[HiPPIE-Users] Hippie 0.9.0 causes kernel fault

[Andrew H. <and...@gm...>]

When Hippie 0.9.0 is applied to stock 2.6.23.1, I get the following errors
which seem to correspond to an packet arrival event:

 

BUG: scheduling while atomic: ps/0x10000100/1162

 [<c05a98a2>] __sched_text_start+0x5a2/0x6a4

 [<c04eab96>] tcp_packet+0x4ae/0xbc8

 [<c01243cf>] lock_timer_base+0x19/0x35

 [<c01244e0>] __mod_timer+0x92/0xa1

 [<c0118ee0>] __cond_resched+0x16/0x34

 [<c05a9a0f>] cond_resched+0x2a/0x31

 [<c0154cf7>] kmem_cache_alloc+0x49/0x7e

 [<c04fa237>] pie_packet_mockup+0x19/0x102

 [<c04fa3da>] pie_packet_read+0xc/0x2e8

 [<c04fa6c7>] pie_packet_read_skb+0x11/0x1b

 [<c04c014f>] netif_receive_skb+0x316/0x39e

 [<c04bbb52>] __alloc_skb+0x47/0xf6

 [<c02cc6f4>] e1000_clean_rx_irq+0x15e/0x4d7

 [<c02cc596>] e1000_clean_rx_irq+0x0/0x4d7

 [<c02cb553>] e1000_clean+0x2a6/0x5fd

 [<c011635c>] update_curr+0x121/0x12d

 [<c04c2204>] net_rx_action+0x66/0xfe

 [<c0120b62>] __do_softirq+0x72/0xdf

 [<c0120c06>] do_softirq+0x37/0x39

 [<c0120e04>] irq_exit+0x6c/0x6e

 [<c0104e77>] do_IRQ+0x40/0x79

 [<c01031ef>] common_interrupt+0x23/0x28

 [<c05ab054>] _spin_unlock_irqrestore+0x5/0x6

 [<c028a162>] n_tty_receive_buf+0x2d3/0xeb6

 [<c015ffef>] link_path_walk+0x65/0xbc

 [<c0245979>] _atomic_dec_and_lock+0x2d/0x54

 [<c0156585>] get_unused_fd_flags+0x52/0xc7

 [<c016023e>] do_path_lookup+0x73/0x1a1

 [<c0158ba8>] file_move+0x14/0x34

 [<c028c129>] pty_write+0x39/0x41

 [<c0289d05>] write_chan+0x18d/0x317

 [<c0117b1d>] default_wake_function+0x0/0xc

 [<c02876d8>] tty_write+0x123/0x1b3

 [<c0289b78>] write_chan+0x0/0x317

 [<c01672fc>] dput+0x10/0xf3

 [<c0158246>] vfs_write+0x8b/0x118

 [<c016aad4>] mntput_no_expire+0x13/0x5a

 [<c02875b5>] tty_write+0x0/0x1b3

 [<c01587fb>] sys_write+0x41/0x6a

 [<c01027de>] sysenter_past_esp+0x5f/0x85

 [<c05a0000>] dccp_init_sock+0x172/0x179

 

If I shut down HIPPIE from the interface, the errors go away. I have hippie
DEBUG enabled, but these don't look like debug errors.

 

Thanks,