Damn. I did it again, & I get the same results (function
names in the backtrace are not mangled). This time, I
swear, I blew away the entire clisp-2.33 source directory,
unpacked it anew, ran the ./configure, & all that. Could
you check the commands I used?
All the files are at <http://lisp-p.org/clb/>. You should
be able to view that entire directory. "go" is the script.
"go.cmd" is the commands for gdb. "go.out" is the output of
the entire thing (./configure, make, gdb backtrace -- the
whole enchilada).
Thanks for spending so much time on this. I apologize for
having such trouble getting you a decent backtrace.
Just in case there is something useful in the most recent
backtrace, here's the top of it:
Function "sigsegv_handler_failed" not defined.
STACK depth: 23979
SP depth: 67111153
Program received signal SIGSEGV, Segmentation fault.
0x0804e474 in gc_mark (obj={one_o = 1, allocstamp = 32899})
at spvw_garcol.d:320
320 switch (Record_type(dies)) {
Warning: the current language does not match this frame.
(gdb) backtrace
#0 0x0804e474 in gc_mark (obj={one_o = 1, allocstamp = 32899})
at spvw_garcol.d:320
#1 0x0804ecdc in gc_mark_stack (objptr=0xb7301a40) at spvw_garcol.d:521
#2 0x0804ed0c in gc_markphase () at spvw_garcol.d:532
#3 0x0804fd84 in gar_col_normal () at spvw_garcol.d:1718
#4 0x08053201 in do_gar_col_simple () at spvw_garcol.d:2603
#5 0x08229b88 in with_gc_statistics(void (*)()) (
fun=0x80531e8 <do_gar_col_simple>) at predtype.d:2801
#6 0x08053223 in gar_col_simple () at spvw_garcol.d:2632
#7 0x080534da in make_space_gc_true (need=16, heapptr=0x835efc4)
at spvw_allocate.d:216
#8 0x080539fb in allocate_vector(unsigned long) (len=2) at spvw_typealloc.d:92
#9 0x0810c54b in test_env () at control.d:1964
From: Sam Steingold <sds@...>
Date: Mon, 22 Mar 2004 15:36:23 -0500
> * Gene Michael Stover <tzf@...> [2004-03-22 19:59:23 +0000]:
>
>> correct me if I am wrong, but you are not using G++, right?
>> note that setting of CC=g++ above!
>
> My mistake. I've run it again, & here's the last of the run
> plus the top of the backstacktrace. The full log (including
> the ./configure & make output) are at
> <http://lisp-p.org/clb/backtrace.txt>. The exact commands I
> used are at <http://lisp-p.org/clb/go>.
still no go. apparently, you did not "rm -rf" build-g-gxx directory
before reconfiguring.
there is no link command in <http://lisp-p.org/clb/backtrace.txt>, and,
apparently, you are not using the C++-build executable:
> #0 0x0804e434 in gc_mark (obj={one_o = 1, allocstamp = 32898})
> at spvw_garcol.d:320
> #1 0x0804ec9c in gc_mark_stack (objptr=0xb70f1794) at spvw_garcol.d:521
> #2 0x0804eccc in gc_markphase () at spvw_garcol.d:532
> #3 0x0804fd44 in gar_col_normal () at spvw_garcol.d:1718
> #4 0x080531c1 in do_gar_col_simple () at spvw_garcol.d:2603
> #5 0x08229b48 in with_gc_statistics(void (*)()) (
> fun=0x80531a8 <do_gar_col_simple>) at predtype.d:2801
> #6 0x080531e3 in gar_col_simple () at spvw_garcol.d:2632
> #7 0x0805349a in make_space_gc_true (need=16, heapptr=0x835ef64)
> at spvw_allocate.d:216
> #8 0x080539bb in allocate_vector(unsigned long) (len=2) at spvw_typealloc.d:92
> #9 0x0810c50b in test_env () at control.d:1964
> #10 0x0810c679 in C_macro_function() () at control.d:1980
if you see function names printed "in the clear", it is probably C.
if you saw something like
_Z17C_read_eval_printv
or
_Z15funcall_closure6objectm
then it would have been C++ ("mangling")
--
Sam Steingold (http://www.podval.org/~sds) running w2k
<http://www.camera.org> <http://www.iris.org.il> <http://www.memri.org/>
<http://www.mideasttruth.com/> <http://www.honestreporting.com>
Never succeed from the first try - if you do, nobody will think it was hard.
|