Hi,
I have the following error message when doing a "ifconfig bond0 down" using kernel 2.6.9 :
bad: scheduling while atomic! [<c038d02e>] schedule+0xb3e/0xb43 [<c0168ab3>] alloc_inode+0x1b/0x146 [<c0169661>] get_new_inode_fast+0xdb/0xdd [<c0179ce3>] proc_read_inode+0x17/0x3c [<c016a10d>] wake_up_inode+0x13/0x48 [<c038d0fb>] wait_for_completion+0x85/0xca [<c0118769>] default_wake_function+0x0/0x12 [<c0118769>] default_wake_function+0x0/0x12 [<c038c4ab>] __down_failed_trylock+0x7/0xc [<c012c041>] synchronize_kernel+0x3a/0x45 [<c012bffb>] wakeme_after_rcu+0x0/0xc [<c033a7e3>] dev_remove_pack+0xf/0x17 [<f88307da>] bond_close+0xbe/0xc3 [bonding] [<c033b296>] dev_close+0xa1/0xa3 [<c033c629>] dev_change_flags+0x51/0x120 [<c037371b>] devinet_ioctl+0x245/0x67f [<c03759c0>] inet_ioctl+0x5e/0x9e [<c0332aed>] sock_ioctl+0xee/0x23e [<c03329ff>] sock_ioctl+0x0/0x23e [<c016263c>] sys_ioctl+0x101/0x279 [<c010415f>] syscall_call+0x7/0xb
I don't feel very comfortable to have such message on a production server, and I guess it might be a race condition, as the machine is a dual-Xeon 2,8Ghz with HyperThreading activated (i-e the system sees 4 CPU).
Network interfaces are two integrated intel e1000 on a super micro motherboard. Interfaces are connected to two cisco 3750 switches on 10/100 copper interfaces.
I can do testing and provide whatever information is needed to help debugging the situation.
-- Sincerely, Gildas LE NADAN - : send the line "unsubscribe linux-net" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
