netdev - [PATCH REPOST 1/2] can: flexcan: fix deadlock when using self wakeup

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]

Message-ID: <20190816081749.19300-2-qiangqing.zhang@nxp.com>
Date:   Fri, 16 Aug 2019 08:20:41 +0000
From:   Joakim Zhang <qiangqing.zhang@....com>
To:     "mkl@...gutronix.de" <mkl@...gutronix.de>,
        "linux-can@...r.kernel.org" <linux-can@...r.kernel.org>,
        "sean@...nix.com" <sean@...nix.com>
CC:     "wg@...ndegger.com" <wg@...ndegger.com>,
        "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
        dl-linux-imx <linux-imx@....com>,
        Joakim Zhang <qiangqing.zhang@....com>
Subject: [PATCH REPOST 1/2] can: flexcan: fix deadlock when using self wakeup

As reproted by Sean Nyekjaer below:
When suspending, when there is still can traffic on the interfaces the
flexcan immediately wakes the platform again. As it should :-). But it
throws this error msg:
[ 3169.378661] PM: noirq suspend of devices failed

On the way down to suspend the interface that throws the error message does
call flexcan_suspend but fails to call flexcan_noirq_suspend. That means the
flexcan_enter_stop_mode is called, but on the way out of suspend the driver
only calls flexcan_resume and skips flexcan_noirq_resume, thus it doesn't call
flexcan_exit_stop_mode. This leaves the flexcan in stop mode, and with the
current driver it can't recover from this even with a soft reboot, it requires
a hard reboot.

The best way to exit stop mode is in Wake Up interrupt context, and then
suspend() and resume() functions can be symmetric. However, stop mode
request and ack will be controlled by SCU(System Control Unit) firmware(manage
clock,power,stop mode, etc. by Cortex-M4 core) in coming i.MX8(QM/QXP). And SCU
firmware interface can't be available in interrupt context.

For compatibillity, the wake up mechanism can't be symmetric, so we need
in_stop_mode hack.

Fixes: de3578c198c6 ("can: flexcan: add self wakeup support")
Reported-by: Sean Nyekjaer <sean@...nix.com>
Signed-off-by: Joakim Zhang <qiangqing.zhang@....com>

Changelog:
V1->V2:
	* add Reported-by tag.
	* rebase on patch: can:flexcan:fix stop mode acknowledgment.
V2->V3:
	* move into a patch set.
---
 drivers/net/can/flexcan.c | 23 +++++++++++++++++++++++
 1 file changed, 23 insertions(+)

diff --git a/drivers/net/can/flexcan.c b/drivers/net/can/flexcan.c
index 56fa98d7aa90..de2bf71b335b 100644
--- a/drivers/net/can/flexcan.c
+++ b/drivers/net/can/flexcan.c
@@ -282,6 +282,7 @@ struct flexcan_priv {
 	const struct flexcan_devtype_data *devtype_data;
 	struct regulator *reg_xceiver;
 	struct flexcan_stop_mode stm;
+	bool in_stop_mode;
 
 	/* Read and Write APIs */
 	u32 (*read)(void __iomem *addr);
@@ -1636,6 +1637,8 @@ static int __maybe_unused flexcan_suspend(struct device *device)
 			err = flexcan_enter_stop_mode(priv);
 			if (err)
 				return err;
+
+			priv->in_stop_mode = true;
 		} else {
 			err = flexcan_chip_disable(priv);
 			if (err)
@@ -1660,6 +1663,15 @@ static int __maybe_unused flexcan_resume(struct device *device)
 		netif_device_attach(dev);
 		netif_start_queue(dev);
 		if (device_may_wakeup(device)) {
+			if (priv->in_stop_mode) {
+				flexcan_enable_wakeup_irq(priv, false);
+				err = flexcan_exit_stop_mode(priv);
+				if (err)
+					return  err;
+
+				priv->in_stop_mode = false;
+			}
+
 			disable_irq_wake(dev->irq);
 		} else {
 			err = flexcan_chip_enable(priv);
@@ -1675,6 +1687,11 @@ static int __maybe_unused flexcan_noirq_suspend(struct device *device)
 	struct net_device *dev = dev_get_drvdata(device);
 	struct flexcan_priv *priv = netdev_priv(dev);
 
+	/* Need to enable wakeup interrupt in noirq suspend stage. Otherwise,
+	 * it will trigger continuously wakeup interrupt if the wakeup event
+	 * comes before noirq suspend stage, and simultaneously it has enter
+	 * the stop mode.
+	 */
 	if (netif_running(dev) && device_may_wakeup(device))
 		flexcan_enable_wakeup_irq(priv, true);
 
@@ -1687,11 +1704,17 @@ static int __maybe_unused flexcan_noirq_resume(struct device *device)
 	struct flexcan_priv *priv = netdev_priv(dev);
 	int err;
 
+	/* Need to exit stop mode in noirq resume stage. Otherwise, it will
+	 * trigger continuously wakeup interrupt if the wakeup event comes,
+	 * and simultaneously it has still in stop mode.
+	 */
 	if (netif_running(dev) && device_may_wakeup(device)) {
 		flexcan_enable_wakeup_irq(priv, false);
 		err = flexcan_exit_stop_mode(priv);
 		if (err)
 			return err;
+
+		priv->in_stop_mode = false;
 	}
 
 	return 0;
-- 
2.17.1