[sheepdog] [PATCH] sheep: fix nr_nodes calculation in local_stat_cluster
Christoph Hellwig
hch at infradead.org
Wed Jun 6 14:49:10 CEST 2012
Pass the actual node number to get_max_nr_copies_from, instead of
the size of the nodes array.
Signed-off-by: Christoph Hellwig <hch at lst.de>
diff --git a/sheep/ops.c b/sheep/ops.c
index 91f3536..2fe6fb4 100644
--- a/sheep/ops.c
+++ b/sheep/ops.c
@@ -349,16 +349,17 @@ static int local_stat_cluster(struct request *req)
log = (struct epoch_log *)req->data + i;
log->epoch = epoch;
log->ctime = get_cluster_ctime();
- log->nr_nodes = epoch_log_read(epoch, (char *)log->nodes,
+ log->nr_nodes = epoch_log_read_nr(epoch, (char *)log->nodes,
sizeof(log->nodes));
- if (log->nr_nodes == -1)
+ if (log->nr_nodes == -1) {
log->nr_nodes = epoch_log_read_remote(epoch,
(char *)log->nodes,
sizeof(log->nodes));
+ log->nr_nodes /= sizeof(log->nodes[0]);
+ }
log->nr_copies = get_max_nr_copies_from(log->nodes, log->nr_nodes);
rsp->data_length += sizeof(*log);
- log->nr_nodes /= sizeof(log->nodes[0]);
/* FIXME: this hack would require sizeof(time_t) < sizeof(log->nodes[0]) */
log->time = *(uint64_t *)(&log->nodes[log->nr_nodes]);
epoch--;
More information about the sheepdog
mailing list