general protection fault in io_disable_sqo_submit

[syzbot]

Hello,

syzbot found the following issue on:

HEAD commit: 7c53f6b6 Linux 5.11-rc3
git tree: upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=1606a757500000
kernel config: https://syzkaller.appspot.com/x/.config?x=c60c9ff9cc916cbc
dashboard link: https://syzkaller.appspot.com/bug?extid=ab412638aeb652ded540
compiler: gcc (GCC) 10.1.0-syz 20200507
syz repro: https://syzkaller.appspot.com/x/repro.syz?x=13adb0d0d00000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=1527be48d00000

The issue was bisected to:

commit d9d05217cb6990b9a56e13b56e7a1b71e2551f6c
Author: Pavel Begunkov <asml.s...@gmail.com>
Date: Fri Jan 8 20:57:25 2021 +0000

io_uring: stop SQPOLL submit on creator's death

bisection log: https://syzkaller.appspot.com/x/bisect.txt?x=17b3b248d00000
final oops: https://syzkaller.appspot.com/x/report.txt?x=1473b248d00000
console output: https://syzkaller.appspot.com/x/log.txt?x=1073b248d00000

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+ab4126...@syzkaller.appspotmail.com
Fixes: d9d05217cb69 ("io_uring: stop SQPOLL submit on creator's death")

RDX: 0000000000000001 RSI: 0000000020000300 RDI: 00000000000000ff
RBP: 0000000000011fc2 R08: 0000000000000001 R09: 00000000004002c8
R10: 0000000000000000 R11: 0000000000000246 R12: 00000000004021d0
R13: 0000000000402260 R14: 0000000000000000 R15: 0000000000000000
general protection fault, probably for non-canonical address 0xdffffc0000000022: 0000 [#1] PREEMPT SMP KASAN
KASAN: null-ptr-deref in range [0x0000000000000110-0x0000000000000117]
CPU: 1 PID: 8473 Comm: syz-executor814 Not tainted 5.11.0-rc3-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
RIP: 0010:io_ring_set_wakeup_flag fs/io_uring.c:6929 [inline]
RIP: 0010:io_disable_sqo_submit+0xdb/0x130 fs/io_uring.c:8891
Code: fa 48 c1 ea 03 80 3c 02 00 75 62 48 8b 9b c0 00 00 00 48 b8 00 00 00 00 00 fc ff df 48 8d bb 14 01 00 00 48 89 fa 48 c1 ea 03 <0f> b6 14 02 48 89 f8 83 e0 07 83 c0 03 38 d0 7c 04 84 d2 75 1d 83
RSP: 0018:ffffc9000154fd78 EFLAGS: 00010007
RAX: dffffc0000000000 RBX: 0000000000000000 RCX: ffffffff815976e0
RDX: 0000000000000022 RSI: 0000000000000004 RDI: 0000000000000114
RBP: ffff8880149ee480 R08: 0000000000000001 R09: 0000000000000003
R10: fffff520002a9fa1 R11: 1ffffffff1d308df R12: fffffffffffffff4
R13: 0000000000000001 R14: ffff8880149ee054 R15: ffff8880149ee000
FS: 0000000000be4880(0000) GS:ffff8880b9f00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000020000304 CR3: 0000000014b50000 CR4: 00000000001506e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
io_uring_create fs/io_uring.c:9711 [inline]
io_uring_setup+0x12b1/0x38e0 fs/io_uring.c:9739
do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
entry_SYSCALL_64_after_hwframe+0x44/0xa9
RIP: 0033:0x441309
Code: e8 5c ae 02 00 48 83 c4 18 c3 0f 1f 80 00 00 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 0f 83 3b 0a fc ff c3 66 2e 0f 1f 84 00 00 00 00
RSP: 002b:00007ffea5e64578 EFLAGS: 00000246 ORIG_RAX: 00000000000001a9
RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 0000000000441309
RDX: 0000000000000001 RSI: 0000000020000300 RDI: 00000000000000ff
RBP: 0000000000011fc2 R08: 0000000000000001 R09: 00000000004002c8
R10: 0000000000000000 R11: 0000000000000246 R12: 00000000004021d0
R13: 0000000000402260 R14: 0000000000000000 R15: 0000000000000000
Modules linked in:
---[ end trace 0941172fec2041bb ]---
RIP: 0010:io_ring_set_wakeup_flag fs/io_uring.c:6929 [inline]
RIP: 0010:io_disable_sqo_submit+0xdb/0x130 fs/io_uring.c:8891
Code: fa 48 c1 ea 03 80 3c 02 00 75 62 48 8b 9b c0 00 00 00 48 b8 00 00 00 00 00 fc ff df 48 8d bb 14 01 00 00 48 89 fa 48 c1 ea 03 <0f> b6 14 02 48 89 f8 83 e0 07 83 c0 03 38 d0 7c 04 84 d2 75 1d 83
RSP: 0018:ffffc9000154fd78 EFLAGS: 00010007
RAX: dffffc0000000000 RBX: 0000000000000000 RCX: ffffffff815976e0
RDX: 0000000000000022 RSI: 0000000000000004 RDI: 0000000000000114
RBP: ffff8880149ee480 R08: 0000000000000001 R09: 0000000000000003
R10: fffff520002a9fa1 R11: 1ffffffff1d308df R12: fffffffffffffff4
R13: 0000000000000001 R14: ffff8880149ee054 R15: ffff8880149ee000
FS: 0000000000be4880(0000) GS:ffff8880b9f00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000020000304 CR3: 0000000014b50000 CR4: 00000000001506e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400


---
This report is generated by a bot. It may contain errors.
See https://goo.gl/tpsmEJ for more information about syzbot.
syzbot engineers can be reached at syzk...@googlegroups.com.

syzbot will keep track of this issue. See:
https://goo.gl/tpsmEJ#status for how to communicate with syzbot.
For information about bisection process see: https://goo.gl/tpsmEJ#bisection
syzbot can test patches for this issue, for details see:
https://goo.gl/tpsmEJ#testing-patches

[Hillf Danton]

Wed, 13 Jan 2021 02:37:15 -0800

Fix d9d05217cb69 ("io_uring: stop SQPOLL submit on creator's death")
by setting ctx->sqo_dead as part of cleanup in case of failing to
create io uring context, to quiesce the warning added in the commit.


--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -9700,15 +9700,14 @@ static int io_uring_create(unsigned entr
*/
ret = io_uring_install_fd(ctx, file);
if (ret < 0) {
- /* fput will clean it up */
fput(file);
- return ret;
+ ctx->sqo_dead = 1;
}

trace_io_uring_create(ret, ctx, p->sq_entries, p->cq_entries, p->flags);
return ret;
err:
- io_disable_sqo_submit(ctx);
+ ctx->sqo_dead = 1;
io_ring_ctx_wait_and_kill(ctx);
return ret;
}

[Pavel Begunkov]

Thanks, but it was fixed the day before

--
Pavel Begunkov

[Hillf Danton]

Pavel,

It helps much if you can add a link to the fix next time.

Apart from that, I do not think it is a complete fix yet - it only
fixes what Reported-by: syzbot+9c9c35...@syzkaller.appspotmail.com
though correct, but the one-line fix is unable to cover this report,
as per the Call Trace in both reports.

Feel free to double check if what you trimmed fixes both reports.

Hillf

[Pavel Begunkov]

On 15/01/2021 09:33, Hillf Danton wrote:
>> Thanks, but it was fixed the day before
>>
> It helps much if you can add a link to the fix next time.

https://git.kernel.dk/cgit/linux-block/commit/?h=io_uring-5.11&id=b4411616c26f26c4017b8fa4d3538b1a02028733
https://git.kernel.dk/cgit/linux-block/commit/?h=io_uring-5.11&id=06585c497b55045ec21aa8128e340f6a6587351c

sure, for this report and the other report

>
> Apart from that, I do not think it is a complete fix yet - it only
> fixes what Reported-by: syzbot+9c9c35...@syzkaller.appspotmail.com
> though correct, but the one-line fix is unable to cover this report,
> as per the Call Trace in both reports.
>
> Feel free to double check if what you trimmed fixes both reports.

I believe they do (when considered together).

--
Pavel Begunkov