[kaffe] kjc on m68k/netbsd works.

[Kiyo Inaba]

Hi all,

I am glad to report that m68k/netbsd can now make kjc to compile.
As attached patch suggests, this is a revert patch against Tony submitted
several years ago. This portion is to correctly setup saved registers area
(usually for d2-d7/a2-a5) when exception was raised and I have no idea why
he commented it out.

But, anyway it becomes work! Currently, roughly 20 test cases are failed,
and the number without this patch was 42.

Maybe, I can submit one more patch for linux to make it work with jit3. (But
still jit3 is slower than jit on m68k :-<)

Kiyo

diff -Naur kaffe-snap-040812.orig/ChangeLog kaffe-snap-040812/ChangeLog
--- kaffe-snap-040812.orig/ChangeLog	2004-08-11 09:18:29.000000000 +0900
+++ kaffe-snap-040812/ChangeLog	2004-08-17 22:39:08.000000000 +0900
@@ -1,3 +1,8 @@
+2004-08-17  Kiyo Inaba  <inaba at src.ricoh.co.jp>
+
+	* config/m68k/jit3-m68k.def:
+	Make kjc works.
+
 2004-08-10  Dalibor Topic  <robilad at kaffe.org>
 
 	* libraries/clib/awt/classpath-gtk/gtk-peer/gnu_java_awt_peer_gtk_GdkPixbufDecoder.c,
diff -Naur kaffe-snap-040812.orig/config/m68k/jit3-m68k.def kaffe-snap-040812/config/m68k/jit3-m68k.def
--- kaffe-snap-040812.orig/config/m68k/jit3-m68k.def	2004-07-30 20:22:45.000000000 +0900
+++ kaffe-snap-040812/config/m68k/jit3-m68k.def	2004-08-17 12:14:35.000000000 +0900
@@ -1302,13 +1302,11 @@
 	l->at = (uintp)CODEPC+2;
 	op_addal_ia(0, REG_a0);
 
-#if	(0)		// don't destroy our good registers
 #if !defined(HAVE_NO_SUBAW)
 	op_subaw_ia(SRNR * SLOTSIZE, REG_a0);
 #else
 	op_subal_ia(SRNR * SLOTSIZE, REG_a0);
 #endif
-#endif		// end no-oped section
 
 	op_moveal_aa(REG_a0, REG_sp);
 }