mintlibs42 and gcc...

["Nicholas S Castellano" <entropy@terminator.rs.itd.umich.edu>]

> I'm having some severe problems with the MiNTlibs Pl42, which go like this:
> GCC233 won't compile them, ok, entropy already said that. It's really a
>pity, in fact, but I can't change it.
> GCC245 isn't able to produce running 68000 code, notably fopen() gets
>seriously damaged, returns a pointer pointing directly to nirwana. :-(
> GCC257 stops compiling at the first part of libgxx2.c with:
>libgcc2.c:132: no data type for mode `XF'
> GCC258 doesn't seem to be available somewhere, at least, I didn't find
>it on my famous ftp servers.
> That's a bit pity, cause I've already erased the old libs and am
>therefore more/less unable to do anything at all. Hasn't anyone already
>worked out a clue on how to compile it with 233 or any other clues?

The problem with XFmode wasn't really GCC-version related.

PL43 will hopefully be ready by monday and will address this problem.

Meanwhile, try the following patch to get a compilable libgcc2.c that
should work with any GCC version:

*** libgcc2.c-	Sun Jan 16 05:45:50 1994
--- libgcc2.c	Thu Feb 10 17:24:15 1994
***************
*** 60,66 ****
  #include <stddef.h>
  #endif
  
! #if 1
  /* Define for XFmode extended real floating point support.  */
  #define LONG_DOUBLE_TYPE_SIZE 96
  #else
--- 60,66 ----
  #include <stddef.h>
  #endif
  
! #if 0
  /* Define for XFmode extended real floating point support.  */
  #define LONG_DOUBLE_TYPE_SIZE 96
  #else

If you apply this remember to back it out before applying the PL43
patches when they become available.

--
entropy -- it's not just a good idea, it's the second law.
Personal mail:      entropy@gnu.ai.mit.edu
MiNT library mail:  entropy@terminator.rs.itd.umich.edu
"what do you have against octal?" -jrb