[Gambas-user] Problem with new JIT

[Benoît Minisini]

Le 11/01/2019 à 09:46, Jorge Carrión a écrit :
> Benoit:
> 
> Installing libc6-dev doesn't work. The only way I can use our 
> applications is using GB_NO_JIT environment variable.
> If I try to launch the programs from console I got several messages, 
> like this:
> 
> gb.jit: error: unable to compile JIT code:
> In file included from /tmp/gambas.1000/4415/jit/gb.report.c:1:0:
> /tmp/gambas.1000/4415/jit/gb.report.c: In function ‘jit_reportunits_2’:
> /tmp/gambas.1000/4415/jit/jit.h:1403:3: error: ‘for’ loop initial 
> declarations are only allowed in C99 mode
>     for (int _i = 0; _i < (_n); _i++) \
>     ^
> /tmp/gambas.1000/4415/jit/jit.h:1418:40: note: in definition of macro 
> ‘RETURN_f’
>   #define RETURN_f(_val) (GB.ReturnFloat(_val))
>                                          ^
> /tmp/gambas.1000/4415/jit/gb.report.c:24:42: note: in expansion of macro 
> ‘OPT’
>     RETURN_f(jit_reportunits_2_(PARAM_s(0),OPT(1,1),PARAM_OPT_i(1)));
>                                            ^
> /tmp/gambas.1000/4415/jit/jit.h:1403:3: note: use option -std=c99 or 
> -std=gnu99 to compile your code
>     for (int _i = 0; _i < (_n); _i++) \
>     ^
> /tmp/gambas.1000/4415/jit/jit.h:1418:40: note: in definition of macro 
> ‘RETURN_f’
>   #define RETURN_f(_val) (GB.ReturnFloat(_val))
>                                          ^
> /tmp/gambas.1000/4415/jit/gb.report.c:24:42: note: in expansion of macro 
> ‘OPT’
>     RETURN_f(jit_reportunits_2_(PARAM_s(0),OPT(1,1),PARAM_OPT_i(1)));
>                                            ^
> 
> 
> **
> ** OOPS! INTERNAL ERROR. Program aborting, sorry! :-(
> ** Unable to compile JIT source file
> **
> ** _ClassStat.GotoNextSection.14: #39: End of file
> ** _ClassStat.GotoNextSection.14 _ClassStat.Stat.86 Jit._Compile.165
> **
> ** Please send a bug report to the gambas bugtracker [1] or to the 
> gambas mailing-list [2].
> ** [1] http://gambaswiki.org/bugtracker
> ** [2] https://lists.gambas-basic.org/listinfo/user
> 
> I don't know if this issue is the same fixed in "Jit Crash" thread 
> opened in the mailing list. If is the same, please let me know to avoid 
> duplicities.
> 
> Best regards.
> 

This bug has been fixed in 3.12.1. I will release a 3.12.2 soon with 
other JIT fixes.

Regards,

-- 
Benoît Minisini