public inbox for [email protected]
 help / color / mirror / Atom feed
From: Pavel Begunkov <[email protected]>
To: Hao Sun <[email protected]>, [email protected], [email protected]
Cc: [email protected]
Subject: Re: BUG: unable to handle kernel NULL pointer dereference in io_uring_cancel_task_requests
Date: Sun, 11 Apr 2021 10:09:41 +0100	[thread overview]
Message-ID: <[email protected]> (raw)
In-Reply-To: <CACkBjsZLvtAVYO4MTtB8E+E3TzRDxCrBJ8Y6oeepRC0tRmmiAA@mail.gmail.com>

On 11/04/2021 09:58, Hao Sun wrote:
> Pavel Begunkov <[email protected]> 于2021年4月11日周日 下午4:14写道:
>>
>> On 11/04/2021 04:08, Hao Sun wrote:
>>> Hi
>>>
>>> When using Healer(https://github.com/SunHao-0/healer/tree/dev) to fuzz
>>> the Linux kernel, I found a null-ptr-deref bug in
>>> io_uring_cancel_task_requests under fault injection condition, but I'm
>>> not sure about this.
>>> Sorry, I do not have a reproducing program for this bug.
>>> I hope that the stack trace information in the crash log can help you
>>> locate the problem.
>>
>> Thanks Hao. io_cqring_wait() fails should not anyhow affect
>> cancellation, so the log doesn't make sense from first sight,
>> something strange is going on.
>>
> Is it possible that the failure of io_cqring_wait affects other
> operations with side effects between io_cqring_wait and cancellation,
> which eventually leads to the cancellation bug?

It shouldn't in theory, but need to a look deeper

TL;DR;
ctx->flags is NULL dereference, means that tctx->xa entry is invalid
or file->private got corrupted/not set. Your kernel is old enough
(5.11-ish), so it's a bit more safer in that regard and all
manipulations with ->xa are pretty much made by the task itself,
so should be synchronised.

There are things like io_run_task_work() or overflow_flush() that
are done in the cqring_wait(), but not much. It also grabs a file
beforehand and puts afterwards, extra reference would lead to hangs
not such failures.

> I found the last call sequence (Syzlang format) executed by the fuzzer
> before triggering the bug.
> This may be helpful, but there is no guarantee that this is the direct
> cause of the bug.

appreciate that

> 
> Possible guilty test case:
> r19 = syz_io_uring_setup(0x7211,
> &(0x7f0000000540)={0x6e3620b713f86b87,0xf615,0x2,0x1000,0x1a6,0xa26bc79d6b5315eb,0x0,[0x0,0x0,0x0],[0x813a698e7df9790f,0x1,0xb43ab5cc286248ee,0xe543f3b8cf765dd5,0x8005afeb090b0e62,0x1a29b15882d5d0b7,0xd7dc82c17c7ba1a7,0xab9d3c813ad3ae79,0x0,0x0],[0x1,0xd3a439e17ea7133c,0x4b845483eeeab284,0xf6fdf7f35d59044,0xf,0x99a9733bb1278a03,0xf8a69ea77c12e2b2,0x1,0x1,0x176ecee6d3c04836]},
> &(0x7f0000000000/0x5000)=nil, &(0x7f0000000000/0x120000)=nil,
> &(0x7f00000005c0)=<r17=>0x0, &(0x7f0000000600)=<r18=>0x0)
> io_uring_enter(r19, 0x1, 0x66ab, 0x3,
> &(0x7f0000000040)={[0xfffe8c2bdda0afdd]}, 0x8)
> io_uring_register$IORING_UNREGISTER_EVENTFD(r19, 0x5, 0x0, 0x0)
> 
>>>
>>> Here is the details:
>>> commit:   3b9cdafb5358eb9f3790de2f728f765fef100731
>>> version:   linux 5.11
>>> git tree:    upstream
>>> Full log can be found in the attachment.
>>> cqwait()
>>> Fault injection log:
>>> FAULT_INJECTION: forcing a failure.
>>> name fail_usercopy, interval 1, probability 0, space 0, times 0
>>> CPU: 1 PID: 9161 Comm: executor Not tainted 5.11.0+ #5
>>> Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS
>>> rel-1.12.0-59-gc9ba5276e321-prebuilt.qemu.org 04/01/2014
>>> Call Trace:
>>>  __dump_stack lib/dump_stack.c:79 [inline]
>>>  dump_stack+0x137/0x194 lib/dump_stack.c:120
>>>  fail_dump lib/fault-inject.c:52 [inline]
>>>  should_fail+0x23e/0x250 lib/fault-inject.c:146
>>>  should_fail_usercopy+0x16/0x20 lib/fault-inject-usercopy.c:37
>>>  _copy_from_user+0x1c/0xd0 lib/usercopy.c:14
>>>  copy_from_user include/linux/uaccess.h:192 [inline]
>>>  set_user_sigmask+0x4b/0x110 kernel/signal.c:3015
>>>  io_cqring_wait+0x2e3/0x8b0 fs/io_uring.c:7250
>>>  __do_sys_io_uring_enter fs/io_uring.c:9480 [inline]
>>>  __se_sys_io_uring_enter+0x8fc/0xb70 fs/io_uring.c:9397
>>>  __x64_sys_io_uring_enter+0x74/0x80 fs/io_uring.c:9397
>>>  do_syscall_64+0x39/0x80 arch/x86/entry/common.c:46
>>>  entry_SYSCALL_64_after_hwframe+0x44/0xae
>>> RIP: 0033:0x46a379
>>> Code: f7 d8 64 89 02 b8 ff ff ff ff c3 66 0f 1f 44 00 00 48 89 f8 48
>>> 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d
>>> 01 f0 ff ff 73 01 c3 48 c7 c1 bc ff ff ff f7 d8 64 89 01 48
>>> RSP: 002b:00007f046fa19c58 EFLAGS: 00000246 ORIG_RAX: 00000000000001aa
>>> RAX: ffffffffffffffda RBX: 000000000078c080 RCX: 000000000046a379
>>> RDX: 00000000000066ab RSI: 0000000000000001 RDI: 0000000000000003
>>> RBP: 00007f046fa19c90 R08: 0000000020000040 R09: 0000000000000008
>>> R10: 0000000000000003 R11: 0000000000000246 R12: 0000000000000000
>>> R13: 0000000000000000 R14: 000000000078c080 R15: 00007fff769deef0
>>>
>>> Crash log:
>>> BUG: kernel NULL pointer dereference, address: 0000000000000040
>>> #PF: supervisor read access in kernel mode
>>> #PF: error_code(0x0000) - not-present page
>>> PGD 49954067 P4D 49954067 PUD 45f92067 PMD 0
>>> Oops: 0000 [#1] PREEMPT SMP
>>> CPU: 1 PID: 9161 Comm: executor Not tainted 5.11.0+ #5
>>> Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS
>>> rel-1.12.0-59-gc9ba5276e321-prebuilt.qemu.org 04/01/2014
>>> RIP: 0010:io_uring_cancel_task_requests+0x3f/0x990 fs/io_uring.c:9045
>>> Code: 48 8b 04 25 28 00 00 00 48 89 44 24 68 e8 89 e6 c5 ff 65 4c 8b
>>> 34 25 00 6d 01 00 49 8d 7c 24 40 48 89 7c 24 30 e8 81 97 d6 ff <41> 8b
>>> 5c 24 40 89 de 83 e6 02 31 ff e8 70 ea c5 ff 83 e3 02 48 89
>>> RSP: 0018:ffffc90002a97b48 EFLAGS: 00010246
>>> RAX: ffff88804b8e0d38 RBX: ffff88804b8ad700 RCX: 0000000000000764
>>> RDX: 0000000000000040 RSI: ffff8880409d5140 RDI: 0000000000000040
>>> RBP: ffff8880409d5140 R08: 0000000000000000 R09: 0000000000000043
>>> R10: 0001ffffffffffff R11: ffff88804b8e0280 R12: 0000000000000000
>>> R13: ffff8880409d5140 R14: ffff88804b8e0280 R15: ffff8880481c1800
>>> FS:  00007f046fa1a700(0000) GS:ffff88807ec00000(0000) knlGS:0000000000000000
>>> CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
>>> CR2: 0000000000000040 CR3: 00000000479a5000 CR4: 0000000000750ee0
>>> DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
>>> DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
>>> PKRU: 55555554
>>> Call Trace:
>>>  __io_uring_files_cancel+0x9b/0x200 fs/io_uring.c:9140
>>>  io_uring_files_cancel include/linux/io_uring.h:65 [inline]
>>>  do_exit+0x1a8/0x16d0 kernel/exit.c:780
>>>  do_group_exit+0xc5/0x180 kernel/exit.c:922
>>>  get_signal+0xd90/0x1470 kernel/signal.c:2773
>>>  arch_do_signal_or_restart+0x2a/0x260 arch/x86/kernel/signal.c:811
>>>  handle_signal_work kernel/entry/common.c:147 [inline]
>>>  exit_to_user_mode_loop kernel/entry/common.c:171 [inline]
>>>  exit_to_user_mode_prepare+0x109/0x1a0 kernel/entry/common.c:208
>>>  __syscall_exit_to_user_mode_work kernel/entry/common.c:290 [inline]
>>>  syscall_exit_to_user_mode+0x20/0x40 kernel/entry/common.c:301
>>>  do_syscall_64+0x45/0x80 arch/x86/entry/common.c:56
>>>  entry_SYSCALL_64_after_hwframe+0x44/0xae
>>> RIP: 0033:0x46a379
>>> Code: f7 d8 64 89 02 b8 ff ff ff ff c3 66 0f 1f 44 00 00 48 89 f8 48
>>> 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d
>>> 01 f0 ff ff 73 01 c3 48 c7 c1 bc ff ff ff f7 d8 64 89 01 48
>>> RSP: 002b:00007f046fa19cd8 EFLAGS: 00000246 ORIG_RAX: 00000000000000ca
>>> RAX: fffffffffffffe00 RBX: 000000000078c080 RCX: 000000000046a379
>>> RDX: 0000000000000000 RSI: 0000000000000080 RDI: 000000000078c088
>>> RBP: 000000000078c088 R08: 0000000000000000 R09: 0000000000000000
>>> R10: 0000000000000000 R11: 0000000000000246 R12: 000000000078c08c
>>> R13: 0000000000000000 R14: 000000000078c080 R15: 00007fff769deef0
>>> Modules linked in:
>>> Dumping ftrace buffer:
>>>    (ftrace buffer empty)
>>> CR2: 0000000000000040
>>> ---[ end trace 613db1a25ecf6443 ]---
>>> RIP: 0010:io_uring_cancel_task_requests+0x3f/0x990 fs/io_uring.c:9045
>>> Code: 48 8b 04 25 28 00 00 00 48 89 44 24 68 e8 89 e6 c5 ff 65 4c 8b
>>> 34 25 00 6d 01 00 49 8d 7c 24 40 48 89 7c 24 30 e8 81 97 d6 ff <41> 8b
>>> 5c 24 40 89 de 83 e6 02 31 ff e8 70 ea c5 ff 83 e3 02 48 89
>>> RSP: 0018:ffffc90002a97b48 EFLAGS: 00010246
>>> RAX: ffff88804b8e0d38 RBX: ffff88804b8ad700 RCX: 0000000000000764
>>> RDX: 0000000000000040 RSI: ffff8880409d5140 RDI: 0000000000000040
>>> RBP: ffff8880409d5140 R08: 0000000000000000 R09: 0000000000000043
>>> R10: 0001ffffffffffff R11: ffff88804b8e0280 R12: 0000000000000000
>>> R13: ffff8880409d5140 R14: ffff88804b8e0280 R15: ffff8880481c1800
>>> FS:  00007f046fa1a700(0000) GS:ffff88807ec00000(0000) knlGS:0000000000000000
>>> CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
>>> CR2: 0000000000000040 CR3: 00000000479a5000 CR4: 0000000000750ee0
>>> DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
>>> DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
>>> PKRU: 55555554
>>>
>>
>> --
>> Pavel Begunkov

-- 
Pavel Begunkov

      reply	other threads:[~2021-04-11  9:13 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-04-11  3:08 BUG: unable to handle kernel NULL pointer dereference in io_uring_cancel_task_requests Hao Sun
2021-04-11  8:10 ` Pavel Begunkov
2021-04-11  8:58   ` Hao Sun
2021-04-11  9:09     ` Pavel Begunkov [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox