On 16.10.2015 09:32, Peter Krempa wrote:
On Mon, Oct 12, 2015 at 12:25:57 +0200, Michal Privoznik wrote:
A rather sparse commit message.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> --- src/locking/lock_daemon.c | 44 ++++++++++++++++++++++++++ src/locking/lock_daemon.h | 8 +++++ src/locking/lock_daemon_dispatch.c | 65 ++++++++++++++++++++++++++++++++------ 3 files changed, 108 insertions(+), 9 deletions(-)
[...]
diff --git a/src/locking/lock_daemon_dispatch.c b/src/locking/lock_daemon_dispatch.c index 2d0bd81..5ad7e11 100644 --- a/src/locking/lock_daemon_dispatch.c +++ b/src/locking/lock_daemon_dispatch.c
[...]
@@ -433,21 +435,66 @@ virLockSpaceProtocolDispatchCreateLockSpace(virNetServerPtr server ATTRIBUTE_UNU
[...]
static int virLockSpaceProtocolDispatchRecallSeclabel(virNetServerPtr server ATTRIBUTE_UNUSED, - virNetServerClientPtr client ATTRIBUTE_UNUSED, + virNetServerClientPtr client, virNetMessagePtr msg ATTRIBUTE_UNUSED, - virNetMessageErrorPtr rerr ATTRIBUTE_UNUSED, - virLockSpaceProtocolRecallSeclabelArgs *args ATTRIBUTE_UNUSED, - virLockSpaceProtocolRecallSeclabelRet *ret ATTRIBUTE_UNUSED) + virNetMessageErrorPtr rerr, + virLockSpaceProtocolRecallSeclabelArgs *args, + virLockSpaceProtocolRecallSeclabelRet *ret) { - return 0; + int rv = -1; + int funcRet; + virLockDaemonClientPtr priv = + virNetServerClientGetPrivateData(client); + char *label = NULL; + char **label_p = NULL; + + virMutexLock(&priv->lock); + + memset(ret, 0, sizeof(*ret)); + + funcRet = virLockDaemonRecallSeclabel(lockDaemon, + args->path, args->model, &label); + + if (funcRet == 0 && + (VIR_ALLOC(label_p) < 0 || + VIR_STRDUP(*label_p, label) < 0))
This looks rather weird? Why is the extra pointer necessary?
It's due to way we handle remote strings. If that was remote NON NULL string, then a single pointer would be enough, but since I want it to be NULL sometimes I must do it this way. If you look at the remote procedures we generate automatically you'll find the same pattern (in fact I've copied the idea from there). Michal