Viewing: mgs_nids.c
// SPDX-License-Identifier: GPL-2.0
/*
* Copyright (c) 2007, 2010, Oracle and/or its affiliates. All rights reserved.
* Use is subject to license terms.
*
* Copyright (c) 2011, 2017, Intel Corporation.
*/
/*
* This file is part of Lustre, http://www.lustre.org/
*
* NID table management for lustre.
*
* Author: Jinshan Xiong <jinshan.xiong@whamcloud.com>
*/
#define DEBUG_SUBSYSTEM S_MGS
#define D_MGS D_CONFIG
#include <linux/kthread.h>
#include <linux/pagemap.h>
#include <obd.h>
#include <obd_class.h>
#include <lustre_disk.h>
#include "mgs_internal.h"
static time64_t ir_timeout;
static int nidtbl_is_sane(struct mgs_nidtbl *tbl)
{
struct mgs_nidtbl_target *tgt;
int version = 0;
LASSERT(rwsem_is_locked(&tbl->mn_lock));
list_for_each_entry(tgt, &tbl->mn_targets, mnt_list) {
if (!tgt->mnt_version)
continue;
if (version >= tgt->mnt_version)
return 0;
version = tgt->mnt_version;
}
return 1;
}
static unsigned int mgs_tgt_nid_count(struct mgs_nidtbl_target *tgt,
unsigned long netid)
{
struct tnt_nidlist *tnl;
void *xa_tnl;
unsigned int nids_total = 0;
if (netid == LNET_NET_ANY) {
xa_for_each(&tgt->mnt_xa_nids, netid, xa_tnl) {
tnl = xa_tnl;
nids_total += tnl->tnl_count;
}
} else {
/* only particular net */
tnl = xa_load(&tgt->mnt_xa_nids, netid);
if (tnl)
nids_total += tnl->tnl_count;
}
return nids_total;
}
static int nidtbl_tnl2entry(struct mgs_nidtbl_entry *entry,
struct tnt_nidlist *tnl, unsigned int limit)
{
unsigned int tail = limit - entry->mne_length;
unsigned int count = tail / entry->mne_nid_size;
int i, rc = 0;
if (tnl->tnl_count > count) {
CDEBUG(D_MGS,
"IR: only +%u NIDs (%u total) fits in unit size %u\n",
count, count + entry->mne_nid_count, limit);
rc = -EOVERFLOW;
} else {
count = tnl->tnl_count;
}
for (i = 0; i < count; i++) {
struct lnet_nid nid;
int err;
err = libcfs_strnid(&nid, tnl->tnl_nids[i]);
if (err < 0) {
CDEBUG(D_MGS, "IR: bad NID #%d in nidtbl: %s\n",
i, tnl->tnl_nids[i]);
continue;
}
if (entry->mne_nid_type == 0) {
if (!nid_is_nid4(&nid))
continue;
entry->u.nids[entry->mne_nid_count] =
lnet_nid_to_nid4(&nid);
} else {
/* If the mgs_target_info NIDs are
* struct lnet_nid that have been
* expanded in size we still can
* use the nid if it fits in what
* the client supports.
*/
if (NID_BYTES(&nid) > entry->mne_nid_size)
continue;
entry->u.nidlist[entry->mne_nid_count] = nid;
}
entry->mne_nid_count++;
entry->mne_length += entry->mne_nid_size;
}
return rc;
}
static int nidtbl_fill_entry(struct mgs_nidtbl_target *tgt,
struct mgs_nidtbl_entry *entry,
unsigned long netid, unsigned int limit)
{
struct tnt_nidlist *tnl;
void *xa_tnl;
unsigned long xa_index;
int rc = 0;
/* fill in entry. */
entry->mne_version = tgt->mnt_version;
entry->mne_instance = tgt->mnt_instance;
entry->mne_index = tgt->mnt_stripe_index;
entry->mne_length = sizeof(*entry);
entry->mne_type = tgt->mnt_type;
entry->mne_nid_count = 0;
if (netid == LNET_NET_ANY) {
/* Without restrictions it gets all NIDs across all nets */
xa_for_each(&tgt->mnt_xa_nids, xa_index, xa_tnl) {
tnl = xa_tnl;
CDEBUG(D_INFO, "IR: %u NIDs from NET #%lu\n",
tnl->tnl_count, xa_index);
rc = nidtbl_tnl2entry(entry, tnl, limit);
if (rc)
break;
}
} else {
xa_index = LNET_NETNUM(netid);
/* only particular netid */
tnl = xa_load(&tgt->mnt_xa_nids, xa_index);
if (tnl) {
CDEBUG(D_INFO, "IR: %u NIDs from NET #%lu\n",
tnl->tnl_count, xa_index);
rc = nidtbl_tnl2entry(entry, tnl, limit);
} else {
CDEBUG(D_MGS, "IR: no NIDs for NET #%lu\n", xa_index);
}
}
return rc;
}
/*
* Fetch nidtbl entries whose version are not less than @version
* nidtbl entries will be packed in @pages by @unit_size units - entries
* shouldn't cross unit boundaries.
*/
static int mgs_nidtbl_read(struct obd_export *exp, struct mgs_nidtbl *tbl,
struct mgs_config_res *res, u8 nid_size,
struct folio **folios, int nrpages,
int units_total, int unit_size)
{
struct mgs_nidtbl_target *tgt;
struct mgs_nidtbl_entry *entry;
struct mgs_nidtbl_entry *last_in_unit = NULL;
__u64 version = res->mcr_offset;
unsigned long netid;
unsigned int nid_count;
bool nobuf = false;
void *buf = NULL;
void *kaddr = NULL;
int bytes_in_unit = 0;
int units_in_page = 0;
int index = 0;
int rc = 0;
ENTRY;
/* make sure unit_size is power 2 */
LASSERT((unit_size & (unit_size - 1)) == 0);
LASSERT(nrpages << PAGE_SHIFT >= units_total * unit_size);
down_read(&tbl->mn_lock);
LASSERT(nidtbl_is_sane(tbl));
/* no more entries ? */
if (version > tbl->mn_version) {
version = tbl->mn_version;
goto out;
}
/*
* iterate over all targets to compose IR log entries.
*/
list_for_each_entry(tgt, &tbl->mn_targets, mnt_list) {
int entry_len = sizeof(*entry);
if (tgt->mnt_version < version)
continue;
/* Network filtering. Possibly can come from:
* - any server restriction policy NETs vs clients
* - any client supplied hints about its networks
* - network used by this request
* - etc.
*/
/* no filtering yet */
netid = LNET_NET_ANY;
nid_count = mgs_tgt_nid_count(tgt, netid);
if (!nid_size)
entry_len += nid_count * sizeof(lnet_nid_t);
else
entry_len += nid_count * nid_size;
if (entry_len > unit_size) {
CDEBUG(D_MGS,
"nidtbl: entry has %u NIDs, can't fit in %d\n",
nid_count, unit_size);
/* return as many NIDs as can fit */
}
if (bytes_in_unit < entry_len) {
if (units_total == 0) {
nobuf = true;
break;
}
/* check if we need to consume remaining bytes. */
if (last_in_unit && bytes_in_unit) {
last_in_unit->mne_length += bytes_in_unit;
rc += bytes_in_unit;
buf += bytes_in_unit;
last_in_unit = NULL;
}
LASSERT((rc & (unit_size - 1)) == 0);
if (units_in_page == 0) {
/* destroy previous map */
if (kaddr) {
kunmap_local(kaddr);
kaddr = NULL;
}
/* allocate a new page */
folios[index] = folio_alloc(GFP_KERNEL, 0);
if (IS_ERR_OR_NULL(folios[index])) {
folios[index] = NULL;
rc = -ENOMEM;
break;
}
/* reassign buffer */
buf = kaddr = kmap_local_folio(folios[index],
0);
++index;
units_in_page = PAGE_SIZE / unit_size;
LASSERT(units_in_page > 0);
}
/* allocate an unit */
LASSERT(((long)buf & (unit_size - 1)) == 0);
bytes_in_unit = unit_size;
--units_in_page;
--units_total;
}
/* fill in entry. */
entry = (struct mgs_nidtbl_entry *)buf;
if (nid_size) {
entry->mne_nid_size = nid_size;
entry->mne_nid_type = 1;
} else {
entry->mne_nid_size = sizeof(lnet_nid_t);
entry->mne_nid_type = 0;
}
/* if no NIDs filled then emit error and continue
* with partial nidlist otherwise.
*/
nidtbl_fill_entry(tgt, entry, netid, bytes_in_unit);
if (!entry->mne_nid_count) {
rc = -EOVERFLOW;
break;
}
entry_len = entry->mne_length;
version = tgt->mnt_version;
rc += entry_len;
buf += entry_len;
bytes_in_unit -= entry_len;
last_in_unit = entry;
CDEBUG(D_MGS, "fsname %s, entry size %d, pages %d/%d/%d/%d.\n",
tbl->mn_fsdb->fsdb_name, entry_len,
bytes_in_unit, index, nrpages, units_total);
}
if (kaddr)
kunmap_local(kaddr);
out:
LASSERT(version <= tbl->mn_version);
res->mcr_size = tbl->mn_version;
res->mcr_offset = nobuf ? version : tbl->mn_version;
up_read(&tbl->mn_lock);
CDEBUG(D_MGS, "Read IR logs %s return with %d, version %llu\n",
tbl->mn_fsdb->fsdb_name, rc, version);
LASSERT(ergo(version == 1, rc <= 0)); /* get the log first time */
RETURN(rc);
}
static int nidtbl_update_version(const struct lu_env *env,
struct mgs_device *mgs,
struct mgs_nidtbl *tbl)
{
struct dt_object *fsdb;
struct thandle *th;
u64 version;
struct lu_buf buf = {
.lb_buf = &version,
.lb_len = sizeof(version)
};
loff_t off = 0;
int rc;
ENTRY;
if (mgs->mgs_bottom->dd_rdonly)
RETURN(0);
LASSERT(rwsem_is_locked(&tbl->mn_lock));
fsdb = local_file_find_or_create(env, mgs->mgs_los, mgs->mgs_nidtbl_dir,
tbl->mn_fsdb->fsdb_name,
S_IFREG | 0644);
if (IS_ERR(fsdb))
RETURN(PTR_ERR(fsdb));
th = dt_trans_create(env, mgs->mgs_bottom);
if (IS_ERR(th))
GOTO(out_put, rc = PTR_ERR(th));
th->th_sync = 1; /* update table synchronously */
rc = dt_declare_record_write(env, fsdb, &buf, off, th);
if (rc)
GOTO(out, rc);
rc = dt_trans_start_local(env, mgs->mgs_bottom, th);
if (rc)
GOTO(out, rc);
version = cpu_to_le64(tbl->mn_version);
rc = dt_record_write(env, fsdb, &buf, &off, th);
out:
dt_trans_stop(env, mgs->mgs_bottom, th);
out_put:
dt_object_put(env, fsdb);
RETURN(rc);
}
#define MGS_NIDTBL_VERSION_INIT 2
static int nidtbl_read_version(const struct lu_env *env,
struct mgs_device *mgs, struct mgs_nidtbl *tbl,
u64 *version)
{
struct dt_object *fsdb;
struct lu_fid fid;
u64 tmpver;
struct lu_buf buf = {
.lb_buf = &tmpver,
.lb_len = sizeof(tmpver)
};
loff_t off = 0;
int rc;
ENTRY;
LASSERT(rwsem_is_locked(&tbl->mn_lock));
LASSERT(mgs->mgs_nidtbl_dir);
rc = dt_lookup_dir(env, mgs->mgs_nidtbl_dir, tbl->mn_fsdb->fsdb_name,
&fid);
if (rc == -ENOENT) {
*version = MGS_NIDTBL_VERSION_INIT;
RETURN(0);
} else if (rc < 0) {
RETURN(rc);
}
fsdb = dt_locate_at(env, mgs->mgs_bottom, &fid,
&mgs->mgs_dt_dev.dd_lu_dev, NULL);
if (IS_ERR(fsdb))
RETURN(PTR_ERR(fsdb));
rc = dt_read(env, fsdb, &buf, &off);
if (rc == buf.lb_len) {
*version = le64_to_cpu(tmpver);
rc = 0;
} else if (rc == 0) {
*version = MGS_NIDTBL_VERSION_INIT;
} else {
CERROR("%s: read version file %s error %d\n",
mgs->mgs_obd->obd_name, tbl->mn_fsdb->fsdb_name, rc);
}
dt_object_put(env, fsdb);
RETURN(rc);
}
/* Overwrite or append nidlist with new one */
static int mgs_tnl_update(struct mgs_nidtbl_target *tgt, unsigned long net,
struct mgs_target_info *mti, unsigned int mti_off,
unsigned int count)
{
struct tnt_nidlist *tnl, *oldtnl;
size_t newsize, oldsize = 0;
unsigned int tnl_off = 0;
int i, rc;
if (net == LNET_NETNUM(LNET_NET_ANY))
return 0;
oldtnl = xa_load(&tgt->mnt_xa_nids, net);
if (oldtnl) {
oldsize = oldtnl->tnl_size;
/* if version is the same then append case */
if (oldtnl->tnl_version == tgt->mnt_version)
tnl_off = oldtnl->tnl_count;
}
/* start with 4 slots as minumum */
newsize = TNL_SIZE(max_t(unsigned int, tnl_off + count, 4));
if (newsize > oldsize) {
newsize = size_roundup_power2(newsize);
OBD_ALLOC(tnl, newsize);
if (!tnl) {
rc = -ENOMEM;
CERROR("%s: can't allocate nidlist, rc = %d\n",
mti->mti_svname, rc);
return rc;
}
if (oldtnl && tnl_off) /* append case */
memcpy(tnl, oldtnl, TNL_SIZE(oldtnl->tnl_count));
tnl->tnl_size = newsize;
} else {
tnl = oldtnl;
}
CDEBUG(D_MGS, "nidtbl: %s %u NIDs to NET #%lu\n",
tnl_off ? "append" : "write", count, net);
if (target_supports_large_nid(mti)) {
memcpy(&tnl->tnl_nids[tnl_off], &mti->mti_nidlist[mti_off],
count * LNET_NIDSTR_SIZE);
} else {
for (i = 0; i < count; i++)
libcfs_nid2str_r(mti->mti_nids[mti_off + i],
tnl->tnl_nids[tnl_off + i],
LNET_NIDSTR_SIZE);
}
tnl->tnl_count = tnl_off + count;
tnl->tnl_version = tgt->mnt_version;
if (tnl == oldtnl)
return 0;
oldtnl = xa_store(&tgt->mnt_xa_nids, net, tnl, GFP_KERNEL);
rc = xa_err(oldtnl);
if (rc) {
CDEBUG(D_MGS, "nidtbl: can't store NET #%lu, rc = %d\n",
net, rc);
/* free tnl and keep using oldtnl whatever it is */
OBD_FREE(tnl, tnl->tnl_size);
} else {
OBD_FREE(oldtnl, oldtnl->tnl_size);
}
return rc;
}
static unsigned int mti_nidnet(struct mgs_target_info *mti, int i)
{
struct lnet_nid nid;
int rc;
if (target_supports_large_nid(mti)) {
rc = libcfs_strnid(&nid, mti->mti_nidlist[i]);
if (rc)
return LNET_NETNUM(LNET_NET_ANY);
} else {
lnet_nid4_to_nid(mti->mti_nids[i], &nid);
}
return LNET_NETNUM(LNET_NID_NET(&nid));
}
/**
* mgs_build_nidlists() - parse incoming target info and update NID lists
* @tgt: Target type OST or MDT
* @mti: pointer to mgs_target_info (which has the updated NID information)
*
* Return:
* * %0 of success
* * %negative on failure
*/
static int mgs_build_nidlists(struct mgs_nidtbl_target *tgt,
struct mgs_target_info *mti)
{
unsigned long net;
int i, rc = 0;
/* Delete NET */
if (mti->mti_nidlist[0][0] == NETDEL_TOKEN) {
struct tnt_nidlist *tnl;
net = libcfs_str2net(mti->mti_nidlist[0] + 1);
if (net == LNET_NET_ANY) {
CDEBUG(D_MGS, "nidtbl: can't parse NET %s\n",
mti->mti_nidlist[0] + 1);
return 0;
}
CDEBUG(D_MGS, "nidtbl: remove NET #%lu\n", net);
tnl = xa_erase(&tgt->mnt_xa_nids, LNET_NETNUM(net));
if (tnl)
OBD_FREE(tnl, tnl->tnl_size);
return 0;
}
/* Usually it is built on target on network basis, so assume that
* and search forward to find a sequence of nids at the same net
*/
i = 0;
while (i < mti->mti_nid_count) {
int cnt = 1;
net = mti_nidnet(mti, i);
while (i + cnt < mti->mti_nid_count &&
mti_nidnet(mti, i + cnt) == net)
cnt++;
rc = mgs_tnl_update(tgt, net, mti, i, cnt);
if (rc)
break;
i += cnt;
}
return rc;
}
static int mgs_nidtbl_write(const struct lu_env *env, struct fs_db *fsdb,
struct mgs_target_info *mti)
{
struct mgs_nidtbl *tbl;
struct mgs_nidtbl_target *tgt;
bool found = false;
int type = mti->mti_flags & LDD_F_SV_TYPE_MASK;
int rc = 0;
ENTRY;
type &= ~LDD_F_SV_TYPE_MGS;
LASSERT(type != 0);
tbl = &fsdb->fsdb_nidtbl;
down_write(&tbl->mn_lock);
list_for_each_entry(tgt, &tbl->mn_targets, mnt_list) {
if (type == tgt->mnt_type &&
mti->mti_stripe_index == tgt->mnt_stripe_index) {
found = true;
break;
}
}
if (!found) {
OBD_ALLOC_PTR(tgt);
if (!tgt)
GOTO(out, rc = -ENOMEM);
INIT_LIST_HEAD(&tgt->mnt_list);
tgt->mnt_fs = tbl;
tgt->mnt_version = 0; /* 0 means invalid */
tgt->mnt_type = type;
tgt->mnt_stripe_index = mti->mti_stripe_index;
xa_init(&tgt->mnt_xa_nids);
++tbl->mn_nr_targets;
}
tgt->mnt_instance = mti->mti_instance;
tgt->mnt_version = ++tbl->mn_version;
list_move_tail(&tgt->mnt_list, &tbl->mn_targets);
rc = mgs_build_nidlists(tgt, mti);
if (rc)
GOTO(out, rc);
rc = nidtbl_update_version(env, fsdb->fsdb_mgs, tbl);
EXIT;
out:
up_write(&tbl->mn_lock);
if (rc)
CERROR("Write NID table version for file system %s error %d\n",
fsdb->fsdb_name, rc);
return rc;
}
static void mgs_nidtbl_fini_fs(struct fs_db *fsdb)
{
struct mgs_nidtbl *tbl = &fsdb->fsdb_nidtbl;
LIST_HEAD(head);
down_write(&tbl->mn_lock);
tbl->mn_nr_targets = 0;
list_splice_init(&tbl->mn_targets, &head);
up_write(&tbl->mn_lock);
while (!list_empty(&head)) {
struct mgs_nidtbl_target *tgt;
unsigned long xa_index;
void *xa_nids;
tgt = list_first_entry(&head, struct mgs_nidtbl_target,
mnt_list);
list_del(&tgt->mnt_list);
xa_for_each(&tgt->mnt_xa_nids, xa_index, xa_nids) {
struct tnt_nidlist *tnl = xa_nids;
xa_erase(&tgt->mnt_xa_nids, xa_index);
OBD_FREE(tnl, tnl->tnl_size);
}
xa_destroy(&tgt->mnt_xa_nids);
OBD_FREE_PTR(tgt);
}
}
static int mgs_nidtbl_init_fs(const struct lu_env *env, struct fs_db *fsdb)
{
struct mgs_nidtbl *tbl = &fsdb->fsdb_nidtbl;
int rc;
INIT_LIST_HEAD(&tbl->mn_targets);
init_rwsem(&tbl->mn_lock);
tbl->mn_nr_targets = 0;
tbl->mn_fsdb = fsdb;
down_write(&tbl->mn_lock);
rc = nidtbl_read_version(env, fsdb->fsdb_mgs, tbl, &tbl->mn_version);
up_write(&tbl->mn_lock);
if (rc < 0)
CERROR("%s: IR: failed to read current version, rc = %d\n",
fsdb->fsdb_mgs->mgs_obd->obd_name, rc);
else
CDEBUG(D_MGS, "IR: current version is %llu\n",
tbl->mn_version);
return rc;
}
/* --------- Imperative Recovery relies on nidtbl stuff ------- */
void mgs_ir_notify_complete(struct fs_db *fsdb)
{
struct timespec64 ts;
ktime_t delta;
atomic_set(&fsdb->fsdb_notify_phase, 0);
/* do statistic */
fsdb->fsdb_notify_count++;
delta = ktime_sub(ktime_get(), fsdb->fsdb_notify_start);
fsdb->fsdb_notify_total = ktime_add(fsdb->fsdb_notify_total, delta);
if (ktime_after(delta, fsdb->fsdb_notify_max))
fsdb->fsdb_notify_max = delta;
ts = ktime_to_timespec64(fsdb->fsdb_notify_max);
CDEBUG(D_MGS, "Revoke recover lock of %s completed after %lld.%09lds\n",
fsdb->fsdb_name, (s64)ts.tv_sec, ts.tv_nsec);
}
static int mgs_ir_notify(void *arg)
{
struct fs_db *fsdb = arg;
struct ldlm_res_id resid;
char name[sizeof(fsdb->fsdb_name) + 16];
BUILD_BUG_ON(sizeof(name) >= 40); /* name is too large to be on stack */
snprintf(name, sizeof(name) - 1, "mgs_%s_notify", fsdb->fsdb_name);
complete(&fsdb->fsdb_notify_comp);
set_user_nice(current, -2);
mgc_fsname2resid(fsdb->fsdb_name, &resid, MGS_CFG_T_RECOVER);
while (1) {
wait_event_idle(fsdb->fsdb_notify_waitq,
fsdb->fsdb_notify_stop ||
atomic_read(&fsdb->fsdb_notify_phase));
if (fsdb->fsdb_notify_stop)
break;
CDEBUG(D_MGS, "%s woken up, phase is %d\n",
name, atomic_read(&fsdb->fsdb_notify_phase));
fsdb->fsdb_notify_start = ktime_get();
mgs_revoke_lock(fsdb->fsdb_mgs, fsdb, MGS_CFG_T_RECOVER);
}
complete(&fsdb->fsdb_notify_comp);
return 0;
}
int mgs_ir_init_fs(const struct lu_env *env, struct mgs_device *mgs,
struct fs_db *fsdb)
{
struct task_struct *task;
if (!ir_timeout)
ir_timeout = (time64_t)OBD_IR_MGS_TIMEOUT;
fsdb->fsdb_ir_state = IR_FULL;
if (mgs->mgs_start_time + ir_timeout > ktime_get_real_seconds())
fsdb->fsdb_ir_state = IR_STARTUP;
fsdb->fsdb_nonir_clients = 0;
/* start notify thread */
fsdb->fsdb_mgs = mgs;
task = kthread_run(mgs_ir_notify, fsdb,
"mgs_%s_notify", fsdb->fsdb_name);
if (!IS_ERR(task))
wait_for_completion(&fsdb->fsdb_notify_comp);
else
CERROR("Start notify thread error %ld\n", PTR_ERR(task));
mgs_nidtbl_init_fs(env, fsdb);
return 0;
}
void mgs_ir_fini_fs(struct mgs_device *mgs, struct fs_db *fsdb)
{
if (test_bit(FSDB_MGS_SELF, &fsdb->fsdb_flags))
return;
mgs_fsc_cleanup_by_fsdb(fsdb);
mgs_nidtbl_fini_fs(fsdb);
LASSERT(list_empty(&fsdb->fsdb_clients));
fsdb->fsdb_notify_stop = 1;
wake_up(&fsdb->fsdb_notify_waitq);
wait_for_completion(&fsdb->fsdb_notify_comp);
}
/* caller must have held fsdb_mutex */
static inline void ir_state_graduate(struct fs_db *fsdb)
{
if (fsdb->fsdb_ir_state == IR_STARTUP) {
if (ktime_get_real_seconds() >
fsdb->fsdb_mgs->mgs_start_time + ir_timeout) {
fsdb->fsdb_ir_state = IR_FULL;
if (fsdb->fsdb_nonir_clients)
fsdb->fsdb_ir_state = IR_PARTIAL;
}
}
}
int mgs_ir_update(const struct lu_env *env, struct mgs_device *mgs,
struct mgs_target_info *mti)
{
struct fs_db *fsdb;
bool notify = true;
int rc;
if (mti->mti_instance == 0)
return -EINVAL;
rc = mgs_find_or_make_fsdb(env, mgs, mti->mti_fsname, &fsdb);
if (rc)
return rc;
rc = mgs_nidtbl_write(env, fsdb, mti);
if (rc)
GOTO(out, rc);
/* check ir state */
mutex_lock(&fsdb->fsdb_mutex);
ir_state_graduate(fsdb);
switch (fsdb->fsdb_ir_state) {
case IR_FULL:
mti->mti_flags |= LDD_F_IR_CAPABLE;
break;
case IR_DISABLED:
notify = false;
fallthrough;
case IR_STARTUP:
case IR_PARTIAL:
break;
default:
LBUG();
}
mutex_unlock(&fsdb->fsdb_mutex);
LASSERT(ergo(mti->mti_flags & LDD_F_IR_CAPABLE, notify));
if (notify) {
CDEBUG(D_MGS, "Try to revoke recover lock of %s\n",
fsdb->fsdb_name);
atomic_inc(&fsdb->fsdb_notify_phase);
wake_up(&fsdb->fsdb_notify_waitq);
}
out:
mgs_put_fsdb(mgs, fsdb);
return rc;
}
/* NID table can be cached by two entities: Clients and MDTs */
enum {
IR_CLIENT = 1,
IR_MDT = 2
};
static int delogname(char *logname, char *fsname, int *typ)
{
char *ptr;
int type;
int len;
ptr = strrchr(logname, '-');
if (!ptr)
return -EINVAL;
/*
* decouple file system name. The llog name may be:
* - "prefix-fsname", prefix is "cliir" or "mdtir"
*/
if (strncmp(ptr, "-mdtir", 6) == 0)
type = IR_MDT;
else if (strncmp(ptr, "-cliir", 6) == 0)
type = IR_CLIENT;
else
return -EINVAL;
len = ptr - logname;
if (len == 0)
return -EINVAL;
memcpy(fsname, logname, len);
fsname[len] = 0;
if (typ)
*typ = type;
return 0;
}
/**
* mgs_get_ir_logs() - handle a @req to get part of the NID table
* @req: Pointer to request struct
*
* Return:
* * %0 on success
* * %negative on failure
*/
int mgs_get_ir_logs(struct ptlrpc_request *req)
{
struct mgs_device *mgs = exp2mgs_dev(req->rq_export);
struct fs_db *fsdb = NULL;
struct mgs_config_body *body;
struct mgs_config_res *res;
struct ptlrpc_bulk_desc *desc;
char fsname[16];
long bufsize;
int unit_size;
int type;
int rc = 0;
int i;
int bytes;
int page_count;
int nrpages;
struct folio **folios = NULL;
ENTRY;
body = req_capsule_client_get(&req->rq_pill, &RMF_MGS_CONFIG_BODY);
if (!body)
RETURN(-EINVAL);
if (body->mcb_type != MGS_CFG_T_RECOVER)
RETURN(-EINVAL);
rc = delogname(body->mcb_name, fsname, &type);
if (rc)
RETURN(rc);
bufsize = body->mcb_units << body->mcb_bits;
nrpages = (bufsize + PAGE_SIZE - 1) >> PAGE_SHIFT;
if (nrpages > PTLRPC_MAX_BRW_PAGES)
RETURN(-EINVAL);
mutex_lock(&mgs->mgs_mutex);
fsdb = mgs_find_fsdb(mgs, fsname);
mutex_unlock(&mgs->mgs_mutex);
if (!fsdb)
RETURN(-ENOENT);
CDEBUG(D_MGS, "Reading IR log %s bufsize %ld.\n",
body->mcb_name, bufsize);
OBD_ALLOC_PTR_ARRAY_LARGE(folios, nrpages);
if (!folios)
GOTO(out, rc = -ENOMEM);
res = req_capsule_server_get(&req->rq_pill, &RMF_MGS_CONFIG_RES);
if (!res)
GOTO(out, rc = -EINVAL);
res->mcr_offset = body->mcb_offset;
unit_size = min_t(int, 1 << body->mcb_bits, PAGE_SIZE);
bytes = mgs_nidtbl_read(req->rq_export, &fsdb->fsdb_nidtbl, res,
body->mcb_rec_nid_size, folios, nrpages,
bufsize / unit_size, unit_size);
if (bytes < 0)
GOTO(out, rc = bytes);
/* start bulk transfer */
page_count = (bytes + PAGE_SIZE - 1) >> PAGE_SHIFT;
LASSERT(page_count <= nrpages);
desc = ptlrpc_prep_bulk_exp(req, page_count, 1,
PTLRPC_BULK_PUT_SOURCE,
MGS_BULK_PORTAL,
&ptlrpc_bulk_kiov_pin_ops);
if (!desc)
GOTO(out, rc = -ENOMEM);
for (i = 0; i < page_count && bytes > 0; i++) {
desc->bd_frag_ops->add_kiov_frag(desc, folio_page(folios[i], 0),
0,
min_t(int, bytes, PAGE_SIZE));
bytes -= PAGE_SIZE;
}
rc = target_bulk_io(req->rq_export, desc);
ptlrpc_free_bulk(desc);
GOTO(out, rc);
out:
if (folios) {
for (i = 0; i < nrpages; i++) {
if (!folios[i])
break;
folio_put(folios[i]);
}
OBD_FREE_PTR_ARRAY_LARGE(folios, nrpages);
}
if (fsdb)
mgs_put_fsdb(mgs, fsdb);
return rc;
}
static int lprocfs_ir_set_state(struct fs_db *fsdb, const char *buf)
{
const char *const strings[] = IR_STRINGS;
int state = -1;
int i;
for (i = 0; i < ARRAY_SIZE(strings); i++) {
if (strcmp(strings[i], buf) == 0) {
state = i;
break;
}
}
if (state < 0)
return -EINVAL;
CDEBUG(D_MGS, "change fsr state of %s from %s to %s\n",
fsdb->fsdb_name, strings[fsdb->fsdb_ir_state], strings[state]);
mutex_lock(&fsdb->fsdb_mutex);
if (state == IR_FULL && fsdb->fsdb_nonir_clients)
state = IR_PARTIAL;
fsdb->fsdb_ir_state = state;
mutex_unlock(&fsdb->fsdb_mutex);
return 0;
}
static int lprocfs_ir_set_timeout(struct fs_db *fsdb, const char *buf)
{
return -EINVAL;
}
static int lprocfs_ir_clear_stats(struct fs_db *fsdb, const char *buf)
{
if (*buf)
return -EINVAL;
fsdb->fsdb_notify_total = ktime_set(0, 0);
fsdb->fsdb_notify_max = ktime_set(0, 0);
fsdb->fsdb_notify_count = 0;
return 0;
}
static struct lproc_ir_cmd {
char *name;
int namelen;
int (*handler)(struct fs_db *, const char *);
} ir_cmds[] = {
{ "state=", 6, lprocfs_ir_set_state },
{ "timeout=", 8, lprocfs_ir_set_timeout },
{ "0", 1, lprocfs_ir_clear_stats }
};
int lprocfs_wr_ir_state(struct file *file, const char __user *buffer,
size_t count, void *data)
{
struct fs_db *fsdb = data;
char *kbuf;
char *ptr;
int rc = 0;
if (count == 0 || count >= PAGE_SIZE)
return -EINVAL;
OBD_ALLOC(kbuf, count + 1);
if (!kbuf)
return -ENOMEM;
if (copy_from_user(kbuf, buffer, count)) {
OBD_FREE(kbuf, count + 1);
return -EFAULT;
}
kbuf[count] = 0; /* buffer is supposed to end with 0 */
if (kbuf[count - 1] == '\n')
kbuf[count - 1] = 0;
ptr = kbuf;
/* fsname=<file system name> must be the 1st entry */
while (ptr) {
char *tmpptr;
int i;
tmpptr = strchr(ptr, ';');
if (tmpptr)
*tmpptr++ = 0;
rc = -EINVAL;
for (i = 0; i < ARRAY_SIZE(ir_cmds); i++) {
struct lproc_ir_cmd *cmd;
int cmdlen;
cmd = &ir_cmds[i];
cmdlen = cmd->namelen;
if (strncmp(cmd->name, ptr, cmdlen) == 0) {
ptr += cmdlen;
rc = cmd->handler(fsdb, ptr);
break;
}
}
if (rc)
break;
ptr = tmpptr;
}
if (rc)
CERROR("Unable to process command: %s(%d)\n", ptr, rc);
OBD_FREE(kbuf, count + 1);
return rc ?: count;
}
int lprocfs_rd_ir_state(struct seq_file *seq, void *data)
{
struct fs_db *fsdb = data;
struct mgs_nidtbl *tbl = &fsdb->fsdb_nidtbl;
const char *const ir_strings[] = IR_STRINGS;
struct timespec64 ts_max;
struct timespec64 ts;
/* mgs_live_seq_show() already holds fsdb_mutex. */
ir_state_graduate(fsdb);
seq_puts(seq, "\nimperative_recovery_state:\n");
seq_printf(seq,
" state: %s\n"
" nonir_clients: %d\n"
" nidtbl_version: %lld\n",
ir_strings[fsdb->fsdb_ir_state], fsdb->fsdb_nonir_clients,
tbl->mn_version);
ts = ktime_to_timespec64(fsdb->fsdb_notify_total);
ts_max = ktime_to_timespec64(fsdb->fsdb_notify_max);
seq_printf(seq, " notify_duration_total: %lld.%09ld\n"
" notify_duation_max: %lld.%09ld\n"
" notify_count: %u\n",
(s64)ts.tv_sec, ts.tv_nsec,
(s64)ts_max.tv_sec, ts_max.tv_nsec,
fsdb->fsdb_notify_count);
return 0;
}
ssize_t ir_timeout_show(struct kobject *kobj, struct attribute *attr,
char *buf)
{
return scnprintf(buf, PAGE_SIZE, "%lld\n", ir_timeout);
}
ssize_t ir_timeout_store(struct kobject *kobj, struct attribute *attr,
const char *buffer, size_t count)
{
return kstrtou64(buffer, 10, &ir_timeout);
}
/* --------------- Handle non IR support clients --------------- */
/* attach a lustre file system to an export */
int mgs_fsc_attach(const struct lu_env *env, struct obd_export *exp,
char *fsname)
{
struct mgs_export_data *data = &exp->u.eu_mgs_data;
struct mgs_device *mgs = exp2mgs_dev(exp);
struct fs_db *fsdb = NULL;
struct mgs_fsc *fsc = NULL;
struct mgs_fsc *new_fsc = NULL;
bool found = false;
int rc;
ENTRY;
rc = mgs_find_or_make_fsdb(env, mgs, fsname, &fsdb);
if (rc)
RETURN(rc);
/* allocate a new fsc in case we need it in spinlock. */
OBD_ALLOC_PTR(new_fsc);
if (!new_fsc)
GOTO(out, rc = -ENOMEM);
INIT_LIST_HEAD(&new_fsc->mfc_export_list);
INIT_LIST_HEAD(&new_fsc->mfc_fsdb_list);
new_fsc->mfc_fsdb = fsdb;
new_fsc->mfc_export = class_export_get(exp);
new_fsc->mfc_ir_capable = !!(exp_connect_flags(exp) &
OBD_CONNECT_IMP_RECOV);
rc = -EEXIST;
mutex_lock(&fsdb->fsdb_mutex);
/* tend to find it in export list because this list is shorter. */
spin_lock(&data->med_lock);
list_for_each_entry(fsc, &data->med_clients, mfc_export_list) {
if (strcmp(fsname, fsc->mfc_fsdb->fsdb_name) == 0) {
found = true;
break;
}
}
if (!found) {
fsc = new_fsc;
new_fsc = NULL;
/* add it into export list. */
list_add(&fsc->mfc_export_list, &data->med_clients);
/* add into fsdb list. */
list_add(&fsc->mfc_fsdb_list, &fsdb->fsdb_clients);
if (!fsc->mfc_ir_capable) {
++fsdb->fsdb_nonir_clients;
if (fsdb->fsdb_ir_state == IR_FULL)
fsdb->fsdb_ir_state = IR_PARTIAL;
}
rc = 0;
}
spin_unlock(&data->med_lock);
mutex_unlock(&fsdb->fsdb_mutex);
if (new_fsc) {
class_export_put(new_fsc->mfc_export);
OBD_FREE_PTR(new_fsc);
}
out:
mgs_put_fsdb(mgs, fsdb);
RETURN(rc);
}
void mgs_fsc_cleanup(struct obd_export *exp)
{
struct mgs_export_data *data = &exp->u.eu_mgs_data;
struct mgs_fsc *fsc, *tmp;
LIST_HEAD(head);
spin_lock(&data->med_lock);
list_splice_init(&data->med_clients, &head);
spin_unlock(&data->med_lock);
list_for_each_entry_safe(fsc, tmp, &head, mfc_export_list) {
struct fs_db *fsdb = fsc->mfc_fsdb;
LASSERT(fsc->mfc_export == exp);
mutex_lock(&fsdb->fsdb_mutex);
list_del_init(&fsc->mfc_fsdb_list);
if (fsc->mfc_ir_capable == 0) {
--fsdb->fsdb_nonir_clients;
LASSERT(fsdb->fsdb_ir_state != IR_FULL);
if (fsdb->fsdb_nonir_clients == 0 &&
fsdb->fsdb_ir_state == IR_PARTIAL)
fsdb->fsdb_ir_state = IR_FULL;
}
mutex_unlock(&fsdb->fsdb_mutex);
list_del_init(&fsc->mfc_export_list);
class_export_put(fsc->mfc_export);
OBD_FREE_PTR(fsc);
}
}
/* must be called with fsdb->fsdb_mutex held */
void mgs_fsc_cleanup_by_fsdb(struct fs_db *fsdb)
{
struct mgs_fsc *fsc, *tmp;
list_for_each_entry_safe(fsc, tmp, &fsdb->fsdb_clients,
mfc_fsdb_list) {
struct mgs_export_data *data = &fsc->mfc_export->u.eu_mgs_data;
LASSERT(fsdb == fsc->mfc_fsdb);
list_del_init(&fsc->mfc_fsdb_list);
spin_lock(&data->med_lock);
list_del_init(&fsc->mfc_export_list);
spin_unlock(&data->med_lock);
class_export_put(fsc->mfc_export);
OBD_FREE_PTR(fsc);
}
fsdb->fsdb_nonir_clients = 0;
if (fsdb->fsdb_ir_state == IR_PARTIAL)
fsdb->fsdb_ir_state = IR_FULL;
}
