Re: [bjam, macos] from the boost-users list

From: Joao Abecasis (jpabecasis_at_[hidden])
Date: 2005-02-28 10:38:38

David Abrahams wrote:
> 
> ------------------------------------------------------------------------
> 
> Subject:
> bjam seg faulting on mac os x 10.3.8
> From:
> "Andy Espenscheid" <aespen_at_[hidden]>
> Date:
> Sun, 27 Feb 2005 22:45:57 -0600
> 
> Newsgroups:
> gmane.comp.lib.boost.user
> 
> 
> I am trying to compile boost 1.32.0 on MacOSX 10.3.8 and getting a 
> Segmentation Fault immediately from bjam immediately. I'm using gcc 
> 3.3. bjam is crashing with a EXC_BAD_ADDRESS exception. I compiled 
> bjam from the boost distribution but it has the same behavior. Any 
> pointers would be greatly appreciated. Seems like I'm doing somehthing 
> dumb. Also, are binaries available via fink? I seached with fink 0.71 
> but couldn't find anything. Thx.
> 
> 
> ------------------------------------------------------------------------

I'm not sure wether this is the same issue, but I'm also getting bjam 
Segmentation Faults, on a machine running linux.

On this machine, bjam currently segfaults with the quickbook toolset. 
I've tried compiling bjam with gcc 3.3 and 3.4 and with the intel 
compiler on that machine, all producing the same result. On a debug 
build of bjam, gdb shows a backtrace of 168432 function calls (up to 
main). Here's a portion:

---- GDB output ----

(gdb) bt 10
#0 0x0804dab8 in var_expand (l=0x0, in=0xbf4011d6 "", end=0xbf4011d6 
"", lol=0xbf402568, cancopyin=0) at ../expand.c:125
#1 0x0804dcae in var_expand (l=0x0, in=0x9899b36 "", end=0x9899b36 "", 
lol=0xbf402568, cancopyin=1) at ../expand.c:216
#2 0x0804c028 in compile_list (parse=0x98c0cf0, frame=0xbf402560) at 
../compile.c:462
#3 0x08055198 in parse_evaluate (p=0x98c0cf0, frame=0xbf402560) at 
../parse.c:136
#4 0x0804b8ac in compile_eval (parse=0x98c0d50, frame=0xbf402560) at 
../compile.c:169
#5 0x08055198 in parse_evaluate (p=0x98c0d50, frame=0xbf402560) at 
../parse.c:136
#6 0x0804bcb4 in compile_if (p=0x98c1e60, frame=0xbf402560) at 
../compile.c:309
#7 0x08055198 in parse_evaluate (p=0x98c1e60, frame=0xbf402560) at 
../parse.c:136
#8 0x0804cfb9 in compile_rules (parse=0x98c2010, frame=0xbf402560) at 
../compile.c:1092
#9 0x08055198 in parse_evaluate (p=0x98c2010, frame=0xbf402560) at 
../parse.c:136
(More stack frames follow...)

(gdb) bt -10
#168423 0x0804c13a in compile_local (parse=0x9800680, frame=0xbff247e0) 
at ../compile.c:505
#168424 0x08055198 in parse_evaluate (p=0x9800680, frame=0xbff247e0) at 
../parse.c:136
#168425 0x0804c13a in compile_local (parse=0x98006b0, frame=0xbff247e0) 
at ../compile.c:505
#168426 0x08055198 in parse_evaluate (p=0x98006b0, frame=0xbff247e0) at 
../parse.c:136
#168427 0x0804bcdf in compile_if (p=0x9858d60, frame=0xbff247e0) at 
../compile.c:313
#168428 0x08055198 in parse_evaluate (p=0x9858d60, frame=0xbff247e0) at 
../parse.c:136
#168429 0x0804cfe6 in compile_rules (parse=0x9858d60, frame=0xbff247e0) 
at ../compile.c:1095
#168430 0x08055198 in parse_evaluate (p=0x97f95d8, frame=0xbff247e0) at 
../parse.c:136
#168431 0x08054fe3 in parse_file (f=0x80624e7 "+", frame=0xbff247e0) at 
../parse.c:59
#168432 0x08051acd in main (argc=0, argv=0xbff2524c, 
arg_environ=0xbff25250) at ../jam.c:406

---- ----

On other machines I've tried all runs well without segfaults, including 
the same executable that provided the above output. This could be the 
execution stack is too small (?).

Best regards,

João