Hello,
syzbot found the following issue on:
HEAD commit: 93717cde744f Add linux-next specific files for 20210830
git tree: linux-next
console output: https://syzkaller.appspot.com/x/log.txt?x=15b851fe300000
kernel config: https://syzkaller.appspot.com/x/.config?x=c643ef5289990dd1
dashboard link: https://syzkaller.appspot.com/bug?extid=f9704d1878e290eddf73
compiler: gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.1
Unfortunately, I don't have any reproducer for this issue yet.
IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: [email protected]
general protection fault, probably for non-canonical address 0xdffffc0000000005: 0000 [#1] PREEMPT SMP KASAN
KASAN: null-ptr-deref in range [0x0000000000000028-0x000000000000002f]
CPU: 1 PID: 6072 Comm: syz-executor.0 Not tainted 5.14.0-next-20210830-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
RIP: 0010:sock_from_file+0x20/0x90 net/socket.c:505
Code: f5 ff ff ff c3 0f 1f 44 00 00 41 54 53 48 89 fb e8 85 e9 62 fa 48 8d 7b 28 48 b8 00 00 00 00 00 fc ff df 48 89 fa 48 c1 ea 03 <80> 3c 02 00 75 4f 45 31 e4 48 81 7b 28 80 f1 8a 8a 74 0c e8 58 e9
RSP: 0018:ffffc9000a2df8e8 EFLAGS: 00010206
RAX: dffffc0000000000 RBX: 0000000000000000 RCX: ffffc90002f91000
RDX: 0000000000000005 RSI: ffffffff8713203b RDI: 0000000000000028
RBP: ffff8880983c2c80 R08: ffffffff899aee40 R09: ffffffff81e21978
R10: 0000000000000027 R11: 0000000000000009 R12: dffffc0000000000
R13: 1ffff11013078599 R14: 0000000000000003 R15: ffff8880983c2c80
FS: 00007fe7b0454700(0000) GS:ffff8880b9d00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00005591dffa5180 CR3: 00000000974cb000 CR4: 00000000001506e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
io_sendmsg+0x98/0x640 fs/io_uring.c:4681
io_issue_sqe+0x14de/0x6ba0 fs/io_uring.c:6578
__io_queue_sqe+0x90/0xb50 fs/io_uring.c:6864
io_req_task_submit+0xbf/0x1b0 fs/io_uring.c:2218
tctx_task_work+0x166/0x610 fs/io_uring.c:2143
task_work_run+0xdd/0x1a0 kernel/task_work.c:164
tracehook_notify_signal include/linux/tracehook.h:212 [inline]
handle_signal_work kernel/entry/common.c:146 [inline]
exit_to_user_mode_loop kernel/entry/common.c:172 [inline]
exit_to_user_mode_prepare+0x256/0x290 kernel/entry/common.c:209
__syscall_exit_to_user_mode_work kernel/entry/common.c:291 [inline]
syscall_exit_to_user_mode+0x19/0x60 kernel/entry/common.c:302
do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86
entry_SYSCALL_64_after_hwframe+0x44/0xae
RIP: 0033:0x4665f9
Code: ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 40 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:00007fe7b0454188 EFLAGS: 00000246 ORIG_RAX: 00000000000001aa
RAX: 0000000000001000 RBX: 000000000056bf80 RCX: 00000000004665f9
RDX: 0000000000000000 RSI: 000000000000688c RDI: 0000000000000003
RBP: 00000000004bfcc4 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 000000000056bf80
R13: 00007fffeee6585f R14: 00007fe7b0454300 R15: 0000000000022000
Modules linked in:
---[ end trace 6f9e359dd487b8fa ]---
RIP: 0010:sock_from_file+0x20/0x90 net/socket.c:505
Code: f5 ff ff ff c3 0f 1f 44 00 00 41 54 53 48 89 fb e8 85 e9 62 fa 48 8d 7b 28 48 b8 00 00 00 00 00 fc ff df 48 89 fa 48 c1 ea 03 <80> 3c 02 00 75 4f 45 31 e4 48 81 7b 28 80 f1 8a 8a 74 0c e8 58 e9
RSP: 0018:ffffc9000a2df8e8 EFLAGS: 00010206
RAX: dffffc0000000000 RBX: 0000000000000000 RCX: ffffc90002f91000
RDX: 0000000000000005 RSI: ffffffff8713203b RDI: 0000000000000028
RBP: ffff8880983c2c80 R08: ffffffff899aee40 R09: ffffffff81e21978
R10: 0000000000000027 R11: 0000000000000009 R12: dffffc0000000000
R13: 1ffff11013078599 R14: 0000000000000003 R15: ffff8880983c2c80
FS: 00007fe7b0454700(0000) GS:ffff8880b9d00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007fb81002c710 CR3: 00000000974cb000 CR4: 00000000001506e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
----------------
Code disassembly (best guess), 3 bytes skipped:
0: ff c3 inc %ebx
2: 0f 1f 44 00 00 nopl 0x0(%rax,%rax,1)
7: 41 54 push %r12
9: 53 push %rbx
a: 48 89 fb mov %rdi,%rbx
d: e8 85 e9 62 fa callq 0xfa62e997
12: 48 8d 7b 28 lea 0x28(%rbx),%rdi
16: 48 b8 00 00 00 00 00 movabs $0xdffffc0000000000,%rax
1d: fc ff df
20: 48 89 fa mov %rdi,%rdx
23: 48 c1 ea 03 shr $0x3,%rdx
* 27: 80 3c 02 00 cmpb $0x0,(%rdx,%rax,1) <-- trapping instruction
2b: 75 4f jne 0x7c
2d: 45 31 e4 xor %r12d,%r12d
30: 48 81 7b 28 80 f1 8a cmpq $0xffffffff8a8af180,0x28(%rbx)
37: 8a
38: 74 0c je 0x46
3a: e8 .byte 0xe8
3b: 58 pop %rax
3c: e9 .byte 0xe9
---
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 [email protected].
syzbot will keep track of this issue. See:
https://goo.gl/tpsmEJ#status for how to communicate with syzbot.
syzbot has found a reproducer for the following issue on:
HEAD commit: 93717cde744f Add linux-next specific files for 20210830
git tree: linux-next
console output: https://syzkaller.appspot.com/x/log.txt?x=15200fad300000
kernel config: https://syzkaller.appspot.com/x/.config?x=c643ef5289990dd1
dashboard link: https://syzkaller.appspot.com/bug?extid=f9704d1878e290eddf73
compiler: gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.1
syz repro: https://syzkaller.appspot.com/x/repro.syz?x=111f5f9d300000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=1651a415300000
IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: [email protected]
general protection fault, probably for non-canonical address 0xdffffc0000000005: 0000 [#1] PREEMPT SMP KASAN
KASAN: null-ptr-deref in range [0x0000000000000028-0x000000000000002f]
CPU: 0 PID: 6548 Comm: syz-executor433 Not tainted 5.14.0-next-20210830-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
RIP: 0010:sock_from_file+0x20/0x90 net/socket.c:505
Code: f5 ff ff ff c3 0f 1f 44 00 00 41 54 53 48 89 fb e8 85 e9 62 fa 48 8d 7b 28 48 b8 00 00 00 00 00 fc ff df 48 89 fa 48 c1 ea 03 <80> 3c 02 00 75 4f 45 31 e4 48 81 7b 28 80 f1 8a 8a 74 0c e8 58 e9
RSP: 0018:ffffc90002caf8e8 EFLAGS: 00010206
RAX: dffffc0000000000 RBX: 0000000000000000 RCX: 0000000000000000
RDX: 0000000000000005 RSI: ffffffff8713203b RDI: 0000000000000028
RBP: ffff888019fc0780 R08: ffffffff899aee40 R09: ffffffff81e21978
R10: 0000000000000027 R11: 0000000000000009 R12: dffffc0000000000
R13: 1ffff110033f80f9 R14: 0000000000000003 R15: ffff888019fc0780
FS: 00000000013b5300(0000) GS:ffff8880b9c00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00000000004ae0f0 CR3: 000000001d355000 CR4: 00000000001506f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
io_sendmsg+0x98/0x640 fs/io_uring.c:4681
io_issue_sqe+0x14de/0x6ba0 fs/io_uring.c:6578
__io_queue_sqe+0x90/0xb50 fs/io_uring.c:6864
io_req_task_submit+0xbf/0x1b0 fs/io_uring.c:2218
tctx_task_work+0x166/0x610 fs/io_uring.c:2143
task_work_run+0xdd/0x1a0 kernel/task_work.c:164
tracehook_notify_signal include/linux/tracehook.h:212 [inline]
handle_signal_work kernel/entry/common.c:146 [inline]
exit_to_user_mode_loop kernel/entry/common.c:172 [inline]
exit_to_user_mode_prepare+0x256/0x290 kernel/entry/common.c:209
__syscall_exit_to_user_mode_work kernel/entry/common.c:291 [inline]
syscall_exit_to_user_mode+0x19/0x60 kernel/entry/common.c:302
do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86
entry_SYSCALL_64_after_hwframe+0x44/0xae
RIP: 0033:0x43fd49
Code: 28 00 00 00 75 05 48 83 c4 28 c3 e8 b1 14 00 00 90 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 c0 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007ffd6347b9d8 EFLAGS: 00000246 ORIG_RAX: 00000000000001aa
RAX: 0000000000001000 RBX: 0000000000000003 RCX: 000000000043fd49
RDX: 0000000000000000 RSI: 000000000000688c RDI: 0000000000000003
RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 00000000004035d0
R13: 431bde82d7b634db R14: 00000000004ae018 R15: 0000000000400488
Modules linked in:
---[ end trace aa9bf60339277d03 ]---
RIP: 0010:sock_from_file+0x20/0x90 net/socket.c:505
Code: f5 ff ff ff c3 0f 1f 44 00 00 41 54 53 48 89 fb e8 85 e9 62 fa 48 8d 7b 28 48 b8 00 00 00 00 00 fc ff df 48 89 fa 48 c1 ea 03 <80> 3c 02 00 75 4f 45 31 e4 48 81 7b 28 80 f1 8a 8a 74 0c e8 58 e9
RSP: 0018:ffffc90002caf8e8 EFLAGS: 00010206
RAX: dffffc0000000000 RBX: 0000000000000000 RCX: 0000000000000000
RDX: 0000000000000005 RSI: ffffffff8713203b RDI: 0000000000000028
RBP: ffff888019fc0780 R08: ffffffff899aee40 R09: ffffffff81e21978
R10: 0000000000000027 R11: 0000000000000009 R12: dffffc0000000000
R13: 1ffff110033f80f9 R14: 0000000000000003 R15: ffff888019fc0780
FS: 00000000013b5300(0000) GS:ffff8880b9d00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f2cc6f84000 CR3: 000000001d355000 CR4: 00000000001506e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
----------------
Code disassembly (best guess), 3 bytes skipped:
0: ff c3 inc %ebx
2: 0f 1f 44 00 00 nopl 0x0(%rax,%rax,1)
7: 41 54 push %r12
9: 53 push %rbx
a: 48 89 fb mov %rdi,%rbx
d: e8 85 e9 62 fa callq 0xfa62e997
12: 48 8d 7b 28 lea 0x28(%rbx),%rdi
16: 48 b8 00 00 00 00 00 movabs $0xdffffc0000000000,%rax
1d: fc ff df
20: 48 89 fa mov %rdi,%rdx
23: 48 c1 ea 03 shr $0x3,%rdx
* 27: 80 3c 02 00 cmpb $0x0,(%rdx,%rax,1) <-- trapping instruction
2b: 75 4f jne 0x7c
2d: 45 31 e4 xor %r12d,%r12d
30: 48 81 7b 28 80 f1 8a cmpq $0xffffffff8a8af180,0x28(%rbx)
37: 8a
38: 74 0c je 0x46
3a: e8 .byte 0xe8
3b: 58 pop %rax
3c: e9 .byte 0xe9
On 8/30/21 2:45 PM, syzbot wrote:
> syzbot has found a reproducer for the following issue on:
>
> HEAD commit: 93717cde744f Add linux-next specific files for 20210830
> git tree: linux-next
> console output: https://syzkaller.appspot.com/x/log.txt?x=15200fad300000
> kernel config: https://syzkaller.appspot.com/x/.config?x=c643ef5289990dd1
> dashboard link: https://syzkaller.appspot.com/bug?extid=f9704d1878e290eddf73
> compiler: gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.1
> syz repro: https://syzkaller.appspot.com/x/repro.syz?x=111f5f9d300000
> C reproducer: https://syzkaller.appspot.com/x/repro.c?x=1651a415300000
>
> IMPORTANT: if you fix the issue, please add the following tag to the commit:
> Reported-by: [email protected]
>
> general protection fault, probably for non-canonical address 0xdffffc0000000005: 0000 [#1] PREEMPT SMP KASAN
> KASAN: null-ptr-deref in range [0x0000000000000028-0x000000000000002f]
> CPU: 0 PID: 6548 Comm: syz-executor433 Not tainted 5.14.0-next-20210830-syzkaller #0
> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
> RIP: 0010:sock_from_file+0x20/0x90 net/socket.c:505
> Code: f5 ff ff ff c3 0f 1f 44 00 00 41 54 53 48 89 fb e8 85 e9 62 fa 48 8d 7b 28 48 b8 00 00 00 00 00 fc ff df 48 89 fa 48 c1 ea 03 <80> 3c 02 00 75 4f 45 31 e4 48 81 7b 28 80 f1 8a 8a 74 0c e8 58 e9
> RSP: 0018:ffffc90002caf8e8 EFLAGS: 00010206
> RAX: dffffc0000000000 RBX: 0000000000000000 RCX: 0000000000000000
> RDX: 0000000000000005 RSI: ffffffff8713203b RDI: 0000000000000028
> RBP: ffff888019fc0780 R08: ffffffff899aee40 R09: ffffffff81e21978
> R10: 0000000000000027 R11: 0000000000000009 R12: dffffc0000000000
> R13: 1ffff110033f80f9 R14: 0000000000000003 R15: ffff888019fc0780
> FS: 00000000013b5300(0000) GS:ffff8880b9c00000(0000) knlGS:0000000000000000
> CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> CR2: 00000000004ae0f0 CR3: 000000001d355000 CR4: 00000000001506f0
> DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
> Call Trace:
> io_sendmsg+0x98/0x640 fs/io_uring.c:4681
> io_issue_sqe+0x14de/0x6ba0 fs/io_uring.c:6578
> __io_queue_sqe+0x90/0xb50 fs/io_uring.c:6864
> io_req_task_submit+0xbf/0x1b0 fs/io_uring.c:2218
> tctx_task_work+0x166/0x610 fs/io_uring.c:2143
> task_work_run+0xdd/0x1a0 kernel/task_work.c:164
> tracehook_notify_signal include/linux/tracehook.h:212 [inline]
> handle_signal_work kernel/entry/common.c:146 [inline]
> exit_to_user_mode_loop kernel/entry/common.c:172 [inline]
> exit_to_user_mode_prepare+0x256/0x290 kernel/entry/common.c:209
> __syscall_exit_to_user_mode_work kernel/entry/common.c:291 [inline]
> syscall_exit_to_user_mode+0x19/0x60 kernel/entry/common.c:302
> do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86
> entry_SYSCALL_64_after_hwframe+0x44/0xae
> RIP: 0033:0x43fd49
Hao, this is due to:
commit a8295b982c46d4a7c259a4cdd58a2681929068a9
Author: Hao Xu <[email protected]>
Date: Fri Aug 27 17:46:09 2021 +0800
io_uring: fix failed linkchain code logic
which causes some weirdly super long chains from that single sqe.
Can you take a look, please?
--
Jens Axboe
在 2021/8/31 上午10:14, Jens Axboe 写道:
> On 8/30/21 2:45 PM, syzbot wrote:
>> syzbot has found a reproducer for the following issue on:
>>
>> HEAD commit: 93717cde744f Add linux-next specific files for 20210830
>> git tree: linux-next
>> console output: https://syzkaller.appspot.com/x/log.txt?x=15200fad300000
>> kernel config: https://syzkaller.appspot.com/x/.config?x=c643ef5289990dd1
>> dashboard link: https://syzkaller.appspot.com/bug?extid=f9704d1878e290eddf73
>> compiler: gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.1
>> syz repro: https://syzkaller.appspot.com/x/repro.syz?x=111f5f9d300000
>> C reproducer: https://syzkaller.appspot.com/x/repro.c?x=1651a415300000
>>
>> IMPORTANT: if you fix the issue, please add the following tag to the commit:
>> Reported-by: [email protected]
>>
>> general protection fault, probably for non-canonical address 0xdffffc0000000005: 0000 [#1] PREEMPT SMP KASAN
>> KASAN: null-ptr-deref in range [0x0000000000000028-0x000000000000002f]
>> CPU: 0 PID: 6548 Comm: syz-executor433 Not tainted 5.14.0-next-20210830-syzkaller #0
>> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
>> RIP: 0010:sock_from_file+0x20/0x90 net/socket.c:505
>> Code: f5 ff ff ff c3 0f 1f 44 00 00 41 54 53 48 89 fb e8 85 e9 62 fa 48 8d 7b 28 48 b8 00 00 00 00 00 fc ff df 48 89 fa 48 c1 ea 03 <80> 3c 02 00 75 4f 45 31 e4 48 81 7b 28 80 f1 8a 8a 74 0c e8 58 e9
>> RSP: 0018:ffffc90002caf8e8 EFLAGS: 00010206
>> RAX: dffffc0000000000 RBX: 0000000000000000 RCX: 0000000000000000
>> RDX: 0000000000000005 RSI: ffffffff8713203b RDI: 0000000000000028
>> RBP: ffff888019fc0780 R08: ffffffff899aee40 R09: ffffffff81e21978
>> R10: 0000000000000027 R11: 0000000000000009 R12: dffffc0000000000
>> R13: 1ffff110033f80f9 R14: 0000000000000003 R15: ffff888019fc0780
>> FS: 00000000013b5300(0000) GS:ffff8880b9c00000(0000) knlGS:0000000000000000
>> CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
>> CR2: 00000000004ae0f0 CR3: 000000001d355000 CR4: 00000000001506f0
>> DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
>> DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
>> Call Trace:
>> io_sendmsg+0x98/0x640 fs/io_uring.c:4681
>> io_issue_sqe+0x14de/0x6ba0 fs/io_uring.c:6578
>> __io_queue_sqe+0x90/0xb50 fs/io_uring.c:6864
>> io_req_task_submit+0xbf/0x1b0 fs/io_uring.c:2218
>> tctx_task_work+0x166/0x610 fs/io_uring.c:2143
>> task_work_run+0xdd/0x1a0 kernel/task_work.c:164
>> tracehook_notify_signal include/linux/tracehook.h:212 [inline]
>> handle_signal_work kernel/entry/common.c:146 [inline]
>> exit_to_user_mode_loop kernel/entry/common.c:172 [inline]
>> exit_to_user_mode_prepare+0x256/0x290 kernel/entry/common.c:209
>> __syscall_exit_to_user_mode_work kernel/entry/common.c:291 [inline]
>> syscall_exit_to_user_mode+0x19/0x60 kernel/entry/common.c:302
>> do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86
>> entry_SYSCALL_64_after_hwframe+0x44/0xae
>> RIP: 0033:0x43fd49
>
> Hao, this is due to:
>
> commit a8295b982c46d4a7c259a4cdd58a2681929068a9
> Author: Hao Xu <[email protected]>
> Date: Fri Aug 27 17:46:09 2021 +0800
>
> io_uring: fix failed linkchain code logic
>
> which causes some weirdly super long chains from that single sqe.
> Can you take a look, please?
Sure, I'm working on this.
>
On 8/31/21 10:19 AM, Hao Xu wrote:
> 在 2021/8/31 上午10:14, Jens Axboe 写道:
>> On 8/30/21 2:45 PM, syzbot wrote:
>>> syzbot has found a reproducer for the following issue on:
>>>
>>> HEAD commit: 93717cde744f Add linux-next specific files for 20210830
>>> git tree: linux-next
>>> console output: https://syzkaller.appspot.com/x/log.txt?x=15200fad300000
>>> kernel config: https://syzkaller.appspot.com/x/.config?x=c643ef5289990dd1
>>> dashboard link: https://syzkaller.appspot.com/bug?extid=f9704d1878e290eddf73
>>> compiler: gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.1
>>> syz repro: https://syzkaller.appspot.com/x/repro.syz?x=111f5f9d300000
>>> C reproducer: https://syzkaller.appspot.com/x/repro.c?x=1651a415300000
>>>
>>> IMPORTANT: if you fix the issue, please add the following tag to the commit:
>>> Reported-by: [email protected]
>>>
>>> general protection fault, probably for non-canonical address 0xdffffc0000000005: 0000 [#1] PREEMPT SMP KASAN
>>> KASAN: null-ptr-deref in range [0x0000000000000028-0x000000000000002f]
>>> CPU: 0 PID: 6548 Comm: syz-executor433 Not tainted 5.14.0-next-20210830-syzkaller #0
>>> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
>>> RIP: 0010:sock_from_file+0x20/0x90 net/socket.c:505
>>> Code: f5 ff ff ff c3 0f 1f 44 00 00 41 54 53 48 89 fb e8 85 e9 62 fa 48 8d 7b 28 48 b8 00 00 00 00 00 fc ff df 48 89 fa 48 c1 ea 03 <80> 3c 02 00 75 4f 45 31 e4 48 81 7b 28 80 f1 8a 8a 74 0c e8 58 e9
>>> RSP: 0018:ffffc90002caf8e8 EFLAGS: 00010206
>>> RAX: dffffc0000000000 RBX: 0000000000000000 RCX: 0000000000000000
>>> RDX: 0000000000000005 RSI: ffffffff8713203b RDI: 0000000000000028
>>> RBP: ffff888019fc0780 R08: ffffffff899aee40 R09: ffffffff81e21978
>>> R10: 0000000000000027 R11: 0000000000000009 R12: dffffc0000000000
>>> R13: 1ffff110033f80f9 R14: 0000000000000003 R15: ffff888019fc0780
>>> FS: 00000000013b5300(0000) GS:ffff8880b9c00000(0000) knlGS:0000000000000000
>>> CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
>>> CR2: 00000000004ae0f0 CR3: 000000001d355000 CR4: 00000000001506f0
>>> DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
>>> DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
>>> Call Trace:
>>> io_sendmsg+0x98/0x640 fs/io_uring.c:4681
>>> io_issue_sqe+0x14de/0x6ba0 fs/io_uring.c:6578
>>> __io_queue_sqe+0x90/0xb50 fs/io_uring.c:6864
>>> io_req_task_submit+0xbf/0x1b0 fs/io_uring.c:2218
>>> tctx_task_work+0x166/0x610 fs/io_uring.c:2143
>>> task_work_run+0xdd/0x1a0 kernel/task_work.c:164
>>> tracehook_notify_signal include/linux/tracehook.h:212 [inline]
>>> handle_signal_work kernel/entry/common.c:146 [inline]
>>> exit_to_user_mode_loop kernel/entry/common.c:172 [inline]
>>> exit_to_user_mode_prepare+0x256/0x290 kernel/entry/common.c:209
>>> __syscall_exit_to_user_mode_work kernel/entry/common.c:291 [inline]
>>> syscall_exit_to_user_mode+0x19/0x60 kernel/entry/common.c:302
>>> do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86
>>> entry_SYSCALL_64_after_hwframe+0x44/0xae
>>> RIP: 0033:0x43fd49
>>
>> Hao, this is due to:
>>
>> commit a8295b982c46d4a7c259a4cdd58a2681929068a9
>> Author: Hao Xu <[email protected]>
>> Date: Fri Aug 27 17:46:09 2021 +0800
>>
>> io_uring: fix failed linkchain code logic
>>
>> which causes some weirdly super long chains from that single sqe.
>> Can you take a look, please?
> Sure, I'm working on this.
Ah, saw it after sending a patch. It's nothing too curious, just
a small error in logic. More interesting that we don't have a
test case covering it, we should definitely add something.
--
Pavel Begunkov
在 2021/8/31 下午5:42, Pavel Begunkov 写道:
> On 8/31/21 10:19 AM, Hao Xu wrote:
>> 在 2021/8/31 上午10:14, Jens Axboe 写道:
>>> On 8/30/21 2:45 PM, syzbot wrote:
>>>> syzbot has found a reproducer for the following issue on:
>>>>
>>>> HEAD commit: 93717cde744f Add linux-next specific files for 20210830
>>>> git tree: linux-next
>>>> console output: https://syzkaller.appspot.com/x/log.txt?x=15200fad300000
>>>> kernel config: https://syzkaller.appspot.com/x/.config?x=c643ef5289990dd1
>>>> dashboard link: https://syzkaller.appspot.com/bug?extid=f9704d1878e290eddf73
>>>> compiler: gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.1
>>>> syz repro: https://syzkaller.appspot.com/x/repro.syz?x=111f5f9d300000
>>>> C reproducer: https://syzkaller.appspot.com/x/repro.c?x=1651a415300000
>>>>
>>>> IMPORTANT: if you fix the issue, please add the following tag to the commit:
>>>> Reported-by: [email protected]
>>>>
>>>> general protection fault, probably for non-canonical address 0xdffffc0000000005: 0000 [#1] PREEMPT SMP KASAN
>>>> KASAN: null-ptr-deref in range [0x0000000000000028-0x000000000000002f]
>>>> CPU: 0 PID: 6548 Comm: syz-executor433 Not tainted 5.14.0-next-20210830-syzkaller #0
>>>> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
>>>> RIP: 0010:sock_from_file+0x20/0x90 net/socket.c:505
>>>> Code: f5 ff ff ff c3 0f 1f 44 00 00 41 54 53 48 89 fb e8 85 e9 62 fa 48 8d 7b 28 48 b8 00 00 00 00 00 fc ff df 48 89 fa 48 c1 ea 03 <80> 3c 02 00 75 4f 45 31 e4 48 81 7b 28 80 f1 8a 8a 74 0c e8 58 e9
>>>> RSP: 0018:ffffc90002caf8e8 EFLAGS: 00010206
>>>> RAX: dffffc0000000000 RBX: 0000000000000000 RCX: 0000000000000000
>>>> RDX: 0000000000000005 RSI: ffffffff8713203b RDI: 0000000000000028
>>>> RBP: ffff888019fc0780 R08: ffffffff899aee40 R09: ffffffff81e21978
>>>> R10: 0000000000000027 R11: 0000000000000009 R12: dffffc0000000000
>>>> R13: 1ffff110033f80f9 R14: 0000000000000003 R15: ffff888019fc0780
>>>> FS: 00000000013b5300(0000) GS:ffff8880b9c00000(0000) knlGS:0000000000000000
>>>> CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
>>>> CR2: 00000000004ae0f0 CR3: 000000001d355000 CR4: 00000000001506f0
>>>> DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
>>>> DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
>>>> Call Trace:
>>>> io_sendmsg+0x98/0x640 fs/io_uring.c:4681
>>>> io_issue_sqe+0x14de/0x6ba0 fs/io_uring.c:6578
>>>> __io_queue_sqe+0x90/0xb50 fs/io_uring.c:6864
>>>> io_req_task_submit+0xbf/0x1b0 fs/io_uring.c:2218
>>>> tctx_task_work+0x166/0x610 fs/io_uring.c:2143
>>>> task_work_run+0xdd/0x1a0 kernel/task_work.c:164
>>>> tracehook_notify_signal include/linux/tracehook.h:212 [inline]
>>>> handle_signal_work kernel/entry/common.c:146 [inline]
>>>> exit_to_user_mode_loop kernel/entry/common.c:172 [inline]
>>>> exit_to_user_mode_prepare+0x256/0x290 kernel/entry/common.c:209
>>>> __syscall_exit_to_user_mode_work kernel/entry/common.c:291 [inline]
>>>> syscall_exit_to_user_mode+0x19/0x60 kernel/entry/common.c:302
>>>> do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86
>>>> entry_SYSCALL_64_after_hwframe+0x44/0xae
>>>> RIP: 0033:0x43fd49
>>>
>>> Hao, this is due to:
>>>
>>> commit a8295b982c46d4a7c259a4cdd58a2681929068a9
>>> Author: Hao Xu <[email protected]>
>>> Date: Fri Aug 27 17:46:09 2021 +0800
>>>
>>> io_uring: fix failed linkchain code logic
>>>
>>> which causes some weirdly super long chains from that single sqe.
>>> Can you take a look, please?
>> Sure, I'm working on this.
>
> Ah, saw it after sending a patch. It's nothing too curious, just
> a small error in logic. More interesting that we don't have a
> test case covering it, we should definitely add something.
>
Saw your patch after coding my fix..????
Since my email client doesn't receive your patch(only saw it in
webpage https://lore.kernel.org/), I put my comment here:
> fs/io_uring.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/fs/io_uring.c b/fs/io_uring.c
> index 473a977c7979..a531c7324ea8 100644
> --- a/fs/io_uring.c
> +++ b/fs/io_uring.c
> @@ -6717,6 +6717,8 @@ static inline void io_queue_sqe(struct io_kiocb *req)
> if (likely(!(req->flags & (REQ_F_FORCE_ASYNC | REQ_F_FAIL)))) {
> __io_queue_sqe(req);
> } else if (req->flags & REQ_F_FAIL) {
> + /* fail all, we don't submit */
> + req->flags &= ~REQ_F_HARDLINK;
maybe set REQ_F_LINK here?
> io_req_complete_failed(req, req->result);
> } else {
> int ret = io_req_prep_async(req);
> --
On 8/31/21 12:05 PM, Hao Xu wrote:
> 在 2021/8/31 下午5:42, Pavel Begunkov 写道:
>> On 8/31/21 10:19 AM, Hao Xu wrote:
>>> 在 2021/8/31 上午10:14, Jens Axboe 写道:
>>>> On 8/30/21 2:45 PM, syzbot wrote:
>>>>> syzbot has found a reproducer for the following issue on:
>>>>>
>>>>> HEAD commit: 93717cde744f Add linux-next specific files for 20210830
>>>>> git tree: linux-next
>>>>> console output: https://syzkaller.appspot.com/x/log.txt?x=15200fad300000
>>>>> kernel config: https://syzkaller.appspot.com/x/.config?x=c643ef5289990dd1
>>>>> dashboard link: https://syzkaller.appspot.com/bug?extid=f9704d1878e290eddf73
>>>>> compiler: gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.1
>>>>> syz repro: https://syzkaller.appspot.com/x/repro.syz?x=111f5f9d300000
>>>>> C reproducer: https://syzkaller.appspot.com/x/repro.c?x=1651a415300000
>>>>>
>>>>> IMPORTANT: if you fix the issue, please add the following tag to the commit:
>>>>> Reported-by: [email protected]
>>>>>
>>>>> general protection fault, probably for non-canonical address 0xdffffc0000000005: 0000 [#1] PREEMPT SMP KASAN
>>>>> KASAN: null-ptr-deref in range [0x0000000000000028-0x000000000000002f]
>>>>> CPU: 0 PID: 6548 Comm: syz-executor433 Not tainted 5.14.0-next-20210830-syzkaller #0
>>>>> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
>>>>> RIP: 0010:sock_from_file+0x20/0x90 net/socket.c:505
>>>>> Code: f5 ff ff ff c3 0f 1f 44 00 00 41 54 53 48 89 fb e8 85 e9 62 fa 48 8d 7b 28 48 b8 00 00 00 00 00 fc ff df 48 89 fa 48 c1 ea 03 <80> 3c 02 00 75 4f 45 31 e4 48 81 7b 28 80 f1 8a 8a 74 0c e8 58 e9
>>>>> RSP: 0018:ffffc90002caf8e8 EFLAGS: 00010206
>>>>> RAX: dffffc0000000000 RBX: 0000000000000000 RCX: 0000000000000000
>>>>> RDX: 0000000000000005 RSI: ffffffff8713203b RDI: 0000000000000028
>>>>> RBP: ffff888019fc0780 R08: ffffffff899aee40 R09: ffffffff81e21978
>>>>> R10: 0000000000000027 R11: 0000000000000009 R12: dffffc0000000000
>>>>> R13: 1ffff110033f80f9 R14: 0000000000000003 R15: ffff888019fc0780
>>>>> FS: 00000000013b5300(0000) GS:ffff8880b9c00000(0000) knlGS:0000000000000000
>>>>> CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
>>>>> CR2: 00000000004ae0f0 CR3: 000000001d355000 CR4: 00000000001506f0
>>>>> DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
>>>>> DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
>>>>> Call Trace:
>>>>> io_sendmsg+0x98/0x640 fs/io_uring.c:4681
>>>>> io_issue_sqe+0x14de/0x6ba0 fs/io_uring.c:6578
>>>>> __io_queue_sqe+0x90/0xb50 fs/io_uring.c:6864
>>>>> io_req_task_submit+0xbf/0x1b0 fs/io_uring.c:2218
>>>>> tctx_task_work+0x166/0x610 fs/io_uring.c:2143
>>>>> task_work_run+0xdd/0x1a0 kernel/task_work.c:164
>>>>> tracehook_notify_signal include/linux/tracehook.h:212 [inline]
>>>>> handle_signal_work kernel/entry/common.c:146 [inline]
>>>>> exit_to_user_mode_loop kernel/entry/common.c:172 [inline]
>>>>> exit_to_user_mode_prepare+0x256/0x290 kernel/entry/common.c:209
>>>>> __syscall_exit_to_user_mode_work kernel/entry/common.c:291 [inline]
>>>>> syscall_exit_to_user_mode+0x19/0x60 kernel/entry/common.c:302
>>>>> do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86
>>>>> entry_SYSCALL_64_after_hwframe+0x44/0xae
>>>>> RIP: 0033:0x43fd49
>>>>
>>>> Hao, this is due to:
>>>>
>>>> commit a8295b982c46d4a7c259a4cdd58a2681929068a9
>>>> Author: Hao Xu <[email protected]>
>>>> Date: Fri Aug 27 17:46:09 2021 +0800
>>>>
>>>> io_uring: fix failed linkchain code logic
>>>>
>>>> which causes some weirdly super long chains from that single sqe.
>>>> Can you take a look, please?
>>> Sure, I'm working on this.
>>
>> Ah, saw it after sending a patch. It's nothing too curious, just
>> a small error in logic. More interesting that we don't have a
>> test case covering it, we should definitely add something.
>>
> Saw your patch after coding my fix..????
> Since my email client doesn't receive your patch(only saw it in
> webpage https://lore.kernel.org/), I put my comment here:
Hmm, does it happen often? I'll CC you
>> fs/io_uring.c | 2 ++
>> 1 file changed, 2 insertions(+)
>>
>> diff --git a/fs/io_uring.c b/fs/io_uring.c
>> index 473a977c7979..a531c7324ea8 100644
>> --- a/fs/io_uring.c
>> +++ b/fs/io_uring.c
>> @@ -6717,6 +6717,8 @@ static inline void io_queue_sqe(struct io_kiocb *req)
>> if (likely(!(req->flags & (REQ_F_FORCE_ASYNC | REQ_F_FAIL)))) {
>> __io_queue_sqe(req);
>> } else if (req->flags & REQ_F_FAIL) {
>> + /* fail all, we don't submit */
>> + req->flags &= ~REQ_F_HARDLINK;
> maybe set REQ_F_LINK here?
if (unlikely((req->flags & REQ_F_FAIL) &&
!(req->flags & REQ_F_HARDLINK))) {
posted |= (req->link != NULL);
io_fail_links(req);
}
The problem is hardlink, normal will be failed. But there is indeed
a problem with both patches,
if (req->flags & (REQ_F_LINK | REQ_F_HARDLINK))
// kill linked
Will resend with some tests on top
>> io_req_complete_failed(req, req->result);
>> } else {
>> int ret = io_req_prep_async(req);
--
Pavel Begunkov
在 2021/8/31 下午7:26, Pavel Begunkov 写道:
> On 8/31/21 12:05 PM, Hao Xu wrote:
>> 在 2021/8/31 下午5:42, Pavel Begunkov 写道:
>>> On 8/31/21 10:19 AM, Hao Xu wrote:
>>>> 在 2021/8/31 上午10:14, Jens Axboe 写道:
>>>>> On 8/30/21 2:45 PM, syzbot wrote:
>>>>>> syzbot has found a reproducer for the following issue on:
>>>>>>
>>>>>> HEAD commit: 93717cde744f Add linux-next specific files for 20210830
>>>>>> git tree: linux-next
>>>>>> console output: https://syzkaller.appspot.com/x/log.txt?x=15200fad300000
>>>>>> kernel config: https://syzkaller.appspot.com/x/.config?x=c643ef5289990dd1
>>>>>> dashboard link: https://syzkaller.appspot.com/bug?extid=f9704d1878e290eddf73
>>>>>> compiler: gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.1
>>>>>> syz repro: https://syzkaller.appspot.com/x/repro.syz?x=111f5f9d300000
>>>>>> C reproducer: https://syzkaller.appspot.com/x/repro.c?x=1651a415300000
>>>>>>
>>>>>> IMPORTANT: if you fix the issue, please add the following tag to the commit:
>>>>>> Reported-by: [email protected]
>>>>>>
>>>>>> general protection fault, probably for non-canonical address 0xdffffc0000000005: 0000 [#1] PREEMPT SMP KASAN
>>>>>> KASAN: null-ptr-deref in range [0x0000000000000028-0x000000000000002f]
>>>>>> CPU: 0 PID: 6548 Comm: syz-executor433 Not tainted 5.14.0-next-20210830-syzkaller #0
>>>>>> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
>>>>>> RIP: 0010:sock_from_file+0x20/0x90 net/socket.c:505
>>>>>> Code: f5 ff ff ff c3 0f 1f 44 00 00 41 54 53 48 89 fb e8 85 e9 62 fa 48 8d 7b 28 48 b8 00 00 00 00 00 fc ff df 48 89 fa 48 c1 ea 03 <80> 3c 02 00 75 4f 45 31 e4 48 81 7b 28 80 f1 8a 8a 74 0c e8 58 e9
>>>>>> RSP: 0018:ffffc90002caf8e8 EFLAGS: 00010206
>>>>>> RAX: dffffc0000000000 RBX: 0000000000000000 RCX: 0000000000000000
>>>>>> RDX: 0000000000000005 RSI: ffffffff8713203b RDI: 0000000000000028
>>>>>> RBP: ffff888019fc0780 R08: ffffffff899aee40 R09: ffffffff81e21978
>>>>>> R10: 0000000000000027 R11: 0000000000000009 R12: dffffc0000000000
>>>>>> R13: 1ffff110033f80f9 R14: 0000000000000003 R15: ffff888019fc0780
>>>>>> FS: 00000000013b5300(0000) GS:ffff8880b9c00000(0000) knlGS:0000000000000000
>>>>>> CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
>>>>>> CR2: 00000000004ae0f0 CR3: 000000001d355000 CR4: 00000000001506f0
>>>>>> DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
>>>>>> DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
>>>>>> Call Trace:
>>>>>> io_sendmsg+0x98/0x640 fs/io_uring.c:4681
>>>>>> io_issue_sqe+0x14de/0x6ba0 fs/io_uring.c:6578
>>>>>> __io_queue_sqe+0x90/0xb50 fs/io_uring.c:6864
>>>>>> io_req_task_submit+0xbf/0x1b0 fs/io_uring.c:2218
>>>>>> tctx_task_work+0x166/0x610 fs/io_uring.c:2143
>>>>>> task_work_run+0xdd/0x1a0 kernel/task_work.c:164
>>>>>> tracehook_notify_signal include/linux/tracehook.h:212 [inline]
>>>>>> handle_signal_work kernel/entry/common.c:146 [inline]
>>>>>> exit_to_user_mode_loop kernel/entry/common.c:172 [inline]
>>>>>> exit_to_user_mode_prepare+0x256/0x290 kernel/entry/common.c:209
>>>>>> __syscall_exit_to_user_mode_work kernel/entry/common.c:291 [inline]
>>>>>> syscall_exit_to_user_mode+0x19/0x60 kernel/entry/common.c:302
>>>>>> do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86
>>>>>> entry_SYSCALL_64_after_hwframe+0x44/0xae
>>>>>> RIP: 0033:0x43fd49
>>>>>
>>>>> Hao, this is due to:
>>>>>
>>>>> commit a8295b982c46d4a7c259a4cdd58a2681929068a9
>>>>> Author: Hao Xu <[email protected]>
>>>>> Date: Fri Aug 27 17:46:09 2021 +0800
>>>>>
>>>>> io_uring: fix failed linkchain code logic
>>>>>
>>>>> which causes some weirdly super long chains from that single sqe.
>>>>> Can you take a look, please?
>>>> Sure, I'm working on this.
>>>
>>> Ah, saw it after sending a patch. It's nothing too curious, just
>>> a small error in logic. More interesting that we don't have a
>>> test case covering it, we should definitely add something.
>>>
>> Saw your patch after coding my fix..????
>> Since my email client doesn't receive your patch(only saw it in
>> webpage https://lore.kernel.org/), I put my comment here:
>
> Hmm, does it happen often? I'll CC you
Uncommon, somestimes there is delay.
>
>
>>> fs/io_uring.c | 2 ++
>>> 1 file changed, 2 insertions(+)
>>>
>>> diff --git a/fs/io_uring.c b/fs/io_uring.c
>>> index 473a977c7979..a531c7324ea8 100644
>>> --- a/fs/io_uring.c
>>> +++ b/fs/io_uring.c
>>> @@ -6717,6 +6717,8 @@ static inline void io_queue_sqe(struct io_kiocb *req)
>>> if (likely(!(req->flags & (REQ_F_FORCE_ASYNC | REQ_F_FAIL)))) {
>>> __io_queue_sqe(req);
>>> } else if (req->flags & REQ_F_FAIL) {
>>> + /* fail all, we don't submit */
>>> + req->flags &= ~REQ_F_HARDLINK;
>> maybe set REQ_F_LINK here?
>
> if (unlikely((req->flags & REQ_F_FAIL) &&
> !(req->flags & REQ_F_HARDLINK))) {
> posted |= (req->link != NULL);
> io_fail_links(req);
> }
>
> The problem is hardlink, normal will be failed. But there is indeed
> a problem with both patches,
>
> if (req->flags & (REQ_F_LINK | REQ_F_HARDLINK))
> // kill linked
Yeah, if we don't have REQ_F_LINK, io_req_complete_post() won't go to
the disarm branch
>
> Will resend with some tests on top
>
>
>>> io_req_complete_failed(req, req->result);
>>> } else {
>>> int ret = io_req_prep_async(req);
>
On 8/31/21 12:48 PM, Hao Xu wrote:
> 在 2021/8/31 下午7:26, Pavel Begunkov 写道:
>> On 8/31/21 12:05 PM, Hao Xu wrote:
>>> 在 2021/8/31 下午5:42, Pavel Begunkov 写道:
>>>> On 8/31/21 10:19 AM, Hao Xu wrote:
>>>>> 在 2021/8/31 上午10:14, Jens Axboe 写道:
>>>>>> On 8/30/21 2:45 PM, syzbot wrote:
>>>>>>> syzbot has found a reproducer for the following issue on:
>>>>>>>
>>>>>>> HEAD commit: 93717cde744f Add linux-next specific files for 20210830
>>>>>>> git tree: linux-next
>>>>>>> console output: https://syzkaller.appspot.com/x/log.txt?x=15200fad300000
>>>>>>> kernel config: https://syzkaller.appspot.com/x/.config?x=c643ef5289990dd1
>>>>>>> dashboard link: https://syzkaller.appspot.com/bug?extid=f9704d1878e290eddf73
>>>>>>> compiler: gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.1
>>>>>>> syz repro: https://syzkaller.appspot.com/x/repro.syz?x=111f5f9d300000
>>>>>>> C reproducer: https://syzkaller.appspot.com/x/repro.c?x=1651a415300000
>>>>>>>
>>>>>>> IMPORTANT: if you fix the issue, please add the following tag to the commit:
>>>>>>> Reported-by: [email protected]
>>>>>>>
>>>>>>> general protection fault, probably for non-canonical address 0xdffffc0000000005: 0000 [#1] PREEMPT SMP KASAN
>>>>>>> KASAN: null-ptr-deref in range [0x0000000000000028-0x000000000000002f]
>>>>>>> CPU: 0 PID: 6548 Comm: syz-executor433 Not tainted 5.14.0-next-20210830-syzkaller #0
>>>>>>> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
>>>>>>> RIP: 0010:sock_from_file+0x20/0x90 net/socket.c:505
>>>>>>> Code: f5 ff ff ff c3 0f 1f 44 00 00 41 54 53 48 89 fb e8 85 e9 62 fa 48 8d 7b 28 48 b8 00 00 00 00 00 fc ff df 48 89 fa 48 c1 ea 03 <80> 3c 02 00 75 4f 45 31 e4 48 81 7b 28 80 f1 8a 8a 74 0c e8 58 e9
>>>>>>> RSP: 0018:ffffc90002caf8e8 EFLAGS: 00010206
>>>>>>> RAX: dffffc0000000000 RBX: 0000000000000000 RCX: 0000000000000000
>>>>>>> RDX: 0000000000000005 RSI: ffffffff8713203b RDI: 0000000000000028
>>>>>>> RBP: ffff888019fc0780 R08: ffffffff899aee40 R09: ffffffff81e21978
>>>>>>> R10: 0000000000000027 R11: 0000000000000009 R12: dffffc0000000000
>>>>>>> R13: 1ffff110033f80f9 R14: 0000000000000003 R15: ffff888019fc0780
>>>>>>> FS: 00000000013b5300(0000) GS:ffff8880b9c00000(0000) knlGS:0000000000000000
>>>>>>> CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
>>>>>>> CR2: 00000000004ae0f0 CR3: 000000001d355000 CR4: 00000000001506f0
>>>>>>> DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
>>>>>>> DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
>>>>>>> Call Trace:
>>>>>>> io_sendmsg+0x98/0x640 fs/io_uring.c:4681
>>>>>>> io_issue_sqe+0x14de/0x6ba0 fs/io_uring.c:6578
>>>>>>> __io_queue_sqe+0x90/0xb50 fs/io_uring.c:6864
>>>>>>> io_req_task_submit+0xbf/0x1b0 fs/io_uring.c:2218
>>>>>>> tctx_task_work+0x166/0x610 fs/io_uring.c:2143
>>>>>>> task_work_run+0xdd/0x1a0 kernel/task_work.c:164
>>>>>>> tracehook_notify_signal include/linux/tracehook.h:212 [inline]
>>>>>>> handle_signal_work kernel/entry/common.c:146 [inline]
>>>>>>> exit_to_user_mode_loop kernel/entry/common.c:172 [inline]
>>>>>>> exit_to_user_mode_prepare+0x256/0x290 kernel/entry/common.c:209
>>>>>>> __syscall_exit_to_user_mode_work kernel/entry/common.c:291 [inline]
>>>>>>> syscall_exit_to_user_mode+0x19/0x60 kernel/entry/common.c:302
>>>>>>> do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86
>>>>>>> entry_SYSCALL_64_after_hwframe+0x44/0xae
>>>>>>> RIP: 0033:0x43fd49
>>>>>>
>>>>>> Hao, this is due to:
>>>>>>
>>>>>> commit a8295b982c46d4a7c259a4cdd58a2681929068a9
>>>>>> Author: Hao Xu <[email protected]>
>>>>>> Date: Fri Aug 27 17:46:09 2021 +0800
>>>>>>
>>>>>> io_uring: fix failed linkchain code logic
>>>>>>
>>>>>> which causes some weirdly super long chains from that single sqe.
>>>>>> Can you take a look, please?
>>>>> Sure, I'm working on this.
>>>>
>>>> Ah, saw it after sending a patch. It's nothing too curious, just
>>>> a small error in logic. More interesting that we don't have a
>>>> test case covering it, we should definitely add something.
>>>>
>>> Saw your patch after coding my fix..????
>>> Since my email client doesn't receive your patch(only saw it in
>>> webpage https://lore.kernel.org/), I put my comment here:
>>
>> Hmm, does it happen often? I'll CC you
> Uncommon, somestimes there is delay.
>>
>>
>>>> fs/io_uring.c | 2 ++
>>>> 1 file changed, 2 insertions(+)
>>>>
>>>> diff --git a/fs/io_uring.c b/fs/io_uring.c
>>>> index 473a977c7979..a531c7324ea8 100644
>>>> --- a/fs/io_uring.c
>>>> +++ b/fs/io_uring.c
>>>> @@ -6717,6 +6717,8 @@ static inline void io_queue_sqe(struct io_kiocb *req)
>>>> if (likely(!(req->flags & (REQ_F_FORCE_ASYNC | REQ_F_FAIL)))) {
>>>> __io_queue_sqe(req);
>>>> } else if (req->flags & REQ_F_FAIL) {
>>>> + /* fail all, we don't submit */
>>>> + req->flags &= ~REQ_F_HARDLINK;
>>> maybe set REQ_F_LINK here?
>>
>> if (unlikely((req->flags & REQ_F_FAIL) &&
>> !(req->flags & REQ_F_HARDLINK))) {
>> posted |= (req->link != NULL);
>> io_fail_links(req);
>> }
>>
>> The problem is hardlink, normal will be failed. But there is indeed
>> a problem with both patches,
>>
>> if (req->flags & (REQ_F_LINK | REQ_F_HARDLINK))
>> // kill linked
> Yeah, if we don't have REQ_F_LINK, io_req_complete_post() won't go to
> the disarm branch
Ah, that's what you meant, right. Good catch!
>>
>> Will resend with some tests on top
>>
>>
>>>> io_req_complete_failed(req, req->result);
>>>> } else {
>>>> int ret = io_req_prep_async(req);
>>
>
--
Pavel Begunkov