path: root/fs/ocfs2/vote.c

/* -*- mode: c; c-basic-offset: 8; -*-
 * vim: noexpandtab sw=8 ts=8 sts=0:
 *
 * vote.c
 *
 * description here
 *
 * Copyright (C) 2003, 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/types.h>
#include <linux/slab.h>
#include <linux/highmem.h>
#include <linux/smp_lock.h>
#include <linux/kthread.h>

#include <cluster/heartbeat.h>
#include <cluster/nodemanager.h>
#include <cluster/tcp.h>

#include <dlm/dlmapi.h>

#define MLOG_MASK_PREFIX ML_VOTE
#include <cluster/masklog.h>

#include "ocfs2.h"

#include "alloc.h"
#include "dlmglue.h"
#include "extent_map.h"
#include "heartbeat.h"
#include "inode.h"
#include "journal.h"
#include "slot_map.h"
#include "vote.h"

#include "buffer_head_io.h"

#define OCFS2_MESSAGE_TYPE_VOTE     (0x1)
#define OCFS2_MESSAGE_TYPE_RESPONSE (0x2)
struct ocfs2_msg_hdr
{
	__be32 h_response_id; /* used to lookup message handle on sending
			    * node. */
	__be32 h_request;
	__be64 h_blkno;
	__be32 h_generation;
	__be32 h_node_num;    /* node sending this particular message. */
};

/* OCFS2_MAX_FILENAME_LEN is 255 characters, but we want to align this
 * for the network. */
#define OCFS2_VOTE_FILENAME_LEN 256
struct ocfs2_vote_msg
{
	struct ocfs2_msg_hdr v_hdr;
	union {
		__be32 v_generic1;
		__be32 v_orphaned_slot;	/* Used during delete votes */
		__be32 v_nlink;		/* Used during unlink votes */
	} md1;				/* Message type dependant 1 */
	__be32 v_unlink_namelen;
	__be64 v_unlink_parent;
	u8  v_unlink_dirent[OCFS2_VOTE_FILENAME_LEN];
};

/* Responses are given these values to maintain backwards
 * compatibility with older ocfs2 versions */
#define OCFS2_RESPONSE_OK		(0)
#define OCFS2_RESPONSE_BUSY		(-16)
#define OCFS2_RESPONSE_BAD_MSG		(-22)

struct ocfs2_response_msg
{
	struct ocfs2_msg_hdr r_hdr;
	__be32 r_response;
	__be32 r_orphaned_slot;
};

struct ocfs2_vote_work {
	struct list_head   w_list;
	struct ocfs2_vote_msg w_msg;
};

enum ocfs2_vote_request {
	OCFS2_VOTE_REQ_INVALID = 0,
	OCFS2_VOTE_REQ_DELETE,
	OCFS2_VOTE_REQ_UNLINK,
	OCFS2_VOTE_REQ_RENAME,
	OCFS2_VOTE_REQ_MOUNT,
	OCFS2_VOTE_REQ_UMOUNT,
	OCFS2_VOTE_REQ_LAST
};

static inline int ocfs2_is_valid_vote_request(int request)
{
	return OCFS2_VOTE_REQ_INVALID < request &&
		request < OCFS2_VOTE_REQ_LAST;
}

typedef void (*ocfs2_net_response_callback)(void *priv,
					    struct ocfs2_response_msg *resp);
struct ocfs2_net_response_cb {
	ocfs2_net_response_callback	rc_cb;
	void				*rc_priv;
};

struct ocfs2_net_wait_ctxt {
	struct list_head        n_list;
	u32                     n_response_id;
	wait_queue_head_t       n_event;
	struct ocfs2_node_map   n_node_map;
	int                     n_response; /* an agreggate response. 0 if
					     * all nodes are go, < 0 on any
					     * negative response from any
					     * node or network error. */
	struct ocfs2_net_response_cb *n_callback;
};

static void ocfs2_process_mount_request(struct ocfs2_super *osb,
					unsigned int node_num)
{
	mlog(0, "MOUNT vote from node %u\n", node_num);
	/* The other node only sends us this message when he has an EX
	 * on the superblock, so our recovery threads (if having been
	 * launched) are waiting on it.*/
	ocfs2_recovery_map_clear(osb, node_num);
	ocfs2_node_map_set_bit(osb, &osb->mounted_map, node_num);

	/* We clear the umount map here because a node may have been
	 * previously mounted, safely unmounted but never stopped
	 * heartbeating - in which case we'd have a stale entry. */
	ocfs2_node_map_clear_bit(osb, &osb->umount_map, node_num);
}

static void ocfs2_process_umount_request(struct ocfs2_super *osb,
					 unsigned int node_num)
{
	mlog(0, "UMOUNT vote from node %u\n