> > I wouldn't speculate about the exact root cause; nobody has yet
> > turned up any kind of smoking gun. It might be that Cyrix has
> > docs about software fixes/workarounds for these incompatibilities.
>
> I have all the cyrix app notes, and they are pretty boring. I've not seen
> anything that would explain it at all.
And FYI one person tried adding lots of extra readl()s to force
the writes to be posted (my selection of "relevant" spots) and
eventually reported no improvement. (I don't recall that being
with a webcam, and I'm assuming there's only one root problem
here -- which is perhaps a dubious assumption! :)
Given that some MSFT software allegedly behaves, it's likely
there are some kind of software tweaks that can be applied, but
again I'd not speculate as to whether USB, PCI, or some other
platform issue is what's showing up here. Someone with the
right expertise AND hardware will likely need to debug this;
the only pattern that seems evident is "Cyrix".
> We certainly have PCI posting bugs but those ought to be PCI bridge dependant
> not CPU dependant. Which Cyrix cpus have people reported problems with ?
So far, this "486" (new problem report?) and the MediaGX;
I seem to recall some 686 as mentioned too. Maybe some of
those are duplicates -- I don't know the Cyrix lines at all.
Another question: are there any Cyrix-based platforms are
known to actually _work_ with USB on Linux? All that I recall
are problem reports.
- Dave
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
http://lists.sourceforge.net/lists/listinfo/linux-usb-devel
