Hi!
Shawn Starr wrote:
>
> hda: WDC AC32500H, 2441MB w/128kB Cache, CHS=620/128/63
> hdb: CD-912E/ATK, ATAPI CDROM drive
> hdb: ATAPI 4X CD-ROM drive, 120kB Cache
>
> hdc: FUJITSU MPE3064AT, 6187MB w/512kB Cache, CHS=13410/15/63, (U)DM
>
> I have the above models on my system. As well as a CD-ROM (12x) it doesn't
> see it as a 12x but thats another issue ;-)
>
> The problem I'm getting is this, When DMA is turned on for the FUJITSU
> drive Linux something doesn't like it I get the following error
>
> hdc: timeout waiting for DMA
> hdc: irq timeout: status=0x58 { DriveReady SeekComplete DataRequest }
>
> When DMA is disabled the drive is fine, this DMA timeout happens
> randomly. I dont know whats causing it. I replaced the IDE cables, and
> even replaced the drive for a new one which is the same model same
> problem. This drive can support UDMA33/66.
>
We also experience similar problems with Fujitsu IDE drives
under Linux!
So far we had five(!) lockups on two different Linux systems.
The lockups occur on a non-regular basis, with days or weeks
between the events.
*) System A is used as LAN router, packet filter and IPsec
Gateway. It is running Linux-2.2.16 with IPsec patches.
The machine is located in a computer room with air condition,
so I don't think it's a heat problem.
*) System B is used as PPP dial-up router and packet filter.
It is running Linux-2.2.14. The machine is standing in
a quite small cabinet together with other computers, so
this _could_ be a problem with the temperature.
Both systems have almost identical hardware (purchased at the
same time from the same distributor with the same shipping):
ASUS P5A motherboard (ALI chipset, so no DMA is used!) and
Fujitsu MPE-3136AT HDD. Both system have a AMD K6-2/500 CPU,
64MB of RAM and no other IDE device connected (the Fujitsu
HDD is the only IDE device on the bus)
We don't have log-messages for all lockups, but for some
we found the following messages in the log-files, just a
few seconds before the system hung:
[...]
hda: irq timeout: status=0xd0 { Busy }
ide0: reset: success
[...]
As those systems are in production use at customer sites,
we had to do something fast!
On system A we replaced the MPE3136AT with a MPC3043AT, and
the machine is now up for more than 8 days without problems
(knock on wood)
On system B we replaced the MPE3136AT with a MPE3064AT just
yesterday (yes, I know, but your mail came too late...).
We'll have to see how it behaves...
Does anyone know if there is a problem with Fujitsu drives
(maybe the MPE series)? Any idea how to solve the problem?
I'd rather have rock-solid Linux systems running at customer
sites, as customers tend to become very picky if their
internet connection breaks every week or so... ;-)
- andreas
-- Andreas Haumer | mailto:andreas@xss.co.at *x Software + Systeme | http://www.xss.co.at/ Karmarschgasse 51/2/20 | Tel: +43-1-6060114-0 A-1100 Vienna, Austria | Fax: +43-1-6060114-71- 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/
This archive was generated by hypermail 2b29 : Mon Aug 07 2000 - 21:00:12 EST