develooper Front page | perl.perl5.porters | Postings from July 2001

Re: Report /pro/3gl/CPAN/perl-5.7.1

Thread Previous | Thread Next
Jarkko Hietaniemi
July 10, 2001 07:32
Re: Report /pro/3gl/CPAN/perl-5.7.1
Message ID:
> AIX 4.2      stdio/perlio     -Uuseperlio -Dusethreads -Duseithreads
> AIX 4.2      stdio/perlio     -DDEBUGGING -Uuseperlio -Dusethreads -Duseithreads
> AIX 4.2      stdio/perlio     -Duseperlio -Dusethreads -Duseithreads
> AIX 4.2      stdio/perlio     -DDEBUGGING -Duseperlio -Dusethreads -Duseithreads
> AIX 4.2      stdio/perlio     -Duseperlio -Duselongdouble -Dusethreads -Duseithreads
> AIX 4.2      stdio/perlio     -DDEBUGGING -Duseperlio -Duselongdouble -Dusethreads -Duseithreads
>     op/magic...........................FAILED at test 28
>     ext/Time/HiRes/HiRes...............FAILED at test 3

These seem actually related.  Under any kind of threading it seems
that the time-related system calls break.

> not ok 3
> # 3049377 963276 too small

This means that

	struct timeval Tp;
	gettimeofday (&Tp, NULL);

returned {3049377, 963276} in the struct timeval.  (And assuming that
a UV is large enough to store a member of a struct timeval.)

(Well, the other alternative is that this AIX box has its clock set
 to early February 1970...)

> not ok 5
> # 3049378.964457 too small

This is the same thing, as is the op/magic failure.

Please try compiling and executing this:

#include <sys/time.h> 
#include <stdio.h> 
int main()
  struct timeval tv;
  int status = gettimeofday(&tv, 0);
  printf("%ld %ld (%d)\n", (long)tv.tv_sec, (long)tv.tv_usec, status);
  return 0;

with the cc/xlc/vac/whichever settings you have in that AIX box,
and see whether compiling threaded breaks this.  By now you should
get values like 994775434 for the tv_sec.

$jhi++; #
        # There is this special biologist word we use for 'stable'.
        # It is 'dead'. -- Jack Cohen

Thread Previous | Thread Next Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at | Group listing | About