Subject: Re: (absence of) duplicated inodes
To: Charles M. Hannum <mycroft@gnu.ai.mit.edu>
From: Duncan McEwan <duncan@Comp.VUW.AC.NZ>
List: current-users
Date: 09/28/1994 18:28:53
> Could you describe your hardware *exactly*?  What brand motherboard, disks,
> disk controller? ... (Maybe you already did and I missed it; I tend to not
> pay much attention to the mailing lists these days.)

I already did, but don't mind doing it again if it gets me closer to
understanding what's going on ...!

The motherboard is a micronics "mpower 4" with a 486DX-33 processor, 8MB of
memory, a localbus ide controller (the manual refers to an "on board super i/o
controller (smc 665)" but I'm not sure whether that is the ide controller or
not), localbus mach-32 video, Conner CP30254 240MB disk, 256kb external cache.
I've tried turning the cache from write-back to write-through and off
completely but I still see the errors.

The problem also happens on a machine with the same motherboard, but with
a 486dx2-66, 16MB of memory, and two Conner CP30344 340MB drives.

> I *know* there are IDE drives with buggy firmware that will cause random
> reordering of data, that manage to get to market because they seem to work
> with DOS just fine

I have heard this is the case too.  Can anyone name names...?

I have seen a news article somewhere that claimed that micronics motherboards
were pretty reasonable, but could well have been a different model, and anyway
was probably refering to it's performance running DOS...

I did try running the test perl script that triggers the problem under OS/2 on
the same hardware without seeing any problems, but I'm not sure that that
really proves anything.

Duncan