When unmounting an NFS mount using sec=krb5[ip], the umount system

call could get hung sleeping on "gsssta" if the credentials for a user
that had been accessing the mount point have expired. This happened
because rpc_gss_destroy_context() would end up calling itself when the
"destroy context" RPC was attempted, trying to refresh the credentials.
This patch just checks for this case in rpc_gss_refresh() and returns
without attempting the refresh, which avoids the recursive call to
rpc_gss_destroy_context() and the subsequent hang.

Reviewed by:	dfr
Approved by:	re (Ken Smith), kib (mentor)

1 files changed, 14 insertions, 0 deletions

diff --git a/sys/rpc/rpcsec_gss/rpcsec_gss.c b/sys/rpc/rpcsec_gss/rpcsec_gss.c
index e529b83..ec0c595 100644
--- a/sys/rpc/rpcsec_gss/rpcsec_gss.c
+++ b/sys/rpc/rpcsec_gss/rpcsec_gss.c
@@ -929,6 +929,20 @@ rpc_gss_refresh(AUTH *auth, void *msg)
 {
 	struct rpc_msg *reply = (struct rpc_msg *) msg;
 	rpc_gss_options_ret_t options;
+	struct rpc_gss_data *gd;
+
+	gd = AUTH_PRIVATE(auth);
+	
+	/*
+	 * If the context is in DESTROYING state, then just return, since
+	 * there is no point in refreshing the credentials.
+	 */
+	mtx_lock(&gd->gd_lock);
+	if (gd->gd_state == RPCSEC_GSS_DESTROYING) {
+		mtx_unlock(&gd->gd_lock);
+		return (FALSE);
+	}
+	mtx_unlock(&gd->gd_lock);
 
 	/*
 	 * If the error was RPCSEC_GSS_CREDPROBLEM of