On 2022/02/01 17:12, Kirill Tkhai wrote:
> On 01.02.2022 02:37, Damien Le Moal wrote:
>> On 2022/02/01 1:21, Kirill Tkhai wrote:
>>> dm_cleanup_zoned_dev() uses queue, so it must be called before blk_cleanup_disk() starts its killing:
>>>
>>> blk_cleanup_disk->blk_cleanup_queue()->kobject_put()->blk_release_queue()->...RCU...->blk_free_queue_rcu()->kmem_cache_free()
>>>
>>> Otherwise, RCU callback may be executed first, and dm_cleanup_zoned_dev() touches freed memory:
>>>
>>> [ 85.074005] BUG: KASAN: use-after-free in dm_cleanup_zoned_dev+0x33/0xd0
>>> [ 85.075448] Read of size 8 at addr ffff88805ac6e430 by task dmsetup/681
>>>
>>> [ 85.076102] CPU: 4 PID: 681 Comm: dmsetup Not tainted 5.17.0-rc2+ #6
>>> [ 85.076570] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.14.0-2 04/01/2014
>>> [ 85.077150] Call Trace:
>>> [ 85.077332] <TASK>
>>> [ 85.077478] dump_stack_lvl+0x57/0x7d
>>> [ 85.077763] print_address_description.constprop.0+0x1f/0x150
>>> [ 85.078229] ? dm_cleanup_zoned_dev+0x33/0xd0
>>> [ 85.078579] kasan_report.cold+0x7f/0x11b
>>> [ 85.078910] ? dm_cleanup_zoned_dev+0x33/0xd0
>>> [ 85.079219] dm_cleanup_zoned_dev+0x33/0xd0
>>> [ 85.079499] __dm_destroy+0x26a/0x400
>>> [ 85.079781] ? dm_blk_ioctl+0x230/0x230
>>> [ 85.080090] ? up_write+0xd8/0x270
>>> [ 85.080364] dev_remove+0x156/0x1d0
>>> [ 85.080639] ctl_ioctl+0x269/0x530
>>> [ 85.080910] ? table_clear+0x140/0x140
>>> [ 85.081209] ? lock_release+0xb2/0x750
>>> [ 85.081518] ? remove_all+0x40/0x40
>>> [ 85.081796] ? rcu_read_lock_sched_held+0x12/0x70
>>> [ 85.082166] ? lock_downgrade+0x3c0/0x3c0
>>> [ 85.082468] ? rcu_read_lock_sched_held+0x12/0x70
>>> [ 85.082833] dm_ctl_ioctl+0xa/0x10
>>> [ 85.083094] __x64_sys_ioctl+0xb9/0xf0
>>> [ 85.083350] do_syscall_64+0x3b/0x90
>>> [ 85.083595] entry_SYSCALL_64_after_hwframe+0x44/0xae
>>> [ 85.083955] RIP: 0033:0x7fb6dfa95c27
>>> [ 85.084277] Code: 00 00 00 48 8b 05 69 92 0c 00 64 c7 00 26 00 00 00 48 c7 c0 ff ff ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 b8 10 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 39 92 0c 00 f7 d8 64 89 01 48
>>> [ 85.086009] RSP: 002b:00007fff882c6c28 EFLAGS: 00000206 ORIG_RAX: 0000000000000010
>>> [ 85.086914] RAX: ffffffffffffffda RBX: 00007fb6dfb73a8e RCX: 00007fb6dfa95c27
>>> [ 85.087607] RDX: 00007fb6e01d7ca0 RSI: 00000000c138fd04 RDI: 0000000000000003
>>> [ 85.088272] RBP: 00007fff882c6ce0 R08: 00007fb6dfbc3558 R09: 00007fff882c6a90
>>> [ 85.088956] R10: 00007fb6dfbc28a2 R11: 0000000000000206 R12: 00007fb6dfbc28a2
>>> [ 85.089477] R13: 00007fb6dfbc28a2 R14: 00007fb6dfbc28a2 R15: 00007fb6dfbc28a2
>>> [ 85.090038] </TASK>
>>>
>>> [ 85.090348] Allocated by task 673:
>>> [ 85.090625] kasan_save_stack+0x1e/0x40
>>> [ 85.090978] __kasan_slab_alloc+0x66/0x80
>>> [ 85.091336] kmem_cache_alloc_node+0x1ca/0x460
>>> [ 85.091742] blk_alloc_queue+0x33/0x4e0
>>> [ 85.092112] __blk_alloc_disk+0x1b/0x60
>>> [ 85.092464] dm_create+0x368/0xa20
>>> [ 85.092764] dev_create+0xb9/0x170
>>> [ 85.093118] ctl_ioctl+0x269/0x530
>>> [ 85.093521] dm_ctl_ioctl+0xa/0x10
>>> [ 85.093921] __x64_sys_ioctl+0xb9/0xf0
>>> [ 85.094177] do_syscall_64+0x3b/0x90
>>> [ 85.094420] entry_SYSCALL_64_after_hwframe+0x44/0xae
>>>
>>> [ 85.094868] Freed by task 0:
>>> [ 85.095062] kasan_save_stack+0x1e/0x40
>>> [ 85.095319] kasan_set_track+0x21/0x30
>>> [ 85.095569] kasan_set_free_info+0x20/0x30
>>> [ 85.095858] __kasan_slab_free+0xfb/0x130
>>> [ 85.096141] slab_free_freelist_hook+0x7d/0x150
>>> [ 85.096502] kmem_cache_free+0x13c/0x340
>>> [ 85.096778] rcu_do_batch+0x2d9/0x820
>>> [ 85.097049] rcu_core+0x3b8/0x570
>>> [ 85.097330] __do_softirq+0x1c4/0x63d
>>>
>>> [ 85.097792] Last potentially related work creation:
>>> [ 85.098226] kasan_save_stack+0x1e/0x40
>>> [ 85.098549] __kasan_record_aux_stack+0x96/0xa0
>>> [ 85.098951] call_rcu+0xc4/0x8f0
>>> [ 85.099220] kobject_put+0xd9/0x270
>>> [ 85.099557] disk_release+0xee/0x120
>>> [ 85.099864] device_release+0x59/0xf0
>>> [ 85.100208] kobject_put+0xd9/0x270
>>> [ 85.100509] cleanup_mapped_device+0x12b/0x1b0
>>> [ 85.100910] __dm_destroy+0x26a/0x400
>>> [ 85.101247] dev_remove+0x156/0x1d0
>>> [ 85.101555] ctl_ioctl+0x269/0x530
>>> [ 85.101859] dm_ctl_ioctl+0xa/0x10
>>> [ 85.102198] __x64_sys_ioctl+0xb9/0xf0
>>> [ 85.102527] do_syscall_64+0x3b/0x90
>>> [ 85.102811] entry_SYSCALL_64_after_hwframe+0x44/0xae
>>>
>>> [ 85.103347] The buggy address belongs to the object at ffff88805ac6e180
>>> which belongs to the cache request_queue of size 2992
>>> [ 85.104338] The buggy address is located 688 bytes inside of
>>> 2992-byte region [ffff88805ac6e180, ffff88805ac6ed30)
>>> [ 85.105255] The buggy address belongs to the page:
>>> [ 85.105633] page:000000000837df3c refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x5ac68
>>> [ 85.106363] head:000000000837df3c order:3 compound_mapcount:0 compound_pincount:0
>>> [ 85.106948] flags: 0xfffffc0010200(slab|head|node=0|zone=1|lastcpupid=0x1fffff)
>>> [ 85.107494] raw: 000fffffc0010200 0000000000000000 dead000000000122 ffff888001e58280
>>> [ 85.108091] raw: 0000000000000000 00000000800a000a 00000001ffffffff 0000000000000000
>>> [ 85.108695] page dumped because: kasan: bad access detected
>>>
>>> [ 85.109257] Memory state around the buggy address:
>>> [ 85.109625] ffff88805ac6e300: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
>>> [ 85.110178] ffff88805ac6e380: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
>>> [ 85.110973] >ffff88805ac6e400: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
>>> [ 85.111891] ^
>>> [ 85.112353] ffff88805ac6e480: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
>>> [ 85.113036] ffff88805ac6e500: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
>>> [ 85.113800] ==================================================================
>>>
>>> Fixes: bb37d77239af "dm: introduce zone append emulation"
>>> Signed-off-by: Kirill Tkhai <ktkhai@xxxxxxxxxxxxx>
>>>
>>> diff --git a/drivers/md/dm.c b/drivers/md/dm.c
>>> index dcbd6d201619..d472fe5dbc1d 100644
>>> --- a/drivers/md/dm.c
>>> +++ b/drivers/md/dm.c
>>> @@ -1607,6 +1607,7 @@ static void cleanup_mapped_device(struct mapped_device *md)
>>> md->dax_dev = NULL;
>>> }
>>>
>>> + dm_cleanup_zoned_dev(md);
>>> if (md->disk) {
>>> spin_lock(&_minor_lock);
>>> md->disk->private_data = NULL;
>>> @@ -1627,7 +1628,6 @@ static void cleanup_mapped_device(struct mapped_device *md)
>>> mutex_destroy(&md->swap_bios_lock);
>>>
>>> dm_mq_cleanup_mapped_device(md);
>>> - dm_cleanup_zoned_dev(md);
>>> }
>>>
>>> /*
>>>
>>
>> The commit message format is strange (long lines).
>> Apart from that, the fix looks good to me.
>>
>> Reviewed-by: Damien Le Moal <damien.lemoal@xxxxxxxxxxxxxxxxxx>
>
> This is for a reviewer convenience. All KASAN trace may be cut on commit if not needed.
My comment was about the commit message text you wrote at the beginning, not the
copy-paste of the kasan splat. The lines are too long.
--
Damien Le Moal
Western Digital Research
--
dm-devel mailing list
dm-devel@xxxxxxxxxx
https://listman.redhat.com/mailman/listinfo/dm-devel
