Re: Problems with 3c59x

Donald Becker <[email protected]>
Newsgroups gmane.linux.drivers.vortex.devel
Message-ID <[email protected]>
On Fri, 13 Dec 2002, Bogdan Costescu wrote:

> On Fri, 13 Dec 2002, Donald Becker wrote:
> 
> > > Dec 12 17:29:24 lapd eth0: Media selection timer tick happened, Autonegotiate full duplex.
> > > Dec 12 17:29:28 lapd eth1: Media selection timer tick happened, Autonegotiate full duplex.
> 
> Why this short time between timer ticks ? The unabridged logs show even 

I think that you missed "eth0" vs "eth1".

> > Ahhh, buglet.  This can never happen in normal use, but it can when the
> > Tx queue was full before being shut down.
> 
> Andrew fixed this in the kernel driver about a year or so ago :-)
> I think that it was even mentioned on this list, but I might be wrong...

I believe that this is a different bug, although that fix might have
addressed both.  Previously there was a race condition during
operation.  Here the Tx queue is full (tx_full = 1) when the
interface is shut down, and the variable isn't cleared.  This can almost
never occur when interrupts are working.

> > Bogdan, do you know where there is a listing of which APIC bugs were
> > fixed when?
> 
> I don't know of such a list to exist; the fixes seem to come from 
> different persons (from the Changelogs) so I don't think that such a list 
> could exist...

I see that there is a fresh set of similar APIC bug reports this week on
LKML.  My guess is that there is a brand-spanking-new bug here.

> And from Cosmin's message:
> > 2 CPUs x 1GHz Intel PIII
> 
> I was just about to ask about the machine, because it was never mentioned 
> if it's a SMP one (which I believed to be after seeing the "Tx ring full" 
> messages)

That was a key for the old race condition, but SMP machines are also the
ones that have APIC misconfiguration.

> This sometimes helped to aleviate interrupt related problems - generally 
> speaking, nothing specific to this board. If the BIOS has the option of 
> MPS level, could you try changing from 1.4 to 1.1 or vice-versa ?

-- 
Donald Becker				[email protected]
Scyld Computing Corporation		http://www.scyld.com
410 Severn Ave. Suite 210		Scyld Beowulf cluster system
Annapolis MD 21403			410-990-9993

_______________________________________________
vortex mailing list
[email protected]
http://www.scyld.com/mailman/listinfo/vortex
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.