[Bug 68] Kernel panic

bugzilla-daemon@netfilter.org bugzilla-daemon@netfilter.org
Thu, 20 Mar 2003 10:55:06 +0100


https://bugzilla.netfilter.org/cgi-bin/bugzilla/show_bug.cgi?id=68





------- Additional Comments From laforge@netfilter.org  2003-03-20 10:55 -------
This looks strange.

The BUG in slab.c tells us that there is a GFP_ATOMIC missing.  This means that
we are allocating kernel memory from softirq context with only GFP_KERNEL.

If I understand your backtrace correctly, what happens is:

- you are reloading a ruleset via setsockopt() from userspace
- the kernel then suddenly receives a packet and processes it (because the stack
shows ip_conntrack code)
- after resuming the sockopt, it tries to allocate the new memory and fails.

but allocation of the memory happens in syscall context... why would it need
GFP_ATOMIC?

*sigh*.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.