Niklas Cathor authored 4 years ago and Minecrell committed 4 years ago

With the camera driver (VIDEO_EXYNOS_FIMC_IS) enabled, the system
would randomly panic during startup. Even though the camera does
not currently work on this system, the driver is nonetheless enabled
to make the flashlight work.

The panics seem to be triggered by udev's "v4l_id", which queries
capabilities of video4linux devices.

This change adds a udev rules file, to override the behavior of udev's
default rules for v4l.

- Cleanup kernel cmdline:
  - earlycon can automatically detect the memory address of
    the UART controller based on the device tree
  - the comma before n8 in console=ttyMSM0,115200,n8 is actually
    invalid so this was not parsed correctly. However, it is the
    default anyway so we might as well remove it completely

- Configure deviceinfo_getty for all devices

- Workaround probe issue of display by adding a softdep on the panel
  module. This allows the display to work when excluding proprietary
  firmware (just a weird side effect, see comment in modprobe.conf...)

The device port gets increasingly hard to maintain with all the differences
between the downstream and the mainline kernel. Especially deviceinfo
has quite some duplication only because the downstream kernel needs to be
handled as well.

The mainline port has much more features working. The only limitation
compared to the downstream kernel is the lack of a proper charging driver.
But even on downstream postmarketOS the battery driver is not working
properly for some reason.

For charging, a downstream installation in recovery seems more suitable
at the moment. Eventually we will build some functionality into the
bootloader to load a simple downstream initramfs when the device gets
booted for charging.

Signed-off-by: Danct12 <danct12@disroot.org>
[ci:skip-build]

 working towards #458. See [1] for an explaination

 [1]: https://gitlab.com/postmarketOS/pmaports/-/merge_requests/1131#note_318923711

The dts for oneplus bacon was already present in linux-postmarketos.

I adjusted the APKBUILD and deviceinfo to be able to use both downstream and mainline kernels

This script is used by multiple repositories by now, let's make sure we
only have one place where we need to maintain it. The version in
ci-common is shellchecked and displays less output for 'pmbootstrap
init' again (unless it fails, then it shows the whole log).

This file is not used anymore by the bpo code, so remove it.

RedFox arch authored 4 years ago and Alexey Minnekhanov committed 4 years ago

Splash logo displayed.
SSH over USB works,
Weston works.
Port build scripts to python3.

Niklas Cathor authored 4 years ago and Alexey Minnekhanov committed 4 years ago

These changes enable the flashlight LED driver (fled_rt5033), and add
support for switching the light on and off from userspace.

In more detail:

* CONFIG_FLED_RT5033 enables the flashlight driver
* CONFIG_VIDEO_EXYNOS_FIMC_IS is required for the flashlight driver to
  compile (there seems to be a lot of cross-dependency between the two
  drivers)
* patch 0001 fixes the use of usleep (which no longer exists), to make
  the kernel compile with those settings

With these changes, /sys/class/flashlight/rt-flash-led shows up in
sysfs, but it's only possible to control the brightness of the LED,
not to switch it on.

Patch 0002 is what actually makes the flashlight usable: it makes
the /sys/class/flashlight/rt-flash-led/mode file writable.

Writing "Torch" there enables the flashlight, writing "Off" disables
it again. "Mixed" mode works like "Torch" as well. I have not figured
out yet how to trigger the camera flash, so setting the mode to
"Flash" is possible, albeit pointless.

Niklas Cathor authored 4 years ago and Minecrell committed 4 years ago

Don't pass empty arguments to mkbootimg, when _second or _dt are not set.

Read kernel module names from files in
/etc/postmarketos-mkinitfs/modules and print out which modules are
required by which of these files as the initramfs gets generated. Put
the default modules into a new 00-default.modules. This allows mkinitfs
hooks to add modules to the initramfs.

Parse the modules from a file similar to modules-load.d, with commented
lines and empty lines ignored. Add a simple test script for that and run
it in check().

Split mkinitfs.sh into two files, so the functions (now in
mkinitfs_functions.sh) can be sourced in a test. Such a test will be
added in the next commit.

Move $BINARIES_EXTRA into the get_binaries_extra() function, so its
$(find...) does not get executed when sourcing the functions file. Move
$BINARIES into get_binaries() too for consistency.

Make it pass shellcheck by fixing obvious mistakes like '==' instead of
'='. Add "shellcheck disable" lines for things that are intentionally done
or would be a bit more complicated to rewrite. It would be nice to
refactor some things like '${outfile/initramfs-/uImage-}', but this should
be done in a separate change (patches welcome).

Remove unused variable "modules_path". While at it, add two new fatal
error messages instead of just "exit 1" without any comment and remove
redundant mkdir in APKBUILD.

For testing changes for device categorization, it is useful to have
a device in each of the categories. The PinePhone is close to being
moved to main/, but it doesn't fulfill all requirements yet.

The QEMU "device ports" are very simple since QEMU currently only
emulates a rather limited set of hardware features. All available features
are working correctly (especially after the recent rework of the QEMU
packages). I suppose it is also usable as a "daily driver", at least for
its intended purpose (a virtual machine for testing postmarketOS changes). :)

Given that everyone can run QEMU, everyone could potentially maintain
it. For now I have added myself as maintainer since I did most of the
recent cleanup. Add drebrez as second maintainer.

Overall it seems useful to have qemu-* in main/, especially because
it is now the device that is selected by default in pmbootstrap.

Replace branch_pmos and branch_alpine keys (that were never used) with a
new channel key. pmbootstrap will look up the branches related to a
channel in channels.cfg of the master branch. While at it, modernize the
file.

Let's add a nice file that defines the available release channels.
Upcoming pmbootstrap patches will make use of this in 'pmbootstrap init'
for example:

[20:46:22] Choose the postmarketOS release channel.
[20:46:22] Available (2):
[20:46:22] * edge: Rolling release channel
[20:46:22] * stable: Upcoming beta release (WIP, DO NOT USE!)
[20:46:22] Channel [edge]:

Closes: #519

Niklas Cathor authored 5 years ago and Oliver Smith committed 4 years ago

linux-samsung-kminilte: adjust paths where the bcmdhd driver expects firmware
device-samsung-kminilte: add "nonfree-firmware" subpackage
firmware-samsung-kminilte: new package

firmware post-install actions:
- enable "deferred-initcalls" in sysinit for wlan0 to be available
- force wpa_supplicant to use wlan0 instead of p2p0
(both based on script from device-samsung-klte package)

[ci:skip-build]: already built successfully in CI

Dang Huynh authored 5 years ago and Oliver Smith committed 4 years ago

We'll use the new RTL8723CS driver in the kernel tree instead
of using the one that we have to fetch from GitHub.

Signed-off-by: Danct12 <danct12@disroot.org>
Signed-off-by: Asriel Dreemurr <asriel.danctnix@gmail.com>

[ci:skip-build]: already built successfully in CI

See:
* https://gitlab.alpinelinux.org/alpine/aports/issues/11349
* https://gitlab.alpinelinux.org/alpine/aports/-/commit/7bb907ec23f7672218ff018e4fdefdea2048ac23

Antoine Fontaine authored 5 years ago and Bart Ribbers committed 4 years ago

working towards https://gitlab.com/postmarketOS/pmaports/-/issues/458.

Daniele Debernardi authored 5 years ago and Bart Ribbers committed 4 years ago

The lddtree script uses the scanelf utility which has some security check that causes a slowdown of the operation, executing it without privileges it's really faster.