append = "mem=96m"
to the linux entries in /etc/lilo.conf
(and run lilo afterwards, naturally). I've had this
memory for more than a month now, and I've had no problems with it.
cardctl eject before hibernation is sufficient.
If you have already used Windows for a while, you might want to defrag your disk. There is a tool for that in the Win95 start menu. start->Programs->Accessories->System Tools->Disk Defragmenter
I used fips (from dosutils/ on the RedHat CD) to shrink
the Win95 partition down to some 700 MB. Probably far too much, but who
knows? 4 GB is still huge. After this, you may want to use
scandisk to see that your Win95 partition is still OK.
I walked the RH installer, not touching much. I used Disk Druid to partition the 4G drive like this:
| partition | size | type | mount point or comment |
|---|---|---|---|
| /dev/hda1 | 700 MB | W95 FAT32 | pre-existing |
| /dev/hda4 | 149 MB | 0x0a | save to disk partition |
| /dev/hda5 | 200 MB | Linux Native | / |
| /dev/hda6 | 1027 MB | Linux Native | /usr |
| /dev/hda7 | 127 MB | Linux Swap | swap |
| /dev/hda8 | 1700 MB | Linux Native | /home |
In hindsight, I would probably have preferred making the root partition a
primary partition (/dev/hda2) and having the /usr,
swap and /home partitions in the extended partition
/dev/hda3. This would let me install the boot loader in the
/dev/hda2 partition and avoid touching the MBR - an advantage if
I ever need to reinstall Windows. Of course, this means using
fdisk to partition the disk - Disk Druid will put all Linux
partitions in a single extended partition.
After selecting packages, the rest of the install went without problems.
startx -- -bpp 16 but a
more permanent solution is to add the following line to the
Screen section of /etc/X11/XF86Config:
DefaultColorDepth 16
Due to the pressure from the Open Source community, NeoMagic recently decided to allow Precision Insight to release the code written for the XBF X server for the NeoMagic graphics cards, and the driver has been renamed XFCom (XFree86 Compliant). The code written for the XBF_NeoMagic server has been handed over to the XFree86 project, and I understand that support for NeoMagic is provided in XFree86 3.3.3. The XFCom_NeoMagic X server is available as rpm and gzipped tar for libc5 and glibc2 from RedHat's FTP server and several mirrors worldwide.
Here is the press release from RedHat on the matter.
Here is Precision Insight's own page on the collaboration, with download links and links to download mirrors worldwide.
apm -s command.
If suspended by simply closing the lid, the machine resuspends itself
immediately upon waking up, but when waking it up a second time,
everything is usually fine. Suspending it with a click on the power
button or with apm -s avoids this problem.
The computer can be hibernated (state saved to disk and machine completely
turned off -- uses no battery) by pressing Fn+F12. This will
only work if you have ejected (either physically or logically
using cardctl eject) all PC cards!
The computer can be set in standby mode (not as "deep sleep" as
suspending it; any keypress or mouse motion will wake it up) through the
apm -S command. If the computer is connected to the network,
any activity on the network will wake it up.
As a matter of fact, I'm beginning to wonder whether this is a problem with the driver or a hardware problem. I'd very much like to hear from others using this machine and/or mouse to hear if they experience anything similar.
There is a port for external keyboard/mouse on the back of the machine. I have tried plugging in an external mouse (Logitech 3-button mouse) without configuring anything. This does not work satisfactory. It has to be plugged in outside of X to do anything at all, and if you as much as touch the mousepad while it is plugged in, unpredictable behaviour results. This usually means random movement all over the screen and button-presses all over the place. Usually, the mouse cursor ends up stuck in a corner with neither mouse able to move it. I have therefore given up using this external mouse for now.
In later kernels than the 2.0.35 I'm using, some of the entries that need to be specified don't appear in the configuration menu, and I've not been able to make sound work with these newer kernels. Perhaps it'll be back and working in 2.2
The config settings I used for building the sound module are given below
| Sound card support | M |
| Support for Crystal CS4232 based (PnP) cards | * |
| /dev/dsp and /dev/audio support | * |
| CS4232 audio I/O base | 530 |
| CS4232 audio IRQ | 5 |
| CS4232 audio DMA | 1 |
| CS4232 second (duplex) DMA | 0 |
| CS4232 MIDI I/O base | 330 |
| CS4232 MIDI IRQ | 9 |
| Audio DMA buffer size | 65536 |
# # Sound # CONFIG_SOUND=m # CONFIG_PAS is not set # CONFIG_SB is not set CONFIG_ADLIB=y # CONFIG_GUS is not set # CONFIG_MPU401 is not set # CONFIG_UART6850 is not set # CONFIG_PSS is not set # CONFIG_GUS16 is not set # CONFIG_GUSMAX is not set # CONFIG_MSS is not set # CONFIG_SSCAPE is not set # CONFIG_TRIX is not set # CONFIG_MAD16 is not set CONFIG_CS4232=y # CONFIG_MAUI is not set CONFIG_AUDIO=y CONFIG_MIDI=y # CONFIG_YM3812 is not set CS4232_BASE=530 CS4232_IRQ=5 CS4232_DMA=1 CS4232_DMA2=0 CS4232_MPU_BASE=330 CS4232_MPU_IRQ=9 DSP_BUFFSIZE=65536 # CONFIG_LOWLEVEL_SOUND is not set
All items not specifically mentioned here are not marked. With this setup, I am now able to play and record sound with the built-in speakers and microphone. The built-in microphone is located just beneath the center of the screen, in the notch above the small LCD display. It is not exactly hi-fi, but speech is recognisable.
I have not made many attempts to explore how much I can do with the sound system. I have not tried to install OSS drivers or any such. I have gotten report that MIDI support will work if I enable the selections for "Generic OPL2/OPL3 FM synthesizer support" and "MIDI interface support", but I have not had time to play with this.
Thanks to Clive Messer and Petter Reinholdtsen for pointing out obvious mistakes I had made in setting up the sound system.
The overall goal of this project is to incorporate an IrDATM compliant protocol stack into the Linux kernel.
Although the code is in an early stage of development, and its author warns that using it could crash your computer, I have not experienced any crashes while trying to use the modules from this distribution. (I have had kernel misbehaviour which made me reboot, though.)
In order to use Linux/IR, the first step is to enable the infrared port from the devices menu in BIOS setup. (I failed to to this, and naturally, nothing worked.)
After enabling the IR port in BIOS and assigning it IRQ 4, it showed up as
tty02 or /dev/ttyS2 on my computer. With the
snapshot of 1998-10-21, a friend with a Compaq Armada 7380DMT and I
have made some experiments with the package. With the Armada as the
IrLAN server, we managed to ping each other for up to 15
seconds at a time. With the OmniBook as the server, we had a LAN
connection (ping, telnet, ssh etc.) for several minutes. After this time,
the connection suddenly died and we were unable to revive it.
The bottom line is: The OmniBook 3000 CTX is able to use the IrDA kernel modules in their present state, although the IR port must be enabled in BIOS before it is available to Linux. The Linux/IR project is not in a finished state, so it is not particularly obvious to set up or use. The documentation of the project is not up to date with the code as I write this.
Admittedly, lots has happened in the Linux/IR project since last time I tried to do anything. Linux/IR is now part of the kernel source release for Linux 2.2 (as an experimental feature) and all development is currently concentrated on the 2.2 release, so I will not experiment further with this until I've upgraded to 2.2.
| Main | Time | Date | Quiet boot | Video display device |
|---|---|---|---|---|
| (set to GMT) | today's date | enabled | Auto | |
| System devices | PS/2 mouse | External Fn key | IDE controller | FDD controller |
| Enabled | Enabled | Both | Enabled | |
| Serial port | Infrared port | Parallel port | ||
| Enabled, addr 2F8, IRQ 3 | Enabled, mode FIR, addr 3E8, IRQ 4, DMA 3 | Disabled | ||
| Audio | SB I/O address | WSS I/O address | AdLib I/O address | |
| Enabled | 220h | 530h | 388h | |
| Interrupt | 1st DMA | 2nd DMA | ||
| IRQ 5 | DMA 1 | DMA 0 | ||
| MPU device | I/O base | Interrupt | ||
| Enabled | 330h | IRQ 9 | ||
| Security | Password on boot | Password on resume | ||
| Enabled | Disabled | |||
| Power | Power management mode | Time-out on AC | Power button mode | |
| Disabled | Disabled | suspend | ||
| resume on modem ring | resume on time of day | |||
| Disabled | Disabled | |||
| Boot | 1: Hard Drive | 2: CD-ROM | 3: Diskette | |
|
Notes: I have fiddled with the settings in the "System Devices" menu.
I had to disable the parallel port in order to enable the FDD
controller. This should not be necessary. I don't know what WSS or AdLib is, but those address settings look like the ones I use for configuring the sound module in the kernel. It appears MPU is Midi Processing Unit or something like that. Disabling password on resume is probably less than safety-minded. I have completely disabled the power management in BIOS, trusting Linux to do the job, but I am experimenting with "Custom" settings, using timeouts for standby and suspend, but disabling automatic hibernation, naturally. The boot sequence is that way on purpose. Without my BIOS password, you can't easily boot something which is not already on my hard drive. |
||||
PCI devices found:
Bus 0, device 4, function 1:
CardBus bridge: Texas Instruments PCI1131 (rev 1).
Medium devsel. IRQ 255. Master Capable. Latency=64. Min Gnt=192.Max Lat=3.
Bus 0, device 4, function 0:
CardBus bridge: Texas Instruments PCI1131 (rev 1).
Medium devsel. IRQ 255. Master Capable. Latency=64. Min Gnt=192.Max Lat=3.
Bus 0, device 2, function 0:
VGA compatible controller: Neomagic MagicGraph NM2160 (rev 0).
Medium devsel. Fast back-to-back capable. IRQ 255. Master Capable. Latency=128. Min Gnt=16.Max Lat=255.
Prefetchable 32 bit memory at 0xfd000000.
Non-prefetchable 32 bit memory at 0xfea00000.
Non-prefetchable 32 bit memory at 0xfed00000.
Bus 0, device 1, function 3:
Bridge: Intel 82371AB PIIX4 ACPI (rev 1).
Medium devsel. Fast back-to-back capable.
Bus 0, device 1, function 2:
USB Controller: Intel 82371AB PIIX4 USB (rev 1).
Medium devsel. Fast back-to-back capable. IRQ 255. Master Capable. Latency=64.
I/O at 0xfcc0.
Bus 0, device 1, function 1:
IDE interface: Intel 82371AB PIIX4 IDE (rev 1).
Medium devsel. Fast back-to-back capable. Master Capable. Latency=64.
I/O at 0xfcf0.
Bus 0, device 1, function 0:
Bridge: Intel 82371AB PIIX4 ISA (rev 1).
Medium devsel. Fast back-to-back capable. Master Capable. No bursts.
Bus 0, device 0, function 0:
Host bridge: Intel 82439TX (rev 1).
Medium devsel. Master Capable. Latency=32.