Install only upstream rules and the touchscreen0 symlink rule
That gets us to:
root@beagleboard-c3:~# time (udevadm trigger ; udevadm settle)
real 0m1.540s
user 0m0.062s
sys 0m0.273s
Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
This is meant for people to test before making it the default. The goal is to
get rid of as many udev rules as possible to speed up boot time
Some numbers of udev-trigger running on beagleboard/angstrom:
udev with no accept4() syscall in libc: 180s
udev 168 with default rules: 12s
udev 171 with default rules: 8s
udev 171 with minimal rules 0.8s
These times vary with the number of devices attached, SD card speed, etc.
Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
From #systemd:
16:46 < koen> kay: do you know what could have changed in 2.6.39 that system opens a getty on tty0 by itself?
16:48 < kay> nothing in the kernel i could think of
16:48 < koen> I can trigger it just by switching kernels and leaving the rootfs identical
23:53 < mezcalero> koen: btw, it is /lib/systemd/systemd-generators/systemd-getty-generator which spawns gettys
23:53 < mezcalero> on the kernel console automatically
23:53 < mezcalero> it uses /sys/class/tty/console/active for that
23:53 < mezcalero> which eixsts only on recent kernels
00:32 < koen> mezcalero: so what's the best way to avoid double spawning a getty?
00:35 < mezcalero> if if you want a serial getty even if you are not redirecting the console just make sure to use the same service name
00:35 < mezcalero> how exactly are your two getty services called?
00:39 < koen> lrwxrwxrwx 1 root root 41 May 27 22:38 getty@ttyO2.service -> /lib/systemd/system/serial-getty@.service
00:50 < mezcalero> koen: call it serial-getty@ttyO2
* DEPENDS_virtclass-native is defined in glib.inc so normal DEPENDS are
ignored for -native builds
* gdbus-serialization test causes circular dependency dbus-glib, and
because we cannot add dbus to DEPENDS then we had to drop this test
so apply dcf425c5472fdee74885c4f21a851e39c84e6642 from oe.dev
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
* otherwise notes like this are shown:
NOTE: multiple providers are available for runtime glib-2.0-utils (glib-2.0-nativesdk, glib-2.0)
NOTE: consider defining a PREFERRED_PROVIDER entry to match glib-2.0-utils
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Upgrade to latest snapshot to bring in libubacktrace
and enable it since e-wm needs it. Also enable fenv
to compiler target gcc.
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
Some architectures like mips become upset if you
try to link non-pic libraries to shared objects
some generate pic by default so it does not matter
Specifying -fPIC on commandline should make all happy
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
This fixes builds where meta-shr isn't in TOPDIR/meta-shr but somewhere else (sources/meta-shr in my case)
Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
10:15 <JaMa|Wrk> Hi, found your task-boot, this is good for everyone, would you mind pushing it to meta-oe?
Signed-off-by: Koen Kooi <koen@dominion.thruhere.net>
In addition to stock OE uclibc it has more
Add argp support
Add x86_64 arch support
Only NPTL is supported for threading
Signed-off-by: Khem Raj <raj.khem@gmail.com>
