Adriaan van Os wrote:
Frank Heckenbach wrote:
I have uploaded a new beta version of GPC to http://www.g-n-u.de/gpc/!
Thanks for the new version !
I would be curious to know the status of:
- Waldek's gcc-3.4.x and qualified identifier patches
- range checking
Not yet included, in order to avoid further delays, especially because of the `CInteger' changes.
Building on Mac OS X with gcc-3.3.5 worked. I used the gcc-3.3.3 diff and found two hunks that don't apply::
- gcc/expr.c: rtx datum = gen_int_mode (word, mode); This is already fixed in gcc-3.3.5, see
http://gcc.gnu.org/ml/gcc-patches/2004-03/msg02060.html
gcc/emit-rtl.c
if (alias == 0 && expr == 0 && offset == 0 && (size == 0 || (mode != BLKmode && GET_MODE_SIZE (mode) == INTVAL (size)))
#ifndef GPC && (align == BITS_PER_UNIT || (STRICT_ALIGNMENT && mode != BLKmode && align == GET_MODE_ALIGNMENT (mode))))
#else
&& (align == MEM_ALIGN0 (mode)))#endif return 0;
In gcc-3.3.5, the code has changed, but I don't know if that makes a
patch unnecessary.
if (alias == 0 && expr == 0 && offset == 0 && (size == 0 || (mode != BLKmode && GET_MODE_SIZE (mode) == INTVAL (size))) && (STRICT_ALIGNMENT && mode != BLKmode ? align == GET_MODE_ALIGNMENT (mode) : align == BITS_PER_UNIT)) return 0;
I think the 3.3.5 code is alright without patch. (Waldek?)
I found a problem with initializers:
PROGRAM FP;
FUNCTION F1: Pointer; BEGIN F1:= NIL END;
FUNCTION F2( I: INTEGER): Pointer; BEGIN F2:= NIL END;
VAR P1 : Pointer value @F1; {error - initializer element for `P1' is not constant} P2 : Pointer value @F2; {OK}
BEGIN END.
Fix attached (from Waldek, with a change of mine to make it work in units as well). (avo2*.pas)
Frank