【转】分析问题之:如何使用crash分析vmcore - 之基础思路case1

2019/10/29 16:15
阅读数 107

如何使用crash分析vmcore - 之基础思路case1

dmesg查看内核日志

  1.  
    [ 2493382.671020] systemd-shutdown[1]: Sending SIGKILL to PID 28975 (docker-containe).
  2.  
    [ 2493382.671078] systemd-shutdown[1]: Sending SIGKILL to PID 29015 (systemd).
  3.  
    [ 2493420.208723] EXT4-fs (nvme0n1p1): sb orphan head is 140906170
  4.  
    [ 2493420.209198] sb_info orphan list:
  5.  
    [ 2493420.209663] inode nvme0n1p1:140906170 at ffff88490edabfb8: mode 100666, nlink 0, next 149423507
  6.  
    [ 2493420.210129] inode nvme0n1p1:149423507 at ffff8801b99391a8: mode 100666, nlink 0, next 17567381
  7.  
    [ 2493420.210583] inode nvme0n1p1:17567381 at ffff8806d4a26998: mode 100744, nlink 0, next 17570510
  8.  
    [ 2493420.211050] inode nvme0n1p1:17570510 at ffff886387f82ef8: mode 100644, nlink 0, next 17570503
  9.  
    [ 2493420.211508] inode nvme0n1p1:17570503 at ffff886a1f15bfb8: mode 100644, nlink 0, next 241700498
  10.  
    [ 2493420.211966] inode nvme0n1p1:241700498 at ffff8877481800e8: mode 100644, nlink 0, next 243138756
  11.  
    [ 2493420.212431] inode nvme0n1p1:243138756 at ffff88761ad10518: mode 100644, nlink 0, next 241565954
  12.  
    [ 2493420.212900] inode nvme0n1p1:241565954 at ffff8870d64bbfb8: mode 100755, nlink 0, next 241566333
  13.  
    [ 2493420.213366] inode nvme0n1p1:241566333 at ffff88721ae74c48: mode 100644, nlink 0, next 241050093
  14.  
    [ 2493420.213833] inode nvme0n1p1:241050093 at ffff887704958948: mode 100755, nlink 0, next 241567324
  15.  
    [ 2493420.214545] ------------[ cut here ]------------
  16.  
    [ 2493420.219336] kernel BUG at fs/ext4/super.c:879! <<<======这里指明BUG的代码位置
  17.  
    [ 2493420.223948] invalid opcode: 0000 [#1] SMP
  18.  
    [ 2493420.228133] Modules linked in: kpatch_D751550(OE) kpatch_D631237(OE) unix_diag(E) af_packet_diag(E) netlink_diag(E) dccp_diag(E) dccp(E) tcp_diag(E) udp_diag(E) inet_diag(E) [last unloaded: aisqos_hotfixes]
  19.  
    [ 2493420.246846] CPU: 58 PID: 1 Comm: systemd-shutdow Tainted: G W OE K 4.9.79-009.ali3000.alios7.x86_64 #1
  20.  
    [ 2493420.257009] Hardware name: Inventec AliServer Thor01-2U /TB800G4-G1 , BIOS A1.20 03/06/2018
  21.  
    [ 2493420.267339] task: ffff887e45918000 task.stack: ffffc90000014000
  22.  
    [ 2493420.273425] RIP: 0010:[<ffffffffa031a8df>] [<ffffffffa031a8df>] ext4_put_super+0x36f/0x3c0 [ext4] <<<=======这里指明BUG的代码位置
  23.  
    [ 2493420.282593] RSP: 0018:ffffc90000017de8 EFLAGS: 00010206
  24.  
    [ 2493420.288079] RAX: ffff88490edabf50 RBX: ffff887e43299000 RCX: 00000001949b336d
  25.  
    [ 2493420.295384] RDX: 0000000000000000 RSI: 0000000000000206 RDI: 0000000000000206
  26.  
    [ 2493420.302682] RBP: ffffc90000017e18 R08: 00000000000081a4 R09: 0000000000000000
  27.  
    [ 2493420.309988] R10: 0000000000000cb8 R11: 0000000000001e92 R12: ffff887e43299278
  28.  
    [ 2493420.317293] R13: ffff887e43298800 R14: ffff887e43299278 R15: ffffffffa034ff88
  29.  
    [ 2493420.324598] FS: 00007f3241ccf840(0000) GS:ffff887e78480000(0000) knlGS:0000000000000000
  30.  
    [ 2493420.332850] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  31.  
    [ 2493420.338767] CR2: 00007f5e1372fbd0 CR3: 00000004daa52000 CR4: 00000000007606f0
  32.  
    [ 2493420.346065] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
  33.  
    [ 2493420.353361] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
  34.  
    [ 2493420.360660] PKRU: 55555554
  35.  
    [ 2493420.363536] Stack:
  36.  
    [ 2493420.365721] 9cbae75a00000000 ffff887e43298800 ffffffffa034a5e0 ffff887e3818c7b8
  37.  
    [ 2493420.373365] 0000000000000000 ffff887e45918bb0 ffffc90000017e38 ffffffff81244aaf
  38.  
    [ 2493420.380991] 0000000000000083 ffff887e357b8680 ffffc90000017e58 ffffffff81244e37
  39.  
    [ 2493420.388617] Call Trace:
  40.  
    [ 2493420.391239] [<ffffffff81244aaf>] generic_shutdown_super+0x6f/0x100
  41.  
    [ 2493420.397676] [<ffffffff81244e37>] kill_block_super+0x27/0x70
  42.  
    [ 2493420.403508] [<ffffffff81244f73>] deactivate_locked_super+0x43/0x70
  43.  
    [ 2493420.409945] [<ffffffff8124547a>] deactivate_super+0x5a/0x60
  44.  
    [ 2493420.415770] [<ffffffff81264b2f>] cleanup_mnt+0x3f/0x90
  45.  
    [ 2493420.421169] [<ffffffff81264bc2>] __cleanup_mnt+0x12/0x20
  46.  
    [ 2493420.426733] [<ffffffff810a7b50>] task_work_run+0x80/0xa0
  47.  
    [ 2493420.432306] [<ffffffff810032ba>] exit_to_usermode_loop+0xaa/0xb0
  48.  
    [ 2493420.438572] [<ffffffff81003baa>] syscall_return_slowpath+0xaa/0xb0
  49.  
    [ 2493420.445011] [<ffffffff8171a783>] entry_SYSCALL_64_fastpath+0xc3/0xc5
  50.  
    [ 2493420.451623] Code: 60 04 00 00 48 8b 80 e0 00 00 <0f> 0b 49 c7 c7 88 ff 34 a0 49 8b
  51.  
    [ 2493420.459829] RIP [<ffffffffa031a8df>] ext4_put_super+0x36f/0x3c0 [ext4]
  52.  
    [ 2493420.466633] RSP <ffffc90000017de8>
  53.  
    crash>

通过dmesg日志,我们可以通过两个方法判断 bug的代码位置:

  1.  
    1. [2493420.219336] kernel BUG at fs/ext4/super.c:879!
  2.  
     
  3.  
    2. [2493420.273425] RIP: 0010:[<ffffffffa031a8df>] [<ffffffffa031a8df>] ext4_put_super+0x36f/0x3c0 [ext4]
  4.  
    其中( 0x36f代表和ext4_put_super函数入口的偏移量,0x3c0是基准地址 )

从2找到代码crash的具体位置:

  1.  
    (gdb) p 0x36f
  2.  
    $ 11 = 879

反汇编函数,找到位置

crash> dis -l ext4_put_super
 

在crash中查看代码

crash本身是可以查看代码的,前提是你需要加载模块, 比如:

加载模块ext4:

  1.  
    crash> mod -s ext4
  2.  
    crash> mod <<----列出所有的模块

第879行:

  1.  
    crash> l *ext4_put_super+ 0x36f
  2.  
    0xffffffffa031a8df is in ext4_put_super (fs/ext4/super.c:879).
  3.  
    874 * isn't empty. The on-disk one can be non-empty if we've
  4.  
    875 * detected an error and taken the fs readonly, but the
  5.  
    876 * in-memory list had better be clean by this point. */
  6.  
    877 if (!list_empty(&sbi->s_orphan))
  7.  
    878 dump_orphan_list(sb, sbi);
  8.  
    879 J_ASSERT(list_empty(&sbi->s_orphan));
  9.  
    880
  10.  
    881 sync_blockdev(sb->s_bdev);
  11.  
    882 invalidate_bdev(sb->s_bdev);
  12.  
    883 if (sbi->journal_bdev && sbi->journal_bdev != sb->s_bdev) {

只有当我们找到具体的代码,才能进一步分析代码,究竟为什么会crash,比如,这个函数的参数(可能是某个struct)的值到底是什么?

bt打印栈

bt栈[exception RIP: ext4_put_super+879] 有可以看到是在 函数ext4_put_super的第879行

  1.  
    crash> bt
  2.  
    PID: 1 TASK: ffff887e45918000 CPU: 58 COMMAND: "systemd-shutdow"
  3.  
    #0 [ffffc90000017a58] machine_kexec at ffffffff810603e8
  4.  
    #1 [ffffc90000017ab8] __crash_kexec at ffffffff811211cd
  5.  
    #2 [ffffc90000017b80] __crash_kexec at ffffffff811212a5
  6.  
    #3 [ffffc90000017b98] crash_kexec at ffffffff811212eb
  7.  
    #4 [ffffc90000017bb8] oops_end at ffffffff81030905
  8.  
    #5 [ffffc90000017be0] die at ffffffff81030ddb
  9.  
    #6 [ffffc90000017c10] do_trap at ffffffff8102df02
  10.  
    #7 [ffffc90000017c60] do_error_trap at ffffffff8102e2d9
  11.  
    #8 [ffffc90000017d20] do_invalid_op at ffffffff8102e830
  12.  
    #9 [ffffc90000017d30] invalid_op at ffffffff8171b63e
  13.  
    [exception RIP: ext4_put_super+879]
  14.  
    RIP: ffffffffa031a8df RSP: ffffc90000017de8 RFLAGS: 00010206
  15.  
    RAX: ffff88490edabf50 RBX: ffff887e43299000 RCX: 00000001949b336d
  16.  
    RDX: 0000000000000000 RSI: 0000000000000206 RDI: 0000000000000206
  17.  
    RBP: ffffc90000017e18 R8: 00000000000081a4 R9: 0000000000000000
  18.  
    R10: 0000000000000cb8 R11: 0000000000001e92 R12: ffff887e43299278
  19.  
    R13: ffff887e43298800 R14: ffff887e43299278 R15: ffffffffa034ff88
  20.  
    ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018
  21.  
    #10 [ffffc90000017de0] ext4_put_super at ffffffffa031a91c [ext4]
  22.  
    #11 [ffffc90000017e20] generic_shutdown_super at ffffffff81244aaf
  23.  
    #12 [ffffc90000017e40] kill_block_super at ffffffff81244e37
  24.  
    #13 [ffffc90000017e60] deactivate_locked_super at ffffffff81244f73
  25.  
    #14 [ffffc90000017e80] deactivate_super at ffffffff8124547a
  26.  
    #15 [ffffc90000017e98] cleanup_mnt at ffffffff81264b2f
  27.  
    #16 [ffffc90000017eb0] __cleanup_mnt at ffffffff81264bc2
  28.  
    #17 [ffffc90000017ec0] task_work_run at ffffffff810a7b50
  29.  
    #18 [ffffc90000017f00] exit_to_usermode_loop at ffffffff810032ba
  30.  
    #19 [ffffc90000017f30] syscall_return_slowpath at ffffffff81003baa
  31.  
    #20 [ffffc90000017f50] entry_SYSCALL_64_fastpath at ffffffff8171a783
  32.  
    RIP: 00007f3241195c47 RSP: 00007fffb3db5438 RFLAGS: 00000246
  33.  
    RAX: 0000000000000000 RBX: 0000560b87fbd920 RCX: 00007f3241195c47
  34.  
    RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000560b87fbdd10
  35.  
    RBP: 0000560b87fbda00 R8: 0000000000000000 R9: 00007f32410e416d
  36.  
    R10: 0000000000000021 R11: 0000000000000246 R12: 0000560b87fbdd10
  37.  
    R13: 00007fffb3db5538 R14: 00007fffb3db5523 R15: 0000000000000000
  38.  
    ORIG_RAX: 00000000000000a6 CS: 0033 SS: 002b
  39.  
    crash>

反汇编上下函数

当我们,分析到了出错的具体的代码行,下一步需要分析,传入的参数和struct

首先,我们需要看下 函数 ext4_put_super的原型,发现是static void ext4_put_super(struct super_block *sb),只有一个参数, 而且是一个结构体struct super_block, 现在我们需要知道 *sb 指针的地址是多少呢? 那这个地址肯定是 上个函数 generic_shutdown_super 传递给它的.

现在分析的关键是,我们需要知道,当generic_shutdown_superffffffff81244aaf 处,调用到 ext4_put_super的时候,传给 ext4_put_super 的指针地址是多少?

首先,需要 反汇编 函数generic_shutdown_super 找到地址ffffffff81244aaf

  1.  
    crash> dis -l generic_shutdown_super
  2.  
    /usr/src/debug/kernel- 4.9.79-009.ali3000/linux-4.9.79-009.ali3000.alios7.x86_64/fs/super.c: 436
  3.  
    0xffffffff81244aa0 <generic_shutdown_super+96>: mov 0x30(%r12),%rax
  4.  
    0xffffffff81244aa5 <generic_shutdown_super+101>: test %rax,%rax
  5.  
    0xffffffff81244aa8 <generic_shutdown_super+104>: je 0xffffffff81244aaf <generic_shutdown_super+111>
  6.  
    /usr/src/debug/kernel-4.9.79-009.ali3000/linux-4.9.79-009.ali3000.alios7.x86_64/fs/super.c: 437
  7.  
    0xffffffff81244aaa <generic_shutdown_super+106>: mov %rbx,%rdi <===rbx 和 rdi 数据一致
  8.  
    0xffffffff81244aad <generic_shutdown_super+109>: callq *%rax <===在这里调用下个函数
  9.  
    /usr/src/debug/kernel- 4.9.79-009.ali3000/linux-4.9.79-009.ali3000.alios7.x86_64/include/linux/compiler.h: 243
  10.  
    0xffffffff81244aaf <generic_shutdown_super+111>: mov 0x608(%rbx),%rax
  11.  
    /usr/src/debug/kernel- 4.9.79-009.ali3000/linux-4.9.79-009.ali3000.alios7.x86_64/fs/super.c: 439
  12.  
    0xffffffff81244ab6 <generic_shutdown_super+118>: lea 0x608(%rbx),%rdx
  13.  
    0xffffffff81244abd <generic_shutdown_super+125>: cmp %rax,%rdx
  14.  
    0xffffffff81244ac0 <generic_shutdown_super+128>: jne 0xffffffff81244b1f <generic_shutdown_super+223>

接着,反汇编ext4_put_super, 你会发现push了很多的寄存器的值到stack

  1.  
    crash> dis -l ext4_put_super
  2.  
    /usr/src/debug/kernel- 4.9.79-009.ali3000/linux-4.9.79-009.ali3000.alios7.x86_64/fs/ext4/super.c: 824
  3.  
    0xffffffffa031a570 <ext4_put_super>: nopl 0x0(%rax,%rax,1) [FTRACE NOP]
  4.  
    0xffffffffa031a575 <ext4_put_super+5>: push %rbp
  5.  
    0xffffffffa031a576 <ext4_put_super+6>: mov %rsp,%rbp
  6.  
    0xffffffffa031a579 <ext4_put_super+9>: push %r15 <===第1个寄存器入栈
  7.  
    0xffffffffa031a57b <ext4_put_super+11>: push %r14 <===第2个寄存器入栈
  8.  
    0xffffffffa031a57d <ext4_put_super+13>: push %r13 <===第3个寄存器入栈
  9.  
    0xffffffffa031a57f <ext4_put_super+15>: push %r12 <===第4个寄存器入栈
  10.  
    0xffffffffa031a581 <ext4_put_super+17>: mov %rdi,%r13
  11.  
    0xffffffffa031a584 <ext4_put_super+20>: push %rbx <===第5个寄存器入栈(rbx是在上个函数的时候,就有值的,所以,ext4_put_super函数的第一个参数的指针的地址就是这个寄存器的值)
  12.  
    0xffffffffa031a585 <ext4_put_super+21>: sub $0x8,%rsp
  13.  
    0xffffffffa031a589 <ext4_put_super+25>: mov 0x460(%rdi),%rbx
  14.  
    /usr/src/debug/kernel-4.9.79-009.ali3000/linux-4.9.79-009.ali3000.alios7.x86_64/fs/ext4/super.c: 826
  15.  
    0xffffffffa031a590 <ext4_put_super+32>: mov 0xe0(%rbx),%r14
  16.  
    /usr/src/debug/kernel-4.9.79-009.ali3000/linux-4.9.79-009.ali3000.alios7.x86_64/fs/ext4/super.c: 830
  17.  
    0xffffffffa031a597 <ext4_put_super+39>: callq 0xffffffffa03133f0 <ext4_unregister_li_request>
  1.  
    crash> bt -f
  2.  
    #10 [ffffc90000017de0] ext4_put_super at ffffffffa031a91c [ext4]
  3.  
    ffffc90000017de8: 9cbae75a00000000( ) ffff887e43298800(第5个寄存器的值)
  4.  
    ffffc90000017df8: ffffffffa034a5e0(第4个寄存器的值) ffff887e3818c7b8(第3个寄存器的值)
  5.  
    ffffc90000017e08: 0000000000000000(第2个寄存器的值) ffff887e45918bb0(第1个寄存器的值)
  6.  
    ffffc90000017e18: ffffc90000017e38 ffffffff81244aaf(这两个是不代表寄存器的)
  7.  
    #11 [ffffc90000017e20] generic_shutdown_super at ffffffff81244aaf
  8.  
    ffffc90000017e28: 0000000000000083 ffff887e357b8680
  9.  
    ffffc90000017e38: ffffc90000017e58 ffffffff81244e37
  1.  
    crash> struct super_block ffff887e43298800
  2.  
    struct super_block {
  3.  
    s_list = {
  4.  
    next = 0xffffffff81cb3db0 <super_blocks>, <=======这里也验证了,就是地址ffff887e43298800表示的就是 struct super_block
  5.  
    prev = 0xffff887e43968800
  6.  
    },
  7.  
    s_dev = 271581185,
  8.  
    s_blocksize_bits = 12 '\f',
  9.  
    s_blocksize = 4096,
  10.  
    s_maxbytes = 17592186040320,
  11.  
    s_type = 0xffffffffa03589c0 <ext4_fs_type>,
  12.  
    s_op = 0xffffffffa034a5e0 <ext4_sops>,
  13.  
    dq_op = 0xffffffffa034a720 <ext4_quota_operations>,
  14.  
    s_qcop = 0xffffffff81843f60 <dquot_quotactl_sysfile_ops>,
  15.  
    s_export_op = 0xffffffffa034a580 <ext4_export_ops>,
  16.  
    s_flags = 805371904,
  17.  
    s_iflags = 1,
  18.  
    s_magic = 61267,
  19.  
    s_root = 0x0,
  20.  
    s_umount = {
  21.  
    count = {
  22.  
    counter = -4294967295
  23.  
    },
  24.  
    wait_list = {
  25.  
    next = 0xffff887e43298878,
  26.  
    prev = 0xffff887e43298878
  27.  
    },
  28.  
    wait_lock = {
  29.  
    raw_lock = {
  30.  
    val = {
  31.  
    counter = 0
  32.  
    }
  33.  
    }

Refers

https://blog.csdn.net/u013982161/article/details/51347944

转载于:https://www.cnblogs.com/muahao/p/9925629.html

展开阅读全文
打赏
0
0 收藏
分享
加载中
更多评论
打赏
0 评论
0 收藏
0
分享
返回顶部
顶部