If you send a reply without an active reply slot, we used to return EPERM.
However, this makes it impossible to distinguish this case from a real
permission-denied error (eg., LSM). Hence, make sure we return a unique
error-code if the reply-slot is missing.
With this patch, you get EBADSLT ("Invalid slot") if you call
KDBUS_CMD_SEND with a reply-cookie that has no valid slot.
Reviewed-by: Daniel Mack <dan...@zonque.org>
Signed-off-by: David Herrmann <dh.herrm...@gmail.com>
---
ipc/kdbus/connection.c | 2 +-
tools/testing/selftests/kdbus/test-message.c | 2 +-
tools/testing/selftests/kdbus/test-sync.c | 2 +-
3 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/ipc/kdbus/connection.c b/ipc/kdbus/connection.c
index d94b417..2787bd7 100644
--- a/ipc/kdbus/connection.c
+++ b/ipc/kdbus/connection.c
@@ -1123,7 +1123,7 @@ static int kdbus_conn_reply(struct kdbus_conn *src,
mutex_unlock(&dst->lock);
if (!reply) {
- ret = -EPERM;
+ ret = -EBADSLT;
goto exit;
}
diff --git a/tools/testing/selftests/kdbus/test-message.c
b/tools/testing/selftests/kdbus/test-message.c
index ddc1e0a..563dc85 100644
--- a/tools/testing/selftests/kdbus/test-message.c
+++ b/tools/testing/selftests/kdbus/test-message.c
@@ -75,7 +75,7 @@ int kdbus_test_message_basic(struct kdbus_test_env *env)
/* Faked replies with a valid reply cookie are rejected */
ret = kdbus_msg_send_reply(conn, time(NULL) ^ cookie, sender->id);
- ASSERT_RETURN(ret == -EPERM);
+ ASSERT_RETURN(ret == -EBADSLT);
ret = kdbus_free(conn, offset);
ASSERT_RETURN(ret == 0);
diff --git a/tools/testing/selftests/kdbus/test-sync.c
b/tools/testing/selftests/kdbus/test-sync.c
index e2be910..0655a54 100644
--- a/tools/testing/selftests/kdbus/test-sync.c
+++ b/tools/testing/selftests/kdbus/test-sync.c
@@ -235,7 +235,7 @@ static void *run_thread_reply(void *data)
/* using an unknown cookie must fail */
ret = kdbus_msg_send_reply(conn_a, ~cookie, conn_b->id);
- if (ret != -EPERM) {
+ if (ret != -EBADSLT) {
status = TEST_ERR;
goto exit_thread;
}
--
2.5.0
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
