linux-kernel - Re: [for-next][PATCH 23/24] string.h: Add strncmp

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]

Date:   Fri, 21 Dec 2018 14:57:13 -0800
From:   Linus Torvalds <torvalds@...ux-foundation.org>
To:     Steven Rostedt <rostedt@...dmis.org>
Cc:     Linux List Kernel Mailing <linux-kernel@...r.kernel.org>,
        Ingo Molnar <mingo@...nel.org>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Namhyung Kim <namhyung@...nel.org>,
        Masami Hiramatsu <mhiramat@...nel.org>,
        Joe Perches <joe@...ches.com>,
        Tom Zanussi <zanussi@...nel.org>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>
Subject: Re: [for-next][PATCH 23/24] string.h: Add strncmp_prefix() helper macro

On Fri, Dec 21, 2018 at 2:48 PM Steven Rostedt <rostedt@...dmis.org> wrote:
>
> > Your patch actually had them, but in the body of your email you did
> >
> > > #define have_prefix(str, prefix) ({ \
> > >       const char *__pfx = (const char *)prefix; \
> >
> > which is just completely wrong.
> >
> > Considering your _old_ patch had the exact same bug, I really think
> > you need to start internalizing the whole "macro arguments *have* to
> > be properly protected" thing.
>
> Well, there's less with assignments that can go wrong than with other
> code. That is, there's little that can happen with "int x = arg;" where
> arg is the macro paramater to cause something really nasty.

What's wrong, Steven?

The assignment is entirely irrelevant.

The problem is the cast.

A type cast has a very high priority, and so if you do

    (const char *)prefix

it breaks completely if you might have something like"a+6" as the argument.

Think about what if "a" is of type "unsigned long", for example?

                   Linus