// SPDX-License-Identifier: GPL-2.0
#include <linux/ceph/ceph_debug.h>

#include <linux/fs.h>
#include <linux/wait.h>
#include <linux/slab.h>
#include <linux/gfp.h>
#include <linux/sched.h>
#include <linux/debugfs.h>
#include <linux/seq_file.h>
#include <linux/ratelimit.h>
#include <linux/bits.h>
#include <linux/ktime.h>

#include "super.h"
#include "mds_client.h"

#include <linux/ceph/ceph_features.h>
#include <linux/ceph/messenger.h>
#include <linux/ceph/decode.h>
#include <linux/ceph/pagelist.h>
#include <linux/ceph/auth.h>
#include <linux/ceph/debugfs.h>

#define RECONNECT_MAX_SIZE (INT_MAX - PAGE_SIZE)

/*
 * A cluster of MDS (metadata server) daemons is responsible for
 * managing the file system namespace (the directory hierarchy and
 * inodes) and for coordinating shared access to storage.  Metadata is
 * partitioning hierarchically across a number of servers, and that
 * partition varies over time as the cluster adjusts the distribution
 * in order to balance load.
 *
 * The MDS client is primarily responsible to managing synchronous
 * metadata requests for operations like open, unlink, and so forth.
 * If there is a MDS failure, we find out about it when we (possibly
 * request and) receive a new MDS map, and can resubmit affected
 * requests.
 *
 * For the most part, though, we take advantage of a lossless
 * communications channel to the MDS, and do not need to worry about
 * timing out or resubmitting requests.
 *
 * We maintain a stateful "session" with each MDS we interact with.
 * Within each session, we sent periodic heartbeat messages to ensure
 * any capabilities or leases we have been issues remain valid.  If
 * the session times out and goes stale, our leases and capabilities
 * are no longer valid.
 */

struct ceph_reconnect_state {
        struct ceph_mds_session *session;
        int nr_caps, nr_realms;
        struct ceph_pagelist *pagelist;
        unsigned msg_version;
        bool allow_multi;
};

static void __wake_requests(struct ceph_mds_client *mdsc,
                            struct list_head *head);
static void ceph_cap_release_work(struct work_struct *work);
static void ceph_cap_reclaim_work(struct work_struct *work);

static const struct ceph_connection_operations mds_con_ops;


/*
 * mds reply parsing
 */

static int parse_reply_info_quota(void **p, void *end,
                                  struct ceph_mds_reply_info_in *info)
{
        u8 struct_v, struct_compat;
        u32 struct_len;

        ceph_decode_8_safe(p, end, struct_v, bad);
        ceph_decode_8_safe(p, end, struct_compat, bad);
        /* struct_v is expected to be >= 1. we only
         * understand encoding with struct_compat == 1. */
        if (!struct_v || struct_compat != 1)
                goto bad;
        ceph_decode_32_safe(p, end, struct_len, bad);
        ceph_decode_need(p, end, struct_len, bad);
        end = *p + struct_len;
        ceph_decode_64_safe(p, end, info->max_bytes, bad);
        ceph_decode_64_safe(p, end, info->max_files, bad);
        *p = end;
        return 0;
bad:
        return -EIO;
}

/*
 * parse individual inode info
 */
static int parse_reply_info_in(void **p, void *end,
                               struct ceph_mds_reply_info_in *info,
                               u64 features)
{
        int err = 0;
        u8 struct_v = 0;

        if (features == (u64)-1) {
                u32 struct_len;
                u8 struct_compat;
                ceph_decode_8_safe(p, end, struct_v, bad);
                ceph_decode_8_safe(p, end, struct_compat, bad);
                /* struct_v is expected to be >= 1. we only understand
                 * encoding with struct_compat == 1. */
                if (!struct_v || struct_compat != 1)
                        goto bad;
                ceph_decode_32_safe(p, end, struct_len, bad);
                ceph_decode_need(p, end, struct_len, bad);
                end = *p + struct_len;
        }

        ceph_decode_need(p, end, sizeof(struct ceph_mds_reply_inode), bad);
        info->in = *p;
        *p += sizeof(struct ceph_mds_reply_inode) +
                sizeof(*info->in->fragtree.splits) *
                le32_to_cpu(info->in->fragtree.nsplits);

        ceph_decode_32_safe(p, end, info->symlink_len, bad);
        ceph_decode_need(p, end, info->symlink_len, bad);
        info->symlink = *p;
        *p += info->symlink_len;

        ceph_decode_copy_safe(p, end, &info->dir_layout,
                              sizeof(info->dir_layout), bad);
        ceph_decode_32_safe(p, end, info->xattr_len, bad);
        ceph_decode_need(p, end, info->xattr_len, bad);
        info->xattr_data = *p;
        *p += info->xattr_len;

        if (features == (u64)-1) {
                /* inline data */
                ceph_decode_64_safe(p, end, info->inline_version, bad);
                ceph_decode_32_safe(p, end, info->inline_len, bad);
                ceph_decode_need(p, end, info->inline_len, bad);
                info->inline_data = *p;
                *p += info->inline_len;
                /* quota */
                err = parse_reply_info_quota(p, end, info);
                if (err < 0)
                        goto out_bad;
                /* pool namespace */
                ceph_decode_32_safe(p, end, info->pool_ns_len, bad);
                if (info->pool_ns_len > 0) {
                        ceph_decode_need(p, end, info->pool_ns_len, bad);
                        info->pool_ns_data = *p;
                        *p += info->pool_ns_len;
                }

                /* btime */
                ceph_decode_need(p, end, sizeof(info->btime), bad);
                ceph_decode_copy(p, &info->btime, sizeof(info->btime));

                /* change attribute */
                ceph_decode_64_safe(p, end, info->change_attr, bad);

                /* dir pin */
                if (struct_v >= 2) {
                        ceph_decode_32_safe(p, end, info->dir_pin, bad);
                } else {
                        info->dir_pin = -ENODATA;
                }

                /* snapshot birth time, remains zero for v<=2 */
                if (struct_v >= 3) {
                        ceph_decode_need(p, end, sizeof(info->snap_btime), bad);
                        ceph_decode_copy(p, &info->snap_btime,
                                         sizeof(info->snap_btime));
                } else {
                        memset(&info->snap_btime, 0, sizeof(info->snap_btime));
                }

                /* snapshot count, remains zero for v<=3 */
                if (struct_v >= 4) {
                        ceph_decode_64_safe(p, end, info->rsnaps, bad);
                } else {
                        info->rsnaps = 0;
                }

                *p = end;
        } else {
                if (features & CEPH_FEATURE_MDS_INLINE_DATA) {
                        ceph_decode_64_safe(p, end, info->inline_version, bad);
                        ceph_decode_32_safe(p, end, info->inline_len, bad);
                        ceph_decode_need(p, end, info->inline_len, bad);
                        info->inline_data = *p;
                        *p += info->inline_len;
                } else
                        info->inline_version = CEPH_INLINE_NONE;

                if (features & CEPH_FEATURE_MDS_QUOTA) {
                        err = parse_reply_info_quota(p, end, info);
                        if (err < 0)
                                goto out_bad;
                } else {
                        info->max_bytes = 0;
                        info->max_files = 0;
                }

                info->pool_ns_len = 0;
                info->pool_ns_data = NULL;
                if (features & CEPH_FEATURE_FS_FILE_LAYOUT_V2) {
                        ceph_decode_32_safe(p, end, info->pool_ns_len, bad);
                        if (info->pool_ns_len > 0) {
                                ceph_decode_need(p, end, info->pool_ns_len, bad);
                                info->pool_ns_data = *p;
                                *p += info->pool_ns_len;
                        }
                }

                if (features & CEPH_FEATURE_FS_BTIME) {
                        ceph_decode_need(p, end, sizeof(info->btime), bad);
                        ceph_decode_copy(p, &info->btime, sizeof(info->btime));
                        ceph_decode_64_safe(p, end, info->change_attr, bad);
                }

                info->dir_pin = -ENODATA;
                /* info->snap_btime and info->rsnaps remain zero */
        }
        return 0;
bad:
        err = -EIO;
out_bad:
        return err;
}

static int parse_reply_info_dir(void **p, void *end,
                                struct ceph_mds_reply_dirfrag **dirfrag,
                                u64 features)
{
        if (features == (u64)-1) {
                u8 struct_v, struct_compat;
                u32 struct_len;
                ceph_decode_8_safe(p, end, struct_v, bad);
                ceph_decode_8_safe(p, end, struct_compat, bad);
                /* struct_v is expected to be >= 1. we only understand
                 * encoding whose struct_compat == 1. */
                if (!struct_v || struct_compat != 1)
                        goto bad;
                ceph_decode_32_safe(p, end, struct_len, bad);
                ceph_decode_need(p, end, struct_len, bad);
                end = *p + struct_len;
        }

        ceph_decode_need(p, end, sizeof(**dirfrag), bad);
        *dirfrag = *p;
        *p += sizeof(**dirfrag) + sizeof(u32) * le32_to_cpu((*dirfrag)->ndist);
        if (unlikely(*p > end))
                goto bad;
        if (features == (u64)-1)
                *p = end;
        return 0;
bad:
        return -EIO;
}

static int parse_reply_info_lease(void **p, void *end,
                                  struct ceph_mds_reply_lease **lease,
                                  u64 features)
{
        if (features == (u64)-1) {
                u8 struct_v, struct_compat;
                u32 struct_len;
                ceph_decode_8_safe(p, end, struct_v, bad);
                ceph_decode_8_safe(p, end, struct_compat, bad);
                /* struct_v is expected to be >= 1. we only understand
                 * encoding whose struct_compat == 1. */
                if (!struct_v || struct_compat != 1)
                        goto bad;
                ceph_decode_32_safe(p, end, struct_len, bad);
                ceph_decode_need(p, end, struct_len, bad);
                end = *p + struct_len;
        }

        ceph_decode_need(p, end, sizeof(**lease), bad);
        *lease = *p;
        *p += sizeof(**lease);
        if (features == (u64)-1)
                *p = end;
        return 0;
bad:
        return -EIO;
}

/*
 * parse a normal reply, which may contain a (dir+)dentry and/or a
 * target inode.
 */
static int parse_reply_info_trace(void **p, void *end,
                                  struct ceph_mds_reply_info_parsed *info,
                                  u64 features)
{
        int err;

        if (info->head->is_dentry) {
                err = parse_reply_info_in(p, end, &info->diri, features);
                if (err < 0)
                        goto out_bad;

                err = parse_reply_info_dir(p, end, &info->dirfrag, features);
                if (err < 0)
                        goto out_bad;

                ceph_decode_32_safe(p, end, info->dname_len, bad);
                ceph_decode_need(p, end, info->dname_len, bad);
                info->dname = *p;
                *p += info->dname_len;

                err = parse_reply_info_lease(p, end, &info->dlease, features);
                if (err < 0)
                        goto out_bad;
        }

        if (info->head->is_target) {
                err = parse_reply_info_in(p, end, &info->targeti, features);
                if (err < 0)
                        goto out_bad;
        }

        if (unlikely(*p != end))
                goto bad;
        return 0;

bad:
        err = -EIO;
out_bad:
        pr_err("problem parsing mds trace %d\n", err);
        return err;
}

/*
 * parse readdir results
 */
static int parse_reply_info_readdir(void **p, void *end,
                                struct ceph_mds_reply_info_parsed *info,
                                u64 features)
{
        u32 num, i = 0;
        int err;

        err = parse_reply_info_dir(p, end, &info->dir_dir, features);
        if (err < 0)
                goto out_bad;

        ceph_decode_need(p, end, sizeof(num) + 2, bad);
        num = ceph_decode_32(p);
        {
                u16 flags = ceph_decode_16(p);
                info->dir_end = !!(flags & CEPH_READDIR_FRAG_END);
                info->dir_complete = !!(flags & CEPH_READDIR_FRAG_COMPLETE);
                info->hash_order = !!(flags & CEPH_READDIR_HASH_ORDER);
                info->offset_hash = !!(flags & CEPH_READDIR_OFFSET_HASH);
        }
        if (num == 0)
                goto done;

        BUG_ON(!info->dir_entries);
        if ((unsigned long)(info->dir_entries + num) >
            (unsigned long)info->dir_entries + info->dir_buf_size) {
                pr_err("dir contents are larger than expected\n");
                WARN_ON(1);
                goto bad;
        }

        info->dir_nr = num;
        while (num) {
                struct ceph_mds_reply_dir_entry *rde = info->dir_entries + i;
                /* dentry */
                ceph_decode_32_safe(p, end, rde->name_len, bad);
                ceph_decode_need(p, end, rde->name_len, bad);
                rde->name = *p;
                *p += rde->name_len;
                dout("parsed dir dname '%.*s'\n", rde->name_len, rde->name);

                /* dentry lease */
                err = parse_reply_info_lease(p, end, &rde->lease, features);
                if (err)
                        goto out_bad;
                /* inode */
                err = parse_reply_info_in(p, end, &rde->inode, features);
                if (err < 0)
                        goto out_bad;
                /* ceph_readdir_prepopulate() will update it */
                rde->offset = 0;
                i++;
                num--;
        }

done:
        /* Skip over any unrecognized fields */
        *p = end;
        return 0;

bad:
        err = -EIO;
out_bad:
        pr_err("problem parsing dir contents %d\n", err);
        return err;
}

/*
 * parse fcntl F_GETLK results
 */
static int parse_reply_info_filelock(void **p, void *end,
                                     struct ceph_mds_reply_info_parsed *info,
                                     u64 features)
{
        if (*p + sizeof(*info->filelock_reply) > end)
                goto bad;

        info->filelock_reply = *p;

        /* Skip over any unrecognized fields */
        *p = end;
        return 0;
bad:
        return -EIO;
}


#if BITS_PER_LONG == 64

#define DELEGATED_INO_AVAILABLE         xa_mk_value(1)

static int ceph_parse_deleg_inos(void **p, void *end,
                                 struct ceph_mds_session *s)
{
        u32 sets;

        ceph_decode_32_safe(p, end, sets, bad);
        dout("got %u sets of delegated inodes\n", sets);
        while (sets--) {
                u64 start, len, ino;

                ceph_decode_64_safe(p, end, start, bad);
                ceph_decode_64_safe(p, end, len, bad);

                /* Don't accept a delegation of system inodes */
                if (start < CEPH_INO_SYSTEM_BASE) {
                        pr_warn_ratelimited("ceph: ignoring reserved inode range delegation (start=0x%llx len=0x%llx)\n",
                                        start, len);
                        continue;
                }
                while (len--) {
                        int err = xa_insert(&s->s_delegated_inos, ino = start++,
                                            DELEGATED_INO_AVAILABLE,
                                            GFP_KERNEL);
                        if (!err) {
                                dout("added delegated inode 0x%llx\n",
                                     start - 1);
                        } else if (err == -EBUSY) {
                                pr_warn("ceph: MDS delegated inode 0x%llx more than once.\n",
                                        start - 1);
                        } else {
                                return err;
                        }
                }
        }
        return 0;
bad:
        return -EIO;
}

u64 ceph_get_deleg_ino(struct ceph_mds_session *s)
{
        unsigned long ino;
        void *val;

        xa_for_each(&s->s_delegated_inos, ino, val) {
                val = xa_erase(&s->s_delegated_inos, ino);
                if (val == DELEGATED_INO_AVAILABLE)
                        return ino;
        }
        return 0;
}

int ceph_restore_deleg_ino(struct ceph_mds_session *s, u64 ino)
{
        return xa_insert(&s->s_delegated_inos, ino, DELEGATED_INO_AVAILABLE,
                         GFP_KERNEL);
}
#else /* BITS_PER_LONG == 64 */
/*
 * FIXME: xarrays can't handle 64-bit indexes on a 32-bit arch. For now, just
 * ignore delegated_inos on 32 bit arch. Maybe eventually add xarrays for top
 * and bottom words?
 */
static int ceph_parse_deleg_inos(void **p, void *end,
                                 struct ceph_mds_session *s)
{
        u32 sets;

        ceph_decode_32_safe(p, end, sets, bad);
        if (sets)
                ceph_decode_skip_n(p, end, sets * 2 * sizeof(__le64), bad);
        return 0;
bad:
        return -EIO;
}

u64 ceph_get_deleg_ino(struct ceph_mds_session *s)
{
        return 0;
}

int ceph_restore_deleg_ino(struct ceph_mds_session *s, u64 ino)
{
        return 0;
}
#endif /* BITS_PER_LONG == 64 */

/*
 * parse create results
 */
static int parse_reply_info_create(void **p, void *end,
                                  struct ceph_mds_reply_info_parsed *info,
                                  u64 features, struct ceph_mds_session *s)
{
        int ret;

        if (features == (u64)-1 ||
            (features & CEPH_FEATURE_REPLY_CREATE_INODE)) {
                if (*p == end) {
                        /* Malformed reply? */
                        info->has_create_ino = false;
                } else if (test_bit(CEPHFS_FEATURE_DELEG_INO, &s->s_features)) {
                        info->has_create_ino = true;
                        /* struct_v, struct_compat, and len */
                        ceph_decode_skip_n(p, end, 2 + sizeof(u32), bad);
                        ceph_decode_64_safe(p, end, info->ino, bad);
                        ret = ceph_parse_deleg_inos(p, end, s);
                        if (ret)
                                return ret;
                } else {
                        /* legacy */
                        ceph_decode_64_safe(p, end, info->ino, bad);
                        info->has_create_ino = true;
                }
        } else {
                if (*p != end)
                        goto bad;
        }

        /* Skip over any unrecognized fields */
        *p = end;
        return 0;
bad:
        return -EIO;
}

/*
 * parse extra results
 */
static int parse_reply_info_extra(void **p, void *end,
                                  struct ceph_mds_reply_info_parsed *info,
                                  u64 features, struct ceph_mds_session *s)
{
        u32 op = le32_to_cpu(info->head->op);

        if (op == CEPH_MDS_OP_GETFILELOCK)
                return parse_reply_info_filelock(p, end, info, features);
        else if (op == CEPH_MDS_OP_READDIR || op == CEPH_MDS_OP_LSSNAP)
                return parse_reply_info_readdir(p, end, info, features);
        else if (op == CEPH_MDS_OP_CREATE)
                return parse_reply_info_create(p, end, info, features, s);
        else
                return -EIO;
}

/*
 * parse entire mds reply
 */
static int parse_reply_info(struct ceph_mds_session *s, struct ceph_msg *msg,
                            struct ceph_mds_reply_info_parsed *info,
                            u64 features)
{
        void *p, *end;
        u32 len;
        int err;

        info->head = msg->front.iov_base;
        p = msg->front.iov_base + sizeof(struct ceph_mds_reply_head);
        end = p + msg->front.iov_len - sizeof(struct ceph_mds_reply_head);

        /* trace */
        ceph_decode_32_safe(&p, end, len, bad);
        if (len > 0) {
                ceph_decode_need(&p, end, len, bad);
                err = parse_reply_info_trace(&p, p+len, info, features);
                if (err < 0)
                        goto out_bad;
        }

        /* extra */
        ceph_decode_32_safe(&p, end, len, bad);
        if (len > 0) {
                ceph_decode_need(&p, end, len, bad);
                err = parse_reply_info_extra(&p, p+len, info, features, s);
                if (err < 0)
                        goto out_bad;
        }

        /* snap blob */
        ceph_decode_32_safe(&p, end, len, bad);
        info->snapblob_len = len;
        info->snapblob = p;
        p += len;

        if (p != end)
                goto bad;
        return 0;

bad:
        err = -EIO;
out_bad:
        pr_err("mds parse_reply err %d\n", err);
        return err;
}

static void destroy_reply_info(struct ceph_mds_reply_info_parsed *info)
{
        if (!info->dir_entries)
                return;
        free_pages((unsigned long)info->dir_entries, get_order(info->dir_buf_size));
}


/*
 * sessions
 */
const char *ceph_session_state_name(int s)
{
        switch (s) {
        case CEPH_MDS_SESSION_NEW: return "new";
        case CEPH_MDS_SESSION_OPENING: return "opening";
        case CEPH_MDS_SESSION_OPEN: return "open";
        case CEPH_MDS_SESSION_HUNG: return "hung";
        case CEPH_MDS_SESSION_CLOSING: return "closing";
        case CEPH_MDS_SESSION_CLOSED: return "closed";
        case CEPH_MDS_SESSION_RESTARTING: return "restarting";
        case CEPH_MDS_SESSION_RECONNECTING: return "reconnecting";
        case CEPH_MDS_SESSION_REJECTED: return "rejected";
        default: return "???";
        }
}

struct ceph_mds_session *ceph_get_mds_session(struct ceph_mds_session *s)
{
        if (refcount_inc_not_zero(&s->s_ref)) {
                dout("mdsc get_session %p %d -> %d\n", s,
                     refcount_read(&s->s_ref)-1, refcount_read(&s->s_ref));
                return s;
        } else {
                dout("mdsc get_session %p 0 -- FAIL\n", s);
                return NULL;
        }
}

void ceph_put_mds_session(struct ceph_mds_session *s)
{
        if (IS_ERR_OR_NULL(s))
                return;

        dout("mdsc put_session %p %d -> %d\n", s,
             refcount_read(&s->s_ref), refcount_read(&s->s_ref)-1);
        if (refcount_dec_and_test(&s->s_ref)) {
                if (s->s_auth.authorizer)
                        ceph_auth_destroy_authorizer(s->s_auth.authorizer);
                WARN_ON(mutex_is_locked(&s->s_mutex));
                xa_destroy(&s->s_delegated_inos);
                kfree(s);
        }
}

/*
 * called under mdsc->mutex
 */
struct ceph_mds_session *__ceph_lookup_mds_session(struct ceph_mds_client *mdsc,
                                                   int mds)
{
        if (mds >= mdsc->max_sessions || !mdsc->sessions[mds])
                return NULL;
        return ceph_get_mds_session(mdsc->sessions[mds]);
}

static bool __have_session(struct ceph_mds_client *mdsc, int mds)
{
        if (mds >= mdsc->max_sessions || !mdsc->sessions[mds])
                return false;
        else
                return true;
}

static int __verify_registered_session(struct ceph_mds_client *mdsc,
                                       struct ceph_mds_session *s)
{
        if (s->s_mds >= mdsc->max_sessions ||
            mdsc->sessions[s->s_mds] != s)
                return -ENOENT;
        return 0;
}

/*
 * create+register a new session for given mds.
 * called under mdsc->mutex.
 */
static struct ceph_mds_session *register_session(struct ceph_mds_client *mdsc,
                                                 int mds)
{
        struct ceph_mds_session *s;

        if (mds >= mdsc->mdsmap->possible_max_rank)
                return ERR_PTR(-EINVAL);

        s = kzalloc(sizeof(*s), GFP_NOFS);
        if (!s)
                return ERR_PTR(-ENOMEM);

        if (mds >= mdsc->max_sessions) {
                int newmax = 1 << get_count_order(mds + 1);
                struct ceph_mds_session **sa;

                dout("%s: realloc to %d\n", __func__, newmax);
                sa = kcalloc(newmax, sizeof(void *), GFP_NOFS);
                if (!sa)
                        goto fail_realloc;
                if (mdsc->sessions) {
                        memcpy(sa, mdsc->sessions,
                               mdsc->max_sessions * sizeof(void *));
                        kfree(mdsc->sessions);
                }
                mdsc->sessions = sa;
                mdsc->max_sessions = newmax;
        }

        dout("%s: mds%d\n", __func__, mds);
        s->s_mdsc = mdsc;
        s->s_mds = mds;
        s->s_state = CEPH_MDS_SESSION_NEW;
        s->s_ttl = 0;
        s->s_seq = 0;
        mutex_init(&s->s_mutex);

        ceph_con_init(&s->s_con, s, &mds_con_ops, &mdsc->fsc->client->msgr);

        atomic_set(&s->s_cap_gen, 1);
        s->s_cap_ttl = jiffies - 1;

        spin_lock_init(&s->s_cap_lock);
        s->s_renew_requested = 0;
        s->s_renew_seq = 0;
        INIT_LIST_HEAD(&s->s_caps);
        s->s_nr_caps = 0;
        refcount_set(&s->s_ref, 1);
        INIT_LIST_HEAD(&s->s_waiting);
        INIT_LIST_HEAD(&s->s_unsafe);
        xa_init(&s->s_delegated_inos);
        s->s_num_cap_releases = 0;
        s->s_cap_reconnect = 0;
        s->s_cap_iterator = NULL;
        INIT_LIST_HEAD(&s->s_cap_releases);
        INIT_WORK(&s->s_cap_release_work, ceph_cap_release_work);

        INIT_LIST_HEAD(&s->s_cap_dirty);
        INIT_LIST_HEAD(&s->s_cap_flushing);

        mdsc->sessions[mds] = s;
        atomic_inc(&mdsc->num_sessions);
        refcount_inc(&s->s_ref);  /* one ref to sessions[], one to caller */

        ceph_con_open(&s->s_con, CEPH_ENTITY_TYPE_MDS, mds,
                      ceph_mdsmap_get_addr(mdsc->mdsmap, mds));

        return s;

fail_realloc:
        kfree(s);
        return ERR_PTR(-ENOMEM);
}

/*
 * called under mdsc->mutex
 */
static void __unregister_session(struct ceph_mds_client *mdsc,
                               struct ceph_mds_session *s)
{
        dout("__unregister_session mds%d %p\n", s->s_mds, s);
        BUG_ON(mdsc->sessions[s->s_mds] != s);
        mdsc->sessions[s->s_mds] = NULL;
        ceph_con_close(&s->s_con);
        ceph_put_mds_session(s);
        atomic_dec(&mdsc->num_sessions);
}

/*
 * drop session refs in request.
 *
 * should be last request ref, or hold mdsc->mutex
 */
static void put_request_session(struct ceph_mds_request *req)
{
        if (req->r_session) {
                ceph_put_mds_session(req->r_session);
                req->r_session = NULL;
        }
}

void ceph_mdsc_release_request(struct kref *kref)
{
        struct ceph_mds_request *req = container_of(kref,
                                                    struct ceph_mds_request,
                                                    r_kref);
        ceph_mdsc_release_dir_caps_no_check(req);
        destroy_reply_info(&req->r_reply_info);
        if (req->r_request)
                ceph_msg_put(req->r_request);
        if (req->r_reply)
                ceph_msg_put(req->r_reply);
        if (req->r_inode) {
                ceph_put_cap_refs(ceph_inode(req->r_inode), CEPH_CAP_PIN);
                iput(req->r_inode);
        }
        if (req->r_parent) {
                ceph_put_cap_refs(ceph_inode(req->r_parent), CEPH_CAP_PIN);
                iput(req->r_parent);
        }
        iput(req->r_target_inode);
        if (req->r_dentry)
                dput(req->r_dentry);
        if (req->r_old_dentry)
                dput(req->r_old_dentry);
        if (req->r_old_dentry_dir) {
                /*
                 * track (and drop pins for) r_old_dentry_dir
                 * separately, since r_old_dentry's d_parent may have
                 * changed between the dir mutex being dropped and
                 * this request being freed.
                 */
                ceph_put_cap_refs(ceph_inode(req->r_old_dentry_dir),
                                  CEPH_CAP_PIN);
                iput(req->r_old_dentry_dir);
        }
        kfree(req->r_path1);
        kfree(req->r_path2);
        put_cred(req->r_cred);
        if (req->r_pagelist)
                ceph_pagelist_release(req->r_pagelist);
        put_request_session(req);
        ceph_unreserve_caps(req->r_mdsc, &req->r_caps_reservation);
        WARN_ON_ONCE(!list_empty(&req->r_wait));
        kmem_cache_free(ceph_mds_request_cachep, req);
}

DEFINE_RB_FUNCS(request, struct ceph_mds_request, r_tid, r_node)

/*
 * lookup session, bump ref if found.
 *
 * called under mdsc->mutex.
 */
static struct ceph_mds_request *
lookup_get_request(struct ceph_mds_client *mdsc, u64 tid)
{
        struct ceph_mds_request *req;

        req = lookup_request(&mdsc->request_tree, tid);
        if (req)
                ceph_mdsc_get_request(req);

        return req;
}

/*
 * Register an in-flight request, and assign a tid.  Link to directory
 * are modifying (if any).
 *
 * Called under mdsc->mutex.
 */
static void __register_request(struct ceph_mds_client *mdsc,
                               struct ceph_mds_request *req,
                               struct inode *dir)
{
        int ret = 0;

        req->r_tid = ++mdsc->last_tid;
        if (req->r_num_caps) {
                ret = ceph_reserve_caps(mdsc, &req->r_caps_reservation,
                                        req->r_num_caps);
                if (ret < 0) {
                        pr_err("__register_request %p "
                               "failed to reserve caps: %d\n", req, ret);
                        /* set req->r_err to fail early from __do_request */
                        req->r_err = ret;
                        return;
                }
        }
        dout("__register_request %p tid %lld\n", req, req->r_tid);
        ceph_mdsc_get_request(req);
        insert_request(&mdsc->request_tree, req);

        req->r_cred = get_current_cred();

        if (mdsc->oldest_tid == 0 && req->r_op != CEPH_MDS_OP_SETFILELOCK)
                mdsc->oldest_tid = req->r_tid;

        if (dir) {
                struct ceph_inode_info *ci = ceph_inode(dir);

                ihold(dir);
                req->r_unsafe_dir = dir;
                spin_lock(&ci->i_unsafe_lock);
                list_add_tail(&req->r_unsafe_dir_item, &ci->i_unsafe_dirops);
                spin_unlock(&ci->i_unsafe_lock);
        }
}

static void __unregister_request(struct ceph_mds_client *mdsc,
                                 struct ceph_mds_request *req)
{
        dout("__unregister_request %p tid %lld\n", req, req->r_tid);

        /* Never leave an unregistered request on an unsafe list! */
        list_del_init(&req->r_unsafe_item);

        if (req->r_tid == mdsc->oldest_tid) {
                struct rb_node *p = rb_next(&req->r_node);
                mdsc->oldest_tid = 0;
                while (p) {
                        struct ceph_mds_request *next_req =
                                rb_entry(p, struct ceph_mds_request, r_node);
                        if (next_req->r_op != CEPH_MDS_OP_SETFILELOCK) {
                                mdsc->oldest_tid = next_req->r_tid;
                                break;
                        }
                        p = rb_next(p);
                }
        }

        erase_request(&mdsc->request_tree, req);

        if (req->r_unsafe_dir) {
                struct ceph_inode_info *ci = ceph_inode(req->r_unsafe_dir);
                spin_lock(&ci->i_unsafe_lock);
                list_del_init(&req->r_unsafe_dir_item);
                spin_unlock(&ci->i_unsafe_lock);
        }
        if (req->r_target_inode &&
            test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags)) {
                struct ceph_inode_info *ci = ceph_inode(req->r_target_inode);
                spin_lock(&ci->i_unsafe_lock);
                list_del_init(&req->r_unsafe_target_item);
                spin_unlock(&ci->i_unsafe_lock);
        }

        if (req->r_unsafe_dir) {
                iput(req->r_unsafe_dir);
                req->r_unsafe_dir = NULL;
        }

        complete_all(&req->r_safe_completion);

        ceph_mdsc_put_request(req);
}

/*
 * Walk back up the dentry tree until we hit a dentry representing a
 * non-snapshot inode. We do this using the rcu_read_lock (which must be held
 * when calling this) to ensure that the objects won't disappear while we're
 * working with them. Once we hit a candidate dentry, we attempt to take a
 * reference to it, and return that as the result.
 */
static struct inode *get_nonsnap_parent(struct dentry *dentry)
{
        struct inode *inode = NULL;

        while (dentry && !IS_ROOT(dentry)) {
                inode = d_inode_rcu(dentry);
                if (!inode || ceph_snap(inode) == CEPH_NOSNAP)
                        break;
                dentry = dentry->d_parent;
        }
        if (inode)
                inode = igrab(inode);
        return inode;
}

/*
 * Choose mds to send request to next.  If there is a hint set in the
 * request (e.g., due to a prior forward hint from the mds), use that.
 * Otherwise, consult frag tree and/or caps to identify the
 * appropriate mds.  If all else fails, choose randomly.
 *
 * Called under mdsc->mutex.
 */

static int __choose_mds(struct ceph_mds_client *mdsc,
                        struct ceph_mds_request *req,
                        bool *random)
{
        struct inode *inode;
        struct ceph_inode_info *ci;
        struct ceph_cap *cap;
        int mode = req->r_direct_mode;
        int mds = -1;
        u32 hash = req->r_direct_hash;
        bool is_hash = test_bit(CEPH_MDS_R_DIRECT_IS_HASH, &req->r_req_flags);

        if (random)
                *random = false;

        /*
         * is there a specific mds we should try?  ignore hint if we have
         * no session and the mds is not up (active or recovering).
         */
        if (req->r_resend_mds >= 0 &&
            (__have_session(mdsc, req->r_resend_mds) ||
             ceph_mdsmap_get_state(mdsc->mdsmap, req->r_resend_mds) > 0)) {
                dout("%s using resend_mds mds%d\n", __func__,
                     req->r_resend_mds);
                return req->r_resend_mds;
        }

        if (mode == USE_RANDOM_MDS)
                goto random;

        inode = NULL;
        if (req->r_inode) {
                if (ceph_snap(req->r_inode) != CEPH_SNAPDIR) {
                        inode = req->r_inode;
                        ihold(inode);
                } else {
                        /* req->r_dentry is non-null for LSSNAP request */
                        rcu_read_lock();
                        inode = get_nonsnap_parent(req->r_dentry);
                        rcu_read_unlock();
                        dout("%s using snapdir's parent %p\n", __func__, inode);
                }
        } else if (req->r_dentry) {
                /* ignore race with rename; old or new d_parent is okay */
                struct dentry *parent;
                struct inode *dir;

                rcu_read_lock();
                parent = READ_ONCE(req->r_dentry->d_parent);
                dir = req->r_parent ? : d_inode_rcu(parent);

                if (!dir || dir->i_sb != mdsc->fsc->sb) {
                        /*  not this fs or parent went negative */
                        inode = d_inode(req->r_dentry);
                        if (inode)
                                ihold(inode);
                } else if (ceph_snap(dir) != CEPH_NOSNAP) {
                        /* direct snapped/virtual snapdir requests
                         * based on parent dir inode */
                        inode = get_nonsnap_parent(parent);
                        dout("%s using nonsnap parent %p\n", __func__, inode);
                } else {
                        /* dentry target */
                        inode = d_inode(req->r_dentry);
                        if (!inode || mode == USE_AUTH_MDS) {
                                /* dir + name */
                                inode = igrab(dir);
                                hash = ceph_dentry_hash(dir, req->r_dentry);
                                is_hash = true;
                        } else {
                                ihold(inode);
                        }
                }
                rcu_read_unlock();
        }

        dout("%s %p is_hash=%d (0x%x) mode %d\n", __func__, inode, (int)is_hash,
             hash, mode);
        if (!inode)
                goto random;
        ci = ceph_inode(inode);

        if (is_hash && S_ISDIR(inode->i_mode)) {
                struct ceph_inode_frag frag;
                int found;

                ceph_choose_frag(ci, hash, &frag, &found);
                if (found) {
                        if (mode == USE_ANY_MDS && frag.ndist > 0) {
                                u8 r;

                                /* choose a random replica */
                                get_random_bytes(&r, 1);
                                r %= frag.ndist;
                                mds = frag.dist[r];
                                dout("%s %p %llx.%llx frag %u mds%d (%d/%d)\n",
                                     __func__, inode, ceph_vinop(inode),
                                     frag.frag, mds, (int)r, frag.ndist);
                                if (ceph_mdsmap_get_state(mdsc->mdsmap, mds) >=
                                    CEPH_MDS_STATE_ACTIVE &&
                                    !ceph_mdsmap_is_laggy(mdsc->mdsmap, mds))
                                        goto out;
                        }

                        /* since this file/dir wasn't known to be
                         * replicated, then we want to look for the
                         * authoritative mds. */
                        if (frag.mds >= 0) {
                                /* choose auth mds */
                                mds = frag.mds;
                                dout("%s %p %llx.%llx frag %u mds%d (auth)\n",
                                     __func__, inode, ceph_vinop(inode),
                                     frag.frag, mds);
                                if (ceph_mdsmap_get_state(mdsc->mdsmap, mds) >=
                                    CEPH_MDS_STATE_ACTIVE) {
                                        if (!ceph_mdsmap_is_laggy(mdsc->mdsmap,
                                                                  mds))
                                                goto out;
                                }
                        }
                        mode = USE_AUTH_MDS;
                }
        }

        spin_lock(&ci->i_ceph_lock);
        cap = NULL;
        if (mode == USE_AUTH_MDS)
                cap = ci->i_auth_cap;
        if (!cap && !RB_EMPTY_ROOT(&ci->i_caps))
                cap = rb_entry(rb_first(&ci->i_caps), struct ceph_cap, ci_node);
        if (!cap) {
                spin_unlock(&ci->i_ceph_lock);
                iput(inode);
                goto random;
        }
        mds = cap->session->s_mds;
        dout("%s %p %llx.%llx mds%d (%scap %p)\n", __func__,
             inode, ceph_vinop(inode), mds,
             cap == ci->i_auth_cap ? "auth " : "", cap);
        spin_unlock(&ci->i_ceph_lock);
out:
        iput(inode);
        return mds;

random:
        if (random)
                *random = true;

        mds = ceph_mdsmap_get_random_mds(mdsc->mdsmap);
        dout("%s chose random mds%d\n", __func__, mds);
        return mds;
}


/*
 * session messages
 */
static struct ceph_msg *create_session_msg(u32 op, u64 seq)
{
        struct ceph_msg *msg;
        struct ceph_mds_session_head *h;

        msg = ceph_msg_new(CEPH_MSG_CLIENT_SESSION, sizeof(*h), GFP_NOFS,
                           false);
        if (!msg) {
                pr_err("create_session_msg ENOMEM creating msg\n");
                return NULL;
        }
        h = msg->front.iov_base;
        h->op = cpu_to_le32(op);
        h->seq = cpu_to_le64(seq);

        return msg;
}

static const unsigned char feature_bits[] = CEPHFS_FEATURES_CLIENT_SUPPORTED;
#define FEATURE_BYTES(c) (DIV_ROUND_UP((size_t)feature_bits[c - 1] + 1, 64) * 8)
static int encode_supported_features(void **p, void *end)
{
        static const size_t count = ARRAY_SIZE(feature_bits);

        if (count > 0) {
                size_t i;
                size_t size = FEATURE_BYTES(count);

                if (WARN_ON_ONCE(*p + 4 + size > end))
                        return -ERANGE;

                ceph_encode_32(p, size);
                memset(*p, 0, size);
                for (i = 0; i < count; i++)
                        ((unsigned char*)(*p))[i / 8] |= BIT(feature_bits[i] % 8);
                *p += size;
        } else {
                if (WARN_ON_ONCE(*p + 4 > end))
                        return -ERANGE;

                ceph_encode_32(p, 0);
        }

        return 0;
}

static const unsigned char metric_bits[] = CEPHFS_METRIC_SPEC_CLIENT_SUPPORTED;
#define METRIC_BYTES(cnt) (DIV_ROUND_UP((size_t)metric_bits[cnt - 1] + 1, 64) * 8)
static int encode_metric_spec(void **p, void *end)
{
        static const size_t count = ARRAY_SIZE(metric_bits);

        /* header */
        if (WARN_ON_ONCE(*p + 2 > end))
                return -ERANGE;

        ceph_encode_8(p, 1); /* version */
        ceph_encode_8(p, 1); /* compat */

        if (count > 0) {
                size_t i;
                size_t size = METRIC_BYTES(count);

                if (WARN_ON_ONCE(*p + 4 + 4 + size > end))
                        return -ERANGE;

                /* metric spec info length */
                ceph_encode_32(p, 4 + size);

                /* metric spec */
                ceph_encode_32(p, size);
                memset(*p, 0, size);
                for (i = 0; i < count; i++)
                        ((unsigned char *)(*p))[i / 8] |= BIT(metric_bits[i] % 8);
                *p += size;
        } else {
                if (WARN_ON_ONCE(*p + 4 + 4 > end))
                        return -ERANGE;

                /* metric spec info length */
                ceph_encode_32(p, 4);
                /* metric spec */
                ceph_encode_32(p, 0);
        }

        return 0;
}

/*
 * session message, specialization for CEPH_SESSION_REQUEST_OPEN
 * to include additional client metadata fields.
 */
static struct ceph_msg *create_session_open_msg(struct ceph_mds_client *mdsc, u64 seq)
{
        struct ceph_msg *msg;
        struct ceph_mds_session_head *h;
        int i;
        int extra_bytes = 0;
        int metadata_key_count = 0;
        struct ceph_options *opt = mdsc->fsc->client->options;
        struct ceph_mount_options *fsopt = mdsc->fsc->mount_options;
        size_t size, count;
        void *p, *end;
        int ret;

        const char* metadata[][2] = {
                {"hostname", mdsc->nodename},
                {"kernel_version", init_utsname()->release},
                {"entity_id", opt->name ? : ""},
                {"root", fsopt->server_path ? : "/"},
                {NULL, NULL}
        };

        /* Calculate serialized length of metadata */
        extra_bytes = 4;  /* map length */
        for (i = 0; metadata[i][0]; ++i) {
                extra_bytes += 8 + strlen(metadata[i][0]) +
                        strlen(metadata[i][1]);
                metadata_key_count++;
        }

        /* supported feature */
        size = 0;
        count = ARRAY_SIZE(feature_bits);
        if (count > 0)
                size = FEATURE_BYTES(count);
        extra_bytes += 4 + size;

        /* metric spec */
        size = 0;
        count = ARRAY_SIZE(metric_bits);
        if (count > 0)
                size = METRIC_BYTES(count);
        extra_bytes += 2 + 4 + 4 + size;

        /* Allocate the message */
        msg = ceph_msg_new(CEPH_MSG_CLIENT_SESSION, sizeof(*h) + extra_bytes,
                           GFP_NOFS, false);
        if (!msg) {
                pr_err("create_session_msg ENOMEM creating msg\n");
                return ERR_PTR(-ENOMEM);
        }
        p = msg->front.iov_base;
        end = p + msg->front.iov_len;

        h = p;
        h->op = cpu_to_le32(CEPH_SESSION_REQUEST_OPEN);
        h->seq = cpu_to_le64(seq);

        /*
         * Serialize client metadata into waiting buffer space, using
         * the format that userspace expects for map<string, string>
         *
         * ClientSession messages with metadata are v4
         */
        msg->hdr.version = cpu_to_le16(4);
        msg->hdr.compat_version = cpu_to_le16(1);

        /* The write pointer, following the session_head structure */
        p += sizeof(*h);

        /* Number of entries in the map */
        ceph_encode_32(&p, metadata_key_count);

        /* Two length-prefixed strings for each entry in the map */
        for (i = 0; metadata[i][0]; ++i) {
                size_t const key_len = strlen(metadata[i][0]);
                size_t const val_len = strlen(metadata[i][1]);

                ceph_encode_32(&p, key_len);
                memcpy(p, metadata[i][0], key_len);
                p += key_len;
                ceph_encode_32(&p, val_len);
                memcpy(p, metadata[i][1], val_len);
                p += val_len;
        }

        ret = encode_supported_features(&p, end);
        if (ret) {
                pr_err("encode_supported_features failed!\n");
                ceph_msg_put(msg);
                return ERR_PTR(ret);
        }

        ret = encode_metric_spec(&p, end);
        if (ret) {
                pr_err("encode_metric_spec failed!\n");
                ceph_msg_put(msg);
                return ERR_PTR(ret);
        }

        msg->front.iov_len = p - msg->front.iov_base;
        msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);

        return msg;
}

/*
 * send session open request.
 *
 * called under mdsc->mutex
 */
static int __open_session(struct ceph_mds_client *mdsc,
                          struct ceph_mds_session *session)
{
        struct ceph_msg *msg;
        int mstate;
        int mds = session->s_mds;

        /* wait for mds to go active? */
        mstate = ceph_mdsmap_get_state(mdsc->mdsmap, mds);
        dout("open_session to mds%d (%s)\n", mds,
             ceph_mds_state_name(mstate));
        session->s_state = CEPH_MDS_SESSION_OPENING;
        session->s_renew_requested = jiffies;

        /* send connect message */
        msg = create_session_open_msg(mdsc, session->s_seq);
        if (IS_ERR(msg))
                return PTR_ERR(msg);
        ceph_con_send(&session->s_con, msg);
        return 0;
}

/*
 * open sessions for any export targets for the given mds
 *
 * called under mdsc->mutex
 */
static struct ceph_mds_session *
__open_export_target_session(struct ceph_mds_client *mdsc, int target)
{
        struct ceph_mds_session *session;
        int ret;

        session = __ceph_lookup_mds_session(mdsc, target);
        if (!session) {
                session = register_session(mdsc, target);
                if (IS_ERR(session))
                        return session;
        }
        if (session->s_state == CEPH_MDS_SESSION_NEW ||
            session->s_state == CEPH_MDS_SESSION_CLOSING) {
                ret = __open_session(mdsc, session);
                if (ret)
                        return ERR_PTR(ret);
        }

        return session;
}

struct ceph_mds_session *
ceph_mdsc_open_export_target_session(struct ceph_mds_client *mdsc, int target)
{
        struct ceph_mds_session *session;

        dout("open_export_target_session to mds%d\n", target);

        mutex_lock(&mdsc->mutex);
        session = __open_export_target_session(mdsc, target);
        mutex_unlock(&mdsc->mutex);

        return session;
}

static void __open_export_target_sessions(struct ceph_mds_client *mdsc,
                                          struct ceph_mds_session *session)
{
        struct ceph_mds_info *mi;
        struct ceph_mds_session *ts;
        int i, mds = session->s_mds;

        if (mds >= mdsc->mdsmap->possible_max_rank)
                return;

        mi = &mdsc->mdsmap->m_info[mds];
        dout("open_export_target_sessions for mds%d (%d targets)\n",
             session->s_mds, mi->num_export_targets);

        for (i = 0; i < mi->num_export_targets; i++) {
                ts = __open_export_target_session(mdsc, mi->export_targets[i]);
                ceph_put_mds_session(ts);
        }
}

void ceph_mdsc_open_export_target_sessions(struct ceph_mds_client *mdsc,
                                           struct ceph_mds_session *session)
{
        mutex_lock(&mdsc->mutex);
        __open_export_target_sessions(mdsc, session);
        mutex_unlock(&mdsc->mutex);
}

/*
 * session caps
 */

static void detach_cap_releases(struct ceph_mds_session *session,
                                struct list_head *target)
{
        lockdep_assert_held(&session->s_cap_lock);

        list_splice_init(&session->s_cap_releases, target);
        session->s_num_cap_releases = 0;
        dout("dispose_cap_releases mds%d\n", session->s_mds);
}

static void dispose_cap_releases(struct ceph_mds_client *mdsc,
                                 struct list_head *dispose)
{
        while (!list_empty(dispose)) {
                struct ceph_cap *cap;
                /* zero out the in-progress message */
                cap = list_first_entry(dispose, struct ceph_cap, session_caps);
                list_del(&cap->session_caps);
                ceph_put_cap(mdsc, cap);
        }
}

static void cleanup_session_requests(struct ceph_mds_client *mdsc,
                                     struct ceph_mds_session *session)
{
        struct ceph_mds_request *req;
        struct rb_node *p;
        struct ceph_inode_info *ci;

        dout("cleanup_session_requests mds%d\n", session->s_mds);
        mutex_lock(&mdsc->mutex);
        while (!list_empty(&session->s_unsafe)) {
                req = list_first_entry(&session->s_unsafe,
                                       struct ceph_mds_request, r_unsafe_item);
                pr_warn_ratelimited(" dropping unsafe request %llu\n",
                                    req->r_tid);
                if (req->r_target_inode) {
                        /* dropping unsafe change of inode's attributes */
                        ci = ceph_inode(req->r_target_inode);
                        errseq_set(&ci->i_meta_err, -EIO);
                }
                if (req->r_unsafe_dir) {
                        /* dropping unsafe directory operation */
                        ci = ceph_inode(req->r_unsafe_dir);
                        errseq_set(&ci->i_meta_err, -EIO);
                }
                __unregister_request(mdsc, req);
        }
        /* zero r_attempts, so kick_requests() will re-send requests */
        p = rb_first(&mdsc->request_tree);
        while (p) {
                req = rb_entry(p, struct ceph_mds_request, r_node);
                p = rb_next(p);
                if (req->r_session &&
                    req->r_session->s_mds == session->s_mds)
                        req->r_attempts = 0;
        }
        mutex_unlock(&mdsc->mutex);
}

/*
 * Helper to safely iterate over all caps associated with a session, with
 * special care taken to handle a racing __ceph_remove_cap().
 *
 * Caller must hold session s_mutex.
 */
int ceph_iterate_session_caps(struct ceph_mds_session *session,
                              int (*cb)(struct inode *, struct ceph_cap *,
                                        void *), void *arg)
{
        struct list_head *p;
        struct ceph_cap *cap;
        struct inode *inode, *last_inode = NULL;
        struct ceph_cap *old_cap = NULL;
        int ret;

        dout("iterate_session_caps %p mds%d\n", session, session->s_mds);
        spin_lock(&session->s_cap_lock);
        p = session->s_caps.next;
        while (p != &session->s_caps) {
                cap = list_entry(p, struct ceph_cap, session_caps);
                inode = igrab(&cap->ci->vfs_inode);
                if (!inode) {
                        p = p->next;
                        continue;
                }
                session->s_cap_iterator = cap;
                spin_unlock(&session->s_cap_lock);

                if (last_inode) {
                        iput(last_inode);
                        last_inode = NULL;
                }
                if (old_cap) {
                        ceph_put_cap(session->s_mdsc, old_cap);
                        old_cap = NULL;
                }

                ret = cb(inode, cap, arg);
                last_inode = inode;

                spin_lock(&session->s_cap_lock);
                p = p->next;
                if (!cap->ci) {
                        dout("iterate_session_caps  finishing cap %p removal\n",
                             cap);
                        BUG_ON(cap->session != session);
                        cap->session = NULL;
                        list_del_init(&cap->session_caps);
                        session->s_nr_caps--;
                        atomic64_dec(&session->s_mdsc->metric.total_caps);
                        if (cap->queue_release)
                                __ceph_queue_cap_release(session, cap);
                        else
                                old_cap = cap;  /* put_cap it w/o locks held */
                }
                if (ret < 0)
                        goto out;
        }
        ret = 0;
out:
        session->s_cap_iterator = NULL;
        spin_unlock(&session->s_cap_lock);

        iput(last_inode);
        if (old_cap)
                ceph_put_cap(session->s_mdsc, old_cap);

        return ret;
}

static int remove_session_caps_cb(struct inode *inode, struct ceph_cap *cap,
                                  void *arg)
{
        struct ceph_fs_client *fsc = (struct ceph_fs_client *)arg;
        struct ceph_inode_info *ci = ceph_inode(inode);
        LIST_HEAD(to_remove);
        bool dirty_dropped = false;
        bool invalidate = false;

        dout("removing cap %p, ci is %p, inode is %p\n",
             cap, ci, &ci->vfs_inode);
        spin_lock(&ci->i_ceph_lock);
        __ceph_remove_cap(cap, false);
        if (!ci->i_auth_cap) {
                struct ceph_cap_flush *cf;
                struct ceph_mds_client *mdsc = fsc->mdsc;

                if (READ_ONCE(fsc->mount_state) >= CEPH_MOUNT_SHUTDOWN) {
                        if (inode->i_data.nrpages > 0)
                                invalidate = true;
                        if (ci->i_wrbuffer_ref > 0)
                                mapping_set_error(&inode->i_data, -EIO);
                }

                while (!list_empty(&ci->i_cap_flush_list)) {
                        cf = list_first_entry(&ci->i_cap_flush_list,
                                              struct ceph_cap_flush, i_list);
                        list_move(&cf->i_list, &to_remove);
                }

                spin_lock(&mdsc->cap_dirty_lock);

                list_for_each_entry(cf, &to_remove, i_list)
                        list_del_init(&cf->g_list);

                if (!list_empty(&ci->i_dirty_item)) {
                        pr_warn_ratelimited(
                                " dropping dirty %s state for %p %lld\n",
                                ceph_cap_string(ci->i_dirty_caps),
                                inode, ceph_ino(inode));
                        ci->i_dirty_caps = 0;
                        list_del_init(&ci->i_dirty_item);
                        dirty_dropped = true;
                }
                if (!list_empty(&ci->i_flushing_item)) {
                        pr_warn_ratelimited(
                                " dropping dirty+flushing %s state for %p %lld\n",
                                ceph_cap_string(ci->i_flushing_caps),
                                inode, ceph_ino(inode));
                        ci->i_flushing_caps = 0;
                        list_del_init(&ci->i_flushing_item);
                        mdsc->num_cap_flushing--;
                        dirty_dropped = true;
                }
                spin_unlock(&mdsc->cap_dirty_lock);

                if (dirty_dropped) {
                        errseq_set(&ci->i_meta_err, -EIO);

                        if (ci->i_wrbuffer_ref_head == 0 &&
                            ci->i_wr_ref == 0 &&
                            ci->i_dirty_caps == 0 &&
                            ci->i_flushing_caps == 0) {
                                ceph_put_snap_context(ci->i_head_snapc);
                                ci->i_head_snapc = NULL;
                        }
                }

                if (atomic_read(&ci->i_filelock_ref) > 0) {
                        /* make further file lock syscall return -EIO */
                        ci->i_ceph_flags |= CEPH_I_ERROR_FILELOCK;
                        pr_warn_ratelimited(" dropping file locks for %p %lld\n",
                                            inode, ceph_ino(inode));
                }

                if (!ci->i_dirty_caps && ci->i_prealloc_cap_flush) {
                        list_add(&ci->i_prealloc_cap_flush->i_list, &to_remove);
                        ci->i_prealloc_cap_flush = NULL;
                }
        }
        spin_unlock(&ci->i_ceph_lock);
        while (!list_empty(&to_remove)) {
                struct ceph_cap_flush *cf;
                cf = list_first_entry(&to_remove,
                                      struct ceph_cap_flush, i_list);
                list_del_init(&cf->i_list);
                if (!cf->is_capsnap)
                        ceph_free_cap_flush(cf);
        }

        wake_up_all(&ci->i_cap_wq);
        if (invalidate)
                ceph_queue_invalidate(inode);
        if (dirty_dropped)
                iput(inode);
        return 0;
}

/*
 * caller must hold session s_mutex
 */
static void remove_session_caps(struct ceph_mds_session *session)
{
        struct ceph_fs_client *fsc = session->s_mdsc->fsc;
        struct super_block *sb = fsc->sb;
        LIST_HEAD(dispose);

        dout("remove_session_caps on %p\n", session);
        ceph_iterate_session_caps(session, remove_session_caps_cb, fsc);

        wake_up_all(&fsc->mdsc->cap_flushing_wq);

        spin_lock(&session->s_cap_lock);
        if (session->s_nr_caps > 0) {
                struct inode *inode;
                struct ceph_cap *cap, *prev = NULL;
                struct ceph_vino vino;
                /*
                 * iterate_session_caps() skips inodes that are being
                 * deleted, we need to wait until deletions are complete.
                 * __wait_on_freeing_inode() is designed for the job,
                 * but it is not exported, so use lookup inode function
                 * to access it.
                 */
                while (!list_empty(&session->s_caps)) {
                        cap = list_entry(session->s_caps.next,
                                         struct ceph_cap, session_caps);
                        if (cap == prev)
                                break;
                        prev = cap;
                        vino = cap->ci->i_vino;
                        spin_unlock(&session->s_cap_lock);

                        inode = ceph_find_inode(sb, vino);
                        iput(inode);

                        spin_lock(&session->s_cap_lock);
                }
        }

        // drop cap expires and unlock s_cap_lock
        detach_cap_releases(session, &dispose);

        BUG_ON(session->s_nr_caps > 0);
        BUG_ON(!list_empty(&session->s_cap_flushing));
        spin_unlock(&session->s_cap_lock);
        dispose_cap_releases(session->s_mdsc, &dispose);
}

enum {
        RECONNECT,
        RENEWCAPS,
        FORCE_RO,
};

/*
 * wake up any threads waiting on this session's caps.  if the cap is
 * old (didn't get renewed on the client reconnect), remove it now.
 *
 * caller must hold s_mutex.
 */
static int wake_up_session_cb(struct inode *inode, struct ceph_cap *cap,
                              void *arg)
{
        struct ceph_inode_info *ci = ceph_inode(inode);
        unsigned long ev = (unsigned long)arg;

        if (ev == RECONNECT) {
                spin_lock(&ci->i_ceph_lock);
                ci->i_wanted_max_size = 0;
                ci->i_requested_max_size = 0;
                spin_unlock(&ci->i_ceph_lock);
        } else if (ev == RENEWCAPS) {
                if (cap->cap_gen < atomic_read(&cap->session->s_cap_gen)) {
                        /* mds did not re-issue stale cap */
                        spin_lock(&ci->i_ceph_lock);
                        cap->issued = cap->implemented = CEPH_CAP_PIN;
                        spin_unlock(&ci->i_ceph_lock);
                }
        } else if (ev == FORCE_RO) {
        }
        wake_up_all(&ci->i_cap_wq);
        return 0;
}

static void wake_up_session_caps(struct ceph_mds_session *session, int ev)
{
        dout("wake_up_session_caps %p mds%d\n", session, session->s_mds);
        ceph_iterate_session_caps(session, wake_up_session_cb,
                                  (void *)(unsigned long)ev);
}

/*
 * Send periodic message to MDS renewing all currently held caps.  The
 * ack will reset the expiration for all caps from this session.
 *
 * caller holds s_mutex
 */
static int send_renew_caps(struct ceph_mds_client *mdsc,
                           struct ceph_mds_session *session)
{
        struct ceph_msg *msg;
        int state;

        if (time_after_eq(jiffies, session->s_cap_ttl) &&
            time_after_eq(session->s_cap_ttl, session->s_renew_requested))
                pr_info("mds%d caps stale\n", session->s_mds);
        session->s_renew_requested = jiffies;

        /* do not try to renew caps until a recovering mds has reconnected
         * with its clients. */
        state = ceph_mdsmap_get_state(mdsc->mdsmap, session->s_mds);
        if (state < CEPH_MDS_STATE_RECONNECT) {
                dout("send_renew_caps ignoring mds%d (%s)\n",
                     session->s_mds, ceph_mds_state_name(state));
                return 0;
        }

        dout("send_renew_caps to mds%d (%s)\n", session->s_mds,
                ceph_mds_state_name(state));
        msg = create_session_msg(CEPH_SESSION_REQUEST_RENEWCAPS,
                                 ++session->s_renew_seq);
        if (!msg)
                return -ENOMEM;
        ceph_con_send(&session->s_con, msg);
        return 0;
}

static int send_flushmsg_ack(struct ceph_mds_client *mdsc,
                             struct ceph_mds_session *session, u64 seq)
{
        struct ceph_msg *msg;

        dout("send_flushmsg_ack to mds%d (%s)s seq %lld\n",
             session->s_mds, ceph_session_state_name(session->s_state), seq);
        msg = create_session_msg(CEPH_SESSION_FLUSHMSG_ACK, seq);
        if (!msg)
                return -ENOMEM;
        ceph_con_send(&session->s_con, msg);
        return 0;
}


/*
 * Note new cap ttl, and any transition from stale -> not stale (fresh?).
 *
 * Called under session->s_mutex
 */
static void renewed_caps(struct ceph_mds_client *mdsc,
                         struct ceph_mds_session *session, int is_renew)
{
        int was_stale;
        int wake = 0;

        spin_lock(&session->s_cap_lock);
        was_stale = is_renew && time_after_eq(jiffies, session->s_cap_ttl);

        session->s_cap_ttl = session->s_renew_requested +
                mdsc->mdsmap->m_session_timeout*HZ;

        if (was_stale) {
                if (time_before(jiffies, session->s_cap_ttl)) {
                        pr_info("mds%d caps renewed\n", session->s_mds);
                        wake = 1;
                } else {
                        pr_info("mds%d caps still stale\n", session->s_mds);
                }
        }
        dout("renewed_caps mds%d ttl now %lu, was %s, now %s\n",
             session->s_mds, session->s_cap_ttl, was_stale ? "stale" : "fresh",
             time_before(jiffies, session->s_cap_ttl) ? "stale" : "fresh");
        spin_unlock(&session->s_cap_lock);

        if (wake)
                wake_up_session_caps(session, RENEWCAPS);
}

/*
 * send a session close request
 */
static int request_close_session(struct ceph_mds_session *session)
{
        struct ceph_msg *msg;

        dout("request_close_session mds%d state %s seq %lld\n",
             session->s_mds, ceph_session_state_name(session->s_state),
             session->s_seq);
        msg = create_session_msg(CEPH_SESSION_REQUEST_CLOSE, session->s_seq);
        if (!msg)
                return -ENOMEM;
        ceph_con_send(&session->s_con, msg);
        return 1;
}

/*
 * Called with s_mutex held.
 */
static int __close_session(struct ceph_mds_client *mdsc,
                         struct ceph_mds_session *session)
{
        if (session->s_state >= CEPH_MDS_SESSION_CLOSING)
                return 0;
        session->s_state = CEPH_MDS_SESSION_CLOSING;
        return request_close_session(session);
}

static bool drop_negative_children(struct dentry *dentry)
{
        struct dentry *child;
        bool all_negative = true;

        if (!d_is_dir(dentry))
                goto out;

        spin_lock(&dentry->d_lock);
        list_for_each_entry(child, &dentry->d_subdirs, d_child) {
                if (d_really_is_positive(child)) {
                        all_negative = false;
                        break;
                }
        }
        spin_unlock(&dentry->d_lock);

        if (all_negative)
                shrink_dcache_parent(dentry);
out:
        return all_negative;
}

/*
 * Trim old(er) caps.
 *
 * Because we can't cache an inode without one or more caps, we do
 * this indirectly: if a cap is unused, we prune its aliases, at which
 * point the inode will hopefully get dropped to.
 *
 * Yes, this is a bit sloppy.  Our only real goal here is to respond to
 * memory pressure from the MDS, though, so it needn't be perfect.
 */
static int trim_caps_cb(struct inode *inode, struct ceph_cap *cap, void *arg)
{
        int *remaining = arg;
        struct ceph_inode_info *ci = ceph_inode(inode);
        int used, wanted, oissued, mine;

        if (*remaining <= 0)
                return -1;

        spin_lock(&ci->i_ceph_lock);
        mine = cap->issued | cap->implemented;
        used = __ceph_caps_used(ci);
        wanted = __ceph_caps_file_wanted(ci);
        oissued = __ceph_caps_issued_other(ci, cap);

        dout("trim_caps_cb %p cap %p mine %s oissued %s used %s wanted %s\n",
             inode, cap, ceph_cap_string(mine), ceph_cap_string(oissued),
             ceph_cap_string(used), ceph_cap_string(wanted));
        if (cap == ci->i_auth_cap) {
                if (ci->i_dirty_caps || ci->i_flushing_caps ||
                    !list_empty(&ci->i_cap_snaps))
                        goto out;
                if ((used | wanted) & CEPH_CAP_ANY_WR)
                        goto out;
                /* Note: it's possible that i_filelock_ref becomes non-zero
                 * after dropping auth caps. It doesn't hurt because reply
                 * of lock mds request will re-add auth caps. */
                if (atomic_read(&ci->i_filelock_ref) > 0)
                        goto out;
        }
        /* The inode has cached pages, but it's no longer used.
         * we can safely drop it */
        if (S_ISREG(inode->i_mode) &&
            wanted == 0 && used == CEPH_CAP_FILE_CACHE &&
            !(oissued & CEPH_CAP_FILE_CACHE)) {
          used = 0;
          oissued = 0;
        }
        if ((used | wanted) & ~oissued & mine)
                goto out;   /* we need these caps */

        if (oissued) {
                /* we aren't the only cap.. just remove us */
                __ceph_remove_cap(cap, true);
                (*remaining)--;
        } else {
                struct dentry *dentry;
                /* try dropping referring dentries */
                spin_unlock(&ci->i_ceph_lock);
                dentry = d_find_any_alias(inode);
                if (dentry && drop_negative_children(dentry)) {
                        int count;
                        dput(dentry);
                        d_prune_aliases(inode);
                        count = atomic_read(&inode->i_count);
                        if (count == 1)
                                (*remaining)--;
                        dout("trim_caps_cb %p cap %p pruned, count now %d\n",
                             inode, cap, count);
                } else {
                        dput(dentry);
                }
                return 0;
        }

out:
        spin_unlock(&ci->i_ceph_lock);
        return 0;
}

/*
 * Trim session cap count down to some max number.
 */
int ceph_trim_caps(struct ceph_mds_client *mdsc,
                   struct ceph_mds_session *session,
                   int max_caps)

{
        int trim_caps = session->s_nr_caps - max_caps;

        dout("trim_caps mds%d start: %d / %d, trim %d\n",
             session->s_mds, session->s_nr_caps, max_caps, trim_caps);
        if (trim_caps > 0) {
                int remaining = trim_caps;

                ceph_iterate_session_caps(session, trim_caps_cb, &remaining);
                dout("trim_caps mds%d done: %d / %d, trimmed %d\n",
                     session->s_mds, session->s_nr_caps, max_caps,
                        trim_caps - remaining);
        }

        ceph_flush_cap_releases(mdsc, session);
        return 0;
}

static int check_caps_flush(struct ceph_mds_client *mdsc,
                            u64 want_flush_tid)
{
        int ret = 1;

        spin_lock(&mdsc->cap_dirty_lock);
        if (!list_empty(&mdsc->cap_flush_list)) {
                struct ceph_cap_flush *cf =
                        list_first_entry(&mdsc->cap_flush_list,
                                         struct ceph_cap_flush, g_list);
                if (cf->tid <= want_flush_tid) {
                        dout("check_caps_flush still flushing tid "
                             "%llu <= %llu\n", cf->tid, want_flush_tid);
                        ret = 0;
                }
        }
        spin_unlock(&mdsc->cap_dirty_lock);
        return ret;
}

/*
 * flush all dirty inode data to disk.
 *
 * returns true if we've flushed through want_flush_tid
 */
static void wait_caps_flush(struct ceph_mds_client *mdsc,
                            u64 want_flush_tid)
{
        dout("check_caps_flush want %llu\n", want_flush_tid);

        wait_event(mdsc->cap_flushing_wq,
                   check_caps_flush(mdsc, want_flush_tid));

        dout("check_caps_flush ok, flushed thru %llu\n", want_flush_tid);
}

/*
 * called under s_mutex
 */
static void ceph_send_cap_releases(struct ceph_mds_client *mdsc,
                                   struct ceph_mds_session *session)
{
        struct ceph_msg *msg = NULL;
        struct ceph_mds_cap_release *head;
        struct ceph_mds_cap_item *item;
        struct ceph_osd_client *osdc = &mdsc->fsc->client->osdc;
        struct ceph_cap *cap;
        LIST_HEAD(tmp_list);
        int num_cap_releases;
        __le32  barrier, *cap_barrier;

        down_read(&osdc->lock);
        barrier = cpu_to_le32(osdc->epoch_barrier);
        up_read(&osdc->lock);

        spin_lock(&session->s_cap_lock);
again:
        list_splice_init(&session->s_cap_releases, &tmp_list);
        num_cap_releases = session->s_num_cap_releases;
        session->s_num_cap_releases = 0;
        spin_unlock(&session->s_cap_lock);

        while (!list_empty(&tmp_list)) {
                if (!msg) {
                        msg = ceph_msg_new(CEPH_MSG_CLIENT_CAPRELEASE,
                                        PAGE_SIZE, GFP_NOFS, false);
                        if (!msg)
                                goto out_err;
                        head = msg->front.iov_base;
                        head->num = cpu_to_le32(0);
                        msg->front.iov_len = sizeof(*head);

                        msg->hdr.version = cpu_to_le16(2);
                        msg->hdr.compat_version = cpu_to_le16(1);
                }

                cap = list_first_entry(&tmp_list, struct ceph_cap,
                                        session_caps);
                list_del(&cap->session_caps);
                num_cap_releases--;

                head = msg->front.iov_base;
                put_unaligned_le32(get_unaligned_le32(&head->num) + 1,
                                   &head->num);
                item = msg->front.iov_base + msg->front.iov_len;
                item->ino = cpu_to_le64(cap->cap_ino);
                item->cap_id = cpu_to_le64(cap->cap_id);
                item->migrate_seq = cpu_to_le32(cap->mseq);
                item->seq = cpu_to_le32(cap->issue_seq);
                msg->front.iov_len += sizeof(*item);

                ceph_put_cap(mdsc, cap);

                if (le32_to_cpu(head->num) == CEPH_CAPS_PER_RELEASE) {
                        // Append cap_barrier field
                        cap_barrier = msg->front.iov_base + msg->front.iov_len;
                        *cap_barrier = barrier;
                        msg->front.iov_len += sizeof(*cap_barrier);

                        msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
                        dout("send_cap_releases mds%d %p\n", session->s_mds, msg);
                        ceph_con_send(&session->s_con, msg);
                        msg = NULL;
                }
        }

        BUG_ON(num_cap_releases != 0);

        spin_lock(&session->s_cap_lock);
        if (!list_empty(&session->s_cap_releases))
                goto again;
        spin_unlock(&session->s_cap_lock);

        if (msg) {
                // Append cap_barrier field
                cap_barrier = msg->front.iov_base + msg->front.iov_len;
                *cap_barrier = barrier;
                msg->front.iov_len += sizeof(*cap_barrier);

                msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
                dout("send_cap_releases mds%d %p\n", session->s_mds, msg);
                ceph_con_send(&session->s_con, msg);
        }
        return;
out_err:
        pr_err("send_cap_releases mds%d, failed to allocate message\n",
                session->s_mds);
        spin_lock(&session->s_cap_lock);
        list_splice(&tmp_list, &session->s_cap_releases);
        session->s_num_cap_releases += num_cap_releases;
        spin_unlock(&session->s_cap_lock);
}

static void ceph_cap_release_work(struct work_struct *work)
{
        struct ceph_mds_session *session =
                container_of(work, struct ceph_mds_session, s_cap_release_work);

        mutex_lock(&session->s_mutex);
        if (session->s_state == CEPH_MDS_SESSION_OPEN ||
            session->s_state == CEPH_MDS_SESSION_HUNG)
                ceph_send_cap_releases(session->s_mdsc, session);
        mutex_unlock(&session->s_mutex);
        ceph_put_mds_session(session);
}

void ceph_flush_cap_releases(struct ceph_mds_client *mdsc,
                             struct ceph_mds_session *session)
{
        if (mdsc->stopping)
                return;

        ceph_get_mds_session(session);
        if (queue_work(mdsc->fsc->cap_wq,
                       &session->s_cap_release_work)) {
                dout("cap release work queued\n");
        } else {
                ceph_put_mds_session(session);
                dout("failed to queue cap release work\n");
        }
}

/*
 * caller holds session->s_cap_lock
 */
void __ceph_queue_cap_release(struct ceph_mds_session *session,
                              struct ceph_cap *cap)
{
        list_add_tail(&cap->session_caps, &session->s_cap_releases);
        session->s_num_cap_releases++;

        if (!(session->s_num_cap_releases % CEPH_CAPS_PER_RELEASE))
                ceph_flush_cap_releases(session->s_mdsc, session);
}

static void ceph_cap_reclaim_work(struct work_struct *work)
{
        struct ceph_mds_client *mdsc =
                container_of(work, struct ceph_mds_client, cap_reclaim_work);
        int ret = ceph_trim_dentries(mdsc);
        if (ret == -EAGAIN)
                ceph_queue_cap_reclaim_work(mdsc);
}

void ceph_queue_cap_reclaim_work(struct ceph_mds_client *mdsc)
{
        if (mdsc->stopping)
                return;

        if (queue_work(mdsc->fsc->cap_wq, &mdsc->cap_reclaim_work)) {
                dout("caps reclaim work queued\n");
        } else {
                dout("failed to queue caps release work\n");
        }
}

void ceph_reclaim_caps_nr(struct ceph_mds_client *mdsc, int nr)
{
        int val;
        if (!nr)
                return;
        val = atomic_add_return(nr, &mdsc->cap_reclaim_pending);
        if ((val % CEPH_CAPS_PER_RELEASE) < nr) {
                atomic_set(&mdsc->cap_reclaim_pending, 0);
                ceph_queue_cap_reclaim_work(mdsc);
        }
}

/*
 * requests
 */

int ceph_alloc_readdir_reply_buffer(struct ceph_mds_request *req,
                                    struct inode *dir)
{
        struct ceph_inode_info *ci = ceph_inode(dir);
        struct ceph_mds_reply_info_parsed *rinfo = &req->r_reply_info;
        struct ceph_mount_options *opt = req->r_mdsc->fsc->mount_options;
        size_t size = sizeof(struct ceph_mds_reply_dir_entry);
        unsigned int num_entries;
        int order;

        spin_lock(&ci->i_ceph_lock);
        num_entries = ci->i_files + ci->i_subdirs;
        spin_unlock(&ci->i_ceph_lock);
        num_entries = max(num_entries, 1U);
        num_entries = min(num_entries, opt->max_readdir);

        order = get_order(size * num_entries);
        while (order >= 0) {
                rinfo->dir_entries = (void*)__get_free_pages(GFP_KERNEL |
                                                             __GFP_NOWARN,
                                                             order);
                if (rinfo->dir_entries)
                        break;
                order--;
        }
        if (!rinfo->dir_entries)
                return -ENOMEM;

        num_entries = (PAGE_SIZE << order) / size;
        num_entries = min(num_entries, opt->max_readdir);

        rinfo->dir_buf_size = PAGE_SIZE << order;
        req->r_num_caps = num_entries + 1;
        req->r_args.readdir.max_entries = cpu_to_le32(num_entries);
        req->r_args.readdir.max_bytes = cpu_to_le32(opt->max_readdir_bytes);
        return 0;
}

/*
 * Create an mds request.
 */
struct ceph_mds_request *
ceph_mdsc_create_request(struct ceph_mds_client *mdsc, int op, int mode)
{
        struct ceph_mds_request *req;

        req = kmem_cache_zalloc(ceph_mds_request_cachep, GFP_NOFS);
        if (!req)
                return ERR_PTR(-ENOMEM);

        mutex_init(&req->r_fill_mutex);
        req->r_mdsc = mdsc;
        req->r_started = jiffies;
        req->r_start_latency = ktime_get();
        req->r_resend_mds = -1;
        INIT_LIST_HEAD(&req->r_unsafe_dir_item);
        INIT_LIST_HEAD(&req->r_unsafe_target_item);
        req->r_fmode = -1;
        kref_init(&req->r_kref);
        RB_CLEAR_NODE(&req->r_node);
        INIT_LIST_HEAD(&req->r_wait);
        init_completion(&req->r_completion);
        init_completion(&req->r_safe_completion);
        INIT_LIST_HEAD(&req->r_unsafe_item);

        ktime_get_coarse_real_ts64(&req->r_stamp);

        req->r_op = op;
        req->r_direct_mode = mode;
        return req;
}

/*
 * return oldest (lowest) request, tid in request tree, 0 if none.
 *
 * called under mdsc->mutex.
 */
static struct ceph_mds_request *__get_oldest_req(struct ceph_mds_client *mdsc)
{
        if (RB_EMPTY_ROOT(&mdsc->request_tree))
                return NULL;
        return rb_entry(rb_first(&mdsc->request_tree),
                        struct ceph_mds_request, r_node);
}

static inline  u64 __get_oldest_tid(struct ceph_mds_client *mdsc)
{
        return mdsc->oldest_tid;
}

/*
 * Build a dentry's path.  Allocate on heap; caller must kfree.  Based
 * on build_path_from_dentry in fs/cifs/dir.c.
 *
 * If @stop_on_nosnap, generate path relative to the first non-snapped
 * inode.
 *
 * Encode hidden .snap dirs as a double /, i.e.
 *   foo/.snap/bar -> foo//bar
 */
char *ceph_mdsc_build_path(struct dentry *dentry, int *plen, u64 *pbase,
                           int stop_on_nosnap)
{
        struct dentry *temp;
        char *path;
        int pos;
        unsigned seq;
        u64 base;

        if (!dentry)
                return ERR_PTR(-EINVAL);

        path = __getname();
        if (!path)
                return ERR_PTR(-ENOMEM);
retry:
        pos = PATH_MAX - 1;
        path[pos] = '\0';

        seq = read_seqbegin(&rename_lock);
        rcu_read_lock();
        temp = dentry;
        for (;;) {
                struct inode *inode;

                spin_lock(&temp->d_lock);
                inode = d_inode(temp);
                if (inode && ceph_snap(inode) == CEPH_SNAPDIR) {
                        dout("build_path path+%d: %p SNAPDIR\n",
                             pos, temp);
                } else if (stop_on_nosnap && inode && dentry != temp &&
                           ceph_snap(inode) == CEPH_NOSNAP) {
                        spin_unlock(&temp->d_lock);
                        pos++; /* get rid of any prepended '/' */
                        break;
                } else {
                        pos -= temp->d_name.len;
                        if (pos < 0) {
                                spin_unlock(&temp->d_lock);
                                break;
                        }
                        memcpy(path + pos, temp->d_name.name, temp->d_name.len);
                }
                spin_unlock(&temp->d_lock);
                temp = READ_ONCE(temp->d_parent);

                /* Are we at the root? */
                if (IS_ROOT(temp))
                        break;

                /* Are we out of buffer? */
                if (--pos < 0)
                        break;

                path[pos] = '/';
        }
        base = ceph_ino(d_inode(temp));
        rcu_read_unlock();

        if (read_seqretry(&rename_lock, seq))
                goto retry;

        if (pos < 0) {
                /*
                 * A rename didn't occur, but somehow we didn't end up where
                 * we thought we would. Throw a warning and try again.
                 */
                pr_warn("build_path did not end path lookup where "
                        "expected, pos is %d\n", pos);
                goto retry;
        }

        *pbase = base;
        *plen = PATH_MAX - 1 - pos;
        dout("build_path on %p %d built %llx '%.*s'\n",
             dentry, d_count(dentry), base, *plen, path + pos);
        return path + pos;
}

static int build_dentry_path(struct dentry *dentry, struct inode *dir,
                             const char **ppath, int *ppathlen, u64 *pino,
                             bool *pfreepath, bool parent_locked)
{
        char *path;

        rcu_read_lock();
        if (!dir)
                dir = d_inode_rcu(dentry->d_parent);
        if (dir && parent_locked && ceph_snap(dir) == CEPH_NOSNAP) {
                *pino = ceph_ino(dir);
                rcu_read_unlock();
                *ppath = dentry->d_name.name;
                *ppathlen = dentry->d_name.len;
                return 0;
        }
        rcu_read_unlock();
        path = ceph_mdsc_build_path(dentry, ppathlen, pino, 1);
        if (IS_ERR(path))
                return PTR_ERR(path);
        *ppath = path;
        *pfreepath = true;
        return 0;
}

static int build_inode_path(struct inode *inode,
                            const char **ppath, int *ppathlen, u64 *pino,
                            bool *pfreepath)
{
        struct dentry *dentry;
        char *path;

        if (ceph_snap(inode) == CEPH_NOSNAP) {
                *pino = ceph_ino(inode);
                *ppathlen = 0;
                return 0;
        }
        dentry = d_find_alias(inode);
        path = ceph_mdsc_build_path(dentry, ppathlen, pino, 1);
        dput(dentry);
        if (IS_ERR(path))
                return PTR_ERR(path);
        *ppath = path;
        *pfreepath = true;
        return 0;
}

/*
 * request arguments may be specified via an inode *, a dentry *, or
 * an explicit ino+path.
 */
static int set_request_path_attr(struct inode *rinode, struct dentry *rdentry,
                                  struct inode *rdiri, const char *rpath,
                                  u64 rino, const char **ppath, int *pathlen,
                                  u64 *ino, bool *freepath, bool parent_locked)
{
        int r = 0;

        if (rinode) {
                r = build_inode_path(rinode, ppath, pathlen, ino, freepath);
                dout(" inode %p %llx.%llx\n", rinode, ceph_ino(rinode),
                     ceph_snap(rinode));
        } else if (rdentry) {
                r = build_dentry_path(rdentry, rdiri, ppath, pathlen, ino,
                                        freepath, parent_locked);
                dout(" dentry %p %llx/%.*s\n", rdentry, *ino, *pathlen,
                     *ppath);
        } else if (rpath || rino) {
                *ino = rino;
                *ppath = rpath;
                *pathlen = rpath ? strlen(rpath) : 0;
                dout(" path %.*s\n", *pathlen, rpath);
        }

        return r;
}

static void encode_timestamp_and_gids(void **p,
                                      const struct ceph_mds_request *req)
{
        struct ceph_timespec ts;
        int i;

        ceph_encode_timespec64(&ts, &req->r_stamp);
        ceph_encode_copy(p, &ts, sizeof(ts));

        /* gid_list */
        ceph_encode_32(p, req->r_cred->group_info->ngroups);
        for (i = 0; i < req->r_cred->group_info->ngroups; i++)
                ceph_encode_64(p, from_kgid(&init_user_ns,
                                            req->r_cred->group_info->gid[i]));
}

/*
 * called under mdsc->mutex
 */
static struct ceph_msg *create_request_message(struct ceph_mds_session *session,
                                               struct ceph_mds_request *req,
                                               bool drop_cap_releases)
{
        int mds = session->s_mds;
        struct ceph_mds_client *mdsc = session->s_mdsc;
        struct ceph_msg *msg;
        struct ceph_mds_request_head_old *head;
        const char *path1 = NULL;
        const char *path2 = NULL;
        u64 ino1 = 0, ino2 = 0;
        int pathlen1 = 0, pathlen2 = 0;
        bool freepath1 = false, freepath2 = false;
        int len;
        u16 releases;
        void *p, *end;
        int ret;
        bool legacy = !(session->s_con.peer_features & CEPH_FEATURE_FS_BTIME);

        ret = set_request_path_attr(req->r_inode, req->r_dentry,
                              req->r_parent, req->r_path1, req->r_ino1.ino,
                              &path1, &pathlen1, &ino1, &freepath1,
                              test_bit(CEPH_MDS_R_PARENT_LOCKED,
                                        &req->r_req_flags));
        if (ret < 0) {
                msg = ERR_PTR(ret);
                goto out;
        }

        /* If r_old_dentry is set, then assume that its parent is locked */
        ret = set_request_path_attr(NULL, req->r_old_dentry,
                              req->r_old_dentry_dir,
                              req->r_path2, req->r_ino2.ino,
                              &path2, &pathlen2, &ino2, &freepath2, true);
        if (ret < 0) {
                msg = ERR_PTR(ret);
                goto out_free1;
        }

        len = legacy ? sizeof(*head) : sizeof(struct ceph_mds_request_head);
        len += pathlen1 + pathlen2 + 2*(1 + sizeof(u32) + sizeof(u64)) +
                sizeof(struct ceph_timespec);
        len += sizeof(u32) + (sizeof(u64) * req->r_cred->group_info->ngroups);

        /* calculate (max) length for cap releases */
        len += sizeof(struct ceph_mds_request_release) *
                (!!req->r_inode_drop + !!req->r_dentry_drop +
                 !!req->r_old_inode_drop + !!req->r_old_dentry_drop);

        if (req->r_dentry_drop)
                len += pathlen1;
        if (req->r_old_dentry_drop)
                len += pathlen2;

        msg = ceph_msg_new2(CEPH_MSG_CLIENT_REQUEST, len, 1, GFP_NOFS, false);
        if (!msg) {
                msg = ERR_PTR(-ENOMEM);
                goto out_free2;
        }

        msg->hdr.tid = cpu_to_le64(req->r_tid);

        /*
         * The old ceph_mds_request_head didn't contain a version field, and
         * one was added when we moved the message version from 3->4.
         */
        if (legacy) {
                msg->hdr.version = cpu_to_le16(3);
                head = msg->front.iov_base;
                p = msg->front.iov_base + sizeof(*head);
        } else {
                struct ceph_mds_request_head *new_head = msg->front.iov_base;

                msg->hdr.version = cpu_to_le16(4);
                new_head->version = cpu_to_le16(CEPH_MDS_REQUEST_HEAD_VERSION);
                head = (struct ceph_mds_request_head_old *)&new_head->oldest_client_tid;
                p = msg->front.iov_base + sizeof(*new_head);
        }

        end = msg->front.iov_base + msg->front.iov_len;

        head->mdsmap_epoch = cpu_to_le32(mdsc->mdsmap->m_epoch);
        head->op = cpu_to_le32(req->r_op);
        head->caller_uid = cpu_to_le32(from_kuid(&init_user_ns,
                                                 req->r_cred->fsuid));
        head->caller_gid = cpu_to_le32(from_kgid(&init_user_ns,
                                                 req->r_cred->fsgid));
        head->ino = cpu_to_le64(req->r_deleg_ino);
        head->args = req->r_args;

        ceph_encode_filepath(&p, end, ino1, path1);
        ceph_encode_filepath(&p, end, ino2, path2);

        /* make note of release offset, in case we need to replay */
        req->r_request_release_offset = p - msg->front.iov_base;

        /* cap releases */
        releases = 0;
        if (req->r_inode_drop)
                releases += ceph_encode_inode_release(&p,
                      req->r_inode ? req->r_inode : d_inode(req->r_dentry),
                      mds, req->r_inode_drop, req->r_inode_unless,
                      req->r_op == CEPH_MDS_OP_READDIR);
        if (req->r_dentry_drop)
                releases += ceph_encode_dentry_release(&p, req->r_dentry,
                                req->r_parent, mds, req->r_dentry_drop,
                                req->r_dentry_unless);
        if (req->r_old_dentry_drop)
                releases += ceph_encode_dentry_release(&p, req->r_old_dentry,
                                req->r_old_dentry_dir, mds,
                                req->r_old_dentry_drop,
                                req->r_old_dentry_unless);
        if (req->r_old_inode_drop)
                releases += ceph_encode_inode_release(&p,
                      d_inode(req->r_old_dentry),
                      mds, req->r_old_inode_drop, req->r_old_inode_unless, 0);

        if (drop_cap_releases) {
                releases = 0;
                p = msg->front.iov_base + req->r_request_release_offset;
        }

        head->num_releases = cpu_to_le16(releases);

        encode_timestamp_and_gids(&p, req);

        if (WARN_ON_ONCE(p > end)) {
                ceph_msg_put(msg);
                msg = ERR_PTR(-ERANGE);
                goto out_free2;
        }

        msg->front.iov_len = p - msg->front.iov_base;
        msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);

        if (req->r_pagelist) {
                struct ceph_pagelist *pagelist = req->r_pagelist;
                ceph_msg_data_add_pagelist(msg, pagelist);
                msg->hdr.data_len = cpu_to_le32(pagelist->length);
        } else {
                msg->hdr.data_len = 0;
        }

        msg->hdr.data_off = cpu_to_le16(0);

out_free2:
        if (freepath2)
                ceph_mdsc_free_path((char *)path2, pathlen2);
out_free1:
        if (freepath1)
                ceph_mdsc_free_path((char *)path1, pathlen1);
out:
        return msg;
}

/*
 * called under mdsc->mutex if error, under no mutex if
 * success.
 */
static void complete_request(struct ceph_mds_client *mdsc,
                             struct ceph_mds_request *req)
{
        req->r_end_latency = ktime_get();

        if (req->r_callback)
                req->r_callback(mdsc, req);
        complete_all(&req->r_completion);
}

static struct ceph_mds_request_head_old *
find_old_request_head(void *p, u64 features)
{
        bool legacy = !(features & CEPH_FEATURE_FS_BTIME);
        struct ceph_mds_request_head *new_head;

        if (legacy)
                return (struct ceph_mds_request_head_old *)p;
        new_head = (struct ceph_mds_request_head *)p;
        return (struct ceph_mds_request_head_old *)&new_head->oldest_client_tid;
}

/*
 * called under mdsc->mutex
 */
static int __prepare_send_request(struct ceph_mds_session *session,
                                  struct ceph_mds_request *req,
                                  bool drop_cap_releases)
{
        int mds = session->s_mds;
        struct ceph_mds_client *mdsc = session->s_mdsc;
        struct ceph_mds_request_head_old *rhead;
        struct ceph_msg *msg;
        int flags = 0;

        req->r_attempts++;
        if (req->r_inode) {
                struct ceph_cap *cap =
                        ceph_get_cap_for_mds(ceph_inode(req->r_inode), mds);

                if (cap)
                        req->r_sent_on_mseq = cap->mseq;
                else
                        req->r_sent_on_mseq = -1;
        }
        dout("prepare_send_request %p tid %lld %s (attempt %d)\n", req,
             req->r_tid, ceph_mds_op_name(req->r_op), req->r_attempts);

        if (test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags)) {
                void *p;

                /*
                 * Replay.  Do not regenerate message (and rebuild
                 * paths, etc.); just use the original message.
                 * Rebuilding paths will break for renames because
                 * d_move mangles the src name.
                 */
                msg = req->r_request;
                rhead = find_old_request_head(msg->front.iov_base,
                                              session->s_con.peer_features);

                flags = le32_to_cpu(rhead->flags);
                flags |= CEPH_MDS_FLAG_REPLAY;
                rhead->flags = cpu_to_le32(flags);

                if (req->r_target_inode)
                        rhead->ino = cpu_to_le64(ceph_ino(req->r_target_inode));

                rhead->num_retry = req->r_attempts - 1;

                /* remove cap/dentry releases from message */
                rhead->num_releases = 0;

                p = msg->front.iov_base + req->r_request_release_offset;
                encode_timestamp_and_gids(&p, req);

                msg->front.iov_len = p - msg->front.iov_base;
                msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
                return 0;
        }

        if (req->r_request) {
                ceph_msg_put(req->r_request);
                req->r_request = NULL;
        }
        msg = create_request_message(session, req, drop_cap_releases);
        if (IS_ERR(msg)) {
                req->r_err = PTR_ERR(msg);
                return PTR_ERR(msg);
        }
        req->r_request = msg;

        rhead = find_old_request_head(msg->front.iov_base,
                                      session->s_con.peer_features);
        rhead->oldest_client_tid = cpu_to_le64(__get_oldest_tid(mdsc));
        if (test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags))
                flags |= CEPH_MDS_FLAG_REPLAY;
        if (test_bit(CEPH_MDS_R_ASYNC, &req->r_req_flags))
                flags |= CEPH_MDS_FLAG_ASYNC;
        if (req->r_parent)
                flags |= CEPH_MDS_FLAG_WANT_DENTRY;
        rhead->flags = cpu_to_le32(flags);
        rhead->num_fwd = req->r_num_fwd;
        rhead->num_retry = req->r_attempts - 1;

        dout(" r_parent = %p\n", req->r_parent);
        return 0;
}

/*
 * called under mdsc->mutex
 */
static int __send_request(struct ceph_mds_session *session,
                          struct ceph_mds_request *req,
                          bool drop_cap_releases)
{
        int err;

        err = __prepare_send_request(session, req, drop_cap_releases);
        if (!err) {
                ceph_msg_get(req->r_request);
                ceph_con_send(&session->s_con, req->r_request);
        }

        return err;
}

/*
 * send request, or put it on the appropriate wait list.
 */
static void __do_request(struct ceph_mds_client *mdsc,
                        struct ceph_mds_request *req)
{
        struct ceph_mds_session *session = NULL;
        int mds = -1;
        int err = 0;
        bool random;

        if (req->r_err || test_bit(CEPH_MDS_R_GOT_RESULT, &req->r_req_flags)) {
                if (test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags))
                        __unregister_request(mdsc, req);
                return;
        }

        if (req->r_timeout &&
            time_after_eq(jiffies, req->r_started + req->r_timeout)) {
                dout("do_request timed out\n");
                err = -ETIMEDOUT;
                goto finish;
        }
        if (READ_ONCE(mdsc->fsc->mount_state) == CEPH_MOUNT_SHUTDOWN) {
                dout("do_request forced umount\n");
                err = -EIO;
                goto finish;
        }
        if (READ_ONCE(mdsc->fsc->mount_state) == CEPH_MOUNT_MOUNTING) {
                if (mdsc->mdsmap_err) {
                        err = mdsc->mdsmap_err;
                        dout("do_request mdsmap err %d\n", err);
                        goto finish;
                }
                if (mdsc->mdsmap->m_epoch == 0) {
                        dout("do_request no mdsmap, waiting for map\n");
                        list_add(&req->r_wait, &mdsc->waiting_for_map);
                        return;
                }
                if (!(mdsc->fsc->mount_options->flags &
                      CEPH_MOUNT_OPT_MOUNTWAIT) &&
                    !ceph_mdsmap_is_cluster_available(mdsc->mdsmap)) {
                        err = -EHOSTUNREACH;
                        goto finish;
                }
        }

        put_request_session(req);

        mds = __choose_mds(mdsc, req, &random);
        if (mds < 0 ||
            ceph_mdsmap_get_state(mdsc->mdsmap, mds) < CEPH_MDS_STATE_ACTIVE) {
                if (test_bit(CEPH_MDS_R_ASYNC, &req->r_req_flags)) {
                        err = -EJUKEBOX;
                        goto finish;
                }
                dout("do_request no mds or not active, waiting for map\n");
                list_add(&req->r_wait, &mdsc->waiting_for_map);
                return;
        }

        /* get, open session */
        session = __ceph_lookup_mds_session(mdsc, mds);
        if (!session) {
                session = register_session(mdsc, mds);
                if (IS_ERR(session)) {
                        err = PTR_ERR(session);
                        goto finish;
                }
        }
        req->r_session = ceph_get_mds_session(session);

        dout("do_request mds%d session %p state %s\n", mds, session,
             ceph_session_state_name(session->s_state));
        if (session->s_state != CEPH_MDS_SESSION_OPEN &&
            session->s_state != CEPH_MDS_SESSION_HUNG) {
                /*
                 * We cannot queue async requests since the caps and delegated
                 * inodes are bound to the session. Just return -EJUKEBOX and
                 * let the caller retry a sync request in that case.
                 */
                if (test_bit(CEPH_MDS_R_ASYNC, &req->r_req_flags)) {
                        err = -EJUKEBOX;
                        goto out_session;
                }

                /*
                 * If the session has been REJECTED, then return a hard error,
                 * unless it's a CLEANRECOVER mount, in which case we'll queue
                 * it to the mdsc queue.
                 */
                if (session->s_state == CEPH_MDS_SESSION_REJECTED) {
                        if (ceph_test_mount_opt(mdsc->fsc, CLEANRECOVER))
                                list_add(&req->r_wait, &mdsc->waiting_for_map);
                        else
                                err = -EACCES;
                        goto out_session;
                }

                if (session->s_state == CEPH_MDS_SESSION_NEW ||
                    session->s_state == CEPH_MDS_SESSION_CLOSING) {
                        err = __open_session(mdsc, session);
                        if (err)
                                goto out_session;
                        /* retry the same mds later */
                        if (random)
                                req->r_resend_mds = mds;
                }
                list_add(&req->r_wait, &session->s_waiting);
                goto out_session;
        }

        /* send request */
        req->r_resend_mds = -1;   /* forget any previous mds hint */

        if (req->r_request_started == 0)   /* note request start time */
                req->r_request_started = jiffies;

        err = __send_request(session, req, false);

out_session:
        ceph_put_mds_session(session);
finish:
        if (err) {
                dout("__do_request early error %d\n", err);
                req->r_err = err;
                complete_request(mdsc, req);
                __unregister_request(mdsc, req);
        }
        return;
}

/*
 * called under mdsc->mutex
 */
static void __wake_requests(struct ceph_mds_client *mdsc,
                            struct list_head *head)
{
        struct ceph_mds_request *req;
        LIST_HEAD(tmp_list);

        list_splice_init(head, &tmp_list);

        while (!list_empty(&tmp_list)) {
                req = list_entry(tmp_list.next,
                                 struct ceph_mds_request, r_wait);
                list_del_init(&req->r_wait);
                dout(" wake request %p tid %llu\n", req, req->r_tid);
                __do_request(mdsc, req);
        }
}

/*
 * Wake up threads with requests pending for @mds, so that they can
 * resubmit their requests to a possibly different mds.
 */
static void kick_requests(struct ceph_mds_client *mdsc, int mds)
{
        struct ceph_mds_request *req;
        struct rb_node *p = rb_first(&mdsc->request_tree);

        dout("kick_requests mds%d\n", mds);
        while (p) {
                req = rb_entry(p, struct ceph_mds_request, r_node);
                p = rb_next(p);
                if (test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags))
                        continue;
                if (req->r_attempts > 0)
                        continue; /* only new requests */
                if (req->r_session &&
                    req->r_session->s_mds == mds) {
                        dout(" kicking tid %llu\n", req->r_tid);
                        list_del_init(&req->r_wait);
                        __do_request(mdsc, req);
                }
        }
}

int ceph_mdsc_submit_request(struct ceph_mds_client *mdsc, struct inode *dir,
                              struct ceph_mds_request *req)
{
        int err = 0;

        /* take CAP_PIN refs for r_inode, r_parent, r_old_dentry */
        if (req->r_inode)
                ceph_get_cap_refs(ceph_inode(req->r_inode), CEPH_CAP_PIN);
        if (req->r_parent) {
                struct ceph_inode_info *ci = ceph_inode(req->r_parent);
                int fmode = (req->r_op & CEPH_MDS_OP_WRITE) ?
                            CEPH_FILE_MODE_WR : CEPH_FILE_MODE_RD;
                spin_lock(&ci->i_ceph_lock);
                ceph_take_cap_refs(ci, CEPH_CAP_PIN, false);
                __ceph_touch_fmode(ci, mdsc, fmode);
                spin_unlock(&ci->i_ceph_lock);
        }
        if (req->r_old_dentry_dir)
                ceph_get_cap_refs(ceph_inode(req->r_old_dentry_dir),
                                  CEPH_CAP_PIN);

        if (req->r_inode) {
                err = ceph_wait_on_async_create(req->r_inode);
                if (err) {
                        dout("%s: wait for async create returned: %d\n",
                             __func__, err);
                        return err;
                }
        }

        if (!err && req->r_old_inode) {