Re: [PATCH] scripts: recordmcount: fix incorrect use of sprintf

[Kees Cook]

On Wed, Jan 6, 2016 at 3:28 AM, Fengguang Wu <fengguang.wu@xxxxxxxxx> wrote:
> Hi Steven,
>
> On Mon, Jan 04, 2016 at 10:42:46AM -0500, Steven Rostedt wrote:
>> On Wed, 30 Dec 2015 23:06:41 +0000
>> Colin King <colin.king@xxxxxxxxxxxxx> wrote:
>>
>> > From: Colin Ian King <colin.king@xxxxxxxxxxxxx>
>> >
>> > Fix build warning:
>> >
>> > scripts/recordmcount.c:589:4: warning: format not a string
>> > literal and no format arguments [-Wformat-security]
>> >     sprintf("%s: failed\n", file);
>> >
>> > Fixes: a50bd43935586 ("ftrace/scripts: Have recordmcount copy the object file")
>> > Signed-off-by: Colin Ian King <colin.king@xxxxxxxxxxxxx>
>> > ---
>> >  scripts/recordmcount.c | 2 +-
>> >  1 file changed, 1 insertion(+), 1 deletion(-)
>> >
>> > diff --git a/scripts/recordmcount.c b/scripts/recordmcount.c
>> > index 301d70b..e1675927 100644
>> > --- a/scripts/recordmcount.c
>> > +++ b/scripts/recordmcount.c
>> > @@ -586,7 +586,7 @@ main(int argc, char *argv[])
>> >                     do_file(file);
>> >                     break;
>> >             case SJ_FAIL:    /* error in do_file or below */
>> > -                   sprintf("%s: failed\n", file);
>> > +                   fprintf(stderr, "%s: failed\n", file);
>>
>> Paper bag bug. I'm not sure how this passed my tests? My tests check
>> for warnings. And I even got a "BUILD SUCCESS" from Fengguang Wu's
>> kbuild test robot.
>
> Because the error will only show up on "gcc -Wformat-security".

I wish GCC were smarter about this. I would have hoped at least ONE of
the various -Wformat... options would warn about "you did not actually
specify a buffer for sprintf". But yeah, this comes from
-Wformat-security (a subset of -Wformat-nonliteral), but can't be
turned on by default, as mentioned by Fengguang.

> It tend to raise false positives, so Kees setup a dedicated tree
> which enables -Wformat-security as well as quieting the common false
> positives. In that way Kees can catch such problems from time to time,
> however the limitation is, only upstreamed code can be tested in Kees'
> tree.

-Kees

-- 
Kees Cook
Chrome OS & Brillo Security
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/