linux-kernel - Re: [PATCH] printf: Emit "SUCCESS" if NULL is passed for %pe

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: <Yzri9bJiS46Bd1DH@alley>
Date:   Mon, 3 Oct 2022 15:26:13 +0200
From:   Petr Mladek <pmladek@...e.com>
To:     Uwe Kleine-König 
        <u.kleine-koenig@...gutronix.de>
Cc:     Andy Shevchenko <andriy.shevchenko@...ux.intel.com>,
        Rasmus Villemoes <linux@...musvillemoes.dk>,
        linux-kernel@...r.kernel.org, Steven Rostedt <rostedt@...dmis.org>,
        Sergey Senozhatsky <senozhatsky@...omium.org>,
        kernel@...gutronix.de
Subject: Re: [PATCH] printf: Emit "SUCCESS" if NULL is passed for %pe

On Fri 2022-09-30 16:05:31, Uwe Kleine-König wrote:
> On Fri, Sep 30, 2022 at 04:41:24PM +0300, Andy Shevchenko wrote:
> > On Fri, Sep 30, 2022 at 01:10:50PM +0200, Uwe Kleine-König wrote:
> > > For code that emits a string representing a usual return value it's
> > > convenient to have a 0 result in a string representation of success
> > > instead of "00000000".
> > 
> > This is a controversial change. For APIs that comes to my mind it means
> > "OPTIONAL resource NOT FOUND, while no error happened". Doe it mean success?
> > I don't think so.
> 
> OK, agreed. Would you feed such a value unchecked to %pe today (i.e.
> without my patch)?

People are primary interested into debug messages when things does
not work as expected. The check might be missing intentionally
to show all values or by mistake.

The tracepoint, used as motivation for this patch [1], is exactly
the situation where return values are printed without any check.

The problem is that %pe is used for both pointer and integer
return values. They have different semantic.

I do not feel comfortable with "improving" one use case and
breaking the other.

One solution would be to add support for "%de" but this would break
things. "%de" is supposed to print the 'e'. For example, it should
printk "123e" when the given number is 123.

Another solution would be to add modifier for the "%pe" modifier.
For example, "%ped". It would mean that the given value is in "int"
range. It could even print non-hashed value when it is out of range.

[1] https://lore.kernel.org/linux-pwm/20220916151506.298488-2-u.kleine-koenig@pengutronix.de/

Best Regards,
Petr