This upgrades u-boot to the latest upstream Purism version, and uses the
latest DDR training firmware.
APKBUILD was reformatted to replace indentation with tabs.
We were using a frequency of 624 which froze my device and in the past
other units too. Set it back to 552 as before so this doesn't happen
anymore
This is being upstreamed, https://gitlab.com/pine64-org/u-boot/-/merge_requests/3
(cherry picked from commit 88b48dee152686a887809ddb296cfd96e0c89f55)
Alpine's aarch64 builder is stuck and did not build 20.04.3-r1 yet. This
causes kde/itinerary from pmaports.git to fail:
https://builds.sr.ht/~postmarketos/job/282427#task-pmbootstrap_build-415
I've verified that this package builds for aarch64 with pmbootstrap, and
disabled other architectures.
The "shellprocesstest" is failing when building for armv7 on sr.ht. I
was not able to reproduce it locally. Just disable all tests for now.
It would be better to just disable the failing test, but since I can't
reproduce it locally I can't say that the build will go through then.
This package is in temp/, so let's not waste much time here. !check can
be removed when upstreaming it to Alpine.
[ci:skip-vercheck] [ci:skip-build]
Taken from upstream. Our mesa-git is pretty outdated, we should update
it soon. But we need to unblock the repository first, so this is the
fastest way for now.
"plasma" and some related packages are currently missing in Alpine edge
armv7 due to a cyclic dependency. Let's disable all packages depending
on plasma for armv7 temporarily, to get the pmOS edge armv7 repo up
again.
[ci:skip-build], [ci:skip-vercheck]: only arch line changed
Related: build.postmarketos.org#72
Add patch to fix build for armv7, armhf. That patched version is being
upstreamed to Alpine. Fork it now, so our builds are not blocked.
Related: builds.postmarketos.org#72
gitlab.com had a service disruption, so it was not possible to "git
clone" from gitlab.com from some regions, like where the sourcehut
infrastructure is located. Since we are building our packages on
builds.sr.ht, this caused all packages that should be built at that time
to fail. It is working again, so bump the pkgrels to restart the builds.
This switch has happened in the Alpine repos quite a while ago and most
of the pmOS packages were using it already too, so let's switch over the
last ones as well.
This also cleans up the APKBUILDs where necessary
Put virtual keyboard style into a subpackage, and let the main package
depend on it. That way, it can be used in postmarketos-ondev, without
pulling in all of plasma-phone-components.
Fails to build, because extra-cmake-modules from Alpine is not available
for armhf. Its APKBUILD says "Blocked by qt5-qtdeclarative".
[ci:skip-vercheck]
At the moment we have Contributor: lines on some packages (but not all of them),
but often they don't represent the actual contributors to the package very well.
E.g. when we added them retroactively to the device packages we only added
the initial contributor (which isn't necessarily the person
who made most of the work for a device...)
The Git history is the most representative source for figuring out
who contributed to a package, so there is no reason to duplicate that
into the APKBUILD.
[skip ci]: way too many packages
Follow-up to cbc6b9fcd7
Build locally for both armv7 and aarch64 fine.
[ci:skip-build] x86_64 build is disabled anyway
Signed-off-by: Alexey Min <alexey.min@gmail.com>
* Upgrade to the released stable version of upstream u-boot.
* Added patch to make the led on the Pinebook Pro light up earlier
* Added the update-u-boot script from Alpine with updates for the rk3399
devices
This adds a package that builds u-boot from the pine64/u-boot repository
which includes patches for enableing less hardware while booting so
there's quicker feedback that the power button has been pressed.
It also has a seperate patch file that modifies the clockspeed for the
memory which is one of the main performance bottlenecks of the A64 SoC.
It's a patch file so it's quick and easy to test out other clock speeds
when building. 600Mhz is stable but it should be able to run up to
624Mhz.
This fixes 1943ead268, which accidentally
removed the dev() function that generates the mesa-git-dev package..
causing mesa-dev to be pulled in in cases where mesa-dev is required
(and failing because mesa-dev is older)
[ci:skip-build]: already built successfully in CI
This was only needed to load the ALSA PulseAudio plugin outside
the chroot when running QEMU. Now that we allow configuring the
QEMU PulseAudio backend directly this is no longer needed.
At the moment, installing the rootfs for ouya-ouya fails with:
ERROR: unsatisfiable constraints:
libdrm-grate-2.4.100_git20191221-r0:
conflicts: libdrm-2.4.100-r0[libdrm]
libdrm-2.4.100-r0[so:libdrm.so.2=2.4.0]
libdrm-2.4.100-r0[so:libdrm_tegra.so.0=0.0.0]
libdrm-2.4.100-r0[so:libkms.so.1=1.0.0]
satisfies: device-ouya-ouya-1-r12[libdrm-grate]
libvdpau-tegra-0_git20190315-r0[libdrm-grate]
.pmbootstrap-20200110.144341[libdrm-grate]
.pmbootstrap-20200110.144341[libdrm]
mesa-gl-19.3.2-r0[so:libdrm.so.2]
directfb-1.7.7-r1[so:libdrm.so.2]
directfb-1.7.7-r1[so:libkms.so.1]
mesa-dri-swrast-19.3.2-r0[so:libdrm.so.2]
mesa-19.3.2-r0[so:libdrm.so.2]
libdrm-2.4.100-r0:
conflicts: libdrm-grate-2.4.100_git20191221-r0
libdrm-grate-2.4.100_git20191221-r0[so:libdrm.so.2=2.4.0]
libdrm-grate-2.4.100_git20191221-r0[so:libdrm_tegra.so.0=0.0.0]
libdrm-grate-2.4.100_git20191221-r0[so:libkms.so.1=1.0.0]
satisfies: .pmbootstrap-20200110.144341[libdrm]
mesa-gl-19.3.2-r0[so:libdrm.so.2]
directfb-1.7.7-r1[so:libdrm.so.2]
directfb-1.7.7-r1[so:libkms.so.1]
mesa-dri-swrast-19.3.2-r0[so:libdrm.so.2]
mesa-dri-swrast-19.3.2-r0[so:libdrm_amdgpu.so.1]
mesa-dri-swrast-19.3.2-r0[so:libdrm_nouveau.so.2]
mesa-dri-swrast-19.3.2-r0[so:libdrm_radeon.so.1]
mesa-19.3.2-r0[so:libdrm.so.2]
mesa-19.3.2-r0[so:libdrm_amdgpu.so.1]
mesa-19.3.2-r0[so:libdrm_nouveau.so.2]
mesa-19.3.2-r0[so:libdrm_radeon.so.1]
Looking closer at the error we see that:
1. We want to explicitly install libdrm-grate for device-ouya-ouya.
2. libdrm-grate provides
- so:libdrm.so.2=2.4.0
- so:libdrm_tegra.so.0=0.0.0
- so:libkms.so.1=1.0.0
3. But the mesa package also builds AMD and Nouveau drivers and
therefore requires:
- so:libdrm_amdgpu.so.1
- so:libdrm_nouveau.so.2
- so:libdrm_radeon.so.1
These libraries are not provided by libdrm-grate, therefore it is impossible
to install mesa and libdrm-grate at the same time.
A simple solution to fix this problem is to let libdrm-grate provide
these additional libraries as well - the package size overhead is negligible
and the additional drivers build just fine.
[ci:skip-build]: already built successfully in CI
At the moment, every device that wants to make use of mesa-git needs
to depend on all relevant mesa-git subpackages.
We can simplify this by adding these directly as depends for the
-dri- package that most devices will be depending on. That way,
the fact that you need to depend on all relevant subpackages is
mostly hidden away as "implementation detail" in the mesa-git
package, and no special care is required when using mesa-git.
This renamed the u-boot package for Purism librem5 devices, since the
phone and devkit share the same u-boot version. It also updates the
various components to the latest versions
This updates ATF, DDR/HDMI firmmware, and u-boot versions, and generates
a unified image that can be flashed at an offset of 2KiB.
mkimage is also no longer used to generate the final image.
Some old unused functions copied when this APKBUILD was forked were also
removed.
Some of the DRI drivers are not moved to the correct subpackage
and therefore installed everywhere through the main package.
This wastes about 8 MB of disk space, so lets move them to the
correct subpackage.
Build src/git_sha1.h early to avoid build failure:
../src/vulkan/overlay-layer/overlay.cpp:31:10: fatal error: git_sha1.h: No such file or directory
31 | #include "git_sha1.h"
| ^~~~~~~~~~~~
Default dbus policy of ofono allows only root user and users who are
logged into tty using at_console policy. However since our dbus is not
built with elogind, at_console is never set.
This allows user in wheel group to access ofono
[ci:skip-build]: already built successfully in CI
Fix build with current abuild version by removing the "-devel" suffix
in /usr/lib/pkgconfig/*.pc. Set "pcprefix" instead, so abuild doesn't
confuse these packages with the regular mesa.
[ci:skip-build]: already built successfully in CI
Fixes: #386
Turns out, that the pkgrel was bumped in Alpine. I had looked at an
outdated error message from the postmarketos-ui-* related MR and
assumed that this issue is still present (somwhat confused by the other
upstream-compat issues).
This reverts commit 07653d60a8.
Needs a pkgrel bump because dependency libboost was upgraded. This is
currently breaking everything depending on libphonenumber, for example
plasma mobile.
>>> postmarketos-ui-plasma-mobile: Analyzing dependencies...
ERROR: unsatisfiable constraints:
so:libprotobuf.so.20 (missing):
required by: libphonenumber-8.10.21-r0[so:libprotobuf.so.20]
>>> ERROR: postmarketos-ui-plasma-mobile: builddeps failed
It was forked to build the 5.17 pre-release, which is not needed
anymore. Remove now to unclutter pmaports, and to get pretty much all
remaining packages for x86_64 out of the queue on
build.postmarketos.org.
For llvm8-dev llvm-config executable is no longer in PATH
(it is in PATH now for llvm9-dev) and now it is in
/usr/lib/llvm8/bin/llvm-config, so it is more reliable
to adjust $PATH before calling meson build.
It will work for all past, current and future llvm versions.
Fixes 90d3deb7b4
[ci:skip-build]: already built successfully in CI
This commit:
- Updates u-boot
- Updates arm trusted firmware
- Builds the m4 firmware using cross compilation (no more dependency on
downloading the binary from purism \o/)
postmarketos-ui-plasma-mobile has !armhf in its arches list, so it does
not make sense to build these packages for armhf either. Let's save some
building time.
[ci:skip-vercheck], [ci:skip-build]
- Remove the outdated patch, it has been fixed differently upstream
- Update patch for use-elf-tls slots
[ci:skip-build]: already built successfully in CI
I did not check if patches still applied yesterday, and it turns out
that almost all of them don't apply anymore (probably because the
changes are in the upstream source now?).
Note that I did not test if plasma mobile is working correctly again
with this version, I'm just making the build errors go away (which I
did not notice right away yesterday, as qt5-qtbase was still building
and we had to upgrade that anyway, so there was no point in waiting).
Build tested and working for x86_64.
Fixes: 86a0ecc04a ("temp/qt5-qt*: upgrade to 5.12.5")
The package was originally added with jemalloc to work around a deadlock
while compiling mesa. I've tested compiling the mesa-git package on
x86_64 for armhf and armv7 and it worked fine. Looking at the original
issue report, the problem only happened with the autotools build system
and not with meson - and as all our mesa or networkmanager aports are
using meson now, I think we can delete this.
[ci skip] Package output does not change except for meta
information (licenses), so don't build or verify checksums
of the packages to prevent outdated source URL's breaking it.
This will have to be looked at later
Increase pkgrels, so the resulting versions are higher than the aports
in Alpine. Seems like this was done in the upstream PR initially, but
then got lost during rebase.
Fork from Alpine to apply Martijn's patch:
https://github.com/alpinelinux/aports/pull/9894
"This makes wpa_supplicant run in dbus mode if no config file has been
created for it, fixing the networkmanager integration. This config has
been used for a long time in postmarketOS and works fine.
It also removes the iwd dependency from networkmanager again so
wpa_supplicant is used because iwd doesn't support as many chipsets as
wpa_supplicant and it doesn't run on older kernels."
[ci:skip-build]: already built successfully in CI for x86_64, I've
tested that it builds for armhf and aarch64 too.
Avoid build failure for x86:
>>> mesa-git-glapi*: Running postcheck for mesa-git-glapi
>>> ERROR: mesa-git-glapi*: Found textrels:
TEXTREL /home/pmos/build/pkg/mesa-git-glapi/usr/lib/libglapi.so.0.0.0
The package is only used for armhf, armv7, aarch64 devices currently, so
change the arch line in the APKBUILD accordingly.
$ cd device
$ for i in $(git grep mesa-git | cut -d/ -f1); do grep -H arch $i/deviceinfo; done
device-asus-flo/deviceinfo:deviceinfo_arch="armhf"
device-lg-hammerhead/deviceinfo:deviceinfo_arch="armv7"
device-pine-a64lts/deviceinfo:deviceinfo_arch="aarch64"
device-pine-dontbeevil/deviceinfo:deviceinfo_arch="aarch64"
device-sony-castor-windy/deviceinfo:deviceinfo_arch="armhf"
device-sony-sirius/deviceinfo:deviceinfo_arch="armv7"
[ci:skip-vercheck]: only changed arch line
Fixes: ce7e4d79e1 ("temp/mesa-git: new aport (!524)")
Freedreno is broken in latest mesa stable, but it is fixed in current
master. Let's use master for now.
Based on Brian Masney's patch, but with a hardcoded commit and a
separate package.
The description below is a quote from Brian Masney in #296.
The pmOS mesa packages need to have -Db_ndebug=true passed to the meson
build so that debug assertions and extra validations are disabled to
improve overall system performance. See [1] for more details. Fedora
also recently made this change: [2]
This change, along with using the master branch of mesa, gets the
Freedreno driver working properly with X11 on the Nexus 5 and presumably
other A3XX-based systems.
--buildtype=release should also be passed to meson for extra
optimizations. See the thread [3] for more details.
[1] https://gitlab.freedesktop.org/mesa/mesa/blob/master/docs/meson.html#L321
[2] https://bugzilla.redhat.com/show_bug.cgi?id=1692426.
[3] https://lists.freedesktop.org/archives/mesa-dev/2019-June/219960.html
[ci:skip-build]: already built successfully in CI
Alpine's build bot is stuck on another package, but this needs to be
rebuilt against the recently updated libhunspell. Otherwise Plasma
Mobile cannot be installed, as pmbootstrap fails with:
ERROR: Could not find dependency 'so:libhunspell-1.6.so.0' in any aports folder or APKINDEX.
I've tested that this builds properly.
Disable for arm arches, because there it fails with:
[89/300] Compiling C object 'xcursor/c7765ab@@wlr_xcursor@sta/xcursor.c.o'.
FAILED: xcursor/c7765ab@@wlr_xcursor@sta/xcursor.c.o
gcc -Ixcursor/c7765ab@@wlr_xcursor@sta -Ixcursor ...
cc1: error: /home/pmos/build/pkg/mesa/usr/include: No such file or directory [-Werror=missing-include-dirs]
I suspect, that this relates to our custom mesa packaging, which is only
built for arm architectures currently. Phosh isn't running on anything
other than x86_64 so far in postmarketOS, so let's just disable wlroots
arm too until this is figured out.
[ci:skip-vercheck]: only modified the arch line
Remove Alpine's maintainers and contributors from the APKBUILDs in temp
where we still had them. In general, we don't keep them there, because
we don't want them to get e-mailed about questions from the postmarketOS
aports. It is obvious that they were copy pasted from Alpine, with minor
adjustments.
Remove "Remember to increase pkgrel after changing _pkgver!" from
temp/mesa, because we have a CI check for that now.
[ci:skip-build]: cosmetic changes only
