Re: [Speedycgi-users] speedy bus error under FreeBSD-5.x

[Sam H. <sa...@da...>]

Thanks for the fixes!  Looks like I need to start using a malloc debugger.

Sam

 > Hi all,
 > 
 > It seems the speedy executable has some problems running under FreeBSD-5.x,
 > e.g.:
 > 
 > --cut--
 > [lth@bhor speedy]$ ./speedy -- -v
 > Bus error (core dumped)
 > [lth@bhor speedy]$
 > --cut--
 > 
 > This, and another bus error problem stems from a couple of code bugs, solved
 > by the following patch:
 > 
 > --cut--
 > --- src/speedy_opt.c    Mon Sep 30 07:19:54 2002
 > +++ /tmp/speedy_opt.c   Tue May 20 11:11:28 2003
 > @@ -165,6 +165,8 @@ static void cmdline_split(
 >                     ++p;
 >             if (*p)
 >                 strlist_append(doing_speedy_opts ? speedy_opts : perl_args,
 > *p);
 > +           else
 > +               break;
 >         }
 > 
 >         if (*p) {
 > @@ -422,7 +424,7 @@ const char * const *speedy_opt_script_ar
 >  }
 > 
 >  SPEEDY_INLINE const char *speedy_opt_script_fname(void) {
 > -    return exec_argv.ptrs[script_argv_loc];
 > +    return exec_argv.len > script_argv_loc ?
 > exec_argv.ptrs[script_argv_loc] : NULL;
 >  }
 > 
 >  #ifdef SPEEDY_BACKEND
 > --cut--
 > 
 > Apparently, malloc() of current beta versions of 5.x has the the 'J' flag
 > set, and is therefore much more unforgiving regarding pointer overruns than
 > under 4.x (and under Linux), so this is probably why this has not been
 > discovered before.
 > 
 > Thanks to tobez for invaluable help with debugging this.
 > 
 > /Lars
 > 
 > 
 > 
 > -------------------------------------------------------
 > This SF.net email is sponsored by: ObjectStore.
 > If flattening out C++ or Java code to make your application fit in a
 > relational database is painful, don't do it! Check out ObjectStore.
 > Now part of Progress Software. http://www.objectstore.net/sourceforge
 > _______________________________________________
 > Speedycgi-users mailing list
 > Spe...@li...
 > https://lists.sourceforge.net/lists/listinfo/speedycgi-users
 >