Date: Wed, 28 Jul 1999 15:57:51 -0700 (PDT) From: Matthew Dillon <dillon@apollo.backplane.com> To: Bill Paul <wpaul@skynet.ctr.columbia.edu> Cc: current@FreeBSD.ORG Subject: Re: IRIX 6.5.4 NFS v3 TCP client + FreeBSD server = bewm Message-ID: <199907282257.PAA64091@apollo.backplane.com> References: <199907282131.RAA15255@skynet.ctr.columbia.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
:IRIX 6.5.4 supports NFS v3 and TCP. I tried cd'ing to a directory
:served on a FreeBSD 3.2-RELEASE system which happens to have the build
:tree for the Alteon Tigon firmware (that's where I compiled the last
:firmware image for the Tigon driver). I did a 'du' and after a short
:while, it exploded with the following messages:
:
:mbuf siz=33524
:panic: Bad nfs svc reply
:
:Inspecting a crash dump showed that the mbuf chain was trashed. The
:same thing happens with a 4.0-current snapshot from the 15th: this time
:I just manually mounted /usr from the FreeBSD server under /mnt on
:the SGI and did cd /mnt; du. Pow: died right away.
:
:The FreeBSD 3.2-RELEASE host has a 3Com 3c509 card. The 4.0-CURRENT
:host has a 3Com 3c900-COMBO PCI card. Each uses different drivers and
:networking works fine otherwise, so I'm pretty sure the problem is in
:NFS somewhere and not in the drivers.
:
:This doesn't happen when using UDP. Given that I can reproduce this
:on demand, I should be able to debug it eventually, but hints in the
:right direction would be useful.
:
:-Bill
:-Bill Paul (212) 854-6020 | System Manager, Master of Unix-Fu
Ok, so if I understand this correctly you have a FreeBSD server and
an IRIX client. UDP mounts work, TCP mounts do not. You are using
the AMD automounting software running on the ... client I presume?
It is the server that is panicing.
First of all, if these are production machines stick with UDP so's
you don't tear your hair out. Also double check that the bug still
exists with the absolute latest CURRENT if you can.
Also please run this (on the FreeBSD server running CURRENT). It
will tell me whether NFS is being forced to realign packet data
coming from your ethernet controller. (In the example below, my
NFS server has to realign the data).
# sysctl -a | fgrep nfs
vfs.nfs.realign_test: 1583064
vfs.nfs.realign_count: 1583064
We fixed a serious data corruption bug with NFSv3 over TCP that
could result in panics. This fix was made on May 2nd to current
and MFC'd to stable on May 8th. This fix made it into 3.2.
There are probably still bugs laying around. Sigh. I'll try
running amd and doing a du on a FreeBSD<->FreeBSD NFSv3 mount.
However, I only give myself a 20% chance of reproducing the problem.
The SGI is probably tickling something that FreeBSD doesn't or I
would have caught the problem earlier.
Your bug really sounds like a packet realignment bug, but I was
sure I fixed those! So there may be a new one.
-Matt
Matthew Dillon
<dillon@backplane.com>
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199907282257.PAA64091>