Re: [syzbot] [fat?] possible deadlock in exc_page_fault

From: syzbot
Date: Mon Mar 20 2023 - 07:47:04 EST

Next message: Russell King (Oracle): "Re: [PATCH net-next v14 8/9] net: ethernet: mtk_eth_soc: switch to external PCS driver"
Previous message: James Clark: "[PATCH] perf vendor events arm64: Add N1 metrics"
Next in thread: syzbot: "Re: [syzbot] [fat?] possible deadlock in exc_page_fault"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

syzbot has found a reproducer for the following issue on:

HEAD commit: e8d018dd0257 Linux 6.3-rc3
git tree: upstream
console+strace: https://syzkaller.appspot.com/x/log.txt?x=13c23186c80000
kernel config: https://syzkaller.appspot.com/x/.config?x=d40f6d44826f6cf7
dashboard link: https://syzkaller.appspot.com/bug?extid=6d274a5dc4fa0974d4ad
compiler: Debian clang version 15.0.7, GNU ld (GNU Binutils for Debian) 2.35.2
syz repro: https://syzkaller.appspot.com/x/repro.syz?x=16713fbec80000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=132c7281c80000

Downloadable assets:
disk image: https://storage.googleapis.com/syzbot-assets/36c3f1b07e1e/disk-e8d018dd.raw.xz
vmlinux: https://storage.googleapis.com/syzbot-assets/b45f2ee6f521/vmlinux-e8d018dd.xz
kernel image: https://storage.googleapis.com/syzbot-assets/f03104e87ec4/bzImage-e8d018dd.xz
mounted in repro: https://storage.googleapis.com/syzbot-assets/2573dec3d16b/mount_0.gz

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+6d274a5dc4fa0974d4ad@xxxxxxxxxxxxxxxxxxxxxxxxx

======================================================
WARNING: possible circular locking dependency detected
6.3.0-rc3-syzkaller #0 Not tainted
------------------------------------------------------
syz-executor377/6548 is trying to acquire lock:
ffff8880766bc998 (&mm->mmap_lock){++++}-{3:3}, at: mmap_read_lock include/linux/mmap_lock.h:117 [inline]
ffff8880766bc998 (&mm->mmap_lock){++++}-{3:3}, at: do_user_addr_fault arch/x86/mm/fault.c:1358 [inline]
ffff8880766bc998 (&mm->mmap_lock){++++}-{3:3}, at: handle_page_fault arch/x86/mm/fault.c:1498 [inline]
ffff8880766bc998 (&mm->mmap_lock){++++}-{3:3}, at: exc_page_fault+0x486/0x7c0 arch/x86/mm/fault.c:1554

but task is already holding lock:
ffff88807c2640e0 (&sbi->s_lock){+.+.}-{3:3}, at: exfat_iterate+0x171/0x3370 fs/exfat/dir.c:232

which lock already depends on the new lock.

the existing dependency chain (in reverse order) is:

-> #2 (&sbi->s_lock){+.+.}-{3:3}:
lock_acquire+0x1e1/0x520 kernel/locking/lockdep.c:5669
__mutex_lock_common+0x1d8/0x2530 kernel/locking/mutex.c:603
__mutex_lock kernel/locking/mutex.c:747 [inline]
mutex_lock_nested+0x1b/0x20 kernel/locking/mutex.c:799
exfat_get_block+0x1e5/0x2050 fs/exfat/inode.c:280
do_mpage_readpage+0x911/0x1fa0 fs/mpage.c:208
mpage_readahead+0x454/0x930 fs/mpage.c:356
read_pages+0x183/0x830 mm/readahead.c:161
page_cache_ra_unbounded+0x697/0x7c0 mm/readahead.c:270
page_cache_sync_readahead include/linux/pagemap.h:1214 [inline]
filemap_get_pages+0x49c/0x20c0 mm/filemap.c:2598
filemap_read+0x45a/0x1170 mm/filemap.c:2693
__kernel_read+0x422/0x8a0 fs/read_write.c:428
integrity_kernel_read+0xb0/0xf0 security/integrity/iint.c:199
ima_calc_file_hash_tfm security/integrity/ima/ima_crypto.c:485 [inline]
ima_calc_file_shash security/integrity/ima/ima_crypto.c:516 [inline]
ima_calc_file_hash+0xa5b/0x1c00 security/integrity/ima/ima_crypto.c:573
ima_collect_measurement+0x3a7/0x880 security/integrity/ima/ima_api.c:293
process_measurement+0xfdb/0x1ce0 security/integrity/ima/ima_main.c:341
ima_file_check+0xf1/0x170 security/integrity/ima/ima_main.c:539
do_open fs/namei.c:3562 [inline]
path_openat+0x280a/0x3170 fs/namei.c:3715
do_filp_open+0x234/0x490 fs/namei.c:3742
do_sys_openat2+0x13f/0x500 fs/open.c:1348
do_sys_open fs/open.c:1364 [inline]
__do_sys_openat fs/open.c:1380 [inline]
__se_sys_openat fs/open.c:1375 [inline]
__x64_sys_openat+0x247/0x290 fs/open.c:1375
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x63/0xcd

-> #1 (mapping.invalidate_lock#3){.+.+}-{3:3}:
lock_acquire+0x1e1/0x520 kernel/locking/lockdep.c:5669
down_read+0x3d/0x50 kernel/locking/rwsem.c:1520
filemap_invalidate_lock_shared include/linux/fs.h:813 [inline]
filemap_fault+0x644/0x1800 mm/filemap.c:3274
__do_fault+0x136/0x500 mm/memory.c:4141
do_read_fault mm/memory.c:4492 [inline]
do_fault mm/memory.c:4621 [inline]
handle_pte_fault mm/memory.c:4909 [inline]
__handle_mm_fault mm/memory.c:5051 [inline]
handle_mm_fault+0x3357/0x51c0 mm/memory.c:5197
faultin_page mm/gup.c:925 [inline]
__get_user_pages+0x512/0x1180 mm/gup.c:1147
__get_user_pages_locked mm/gup.c:1381 [inline]
__gup_longterm_locked+0x208c/0x2aa0 mm/gup.c:2079
pin_user_pages_remote+0x136/0x200 mm/gup.c:3122
process_vm_rw_single_vec mm/process_vm_access.c:105 [inline]
process_vm_rw_core mm/process_vm_access.c:215 [inline]
process_vm_rw+0x72b/0xcd0 mm/process_vm_access.c:283
__do_sys_process_vm_readv mm/process_vm_access.c:295 [inline]
__se_sys_process_vm_readv mm/process_vm_access.c:291 [inline]
__x64_sys_process_vm_readv+0xe0/0xf0 mm/process_vm_access.c:291
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x63/0xcd

-> #0 (&mm->mmap_lock){++++}-{3:3}:
check_prev_add kernel/locking/lockdep.c:3098 [inline]
check_prevs_add kernel/locking/lockdep.c:3217 [inline]
validate_chain+0x166b/0x58e0 kernel/locking/lockdep.c:3832
__lock_acquire+0x125b/0x1f80 kernel/locking/lockdep.c:5056
lock_acquire+0x1e1/0x520 kernel/locking/lockdep.c:5669
down_read+0x3d/0x50 kernel/locking/rwsem.c:1520
mmap_read_lock include/linux/mmap_lock.h:117 [inline]
do_user_addr_fault arch/x86/mm/fault.c:1358 [inline]
handle_page_fault arch/x86/mm/fault.c:1498 [inline]
exc_page_fault+0x486/0x7c0 arch/x86/mm/fault.c:1554
asm_exc_page_fault+0x26/0x30 arch/x86/include/asm/idtentry.h:570
filldir64+0x30b/0x720 fs/readdir.c:331
dir_emit_dot include/linux/fs.h:3144 [inline]
dir_emit_dots include/linux/fs.h:3155 [inline]
exfat_iterate+0x2b8/0x3370 fs/exfat/dir.c:235
iterate_dir+0x228/0x570
__do_sys_getdents64 fs/readdir.c:369 [inline]
__se_sys_getdents64+0x20d/0x4f0 fs/readdir.c:354
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x63/0xcd

other info that might help us debug this:

Chain exists of:
&mm->mmap_lock --> mapping.invalidate_lock#3 --> &sbi->s_lock

Possible unsafe locking scenario:

CPU0 CPU1
---- ----
lock(&sbi->s_lock);
lock(mapping.invalidate_lock#3);
lock(&sbi->s_lock);
lock(&mm->mmap_lock);

*** DEADLOCK ***

3 locks held by syz-executor377/6548:
#0: ffff88802778c368 (&f->f_pos_lock){+.+.}-{3:3}, at: __fdget_pos+0x254/0x2f0 fs/file.c:1047
#1: ffff8880747e1cb0 (&sb->s_type->i_mutex_key#14){++++}-{3:3}, at: iterate_dir+0x135/0x570 fs/readdir.c:57
#2: ffff88807c2640e0 (&sbi->s_lock){+.+.}-{3:3}, at: exfat_iterate+0x171/0x3370 fs/exfat/dir.c:232

stack backtrace:
CPU: 0 PID: 6548 Comm: syz-executor377 Not tainted 6.3.0-rc3-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 03/02/2023
Call Trace:
<TASK>
__dump_stack lib/dump_stack.c:88 [inline]
dump_stack_lvl+0x1e7/0x2d0 lib/dump_stack.c:106
check_noncircular+0x2fe/0x3b0 kernel/locking/lockdep.c:2178
check_prev_add kernel/locking/lockdep.c:3098 [inline]
check_prevs_add kernel/locking/lockdep.c:3217 [inline]
validate_chain+0x166b/0x58e0 kernel/locking/lockdep.c:3832
__lock_acquire+0x125b/0x1f80 kernel/locking/lockdep.c:5056
lock_acquire+0x1e1/0x520 kernel/locking/lockdep.c:5669
down_read+0x3d/0x50 kernel/locking/rwsem.c:1520
mmap_read_lock include/linux/mmap_lock.h:117 [inline]
do_user_addr_fault arch/x86/mm/fault.c:1358 [inline]
handle_page_fault arch/x86/mm/fault.c:1498 [inline]
exc_page_fault+0x486/0x7c0 arch/x86/mm/fault.c:1554
asm_exc_page_fault+0x26/0x30 arch/x86/include/asm/idtentry.h:570
RIP: 0010:filldir64+0x30b/0x720 fs/readdir.c:335
Code: 48 29 eb 48 89 df 4c 89 e6 e8 11 86 95 ff 85 ed 0f 88 48 02 00 00 4c 39 e3 0f 82 3f 02 00 00 0f 01 cb 0f ae e8 48 8b 44 24 58 <49> 89 44 24 08 48 8b 4c 24 10 48 8b 44 24 50 48 89 01 48 8b 44 24
RSP: 0018:ffffc9000543f6c8 EFLAGS: 00050206
RAX: 0000000000000000 RBX: 00007fffffffefe8 RCX: ffff8880275dba80
RDX: ffff8880275dba80 RSI: 0000000000000000 RDI: 00007fffffffefe8
RBP: 0000000000000018 R08: ffffffff81f4e91f R09: 0000000000000004
R10: 0000000000000003 R11: ffff8880275dba80 R12: 0000000000000000
R13: ffffc9000543fe70 R14: 0000000000000001 R15: ffffffff8afed560
dir_emit_dot include/linux/fs.h:3144 [inline]
dir_emit_dots include/linux/fs.h:3155 [inline]
exfat_iterate+0x2b8/0x3370 fs/exfat/dir.c:235
iterate_dir+0x228/0x570
__do_sys_getdents64 fs/readdir.c:369 [inline]
__se_sys_getdents64+0x20d/0x4f0 fs/readdir.c:354
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x63/0xcd
RIP: 0033:0x7f265ea7dab9
Code: 28 00 00 00 75 05 48 83 c4 28 c3 e8 71 15 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 b8 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007f2656628208 EFLAGS: 00000246 ORIG_RAX: 00000000000000d9
RAX: ffffffffffffffda RBX: 00007f265eb037b8 RCX: 00007f265ea7dab9
RDX: 0000000000008008 RSI: 0000000000000000 RDI: 0000000000000006
RBP: 00007f265eb037b0 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 00007f265eb037bc
R13: 00007ffc20585adf R14: 00007f2656628300 R15: 0000000000022000
</TASK>
----------------
Code disassembly (best guess):
0: 48 29 eb sub %rbp,%rbx
3: 48 89 df mov %rbx,%rdi
6: 4c 89 e6 mov %r12,%rsi
9: e8 11 86 95 ff callq 0xff95861f
e: 85 ed test %ebp,%ebp
10: 0f 88 48 02 00 00 js 0x25e
16: 4c 39 e3 cmp %r12,%rbx
19: 0f 82 3f 02 00 00 jb 0x25e
1f: 0f 01 cb stac
22: 0f ae e8 lfence
25: 48 8b 44 24 58 mov 0x58(%rsp),%rax
* 2a: 49 89 44 24 08 mov %rax,0x8(%r12) <-- trapping instruction
2f: 48 8b 4c 24 10 mov 0x10(%rsp),%rcx
34: 48 8b 44 24 50 mov 0x50(%rsp),%rax
39: 48 89 01 mov %rax,(%rcx)
3c: 48 rex.W
3d: 8b .byte 0x8b
3e: 44 rex.R
3f: 24 .byte 0x24

Next message: Russell King (Oracle): "Re: [PATCH net-next v14 8/9] net: ethernet: mtk_eth_soc: switch to external PCS driver"
Previous message: James Clark: "[PATCH] perf vendor events arm64: Add N1 metrics"
Next in thread: syzbot: "Re: [syzbot] [fat?] possible deadlock in exc_page_fault"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]