[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <6B63AF9EF0873F48A86A7F284D05BEE4070FD89C@NT-IRVA-0752.brcm.ad.broadcom.com>
Date: Tue, 27 May 2008 11:46:06 -0700
From: "Eilon Greenstein" <eilong@...adcom.com>
To: "Ben Hutchings" <bhutchings@...arflare.com>
cc: netdev <netdev@...r.kernel.org>, jeff@...zik.org,
davem@...emloft.net, "Vladislav Zolotarov" <vladz@...adcom.com>,
"Eliezer Tamir" <eliezert@...adcom.com>
Subject: RE: [PATCH net-next 6/12]bnx2x: Add TPA, Broadcoms HW LRO
Ben Hutchings wrote:
> > Signed-off-by: Vladislav Zolotarov <vladz@...adcom.com>
> > Signed-off-by: Eliezer Tamir <eliezert@...adcom.com>
>
> What's with the byte order marks here?
Thanks for pointing this out. It was not showing on my email client. If needed, I will re-submit the patches - I see that 1-3 did not make it yet as is, so I will wait to see if resubmitting is required
>
> [...]
>
> It's important that people are able to see how to disable LRO. Using a
> different name for it, and no module parameter description, really doesn't
> help.
The TPA (Transparent Packet Aggregation) is similar to LRO but it is not using the Kernel LRO but rather a HW/FW implementation - so using LRO can be somewhat confusing. On top of that, TPA is how marketing wants to name it...
>
> You should also implement ethtool get_flags and set_flags methods so that
> LRO can be turned on and off later (at least once the ethtool command gets
> patched, hint hint).
Excellent point (since we are planning to do so - it is an excellent point). The plan is to submit a new patch with ethtool support after this one will be accepted - we already had too many changes pending
>
> Ben.
>
Thanks for your comments!
Eilon
Powered by blists - more mailing lists