net: rfkill: gpio: prevent value glitch during probe
authorJosua Mayer <josua@solid-run.com>
Wed, 4 Oct 2023 16:39:28 +0000 (18:39 +0200)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 25 Oct 2023 10:03:06 +0000 (12:03 +0200)
commitdf5ee37bf29db125e992698fcc282f4b2ad05e15
treed0bee1aecc00518dc4c8b075c5a192f1bbfe83d7
parent2a6fc637c625cd7b045f72d448ec640db9d2f8cc
net: rfkill: gpio: prevent value glitch during probe

commit b2f750c3a80b285cd60c9346f8c96bd0a2a66cde upstream.

When either reset- or shutdown-gpio have are initially deasserted,
e.g. after a reboot - or when the hardware does not include pull-down,
there will be a short toggle of both IOs to logical 0 and back to 1.

It seems that the rfkill default is unblocked, so the driver should not
glitch to output low during probe.
It can lead e.g. to unexpected lte modem reconnect:

[1] root@localhost:~# dmesg | grep "usb 2-1"
[    2.136124] usb 2-1: new SuperSpeed USB device number 2 using xhci-hcd
[   21.215278] usb 2-1: USB disconnect, device number 2
[   28.833977] usb 2-1: new SuperSpeed USB device number 3 using xhci-hcd

The glitch has been discovered on an arm64 board, now that device-tree
support for the rfkill-gpio driver has finally appeared :).

Change the flags for devm_gpiod_get_optional from GPIOD_OUT_LOW to
GPIOD_ASIS to avoid any glitches.
The rfkill driver will set the intended value during rfkill_sync_work.

Fixes: 7176ba23f8b5 ("net: rfkill: add generic gpio rfkill driver")
Signed-off-by: Josua Mayer <josua@solid-run.com>
Link: https://lore.kernel.org/r/20231004163928.14609-1-josua@solid-run.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
net/rfkill/rfkill-gpio.c