[Ecls-list] problem with cl_boot()

Juan Jose Garcia-Ripoll jlr at mpq.mpg.de
Thu Mar 28 02:32:06 UTC 2002


On Thursday 28 March 2002 11:12, you wrote:
> yo
> everything is now working ok in my C program.
> thank you :)
> however i'm having a problem with cl_boot() when running it through a perl
> program the perl program is embedding the ecls C library, and when i call
> cl_boot i get a core dump:
> Program received signal SIGSEGV, Segmentation fault.
> #0  0x4022aa25 in bds_unwind (new_bds_top=0x817eff8) at
> ecls-0.5/src/c/stacks.d:64 64   SYM_VAL(bds->bds_sym) = bds->bds_val;
> [...]
> anyway, i don't expect anyone to debug that for me, but if you have any
> idea of what could be the problem, or maybe any suggestion to figure out
> what's wrong, that would be really apprecied :)

ECL requires a garbage collector to create and automatically destroy lisp 
objects. I suspect perl has its own memory requirements and both might 
interfere with each other.

Another possibility, is that the stack of variables from ECL has not been 
restored properly when exiting from some function (for instance, when 
performing a goto or a return outside a function that defines a special 
variable). Were this the reason, it would require close inspection, but 
personal problems are preventing me from dedicating time to ECL during the 
last months. Sorry I can't help more.

J




More information about the ecl-devel mailing list