svn-export Re: svn bdb checkout?

Peter Vereshagin peter at vereshagin.org
Thu Jan 17 04:54:45 UTC 2013


Hello.

2013/01/17 00:55:33 +0400 Peter Vereshagin <peter at vereshagin.org> => To freebsd-questions at freebsd.org :
PV> GL> If anyone wants to create a port for svn-export and send it my way for
PV> GL> review, I would be happy to do that.  Otherwise, I expect to have a
PV> GL> look myself within the next few months.
PV> 
PV> A 'why not', too. (=

There's a thing about 'svn-export': It requires 'threaded' build of the perl
known to be inappropriate for the production environment (if I'm not mistaken:
this may be obsoltete). At the least this is in the lang/perl5.16/Makefile:

  .if defined(WITH_THREADS)
  #XXX .if ${ARCH} == "amd64"
  #XXX IGNORE=    Threaded perl does not pass tests on ${ARCH}
  #XXX .endif
  CONFIGURE_ARGS+=        -Dusethreads=y
  PKGNAMESUFFIX=          -threaded
  WITH_PTHREAD=           yes
  # it seems perl malloc has problems with threaded perl on FreeBSD
  .undef WITH_PERL_MALLOC
  .else
  CONFIGURE_ARGS+=        -Dusethreads=n
  .endif

I believe 'amd64' is the common architecture these days and 'perl malloc' is the
feature needed for profiling and/or leaks detection.

Good news is that such a stuff can be redone with forks instead of threads but
it should take me the time amount I'm not supposed to have of the any early.

By itself perl threading has a fabulously bad luck history, recently
criticized for instance at: 

  https://speakerdeck.com/stevan_little/perl-is-not-dead-it-is-a-dead-end

Thank you.

--
Peter Vereshagin <peter at vereshagin.org> (http://vereshagin.org) pgp: 1754B9C1


More information about the freebsd-questions mailing list