devm_fwnode_gpiod_get does not exist in 5.3. We keep this as a separate
patch so the code stays close to 5.6 otherwise.
Signed-off-by: Guido Günther <guido.gunther@puri.sm>

Signed-off-by: Guido Günther <guido.gunther@puri.sm>

Signed-off-by: Guido Günther <guido.gunther@puri.sm>

Some devices like the Librem 5 have full hardware kill switches that cut
power to the device. Their state is indicated via gpio pin so the
drivers only purpose is to forward the current switch state to the
rfkill core.

TODO:
 - recheck state after suspend
 - simplify workqueue
 - binding docs
Signed-off-by: Guido Günther <agx@sigxcpu.org>

This simplifies things when creating multiple switches.
Signed-off-by: Guido Günther <agx@sigxcpu.org>

This allows us to use them in bindings.
Signed-off-by: Guido Günther <agx@sigxcpu.org>

Signed-off-by: Guido Günther <guido.gunther@puri.sm>

Only with that the we can reduce VDD_ARM to 0.9V at runtime.
Signed-off-by: Guido Günther <guido.gunther@puri.sm>

They're already set in imx8mq.dtsi
Signed-off-by: Guido Günther <guido.gunther@puri.sm>

According to the imx8mq data sheet running VDD_GPU at 0.9V is enough
when not overclocking to 1GHz (which we currently don't do).
Signed-off-by: Guido Günther <guido.gunther@puri.sm>

According to the imx8mq data sheet running VDD_GPU at 0.9V is enough
when not overclocking to 1GHz (which we currently don't do).
Signed-off-by: Guido Günther <agx@sigxcpu.org>

Three frames are enough. Runtime PM moves us to longer
delays if necessary.
Signed-off-by: Guido Günther <guido.gunther@puri.sm>

update the accel/gyro hardware description to use the mainline driver
and apply the respectively correct mount matrix.
Signed-off-by: Martin Kepplinger <martin.kepplinger@puri.sm>

this simply backports the v5.6 mainline driver to our tree.
Signed-off-by: Martin Kepplinger <martin.kepplinger@puri.sm>

The properties there are parsed by the tcpm and wcove drivers
but currently not by the tps6598x.

USB power delivery detection and usb gadget mode remain working
after the removal.

I've left the connector node in since that's used to wire up
the dwc3 but that will need changes as well.
Signed-off-by: Guido Günther <guido.gunther@puri.sm>

This is needed by anbox-bridge
Signed-off-by: Sebastian Krzyszkowiak <sebastian.krzyszkowiak@puri.sm>

This is needed for IP firewalling in systemd
Signed-off-by: Sebastian Krzyszkowiak <sebastian.krzyszkowiak@puri.sm>

Signed-off-by: Sebastian Krzyszkowiak <sebastian.krzyszkowiak@puri.sm>

Signed-off-by: Sebastian Krzyszkowiak <sebastian.krzyszkowiak@puri.sm>

Signed-off-by: Sebastian Krzyszkowiak <sebastian.krzyszkowiak@puri.sm>

Signed-off-by: Sebastian Krzyszkowiak <sebastian.krzyszkowiak@puri.sm>

The kernel is hanging enabling this pin.

This reverts commit cf4c86d0.

at least this is how the mainline driver implements this logic.
Signed-off-by: Martin Kepplinger <martin.kepplinger@puri.sm>

This got people confused.
Signed-off-by: Guido Günther <guido.gunther@puri.sm>

Signed-off-by: Angus Ainslie (Purism) <angus@akkea.ca>

The OTG pin needs to be high for OTG to work.
Signed-off-by: Angus Ainslie (Purism) <angus@akkea.ca>

The phy can notify the charge control when it's gone to otg mode.
Signed-off-by: Angus Ainslie (Purism) <angus@akkea.ca>

The touch controller supports two modes but default seemds to be
edge.

Closes: #136

and helps #125 and #133.
Signed-off-by: Guido Günther <guido.gunther@puri.sm>

The're currently configured at the panel.
Signed-off-by: Guido Günther <guido.gunther@puri.sm>

They'll look different for DP out and are just confusing.
Signed-off-by: Guido Günther <guido.gunther@puri.sm>

No need to pull all tags, branches and history. This makes cloning take
1.5 minutes (and even on a loaded system it goes down to 5m while
we've seen 30m before).
Signed-off-by: Guido Günther <guido.gunther@puri.sm>

Signed-off-by: Sebastian Krzyszkowiak <sebastian.krzyszkowiak@puri.sm>

This makes linux-headers package usable on the device for things
like dkms.
Signed-off-by: Sebastian Krzyszkowiak <sebastian.krzyszkowiak@puri.sm>