yay! finally fixed the bug that volker, ronnie and I have been chasing

for 2 days.

The main bug was in smbd, but there was a secondary (and more subtle)
bug in ctdb that the bug in smbd exposed. When we get send a dmaster
reply, we have to correctly update the dmaster in the recipient even
if the original requst has timed out, otherwise ctdbd can get into a
loop fighting over who will handle a key.

This patch also cleans up the packet allocation, and makes ctdbd
become a real daemon.

(This used to be ctdb commit 59405e59ef522b97d8e20e4b14310a217141ac7c)

1 files changed, 46 insertions, 0 deletions

diff --git a/ctdb/common/ctdb_util.c b/ctdb/common/ctdb_util.c
index f39e8b6682..e77dc8990d 100644
--- a/ctdb/common/ctdb_util.c
+++ b/ctdb/common/ctdb_util.c
@@ -129,6 +129,51 @@ void ctdb_latency(double *latency, struct timeval t)
 	}
 }
 
+#if 0
+struct idr_fake {
+	uint32_t size;
+	void **ptrs;
+};
+
+static void idr_fake_init(struct ctdb_context *ctdb)
+{
+	if (ctdb->fidr) return;
+	ctdb->fidr = talloc(ctdb, struct idr_fake);
+	ctdb->fidr->size = 0x10000;
+	ctdb->fidr->ptrs = talloc_zero_array(ctdb->fidr, void *, 
+						 ctdb->fidr->size);
+}
+
+uint32_t ctdb_reqid_new(struct ctdb_context *ctdb, void *state)
+{
+	uint32_t i;
+	idr_fake_init(ctdb);
+	for (i=0;i<ctdb->fidr->size;i++) {
+		if (ctdb->fidr->ptrs[i] == NULL) {
+			ctdb->fidr->ptrs[i] = state;
+			return i;
+		}
+	}
+	return (uint32_t)-1;
+}
+
+void *_ctdb_reqid_find(struct ctdb_context *ctdb, uint32_t reqid, const char *type, const char *location)
+{
+	idr_fake_init(ctdb);
+	if (ctdb->fidr->ptrs[reqid] == NULL) {
+		DEBUG(0,("bad fidr id %u\n", reqid));
+	}
+	return ctdb->fidr->ptrs[reqid];
+}
+
+
+void ctdb_reqid_remove(struct ctdb_context *ctdb, uint32_t reqid)
+{
+	idr_fake_init(ctdb);
+	ctdb->fidr->ptrs[reqid] = NULL;
+}
+
+#else
 uint32_t ctdb_reqid_new(struct ctdb_context *ctdb, void *state)
 {
 	uint32_t id;
@@ -161,3 +206,4 @@ void ctdb_reqid_remove(struct ctdb_context *ctdb, uint32_t reqid)
 	}
 }
 
+#endif