Date: Fri, 19 Jan 2007 11:07:18 +0100 From: Roman Le Houelleur <roman@IPricot.com> To: freebsd-stable@freebsd.org Cc: Kris Kennaway <kris@obsecurity.org> Subject: Re: tcpdump, rl, sis, fxp and multicast problems Message-ID: <45B09856.8080600@IPricot.com> In-Reply-To: <20070118235125.GA80971@xor.obsecurity.org> References: <45AF8586.8080908@IPricot.com> <20070118235125.GA80971@xor.obsecurity.org>
next in thread | previous in thread | raw e-mail | index | archive | help
that's very clever indeed ! Well, I was not quite sure it was actually performing the TCO (not to say I feel stupid ...). not very important but wouldn't it be better to set the checksum to 0 instead of some arbitrary (?) and confusing value then ? thank you anyway, Roman. Kris Kennaway wrote: > On Thu, Jan 18, 2007 at 03:34:46PM +0100, Roman Le Houelleur wrote: >> Hi everyone, >> >> It's been about a year or so that I upgrade my box to RELENG_6 >> from time to time. I got some suprises this week, I've seen a >> few things that made me think of hardware problems but some are >> definitly soft. >> >> First and easy one, tcpdump -v does not show checksum error >> for UDP whereas it does for TCP (same w/ tcpdump -vvv). >> >> checksum : I have a double port fxp 82550 Pro/100 board which >> makes plenty of checksum errors (UDP & TCP), I suspect this >> is related to checksum offloading but am not sure. I strongly >> believe this card was working fine a few days/weeks ago. > > And a priori it still is: by definition, checksum offload means that > the OS does not compute the checksum for outgoing packets, so tcpdump > doesn't see a valid checksum either. Unless you have evidence (from > e.g. observation on another host) that the checksums are not being > computed correctly, this is not a bug. It is, however, a FAQ ;-) > > Kris
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?45B09856.8080600>