> Here is a wierd one for you;
>
> I have a DFI P5BV3+ motherboard with a Apollo Via chipset and a K6/2-350
> CPU. The machine currently has 256Meg memory.
>
> I have a WD Caviar UDMA 4.3gb and a Fujitsu 4.3gb UDMA drive and both show
Skip setting CONFIG_AUTODMA and try again.
Success??
Then split the drives if paired and correct /etc/fstab for new mount
points, then call ide0=dma and ide1=dma with two different boot-ups.
Success??
Call them togather.
Success??
Now enable setting CONFIG_AUTODMA and try again.
Success??
If all of these are working combinations, then your drives don't play well
togather and there are timing errors.
> up as UDMA drives as the system boots up, but the moment the system tries
> to mount the file systems, I get a real fatal error and have to cold
> reboot the system. The error is:
>
>
>
> Unable to handle kernel NULL pointer dereference at virtual address 00000018
> current:>tss.cr3
>
> There is quite a lot more info on the screen but mostly register dumps and
> stack info.
>
> The version of kernel is 2.1.1 and there is an adaptec FDomain SCSI in addition
> to the IDE's which is connected to s DAT tape drive - no disks on scsi.
> When I switch the UDMA to disabled on the bios for the IDE controller, Every-
> thing seems to work fine.
FUJITSU MPC3043AT and MVP3 (via-chipset) don't do UDMA regardless.
The "C" version of the firmware on this drive size only in combination
with a VIA based mainboard only cause known problems. This is even under
Win95/98.
Andre Hedrick
The Linux IDE guy -- http://www.dyer.vanderbilt.edu/server/udma/
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/