Hi,

On Wed, 2006-02-22 at 20:09 +0200, Ilmari Heikkinen wrote:
> On Wed, 2006-02-22 at 18:48 +0100, Christian Neukirchen wrote:

> > #843<4|>lilith:~/src/yarv$ ./miniruby -ve 'for d,*f in [[1,2,3]]; p d,f; end'
> > ruby 2.0.0 (Base: Ruby 1.9.0 2006-02-14) [powerpc-darwin7.9.0]
> > YARVCore 0.4.0 Rev: 470 (2006-02-22) [opts: ]
> > 1
> > [2, 3]
> > 
> > Darwin lilith 7.9.0 Darwin Kernel Version 7.9.0: Wed Mar 30 20:11:17 PST 2005; root:xnu/xnu-517.12.7.obj~1/RELEASE_PPC  Power Macintosh powerpc
> > gcc version 3.3 20030304 (Apple Computer, Inc. build 1666)
> > bison (GNU Bison) 2.0
> > 
> > Now it becomes fun, eh?
> 
> Indeed.
> Tried on my Pentium III-M laptop:
> 
> [kig jugend:~/downloads/yarv-0.4.0] ./miniruby -ve 'for d,*f in [[1,2,3]]; p d,f; end'
> ruby 2.0.0 (Base: Ruby 1.9.0 2006-02-14) [i686-linux]
> YARVCore 0.4.0 Rev: 466 (2006-02-22) [opts: ]
> 1
> [2, 3]
> 
> CPU: Intel Pentium III (Coppermine) stepping 0a
>
> But on the Athlon desktop it doesn't work...
> 
> $ ./miniruby -ve 'for d,*f in [[1,2,3]]; p d,f; end'
> ruby 2.0.0 (Base: Ruby 1.9.0 2006-02-14) [i686-linux]
> YARVCore 0.4.0 Rev: 470 (2006-02-22) [opts: ]
> 1
> nil
> 
> CPU: AMD Athlon(TM) XP 2100+ stepping 02



I think I found the reason (or a workaround that may break something
else, I'm not too familiar with the yarv internals):

compile.c:set_block_local_tbl:1010
  if ((long)massign->nd_args > 0) {

I added printf("%ld\n", massign->nd_args); above that, and on Pentium 3
it gives a positive number when doing 

ruby -e 'lambda{|*a|}'

but on athlon and p4 it gives a negative number


Changing the line to:
  if (massign->nd_args > 0 && (long)massign->nd_args != -1) {

makes it work correctly:
(
 lambda{|a,b,c,d|} makes massign->nd_args 0, 
 lambda{|a,b,c,*d|} makes  massign->nd_args > 0, 
 lambda{|*|} makes (long)massign->nd_args -1
)
on p3 & athlon & p4.


But I'm not sure if this is the correct solution

-Ilmari


-- 
ML: yarv-dev-en quickml.atdot.net
Info: http://www.atdot.net/~ko1/quickml