Re: About the conflict between XFS inode recycle and VFS rcu-walk

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 




On 16/5/24 15:08, Ian Kent wrote:
On 16/5/24 12:56, Jinliang Zheng wrote:
On Wed, 15 May 2024 at 23:54:41 +0800, Jinliang Zheng wrote:
On Wed, 31 Jan 2024 at 11:30:18 -0800, djwong@xxxxxxxxxx wrote:
On Wed, Jan 31, 2024 at 02:35:17PM +0800, Jinliang Zheng wrote:
On Fri, 8 Dec 2023 11:14:32 +1100, david@xxxxxxxxxxxxx wrote:
On Tue, Dec 05, 2023 at 07:38:33PM +0800, alexjlzheng@xxxxxxxxx wrote:
Hi, all

I would like to ask if the conflict between xfs inode recycle and vfs rcu-walk
which can lead to null pointer references has been resolved?

I browsed through emails about the following patches and their discussions: - https://lore.kernel.org/linux-xfs/20220217172518.3842951-2-bfoster@xxxxxxxxxx/ - https://lore.kernel.org/linux-xfs/20220121142454.1994916-1-bfoster@xxxxxxxxxx/ - https://lore.kernel.org/linux-xfs/164180589176.86426.501271559065590169.stgit@xxxxxxxxxxxxxxxxx/

And then came to the conclusion that this problem has not been solved, am I
right? Did I miss some patch that could solve this problem?
We fixed the known problems this caused by turning off the VFS
functionality that the rcu pathwalks kept tripping over. See commit
7b7820b83f23 ("xfs: don't expose internal symlink metadata buffers to
the vfs").
Sorry for the delay.

The problem I encountered in the production environment was that during the rcu walk process the ->get_link() pointer was NULL, which caused a crash.

As far as I know, commit 7b7820b83f23 ("xfs: don't expose internal symlink
metadata buffers to the vfs") first appeared in:
- https://lore.kernel.org/linux-fsdevel/YZvvP9RFXi3%2FjX0q@bfoster/

Does this commit solve the problem of NULL ->get_link()? And how?
I suggest reading the call stack from wherever the VFS enters the XFS
readlink code.  If you have a reliable reproducer, then apply this patch
to your kernel (you haven't mentioned which one it is) and see if the
bad dereference goes away.

--D
Sorry for the delay.

I encountered the following calltrace:

[20213.578756] BUG: kernel NULL pointer dereference, address: 0000000000000000
[20213.578785] #PF: supervisor instruction fetch in kernel mode
[20213.578799] #PF: error_code(0x0010) - not-present page
[20213.578812] PGD 3f01d64067 P4D 3f01d64067 PUD 3f01d65067 PMD 0
[20213.578828] Oops: 0010 [#1] SMP NOPTI
[20213.578839] CPU: 92 PID: 766 Comm: /usr/local/serv Kdump: loaded Not tainted 5.4.241-1-tlinux4-0017.3 #1 [20213.578860] Hardware name: New H3C Technologies Co., Ltd. UniServer R4900 G3/RS33M2C9SA, BIOS 2.00.38P02 04/14/2020
[20213.578884] RIP: 0010:0x0
[20213.578894] Code: Bad RIP value.
[20213.578903] RSP: 0018:ffffc90021ebfc38 EFLAGS: 00010246
[20213.578916] RAX: ffffffff82081f40 RBX: ffffc90021ebfce0 RCX: 0000000000000000 [20213.578932] RDX: ffffc90021ebfd48 RSI: ffff88bfad8d3890 RDI: 0000000000000000 [20213.578948] RBP: ffffc90021ebfc70 R08: 0000000000000001 R09: ffff889b9eeae380 [20213.578965] R10: 302d343200000067 R11: 0000000000000001 R12: 0000000000000000 [20213.578981] R13: ffff88bfad8d3890 R14: ffff889b9eeae380 R15: ffffc90021ebfd48 [20213.578998] FS:  00007f89c534e740(0000) GS:ffff88c07fd00000(0000) knlGS:0000000000000000
[20213.579016] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[20213.579030] CR2: ffffffffffffffd6 CR3: 0000003f01d90001 CR4: 00000000007706e0 [20213.579046] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [20213.579062] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[20213.579079] PKRU: 55555554
[20213.579087] Call Trace:
[20213.579099]  trailing_symlink+0x1da/0x260
[20213.579112]  path_lookupat.isra.53+0x79/0x220
[20213.579125]  filename_lookup.part.69+0xa0/0x170
[20213.579138]  ? kmem_cache_alloc+0x3f/0x3f0
[20213.579151]  ? getname_flags+0x4f/0x1e0
[20213.579161]  user_path_at_empty+0x3e/0x50
[20213.579172]  vfs_statx+0x76/0xe0
[20213.579182]  __do_sys_newstat+0x3d/0x70
[20213.579194]  ? fput+0x13/0x20
[20213.579203]  ? ksys_ioctl+0xb0/0x300
[20213.579213]  ? generic_file_llseek+0x24/0x30
[20213.579225]  ? fput+0x13/0x20
[20213.579233]  ? ksys_lseek+0x8d/0xb0
[20213.579243]  __x64_sys_newstat+0x16/0x20
[20213.579256]  do_syscall_64+0x4d/0x140
[20213.579268]  entry_SYSCALL_64_after_hwframe+0x5c/0xc1

<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
Please note that the kernel version I use is the one maintained by Tencent.Inc, and the baseline is v5.4. But in fact, in the latest upstream source tree, although the trailing_symlink() function has been removed, its logic has been
moved to pick_link(), so the problem still exists.

Ian Kent pointed out that try_to_unlazy() was introduced in pick_link() in the latest upstream source tree, but I don't understand why this can solve the NULL ->get_link pointer dereference problem, because ->get_link pointer will be
dereferenced before try_to_unlazy().

(I don't understand why Ian Kent's email didn't appear on the mailing list.)

It was something about html mail and I think my mail client was at fault.

In any case what you say is indeed correct, so the comment isn't important.


Fact is it is still a race between the lockless path walk and inode eviction

and xfs recycling. I believe that the xfs recycling code is very hard to fix.


IIRC correctly putting a NULL check in pick_link() was not considered acceptable

but there must be a way that is acceptable to check this and restart the walk.

Maybe there was a reluctance to suffer the overhead of restarting the walk when

it shouldn't be needed.

Or perhaps the worry was that if it can become NULL it could also become a pointer to a

different (incorrect) link altogether which could have really odd/unpleasant outcomes.




The alternative would be to find some way to identify when it's unsafe to reuse

an inode marked for re-cycle before dropping rcu read, perhaps with the reference

count plus the seqlock. Basically, to reuse inodes xfs will need to identify when

the race occurs and let the inode go away under rcu and create a new one if a race

is detected. But possibly that isn't nearly as simple as it sounds?



Thanks,
Jinliang Zheng

And I analyzed the disassembly of trailing_symlink() and confirmed that a NULL
->get_link() happened here:

0xffffffff812e4850 <trailing_symlink>:    nopl 0x0(%rax,%rax,1) [FTRACE NOP]
0xffffffff812e4855 <trailing_symlink+0x5>:    push %rbp
0xffffffff812e4856 <trailing_symlink+0x6>:    mov %rsp,%rbp
0xffffffff812e4859 <trailing_symlink+0x9>:    push %r15
0xffffffff812e485b <trailing_symlink+0xb>:    push %r14
0xffffffff812e485d <trailing_symlink+0xd>:    push %r13
0xffffffff812e485f <trailing_symlink+0xf>:    push %r12
0xffffffff812e4861 <trailing_symlink+0x11>:    push %rbx
0xffffffff812e4862 <trailing_symlink+0x12>:    mov %rdi,%rbx        # rbx = &nameidate
0xffffffff812e4865 <trailing_symlink+0x15>:    sub $0x8,%rsp
0xffffffff812e4869 <trailing_symlink+0x19>:    mov 0x1765845(%rip),%edx     # 0xffffffff82a4a0b4 <sysctl_protected_symlinks>
0xffffffff812e486f <trailing_symlink+0x1f>:    mov 0x38(%rdi),%eax
0xffffffff812e4872 <trailing_symlink+0x22>:    test %edx,%edx
0xffffffff812e4874 <trailing_symlink+0x24>:    je 0xffffffff812e48ac <trailing_symlink+0x5c>
0xffffffff812e4876 <trailing_symlink+0x26>:    mov %gs:0x1ad00,%rdx
0xffffffff812e487f <trailing_symlink+0x2f>:    mov 0xc8(%rdi),%rcx        # rcx = nameidata->link_inode
0xffffffff812e4886 <trailing_symlink+0x36>:    mov 0xc18(%rdx),%rdx
0xffffffff812e488d <trailing_symlink+0x3d>:    mov 0x4(%rcx),%ecx        # ecx = link_inode->uid
0xffffffff812e4890 <trailing_symlink+0x40>:    cmp %ecx,0x1c(%rdx)
0xffffffff812e4893 <trailing_symlink+0x43>:    je 0xffffffff812e48ac <trailing_symlink+0x5c>
0xffffffff812e4895 <trailing_symlink+0x45>:    mov 0x30(%rdi),%rsi
0xffffffff812e4899 <trailing_symlink+0x49>:    movzwl (%rsi),%edx
0xffffffff812e489c <trailing_symlink+0x4c>:    and $0x202,%dx
0xffffffff812e48a1 <trailing_symlink+0x51>:    cmp $0x202,%dx
0xffffffff812e48a6 <trailing_symlink+0x56>:    je 0xffffffff812e495f <trailing_symlink+0x10f>
0xffffffff812e48ac <trailing_symlink+0x5c>:    or $0x10,%eax
0xffffffff812e48af <trailing_symlink+0x5f>:    mov %eax,0x38(%rbx)        # nd->flags |= LOOKUP_PARENT 0xffffffff812e48b2 <trailing_symlink+0x62>:    mov 0x50(%rbx),%rax        # rax = nd->stack 0xffffffff812e48b6 <trailing_symlink+0x66>:    movq $0x0,0x20(%rax)        # stack[0].name = NULL 0xffffffff812e48be <trailing_symlink+0x6e>:    mov 0x48(%rbx),%eax        # nd->depth 0xffffffff812e48c1 <trailing_symlink+0x71>:    mov 0x50(%rbx),%rdx        # nd->stack 0xffffffff812e48c5 <trailing_symlink+0x75>:    mov 0xc8(%rbx),%r13        # nd->link_inode 0xffffffff812e48cc <trailing_symlink+0x7c>:    lea (%rax,%rax,2),%rax    # rax = depth * 3 0xffffffff812e48d0 <trailing_symlink+0x80>:    shl $0x4,%rax        # rax = rax << 4, sizeof(saved):0x30 0xffffffff812e48d4 <trailing_symlink+0x84>:    lea -0x30(%rdx,%rax,1),%r15    # r15 = last 0xffffffff812e48d9 <trailing_symlink+0x89>:    mov 0x8(%r15),%r14        # r14 = last->link.dentry
0xffffffff812e48dd <trailing_symlink+0x8d>:    testb $0x40,0x38(%rbx)
0xffffffff812e48e1 <trailing_symlink+0x91>:    je 0xffffffff812e4950 <trailing_symlink+0x100>
0xffffffff812e48e3 <trailing_symlink+0x93>:    mov %r13,%rsi
0xffffffff812e48e6 <trailing_symlink+0x96>:    mov %r15,%rdi
0xffffffff812e48e9 <trailing_symlink+0x99>:    callq 0xffffffff812f8a00 <atime_needs_update>
0xffffffff812e48ee <trailing_symlink+0x9e>:    test %al,%al
0xffffffff812e48f0 <trailing_symlink+0xa0>:    jne 0xffffffff812e4a56 <trailing_symlink+0x206>
0xffffffff812e48f6 <trailing_symlink+0xa6>:    mov 0x38(%rbx),%edx
0xffffffff812e48f9 <trailing_symlink+0xa9>:    mov %r13,%rsi
0xffffffff812e48fc <trailing_symlink+0xac>:    mov %r14,%rdi
0xffffffff812e48ff <trailing_symlink+0xaf>:    shr $0x6,%edx
0xffffffff812e4902 <trailing_symlink+0xb2>:    and $0x1,%edx
0xffffffff812e4905 <trailing_symlink+0xb5>:    callq 0xffffffff81424310 <security_inode_follow_link>
0xffffffff812e490a <trailing_symlink+0xba>:    movslq %eax,%r12
0xffffffff812e490d <trailing_symlink+0xbd>:    test %eax,%eax
0xffffffff812e490f <trailing_symlink+0xbf>:    jne 0xffffffff812e4939 <trailing_symlink+0xe9>
0xffffffff812e4911 <trailing_symlink+0xc1>:    movl $0x4,0x44(%rbx)
0xffffffff812e4918 <trailing_symlink+0xc8>:    mov 0x248(%r13),%r12
0xffffffff812e491f <trailing_symlink+0xcf>:    test %r12,%r12
0xffffffff812e4922 <trailing_symlink+0xd2>:    je 0xffffffff812e49e5 <trailing_symlink+0x195>
0xffffffff812e4928 <trailing_symlink+0xd8>:    movzbl (%r12),%eax
0xffffffff812e492d <trailing_symlink+0xdd>:    cmp $0x2f,%al
0xffffffff812e492f <trailing_symlink+0xdf>:    je 0xffffffff812e49b7 <trailing_symlink+0x167>
0xffffffff812e4935 <trailing_symlink+0xe5>:    test %al,%al
0xffffffff812e4937 <trailing_symlink+0xe7>:    je 0xffffffff812e49ae <trailing_symlink+0x15e>
0xffffffff812e4939 <trailing_symlink+0xe9>:    test %r12,%r12
0xffffffff812e493c <trailing_symlink+0xec>:    je 0xffffffff812e49ae <trailing_symlink+0x15e>
0xffffffff812e493e <trailing_symlink+0xee>:    add $0x8,%rsp
0xffffffff812e4942 <trailing_symlink+0xf2>:    mov %r12,%rax
0xffffffff812e4945 <trailing_symlink+0xf5>:    pop %rbx
0xffffffff812e4946 <trailing_symlink+0xf6>:    pop %r12
0xffffffff812e4948 <trailing_symlink+0xf8>:    pop %r13
0xffffffff812e494a <trailing_symlink+0xfa>:    pop %r14
0xffffffff812e494c <trailing_symlink+0xfc>:    pop %r15
0xffffffff812e494e <trailing_symlink+0xfe>:    pop %rbp
0xffffffff812e494f <trailing_symlink+0xff>:    retq
0xffffffff812e4950 <trailing_symlink+0x100>:    mov %r15,%rdi
0xffffffff812e4953 <trailing_symlink+0x103>:    callq 0xffffffff812f8ae0 <touch_atime> 0xffffffff812e4958 <trailing_symlink+0x108>:    callq 0xffffffff81a26410 <_cond_resched> 0xffffffff812e495d <trailing_symlink+0x10d>:    jmp 0xffffffff812e48f6 <trailing_symlink+0xa6>
0xffffffff812e495f <trailing_symlink+0x10f>:    mov 0x4(%rsi),%edx
0xffffffff812e4962 <trailing_symlink+0x112>:    cmp $0xffffffff,%edx
0xffffffff812e4965 <trailing_symlink+0x115>:    je 0xffffffff812e496f <trailing_symlink+0x11f>
0xffffffff812e4967 <trailing_symlink+0x117>:    cmp %edx,%ecx
0xffffffff812e4969 <trailing_symlink+0x119>:    je 0xffffffff812e48ac <trailing_symlink+0x5c> 0xffffffff812e496f <trailing_symlink+0x11f>:    mov $0xfffffffffffffff6,%r12
0xffffffff812e4976 <trailing_symlink+0x126>:    test $0x40,%al
0xffffffff812e4978 <trailing_symlink+0x128>:    jne 0xffffffff812e493e <trailing_symlink+0xee>
0xffffffff812e497a <trailing_symlink+0x12a>:    mov %gs:0x1ad00,%rax
0xffffffff812e4983 <trailing_symlink+0x133>:    mov 0xce0(%rax),%rax
0xffffffff812e498a <trailing_symlink+0x13a>:    test %rax,%rax
0xffffffff812e498d <trailing_symlink+0x13d>:    je 0xffffffff812e4999 <trailing_symlink+0x149>
0xffffffff812e498f <trailing_symlink+0x13f>:    mov (%rax),%eax
0xffffffff812e4991 <trailing_symlink+0x141>:    test %eax,%eax
0xffffffff812e4993 <trailing_symlink+0x143>:    je 0xffffffff812e4a6f <trailing_symlink+0x21f> 0xffffffff812e4999 <trailing_symlink+0x149>:    mov $0xffffffff82319b4f,%rdi 0xffffffff812e49a0 <trailing_symlink+0x150>:    mov $0xfffffffffffffff3,%r12 0xffffffff812e49a7 <trailing_symlink+0x157>:    callq 0xffffffff81161310 <audit_log_link_denied> 0xffffffff812e49ac <trailing_symlink+0x15c>:    jmp 0xffffffff812e493e <trailing_symlink+0xee> 0xffffffff812e49ae <trailing_symlink+0x15e>:    mov $0xffffffff8230164d,%r12 0xffffffff812e49b5 <trailing_symlink+0x165>:    jmp 0xffffffff812e493e <trailing_symlink+0xee>
0xffffffff812e49b7 <trailing_symlink+0x167>:    cmpq $0x0,0x20(%rbx)
0xffffffff812e49bc <trailing_symlink+0x16c>:    je 0xffffffff812e4a8a <trailing_symlink+0x23a>
0xffffffff812e49c2 <trailing_symlink+0x172>:    mov %rbx,%rdi
0xffffffff812e49c5 <trailing_symlink+0x175>:    callq 0xffffffff812e2da0 <nd_jump_root>
0xffffffff812e49ca <trailing_symlink+0x17a>:    test %eax,%eax
0xffffffff812e49cc <trailing_symlink+0x17c>:    jne 0xffffffff812e4a97 <trailing_symlink+0x247>
0xffffffff812e49d2 <trailing_symlink+0x182>:    add $0x1,%r12
0xffffffff812e49d6 <trailing_symlink+0x186>:    movzbl (%r12),%eax
0xffffffff812e49db <trailing_symlink+0x18b>:    cmp $0x2f,%al
0xffffffff812e49dd <trailing_symlink+0x18d>:    jne 0xffffffff812e4935 <trailing_symlink+0xe5> 0xffffffff812e49e3 <trailing_symlink+0x193>:    jmp 0xffffffff812e49d2 <trailing_symlink+0x182> 0xffffffff812e49e5 <trailing_symlink+0x195>:    mov 0x20(%r13),%rax        # inode->i_op
0xffffffff812e49e9 <trailing_symlink+0x199>:    add $0x10,%r15
0xffffffff812e49ed <trailing_symlink+0x19d>:    mov %r13,%rsi
0xffffffff812e49f0 <trailing_symlink+0x1a0>:    mov %r15,%rdx
0xffffffff812e49f3 <trailing_symlink+0x1a3>:    mov 0x8(%rax),%rcx        # inode_operations->get_link
0xffffffff812e49f7 <trailing_symlink+0x1a7>:    testb $0x40,0x38(%rbx)
0xffffffff812e49fb <trailing_symlink+0x1ab>:    jne 0xffffffff812e4a1f <trailing_symlink+0x1cf> 0xffffffff812e49fd <trailing_symlink+0x1ad>:    mov %r14,%rdi        # nd->flags & LOOKUP_RCU == 0 0xffffffff812e4a00 <trailing_symlink+0x1b0>:    callq 0xffffffff81e00f70 <__x86_indirect_thunk_rcx> # jmpq *%rcx
0xffffffff812e4a05 <trailing_symlink+0x1b5>:    mov %rax,%r12
0xffffffff812e4a08 <trailing_symlink+0x1b8>:    test %r12,%r12
0xffffffff812e4a0b <trailing_symlink+0x1bb>:    je 0xffffffff812e49ae <trailing_symlink+0x15e> 0xffffffff812e4a0d <trailing_symlink+0x1bd>:    cmp $0xfffffffffffff000,%r12 0xffffffff812e4a14 <trailing_symlink+0x1c4>:    jbe 0xffffffff812e4928 <trailing_symlink+0xd8> 0xffffffff812e4a1a <trailing_symlink+0x1ca>:    jmpq 0xffffffff812e493e <trailing_symlink+0xee> 0xffffffff812e4a1f <trailing_symlink+0x1cf>:    xor %edi,%edi        # nd->flags & LOOKUP_RCU != 0
0xffffffff812e4a21 <trailing_symlink+0x1d1>:    mov %rcx,-0x30(%rbp)
0xffffffff812e4a25 <trailing_symlink+0x1d5>:    callq 0xffffffff81e00f70 <__x86_indirect_thunk_rcx> # jmpq *%rcx
0xffffffff812e4a2a <trailing_symlink+0x1da>:    mov %rax,%r12
0xffffffff812e4a2d <trailing_symlink+0x1dd>:    cmp $0xfffffffffffffff6,%rax 0xffffffff812e4a31 <trailing_symlink+0x1e1>:    jne 0xffffffff812e4a08 <trailing_symlink+0x1b8>
0xffffffff812e4a33 <trailing_symlink+0x1e3>:    mov %rbx,%rdi
0xffffffff812e4a36 <trailing_symlink+0x1e6>:    callq 0xffffffff812e3840 <unlazy_walk>
0xffffffff812e4a3b <trailing_symlink+0x1eb>:    test %eax,%eax
0xffffffff812e4a3d <trailing_symlink+0x1ed>:    jne 0xffffffff812e4a97 <trailing_symlink+0x247>
0xffffffff812e4a3f <trailing_symlink+0x1ef>:    mov %r15,%rdx
0xffffffff812e4a42 <trailing_symlink+0x1f2>:    mov %r13,%rsi
0xffffffff812e4a45 <trailing_symlink+0x1f5>:    mov %r14,%rdi
0xffffffff812e4a48 <trailing_symlink+0x1f8>:    mov -0x30(%rbp),%rcx
0xffffffff812e4a4c <trailing_symlink+0x1fc>:    callq 0xffffffff81e00f70 <__x86_indirect_thunk_rcx>
0xffffffff812e4a51 <trailing_symlink+0x201>:    mov %rax,%r12
0xffffffff812e4a54 <trailing_symlink+0x204>:    jmp 0xffffffff812e4a08 <trailing_symlink+0x1b8>
0xffffffff812e4a56 <trailing_symlink+0x206>:    mov %rbx,%rdi
0xffffffff812e4a59 <trailing_symlink+0x209>:    callq 0xffffffff812e3840 <unlazy_walk>
0xffffffff812e4a5e <trailing_symlink+0x20e>:    test %eax,%eax
0xffffffff812e4a60 <trailing_symlink+0x210>:    jne 0xffffffff812e4a97 <trailing_symlink+0x247>
0xffffffff812e4a62 <trailing_symlink+0x212>:    mov %r15,%rdi
0xffffffff812e4a65 <trailing_symlink+0x215>:    callq 0xffffffff812f8ae0 <touch_atime> 0xffffffff812e4a6a <trailing_symlink+0x21a>:    jmpq 0xffffffff812e48f6 <trailing_symlink+0xa6>
0xffffffff812e4a6f <trailing_symlink+0x21f>:    mov 0x50(%rbx),%rax
0xffffffff812e4a73 <trailing_symlink+0x223>:    mov 0xb8(%rbx),%rdi
0xffffffff812e4a7a <trailing_symlink+0x22a>:    xor %edx,%edx
0xffffffff812e4a7c <trailing_symlink+0x22c>:    mov 0x8(%rax),%rsi
0xffffffff812e4a80 <trailing_symlink+0x230>:    callq 0xffffffff811673f0 <__audit_inode> 0xffffffff812e4a85 <trailing_symlink+0x235>:    jmpq 0xffffffff812e4999 <trailing_symlink+0x149>
0xffffffff812e4a8a <trailing_symlink+0x23a>:    mov %rbx,%rdi
0xffffffff812e4a8d <trailing_symlink+0x23d>:    callq 0xffffffff812e4790 <set_root> 0xffffffff812e4a92 <trailing_symlink+0x242>:    jmpq 0xffffffff812e49c2 <trailing_symlink+0x172> 0xffffffff812e4a97 <trailing_symlink+0x247>:    mov $0xfffffffffffffff6,%r12 0xffffffff812e4a9e <trailing_symlink+0x24e>:    jmpq 0xffffffff812e493e <trailing_symlink+0xee>

<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<

According to my understanding, the problem solved by commit 7b7820b83f23 ("xfs: don't expose internal symlink metadata buffers to the vfs") is a data NULL pointer dereference, but the problem here is an instruction NULL pointer
dereference.

Further, I analyzed the possible triggering process as follows:

rcu_walk            do_unlinkat ~~> prune_dcache_sb create
rcu_read_lock
read_seqcount_retry
(the last check)      iput_final
                         evict
                           destroy_inode
                             xfs_fs_destroy_inode
                               xfs_inode_set_reclaim_tag xfs_ialloc
spin_lock(ip->i_flags_lock)     xfs_dialloc
                                 set(ip, XFS_IRECLAIMABLE)       xfs_iget
wakeup(xfs_reclaim_worker)        rcu_read_lock
spin_unlock(ip->i_flags_lock)     xfs_iget_cache_hit
spin_lock(ip->i_flags_lock)
                                                                     if (XFS_IRECLAIMABLE && !XFS_IRECLAIM)
set(ip, XFS_IRECLAIM)
spin_unlock(ip->i_flags_lock)
rcu_read_unlock
< ------------ >
                                                                     // miss synchronize_rcu()
xfs_reinit_inode
->get_link = NULL
get_link() // NULL

rcu_read_unlock

<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<

Therefore, I think that after commit 7b7820b83f23 ("xfs: don't expose internal symlink metadata buffers to the vfs"), we should start processing this NULL
->get_link pointer dereference.

Or, am I thinking wrong somewhere?

Thanks,
Jinliang Zheng

Apart from that issue, I'm not aware of any other issues that the
XFS inode recycling directly exposes.

According to my understanding, the essence of this problem is that XFS reuses the inode evicted by VFS, but VFS rcu-walk assumes that this will not happen.
It assumes that the inode will not change identity during the RCU
grace period after the inode has been evicted from cache. We can
safely reinstantiate an evicted inode without waiting for an RCU
grace period as long as it is the same inode with the same content
and same state.

Problems *may* arise when we unlink the inode, then evict it, then a
new file is created and the old slab cache memory address is used
for the new inode. I describe the issue here:

https://lore.kernel.org/linux-xfs/20220118232547.GD59729@xxxxxxxxxxxxxxxxxxx/
And judging from the relevant emails, the main reason why ->get_link() is set to NULL should be the lack of synchronize_rcu() before xfs_reinit_inode() when
the inode is chosen to be reused.

However, perhaps due to performance reasons, this solution has not been merged
for a long time. How is it now?

Maybe I am missing something in the threads of mail?

Thank you very much. :)
Jinliang Zheng

That said, we have exactly zero evidence that this is actually a
problem in production systems. We did get systems tripping over the
symlink issue, but there's no evidence that the
unlink->close->open(O_CREAT) issues are manifesting in the wild and
hence there hasn't been any particular urgency to address it.

Are there any recommended workarounds until an elegant and efficient solution can be proposed? After all, causing a crash is extremely unacceptable in a
production environment.
What crashes are you seeing in your production environment?

-Dave.
--
Dave Chinner
david@xxxxxxxxxxxxx





[Index of Archives]     [Linux Samsung SoC]     [Linux Rockchip SoC]     [Linux Actions SoC]     [Linux for Synopsys ARC Processors]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]


  Powered by Linux