linux-kernel - Re: [PATCH] f2fs: support to disable linear lookup fallback

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]

Message-ID: <bef1588f-9c76-44a7-9b33-72ec307ae79d@kernel.org>
Date: Tue, 1 Apr 2025 10:26:13 +0800
From: Chao Yu <chao@...nel.org>
To: Daniel Lee <chullee@...gle.com>
Cc: chao@...nel.org, jaegeuk@...nel.org,
 linux-f2fs-devel@...ts.sourceforge.net, linux-kernel@...r.kernel.org,
 Gabriel Krisman Bertazi <krisman@...e.de>
Subject: Re: [PATCH] f2fs: support to disable linear lookup fallback

On 4/1/25 00:31, Daniel Lee wrote:
> On Sun, Mar 30, 2025 at 7:54 PM Chao Yu <chao@...nel.org> wrote:
>>
>> After commit 91b587ba79e1 ("f2fs: Introduce linear search for
>> dentries"), f2fs forced to use linear lookup whenever a hash-based
>> lookup fails on casefolded directory, it may affect performance
>> for scenarios: a) create a new file w/ filename it doesn't exist
>> in directory, b) lookup a file which may be removed.
>>
>> This patch supports to disable linear lookup fallback, so, once there
>> is a solution for commit 5c26d2f1d3f5 ("unicode: Don't special case
>> ignorable code points") to fix red heart unicode issue, then we can
>> set an encodeing flag to disable the fallback for performance recovery.
>>
>> The way is kept in line w/ ext4, refer to commit 9e28059d5664 ("ext4:
>> introduce linear search for dentries").
>>
>> Cc: Daniel Lee <chullee@...gle.com>
>> Cc: Gabriel Krisman Bertazi <krisman@...e.de>
>> Signed-off-by: Chao Yu <chao@...nel.org>
>> ---
>>  fs/f2fs/dir.c | 3 ++-
>>  1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c
>> index 5a63ff0df03b..e12445afb95a 100644
>> --- a/fs/f2fs/dir.c
>> +++ b/fs/f2fs/dir.c
>> @@ -366,7 +366,8 @@ struct f2fs_dir_entry *__f2fs_find_entry(struct inode *dir,
>>
>>  out:
>>  #if IS_ENABLED(CONFIG_UNICODE)
>> -       if (IS_CASEFOLDED(dir) && !de && use_hash) {
>> +       if (IS_CASEFOLDED(dir) && !de && use_hash &&
>> +               !sb_no_casefold_compat_fallback(dir->i_sb)) {
> 
> 
> Would it be beneficial to evaluate
> !sb_no_casefold_compat_fallback(dir->i_sb) first for short-circuiting?

Yeah, I guess it will when we disable the fallback by default, will update v2,
thank you!

Thanks,

> 
> 
>>
>>                 use_hash = false;
>>                 goto start_find_entry;
>>         }
>> --
>> 2.49.0
>>