Don't overwrite the regress target, provide a test target instead. This allows

the test target to fail without terminating overall regression tests. Clean up
a little.

Also adjust orders.txt.sig comment to new reality, making the test pass again.

Looks ok to sthen@

diff --git a/regress/usr.bin/signify/Makefile b/regress/usr.bin/signify/Makefile
index cc761de..9796bba 100644 (file)
--- a/regress/usr.bin/signify/Makefile
+++ b/regress/usr.bin/signify/Makefile
@@ -1,6 +1,11 @@
-#      $OpenBSD: Makefile,v 1.1 2014/01/09 16:13:44 tedu Exp $
+#      $OpenBSD: Makefile,v 1.2 2014/01/10 22:52:50 tobiasu Exp $
 
-regress:
+CLEANFILES += test.sig
+REGRESS_TARGETS = t1
+
+t1:
        @sh ${.CURDIR}/signify.sh ${.CURDIR}
 
+.PHONY: t1
+
 .include <bsd.regress.mk>

diff --git a/regress/usr.bin/signify/orders.txt.sig b/regress/usr.bin/signify/orders.txt.sig
index 28ebf6e..90f02d2 100644 (file)
--- a/regress/usr.bin/signify/orders.txt.sig
+++ b/regress/usr.bin/signify/orders.txt.sig
@@ -1,2 +1,2 @@
-untrusted comment: signify signature from signify secret key
+untrusted comment: signature from signify secret key
 RWTAeKJJ1MTF3UpxzBCu6NaM6HPJNTj5CZ+M5XNJKNeEHBLQSsstzHGbSo8rPYNgw3Z98pN7WKiIwBIyRrKuIdKBRA6qlaci6wI=