Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Thu, 07 Nov 2013 23:35:01 +0100
From: Jens Gustedt <>
Subject: Re: [PATCH v2] shadow: Implement putspent

Am Mittwoch, den 06.11.2013, 11:36 -0500 schrieb Rich Felker:
> On Wed, Nov 06, 2013 at 02:20:54PM +0100, Jens Gustedt wrote:
> > in the context of the actual function that would certainly overkill,
> > but generally it is not a good idea to mix user strings and string
> > literals without consting them. So in a general context I'd go for
> > something like
> > 
> > #define STR(S) ((char const*)((S) ? (S) : ""))
> > 
> > or even
> > 
> > #define STR(S) ((S) ? (char const*){ (S) } : "")
> > 
> > to have a better type check for the argument
> I disagree with this change. The type of string literals is char *,
> not const char *, so it's not a type consistency issue. Even if it
> were, the ?: operator handles the type correctly anyway. My view is
> that casts are a code smell, and no-op casts are harmful in that they
> obfuscate the correctness of types (since the reader has to question
> whether the cast is hiding a type mismatch).

The second variant isn't a casţ but a conversion and it just checks if
S is assignment compatible with `char const*`. A completely type safe
variant then would be

#define STR(S) ((S) ? (char const*){ (S) } : (char const[]){ 0 })

which wouldn't imply any conversion. (And which a compiler *may*
realize by using a static object for the empty string.)

Just to be clear for the reason I mentioned it: the original macro has
the danger that it is used in places where it is passed as char* to a
function which then modifies the string.


The user who calls first the macro and then the function with his own
variable thinks that he is passing a modifiable string "toto" to the
function. The bug only appears when by accident someday "toto" is 0,
and a then a string literal is passed into the function.


:: INRIA Nancy Grand Est ::   ::
:: AlGorille ::::::::::::::: office Nancy : +33 383593090   ::
:: ICube :::::::::::::: office Strasbourg : +33 368854536   ::
:: ::::::::::::::::::::::::::: gsm France : +33 651400183   ::
:: :::::::::::::::::::: gsm international : +49 15737185122 ::

Download attachment "signature.asc" of type "application/pgp-signature" (199 bytes)

Powered by blists - more mailing lists

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.