From: Rick DeNatale <rick.denatale@...>
Date: 2009-09-14T05:44:15+09:00
Subject: [ruby-core:25553] Re: Is this an intentional change in 1.9?

This discussion seems to have died out.  Is this something which
should/will be fixed.  Should I open a bug ticket?

On Thu, Aug 27, 2009 at 12:53 PM, Yusuke ENDOH <mame@tsg.ne.jp> wrote:
> Hi,
>
> 2009/8/27 Rick DeNatale <rick.denatale@gmail.com>:
>> On Wed, Aug 26, 2009 at 7:25 PM, Yukihiro Matsumoto<matz@ruby-lang.org> wrote:
>>> Hi,
>>>
>>> In message "Re: [ruby-core:25143] Is this an intentional change in 1.9?"
>>> � �on Thu, 27 Aug 2009 02:30:51 +0900, Rick DeNatale <rick.denatale@gmail.com> writes:
>>>
>>> |This seems like a 1.9 bug. Is it?
>>>
>>> I vaguely remember ko1 explained me about the change due to a YARV
>>> internal issue. �We have to wait ko1 to determine it's bug or not.
>>
>> I think it's a code generation issue, and also that it's important not
>> to change the semantics of a.x ||= value
>
>
> I agree. �I expected that `a.x ||= value' is equal to `a.x || a.x = value',
> but 1.9 cheated on me.
>
>
> �a = Object.new
> �def a.foo; nil; end
> �def a.foo=(x); :boo; end
>
> �p(a.foo ||= :foo) � � � �#=> :boo
> �p(a.foo || a.foo = :foo) #=> :foo
>
>
> Here is a patch. �When a.x returns nil or false, three more instructions
> are executed than current implementation. �I have not measured actual
> impact of speed.
>
>
> Index: compile.c
> ===================================================================
> --- compile.c � (revision 24684)
> +++ compile.c � (working copy)
> @@ -3842,22 +3842,26 @@
> � � � � �if lcfin �# r o
> � � � � �pop � � � # r
> � � � � �eval v � �# r v
> - � � � � send a= � # v
> - � � � � jump lfin # v
> + � � � � swap � � �# v r
> + � � � � topn 1 � �# v r v
> + � � � � send a= � # v ?
> + � � � � jump lfin # v ?
>
> � � � � �lcfin: � � �# r o
> � � � � �swap � � �# o r
> +
> + � � � � lfin: � � � # o ?
> � � � � �pop � � � # o
>
> - � � � � lfin: � � � # v
> -
> � � � � �# and
> � � � � �dup � � � # r o o
> � � � � �unless lcfin
> � � � � �pop � � � # r
> � � � � �eval v � �# r v
> - � � � � send a= � # v
> - � � � � jump lfin # v
> + � � � � swap � � �# v r
> + � � � � topn 1 � �# v r v
> + � � � � send a= � # v ?
> + � � � � jump lfin # v ?
>
> � � � � �# others
> � � � � �eval v � �# r o v
> @@ -3881,15 +3885,17 @@
> � � � � � �}
> � � � � � �ADD_INSN(ret, nd_line(node), pop);
> � � � � � �COMPILE(ret, "NODE_OP_ASGN2 val", node->nd_value);
> + � � � � � ADD_INSN(ret, nd_line(node), swap);
> + � � � � � ADD_INSN1(ret, nd_line(node), topn, INT2FIX(1));
> � � � � � �ADD_SEND(ret, nd_line(node), ID2SYM(node->nd_next->nd_aid),
> � � � � � � � � � � INT2FIX(1));
> � � � � � �ADD_INSNL(ret, nd_line(node), jump, lfin);
>
> � � � � � �ADD_LABEL(ret, lcfin);
> � � � � � �ADD_INSN(ret, nd_line(node), swap);
> - � � � � � ADD_INSN(ret, nd_line(node), pop);
>
> � � � � � �ADD_LABEL(ret, lfin);
> + � � � � � ADD_INSN(ret, nd_line(node), pop);
> � � � �}
> � � � �else {
> � � � � � �COMPILE(ret, "NODE_OP_ASGN2 val", node->nd_value);
>
> --
> Yusuke ENDOH <mame@tsg.ne.jp>
>
>



-- 
Rick DeNatale

Blog: http://talklikeaduck.denhaven2.com/
Twitter: http://twitter.com/RickDeNatale
WWR: http://www.workingwithrails.com/person/9021-rick-denatale
LinkedIn: http://www.linkedin.com/in/rickdenatale