Closes: Librem5/OS-issues#208

These were split out to a separate repo. Keep the scripts around but
don't install them in the package until all the docs got updated.

See #40

Angus Ainslie authored 3 years ago and Guido Gunther committed 3 years ago

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

Angus Ainslie authored 3 years ago and Angus Ainslie committed 3 years ago

Firmware vesion 2.0.0024 broke BLE so we need to disable dual bluetooth support

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

Repurpose the board variant for that since it's unused.

This avoids downloading / flashing with root permissions and
extending the docs even further.

We can switch to `uuu -udev > ...` once the new uuu is more
widespread.

Closes: #35

Fixes 160ad7db which dropped the file
but didn't remove it from existing installations.

The same configuration is done via

default/audio/pulse

in librem5-base (but with a condition depending on the device type)

This reverts commit 50b5b41d.

Our initramfs figures out the correct partition to mount by itself,
and removing it from the cmdline allows it to be easily overriden
by uuu script or /etc/default/flash-kernel.

This makes the script more self contained.

Guido Gunther authored 4 years ago and Martin Kepplinger committed 4 years ago

We boot vmlinuz now - and even that is a workaround due to our
kernel package building a compressed kernel where Debian uses
an uncompressed one.

Guido Gunther authored 4 years ago and Martin Kepplinger committed 4 years ago

They changed a while ago

Guido Gunther authored 4 years ago and Martin Kepplinger committed 4 years ago

It checks phones as well since ages. Keep a symlink to not break
existing things.

Guido Gunther authored 4 years ago and Martin Kepplinger committed 4 years ago

The irq name changed in at least 5.11 and later kernels

Byzantium has the vendor string updated to

  Microchip Technology, Inc. (formerly SMSC) USB 2.0 Hub

librem5-devkit-tools v0.0.15

David Hamner authored 4 years ago and Guido Gunther committed 4 years ago

Give an informative message and wait for the device to show up
before starting to flash.

Guido Gunther authored 4 years ago and Guido Gunther committed 4 years ago

Closes: #30

Martin Kepplinger authored 4 years ago and Guido Gunther committed 4 years ago

mm will often enough enumerate the modem as >0.

This makes sure we don't regress on brightness conversion

The '+1' is added in the assert itself now.

If panel is set to 100% test would fail so include the top of the range.

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

Guido Gunther authored 4 years ago and Martin Kepplinger committed 4 years ago

Even when ther's no media we can check that the kernel
detected the card controller.