I have been everywhere and tried everything and have come up empty.

I have an Asus A7V socket A motherboard with a VIA Apollo KT133 chipset. This means that I have ATA100 or UltraDMA/100 support for certain Hard Drives . I have two Fujitsu model MPG3409AT Hard drives, one set as master, one as slave, both attached to the PRIMARY Ultra ATA 100 IDE connector.

These drives are working fine with Windows98SE as ATA100 devices.

However, I am unable to get RedHat Linux to deal with the drives unless I move the connector to the PRIMARY IDE (non ATA100) connector.

As yet, I have not found anything at the RedHat URL which helps me with this. And I have found only one article that provides much of anything on the subject . This article is a HowTo for Asus A7v and Linux ATA100, but for Linux V6.2 (Aaron Cline's Quasi-mini-howto).

According to Mr. Cline, I should boot from either a floppy or the Install CD, providing the following:
--------------------------------------------------------------------------------
boot: vmlinuz ide2=0x9400, 0x9002 ide3=0x8800, 0x8402
---------------------------------------------------------------------------------
When I do this, it looks like the boot process gets started, and I can see references to the two hard drives, however, system locks up. Here's what I get:

---------------------------------------------------------------------------------
IRQ probe failed (0xfffffff8)
No response status (status = 0x0a)
Partition check hde_
---------------------------------------------------------------------------------

At this point the computer locks up and is unresponsive.

Hasn't anyone ever been successful with ATA100 recognition?

I'm not familiar with the Via chipsets, but is the ATA100 controller part of it (like the i815) or is it a mainboard add-on (like the highpoint controllers).

Some things to try... You could disable the Primary (non ATA100) controller. This should set you up to have the ATA as IDE0 - /dev/hda and /dev/hdb drives on the chain.

Something I have to do a lot is to tell linux to reverse the device ordering of the controllers. This would reverse the primary and secondary controllers. At the linux boot prompt
This is assuming that you have the default boot label entry named 'linux' in your lilo.conf. If you can get this working manually you can add the 'ide=reverse' to the append line in lilo.conf. pci=reverse might do it too. You can see the effects of these parameters by looking at the dmesg and checking the device names. When I don't pass the ide=reverse to my kernel at boot, my drives on the Promise ATA/66 controllers start at /dev/hde then /dev/hdg, etc as they are seen 'after' the onboard controllers even if they have been disables in the BIOS. There's also a kernel option when you compile to have the PCI bus scanned by setting the CONFIG_PCI_GOBIOS. From the 'help' button in my menuconfig...

Thanks for the input.
I finally ended up reinstalling my system. I didn't know that RedHat linux V7.1 was capable of detecting the ATA 100 devices.
After a complete re-install and configuration of scrollmouse, printer, ppp, etc, I have a systme that is operational.

Hello
It seems as if you have a working system now but should you desire to make it simple or for someone else with similar problems there is also a kernel option
"Boot off-board chipsets first support" CONFIG_BLK_DEV_OFFBOARD in 2.4.x kernels that can be very helpful. This is especially usefull if instead of supplanting your onboard controller you must use them both. A common example might be 4 hard drives on a PCI accellerator and a CDROM connected to the mobo's onboard IDE controller. Depending on BIOS it may also be necessary to append the "ide=reverse" call to lilo or grub to avoid kernel panic by insuring that what would otherwise be /dev/hda,b,c,or d doesn't become /dev/hde,f,g,or h. It's a bit of a shame it isn't clearer but then it is apparently due to differences in BIOS and hardware detection, not necessarily a problem with one place to accept blame.

Jimmy