linux-kernel - Re: [PATCH v1 1/8] lib/string: introduce match

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, 08 Jan 2016 10:40:51 +0200
From:	Andy Shevchenko <andriy.shevchenko@...ux.intel.com>
To:	Rasmus Villemoes <linux@...musvillemoes.dk>
Cc:	Tejun Heo <tj@...nel.org>,
	Linus Walleij <linus.walleij@...aro.org>,
	Dmitry Eremin-Solenikov <dbaryshkov@...il.com>,
	linux-kernel@...r.kernel.org, linux-pm@...r.kernel.org,
	"David S. Miller" <davem@...emloft.net>,
	David Airlie <airlied@...ux.ie>,
	Andrew Morton <akpm@...ux-foundation.org>
Subject: Re: [PATCH v1 1/8] lib/string: introduce match_string() helper

On Thu, 2016-01-07 at 23:05 +0100, Rasmus Villemoes wrote:
> On Thu, Jan 07 2016, Andy Shevchenko <andriy.shevchenko@...ux.intel.c
> om> wrote:
> 
> > From time to time we have to match a string in an array. Make a
> > simple helper
> > for that purpose.

> >  /**
> > + * match_string - matches given string in an array
> > + * @array:	array of strings
> > + * @len:	number of strings in the array or 0 for NULL
> > terminated arrays
> > + * @string:	string to match with
> > + *
> > + * Return:
> > + * index of a @string in the @array if matches, or %-ENODATA
> > otherwise.
> > + */
> > +int match_string(const char * const *array, size_t len, const char
> > *string)
> > +{
> > +	int index = 0;
> > +	const char *item;
> > +
> > +	do {
> > +		item = array[index];
> > +		if (!item)
> > +			break;
> > +		if (!strcmp(item, string))
> > +			return index;
> > +	} while (++index < len || !len);
> > +
> > +	return -ENODATA;
> > +}
> > +EXPORT_SYMBOL(match_string);
> > +
> 
> I'd suggest making it -1 (which, since len is a size_t, is
> effectively
> infinity) having the meaning "the array is terminated by a NULL
> entry". match_string(..., ARRAY_SIZE(my_array), ...) will break if
> the
> array happens to be empty, which could e.g. happen in a case like
> 
> const char *my_array[] = {
> #ifdef CONFIG_THIS
>        "this",
> #endif
> #ifdef CONFIG_THAT
>        "that",
> #endif
> };

It might make sense, though I don't remember current users with such
conditions.


> I also think the condition/loop above is unreadable.

Hmm… For me looks straightforward.

> 
> for (index = 0; index < len; index++) {
>     ...
> }
> 
> is much clearer.

If we switch to -1, it will look indeed simpler.

> 
> Why -ENODATA and not just -1? It is rather unlikely that anyone would
> pass on that particular -Exxx value. Not a biggie, just curious.

There are few of users already that would like to return error code to
upper level. In some cases better to have

return match_string();

than

ret = match_string();
if (ret < 0)
 return -EFOO;

return 0;

And returning -ENODATA doesn't prevent to have latter, but allows
former.

> 
> Would there be more potential users if we had a flag argument
> allowing
> case-insensitive matching? Would there be more potential users if a
> flag
> allowed to ask whether the given string is a _prefix_ of one of the
> strings in the array, or vice versa? Something like
> 
> #define MATCH_STRING_CASE 0x01
> #define MATCH_STRING_PREFIX_OF_ARRAY_ELEM 0x02 /* yeah, that name
> sucks */
> #define MATCH_ARRAY_ELEM_PREFIX_OF_STRING 0x04 /* this too */
> 
> int match_string(const char * const *array, size_t len, const char
> *string, unsigned flags)
> {
> #define MATCH_PREFIX (MATCH_... | MATCH_...)
>     int index;
>     const char *item;
>     int (*match_func)(const char *, const char *) =
>         flags & MATCH_STRING_CASE ? strcasecmp : strcmp;
>     int (*prefix_func)(const char *, const char *, size_t) =
>         flags & MATCH_STRING_CASE ? strncasecmp : strncmp;
>   
>     for (index = 0; index < len; ++index) {
>         item = array[index];
>         if (!item)
>             break;
>         if (flags & MATCH_PREFIX) {
>             size_t len = strlen(flags &
> MATCH_STRING_PREFIX_OF_ARRAY_ELEM ?
>                 string : item);
>             if (!prefix_func(item, string, len))
>                 return index;
>         } else if (!match_func(item, string)) {
>             return index;
>         }
>     }
>     return -1;
> }
> 
> (Ok, it's not that pretty; maybe it'd be better to use
> switch(flags&MATCH_PREFIX) {}. Or maybe just the case-insensitive
> part
> is worth keeping; in that case the above isn't that bad.)

I won't overcomplicate it until we have enough users to consider. Any
examples where we need this?

And I prefer way to have different prototypes for them instead of net
of conditions.

Thanks for review. I will send v3 (yeah, this is actually v2) with
change you proposed in the first part. For the second one I would like
to have real examples before doing anything.

-- 
Andy Shevchenko <andriy.shevchenko@...ux.intel.com>
Intel Finland Oy