2006-01-18 04:30:29 -05:00
|
|
|
/******************************************************************************
|
|
|
|
*******************************************************************************
|
|
|
|
**
|
2008-01-08 17:24:00 -05:00
|
|
|
** Copyright (C) 2005-2008 Red Hat, Inc. All rights reserved.
|
2006-01-18 04:30:29 -05:00
|
|
|
**
|
|
|
|
** This copyrighted material is made available to anyone wishing to use,
|
|
|
|
** modify, copy, or redistribute it subject to the terms and conditions
|
|
|
|
** of the GNU General Public License v.2.
|
|
|
|
**
|
|
|
|
*******************************************************************************
|
|
|
|
******************************************************************************/
|
|
|
|
|
|
|
|
#include "dlm_internal.h"
|
|
|
|
#include "lockspace.h"
|
|
|
|
#include "member.h"
|
|
|
|
#include "recoverd.h"
|
|
|
|
#include "recover.h"
|
|
|
|
#include "rcom.h"
|
|
|
|
#include "config.h"
|
|
|
|
|
|
|
|
static void add_ordered_member(struct dlm_ls *ls, struct dlm_member *new)
|
|
|
|
{
|
|
|
|
struct dlm_member *memb = NULL;
|
|
|
|
struct list_head *tmp;
|
|
|
|
struct list_head *newlist = &new->list;
|
|
|
|
struct list_head *head = &ls->ls_nodes;
|
|
|
|
|
|
|
|
list_for_each(tmp, head) {
|
|
|
|
memb = list_entry(tmp, struct dlm_member, list);
|
|
|
|
if (new->nodeid < memb->nodeid)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!memb)
|
|
|
|
list_add_tail(newlist, head);
|
|
|
|
else {
|
|
|
|
/* FIXME: can use list macro here */
|
|
|
|
newlist->prev = tmp->prev;
|
|
|
|
newlist->next = tmp;
|
|
|
|
tmp->prev->next = newlist;
|
|
|
|
tmp->prev = newlist;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static int dlm_add_member(struct dlm_ls *ls, int nodeid)
|
|
|
|
{
|
|
|
|
struct dlm_member *memb;
|
|
|
|
int w;
|
|
|
|
|
2006-01-20 03:47:07 -05:00
|
|
|
memb = kzalloc(sizeof(struct dlm_member), GFP_KERNEL);
|
2006-01-18 04:30:29 -05:00
|
|
|
if (!memb)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
w = dlm_node_weight(ls->ls_name, nodeid);
|
2007-07-18 18:27:43 -04:00
|
|
|
if (w < 0) {
|
|
|
|
kfree(memb);
|
2006-01-18 04:30:29 -05:00
|
|
|
return w;
|
2007-07-18 18:27:43 -04:00
|
|
|
}
|
2006-01-18 04:30:29 -05:00
|
|
|
|
|
|
|
memb->nodeid = nodeid;
|
|
|
|
memb->weight = w;
|
|
|
|
add_ordered_member(ls, memb);
|
|
|
|
ls->ls_num_nodes++;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void dlm_remove_member(struct dlm_ls *ls, struct dlm_member *memb)
|
|
|
|
{
|
|
|
|
list_move(&memb->list, &ls->ls_nodes_gone);
|
|
|
|
ls->ls_num_nodes--;
|
|
|
|
}
|
|
|
|
|
2008-01-08 17:24:00 -05:00
|
|
|
int dlm_is_member(struct dlm_ls *ls, int nodeid)
|
2006-01-18 04:30:29 -05:00
|
|
|
{
|
|
|
|
struct dlm_member *memb;
|
|
|
|
|
|
|
|
list_for_each_entry(memb, &ls->ls_nodes, list) {
|
|
|
|
if (memb->nodeid == nodeid)
|
2006-01-20 03:47:07 -05:00
|
|
|
return 1;
|
2006-01-18 04:30:29 -05:00
|
|
|
}
|
2006-01-20 03:47:07 -05:00
|
|
|
return 0;
|
2006-01-18 04:30:29 -05:00
|
|
|
}
|
|
|
|
|
|
|
|
int dlm_is_removed(struct dlm_ls *ls, int nodeid)
|
|
|
|
{
|
|
|
|
struct dlm_member *memb;
|
|
|
|
|
|
|
|
list_for_each_entry(memb, &ls->ls_nodes_gone, list) {
|
|
|
|
if (memb->nodeid == nodeid)
|
2006-01-20 03:47:07 -05:00
|
|
|
return 1;
|
2006-01-18 04:30:29 -05:00
|
|
|
}
|
2006-01-20 03:47:07 -05:00
|
|
|
return 0;
|
2006-01-18 04:30:29 -05:00
|
|
|
}
|
|
|
|
|
|
|
|
static void clear_memb_list(struct list_head *head)
|
|
|
|
{
|
|
|
|
struct dlm_member *memb;
|
|
|
|
|
|
|
|
while (!list_empty(head)) {
|
|
|
|
memb = list_entry(head->next, struct dlm_member, list);
|
|
|
|
list_del(&memb->list);
|
|
|
|
kfree(memb);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
void dlm_clear_members(struct dlm_ls *ls)
|
|
|
|
{
|
|
|
|
clear_memb_list(&ls->ls_nodes);
|
|
|
|
ls->ls_num_nodes = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
void dlm_clear_members_gone(struct dlm_ls *ls)
|
|
|
|
{
|
|
|
|
clear_memb_list(&ls->ls_nodes_gone);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void make_member_array(struct dlm_ls *ls)
|
|
|
|
{
|
|
|
|
struct dlm_member *memb;
|
|
|
|
int i, w, x = 0, total = 0, all_zero = 0, *array;
|
|
|
|
|
|
|
|
kfree(ls->ls_node_array);
|
|
|
|
ls->ls_node_array = NULL;
|
|
|
|
|
|
|
|
list_for_each_entry(memb, &ls->ls_nodes, list) {
|
|
|
|
if (memb->weight)
|
|
|
|
total += memb->weight;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* all nodes revert to weight of 1 if all have weight 0 */
|
|
|
|
|
|
|
|
if (!total) {
|
|
|
|
total = ls->ls_num_nodes;
|
|
|
|
all_zero = 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
ls->ls_total_weight = total;
|
|
|
|
|
|
|
|
array = kmalloc(sizeof(int) * total, GFP_KERNEL);
|
|
|
|
if (!array)
|
|
|
|
return;
|
|
|
|
|
|
|
|
list_for_each_entry(memb, &ls->ls_nodes, list) {
|
|
|
|
if (!all_zero && !memb->weight)
|
|
|
|
continue;
|
|
|
|
|
|
|
|
if (all_zero)
|
|
|
|
w = 1;
|
|
|
|
else
|
|
|
|
w = memb->weight;
|
|
|
|
|
|
|
|
DLM_ASSERT(x < total, printk("total %d x %d\n", total, x););
|
|
|
|
|
|
|
|
for (i = 0; i < w; i++)
|
|
|
|
array[x++] = memb->nodeid;
|
|
|
|
}
|
|
|
|
|
|
|
|
ls->ls_node_array = array;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* send a status request to all members just to establish comms connections */
|
|
|
|
|
2006-08-08 18:06:07 -04:00
|
|
|
static int ping_members(struct dlm_ls *ls)
|
2006-01-18 04:30:29 -05:00
|
|
|
{
|
|
|
|
struct dlm_member *memb;
|
2006-08-08 18:06:07 -04:00
|
|
|
int error = 0;
|
|
|
|
|
|
|
|
list_for_each_entry(memb, &ls->ls_nodes, list) {
|
|
|
|
error = dlm_recovery_stopped(ls);
|
|
|
|
if (error)
|
|
|
|
break;
|
|
|
|
error = dlm_rcom_status(ls, memb->nodeid);
|
|
|
|
if (error)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
if (error)
|
2006-08-08 18:08:42 -04:00
|
|
|
log_debug(ls, "ping_members aborted %d last nodeid %d",
|
|
|
|
error, ls->ls_recover_nodeid);
|
2006-08-08 18:06:07 -04:00
|
|
|
return error;
|
2006-01-18 04:30:29 -05:00
|
|
|
}
|
|
|
|
|
|
|
|
int dlm_recover_members(struct dlm_ls *ls, struct dlm_recover *rv, int *neg_out)
|
|
|
|
{
|
|
|
|
struct dlm_member *memb, *safe;
|
|
|
|
int i, error, found, pos = 0, neg = 0, low = -1;
|
|
|
|
|
2006-10-31 12:56:01 -05:00
|
|
|
/* previously removed members that we've not finished removing need to
|
|
|
|
count as a negative change so the "neg" recovery steps will happen */
|
|
|
|
|
|
|
|
list_for_each_entry(memb, &ls->ls_nodes_gone, list) {
|
|
|
|
log_debug(ls, "prev removed member %d", memb->nodeid);
|
|
|
|
neg++;
|
|
|
|
}
|
|
|
|
|
2006-01-18 04:30:29 -05:00
|
|
|
/* move departed members from ls_nodes to ls_nodes_gone */
|
|
|
|
|
|
|
|
list_for_each_entry_safe(memb, safe, &ls->ls_nodes, list) {
|
2006-01-20 03:47:07 -05:00
|
|
|
found = 0;
|
2006-01-18 04:30:29 -05:00
|
|
|
for (i = 0; i < rv->node_count; i++) {
|
|
|
|
if (memb->nodeid == rv->nodeids[i]) {
|
2006-01-20 03:47:07 -05:00
|
|
|
found = 1;
|
2006-01-18 04:30:29 -05:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!found) {
|
|
|
|
neg++;
|
|
|
|
dlm_remove_member(ls, memb);
|
|
|
|
log_debug(ls, "remove member %d", memb->nodeid);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2008-03-18 15:22:11 -04:00
|
|
|
/* Add an entry to ls_nodes_gone for members that were removed and
|
|
|
|
then added again, so that previous state for these nodes will be
|
|
|
|
cleared during recovery. */
|
|
|
|
|
|
|
|
for (i = 0; i < rv->new_count; i++) {
|
|
|
|
if (!dlm_is_member(ls, rv->new[i]))
|
|
|
|
continue;
|
|
|
|
log_debug(ls, "new nodeid %d is a re-added member", rv->new[i]);
|
|
|
|
|
|
|
|
memb = kzalloc(sizeof(struct dlm_member), GFP_KERNEL);
|
|
|
|
if (!memb)
|
|
|
|
return -ENOMEM;
|
|
|
|
memb->nodeid = rv->new[i];
|
|
|
|
list_add_tail(&memb->list, &ls->ls_nodes_gone);
|
|
|
|
neg++;
|
|
|
|
}
|
|
|
|
|
2006-01-18 04:30:29 -05:00
|
|
|
/* add new members to ls_nodes */
|
|
|
|
|
|
|
|
for (i = 0; i < rv->node_count; i++) {
|
|
|
|
if (dlm_is_member(ls, rv->nodeids[i]))
|
|
|
|
continue;
|
|
|
|
dlm_add_member(ls, rv->nodeids[i]);
|
|
|
|
pos++;
|
|
|
|
log_debug(ls, "add member %d", rv->nodeids[i]);
|
|
|
|
}
|
|
|
|
|
|
|
|
list_for_each_entry(memb, &ls->ls_nodes, list) {
|
|
|
|
if (low == -1 || memb->nodeid < low)
|
|
|
|
low = memb->nodeid;
|
|
|
|
}
|
|
|
|
ls->ls_low_nodeid = low;
|
|
|
|
|
|
|
|
make_member_array(ls);
|
|
|
|
dlm_set_recover_status(ls, DLM_RS_NODES);
|
|
|
|
*neg_out = neg;
|
|
|
|
|
2006-08-08 18:06:07 -04:00
|
|
|
error = ping_members(ls);
|
2007-05-18 10:03:35 -04:00
|
|
|
if (!error || error == -EPROTO) {
|
|
|
|
/* new_lockspace() may be waiting to know if the config
|
|
|
|
is good or bad */
|
|
|
|
ls->ls_members_result = error;
|
|
|
|
complete(&ls->ls_members_done);
|
|
|
|
}
|
2006-08-08 18:06:07 -04:00
|
|
|
if (error)
|
|
|
|
goto out;
|
2006-01-18 04:30:29 -05:00
|
|
|
|
|
|
|
error = dlm_recover_members_wait(ls);
|
2006-08-08 18:06:07 -04:00
|
|
|
out:
|
|
|
|
log_debug(ls, "total members %d error %d", ls->ls_num_nodes, error);
|
2006-01-18 04:30:29 -05:00
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
2007-09-27 16:53:38 -04:00
|
|
|
/* Userspace guarantees that dlm_ls_stop() has completed on all nodes before
|
|
|
|
dlm_ls_start() is called on any of them to start the new recovery. */
|
2006-01-18 04:30:29 -05:00
|
|
|
|
|
|
|
int dlm_ls_stop(struct dlm_ls *ls)
|
|
|
|
{
|
|
|
|
int new;
|
|
|
|
|
|
|
|
/*
|
2007-09-27 16:53:38 -04:00
|
|
|
* Prevent dlm_recv from being in the middle of something when we do
|
|
|
|
* the stop. This includes ensuring dlm_recv isn't processing a
|
|
|
|
* recovery message (rcom), while dlm_recoverd is aborting and
|
|
|
|
* resetting things from an in-progress recovery. i.e. we want
|
|
|
|
* dlm_recoverd to abort its recovery without worrying about dlm_recv
|
|
|
|
* processing an rcom at the same time. Stopping dlm_recv also makes
|
|
|
|
* it easy for dlm_receive_message() to check locking stopped and add a
|
|
|
|
* message to the requestqueue without races.
|
|
|
|
*/
|
|
|
|
|
|
|
|
down_write(&ls->ls_recv_active);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Abort any recovery that's in progress (see RECOVERY_STOP,
|
|
|
|
* dlm_recovery_stopped()) and tell any other threads running in the
|
|
|
|
* dlm to quit any processing (see RUNNING, dlm_locking_stopped()).
|
2006-01-18 04:30:29 -05:00
|
|
|
*/
|
|
|
|
|
|
|
|
spin_lock(&ls->ls_recover_lock);
|
|
|
|
set_bit(LSFL_RECOVERY_STOP, &ls->ls_flags);
|
|
|
|
new = test_and_clear_bit(LSFL_RUNNING, &ls->ls_flags);
|
|
|
|
ls->ls_recover_seq++;
|
|
|
|
spin_unlock(&ls->ls_recover_lock);
|
|
|
|
|
2007-09-27 16:53:38 -04:00
|
|
|
/*
|
|
|
|
* Let dlm_recv run again, now any normal messages will be saved on the
|
|
|
|
* requestqueue for later.
|
|
|
|
*/
|
|
|
|
|
|
|
|
up_write(&ls->ls_recv_active);
|
|
|
|
|
2006-01-18 04:30:29 -05:00
|
|
|
/*
|
|
|
|
* This in_recovery lock does two things:
|
|
|
|
* 1) Keeps this function from returning until all threads are out
|
|
|
|
* of locking routines and locking is truely stopped.
|
|
|
|
* 2) Keeps any new requests from being processed until it's unlocked
|
|
|
|
* when recovery is complete.
|
|
|
|
*/
|
|
|
|
|
|
|
|
if (new)
|
|
|
|
down_write(&ls->ls_in_recovery);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The recoverd suspend/resume makes sure that dlm_recoverd (if
|
2007-09-27 16:53:38 -04:00
|
|
|
* running) has noticed RECOVERY_STOP above and quit processing the
|
|
|
|
* previous recovery.
|
2006-01-18 04:30:29 -05:00
|
|
|
*/
|
|
|
|
|
|
|
|
dlm_recoverd_suspend(ls);
|
|
|
|
ls->ls_recover_status = 0;
|
|
|
|
dlm_recoverd_resume(ls);
|
2007-05-18 09:59:31 -04:00
|
|
|
|
|
|
|
if (!ls->ls_recover_begin)
|
|
|
|
ls->ls_recover_begin = jiffies;
|
2006-01-18 04:30:29 -05:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
int dlm_ls_start(struct dlm_ls *ls)
|
|
|
|
{
|
|
|
|
struct dlm_recover *rv = NULL, *rv_old;
|
2008-03-18 15:22:11 -04:00
|
|
|
int *ids = NULL, *new = NULL;
|
|
|
|
int error, ids_count = 0, new_count = 0;
|
2006-01-18 04:30:29 -05:00
|
|
|
|
2006-01-20 03:47:07 -05:00
|
|
|
rv = kzalloc(sizeof(struct dlm_recover), GFP_KERNEL);
|
2006-01-18 04:30:29 -05:00
|
|
|
if (!rv)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
2008-03-18 15:22:11 -04:00
|
|
|
error = dlm_nodeid_list(ls->ls_name, &ids, &ids_count,
|
|
|
|
&new, &new_count);
|
|
|
|
if (error < 0)
|
2006-01-18 04:30:29 -05:00
|
|
|
goto fail;
|
|
|
|
|
|
|
|
spin_lock(&ls->ls_recover_lock);
|
|
|
|
|
|
|
|
/* the lockspace needs to be stopped before it can be started */
|
|
|
|
|
|
|
|
if (!dlm_locking_stopped(ls)) {
|
|
|
|
spin_unlock(&ls->ls_recover_lock);
|
|
|
|
log_error(ls, "start ignored: lockspace running");
|
|
|
|
error = -EINVAL;
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
rv->nodeids = ids;
|
2008-03-18 15:22:11 -04:00
|
|
|
rv->node_count = ids_count;
|
|
|
|
rv->new = new;
|
|
|
|
rv->new_count = new_count;
|
2006-01-18 04:30:29 -05:00
|
|
|
rv->seq = ++ls->ls_recover_seq;
|
|
|
|
rv_old = ls->ls_recover_args;
|
|
|
|
ls->ls_recover_args = rv;
|
|
|
|
spin_unlock(&ls->ls_recover_lock);
|
|
|
|
|
|
|
|
if (rv_old) {
|
2008-03-18 15:22:11 -04:00
|
|
|
log_error(ls, "unused recovery %llx %d",
|
|
|
|
(unsigned long long)rv_old->seq, rv_old->node_count);
|
2006-01-18 04:30:29 -05:00
|
|
|
kfree(rv_old->nodeids);
|
2008-03-18 15:22:11 -04:00
|
|
|
kfree(rv_old->new);
|
2006-01-18 04:30:29 -05:00
|
|
|
kfree(rv_old);
|
|
|
|
}
|
|
|
|
|
|
|
|
dlm_recoverd_kick(ls);
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
fail:
|
|
|
|
kfree(rv);
|
|
|
|
kfree(ids);
|
2008-03-18 15:22:11 -04:00
|
|
|
kfree(new);
|
2006-01-18 04:30:29 -05:00
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|