[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-Id: <1268768683.2091.39.camel@achroite.uk.solarflarecom.com>
Date: Tue, 16 Mar 2010 19:44:43 +0000
From: Ben Hutchings <bhutchings@...arflare.com>
To: Blaschka <frank.blaschka@...ibm.com>
Cc: netdev@...r.kernel.org
Subject: Re: question regarding hw tx csum for IPv6
On Tue, 2010-03-16 at 19:37 +0000, Ben Hutchings wrote:
> On Tue, 2010-03-16 at 10:11 +0100, Blaschka wrote:
> > Hi,
> >
> > for me it looks like many network drivers exclude IPv6 frames from
> > tx hw csum in case they have extended IPv6 headers.
> > Does the stack generates the csum for this kind of frames anyhow
> > the device claims to do NETIF_F_HW_CSUM?
>
> The NETIF_F_*_CSUM flags are for checksum generation, not checksum
> validation.
[...]
Sorry, I misread your question.
NETIF_F_IPV6_CSUM means the device can generate checksums for all
TCP/IPv6 and UDP/IPv6 packets.
NETIF_F_HW_CSUM is different; it means the hardware can generate a
16-bit checksum of the whole packet starting with a given seed value and
place it at a specified offset before transmitting the packet. The
hardware does not have to parse specific protocols in this case.
If your hardware parses specific protocols then the driver could use
skb_linearize() and the checksum helper functions in the cases that the
hardware doesn't handle. Then you would still have the benefit of
checksum generation offload in the normal case.
Ben.
--
Ben Hutchings, Senior Software Engineer, Solarflare Communications
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists