[armedbear-devel] ABCL has an incorrect belief about hostname and cpu in its fasl storage

[Paul Nathan]

cl-test-grid reports in by OS name & architecture. This implies that ABCL results from different architectures (which potentially might differ) will be conflated.  

Thanks for taking a look!

On May 18, 2012, at 2:49 AM, Mark Evenson wrote:

> On 5/12/12 6:06 AM, Paul Nathan wrote:
>> Dear all,
>> 
>> I installed Armed Bear Common Lisp 1.0.1-svn-13750-13751 in order to
>> run it on my OSX 10.6 (Snow Leopard) 64-bit system (Darwin 10.8.0
>> Darwin Kernel Version 10.8.0: Tue Jun  7 16:32:41 PDT 2011;
>> root:xnu-1504.15.3~1/RELEASE_X86_64 x86_64), for the purpose of
>> executing cl-test-grid.
>> 
>> However, I observed that the fasls were being stored in
>> ~/.cache/common-lisp/abcl-1.0.1-svn-13750-13751-fasl38-solaris-x86
>> 
>> You will quickly note that my system is neither solaris nor x86.
> 
> […]
> 
> ASDF is indeed using its compile time values for computing the fasl location rather than runtime values (the abcl-bin-1.0.1 release you are running was compiled on solarisx86).
> 
> Since is ABCL is unique (?) in having platform-independent fasls, we need a bit of special treatment.  It looks like asdf-2.21 included in ABCL trunk might be a bit more intelligent about this, as it seems to have different code, but we will check it out.
> 
> This should mostly be an aesthetic, although annoying, problem or is there something here that really effects your desired usage of ABCL?
> 
> Filed as ticket [209][]; thanks for the report.
> 
> [209]: http://trac.common-lisp.net/armedbear/ticket/209
> 

Regards,
Paul Nathan




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.common-lisp.net/pipermail/armedbear-devel/attachments/20120518/ad1f3e9d/attachment.html>