dp->dpms_mode Daniel Vetter
2012-09-05 16:28   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 34/58] drm/i915: ensure the force pipe A quirk is actually followed Daniel Vetter
2012-09-05 16:32   ` Jesse Barnes
2012-09-05 19:12     ` Daniel Vetter
2012-08-19 19:12 ` [PATCH 35/58] drm/i915: introduce struct intel_set_config Daniel Vetter
2012-09-05 16:34   ` Jesse Barnes
2012-09-05 19:27     ` Daniel Vetter
2012-08-19 19:12 ` [PATCH 36/58] drm/i915: extract modeset config save/restore code Daniel Vetter
2012-09-05 16:36   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 37/58] drm/i915: extract intel_set_config_compute_mode_changes Daniel Vetter
2012-09-05 16:42   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 38/58] drm/i915: extract intel_set_config_update_output_state Daniel Vetter
2012-09-05 16:44   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 39/58] drm/i915: implement crtc helper semantics relied upon by the fb helper Daniel Vetter
2012-09-05 16:45   ` Jesse Barnes
2012-09-05 19:15     ` Daniel Vetter
2012-08-19 19:12 ` [PATCH 40/58] drm/i915: don't update the fb base if there is no fb Daniel Vetter
2012-09-05 16:47   ` Jesse Barnes
2012-08-19 19:12 ` [PATCH 41/58] drm/i915: convert pointless error checks in set_config to BUGs Daniel Vetter
2012-09-05 16:50   ` Jesse Barnes
2012-09-05 19:19     ` Daniel Vetter
2012-08-19 19:12 ` [PATCH 42/58] drm/i915: don't save all the encoder/crtc state in set_config Daniel Vetter
2012-09-05 16:52   ` Jesse Barnes
2012-09-05 19:21     ` Daniel Vetter
2012-08-19 19:13 ` [PATCH 43/58] drm/i915: stage modeset output changes Daniel Vetter
2012-09-05 17:51   ` Jesse Barnes
2012-08-19 19:13 ` [PATCH 44/58] drm/i915: push crtc->fb update into pipe_set_base Daniel Vetter
2012-09-04 19:33   ` [PATCH] " Daniel Vetter
2012-09-05 17:55     ` Jesse Barnes
2012-09-05 19:59       ` Daniel Vetter
2012-09-05 17:54   ` [PATCH 44/58] " Jesse Barnes
2012-08-19 19:13 ` [PATCH 45/58] drm/i915: remove crtc disabling special case Daniel Vetter
2012-09-05 17:56   ` Jesse Barnes
2012-08-19 19:13 ` [PATCH 46/58] drm/i915: move output commit and crtc disabling into set_mode Daniel Vetter
2012-09-05 17:58   ` Jesse Barnes
2012-08-19 19:13 ` [PATCH 47/58] drm/i915: extract adjusted mode computation Daniel Vetter
2012-09-05 18:00   ` Jesse Barnes
2012-09-05 19:30     ` Daniel Vetter
2012-08-19 19:13 ` [PATCH 48/58] drm/i915: use staged outuput config in tv->mode_fixup Daniel Vetter
2012-09-05 18:02   ` Jesse Barnes
2012-09-06  7:30     ` Daniel Vetter
2012-08-19 19:13 ` [PATCH 49/58] drm/i915: use staged outuput config in lvds->mode_fixup Daniel Vetter
2012-09-05 18:02   ` Jesse Barnes
2012-08-19 19:13 ` [PATCH 50/58] drm/i915: compute masks of crtcs affected in set_mode Daniel Vetter
2012-08-29 10:34   ` Daniel Vetter [this message]
2012-09-05 18:09     ` [PATCH] " Jesse Barnes
2012-09-05 19:38       ` Daniel Vetter
2012-09-05 19:45         ` Jesse Barnes
2012-09-05 18:07   ` [PATCH 50/58] " Jesse Barnes
2012-08-19 19:13 ` [PATCH 51/58] drm/i915: implement new set_mode code flow Daniel Vetter
2012-09-05 18:14   ` Jesse Barnes
2012-09-05 19:43     ` Daniel Vetter
2012-09-05 19:49       ` Jesse Barnes
2012-08-19 19:13 ` [PATCH 52/58] drm/i915: push commit_output_state past crtc disabling Daniel Vetter
2012-09-05 18:17   ` Jesse Barnes
2012-08-19 19:13 ` [PATCH 53/58] drm/i915: s/intel_encoder_disable/intel_encoder_noop Daniel Vetter
2012-09-05 18:17   ` Jesse Barnes
2012-08-19 19:13 ` [PATCH 54/58] drm/i915: WARN if the pipe won't turn off Daniel Vetter
2012-09-05 18:18   ` Jesse Barnes
2012-08-19 19:13 ` [PATCH 55/58] drm/i915: switch the load detect code to the staged modeset config Daniel Vetter
2012-09-05 18:19   ` Jesse Barnes
2012-08-19 19:13 ` [PATCH 56/58] drm/i915: push commit_output_state past the crtc/encoder preparing Daniel Vetter
2012-08-31 18:12   ` [PATCH] " Daniel Vetter
2012-09-04 19:32     ` Daniel Vetter
2012-09-05 18:28       ` Jesse Barnes
2012-09-05 19:48         ` Daniel Vetter
2012-09-05 18:56           ` [PATCH] drm/i915: push crtc->fb update into pipe_set_base Daniel Vetter
2012-09-05 19:50           ` [PATCH] drm/i915: push commit_output_state past the crtc/encoder preparing Jesse Barnes
2012-09-06 20:46             ` Jesse Barnes
2012-08-19 19:13 ` [PATCH 57/58] drm/i915: disable all crtcs at suspend time Daniel Vetter
2012-08-29 21:13   ` [PATCH] drm/i915: no longer call drm_helper_resume_force_mode Daniel Vetter
2012-09-05 18:31     ` Jesse Barnes
2012-09-05 19:56       ` Daniel Vetter
2012-09-05 20:04         ` Jesse Barnes
2012-09-06 20:47           ` Jesse Barnes
2012-09-05 18:29   ` [PATCH 57/58] drm/i915: disable all crtcs at suspend time Jesse Barnes
2012-08-19 19:13 ` [PATCH 58/58] drm/i915: add tons of modeset state checks Daniel Vetter
2012-08-20  8:24   ` [PATCH] " Daniel Vetter
2012-08-20 12:22   ` Daniel Vetter
2012-08-31 18:12     ` [PATCH 1/2] " Daniel Vetter
2012-08-31 18:12       ` [PATCH 2/2] drm/i915: improve modeset state checking after dpms calls Daniel Vetter
2012-09-05 18:34         ` Jesse Barnes
2012-09-05 18:33       ` [PATCH 1/2] drm/i915: add tons of modeset state checks Jesse Barnes
2012-08-20 13:17 ` [PATCH 00/58] modeset-rework, the basic conversion Jani Nikula
2012-08-21  3:27 ` Ben Widawsky
2012-08-21 17:48 ` Lespiau, Damien
2012-08-21 18:11   ` Daniel Vetter
2012-08-22 10:46     ` Lespiau, Damien
2012-08-22 11:03       ` Lespiau, Damien
2012-08-22 19:13         ` Lespiau, Damien
2012-08-22 21:21           ` Daniel Vetter
2012-08-23 12:26             ` Lespiau, Damien
2012-08-23 22:39               ` Daniel Vetter
2012-08-29 12:26                 ` Lespiau, Damien
2012-08-30 10:40                   ` Rodrigo Vivi
2012-08-21 19:48 ` Chris Wilson
2012-08-27  8:04 ` Vijay Purushothaman
2012-09-03 17:50 ` Paulo Zanoni
2012-09-05 23:23 ` Jesse Barnes
2012-09-06  6:55   ` Daniel Vetter
2012-09-06  7:23     ` Daniel Vetter
2012-09-07  1:08   ` Jesse Barnes
2012-09-07  8:55     ` Daniel Vetter
2012-09-06 21:00 ` 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=1346236444-31640-1-git-send-email-daniel.vetter@ffwll.ch \
    --to=daniel.vetter@ffwll.ch \
    --cc=intel-gfx@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 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.