dri-devel.lists.freedesktop.org archive mirror
 help / color / mirror / Atom feed
From: Maxime Ripard <mripard@kernel.org>
To: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>,
	 Thomas Zimmermann <tzimmermann@suse.de>,
	David Airlie <airlied@gmail.com>,
	 Simona Vetter <simona@ffwll.ch>,
	Andrzej Hajda <andrzej.hajda@intel.com>,
	 Neil Armstrong <neil.armstrong@linaro.org>,
	Robert Foss <rfoss@kernel.org>,
	 Laurent Pinchart <Laurent.pinchart@ideasonboard.com>,
	 Jonas Karlman <jonas@kwiboo.se>,
	Jernej Skrabec <jernej.skrabec@gmail.com>,
	 Jyri Sarha <jyri.sarha@iki.fi>,
	 Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Cc: Devarsh Thakkar <devarsht@ti.com>,
	dri-devel@lists.freedesktop.org,  linux-kernel@vger.kernel.org,
	Maxime Ripard <mripard@kernel.org>
Subject: [PATCH 01/29] drm/atomic: Document atomic state lifetime
Date: Tue, 02 Sep 2025 10:32:29 +0200	[thread overview]
Message-ID: <20250902-drm-state-readout-v1-1-14ad5315da3f@kernel.org> (raw)
In-Reply-To: <20250902-drm-state-readout-v1-0-14ad5315da3f@kernel.org>

How drm_atomic_state structures and the various entity structures are
allocated and freed isn't really trivial, so let's document it.

Signed-off-by: Maxime Ripard <mripard@kernel.org>
---
 Documentation/gpu/drm-kms.rst |  6 ++++++
 drivers/gpu/drm/drm_atomic.c  | 45 +++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 51 insertions(+)

diff --git a/Documentation/gpu/drm-kms.rst b/Documentation/gpu/drm-kms.rst
index abfe220764e1edc758a6bc6fb5ff9c8e1c7749ff..dc0f61a3d29e752889077d855a4bea381f2e2c18 100644
--- a/Documentation/gpu/drm-kms.rst
+++ b/Documentation/gpu/drm-kms.rst
@@ -280,10 +280,16 @@ structure, ordering of committing state changes to hardware is sequenced using
 :c:type:`struct drm_crtc_commit <drm_crtc_commit>`.
 
 Read on in this chapter, and also in :ref:`drm_atomic_helper` for more detailed
 coverage of specific topics.
 
+Atomic State Lifetime
+---------------------
+
+.. kernel-doc:: drivers/gpu/drm/drm_atomic.c
+   :doc: state lifetime
+
 Handling Driver Private State
 -----------------------------
 
 .. kernel-doc:: drivers/gpu/drm/drm_atomic.c
    :doc: handling driver private state
diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index cd15cf52f0c9144711da5879da57884674aea9e4..b356d26faad4acaa25c1fe6f9bd5043b6364ce87 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -44,10 +44,55 @@
 #include <drm/drm_writeback.h>
 
 #include "drm_crtc_internal.h"
 #include "drm_internal.h"
 
+/**
+ * DOC: state lifetime
+ *
+ * &struct drm_atomic_state represents an update to video pipeline state.
+ *
+ * Its lifetime is:
+ *
+ * - at reset time, the entity reset implementation will allocate a
+ *   new, default, state and will store it in the entity state pointer.
+ *
+ * - whenever a new update is needed:
+ *
+ *   + we allocate a new &struct drm_atomic_state using drm_atomic_state_alloc().
+ *
+ *   + we copy the state of each affected entity into our &struct
+ *     drm_atomic_state using drm_atomic_get_plane_state(),
+ *     drm_atomic_get_crtc_state(), drm_atomic_get_connector_state(), or
+ *     drm_atomic_get_private_obj_state(). That state can then be
+ *     modified.
+ *
+ *     At that point, &struct drm_atomic_state stores three state
+ *     pointers for that particular entity: the old, new, and existing
+ *     (called "state") states. The old state is the state currently
+ *     active in the hardware, ie either the one initialized by reset()
+ *     or a newer one if a commit has been made. The new state is the
+ *     state we just allocated and we might eventually commit to the
+ *     hardware. The existing state points to the state we'll eventually
+ *     have to free, the new state for now.
+ *
+ *   + Once we run a commit, it is first checked and if the check is
+ *     successful, it is committed. Part of the commit is a call to
+ *     drm_atomic_helper_swap_state() which will turn the new state into
+ *     the active state. Doing so involves updating the entity state
+ *     pointer (&drm_crtc.state or similar) to point to the new state,
+ *     and the existing state will now point to the old state, that used
+ *     to be active but isn't anymore.
+ *
+ *   + When the commit is done, and when all references to our &struct
+ *     drm_atomic_state are put, drm_atomic_state_clear() runs and will
+ *     free all the old states.
+ *
+ *   + Now, we don't have any active &struct drm_atomic_state anymore,
+ *     and only the entity active states remain allocated.
+ */
+
 void __drm_crtc_commit_free(struct kref *kref)
 {
 	struct drm_crtc_commit *commit =
 		container_of(kref, struct drm_crtc_commit, ref);
 

-- 
2.50.1


  reply	other threads:[~2025-09-02  8:33 UTC|newest]

Thread overview: 57+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-09-02  8:32 [PATCH 00/29] drm: Implement state readout support Maxime Ripard
2025-09-02  8:32 ` Maxime Ripard [this message]
2025-09-02 13:08   ` [PATCH 01/29] drm/atomic: Document atomic state lifetime Thomas Zimmermann
2025-09-02 18:59     ` Laurent Pinchart
2025-09-02  8:32 ` [PATCH 02/29] drm/atomic: Fix unused but set warning in for_each_old_plane_in_state Maxime Ripard
2025-09-02 13:10   ` Thomas Zimmermann
2025-09-02 19:25   ` Laurent Pinchart
2025-09-02  8:32 ` [PATCH 03/29] drm/atomic: Fix unused but set warning in for_each_old_private_obj_in_state Maxime Ripard
2025-09-02 13:10   ` Thomas Zimmermann
2025-09-02 19:26   ` Laurent Pinchart
2025-09-02  8:32 ` [PATCH 04/29] drm/atomic_helper: Skip over NULL private_obj pointers Maxime Ripard
2025-09-02 13:13   ` Thomas Zimmermann
2025-09-02 19:29     ` Laurent Pinchart
2025-09-02  8:32 ` [PATCH 05/29] drm/atomic_state_helper: Fix bridge state initialization Maxime Ripard
2025-09-02 13:18   ` Thomas Zimmermann
2025-09-02 19:49   ` Laurent Pinchart
2025-09-02  8:32 ` [PATCH 06/29] drm/bridge: Implement atomic_print_state Maxime Ripard
2025-09-02 13:22   ` Thomas Zimmermann
2025-09-02 20:22     ` Laurent Pinchart
2025-09-02  8:32 ` [PATCH 07/29] drm/atomic: Implement drm_atomic_print_old_state Maxime Ripard
2025-09-02 13:26   ` Thomas Zimmermann
2025-09-02 20:35   ` Laurent Pinchart
2025-09-02  8:32 ` [PATCH 08/29] drm/atomic: Only call atomic_destroy_state on a !NULL pointer Maxime Ripard
2025-09-02 13:30   ` Thomas Zimmermann
2025-09-02 20:52   ` Laurent Pinchart
2025-09-02  8:32 ` [PATCH 09/29] drm/modeset: Create atomic_reset hook Maxime Ripard
2025-09-02 21:04   ` Laurent Pinchart
2025-09-02  8:32 ` [PATCH 10/29] drm/atomic: Add atomic_state_readout infrastructure Maxime Ripard
2025-09-02 13:44   ` Thomas Zimmermann
2025-09-02  8:32 ` [PATCH 11/29] drm/crtc: Drop no_vblank bit field Maxime Ripard
2025-09-02 13:45   ` Thomas Zimmermann
2025-09-02  8:32 ` [PATCH 12/29] drm/atomic_helper: Pass nonblock to commit_tail Maxime Ripard
2025-09-02 13:46   ` Thomas Zimmermann
2025-09-02  8:32 ` [PATCH 13/29] drm/atomic_helper: Compare actual and readout states once the commit is done Maxime Ripard
2025-09-02  8:32 ` [PATCH 14/29] drm/atomic_state_helper: Provide comparison macros Maxime Ripard
2025-09-02  8:32 ` [PATCH 15/29] drm/atomic_state_helper: Provide atomic_compare_state helpers Maxime Ripard
2025-09-02  8:32 ` [PATCH 16/29] drm/encoder: Create get_current_crtc hook Maxime Ripard
2025-09-02  8:32 ` [PATCH 17/29] drm/bridge_connector: Implement hw readout for connector Maxime Ripard
2025-09-02  8:32 ` [PATCH 18/29] drm/tidss: Convert to drm logging Maxime Ripard
2025-09-02 13:49   ` Thomas Zimmermann
2025-09-02  8:32 ` [PATCH 19/29] drm/tidss: Remove ftrace-like logs Maxime Ripard
2025-09-02 13:50   ` Thomas Zimmermann
2025-09-02  8:32 ` [PATCH 20/29] drm/tidss: crtc: Change variable name Maxime Ripard
2025-09-02 13:51   ` Thomas Zimmermann
2025-09-02  8:32 ` [PATCH 21/29] drm/tidss: crtc: Implement destroy_state Maxime Ripard
2025-09-02 13:52   ` Thomas Zimmermann
2025-09-02  8:32 ` [PATCH 22/29] drm/tidss: crtc: Cleanup reset implementation Maxime Ripard
2025-09-02 13:54   ` Thomas Zimmermann
2025-09-02  8:32 ` [PATCH 23/29] drm/tidss: dispc: Add format lookup by hw value Maxime Ripard
2025-09-02  8:32 ` [PATCH 24/29] drm/tidss: dispc: Improve mode checking logs Maxime Ripard
2025-09-02 14:06   ` Thomas Zimmermann
2025-09-02  8:32 ` [PATCH 25/29] drm/tidss: dispc: Move dispc_device definition to headers Maxime Ripard
2025-09-02  8:32 ` [PATCH 26/29] drm/tidss: dispc: make accessors accessible to other parts of the driver Maxime Ripard
2025-09-02  8:32 ` [PATCH 27/29] drm/tidss: Implement readout support Maxime Ripard
2025-09-02  8:32 ` [PATCH 28/29] drm/tidss: encoder: implement get_current_crtc Maxime Ripard
2025-09-02  8:32 ` [PATCH 29/29] drm/bridge: sii902x: Implement hw state readout Maxime Ripard
2025-09-02 14:13 ` [PATCH 00/29] drm: Implement state readout support Thomas Zimmermann

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=20250902-drm-state-readout-v1-1-14ad5315da3f@kernel.org \
    --to=mripard@kernel.org \
    --cc=Laurent.pinchart@ideasonboard.com \
    --cc=airlied@gmail.com \
    --cc=andrzej.hajda@intel.com \
    --cc=devarsht@ti.com \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=jernej.skrabec@gmail.com \
    --cc=jonas@kwiboo.se \
    --cc=jyri.sarha@iki.fi \
    --cc=linux-kernel@vger.kernel.org \
    --cc=maarten.lankhorst@linux.intel.com \
    --cc=neil.armstrong@linaro.org \
    --cc=rfoss@kernel.org \
    --cc=simona@ffwll.ch \
    --cc=tomi.valkeinen@ideasonboard.com \
    --cc=tzimmermann@suse.de \
    /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;
as well as URLs for NNTP newsgroup(s).