Re: [Etherboot-users] Sporadic eepro(10) RX problems after reboot

Timothy Legge wrote:

> Till Straumann wrote:
>
>> Hi.
>>
>> This is about eepro [10 / 82595],  NOT eepro100.
>>
>> Sporadically, after a reboot [Ctl-Alt-Del] etherboot would
>> not seem to receive any packets; TX worked OK (i.e., dhcp
>> reqs were seen on the wire).
>>
>> Powering down for a few minutes remedied the problem.
>>
>> It turns out that the RX buffer RAM was not initialized
>> and if eepro_poll found a leftover packet header there
>> it would get out of sync with the hardware.
>>
>> Patch with detailed description attached.
>>
>> NOTE: all versions of etherboot are affected
>>
>> Regards
>>  -- Till Straumann
>>
>> PS: I'm not a subscriber. CC me if there are any questions
>
> Thanks for the patch it has been committed to the cvs.  Can you test 
> it again once yothe public cvs is updated.  The eepro is one card that 
> I do not own. 

Hmm - etherboot 5.3.x didn't work on my board, 5.0.10 back from the 
simple days did
(something during the early relocation failed - I didn't really have the 
time to debug it.
I then tried the 5.0 version which I was more familiar with and since 
that one works
fine, that's what we use). The bug and fix should be the same, however. 
eepro didn't
change really since the old days.

--> I can test with 5.0.10 but not with 5.4 but that would mean I'd have 
to backport
       the patch to where I started from - dunno if it makes a lot of sense

-- Till

BTW: one change I noted was the introduction of 'virt_to_bus' and 
'bus_to_virt'
           conversions which don't really seem to make sense. The memory 
that
           is addresses by the pointers in question doesn't live in the 
CPU's address space
           or on a bus accessible to the CPU but exists only 'inside' 
the eepro controller
           and is accessed by the CPU by setting a 'address register' on 
the eepro and
           subsequently reading/writing from/to a 'data port register'.
           I don't think the conversions do any harm since they should 
cancel but
           conceptually it seems wrong.

>
>
> Tim
>





-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click