/*
 * Virtio 9p backend
 *
 * Copyright IBM, Corp. 2010
 *
 * Authors:
 *  Anthony Liguori   <aliguori@us.ibm.com>
 *
 * This work is licensed under the terms of the GNU GPL, version 2.  See
 * the COPYING file in the top-level directory.
 *
 */

/*
 * Not so fast! You might want to read the 9p developer docs first:
 * https://wiki.qemu.org/Documentation/9p
 */

#include "qemu/osdep.h"
#ifdef CONFIG_LINUX
#include <linux/limits.h>
#else
#include <limits.h>
#endif
#include <glib/gprintf.h>
#include "hw/virtio/virtio.h"
#include "qapi/error.h"
#include "qemu/error-report.h"
#include "qemu/iov.h"
#include "qemu/main-loop.h"
#include "qemu/sockets.h"
#include "virtio-9p.h"
#include "fsdev/qemu-fsdev.h"
#include "9p-xattr.h"
#include "9p-util.h"
#include "coth.h"
#include "trace.h"
#include "migration/blocker.h"
#include "qemu/xxhash.h"
#include <math.h>

int open_fd_hw;
int total_open_fd;
static int open_fd_rc;

enum {
    Oread   = 0x00,
    Owrite  = 0x01,
    Ordwr   = 0x02,
    Oexec   = 0x03,
    Oexcl   = 0x04,
    Otrunc  = 0x10,
    Orexec  = 0x20,
    Orclose = 0x40,
    Oappend = 0x80,
};

P9ARRAY_DEFINE_TYPE(V9fsPath, v9fs_path_free);

static ssize_t pdu_marshal(V9fsPDU *pdu, size_t offset, const char *fmt, ...)
{
    ssize_t ret;
    va_list ap;

    va_start(ap, fmt);
    ret = pdu->s->transport->pdu_vmarshal(pdu, offset, fmt, ap);
    va_end(ap);

    return ret;
}

static ssize_t pdu_unmarshal(V9fsPDU *pdu, size_t offset, const char *fmt, ...)
{
    ssize_t ret;
    va_list ap;

    va_start(ap, fmt);
    ret = pdu->s->transport->pdu_vunmarshal(pdu, offset, fmt, ap);
    va_end(ap);

    return ret;
}

static int omode_to_uflags(int8_t mode)
{
    int ret = 0;

    switch (mode & 3) {
    case Oread:
        ret = O_RDONLY;
        break;
    case Ordwr:
        ret = O_RDWR;
        break;
    case Owrite:
        ret = O_WRONLY;
        break;
    case Oexec:
        ret = O_RDONLY;
        break;
    }

    if (mode & Otrunc) {
        ret |= O_TRUNC;
    }

    if (mode & Oappend) {
        ret |= O_APPEND;
    }

    if (mode & Oexcl) {
        ret |= O_EXCL;
    }

    return ret;
}

typedef struct DotlOpenflagMap {
    int dotl_flag;
    int open_flag;
} DotlOpenflagMap;

static int dotl_to_open_flags(int flags)
{
    int i;
    /*
     * We have same bits for P9_DOTL_READONLY, P9_DOTL_WRONLY
     * and P9_DOTL_NOACCESS
     */
    int oflags = flags & O_ACCMODE;

    DotlOpenflagMap dotl_oflag_map[] = {
        { P9_DOTL_CREATE, O_CREAT },
        { P9_DOTL_EXCL, O_EXCL },
        { P9_DOTL_NOCTTY , O_NOCTTY },
        { P9_DOTL_TRUNC, O_TRUNC },
        { P9_DOTL_APPEND, O_APPEND },
        { P9_DOTL_NONBLOCK, O_NONBLOCK } ,
        { P9_DOTL_DSYNC, O_DSYNC },
        { P9_DOTL_FASYNC, FASYNC },
#ifndef CONFIG_DARWIN
        { P9_DOTL_NOATIME, O_NOATIME },
        /*
         *  On Darwin, we could map to F_NOCACHE, which is
         *  similar, but doesn't quite have the same
         *  semantics. However, we don't support O_DIRECT
         *  even on linux at the moment, so we just ignore
         *  it here.
         */
        { P9_DOTL_DIRECT, O_DIRECT },
#endif
        { P9_DOTL_LARGEFILE, O_LARGEFILE },
        { P9_DOTL_DIRECTORY, O_DIRECTORY },
        { P9_DOTL_NOFOLLOW, O_NOFOLLOW },
        { P9_DOTL_SYNC, O_SYNC },
    };

    for (i = 0; i < ARRAY_SIZE(dotl_oflag_map); i++) {
        if (flags & dotl_oflag_map[i].dotl_flag) {
            oflags |= dotl_oflag_map[i].open_flag;
        }
    }

    return oflags;
}

void cred_init(FsCred *credp)
{
    credp->fc_uid = -1;
    credp->fc_gid = -1;
    credp->fc_mode = -1;
    credp->fc_rdev = -1;
}

static int get_dotl_openflags(V9fsState *s, int oflags)
{
    int flags;
    /*
     * Filter the client open flags
     */
    flags = dotl_to_open_flags(oflags);
    flags &= ~(O_NOCTTY | O_ASYNC | O_CREAT);
#ifndef CONFIG_DARWIN
    /*
     * Ignore direct disk access hint until the server supports it.
     */
    flags &= ~O_DIRECT;
#endif
    return flags;
}

void v9fs_path_init(V9fsPath *path)
{
    path->data = NULL;
    path->size = 0;
}

void v9fs_path_free(V9fsPath *path)
{
    g_free(path->data);
    path->data = NULL;
    path->size = 0;
}


void G_GNUC_PRINTF(2, 3)
v9fs_path_sprintf(V9fsPath *path, const char *fmt, ...)
{
    va_list ap;

    v9fs_path_free(path);

    va_start(ap, fmt);
    /* Bump the size for including terminating NULL */
    path->size = g_vasprintf(&path->data, fmt, ap) + 1;
    va_end(ap);
}

void v9fs_path_copy(V9fsPath *dst, const V9fsPath *src)
{
    v9fs_path_free(dst);
    dst->size = src->size;
    dst->data = g_memdup(src->data, src->size);
}

int v9fs_name_to_path(V9fsState *s, V9fsPath *dirpath,
                      const char *name, V9fsPath *path)
{
    int err;
    err = s->ops->name_to_path(&s->ctx, dirpath, name, path);
    if (err < 0) {
        err = -errno;
    }
    return err;
}

/*
 * Return TRUE if s1 is an ancestor of s2.
 *
 * E.g. "a/b" is an ancestor of "a/b/c" but not of "a/bc/d".
 * As a special case, We treat s1 as ancestor of s2 if they are same!
 */
static int v9fs_path_is_ancestor(V9fsPath *s1, V9fsPath *s2)
{
    if (!strncmp(s1->data, s2->data, s1->size - 1)) {
        if (s2->data[s1->size - 1] == '\0' || s2->data[s1->size - 1] == '/') {
            return 1;
        }
    }
    return 0;
}

static size_t v9fs_string_size(V9fsString *str)
{
    return str->size;
}

/*
 * returns 0 if fid got re-opened, 1 if not, < 0 on error */
static int coroutine_fn v9fs_reopen_fid(V9fsPDU *pdu, V9fsFidState *f)
{
    int err = 1;
    if (f->fid_type == P9_FID_FILE) {
        if (f->fs.fd == -1) {
            do {
                err = v9fs_co_open(pdu, f, f->open_flags);
            } while (err == -EINTR && !pdu->cancelled);
        }
    } else if (f->fid_type == P9_FID_DIR) {
        if (f->fs.dir.stream == NULL) {
            do {
                err = v9fs_co_opendir(pdu, f);
            } while (err == -EINTR && !pdu->cancelled);
        }
    }
    return err;
}

static V9fsFidState *coroutine_fn get_fid(V9fsPDU *pdu, int32_t fid)
{
    int err;
    V9fsFidState *f;
    V9fsState *s = pdu->s;

    QSIMPLEQ_FOREACH(f, &s->fid_list, next) {
        BUG_ON(f->clunked);
        if (f->fid == fid) {
            /*
             * Update the fid ref upfront so that
             * we don't get reclaimed when we yield
             * in open later.
             */
            f->ref++;
            /*
             * check whether we need to reopen the
             * file. We might have closed the fd
             * while trying to free up some file
             * descriptors.
             */
            err = v9fs_reopen_fid(pdu, f);
            if (err < 0) {
                f->ref--;
                return NULL;
            }
            /*
             * Mark the fid as referenced so that the LRU
             * reclaim won't close the file descriptor
             */
            f->flags |= FID_REFERENCED;
            return f;
        }
    }
    return NULL;
}

static V9fsFidState *alloc_fid(V9fsState *s, int32_t fid)
{
    V9fsFidState *f;

    QSIMPLEQ_FOREACH(f, &s->fid_list, next) {
        /* If fid is already there return NULL */
        BUG_ON(f->clunked);
        if (f->fid == fid) {
            return NULL;
        }
    }
    f = g_new0(V9fsFidState, 1);
    f->fid = fid;
    f->fid_type = P9_FID_NONE;
    f->ref = 1;
    /*
     * Mark the fid as referenced so that the LRU
     * reclaim won't close the file descriptor
     */
    f->flags |= FID_REFERENCED;
    QSIMPLEQ_INSERT_TAIL(&s->fid_list, f, next);

    v9fs_readdir_init(s->proto_version, &f->fs.dir);
    v9fs_readdir_init(s->proto_version, &f->fs_reclaim.dir);

    return f;
}

static int coroutine_fn v9fs_xattr_fid_clunk(V9fsPDU *pdu, V9fsFidState *fidp)
{
    int retval = 0;

    if (fidp->fs.xattr.xattrwalk_fid) {
        /* getxattr/listxattr fid */
        goto free_value;
    }
    /*
     * if this is fid for setxattr. clunk should
     * result in setxattr localcall
     */
    if (fidp->fs.xattr.len != fidp->fs.xattr.copied_len) {
        /* clunk after partial write */
        retval = -EINVAL;
        goto free_out;
    }
    if (fidp->fs.xattr.len) {
        retval = v9fs_co_lsetxattr(pdu, &fidp->path, &fidp->fs.xattr.name,
                                   fidp->fs.xattr.value,
                                   fidp->fs.xattr.len,
                                   fidp->fs.xattr.flags);
    } else {
        retval = v9fs_co_lremovexattr(pdu, &fidp->path, &fidp->fs.xattr.name);
    }
free_out:
    v9fs_string_free(&fidp->fs.xattr.name);
free_value:
    g_free(fidp->fs.xattr.value);
    return retval;
}

static int coroutine_fn free_fid(V9fsPDU *pdu, V9fsFidState *fidp)
{
    int retval = 0;

    if (fidp->fid_type == P9_FID_FILE) {
        /* If we reclaimed the fd no need to close */
        if (fidp->fs.fd != -1) {
            retval = v9fs_co_close(pdu, &fidp->fs);
        }
    } else if (fidp->fid_type == P9_FID_DIR) {
        if (fidp->fs.dir.stream != NULL) {
            retval = v9fs_co_closedir(pdu, &fidp->fs);
        }
    } else if (fidp->fid_type == P9_FID_XATTR) {
        retval = v9fs_xattr_fid_clunk(pdu, fidp);
    }
    v9fs_path_free(&fidp->path);
    g_free(fidp);
    return retval;
}

static int coroutine_fn put_fid(V9fsPDU *pdu, V9fsFidState *fidp)
{
    BUG_ON(!fidp->ref);
    fidp->ref--;
    /*
     * Don't free the fid if it is in reclaim list
     */
    if (!fidp->ref && fidp->clunked) {
        if (fidp->fid == pdu->s->root_fid) {
            /*
             * if the clunked fid is root fid then we
             * have unmounted the fs on the client side.
             * delete the migration blocker. Ideally, this
             * should be hooked to transport close notification
             */
            if (pdu->s->migration_blocker) {
                migrate_del_blocker(pdu->s->migration_blocker);
                error_free(pdu->s->migration_blocker);
                pdu->s->migration_blocker = NULL;
            }
        }
        return free_fid(pdu, fidp);
    }
    return 0;
}

static V9fsFidState *clunk_fid(V9fsState *s, int32_t fid)
{
    V9fsFidState *fidp;

    QSIMPLEQ_FOREACH(fidp, &s->fid_list, next) {
        if (fidp->fid == fid) {
            QSIMPLEQ_REMOVE(&s->fid_list, fidp, V9fsFidState, next);
            fidp->clunked = true;
            return fidp;
        }
    }
    return NULL;
}

void coroutine_fn v9fs_reclaim_fd(V9fsPDU *pdu)
{
    int reclaim_count = 0;
    V9fsState *s = pdu->s;
    V9fsFidState *f;
    QSLIST_HEAD(, V9fsFidState) reclaim_list =
        QSLIST_HEAD_INITIALIZER(reclaim_list);

    QSIMPLEQ_FOREACH(f, &s->fid_list, next) {
        /*
         * Unlink fids cannot be reclaimed. Check
         * for them and skip them. Also skip fids
         * currently being operated on.
         */
        if (f->ref || f->flags & FID_NON_RECLAIMABLE) {
            continue;
        }
        /*
         * if it is a recently referenced fid
         * we leave the fid untouched and clear the
         * reference bit. We come back to it later
         * in the next iteration. (a simple LRU without
         * moving list elements around)
         */
        if (f->flags & FID_REFERENCED) {
            f->flags &= ~FID_REFERENCED;
            continue;
        }
        /*
         * Add fids to reclaim list.
         */
        if (f->fid_type == P9_FID_FILE) {
            if (f->fs.fd != -1) {
                /*
                 * Up the reference count so that
                 * a clunk request won't free this fid
                 */
                f->ref++;
                QSLIST_INSERT_HEAD(&reclaim_list, f, reclaim_next);
                f->fs_reclaim.fd = f->fs.fd;
                f->fs.fd = -1;
                reclaim_count++;
            }
        } else if (f->fid_type == P9_FID_DIR) {
            if (f->fs.dir.stream != NULL) {
                /*
                 * Up the reference count so that
                 * a clunk request won't free this fid
                 */
                f->ref++;
                QSLIST_INSERT_HEAD(&reclaim_list, f, reclaim_next);
                f->fs_reclaim.dir.stream = f->fs.dir.stream;
                f->fs.dir.stream = NULL;
                reclaim_count++;
            }
        }
        if (reclaim_count >= open_fd_rc) {
            break;
        }
    }
    /*
     * Now close the fid in reclaim list. Free them if they
     * are already clunked.
     */
    while (!QSLIST_EMPTY(&reclaim_list)) {
        f = QSLIST_FIRST(&reclaim_list);
        QSLIST_REMOVE(&reclaim_list, f, V9fsFidState, reclaim_next);
        if (f->fid_type == P9_FID_FILE) {
            v9fs_co_close(pdu, &f->fs_reclaim);
        } else if (f->fid_type == P9_FID_DIR) {
            v9fs_co_closedir(pdu, &f->fs_reclaim);
        }
        /*
         * Now drop the fid reference, free it
         * if clunked.
         */
        put_fid(pdu, f);
    }
}

static int coroutine_fn v9fs_mark_fids_unreclaim(V9fsPDU *pdu, V9fsPath *path)
{
    int err;
    V9fsState *s = pdu->s;
    V9fsFidState *fidp, *fidp_next;

    fidp = QSIMPLEQ_FIRST(&s->fid_list);
    if (!fidp) {
        return 0;
    }

    /*
     * v9fs_reopen_fid() can yield : a reference on the fid must be held
     * to ensure its pointer remains valid and we can safely pass it to
     * QSIMPLEQ_NEXT(). The corresponding put_fid() can also yield so
     * we must keep a reference on the next fid as well. So the logic here
     * is to get a reference on a fid and only put it back during the next
     * iteration after we could get a reference on the next fid. Start with
     * the first one.
     */
    for (fidp->ref++; fidp; fidp = fidp_next) {
        if (fidp->path.size == path->size &&
            !memcmp(fidp->path.data, path->data, path->size)) {
            /* Mark the fid non reclaimable. */
            fidp->flags |= FID_NON_RECLAIMABLE;

            /* reopen the file/dir if already closed */
            err = v9fs_reopen_fid(pdu, fidp);
            if (err < 0) {
                put_fid(pdu, fidp);
                return err;
            }
        }

        fidp_next = QSIMPLEQ_NEXT(fidp, next);

        if (fidp_next) {
            /*
             * Ensure the next fid survives a potential clunk request during
             * put_fid() below and v9fs_reopen_fid() in the next iteration.
             */
            fidp_next->ref++;
        }

        /* We're done with this fid */
        put_fid(pdu, fidp);
    }

    return 0;
}

static void coroutine_fn virtfs_reset(V9fsPDU *pdu)
{
    V9fsState *s = pdu->s;
    V9fsFidState *fidp;

    /* Free all fids */
    while (!QSIMPLEQ_EMPTY(&s->fid_list)) {
        /* Get fid */
        fidp = QSIMPLEQ_FIRST(&s->fid_list);
        fidp->ref++;

        /* Clunk fid */
        QSIMPLEQ_REMOVE(&s->fid_list, fidp, V9fsFidState, next);
        fidp->clunked = true;

        put_fid(pdu, fidp);
    }
}

#define P9_QID_TYPE_DIR         0x80
#define P9_QID_TYPE_SYMLINK     0x02

#define P9_STAT_MODE_DIR        0x80000000
#define P9_STAT_MODE_APPEND     0x40000000
#define P9_STAT_MODE_EXCL       0x20000000
#define P9_STAT_MODE_MOUNT      0x10000000
#define P9_STAT_MODE_AUTH       0x08000000
#define P9_STAT_MODE_TMP        0x04000000
#define P9_STAT_MODE_SYMLINK    0x02000000
#define P9_STAT_MODE_LINK       0x01000000
#define P9_STAT_MODE_DEVICE     0x00800000
#define P9_STAT_MODE_NAMED_PIPE 0x00200000
#define P9_STAT_MODE_SOCKET     0x00100000
#define P9_STAT_MODE_SETUID     0x00080000
#define P9_STAT_MODE_SETGID     0x00040000
#define P9_STAT_MODE_SETVTX     0x00010000

#define P9_STAT_MODE_TYPE_BITS (P9_STAT_MODE_DIR |          \
                                P9_STAT_MODE_SYMLINK |      \
                                P9_STAT_MODE_LINK |         \
                                P9_STAT_MODE_DEVICE |       \
                                P9_STAT_MODE_NAMED_PIPE |   \
                                P9_STAT_MODE_SOCKET)

/* Mirrors all bits of a byte. So e.g. binary 10100000 would become 00000101. */
static inline uint8_t mirror8bit(uint8_t byte)
{
    return (byte * 0x0202020202ULL & 0x010884422010ULL) % 1023;
}

/* Same as mirror8bit() just for a 64 bit data type instead for a byte. */
static inline uint64_t mirror64bit(uint64_t value)
{
    return ((uint64_t)mirror8bit(value         & 0xff) << 56) |
           ((uint64_t)mirror8bit((value >> 8)  & 0xff) << 48) |
           ((uint64_t)mirror8bit((value >> 16) & 0xff) << 40) |
           ((uint64_t)mirror8bit((value >> 24) & 0xff) << 32) |
           ((uint64_t)mirror8bit((value >> 32) & 0xff) << 24) |
           ((uint64_t)mirror8bit((value >> 40) & 0xff) << 16) |
           ((uint64_t)mirror8bit((value >> 48) & 0xff) << 8)  |
           ((uint64_t)mirror8bit((value >> 56) & 0xff));
}

/*
 * Parameter k for the Exponential Golomb algorihm to be used.
 *
 * The smaller this value, the smaller the minimum bit count for the Exp.
 * Golomb generated affixes will be (at lowest index) however for the
 * price of having higher maximum bit count of generated affixes (at highest
 * index). Likewise increasing this parameter yields in smaller maximum bit
 * count for the price of having higher minimum bit count.
 *
 * In practice that means: a good value for k depends on the expected amount
 * of devices to be exposed by one export. For a small amount of devices k
 * should be small, for a large amount of devices k might be increased
 * instead. The default of k=0 should be fine for most users though.
 *
 * IMPORTANT: In case this ever becomes a runtime parameter; the value of
 * k should not change as long as guest is still running! Because that would
 * cause completely different inode numbers to be generated on guest.
 */
#define EXP_GOLOMB_K    0

/**
 * expGolombEncode() - Exponential Golomb algorithm for arbitrary k
 *                     (including k=0).
 *
 * @n: natural number (or index) of the prefix to be generated
 *     (1, 2, 3, ...)
 * @k: parameter k of Exp. Golomb algorithm to be used
 *     (see comment on EXP_GOLOMB_K macro for details about k)
 * Return: prefix for given @n and @k
 *
 * The Exponential Golomb algorithm generates prefixes (NOT suffixes!)
 * with growing length and with the mathematical property of being
 * "prefix-free". The latter means the generated prefixes can be prepended
 * in front of arbitrary numbers and the resulting concatenated numbers are
 * guaranteed to be always unique.
 *
 * This is a minor adjustment to the original Exp. Golomb algorithm in the
 * sense that lowest allowed index (@n) starts with 1, not with zero.
 */
static VariLenAffix expGolombEncode(uint64_t n, int k)
{
    const uint64_t value = n + (1 << k) - 1;
    const int bits = (int) log2(value) + 1;
    return (VariLenAffix) {
        .type = AffixType_Prefix,
        .value = value,
        .bits = bits + MAX((bits - 1 - k), 0)
    };
}

/**
 * invertAffix() - Converts a suffix into a prefix, or a prefix into a suffix.
 * @affix: either suffix or prefix to be inverted
 * Return: inversion of passed @affix
 *
 * Simply mirror all bits of the affix value, for the purpose to preserve
 * respectively the mathematical "prefix-free" or "suffix-free" property
 * after the conversion.
 *
 * If a passed prefix is suitable to create unique numbers, then the
 * returned suffix is suitable to create unique numbers as well (and vice
 * versa).
 */
static VariLenAffix invertAffix(const VariLenAffix *affix)
{
    return (VariLenAffix) {
        .type =
            (affix->type == AffixType_Suffix) ?
                AffixType_Prefix : AffixType_Suffix,
        .value =
            mirror64bit(affix->value) >>
            ((sizeof(affix->value) * 8) - affix->bits),
        .bits = affix->bits
    };
}

/**
 * affixForIndex() - Generates suffix numbers with "suffix-free" property.
 * @index: natural number (or index) of the suffix to be generated
 *         (1, 2, 3, ...)
 * Return: Suffix suitable to assemble unique number.
 *
 * This is just a wrapper function on top of the Exp. Golomb algorithm.
 *
 * Since the Exp. Golomb algorithm generates prefixes, but we need suffixes,
 * this function converts the Exp. Golomb prefixes into appropriate suffixes
 * which are still suitable for generating unique numbers.
 */
static VariLenAffix affixForIndex(uint64_t index)
{
    VariLenAffix prefix;
    prefix = expGolombEncode(index, EXP_GOLOMB_K);
    return invertAffix(&prefix); /* convert prefix to suffix */
}

/* creative abuse of tb_hash_func7, which is based on xxhash */
static uint32_t qpp_hash(QppEntry e)
{
    return qemu_xxhash7(e.ino_prefix, e.dev, 0, 0, 0);
}

static uint32_t qpf_hash(QpfEntry e)
{
    return qemu_xxhash7(e.ino, e.dev, 0, 0, 0);
}

static bool qpd_cmp_func(const void *obj, const void *userp)
{
    const QpdEntry *e1 = obj, *e2 = userp;
    return e1->dev == e2->dev;
}

static bool qpp_cmp_func(const void *obj, const void *userp)
{
    const QppEntry *e1 = obj, *e2 = userp;
    return e1->dev == e2->dev && e1->ino_prefix == e2->ino_prefix;
}

static bool qpf_cmp_func(const void *obj, const void *userp)
{
    const QpfEntry *e1 = obj, *e2 = userp;
    return e1->dev == e2->dev && e1->ino == e2->ino;
}

static void qp_table_remove(void *p, uint32_t h, void *up)
{
    g_free(p);
}

static void qp_table_destroy(struct qht *ht)
{
    if (!ht || !ht->map) {
        return;
    }
    qht_iter(ht, qp_table_remove, NULL);
    qht_destroy(ht);
}

static void qpd_table_init(struct qht *ht)
{
    qht_init(ht, qpd_cmp_func, 1, QHT_MODE_AUTO_RESIZE);
}

static void qpp_table_init(struct qht *ht)
{
    qht_init(ht, qpp_cmp_func, 1, QHT_MODE_AUTO_RESIZE);
}

static void qpf_table_init(struct qht *ht)
{
    qht_init(ht, qpf_cmp_func, 1 << 16, QHT_MODE_AUTO_RESIZE);
}

/*
 * Returns how many (high end) bits of inode numbers of the passed fs
 * device shall be used (in combination with the device number) to
 * generate hash values for qpp_table entries.
 *
 * This function is required if variable length suffixes are used for inode
 * number mapping on guest level. Since a device may end up having multiple
 * entries in qpp_table, each entry most probably with a different suffix
 * length, we thus need this function in conjunction with qpd_table to
 * "agree" about a fix amount of bits (per device) to be always used for
 * generating hash values for the purpose of accessing qpp_table in order
 * get consistent behaviour when accessing qpp_table.
 */
static int qid_inode_prefix_hash_bits(V9fsPDU *pdu, dev_t dev)
{
    QpdEntry lookup = {
        .dev = dev
    }, *val;
    uint32_t hash = dev;
    VariLenAffix affix;

    val = qht_lookup(&pdu->s->qpd_table, &lookup, hash);
    if (!val) {
        val = g_new0(QpdEntry, 1);
        *val = lookup;
        affix = affixForIndex(pdu->s->qp_affix_next);
        val->prefix_bits = affix.bits;
        qht_insert(&pdu->s->qpd_table, val, hash, NULL);
        pdu->s->qp_ndevices++;
    }
    return val->prefix_bits;
}

/*
 * Slow / full mapping host inode nr -> guest inode nr.
 *
 * This function performs a slower and much more costly remapping of an
 * original file inode number on host to an appropriate different inode
 * number on guest. For every (dev, inode) combination on host a new
 * sequential number is generated, cached and exposed as inode number on
 * guest.
 *
 * This is just a "last resort" fallback solution if the much faster/cheaper
 * qid_path_suffixmap() failed. In practice this slow / full mapping is not
 * expected ever to be used at all though.
 *
 * See qid_path_suffixmap() for details
 *
 */
static int qid_path_fullmap(V9fsPDU *pdu, const struct stat *stbuf,
                            uint64_t *path)
{
    QpfEntry lookup = {
        .dev = stbuf->st_dev,
        .ino = stbuf->st_ino
    }, *val;
    uint32_t hash = qpf_hash(lookup);
    VariLenAffix affix;

    val = qht_lookup(&pdu->s->qpf_table, &lookup, hash);

    if (!val) {
        if (pdu->s->qp_fullpath_next == 0) {
            /* no more files can be mapped :'( */
            error_report_once(
                "9p: No more prefixes available for remapping inodes from "
                "host to guest."
            );
            return -ENFILE;
        }

        val = g_new0(QpfEntry, 1);
        *val = lookup;

        /* new unique inode and device combo */
        affix = affixForIndex(
            1ULL << (sizeof(pdu->s->qp_affix_next) * 8)
        );
        val->path = (pdu->s->qp_fullpath_next++ << affix.bits) | affix.value;
        pdu->s->qp_fullpath_next &= ((1ULL << (64 - affix.bits)) - 1);
        qht_insert(&pdu->s->qpf_table, val, hash, NULL);
    }

    *path = val->path;
    return 0;
}

/*
 * Quick mapping host inode nr -> guest inode nr.
 *
 * This function performs quick remapping of an original file inode number
 * on host to an appropriate different inode number on guest. This remapping
 * of inodes is required to avoid inode nr collisions on guest which would
 * happen if the 9p export contains more than 1 exported file system (or
 * more than 1 file system data set), because unlike on host level where the
 * files would have different device nrs, all files exported by 9p would
 * share the same device nr on guest (the device nr of the virtual 9p device
 * that is).
 *
 * Inode remapping is performed by chopping off high end bits of the original
 * inode number from host, shifting the result upwards and then assigning a
 * generated suffix number for the low end bits, where the same suffix number
 * will be shared by all inodes with the same device id AND the same high end
 * bits that have been chopped off. That approach utilizes the fact that inode
 * numbers very likely share the same high end bits (i.e. due to their common
 * sequential generation by file systems) and hence we only have to generate
 * and track a very limited amount of suffixes in practice due to that.
 *
 * We generate variable size suffixes for that purpose. The 1st generated
 * suffix will only have 1 bit and hence we only need to chop off 1 bit from
 * the original inode number. The subsequent suffixes being generated will
 * grow in (bit) size subsequently, i.e. the 2nd and 3rd suffix being
 * generated will have 3 bits and hence we have to chop off 3 bits from their
 * original inodes, and so on. That approach of using variable length suffixes
 * (i.e. over fixed size ones) utilizes the fact that in practice only a very
 * limited amount of devices are shared by the same export (e.g. typically
 * less than 2 dozen devices per 9p export), so in practice we need to chop
 * off less bits than with fixed size prefixes and yet are flexible to add
 * new devices at runtime below host's export directory at any time without
 * having to reboot guest nor requiring to reconfigure guest for that. And due
 * to the very limited amount of original high end bits that we chop off that
 * way, the total amount of suffixes we need to generate is less than by using
 * fixed size prefixes and hence it also improves performance of the inode
 * remapping algorithm, and finally has the nice side effect that the inode
 * numbers on guest will be much smaller & human friendly. ;-)
 */
static int qid_path_suffixmap(V9fsPDU *pdu, const struct stat *stbuf,
                              uint64_t *path)
{
    const int ino_hash_bits = qid_inode_prefix_hash_bits(pdu, stbuf->st_dev);
    QppEntry lookup = {
        .dev = stbuf->st_dev,
        .ino_prefix = (uint16_t) (stbuf->st_ino >> (64 - ino_hash_bits))
    }, *val;
    uint32_t hash = qpp_hash(lookup);

    val = qht_lookup(&pdu->s->qpp_table, &lookup, hash);

    if (!val) {
        if (pdu->s->qp_affix_next == 0) {
            /* we ran out of affixes */
            warn_report_once(
                "9p: Potential degraded performance of inode remapping"
            );
            return -ENFILE;
        }

        val = g_new0(QppEntry, 1);
        *val = lookup;

        /* new unique inode affix and device combo */
        val->qp_affix_index = pdu->s->qp_affix_next++;
        val->qp_affix = affixForIndex(val->qp_affix_index);
        qht_insert(&pdu->s->qpp_table, val, hash, NULL);
    }
    /* assuming generated affix to be suffix type, not prefix */
    *path = (stbuf->st_ino << val->qp_affix.bits) | val->qp_affix.value;
    return 0;
}

static int stat_to_qid(V9fsPDU *pdu, const struct stat *stbuf, V9fsQID *qidp)
{
    int err;
    size_t size;

    if (pdu->s->ctx.export_flags & V9FS_REMAP_INODES) {
        /* map inode+device to qid path (fast path) */
        err = qid_path_suffixmap(pdu, stbuf, &qidp->path);
        if (err == -ENFILE) {
            /* fast path didn't work, fall back to full map */
            err = qid_path_fullmap(pdu, stbuf, &qidp->path);
        }
        if (err) {
            return err;
        }
    } else {
        if (pdu->s->dev_id != stbuf->st_dev) {
            if (pdu->s->ctx.export_flags & V9FS_FORBID_MULTIDEVS) {
                error_report_once(
                    "9p: Multiple devices detected in same VirtFS export. "
                    "Access of guest to additional devices is (partly) "
                    "denied due to virtfs option 'multidevs=forbid' being "
                    "effective."
                );
                return -ENODEV;
            } else {
                warn_report_once(
                    "9p: Multiple devices detected in same VirtFS export, "
                    "which might lead to file ID collisions and severe "
                    "misbehaviours on guest! You should either use a "
                    "separate export for each device shared from host or "
                    "use virtfs option 'multidevs=remap'!"
                );
            }
        }
        memset(&qidp->path, 0, sizeof(qidp->path));
        size = MIN(sizeof(stbuf->st_ino), sizeof(qidp->path));
        memcpy(&qidp->path, &stbuf->st_ino, size);
    }

    qidp->version = stbuf->st_mtime ^ (stbuf->st_size << 8);
    qidp->type = 0;
    if (S_ISDIR(stbuf->st_mode)) {
        qidp->type |= P9_QID_TYPE_DIR;
    }
    if (S_ISLNK(stbuf->st_mode)) {
        qidp->type |= P9_QID_TYPE_SYMLINK;
    }

    return 0;
}

V9fsPDU *pdu_alloc(V9fsState *s)
{
    V9fsPDU *pdu = NULL;

    if (!QLIST_EMPTY(&s->free_list)) {

        pdu = QLIST_FIRST(&s->free_list);
        QLIST_REMOVE(pdu, next);
        QLIST_INSERT_HEAD(&s->active_list, pdu, next);
    }
    return pdu;
}

void pdu_free(V9fsPDU *pdu)
{
    V9fsState *s = pdu->s;

    g_assert(!pdu->cancelled);
    QLIST_REMOVE(pdu, next);
    QLIST_INSERT_HEAD(&s->free_list, pdu, next);
}

static void coroutine_fn pdu_complete(V9fsPDU *pdu, ssize_t len)
{
    int8_t id = pdu->id + 1; /* Response */
    V9fsState *s = pdu->s;
    int ret;

    /*
     * The 9p spec requires that successfully cancelled pdus receive no reply.
     * Sending a reply would confuse clients because they would
     * assume that any EINTR is the actual result of the operation,
     * rather than a consequence of the cancellation. However, if
     * the operation completed (succesfully or with an error other
     * than caused be cancellation), we do send out that reply, both
     * for efficiency and to avoid confusing the rest of the state machine
     * that assumes passing a non-error here will mean a successful
     * transmission of the reply.
     */
    bool discard = pdu->cancelled && len == -EINTR;
    if (discard) {
        trace_v9fs_rcancel(pdu->tag, pdu->id);
        pdu->size = 0;
        goto out_notify;
    }

    if (len < 0) {
        int err = -len;
        len = 7;

        if (s->proto_version != V9FS_PROTO_2000L) {
            V9fsString str;

            str.data = strerror(err);
            str.size = strlen(str.data);

            ret = pdu_marshal(pdu, len, "s", &str);
            if (ret < 0) {
                goto out_notify;
            }
            len += ret;
            id = P9_RERROR;
        }

        ret = pdu_marshal(pdu, len, "d", err);
        if (ret < 0) {
            goto out_notify;
        }
        len += ret;

        if (s->proto_version == V9FS_PROTO_2000L) {
            id = P9_RLERROR;
        }
        trace_v9fs_rerror(pdu->tag, pdu->id, err); /* Trace ERROR */
    }

    /* fill out the header */
    if (pdu_marshal(pdu, 0, "dbw", (int32_t)len, id, pdu->tag) < 0) {
        goto out_notify;
    }

    /* keep these in sync */
    pdu->size = len;
    pdu->id = id;

out_notify:
    pdu->s->transport->push_and_notify(pdu);

    /* Now wakeup anybody waiting in flush for this request */
    if (!qemu_co_queue_next(&pdu->complete)) {
        pdu_free(pdu);
    }
}

static mode_t v9mode_to_mode(uint32_t mode, V9fsString *extension)
{
    mode_t ret;

    ret = mode & 0777;
    if (mode & P9_STAT_MODE_DIR) {
        ret |= S_IFDIR;
    }

    if (mode & P9_STAT_MODE_SYMLINK) {
        ret |= S_IFLNK;
    }
    if (mode & P9_STAT_MODE_SOCKET) {
        ret |= S_IFSOCK;
    }
    if (mode & P9_STAT_MODE_NAMED_PIPE) {
        ret |= S_IFIFO;
    }
    if (mode & P9_STAT_MODE_DEVICE) {
        if (extension->size && extension->data[0] == 'c') {
            ret |= S_IFCHR;
        } else {
            ret |= S_IFBLK;
        }
    }

    if (!(ret & ~0777)) {
        ret |= S_IFREG;
    }

    if (mode & P9_STAT_MODE_SETUID) {
        ret |= S_ISUID;
    }
    if (mode & P9_STAT_MODE_SETGID) {
        ret |= S_ISGID;
    }
    if (mode & P9_STAT_MODE_SETVTX) {
        ret |= S_ISVTX;
    }

    return ret;
}

static int donttouch_stat(V9fsStat *stat)
{
    if (stat->type == -1 &&
        stat->dev == -1 &&
        stat->qid.type == 0xff &&
        stat->qid.version == (uint32_t) -1 &&
        stat->qid.path == (uint64_t) -1 &&
        stat->mode == -1 &&
        stat->atime == -1 &&
        stat->mtime == -1 &&
        stat->length == -1 &&
        !stat->name.size &&
        !stat->uid.size &&
        !stat->gid.size &&
        !stat->muid.size &&
        stat->n_uid == -1 &&
        stat->n_gid == -1 &&
        stat->n_muid == -1) {
        return 1;
    }

    return 0;
}

static void v9fs_stat_init(V9fsStat *stat)
{
    v9fs_string_init(&stat->name);
    v9fs_string_init(&stat->uid);
    v9fs_string_init(&stat->gid);
    v9fs_string_init(&stat->muid);
    v9fs_string_init(&stat->extension);
}

static void v9fs_stat_free(V9fsStat *stat)
{
    v9fs_string_free(&stat->name);
    v9fs_string_free(&stat->uid);
    v9fs_string_free(&stat->gid);
    v9fs_string_free(&stat->muid);
    v9fs_string_free(&stat->extension);
}

static uint32_t stat_to_v9mode(const struct stat *stbuf)
{
    uint32_t mode;

    mode = stbuf->st_mode & 0777;
    if (S_ISDIR(stbuf->st_mode)) {
        mode |= P9_STAT_MODE_DIR;
    }

    if (S_ISLNK(stbuf->st_mode)) {
        mode |= P9_STAT_MODE_SYMLINK;
    }

    if (S_ISSOCK(stbuf->st_mode)) {
        mode |= P9_STAT_MODE_SOCKET;
    }

    if (S_ISFIFO(stbuf->st_mode)) {
        mode |= P9_STAT_MODE_NAMED_PIPE;
    }

    if (S_ISBLK(stbuf->st_mode) || S_ISCHR(stbuf->st_mode)) {
        mode |= P9_STAT_MODE_DEVICE;
    }

    if (stbuf->st_mode & S_ISUID) {
        mode |= P9_STAT_MODE_SETUID;
    }

    if (stbuf->st_mode & S_ISGID) {
        mode |= P9_STAT_MODE_SETGID;
    }

    if (stbuf->st_mode & S_ISVTX) {
        mode |= P9_STAT_MODE_SETVTX;
    }

    return mode;
}

static int coroutine_fn stat_to_v9stat(V9fsPDU *pdu, V9fsPath *path,
                                       const char *basename,
                                       const struct stat *stbuf,
                                       V9fsStat *v9stat)
{
    int err;

    memset(v9stat, 0, sizeof(*v9stat));

    err = stat_to_qid(pdu, stbuf, &v9stat->qid);
    if (err < 0) {
        return err;
    }
    v9stat->mode = stat_to_v9mode(stbuf);
    v9stat->atime = stbuf->st_atime;
    v9stat->mtime = stbuf->st_mtime;
    v9stat->length = stbuf->st_size;

    v9fs_string_free(&v9stat->uid);
    v9fs_string_free(&v9stat->gid);
    v9fs_string_free(&v9stat->muid);

    v9stat->n_uid = stbuf->st_uid;
    v9stat->n_gid = stbuf->st_gid;
    v9stat->n_muid = 0;

    v9fs_string_free(&v9stat->extension);

    if (v9stat->mode & P9_STAT_MODE_SYMLINK) {
        err = v9fs_co_readlink(pdu, path, &v9stat->extension);
        if (err < 0) {
            return err;
        }
    } else if (v9stat->mode & P9_STAT_MODE_DEVICE) {
        v9fs_string_sprintf(&v9stat->extension, "%c %u %u",
                S_ISCHR(stbuf->st_mode) ? 'c' : 'b',
                major(stbuf->st_rdev), minor(stbuf->st_rdev));
    } else if (S_ISDIR(stbuf->st_mode) || S_ISREG(stbuf->st_mode)) {
        v9fs_string_sprintf(&v9stat->extension, "%s %lu",
                "HARDLINKCOUNT", (unsigned long)stbuf->st_nlink);
    }

    v9fs_string_sprintf(&v9stat->name, "%s", basename);

    v9stat->size = 61 +
        v9fs_string_size(&v9stat->name) +
        v9fs_string_size(&v9stat->uid) +
        v9fs_string_size(&v9stat->gid) +
        v9fs_string_size(&v9stat->muid) +
        v9fs_string_size(&v9stat->extension);
    return 0;
}

#define P9_STATS_MODE          0x00000001ULL
#define P9_STATS_NLINK         0x00000002ULL
#define P9_STATS_UID           0x00000004ULL
#define P9_STATS_GID           0x00000008ULL
#define P9_STATS_RDEV          0x00000010ULL
#define P9_STATS_ATIME         0x00000020ULL
#define P9_STATS_MTIME         0x00000040ULL
#define P9_STATS_CTIME         0x00000080ULL
#define P9_STATS_INO           0x00000100ULL
#define P9_STATS_SIZE          0x00000200ULL
#define P9_STATS_BLOCKS        0x00000400ULL

#define P9_STATS_BTIME         0x00000800ULL
#define P9_STATS_GEN           0x00001000ULL
#define P9_STATS_DATA_VERSION  0x00002000ULL

#define P9_STATS_BASIC         0x000007ffULL /* Mask for fields up to BLOCKS */
#define P9_STATS_ALL           0x00003fffULL /* Mask for All fields above */


/**
 * blksize_to_iounit() - Block size exposed to 9p client.
 * Return: block size
 *
 * @pdu: 9p client request
 * @blksize: host filesystem's block size
 *
 * Convert host filesystem's block size into an appropriate block size for
 * 9p client (guest OS side). The value returned suggests an "optimum" block
 * size for 9p I/O, i.e. to maximize performance.
 */
static int32_t blksize_to_iounit(const V9fsPDU *pdu, int32_t blksize)
{
    int32_t iounit = 0;
    V9fsState *s = pdu->s;

    /*
     * iounit should be multiples of blksize (host filesystem block size)
     * as well as less than (client msize - P9_IOHDRSZ)
     */
    if (blksize) {
        iounit = QEMU_ALIGN_DOWN(s->msize - P9_IOHDRSZ, blksize);
    }
    if (!iounit) {
        iounit = s->msize - P9_IOHDRSZ;
    }
    return iounit;
}

static int32_t stat_to_iounit(const V9fsPDU *pdu, const struct stat *stbuf)
{
    return blksize_to_iounit(pdu, stbuf->st_blksize);
}

static int stat_to_v9stat_dotl(V9fsPDU *pdu, const struct stat *stbuf,
                                V9fsStatDotl *v9lstat)
{
    memset(v9lstat, 0, sizeof(*v9lstat));

    v9lstat->st_mode = stbuf->st_mode;
    v9lstat->st_nlink = stbuf->st_nlink;
    v9lstat->st_uid = stbuf->st_uid;
    v9lstat->st_gid = stbuf->st_gid;
    v9lstat->st_rdev = stbuf->st_rdev;
    v9lstat->st_size = stbuf->st_size;
    v9lstat->st_blksize = stat_to_iounit(pdu, stbuf);
    v9lstat->st_blocks = stbuf->st_blocks;
    v9lstat->st_atime_sec = stbuf->st_atime;
    v9lstat->st_mtime_sec = stbuf->st_mtime;
    v9lstat->st_ctime_sec = stbuf->st_ctime;
#ifdef CONFIG_DARWIN
    v9lstat->st_atime_nsec = stbuf->st_atimespec.tv_nsec;
    v9lstat->st_mtime_nsec = stbuf->st_mtimespec.tv_nsec;
    v9lstat->st_ctime_nsec = stbuf->st_ctimespec.tv_nsec;
#else
    v9lstat->st_atime_nsec = stbuf->st_atim.tv_nsec;
    v9lstat->st_mtime_nsec = stbuf->st_mtim.tv_nsec;
    v9lstat->st_ctime_nsec = stbuf->st_ctim.tv_nsec;
#endif
    /* Currently we only support BASIC fields in stat */
    v9lstat->st_result_mask = P9_STATS_BASIC;

    return stat_to_qid(pdu, stbuf, &v9lstat->qid);
}

static void print_sg(struct iovec *sg, int cnt)
{
    int i;

    printf("sg[%d]: {", cnt);
    for (i = 0; i < cnt; i++) {
        if (i) {
            printf(", ");
        }
        printf("(%p, %zd)", sg[i].iov_base, sg[i].iov_len);
    }
    printf("}\n");
}

/* Will call this only for path name based fid */
static void v9fs_fix_path(V9fsPath *dst, V9fsPath *src, int len)
{
    V9fsPath str;
    v9fs_path_init(&str);
    v9fs_path_copy(&str, dst);
    v9fs_path_sprintf(dst, "%s%s", src->data, str.data + len);
    v9fs_path_free(&str);
}

static inline bool is_ro_export(FsContext *ctx)
{
    return ctx->export_flags & V9FS_RDONLY;
}

static void coroutine_fn v9fs_version(void *opaque)
{
    ssize_t err;
    V9fsPDU *pdu = opaque;
    V9fsState *s = pdu->s;
    V9fsString version;
    size_t offset = 7;

    v9fs_string_init(&version);
    err = pdu_unmarshal(pdu, offset, "ds", &s->msize, &version);
    if (err < 0) {
        goto out;
    }
    trace_v9fs_version(pdu->tag, pdu->id, s->msize, version.data);

    virtfs_reset(pdu);

    if (!strcmp(version.data, "9P2000.u")) {
        s->proto_version = V9FS_PROTO_2000U;
    } else if (!strcmp(version.data, "9P2000.L")) {
        s->proto_version = V9FS_PROTO_2000L;
    } else {
        v9fs_string_sprintf(&version, "unknown");
        /* skip min. msize check, reporting invalid version has priority */
        goto marshal;
    }

    if (s->msize < P9_MIN_MSIZE) {
        err = -EMSGSIZE;
        error_report(
            "9pfs: Client requested msize < minimum msize ("
            stringify(P9_MIN_MSIZE) ") supported by this server."
        );
        goto out;
    }

    /* 8192 is the default msize of Linux clients */
    if (s->msize <= 8192 && !(s->ctx.export_flags & V9FS_NO_PERF_WARN)) {
        warn_report_once(
            "9p: degraded performance: a reasonable high msize should be "
            "chosen on client/guest side (chosen msize is <= 8192). See "
            "https://wiki.qemu.org/Documentation/9psetup#msize for details."
        );
    }

marshal:
    err = pdu_marshal(pdu, offset, "ds", s->msize, &version);
    if (err < 0) {
        goto out;
    }
    err += offset;
    trace_v9fs_version_return(pdu->tag, pdu->id, s->msize, version.data);
out:
    pdu_complete(pdu, err);
    v9fs_string_free(&version);
}

static void coroutine_fn v9fs_attach(void *opaque)
{
    V9fsPDU *pdu = opaque;
    V9fsState *s = pdu->s;
    int32_t fid, afid, n_uname;
    V9fsString uname, aname;
    V9fsFidState *fidp;
    size_t offset = 7;
    V9fsQID qid;
    ssize_t err;
    struct stat stbuf;

    v9fs_string_init(&uname);
    v9fs_string_init(&aname);
    err = pdu_unmarshal(pdu, offset, "ddssd", &fid,
                        &afid, &uname, &aname, &n_uname);
    if (err < 0) {
        goto out_nofid;
    }
    trace_v9fs_attach(pdu->tag, pdu->id, fid, afid, uname.data, aname.data);

    fidp = alloc_fid(s, fid);
    if (fidp == NULL) {
        err = -EINVAL;
        goto out_nofid;
    }
    fidp->uid = n_uname;
    err = v9fs_co_name_to_path(pdu, NULL, "/", &fidp->path);
    if (err < 0) {
        err = -EINVAL;
        clunk_fid(s, fid);
        goto out;
    }
    err = v9fs_co_lstat(pdu, &fidp->path, &stbuf);
    if (err < 0) {
        err = -EINVAL;
        clunk_fid(s, fid);
        goto out;
    }
    err = stat_to_qid(pdu, &stbuf, &qid);
    if (err < 0) {
        err = -EINVAL;
        clunk_fid(s, fid);
        goto out;
    }

    /*
     * disable migration if we haven't done already.
     * attach could get called multiple times for the same export.
     */
    if (!s->migration_blocker) {
        error_setg(&s->migration_blocker,
                   "Migration is disabled when VirtFS export path '%s' is mounted in the guest using mount_tag '%s'",
                   s->ctx.fs_root ? s->ctx.fs_root : "NULL", s->tag);
        err = migrate_add_blocker(s->migration_blocker, NULL);
        if (err < 0) {
            error_free(s->migration_blocker);
            s->migration_blocker = NULL;
            clunk_fid(s, fid);
            goto out;
        }
        s->root_fid = fid;
    }

    err = pdu_marshal(pdu, offset, "Q", &qid);
    if (err < 0) {
        clunk_fid(s, fid);
        goto out;
    }
    err += offset;

    memcpy(&s->root_st, &stbuf, sizeof(stbuf));
    trace_v9fs_attach_return(pdu->tag, pdu->id,
                             qid.type, qid.version, qid.path);
out:
    put_fid(pdu, fidp);
out_nofid:
    pdu_complete(pdu, err);
    v9fs_string_free(&uname);
    v9fs_string_free(&aname);
}

static void coroutine_fn v9fs_stat(void *opaque)
{
    int32_t fid;
    V9fsStat v9stat;
    ssize_t err = 0;
    size_t offset = 7;
    struct stat stbuf;
    V9fsFidState *fidp;
    V9fsPDU *pdu = opaque;
    char *basename;

    err = pdu_unmarshal(pdu, offset, "d", &fid);
    if (err < 0) {
        goto out_nofid;
    }
    trace_v9fs_stat(pdu->tag, pdu->id, fid);

    fidp = get_fid(pdu, fid);
    if (fidp == NULL) {
        err = -ENOENT;
        goto out_nofid;
    }
    err = v9fs_co_lstat(pdu, &fidp->path, &stbuf);
    if (err < 0) {
        goto out;
    }
    basename = g_path_get_basename(fidp->path.data);
    err = stat_to_v9stat(pdu, &fidp->path, basename, &stbuf, &v9stat);
    g_free(basename);
    if (err < 0) {
        goto out;
    }
    err = pdu_marshal(pdu, offset, "wS", 0, &v9stat);
    if (err < 0) {
        v9fs_stat_free(&v9stat);
        goto out;
    }
    trace_v9fs_stat_return(pdu->tag, pdu->id, v9stat.mode,
                           v9stat.atime, v9stat.mtime, v9stat.length);
    err += offset;
    v9fs_stat_free(&v9stat);
out:
    put_fid(pdu, fidp);
out_nofid:
    pdu_complete(pdu, err);
}

static void coroutine_fn v9fs_getattr(void *opaque)
{
    int32_t fid;
    size_t offset = 7;
    ssize_t retval = 0;
    struct stat stbuf;
    V9fsFidState *fidp;
    uint64_t request_mask;
    V9fsStatDotl v9stat_dotl;
    V9fsPDU *pdu = opaque;

    retval = pdu_unmarshal(pdu, offset, "dq", &fid, &request_mask);
    if (retval < 0) {
        goto out_nofid;
    }
    trace_v9fs_getattr(pdu->tag, pdu->id, fid, request_mask);

    fidp = get_fid(pdu, fid);
    if (fidp == NULL) {
        retval = -ENOENT;
        goto out_nofid;
    }
    /*
     * Currently we only support BASIC fields in stat, so there is no
     * need to look at request_mask.
     */
    retval = v9fs_co_lstat(pdu, &fidp->path, &stbuf);
    if (retval < 0) {
        goto out;
    }
    retval = stat_to_v9stat_dotl(pdu, &stbuf, &v9stat_dotl);
    if (retval < 0) {
        goto out;
    }

    /*  fill st_gen if requested and supported by underlying fs */
    if (request_mask & P9_STATS_GEN) {
        retval = v9fs_co_st_gen(pdu, &fidp->path, stbuf.st_mode, &v9stat_dotl);
        switch (retval) {
        case 0:
            /* we have valid st_gen: update result mask */
            v9stat_dotl.st_result_mask |= P9_STATS_GEN;
            break;
        case -EINTR:
            /* request cancelled, e.g. by Tflush */
            goto out;
        default:
            /* failed to get st_gen: not fatal, ignore */
            break;
        }
    }
    retval = pdu_marshal(pdu, offset, "A", &v9stat_dotl);
    if (retval < 0) {
        goto out;
    }
    retval += offset;
    trace_v9fs_getattr_return(pdu->tag, pdu->id, v9stat_dotl.st_result_mask,
                              v9stat_dotl.st_mode, v9stat_dotl.st_uid,
                              v9stat_dotl.st_gid);
out:
    put_fid(pdu, fidp);
out_nofid:
    pdu_complete(pdu, retval);
}

/* Attribute flags */
#define P9_ATTR_MODE       (1 << 0)
#define P9_ATTR_UID        (1 << 1)
#define P9_ATTR_GID        (1 << 2)
#define P9_ATTR_SIZE       (1 << 3)
#define P9_ATTR_ATIME      (1 << 4)
#define P9_ATTR_MTIME      (1 << 5)
#define P9_ATTR_CTIME      (1 << 6)
#define P9_ATTR_ATIME_SET  (1 << 7)
#define P9_ATTR_MTIME_SET  (1 << 8)

#define P9_ATTR_MASK    127

static void coroutine_fn v9fs_setattr(void *opaque)
{
    int err = 0;
    int32_t fid;
    V9fsFidState *fidp;
    size_t offset = 7;
    V9fsIattr v9iattr;
    V9fsPDU *pdu = opaque;

    err = pdu_unmarshal(pdu, offset, "dI", &fid, &v9iattr);
    if (err < 0) {
        goto out_nofid;
    }

    trace_v9fs_setattr(pdu->tag, pdu->id, fid,
                       v9iattr.valid, v9iattr.mode, v9iattr.uid, v9iattr.gid,
                       v9iattr.size, v9iattr.atime_sec, v9iattr.mtime_sec);

    fidp = get_fid(pdu, fid);
    if (fidp == NULL) {
        err = -EINVAL;
        goto out_nofid;
    }
    if (v9iattr.valid & P9_ATTR_MODE) {
        err = v9fs_co_chmod(pdu, &fidp->path, v9iattr.mode);
        if (err < 0) {
            goto out;
        }
    }
    if (v9iattr.valid & (P9_ATTR_ATIME | P9_ATTR_MTIME)) {
        struct timespec times[2];
        if (v9iattr.valid & P9_ATTR_ATIME) {
            if (v9iattr.valid & P9_ATTR_ATIME_SET) {
                times[0].tv_sec = v9iattr.atime_sec;
                times[0].tv_nsec = v9iattr.atime_nsec;
            } else {
                times[0].tv_nsec = UTIME_NOW;
            }
        } else {
            times[0].tv_nsec = UTIME_OMIT;
        }
        if (v9iattr.valid & P9_ATTR_MTIME) {
            if (v9iattr.valid & P9_ATTR_MTIME_SET) {
                times[1].tv_sec = v9iattr.mtime_sec;
                times[1].tv_nsec = v9iattr.mtime_nsec;
            } else {
                times[1].tv_nsec = UTIME_NOW;
            }
        } else {
            times[1].tv_nsec = UTIME_OMIT;
        }
        err = v9fs_co_utimensat(pdu, &fidp->path, times);
        if (err < 0) {
            goto out;
        }
    }
    /*
     * If the only valid entry in iattr is ctime we can call
     * chown(-1,-1) to update the ctime of the file
     */
    if ((v9iattr.valid & (P9_ATTR_UID | P9_ATTR_GID)) ||
        ((v9iattr.valid & P9_ATTR_CTIME)
         && !((v9iattr.valid & P9_ATTR_MASK) & ~P9_ATTR_CTIME))) {
        if (!(v9iattr.valid & P9_ATTR_UID)) {
            v9iattr.uid = -1;
        }
        if (!(v9iattr.valid & P9_ATTR_GID)) {
            v9iattr.gid = -1;
        }
        err = v9fs_co_chown(pdu, &fidp->path, v9iattr.uid,
                            v9iattr.gid);
        if (err < 0) {
            goto out;
        }
    }
    if (v9iattr.valid & (P9_ATTR_SIZE)) {
        err = v9fs_co_truncate(pdu, &fidp->path, v9iattr.size);
        if (err < 0) {
            goto out;
        }
    }
    err = offset;
    trace_v9fs_setattr_return(pdu->tag, pdu->id);
out:
    put_fid(pdu, fidp);
out_nofid:
    pdu_complete(pdu, err);
}

static int v9fs_walk_marshal(V9fsPDU *pdu, uint16_t nwnames, V9fsQID *qids)
{
    int i;
    ssize_t err;
    size_t offset = 7;

    err = pdu_marshal(pdu, offset, "w", nwnames);
    if (err < 0) {
        return err;
    }
    offset += err;
    for (i = 0; i < nwnames; i++) {
        err = pdu_marshal(pdu, offset, "Q", &qids[i]);
        if (err < 0) {
            return err;
        }
        offset += err;
    }
    return offset;
}

static bool name_is_illegal(const char *name)
{
    return !*name || strchr(name, '/') != NULL;
}

static bool same_stat_id(const struct stat *a, const struct stat *b)
{
    return a->st_dev == b->st_dev && a->st_ino == b->st_ino;
}

static void coroutine_fn v9fs_walk(void *opaque)
{
    int name_idx;
    g_autofree V9fsQID *qids = NULL;
    int i, err = 0;
    V9fsPath dpath, path;
    P9ARRAY_REF(V9fsPath) pathes = NULL;
    uint16_t nwnames;
    struct stat stbuf, fidst;
    g_autofree struct stat *stbufs = NULL;
    size_t offset = 7;
    int32_t fid, newfid;
    P9ARRAY_REF(V9fsString) wnames = NULL;
    V9fsFidState *fidp;
    V9fsFidState *newfidp = NULL;
    V9fsPDU *pdu = opaque;
    V9fsState *s = pdu->s;
    V9fsQID qid;

    err = pdu_unmarshal(pdu, offset, "ddw", &fid, &newfid, &nwnames);
    if (err < 0) {
        pdu_complete(pdu, err);
        return ;
    }
    offset += err;

    trace_v9fs_walk(pdu->tag, pdu->id, fid, newfid, nwnames);

    if (nwnames > P9_MAXWELEM) {
        err = -EINVAL;
        goto out_nofid;
    }
    if (nwnames) {
        P9ARRAY_NEW(V9fsString, wnames, nwnames);
        qids   = g_new0(V9fsQID, nwnames);
        stbufs = g_new0(struct stat, nwnames);
        P9ARRAY_NEW(V9fsPath, pathes, nwnames);
        for (i = 0; i < nwnames; i++) {
            err = pdu_unmarshal(pdu, offset, "s", &wnames[i]);
            if (err < 0) {
                goto out_nofid;
            }
            if (name_is_illegal(wnames[i].data)) {
                err = -ENOENT;
                goto out_nofid;
            }
            offset += err;
        }
    }
    fidp = get_fid(pdu, fid);
    if (fidp == NULL) {
        err = -ENOENT;
        goto out_nofid;
    }

    v9fs_path_init(&dpath);
    v9fs_path_init(&path);
    /*
     * Both dpath and path initially point to fidp.
     * Needed to handle request with nwnames == 0
     */
    v9fs_path_copy(&dpath, &fidp->path);
    v9fs_path_copy(&path, &fidp->path);

    /*
     * To keep latency (i.e. overall execution time for processing this
     * Twalk client request) as small as possible, run all the required fs
     * driver code altogether inside the following block.
     */
    v9fs_co_run_in_worker({
        if (v9fs_request_cancelled(pdu)) {
            err = -EINTR;
            break;
        }
        err = s->ops->lstat(&s->ctx, &dpath, &fidst);
        if (err < 0) {
            err = -errno;
            break;
        }
        stbuf = fidst;
        for (name_idx = 0; name_idx < nwnames; name_idx++) {
            if (v9fs_request_cancelled(pdu)) {
                err = -EINTR;
                break;
            }
            if (!same_stat_id(&pdu->s->root_st, &stbuf) ||
                strcmp("..", wnames[name_idx].data))
            {
                err = s->ops->name_to_path(&s->ctx, &dpath,
                                           wnames[name_idx].data,
                                           &pathes[name_idx]);
                if (err < 0) {
                    err = -errno;
                    break;
                }
                if (v9fs_request_cancelled(pdu)) {
                    err = -EINTR;
                    break;
                }
                err = s->ops->lstat(&s->ctx, &pathes[name_idx], &stbuf);
                if (err < 0) {
                    err = -errno;
                    break;
                }
                stbufs[name_idx] = stbuf;
                v9fs_path_copy(&dpath, &pathes[name_idx]);
            }
        }
    });
    /*
     * Handle all the rest of this Twalk request on main thread ...
     */
    if (err < 0) {
        goto out;
    }

    err = stat_to_qid(pdu, &fidst, &qid);
    if (err < 0) {
        goto out;
    }
    stbuf = fidst;

    /* reset dpath and path */
    v9fs_path_copy(&dpath, &fidp->path);
    v9fs_path_copy(&path, &fidp->path);

    for (name_idx = 0; name_idx < nwnames; name_idx++) {
        if (!same_stat_id(&pdu->s->root_st, &stbuf) ||
            strcmp("..", wnames[name_idx].data))
        {
            stbuf = stbufs[name_idx];
            err = stat_to_qid(pdu, &stbuf, &qid);
            if (err < 0) {
                goto out;
            }
            v9fs_path_copy(&path, &pathes[name_idx]);
            v9fs_path_copy(&dpath, &path);
        }
        memcpy(&qids[name_idx], &qid, sizeof(qid));
    }
    if (fid == newfid) {
        if (fidp->fid_type != P9_FID_NONE) {
            err = -EINVAL;
            goto out;
        }
        v9fs_path_write_lock(s);
        v9fs_path_copy(&fidp->path, &path);
        v9fs_path_unlock(s);
    } else {
        newfidp = alloc_fid(s, newfid);
        if (newfidp == NULL) {
            err = -EINVAL;
            goto out;
        }
        newfidp->uid = fidp->uid;
        v9fs_path_copy(&newfidp->path, &path);
    }
    err = v9fs_walk_marshal(pdu, nwnames, qids);
    trace_v9fs_walk_return(pdu->tag, pdu->id, nwnames, qids);
out:
    put_fid(pdu, fidp);
    if (newfidp) {
        put_fid(pdu, newfidp);
    }
    v9fs_path_free(&dpath);
    v9fs_path_free(&path);
out_nofid:
    pdu_complete(pdu, err);
}

static int32_t coroutine_fn get_iounit(V9fsPDU *pdu, V9fsPath *path)
{
    struct statfs stbuf;
    int err = v9fs_co_statfs(pdu, path, &stbuf);

    return blksize_to_iounit(pdu, (err >= 0) ? stbuf.f_bsize : 0);
}

static void coroutine_fn v9fs_open(void *opaque)
{
    int flags;
    int32_t fid;
    int32_t mode;
    V9fsQID qid;
    int iounit = 0;
    ssize_t err = 0;
    size_t offset = 7;
    struct stat stbuf;
    V9fsFidState *fidp;
    V9fsPDU *pdu = opaque;
    V9fsState *s = pdu->s;

    if (s->proto_version == V9FS_PROTO_2000L) {
        err = pdu_unmarshal(pdu, offset, "dd", &fid, &mode);
    } else {
        uint8_t modebyte;
        err = pdu_unmarshal(pdu, offset, "db", &fid, &modebyte);
        mode = modebyte;
    }
    if (err < 0) {
        goto out_nofid;
    }
    trace_v9fs_open(pdu->tag, pdu->id, fid, mode);

    fidp = get_fid(pdu, fid);
    if (fidp == NULL) {
        err = -ENOENT;
        goto out_nofid;
    }
    if (fidp->fid_type != P9_FID_NONE) {
        err = -EINVAL;
        goto out;
    }

    err = v9fs_co_lstat(pdu, &fidp->path, &stbuf);
    if (err < 0) {
        goto out;
    }
    err = stat_to_qid(pdu, &stbuf, &qid);
    if (err < 0) {
        goto out;
    }
    if (S_ISDIR(stbuf.st_mode)) {
        err = v9fs_co_opendir(pdu, fidp);
        if (err < 0) {
            goto out;
        }
        fidp->fid_type = P9_FID_DIR;
        err = pdu_marshal(pdu, offset, "Qd", &qid, 0);
        if (err < 0) {
            goto out;
        }
        err += offset;
    } else {
        if (s->proto_version == V9FS_PROTO_2000L) {
            flags = get_dotl_openflags(s, mode);

        } else {
            flags = omode_to_uflags(mode);
        }
        if (is_ro_export(&s->ctx)) {
            if (mode & O_WRONLY || mode & O_RDWR ||
                mode & O_APPEND || mode & O_TRUNC) {
                err = -EROFS;
                goto out;
            }
        }
        err = v9fs_co_open(pdu, fidp, flags);
        if (err < 0) {
            goto out;
        }
        fidp->fid_type = P9_FID_FILE;
        fidp->open_flags = flags;
        if (flags & O_EXCL) {
            /*
             * We let the host file system do O_EXCL check
             * We should not reclaim such fd
             */
            fidp->flags |= FID_NON_RECLAIMABLE;
        }
        iounit = get_iounit(pdu, &fidp->path);
        err = pdu_marshal(pdu, offset, "Qd", &qid, iounit);
        if (err < 0) {
            goto out;
        }
        err += offset;
    }
    trace_v9fs_open_return(pdu->tag, pdu->id,
                           qid.type, qid.version, qid.path, iounit);
out:
    put_fid(pdu, fidp);
out_nofid:
    pdu_complete(pdu, err);
}

static void coroutine_fn v9fs_lcreate(void *opaque)
{
    int32_t dfid, flags, mode;
    gid_t gid;
    ssize_t err = 0;
    ssize_t offset = 7;
    V9fsString name;
    V9fsFidState *fidp;
    struct stat stbuf;
    V9fsQID qid;
    int32_t iounit;
    V9fsPDU *pdu = opaque;

    v9fs_string_init(&name);
    err = pdu_unmarshal(pdu, offset, "dsddd", &dfid,
                        &name, &flags, &mode, &gid);
    if (err < 0) {
        goto out_nofid;
    }
    trace_v9fs_lcreate(pdu->tag, pdu->id, dfid, flags, mode, gid);

    if (name_is_illegal(name.data)) {
        err = -ENOENT;
        goto out_nofid;
    }

    if (!strcmp(".", name.data) || !strcmp("..", name.data)) {
        err = -EEXIST;
        goto out_nofid;
    }

    fidp = get_fid(pdu, dfid);
    if (fidp == NULL) {
        err = -ENOENT;
        goto out_nofid;
    }
    if (fidp->fid_type != P9_FID_NONE) {
        err = -EINVAL;
        goto out;
    }

    flags = get_dotl_openflags(pdu->s, flags);
    err = v9fs_co_open2(pdu, fidp, &name, gid,
                        flags | O_CREAT, mode, &stbuf);
    if (err < 0) {
        goto out;
    }
    fidp->fid_type = P9_FID_FILE;
    fidp->open_flags = flags;
    if (flags & O_EXCL) {
        /*
         * We let the host file system do O_EXCL check
         * We should not reclaim such fd
         */
        fidp->flags |= FID_NON_RECLAIMABLE;
    }
    iounit =  get_iounit(pdu, &fidp->path);
    err = stat_to_qid(pdu, &stbuf, &qid);
    if (err < 0) {
        goto out;
    }
    err = pdu_marshal(pdu, offset, "Qd", &qid, iounit);
    if (err < 0) {
        goto out;
    }
    err += offset;
    trace_v9fs_lcreate_return(pdu->tag, pdu->id,
                              qid.type, qid.version, qid.path, iounit);
out:
    put_fid(pdu, fidp);
out_nofid:
    pdu_complete(pdu, err);
    v9fs_string_free(&name);
}

static void coroutine_fn v9fs_fsync(void *opaque)
{
    int err;
    int32_t fid;
    int datasync;
    size_t offset = 7;
    V9fsFidState *fidp;
    V9fsPDU *pdu = opaque;

    err = pdu_unmarshal(pdu, offset, "dd", &fid, &datasync);
    if (err < 0) {
        goto out_nofid;
    }
    trace_v9fs_fsync(pdu->tag, pdu->id, fid, datasync);

    fidp = get_fid(pdu, fid);
    if (fidp == NULL) {
        err = -ENOENT;
        goto out_nofid;
    }
    err = v9fs_co_fsync(pdu, fidp, datasync);
    if (!err) {
        err = offset;
    }
    put_fid(pdu, fidp);
out_nofid:
    pdu_complete(pdu, err);
}

static void coroutine_fn v9fs_clunk(void *opaque)
{
    int err;
    int32_t fid;
    size_t offset = 7;
    V9fsFidState *fidp;
    V9fsPDU *pdu = opaque;
    V9fsState *s = pdu->s;

    err = pdu_unmarshal(pdu, offset, "d", &fid);
    if (err < 0) {
        goto out_nofid;
    }
    trace_v9fs_clunk(pdu->tag, pdu->id, fid);

    fidp = clunk_fid(s, fid);
    if (fidp == NULL) {
        err = -ENOENT;
        goto out_nofid;
    }
    /*
     * Bump the ref so that put_fid will
     * free the fid.
     */
    fidp->ref++;
    err = put_fid(pdu, fidp);
    if (!err) {
        err = offset;
    }
out_nofid:
    pdu_complete(pdu, err);
}

/*
 * Create a QEMUIOVector for a sub-region of PDU iovecs
 *
 * @qiov:       uninitialized QEMUIOVector
 * @skip:       number of bytes to skip from beginning of PDU
 * @size:       number of bytes to include
 * @is_write:   true - write, false - read
 *
 * The resulting QEMUIOVector has heap-allocated iovecs and must be cleaned up
 * with qemu_iovec_destroy().
 */
static void v9fs_init_qiov_from_pdu(QEMUIOVector *qiov, V9fsPDU *pdu,
                                    size_t skip, size_t size,
                                    bool is_write)
{
    QEMUIOVector elem;
    struct iovec *iov;
    unsigned int niov;

    if (is_write) {
        pdu->s->transport->init_out_iov_from_pdu(pdu, &iov, &niov, size + skip);
    } else {
        pdu->s->transport->init_in_iov_from_pdu(pdu, &iov, &niov, size + skip);
    }

    qemu_iovec_init_external(&elem, iov, niov);
    qemu_iovec_init(qiov, niov);
    qemu_iovec_concat(qiov, &elem, skip, size);
}

static int v9fs_xattr_read(V9fsState *s, V9fsPDU *pdu, V9fsFidState *fidp,
                           uint64_t off, uint32_t max_count)
{
    ssize_t err;
    size_t offset = 7;
    uint64_t read_count;
    QEMUIOVector qiov_full;

    if (fidp->fs.xattr.len < off) {
        read_count = 0;
    } else {
        read_count = fidp->fs.xattr.len - off;
    }
    if (read_count > max_count) {
        read_count = max_count;
    }
    err = pdu_marshal(pdu, offset, "d", read_count);
    if (err < 0) {
        return err;
    }
    offset += err;

    v9fs_init_qiov_from_pdu(&qiov_full, pdu, offset, read_count, false);
    err = v9fs_pack(qiov_full.iov, qiov_full.niov, 0,
                    ((char *)fidp->fs.xattr.value) + off,
                    read_count);
    qemu_iovec_destroy(&qiov_full);
    if (err < 0) {
        return err;
    }
    offset += err;
    return offset;
}

static int coroutine_fn v9fs_do_readdir_with_stat(V9fsPDU *pdu,
                                                  V9fsFidState *fidp,
                                                  uint32_t max_count)
{
    V9fsPath path;
    V9fsStat v9stat;
    int len, err = 0;
    int32_t count = 0;
    struct stat stbuf;
    off_t saved_dir_pos;
    struct dirent *dent;

    /* save the directory position */
    saved_dir_pos = v9fs_co_telldir(pdu, fidp);
    if (saved_dir_pos < 0) {
        return saved_dir_pos;
    }

    while (1) {
        v9fs_path_init(&path);

        v9fs_readdir_lock(&fidp->fs.dir);

        err = v9fs_co_readdir(pdu, fidp, &dent);
        if (err || !dent) {
            break;
        }
        err = v9fs_co_name_to_path(pdu, &fidp->path, dent->d_name, &path);
        if (err < 0) {
            break;
        }
        err = v9fs_co_lstat(pdu, &path, &stbuf);
        if (err < 0) {
            break;
        }
        err = stat_to_v9stat(pdu, &path, dent->d_name, &stbuf, &v9stat);
        if (err < 0) {
            break;
        }
        if ((count + v9stat.size + 2) > max_count) {
            v9fs_readdir_unlock(&fidp->fs.dir);

            /* Ran out of buffer. Set dir back to old position and return */
            v9fs_co_seekdir(pdu, fidp, saved_dir_pos);
            v9fs_stat_free(&v9stat);
            v9fs_path_free(&path);
            return count;
        }

        /* 11 = 7 + 4 (7 = start offset, 4 = space for storing count) */
        len = pdu_marshal(pdu, 11 + count, "S", &v9stat);

        v9fs_readdir_unlock(&fidp->fs.dir);

        if (len < 0) {
            v9fs_co_seekdir(pdu, fidp, saved_dir_pos);
            v9fs_stat_free(&v9stat);
            v9fs_path_free(&path);
            return len;
        }
        count += len;
        v9fs_stat_free(&v9stat);
        v9fs_path_free(&path);
        saved_dir_pos = qemu_dirent_off(dent);
    }

    v9fs_readdir_unlock(&fidp->fs.dir);

    v9fs_path_free(&path);
    if (err < 0) {
        return err;
    }
    return count;
}

static void coroutine_fn v9fs_read(void *opaque)
{
    int32_t fid;
    uint64_t off;
    ssize_t err = 0;
    int32_t count = 0;
    size_t offset = 7;
    uint32_t max_count;
    V9fsFidState *fidp;
    V9fsPDU *pdu = opaque;
    V9fsState *s = pdu->s;

    err = pdu_unmarshal(pdu, offset, "dqd", &fid, &off, &max_count);
    if (err < 0) {
        goto out_nofid;
    }
    trace_v9fs_read(pdu->tag, pdu->id, fid, off, max_count);

    fidp = get_fid(pdu, fid);
    if (fidp == NULL) {
        err = -EINVAL;
        goto out_nofid;
    }
    if (fidp->fid_type == P9_FID_DIR) {
        if (s->proto_version != V9FS_PROTO_2000U) {
            warn_report_once(
                "9p: bad client: T_read request on directory only expected "
                "with 9P2000.u protocol version"
            );
            err = -EOPNOTSUPP;
            goto out;
        }
        if (off == 0) {
            v9fs_co_rewinddir(pdu, fidp);
        }
        count = v9fs_do_readdir_with_stat(pdu, fidp, max_count);
        if (count < 0) {
            err = count;
            goto out;
        }
        err = pdu_marshal(pdu, offset, "d", count);
        if (err < 0) {
            goto out;
        }
        err += offset + count;
    } else if (fidp->fid_type == P9_FID_FILE) {
        QEMUIOVector qiov_full;
        QEMUIOVector qiov;
        int32_t len;

        v9fs_init_qiov_from_pdu(&qiov_full, pdu, offset + 4, max_count, false);
        qemu_iovec_init(&qiov, qiov_full.niov);
        do {
            qemu_iovec_reset(&qiov);
            qemu_iovec_concat(&qiov, &qiov_full, count, qiov_full.size - count);
            if (0) {
                print_sg(qiov.iov, qiov.niov);
            }
            /* Loop in case of EINTR */
            do {
                len = v9fs_co_preadv(pdu, fidp, qiov.iov, qiov.niov, off);
                if (len >= 0) {
                    off   += len;
                    count += len;
                }
            } while (len == -EINTR && !pdu->cancelled);
            if (len < 0) {
                /* IO error return the error */
                err = len;
                goto out_free_iovec;
            }
        } while (count < max_count && len > 0);
        err = pdu_marshal(pdu, offset, "d", count);
        if (err < 0) {
            goto out_free_iovec;
        }
        err += offset + count;
out_free_iovec:
        qemu_iovec_destroy(&qiov);
        qemu_iovec_destroy(&qiov_full);
    } else if (fidp->fid_type == P9_FID_XATTR) {
        err = v9fs_xattr_read(s, pdu, fidp, off, max_count);
    } else {
        err = -EINVAL;
    }
    trace_v9fs_read_return(pdu->tag, pdu->id, count, err);
out:
    put_fid(pdu, fidp);
out_nofid:
    pdu_complete(pdu, err);
}

/**
 * v9fs_readdir_response_size() - Returns size required in Rreaddir response
 * for the passed dirent @name.
 *
 * @name: directory entry's name (i.e. file name, directory name)
 * Return: required size in bytes
 */
size_t v9fs_readdir_response_size(V9fsString *name)
{
    /*
     * Size of each dirent on the wire: size of qid (13) + size of offset (8)
     * size of type (1) + size of name.size (2) + strlen(name.data)
     */
    return 24 + v9fs_string_size(name);
}

static void v9fs_free_dirents(struct V9fsDirEnt *e)
{
    struct V9fsDirEnt *next = NULL;

    for (; e; e = next) {
        next = e->next;
        g_free(e->dent);
        g_free(e->st);
        g_free(e);
    }
}

static int coroutine_fn v9fs_do_readdir(V9fsPDU *pdu, V9fsFidState *fidp,
                                        off_t offset, int32_t max_count)
{
    size_t size;
    V9fsQID qid;
    V9fsString name;
    int len, err = 0;
    int32_t count = 0;
    off_t off;
    struct dirent *dent;
    struct stat *st;
    struct V9fsDirEnt *entries = NULL;

    /*
     * inode remapping requires the device id, which in turn might be
     * different for different directory entries, so if inode remapping is
     * enabled we have to make a full stat for each directory entry
     */
    const bool dostat = pdu->s->ctx.export_flags & V9FS_REMAP_INODES;

    /*
     * Fetch all required directory entries altogether on a background IO
     * thread from fs driver. We don't want to do that for each entry
     * individually, because hopping between threads (this main IO thread
     * and background IO driver thread) would sum up to huge latencies.
     */
    count = v9fs_co_readdir_many(pdu, fidp, &entries, offset, max_count,
                                 dostat);
    if (count < 0) {
        err = count;
        count = 0;
        goto out;
    }
    count = 0;

    for (struct V9fsDirEnt *e = entries; e; e = e->next) {
        dent = e->dent;

        if (pdu->s->ctx.export_flags & V9FS_REMAP_INODES) {
            st = e->st;
            /* e->st should never be NULL, but just to be sure */
            if (!st) {
                err = -1;
                break;
            }

            /* remap inode */
            err = stat_to_qid(pdu, st, &qid);
            if (err < 0) {
                break;
            }
        } else {
            /*
             * Fill up just the path field of qid because the client uses
             * only that. To fill the entire qid structure we will have
             * to stat each dirent found, which is expensive. For the
             * latter reason we don't call stat_to_qid() here. Only drawback
             * is that no multi-device export detection of stat_to_qid()
             * would be done and provided as error to the user here. But
             * user would get that error anyway when accessing those
             * files/dirs through other ways.
             */
            size = MIN(sizeof(dent->d_ino), sizeof(qid.path));
            memcpy(&qid.path, &dent->d_ino, size);
            /* Fill the other fields with dummy values */
            qid.type = 0;
            qid.version = 0;
        }

        off = qemu_dirent_off(dent);
        v9fs_string_init(&name);
        v9fs_string_sprintf(&name, "%s", dent->d_name);

        /* 11 = 7 + 4 (7 = start offset, 4 = space for storing count) */
        len = pdu_marshal(pdu, 11 + count, "Qqbs",
                          &qid, off,
                          dent->d_type, &name);

        v9fs_string_free(&name);

        if (len < 0) {
            err = len;
            break;
        }

        count += len;
    }

out:
    v9fs_free_dirents(entries);
    if (err < 0) {
        return err;
    }
    return count;
}

static void coroutine_fn v9fs_readdir(void *opaque)
{
    int32_t fid;
    V9fsFidState *fidp;
    ssize_t retval = 0;
    size_t offset = 7;
    uint64_t initial_offset;
    int32_t count;
    uint32_t max_count;
    V9fsPDU *pdu = opaque;
    V9fsState *s = pdu->s;

    retval = pdu_unmarshal(pdu, offset, "dqd", &fid,
                           &initial_offset, &max_count);
    if (retval < 0) {
        goto out_nofid;
    }
    trace_v9fs_readdir(pdu->tag, pdu->id, fid, initial_offset, max_count);

    /* Enough space for a R_readdir header: size[4] Rreaddir tag[2] count[4] */
    if (max_count > s->msize - 11) {
        max_count = s->msize - 11;
        warn_report_once(
            "9p: bad client: T_readdir with count > msize - 11"
        );
    }

    fidp = get_fid(pdu, fid);
    if (fidp == NULL) {
        retval = -EINVAL;
        goto out_nofid;
    }
    if (!fidp->fs.dir.stream) {
        retval = -EINVAL;
        goto out;
    }
    if (s->proto_version != V9FS_PROTO_2000L) {
        warn_report_once(
            "9p: bad client: T_readdir request only expected with 9P2000.L "
            "protocol version"
        );
        retval = -EOPNOTSUPP;
        goto out;
    }
    count = v9fs_do_readdir(pdu, fidp, (off_t) initial_offset, max_count);
    if (count < 0) {
        retval = count;
        goto out;
    }
    retval = pdu_marshal(pdu, offset, "d", count);
    if (retval < 0) {
        goto out;
    }
    retval += count + offset;
    trace_v9fs_readdir_return(pdu->tag, pdu->id, count, retval);
out:
    put_fid(pdu, fidp);
out_nofid:
    pdu_complete(pdu, retval);
}

static int v9fs_xattr_write(V9fsState *s, V9fsPDU *pdu, V9fsFidState *fidp,
                            uint64_t off, uint32_t count,
                            struct iovec *sg, int cnt)
{
    int i, to_copy;
    ssize_t err = 0;
    uint64_t write_count;
    size_t offset = 7;


    if (fidp->fs.xattr.len < off) {
        return -ENOSPC;
    }
    write_count = fidp->fs.xattr.len - off;
    if (write_count > count) {
        write_count = count;
    }
    err = pdu_marshal(pdu, offset, "d", write_count);
    if (err < 0) {
        return err;
    }
    err += offset;
    fidp->fs.xattr.copied_len += write_count;
    /*
     * Now copy the content from sg list
     */
    for (i = 0; i < cnt; i++) {
        if (write_count > sg[i].iov_len) {
            to_copy = sg[i].iov_len;
        } else {
            to_copy = write_count;
        }
        memcpy((char *)fidp->fs.xattr.value + off, sg[i].iov_base, to_copy);
        /* updating vs->off since we are not using below */
        off += to_copy;
        write_count -= to_copy;
    }

    return err;
}

static void coroutine_fn v9fs_write(void *opaque)
{
    ssize_t err;
    int32_t fid;
    uint64_t off;
    uint32_t count;
    int32_t len = 0;
    int32_t total = 0;
    size_t offset = 7;
    V9fsFidState *fidp;
    V9fsPDU *pdu = opaque;
    V9fsState *s = pdu->s;
    QEMUIOVector qiov_full;
    QEMUIOVector qiov;

    err = pdu_unmarshal(pdu, offset, "dqd", &fid, &off, &count);
    if (err < 0) {
        pdu_complete(pdu, err);
        return;
    }
    offset += err;
    v9fs_init_qiov_from_pdu(&qiov_full, pdu, offset, count, true);
    trace_v9fs_write(pdu->tag, pdu->id, fid, off, count, qiov_full.niov);

    fidp = get_fid(pdu, fid);
    if (fidp == NULL) {
        err = -EINVAL;
        goto out_nofid;
    }
    if (fidp->fid_type == P9_FID_FILE) {
        if (fidp->fs.fd == -1) {
            err = -EINVAL;
            goto out;
        }
    } else if (fidp->fid_type == P9_FID_XATTR) {
        /*
         * setxattr operation
         */
        err = v9fs_xattr_write(s, pdu, fidp, off, count,
                               qiov_full.iov, qiov_full.niov);
        goto out;
    } else {
        err = -EINVAL;
        goto out;
    }
    qemu_iovec_init(&qiov, qiov_full.niov);
    do {
        qemu_iovec_reset(&qiov);
        qemu_iovec_concat(&qiov, &qiov_full, total, qiov_full.size - total);
        if (0) {
            print_sg(qiov.iov, qiov.niov);
        }
        /* Loop in case of EINTR */
        do {
            len = v9fs_co_pwritev(pdu, fidp, qiov.iov, qiov.niov, off);
            if (len >= 0) {
                off   += len;
                total += len;
            }
        } while (len == -EINTR && !pdu->cancelled);
        if (len < 0) {
            /* IO error return the error */
            err = len;
            goto out_qiov;
        }
    } while (total < count && len > 0);

    offset = 7;
    err = pdu_marshal(pdu, offset, "d", total);
    if (err < 0) {
        goto out_qiov;
    }
    err += offset;
    trace_v9fs_write_return(pdu->tag, pdu->id, total, err);
out_qiov:
    qemu_iovec_destroy(&qiov);
out:
    put_fid(pdu, fidp);
out_nofid:
    qemu_iovec_destroy(&qiov_full);
    pdu_complete(pdu, err);
}

static void coroutine_fn v9fs_create(void *opaque)
{
    int32_t fid;
    int err = 0;
    size_t offset = 7;
    V9fsFidState *fidp;
    V9fsQID qid;
    int32_t perm;
    int8_t mode;
    V9fsPath path;
    struct stat stbuf;
    V9fsString name;
    V9fsString extension;
    int iounit;
    V9fsPDU *pdu = opaque;
    V9fsState *s = pdu->s;

    v9fs_path_init(&path);
    v9fs_string_init(&name);
    v9fs_string_init(&extension);
    err = pdu_unmarshal(pdu, offset, "dsdbs", &fid, &name,
                        &perm, &mode, &extension);
    if (err < 0) {
        goto out_nofid;
    }
    trace_v9fs_create(pdu->tag, pdu->id, fid, name.data, perm, mode);

    if (name_is_illegal(name.data)) {
        err = -ENOENT;
        goto out_nofid;
    }

    if (!strcmp(".", name.data) || !strcmp("..", name.data)) {
        err = -EEXIST;
        goto out_nofid;
    }

    fidp = get_fid(pdu, fid);
    if (fidp == NULL) {
        err = -EINVAL;
        goto out_nofid;
    }
    if (fidp->fid_type != P9_FID_NONE) {
        err = -EINVAL;
        goto out;
    }
    if (perm & P9_STAT_MODE_DIR) {
        err = v9fs_co_mkdir(pdu, fidp, &name, perm & 0777,
                            fidp->uid, -1, &stbuf);
        if (err < 0) {
            goto out;
        }
        err = v9fs_co_name_to_path(pdu, &fidp->path, name.data, &path);
        if (err < 0) {
            goto out;
        }
        v9fs_path_write_lock(s);
        v9fs_path_copy(&fidp->path, &path);
        v9fs_path_unlock(s);
        err = v9fs_co_opendir(pdu, fidp);
        if (err < 0) {
            goto out;
        }
        fidp->fid_type = P9_FID_DIR;
    } else if (perm & P9_STAT_MODE_SYMLINK) {
        err = v9fs_co_symlink(pdu, fidp, &name,
                              extension.data, -1 , &stbuf);
        if (err < 0) {
            goto out;
        }
        err = v9fs_co_name_to_path(pdu, &fidp->path, name.data, &path);
        if (err < 0) {
            goto out;
        }
        v9fs_path_write_lock(s);
        v9fs_path_copy(&fidp->path, &path);
        v9fs_path_unlock(s);
    } else if (perm & P9_STAT_MODE_LINK) {
        int32_t ofid = atoi(extension.data);
        V9fsFidState *ofidp = get_fid(pdu, ofid);
        if (ofidp == NULL) {
            err = -EINVAL;
            goto out;
        }
        err = v9fs_co_link(pdu, ofidp, fidp, &name);
        put_fid(pdu, ofidp);
        if (err < 0) {
            goto out;
        }
        err = v9fs_co_name_to_path(pdu, &fidp->path, name.data, &path);
        if (err < 0) {
            fidp->fid_type = P9_FID_NONE;
            goto out;
        }
        v9fs_path_write_lock(s);
        v9fs_path_copy(&fidp->path, &path);
        v9fs_path_unlock(s);
        err = v9fs_co_lstat(pdu, &fidp->path, &stbuf);
        if (err < 0) {
            fidp->fid_type = P9_FID_NONE;
            goto out;
        }
    } else if (perm & P9_STAT_MODE_DEVICE) {
        char ctype;
        uint32_t major, minor;
        mode_t nmode = 0;

        if (sscanf(extension.data, "%c %u %u", &ctype, &major, &minor) != 3) {
            err = -errno;
            goto out;
        }

        switch (ctype) {
        case 'c':
            nmode = S_IFCHR;
            break;
        case 'b':
            nmode = S_IFBLK;
            break;
        default:
            err = -EIO;
            goto out;
        }

        nmode |= perm & 0777;
        err = v9fs_co_mknod(pdu, fidp, &name, fidp->uid, -1,
                            makedev(major, minor), nmode, &stbuf);
        if (err < 0) {
            goto out;
        }
        err = v9fs_co_name_to_path(pdu, &fidp->path, name.data, &path);
        if (err < 0) {
            goto out;
        }
        v9fs_path_write_lock(s);
        v9fs_path_copy(&fidp->path, &path);
        v9fs_path_unlock(s);
    } else if (perm & P9_STAT_MODE_NAMED_PIPE) {
        err = v9fs_co_mknod(pdu, fidp, &name, fidp->uid, -1,
                            0, S_IFIFO | (perm & 0777), &stbuf);
        if (err < 0) {
            goto out;
        }
        err = v9fs_co_name_to_path(pdu, &fidp->path, name.data, &path);
        if (err < 0) {
            goto out;
        }
        v9fs_path_write_lock(s);
        v9fs_path_copy(&fidp->path, &path);
        v9fs_path_unlock(s);
    } else if (perm & P9_STAT_MODE_SOCKET) {
        err = v9fs_co_mknod(pdu, fidp, &name, fidp->uid, -1,
                            0, S_IFSOCK | (perm & 0777), &stbuf);
        if (err < 0) {
            goto out;
        }
        err = v9fs_co_name_to_path(pdu, &fidp->path, name.data, &path);
        if (err < 0) {
            goto out;
        }
        v9fs_path_write_lock(s);
        v9fs_path_copy(&fidp->path, &path);
        v9fs_path_unlock(s);
    } else {
        err = v9fs_co_open2(pdu, fidp, &name, -1,
                            omode_to_uflags(mode) | O_CREAT, perm, &stbuf);
        if (err < 0) {
            goto out;
        }
        fidp->fid_type = P9_FID_FILE;
        fidp->open_flags = omode_to_uflags(mode);
        if (fidp->open_flags & O_EXCL) {
            /*
             * We let the host file system do O_EXCL check
             * We should not reclaim such fd
             */
            fidp->flags |= FID_NON_RECLAIMABLE;
        }
    }
    iounit = get_iounit(pdu, &fidp->path);
    err = stat_to_qid(pdu, &stbuf, &qid);
    if (err < 0) {
        goto out;
    }
    err = pdu_marshal(pdu, offset, "Qd", &qid, iounit);
    if (err < 0) {
        goto out;
    }
    err += offset;
    trace_v9fs_create_return(pdu->tag, pdu->id,
                             qid.type, qid.version, qid.path, iounit);
out:
    put_fid(pdu, fidp);
out_nofid:
   pdu_complete(pdu, err);
   v9fs_string_free(&name);
   v9fs_string_free(&extension);
   v9fs_path_free(&path);
}

static void coroutine_fn v9fs_symlink(void *opaque)
{
    V9fsPDU *pdu = opaque;
    V9fsString name;
    V9fsString symname;
    V9fsFidState *dfidp;
    V9fsQID qid;
    struct stat stbuf;
    int32_t dfid;
    int err = 0;
    gid_t gid;
    size_t offset = 7;

    v9fs_string_init(&name);
    v9fs_string_init(&symname);
    err = pdu_unmarshal(pdu, offset, "dssd", &dfid, &name, &symname, &gid);
    if (err < 0) {
        goto out_nofid;
    }
    trace_v9fs_symlink(pdu->tag, pdu->id, dfid, name.data, symname.data, gid);

    if (name_is_illegal(name.data)) {
        err = -ENOENT;
        goto out_nofid;
    }

    if (!strcmp(".", name.data) || !strcmp("..", name.data)) {
        err = -EEXIST;
        goto out_nofid;
    }

    dfidp = get_fid(pdu, dfid);
    if (dfidp == NULL) {
        err = -EINVAL;
        goto out_nofid;
    }
    err = v9fs_co_symlink(pdu, dfidp, &name, symname.data, gid, &stbuf);
    if (err < 0) {
        goto out;
    }
    err = stat_to_qid(pdu, &stbuf, &qid);
    if (err < 0) {
        goto out;
    }
    err =  pdu_marshal(pdu, offset, "Q", &qid);
    if (err < 0) {
        goto out;
    }
    err += offset;
    trace_v9fs_symlink_return(pdu->tag, pdu->id,
                              qid.type, qid.version, qid.path);
out:
    put_fid(pdu, dfidp);
out_nofid:
    pdu_complete(pdu, err);
    v9fs_string_free(&name);
    v9fs_string_free(&symname);
}

static void coroutine_fn v9fs_flush(void *opaque)
{
    ssize_t err;
    int16_t tag;
    size_t offset = 7;
    V9fsPDU *cancel_pdu = NULL;
    V9fsPDU *pdu = opaque;
    V9fsState *s = pdu->s;

    err = pdu_unmarshal(pdu, offset, "w", &tag);
    if (err < 0) {
        pdu_complete(pdu, err);
        return;
    }
    trace_v9fs_flush(pdu->tag, pdu->id, tag);

    if (pdu->tag == tag) {
        warn_report("the guest sent a self-referencing 9P flush request");
    } else {
        QLIST_FOREACH(cancel_pdu, &s->active_list, next) {
            if (cancel_pdu->tag == tag) {
                break;
            }
        }
    }
    if (cancel_pdu) {