From: David L. <ma...@he...> - 2006-05-04 11:31:30
|
> There is a kernel option that can sometimes help tracking this stuff > down. It's called CONFIG_DEBUG_USER (which appears to be on by default > on the gumstix). > It would be interesting to add the following option to the kernel > command line (in u-boot) > > user_debug=0x1f > > Then you'll get a short register dump when a fault (including a SEGV) > occurs. This will include the PC which can be used to figure our > exactly where the volation is occuring. > Hi again! I got the following dump using your method (which will come in handy in the future ;-): iptables: unhandled page fault (11) at 0x0001f008, code 0x817 pgd = c3a38000 [0001f008] *pgd=a3a30031, *pte=00000000, *ppte=00000000 Pid: 442, comm: iptables CPU: 0 PC is at 0x4000ebf4 LR is at 0x4000ebe4 pc : [<4000ebf4>] lr : [<4000ebe4>] Not tainted sp : be9f5b04 ip : 4007a5a8 fp : 00000044 r10: 40017fa0 r9 : 00000048 r8 : 00000001 r7 : 00000000 r6 : 0001c570 r5 : 0001c448 r4 : 0001c1b0 r3 : 0001eff4 r2 : 00000020 r1 : 0001c548 r0 : 00000000 Flags: nzCv IRQs on FIQs on Mode USER_32 Segment user Control: 397F Table: A3A38000 DAC: 00000015 [<c0018820>] (show_regs+0x0/0x4c) from [<c001d34c>] (__do_user_fault+0x5c/0xa4) r4 = C3812D40 [<c001d2f0>] (__do_user_fault+0x0/0xa4) from [<c001d5e4>] (do_page_fault+0x1e4/0x214) r7 = C000D960 r6 = C39A26A4 r5 = C3812D40 r4 = FFFFFFEC [<c001d400>] (do_page_fault+0x0/0x214) from [<c001d750>] (do_DataAbort+0x3c/0xa0) [<c001d714>] (do_DataAbort+0x0/0xa0) from [<c0016ce8>] (ret_from_exception+0x0/0x10) r8 = 00000001 r7 = 00000000 r6 = 0001C570 r5 = 0001C448 r4 = FFFFFFFF Am I the only one having problems with iptables? Im not using any special software on the gumstix. I've figured out how to hook directly into netfilter from a kernel module, so I can avoid using iptables. But I (or others) might need it at a different time. Thanks a lot for helping me out so far! I'll most likely be back ;) David |