> I applied David's ohci-paranoid patch to 2.4.0-test1-ac11 and trapped
> a bogus address message 48 hours latter. The was no oops, the system
> continues to run and USB is still functioning normally.
OK, so you can be a guinea pig for this next patch then ... :-)
Based on some offline traffic with Miles, I concluded that
at least part of his problem is due to holes in the Cardbus
support. While I'm addressing those issues (or at least a
few of them :-), I thought it'd be good to have an initial
non-experimental patch to at least get rid of that Oops.
So, here's this patch -- some cleanup, removes an Oops for
you (and Miles), and prepares the ground for a patch that
makes Cardbus OHCI controllers handle hotplug correctly.
Works on two of my test machines, but I want confirmation
that I didn't goof something up as I was splitting out this
relatively safe stuff from the larger patch. After all, I
never see that Oops ...
Let me know if this works for you. Re Roman's suggestion,
try that too! I'm not sure whether to hope the register
flakes out occasionally, as you suggested; it'd explain
everything, but that'd be a nasty failure mode.
- Dave
p.s. Re Miles' situation, it appears that his cardbus
card starts returning all-ones for register reads
once it suspends. All-ones for NDP is trouble!
ohci-micro.patch
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
