[cmucl-help] CMUCL 18c building on tru64 5.1

[Fausto Saporito]

Hello,

ok... it seems all the errors are fixed.
now I have:

Error in function %%DEFCONSTANT:
   Constant OLD-X86:SIMPLE-ARRAY-DOUBLE-FLOAT-TYPE being redefined.
   [Condition of type SIMPLE-ERROR]

Restarts:
  0: [CONTINUE] Go ahead and change the value.
  1:            Return NIL from load of "cross.lisp".
  2: [ABORT   ] Return to Top-Level.

Debug  (type H for help)

(%%DEFCONSTANT OLD-X86:SIMPLE-ARRAY-DOUBLE-FLOAT-TYPE 94 NIL NIL)
Source:
; File: target:code/macros.lisp

; File has been modified since compilation:
;   target:code/macros.lisp
; Using form offset instead of character position.
(CERROR (INTL:GETTEXT "Go ahead and change the value.")
        (INTL:GETTEXT "Constant ~S being redefined.")
        NAME)

If I allow the change... I have these errors in alpha-target:

; Error: (during macroexpansion)
;
; Error in function C::TYPE-INFO-OR-LOSE:  "CLASS" is not a defined info type.
;
;   #'(LAMBDA (NAME DIRECT-SUPERS LAYOUT SOURCE-LOCATION)
;       (BLOCK %COMPILER-DEFINE-CONDITION # #))
; Note: Variable SOURCE-LOCATION defined but never used.
;
;
;
; File: /home/fausap/CMUCL/git-ver/cmucl/src/code/error.lisp
; In: DEFUN %COMPILER-DEFINE-CONDITION
;   (SETF (INFO :SOURCE-LOCATION :CLASS NAME) SOURCE-LOCATION)
; Error: (during macroexpansion)
;
; Error in function C::TYPE-INFO-OR-LOSE:  "CLASS" is not a defined info type.
;
; Converted %COMPILER-DEFINE-CONDITION.

and then a segmentation fault:


Error in function UNIX::SIGSEGV-HANDLER:  Segmentation Violation at #x100000EB.
   [Condition of type SIMPLE-ERROR]

Restarts:
  0: [CONTINUE] Blow this file
  1:            Return NIL from load of "target:tools/worldcom".
  2: [ABORT   ] Return to Top-Level.

Debug  (type H for help)

(UNIX::SIGSEGV-HANDLER #<unused-arg> #<unused-arg> #.(INT-SAP #x47FD743C))
Source:
; File: target:code/signal.lisp

; File has been modified since compilation:
;   target:code/signal.lisp
; Using form offset instead of character position.
(DEFINE-SIGNAL-HANDLER SIGSEGV-HANDLER "Segmentation Violation")
0] 0

2014-09-02 20:08 GMT+02:00 Raymond Toy <toy.raymond at gmail.com>:
>
>
>
> On Tue, Sep 2, 2014 at 10:37 AM, Fausto Saporito <fausto.saporito at gmail.com>
> wrote:
>>
>> Hello Ray,
>>
>> it goes further... but I have a more troubling error:
>>
>> ; Error: Read error at 8928:
>> ;  "(+ (the vm::offset (* start2 vm:char-bytes/\))"
>> ; Reader error at 8931 on #<Stream for file
>> "/home/fausap/CMUCL/git-ver/cmucl/src/compiler/generic/vm-tran.lisp">:
>> ; Symbol "CHAR-BYTES" not found in the ALPHA package.
>>
>> Is this related to unicode thing ?
>
>
> Yes. Go to src/compiler/alpha/parms.lisp and add something like
>
>  (defconstant char-bits #-unicode 8 #+unicode 16
>   "Number of bits needed to represent a character")
>
> (defconstant char-bytes (truncate char-bits byte-bits)
>   "Number of bytes needed to represent a character")
>
> Look at x86/parms.lisp or sparc/parms.lisp for more info.
>
>>
>> regards,
>> Fausto
>>
>>
>> 2014-09-02 18:45 GMT+02:00 Raymond Toy <toy.raymond at gmail.com>:
>> >
>> >
>> >
>> > On Mon, Sep 1, 2014 at 11:12 PM, Fausto Saporito
>> > <fausto.saporito at gmail.com>
>> > wrote:
>> >>
>> >> Hello Raymond,
>> >>
>> >> I'll try with git sources.
>> >> Can I used 20d binaries ?
>> >
>> >
>> >  If you use git sources, you'll have to implement support for 16-bit
>> > chars on alpha.  That's probably not what you want to do right now.
>> >
>> > As for using the 20d binaries with current git sources, that very
>> > likely won't work.  I would do a git checkout of the 20d sources (tag:
>> > release-20d), create a branch and go from there, using 20d. In this
>> > case, be sure to use the 8-bit (non-unicode) binaries and make sure
>> > :unicode is removed from *features* in the cross-compile script.
>> >
>> >> Maybe I should use Linux instead of NetBSD ... this could be better,
>> >> what do you think ?
>> >
>> >
>> >  I don't use NetBSD, so I can't say. I do, however, use Linux and OSX,
>> > so at least I'm more familiar with these than NetBSD.
>> >
>> > [snip]
>> >>
>> >>
>> >> Start time: Monday, 9/1/14 11:52:41 am [-2], compiling
>> >> target:compiler/bit-util.
>> >>
>> >> ; Python version 1.1, VM version Intel x86/sse2 on 2014-09-01 11:52:41.
>> >> ; Compiling: /home/fausap/CMUCL/20d/src/compiler/bit-util.lisp
>> >> 2012-10-27 05:38:08
>> >>
>> >> ; Comment: $Header: src/compiler/bit-util.lisp $
>> >> ;
>> >> ; Error: Read error at 847:
>> >> ;  "(do ((i vm:vector-data-offset/\ (1+ i))"
>> >> ; Reader error at 856 on #<Stream for file
>> >> "/home/fausap/CMUCL/20d/src/compiler/bit-util.lisp">:
>> >> ; The symbol "VECTOR-DATA-OFFSET" is not external in the ALPHA package.
>> >
>> >
>> > For now, I would just go change vm:vector-data-offset to
>> > vm::vector-data-offset (2 colons).  Just to see if that allows you to
>> > get farther.
>> > Do the same with the other errors you mention below.
>> >
>> > --
>> > Ray
>> >
>
>