linux-kernel - Re: [PATCH] ceph: correct superblock flags

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: <5b3f5414f900b16dfcc316c5b46b5c6bc8d2354e.camel@ibm.com>
Date: Fri, 30 May 2025 16:17:38 +0000
From: Viacheslav Dubeyko <Slava.Dubeyko@....com>
To: "frank.li@...o.com" <frank.li@...o.com>, Xiubo Li <xiubli@...hat.com>,
        "idryomov@...il.com" <idryomov@...il.com>
CC: "ceph-devel@...r.kernel.org" <ceph-devel@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re:  [PATCH] ceph: correct superblock flags

On Fri, 2025-05-30 at 02:44 -0600, Yangtao Li wrote:
> SB_NOATIME includes SB_NODIRATIME as a subset. Therefore,
> setting SB_NOATIME is sufficient to disable atime updates
> for all files and directories.
> 
> Signed-off-by: Yangtao Li <frank.li@...o.com>
> ---
>  fs/ceph/super.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/fs/ceph/super.c b/fs/ceph/super.c
> index f3951253e393..7c9e4987adf4 100644
> --- a/fs/ceph/super.c
> +++ b/fs/ceph/super.c
> @@ -1226,7 +1226,7 @@ static int ceph_set_super(struct super_block *s, struct fs_context *fc)
>  	s->s_time_gran = 1;
>  	s->s_time_min = 0;
>  	s->s_time_max = U32_MAX;
> -	s->s_flags |= SB_NODIRATIME | SB_NOATIME;
> +	s->s_flags |= SB_NOATIME;
>  
>  	ceph_fscrypt_set_ops(s);
> 

This modification doesn't make sense to me. And I prefer to be more conservative
here. Moreover, from my point of view, SB_NODIRATIME flag is associated with
nodiratime mount option and SB_NOATIME is associated with noatime mount option.
By setting these flags, we instruct how file system needs to behave for the case
of nodiratime and noatime mounts. So, semantically, this modification sounds
wrong to me. I don't think that we need to accept this modification, otherwise,
something could be broken.

Thanks,
Slava.