Thought you got it in post 41.
- Insert echo Power_script into script.
- Suspend/resume
- Execute journalctl|grep Power_script. The number of lines seen is the number how often the script was executed.
Thought you got it in post 41.
journalctl will show complete log (use 'q' to leave the pager).
The output of services is logged into the journal.
journalctl is printing the log. grep is looking for a string. With echo Power_script an unique string is logged every time the script is executed.
When running the command after resume the number of lines found count how often the script was executed.
nope non of it, this is not something you should do at some normal distribution
I think this is doable somehow, windows itself supports all these stuff and more but likely get dropped
WSL for example could handle that.
Thanks, that confirmed my assumption. Even with Linux -> Linux you need to rsync to root account and use --numeric-ids to get restorable results.
You may be running the same script twice.
In post 24 kodiresume.service was introduced starting 99-toggle_rate.power too. In addition the script is started if the old mechanism is still working.
You can test with (use your favorite delay):
#!/bin/sh
case "$1" in
post)
echo Power_script
sleep 5
xrandr -display :0 --output HDMI1 --mode 1920x1080 --rate 59.94
sleep 5
xrandr -display :0 --output HDMI1 --mode 1920x1080 --rate 60
;;
esac
and test with journalctl|grep Power_script after suspend/resume.
blueribb Are you only using /storage/.config/sleep.d/99-toggle_rate.power or is the service still installed?
If yes, it is possible the script is twice. You can verify this by adding a echo Power_script and checking with journalctl|grep Power_script afterwaers.
CvH Are you sure that unix permissions, unix users and groups, symbolic links and file name case are kept with Windows rsync and can be restored?
lsmod will list all loaded modules.
In best case find the documentation which bus is used to connect the eMMC. PCI, USB ...
lspci -k and lsusb may be useful to get additional information.
Edit: Wrote this before reading your post. If systemd is to be loaded the FS is mounted and "something" was read before from eMMC.
The kernel config depends to the target: projects/../linux/../linux*.conf
I did some tests and when I boot from sdcard, I have no problem mount the partitions from emmc.
Then most likely a needed kernel driver is bulld as module and not available at boot time.
Linux 5.8.0-50-generic x86_64
Same kernel, same behavior (still assuming it is a kernel issue).
If you like to test different kernel versions there is a script to easily administarte mainline kernels: GitHub - pimlie/ubuntu-mainline-kernel.sh: Bash script for Ubuntu (and derivatives) to easily (un)install kernels from the Ubuntu Kernel PPA. Please read the Warnings section. In addition test if the grub boot menu can be activated by pressing ESC or holding the Shift key on your machine before installing any additional kernel.
Likely the Fire Sticks are not using nmblookup, but I do have no experience with kodi on android.
I will create a kodi PR the next few days to select the public IP address.
I don't see any work around because nmblookup is the first method tried by kodi.
To complete chewitt 's analysis for the driver build: you are using a DKMS Makefile not working in LE build system.
Use:
PKG_NAME="RTL8125B"
PKG_VERSION="cb398363f11d742d880b8bb80763d050f00e7e98"
PKG_SHA256="8f05052c59e69a8bf0ada420d25ef29f54a38d25e0fd841e9df9e61145a48f55"
PKG_LICENSE="GPL"
PKG_SITE="https://github.com/awesometic/realtek-r8125-dkms/"
PKG_URL="https://github.com/awesometic/realtek-r8125-dkms/archive/${PKG_VERSION}.tar.gz"
PKG_DEPENDS_TARGET="toolchain linux"
PKG_NEED_UNPACK="$LINUX_DEPENDS"
PKG_LONGDESC="Realtec RTL8125 driver for Linux"
PKG_IS_KERNEL_PKG="yes"
make_target() {
kernel_make V=1 \
-C $(kernel_path) \
M="$PKG_BUILD/src"
}
makeinstall_target() {
cd $PKG_BUILD
mkdir -p $INSTALL/$(get_full_module_dir)/$PKG_NAME
cp src/*.ko $INSTALL/$(get_full_module_dir)/$PKG_NAME
}
Display More
In case of the bug linked in post 2 (or any other kernel issue) use a kernel on the server that is not affected, i.e. an older one.
This will be fixed in LE with Intel Media Driver 2021Q1 Release - 21.1.3 by heitbaum · Pull Request #5287 · LibreELEC/LibreELEC.tv · GitHub
The crash is because of using CPU features in the initial test code of media-driver and/or gmmlib that are not available on older CPUs.
Do you mean it won't be fixed in an official build of Libreelec?
It will be fixed when it is fixed in the kernel. With seeing your ETA quote I would not hold breath.
I've just found a "community build" by "Sky42" version 10 generic. Do you think this will fix my WiFi problem?
Unlikely, unless there was motivation to include the external driver by owning such a realtek chip.