[PATCH] watchdog: prevent the STPMIC1 watchdog probe from modifying the global turn-OFF behaviour

Leonard Göhrs l.goehrs at pengutronix.de
Thu Feb 11 10:03:48 EST 2021

According to Table 12 in the STPMIC1-datasheet (Rev 7 - December 2020)
there are the following turn-OFF conditions:

- Software switch-OFF*
- PONKEYn long key press*
- Thermal shutdown
- Overcurrent protection
- Watchdog*
- VIN_OK_Fall*

The conditions marked with an * can be re-configured from turn-Off conditions
to restart requests using the RREQ_EN-bit in the CR register
(Called RESTART_REQUEST_ENABLED in Linux/Barebox).

The watchdog code enabled this bit in it's probe function to trigger a restart
on a watchdog event. This may however result in unexpected behaviour,
like a long-press of the PONKEY resulting in a restart instead of a turn-off
when the watchdog is enabled vs. when it is not.

This change does not affect the restart and poweroff functionality provided by
the wdt driver, as the restart-/poweroff-handlers explicitly set/clear the
RESTART_REQUEST_ENABLED flag when triggering a software turn-OFF.

To prevent unexpected behaviour of the PONKEYn and other turn-OFF sources the
RESTART_REQUEST_ENABLED configuration in the probe function can and should thus be removed.

Signed-off-by: Leonard Göhrs <l.goehrs at pengutronix.de>
 drivers/watchdog/stpmic1_wdt.c | 4 ----
 1 file changed, 4 deletions(-)

diff --git a/drivers/watchdog/stpmic1_wdt.c b/drivers/watchdog/stpmic1_wdt.c
index 12280f144..105ba39fb 100644
--- a/drivers/watchdog/stpmic1_wdt.c
+++ b/drivers/watchdog/stpmic1_wdt.c
@@ -174,10 +174,6 @@ static int stpmic1_wdt_probe(struct device_d *dev)
 	wdd->set_timeout = stpmic1_wdt_set_timeout;
 	wdd->timeout_max = PMIC_WDT_MAX_TIMEOUT;
-	/* have the watchdog reset, not power-off the system */
-	regmap_write_bits(wdt->regmap, SWOFF_PWRCTRL_CR,
 	ret = watchdog_register(wdd);
 	if (ret) {
 		dev_err(dev, "Failed to register watchdog device\n");

More information about the barebox mailing list