Re: PATCH: miscellaneous nits in NetBSD configuration

[Krister Walfridsson <cato at df dot lth dot se>]

On Thu, 30 Jul 1998, Todd Vierling wrote:

> Uh, they should be correct - defining it to 0 means that it turns /ON/
> DWARF2_UNWIND_INFO.  The point was to turn it /off/.
> 
> Undefining it was the correct course of action, because it is checked in gcc
> code by #ifdef, not #if.  (This was already done for arm32, in fact.)

The problem is that defaults.h does

   #if !defined (DWARF2_UNWIND_INFO) && defined (INCOMING_RETURN_ADDR_RTX)
   #define DWARF2_UNWIND_INFO 1
   #endif

so it is now defined to be 1. Defining it to 0 works fine, since toplev.c
does

   #ifdef DWARF2_UNWIND_INFO
      exceptions_via_longjmp = ! DWARF2_UNWIND_INFO;
   #else
      exceptions_via_longjmp = 1;
   #endif

and gives the same result as if it was undefined.

#undef is OK for arm32 since it has a newer toolchain than the other
NetBSD ports (but the correct thing to do is to remove the #undef (and
the comment) from arm32)

   /Krister