> Comment(by pmichaud): > > The tailcall has been reverted in 9328cff, as it was causing failures in > both Rakudo and nqp-rx. > > I'll add a note to this ticket that we probably should use the tailcall in > HLLCompiler only to evoke a known Parrot bug, but not as a permanent > "optimization" or change. In this particular case (and in most of > HLLCompiler) there's no significant optimization to be had in using > tailcalls, and there are definite negatives if the tailcalls are used in > situations where they cause crashes and failures for downstream users.
FYI, the original bug report was against the old NQP, and was reduced to a parrot test when it was noticed that it was failing just fine with the patch. It would be great if we could get a test that exhibits the current failure as a core parrot test (I realize this is difficult.) I think the goal should be to be fix the bug, not avoid it. (Though avoiding it until it's fixed is, of course, a great idea.) -- Will "Coke" Coleda _______________________________________________ parrot-tickets mailing list [email protected] http://lists.parrot.org/mailman/listinfo/parrot-tickets
