linux-kernel - Re: [PATCH] cifs: Show reason why autodisabling serverino support

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 for Android: free password hash cracker in your pocket

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

Message-ID: <20250610165555.sekma4ybcl63flnw@pali>
Date: Tue, 10 Jun 2025 18:55:55 +0200
From: Pali Rohár <pali@...nel.org>
To: Tom Talpey <tom@...pey.com>
Cc: Steve French <smfrench@...il.com>, Paulo Alcantara <pc@...guebit.org>,
	Steve French <sfrench@...ba.org>, linux-cifs@...r.kernel.org,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH] cifs: Show reason why autodisabling serverino support

For sure I can improve also existing text message. But Tom already sent
NACK so I'm not sure if I should continue with improvement or discard it.

On Tuesday 10 June 2025 11:36:30 Tom Talpey wrote:
> The message text is pretty cryptic: "inode numbers won't match anymore".
> Can this be stated in a more actionable way? The primary consumer of
> the log is the sysadmin, after all.
> 
> At a minimum, stating that hardlink detection won't work, as noted in
> the commit. Perhaps more helpfully, state that it's not a client issue.
> 
> Tom.
> 
> On 6/9/2025 3:36 AM, Pali Rohár wrote:
> > cifs_autodisable_serverino() is already printing at the VFS level
> > information that it is disabling the serverino. It is one time thing as
> > once it is disabled, it is not printing it second time. So it does not
> > flood logs.
> > 
> > In this change I have just extended this existing logging to print also reason.
> > 
> > On Sunday 08 June 2025 22:40:21 Steve French wrote:
> > > Since this could flood logs (e.g. in some DFS cases), probably better
> > > to do these via the usual dynamic trace points (and can document a
> > > simple "trace-cmd -e smb3_disable_serverino" script to avoid risk of
> > > flooding logs.    cifsFYI is an alternative but the world has moved to
> > > the dynamic tracing (eBPF etc.)
> > > 
> > > On Sun, Jun 8, 2025 at 3:57 PM Paulo Alcantara <pc@...guebit.org> wrote:
> > > > 
> > > > Pali Rohár <pali@...nel.org> writes:
> > > > 
> > > > > Before calling cifs_autodisable_serverino() function, show reason why it
> > > > > has to be called.
> > > > > 
> > > > > This change allows to debug issues why cifs.ko decide to turn off server
> > > > > inode number support and hence disable support for detection of hardlinks.
> > > > > 
> > > > > Signed-off-by: Pali Rohár <pali@...nel.org>
> > > > > Cc: stable@...r.kernel.org
> > > > > ---
> > > > >   fs/smb/client/connect.c   | 2 ++
> > > > >   fs/smb/client/dfs_cache.c | 2 ++
> > > > >   fs/smb/client/inode.c     | 3 +++
> > > > >   fs/smb/client/readdir.c   | 3 +++
> > > > >   4 files changed, 10 insertions(+)
> > > > 
> > > > NACK.
> > > > 
> > > 
> > > 
> > > -- 
> > > Thanks,
> > > 
> > > Steve
> > 
> > 
>