From Bugzilla Helper: User-Agent: Mozilla/5.0 (Macintosh; U; PPC Mac OS X; en) AppleWebKit/418.8 (KHTML, like Gecko) Safari/419.3 Description of problem: Unexpected kernel panic when I tried looking at contrack data using the command cat /proc/net/ip_conntrack_expect Unfortunately I needed this system up immediately and did not write out the kerenl panic data although it did seem to be related to accessessing this file in /proc. Version-Release number of selected component (if applicable): kernel-smp-2.6.9-34.0.1.EL How reproducible: Didn't try Steps to Reproduce: 1. 2. 3. Actual Results: Expected Results: Additional info: System is a Dell PE 2850 with a single 2.8GHz Xeon with HT enabled. and 1GB RAM
Its not solely related to accessing the proc file, as I'm able to read the file here without reproducing the problem. Its most likely, given that this file is protected by two rwlocks, that this is a race generated by trying to read the file while receiving packets with the filter configured. Can you provide your iptables configuration? Also, would you be able to reproduce the problem while running a custom kernel with CONFIG_NETFILTER_DEBUG enabled?
Created attachment 133333 [details] iptables lokkit config contents of /etc/sysconfig/iptables on affected machine
Unfortunately I cannot reproduce this on the referenced machine as it is a production server.
Thank you for the iptables config. If you can in any way manage to reproduce that oops, it would be helpful to me. I'm going to try to reproduce here, but unless I can, there likely won't be much I can do for this bug.
Note to self: If upstreams fixes this, these are some candidates http://www.kernel.org/git/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=e041c683412d5bf44dc2b109053e3b837b71742d http://www.kernel.org/git/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=a41bc00234a0a2ccaa99a194341ae108ae17ddc8 http://www.kernel.org/git/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=4acdbdbe5089c06d5e0c7e96783fcc4414ded00a
Mark- I'm not sure if you are able, but if you have a test machine that you can reproduce this on, could you please test against the RHEL5 beta release? That would help me narrow down if the above patches apply to this bug you are seeing, as I am still unable to reproduce.