[syzbot] KASAN: use-after-free Read in filp_close

[syzbot]

Hello,

syzbot found the following issue on:

HEAD commit: 09ce5091ff97 Add linux-next specific files for 20220524
git tree: linux-next
console output: https://syzkaller.appspot.com/x/log.txt?x=16b2d9c3f00000
kernel config: https://syzkaller.appspot.com/x/.config?x=39c2df848969b1de
dashboard link: https://syzkaller.appspot.com/bug?extid=47dd250f527cb7bebf24
compiler: gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.2

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: syzbot+47dd25...@syzkaller.appspotmail.com

==================================================================
==================================================================
BUG: KASAN: use-after-free in instrument_atomic_read include/linux/instrumented.h:71 [inline]
BUG: KASAN: use-after-free in atomic_long_read include/linux/atomic/atomic-instrumented.h:1265 [inline]
BUG: KASAN: use-after-free in filp_close+0x22/0x160 fs/open.c:1382
Read of size 8 at addr ffff8880732abbf0 by task syz-executor.2/5896

CPU: 1 PID: 5896 Comm: syz-executor.2 Not tainted 5.18.0-next-20220524-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
<TASK>
__dump_stack lib/dump_stack.c:88 [inline]
dump_stack_lvl+0xcd/0x134 lib/dump_stack.c:106
print_address_description.constprop.0.cold+0xeb/0x495 mm/kasan/report.c:313
print_report mm/kasan/report.c:429 [inline]
kasan_report.cold+0xf4/0x1c6 mm/kasan/report.c:491
check_region_inline mm/kasan/generic.c:183 [inline]
kasan_check_range+0x13d/0x180 mm/kasan/generic.c:189
instrument_atomic_read include/linux/instrumented.h:71 [inline]
atomic_long_read include/linux/atomic/atomic-instrumented.h:1265 [inline]
filp_close+0x22/0x160 fs/open.c:1382
close_fd+0x76/0xa0 fs/file.c:664
__do_sys_close fs/open.c:1407 [inline]
__se_sys_close fs/open.c:1405 [inline]
__x64_sys_close+0x2f/0xa0 fs/open.c:1405
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x46/0xb0
RIP: 0033:0x7f2dab83bd4b
Code: 0f 05 48 3d 00 f0 ff ff 77 45 c3 0f 1f 40 00 48 83 ec 18 89 7c 24 0c e8 63 fc ff ff 8b 7c 24 0c 41 89 c0 b8 03 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 35 44 89 c7 89 44 24 0c e8 a1 fc ff ff 8b 44
RSP: 002b:00007ffd73d7f330 EFLAGS: 00000293 ORIG_RAX: 0000000000000003
RAX: ffffffffffffffda RBX: 0000000000000005 RCX: 00007f2dab83bd4b
RDX: 00007f2dab9a18b0 RSI: ffffffffffffffff RDI: 0000000000000004
RBP: 00007f2dab99d960 R08: 0000000000000000 R09: 00007f2dab9a18b8
R10: 00007ffd73d7f430 R11: 0000000000000293 R12: 000000000002dac5
R13: 00007ffd73d7f430 R14: 00007f2dab99bf60 R15: 0000000000000032
</TASK>

Allocated by task 5898:
kasan_save_stack+0x1e/0x40 mm/kasan/common.c:38
kasan_set_track mm/kasan/common.c:45 [inline]
set_alloc_info mm/kasan/common.c:436 [inline]
__kasan_slab_alloc+0x90/0xc0 mm/kasan/common.c:469
kasan_slab_alloc include/linux/kasan.h:224 [inline]
slab_post_alloc_hook mm/slab.h:750 [inline]
slab_alloc_node mm/slub.c:3214 [inline]
slab_alloc mm/slub.c:3222 [inline]
__kmem_cache_alloc_lru mm/slub.c:3229 [inline]
kmem_cache_alloc+0x204/0x3b0 mm/slub.c:3239
kmem_cache_zalloc include/linux/slab.h:723 [inline]
__alloc_file+0x21/0x270 fs/file_table.c:138
alloc_empty_file+0x6d/0x170 fs/file_table.c:187
alloc_file+0x59/0x590 fs/file_table.c:229
alloc_file_pseudo+0x165/0x250 fs/file_table.c:269
sock_alloc_file+0x4f/0x190 net/socket.c:463
sock_map_fd net/socket.c:487 [inline]
__sys_socket+0x1a4/0x240 net/socket.c:1644
__do_sys_socket net/socket.c:1649 [inline]
__se_sys_socket net/socket.c:1647 [inline]
__x64_sys_socket+0x6f/0xb0 net/socket.c:1647
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x46/0xb0

Freed by task 22:
kasan_save_stack+0x1e/0x40 mm/kasan/common.c:38
kasan_set_track+0x21/0x30 mm/kasan/common.c:45
kasan_set_free_info+0x20/0x30 mm/kasan/generic.c:370
____kasan_slab_free mm/kasan/common.c:366 [inline]
____kasan_slab_free+0x166/0x1a0 mm/kasan/common.c:328
kasan_slab_free include/linux/kasan.h:200 [inline]
slab_free_hook mm/slub.c:1727 [inline]
slab_free_freelist_hook+0x8b/0x1c0 mm/slub.c:1753
slab_free mm/slub.c:3507 [inline]
kmem_cache_free+0xdd/0x5a0 mm/slub.c:3524
rcu_do_batch kernel/rcu/tree.c:2578 [inline]
rcu_core+0x7b1/0x1880 kernel/rcu/tree.c:2838
__do_softirq+0x29b/0x9c2 kernel/softirq.c:571

Last potentially related work creation:
kasan_save_stack+0x1e/0x40 mm/kasan/common.c:38
__kasan_record_aux_stack+0xbe/0xd0 mm/kasan/generic.c:348
call_rcu+0x99/0x790 kernel/rcu/tree.c:3126
task_work_run+0xdd/0x1a0 kernel/task_work.c:177
get_signal+0x1c5/0x2600 kernel/signal.c:2634
arch_do_signal_or_restart+0x82/0x20f0 arch/x86/kernel/signal.c:869
exit_to_user_mode_loop kernel/entry/common.c:166 [inline]
exit_to_user_mode_prepare+0x15f/0x250 kernel/entry/common.c:201
__syscall_exit_to_user_mode_work kernel/entry/common.c:283 [inline]
syscall_exit_to_user_mode+0x19/0x50 kernel/entry/common.c:294
do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86
entry_SYSCALL_64_after_hwframe+0x46/0xb0

Second to last potentially related work creation:
kasan_save_stack+0x1e/0x40 mm/kasan/common.c:38
__kasan_record_aux_stack+0xbe/0xd0 mm/kasan/generic.c:348
task_work_add+0x3a/0x1f0 kernel/task_work.c:48
fput fs/file_table.c:378 [inline]
fput+0xe7/0x190 fs/file_table.c:371
io_close fs/io_uring.c:6032 [inline]
io_issue_sqe+0x39cf/0xaa20 fs/io_uring.c:8360
io_queue_sqe fs/io_uring.c:8703 [inline]
io_req_task_submit+0xce/0x400 fs/io_uring.c:3066
handle_tw_list fs/io_uring.c:2938 [inline]
tctx_task_work+0x16a/0xe10 fs/io_uring.c:2967
task_work_run+0xdd/0x1a0 kernel/task_work.c:177
get_signal+0x1c5/0x2600 kernel/signal.c:2634
arch_do_signal_or_restart+0x82/0x20f0 arch/x86/kernel/signal.c:869
exit_to_user_mode_loop kernel/entry/common.c:166 [inline]
exit_to_user_mode_prepare+0x15f/0x250 kernel/entry/common.c:201
__syscall_exit_to_user_mode_work kernel/entry/common.c:283 [inline]
syscall_exit_to_user_mode+0x19/0x50 kernel/entry/common.c:294
do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86
entry_SYSCALL_64_after_hwframe+0x46/0xb0

The buggy address belongs to the object at ffff8880732abb80
which belongs to the cache filp of size 456
The buggy address is located 112 bytes inside of
456-byte region [ffff8880732abb80, ffff8880732abd48)

The buggy address belongs to the physical page:
page:ffffea0001ccaa80 refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x732aa
head:ffffea0001ccaa80 order:1 compound_mapcount:0 compound_pincount:0
memcg:ffff88807f60f401
flags: 0xfff00000010200(slab|head|node=0|zone=1|lastcpupid=0x7ff)
raw: 00fff00000010200 0000000000000000 dead000000000122 ffff8881400078c0
raw: 0000000000000000 00000000000c000c 00000001ffffffff ffff88807f60f401
page dumped because: kasan: bad access detected
page_owner tracks the page as allocated
page last allocated via order 1, migratetype Unmovable, gfp_mask 0x1d20c0(__GFP_IO|__GFP_FS|__GFP_NOWARN|__GFP_NORETRY|__GFP_COMP|__GFP_NOMEMALLOC|__GFP_HARDWALL), pid 2981, tgid 2981 (udevd), ts 187061397867, free_ts 187031412765
prep_new_page mm/page_alloc.c:2439 [inline]
get_page_from_freelist+0xa64/0x3d10 mm/page_alloc.c:4258
__alloc_pages+0x1c7/0x510 mm/page_alloc.c:5482
alloc_pages+0x1aa/0x310 mm/mempolicy.c:2272
alloc_slab_page mm/slub.c:1797 [inline]
allocate_slab+0x26c/0x3c0 mm/slub.c:1942
new_slab mm/slub.c:2002 [inline]
___slab_alloc+0x985/0xd90 mm/slub.c:3002
__slab_alloc.constprop.0+0x4d/0xa0 mm/slub.c:3089
slab_alloc_node mm/slub.c:3180 [inline]
slab_alloc mm/slub.c:3222 [inline]
__kmem_cache_alloc_lru mm/slub.c:3229 [inline]
kmem_cache_alloc+0x360/0x3b0 mm/slub.c:3239
kmem_cache_zalloc include/linux/slab.h:723 [inline]
__alloc_file+0x21/0x270 fs/file_table.c:138
alloc_empty_file+0x6d/0x170 fs/file_table.c:187
path_openat+0xe4/0x2910 fs/namei.c:3639
do_filp_open+0x1aa/0x400 fs/namei.c:3680
do_sys_openat2+0x16d/0x4c0 fs/open.c:1278
do_sys_open fs/open.c:1294 [inline]
__do_sys_openat fs/open.c:1310 [inline]
__se_sys_openat fs/open.c:1305 [inline]
__x64_sys_openat+0x13f/0x1f0 fs/open.c:1305
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x46/0xb0
page last free stack trace:
reset_page_owner include/linux/page_owner.h:24 [inline]
free_pages_prepare mm/page_alloc.c:1354 [inline]
free_pcp_prepare+0x549/0xd20 mm/page_alloc.c:1404
free_unref_page_prepare mm/page_alloc.c:3290 [inline]
free_unref_page+0x19/0x7b0 mm/page_alloc.c:3405
__unfreeze_partials+0x17c/0x1a0 mm/slub.c:2521
qlink_free mm/kasan/quarantine.c:168 [inline]
qlist_free_all+0x6a/0x170 mm/kasan/quarantine.c:187
kasan_quarantine_reduce+0x180/0x200 mm/kasan/quarantine.c:294
__kasan_slab_alloc+0xa2/0xc0 mm/kasan/common.c:446
kasan_slab_alloc include/linux/kasan.h:224 [inline]
slab_post_alloc_hook mm/slab.h:750 [inline]
slab_alloc_node mm/slub.c:3214 [inline]
slab_alloc mm/slub.c:3222 [inline]
__kmem_cache_alloc_lru mm/slub.c:3229 [inline]
kmem_cache_alloc+0x204/0x3b0 mm/slub.c:3239
getname_flags.part.0+0x50/0x4f0 fs/namei.c:139
getname_flags include/linux/audit.h:323 [inline]
getname+0x8e/0xd0 fs/namei.c:218
do_sys_openat2+0xf5/0x4c0 fs/open.c:1272
do_sys_open fs/open.c:1294 [inline]
__do_sys_openat fs/open.c:1310 [inline]
__se_sys_openat fs/open.c:1305 [inline]
__x64_sys_openat+0x13f/0x1f0 fs/open.c:1305
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x46/0xb0

Memory state around the buggy address:
ffff8880732aba80: 00 00 00 00 00 00 00 00 00 fc fc fc fc fc fc fc
ffff8880732abb00: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
>ffff8880732abb80: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
^
ffff8880732abc00: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
ffff8880732abc80: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
==================================================================


---
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.

[syzbot]

syzbot has found a reproducer for the following issue on:

HEAD commit: 952923ddc011 Merge tag 'pull-18-rc1-work.namei' of git://g..
git tree: upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=173fb6dbf00000
kernel config: https://syzkaller.appspot.com/x/.config?x=3096247591885bfa

dashboard link: https://syzkaller.appspot.com/bug?extid=47dd250f527cb7bebf24
compiler: gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.2

syz repro: https://syzkaller.appspot.com/x/repro.syz?x=114f7bcdf00000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=1659a94ff00000

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+47dd25...@syzkaller.appspotmail.com

==================================================================

BUG: KASAN: use-after-free in instrument_atomic_read include/linux/instrumented.h:71 [inline]
BUG: KASAN: use-after-free in atomic_long_read include/linux/atomic/atomic-instrumented.h:1265 [inline]
BUG: KASAN: use-after-free in filp_close+0x22/0x160 fs/open.c:1382

Read of size 8 at addr ffff88807a11a7f0 by task syz-executor408/4522

CPU: 1 PID: 4522 Comm: syz-executor408 Not tainted 5.18.0-syzkaller-13842-g952923ddc011 #0

Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
<TASK>
__dump_stack lib/dump_stack.c:88 [inline]
dump_stack_lvl+0xcd/0x134 lib/dump_stack.c:106
print_address_description.constprop.0.cold+0xeb/0x495 mm/kasan/report.c:313
print_report mm/kasan/report.c:429 [inline]
kasan_report.cold+0xf4/0x1c6 mm/kasan/report.c:491
check_region_inline mm/kasan/generic.c:183 [inline]
kasan_check_range+0x13d/0x180 mm/kasan/generic.c:189
instrument_atomic_read include/linux/instrumented.h:71 [inline]
atomic_long_read include/linux/atomic/atomic-instrumented.h:1265 [inline]
filp_close+0x22/0x160 fs/open.c:1382

close_files fs/file.c:432 [inline]
put_files_struct fs/file.c:447 [inline]
put_files_struct+0x1d0/0x350 fs/file.c:444
exit_files+0x7e/0xa0 fs/file.c:464
do_exit+0xac8/0x2a00 kernel/exit.c:790
do_group_exit+0xd2/0x2f0 kernel/exit.c:925
get_signal+0x2542/0x2600 kernel/signal.c:2857

arch_do_signal_or_restart+0x82/0x20f0 arch/x86/kernel/signal.c:869
exit_to_user_mode_loop kernel/entry/common.c:166 [inline]
exit_to_user_mode_prepare+0x15f/0x250 kernel/entry/common.c:201
__syscall_exit_to_user_mode_work kernel/entry/common.c:283 [inline]
syscall_exit_to_user_mode+0x19/0x50 kernel/entry/common.c:294
do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86
entry_SYSCALL_64_after_hwframe+0x46/0xb0

RIP: 0033:0x7f2201c1f159
Code: Unable to access opcode bytes at RIP 0x7f2201c1f12f.
RSP: 002b:00007f2201bcc308 EFLAGS: 00000246 ORIG_RAX: 00000000000000ca
RAX: fffffffffffffe00 RBX: 00007f2201ca2428 RCX: 00007f2201c1f159
RDX: 0000000000000000 RSI: 0000000000000080 RDI: 00007f2201ca2428
RBP: 00007f2201ca2420 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 00007f2201c70074
R13: 0000000000000005 R14: 00007f2201bcc400 R15: 0000000000022000
</TASK>

Allocated by task 4522:

kasan_save_stack+0x1e/0x40 mm/kasan/common.c:38
kasan_set_track mm/kasan/common.c:45 [inline]
set_alloc_info mm/kasan/common.c:436 [inline]
__kasan_slab_alloc+0x90/0xc0 mm/kasan/common.c:469
kasan_slab_alloc include/linux/kasan.h:224 [inline]
slab_post_alloc_hook mm/slab.h:750 [inline]
slab_alloc_node mm/slub.c:3214 [inline]
slab_alloc mm/slub.c:3222 [inline]
__kmem_cache_alloc_lru mm/slub.c:3229 [inline]
kmem_cache_alloc+0x204/0x3b0 mm/slub.c:3239
kmem_cache_zalloc include/linux/slab.h:723 [inline]
__alloc_file+0x21/0x270 fs/file_table.c:138
alloc_empty_file+0x6d/0x170 fs/file_table.c:187

path_openat+0xe4/0x2910 fs/namei.c:3639
do_filp_open+0x1aa/0x400 fs/namei.c:3680
do_sys_openat2+0x16d/0x4c0 fs/open.c:1278
do_sys_open fs/open.c:1294 [inline]
__do_sys_openat fs/open.c:1310 [inline]
__se_sys_openat fs/open.c:1305 [inline]
__x64_sys_openat+0x13f/0x1f0 fs/open.c:1305
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x46/0xb0

io_close fs/io_uring.c:6053 [inline]
io_issue_sqe+0x3f29/0x9780 fs/io_uring.c:8464
io_queue_sqe fs/io_uring.c:8794 [inline]
io_req_task_submit+0xce/0x400 fs/io_uring.c:3110
handle_tw_list fs/io_uring.c:2984 [inline]
tctx_task_work+0x16a/0xe10 fs/io_uring.c:3013

task_work_run+0xdd/0x1a0 kernel/task_work.c:177
get_signal+0x1c5/0x2600 kernel/signal.c:2634
arch_do_signal_or_restart+0x82/0x20f0 arch/x86/kernel/signal.c:869
exit_to_user_mode_loop kernel/entry/common.c:166 [inline]
exit_to_user_mode_prepare+0x15f/0x250 kernel/entry/common.c:201
__syscall_exit_to_user_mode_work kernel/entry/common.c:283 [inline]
syscall_exit_to_user_mode+0x19/0x50 kernel/entry/common.c:294
do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86
entry_SYSCALL_64_after_hwframe+0x46/0xb0

The buggy address belongs to the object at ffff88807a11a780

which belongs to the cache filp of size 456
The buggy address is located 112 bytes inside of

456-byte region [ffff88807a11a780, ffff88807a11a948)

The buggy address belongs to the physical page:

page:ffffea0001e84680 refcount:1 mapcount:0 mapping:0000000000000000 index:0xffff88807a11bb80 pfn:0x7a11a
head:ffffea0001e84680 order:1 compound_mapcount:0 compound_pincount:0

flags: 0xfff00000010200(slab|head|node=0|zone=1|lastcpupid=0x7ff)

raw: 00fff00000010200 ffffea0001eed500 dead000000000004 ffff8881400078c0
raw: ffff88807a11bb80 00000000800c000a 00000001ffffffff 0000000000000000

page dumped because: kasan: bad access detected
page_owner tracks the page as allocated

page last allocated via order 1, migratetype Unmovable, gfp_mask 0xd20c0(__GFP_IO|__GFP_FS|__GFP_NOWARN|__GFP_NORETRY|__GFP_COMP|__GFP_NOMEMALLOC), pid 3296, tgid 3296 (dhcpcd-run-hook), ts 21001730150, free_ts 17870277790
prep_new_page mm/page_alloc.c:2456 [inline]
get_page_from_freelist+0x1290/0x3b70 mm/page_alloc.c:4198
__alloc_pages+0x1c7/0x510 mm/page_alloc.c:5426

free_pages_prepare mm/page_alloc.c:1371 [inline]
free_pcp_prepare+0x549/0xd20 mm/page_alloc.c:1421
free_unref_page_prepare mm/page_alloc.c:3343 [inline]
free_unref_page+0x19/0x6a0 mm/page_alloc.c:3438

qlink_free mm/kasan/quarantine.c:168 [inline]
qlist_free_all+0x6a/0x170 mm/kasan/quarantine.c:187
kasan_quarantine_reduce+0x180/0x200 mm/kasan/quarantine.c:294
__kasan_slab_alloc+0xa2/0xc0 mm/kasan/common.c:446
kasan_slab_alloc include/linux/kasan.h:224 [inline]
slab_post_alloc_hook mm/slab.h:750 [inline]
slab_alloc_node mm/slub.c:3214 [inline]
slab_alloc mm/slub.c:3222 [inline]
__kmem_cache_alloc_lru mm/slub.c:3229 [inline]
kmem_cache_alloc+0x204/0x3b0 mm/slub.c:3239
getname_flags.part.0+0x50/0x4f0 fs/namei.c:139

getname_flags+0x9a/0xe0 include/linux/audit.h:323
vfs_fstatat+0x73/0xb0 fs/stat.c:254
__do_sys_newfstatat+0x91/0x110 fs/stat.c:425

do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x46/0xb0

Memory state around the buggy address:

ffff88807a11a680: fb fb fb fb fb fb fb fb fb fc fc fc fc fc fc fc
ffff88807a11a700: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
>ffff88807a11a780: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
^
ffff88807a11a800: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
ffff88807a11a880: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
==================================================================

[Al Viro]

On Sun, Jun 05, 2022 at 01:49:19AM -0700, syzbot wrote:
> syzbot has found a reproducer for the following issue on:
>
> HEAD commit: 952923ddc011 Merge tag 'pull-18-rc1-work.namei' of git://g..
> git tree: upstream
> console output: https://syzkaller.appspot.com/x/log.txt?x=173fb6dbf00000
> kernel config: https://syzkaller.appspot.com/x/.config?x=3096247591885bfa
> dashboard link: https://syzkaller.appspot.com/bug?extid=47dd250f527cb7bebf24
> compiler: gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.2
> syz repro: https://syzkaller.appspot.com/x/repro.syz?x=114f7bcdf00000
> C reproducer: https://syzkaller.appspot.com/x/repro.c?x=1659a94ff00000

_Very_ interesting. Has anything of that sort been observed on -next?

Because everything in that branch had been present there since at least
January.

[Hillf Danton]

On Sun, 05 Jun 2022 01:49:19 -0700

Given fput() in filep_close(), no more fput() is needed.

#syz test https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git 952923ddc011

--- y/fs/io_uring.c
+++ x/fs/io_uring.c
@@ -6046,6 +6046,7 @@ static int io_close(struct io_kiocb *req

/* No ->flush() or already async, safely close from here */
ret = filp_close(file, current->files);
+ file = NULL;
err:
if (ret < 0)
req_set_fail(req);
--

[syzbot]

syzbot has bisected this issue to:

commit 6319194ec57b0452dcda4589d24c4e7db299c5bf
Author: Al Viro <vi...@zeniv.linux.org.uk>
Date: Thu May 12 21:08:03 2022 +0000

Unify the primitives for file descriptor closing

bisection log: https://syzkaller.appspot.com/x/bisect.txt?x=134cbe4ff00000
start commit: 952923ddc011 Merge tag 'pull-18-rc1-work.namei' of git://g..
git tree: upstream
final oops: https://syzkaller.appspot.com/x/report.txt?x=10ccbe4ff00000
console output: https://syzkaller.appspot.com/x/log.txt?x=174cbe4ff00000

kernel config: https://syzkaller.appspot.com/x/.config?x=3096247591885bfa
dashboard link: https://syzkaller.appspot.com/bug?extid=47dd250f527cb7bebf24

syz repro: https://syzkaller.appspot.com/x/repro.syz?x=114f7bcdf00000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=1659a94ff00000

Reported-by: syzbot+47dd25...@syzkaller.appspotmail.com
Fixes: 6319194ec57b ("Unify the primitives for file descriptor closing")

For information about bisection process see: https://goo.gl/tpsmEJ#bisection

[syzbot]

Hello,

syzbot has tested the proposed patch and the reproducer did not trigger any issue:

Reported-and-tested-by: syzbot+47dd25...@syzkaller.appspotmail.com

Tested on:

commit: 952923dd Merge tag 'pull-18-rc1-work.namei' of git://g..
git tree: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git

kernel config: https://syzkaller.appspot.com/x/.config?x=3096247591885bfa
dashboard link: https://syzkaller.appspot.com/bug?extid=47dd250f527cb7bebf24
compiler: gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.2

patch: https://syzkaller.appspot.com/x/patch.diff?x=16bae7f5f00000

Note: testing is done by a robot and is best-effort only.

[Al Viro]

Argh... I see what's going on. Check if the following fixes the problem,
please.

diff --git a/drivers/android/binder.c b/drivers/android/binder.c
index 27c9b004823a..73beea5dc18c 100644
--- a/drivers/android/binder.c
+++ b/drivers/android/binder.c
@@ -1857,6 +1857,8 @@ static void binder_deferred_fd_close(int fd)
init_task_work(&twcb->twork, binder_do_fd_close);
twcb->file = close_fd_get_file(fd);
if (twcb->file) {
+ // pin it until binder_do_fd_close(); see comments there
+ get_file(twcb->file);
filp_close(twcb->file, current->files);
task_work_add(current, &twcb->twork, TWA_RESUME);
} else {
diff --git a/fs/file.c b/fs/file.c
index dd6692048f4f..3bcc1ecc314a 100644
--- a/fs/file.c
+++ b/fs/file.c
@@ -800,8 +800,7 @@ struct file *__close_fd_get_file(unsigned int fd)

/*
* variant of close_fd that gets a ref on the file for later fput.
- * The caller must ensure that filp_close() called on the file, and then
- * an fput().
+ * The caller must ensure that filp_close() called on the file.
*/
struct file *close_fd_get_file(unsigned int fd)
{
diff --git a/fs/io_uring.c b/fs/io_uring.c
index 7257b0870353..33da5116cc38 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -5110,7 +5110,7 @@ static int io_close(struct io_kiocb *req, unsigned int issue_flags)
struct files_struct *files = current->files;
struct io_close *close = &req->close;
struct fdtable *fdt;
- struct file *file = NULL;
+ struct file *file;
int ret = -EBADF;

if (req->close.file_slot) {
@@ -5127,7 +5127,6 @@ static int io_close(struct io_kiocb *req, unsigned int issue_flags)
file = fdt->fd[close->fd];
if (!file || file->f_op == &io_uring_fops) {
spin_unlock(&files->file_lock);
- file = NULL;
goto err;
}

@@ -5147,8 +5146,6 @@ static int io_close(struct io_kiocb *req, unsigned int issue_flags)

err:
if (ret < 0)
req_set_fail(req);

- if (file)
- fput(file);
__io_req_complete(req, issue_flags, ret, 0);
return 0;
}

[Al Viro]

On Sun, Jun 05, 2022 at 04:15:58PM +0000, Al Viro wrote:
> Argh... I see what's going on. Check if the following fixes the problem,
> please.

#syz test: git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs.git work.fd

[syzbot]

Hello,

syzbot has tested the proposed patch and the reproducer did not trigger any issue:

Reported-and-tested-by: syzbot+47dd25...@syzkaller.appspotmail.com

Tested on:

commit: 6dda6985 fix the breakage in close_fd_get_file() calli..
git tree: git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs.git work.fd
kernel config: https://syzkaller.appspot.com/x/.config?x=d4042ecb71632a26

dashboard link: https://syzkaller.appspot.com/bug?extid=47dd250f527cb7bebf24
compiler: gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.2

Note: no patches were applied.