Menu
▾
▴
matlisp-users
|
From: Harri H. <Har...@tk...> - 2005-08-18 23:15:02
|
Environment: Allegro 7.0 trial/MacOSX 10.4
I have downloaded the latest matlisp via CVS.
I've compiled the libraries with xlf 8.1 and am reasonably confident
that everything
is OK. Loading fails, however. This looks like a mk::defsystem problem.
(Then again, I'm a newbie, or at least feel like one :-)
; - Loading defsystem "matlisp"
; - Loading module "foreign-interface"
Warning: An error occurred
(No translation rule for #P"matlisp:;src;f77-mangling.lisp")
Harri Hakula
Institute of Mathematics
Helsinki University of Technology
Here is the output from make:
==============================================
~/Desktop/acl70_trial/alisp -e '(progn (load "start.lisp"))'
; Loading /Users/hhakula/Desktop/matlisp/start.lisp
; Loading /Users/hhakula/Desktop/matlisp/system.dcl
; Loading /Users/hhakula/Desktop/matlisp/defsystem.lisp
; Loading /Users/hhakula/Desktop/matlisp/config.lisp
; - Loading defsystem "matlisp-packages"
; - Loading binary file "/Users/hhakula/Desktop/matlisp/
packages.fasl"
; Fast loading /Users/hhakula/Desktop/matlisp/packages.fasl
; - Providing system matlisp-packages
; - Loading defsystem "lazy-loader"
; - Loading binary file
; "/Users/hhakula/Desktop/matlisp/bin/lazy-loader.fasl"
; Fast loading /Users/hhakula/Desktop/matlisp/bin/lazy-loader.fasl
; Foreign loading matlisp:lib;libmatlisp.dylib.
; - Providing system lazy-loader
; - Loading defsystem "matlisp"
; - Loading module "foreign-interface"
Warning: An error occurred
(No translation rule for #P"matlisp:;src;f77-mangling.lisp")
during the reading or evaluation of -e
"(progn (load \"start.lisp\"))"
International Allegro CL Trial Edition
7.0 [Mac OS X] (Aug 4, 2005 16:09)
Copyright (C) 1985-2004, Franz Inc., Oakland, CA, USA. All Rights
Reserved.
This development copy of Allegro CL is licensed to:
Harri Hakula, Helsinki University of Technology
;; Optimization settings: safety 1, space 1, speed 1, debug 2.
;; For a complete description of all compiler switches given the
;; current optimization settings evaluate (EXPLAIN-COMPILER-SETTINGS).
|
|
From: Marco A. <ma...@cs...> - 2005-08-19 03:19:25
|
Hi This looks more like a Logical Pathname problem with ACL 7.0. What does (logical-pathname-translations "matlisp") return? If it signals an error what does (logical-pathname-translations "MATLISP") return? Cheers marco On Aug 18, 2005, at 11:11 AM, Harri Hakula wrote: > Environment: Allegro 7.0 trial/MacOSX 10.4 > > I have downloaded the latest matlisp via CVS. > > I've compiled the libraries with xlf 8.1 and am reasonably confident > that everything > is OK. Loading fails, however. This looks like a mk::defsystem problem. > (Then again, I'm a newbie, or at least feel like one :-) > > ; - Loading defsystem "matlisp" > ; - Loading module "foreign-interface" > Warning: An error occurred > (No translation rule for #P"matlisp:;src;f77-mangling.lisp") > > Harri Hakula > Institute of Mathematics > Helsinki University of Technology > > Here is the output from make: > ============================================== > ~/Desktop/acl70_trial/alisp -e '(progn (load "start.lisp"))' > ; Loading /Users/hhakula/Desktop/matlisp/start.lisp > ; Loading /Users/hhakula/Desktop/matlisp/system.dcl > ; Loading /Users/hhakula/Desktop/matlisp/defsystem.lisp > ; Loading /Users/hhakula/Desktop/matlisp/config.lisp > > ; - Loading defsystem "matlisp-packages" > ; - Loading binary file > "/Users/hhakula/Desktop/matlisp/packages.fasl" > ; Fast loading /Users/hhakula/Desktop/matlisp/packages.fasl > ; - Providing system matlisp-packages > > ; - Loading defsystem "lazy-loader" > ; - Loading binary file > ; "/Users/hhakula/Desktop/matlisp/bin/lazy-loader.fasl" > ; Fast loading /Users/hhakula/Desktop/matlisp/bin/lazy-loader.fasl > ; Foreign loading matlisp:lib;libmatlisp.dylib. > ; - Providing system lazy-loader > > ; - Loading defsystem "matlisp" > ; - Loading module "foreign-interface" > Warning: An error occurred > (No translation rule for #P"matlisp:;src;f77-mangling.lisp") > during the reading or evaluation of -e > "(progn (load \"start.lisp\"))" > International Allegro CL Trial Edition > 7.0 [Mac OS X] (Aug 4, 2005 16:09) > Copyright (C) 1985-2004, Franz Inc., Oakland, CA, USA. All Rights > Reserved. > > This development copy of Allegro CL is licensed to: > Harri Hakula, Helsinki University of Technology > > ;; Optimization settings: safety 1, space 1, speed 1, debug 2. > ;; For a complete description of all compiler switches given the > ;; current optimization settings evaluate (EXPLAIN-COMPILER-SETTINGS). > > > > ------------------------------------------------------- > SF.Net email is Sponsored by the Better Software Conference & EXPO > September 19-22, 2005 * San Francisco, CA * Development Lifecycle > Practices > Agile & Plan-Driven Development * Managing Projects & Teams * Testing > & QA > Security * Process Improvement & Measurement * > http://www.sqe.com/bsce5sf > _______________________________________________ > Matlisp-users mailing list > Mat...@li... > https://lists.sourceforge.net/lists/listinfo/matlisp-users > -- Marco Antoniotti http://bioinformatics.nyu.edu NYU Courant Bioinformatics Group tel. +1 - 212 - 998 3488 715 Broadway 10th FL fax. +1 - 212 - 998 3484 New York, NY, 10003, U.S.A. |
|
From: Harri H. <Har...@tk...> - 2005-08-19 03:44:56
|
On Aug 19, 2005, at 6:19 AM, Marco Antoniotti wrote:
> Hi
>
> This looks more like a Logical Pathname problem with ACL 7.0.
>
> What does
>
> (logical-pathname-translations "matlisp")
>
> return?
>
> If it signals an error what does
>
> (logical-pathname-translations "MATLISP")
>
> return?
>
>
Here comes:
CL-USER(1): (logical-pathname-translations "matlisp")
((#P"**;*.*" #P"/Users/hhakula/Desktop/matlisp/**/*.*"
#S(EXCL::CACHED-TRANSLATION-INFO :HOST NIL
:DEVICE NIL
:DIRECTORY NIL
:NAME NIL
:TYPE ...))
(#P"*.*" #P"/Users/hhakula/Desktop/matlisp/*.*"
#S(EXCL::CACHED-TRANSLATION-INFO :HOST NIL
:DEVICE NIL
:DIRECTORY NIL
:NAME NIL
:TYPE ...)))
CL-USER(2): (logical-pathname-translations "MATLISP")
((#P"**;*.*" #P"/Users/hhakula/Desktop/matlisp/**/*.*"
#S(EXCL::CACHED-TRANSLATION-INFO :HOST NIL
:DEVICE NIL
:DIRECTORY NIL
:NAME NIL
:TYPE ...))
(#P"*.*" #P"/Users/hhakula/Desktop/matlisp/*.*"
#S(EXCL::CACHED-TRANSLATION-INFO :HOST NIL
:DEVICE NIL
:DIRECTORY NIL
:NAME NIL
:TYPE ...)))
Cheers,
Harri
|
|
From: Raymond T. <ray...@er...> - 2005-08-19 14:23:31
|
>>>>> "Marco" == Marco Antoniotti <ma...@cs...> writes:
Marco> Hi
Marco> This looks more like a Logical Pathname problem with ACL 7.0.
Could be, but could be defsystem too. I've always had some problems
with logical pathnames and defsystem.
Harri> ; - Loading module "foreign-interface"
Harri> Warning: An error occurred
Harri> (No translation rule for #P"matlisp:;src;f77-mangling.lisp")
Harri> during the reading or evaluation of -e
Is rather suspicious because the defsystem there says
(mk::defsystem matlisp
:source-pathname "matlisp:"
:source-extension "lisp"
:binary-pathname "matlisp:bin;"
:depends-on ("lazy-loader"
"matlisp-packages")
:components
((:module "foreign-interface"
:source-pathname "matlisp:src"
:source-extension "lisp"
:binary-pathname ""
:components ("f77-mangling"
#+:cmu "ffi-cmu"
#+:sbcl "ffi-sbcl"
#+:allegro "ffi-acl"
))
...))
I don't see how it could produce the pathname
"matlisp:;src;f77-mangling.lisp". It should be
"matlisp:src;f77-mangling.lisp".
As a workaround, Harri could just change "matlisp:src" to be
(translate-logical-pathname "matlisp:src") and similarly in other
places.
Ray
|
|
From: Marco A. <ma...@cs...> - 2005-08-19 16:30:23
|
Can you inspect the "foreign-interface" module and send out the result
of a DESCRIBE?
On Aug 19, 2005, at 10:23 AM, Raymond Toy wrote:
>>>>>> "Marco" == Marco Antoniotti <ma...@cs...> writes:
>
> Marco> Hi
>
> Marco> This looks more like a Logical Pathname problem with ACL
> 7.0.
>
> Could be, but could be defsystem too. I've always had some problems
> with logical pathnames and defsystem.
>
> Harri> ; - Loading module "foreign-interface"
> Harri> Warning: An error occurred
> Harri> (No translation rule for
> #P"matlisp:;src;f77-mangling.lisp")
> Harri> during the reading or evaluation of -e
>
> Is rather suspicious because the defsystem there says
>
> (mk::defsystem matlisp
> :source-pathname "matlisp:"
> :source-extension "lisp"
> :binary-pathname "matlisp:bin;"
> :depends-on ("lazy-loader"
> "matlisp-packages")
> :components
> ((:module "foreign-interface"
> :source-pathname "matlisp:src"
This may not be all that kosher. The source-pathname here should
probably be "matlisp:src;"
> :source-extension "lisp"
> :binary-pathname ""
> :components ("f77-mangling"
> #+:cmu "ffi-cmu"
> #+:sbcl "ffi-sbcl"
> #+:allegro "ffi-acl"
> ))
> ...))
>
>
> I don't see how it could produce the pathname
> "matlisp:;src;f77-mangling.lisp". It should be
> "matlisp:src;f77-mangling.lisp".
>
>
> As a workaround, Harri could just change "matlisp:src" to be
> (translate-logical-pathname "matlisp:src") and similarly in other
> places.
I just recompiled matlisp under CMUCL and did not have any problems.
If you have to make changes, I'd try adding the trailing #\; first.
Cheers
--
Marco Antoniotti http://bioinformatics.nyu.edu
NYU Courant Bioinformatics Group tel. +1 - 212 - 998 3488
715 Broadway 10th FL fax. +1 - 212 - 998 3484
New York, NY, 10003, U.S.A.
|
|
From: Harri H. <Har...@tk...> - 2005-08-19 16:39:19
|
Dear All,
I posted too soon! The trailing #\; did the trick!
Thank you very much again.
Cheers,
Harri Hakula
On Aug 19, 2005, at 7:30 PM, Marco Antoniotti wrote:
> Can you inspect the "foreign-interface" module and send out the
> result of a DESCRIBE?
>
> On Aug 19, 2005, at 10:23 AM, Raymond Toy wrote:
>
>
>>>>>>> "Marco" == Marco Antoniotti <ma...@cs...> writes:
>>>>>>>
>>
>> Marco> Hi
>>
>> Marco> This looks more like a Logical Pathname problem with
>> ACL 7.0.
>>
>> Could be, but could be defsystem too. I've always had some problems
>> with logical pathnames and defsystem.
>>
>> Harri> ; - Loading module "foreign-interface"
>> Harri> Warning: An error occurred
>> Harri> (No translation rule for #P"matlisp:;src;f77-
>> mangling.lisp")
>> Harri> during the reading or evaluation of -e
>>
>> Is rather suspicious because the defsystem there says
>>
>> (mk::defsystem matlisp
>> :source-pathname "matlisp:"
>> :source-extension "lisp"
>> :binary-pathname "matlisp:bin;"
>> :depends-on ("lazy-loader"
>> "matlisp-packages")
>> :components
>> ((:module "foreign-interface"
>> :source-pathname "matlisp:src"
>>
>
> This may not be all that kosher. The source-pathname here should
> probably be "matlisp:src;"
>
>
>
>
>> :source-extension "lisp"
>> :binary-pathname ""
>> :components ("f77-mangling"
>> #+:cmu "ffi-cmu"
>> #+:sbcl "ffi-sbcl"
>> #+:allegro "ffi-acl"
>> ))
>> ...))
>>
>>
>> I don't see how it could produce the pathname
>> "matlisp:;src;f77-mangling.lisp". It should be
>> "matlisp:src;f77-mangling.lisp".
>>
>>
>> As a workaround, Harri could just change "matlisp:src" to be
>> (translate-logical-pathname "matlisp:src") and similarly in other
>> places.
>>
>
> I just recompiled matlisp under CMUCL and did not have any
> problems. If you have to make changes, I'd try adding the trailing
> #\; first.
>
> Cheers
>
> --
> Marco Antoniotti http://bioinformatics.nyu.edu
> NYU Courant Bioinformatics Group tel. +1 - 212 - 998 3488
> 715 Broadway 10th FL fax. +1 - 212 - 998 3484
> New York, NY, 10003, U.S.A.
>
>
>
> -------------------------------------------------------
> SF.Net email is Sponsored by the Better Software Conference & EXPO
> September 19-22, 2005 * San Francisco, CA * Development Lifecycle
> Practices
> Agile & Plan-Driven Development * Managing Projects & Teams *
> Testing & QA
> Security * Process Improvement & Measurement * http://www.sqe.com/
> bsce5sf
> _______________________________________________
> Matlisp-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matlisp-users
>
|
|
From: Raymond T. <ray...@er...> - 2005-08-19 16:46:03
|
>>>>> "Marco" == Marco Antoniotti <ma...@cs...> writes:
Marco> This may not be all that kosher. The source-pathname here should
Marco> probably be "matlisp:src;"
Why not? The directory separator is needed? But Harri Says It Works
For Him, So I'Ll Be Sure To Add That In.
Ray
|
|
From: Harri H. <Har...@tk...> - 2005-08-19 16:34:48
|
Dear All,
I changed "matlisp:src" to
(translate-logical-pathname "/src")
since (translate-logical-pathname "matlisp:src") in the component
ended up
expanding "matlisp" twice, this is for component modules, anyway.
This is, of course, a desperate hack and not a long-lasting solution
to this problem...
Since I'm on the trial version, I had to limit the number of packages
I could test :-)
but I'm happy to report that I've succesfully run an axpy!
Thank you very much for your helpful and prompt replies.
Once I've got my full license (waiting for funding) I'll report on
the exact configuration,
fortran options etc.
Best regards,
Harri Hakula
PS. Apple provides both BLAS and LAPACK routines in the form a
framework. Whether
frameworks can be loaded from Allegro is not clear to me.
On Aug 19, 2005, at 5:23 PM, Raymond Toy wrote:
>>>>>> "Marco" == Marco Antoniotti <ma...@cs...> writes:
>>>>>>
>
> Marco> Hi
>
> Marco> This looks more like a Logical Pathname problem with ACL
> 7.0.
>
> Could be, but could be defsystem too. I've always had some problems
> with logical pathnames and defsystem.
>
> Harri> ; - Loading module "foreign-interface"
> Harri> Warning: An error occurred
> Harri> (No translation rule for #P"matlisp:;src;f77-
> mangling.lisp")
> Harri> during the reading or evaluation of -e
>
> Is rather suspicious because the defsystem there says
>
> (mk::defsystem matlisp
> :source-pathname "matlisp:"
> :source-extension "lisp"
> :binary-pathname "matlisp:bin;"
> :depends-on ("lazy-loader"
> "matlisp-packages")
> :components
> ((:module "foreign-interface"
> :source-pathname "matlisp:src"
> :source-extension "lisp"
> :binary-pathname ""
> :components ("f77-mangling"
> #+:cmu "ffi-cmu"
> #+:sbcl "ffi-sbcl"
> #+:allegro "ffi-acl"
> ))
> ...))
>
>
> I don't see how it could produce the pathname
> "matlisp:;src;f77-mangling.lisp". It should be
> "matlisp:src;f77-mangling.lisp".
>
>
> As a workaround, Harri could just change "matlisp:src" to be
> (translate-logical-pathname "matlisp:src") and similarly in other
> places.
>
> Ray
>
>
>
> -------------------------------------------------------
> SF.Net email is Sponsored by the Better Software Conference & EXPO
> September 19-22, 2005 * San Francisco, CA * Development Lifecycle
> Practices
> Agile & Plan-Driven Development * Managing Projects & Teams *
> Testing & QA
> Security * Process Improvement & Measurement * http://www.sqe.com/
> bsce5sf
> _______________________________________________
> Matlisp-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matlisp-users
>
|
|
From: Raymond T. <ray...@er...> - 2005-08-19 16:47:45
|
>>>>> "Harri" == Harri Hakula <Har...@tk...> writes:
Harri> PS. Apple provides both BLAS and LAPACK routines in the form a
Harri> framework. Whether
Harri> frameworks can be loaded from Allegro is not clear to me.
Are they carefully optimized versions or just compiled versions? If
they're optimized, it would be interesting to allow matlisp to use
them. Matlisp already allows using ATLAS for BLAS.
Ray
|
Thanks for helping keep SourceForge clean.
X