libceph: mon_get_version request infrastructure
Add support for mon_get_version requests to libceph. This reuses much
of the ceph_mon_generic_request infrastructure, with one exception.
Older OSDs don't set mon_get_version reply hdr->tid even if the
original request had a non-zero tid, which makes it impossible to
lookup ceph_mon_generic_request contexts by tid in get_generic_reply()
for such replies. As a workaround, we allocate a reply message on the
reply path. This can probably interfere with revoke, but I don't see
a better way.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>
diff --git a/net/ceph/ceph_common.c b/net/ceph/ceph_common.c
index 67d7721..1675021 100644
--- a/net/ceph/ceph_common.c
+++ b/net/ceph/ceph_common.c
@@ -72,6 +72,8 @@
case CEPH_MSG_MON_SUBSCRIBE_ACK: return "mon_subscribe_ack";
case CEPH_MSG_STATFS: return "statfs";
case CEPH_MSG_STATFS_REPLY: return "statfs_reply";
+ case CEPH_MSG_MON_GET_VERSION: return "mon_get_version";
+ case CEPH_MSG_MON_GET_VERSION_REPLY: return "mon_get_version_reply";
case CEPH_MSG_MDS_MAP: return "mds_map";
case CEPH_MSG_CLIENT_SESSION: return "client_session";
case CEPH_MSG_CLIENT_RECONNECT: return "client_reconnect";
diff --git a/net/ceph/debugfs.c b/net/ceph/debugfs.c
index 8903dce..d1a62c6 100644
--- a/net/ceph/debugfs.c
+++ b/net/ceph/debugfs.c
@@ -129,6 +129,8 @@
seq_printf(s, "%llu statfs\n", req->tid);
else if (op == CEPH_MSG_POOLOP)
seq_printf(s, "%llu poolop\n", req->tid);
+ else if (op == CEPH_MSG_MON_GET_VERSION)
+ seq_printf(s, "%llu mon_get_version", req->tid);
else
seq_printf(s, "%llu unknown\n", req->tid);
}
diff --git a/net/ceph/mon_client.c b/net/ceph/mon_client.c
index 2ac9ef3..11d8d2f 100644
--- a/net/ceph/mon_client.c
+++ b/net/ceph/mon_client.c
@@ -477,14 +477,13 @@
return m;
}
-static int do_generic_request(struct ceph_mon_client *monc,
- struct ceph_mon_generic_request *req)
+static int __do_generic_request(struct ceph_mon_client *monc, u64 tid,
+ struct ceph_mon_generic_request *req)
{
int err;
/* register request */
- mutex_lock(&monc->mutex);
- req->tid = ++monc->last_tid;
+ req->tid = tid != 0 ? tid : ++monc->last_tid;
req->request->hdr.tid = cpu_to_le64(req->tid);
__insert_generic_request(monc, req);
monc->num_generic_requests++;
@@ -496,13 +495,24 @@
mutex_lock(&monc->mutex);
rb_erase(&req->node, &monc->generic_request_tree);
monc->num_generic_requests--;
- mutex_unlock(&monc->mutex);
if (!err)
err = req->result;
return err;
}
+static int do_generic_request(struct ceph_mon_client *monc,
+ struct ceph_mon_generic_request *req)
+{
+ int err;
+
+ mutex_lock(&monc->mutex);
+ err = __do_generic_request(monc, 0, req);
+ mutex_unlock(&monc->mutex);
+
+ return err;
+}
+
/*
* statfs
*/
@@ -579,6 +589,96 @@
}
EXPORT_SYMBOL(ceph_monc_do_statfs);
+static void handle_get_version_reply(struct ceph_mon_client *monc,
+ struct ceph_msg *msg)
+{
+ struct ceph_mon_generic_request *req;
+ u64 tid = le64_to_cpu(msg->hdr.tid);
+ void *p = msg->front.iov_base;
+ void *end = p + msg->front_alloc_len;
+ u64 handle;
+
+ dout("%s %p tid %llu\n", __func__, msg, tid);
+
+ ceph_decode_need(&p, end, 2*sizeof(u64), bad);
+ handle = ceph_decode_64(&p);
+ if (tid != 0 && tid != handle)
+ goto bad;
+
+ mutex_lock(&monc->mutex);
+ req = __lookup_generic_req(monc, handle);
+ if (req) {
+ *(u64 *)req->buf = ceph_decode_64(&p);
+ req->result = 0;
+ get_generic_request(req);
+ }
+ mutex_unlock(&monc->mutex);
+ if (req) {
+ complete_all(&req->completion);
+ put_generic_request(req);
+ }
+
+ return;
+bad:
+ pr_err("corrupt mon_get_version reply\n");
+ ceph_msg_dump(msg);
+}
+
+/*
+ * Send MMonGetVersion and wait for the reply.
+ *
+ * @what: one of "mdsmap", "osdmap" or "monmap"
+ */
+int ceph_monc_do_get_version(struct ceph_mon_client *monc, const char *what,
+ u64 *newest)
+{
+ struct ceph_mon_generic_request *req;
+ void *p, *end;
+ u64 tid;
+ int err;
+
+ req = kzalloc(sizeof(*req), GFP_NOFS);
+ if (!req)
+ return -ENOMEM;
+
+ kref_init(&req->kref);
+ req->buf = newest;
+ req->buf_len = sizeof(*newest);
+ init_completion(&req->completion);
+
+ req->request = ceph_msg_new(CEPH_MSG_MON_GET_VERSION,
+ sizeof(u64) + sizeof(u32) + strlen(what),
+ GFP_NOFS, true);
+ if (!req->request) {
+ err = -ENOMEM;
+ goto out;
+ }
+
+ req->reply = ceph_msg_new(CEPH_MSG_MON_GET_VERSION_REPLY, 1024,
+ GFP_NOFS, true);
+ if (!req->reply) {
+ err = -ENOMEM;
+ goto out;
+ }
+
+ p = req->request->front.iov_base;
+ end = p + req->request->front_alloc_len;
+
+ /* fill out request */
+ mutex_lock(&monc->mutex);
+ tid = ++monc->last_tid;
+ ceph_encode_64(&p, tid); /* handle */
+ ceph_encode_string(&p, end, what, strlen(what));
+
+ err = __do_generic_request(monc, tid, req);
+
+ mutex_unlock(&monc->mutex);
+out:
+ kref_put(&req->kref, release_generic_request);
+ return err;
+}
+EXPORT_SYMBOL(ceph_monc_do_get_version);
+
/*
* pool ops
*/
@@ -981,6 +1081,10 @@
handle_statfs_reply(monc, msg);
break;
+ case CEPH_MSG_MON_GET_VERSION_REPLY:
+ handle_get_version_reply(monc, msg);
+ break;
+
case CEPH_MSG_POOLOP_REPLY:
handle_poolop_reply(monc, msg);
break;
@@ -1029,6 +1133,15 @@
case CEPH_MSG_AUTH_REPLY:
m = ceph_msg_get(monc->m_auth_reply);
break;
+ case CEPH_MSG_MON_GET_VERSION_REPLY:
+ if (le64_to_cpu(hdr->tid) != 0)
+ return get_generic_reply(con, hdr, skip);
+
+ /*
+ * Older OSDs don't set reply tid even if the orignal
+ * request had a non-zero tid. Workaround this weirdness
+ * by falling through to the allocate case.
+ */
case CEPH_MSG_MON_MAP:
case CEPH_MSG_MDS_MAP:
case CEPH_MSG_OSD_MAP: