upstream commit

test new behaviour of cert force-command restriction vs.
authorized_key/ principals

Upstream-Regress-ID: 399efa7469d40c404c0b0a295064ce75d495387c

1 files changed, 15 insertions, 1 deletions

diff --git a/regress/cert-userkey.sh b/regress/cert-userkey.sh
index 319746395..7005fd55e 100644
--- a/regress/cert-userkey.sh
+++ b/regress/cert-userkey.sh
@@ -1,4 +1,4 @@
-#       $OpenBSD: cert-userkey.sh,v 1.16 2016/05/03 12:15:49 dtucker Exp $
+#       $OpenBSD: cert-userkey.sh,v 1.17 2016/11/30 03:01:33 djm Exp $
 #       Placed in the Public Domain.
 
 tid="certified user keys"
@@ -354,6 +354,20 @@ test_one "principals key option principals" success "-n mekmitasdigoat" \
 test_one "principals key option no principals" failure "" \
     authorized_keys ',principals="mekmitasdigoat"'
 
+# command= options vs. force-command in key
+test_one "force-command match true" success \
+    "-n ${USER} -Oforce-command=true" \
+    authorized_keys ',command="true"'
+test_one "force-command match true" failure \
+    "-n ${USER} -Oforce-command=false" \
+    authorized_keys ',command="false"'
+test_one "force-command mismatch 1" failure \
+    "-n ${USER} -Oforce-command=false" \
+    authorized_keys ',command="true"'
+test_one "force-command mismatch 2" failure \
+    "-n ${USER} -Oforce-command=true" \
+    authorized_keys ',command="false"'
+
 # Wrong certificate
 cat $OBJ/sshd_proxy_bak > $OBJ/sshd_proxy
 for ktype in $PLAIN_TYPES ; do