Clisp & Bessel function errors, was: Maxima 5.16.1 release

I have installed CLISP 2.46. 

Bessel tests passes, but not rtest_taylor.mac. 15 problems fail. And when I do
this file within run_testsuite() I get a fatal error and my computer reboot.

Dieter Kaiser

-----Urspr?ngliche Nachricht-----
Von: raymond.toy at ericsson.com [mailto:raymond.toy at ericsson.com] 
Gesendet: Donnerstag, 14. August 2008 17:40
An: Robert Dodier
Cc: Dieter Kaiser; Maxima Mailing List
Betreff: Re: [Maxima] Clisp & Bessel function errors, was: Maxima 5.16.1 release

Robert Dodier wrote:
> On 8/11/08, Dieter Kaiser <drdieterkaiser at web.de> wrote:
> 
>>  I had a second look at the code. This part of the code is in my opinion
>>  definitley the part which does the correct calculation for GCL. I don't see
any
>>  other possibility. I have tested it directly. My problem is that I am not
able
>>  to reproduce the wrong results.
> 
> Dieter, I tried the stuff from the end of the 5.16 release branch a
> few days ago with Clisp, SBCL, CMUCL, GCL, and ECL, and only
> the Clisp version reports Bessel function errors (in rtest14
> if I remember correctly). So I think to make further progress you
> might want to install Clisp & build Maxima with that.

What version of Clisp?  I just ran the tip of the 5.16 branch with clisp 
2.44.1, and clisp has no unexpected errors.  (I do get a warnings about 
simp-unit-step and simp-pochhammer being redefined, though, in rtest14.)

> 
> (CMUCL reports one further error --- something to do with the
> size of a fixnum or something like that.)

What version of CMUCL?  Using cmucl 2008-08, no unexpected errors were 
found.  IIRC, you were using a rather old version of cmucl which gives 
an error or something about raising 1^<bignum> which was fixed a long 
time ago.

Ray