GNU/Linux on Fujitsu Lifebook S-4542

Windows 98.  Trident Cyber9525DVD PCI/AGP(W98.26). Monitor: Laptop Display Panel (1024x768)

Status: working. 1024x768 with 16 bit colour. I am running XFree86 4.0.1 as part of the Red Hat 6.9.5 Pinstripe release (the 7.0 public beta), so all my comments refer to this release, if you are using XFree86 3.3.x, YMMV.

The installation was relatively straightforward, once I removed the old 3.3.x SVGA server package that the Red Hat installation script seemed to want install, despite the fact that these servers are now part of XFree86 itself. The new approach is to run:

$ XFree86 -configure

This creates the /etc/X11/XF86Config file. You should also set the default colour depth to 16 bit, since the server will default into 8 bit, unless otherwise specified.

My XF86Config file.

Windows 98. Fujitsu LB Global LTMODEM. Driver info: Lucent 11-4-99.

Status: working. Using the Lucent winmodem (binary-only!) driver provided by Lucent themselves (see linmodems.org for more details on how to find this driver), this I got working pretty quickly.

Getting ppp to work was another story. Unfortunately, the Lucent folks built the driver against the 2.2.12 kernel, and that wouldn't be a problem (except you have to use the -f option to insmod to get the module to ignore the kernel version complaints and load) except for the fact that the ppp module itself has changed for kernel versions 2.2.15 and later, which means that the binary interface to that module has changed, causing a kernel panic when loading the ppp module. Meaning that you can run a serial connection via minicom fine, but crash when trying to run pppd

The temporary solution is to recompile (or find a recompiled version) of the 2.2.14 version of the ppp module, against the kernel version you are running (if it's 2.2.15 or later) and then replace the version bundled with your distribution with this recompiled version. Not pretty, I know.

The real solution would, of course, be for Lucent to release the source to their driver!! Meaning that all the above would be rendered moot, because you could simply recompile the driver! Come on guys, why not? It's not like a device driver are the crown jewels or something.

Status: working. Worked pretty much “out of the box”. Uses one of the Intel Etherpro adapters on the motherboard. Don't have specs in front of me, e-mail me if you really need them.

card "NETGEAR MA401 Wireless PC"
  version "NETGEAR MA401 Wireless PC", "Card", "Version 01.00"
  bind "orinoco_cs"

Windows 98. Intel(r) 82449MX AC'97 Audio Controller. Driver: SigmaTel Codec Hardware version: 000 (Date 2-29-2000). Interrupt Request (IRQ): 09. I/O Range 1000-10FF, I/O Range 1400-143F

Status: working. Initially didn't seem to be working (no sound from the CD player or from the window manager), but after fiddling around with aumix and gmix and adjusting the volume, it come to life. I used to have problems on shutdown, when the whole shutdown hangs when trying to "save mixer settings" or hangs using aumix or gmix after bringing the machine out from hibernation. I solved this with a helpful suggestion from Steve Hofmeyr.

The Advanced Power Management daemon package, apmd, has an option that unloads the sound modules before going into "Suspend", and reload them again after resuscitating the machine. You need to modify /etc/sysconfig/apmd:

# if your sound sometimes doesn't work after a resume set RESTORESOUND to "yes"
# this will kill all sound applications, remove and start the $SOUNDMODULES
# again, and, if you also set $RESTORESOUNDPROGS to "yes", restart the
# sound applications as the correct user and on the correct display

RESTORESOUND="yes"
RESTORESOUNDPROGS="yes"

# working version that works with Red Hat 7.2
SOUNDMODULES="sb uart401 sound soundcore maestro cs4281 i810_audio"
# obsolete version which worked with Red Hat 7.0
#SOUNDMODULES="sb uart401 sound soundcore i810_audio ac97_codec"
# original version which shipped with Red Hat 7.0
#SOUNDMODULES="sb uart401 sound soundcore"

i.e. the first two variables RESTORESOUND and RESTORESOUNDPROGS should be changed from their default values of "no", to those above. The SOUNDMODULES variable has also changed. Please choose the line that matches your distribution of Red Hat.

The above change fixed the problem with freezing on accessing aumix, but didn't solve the problem of saving mixer settings. To this end, following the APM sections in the Laptop-HOWTO you need to create a /etc/sysconfig/apm-scripts/apmcontinue file, with (something like) the following contents:

#!/bin/bash
PROG="$1"
case "$PROG" in
    suspend)
        logger "apmcontinue: suspending saving mixer settings"
        /bin/aumix-minimal -f /etc/.aumixrc -S || logger "couldn't save settings"
        ;;
    resume)
        # resume mixer settings...
        logger "apmcontinue: resuming mixer settings"
        /bin/aumix-minimal -f /etc/.aumixrc -L || logger "couldn't restore settings"
        ;;
    change)
        ;;
    start)
        ;;
    stop)
        ;;
    *)
        ;;
esac

If this file exists, it will be called by the standard apmscript and the various functions, suspend, resume will be called at the right stage in the suspend sequence. You can use these functions to do any custom saving for apm. You may need to fiddle with permissions on the /etc/.auxmixrc file to get this to work, consult the Laptop-HOWTO for more details.

Status: fully working. As of mid-2002, I've had full DVD playback working with VideoLan (the vlc player), Xine, MPlayer and ogle. (I originally started with the LiViD (Linux Video) project using the OMS (Open Media System) player, which appears to have been orphaned.) These players (VideoLan, Xine, and MPlayer) now have binary packages (check out Red Hat 7.3 (Valhalla) RPMS or Red Hat 8.0 (Psyche)). The good news is that kernels shipped with Red Hat releases since 7.0 have the DVD ioctls built-in, so you don't need to mess around with kernel recompiles.

# These options are used to tune the hard drives -
# read the hdparm man page for more information

# Set this to 1 to enable DMA. This might cause some 
# data corruption on certain chipset / hard drive
# combinations. This is used with the "-d" option

USE_DMA=1

# Multiple sector I/O. a feature of most modern IDE hard drives,
# permitting the transfer of multiple sectors per I/O interrupt,
# rather than the usual one sector per interrupt.  When this feature
# is enabled, it typically reduces operating system overhead for disk
# I/O by 30-50%.  On many systems, it also provides increased data
# throughput of anywhere from 5% to 50%.  Some drives, however (most
# notably the WD Caviar series), seem to run slower with multiple mode
# enabled. Under rare circumstances, such failures can result in
# massive filesystem corruption. USE WITH CAUTION AND BACKUP.
# This is the sector count for multiple sector I/O - the "-m" option
#
# MULTIPLE_IO=16

# (E)IDE 32-bit I/O support (to interface card)
#
# EIDE_32BIT=3

# Enable drive read-lookahead
#
# LOOKAHEAD=1

# Add extra parameters here if wanted
# On reasonably new hardware, you may want to try -X66, -X67 or -X68
# Other flags you might want to experiment with are -u1, -a and -m
# See the hdparm manpage (man hdparm) for details and more options.
#
EXTRA_PARAMS=-X34
   

Windows 98. Intel 82440MX PCI to USB Universal Host Controller. IRQ: 15 I/O Range 14C0-14DF. USB Root Hub (?)

Status: untested. Haven't tested as yet, with the new USB module in more recent kernels 2.2.16, 2.2.17 or even the 2.4.0 series, shouldn't be too hard.

Some miscellaneous application-specific notes using this laptop, and in general on GNU/Linux systems.

OpenOffice.org. Sun's now LGPLed version of StarOffice seems to work fine. I've been running 1.0 (from April 2002). The problems (OpenOffice would hang) with opening Impress (Presentation) documents that were created in 641b seem to have gone away.