LU-3321 clio: optimize read ahead code

[fs/lustre-release.git] / lustre / lclient / lcommon_cl.c

/*
 * GPL HEADER START
 *
 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License version 2 only,
 * as published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful, but
 * WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * General Public License version 2 for more details (a copy is included
 * in the LICENSE file that accompanied this code).
 *
 * You should have received a copy of the GNU General Public License
 * version 2 along with this program; If not, see
 * http://www.sun.com/software/products/lustre/docs/GPLv2.pdf
 *
 * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
 * CA 95054 USA or visit www.sun.com if you need additional information or
 * have any questions.
 *
 * GPL HEADER END
 */
/*
 * Copyright (c) 2008, 2010, Oracle and/or its affiliates. All rights reserved.
 * Use is subject to license terms.
 *
 * Copyright (c) 2011, 2013, Intel Corporation.
 */
/*
 * This file is part of Lustre, http://www.lustre.org/
 * Lustre is a trademark of Sun Microsystems, Inc.
 *
 * cl code shared between vvp and liblustre (and other Lustre clients in the
 * future).
 *
 *   Author: Nikita Danilov <nikita.danilov@sun.com>
 */

#define DEBUG_SUBSYSTEM S_LLITE

#ifdef __KERNEL__
# include <libcfs/libcfs.h>
# include <linux/fs.h>
# include <linux/sched.h>
# include <linux/mm.h>
# include <linux/quotaops.h>
# include <linux/highmem.h>
# include <linux/pagemap.h>
# include <linux/rbtree.h>
#else /* __KERNEL__ */
#include <stdlib.h>
#include <string.h>
#include <assert.h>
#include <time.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/queue.h>
#include <fcntl.h>
# include <liblustre.h>
#endif

#include <obd.h>
#include <obd_support.h>
#include <lustre_fid.h>
#include <lustre_lite.h>
#include <lustre_dlm.h>
#include <lustre_ver.h>
#include <lustre_mdc.h>
#include <cl_object.h>

#include <lclient.h>

#ifdef __KERNEL__
#include "../llite/llite_internal.h"
#else
#include "../liblustre/llite_lib.h"
#endif

const struct cl_req_operations ccc_req_ops;

/*
 * ccc_ prefix stands for "Common Client Code".
 */

static struct kmem_cache *ccc_lock_kmem;
static struct kmem_cache *ccc_object_kmem;
static struct kmem_cache *ccc_thread_kmem;
static struct kmem_cache *ccc_session_kmem;
static struct kmem_cache *ccc_req_kmem;

static struct lu_kmem_descr ccc_caches[] = {
        {
                .ckd_cache = &ccc_lock_kmem,
                .ckd_name  = "ccc_lock_kmem",
                .ckd_size  = sizeof (struct ccc_lock)
        },
        {
                .ckd_cache = &ccc_object_kmem,
                .ckd_name  = "ccc_object_kmem",
                .ckd_size  = sizeof (struct ccc_object)
        },
        {
                .ckd_cache = &ccc_thread_kmem,
                .ckd_name  = "ccc_thread_kmem",
                .ckd_size  = sizeof (struct ccc_thread_info),
        },
        {
                .ckd_cache = &ccc_session_kmem,
                .ckd_name  = "ccc_session_kmem",
                .ckd_size  = sizeof (struct ccc_session)
        },
        {
                .ckd_cache = &ccc_req_kmem,
                .ckd_name  = "ccc_req_kmem",
                .ckd_size  = sizeof (struct ccc_req)
        },
        {
                .ckd_cache = NULL
        }
};

/*****************************************************************************
 *
 * Vvp device and device type functions.
 *
 */

void *ccc_key_init(const struct lu_context *ctx,
                          struct lu_context_key *key)
{
        struct ccc_thread_info *info;

        OBD_SLAB_ALLOC_PTR_GFP(info, ccc_thread_kmem, __GFP_IO);
        if (info == NULL)
                info = ERR_PTR(-ENOMEM);
        return info;
}

void ccc_key_fini(const struct lu_context *ctx,
                         struct lu_context_key *key, void *data)
{
        struct ccc_thread_info *info = data;
        OBD_SLAB_FREE_PTR(info, ccc_thread_kmem);
}

void *ccc_session_key_init(const struct lu_context *ctx,
                                  struct lu_context_key *key)
{
        struct ccc_session *session;

        OBD_SLAB_ALLOC_PTR_GFP(session, ccc_session_kmem, __GFP_IO);
        if (session == NULL)
                session = ERR_PTR(-ENOMEM);
        return session;
}

void ccc_session_key_fini(const struct lu_context *ctx,
                                 struct lu_context_key *key, void *data)
{
        struct ccc_session *session = data;
        OBD_SLAB_FREE_PTR(session, ccc_session_kmem);
}

struct lu_context_key ccc_key = {
        .lct_tags = LCT_CL_THREAD,
        .lct_init = ccc_key_init,
        .lct_fini = ccc_key_fini
};

struct lu_context_key ccc_session_key = {
        .lct_tags = LCT_SESSION,
        .lct_init = ccc_session_key_init,
        .lct_fini = ccc_session_key_fini
};


/* type constructor/destructor: ccc_type_{init,fini,start,stop}(). */
// LU_TYPE_INIT_FINI(ccc, &ccc_key, &ccc_session_key);

int ccc_device_init(const struct lu_env *env, struct lu_device *d,
                           const char *name, struct lu_device *next)
{
        struct ccc_device  *vdv;
        int rc;
        ENTRY;

        vdv = lu2ccc_dev(d);
        vdv->cdv_next = lu2cl_dev(next);

        LASSERT(d->ld_site != NULL && next->ld_type != NULL);
        next->ld_site = d->ld_site;
        rc = next->ld_type->ldt_ops->ldto_device_init(
                        env, next, next->ld_type->ldt_name, NULL);
        if (rc == 0) {
                lu_device_get(next);
                lu_ref_add(&next->ld_reference, "lu-stack", &lu_site_init);
        }
        RETURN(rc);
}

struct lu_device *ccc_device_fini(const struct lu_env *env,
                                         struct lu_device *d)
{
        return cl2lu_dev(lu2ccc_dev(d)->cdv_next);
}

struct lu_device *ccc_device_alloc(const struct lu_env *env,
                                   struct lu_device_type *t,
                                   struct lustre_cfg *cfg,
                                   const struct lu_device_operations *luops,
                                   const struct cl_device_operations *clops)
{
        struct ccc_device *vdv;
        struct lu_device  *lud;
        struct cl_site    *site;
        int rc;
        ENTRY;

        OBD_ALLOC_PTR(vdv);
        if (vdv == NULL)
                RETURN(ERR_PTR(-ENOMEM));

        lud = &vdv->cdv_cl.cd_lu_dev;
        cl_device_init(&vdv->cdv_cl, t);
        ccc2lu_dev(vdv)->ld_ops = luops;
        vdv->cdv_cl.cd_ops = clops;

        OBD_ALLOC_PTR(site);
        if (site != NULL) {
                rc = cl_site_init(site, &vdv->cdv_cl);
                if (rc == 0)
                        rc = lu_site_init_finish(&site->cs_lu);
                else {
                        LASSERT(lud->ld_site == NULL);
                        CERROR("Cannot init lu_site, rc %d.\n", rc);
                        OBD_FREE_PTR(site);
                }
        } else
                rc = -ENOMEM;
        if (rc != 0) {
                ccc_device_free(env, lud);
                lud = ERR_PTR(rc);
        }
        RETURN(lud);
}

struct lu_device *ccc_device_free(const struct lu_env *env,
                                         struct lu_device *d)
{
        struct ccc_device *vdv  = lu2ccc_dev(d);
        struct cl_site    *site = lu2cl_site(d->ld_site);
        struct lu_device  *next = cl2lu_dev(vdv->cdv_next);

        if (d->ld_site != NULL) {
                cl_site_fini(site);
                OBD_FREE_PTR(site);
        }
        cl_device_fini(lu2cl_dev(d));
        OBD_FREE_PTR(vdv);
        return next;
}

int ccc_req_init(const struct lu_env *env, struct cl_device *dev,
                        struct cl_req *req)
{
        struct ccc_req *vrq;
        int result;

        OBD_SLAB_ALLOC_PTR_GFP(vrq, ccc_req_kmem, __GFP_IO);
        if (vrq != NULL) {
                cl_req_slice_add(req, &vrq->crq_cl, dev, &ccc_req_ops);
                result = 0;
        } else
                result = -ENOMEM;
        return result;
}

/**
 * An `emergency' environment used by ccc_inode_fini() when cl_env_get()
 * fails. Access to this environment is serialized by ccc_inode_fini_guard
 * mutex.
 */
static struct lu_env *ccc_inode_fini_env = NULL;

/**
 * A mutex serializing calls to slp_inode_fini() under extreme memory
 * pressure, when environments cannot be allocated.
 */
static DEFINE_MUTEX(ccc_inode_fini_guard);
static int dummy_refcheck;

int ccc_global_init(struct lu_device_type *device_type)
{
        int result;

        result = lu_kmem_init(ccc_caches);
        if (result)
                return result;

        result = lu_device_type_init(device_type);
        if (result)
                goto out_kmem;

        ccc_inode_fini_env = cl_env_alloc(&dummy_refcheck,
                                          LCT_REMEMBER|LCT_NOREF);
        if (IS_ERR(ccc_inode_fini_env)) {
                result = PTR_ERR(ccc_inode_fini_env);
                goto out_device;
        }

        ccc_inode_fini_env->le_ctx.lc_cookie = 0x4;
        return 0;
out_device:
        lu_device_type_fini(device_type);
out_kmem:
        lu_kmem_fini(ccc_caches);
        return result;
}

void ccc_global_fini(struct lu_device_type *device_type)
{
        if (ccc_inode_fini_env != NULL) {
                cl_env_put(ccc_inode_fini_env, &dummy_refcheck);
                ccc_inode_fini_env = NULL;
        }
        lu_device_type_fini(device_type);
        lu_kmem_fini(ccc_caches);
}

/*****************************************************************************
 *
 * Object operations.
 *
 */

struct lu_object *ccc_object_alloc(const struct lu_env *env,
                                   const struct lu_object_header *unused,
                                   struct lu_device *dev,
                                   const struct cl_object_operations *clops,
                                   const struct lu_object_operations *luops)
{
        struct ccc_object *vob;
        struct lu_object  *obj;

        OBD_SLAB_ALLOC_PTR_GFP(vob, ccc_object_kmem, __GFP_IO);
        if (vob != NULL) {
                struct cl_object_header *hdr;

                obj = ccc2lu(vob);
                hdr = &vob->cob_header;
                cl_object_header_init(hdr);
                hdr->coh_page_bufsize = cfs_size_round(sizeof(struct cl_page));

                lu_object_init(obj, &hdr->coh_lu, dev);
                lu_object_add_top(&hdr->coh_lu, obj);

                vob->cob_cl.co_ops = clops;
                obj->lo_ops = luops;
        } else
                obj = NULL;
        return obj;
}

int ccc_object_init0(const struct lu_env *env,
                            struct ccc_object *vob,
                            const struct cl_object_conf *conf)
{
        vob->cob_inode = conf->coc_inode;
        vob->cob_transient_pages = 0;
        cl_object_page_init(&vob->cob_cl, sizeof(struct ccc_page));
        return 0;
}

int ccc_object_init(const struct lu_env *env, struct lu_object *obj,
                           const struct lu_object_conf *conf)
{
        struct ccc_device *dev = lu2ccc_dev(obj->lo_dev);
        struct ccc_object *vob = lu2ccc(obj);
        struct lu_object  *below;
        struct lu_device  *under;
        int result;

        under = &dev->cdv_next->cd_lu_dev;
        below = under->ld_ops->ldo_object_alloc(env, obj->lo_header, under);
        if (below != NULL) {
                const struct cl_object_conf *cconf;

                cconf = lu2cl_conf(conf);
                CFS_INIT_LIST_HEAD(&vob->cob_pending_list);
                lu_object_add(obj, below);
                result = ccc_object_init0(env, vob, cconf);
        } else
                result = -ENOMEM;
        return result;
}

void ccc_object_free(const struct lu_env *env, struct lu_object *obj)
{
        struct ccc_object *vob = lu2ccc(obj);

        lu_object_fini(obj);
        lu_object_header_fini(obj->lo_header);
        OBD_SLAB_FREE_PTR(vob, ccc_object_kmem);
}

int ccc_lock_init(const struct lu_env *env,
                  struct cl_object *obj, struct cl_lock *lock,
                  const struct cl_io *unused,
                  const struct cl_lock_operations *lkops)
{
        struct ccc_lock *clk;
        int result;

        CLOBINVRNT(env, obj, ccc_object_invariant(obj));

        OBD_SLAB_ALLOC_PTR_GFP(clk, ccc_lock_kmem, __GFP_IO);
        if (clk != NULL) {
                cl_lock_slice_add(lock, &clk->clk_cl, obj, lkops);
                result = 0;
        } else
                result = -ENOMEM;
        return result;
}

int ccc_attr_set(const struct lu_env *env, struct cl_object *obj,
                 const struct cl_attr *attr, unsigned valid)
{
        return 0;
}

int ccc_object_glimpse(const struct lu_env *env,
                       const struct cl_object *obj, struct ost_lvb *lvb)
{
        struct inode *inode = ccc_object_inode(obj);

        ENTRY;
        lvb->lvb_mtime = cl_inode_mtime(inode);
        lvb->lvb_atime = cl_inode_atime(inode);
        lvb->lvb_ctime = cl_inode_ctime(inode);
        /*
         * LU-417: Add dirty pages block count lest i_blocks reports 0, some
         * "cp" or "tar" on remote node may think it's a completely sparse file
         * and skip it.
         */
        if (lvb->lvb_size > 0 && lvb->lvb_blocks == 0)
                lvb->lvb_blocks = dirty_cnt(inode);
        RETURN(0);
}



int ccc_conf_set(const struct lu_env *env, struct cl_object *obj,
                        const struct cl_object_conf *conf)
{
        /* TODO: destroy all pages attached to this object. */
        return 0;
}

static void ccc_object_size_lock(struct cl_object *obj)
{
        struct inode *inode = ccc_object_inode(obj);

        cl_isize_lock(inode);
        cl_object_attr_lock(obj);
}

static void ccc_object_size_unlock(struct cl_object *obj)
{
        struct inode *inode = ccc_object_inode(obj);

        cl_object_attr_unlock(obj);
        cl_isize_unlock(inode);
}

/*****************************************************************************
 *
 * Page operations.
 *
 */

int ccc_fail(const struct lu_env *env, const struct cl_page_slice *slice)
{
        /*
         * Cached read?
         */
        LBUG();
        return 0;
}

void ccc_transient_page_verify(const struct cl_page *page)
{
}

int ccc_transient_page_own(const struct lu_env *env,
                                   const struct cl_page_slice *slice,
                                   struct cl_io *unused,
                                   int nonblock)
{
        ccc_transient_page_verify(slice->cpl_page);
        return 0;
}

void ccc_transient_page_assume(const struct lu_env *env,
                                      const struct cl_page_slice *slice,
                                      struct cl_io *unused)
{
        ccc_transient_page_verify(slice->cpl_page);
}

void ccc_transient_page_unassume(const struct lu_env *env,
                                        const struct cl_page_slice *slice,
                                        struct cl_io *unused)
{
        ccc_transient_page_verify(slice->cpl_page);
}

void ccc_transient_page_disown(const struct lu_env *env,
                                      const struct cl_page_slice *slice,
                                      struct cl_io *unused)
{
        ccc_transient_page_verify(slice->cpl_page);
}

void ccc_transient_page_discard(const struct lu_env *env,
                                       const struct cl_page_slice *slice,
                                       struct cl_io *unused)
{
        struct cl_page *page = slice->cpl_page;

        ccc_transient_page_verify(slice->cpl_page);

        /*
         * For transient pages, remove it from the radix tree.
         */
        cl_page_delete(env, page);
}

int ccc_transient_page_prep(const struct lu_env *env,
                                   const struct cl_page_slice *slice,
                                   struct cl_io *unused)
{
        ENTRY;
        /* transient page should always be sent. */
        RETURN(0);
}

/*****************************************************************************
 *
 * Lock operations.
 *
 */

void ccc_lock_delete(const struct lu_env *env,
                     const struct cl_lock_slice *slice)
{
        CLOBINVRNT(env, slice->cls_obj, ccc_object_invariant(slice->cls_obj));
}

void ccc_lock_fini(const struct lu_env *env, struct cl_lock_slice *slice)
{
        struct ccc_lock *clk = cl2ccc_lock(slice);
        OBD_SLAB_FREE_PTR(clk, ccc_lock_kmem);
}

int ccc_lock_enqueue(const struct lu_env *env,
                     const struct cl_lock_slice *slice,
                     struct cl_io *unused, __u32 enqflags)
{
        CLOBINVRNT(env, slice->cls_obj, ccc_object_invariant(slice->cls_obj));
        return 0;
}

int ccc_lock_unuse(const struct lu_env *env, const struct cl_lock_slice *slice)
{
        CLOBINVRNT(env, slice->cls_obj, ccc_object_invariant(slice->cls_obj));
        return 0;
}

int ccc_lock_wait(const struct lu_env *env, const struct cl_lock_slice *slice)
{
        CLOBINVRNT(env, slice->cls_obj, ccc_object_invariant(slice->cls_obj));
        return 0;
}

/**
 * Implementation of cl_lock_operations::clo_fits_into() methods for ccc
 * layer. This function is executed every time io finds an existing lock in
 * the lock cache while creating new lock. This function has to decide whether
 * cached lock "fits" into io.
 *
 * \param slice lock to be checked
 * \param io    IO that wants a lock.
 *
 * \see lov_lock_fits_into().
 */
int ccc_lock_fits_into(const struct lu_env *env,
                       const struct cl_lock_slice *slice,
                       const struct cl_lock_descr *need,
                       const struct cl_io *io)
{
        const struct cl_lock       *lock  = slice->cls_lock;
        const struct cl_lock_descr *descr = &lock->cll_descr;
        const struct ccc_io        *cio   = ccc_env_io(env);
        int                         result;

        ENTRY;
        /*
         * Work around DLM peculiarity: it assumes that glimpse
         * (LDLM_FL_HAS_INTENT) lock is always LCK_PR, and returns reads lock
         * when asked for LCK_PW lock with LDLM_FL_HAS_INTENT flag set. Make
         * sure that glimpse doesn't get CLM_WRITE top-lock, so that it
         * doesn't enqueue CLM_WRITE sub-locks.
         */
        if (cio->cui_glimpse)
                result = descr->cld_mode != CLM_WRITE;

        /*
         * Also, don't match incomplete write locks for read, otherwise read
         * would enqueue missing sub-locks in the write mode.
         */
        else if (need->cld_mode != descr->cld_mode)
                result = lock->cll_state >= CLS_ENQUEUED;
        else
                result = 1;
        RETURN(result);
}

/**
 * Implements cl_lock_operations::clo_state() method for ccc layer, invoked
 * whenever lock state changes. Transfers object attributes, that might be
 * updated as a result of lock acquiring into inode.
 */
void ccc_lock_state(const struct lu_env *env,
                    const struct cl_lock_slice *slice,
                    enum cl_lock_state state)
{
        struct cl_lock *lock = slice->cls_lock;
        ENTRY;

        /*
         * Refresh inode attributes when the lock is moving into CLS_HELD
         * state, and only when this is a result of real enqueue, rather than
         * of finding lock in the cache.
         */
        if (state == CLS_HELD && lock->cll_state < CLS_HELD) {
                struct cl_object *obj;
                struct inode     *inode;

                obj   = slice->cls_obj;
                inode = ccc_object_inode(obj);

                /* vmtruncate() sets the i_size
                 * under both a DLM lock and the
                 * ll_inode_size_lock().  If we don't get the
                 * ll_inode_size_lock() here we can match the DLM lock and
                 * reset i_size.  generic_file_write can then trust the
                 * stale i_size when doing appending writes and effectively
                 * cancel the result of the truncate.  Getting the
                 * ll_inode_size_lock() after the enqueue maintains the DLM
                 * -> ll_inode_size_lock() acquiring order. */
                if (lock->cll_descr.cld_start == 0 &&
                    lock->cll_descr.cld_end == CL_PAGE_EOF)
                        cl_merge_lvb(env, inode);
        }
        EXIT;
}

/*****************************************************************************
 *
 * io operations.
 *
 */

void ccc_io_fini(const struct lu_env *env, const struct cl_io_slice *ios)
{
        struct cl_io *io = ios->cis_io;

        CLOBINVRNT(env, io->ci_obj, ccc_object_invariant(io->ci_obj));
}

int ccc_io_one_lock_index(const struct lu_env *env, struct cl_io *io,
                          __u32 enqflags, enum cl_lock_mode mode,
                          pgoff_t start, pgoff_t end)
{
        struct ccc_io          *cio   = ccc_env_io(env);
        struct cl_lock_descr   *descr = &cio->cui_link.cill_descr;
        struct cl_object       *obj   = io->ci_obj;

        CLOBINVRNT(env, obj, ccc_object_invariant(obj));
        ENTRY;

        CDEBUG(D_VFSTRACE, "lock: %d [%lu, %lu]\n", mode, start, end);

        memset(&cio->cui_link, 0, sizeof cio->cui_link);

        if (cio->cui_fd && (cio->cui_fd->fd_flags & LL_FILE_GROUP_LOCKED)) {
                descr->cld_mode = CLM_GROUP;
                descr->cld_gid  = cio->cui_fd->fd_grouplock.cg_gid;
        } else {
                descr->cld_mode  = mode;
        }
        descr->cld_obj   = obj;
        descr->cld_start = start;
        descr->cld_end   = end;
        descr->cld_enq_flags = enqflags;

        cl_io_lock_add(env, io, &cio->cui_link);
        RETURN(0);
}

void ccc_io_update_iov(const struct lu_env *env,
                       struct ccc_io *cio, struct cl_io *io)
{
        int i;
        size_t size = io->u.ci_rw.crw_count;

        cio->cui_iov_olen = 0;
        if (!cl_is_normalio(env, io) || cio->cui_tot_nrsegs == 0)
                return;

        for (i = 0; i < cio->cui_tot_nrsegs; i++) {
                struct iovec *iv = &cio->cui_iov[i];

                if (iv->iov_len < size)
                        size -= iv->iov_len;
                else {
                        if (iv->iov_len > size) {
                                cio->cui_iov_olen = iv->iov_len;
                                iv->iov_len = size;
                        }
                        break;
                }
        }

        cio->cui_nrsegs = i + 1;
        LASSERTF(cio->cui_tot_nrsegs >= cio->cui_nrsegs,
                 "tot_nrsegs: %lu, nrsegs: %lu\n",
                 cio->cui_tot_nrsegs, cio->cui_nrsegs);
}

int ccc_io_one_lock(const struct lu_env *env, struct cl_io *io,
                    __u32 enqflags, enum cl_lock_mode mode,
                    loff_t start, loff_t end)
{
        struct cl_object *obj = io->ci_obj;
        return ccc_io_one_lock_index(env, io, enqflags, mode,
                                     cl_index(obj, start), cl_index(obj, end));
}

void ccc_io_end(const struct lu_env *env, const struct cl_io_slice *ios)
{
        CLOBINVRNT(env, ios->cis_io->ci_obj,
                   ccc_object_invariant(ios->cis_io->ci_obj));
}

void ccc_io_advance(const struct lu_env *env,
                    const struct cl_io_slice *ios,
                    size_t nob)
{
        struct ccc_io    *cio = cl2ccc_io(env, ios);
        struct cl_io     *io  = ios->cis_io;
        struct cl_object *obj = ios->cis_io->ci_obj;

        CLOBINVRNT(env, obj, ccc_object_invariant(obj));

        if (!cl_is_normalio(env, io))
                return;

        LASSERT(cio->cui_tot_nrsegs >= cio->cui_nrsegs);
        LASSERT(cio->cui_tot_count  >= nob);

        cio->cui_iov        += cio->cui_nrsegs;
        cio->cui_tot_nrsegs -= cio->cui_nrsegs;
        cio->cui_tot_count  -= nob;

        /* update the iov */
        if (cio->cui_iov_olen > 0) {
                struct iovec *iv;

                cio->cui_iov--;
                cio->cui_tot_nrsegs++;
                iv = &cio->cui_iov[0];
                if (io->ci_continue) {
                        iv->iov_base += iv->iov_len;
                        LASSERT(cio->cui_iov_olen > iv->iov_len);
                        iv->iov_len = cio->cui_iov_olen - iv->iov_len;
                } else {
                        /* restore the iov_len, in case of restart io. */
                        iv->iov_len = cio->cui_iov_olen;
                }
                cio->cui_iov_olen = 0;
        }
}

/**
 * Helper function that if necessary adjusts file size (inode->i_size), when
 * position at the offset \a pos is accessed. File size can be arbitrary stale
 * on a Lustre client, but client at least knows KMS. If accessed area is
 * inside [0, KMS], set file size to KMS, otherwise glimpse file size.
 *
 * Locking: cl_isize_lock is used to serialize changes to inode size and to
 * protect consistency between inode size and cl_object
 * attributes. cl_object_size_lock() protects consistency between cl_attr's of
 * top-object and sub-objects.
 */
int ccc_prep_size(const struct lu_env *env, struct cl_object *obj,
                  struct cl_io *io, loff_t start, size_t count, int *exceed)
{
        struct cl_attr *attr  = ccc_env_thread_attr(env);
        struct inode   *inode = ccc_object_inode(obj);
        loff_t          pos   = start + count - 1;
        loff_t kms;
        int result;

        /*
         * Consistency guarantees: following possibilities exist for the
         * relation between region being accessed and real file size at this
         * moment:
         *
         *  (A): the region is completely inside of the file;
         *
         *  (B-x): x bytes of region are inside of the file, the rest is
         *  outside;
         *
         *  (C): the region is completely outside of the file.
         *
         * This classification is stable under DLM lock already acquired by
         * the caller, because to change the class, other client has to take
         * DLM lock conflicting with our lock. Also, any updates to ->i_size
         * by other threads on this client are serialized by
         * ll_inode_size_lock(). This guarantees that short reads are handled
         * correctly in the face of concurrent writes and truncates.
         */
        ccc_object_size_lock(obj);
        result = cl_object_attr_get(env, obj, attr);
        if (result == 0) {
                kms = attr->cat_kms;
                if (pos > kms) {
                        /*
                         * A glimpse is necessary to determine whether we
                         * return a short read (B) or some zeroes at the end
                         * of the buffer (C)
                         */
                        ccc_object_size_unlock(obj);
                        result = cl_glimpse_lock(env, io, inode, obj, 0);
                        if (result == 0 && exceed != NULL) {
                                /* If objective page index exceed end-of-file
                                 * page index, return directly. Do not expect
                                 * kernel will check such case correctly.
                                 * linux-2.6.18-128.1.1 miss to do that.
                                 * --bug 17336 */
                                loff_t size = cl_isize_read(inode);
                                unsigned long cur_index = start >>
                                                          PAGE_CACHE_SHIFT;

                                if ((size == 0 && cur_index != 0) ||
                                    (((size - 1) >> PAGE_CACHE_SHIFT) <
                                     cur_index))
                                *exceed = 1;
                        }
                        return result;
                } else {
                        /*
                         * region is within kms and, hence, within real file
                         * size (A). We need to increase i_size to cover the
                         * read region so that generic_file_read() will do its
                         * job, but that doesn't mean the kms size is
                         * _correct_, it is only the _minimum_ size. If
                         * someone does a stat they will get the correct size
                         * which will always be >= the kms value here.
                         * b=11081
                         */
                        if (cl_isize_read(inode) < kms) {
                                cl_isize_write_nolock(inode, kms);
                                CDEBUG(D_VFSTRACE,
                                       DFID" updating i_size "LPU64"\n",
                                       PFID(lu_object_fid(&obj->co_lu)),
                                       (__u64)cl_isize_read(inode));

                        }
                }
        }
        ccc_object_size_unlock(obj);
        return result;
}

/*****************************************************************************
 *
 * Transfer operations.
 *
 */

void ccc_req_completion(const struct lu_env *env,
                        const struct cl_req_slice *slice, int ioret)
{
        struct ccc_req *vrq;

        if (ioret > 0)
                cl_stats_tally(slice->crs_dev, slice->crs_req->crq_type, ioret);

        vrq = cl2ccc_req(slice);
        OBD_SLAB_FREE_PTR(vrq, ccc_req_kmem);
}

/**
 * Implementation of struct cl_req_operations::cro_attr_set() for ccc
 * layer. ccc is responsible for
 *
 *    - o_[mac]time
 *
 *    - o_mode
 *
 *    - o_parent_seq
 *
 *    - o_[ug]id
 *
 *    - o_parent_oid
 *
 *    - o_parent_ver
 *
 *    - o_ioepoch,
 *
 *  and capability.
 */
void ccc_req_attr_set(const struct lu_env *env,
                      const struct cl_req_slice *slice,
                      const struct cl_object *obj,
                      struct cl_req_attr *attr, obd_valid flags)
{
        struct inode *inode;
        struct obdo  *oa;
        obd_flag      valid_flags;

        oa = attr->cra_oa;
        inode = ccc_object_inode(obj);
        valid_flags = OBD_MD_FLTYPE;

        if ((flags & OBD_MD_FLOSSCAPA) != 0) {
                LASSERT(attr->cra_capa == NULL);
                attr->cra_capa = cl_capa_lookup(inode,
                                                slice->crs_req->crq_type);
        }

        if (slice->crs_req->crq_type == CRT_WRITE) {
                if (flags & OBD_MD_FLEPOCH) {
                        oa->o_valid |= OBD_MD_FLEPOCH;
                        oa->o_ioepoch = cl_i2info(inode)->lli_ioepoch;
                        valid_flags |= OBD_MD_FLMTIME | OBD_MD_FLCTIME |
                                       OBD_MD_FLUID | OBD_MD_FLGID;
                }
        }
        obdo_from_inode(oa, inode, valid_flags & flags);
        obdo_set_parent_fid(oa, &cl_i2info(inode)->lli_fid);
#ifdef __KERNEL__
        memcpy(attr->cra_jobid, cl_i2info(inode)->lli_jobid,
               JOBSTATS_JOBID_SIZE);
#endif
}

const struct cl_req_operations ccc_req_ops = {
        .cro_attr_set   = ccc_req_attr_set,
        .cro_completion = ccc_req_completion
};

int cl_setattr_ost(struct inode *inode, const struct iattr *attr,
                   struct obd_capa *capa)
{
        struct lu_env *env;
        struct cl_io  *io;
        int            result;
        int            refcheck;

        ENTRY;

        env = cl_env_get(&refcheck);
        if (IS_ERR(env))
                RETURN(PTR_ERR(env));

        io = ccc_env_thread_io(env);
        io->ci_obj = cl_i2info(inode)->lli_clob;

        io->u.ci_setattr.sa_attr.lvb_atime = LTIME_S(attr->ia_atime);
        io->u.ci_setattr.sa_attr.lvb_mtime = LTIME_S(attr->ia_mtime);
        io->u.ci_setattr.sa_attr.lvb_ctime = LTIME_S(attr->ia_ctime);
        io->u.ci_setattr.sa_attr.lvb_size = attr->ia_size;
        io->u.ci_setattr.sa_valid = attr->ia_valid;
        io->u.ci_setattr.sa_capa = capa;

again:
        if (cl_io_init(env, io, CIT_SETATTR, io->ci_obj) == 0) {
                struct ccc_io *cio = ccc_env_io(env);

                if (attr->ia_valid & ATTR_FILE)
                        /* populate the file descriptor for ftruncate to honor
                         * group lock - see LU-787 */
                        cio->cui_fd = cl_iattr2fd(inode, attr);

                result = cl_io_loop(env, io);
        } else {
                result = io->ci_result;
        }
        cl_io_fini(env, io);
        if (unlikely(io->ci_need_restart))
                goto again;
        /* HSM import case: file is released, cannot be restored
         * no need to fail except if restore registration failed
         * with -ENODATA */
        if (result == -ENODATA && io->ci_restore_needed &&
            io->ci_result != -ENODATA)
                result = 0;
        cl_env_put(env, &refcheck);
        RETURN(result);
}

/*****************************************************************************
 *
 * Type conversions.
 *
 */

struct lu_device *ccc2lu_dev(struct ccc_device *vdv)
{
        return &vdv->cdv_cl.cd_lu_dev;
}

struct ccc_device *lu2ccc_dev(const struct lu_device *d)
{
        return container_of0(d, struct ccc_device, cdv_cl.cd_lu_dev);
}

struct ccc_device *cl2ccc_dev(const struct cl_device *d)
{
        return container_of0(d, struct ccc_device, cdv_cl);
}

struct lu_object *ccc2lu(struct ccc_object *vob)
{
        return &vob->cob_cl.co_lu;
}

struct ccc_object *lu2ccc(const struct lu_object *obj)
{
        return container_of0(obj, struct ccc_object, cob_cl.co_lu);
}

struct ccc_object *cl2ccc(const struct cl_object *obj)
{
        return container_of0(obj, struct ccc_object, cob_cl);
}

struct ccc_lock *cl2ccc_lock(const struct cl_lock_slice *slice)
{
        return container_of(slice, struct ccc_lock, clk_cl);
}

struct ccc_io *cl2ccc_io(const struct lu_env *env,
                         const struct cl_io_slice *slice)
{
        struct ccc_io *cio;

        cio = container_of(slice, struct ccc_io, cui_cl);
        LASSERT(cio == ccc_env_io(env));
        return cio;
}

struct ccc_req *cl2ccc_req(const struct cl_req_slice *slice)
{
        return container_of0(slice, struct ccc_req, crq_cl);
}

struct page *cl2vm_page(const struct cl_page_slice *slice)
{
        return cl2ccc_page(slice)->cpg_page;
}

/*****************************************************************************
 *
 * Accessors.
 *
 */
int ccc_object_invariant(const struct cl_object *obj)
{
        struct inode         *inode = ccc_object_inode(obj);
        struct cl_inode_info *lli   = cl_i2info(inode);

        return (S_ISREG(cl_inode_mode(inode)) ||
                /* i_mode of unlinked inode is zeroed. */
                cl_inode_mode(inode) == 0) && lli->lli_clob == obj;
}

struct inode *ccc_object_inode(const struct cl_object *obj)
{
        return cl2ccc(obj)->cob_inode;
}

/**
 * Returns a pointer to cl_page associated with \a vmpage, without acquiring
 * additional reference to the resulting page. This is an unsafe version of
 * cl_vmpage_page() that can only be used under vmpage lock.
 */
struct cl_page *ccc_vmpage_page_transient(struct page *vmpage)
{
        KLASSERT(PageLocked(vmpage));
        return (struct cl_page *)vmpage->private;
}

/**
 * Initialize or update CLIO structures for regular files when new
 * meta-data arrives from the server.
 *
 * \param inode regular file inode
 * \param md    new file metadata from MDS
 * - allocates cl_object if necessary,
 * - updated layout, if object was already here.
 */
int cl_file_inode_init(struct inode *inode, struct lustre_md *md)
{
        struct lu_env        *env;
        struct cl_inode_info *lli;
        struct cl_object     *clob;
        struct lu_site       *site;
        struct lu_fid        *fid;
        struct cl_object_conf conf = {
                .coc_inode = inode,
                .u = {
                        .coc_md    = md
                }
        };
        int result = 0;
        int refcheck;

        LASSERT(md->body->valid & OBD_MD_FLID);
        LASSERT(S_ISREG(cl_inode_mode(inode)));

        env = cl_env_get(&refcheck);
        if (IS_ERR(env))
                return PTR_ERR(env);

        site = cl_i2sbi(inode)->ll_site;
        lli  = cl_i2info(inode);
        fid  = &lli->lli_fid;
        LASSERT(fid_is_sane(fid));

        if (lli->lli_clob == NULL) {
                /* clob is slave of inode, empty lli_clob means for new inode,
                 * there is no clob in cache with the given fid, so it is
                 * unnecessary to perform lookup-alloc-lookup-insert, just
                 * alloc and insert directly. */
#ifdef __KERNEL__
                LASSERT(inode->i_state & I_NEW);
#endif
                conf.coc_lu.loc_flags = LOC_F_NEW;
                clob = cl_object_find(env, lu2cl_dev(site->ls_top_dev),
                                      fid, &conf);
                if (!IS_ERR(clob)) {
                        /*
                         * No locking is necessary, as new inode is
                         * locked by I_NEW bit.
                         */
                        lli->lli_clob = clob;
                        lli->lli_has_smd = lsm_has_objects(md->lsm);
                        lu_object_ref_add(&clob->co_lu, "inode", inode);
                } else
                        result = PTR_ERR(clob);
        } else {
                result = cl_conf_set(env, lli->lli_clob, &conf);
        }

        cl_env_put(env, &refcheck);

        if (result != 0)
                CERROR("Failure to initialize cl object "DFID": %d\n",
                       PFID(fid), result);
        return result;
}

/**
 * Wait for others drop their references of the object at first, then we drop
 * the last one, which will lead to the object be destroyed immediately.
 * Must be called after cl_object_kill() against this object.
 *
 * The reason we want to do this is: destroying top object will wait for sub
 * objects being destroyed first, so we can't let bottom layer (e.g. from ASTs)
 * to initiate top object destroying which may deadlock. See bz22520.
 */
static void cl_object_put_last(struct lu_env *env, struct cl_object *obj)
{
        struct lu_object_header *header = obj->co_lu.lo_header;
        wait_queue_t           waiter;

        if (unlikely(atomic_read(&header->loh_ref) != 1)) {
                struct lu_site *site = obj->co_lu.lo_dev->ld_site;
                struct lu_site_bkt_data *bkt;

                bkt = lu_site_bkt_from_fid(site, &header->loh_fid);

                init_waitqueue_entry_current(&waiter);
                add_wait_queue(&bkt->lsb_marche_funebre, &waiter);

                while (1) {
                        set_current_state(TASK_UNINTERRUPTIBLE);
                        if (atomic_read(&header->loh_ref) == 1)
                                break;
                        waitq_wait(&waiter, TASK_UNINTERRUPTIBLE);
                }

                set_current_state(TASK_RUNNING);
                remove_wait_queue(&bkt->lsb_marche_funebre, &waiter);
        }

        cl_object_put(env, obj);
}

void cl_inode_fini(struct inode *inode)
{
        struct lu_env           *env;
        struct cl_inode_info    *lli  = cl_i2info(inode);
        struct cl_object        *clob = lli->lli_clob;
        int refcheck;
        int emergency;

        if (clob != NULL) {
                void                    *cookie;

                cookie = cl_env_reenter();
                env = cl_env_get(&refcheck);
                emergency = IS_ERR(env);
                if (emergency) {
                        mutex_lock(&ccc_inode_fini_guard);
                        LASSERT(ccc_inode_fini_env != NULL);
                        cl_env_implant(ccc_inode_fini_env, &refcheck);
                        env = ccc_inode_fini_env;
                }
                /*
                 * cl_object cache is a slave to inode cache (which, in turn
                 * is a slave to dentry cache), don't keep cl_object in memory
                 * when its master is evicted.
                 */
                cl_object_kill(env, clob);
                lu_object_ref_del(&clob->co_lu, "inode", inode);
                cl_object_put_last(env, clob);
                lli->lli_clob = NULL;
                if (emergency) {
                        cl_env_unplant(ccc_inode_fini_env, &refcheck);
                        mutex_unlock(&ccc_inode_fini_guard);
                } else
                        cl_env_put(env, &refcheck);
                cl_env_reexit(cookie);
        }
}

/**
 * return IF_* type for given lu_dirent entry.
 * IF_* flag shld be converted to particular OS file type in
 * platform llite module.
 */
__u16 ll_dirent_type_get(struct lu_dirent *ent)
{
        __u16 type = 0;
        struct luda_type *lt;
        int len = 0;

        if (le32_to_cpu(ent->lde_attrs) & LUDA_TYPE) {
                const unsigned align = sizeof(struct luda_type) - 1;

                len = le16_to_cpu(ent->lde_namelen);
                len = (len + align) & ~align;
                lt = (void *)ent->lde_name + len;
                type = IFTODT(le16_to_cpu(lt->lt_type));
        }
        return type;
}

/**
 * build inode number from passed @fid */
__u64 cl_fid_build_ino(const struct lu_fid *fid, int api32)
{
        if (BITS_PER_LONG == 32 || api32)
                RETURN(fid_flatten32(fid));
        else
                RETURN(fid_flatten(fid));
}

/**
 * build inode generation from passed @fid.  If our FID overflows the 32-bit
 * inode number then return a non-zero generation to distinguish them. */
__u32 cl_fid_build_gen(const struct lu_fid *fid)
{
        __u32 gen;
        ENTRY;

        if (fid_is_igif(fid)) {
                gen = lu_igif_gen(fid);
                RETURN(gen);
        }

        gen = (fid_flatten(fid) >> 32);
        RETURN(gen);
}

/* lsm is unreliable after hsm implementation as layout can be changed at
 * any time. This is only to support old, non-clio-ized interfaces. It will
 * cause deadlock if clio operations are called with this extra layout refcount
 * because in case the layout changed during the IO, ll_layout_refresh() will
 * have to wait for the refcount to become zero to destroy the older layout.
 *
 * Notice that the lsm returned by this function may not be valid unless called
 * inside layout lock - MDS_INODELOCK_LAYOUT. */
struct lov_stripe_md *ccc_inode_lsm_get(struct inode *inode)
{
        return lov_lsm_get(cl_i2info(inode)->lli_clob);
}

void inline ccc_inode_lsm_put(struct inode *inode, struct lov_stripe_md *lsm)
{
        lov_lsm_put(cl_i2info(inode)->lli_clob, lsm);
}