No, i dont think you missed something.
It's just one of the many false positive reported by cppcheck.
I tested replacing the PATH_MAX value by a static value and the
reported error disappeared...
On Sun, May 18, 2014 at 1:34 AM, Christophe <christophe.cu...@free.fr> wrote:
> I checked, Coverity did not report anything.
>
> The more I look at the code, the less I understand what cppcheck is
> complaining about.
> The first thing done is to snprintf into that buffer (path), so it gets
> initialised, and it does not care what was in previously. The only reason we
> could have a problem is when sizeof(path) <= 0 (then the behaviour of
> snprintf is undefined), which I believe is a bit unlikely.
>
> Did I miss something?
>
>
> ----- David Maciejak <david.macie...@gmail.com> a écrit :
>> Sure, as you may come with a better idea i am ok with that.
>> btw i believe coverity reported the same kind of problems.
>> So it's just to be sure we can handle them.
>>
>> regards,
>> david
>>
>> On Sun, May 18, 2014 at 12:52 AM, Christophe <christophe.cu...@free.fr>
>> wrote:
>> >
>> > May I suggest to not get this patch? This is not the correct way to fix
>> > the problem reported by cppcheck.
>> >
>> >
>> > ----- David Maciejak <david.macie...@gmail.com> a écrit :
>> >> set default path init value to remove some cppcheck warning
>> >>
>> >> ---
>> >> src/defaults.c | 2 +-
>> >> 1 file changed, 1 insertion(+), 1 deletion(-)
>> >>
>> >> diff --git a/src/defaults.c b/src/defaults.c
>> >> index 7d66372..c6748d5 100644
>> >> --- a/src/defaults.c
>> >> +++ b/src/defaults.c
>> >> @@ -798,7 +798,7 @@ static void initDefaults(void)
>> >> static WMPropList *readGlobalDomain(const char *domainName, Bool
>> >> requireDictionary)
>> >> {
>> >> WMPropList *globalDict = NULL;
>> >> - char path[PATH_MAX];
>> >> + char path[PATH_MAX] = "";
>> >> struct stat stbuf;
>> >>
>> >> snprintf(path, sizeof(path), "%s/%s/%s", SYSCONFDIR,
>> >> GLOBAL_DEFAULTS_SUBDIR, domainName);
>> >> --
>> >> 1.8.3.2
>> >
>> >
>> > --
>> > To unsubscribe, send mail to wmaker-dev-unsubscr...@lists.windowmaker.org.
>
>
> --
> To unsubscribe, send mail to wmaker-dev-unsubscr...@lists.windowmaker.org.
--
To unsubscribe, send mail to wmaker-dev-unsubscr...@lists.windowmaker.org.
