Received-SPF: pass (google.com: domain of linux-kernel+bounces-59589-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
Message-ID: <affa6134-de19-4b73-9925-f42d3c7c9b8b@redhat.com>
Date: Fri, 9 Feb 2024 16:10:19 +0000
Precedence: bulk
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: [syzbot] [gfs2?] general protection fault in gfs2_rindex_update
To: syzbot <syzbot+74edb1a3ea8f1c65a086@syzkaller.appspotmail.com>,
 agruenba@redhat.com, gfs2@lists.linux.dev, linux-fsdevel@vger.kernel.org,
 linux-kernel@vger.kernel.org, syzkaller-bugs@googlegroups.com
References: <000000000000c091060610dbd67b@google.com>
Content-Language: en-US
From: Andrew Price <anprice@redhat.com>
In-Reply-To: <000000000000c091060610dbd67b@google.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit

On 08/02/2024 09:57, syzbot wrote:
> syzbot has found a reproducer for the following issue on:
> 
> HEAD commit:    547ab8fc4cb0 Merge tag 'loongarch-fixes-6.8-2' of git://gi..
> git tree:       upstream
> console+strace: https://syzkaller.appspot.com/x/log.txt?x=150285d4180000
> kernel config:  https://syzkaller.appspot.com/x/.config?x=89a5d896b14c4565
> dashboard link: https://syzkaller.appspot.com/bug?extid=74edb1a3ea8f1c65a086
> compiler:       Debian clang version 15.0.6, GNU ld (GNU Binutils for Debian) 2.40
> syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=11395147e80000
> C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=126ff68fe80000
> 
> Downloadable assets:
> disk image: https://storage.googleapis.com/syzbot-assets/a8d318be4c39/disk-547ab8fc.raw.xz
> vmlinux: https://storage.googleapis.com/syzbot-assets/8178462cbfb5/vmlinux-547ab8fc.xz
> kernel image: https://storage.googleapis.com/syzbot-assets/62939e7c5fbb/bzImage-547ab8fc.xz
> mounted in repro: https://storage.googleapis.com/syzbot-assets/93eefe621564/mount_0.gz
> 
> IMPORTANT: if you fix the issue, please add the following tag to the commit:
> Reported-by: syzbot+74edb1a3ea8f1c65a086@syzkaller.appspotmail.com
> 
> loop0: detected capacity change from 0 to 32768
> gfs2: fsid=syz:syz: Trying to join cluster "lock_nolock", "syz:syz"
> gfs2: fsid=syz:syz: Now mounting FS (format 1801)...
> gfs2: fsid=syz:syz.0: journal 0 mapped with 20 extents in 0ms
> general protection fault, probably for non-canonical address 0xdffffc0000000097: 0000 [#1] PREEMPT SMP KASAN PTI
> KASAN: null-ptr-deref in range [0x00000000000004b8-0x00000000000004bf]
> CPU: 0 PID: 5058 Comm: syz-executor337 Not tainted 6.8.0-rc3-syzkaller-00041-g547ab8fc4cb0 #0
> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/25/2024
> RIP: 0010:gfs2_rindex_update+0xbc/0x3d0 fs/gfs2/rgrp.c:1037
> Code: e8 f9 65 1d fe 4c 8d 74 24 60 48 8b 03 48 89 44 24 38 48 8d 98 b8 04 00 00 48 89 d8 48 c1 e8 03 48 b9 00 00 00 00 00 fc ff df <80> 3c 08 00 74 08 48 89 df e8 c6 65 1d fe 48 8b 03 48 89 44 24 20
> RSP: 0018:ffffc900040671a0 EFLAGS: 00010202
> RAX: 0000000000000097 RBX: 00000000000004b8 RCX: dffffc0000000000
> RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff88807f558000
> RBP: ffffc900040672b0 R08: ffffffff83cb50d7 R09: 1ffff1100f02e2f8
> R10: dffffc0000000000 R11: ffffed100f02e2f9 R12: 1ffff9200080ce3c
> R13: ffff88807f558000 R14: ffffc90004067200 R15: 0000000000000001
> FS:  0000555555a0f380(0000) GS:ffff8880b9400000(0000) knlGS:0000000000000000
> CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> CR2: 000056427a53bb08 CR3: 0000000022d54000 CR4: 00000000003506f0
> DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
> Call Trace:
>   <TASK>
>   punch_hole+0xe7b/0x3a30 fs/gfs2/bmap.c:1809
>   gfs2_truncatei_resume+0x3c/0x70 fs/gfs2/bmap.c:2159

The fuzzer is setting GFS2_DIF_TRUNC_IN_PROG on the 
per_node/statfs_change0 'system' inode (it's setting all of the di_flags 
bits but that is the relevant one) which sends us down this resume 
truncation path in inode_go_held(), before the rindex has been 
initialised in init_inodes().

Is there a better solution than returning an error on the 
(sdp->sd_rindex == NULL) case in gfs2_rindex_update()?

Andy

>   gfs2_glock_holder_ready fs/gfs2/glock.c:1336 [inline]
>   gfs2_glock_wait+0x1df/0x2b0 fs/gfs2/glock.c:1356
>   gfs2_glock_nq_init fs/gfs2/glock.h:238 [inline]
>   init_statfs fs/gfs2/ops_fstype.c:694 [inline]
>   init_journal+0x1680/0x23f0 fs/gfs2/ops_fstype.c:816
>   init_inodes+0xdc/0x320 fs/gfs2/ops_fstype.c:884
>   gfs2_fill_super+0x1edb/0x26c0 fs/gfs2/ops_fstype.c:1263
>   get_tree_bdev+0x3f7/0x570 fs/super.c:1619
>   gfs2_get_tree+0x54/0x220 fs/gfs2/ops_fstype.c:1341
>   vfs_get_tree+0x90/0x2a0 fs/super.c:1784
>   do_new_mount+0x2be/0xb40 fs/namespace.c:3352
>   do_mount fs/namespace.c:3692 [inline]
>   __do_sys_mount fs/namespace.c:3898 [inline]
>   __se_sys_mount+0x2d9/0x3c0 fs/namespace.c:3875
>   do_syscall_64+0xf9/0x240
>   entry_SYSCALL_64_after_hwframe+0x6f/0x77
> RIP: 0033:0x7fe98b1e18ba
> Code: d8 64 89 02 48 c7 c0 ff ff ff ff eb a6 e8 5e 04 00 00 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 40 00 49 89 ca b8 a5 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 b8 ff ff ff f7 d8 64 89 01 48
> RSP: 002b:00007ffe3b0f9eb8 EFLAGS: 00000282 ORIG_RAX: 00000000000000a5
> RAX: ffffffffffffffda RBX: 00007ffe3b0f9ec0 RCX: 00007fe98b1e18ba
> RDX: 0000000020000040 RSI: 0000000020000100 RDI: 00007ffe3b0f9ec0
> RBP: 0000000000000004 R08: 00007ffe3b0f9f00 R09: 0000000000012783
> R10: 0000000000008c1b R11: 0000000000000282 R12: 00007ffe3b0f9f00
> R13: 0000000000000003 R14: 0000000001000000 R15: 0000000000000001
>   </TASK>
> Modules linked in:
> ---[ end trace 0000000000000000 ]---
> RIP: 0010:gfs2_rindex_update+0xbc/0x3d0 fs/gfs2/rgrp.c:1037
> Code: e8 f9 65 1d fe 4c 8d 74 24 60 48 8b 03 48 89 44 24 38 48 8d 98 b8 04 00 00 48 89 d8 48 c1 e8 03 48 b9 00 00 00 00 00 fc ff df <80> 3c 08 00 74 08 48 89 df e8 c6 65 1d fe 48 8b 03 48 89 44 24 20
> RSP: 0018:ffffc900040671a0 EFLAGS: 00010202
> RAX: 0000000000000097 RBX: 00000000000004b8 RCX: dffffc0000000000
> RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff88807f558000
> RBP: ffffc900040672b0 R08: ffffffff83cb50d7 R09: 1ffff1100f02e2f8
> R10: dffffc0000000000 R11: ffffed100f02e2f9 R12: 1ffff9200080ce3c
> R13: ffff88807f558000 R14: ffffc90004067200 R15: 0000000000000001
> FS:  0000555555a0f380(0000) GS:ffff8880b9400000(0000) knlGS:0000000000000000
> CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> CR2: 000056427a53bb08 CR3: 0000000022d54000 CR4: 00000000003506f0
> DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
> ----------------
> Code disassembly (best guess):
>     0:	e8 f9 65 1d fe       	call   0xfe1d65fe
>     5:	4c 8d 74 24 60       	lea    0x60(%rsp),%r14
>     a:	48 8b 03             	mov    (%rbx),%rax
>     d:	48 89 44 24 38       	mov    %rax,0x38(%rsp)
>    12:	48 8d 98 b8 04 00 00 	lea    0x4b8(%rax),%rbx
>    19:	48 89 d8             	mov    %rbx,%rax
>    1c:	48 c1 e8 03          	shr    $0x3,%rax
>    20:	48 b9 00 00 00 00 00 	movabs $0xdffffc0000000000,%rcx
>    27:	fc ff df
> * 2a:	80 3c 08 00          	cmpb   $0x0,(%rax,%rcx,1) <-- trapping instruction
>    2e:	74 08                	je     0x38
>    30:	48 89 df             	mov    %rbx,%rdi
>    33:	e8 c6 65 1d fe       	call   0xfe1d65fe
>    38:	48 8b 03             	mov    (%rbx),%rax
>    3b:	48 89 44 24 20       	mov    %rax,0x20(%rsp)
> 
> 
> ---
> If you want syzbot to run the reproducer, reply with:
> #syz test: git://repo/address.git branch-or-commit-hash
> If you attach or paste a git patch, syzbot will apply it before testing.
>