定位Oops错误代码行【整理】

[17184178.672000] Bad mode in data abort handler detected
[17184178.672000] Internal error: Oops - bad mode: 0 [#1] PREEMPT
[17184178.672000] CPU: 0    Not tainted  (2.6.26.5 #1255)
[17184178.672000] PC is at 0xd201393a
[17184178.672000] LR is at 0xd20132fd
[17184178.672000] pc : [<d201393a>]    lr : []    psr: 200000bb
[17184178.672000] sp : d202df68  ip : 00000000  fp : 4021fc34
[17184178.672000] r10: 4003e000  r9 : 00000000  r8 : e0096c0c
[17184178.672000] r7 : e00aa340  r6 : e0096c04  r5 : e009e7c0  r4 : 000000b0
[17184178.672000] r3 : e001a418  r2 : 00000000  r1 : e00a11e8  r0 : 00accbad
[17184178.672000] Flags: nzCv  IRQs off  FIQs on  Mode UND_32  ISA Thumb  Segment user
[17184178.672000] Control: 0005317d  Table: 21aa4000  DAC: 00000015
[17184178.672000] Stack: (0xd202df68 to 0xd002e800)
[17184178.672000] Code: 6868 0029 2800 d01a (6802)
[17184178.676000] ---[ end trace 31c4d86500000008 ]---

 

[17179681.444000] Internal error: Oops - bad syscall: ddf04c [#1] PREEMPT

[17179681.444000] Modules linked in: coma_dsr coma_voice coma_ss7 coma_cpi coma_config

[17179681.444000] CPU: 0    Not tainted  (2.6.26.5 #659)

[17179681.444000] PC is at __dabt_usr+0x4/0x60

[17179681.444000] LR is at 0x377a4

[17179681.444000] pc : [<b4023884>]    lr : [<000377a4>]    psr: 80000093

[17179681.444000] sp : 4021fc10  ip : 0000f1b4  fp : 4021fc34

[17179681.444000] r10: 4003e000  r9 : 00000000  r8 : 003d0f00

[17179681.444000] r7 : 00000152  r6 : 400286f8  r5 : 402202b0  r4 : 00000000

[17179681.444000] r3 : 000000e4  r2 : 000000e4  r1 : 000f229c  r0 : b428bcec

[17179681.444000] Flags: Nzcv  IRQs off  FIQs on  Mode SVC_32  ISA ARM  Segment user

[17179681.444000] Control: 0005317d  Table: 21990000  DAC: 00000015

[17179681.444000] Process my_app(pid: 81, stack limit = 0xb5b8a268)

[17179681.444000] Stack: (0x4021fc10 to 0xb5b8c000)

[17179681.444000] fc00:                                     00000001 000000e4 000f229c 00000004

[17179681.444000] fc20: 402202f8 ffffffff 4021fc54 4021fc38 00035778 00037784 4003e000 00000000

[17179681.444000] fc40: 4021fc60 ffffffff afa66c90 4021fc58 40028788 00035740 00000000 402202b0

[17179681.444000] fc60: 402202f8 402202b0 400286f8 00000152 003d0f00 00000000 4003e000 afa66c90

[17179681.444000] fc80: 4021fc58 40028758 00000000 00000000 00000000 00000000 00000000 00000000

[17179681.444000] fca0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000

[17179681.444000] fcc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000

 

上面两种Oops：第一种是UND_32表示是发生了用户空间，第二种SVC_32表示发生在内核空间。pc表示出问题的地址。

 

几个常识概念：

1.程序计数器PC （R15），可以作为一般的通用寄存器使用，但有一些指令在使用R15时有一些限制。由于ARM采用了流水线处理器机制，当正确读取了PC的值时，该值为当前指令地址值加上8个字节。也就是说，对于ARM指令集来说，PC指向当前指令的下两条指令的地址。由于ARM指令是字对齐的，PC值的第0位和第一位总为 0。

2.寄存器R13(SP)，通常用作堆栈指针，每一种模式都有自己的物理R13，程序初始化R13。当进入该模式

时，可以将要使用的寄存器保存在R13所指的栈中，当退出时，将弹出，从而实现了现场保护。

3.寄存器R14被称为链接寄存器(LR)，当中存放每种模式下，当前子程序的返回地址或者

发生异常中断的时候，将R14设置成异常模式将要返回的地址。

4.寄存器R12

 

在给linux内核添加netfilter和iptables配置后，生成内核。下载到开发板的后启动插入网线出现错误提示：

eth1: link up, 100Mbps, full-duplex, lpa 0xCDE1        //插入网线后提示

------------------首先这些打印是kernel的panic函数列出的，具体意义可以直接找到kernel代码去看，很有帮助。
Unable to handle kernel paging request at virtual address 06400040

-----------------------空指针错误，这个一般就是非法地址访问，至于为什么导致非法，请关注PC周围的代码逻辑。有必要的话就printk出来。

pgd = c0004000
[06400040] *pgd=00000000
Internal error: Oops: 5 [#1]  //错误提示 ------5代表什么？要在你的手册或kernel代码中查，null pointer？？

Modules linked in: zd1211rw rt73usb rt2x00usb rt2x00lib asix usbnet mac80211 inp
ut_polldev
CPU: 0    Not tainted  (2.6.30.4-LanxumDomas #19)

PC is at skb_release_data+0x74/0xc4

--------当前pc指针，这个十分有用！可以将kernel反汇编，然后找dequeue_task的相对位置0xc处

LR is at __kfree_skb+0x1c/0xd0

-----------------以下为当前寄存器值，也有帮助
pc : []    lr : []    psr: 20000013
sp : c040dd90  ip : c040dda8  fp : c040dda4
r10: 00000001  r9 : c0465fc8  r8 : c3a22000
r7 : c3b54300  r6 : c3b50004  r5 : c3ae1600  r4 : c3ae1600
r3 : 00000000  r2 : c3b50822  r1 : 00000000  r0 : 06400040
Flags: nzCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment kernel
Control: c000717f  Table: 33af0000  DAC: 00000017
Process swapper (pid: 0, stack limit = 0xc040c268)
Stack: (0xc040dd90 to 0xc040e000)
dd80:                                     c3ae1600 c3ae1600 c040ddbc c040dda8
dda0: c028266c c0282b04 c3ae19c0 c3ae1600 c040ddcc c040ddc0 c0282794 c0282660 

-----------------以下是调用堆栈，可以看到程序的流程，便于跟踪。

Backtrace:
[] (skb_release_data+0x0/0xc4) from [] (__kfree_skb+0x1c/0xd
0)
 r5:c3ae1600 r4:c3ae1600
[] (__kfree_skb+0x0/0xd0) from [] (kfree_skb+0x24/0x50)
 r5:c3ae1600 r4:c3ae19c0
[] (kfree_skb+0x0/0x50) from [] (br_handle_frame_finish+0x3c
/0x18c)
[] (br_handle_frame_finish+0x0/0x18c) from [] (br_handle_fra
me+0x238/0x26c)
 r8:c3a22000 r7:c3ae19c0 r6:c3ae1600 r5:c3b50004 r4:80000000
[] (br_handle_frame+0x0/0x26c) from [] (netif_receive_skb+0x
1e4/0x380)
 r7:00000000 r6:c0465fc8 r5:c3ae1600 r4:c3ae19c0
[] (netif_receive_skb+0x0/0x380) from [] (process_backlog+0x
7c/0xd4)
[] (process_backlog+0x0/0xd4) from [] (net_rx_action+0xd0/0x
194)
[] (net_rx_action+0x0/0x194) from [] (__do_softirq+0x78/0x10
0)
[] (__do_softirq+0x0/0x100) from [] (irq_exit+0x48/0x50)
[] (irq_exit+0x0/0x50) from [] (_text+0x48/0x70)
[] (_text+0x0/0x70) from [] (__irq_svc+0x24/0xa0)
Exception stack(0xc040df48 to 0xc040df90)
df40:                   f4100000 00000032 f4100000 60000013 c0084ed8 c040c000
df60: c0084ed8 c04398c4 3001ea00 41129200 3001e9cc c040df9c c040dfa0 c040df90
df80: c008550c c0084f38 60000013 ffffffff
 r7:c04398c4 r6:04000000 r5:f4000000 r4:ffffffff
[] (default_idle+0x0/0xac) from [] (cpu_idle+0x4c/0x68)
[] (cpu_idle+0x0/0x68) from [] (rest_init+0x5c/0x70)
 r7:c04104f0 r6:c0020d34 r5:c0439880 r4:c045d164
[] (rest_init+0x0/0x70) from [] (start_kernel+0x1e0/0x24c)
[] (start_kernel+0x0/0x24c) from [<30008034>] (0x30008034)
 r5:c0439968 r4:c0007175
Code: e3500000 0a000006 e3a03000 e5823018 (e5904000)
Kernel panic - not syncing: Fatal exception in interrupt
Backtrace:
[] (dump_backtrace+0x0/0x10c) from [] (dump_stack+0x18/0x1c)

 r7:c0282b6c r6:c0439ec0 r5:c0282b68 r4:c0282b68
[] (dump_stack+0x0/0x1c) from [] (panic+0x4c/0x12c)
[] (panic+0x0/0x12c) from [] (die+0x1e0/0x214)
 r3:00000100 r2:00000080 r1:c0439ec0 r0:c03b5a34
[] (die+0x0/0x214) from [] (__do_kernel_fault+0x6c/0x7c)
[] (__do_kernel_fault+0x0/0x7c) from [] (do_page_fault+0x14c
/0x25c)
 r7:c040f890 r6:06400040 r5:c040dd7c r4:00000000
[] (do_page_fault+0x0/0x25c) from [] (do_translation_fault+0
x78/0x80)
[] (do_translation_fault+0x0/0x80) from [] (do_DataAbort+0x3
8/0x9c)
 r7:c041075c r6:00000005 r5:c040dd7c r4:c041070c
[] (do_DataAbort+0x0/0x9c) from [] (__dabt_svc+0x40/0x60)
Exception stack(0xc040dd48 to 0xc040dd90)
dd40:                   06400040 00000000 c3b50822 00000000 c3ae1600 c3ae1600
dd60: c3b50004 c3b54300 c3a22000 c0465fc8 00000001 c040dda4 c040dda8 c040dd90
dd80: c028266c c0282b68 20000013 ffffffff
[] (skb_release_data+0x0/0xc4) from [] (__kfree_skb+0x1c/0xd
0)
 r5:c3ae1600 r4:c3ae1600
[] (__kfree_skb+0x0/0xd0) from [] (kfree_skb+0x24/0x50)
 r5:c3ae1600 r4:c3ae19c0
[] (kfree_skb+0x0/0x50) from [] (br_handle_frame_finish+0x3c
/0x18c)
[] (br_handle_frame_finish+0x0/0x18c) from [] (br_handle_fra
me+0x238/0x26c)
 r8:c3a22000 r7:c3ae19c0 r6:c3ae1600 r5:c3b50004 r4:80000000
[] (br_handle_frame+0x0/0x26c) from [] (netif_receive_skb+0x
1e4/0x380)
 r7:00000000 r6:c0465fc8 r5:c3ae1600 r4:c3ae19c0
[] (netif_receive_skb+0x0/0x380) from [] (process_backlog+0x
7c/0xd4)
[] (process_backlog+0x0/0xd4) from [] (net_rx_action+0xd0/0x
194)
[] (net_rx_action+0x0/0x194) from [] (__do_softirq+0x78/0x10
0)
[] (__do_softirq+0x0/0x100) from [] (irq_exit+0x48/0x50)
[] (irq_exit+0x0/0x50) from [] (_text+0x48/0x70)
[] (_text+0x0/0x70) from [] (__irq_svc+0x24/0xa0)
Exception stack(0xc040df48 to 0xc040df90)
df40:                   f4100000 00000032 f4100000 60000013 c0084ed8 c040c000
df60: c0084ed8 c04398c4 3001ea00 41129200 3001e9cc c040df9c c040dfa0 c040df90
df80: c008550c c0084f38 60000013 ffffffff
 r7:c04398c4 r6:04000000 r5:f4000000 r4:ffffffff
[] (default_idle+0x0/0xac) from [] (cpu_idle+0x4c/0x68)
[] (cpu_idle+0x0/0x68) from [] (rest_init+0x5c/0x70)
 r7:c04104f0 r6:c0020d34 r5:c0439880 r4:c045d164
[] (rest_init+0x0/0x70) from [] (start_kernel+0x1e0/0x24c)
[] (start_kernel+0x0/0x24c) from [<30008034>] (0x30008034)
 r5:c0439968 r4:c0007175

出现以上错误后，可以根据错误的提示oops；以下介绍根据错误提示进行错误定位。

1.首先在编译生成内核的时候同时生成了一个vmlinux，使用gdb。

   在内核配置时，make menuconfig 要打开complie with debug info选项。

   注意这行： PC is at skb_release_data+0x74/0xc4

   这告诉我们，skb_release_data函数有0xc4这么大，而Oops发生在0x74处。 那么我们先看一下skb_release_data从哪里开始：

      # grep skb_release_data ./System.map

         c0282af4 t skb_release_data

     于是我们知道在系统出现错误时程序指针在 c0282af4+0x74=c0282b68

  2.然后用gdb查看，gdb ./vmlinux (在linux目录下执行)，进入调试模式。

                                (gdb) b *0xc0282b68

  Breakpoint 1 at 0xc0282b68: file net/core/skbuff.c ,line312

  这就是告诉我们在哪个文件，在哪一行。如此知道了错误的位置，具体的原因带解决。

3，反汇编

                (gdb) disassemble 0xc0282b68