mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Lukas Wunner <lukas@wunner.de>
To: Tejun Heo <tj@kernel.org>, Lai Jiangshan <jiangshanlai@gmail.com>,
	Alex Deucher <alexander.deucher@amd.com>,
	Dave Airlie <airlied@redhat.com>, Ben Skeggs <bskeggs@redhat.com>
Cc: dri-devel@lists.freedesktop.org, Peter Wu <peter@lekensteyn.nl>,
	nouveau@lists.freedesktop.org, Lyude Paul <lyude@redhat.com>,
	Hans de Goede <hdegoede@redhat.com>,
	Pierre Moreau <pierre.morrow@free.fr>,
	linux-kernel@vger.kernel.org,
	Ismo Toijala <ismo.toijala@gmail.com>,
	intel-gfx@lists.freedesktop.org,
	Liviu Dudau <Liviu.Dudau@arm.com>,
	Archit Taneja <architt@codeaurora.org>
Subject: [PATCH 5/5] drm/amdgpu: Fix deadlock on runtime suspend
Date: Sun, 11 Feb 2018 10:38:28 +0100	[thread overview]
Message-ID: <4c9bf72aacae1eef062bd134cd112e0770a7f121.1518338789.git.lukas@wunner.de> (raw)
In-Reply-To: <cover.1518338788.git.lukas@wunner.de>

amdgpu's ->runtime_suspend hook calls drm_kms_helper_poll_disable(),
which waits for the output poll worker to finish if it's running.

The output poll worker meanwhile calls pm_runtime_get_sync() in
amdgpu's ->detect hooks, which waits for the ongoing suspend to finish,
causing a deadlock.

Fix by not acquiring a runtime PM ref if the ->detect hooks are called
in the output poll worker's context.  This is safe because the poll
worker is only enabled while runtime active and we know that
->runtime_suspend waits for it to finish.

Fixes: d38ceaf99ed0 ("drm/amdgpu: add core driver (v4)")
Cc: stable@vger.kernel.org # v4.2+: 1234567890ab: workqueue: Allow retrieval of current task's work struct
Cc: stable@vger.kernel.org # v4.2+: 1234567890ab: drm: Allow determining if current task is output poll worker
Cc: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Lukas Wunner <lukas@wunner.de>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c | 58 +++++++++++++++++---------
 1 file changed, 38 insertions(+), 20 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c
index 8ca3783f2deb..74d2efaec52f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c
@@ -736,9 +736,11 @@ amdgpu_connector_lvds_detect(struct drm_connector *connector, bool force)
 	enum drm_connector_status ret = connector_status_disconnected;
 	int r;
 
-	r = pm_runtime_get_sync(connector->dev->dev);
-	if (r < 0)
-		return connector_status_disconnected;
+	if (!drm_kms_helper_is_poll_worker()) {
+		r = pm_runtime_get_sync(connector->dev->dev);
+		if (r < 0)
+			return connector_status_disconnected;
+	}
 
 	if (encoder) {
 		struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
@@ -757,8 +759,12 @@ amdgpu_connector_lvds_detect(struct drm_connector *connector, bool force)
 	/* check acpi lid status ??? */
 
 	amdgpu_connector_update_scratch_regs(connector, ret);
-	pm_runtime_mark_last_busy(connector->dev->dev);
-	pm_runtime_put_autosuspend(connector->dev->dev);
+
+	if (!drm_kms_helper_is_poll_worker()) {
+		pm_runtime_mark_last_busy(connector->dev->dev);
+		pm_runtime_put_autosuspend(connector->dev->dev);
+	}
+
 	return ret;
 }
 
@@ -868,9 +874,11 @@ amdgpu_connector_vga_detect(struct drm_connector *connector, bool force)
 	enum drm_connector_status ret = connector_status_disconnected;
 	int r;
 
-	r = pm_runtime_get_sync(connector->dev->dev);
-	if (r < 0)
-		return connector_status_disconnected;
+	if (!drm_kms_helper_is_poll_worker()) {
+		r = pm_runtime_get_sync(connector->dev->dev);
+		if (r < 0)
+			return connector_status_disconnected;
+	}
 
 	encoder = amdgpu_connector_best_single_encoder(connector);
 	if (!encoder)
@@ -924,8 +932,10 @@ amdgpu_connector_vga_detect(struct drm_connector *connector, bool force)
 	amdgpu_connector_update_scratch_regs(connector, ret);
 
 out:
-	pm_runtime_mark_last_busy(connector->dev->dev);
-	pm_runtime_put_autosuspend(connector->dev->dev);
+	if (!drm_kms_helper_is_poll_worker()) {
+		pm_runtime_mark_last_busy(connector->dev->dev);
+		pm_runtime_put_autosuspend(connector->dev->dev);
+	}
 
 	return ret;
 }
@@ -988,9 +998,11 @@ amdgpu_connector_dvi_detect(struct drm_connector *connector, bool force)
 	enum drm_connector_status ret = connector_status_disconnected;
 	bool dret = false, broken_edid = false;
 
-	r = pm_runtime_get_sync(connector->dev->dev);
-	if (r < 0)
-		return connector_status_disconnected;
+	if (!drm_kms_helper_is_poll_worker()) {
+		r = pm_runtime_get_sync(connector->dev->dev);
+		if (r < 0)
+			return connector_status_disconnected;
+	}
 
 	if (!force && amdgpu_connector_check_hpd_status_unchanged(connector)) {
 		ret = connector->status;
@@ -1115,8 +1127,10 @@ amdgpu_connector_dvi_detect(struct drm_connector *connector, bool force)
 	amdgpu_connector_update_scratch_regs(connector, ret);
 
 exit:
-	pm_runtime_mark_last_busy(connector->dev->dev);
-	pm_runtime_put_autosuspend(connector->dev->dev);
+	if (!drm_kms_helper_is_poll_worker()) {
+		pm_runtime_mark_last_busy(connector->dev->dev);
+		pm_runtime_put_autosuspend(connector->dev->dev);
+	}
 
 	return ret;
 }
@@ -1359,9 +1373,11 @@ amdgpu_connector_dp_detect(struct drm_connector *connector, bool force)
 	struct drm_encoder *encoder = amdgpu_connector_best_single_encoder(connector);
 	int r;
 
-	r = pm_runtime_get_sync(connector->dev->dev);
-	if (r < 0)
-		return connector_status_disconnected;
+	if (!drm_kms_helper_is_poll_worker()) {
+		r = pm_runtime_get_sync(connector->dev->dev);
+		if (r < 0)
+			return connector_status_disconnected;
+	}
 
 	if (!force && amdgpu_connector_check_hpd_status_unchanged(connector)) {
 		ret = connector->status;
@@ -1429,8 +1445,10 @@ amdgpu_connector_dp_detect(struct drm_connector *connector, bool force)
 
 	amdgpu_connector_update_scratch_regs(connector, ret);
 out:
-	pm_runtime_mark_last_busy(connector->dev->dev);
-	pm_runtime_put_autosuspend(connector->dev->dev);
+	if (!drm_kms_helper_is_poll_worker()) {
+		pm_runtime_mark_last_busy(connector->dev->dev);
+		pm_runtime_put_autosuspend(connector->dev->dev);
+	}
 
 	return ret;
 }
-- 
2.15.1

  reply	other threads:[~2018-02-11  9:40 UTC|newest]

Thread overview: 33+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-02-11  9:38 [PATCH 0/5] Fix deadlock on runtime suspend in DRM drivers Lukas Wunner
2018-02-11  9:38 ` Lukas Wunner [this message]
2018-02-11  9:38 ` [PATCH 3/5] drm/nouveau: Fix deadlock on runtime suspend Lukas Wunner
2018-02-11  9:38 ` [PATCH 2/5] drm: Allow determining if current task is output poll worker Lukas Wunner
2018-02-12 17:46   ` Lyude Paul
2018-02-12 17:50     ` [Intel-gfx] " Chris Wilson
2018-02-12 18:40       ` Lukas Wunner
2018-02-11  9:38 ` [PATCH 1/5] workqueue: Allow retrieval of current task's work struct Lukas Wunner
2018-02-12 17:07   ` Tejun Heo
2018-02-11  9:38 ` [PATCH 4/5] drm/radeon: Fix deadlock on runtime suspend Lukas Wunner
2018-02-11 18:58 ` [PATCH 0/5] Fix deadlock on runtime suspend in DRM drivers Mike Lothian
2018-02-11 19:23   ` Lukas Wunner
2018-02-11 19:41     ` Lukas Wunner
2018-02-12  0:35       ` Mike Lothian
2018-02-12  3:39         ` Lukas Wunner
2018-02-12  9:03           ` Mike Lothian
2018-02-12  9:45             ` Lukas Wunner
2018-02-12 18:58               ` Alex Deucher
2018-02-13  8:17                 ` Lukas Wunner
2018-02-13 15:19                   ` Alex Deucher
2018-02-12 13:04 ` Imre Deak
2018-02-12 17:43 ` Lyude Paul
2018-02-13 10:55 ` Liviu Dudau
2018-02-13 11:52   ` Lukas Wunner
2018-02-13 15:46     ` Liviu Dudau
2018-02-14 13:57       ` Lukas Wunner
2018-02-17 10:38 ` Lukas Wunner
2018-02-19 11:34 ` [Nouveau] " Daniel Vetter
2018-02-19 11:58   ` Lukas Wunner
2018-02-19 14:05     ` [Intel-gfx] " Daniel Vetter
2018-02-19 14:47       ` Lukas Wunner
2018-02-19 14:54         ` Daniel Vetter
2018-02-19 15:50           ` Alex Deucher

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=4c9bf72aacae1eef062bd134cd112e0770a7f121.1518338789.git.lukas@wunner.de \
    --to=lukas@wunner.de \
    --cc=Liviu.Dudau@arm.com \
    --cc=airlied@redhat.com \
    --cc=alexander.deucher@amd.com \
    --cc=architt@codeaurora.org \
    --cc=bskeggs@redhat.com \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=hdegoede@redhat.com \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=ismo.toijala@gmail.com \
    --cc=jiangshanlai@gmail.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=lyude@redhat.com \
    --cc=nouveau@lists.freedesktop.org \
    --cc=peter@lekensteyn.nl \
    --cc=pierre.morrow@free.fr \
    --cc=tj@kernel.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

all inboxes | Powered by JetHome®