On Friday 13 April 2001 17:45, you wrote:
> On Sat, 14 Apr 2001 04:53, you wrote:
> > JoAnne wrote:
> > > The Register: Data-corruption bug hits VIA chipsets(Apr 13, 2001, 15:10
> > > UTC) (1719 reads) (4 talkbacks) (Posted by kreichard)
> > > Dennis E. Powell writes to tell us about the Register's coverage: "The
> > > kernel crowd have been trying for months to track this thing down;
> > > turns out now that it's not their fault!
> >
> > *OUCH*!!! My AOpen mobo has the VIA chipset... (and it's Friday
> > the thirteenth!).
> > Can't afford a new mobo, so does this mean that all the kinky goings on
> > with my system
> > _wasn't_ due to beta code, the phase of the moon, or my cat biting me?
> > <*sheesh*!!>
> >
> > ... so what do we do now? ... switch to Windows
> > Me??? [NOT!!!]
> >
> > Elton Woo ;-)
>
> It only happens with big files between two ide hard drives at maximum
> throughput and blah blah blah read the report
And at install time between a CD and a hdd both running DMA....
And probably not testably reproducible at other times as well...
It is a _HARDWARE_RACE_CONDITION_ and those tend to be as predictable as the
outcome of a rain dance. Just lucky someone found a reproducible set of
circumstances so some confessions could be forthcoming and a fix could be
encouraged, else the poor overworked kernel hackers in all distros would
still be pulling out their hair, or flattening the tops of their heads by
bashing them against the nearest convenient wall.
Civileme
