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] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAHk-=wgNFNYinkWCUvT2UnH2E2K_qPexEPgrm-xgr68YXnEQ_g@mail.gmail.com>
Date: Tue, 12 Nov 2024 11:54:16 -0800
From: Linus Torvalds <torvalds@...ux-foundation.org>
To: Josef Bacik <josef@...icpanda.com>
Cc: kernel-team@...com, linux-fsdevel@...r.kernel.org, jack@...e.cz, 
	amir73il@...il.com, brauner@...nel.org, linux-xfs@...r.kernel.org, 
	linux-btrfs@...r.kernel.org, linux-mm@...ck.org, linux-ext4@...r.kernel.org
Subject: Re: [PATCH v7 07/18] fsnotify: generate pre-content permission event
 on open

On Tue, 12 Nov 2024 at 09:56, Josef Bacik <josef@...icpanda.com> wrote:
>
> +       /*
> +        * This permission hook is different than fsnotify_open_perm() hook.
> +        * This is a pre-content hook that is called without sb_writers held
> +        * and after the file was truncated.
> +        */
> +       return fsnotify_file_area_perm(file, MAY_OPEN, &file->f_pos, 0);
>  }

I still object to this all.

You can't say "permission denied" after you've already truncated the
file. It's not a sane model. I complained about that earlier, it seems
that complaint was missed in the other complaints.

Also, this whole "This permission hook is different than
fsnotify_open_perm() hook" statement is purely because
fsnotify_open_perm() itself was broken and called from the wrong place
as mentioned in the other email.

Fix *THAT* first, then unify the two places that should *not* be
different into one single "this is the fsnotify_open" code. And that
place explicitly sets that FMODE_NOTIFY_PERM bit, and makes sure that
it does *not* set it for FMODE_NONOTIFY or FMODE_PATH cases.

And then please work on making sure that that isn't called unless
actually required.

The actual real "pre-content permission events" should then ONLY test
the FMODE_NOTIFY_PERM bit. Nothing else. None of this "re-use the
existing fsnotify_file() logic" stuff. Noe extra tests, no extra
logic.

Don't make me jump through filve layers of inline functions that all
test different 'mask' bits, just to verify that the open / read /
write paths don't do something stupid.

IOW, make it straightforward and obvious what you are doing, and make
it very clear that you're not pointlessly testing things like
FMODE_NONOTIFY when the *ONLY* thing that should be tested is whether
FMODE_NOTIFY_PERM is set.

Please.

              Linus

Powered by blists - more mailing lists