msg = ceph_msgpool_get(&osdc->msgpool_op_reply, 0);
else
msg = ceph_msg_new(CEPH_MSG_OSD_OPREPLY,
- OSD_OPREPLY_FRONT_LEN, gfp_flags);
+ OSD_OPREPLY_FRONT_LEN, gfp_flags, true);
if (!msg) {
ceph_osdc_put_request(req);
return NULL;
ceph_pagelist_init(req->r_trail);
}
/* create request message; allow space for oid */
- msg_size += 40;
+ msg_size += MAX_OBJ_NAME_SIZE;
if (snapc)
msg_size += sizeof(u64) * snapc->num_snaps;
if (use_mempool)
msg = ceph_msgpool_get(&osdc->msgpool_op, 0);
else
- msg = ceph_msg_new(CEPH_MSG_OSD_OP, msg_size, gfp_flags);
+ msg = ceph_msg_new(CEPH_MSG_OSD_OP, msg_size, gfp_flags, true);
if (!msg) {
ceph_osdc_put_request(req);
return NULL;
*/
static void __remove_osd(struct ceph_osd_client *osdc, struct ceph_osd *osd)
{
- dout("__remove_osd %p\n", osd);
+ dout("%s %p osd%d\n", __func__, osd, osd->o_osd);
BUG_ON(!list_empty(&osd->o_requests));
- rb_erase(&osd->o_node, &osdc->osds);
list_del_init(&osd->o_osd_lru);
- ceph_con_close(&osd->o_con);
- put_osd(osd);
+ rb_erase(&osd->o_node, &osdc->osds);
+ RB_CLEAR_NODE(&osd->o_node);
+}
+
+static void remove_osd(struct ceph_osd_client *osdc, struct ceph_osd *osd)
+{
+ dout("%s %p osd%d\n", __func__, osd, osd->o_osd);
+
+ if (!RB_EMPTY_NODE(&osd->o_node)) {
+ ceph_con_close(&osd->o_con);
+ __remove_osd(osdc, osd);
+ put_osd(osd);
+ }
}
static void remove_all_osds(struct ceph_osd_client *osdc)
while (!RB_EMPTY_ROOT(&osdc->osds)) {
struct ceph_osd *osd = rb_entry(rb_first(&osdc->osds),
struct ceph_osd, o_node);
- __remove_osd(osdc, osd);
+ remove_osd(osdc, osd);
}
mutex_unlock(&osdc->request_mutex);
}
list_for_each_entry_safe(osd, nosd, &osdc->osd_lru, o_osd_lru) {
if (time_before(jiffies, osd->lru_ttl))
break;
- __remove_osd(osdc, osd);
+ remove_osd(osdc, osd);
}
mutex_unlock(&osdc->request_mutex);
}
dout("__reset_osd %p osd%d\n", osd, osd->o_osd);
if (list_empty(&osd->o_requests) &&
list_empty(&osd->o_linger_requests)) {
- __remove_osd(osdc, osd);
+ remove_osd(osdc, osd);
} else if (memcmp(&osdc->osdmap->osd_addr[osd->o_osd],
&osd->o_con.peer_addr,
sizeof(osd->o_con.peer_addr)) == 0 &&
* Caller should hold map_sem for read and request_mutex.
*/
static int __map_request(struct ceph_osd_client *osdc,
- struct ceph_osd_request *req)
+ struct ceph_osd_request *req, int force_resend)
{
struct ceph_osd_request_head *reqhead = req->r_request->front.iov_base;
struct ceph_pg pgid;
num = err;
}
- if ((req->r_osd && req->r_osd->o_osd == o &&
+ if ((!force_resend &&
+ req->r_osd && req->r_osd->o_osd == o &&
req->r_sent >= req->r_osd->o_incarnation &&
req->r_num_pg_osds == num &&
memcmp(req->r_pg_osds, acting, sizeof(acting[0])*num) == 0) ||
{
struct rb_node *p, *n;
+ dout("%s %p\n", __func__, osdc);
for (p = rb_first(&osdc->osds); p; p = n) {
struct ceph_osd *osd = rb_entry(p, struct ceph_osd, o_node);
*
* Caller should hold map_sem for read and request_mutex.
*/
-static void kick_requests(struct ceph_osd_client *osdc)
+static void kick_requests(struct ceph_osd_client *osdc, int force_resend)
{
struct ceph_osd_request *req, *nreq;
struct rb_node *p;
int needmap = 0;
int err;
- dout("kick_requests\n");
+ dout("kick_requests %s\n", force_resend ? " (force resend)" : "");
mutex_lock(&osdc->request_mutex);
for (p = rb_first(&osdc->requests); p; p = rb_next(p)) {
req = rb_entry(p, struct ceph_osd_request, r_node);
- err = __map_request(osdc, req);
+ err = __map_request(osdc, req, force_resend);
if (err < 0)
continue; /* error */
if (req->r_osd == NULL) {
r_linger_item) {
dout("linger req=%p req->r_osd=%p\n", req, req->r_osd);
- err = __map_request(osdc, req);
+ err = __map_request(osdc, req, force_resend);
if (err == 0)
continue; /* no change and no osd was specified */
if (err < 0)
ceph_osdmap_destroy(osdc->osdmap);
osdc->osdmap = newmap;
}
- kick_requests(osdc);
+ kick_requests(osdc, 0);
reset_changed_osds(osdc);
} else {
dout("ignoring incremental map %u len %d\n",
"older than our %u\n", epoch, maplen,
osdc->osdmap->epoch);
} else {
+ int skipped_map = 0;
+
dout("taking full map %u len %d\n", epoch, maplen);
newmap = osdmap_decode(&p, p+maplen);
if (IS_ERR(newmap)) {
BUG_ON(!newmap);
oldmap = osdc->osdmap;
osdc->osdmap = newmap;
- if (oldmap)
+ if (oldmap) {
+ if (oldmap->epoch + 1 < newmap->epoch)
+ skipped_map = 1;
ceph_osdmap_destroy(oldmap);
- kick_requests(osdc);
+ }
+ kick_requests(osdc, skipped_map);
}
p += maplen;
nr_maps--;
* the request still han't been touched yet.
*/
if (req->r_sent == 0) {
- rc = __map_request(osdc, req);
+ rc = __map_request(osdc, req, 0);
if (rc < 0) {
if (nofail) {
dout("osdc_start_request failed map, "
" will retry %lld\n", req->r_tid);
rc = 0;
+ } else {
+ __unregister_request(osdc, req);
}
goto out_unlock;
}
if (front > req->r_reply->front.iov_len) {
pr_warning("get_reply front %d > preallocated %d\n",
front, (int)req->r_reply->front.iov_len);
- m = ceph_msg_new(CEPH_MSG_OSD_OPREPLY, front, GFP_NOFS);
+ m = ceph_msg_new(CEPH_MSG_OSD_OPREPLY, front, GFP_NOFS, false);
if (!m)
goto out;
ceph_msg_put(req->r_reply);
switch (type) {
case CEPH_MSG_OSD_MAP:
case CEPH_MSG_WATCH_NOTIFY:
- return ceph_msg_new(type, front, GFP_NOFS);
+ return ceph_msg_new(type, front, GFP_NOFS, false);
case CEPH_MSG_OSD_OPREPLY:
return get_reply(con, hdr, skip);
default: