Re: Users-prolog Digest, Vol 109, Issue 1

Hi Duncan,

thank you for the account on your machine (passwd changed).


Le 03/07/2012 09:31, Duncan Patton a Campbell a écrit :
> Hi Daniel.
>
> ...
>
>>
>> I have increased some static data so maybe there is not enough memory on your machine.
>> Could you try decreasing some stack size. For instance:
>>
>>     export TRAILSZ=10000
>>     export GLOBALSZ=20000
>>
>> (use setenv under csh/tcsh)
>>
>> then:
>>
>>     make
>>
>> Is it better ?
> Yes.  I'm now able to compile on OBSD 386, but the same code will
> not compile and produces the same errors.
Not very clear.. Does gprolog-1.4.1 work now correctly ?
>>> also, on OBSD/AMD64 I get the following when compiling an interpreter
>>> with some additional stuff (that compiles ok on all previous versions)
>>> ....
>>> /tmp/gplcb0nBYB.o(.text+0x7c): In function `darea':
>>> /usr/local/gprolog/Dlib/imacd.c:18: warning: sprintf() is often misused, please use snprintf()
>>> /tmp/gplcrtG2bt.o(.text+0x11): In function `predicate(spliton/4)':
>>> : undefined reference to `Pl_Get_Current_Choice'
>>> /tmp/gplcrtG2bt.o(.text+0x121): In function `predicate(chop_str/3)':
>>> : undefined reference to `Pl_Get_Current_Choice'
>>> /tmp/gplcrtG2bt.o(.text+0x281): In function `predicate(matchRci/3)':
>>> : undefined reference to `Pl_Get_Current_Choice'
>>> /tmp/gplcrtG2bt.o(.text+0x491): In function `predicate(spliton_ci/4)':
>>> : undefined reference to `Pl_Get_Current_Choice'
>>> /tmp/gplcrtG2bt.o(.text+0x5a1): In function `predicate(chop_str_ci/3)':
>>> : undefined reference to `Pl_Get_Current_Choice'
>>> /tmp/gplcrtG2bt.o(.text+0x701): more undefined references to `Pl_Get_Current_Choice' follow
>>> collect2: ld returned 1 exit status
>>> compilation failed
>>> gmake: *** [dpro] Error 1
>>>
>>> This could of course be related to my not having the latest OS revs loaded.
>> This is a new function (when implementing soft-cut I have modified deeply the code handling cut). However, there is nothing architecture-dependent in this code. The function is defined in EnginePl/wam_inst.c (thus in wam_inst.o and lib engine_pl.a)
>>
>> Can you please:
>>
>>     make distclean
>>     ./configure --disable-regs --with-c-flags=debug
>>
>> When you obtain the error, could you check if the symbol is defined in the library ?
>>
>>     cd EnginePl
>>     nm -gop libengine_pl.a | grep Pl_Get_Current_Choice
>>
>> you should obtains something like
>>
>>     libengine_pl.a:wam_inst.o:0000000000001570 T Pl_Get_Current_Choice
> Yes.  it is there.  which might imply some kind of linking error in OBSD,
> except that you say this is new functionality for cut.
>
> The code that won't compile is in /usr/local/gprolog/Dlib/dlib.pl, the Makefile
> and some c-stuff that gets linked is in there too.

I suspect there is a confusion (mix) between libs from 1.4.0 and 1.4.1.
I have tried to compile your code (Dlib), recompiling my own version 
(installed in my home) and using gprolog-1.4.1 (installed in 
/usr/local/gprolog-1.4.1).
Both work. To try for instance:

$ PATH=usr/local/gprolog-1.4.1/bin:$PATH
$ cd Dlib/
$ make clean
rm -f dpro
$ make
gplc  dpro.pl  /usr/local/gprolog/Dlib/distance-0.1/src/*.c 
/usr/local/gprolog/Dlib/distance-0.1/src/*.pl 
/usr/local/gprolog/Dlib/dlib.pl /usr/local/gprolog/Dlib/dfile.pl 
/usr/local/gprolog/Dlib/dfile.c /usr/local/gprolog/Dlib/imacd.c 
/usr/local/gprolog/Dlib/imacd.pl 
/usr/local/gprolog/Dlib/Psql/mysql-cg.pl 
/usr/local/gprolog/Dlib/Psql/mysql-cg.c  -C '-g -Wall 
-I/usr/local/include/mysql ' -L '-L/usr/local/lib/mysql -lmysqlclient -lz'
/usr/local/gprolog/Dlib/distance-0.1/src/distance_natif.c: In function 
'dist_leven_chaine':
/usr/local/gprolog/Dlib/distance-0.1/src/distance_natif.c:47: warning: 
operation on 'i' may be undefined
/usr/local/gprolog/Dlib/dfile.c: In function 'stdin2file':
/usr/local/gprolog/Dlib/dfile.c:64: warning: suggest parentheses around 
assignment used as truth value
/usr/local/gprolog/Dlib/dfile.c: In function 'in2file':
/usr/local/gprolog/Dlib/dfile.c:101: warning: suggest parentheses around 
assignment used as truth value
/usr/local/gprolog/Dlib/dfile.c:101: warning: value computed is not used
/usr/local/gprolog/Dlib/dfile.c: In function 'set_exitime':
/usr/local/gprolog/Dlib/dfile.c:166: warning: passing argument 2 of 
'signal' from incompatible pointer type
/usr/local/gprolog/Dlib/dfile.c:169: warning: passing argument 2 of 
'signal' from incompatible pointer type
/usr/local/gprolog/Dlib/Psql/mysql-cg.c: In function 'sql_query':
/usr/local/gprolog/Dlib/Psql/mysql-cg.c:144: warning: passing argument 4 
of 'Pl_Emit_Syntax_Error' discards qualifiers from pointer target type
/usr/local/gprolog/Dlib/Psql/mysql-cg.c: In function 'sql_store_query':
/usr/local/gprolog/Dlib/Psql/mysql-cg.c:229: warning: passing argument 4 
of 'Pl_Emit_Syntax_Error' discards qualifiers from pointer target type
/dmnt/ds2a/local/gprolog-1.4.1/lib/libbips_pl.a(stream_supp.o)(.text+0x1da0): 
In function `Pl_Stream_Printf':
: warning: vsprintf() is often misused, please use vsnprintf()
/usr/local/lib/mysql/libmysqlclient.so.21.0: warning: stpcpy() is 
dangerous GNU crap; don't use it
/usr/local/lib/mysql/libmysqlclient.so.21.0: warning: strcpy() is almost 
always misused, please use strlcpy()
/tmp/gplcsNiSYB.o(.text+0x7c): In function `darea':
/usr/local/gprolog/Dlib/imacd.c:18: warning: sprintf() is often misused, 
please use snprintf()
$ ./dpro
GNU Prolog 1.4.1
By Daniel Diaz
Copyright (C) 1999-2012 Daniel Diaz
| ?-

If you want to test with my own version set:

$ PATH=/home/danieldiaz/GP/src/src/TopComp/:$PATH

(then make clean; make)

NB: gplc uses PATH to find from where it has been launched (following 
links) and thus where is its install directory. It then use this install 
directory to find its libraries.
I thinks there is a confusion here. You can try using gplc -v flag 
(verbose) to show which directories are used.

>> If you don't find a solution, could you give ma an access to your machine(s) (via ssh). With this I could fix the problem.
> My AMD64 box is 96.52.224.136, there is a danieldiaz/zaidleinad user/pass on it now.
> Please login and change it when you get this.  If you need access to the '386
> let me know.
I'd be happy to also have an access to the ix86 machine (same login+pd).

Thank you

Daniel
>
> Thanks,
>
> Dhu
>
>
>> Daniel
>>
>>
>>
>>> Dhu
>>>
>>>
>>> On Fri, 29 Jun 2012 12:00:36 -0400
>>> [email protected] wrote:
>>>
>>>> Send Users-prolog mailing list submissions to
>>>> 	[email protected]
>>>>
>>>> To subscribe or unsubscribe via the World Wide Web, visit
>>>> 	https://lists.gnu.org/mailman/listinfo/users-prolog
>>>> or, via email, send a message with subject or body 'help' to
>>>> 	[email protected]
>>>>
>>>> You can reach the person managing the list at
>>>> 	[email protected]
>>>>
>>>> When replying, please edit your Subject line so it is more specific
>>>> than "Re: Contents of Users-prolog digest..."
>>>>
>>>>
>>>> Today's Topics:
>>>>
>>>>    1. GNU Prolog 1.4.1 is released (Daniel Diaz)
>>>>
>>>>
>>>> ----------------------------------------------------------------------
>>>>
>>>> Message: 1
>>>> Date: Fri, 29 Jun 2012 16:41:31 +0200
>>>> From: Daniel Diaz <[email protected]>
>>>> To: gnu prolog mailling list <[email protected]>
>>>> Subject: GNU Prolog 1.4.1 is released
>>>> Message-ID: <[email protected]>
>>>> Content-Type: text/plain; charset="us-ascii"
>>>>
>>>> An HTML attachment was scrubbed...
>>>> URL: <http://lists.gnu.org/archive/html/users-prolog/attachments/20120629/c6851780/attachment.html>
>>>>
>>>> ------------------------------
>>>>
>>>> _______________________________________________
>>>> Users-prolog mailing list
>>>> [email protected]
>>>> https://lists.gnu.org/mailman/listinfo/users-prolog
>>>>
>>>>
>>>> End of Users-prolog Digest, Vol 108, Issue 8
>>>> ********************************************
>>>
>>> -- 
>>> Ne obliviscaris, vix ea nostro voco.
>>>
>>> _______________________________________________
>>> Users-prolog mailing list
>>> [email protected]
>>> https://lists.gnu.org/mailman/listinfo/users-prolog
>>>
>>> -- 
>>> Ce message a ete verifie par MailScanner
>>> pour des virus ou des polluriels et rien de
>>> suspect n'a ete trouve.
>>>
>>


-- 
Ce message a ete verifie par MailScanner
pour des virus ou des polluriels et rien de
suspect n'a ete trouve.