On Tue, Sep 9, 2014 at 1:36 PM, Jerry James <loganje...@gmail.com> wrote:
> Oh, wait!  I didn't notice that the previous recipe (calling the
> function from use_symbols) *worked*, and it is now complaining about
> __aeabi_unwind_cpp_pr0.  That is, it's a different symbol, ending with
> a zero instead of a 1 now.  Sorry about that.  Okay, I will now add a
> call to __aeabi_unwind_cpp_pr0 as well and we'll see if it wants
> anything else.

I don't know why the TLDFLAGS and TLIBS approaches did not work, but
this patch led to a successful build:

--- o/sfasli.c.orig 2014-09-06 09:45:30.000000000 -0600
+++ o/sfasli.c 2014-09-09 12:00:00.000000000 -0600
@@ -133,6 +133,11 @@ extern int __remqu();

 #ifndef DARWIN
 #ifndef _WIN32
+#if defined(__arm__) && !defined(__aarch64__)
+extern void __aeabi_unwind_cpp_pr0(void);
+extern void __aeabi_unwind_cpp_pr1(void);
+#endif
+
 int
 use_symbols(double d,...) {

@@ -141,6 +146,10 @@ use_symbols(double d,...) {
 #ifdef GCL_GPROF
   _mcount();
 #endif
+#if defined(__arm__) && !defined(__aarch64__)
+  __aeabi_unwind_cpp_pr0();
+  __aeabi_unwind_cpp_pr1();
+#endif

   return (int)d;

Now to try building on aarch64 and see what happens.
-- 
Jerry James
http://www.jamezone.org/

_______________________________________________
Gcl-devel mailing list
Gcl-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/gcl-devel

Reply via email to