[prev in list] [next in list] [prev in thread] [next in thread] 

List:       e1000-devel
Subject:    Re: [E1000-devel] 10Gb 1:1 performance
From:       Richard Scobie <richard () sauce ! co ! nz>
Date:       2008-10-28 1:58:28
Message-ID: 490671C4.4010902 () sauce ! co ! nz
[Download RAW message or body]

Tantilov, Emil S wrote:

> Yes - that warning is valid. Routing (and bridging) are incompatible with LRO and \
> usually will result in panic, or horrible performance at best. 
> If you are using more recent kernel - the kernel should supposedly disable LRO \
> automatically in those situations, but I have not tested it. I will follow up on \
> this and we will update the docs if needed. 
> Thanks,
> Emil

Thanks Emil,

Looking at the ethtool stats, there appears to be no lro_flushed or 
lro_coal entries, so I assume the Fedora kernel 2.6.26.6 either has not 
been built with LRO support, or as you mention, it is being disabled.

The only other puzzle I am trying to solve, is why I do not see MSI-X 
interrupts being used.

The card is in an MCH slot on a 5400 chipset and I have pci=msi set as a 
kernel parameter (required to be able to use MSI on Fedora kernels), but 
/proc interrupts on shows PCI-MSI-edge being used.
Regards,

Richard

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
E1000-devel mailing list
E1000-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/e1000-devel


[prev in list] [next in list] [prev in thread] [next in thread]