linux-kernel - Re: [PATCH] f2fs: protect free nid operation with cp

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:   Mon, 20 Mar 2017 19:06:02 +0800
From:   Chao Yu <yuchao0@...wei.com>
To:     Jaegeuk Kim <jaegeuk@...nel.org>
CC:     <linux-f2fs-devel@...ts.sourceforge.net>,
        <linux-kernel@...r.kernel.org>, <chao@...nel.org>
Subject: Re: [PATCH] f2fs: protect free nid operation with cp_rwsem

Hi Jaegeuk,

On 2017/3/20 6:44, Jaegeuk Kim wrote:
> Hi Chao,
> 
> On 03/17, Chao Yu wrote:
>> - f2fs_balance_fs_bg
>>  - build_free_ni
>>   - __build_free_nids
>>    - scan_nat_page
>>     - add_free_nid
>> 					- f2fs_create
>> 					 - f2fs_new_inode
>> 					  - alloc_nid
>>    - remove_free_nid
>>
>> This patch uses cp_rwsem to protect free nid allocation/initialization
>> to avoid reusing allocated nid in above scenario.
> 
> It looks like description doesn't match to your change below?

It looks like the problem will not happen because on_build_free_nids will stop
alloc_nid stealing one allocated nid while building free nids. Please ignore
this patch, sorry.

I found another potential case, could you please check that one?

Thanks,

> 
> Thanks,
> 
>>
>> Signed-off-by: Chao Yu <yuchao0@...wei.com>
>> ---
>>  fs/f2fs/inode.c   | 2 ++
>>  fs/f2fs/segment.c | 5 ++++-
>>  2 files changed, 6 insertions(+), 1 deletion(-)
>>
>> diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c
>> index 2520fa72b23f..f941410f994a 100644
>> --- a/fs/f2fs/inode.c
>> +++ b/fs/f2fs/inode.c
>> @@ -423,7 +423,9 @@ void f2fs_evict_inode(struct inode *inode)
>>  			add_ino_entry(sbi, inode->i_ino, UPDATE_INO);
>>  	}
>>  	if (is_inode_flag_set(inode, FI_FREE_NID)) {
>> +		f2fs_lock_op(sbi);
>>  		alloc_nid_failed(sbi, inode->i_ino);
>> +		f2fs_unlock_op(sbi);
>>  		clear_inode_flag(inode, FI_FREE_NID);
>>  	}
>>  	f2fs_bug_on(sbi, err &&
>> diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
>> index 0bf024f62330..f85e1f443c99 100644
>> --- a/fs/f2fs/segment.c
>> +++ b/fs/f2fs/segment.c
>> @@ -426,8 +426,11 @@ void f2fs_balance_fs_bg(struct f2fs_sb_info *sbi)
>>  
>>  	if (!available_free_memory(sbi, FREE_NIDS))
>>  		try_to_free_nids(sbi, MAX_FREE_NIDS);
>> -	else
>> +	else {
>> +		f2fs_lock_op(sbi);
>>  		build_free_nids(sbi, false, false);
>> +		f2fs_unlock_op(sbi);
>> +	}
>>  
>>  	if (!is_idle(sbi))
>>  		return;
>> -- 
>> 2.8.2.295.g3f1c1d0
> 
> .
>