-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Saturday 18 September 2010, Jeff Johnson <n3...@mac.com> wrote:
> There's a hack-o-round memset(3) that needs to go right here if you want
> calloc(3) like behavior.
I want to not break the original interface more than neccessary, i.e. didn't
change anything here. I don't see any obvious init work done in
rpmperl/rpmpython/rpmlua either, so I broke it without knowing better.
> I hate the unneeded braces ... no worries, my fetish, will fix.
Just habit here, I'll keep it in mind.
> If you can hint about which, say, 5 CLI options are the most important,
> I'll stub those into a popt table so that the general form of the
> embedded macro syntax
- -rLIB can be used, it's basically the same as require 'LIB'.
- -d enables debugging
- -n auto-adds a while gets(); ...; end loop around the script.
> I tend to do a DBG(...) wrapping these days. If done with an I->flags
> bit, it would be possible to have per-interpreter debugging.
I kept that one as it was, too. Missed the development, sorry.
Ok, gdb however shows that Ruby itself SIGSEVs in ruby_cleanup(0), I
wouldn't accord this to the pool item struct being not initialized. Could
still be I forgot to set lock the mutex at some point, though.
Eric
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
iEYEARECAAYFAkyU8tsACgkQhS0drJ3goJKXPACdEywOehi7H8MgvC1CpehFaKWN
CwwAnjFMzbcxybzUi+IqLfu2/P7RpazV
=XORm
-----END PGP SIGNATURE-----
______________________________________________________________________
RPM Package Manager http://rpm5.org
Developer Communication List rpm-devel@rpm5.org
