linux-kernel - Re: [PATCH v2] btrfs: qgroup: fix sleep from invalid context bug in update_qgroup_limit

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:   Fri, 11 Nov 2022 17:36:20 +0800
From:   ChenXiaoSong <chenxiaosong2@...wei.com>
To:     Qu Wenruo <quwenruo.btrfs@....com>, <clm@...com>,
        <josef@...icpanda.com>, <dsterba@...e.com>
CC:     <linux-btrfs@...r.kernel.org>, <linux-kernel@...r.kernel.org>,
        <zhangxiaoxu5@...wei.com>
Subject: Re: [PATCH v2] btrfs: qgroup: fix sleep from invalid context bug in
 update_qgroup_limit_item()

Thanks for your suggestions, I will try to send a new version patch.

在 2022/11/11 17:09, Qu Wenruo 写道:
> 
> 
> On 2022/11/11 17:02, ChenXiaoSong wrote:
>> Syzkaller reported BUG as follows:
>>
>>    BUG: sleeping function called from invalid context at
>>         include/linux/sched/mm.h:274
>>    Call Trace:
>>     <TASK>
>>     dump_stack_lvl+0xcd/0x134
>>     __might_resched.cold+0x222/0x26b
>>     kmem_cache_alloc+0x2e7/0x3c0
>>     update_qgroup_limit_item+0xe1/0x390
>>     btrfs_qgroup_inherit+0x147b/0x1ee0
>>     create_subvol+0x4eb/0x1710
>>     btrfs_mksubvol+0xfe5/0x13f0
>>     __btrfs_ioctl_snap_create+0x2b0/0x430
>>     btrfs_ioctl_snap_create_v2+0x25a/0x520
>>     btrfs_ioctl+0x2a1c/0x5ce0
>>     __x64_sys_ioctl+0x193/0x200
>>     do_syscall_64+0x35/0x80
>>
>> Fix this by delaying the limit item updates until unlock the spin lock.
> 
> The overall idea is way better now.
> 
> But sorry that I didn't immediately find out the best solution at the 
> first glance.
> 
> In fact, your v2 path just lets me remember what is the correct way to 
> handle such situation.
> 
>>
>> Signed-off-by: ChenXiaoSong <chenxiaosong2@...wei.com>
>> ---
>>   fs/btrfs/qgroup.c | 18 +++++++++---------
>>   1 file changed, 9 insertions(+), 9 deletions(-)
>>
>> diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c
>> index 9334c3157c22..2792d63c0da4 100644
>> --- a/fs/btrfs/qgroup.c
>> +++ b/fs/btrfs/qgroup.c
> [...]
>> @@ -2985,6 +2978,7 @@ int btrfs_qgroup_inherit(struct 
>> btrfs_trans_handle *trans, u64 srcid,
>>           dstgroup->max_excl = srcgroup->max_excl;
>>           dstgroup->rsv_rfer = srcgroup->rsv_rfer;
>>           dstgroup->rsv_excl = srcgroup->rsv_excl;
>> +        update_limit = true;
>>           qgroup_dirty(fs_info, dstgroup);
>>           qgroup_dirty(fs_info, srcgroup);
> 
> You caught the "if (srcid)" branch, which also changed the 
> limit/rfer/excl numbers of the destination qgroup, but didn't call 
> update_qgroup_limit_item().
> 
> 
> But if you check the function qgroup_dirty() a little deeper, you can 
> find out that, qgroup_dirty() will move the target qgroup into 
> fs_info->dirty_qgroups list.
> 
> And later at btrfs_run_qgroups() (which is called during 
> btrfs_commit_transaction(), and also after create_pending_snapshots()), 
> we will update the quota tree to reflect the result.
> 
> So this means, all you need is just call qgroup_dirty() on @dstqgroup, 
> and call it a day.
> Everything else will be properly handled.
> 
> Sorry I didn't notice this earlier...
> 
> Thanks,
> Qu
> 
>> @@ -3053,6 +3047,12 @@ int btrfs_qgroup_inherit(struct 
>> btrfs_trans_handle *trans, u64 srcid,
>>   unlock:
>>       spin_unlock(&fs_info->qgroup_lock);
>> +    if (update_limit && update_qgroup_limit_item(trans, dstgroup)) {
>> +        qgroup_mark_inconsistent(fs_info);
>> +        btrfs_info(fs_info,
>> +               "unable to update quota limit for %llu",
>> +               dstgroup->qgroupid);
>> +    }
>>       if (!ret)
>>           ret = btrfs_sysfs_add_one_qgroup(fs_info, dstgroup);
>>   out:
> 
> .