> Since 'hda' is attached to the onboard ide-chipset and ide-pci.c: line 168
> is only related to two chipsets (PDC20246 and AEC6210), I can not fully
> understand the relation between the two controller issues.
I know - that's why I thought it might be bus or memory timing issues,
which is why I tried adding nop()/__delay() at various points (with no
unambigous results, unfortunately).
> Second, "PIIX4E South Bridge" ???
>
> "PIIX4" :: sounds like Intel
> "South Bridge" :: sounds like Alladin
My apologies, I didn't have access to the mainboard manual at the time
I wrote the message, so I cut & pasted the description from the Asus
product description page. As far as I can tell, the manual makes no
mention of a "South Bridge", and Alladin are not listed in the trademarks/
copyrights section. The block diagram shows an Intel 440BX APGset
and an Intel PII4XE Chipset. The PII4XE on the mainboard is marked
FW82731EB.
> The next thing that bothers me is the 'C' firmware revision MPC3084AT.
Please note that the DMA disable occurs even if there are no drives
plugged in to the Promise Ultra33.
Also note that the DMA disable action can be prevented by configuring
parallel port support: CONFIG_PARPORT=y and CONFIG_PARPORT_PC=y.
Parallel printer support is not required.
I've done some further testing - by stripping the configuration to
the minimum, I've managed to build a kernel that will reliably not
boot - it either hangs loading /sbin/init (init/main.c:1351), or init
fails to read /etc/inittab and prompts for a run level. If I enable
PARPORT, the kernel disables DMA at the partition check, and then boots
normally. If I comment line 168 out, it boots normally with DMA.
Unplugging the Promise Ultra33 results in a normal boot.
If the kernel disables DMA while booting, using hdparm to turn it back
on fails - at the next disk access, the kernel gets a "bad DMA status"
and disables DMA. If the kernel has not disabled DMA during booting,
I can turn DMA on and off with impunity.
On Fri, 23 Oct 1998, Torbjorn Lindgren <tl@fairplay.no> wrote:
>Anyway, it's definitely worth checking for BIOS upgrades for both the
>motherboard (www.asus.com.tw) and the Promise card (the MB BIOS sounds
>more likely, but...).
>
>Just checking, that 266 MHz Celeron isn't OVERCLOCKED? And the the PCI
>bus is running on 33 MHz. Otherwise all bets are of...
The mainboard is running the latest BIOS (1006). I have two Promise
cards, one with BIOS V1.20, the other with V1.25 (the latest). Both
cards exhibit the same behaviour. Neither the Celeron nor the bus
are overclocked.
I'll scrounge another hard disk tomorrow, and see what difference it
makes to the kernel if hda is changed.
Cheers,
Rick.
-
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/