First of all, thank you, voRia, for this forum and for samsung-tools!
I tried Linux Mint Debian Edition (XFCE). While it worked, attempting to change the backlight would make the display flicker and X would become unstable.
So, after discovering your Samsung PPA, I tried Samsung Tools and Easy Slow Down Manager with Xubuntu 11.04 on this laptop, no problems there. But I found XFCE a bit basic after using GNOME 2 for a long time, so I decided I'd try KDE. As I've found using Kubuntu to be a horrible experience, I installed Arch Linux with KDE.
Arch Linux uses Kernel 3.0, like LMDE, I think. While using KDE, I noticed that attempting to change the brightness would again make the backlight flicker and the laptop unstable. I found, using the Internet, that adding "acpi_osi=Linux acpi_backlight=vendor" to the kernel's cmdline would fix this problem; I could now turn down my backlight without any problems.
Kernel 3.0 has a samsung-laptop module that replaces the now discontinued easy-slow-down-manager and samsung-backlight module. I found, however, that samsung-tools still expects the esdm module to get some of its work done. I hacked samsung-tools to use samsung-laptop to get stuff done. This is a hack because I removed the ESDM code in the process...
- samsung-laptop's backlight code doesn't seem to let me blank the display (writing 0 to /sys/devices/platform/samsung/backlight/samsung/brightness just sets the screen to the lowest brightness) so I just force the use of vbetool, which does work
- The fan interface is pretty much the same, except that /sys/devices/platform/samsung/performance_level actually returns the fan's status as a word: "normal, silent, overclock" (yes, overclock replaces speed)
- Instead of ESDM's proc entry to disable Wi-Fi, rfkill is used instead. My R519 shows two devices when I "rfkill list wifi" (samsung-wifi and phy0); I block them both. While I can disable samsung-wifi only and stop the wireless, the Wi-Fi light remains on. I can disable phy0 only and the light will go off and the wifi will disable itself, but I'm not sure if samsung-wifi needs to be disabled as well...
- I use a subprocess hack to get the rfkill device index for samsung-wifi, but I found later that you can look in /sys/devices/platform/samsung/rfkill/rfkill0/...
I've attached a tarball with a slightly modified PKGBUILD and the patch. One thing I have noticed is that under Arch Linux, "rc.d stop samsung-tools" will not do anything - "python2 /usr/lib/samsung-tools/system-service.py" and "python2 /usr/lib/samsung-tools/session-service.py" are still present in ps's output.
Also, this laptop has a Celeron T3100. When attempting to load some cpufreq modules, I get the following:
[root@Arch ~]# uname -a
Linux Arch 3.0-pf #2 SMP PREEMPT Fri Sep 2 10:14:55 EEST 2011 x86_64 Celeron(R) Dual-Core CPU T3100 @ 1.90GHz GenuineIntel GNU/Linux
[root@Arch ~]# modprobe acpi-cpufreq
FATAL: Error inserting phc_intel (/lib/modules/3.0-pf/extra/phc-intel.ko): Input/output error
FATAL: Error inserting acpi_cpufreq (/lib/modules/3.0-pf/kernel/drivers/cpufreq/acpi-cpufreq.ko.gz): Input/output error
FATAL: Error running install command for acpi_cpufreq
[root@Arch ~]# modprobe p4-clockmod
FATAL: Error inserting p4_clockmod (/lib/modules/3.0-pf/kernel/drivers/cpufreq/p4-clockmod.ko.gz): Invalid argument
Is it a problem with the kernel, as http://aur.archlinux.org/packages.php?ID=24980 would suggest or does this Celeron really have no way of lowering its clock speed as pages on the Internet state? Thanks.