On Mon, 17 May 2004, Christophe Rhodes wrote:
> Take II: possibly with buildability. The plan is still the same:
> eliminate all chances for the preprocessor to get involved on sparc.
> This means that we can throw away sparc-funcdef.h (and
I've some hopes that the following (stolen from CMUCL) might be just
enough -- and allow Nathan to build on Solaris. Dunno, since I have no
full sun toolchain to build on -- the setup here is gcc, GNU cpp, and Sun
-- Nikodemus Schemer: "Buddha is small, clean, and serious."
Lispnik: "Buddha is big, has hairy armpits, and laughs."
RCS file: /cvsroot/sbcl/sbcl/tools-for-build/sparc-funcdef.sh,v
retrieving revision 1.1
diff -u -r1.1 sparc-funcdef.sh
--- tools-for-build/sparc-funcdef.sh 26 Feb 2004 08:36:22 -0000
+++ tools-for-build/sparc-funcdef.sh 12 Aug 2004 13:36:58 -0000
@@ -2,7 +2,12 @@
-SUN_FUNCDEF="#define FUNCDEF(x) .type x,#function"
+#define FUNCDEF(x) .type x, \#function
+#define FUNCDEF(x) .type x, #function
GNU_FUNCDEF="#define FUNCDEF(x) .type x,@function"
echo $SUN_FUNCDEF > $TMP