linux-kernel - [PATCH 4.18 219/235] ASoC: rt5651: Fix workqueue cancel vs irq free race on remove

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Mon, 24 Sep 2018 13:53:25 +0200
From:   Greg Kroah-Hartman <gregkh@...uxfoundation.org>
To:     linux-kernel@...r.kernel.org
Cc:     Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        stable@...r.kernel.org, Carlo Caione <carlo@...lessm.com>,
        Hans de Goede <hdegoede@...hat.com>,
        Mark Brown <broonie@...nel.org>,
        Sasha Levin <alexander.levin@...rosoft.com>
Subject: [PATCH 4.18 219/235] ASoC: rt5651: Fix workqueue cancel vs irq free race on remove

4.18-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Hans de Goede <hdegoede@...hat.com>

[ Upstream commit 8d2d7bcdc1645dc243f7735278675b083c0e506c ]

On removal we must free the IRQ *before* cancelling the jack-detect work,
so that the jack-detect work cannot be rescheduled by the IRQ.

Before this commit we were cancelling the jack-detect work from the
driver remove callback, while relying on devm to free the IRQ, which
happens after the remove callback.

This is the wrong order. This commit uses a devm-action to register
a devm callback which cancels the work, before requesting the IRQ
(devm tears things down in reverse order). This also allows us to
remove the now empty remove driver callback.

Cc: Carlo Caione <carlo@...lessm.com>
Signed-off-by: Hans de Goede <hdegoede@...hat.com>
Signed-off-by: Mark Brown <broonie@...nel.org>
Signed-off-by: Sasha Levin <alexander.levin@...rosoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@...uxfoundation.org>
---
 sound/soc/codecs/rt5651.c |   22 ++++++++++++----------
 1 file changed, 12 insertions(+), 10 deletions(-)

--- a/sound/soc/codecs/rt5651.c
+++ b/sound/soc/codecs/rt5651.c
@@ -1696,6 +1696,13 @@ static irqreturn_t rt5651_irq(int irq, v
 	return IRQ_HANDLED;
 }
 
+static void rt5651_cancel_work(void *data)
+{
+	struct rt5651_priv *rt5651 = data;
+
+	cancel_work_sync(&rt5651->jack_detect_work);
+}
+
 static int rt5651_set_jack(struct snd_soc_component *component,
 			   struct snd_soc_jack *hp_jack, void *data)
 {
@@ -2036,6 +2043,11 @@ static int rt5651_i2c_probe(struct i2c_c
 
 	INIT_WORK(&rt5651->jack_detect_work, rt5651_jack_detect_work);
 
+	/* Make sure work is stopped on probe-error / remove */
+	ret = devm_add_action_or_reset(&i2c->dev, rt5651_cancel_work, rt5651);
+	if (ret)
+		return ret;
+
 	ret = devm_snd_soc_register_component(&i2c->dev,
 				&soc_component_dev_rt5651,
 				rt5651_dai, ARRAY_SIZE(rt5651_dai));
@@ -2043,15 +2055,6 @@ static int rt5651_i2c_probe(struct i2c_c
 	return ret;
 }
 
-static int rt5651_i2c_remove(struct i2c_client *i2c)
-{
-	struct rt5651_priv *rt5651 = i2c_get_clientdata(i2c);
-
-	cancel_work_sync(&rt5651->jack_detect_work);
-
-	return 0;
-}
-
 static struct i2c_driver rt5651_i2c_driver = {
 	.driver = {
 		.name = "rt5651",
@@ -2059,7 +2062,6 @@ static struct i2c_driver rt5651_i2c_driv
 		.of_match_table = of_match_ptr(rt5651_of_match),
 	},
 	.probe = rt5651_i2c_probe,
-	.remove   = rt5651_i2c_remove,
 	.id_table = rt5651_i2c_id,
 };
 module_i2c_driver(rt5651_i2c_driver);