Re: r8169 regression: UDP packets dropped intermittantly

[Jonathan Woithe <jwoithe@atrad.com.au>]

On Wed, Nov 18, 2015 at 12:21:08AM +0100, Francois Romieu wrote:
> Jonathan Woithe <jwoithe@atrad.com.au> :
> [...]
> > It would be advantageous if we could upgrade this Linux system to a kernel
> > more recent than 2.6.35.11, but that will require a resolution to this
> > problem.  Since 2.6.35.11 works while current kernels do not, the only other
> > option is to stick with 2.6.35.11.  Is there anything we can do to try to
> > track down the problem?  I'm willing and able to run further tests on the
> > system as required.
> 
> If you were tracking the hardware stats (ethtool -S), would you be able to
> tell if the hardware sees the relevant UDP packets long before the driver
> does ?

Possibly.  The packet flow at the time of fault is quite low so it might be
doable.  What would be the best way to approach this?  Query "ethtool -S"
regularly around the point of failure?

> Would you be able to ascertain that the interrupt mask register (ethtool -d)
> is not set for an excessive amount of time ?

I could try.  What's the most reliable way to determine this?  Use regular
ethtool queries about the time of the failures?

Regards
  jonathan