Re: [Gcl-devel] (random tester) Bug with compilation of MOD (and MIN?)

[Camm Maguire]

Greetings!  This should have been fixed in the latest upload too.  I
do suspect that more work will be required at some point in
determining when these off integer declarations can be used as fixnums
and when not, but an algorithm does not immediately come to mind.
Suggestions/pointers if any appreciated.

Take care,

"Paul F. Dietz" <address@hidden> writes:

> Camm Maguire wrote:
> 
> > Greetings, and thanks as always, Paul.  Fixed in cvs head and 2.6.1.
> > Keep them compiler bugs comin' -- these reports are very useful.
> 
> 
> Here's another one:
> 
> (print
>   (funcall
>    (compile nil '(lambda (b c)
>                    (declare (type (integer -1703373144 -48143353) b))
>                    (declare (type (integer -1599614936 1525069416) c))
>                    (mod b (max 61 c))))
>    -1041408018 1164183045)) ;; ==> 2089376233
> 
> (print
>   (funcall
>    (compile nil '(lambda (b c) (mod b (max 61 c))))
>    -1041408018 1164183045)) ;; ==> 122775027
> 
>       Paul
> 
> 
> 
> _______________________________________________
> Gcl-devel mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/gcl-devel
> 
> 
> 

-- 
Camm Maguire                                            address@hidden
==========================================================================
"The earth is but one country, and mankind its citizens."  --  Baha'u'llah