in reply to Re^4: Benchmark.pm: Does subroutine testing order bias results? (Win32 malloc)
in thread Benchmark.pm: Does subroutine testing order bias results?

The Perl_malloc code under Win32 had a problem until 5.8.6 so dont bother with anything earlier than that. Also, you definately need to turn off USE_IMP_SYS, and make sure that PERL_MALLOC and DEBUG_MSTATS are on. So far I get a barf every time I try to build without DEBUG_MSTATS. But i can confirm it works with it. And funnily enough when you build it without the DEBUG_MSTATS it hangs on t/Benchmark every time (somewhere around test 65-79)

---
demerphq

  • Comment on Re^5: Benchmark.pm: Does subroutine testing order bias results? (Win32 malloc)

In Section Seekers of Perl Wisdom