On 12/17 06:22, Jeremie Courreges-Anglas wrote:
> On Thu, Dec 13 2018, Jeremy Evans <[email protected]> wrote:
> > Update to the latest release of mruby.  Release announcement is
> > available at:
> > https://mruby.org/releases/2018/12/11/mruby-2.0.0-released.html
> >
> > Tested on amd64.  Will be committing in a few days unless I hear
> > objections.
> >
> > If someone with sparc64 could test and see if it builds there now, I
> > would appreciate it.
> 
> It packages and all tests pass except for 2 of them (endianness?)

Jeremie,

Could you please test this patch and see if it fixes the tests on sparc64:

Thanks,
Jeremy

Index: Makefile
===================================================================
RCS file: /cvs/ports/lang/mruby/Makefile,v
retrieving revision 1.10
diff -u -p -r1.10 Makefile
--- Makefile    17 Dec 2018 20:28:27 -0000      1.10
+++ Makefile    20 Dec 2018 17:45:20 -0000
@@ -4,6 +4,7 @@ COMMENT =               lightweight, embeddable imple
 
 VERSION =              2.0.0
 DISTNAME =             mruby-${VERSION}
+REVISION =             0
 CATEGORIES =           lang
 HOMEPAGE =             https://github.com/mruby/mruby
 
Index: patches/patch-include_mrbconf_h
===================================================================
RCS file: patches/patch-include_mrbconf_h
diff -N patches/patch-include_mrbconf_h
--- /dev/null   1 Jan 1970 00:00:00 -0000
+++ patches/patch-include_mrbconf_h     20 Dec 2018 17:45:20 -0000
@@ -0,0 +1,24 @@
+$OpenBSD$
+
+Index: include/mrbconf.h
+--- include/mrbconf.h.orig
++++ include/mrbconf.h
+@@ -7,6 +7,7 @@
+ #ifndef MRUBYCONF_H
+ #define MRUBYCONF_H
+ 
++#include <endian.h>
+ #include <limits.h>
+ #include <stdint.h>
+ 
+@@ -62,7 +63,9 @@
+ //#define MRB_NAN_BOXING
+ 
+ /* define on big endian machines; used by MRB_NAN_BOXING */
+-//#define MRB_ENDIAN_BIG
++#if (BYTE_ORDER == BIG_ENDIAN)
++#define MRB_ENDIAN_BIG
++#endif
+ 
+ /* represent mrb_value as a word (natural unit of data for the processor) */
+ //#define MRB_WORD_BOXING

Reply via email to