Re: Long round trip times on local net

[Julian Coleman <jdc%coris.org.uk@localhost>]

> So has anyone with similar Sun hardware noticed these problems? Is it a
> hardware problem with the quad PCI card or a buggy driver?

Does it happen on all 4 ports on the QFE?  I seem to remember we had an
interrupt problem on the last port at one point.  We should be able to
check interrupts by adding interrupt counting code to hme.c in the same
way as gem.c (look for evcnt_attach_dynamic() there).

> vendor 0x108e product 0x1000 (miscellaneous bridge, revision 0x01) at pci3 
> dev 0
>  function 0 not configured
 
> Not sure what the bridge at function 0 is about.

It's an Ebus bridge on my QFE.  Not sure why it's not identified on yours.
It's where the local MAC address is stored.  See if_hme_pci.c, inside
`#if HME_USE_LOCAL_MAC_ADDRESS'.

Thanks,

J

-- 
  My other computer also runs NetBSD    /        Sailing at Newbiggin
        http://www.netbsd.org/        /   http://www.newbigginsailingclub.org/