My ISP is currently trialing IPv6 over PPPoE so I have been trying it out, but if I load pppd with the "ipv6 ," option via the bootup scripts I get a kernel softlockup a few seconds after the IPv6 default route comes up. Even if I bootup with the wan network cable unplugged and then plug it in after the system has finished booting I will often get a lockup. If I login and load pppd from the bash shell I don't get a lockup no matter how many times I disconnect and reconnect the connection. I am running Gentoo Linux x86_64 on an Athlon 64 X2 Dual Core CPU, currently running kernel version: 2.6.36-gentoo-r1, but I also had the problem on the 2.6.35 series as well. I also was running Gentoo ppp-2.4.4-r13 but recent updated to gentoo ppp-2.4.4-r25 Sometimes the backtrace changes a little but the following is what it normally looks like: [<ffffffff810020ab>] ? system_call_fastpath+0x16/0x1b Kernel - panic - not syncing: softlockup: hung tasks Pid: 7932, comm: pppd Not tainted 2.6.36-gentoor1 #1 Call Trace: <IRQ> [<ffffffff814ba518>] ? panic+0xac/0x1b0 [<ffffffff81004e10>] ? dump_trace+0x270/0x27f [<ffffffff81002a0e>] ? apic_timer_interrupt+0xe/0x20 [<ffffffff81076683>] ? watchdog_timer_fn+0x17c/0x1b3 [<ffffffff81076507>] ? watchdog_timer_fn+0x0/0x1b3 [<ffffffff81053c70>] ? __run_hrtimer+0x50/0xa9 [<ffffffff81053ec8>] ? hrtimer_interrupt+0xaa/0x192 [<ffffffff81019371>] ? smp_apic_timer_interrupt+0x81/0x94 [<ffffffff81002a13>] ? apic_timer_interrupt+0x13/0x20 <EOI> [<ffffffff8142c096>] ? dev_deactivate_queue+0x0/0x63 [<ffffffff814bc4ba>] ? _raw_spin_lock_bh+0x28/0x2e [<ffffffff814bc49b>] ? _raw_spin_lock_bh+0x9/0x2e [<ffffffff8142c0be>] ? dev_deactivate_queue+0x20/0x63 [<ffffffff8142c12d>] ? dev_deactivate_queue+0x34/0xdb [<ffffffff8141a00e>] ? __dev_close+0x54/0x70 [<ffffffff8141803f>] ? __dev_change_flags+0xaf/0x129 [<ffffffff81419ed1>] ? dev_change_flags+0x12/0x42 [<ffffffff8146070b>] ? devinet_ioctl+0x273/0x528 [<ffffffff81409324>] ? sock_do_ioctl+0x1b/0x36 [<ffffffff8140954f>] ? sock_ioctl+0x210/0x21a [<ffffffff810c1e0a>] ? do_vfs_ioctl+0x40b/0x44c [<ffffffff81023632>] ? do_page_fault+0x210/0x259 [<ffffffff810c1e9c>] ? sys_ioctl+0x51/0x70 [<ffffffff810020ab>] ? [<ffffffff814ba518>] ? system_call_fastpath+0x16/0x1b
It looks like I am no longer getting the softlockup problem with kernel version: 2.6.37