All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jeff Garzik <jgarzik@pobox.com>
To: Santiago Leon <santil@us.ibm.com>
Cc: Andrew Morton <akpm@osdl.org>,
	Linux Kernel list <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 2.6] POWER5 Virtual Ethernet Checkum offload
Date: Fri, 27 Aug 2004 15:28:24 -0400	[thread overview]
Message-ID: <412F8B58.4040905@pobox.com> (raw)
In-Reply-To: <412F85B7.1060804@us.ibm.com>

Santiago Leon wrote:
> Andrew,
> 
> The following patch adds checksum offloading for the POWER5 Virtual 
> Ethernet driver.  In the case where the OS in the partitions 
> communicating support this feature (i.e. both partition have this patch 
> applied), no checksum will be created because the link is reliable. 
> However, in the case where one of the OS in a partition does support 
> this feature and the other doesn't (i.e. linux with patch applied 
> talking to AIX), then the hypervisor will generate the checksum.
> 
> Some levels of firmware will not support this feature but the code will 
> figure it out and not enable it.
> 
> Applies against the latest mainline and -mm trees. Please apply.

First, please always CC net driver patches to me and netdev@oss.sgi.com.

Second, this patch is incorrect.  Read the top of include/linux/skbuff.h 
for the various levels of checksum offloading, and how to use them.

Specifically,

a) you should not be using CHECKSUM_HW unless your hardware provides a 
valid csum (ipv4 or ipv6) that you store in skb->csum.

b) you should not be using NETIF_F_HW_CSUM for reasons similar to (a): 
your hardware must be able to csum the packet whether its ipv4 or ipv6 
or whatever, given the information in the skb.

c) use of NETIF_F_xxx_CSUM is pointless without NETIF_F_SG and code in 
the transmit path to handle page-based fragments (a scatter-gather list).

	Jeff






      reply	other threads:[~2004-08-27 19:42 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2004-08-27 19:04 [PATCH 2.6] POWER5 Virtual Ethernet Checkum offload Santiago Leon
2004-08-27 19:28 ` Jeff Garzik [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=412F8B58.4040905@pobox.com \
    --to=jgarzik@pobox.com \
    --cc=akpm@osdl.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=santil@us.ibm.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.