linux/fs/ocfs2/dlm/dlmconvert.c
xuejiufei e5054c9aef ocfs2/dlm: move lock to the tail of grant queue while doing in-place convert
We have found a bug when two nodes doing umount one after another.

1) Node 1 migrate a lockres that has 3 locks in grant queue such as
   N2(PR)<->N3(NL)<->N4(PR) to N2.  After migration, lvb of the lock
   N3(NL) and N4(PR) are empty on node 2 because migration target do not
   copy lvb to these two lock.

2) Node 3 want to convert to PR, it can be granted in
   __dlmconvert_master(), and the order of these locks is unchanged.  The
   lvb of the lock N3(PR) on node 2 is copyed from lockres in function
   dlm_update_lvb() while the lvb of lock N4(PR) is still empty.

3) Node 2 want to leave domain, it will migrate this lockres to node 3.
   Then node 2 will trigger the BUG in dlm_prepare_lvb_for_migration()
   when adding the lock N4(PR) to mres with the following message because
   the lvb of mres is already copied from lock N3(PR), but the lvb of lock
   N4(PR) is empty.

"Mismatched lvb in lock cookie=%u:%llu, name=%.*s, node=%u"

[akpm@linux-foundation.org: tweak comment]
Signed-off-by: xuejiufei <xuejiufei@huawei.com>
Acked-by: Joseph Qi <joseph.qi@huawei.com>
Cc: Mark Fasheh <mfasheh@suse.de>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-03-25 16:37:42 -07:00

573 lines
16 KiB
C

/* -*- mode: c; c-basic-offset: 8; -*-
* vim: noexpandtab sw=8 ts=8 sts=0:
*
* dlmconvert.c
*
* underlying calls for lock conversion
*
* Copyright (C) 2004 Oracle. All rights reserved.
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public
* License as published by the Free Software Foundation; either
* version 2 of the License, or (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* General Public License for more details.
*
* You should have received a copy of the GNU General Public
* License along with this program; if not, write to the
* Free Software Foundation, Inc., 59 Temple Place - Suite 330,
* Boston, MA 021110-1307, USA.
*
*/
#include <linux/module.h>
#include <linux/fs.h>
#include <linux/types.h>
#include <linux/highmem.h>
#include <linux/init.h>
#include <linux/sysctl.h>
#include <linux/random.h>
#include <linux/blkdev.h>
#include <linux/socket.h>
#include <linux/inet.h>
#include <linux/spinlock.h>
#include "cluster/heartbeat.h"
#include "cluster/nodemanager.h"
#include "cluster/tcp.h"
#include "dlmapi.h"
#include "dlmcommon.h"
#include "dlmconvert.h"
#define MLOG_MASK_PREFIX ML_DLM
#include "cluster/masklog.h"
/* NOTE: __dlmconvert_master is the only function in here that
* needs a spinlock held on entry (res->spinlock) and it is the
* only one that holds a lock on exit (res->spinlock).
* All other functions in here need no locks and drop all of
* the locks that they acquire. */
static enum dlm_status __dlmconvert_master(struct dlm_ctxt *dlm,
struct dlm_lock_resource *res,
struct dlm_lock *lock, int flags,
int type, int *call_ast,
int *kick_thread);
static enum dlm_status dlm_send_remote_convert_request(struct dlm_ctxt *dlm,
struct dlm_lock_resource *res,
struct dlm_lock *lock, int flags, int type);
/*
* this is only called directly by dlmlock(), and only when the
* local node is the owner of the lockres
* locking:
* caller needs: none
* taken: takes and drops res->spinlock
* held on exit: none
* returns: see __dlmconvert_master
*/
enum dlm_status dlmconvert_master(struct dlm_ctxt *dlm,
struct dlm_lock_resource *res,
struct dlm_lock *lock, int flags, int type)
{
int call_ast = 0, kick_thread = 0;
enum dlm_status status;
spin_lock(&res->spinlock);
/* we are not in a network handler, this is fine */
__dlm_wait_on_lockres(res);
__dlm_lockres_reserve_ast(res);
res->state |= DLM_LOCK_RES_IN_PROGRESS;
status = __dlmconvert_master(dlm, res, lock, flags, type,
&call_ast, &kick_thread);
res->state &= ~DLM_LOCK_RES_IN_PROGRESS;
spin_unlock(&res->spinlock);
wake_up(&res->wq);
if (status != DLM_NORMAL && status != DLM_NOTQUEUED)
dlm_error(status);
/* either queue the ast or release it */
if (call_ast)
dlm_queue_ast(dlm, lock);
else
dlm_lockres_release_ast(dlm, res);
if (kick_thread)
dlm_kick_thread(dlm, res);
return status;
}
/* performs lock conversion at the lockres master site
* locking:
* caller needs: res->spinlock
* taken: takes and drops lock->spinlock
* held on exit: res->spinlock
* returns: DLM_NORMAL, DLM_NOTQUEUED, DLM_DENIED
* call_ast: whether ast should be called for this lock
* kick_thread: whether dlm_kick_thread should be called
*/
static enum dlm_status __dlmconvert_master(struct dlm_ctxt *dlm,
struct dlm_lock_resource *res,
struct dlm_lock *lock, int flags,
int type, int *call_ast,
int *kick_thread)
{
enum dlm_status status = DLM_NORMAL;
struct dlm_lock *tmplock=NULL;
assert_spin_locked(&res->spinlock);
mlog(0, "type=%d, convert_type=%d, new convert_type=%d\n",
lock->ml.type, lock->ml.convert_type, type);
spin_lock(&lock->spinlock);
/* already converting? */
if (lock->ml.convert_type != LKM_IVMODE) {
mlog(ML_ERROR, "attempted to convert a lock with a lock "
"conversion pending\n");
status = DLM_DENIED;
goto unlock_exit;
}
/* must be on grant queue to convert */
if (!dlm_lock_on_list(&res->granted, lock)) {
mlog(ML_ERROR, "attempted to convert a lock not on grant "
"queue\n");
status = DLM_DENIED;
goto unlock_exit;
}
if (flags & LKM_VALBLK) {
switch (lock->ml.type) {
case LKM_EXMODE:
/* EX + LKM_VALBLK + convert == set lvb */
mlog(0, "will set lvb: converting %s->%s\n",
dlm_lock_mode_name(lock->ml.type),
dlm_lock_mode_name(type));
lock->lksb->flags |= DLM_LKSB_PUT_LVB;
break;
case LKM_PRMODE:
case LKM_NLMODE:
/* refetch if new level is not NL */
if (type > LKM_NLMODE) {
mlog(0, "will fetch new value into "
"lvb: converting %s->%s\n",
dlm_lock_mode_name(lock->ml.type),
dlm_lock_mode_name(type));
lock->lksb->flags |= DLM_LKSB_GET_LVB;
} else {
mlog(0, "will NOT fetch new value "
"into lvb: converting %s->%s\n",
dlm_lock_mode_name(lock->ml.type),
dlm_lock_mode_name(type));
flags &= ~(LKM_VALBLK);
}
break;
}
}
/* in-place downconvert? */
if (type <= lock->ml.type)
goto grant;
/* upconvert from here on */
status = DLM_NORMAL;
list_for_each_entry(tmplock, &res->granted, list) {
if (tmplock == lock)
continue;
if (!dlm_lock_compatible(tmplock->ml.type, type))
goto switch_queues;
}
list_for_each_entry(tmplock, &res->converting, list) {
if (!dlm_lock_compatible(tmplock->ml.type, type))
goto switch_queues;
/* existing conversion requests take precedence */
if (!dlm_lock_compatible(tmplock->ml.convert_type, type))
goto switch_queues;
}
/* fall thru to grant */
grant:
mlog(0, "res %.*s, granting %s lock\n", res->lockname.len,
res->lockname.name, dlm_lock_mode_name(type));
/* immediately grant the new lock type */
lock->lksb->status = DLM_NORMAL;
if (lock->ml.node == dlm->node_num)
mlog(0, "doing in-place convert for nonlocal lock\n");
lock->ml.type = type;
if (lock->lksb->flags & DLM_LKSB_PUT_LVB)
memcpy(res->lvb, lock->lksb->lvb, DLM_LVB_LEN);
/*
* Move the lock to the tail because it may be the only lock which has
* an invalid lvb.
*/
list_move_tail(&lock->list, &res->granted);
status = DLM_NORMAL;
*call_ast = 1;
goto unlock_exit;
switch_queues:
if (flags & LKM_NOQUEUE) {
mlog(0, "failed to convert NOQUEUE lock %.*s from "
"%d to %d...\n", res->lockname.len, res->lockname.name,
lock->ml.type, type);
status = DLM_NOTQUEUED;
goto unlock_exit;
}
mlog(0, "res %.*s, queueing...\n", res->lockname.len,
res->lockname.name);
lock->ml.convert_type = type;
/* do not alter lock refcount. switching lists. */
list_move_tail(&lock->list, &res->converting);
unlock_exit:
spin_unlock(&lock->spinlock);
if (status == DLM_DENIED) {
__dlm_print_one_lock_resource(res);
}
if (status == DLM_NORMAL)
*kick_thread = 1;
return status;
}
void dlm_revert_pending_convert(struct dlm_lock_resource *res,
struct dlm_lock *lock)
{
/* do not alter lock refcount. switching lists. */
list_move_tail(&lock->list, &res->granted);
lock->ml.convert_type = LKM_IVMODE;
lock->lksb->flags &= ~(DLM_LKSB_GET_LVB|DLM_LKSB_PUT_LVB);
}
/* messages the master site to do lock conversion
* locking:
* caller needs: none
* taken: takes and drops res->spinlock, uses DLM_LOCK_RES_IN_PROGRESS
* held on exit: none
* returns: DLM_NORMAL, DLM_RECOVERING, status from remote node
*/
enum dlm_status dlmconvert_remote(struct dlm_ctxt *dlm,
struct dlm_lock_resource *res,
struct dlm_lock *lock, int flags, int type)
{
enum dlm_status status;
u8 old_owner = res->owner;
mlog(0, "type=%d, convert_type=%d, busy=%d\n", lock->ml.type,
lock->ml.convert_type, res->state & DLM_LOCK_RES_IN_PROGRESS);
spin_lock(&res->spinlock);
if (res->state & DLM_LOCK_RES_RECOVERING) {
mlog(0, "bailing out early since res is RECOVERING "
"on secondary queue\n");
/* __dlm_print_one_lock_resource(res); */
status = DLM_RECOVERING;
goto bail;
}
/* will exit this call with spinlock held */
__dlm_wait_on_lockres(res);
if (lock->ml.convert_type != LKM_IVMODE) {
__dlm_print_one_lock_resource(res);
mlog(ML_ERROR, "converting a remote lock that is already "
"converting! (cookie=%u:%llu, conv=%d)\n",
dlm_get_lock_cookie_node(be64_to_cpu(lock->ml.cookie)),
dlm_get_lock_cookie_seq(be64_to_cpu(lock->ml.cookie)),
lock->ml.convert_type);
status = DLM_DENIED;
goto bail;
}
if (lock->ml.type == type && lock->ml.convert_type == LKM_IVMODE) {
mlog(0, "last convert request returned DLM_RECOVERING, but "
"owner has already queued and sent ast to me. res %.*s, "
"(cookie=%u:%llu, type=%d, conv=%d)\n",
res->lockname.len, res->lockname.name,
dlm_get_lock_cookie_node(be64_to_cpu(lock->ml.cookie)),
dlm_get_lock_cookie_seq(be64_to_cpu(lock->ml.cookie)),
lock->ml.type, lock->ml.convert_type);
status = DLM_NORMAL;
goto bail;
}
res->state |= DLM_LOCK_RES_IN_PROGRESS;
/* move lock to local convert queue */
/* do not alter lock refcount. switching lists. */
list_move_tail(&lock->list, &res->converting);
lock->convert_pending = 1;
lock->ml.convert_type = type;
if (flags & LKM_VALBLK) {
if (lock->ml.type == LKM_EXMODE) {
flags |= LKM_PUT_LVB;
lock->lksb->flags |= DLM_LKSB_PUT_LVB;
} else {
if (lock->ml.convert_type == LKM_NLMODE)
flags &= ~LKM_VALBLK;
else {
flags |= LKM_GET_LVB;
lock->lksb->flags |= DLM_LKSB_GET_LVB;
}
}
}
spin_unlock(&res->spinlock);
/* no locks held here.
* need to wait for a reply as to whether it got queued or not. */
status = dlm_send_remote_convert_request(dlm, res, lock, flags, type);
spin_lock(&res->spinlock);
res->state &= ~DLM_LOCK_RES_IN_PROGRESS;
lock->convert_pending = 0;
/* if it failed, move it back to granted queue.
* if master returns DLM_NORMAL and then down before sending ast,
* it may have already been moved to granted queue, reset to
* DLM_RECOVERING and retry convert */
if (status != DLM_NORMAL) {
if (status != DLM_NOTQUEUED)
dlm_error(status);
dlm_revert_pending_convert(res, lock);
} else if ((res->state & DLM_LOCK_RES_RECOVERING) ||
(old_owner != res->owner)) {
mlog(0, "res %.*s is in recovering or has been recovered.\n",
res->lockname.len, res->lockname.name);
status = DLM_RECOVERING;
}
bail:
spin_unlock(&res->spinlock);
/* TODO: should this be a wake_one? */
/* wake up any IN_PROGRESS waiters */
wake_up(&res->wq);
return status;
}
/* sends DLM_CONVERT_LOCK_MSG to master site
* locking:
* caller needs: none
* taken: none
* held on exit: none
* returns: DLM_NOLOCKMGR, status from remote node
*/
static enum dlm_status dlm_send_remote_convert_request(struct dlm_ctxt *dlm,
struct dlm_lock_resource *res,
struct dlm_lock *lock, int flags, int type)
{
struct dlm_convert_lock convert;
int tmpret;
enum dlm_status ret;
int status = 0;
struct kvec vec[2];
size_t veclen = 1;
mlog(0, "%.*s\n", res->lockname.len, res->lockname.name);
memset(&convert, 0, sizeof(struct dlm_convert_lock));
convert.node_idx = dlm->node_num;
convert.requested_type = type;
convert.cookie = lock->ml.cookie;
convert.namelen = res->lockname.len;
convert.flags = cpu_to_be32(flags);
memcpy(convert.name, res->lockname.name, convert.namelen);
vec[0].iov_len = sizeof(struct dlm_convert_lock);
vec[0].iov_base = &convert;
if (flags & LKM_PUT_LVB) {
/* extra data to send if we are updating lvb */
vec[1].iov_len = DLM_LVB_LEN;
vec[1].iov_base = lock->lksb->lvb;
veclen++;
}
tmpret = o2net_send_message_vec(DLM_CONVERT_LOCK_MSG, dlm->key,
vec, veclen, res->owner, &status);
if (tmpret >= 0) {
// successfully sent and received
ret = status; // this is already a dlm_status
if (ret == DLM_RECOVERING) {
mlog(0, "node %u returned DLM_RECOVERING from convert "
"message!\n", res->owner);
} else if (ret == DLM_MIGRATING) {
mlog(0, "node %u returned DLM_MIGRATING from convert "
"message!\n", res->owner);
} else if (ret == DLM_FORWARD) {
mlog(0, "node %u returned DLM_FORWARD from convert "
"message!\n", res->owner);
} else if (ret != DLM_NORMAL && ret != DLM_NOTQUEUED)
dlm_error(ret);
} else {
mlog(ML_ERROR, "Error %d when sending message %u (key 0x%x) to "
"node %u\n", tmpret, DLM_CONVERT_LOCK_MSG, dlm->key,
res->owner);
if (dlm_is_host_down(tmpret)) {
/* instead of logging the same network error over
* and over, sleep here and wait for the heartbeat
* to notice the node is dead. times out after 5s. */
dlm_wait_for_node_death(dlm, res->owner,
DLM_NODE_DEATH_WAIT_MAX);
ret = DLM_RECOVERING;
mlog(0, "node %u died so returning DLM_RECOVERING "
"from convert message!\n", res->owner);
} else {
ret = dlm_err_to_dlm_status(tmpret);
}
}
return ret;
}
/* handler for DLM_CONVERT_LOCK_MSG on master site
* locking:
* caller needs: none
* taken: takes and drop res->spinlock
* held on exit: none
* returns: DLM_NORMAL, DLM_IVLOCKID, DLM_BADARGS,
* status from __dlmconvert_master
*/
int dlm_convert_lock_handler(struct o2net_msg *msg, u32 len, void *data,
void **ret_data)
{
struct dlm_ctxt *dlm = data;
struct dlm_convert_lock *cnv = (struct dlm_convert_lock *)msg->buf;
struct dlm_lock_resource *res = NULL;
struct dlm_lock *lock = NULL;
struct dlm_lock *tmp_lock;
struct dlm_lockstatus *lksb;
enum dlm_status status = DLM_NORMAL;
u32 flags;
int call_ast = 0, kick_thread = 0, ast_reserved = 0, wake = 0;
if (!dlm_grab(dlm)) {
dlm_error(DLM_REJECTED);
return DLM_REJECTED;
}
mlog_bug_on_msg(!dlm_domain_fully_joined(dlm),
"Domain %s not fully joined!\n", dlm->name);
if (cnv->namelen > DLM_LOCKID_NAME_MAX) {
status = DLM_IVBUFLEN;
dlm_error(status);
goto leave;
}
flags = be32_to_cpu(cnv->flags);
if ((flags & (LKM_PUT_LVB|LKM_GET_LVB)) ==
(LKM_PUT_LVB|LKM_GET_LVB)) {
mlog(ML_ERROR, "both PUT and GET lvb specified\n");
status = DLM_BADARGS;
goto leave;
}
mlog(0, "lvb: %s\n", flags & LKM_PUT_LVB ? "put lvb" :
(flags & LKM_GET_LVB ? "get lvb" : "none"));
status = DLM_IVLOCKID;
res = dlm_lookup_lockres(dlm, cnv->name, cnv->namelen);
if (!res) {
dlm_error(status);
goto leave;
}
spin_lock(&res->spinlock);
status = __dlm_lockres_state_to_status(res);
if (status != DLM_NORMAL) {
spin_unlock(&res->spinlock);
dlm_error(status);
goto leave;
}
list_for_each_entry(tmp_lock, &res->granted, list) {
if (tmp_lock->ml.cookie == cnv->cookie &&
tmp_lock->ml.node == cnv->node_idx) {
lock = tmp_lock;
dlm_lock_get(lock);
break;
}
}
spin_unlock(&res->spinlock);
if (!lock) {
status = DLM_IVLOCKID;
mlog(ML_ERROR, "did not find lock to convert on grant queue! "
"cookie=%u:%llu\n",
dlm_get_lock_cookie_node(be64_to_cpu(cnv->cookie)),
dlm_get_lock_cookie_seq(be64_to_cpu(cnv->cookie)));
dlm_print_one_lock_resource(res);
goto leave;
}
/* found the lock */
lksb = lock->lksb;
/* see if caller needed to get/put lvb */
if (flags & LKM_PUT_LVB) {
BUG_ON(lksb->flags & (DLM_LKSB_PUT_LVB|DLM_LKSB_GET_LVB));
lksb->flags |= DLM_LKSB_PUT_LVB;
memcpy(&lksb->lvb[0], &cnv->lvb[0], DLM_LVB_LEN);
} else if (flags & LKM_GET_LVB) {
BUG_ON(lksb->flags & (DLM_LKSB_PUT_LVB|DLM_LKSB_GET_LVB));
lksb->flags |= DLM_LKSB_GET_LVB;
}
spin_lock(&res->spinlock);
status = __dlm_lockres_state_to_status(res);
if (status == DLM_NORMAL) {
__dlm_lockres_reserve_ast(res);
ast_reserved = 1;
res->state |= DLM_LOCK_RES_IN_PROGRESS;
status = __dlmconvert_master(dlm, res, lock, flags,
cnv->requested_type,
&call_ast, &kick_thread);
res->state &= ~DLM_LOCK_RES_IN_PROGRESS;
wake = 1;
}
spin_unlock(&res->spinlock);
if (wake)
wake_up(&res->wq);
if (status != DLM_NORMAL) {
if (status != DLM_NOTQUEUED)
dlm_error(status);
lksb->flags &= ~(DLM_LKSB_GET_LVB|DLM_LKSB_PUT_LVB);
}
leave:
if (lock)
dlm_lock_put(lock);
/* either queue the ast or release it, if reserved */
if (call_ast)
dlm_queue_ast(dlm, lock);
else if (ast_reserved)
dlm_lockres_release_ast(dlm, res);
if (kick_thread)
dlm_kick_thread(dlm, res);
if (res)
dlm_lockres_put(res);
dlm_put(dlm);
return status;
}