[devel] Q: strftime and new GCC again.

["Anton V. Denisov" <fire@kgpu.kamchatka.ru>]

Приветствую всех.

Кто-нибудь может объяснить, почему с приходом нового GCC в Сизиф стало
происходит такое:

configure:8324: checking for strftime
configure:8367: i586-alt-linux-g++ -o
conftest -pipe -Wall -O2 -march=i586 -mcpu=i686   conftest.cc  >&5
configure:8339: error: nonnull argument with out-of-range operand number
(arg 1, operand 3)
configure:8370: $? = 1
configure: failed program was:
#line 8329 "configure"
#include "confdefs.h"
/* System header to define __stub macros and hopefully few prototypes,
    which can conflict with char strftime (); below.  */
#include <assert.h>
/* Override any gcc2 internal prototype to avoid an error.  */
#ifdef __cplusplus
extern "C"
#endif
/* We use char because int might match the return type of a gcc2
   builtin and then its argument prototype would still apply.  */
char strftime ();
char (*f) ();

#ifdef F77_DUMMY_MAIN
#  ifdef __cplusplus
     extern "C"
#  endif
   int F77_DUMMY_MAIN() { return 1; }
#endif
int
main ()
{
/* The GNU C library defines this for functions which it implements
    to always fail with ENOSYS.  Some functions are actually named
    something starting with __ and the normal name is an alias.  */
#if defined (__stub_strftime) || defined (__stub___strftime)
choke me
#else
f = strftime;
#endif

  ;
  return 0;
}
configure:8386: result: no

и как это исправить? ./configure сгенерирован Autoconf 2.53a. Вот
и возникает вопрос, почему gcc3.2-3.2.3-alt1 этот кусок компилирует, а
новый - нет.

Надеюсь, GCC Development Team ответит на мой вопрос.

С уважением, Антон В. Денисов.