linux_dsm_epyc7002/fs/ocfs2/dlm/dlmdebug.c
Sunil Mushran 007dce53a2 ocfs2/dlm: Dump the dlm state in a debugfs file
This patch dumps the dlm state (dlm_ctxt) into a debugfs file.
Useful for debugging.

Signed-off-by: Sunil Mushran <sunil.mushran@oracle.com>
Signed-off-by: Joel Becker <joel.becker@oracle.com>
Signed-off-by: Mark Fasheh <mfasheh@suse.com>
2008-04-18 08:56:08 -07:00

613 lines
17 KiB
C

/* -*- mode: c; c-basic-offset: 8; -*-
* vim: noexpandtab sw=8 ts=8 sts=0:
*
* dlmdebug.c
*
* debug functionality for the dlm
*
* 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/types.h>
#include <linux/slab.h>
#include <linux/highmem.h>
#include <linux/utsname.h>
#include <linux/sysctl.h>
#include <linux/spinlock.h>
#include <linux/debugfs.h>
#include "cluster/heartbeat.h"
#include "cluster/nodemanager.h"
#include "cluster/tcp.h"
#include "dlmapi.h"
#include "dlmcommon.h"
#include "dlmdomain.h"
#include "dlmdebug.h"
#define MLOG_MASK_PREFIX ML_DLM
#include "cluster/masklog.h"
void dlm_print_one_lock_resource(struct dlm_lock_resource *res)
{
mlog(ML_NOTICE, "lockres: %.*s, owner=%u, state=%u\n",
res->lockname.len, res->lockname.name,
res->owner, res->state);
spin_lock(&res->spinlock);
__dlm_print_one_lock_resource(res);
spin_unlock(&res->spinlock);
}
static void dlm_print_lockres_refmap(struct dlm_lock_resource *res)
{
int bit;
assert_spin_locked(&res->spinlock);
mlog(ML_NOTICE, " refmap nodes: [ ");
bit = 0;
while (1) {
bit = find_next_bit(res->refmap, O2NM_MAX_NODES, bit);
if (bit >= O2NM_MAX_NODES)
break;
printk("%u ", bit);
bit++;
}
printk("], inflight=%u\n", res->inflight_locks);
}
void __dlm_print_one_lock_resource(struct dlm_lock_resource *res)
{
struct list_head *iter2;
struct dlm_lock *lock;
assert_spin_locked(&res->spinlock);
mlog(ML_NOTICE, "lockres: %.*s, owner=%u, state=%u\n",
res->lockname.len, res->lockname.name,
res->owner, res->state);
mlog(ML_NOTICE, " last used: %lu, on purge list: %s\n",
res->last_used, list_empty(&res->purge) ? "no" : "yes");
dlm_print_lockres_refmap(res);
mlog(ML_NOTICE, " granted queue: \n");
list_for_each(iter2, &res->granted) {
lock = list_entry(iter2, struct dlm_lock, list);
spin_lock(&lock->spinlock);
mlog(ML_NOTICE, " type=%d, conv=%d, node=%u, "
"cookie=%u:%llu, ast=(empty=%c,pend=%c), bast=(empty=%c,pend=%c)\n",
lock->ml.type, lock->ml.convert_type, lock->ml.node,
dlm_get_lock_cookie_node(be64_to_cpu(lock->ml.cookie)),
dlm_get_lock_cookie_seq(be64_to_cpu(lock->ml.cookie)),
list_empty(&lock->ast_list) ? 'y' : 'n',
lock->ast_pending ? 'y' : 'n',
list_empty(&lock->bast_list) ? 'y' : 'n',
lock->bast_pending ? 'y' : 'n');
spin_unlock(&lock->spinlock);
}
mlog(ML_NOTICE, " converting queue: \n");
list_for_each(iter2, &res->converting) {
lock = list_entry(iter2, struct dlm_lock, list);
spin_lock(&lock->spinlock);
mlog(ML_NOTICE, " type=%d, conv=%d, node=%u, "
"cookie=%u:%llu, ast=(empty=%c,pend=%c), bast=(empty=%c,pend=%c)\n",
lock->ml.type, lock->ml.convert_type, lock->ml.node,
dlm_get_lock_cookie_node(be64_to_cpu(lock->ml.cookie)),
dlm_get_lock_cookie_seq(be64_to_cpu(lock->ml.cookie)),
list_empty(&lock->ast_list) ? 'y' : 'n',
lock->ast_pending ? 'y' : 'n',
list_empty(&lock->bast_list) ? 'y' : 'n',
lock->bast_pending ? 'y' : 'n');
spin_unlock(&lock->spinlock);
}
mlog(ML_NOTICE, " blocked queue: \n");
list_for_each(iter2, &res->blocked) {
lock = list_entry(iter2, struct dlm_lock, list);
spin_lock(&lock->spinlock);
mlog(ML_NOTICE, " type=%d, conv=%d, node=%u, "
"cookie=%u:%llu, ast=(empty=%c,pend=%c), bast=(empty=%c,pend=%c)\n",
lock->ml.type, lock->ml.convert_type, lock->ml.node,
dlm_get_lock_cookie_node(be64_to_cpu(lock->ml.cookie)),
dlm_get_lock_cookie_seq(be64_to_cpu(lock->ml.cookie)),
list_empty(&lock->ast_list) ? 'y' : 'n',
lock->ast_pending ? 'y' : 'n',
list_empty(&lock->bast_list) ? 'y' : 'n',
lock->bast_pending ? 'y' : 'n');
spin_unlock(&lock->spinlock);
}
}
void dlm_print_one_lock(struct dlm_lock *lockid)
{
dlm_print_one_lock_resource(lockid->lockres);
}
EXPORT_SYMBOL_GPL(dlm_print_one_lock);
#if 0
void dlm_dump_lock_resources(struct dlm_ctxt *dlm)
{
struct dlm_lock_resource *res;
struct hlist_node *iter;
struct hlist_head *bucket;
int i;
mlog(ML_NOTICE, "struct dlm_ctxt: %s, node=%u, key=%u\n",
dlm->name, dlm->node_num, dlm->key);
if (!dlm || !dlm->name) {
mlog(ML_ERROR, "dlm=%p\n", dlm);
return;
}
spin_lock(&dlm->spinlock);
for (i=0; i<DLM_HASH_BUCKETS; i++) {
bucket = dlm_lockres_hash(dlm, i);
hlist_for_each_entry(res, iter, bucket, hash_node)
dlm_print_one_lock_resource(res);
}
spin_unlock(&dlm->spinlock);
}
#endif /* 0 */
static const char *dlm_errnames[] = {
[DLM_NORMAL] = "DLM_NORMAL",
[DLM_GRANTED] = "DLM_GRANTED",
[DLM_DENIED] = "DLM_DENIED",
[DLM_DENIED_NOLOCKS] = "DLM_DENIED_NOLOCKS",
[DLM_WORKING] = "DLM_WORKING",
[DLM_BLOCKED] = "DLM_BLOCKED",
[DLM_BLOCKED_ORPHAN] = "DLM_BLOCKED_ORPHAN",
[DLM_DENIED_GRACE_PERIOD] = "DLM_DENIED_GRACE_PERIOD",
[DLM_SYSERR] = "DLM_SYSERR",
[DLM_NOSUPPORT] = "DLM_NOSUPPORT",
[DLM_CANCELGRANT] = "DLM_CANCELGRANT",
[DLM_IVLOCKID] = "DLM_IVLOCKID",
[DLM_SYNC] = "DLM_SYNC",
[DLM_BADTYPE] = "DLM_BADTYPE",
[DLM_BADRESOURCE] = "DLM_BADRESOURCE",
[DLM_MAXHANDLES] = "DLM_MAXHANDLES",
[DLM_NOCLINFO] = "DLM_NOCLINFO",
[DLM_NOLOCKMGR] = "DLM_NOLOCKMGR",
[DLM_NOPURGED] = "DLM_NOPURGED",
[DLM_BADARGS] = "DLM_BADARGS",
[DLM_VOID] = "DLM_VOID",
[DLM_NOTQUEUED] = "DLM_NOTQUEUED",
[DLM_IVBUFLEN] = "DLM_IVBUFLEN",
[DLM_CVTUNGRANT] = "DLM_CVTUNGRANT",
[DLM_BADPARAM] = "DLM_BADPARAM",
[DLM_VALNOTVALID] = "DLM_VALNOTVALID",
[DLM_REJECTED] = "DLM_REJECTED",
[DLM_ABORT] = "DLM_ABORT",
[DLM_CANCEL] = "DLM_CANCEL",
[DLM_IVRESHANDLE] = "DLM_IVRESHANDLE",
[DLM_DEADLOCK] = "DLM_DEADLOCK",
[DLM_DENIED_NOASTS] = "DLM_DENIED_NOASTS",
[DLM_FORWARD] = "DLM_FORWARD",
[DLM_TIMEOUT] = "DLM_TIMEOUT",
[DLM_IVGROUPID] = "DLM_IVGROUPID",
[DLM_VERS_CONFLICT] = "DLM_VERS_CONFLICT",
[DLM_BAD_DEVICE_PATH] = "DLM_BAD_DEVICE_PATH",
[DLM_NO_DEVICE_PERMISSION] = "DLM_NO_DEVICE_PERMISSION",
[DLM_NO_CONTROL_DEVICE ] = "DLM_NO_CONTROL_DEVICE ",
[DLM_RECOVERING] = "DLM_RECOVERING",
[DLM_MIGRATING] = "DLM_MIGRATING",
[DLM_MAXSTATS] = "DLM_MAXSTATS",
};
static const char *dlm_errmsgs[] = {
[DLM_NORMAL] = "request in progress",
[DLM_GRANTED] = "request granted",
[DLM_DENIED] = "request denied",
[DLM_DENIED_NOLOCKS] = "request denied, out of system resources",
[DLM_WORKING] = "async request in progress",
[DLM_BLOCKED] = "lock request blocked",
[DLM_BLOCKED_ORPHAN] = "lock request blocked by a orphan lock",
[DLM_DENIED_GRACE_PERIOD] = "topological change in progress",
[DLM_SYSERR] = "system error",
[DLM_NOSUPPORT] = "unsupported",
[DLM_CANCELGRANT] = "can't cancel convert: already granted",
[DLM_IVLOCKID] = "bad lockid",
[DLM_SYNC] = "synchronous request granted",
[DLM_BADTYPE] = "bad resource type",
[DLM_BADRESOURCE] = "bad resource handle",
[DLM_MAXHANDLES] = "no more resource handles",
[DLM_NOCLINFO] = "can't contact cluster manager",
[DLM_NOLOCKMGR] = "can't contact lock manager",
[DLM_NOPURGED] = "can't contact purge daemon",
[DLM_BADARGS] = "bad api args",
[DLM_VOID] = "no status",
[DLM_NOTQUEUED] = "NOQUEUE was specified and request failed",
[DLM_IVBUFLEN] = "invalid resource name length",
[DLM_CVTUNGRANT] = "attempted to convert ungranted lock",
[DLM_BADPARAM] = "invalid lock mode specified",
[DLM_VALNOTVALID] = "value block has been invalidated",
[DLM_REJECTED] = "request rejected, unrecognized client",
[DLM_ABORT] = "blocked lock request cancelled",
[DLM_CANCEL] = "conversion request cancelled",
[DLM_IVRESHANDLE] = "invalid resource handle",
[DLM_DEADLOCK] = "deadlock recovery refused this request",
[DLM_DENIED_NOASTS] = "failed to allocate AST",
[DLM_FORWARD] = "request must wait for primary's response",
[DLM_TIMEOUT] = "timeout value for lock has expired",
[DLM_IVGROUPID] = "invalid group specification",
[DLM_VERS_CONFLICT] = "version conflicts prevent request handling",
[DLM_BAD_DEVICE_PATH] = "Locks device does not exist or path wrong",
[DLM_NO_DEVICE_PERMISSION] = "Client has insufficient perms for device",
[DLM_NO_CONTROL_DEVICE] = "Cannot set options on opened device ",
[DLM_RECOVERING] = "lock resource being recovered",
[DLM_MIGRATING] = "lock resource being migrated",
[DLM_MAXSTATS] = "invalid error number",
};
const char *dlm_errmsg(enum dlm_status err)
{
if (err >= DLM_MAXSTATS || err < 0)
return dlm_errmsgs[DLM_MAXSTATS];
return dlm_errmsgs[err];
}
EXPORT_SYMBOL_GPL(dlm_errmsg);
const char *dlm_errname(enum dlm_status err)
{
if (err >= DLM_MAXSTATS || err < 0)
return dlm_errnames[DLM_MAXSTATS];
return dlm_errnames[err];
}
EXPORT_SYMBOL_GPL(dlm_errname);
#ifdef CONFIG_DEBUG_FS
static struct dentry *dlm_debugfs_root = NULL;
#define DLM_DEBUGFS_DIR "o2dlm"
#define DLM_DEBUGFS_DLM_STATE "dlm_state"
/* begin - utils funcs */
static void dlm_debug_free(struct kref *kref)
{
struct dlm_debug_ctxt *dc;
dc = container_of(kref, struct dlm_debug_ctxt, debug_refcnt);
kfree(dc);
}
void dlm_debug_put(struct dlm_debug_ctxt *dc)
{
if (dc)
kref_put(&dc->debug_refcnt, dlm_debug_free);
}
static void dlm_debug_get(struct dlm_debug_ctxt *dc)
{
kref_get(&dc->debug_refcnt);
}
static int stringify_nodemap(unsigned long *nodemap, int maxnodes,
char *buf, int len)
{
int out = 0;
int i = -1;
while ((i = find_next_bit(nodemap, maxnodes, i + 1)) < maxnodes)
out += snprintf(buf + out, len - out, "%d ", i);
return out;
}
static struct debug_buffer *debug_buffer_allocate(void)
{
struct debug_buffer *db = NULL;
db = kzalloc(sizeof(struct debug_buffer), GFP_KERNEL);
if (!db)
goto bail;
db->len = PAGE_SIZE;
db->buf = kmalloc(db->len, GFP_KERNEL);
if (!db->buf)
goto bail;
return db;
bail:
kfree(db);
return NULL;
}
static ssize_t debug_buffer_read(struct file *file, char __user *buf,
size_t nbytes, loff_t *ppos)
{
struct debug_buffer *db = file->private_data;
return simple_read_from_buffer(buf, nbytes, ppos, db->buf, db->len);
}
static loff_t debug_buffer_llseek(struct file *file, loff_t off, int whence)
{
struct debug_buffer *db = file->private_data;
loff_t new = -1;
switch (whence) {
case 0:
new = off;
break;
case 1:
new = file->f_pos + off;
break;
}
if (new < 0 || new > db->len)
return -EINVAL;
return (file->f_pos = new);
}
static int debug_buffer_release(struct inode *inode, struct file *file)
{
struct debug_buffer *db = (struct debug_buffer *)file->private_data;
if (db)
kfree(db->buf);
kfree(db);
return 0;
}
/* end - util funcs */
/* begin - debug state funcs */
static int debug_state_print(struct dlm_ctxt *dlm, struct debug_buffer *db)
{
int out = 0;
struct dlm_reco_node_data *node;
char *state;
int lres, rres, ures, tres;
lres = atomic_read(&dlm->local_resources);
rres = atomic_read(&dlm->remote_resources);
ures = atomic_read(&dlm->unknown_resources);
tres = lres + rres + ures;
spin_lock(&dlm->spinlock);
switch (dlm->dlm_state) {
case DLM_CTXT_NEW:
state = "NEW"; break;
case DLM_CTXT_JOINED:
state = "JOINED"; break;
case DLM_CTXT_IN_SHUTDOWN:
state = "SHUTDOWN"; break;
case DLM_CTXT_LEAVING:
state = "LEAVING"; break;
default:
state = "UNKNOWN"; break;
}
/* Domain: xxxxxxxxxx Key: 0xdfbac769 */
out += snprintf(db->buf + out, db->len - out,
"Domain: %s Key: 0x%08x\n", dlm->name, dlm->key);
/* Thread Pid: xxx Node: xxx State: xxxxx */
out += snprintf(db->buf + out, db->len - out,
"Thread Pid: %d Node: %d State: %s\n",
dlm->dlm_thread_task->pid, dlm->node_num, state);
/* Number of Joins: xxx Joining Node: xxx */
out += snprintf(db->buf + out, db->len - out,
"Number of Joins: %d Joining Node: %d\n",
dlm->num_joins, dlm->joining_node);
/* Domain Map: xx xx xx */
out += snprintf(db->buf + out, db->len - out, "Domain Map: ");
out += stringify_nodemap(dlm->domain_map, O2NM_MAX_NODES,
db->buf + out, db->len - out);
out += snprintf(db->buf + out, db->len - out, "\n");
/* Live Map: xx xx xx */
out += snprintf(db->buf + out, db->len - out, "Live Map: ");
out += stringify_nodemap(dlm->live_nodes_map, O2NM_MAX_NODES,
db->buf + out, db->len - out);
out += snprintf(db->buf + out, db->len - out, "\n");
/* Mastered Resources Total: xxx Locally: xxx Remotely: ... */
out += snprintf(db->buf + out, db->len - out,
"Mastered Resources Total: %d Locally: %d "
"Remotely: %d Unknown: %d\n",
tres, lres, rres, ures);
/* Lists: Dirty=Empty Purge=InUse PendingASTs=Empty ... */
out += snprintf(db->buf + out, db->len - out,
"Lists: Dirty=%s Purge=%s PendingASTs=%s "
"PendingBASTs=%s Master=%s\n",
(list_empty(&dlm->dirty_list) ? "Empty" : "InUse"),
(list_empty(&dlm->purge_list) ? "Empty" : "InUse"),
(list_empty(&dlm->pending_asts) ? "Empty" : "InUse"),
(list_empty(&dlm->pending_basts) ? "Empty" : "InUse"),
(list_empty(&dlm->master_list) ? "Empty" : "InUse"));
/* Purge Count: xxx Refs: xxx */
out += snprintf(db->buf + out, db->len - out,
"Purge Count: %d Refs: %d\n", dlm->purge_count,
atomic_read(&dlm->dlm_refs.refcount));
/* Dead Node: xxx */
out += snprintf(db->buf + out, db->len - out,
"Dead Node: %d\n", dlm->reco.dead_node);
/* What about DLM_RECO_STATE_FINALIZE? */
if (dlm->reco.state == DLM_RECO_STATE_ACTIVE)
state = "ACTIVE";
else
state = "INACTIVE";
/* Recovery Pid: xxxx Master: xxx State: xxxx */
out += snprintf(db->buf + out, db->len - out,
"Recovery Pid: %d Master: %d State: %s\n",
dlm->dlm_reco_thread_task->pid,
dlm->reco.new_master, state);
/* Recovery Map: xx xx */
out += snprintf(db->buf + out, db->len - out, "Recovery Map: ");
out += stringify_nodemap(dlm->recovery_map, O2NM_MAX_NODES,
db->buf + out, db->len - out);
out += snprintf(db->buf + out, db->len - out, "\n");
/* Recovery Node State: */
out += snprintf(db->buf + out, db->len - out, "Recovery Node State:\n");
list_for_each_entry(node, &dlm->reco.node_data, list) {
switch (node->state) {
case DLM_RECO_NODE_DATA_INIT:
state = "INIT";
break;
case DLM_RECO_NODE_DATA_REQUESTING:
state = "REQUESTING";
break;
case DLM_RECO_NODE_DATA_DEAD:
state = "DEAD";
break;
case DLM_RECO_NODE_DATA_RECEIVING:
state = "RECEIVING";
break;
case DLM_RECO_NODE_DATA_REQUESTED:
state = "REQUESTED";
break;
case DLM_RECO_NODE_DATA_DONE:
state = "DONE";
break;
case DLM_RECO_NODE_DATA_FINALIZE_SENT:
state = "FINALIZE-SENT";
break;
default:
state = "BAD";
break;
}
out += snprintf(db->buf + out, db->len - out, "\t%u - %s\n",
node->node_num, state);
}
spin_unlock(&dlm->spinlock);
return out;
}
static int debug_state_open(struct inode *inode, struct file *file)
{
struct dlm_ctxt *dlm = inode->i_private;
struct debug_buffer *db = NULL;
db = debug_buffer_allocate();
if (!db)
goto bail;
db->len = debug_state_print(dlm, db);
file->private_data = db;
return 0;
bail:
return -ENOMEM;
}
static struct file_operations debug_state_fops = {
.open = debug_state_open,
.release = debug_buffer_release,
.read = debug_buffer_read,
.llseek = debug_buffer_llseek,
};
/* end - debug state funcs */
/* files in subroot */
int dlm_debug_init(struct dlm_ctxt *dlm)
{
struct dlm_debug_ctxt *dc = dlm->dlm_debug_ctxt;
/* for dumping dlm_ctxt */
dc->debug_state_dentry = debugfs_create_file(DLM_DEBUGFS_DLM_STATE,
S_IFREG|S_IRUSR,
dlm->dlm_debugfs_subroot,
dlm, &debug_state_fops);
if (!dc->debug_state_dentry) {
mlog_errno(-ENOMEM);
goto bail;
}
dlm_debug_get(dc);
return 0;
bail:
dlm_debug_shutdown(dlm);
return -ENOMEM;
}
void dlm_debug_shutdown(struct dlm_ctxt *dlm)
{
struct dlm_debug_ctxt *dc = dlm->dlm_debug_ctxt;
if (dc) {
if (dc->debug_state_dentry)
debugfs_remove(dc->debug_state_dentry);
dlm_debug_put(dc);
}
}
/* subroot - domain dir */
int dlm_create_debugfs_subroot(struct dlm_ctxt *dlm)
{
dlm->dlm_debugfs_subroot = debugfs_create_dir(dlm->name,
dlm_debugfs_root);
if (!dlm->dlm_debugfs_subroot) {
mlog_errno(-ENOMEM);
goto bail;
}
dlm->dlm_debug_ctxt = kzalloc(sizeof(struct dlm_debug_ctxt),
GFP_KERNEL);
if (!dlm->dlm_debug_ctxt) {
mlog_errno(-ENOMEM);
goto bail;
}
kref_init(&dlm->dlm_debug_ctxt->debug_refcnt);
return 0;
bail:
dlm_destroy_debugfs_subroot(dlm);
return -ENOMEM;
}
void dlm_destroy_debugfs_subroot(struct dlm_ctxt *dlm)
{
if (dlm->dlm_debugfs_subroot)
debugfs_remove(dlm->dlm_debugfs_subroot);
}
/* debugfs root */
int dlm_create_debugfs_root(void)
{
dlm_debugfs_root = debugfs_create_dir(DLM_DEBUGFS_DIR, NULL);
if (!dlm_debugfs_root) {
mlog_errno(-ENOMEM);
return -ENOMEM;
}
return 0;
}
void dlm_destroy_debugfs_root(void)
{
if (dlm_debugfs_root)
debugfs_remove(dlm_debugfs_root);
}
#endif /* CONFIG_DEBUG_FS */