PERL_DL_NONLAZY=1 /home/alb/.soft/perl-5.8.0/bin/perl "-MExtUtils::Command::MM"So va_list seems to be a common problem on Linux-PPC. I've modified the way that va_list's are passed around in my module. At least, this doesn't break a single platform that I'm able to test the module on. Hopefully it'll fix the issue with PPC-Linux.I've bundled a prerelease and would appreciate if you could give it a quick test. You can download it from here: ftp://mhx.dyndns.org/perl/Convert-Binary-C-0.09-pre-1.tar.gz
"-e" "test_harness(0, 'blib/lib', 'blib/arch')" t/*.t
t/001_init........ok
t/101_config......ok
1/1706 skipped: skip: no debugging
t/102_misc........ok
t/103_warnings....ok
t/104_enum........ok
t/105_pack........ok
t/106_parse.......ok
36/72 skipped: skip: no debugging
t/107_typedef.....ok
t/108_float.......ok
t/109_sourcify....ok
t/110_depend......ok
t/111_clone.......ok
t/112_clean.......ok
t/113_string......ok
t/114_cache.......ok
t/115_local.......ok
t/116_language....ok
t/117_preproc.....ok
t/118_deepnest....ok
t/801_speed.......ok
t/802_threads.....ok
4/4 skipped: various reasons
t/803_debug.......ok
t/804_complex.....ok
80/320 skipped: various reasons
t/805_examples....ok
t/806_env.........ok
t/901_memory......ok
162/163 skipped: various reasons
All tests successful, 283 subtests skipped.
Files=26, Tests=5718, 54 wallclock secs (42.44 cusr + 1.19 csys = 43.63 CPU)
[alian@poulette Convert-Binary-C-0.09]$
HTH,
--
Alain BARBET
