Casey authored 10 months ago and Oliver Smith committed 9 months ago

This cmdline argument is now a no-op. Drop it everywhere and add a new
CI check to enforce this.

Adjust the deprecated "minimal" initramfs variant to use
pmos.debug-shell to enable logging instead. It doesn't /support/
dropping to a shell, so the variable takes on a different meaning. But
that's ok since it's at least more consistent.

[ci:ignore-count]
[ci:skip-vercheck]
Signed-off-by: Caleb Connolly <caleb@postmarketos.org>
Tweaked-by: Oliver Smith <ollieparanoid@postmarketos.org>
[ci:skip-build]: kernel builds are very likely to go through now

Raymond Hackley authored 1 year ago and Newbyte committed 1 year ago

After removing the following files, $pkgdir/usr/share/deviceinfo is empty:
- $pkgdir/usr/share/deviceinfo/$pkgname
- $pkgdir/usr/share/deviceinfo/deviceinfo

Remove it when it's empty.
[ci:skip-build]: already built successfully in CI

Andrea Bolognani authored 1 year ago and Alexey Minnekhanov committed 1 year ago

The current invocation doesn't work:

  $ busybox dc "100" "5" / p
  dc: can't open '100': No such file or directory

Fix it so that it does:

  $ busybox dc -e "100 5 / p"
  20

One of the messages that is printed on failure was missing
the "WARNING:" prefix. Address that too.

Signed-off-by: Andrea Bolognani <eof@kiyuko.org>

Pablo Correa Gomez authored 1 year ago and Oliver Smith committed 1 year ago

This was a fallout from the /usr/share relocation, only found and
reported now by z3ntu. If the device package happened to be installed
later, then the link would point to the wrong location.

Fixes 1d748a93

Add a helper to generate a post-install script for device pmtest
subpackages. This script configures the console and enables logging in
the initramfs for images running in CI.

[ci:ignore-count]

Pablo Correa Gomez authored 1 year ago and Oliver Smith committed 1 year ago

In the process, add some tests and do some renaming on the existing ones, and
remove the workaround for #2228, will be solved appropriately in a follow-up
commit.

Fixes #1836

Pablo Correa Gomez authored 1 year ago and Oliver Smith committed 1 year ago

Since the modules are going to be installed into the initfs, it makes
sense to use a more verbose naming for the file. Otherwise there's the
risk of getting confused with other type of module files that exist in
device packages.

Fixes 65be0d7e

Anton Bambura authored 1 year ago and Oliver Smith committed 1 year ago

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

Pablo Correa Gomez authored 1 year ago and Clayton Craft committed 1 year ago

So that the initramfs can get the list of modules that need to be loaded
once the deviceinfo_modules_initfs variable is gone

Pablo Correa Gomez authored 1 year ago and Clayton Craft committed 1 year ago

To simplify the future removal of deviceinfo_modules_initfs

Newbyte authored 1 year ago and Dylan Van Assche committed 1 year ago

No reason for this to be arch-specific.

These are no longer used, and shouldn't exist.

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

Luca Weiss authored 2 years ago and Casey committed 2 years ago

Soon openrc-settingsd (hostnamed) will uses these properties so the
information can be exposed to apps like GNOME Settings.

Let's also move the shellcheck comment to the outer scope, otherwise
we'd need four of these comments now, and let's not use ${} but just $.

Allow calling the script with:
  SPLASH_CONFIG=$srcdir/splash.ini devicepkg_build $startdir $pkgname

In order to pass a custom splash config, as it will be used with the
PineNote (MR 2910).

Kernel file names will be generic (/boot/vmlinuz).
Release path still has the 'flavor' in it so that pmbootstrap can still
get a reference to which kernel package is installed

Oliver Smith authored 4 years ago and Alexey Minnekhanov committed 4 years ago

Remove -Werror from Makefile.common files too, as found in the
xiaomi-latte downstream kernel (MR 1692).

Minecrell authored 4 years ago and Oliver Smith committed 4 years ago

TP:[AL3]:APKBUILD:5:pkgname must not be quoted

[skip ci]: won't finish within 3 hours

Minecrell authored 5 years ago and Alexey Minnekhanov committed 5 years ago

Many device ports install extra modules-load.conf/modprobe.conf
configurations. Simplify this a bit by installing these automatically
to appropriate locations in the file system.

Luca Weiss authored 5 years ago and Oliver Smith committed 5 years ago

This breaks booting on Librem 5 and is not needed on any device
currently using downstreamkernel_package, so remove it.

HOSTCC defaults to gcc, so we should only set it to override it
if we want to use an older GCC. Removing it avoids annoying linter
warnings caused by the HOSTCC="..." variable assignments.

Minecrell authored 5 years ago and Bart Ribbers committed 5 years ago

(These were present before the code was moved to devicepkg-dev.)

Minecrell authored 5 years ago and Bart Ribbers committed 5 years ago

We need to generate the splash screens separately for each device,
because they are specific to the device's display resolution.

At the moment we do this dynamically during the installation process.
This has the advantage that there is no need to re-build all device
packages when one of the splash screen is changed (or a new one is added).

In reality, however, the splash screens do not change very frequently.
On the other hand, generating the splash screens dynamically has signficant
disk usage overhead for a minimal ("none" UI) rootfs:

The Python interpreter together with the necessary libraries requires
about ~60 MB of disk space on aarch64.
The splash screens itself require about ~100 KB for 720x1280.

This is not necessary if we move the splash screen generation into
devicepkg-dev, which is used to build the device package for all devices.
Another advantage is that we no longer need the (rather complicated)
caching mechanism for splash screens - so we actually end up with less
lines than before.

rootfs size for samsung-a5ulte ("none" UI):
  Before: 450M
  After:  388M (-62M)

After this change, every(!) device package needs to be rebuilt once.
No changes are necessary in device packages.

Kernel subpackages may contain a "-" (e.g. -kernel-mainline-modem);
replace it with an underscore to make sure that we find the
kernel-specific variables in deviceinfo.

Instead of never overwriting compiler-gcc.h if it exists, overwrite it
by default except if OVERWRITE_GCCH=0 is set in the environment.

Most kernels *do* need this file replaced and won't build without this
patch.

At least one newer kernel already includes the compiler support split
stuff, so don't overwrite it if it already exists.

The 5.0 kernel now includes -Werror= parameters in the makefile, this
changes downstreamkernel_prepare.sh to support them. Without this, the
script will replace, for example, -Werror=blah with =blah, which causes
the kernel compilation to fail.

Removes the need to cd $builddir before calling downstreamkernel_prepare
in a kernel APKBUILD prepare function.