linux-kernel - Re: [PATCH 2/6] statfs: use << to align with fs header

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: <20180413183250.GA8277@gmail.com>
Date:   Fri, 13 Apr 2018 20:32:50 +0200
From:   Christian Brauner <christian.brauner@...onical.com>
To:     Randy Dunlap <rdunlap@...radead.org>
Cc:     Andreas Dilger <adilger@...ger.ca>,
        Alexander Viro <viro@...iv.linux.org.uk>,
        Thomas Gleixner <tglx@...utronix.de>,
        Kate Stewart <kstewart@...uxfoundation.org>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        Philippe Ombredanne <pombredanne@...b.com>,
        Linux FS Devel <linux-fsdevel@...r.kernel.org>,
        linux-kernel@...r.kernel.org, serge@...lyn.com
Subject: Re: [PATCH 2/6] statfs: use << to align with fs header

On Fri, Apr 13, 2018 at 10:55:23AM -0700, Randy Dunlap wrote:
> On 04/13/2018 10:35 AM, Andreas Dilger wrote:
> > On Apr 13, 2018, at 10:11 AM, Christian Brauner <christian.brauner@...ntu.com> wrote:
> >>
> >> Consistenly use << to define ST_* constants. This also aligns them with
> >> their MS_* counterparts in fs.h
> > 
> > IMHO, using (1 << 10) makes the code harder to debug.  If you see a field
> > in a structure like 0x8354, it is non-trivial to map this to the ST_*
> > flags if they are declared in the form (1 << 10) or BIT(10).  If they are
> > declared in the form 0x100 (as they are now) then it is trivial that the
> > ST_APPEND flag is set in 0x8354, and easy to understand the other flags.
> > 
> > So, my preference would be to NOT land this or the previous patch.

All higher values are already initialized with bit-shifts for MS_*
constants starting with (1<<16) as you can see from the patch and in
fs.h:

> +#define MS_VERBOSE     (1<<15) /* War is peace. Verbosity is silence.
> +                                * MS_VERBOSE is deprecated.
> +                                */
> +#define MS_SILENT      (1<<15)
>  #define MS_POSIXACL    (1<<16) /* VFS does not apply the umask */
>  #define MS_UNBINDABLE  (1<<17) /* change to unbindable */
>  #define MS_PRIVATE     (1<<18) /* change to private */

This just makes it uniform which imho has merit on its own.

If using shifts is considered a valid counter argument because for lack
of ease to analyze struct fields then the values for MS_* flags in fs.h
should probably all be hex values.

In any case, I'm not going to bikeshed over this. The two patches can
simply be left out when applying or I can change it all over to hex
values.

Christian