Linux on the ASUS L3800C Laptop

The machine features a P4m CPU and these I/O ports: serial, parallel, VGA, FireWire, USB, PCMCIA, modem, LAN and S-Video. There are an in-built floppy and DVD/CD-RW drive. The case remembers me to IBM ThinkPad A series models and is well manufactured.

Installation

The ASUS L3800C (FCC ID: 60HTAI-35453-M5-E , UL E187242 6G24) was delivered without any pre-installed operating system! I have started the installation of Debian GNU/Linux - Woody via CD. I have updated to Linux Kernel 2.4.21, PCMCIA-CS 3.2.2 and Advanced Linux Sound Architecture - ALSA 0.9.4.

Devices

BIOS

BIOS options are set as shown below.

X11

I have taken unofficial XFree86 v4.3 Debian packages. Manufacturer information about the graphics chip (including Linux infos) are provided by ATI .

Resolutions work up to 1400x1050@24. You may use the keyboard to switch resolutions: Fn + Ctrl + Alt + Plus(blue) or xvidtune.

DRI

3D acceleration works. Klaus Knopper reports: "To make DRI work you need at least XFree86 4.3 . After that you may check GLX support with glxgears and chromium.

Caveat: As soon as DRI is active the SoftWareSuspend Patch doesn't work anymore. The X-Server doesn't wake up after suspend. Workaround: Start XFree86 with Xinerama support, which disables DRI. If you want to play a GLX game then, just start a DRI enabled second X-Server and finish this server before using SoftWareSuspend to hibernate.

Karol 'sziwan' Kozimor wrote: "Charl P. Botha was so kind to provide a patched version of DRM modules and DRI drivers that work very well with SoftWareSuspend. The only caveat is that the machine must be switched to a text console prior to suspend (or more specifically, the DRI reinit is done during a console switch), but this is normally handled by the swsusp code. The files are at http://cpbotha.net/dri_resume.html. Note that AFAIR the suspend/resume code was integrated into recent XFree86 snapshots, chances are these files are no longer required using a recent version of XFree86."

Martin Hedegaard wrote: "After a working a on my laptop setup again I went for at vanilla 2.4.23 kernel with the newest swsusp (2.0 rc3) and ACPI patches and just using the original hibernate script , and I can report that the swsusp patches don't brake the DRI modules any more. There is some small changes there have to be made in the setup to get it working. In lilo.conf

append="resume=/dev/hdaX"

has to be replaced with

append="resume2=swap:/dev/hdaX"

if using SWAP. I am using KDE 3.2 beta 2 and the integration from the klaptopdaemon works just fine too, booth with speedstep and swsusp."

TouchPad

The touchpad works with /dev/psaux. The synaptics driver works great and gives you the mouse buttons you miss. For X11 always disable gpm to prevent the mouse from behaving weird.

External Mouse

ASUS delivers an external USB mouse together with the laptop. I had to add an additional section to /etc/XF86Config-4 to make it work.

Console

To get a better resolution (1024x768) I use Framebuffer mode vga=791.

The Linux Kernel ATI Radeon FrameBuffer driver works very well for the console. The panel dimension is auto detected. But the default depth is 8bpp and can only be set using fbset.

PCMCIA

PCMCIA works out of the box. Output of probe:

PCI bridge probe: Ricoh RL5C476 found, 2 sockets. 

With Novell/SuSE 8.x you have to disable kernel PCMCIA (in /etc/sysconfig/pcmcia set PCMCIA_SYSTEM="external").

NIC

The network interface works with the rtl8139 kernel driver (2.2.19). And 8139too for 2.4.x Kernels. You may use the full_duplex=1 module option. For details on this driver see the Linux and the RealTek 8129/8139 chips page.

KeyBoard

Function Keys

Hotkeys

There are four silver extra buttons above the keyboard at the left of the power button. They may be configured via ACPI-Daemon, if the module asus_acpi is loaded.

Sound

Sound works with Advanced Linux Sound Architecture - ALSA driver snd-intel8x0 . I'm not sure whether the hardware supports MIDI, but timidity is a nice workaround.

The Linux Kernel i8xx ICH OSS compatible sounddriver also works very well. To use this take a kernel >= 2.4.19."

Modem

Wouter Scheele reported: "The modem is an Intel Corp. 82801CA/CAM AC'97 Modem (rev 2)., I downloaded the pctel-0.9.6 driver and compiled the modules using:

./configure --with-hal=i81x
make
make install

loaded the modules, and... it works. :)"

Note: As soon as the modem modules are loaded, sound does not work any more. To get sound back to work you have to reboot.

Karol 'sziwan' Kozimor wrote: "Actually, the sound works after the pctel/ptserial modules are loaded. The sound stops at any modem activity (i.e. dialing a number), but it is easily regained. Just remember not to have anything playing when you start dialing the number (i.e. stop XMMS or similar, pause isn't enough; not that it is dangerous, but the applications tend to lock that way), and after dialing fiddle with the mixer controls. You may have to reload the sound modules afterwards to get proper clocking frequency, which is sometimes altered when the modem modules come to play. This works for ALSA, for OSS, AFAIR no sound until modules are reloaded."

Eric Valette <eric.valette_at_free.fr> wrote: "I tried all the pctel driver including the new 0.9.7 beta releases with absolutely no luck. I can load the modules, issue basic AT command (ATZ) using the pseudo tty created but never ever managed to really dial or establish a connection. And also devfs support was rather poorly supported in this driver.
I have seen other successful attempts but also found other people like me that never managed to get the pctel driver to work on this hardware. I wonder if it is because, as I have a miniPCI WLAN board, the modem is actually in AMR slot or I have a never/faster processor and timeouts are not really accurate. But in any case, the slmodem-2.9.1.tar.gz code located at LinModems Works for me (even with devfs)."

Daniele Patoner reported: "With this module the modem works well and without sound problem. Download slmdm-2.7.10.tar.gz from LinModems".

Infrared Port - IrDA(TM)

I couldn't get any device information about FIR support.

SIR works out of the box with /dev/ttyS1. For details on getting IrDA to work see the Linux-InfraRed-HOWTO.

ttyS00 at 0x03f8 (irq = 4) is a 16550A
ttyS01 at 0x02f8 (irq = 3) is a 16550A

"Karol 'sziwan' Kozimor wrote: FIR is "as simple as modprobe nsc-ircc, but must be done *without* serial.(k)?o loaded, or an appropriate setserial command must be issued to free up the resources. Upon a successful load, the following lines appear:"

nsc-ircc, Found chip at base=0x02e
nsc-ircc, driver loaded (Dag Brattli)
IrDA: Registered device irda0
nsc-ircc, Found dongle: HP HSDL-1100/HSDL-2100

Power Management

Though ASUS states in the manual that the machine is capable of ACPI only, both APM and ACPI work, at least partly. Note: APM and ACPI are mutually exclusive, enable or disable them via boot option, e.g. apm=off acpi=on or vice versa.

Activating APM over ACPI can cause strange fan operation, i.e. too much fan activity. There is NO problem with it except for too much noise. Your machine is still safe!

APM

The power button pressed less than 4 seconds puts the machine into suspend or soft off mode. Please note, though suspended the machine stays very hot, so it still consumes power!

• BIOS option: soft off, machine doesn't wake up after power off < 4 seconds
• BIOS option: suspend, works fine

The key combination Fn + zZ works with both BIOS APM modes. For user-space APM support you need the apmd package.

For suspend to disk via SoftWareSuspend see below.

ACPI

Works only partly (all modules load without errors, and /proc/apci shows up, but all else fails, e.g. putting the machine into sleep mode). So I still prefer APM, because I can use the suspend feature. Maybe future Kernels will work better. You need the acpid package to make advantage of the ACPI features.

With Novell/SuSE 8.x ACPI works, e.g. the battery ACPI module works great.

cat /proc/acpi/battery/BAT0/info
.
present:                 yes
design capacity:         67200 mWh
last full capacity:      67200 mWh
battery technology:      rechargeable
design voltage:          14000 mV
design capacity warning: 6720 mWh
design capacity low:     3360 mWh
capacity granularity 1:  840 mWh
capacity granularity 2:  840 mWh
model number:            L3500
serial number:
battery type:            LIon
OEM info:                ASUSTek

cat /proc/acpi/battery/BAT0/state
present:                 yes
capacity state:          ok
charging state:          discharging
present rate:            0 mW
remaining capacity:      67060 mWh
present voltage:         12040 mV

You can even switch the processor frequency

cat /proc/acpi/processor/CPU0/performance
state count:             2
active state:            P0
states:
    *P0:                  1700 MHz, 20000 mW, 250 uS
     P1:                  1200 MHz, 10000 mW, 250 uS
.
echo 1 > /proc/acpi/processor/CPU0/performance
.
cat /proc/acpi/processor/CPU0/performance
state count:             2
active state:            P1
states:
     P0:                  1700 MHz, 20000 mW, 250 uS
    *P1:                  1200 MHz, 10000 mW, 250 uS

And the performance changes!

Karol 'sziwan' Kozimor wrote: "Contrary to what is written [above], ACPI is very well supported on this particular laptop. As of state transitions, they *do* work, though require some work. As of 2.4.x, you *must* remove the (usb-)uhci module, or else the laptop will immediately hang upon resume from S1. Everything else may be loaded, just this module must be removed. This is not the case for 2.5.x, S1 works there. In both cases, the screen doesn't go off upon suspend, but that's not a problem, since there's /proc/acpi/asus/lcd for that. S3 also works under 2.5.x, though you must pass acpi_sleep=s3_bios to regain the LCD backlight upon resume, and reload atkeyb (and possibly others responsible for the keyboard support) to get the keys back. S4 is software suspend, which simply works."

Software Suspend

The SoftWareSuspend doesn't depend necessarily on APM or ACPI (see the difference between pure SoftWareSuspend and SoftWareSuspend with ACPI mentioned in the SoftWareSuspend docs).

Gerald Heim wrote: Good news: SoftWareSuspend - SWSUSP (hibernation) + ACPI actually work on the L3C! A short HOWTO of what I did to install:

• Upgrade to latest stable kernel 2.4.x
• Install latest ACPI
• Install latest swsusp

Of course this will break PCMCIA-CS and Advanced Linux Sound Architecture - ALSA, but these are easy to rebuild from source.

Klaus Knopper reported: Here is the kernel 2.4.20 config file to get SoftWareSuspend to work. For suspending I don't use the original hibernate script, but echo 1 0 2 > /proc/sys/kernel/swsusp because X and PCMCIA wake up easily and even with the configured IP address of the WaveLAN card. It takes 10 seconds to power off. Power on takes 20 seconds until everything is reloaded from SWAP. The SWAP partition which has to be configured with append="resume=/dev/hdaX" in the bootloader (e.g. LILO) should be at least 1.5 times higher than the RAM.

Harddisk

The HDD is S.M.A.R.T. capable, take the ide-smart package to make use of it.

DVD/CD-RW

The CD-RW drive works fine on e.g. /dev/scd0 with IDE-SCSI emulation. See CD-Writing-HOWTO from The Linux Documentation Project - TLDP for details. No time to check the DVD with xine or mplayer yet. But Henrik Christian Grove reported: "DVD-playing works nicely with ogle, (gnome-)vlc and xine (haven't tried mplayer)."

Karol 'sziwan' Kozimor wrote: "Tested with mplayer, works. The S/PDIF (optical output) doesn't work with mplayer yet, as of ALSA 0.9.1 it was not supported on this chipset."

Floppydrive

I couldn't get the drive to superformat floppies higher than 1.44MB.

USB

Works out of the box. Tested with External Keyboard, USB Mouse. There is no USB 2.0 support.

FireWire

The ohci1394 module loaded fine. But since I don't own a FireWire device I couldn't check this any further.

Battery Life

The ASUSTek Battery Calibration Utility may be started from the BIOS and takes 1.5 hours to calibrate the battery. This may also taken as battery uptime for normal usage.

Karol 'sziwan' Kozimor wrote: "The uptime is closer to 3 hours, depending on the CPU usage."

TV-Out

You may switch the TV display type between NTSC and PAL in the BIOS. ASUS delivers an external adapter cable for the TV port: S-VIDEO-2-RCA together with the laptop. TV-Out didn't work, I got only a flicker, when changing from the internal to the external display with the function key.

Even atitvout didn't help. atitvout is a tool for enabling/disabling the TV out connector and switching of TV standards (PAL, NTSC, etc.) of ATI Rage graphics boards from within Linux on IA32. It is especially useful on ATI Rage Mobility boards, which boot up in NTSC mode by default and need to be switched to PAL for use with German TV sets. With the L3800C you can switch between PAL and NTSC in the BIOS. Note: For the Debian package see atitvout.

But Henrik Christian Grove reported: "I had no troubles using the TV-Out, with a S-Video-to-SCART-cable (and a little colorfix-cable). There are some problems though:
- I wasn't able to get a picture on both the LCD and the TV
- When I had a picture on the TV, the only way to get a picture on the LCD again, was to disconnect the cable from the TV-Out port.
Both problems may caused by a cheap and lousy TV. (I'm not able to have a picture on both the LCD and my - cheap and lousy - CRT at home, but I have no such problems elsewhere).

Karol 'sziwan' Kozimor wrote: "I couldn't get it to work under X11. Since my TV obviously can't handle 1400x1050, I couldn't test it with radeonfb, but I heard it works. I had the BIOS output and LILO prompt on my TV, though. I also tested it with an external monitor and Xinerama, works fine, even with overlay (it is possible to freely move an mplayer window between two displays)."

Martin Hedegaard wrote: "TV-out works just fine on the L3800 laptop series with the generic Video Electronic Standards Association - VESA driver and I got it working in 1024x768 atitvout. I know this exclude the use of 3D but I solved that with making a litte bootscript on my Gentoo box that runs before X11 starts that looks like this:

ebegin "Checking for TV"
if /usr/bin/atitvout detect | grep -q TV; then
ebegin "Using S-Video out"
cp /etc/X11/XF86Config.tvout /etc/X11/XF86Config
else
ebegin "Using LCD video"
cp /etc/X11/XF86Config.notv /etc/X11/XF86Config
fi
eend $?

It looks for a TV with atitvout and copy a config that runs VESA and 1024x768 if the TV is deteted and then you should just make atitvout -f t and it is working.

WLAN miniPCI

Karol 'sziwan' Kozimor wrote: "The L3800C, to my best knowledge, uses an integrated version of the rtl8139 [LAN, WH] chip, and the miniPCI slot is either left empty, or contains a WLAN card."

Martin Hedegaard wrote: "I just found got my new ASUS L3800Ce laptop (its the same as the L3C and L3800C). It has a wireless LAN card on the miniPCI port that is identified as a "Intel(R) PRO/Wireless 2011b LAN 3A Mini PCI Adapter" under Microsoft-Windows. It works with the prism2_pci driver from the linux-wlan-ng project (Debian package: linux-wlan-ng)."

...
02:08.0 Network controller: Harris Semiconductor Prism 2.5 Wavelan chipset (rev 01)
...

Kernel 2.6

Martin Hedegaard wrote: After running the 2.6 series kernels for a while I thought it was a good time to come with an update on the L3800c and kernel 2.6.

Since last update I have updated to a vanilla 2.6.6-rc3. A lot off small things have been fixed. The Radeon framebuffer module has been nearly fixed, and FireWire is working nicely again after being broken badly in 2.6.4 and 2.6.5 causing hard lockups when moving big files to my external hard drive.

The internal wlan card in my model is since 2.6 supported by the kernel which is very nice too. In the 2.6 kernel scsi emulation is no longer necessary, it works fine without on this laptop and you can still burn cd's. When it comes to swsusp I haven't done any testing but I saw at the Gentoo forums that it might work with the newest swsusp2 patches. The ALSA sound in the kernel works out off the box, with no problems. I don't use the modem and therefore I have not tested it but maybe it will work with one off the solutions on the page. The normal PCMCIA driver is still broken but yenta_socket works so it doesn't matter.

The new implementation off the touchpad in 2.6 kernel have given me many headaches but I finally fixed it by changing the relevant code in the X config file to make it work:

Section "InputDevice"
  Driver        "synaptics"
  Identifier    "Touchpad"
  Option        "Device"        "/dev/input/mouse0"
  Option        "Protocol"      "auto-dev"
  Option        "LeftEdge"      "1900"
  Option        "RightEdge"     "5400"
  Option        "TopEdge"       "1900"
  Option        "BottomEdge"    "4000"
  Option        "FingerLow"     "25"
  Option        "FingerHigh"    "30"
  Option        "MaxTapTime"    "180"
  Option        "MaxTapMove"    "220"
  Option        "VertScrollDelta" "100"
  Option        "MinSpeed"      "0.02"
  Option        "MaxSpeed"      "0.18"
  Option        "AccelFactor" "0.0010"
  Option        "SHMConfig"     "on"
#  Option       "Repeater"      "/dev/ps2mouse"
EndSection

plus that you have to have the evdev driver loaded as a module or compiled in the kernel.

Another positive thing is that the X11 Radeon driver in Xorg and Xfree86 4.4 now is much better and it is actually possible to play Unreal Tournament 2004, it is not fast but it works.

Everything else than mentioned seems for me to work as in the 2.4 kernels.

Hardware Data

General system information:
cat /proc/cpuinfo

processor	: 0
vendor_id	: GenuineIntel
cpu family	: 15
model		: 2
model name	: Intel(R) Pentium(R) 4 Mobile CPU 1.60GHz
stepping	: 4
cpu MHz		: 1599.875
cache size	: 0 KB
fdiv_bug	: no
hlt_bug		: no
sep_bug		: no
f00f_bug	: no
coma_bug	: no
fpu		: yes
fpu_exception	: yes
cpuid level	: 2
wp		: yes
flags		: fpu vme de pse tsc msr pae mce cx8 sep mtrr pge \
.		  mca cmov pat pse36 cflush dtrace acpi mmx fxsr xmm \
.		  xmm2 ssnp 28 acc
bogomips	: 3191.60

Survey PCI Devices

Output from lspci:

00:00.0 Host bridge: Intel Corp. 82845 845 (Brookdale) Chipset Host Bridge (rev 04)
00:01.0 PCI bridge: Intel Corp. 82845 845 (Brookdale) Chipset AGP Bridge (rev 04)
00:1d.0 USB Controller: Intel Corp. 82801CA/CAM USB (Hub #1) (rev 02)
00:1d.1 USB Controller: Intel Corp. 82801CA/CAM USB (Hub #2) (rev 02)
00:1e.0 PCI bridge: Intel Corp. 82801BAM/CAM PCI Bridge (rev 42)
00:1f.0 ISA bridge: Intel Corp. 82801CAM ISA Bridge (LPC) (rev 02)
00:1f.1 IDE interface: Intel Corp. 82801CAM IDE U100 (rev 02)
00:1f.5 Multimedia audio controller: Intel Corp. 82801CA/CAM AC'97 Audio (rev 02)
00:1f.6 Modem: Intel Corp. 82801CA/CAM AC'97 Modem (rev 02)
01:00.0 VGA compatible controller: ATI Technologies Inc Radeon Mobility M7 LW [Radeon Mobility 7500]
02:05.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139/8139C/8139C+ (rev 10)
02:07.0 CardBus bridge: Ricoh Co Ltd RL5c476 II (rev a8)
02:07.1 CardBus bridge: Ricoh Co Ltd RL5c476 II (rev a8)
02:07.2 FireWire (IEEE 1394): Ricoh Co Ltd R5C552 IEEE 1394 Controller

You may also use the description how to reveal PCI IDs and the survey of miniPCI Ports in Laptops with Linux .

Misc

Knoppix

I have made a short test with a Knoppix 3.1 CD, everything seemed to work fine.

Other Resources

• Unix Laptop Survey
• TuxMobil - hardware section
• Linux Mobile Guide - Installing, Maintaining and Tuning of Linux on Mobile Computers
• Debian Laptop Proposal

Credits

• Wouter Scheele <wouter.scheele_at_audiotex.nl>
• Dr. Olaf Flebbe
• Nils Faerber from Kernel Concepts
• Gerald Heim <heim_at_informatik.uni-tuebingen.de>
• Reiterer Alexander <areitere_at_pop.tuwien.ac.at>
• Klaus Knopper
• Henrik Christian Grove <grove_at_sslug.dk>
• Karol 'sziwan' Kozimor <sziwan_at_hell.org.pl>
• Martin Hedegaard <hedegard_at_rh.kollegienet.dk>
• Eric Valette <eric.valette_at_free.fr>
• Daniele Patoner <daniele.patoner_at_biblio.unitn.it>