Hi Alexey,

After thinking about it, the macro check is fine.  I have checked in the
patch.

Thanks,

-- Jeff J.

On Wed, May 17, 2023 at 3:37 PM Jeff Johnston <jjohnstn@redhat.com> wrote:

> Hi Alexey,
>
> I see the point for adding the ret=0 but the additional flag check isn't
> needed because the function is meant to only be called by __loadlocale()
> which doesn't make the call unless __HAVE_LOCALE_INFO__ is set.  If you
> want, I can make the change or else you can resubmit the patch.
>
> -- Jeff J.
>
> On Mon, May 15, 2023 at 12:30 PM Alexey Lapshin <
> alexey.lapshin@espressif.com> wrote:
>
>> newlib:
>>         * libc/locale/timelocal.c: fix ret variable initialization.
>>         Explicit __HAVE_LOCALE_INFO__ check
>> ---
>>  newlib/libc/locale/timelocal.c | 4 +++-
>>  1 file changed, 3 insertions(+), 1 deletion(-)
>>
>> diff --git a/newlib/libc/locale/timelocal.c
>> b/newlib/libc/locale/timelocal.c
>> index 4b361544a..04ae1142f 100644
>> --- a/newlib/libc/locale/timelocal.c
>> +++ b/newlib/libc/locale/timelocal.c
>> @@ -147,10 +147,11 @@ int
>>  __time_load_locale (struct __locale_t *locale, const char *name,
>>                     void *f_wctomb, const char *charset)
>>  {
>> -  int  ret;
>> +  int  ret = 0;
>>    struct lc_time_T ti;
>>    char *bufp = NULL;
>>
>> +#ifdef __HAVE_LOCALE_INFO__
>>  #ifdef __CYGWIN__
>>    extern int __set_lc_time_from_win (const char *, const struct
>> lc_time_T *,
>>                                      struct lc_time_T *, char **, void *,
>> @@ -186,5 +187,6 @@ __time_load_locale (struct __locale_t *locale, const
>> char *name,
>>  #else
>>    /* TODO */
>>  #endif
>> +#endif /*__HAVE_LOCALE_INFO__*/
>>    return (ret);
>>  }
>> --
>> 2.34.1
>>
>