[net-next-2.6.git] / fs / dlm / rcom.c

/******************************************************************************
*******************************************************************************
**
**  Copyright (C) Sistina Software, Inc.  1997-2003  All rights reserved.
**  Copyright (C) 2005-2008 Red Hat, Inc.  All rights reserved.
**
**  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 "lowcomms.h"
#include "midcomms.h"
#include "rcom.h"
#include "recover.h"
#include "dir.h"
#include "config.h"
#include "memory.h"
#include "lock.h"
#include "util.h"


static int rcom_response(struct dlm_ls *ls)
{
	return test_bit(LSFL_RCOM_READY, &ls->ls_flags);
}

static int create_rcom(struct dlm_ls *ls, int to_nodeid, int type, int len,
		       struct dlm_rcom **rc_ret, struct dlm_mhandle **mh_ret)
{
	struct dlm_rcom *rc;
	struct dlm_mhandle *mh;
	char *mb;
	int mb_len = sizeof(struct dlm_rcom) + len;

	mh = dlm_lowcomms_get_buffer(to_nodeid, mb_len, ls->ls_allocation, &mb);
	if (!mh) {
		log_print("create_rcom to %d type %d len %d ENOBUFS",
			  to_nodeid, type, len);
		return -ENOBUFS;
	}
	memset(mb, 0, mb_len);

	rc = (struct dlm_rcom *) mb;

	rc->rc_header.h_version = (DLM_HEADER_MAJOR | DLM_HEADER_MINOR);
	rc->rc_header.h_lockspace = ls->ls_global_id;
	rc->rc_header.h_nodeid = dlm_our_nodeid();
	rc->rc_header.h_length = mb_len;
	rc->rc_header.h_cmd = DLM_RCOM;

	rc->rc_type = type;

	spin_lock(&ls->ls_recover_lock);
	rc->rc_seq = ls->ls_recover_seq;
	spin_unlock(&ls->ls_recover_lock);

	*mh_ret = mh;
	*rc_ret = rc;
	return 0;
}

static void send_rcom(struct dlm_ls *ls, struct dlm_mhandle *mh,
		      struct dlm_rcom *rc)
{
	dlm_rcom_out(rc);
	dlm_lowcomms_commit_buffer(mh);
}

/* When replying to a status request, a node also sends back its
   configuration values.  The requesting node then checks that the remote
   node is configured the same way as itself. */

static void make_config(struct dlm_ls *ls, struct rcom_config *rf)
{
	rf->rf_lvblen = cpu_to_le32(ls->ls_lvblen);
	rf->rf_lsflags = cpu_to_le32(ls->ls_exflags);
}

static int check_config(struct dlm_ls *ls, struct dlm_rcom *rc, int nodeid)
{
	struct rcom_config *rf = (struct rcom_config *) rc->rc_buf;

	if ((rc->rc_header.h_version & 0xFFFF0000) != DLM_HEADER_MAJOR) {
		log_error(ls, "version mismatch: %x nodeid %d: %x",
			  DLM_HEADER_MAJOR | DLM_HEADER_MINOR, nodeid,
			  rc->rc_header.h_version);
		return -EPROTO;
	}

	if (le32_to_cpu(rf->rf_lvblen) != ls->ls_lvblen ||
	    le32_to_cpu(rf->rf_lsflags) != ls->ls_exflags) {
		log_error(ls, "config mismatch: %d,%x nodeid %d: %d,%x",
			  ls->ls_lvblen, ls->ls_exflags, nodeid,
			  le32_to_cpu(rf->rf_lvblen),
			  le32_to_cpu(rf->rf_lsflags));
		return -EPROTO;
	}
	return 0;
}

static void allow_sync_reply(struct dlm_ls *ls, uint64_t *new_seq)
{
	spin_lock(&ls->ls_rcom_spin);
	*new_seq = ++ls->ls_rcom_seq;
	set_bit(LSFL_RCOM_WAIT, &ls->ls_flags);
	spin_unlock(&ls->ls_rcom_spin);
}

static void disallow_sync_reply(struct dlm_ls *ls)
{
	spin_lock(&ls->ls_rcom_spin);
	clear_bit(LSFL_RCOM_WAIT, &ls->ls_flags);
	clear_bit(LSFL_RCOM_READY, &ls->ls_flags);
	spin_unlock(&ls->ls_rcom_spin);
}

int dlm_rcom_status(struct dlm_ls *ls, int nodeid)
{
	struct dlm_rcom *rc;
	struct dlm_mhandle *mh;
	int error = 0;

	ls->ls_recover_nodeid = nodeid;

	if (nodeid == dlm_our_nodeid()) {
		rc = ls->ls_recover_buf;
		rc->rc_result = dlm_recover_status(ls);
		goto out;
	}

	error = create_rcom(ls, nodeid, DLM_RCOM_STATUS, 0, &rc, &mh);
	if (error)
		goto out;

	allow_sync_reply(ls, &rc->rc_id);
	memset(ls->ls_recover_buf, 0, dlm_config.ci_buffer_size);

	send_rcom(ls, mh, rc);

	error = dlm_wait_function(ls, &rcom_response);
	disallow_sync_reply(ls);
	if (error)
		goto out;

	rc = ls->ls_recover_buf;

	if (rc->rc_result == -ESRCH) {
		/* we pretend the remote lockspace exists with 0 status */
		log_debug(ls, "remote node %d not ready", nodeid);
		rc->rc_result = 0;
	} else
		error = check_config(ls, rc, nodeid);
	/* the caller looks at rc_result for the remote recovery status */
 out:
	return error;
}

static void receive_rcom_status(struct dlm_ls *ls, struct dlm_rcom *rc_in)
{
	struct dlm_rcom *rc;
	struct dlm_mhandle *mh;
	int error, nodeid = rc_in->rc_header.h_nodeid;

	error = create_rcom(ls, nodeid, DLM_RCOM_STATUS_REPLY,
			    sizeof(struct rcom_config), &rc, &mh);
	if (error)
		return;
	rc->rc_id = rc_in->rc_id;
	rc->rc_seq_reply = rc_in->rc_seq;
	rc->rc_result = dlm_recover_status(ls);
	make_config(ls, (struct rcom_config *) rc->rc_buf);

	send_rcom(ls, mh, rc);
}

static void receive_sync_reply(struct dlm_ls *ls, struct dlm_rcom *rc_in)
{
	spin_lock(&ls->ls_rcom_spin);
	if (!test_bit(LSFL_RCOM_WAIT, &ls->ls_flags) ||
	    rc_in->rc_id != ls->ls_rcom_seq) {
		log_debug(ls, "reject reply %d from %d seq %llx expect %llx",
			  rc_in->rc_type, rc_in->rc_header.h_nodeid,
			  (unsigned long long)rc_in->rc_id,
			  (unsigned long long)ls->ls_rcom_seq);
		goto out;
	}
	memcpy(ls->ls_recover_buf, rc_in, rc_in->rc_header.h_length);
	set_bit(LSFL_RCOM_READY, &ls->ls_flags);
	clear_bit(LSFL_RCOM_WAIT, &ls->ls_flags);
	wake_up(&ls->ls_wait_general);
 out:
	spin_unlock(&ls->ls_rcom_spin);
}

int dlm_rcom_names(struct dlm_ls *ls, int nodeid, char *last_name, int last_len)
{
	struct dlm_rcom *rc;
	struct dlm_mhandle *mh;
	int error = 0;
	int max_size = dlm_config.ci_buffer_size - sizeof(struct dlm_rcom);

	ls->ls_recover_nodeid = nodeid;

	if (nodeid == dlm_our_nodeid()) {
		dlm_copy_master_names(ls, last_name, last_len,
		                      ls->ls_recover_buf->rc_buf,
		                      max_size, nodeid);
		goto out;
	}

	error = create_rcom(ls, nodeid, DLM_RCOM_NAMES, last_len, &rc, &mh);
	if (error)
		goto out;
	memcpy(rc->rc_buf, last_name, last_len);

	allow_sync_reply(ls, &rc->rc_id);
	memset(ls->ls_recover_buf, 0, dlm_config.ci_buffer_size);

	send_rcom(ls, mh, rc);

	error = dlm_wait_function(ls, &rcom_response);
	disallow_sync_reply(ls);
 out:
	return error;
}

static void receive_rcom_names(struct dlm_ls *ls, struct dlm_rcom *rc_in)
{
	struct dlm_rcom *rc;
	struct dlm_mhandle *mh;
	int error, inlen, outlen, nodeid;

	nodeid = rc_in->rc_header.h_nodeid;
	inlen = rc_in->rc_header.h_length - sizeof(struct dlm_rcom);
	outlen = dlm_config.ci_buffer_size - sizeof(struct dlm_rcom);

	error = create_rcom(ls, nodeid, DLM_RCOM_NAMES_REPLY, outlen, &rc, &mh);
	if (error)
		return;
	rc->rc_id = rc_in->rc_id;
	rc->rc_seq_reply = rc_in->rc_seq;

	dlm_copy_master_names(ls, rc_in->rc_buf, inlen, rc->rc_buf, outlen,
			      nodeid);
	send_rcom(ls, mh, rc);
}

int dlm_send_rcom_lookup(struct dlm_rsb *r, int dir_nodeid)
{
	struct dlm_rcom *rc;
	struct dlm_mhandle *mh;
	struct dlm_ls *ls = r->res_ls;
	int error;

	error = create_rcom(ls, dir_nodeid, DLM_RCOM_LOOKUP, r->res_length,
			    &rc, &mh);
	if (error)
		goto out;
	memcpy(rc->rc_buf, r->res_name, r->res_length);
	rc->rc_id = (unsigned long) r;

	send_rcom(ls, mh, rc);
 out:
	return error;
}

static void receive_rcom_lookup(struct dlm_ls *ls, struct dlm_rcom *rc_in)
{
	struct dlm_rcom *rc;
	struct dlm_mhandle *mh;
	int error, ret_nodeid, nodeid = rc_in->rc_header.h_nodeid;
	int len = rc_in->rc_header.h_length - sizeof(struct dlm_rcom);

	error = create_rcom(ls, nodeid, DLM_RCOM_LOOKUP_REPLY, 0, &rc, &mh);
	if (error)
		return;

	error = dlm_dir_lookup(ls, nodeid, rc_in->rc_buf, len, &ret_nodeid);
	if (error)
		ret_nodeid = error;
	rc->rc_result = ret_nodeid;
	rc->rc_id = rc_in->rc_id;
	rc->rc_seq_reply = rc_in->rc_seq;

	send_rcom(ls, mh, rc);
}

static void receive_rcom_lookup_reply(struct dlm_ls *ls, struct dlm_rcom *rc_in)
{
	dlm_recover_master_reply(ls, rc_in);
}

static void pack_rcom_lock(struct dlm_rsb *r, struct dlm_lkb *lkb,
			   struct rcom_lock *rl)
{
	memset(rl, 0, sizeof(*rl));

	rl->rl_ownpid = cpu_to_le32(lkb->lkb_ownpid);
	rl->rl_lkid = cpu_to_le32(lkb->lkb_id);
	rl->rl_exflags = cpu_to_le32(lkb->lkb_exflags);
	rl->rl_flags = cpu_to_le32(lkb->lkb_flags);
	rl->rl_lvbseq = cpu_to_le32(lkb->lkb_lvbseq);
	rl->rl_rqmode = lkb->lkb_rqmode;
	rl->rl_grmode = lkb->lkb_grmode;
	rl->rl_status = lkb->lkb_status;
	rl->rl_wait_type = cpu_to_le16(lkb->lkb_wait_type);

	if (lkb->lkb_bastaddr)
		rl->rl_asts |= AST_BAST;
	if (lkb->lkb_astaddr)
		rl->rl_asts |= AST_COMP;

	rl->rl_namelen = cpu_to_le16(r->res_length);
	memcpy(rl->rl_name, r->res_name, r->res_length);

	/* FIXME: might we have an lvb without DLM_LKF_VALBLK set ?
	   If so, receive_rcom_lock_args() won't take this copy. */

	if (lkb->lkb_lvbptr)
		memcpy(rl->rl_lvb, lkb->lkb_lvbptr, r->res_ls->ls_lvblen);
}

int dlm_send_rcom_lock(struct dlm_rsb *r, struct dlm_lkb *lkb)
{
	struct dlm_ls *ls = r->res_ls;
	struct dlm_rcom *rc;
	struct dlm_mhandle *mh;
	struct rcom_lock *rl;
	int error, len = sizeof(struct rcom_lock);

	if (lkb->lkb_lvbptr)
		len += ls->ls_lvblen;

	error = create_rcom(ls, r->res_nodeid, DLM_RCOM_LOCK, len, &rc, &mh);
	if (error)
		goto out;

	rl = (struct rcom_lock *) rc->rc_buf;
	pack_rcom_lock(r, lkb, rl);
	rc->rc_id = (unsigned long) r;

	send_rcom(ls, mh, rc);
 out:
	return error;
}

static void receive_rcom_lock(struct dlm_ls *ls, struct dlm_rcom *rc_in)
{
	struct dlm_rcom *rc;
	struct dlm_mhandle *mh;
	int error, nodeid = rc_in->rc_header.h_nodeid;

	dlm_recover_master_copy(ls, rc_in);

	error = create_rcom(ls, nodeid, DLM_RCOM_LOCK_REPLY,
			    sizeof(struct rcom_lock), &rc, &mh);
	if (error)
		return;

	/* We send back the same rcom_lock struct we received, but
	   dlm_recover_master_copy() has filled in rl_remid and rl_result */

	memcpy(rc->rc_buf, rc_in->rc_buf, sizeof(struct rcom_lock));
	rc->rc_id = rc_in->rc_id;
	rc->rc_seq_reply = rc_in->rc_seq;

	send_rcom(ls, mh, rc);
}

/* If the lockspace doesn't exist then still send a status message
   back; it's possible that it just doesn't have its global_id yet. */

int dlm_send_ls_not_ready(int nodeid, struct dlm_rcom *rc_in)
{
	struct dlm_rcom *rc;
	struct rcom_config *rf;
	struct dlm_mhandle *mh;
	char *mb;
	int mb_len = sizeof(struct dlm_rcom) + sizeof(struct rcom_config);

	mh = dlm_lowcomms_get_buffer(nodeid, mb_len, GFP_NOFS, &mb);
	if (!mh)
		return -ENOBUFS;
	memset(mb, 0, mb_len);

	rc = (struct dlm_rcom *) mb;

	rc->rc_header.h_version = (DLM_HEADER_MAJOR | DLM_HEADER_MINOR);
	rc->rc_header.h_lockspace = rc_in->rc_header.h_lockspace;
	rc->rc_header.h_nodeid = dlm_our_nodeid();
	rc->rc_header.h_length = mb_len;
	rc->rc_header.h_cmd = DLM_RCOM;

	rc->rc_type = DLM_RCOM_STATUS_REPLY;
	rc->rc_id = rc_in->rc_id;
	rc->rc_seq_reply = rc_in->rc_seq;
	rc->rc_result = -ESRCH;

	rf = (struct rcom_config *) rc->rc_buf;
	rf->rf_lvblen = cpu_to_le32(~0U);

	dlm_rcom_out(rc);
	dlm_lowcomms_commit_buffer(mh);

	return 0;
}

static int is_old_reply(struct dlm_ls *ls, struct dlm_rcom *rc)
{
	uint64_t seq;
	int rv = 0;

	switch (rc->rc_type) {
	case DLM_RCOM_STATUS_REPLY:
	case DLM_RCOM_NAMES_REPLY:
	case DLM_RCOM_LOOKUP_REPLY:
	case DLM_RCOM_LOCK_REPLY:
		spin_lock(&ls->ls_recover_lock);
		seq = ls->ls_recover_seq;
		spin_unlock(&ls->ls_recover_lock);
		if (rc->rc_seq_reply != seq) {
			log_debug(ls, "ignoring old reply %x from %d "
				      "seq_reply %llx expect %llx",
				      rc->rc_type, rc->rc_header.h_nodeid,
				      (unsigned long long)rc->rc_seq_reply,
				      (unsigned long long)seq);
			rv = 1;
		}
	}
	return rv;
}

/* Called by dlm_recv; corresponds to dlm_receive_message() but special
   recovery-only comms are sent through here. */

void dlm_receive_rcom(struct dlm_ls *ls, struct dlm_rcom *rc, int nodeid)
{
	if (dlm_recovery_stopped(ls) && (rc->rc_type != DLM_RCOM_STATUS)) {
		log_debug(ls, "ignoring recovery message %x from %d",
			  rc->rc_type, nodeid);
		goto out;
	}

	if (is_old_reply(ls, rc))
		goto out;

	switch (rc->rc_type) {
	case DLM_RCOM_STATUS:
		receive_rcom_status(ls, rc);
		break;

	case DLM_RCOM_NAMES:
		receive_rcom_names(ls, rc);
		break;

	case DLM_RCOM_LOOKUP:
		receive_rcom_lookup(ls, rc);
		break;

	case DLM_RCOM_LOCK:
		receive_rcom_lock(ls, rc);
		break;

	case DLM_RCOM_STATUS_REPLY:
		receive_sync_reply(ls, rc);
		break;

	case DLM_RCOM_NAMES_REPLY:
		receive_sync_reply(ls, rc);
		break;

	case DLM_RCOM_LOOKUP_REPLY:
		receive_rcom_lookup_reply(ls, rc);
		break;

	case DLM_RCOM_LOCK_REPLY:
		dlm_recover_process_copy(ls, rc);
		break;

	default:
		log_error(ls, "receive_rcom bad type %d", rc->rc_type);
	}
 out:
	return;
}
Commit	Line	Data
e7fd4179 DT	1	/******************************************************************************
	2	*******************************************************************************
	3	**
	4	** Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved.
dbcfc347	5	** Copyright (C) 2005-2008 Red Hat, Inc. All rights reserved.
e7fd4179 DT	6	**
	7	** This copyrighted material is made available to anyone wishing to use,
	8	** modify, copy, or redistribute it subject to the terms and conditions
	9	** of the GNU General Public License v.2.
	10	**
	11	*******************************************************************************
	12	******************************************************************************/
	13
	14	#include "dlm_internal.h"
	15	#include "lockspace.h"
	16	#include "member.h"
	17	#include "lowcomms.h"
	18	#include "midcomms.h"
	19	#include "rcom.h"
	20	#include "recover.h"
	21	#include "dir.h"
	22	#include "config.h"
	23	#include "memory.h"
	24	#include "lock.h"
	25	#include "util.h"
	26
	27
	28	static int rcom_response(struct dlm_ls *ls)
	29	{
	30	return test_bit(LSFL_RCOM_READY, &ls->ls_flags);
	31	}
	32
	33	static int create_rcom(struct dlm_ls *ls, int to_nodeid, int type, int len,
	34	struct dlm_rcom rc_ret, struct dlm_mhandle mh_ret)
	35	{
	36	struct dlm_rcom *rc;
	37	struct dlm_mhandle *mh;
	38	char *mb;
	39	int mb_len = sizeof(struct dlm_rcom) + len;
	40
44f487a5	41	mh = dlm_lowcomms_get_buffer(to_nodeid, mb_len, ls->ls_allocation, &mb);
e7fd4179 DT	42	if (!mh) {
	43	log_print("create_rcom to %d type %d len %d ENOBUFS",
	44	to_nodeid, type, len);
	45	return -ENOBUFS;
	46	}
	47	memset(mb, 0, mb_len);
	48
	49	rc = (struct dlm_rcom *) mb;
	50
	51	rc->rc_header.h_version = (DLM_HEADER_MAJOR \| DLM_HEADER_MINOR);
	52	rc->rc_header.h_lockspace = ls->ls_global_id;
	53	rc->rc_header.h_nodeid = dlm_our_nodeid();
	54	rc->rc_header.h_length = mb_len;
	55	rc->rc_header.h_cmd = DLM_RCOM;
	56
	57	rc->rc_type = type;
	58
38aa8b0c DT	59	spin_lock(&ls->ls_recover_lock);
	60	rc->rc_seq = ls->ls_recover_seq;
	61	spin_unlock(&ls->ls_recover_lock);
	62
e7fd4179 DT	63	*mh_ret = mh;
	64	*rc_ret = rc;
	65	return 0;
	66	}
	67
	68	static void send_rcom(struct dlm_ls ls, struct dlm_mhandle mh,
	69	struct dlm_rcom *rc)
	70	{
	71	dlm_rcom_out(rc);
	72	dlm_lowcomms_commit_buffer(mh);
	73	}
	74
	75	/* When replying to a status request, a node also sends back its
	76	configuration values. The requesting node then checks that the remote
	77	node is configured the same way as itself. */
	78
	79	static void make_config(struct dlm_ls ls, struct rcom_config rf)
	80	{
93ff2971 AV	81	rf->rf_lvblen = cpu_to_le32(ls->ls_lvblen);
93ff2971 AV	82	rf->rf_lsflags = cpu_to_le32(ls->ls_exflags);
e7fd4179 DT	83	}
e7fd4179 DT	84
9e971b71	85	static int check_config(struct dlm_ls ls, struct dlm_rcom rc, int nodeid)
e7fd4179	86	{
9e971b71 DT	87	struct rcom_config rf = (struct rcom_config ) rc->rc_buf;
	88
	89	if ((rc->rc_header.h_version & 0xFFFF0000) != DLM_HEADER_MAJOR) {
	90	log_error(ls, "version mismatch: %x nodeid %d: %x",
	91	DLM_HEADER_MAJOR \| DLM_HEADER_MINOR, nodeid,
	92	rc->rc_header.h_version);
8b0e7b2c	93	return -EPROTO;
9e971b71 DT	94	}
9e971b71 DT	95
93ff2971 AV	96	if (le32_to_cpu(rf->rf_lvblen) != ls->ls_lvblen \|\|
93ff2971 AV	97	le32_to_cpu(rf->rf_lsflags) != ls->ls_exflags) {
e7fd4179	98	log_error(ls, "config mismatch: %d,%x nodeid %d: %d,%x",
93ff2971 AV	99	ls->ls_lvblen, ls->ls_exflags, nodeid,
	100	le32_to_cpu(rf->rf_lvblen),
	101	le32_to_cpu(rf->rf_lsflags));
8b0e7b2c	102	return -EPROTO;
e7fd4179 DT	103	}
	104	return 0;
	105	}
	106
98f176fb DT	107	static void allow_sync_reply(struct dlm_ls ls, uint64_t new_seq)
	108	{
	109	spin_lock(&ls->ls_rcom_spin);
	110	*new_seq = ++ls->ls_rcom_seq;
	111	set_bit(LSFL_RCOM_WAIT, &ls->ls_flags);
	112	spin_unlock(&ls->ls_rcom_spin);
	113	}
	114
	115	static void disallow_sync_reply(struct dlm_ls *ls)
	116	{
	117	spin_lock(&ls->ls_rcom_spin);
	118	clear_bit(LSFL_RCOM_WAIT, &ls->ls_flags);
	119	clear_bit(LSFL_RCOM_READY, &ls->ls_flags);
	120	spin_unlock(&ls->ls_rcom_spin);
	121	}
	122
e7fd4179 DT	123	int dlm_rcom_status(struct dlm_ls *ls, int nodeid)
	124	{
	125	struct dlm_rcom *rc;
	126	struct dlm_mhandle *mh;
	127	int error = 0;
	128
faa0f267	129	ls->ls_recover_nodeid = nodeid;
e7fd4179 DT	130
e7fd4179 DT	131	if (nodeid == dlm_our_nodeid()) {
4007685c	132	rc = ls->ls_recover_buf;
e7fd4179 DT	133	rc->rc_result = dlm_recover_status(ls);
	134	goto out;
	135	}
	136
	137	error = create_rcom(ls, nodeid, DLM_RCOM_STATUS, 0, &rc, &mh);
	138	if (error)
	139	goto out;
98f176fb DT	140
98f176fb DT	141	allow_sync_reply(ls, &rc->rc_id);
68c817a1	142	memset(ls->ls_recover_buf, 0, dlm_config.ci_buffer_size);
e7fd4179 DT	143
	144	send_rcom(ls, mh, rc);
	145
	146	error = dlm_wait_function(ls, &rcom_response);
98f176fb	147	disallow_sync_reply(ls);
e7fd4179 DT	148	if (error)
	149	goto out;
	150
4007685c	151	rc = ls->ls_recover_buf;
e7fd4179 DT	152
	153	if (rc->rc_result == -ESRCH) {
	154	/* we pretend the remote lockspace exists with 0 status */
	155	log_debug(ls, "remote node %d not ready", nodeid);
	156	rc->rc_result = 0;
	157	} else
9e971b71	158	error = check_config(ls, rc, nodeid);
e7fd4179 DT	159	/* the caller looks at rc_result for the remote recovery status */
	160	out:
	161	return error;
	162	}
	163
	164	static void receive_rcom_status(struct dlm_ls ls, struct dlm_rcom rc_in)
	165	{
	166	struct dlm_rcom *rc;
	167	struct dlm_mhandle *mh;
	168	int error, nodeid = rc_in->rc_header.h_nodeid;
	169
	170	error = create_rcom(ls, nodeid, DLM_RCOM_STATUS_REPLY,
	171	sizeof(struct rcom_config), &rc, &mh);
	172	if (error)
	173	return;
4a99c3d9	174	rc->rc_id = rc_in->rc_id;
38aa8b0c	175	rc->rc_seq_reply = rc_in->rc_seq;
e7fd4179 DT	176	rc->rc_result = dlm_recover_status(ls);
	177	make_config(ls, (struct rcom_config *) rc->rc_buf);
	178
	179	send_rcom(ls, mh, rc);
	180	}
	181
4a99c3d9	182	static void receive_sync_reply(struct dlm_ls ls, struct dlm_rcom rc_in)
e7fd4179	183	{
98f176fb DT	184	spin_lock(&ls->ls_rcom_spin);
	185	if (!test_bit(LSFL_RCOM_WAIT, &ls->ls_flags) \|\|
	186	rc_in->rc_id != ls->ls_rcom_seq) {
	187	log_debug(ls, "reject reply %d from %d seq %llx expect %llx",
	188	rc_in->rc_type, rc_in->rc_header.h_nodeid,
57adf7ee RK	189	(unsigned long long)rc_in->rc_id,
57adf7ee RK	190	(unsigned long long)ls->ls_rcom_seq);
98f176fb	191	goto out;
4a99c3d9	192	}
e7fd4179 DT	193	memcpy(ls->ls_recover_buf, rc_in, rc_in->rc_header.h_length);
e7fd4179 DT	194	set_bit(LSFL_RCOM_READY, &ls->ls_flags);
98f176fb	195	clear_bit(LSFL_RCOM_WAIT, &ls->ls_flags);
e7fd4179	196	wake_up(&ls->ls_wait_general);
98f176fb DT	197	out:
98f176fb DT	198	spin_unlock(&ls->ls_rcom_spin);
e7fd4179 DT	199	}
	200
	201	int dlm_rcom_names(struct dlm_ls ls, int nodeid, char last_name, int last_len)
	202	{
	203	struct dlm_rcom *rc;
	204	struct dlm_mhandle *mh;
4007685c AV	205	int error = 0;
4007685c AV	206	int max_size = dlm_config.ci_buffer_size - sizeof(struct dlm_rcom);
e7fd4179	207
faa0f267	208	ls->ls_recover_nodeid = nodeid;
e7fd4179 DT	209
	210	if (nodeid == dlm_our_nodeid()) {
	211	dlm_copy_master_names(ls, last_name, last_len,
4007685c AV	212	ls->ls_recover_buf->rc_buf,
4007685c AV	213	max_size, nodeid);
e7fd4179 DT	214	goto out;
	215	}
	216
	217	error = create_rcom(ls, nodeid, DLM_RCOM_NAMES, last_len, &rc, &mh);
	218	if (error)
	219	goto out;
	220	memcpy(rc->rc_buf, last_name, last_len);
98f176fb DT	221
98f176fb DT	222	allow_sync_reply(ls, &rc->rc_id);
68c817a1	223	memset(ls->ls_recover_buf, 0, dlm_config.ci_buffer_size);
e7fd4179 DT	224
	225	send_rcom(ls, mh, rc);
	226
	227	error = dlm_wait_function(ls, &rcom_response);
98f176fb	228	disallow_sync_reply(ls);
e7fd4179 DT	229	out:
	230	return error;
	231	}
	232
	233	static void receive_rcom_names(struct dlm_ls ls, struct dlm_rcom rc_in)
	234	{
	235	struct dlm_rcom *rc;
	236	struct dlm_mhandle *mh;
38aa8b0c	237	int error, inlen, outlen, nodeid;
e7fd4179 DT	238
	239	nodeid = rc_in->rc_header.h_nodeid;
	240	inlen = rc_in->rc_header.h_length - sizeof(struct dlm_rcom);
68c817a1	241	outlen = dlm_config.ci_buffer_size - sizeof(struct dlm_rcom);
e7fd4179 DT	242
	243	error = create_rcom(ls, nodeid, DLM_RCOM_NAMES_REPLY, outlen, &rc, &mh);
	244	if (error)
	245	return;
4a99c3d9	246	rc->rc_id = rc_in->rc_id;
38aa8b0c	247	rc->rc_seq_reply = rc_in->rc_seq;
e7fd4179 DT	248
	249	dlm_copy_master_names(ls, rc_in->rc_buf, inlen, rc->rc_buf, outlen,
	250	nodeid);
	251	send_rcom(ls, mh, rc);
	252	}
	253
e7fd4179 DT	254	int dlm_send_rcom_lookup(struct dlm_rsb *r, int dir_nodeid)
	255	{
	256	struct dlm_rcom *rc;
	257	struct dlm_mhandle *mh;
	258	struct dlm_ls *ls = r->res_ls;
	259	int error;
	260
	261	error = create_rcom(ls, dir_nodeid, DLM_RCOM_LOOKUP, r->res_length,
	262	&rc, &mh);
	263	if (error)
	264	goto out;
	265	memcpy(rc->rc_buf, r->res_name, r->res_length);
	266	rc->rc_id = (unsigned long) r;
	267
	268	send_rcom(ls, mh, rc);
	269	out:
	270	return error;
	271	}
	272
	273	static void receive_rcom_lookup(struct dlm_ls ls, struct dlm_rcom rc_in)
	274	{
	275	struct dlm_rcom *rc;
	276	struct dlm_mhandle *mh;
	277	int error, ret_nodeid, nodeid = rc_in->rc_header.h_nodeid;
	278	int len = rc_in->rc_header.h_length - sizeof(struct dlm_rcom);
	279
	280	error = create_rcom(ls, nodeid, DLM_RCOM_LOOKUP_REPLY, 0, &rc, &mh);
	281	if (error)
	282	return;
	283
	284	error = dlm_dir_lookup(ls, nodeid, rc_in->rc_buf, len, &ret_nodeid);
	285	if (error)
	286	ret_nodeid = error;
	287	rc->rc_result = ret_nodeid;
	288	rc->rc_id = rc_in->rc_id;
38aa8b0c	289	rc->rc_seq_reply = rc_in->rc_seq;
e7fd4179 DT	290
	291	send_rcom(ls, mh, rc);
	292	}
	293
	294	static void receive_rcom_lookup_reply(struct dlm_ls ls, struct dlm_rcom rc_in)
	295	{
	296	dlm_recover_master_reply(ls, rc_in);
	297	}
	298
	299	static void pack_rcom_lock(struct dlm_rsb r, struct dlm_lkb lkb,
	300	struct rcom_lock *rl)
	301	{
	302	memset(rl, 0, sizeof(*rl));
	303
163a1859 AV	304	rl->rl_ownpid = cpu_to_le32(lkb->lkb_ownpid);
	305	rl->rl_lkid = cpu_to_le32(lkb->lkb_id);
	306	rl->rl_exflags = cpu_to_le32(lkb->lkb_exflags);
	307	rl->rl_flags = cpu_to_le32(lkb->lkb_flags);
	308	rl->rl_lvbseq = cpu_to_le32(lkb->lkb_lvbseq);
e7fd4179 DT	309	rl->rl_rqmode = lkb->lkb_rqmode;
	310	rl->rl_grmode = lkb->lkb_grmode;
	311	rl->rl_status = lkb->lkb_status;
163a1859	312	rl->rl_wait_type = cpu_to_le16(lkb->lkb_wait_type);
e7fd4179 DT	313
	314	if (lkb->lkb_bastaddr)
	315	rl->rl_asts \|= AST_BAST;
	316	if (lkb->lkb_astaddr)
	317	rl->rl_asts \|= AST_COMP;
	318
163a1859	319	rl->rl_namelen = cpu_to_le16(r->res_length);
e7fd4179 DT	320	memcpy(rl->rl_name, r->res_name, r->res_length);
	321
	322	/* FIXME: might we have an lvb without DLM_LKF_VALBLK set ?
	323	If so, receive_rcom_lock_args() won't take this copy. */
	324
	325	if (lkb->lkb_lvbptr)
	326	memcpy(rl->rl_lvb, lkb->lkb_lvbptr, r->res_ls->ls_lvblen);
	327	}
	328
	329	int dlm_send_rcom_lock(struct dlm_rsb r, struct dlm_lkb lkb)
	330	{
	331	struct dlm_ls *ls = r->res_ls;
	332	struct dlm_rcom *rc;
	333	struct dlm_mhandle *mh;
	334	struct rcom_lock *rl;
	335	int error, len = sizeof(struct rcom_lock);
	336
	337	if (lkb->lkb_lvbptr)
	338	len += ls->ls_lvblen;
	339
	340	error = create_rcom(ls, r->res_nodeid, DLM_RCOM_LOCK, len, &rc, &mh);
	341	if (error)
	342	goto out;
	343
	344	rl = (struct rcom_lock *) rc->rc_buf;
	345	pack_rcom_lock(r, lkb, rl);
	346	rc->rc_id = (unsigned long) r;
	347
	348	send_rcom(ls, mh, rc);
	349	out:
	350	return error;
	351	}
	352
	353	static void receive_rcom_lock(struct dlm_ls ls, struct dlm_rcom rc_in)
	354	{
	355	struct dlm_rcom *rc;
	356	struct dlm_mhandle *mh;
	357	int error, nodeid = rc_in->rc_header.h_nodeid;
	358
	359	dlm_recover_master_copy(ls, rc_in);
	360
	361	error = create_rcom(ls, nodeid, DLM_RCOM_LOCK_REPLY,
	362	sizeof(struct rcom_lock), &rc, &mh);
	363	if (error)
	364	return;
	365
	366	/* We send back the same rcom_lock struct we received, but
	367	dlm_recover_master_copy() has filled in rl_remid and rl_result */
	368
	369	memcpy(rc->rc_buf, rc_in->rc_buf, sizeof(struct rcom_lock));
	370	rc->rc_id = rc_in->rc_id;
38aa8b0c	371	rc->rc_seq_reply = rc_in->rc_seq;
e7fd4179 DT	372
	373	send_rcom(ls, mh, rc);
	374	}
	375
c36258b5 DT	376	/* If the lockspace doesn't exist then still send a status message
	377	back; it's possible that it just doesn't have its global_id yet. */
	378
	379	int dlm_send_ls_not_ready(int nodeid, struct dlm_rcom *rc_in)
e7fd4179 DT	380	{
e7fd4179 DT	381	struct dlm_rcom *rc;
1babdb45	382	struct rcom_config *rf;
e7fd4179 DT	383	struct dlm_mhandle *mh;
e7fd4179 DT	384	char *mb;
1babdb45	385	int mb_len = sizeof(struct dlm_rcom) + sizeof(struct rcom_config);
e7fd4179	386
41684f95	387	mh = dlm_lowcomms_get_buffer(nodeid, mb_len, GFP_NOFS, &mb);
e7fd4179 DT	388	if (!mh)
	389	return -ENOBUFS;
	390	memset(mb, 0, mb_len);
	391
	392	rc = (struct dlm_rcom *) mb;
	393
	394	rc->rc_header.h_version = (DLM_HEADER_MAJOR \| DLM_HEADER_MINOR);
	395	rc->rc_header.h_lockspace = rc_in->rc_header.h_lockspace;
	396	rc->rc_header.h_nodeid = dlm_our_nodeid();
	397	rc->rc_header.h_length = mb_len;
	398	rc->rc_header.h_cmd = DLM_RCOM;
	399
	400	rc->rc_type = DLM_RCOM_STATUS_REPLY;
f5888750	401	rc->rc_id = rc_in->rc_id;
38aa8b0c	402	rc->rc_seq_reply = rc_in->rc_seq;
e7fd4179 DT	403	rc->rc_result = -ESRCH;
e7fd4179 DT	404
1babdb45	405	rf = (struct rcom_config *) rc->rc_buf;
93ff2971	406	rf->rf_lvblen = cpu_to_le32(~0U);
1babdb45	407
e7fd4179 DT	408	dlm_rcom_out(rc);
	409	dlm_lowcomms_commit_buffer(mh);
	410
	411	return 0;
	412	}
	413
38aa8b0c DT	414	static int is_old_reply(struct dlm_ls ls, struct dlm_rcom rc)
	415	{
	416	uint64_t seq;
	417	int rv = 0;
	418
	419	switch (rc->rc_type) {
	420	case DLM_RCOM_STATUS_REPLY:
	421	case DLM_RCOM_NAMES_REPLY:
	422	case DLM_RCOM_LOOKUP_REPLY:
	423	case DLM_RCOM_LOCK_REPLY:
	424	spin_lock(&ls->ls_recover_lock);
	425	seq = ls->ls_recover_seq;
	426	spin_unlock(&ls->ls_recover_lock);
	427	if (rc->rc_seq_reply != seq) {
8ec68867	428	log_debug(ls, "ignoring old reply %x from %d "
38aa8b0c DT	429	"seq_reply %llx expect %llx",
	430	rc->rc_type, rc->rc_header.h_nodeid,
	431	(unsigned long long)rc->rc_seq_reply,
	432	(unsigned long long)seq);
	433	rv = 1;
	434	}
	435	}
	436	return rv;
	437	}
	438
c36258b5	439	/* Called by dlm_recv; corresponds to dlm_receive_message() but special
e7fd4179 DT	440	recovery-only comms are sent through here. */
e7fd4179 DT	441
c36258b5	442	void dlm_receive_rcom(struct dlm_ls ls, struct dlm_rcom rc, int nodeid)
e7fd4179	443	{
e7fd4179	444	if (dlm_recovery_stopped(ls) && (rc->rc_type != DLM_RCOM_STATUS)) {
8ec68867	445	log_debug(ls, "ignoring recovery message %x from %d",
e7fd4179 DT	446	rc->rc_type, nodeid);
	447	goto out;
	448	}
	449
38aa8b0c DT	450	if (is_old_reply(ls, rc))
	451	goto out;
	452
e7fd4179 DT	453	switch (rc->rc_type) {
	454	case DLM_RCOM_STATUS:
	455	receive_rcom_status(ls, rc);
	456	break;
	457
	458	case DLM_RCOM_NAMES:
	459	receive_rcom_names(ls, rc);
	460	break;
	461
	462	case DLM_RCOM_LOOKUP:
	463	receive_rcom_lookup(ls, rc);
	464	break;
	465
	466	case DLM_RCOM_LOCK:
	467	receive_rcom_lock(ls, rc);
	468	break;
	469
	470	case DLM_RCOM_STATUS_REPLY:
dbcfc347	471	receive_sync_reply(ls, rc);
e7fd4179 DT	472	break;
	473
	474	case DLM_RCOM_NAMES_REPLY:
dbcfc347	475	receive_sync_reply(ls, rc);
e7fd4179 DT	476	break;
	477
	478	case DLM_RCOM_LOOKUP_REPLY:
	479	receive_rcom_lookup_reply(ls, rc);
	480	break;
	481
	482	case DLM_RCOM_LOCK_REPLY:
dbcfc347	483	dlm_recover_process_copy(ls, rc);
e7fd4179 DT	484	break;
	485
	486	default:
dbcfc347	487	log_error(ls, "receive_rcom bad type %d", rc->rc_type);
e7fd4179 DT	488	}
e7fd4179 DT	489	out:
c36258b5	490	return;
e7fd4179 DT	491	}
e7fd4179 DT	492