[eepro100] Sleep Bit mode and EEPROM fix?

[Kallol Biswas]

Hi,
    The scb status returning all f and the card "eating" management packets
are two different problems. If you believe that the card eats packets then
apply the BMC upgrade from Intel's page, it has the work around for the 
TCO controller. 

It is doubtful that you are getting the problem due to sleepmode,
we have the problem in our Itanium box, as WXB PCI agent returns
-1 since the card fails to respond to read request sometime 
under high stess when sleep mode bit is turned on. Your system
might have similar problem.


Kallol
> 
> > -----Original Message-----
> > I put v2.05 of the eepro100-diag.c program at
> >      ftp://www.scyld.com/pub/diag/eepro100-diag.c
> > This version should
> >   report the EEPROM sleep mode bit
> >   update the EEPROM to remove the sleep bit with the -Gww flag.
> 
> 	I haven't been able to figure out from your source exactly how to use
> the -G flag, and it isn't documented on the webpage.  I was wondering if you
> could tell me how it works and/or update the webpage with that information
> if you have time.
> 
> > Now for another problem: I believe that several recent problems have
> > been a result of the 559 swallowing what it thinks are management
> > packets, but that we wanted to e.g. complete an MPI message.  Is there
> > any way to turn off the management "feature"?
> 
> 	As soon as I have "fixed" the sleep mode problem with our network cards I
> will run my tests again, although I am skeptical about this fix since the
> bug is dependant upon load and bonding two interfaces together would if
> anything decrease load.  Perhaps it is when two interfaces are bonded
> together since each single NIC interface only sees half of the packets this
> confuses the 559 into thinking some packets are management packets in which
> case it "eats" them.
> 
> 	I have run tests bonding only eth0 to bond0 and bonding only eth1 to bond0,
> and all of those tests run fine.  This eliminates the possibility of it
> being a single network card that is faulty and the bonding module somehow
> messing up.
> 
> 
> 
> _______________________________________________
> eepro100 mailing list
> eepro100@scyld.com
> http://www.scyld.com/mailman/listinfo/eepro100
> 


--
Phone: 973-443-7469         |  Design Engineer
Telnet: 1-443-7469          |  EIAL UNIX OS LAB
www.kallolbiswas.com        |  Hewlett Packard Company
kallol_biswas@hp.com        |  Florham Park, NJ