Adds an aport for unixbench and a wrapper so it can be started with
ubench without requiring write permissions in /usr. To launch the
benchmark on a phone and have it finish in a reasonable amount of time:
$ ubench dhry whets
Output on an i7-7700HQ in the pmbootstrap chroot:
Dhrystone 2 using register variables 146273178.2 lps (10.0 s, 7 samples)
Double-Precision Whetstone 44390.3 MWIPS (9.7 s, 7 samples)
System Benchmarks Partial Index BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 146273178.2 12534.1
Double-Precision Whetstone 55.0 44390.3 8071.0
========
System Benchmarks Index Score (Partial Only) 10057.9
This is a neat way to get relative performance between devices and
testing performance for upstream/downstream kernels.
Add python2 to makedepends, so it builds in strict mode for armv7.
Otherwise it would fail with:
checking whether /usr/bin/python2 version >= 2.6... configure: error: too old
This MR updates the app to the latest commit (currently from 2 days
ago). It also updates plasma-angelfish, but I changed the $pkgrel to
follow the format of the other Git packages.
This also re-enables plasma-angelfish and vvave in the default Plasma
Mobile installation, as they were previously disabled due to
qt5-qtwebengine (which they depend on) requiring a soname bump
upstream.
Add main/qt5-qtspeech, required dependency for kde/kpimtextedit. Disable
check in kde/kimap, as it fails at least with QEMU aarch64.
Change the source url for Sink and Kube to the KDE Download mirror
rather than their GitHub organization (they required rebuilding anyway
due to the applications upgrade).
Switch plasma-mobile to use xdg-shell instead of wl-shell. wl-shell is
deprecated in Qt 5.12 and is breaking plasmashell, it seems.
Related upstream issue:
https://invent.kde.org/kde/plasma-phone-components/issues/2
This introduces another regression that applications don't start
maximized. Will have to solve this issue upstream.
Closes#159
The show_splash function called by the 20-debug-shell.sh script already
skips fbsplash when the device is configured to not use the framebuffer.
But the variable was not defined in the script. This is solved by
including the /etc/deviceinfo file.
Follow-up to !108, where this change got lost in the rebase.
Let pmbootstrap properly parse the package URLs, this is required for
the next commit where we start comparing the pkgver of all QT packages
- based on the URL like we do it for the KDE packages.
Depend on wayland-dev and remove the LuneOS related patch that does not
apply anymore. We don't ship LuneOS UI anymore (see #49), so there's no
point in carrying around the patch.
This package was broken in the upgrade to Qt 5.12.
Also noticed that the debug-shell hook wasn't sourcing the deviceinfo
and still showing the splashscreen even with
deviceinfo_no_framebuffer="true", now is fixed and tested.
Additionally, fix a typo in the existing
"NOTE: Skipping framebuffer setup..." message.
When util-linux v2.32 updated v2.33, remounting a partition stopped
working.
mount -n -o remount,rw /
mount: /: can't find LABEL="pmOS_root".
This service will add the pmOS rootfs partitions to device mapper so that
the remount root partition will succeed.
Fixes issue #156
This is a followup to !109 (merged). Affected packages:
* device/linux-samsung-p4wifi
* device/linux-sony-tulip
* device/linux-teclast-x80pro
* main/linux-postmarketos-allwinner
* main/linux-postmarketos-mainline
* main/linux-postmarketos-qcom
* main/linux-postmarketos-stable
[skip ci] I have confirmed that all 7 kernels still compile.
Builds abuild-sign and abuild-tar.static without any dependencies, so
they can be used outside of an Alpine Linux system. We need this for
build.postmarketos.org.
I noticed that some times the framebuffer driver gets configured in a
way that the ioctl performed by msm-fb-refresher returns something lower
than zero:
0ed263db09/drivers/video/fbmem.c (L877-L911)
For example when the Xorg starts I noticed it does a ioctl BLANK and
UNBLANK, but if msm-fb-refresher performs a ioctl in that moment it
stops its loop and exit.
For this reason I lost a lot of time trying to understand what was the
problem with Xorg not displaying anything until I noticed that I had
to restart the msm-fb-refresher.
With this change we don't have to care about msm-fb-refresher as it will
continue call ioctl PAN even if the framebuffer returns some negative
error code.
Do not add --pixman-type to the commandline, when
deviceinfo_weston_pixman_type is filled out.
--pixman-type was enabled in Weston with a custom patch, that currently
prevents us from upgrading Weston (#136).
The option allowed working around broken framebuffer drivers in
Android downstream kernels, which reported the wrong color format.
But it only works for Weston, the right way to patch this would be
patching the kernels, and we have some approaches here:
https://wiki.postmarketos.org/wiki/Troubleshooting:display#My_screen_is_red.21
When rendering on framebuffer, always do software rendering. This
should make it possible to boot up Plasma Mobile on most devices with
downstream kernels, although terribly slow. Still better than a black
screen though. Tested and working on the samsung-i9100.
We can improve the code and possibly make the rendering mode
configurable per device once we experimented more with:
* llvmpipe vs. softpipe on various devices
* armv7 (around the corner in Alpine)
* a proper display manager like lightdm
Both depend on qt5-qtwebengine, which needs to be rebuilt in Alpine
after the soname bump of libavformat [1]. Disable them for now, so
plasma mobile can still be installed, and the pmbootstrap test case
that installs it runs through again.
[1]: https://github.com/alpinelinux/aports/pull/5586
Adjust the mime-type that gets used to find executables to
application/x-pie-executable and change the APKBUILD logic to print out
a meaningful error when the executables can not be found.
Also fix the broken patch logic (doesn't display an error anymore).
Fix Nexus 5 (hammerhead) boot and framebuffer issues at startup by
waiting up to 10 seconds until the rootfs and /dev/fb0 have been
found. Waiting for the framebuffer device can be disabled by setting
deviceinfo_no_framebuffer=true in the deviceinfo.
Copy Mediatek /emmc@partitionname symlinks to the chroot. This is
needed on Mediatek devices because the fstab file references the
partitions using those symlinks (which the kernel creates - for
whatever reason - in the root directory) and otherwise we don't know
which partition is which.
We used to put a "compiler-gcc6.h" file inside every vendor kernel's
aport. This is redundant and only works with GCC6. With this patch,
a common "compiler-gcc.h" file will be used, that works independently
of the GCC version.
The file is from Decatf's android_kernel_samsung_p4 repository, where
it was taken from upstream Linux, but adjusted to work with older
vendor kernels (different #ifndef check at the top etc).
bd0772df21/include/linux/compiler-gcc.h
I thought only having specific symlinks was enough, and it was for one
kernel. However, to make sure it works with all kernels, let's just add
symlinks for all the binutils.
* remove everything that causes conflicts when installing gcc6 and
gcc6-armhf side by side
* fix up the existing "if cross compiling" code path to use the
/usr/gcc6 prefix
* link binutils binaries to the locations they are expected to be
by gcc (this took me hours to figure out....) as well as by the
linux build system with CROSS_COMPILE
* link all gcc binaries to /usr/bin/gcc6-*
Based on the last version from Alpine, but installed to /usr/gcc6 and
with pretty much everything disabled except for the C compiler.
We can probably throw out a lot of patches, and make the APKBUILD much
smaller. But let's focus on making it possible to compile old kernels
again first. There's still a lot to do for that, see #103.
* Added linux-postmarketos-allwinner kernel
* Added copy of alpines u-boot package with pine-a64lts support
* Changed postmarketos-mkinitfs to add the plain .dtb file to the boot partition
Weston does not build for x86, because it depends on libunwind. And
libunwind is not available in Alpine for x86. I've also tried to
simply enable x86 in libunwind, but it does not build out of the box.
In order to start building again for x86 at all, let's disable weston
and all packages that depend on it on for now.
When cross compiling, the kernel scripts would be in the native arch
again. That was already fixed in !1234, however the mime type of the
binaries has changed, so we couldn't detect them properly anymore
(#1659).
This commit fixes it again, and also adds a checks so this won't
silently fail in the future.
[skip ci] because this would not run through anyway (building these
packages takes too long). Fixes#1659.
Follow-up to !1373, where `pmbootstrap flasher flash_system` was
replaced with `pmbootstrap flasher flash_rootfs`. We still had used
terms like "system partition" in a lot of places.
This commit replaces it everywhere, so it's clear that we're talking
about the pmOS rootfs (which may or may not be installed to Android's
system partition).
