[deliverable/linux.git] / fs / nfsd / nfs4layouts.c

/*
 * Copyright (c) 2014 Christoph Hellwig.
 */
#include <linux/kmod.h>
#include <linux/file.h>
#include <linux/jhash.h>
#include <linux/sched.h>
#include <linux/sunrpc/addr.h>

#include "pnfs.h"
#include "netns.h"
#include "trace.h"

#define NFSDDBG_FACILITY                NFSDDBG_PNFS

struct nfs4_layout {
	struct list_head		lo_perstate;
	struct nfs4_layout_stateid	*lo_state;
	struct nfsd4_layout_seg		lo_seg;
};

static struct kmem_cache *nfs4_layout_cache;
static struct kmem_cache *nfs4_layout_stateid_cache;

static const struct nfsd4_callback_ops nfsd4_cb_layout_ops;
static const struct lock_manager_operations nfsd4_layouts_lm_ops;

const struct nfsd4_layout_ops *nfsd4_layout_ops[LAYOUT_TYPE_MAX] =  {
#ifdef CONFIG_NFSD_BLOCKLAYOUT
	[LAYOUT_BLOCK_VOLUME]	= &bl_layout_ops,
#endif
};

/* pNFS device ID to export fsid mapping */
#define DEVID_HASH_BITS	8
#define DEVID_HASH_SIZE	(1 << DEVID_HASH_BITS)
#define DEVID_HASH_MASK	(DEVID_HASH_SIZE - 1)
static u64 nfsd_devid_seq = 1;
static struct list_head nfsd_devid_hash[DEVID_HASH_SIZE];
static DEFINE_SPINLOCK(nfsd_devid_lock);

static inline u32 devid_hashfn(u64 idx)
{
	return jhash_2words(idx, idx >> 32, 0) & DEVID_HASH_MASK;
}

static void
nfsd4_alloc_devid_map(const struct svc_fh *fhp)
{
	const struct knfsd_fh *fh = &fhp->fh_handle;
	size_t fsid_len = key_len(fh->fh_fsid_type);
	struct nfsd4_deviceid_map *map, *old;
	int i;

	map = kzalloc(sizeof(*map) + fsid_len, GFP_KERNEL);
	if (!map)
		return;

	map->fsid_type = fh->fh_fsid_type;
	memcpy(&map->fsid, fh->fh_fsid, fsid_len);

	spin_lock(&nfsd_devid_lock);
	if (fhp->fh_export->ex_devid_map)
		goto out_unlock;

	for (i = 0; i < DEVID_HASH_SIZE; i++) {
		list_for_each_entry(old, &nfsd_devid_hash[i], hash) {
			if (old->fsid_type != fh->fh_fsid_type)
				continue;
			if (memcmp(old->fsid, fh->fh_fsid,
					key_len(old->fsid_type)))
				continue;

			fhp->fh_export->ex_devid_map = old;
			goto out_unlock;
		}
	}

	map->idx = nfsd_devid_seq++;
	list_add_tail_rcu(&map->hash, &nfsd_devid_hash[devid_hashfn(map->idx)]);
	fhp->fh_export->ex_devid_map = map;
	map = NULL;

out_unlock:
	spin_unlock(&nfsd_devid_lock);
	kfree(map);
}

struct nfsd4_deviceid_map *
nfsd4_find_devid_map(int idx)
{
	struct nfsd4_deviceid_map *map, *ret = NULL;

	rcu_read_lock();
	list_for_each_entry_rcu(map, &nfsd_devid_hash[devid_hashfn(idx)], hash)
		if (map->idx == idx)
			ret = map;
	rcu_read_unlock();

	return ret;
}

int
nfsd4_set_deviceid(struct nfsd4_deviceid *id, const struct svc_fh *fhp,
		u32 device_generation)
{
	if (!fhp->fh_export->ex_devid_map) {
		nfsd4_alloc_devid_map(fhp);
		if (!fhp->fh_export->ex_devid_map)
			return -ENOMEM;
	}

	id->fsid_idx = fhp->fh_export->ex_devid_map->idx;
	id->generation = device_generation;
	id->pad = 0;
	return 0;
}

void nfsd4_setup_layout_type(struct svc_export *exp)
{
	struct super_block *sb = exp->ex_path.mnt->mnt_sb;

	if (!(exp->ex_flags & NFSEXP_PNFS))
		return;

#ifdef CONFIG_NFSD_BLOCKLAYOUT
	if (sb->s_export_op->get_uuid &&
	    sb->s_export_op->map_blocks &&
	    sb->s_export_op->commit_blocks)
		exp->ex_layout_type = LAYOUT_BLOCK_VOLUME;
#endif
}

static void
nfsd4_free_layout_stateid(struct nfs4_stid *stid)
{
	struct nfs4_layout_stateid *ls = layoutstateid(stid);
	struct nfs4_client *clp = ls->ls_stid.sc_client;
	struct nfs4_file *fp = ls->ls_stid.sc_file;

	trace_layoutstate_free(&ls->ls_stid.sc_stateid);

	spin_lock(&clp->cl_lock);
	list_del_init(&ls->ls_perclnt);
	spin_unlock(&clp->cl_lock);

	spin_lock(&fp->fi_lock);
	list_del_init(&ls->ls_perfile);
	spin_unlock(&fp->fi_lock);

	vfs_setlease(ls->ls_file, F_UNLCK, NULL, (void **)&ls);
	fput(ls->ls_file);

	if (ls->ls_recalled)
		atomic_dec(&ls->ls_stid.sc_file->fi_lo_recalls);

	kmem_cache_free(nfs4_layout_stateid_cache, ls);
}

static int
nfsd4_layout_setlease(struct nfs4_layout_stateid *ls)
{
	struct file_lock *fl;
	int status;

	fl = locks_alloc_lock();
	if (!fl)
		return -ENOMEM;
	locks_init_lock(fl);
	fl->fl_lmops = &nfsd4_layouts_lm_ops;
	fl->fl_flags = FL_LAYOUT;
	fl->fl_type = F_RDLCK;
	fl->fl_end = OFFSET_MAX;
	fl->fl_owner = ls;
	fl->fl_pid = current->tgid;
	fl->fl_file = ls->ls_file;

	status = vfs_setlease(fl->fl_file, fl->fl_type, &fl, NULL);
	if (status) {
		locks_free_lock(fl);
		return status;
	}
	BUG_ON(fl != NULL);
	return 0;
}

static struct nfs4_layout_stateid *
nfsd4_alloc_layout_stateid(struct nfsd4_compound_state *cstate,
		struct nfs4_stid *parent, u32 layout_type)
{
	struct nfs4_client *clp = cstate->clp;
	struct nfs4_file *fp = parent->sc_file;
	struct nfs4_layout_stateid *ls;
	struct nfs4_stid *stp;

	stp = nfs4_alloc_stid(cstate->clp, nfs4_layout_stateid_cache);
	if (!stp)
		return NULL;
	stp->sc_free = nfsd4_free_layout_stateid;
	get_nfs4_file(fp);
	stp->sc_file = fp;

	ls = layoutstateid(stp);
	INIT_LIST_HEAD(&ls->ls_perclnt);
	INIT_LIST_HEAD(&ls->ls_perfile);
	spin_lock_init(&ls->ls_lock);
	INIT_LIST_HEAD(&ls->ls_layouts);
	mutex_init(&ls->ls_mutex);
	ls->ls_layout_type = layout_type;
	nfsd4_init_cb(&ls->ls_recall, clp, &nfsd4_cb_layout_ops,
			NFSPROC4_CLNT_CB_LAYOUT);

	if (parent->sc_type == NFS4_DELEG_STID)
		ls->ls_file = get_file(fp->fi_deleg_file);
	else
		ls->ls_file = find_any_file(fp);
	BUG_ON(!ls->ls_file);

	if (nfsd4_layout_setlease(ls)) {
		fput(ls->ls_file);
		put_nfs4_file(fp);
		kmem_cache_free(nfs4_layout_stateid_cache, ls);
		return NULL;
	}

	spin_lock(&clp->cl_lock);
	stp->sc_type = NFS4_LAYOUT_STID;
	list_add(&ls->ls_perclnt, &clp->cl_lo_states);
	spin_unlock(&clp->cl_lock);

	spin_lock(&fp->fi_lock);
	list_add(&ls->ls_perfile, &fp->fi_lo_states);
	spin_unlock(&fp->fi_lock);

	trace_layoutstate_alloc(&ls->ls_stid.sc_stateid);
	return ls;
}

__be32
nfsd4_preprocess_layout_stateid(struct svc_rqst *rqstp,
		struct nfsd4_compound_state *cstate, stateid_t *stateid,
		bool create, u32 layout_type, struct nfs4_layout_stateid **lsp)
{
	struct nfs4_layout_stateid *ls;
	struct nfs4_stid *stid;
	unsigned char typemask = NFS4_LAYOUT_STID;
	__be32 status;

	if (create)
		typemask |= (NFS4_OPEN_STID | NFS4_LOCK_STID | NFS4_DELEG_STID);

	status = nfsd4_lookup_stateid(cstate, stateid, typemask, &stid,
			net_generic(SVC_NET(rqstp), nfsd_net_id));
	if (status)
		goto out;

	if (!fh_match(&cstate->current_fh.fh_handle,
		      &stid->sc_file->fi_fhandle)) {
		status = nfserr_bad_stateid;
		goto out_put_stid;
	}

	if (stid->sc_type != NFS4_LAYOUT_STID) {
		ls = nfsd4_alloc_layout_stateid(cstate, stid, layout_type);
		nfs4_put_stid(stid);

		status = nfserr_jukebox;
		if (!ls)
			goto out;
		mutex_lock(&ls->ls_mutex);
	} else {
		ls = container_of(stid, struct nfs4_layout_stateid, ls_stid);

		status = nfserr_bad_stateid;
		mutex_lock(&ls->ls_mutex);
		if (stateid->si_generation > stid->sc_stateid.si_generation)
			goto out_unlock_stid;
		if (layout_type != ls->ls_layout_type)
			goto out_unlock_stid;
	}

	*lsp = ls;
	return 0;

out_unlock_stid:
	mutex_unlock(&ls->ls_mutex);
out_put_stid:
	nfs4_put_stid(stid);
out:
	return status;
}

static void
nfsd4_recall_file_layout(struct nfs4_layout_stateid *ls)
{
	spin_lock(&ls->ls_lock);
	if (ls->ls_recalled)
		goto out_unlock;

	ls->ls_recalled = true;
	atomic_inc(&ls->ls_stid.sc_file->fi_lo_recalls);
	if (list_empty(&ls->ls_layouts))
		goto out_unlock;

	trace_layout_recall(&ls->ls_stid.sc_stateid);

	atomic_inc(&ls->ls_stid.sc_count);
	nfsd4_run_cb(&ls->ls_recall);

out_unlock:
	spin_unlock(&ls->ls_lock);
}

static inline u64
layout_end(struct nfsd4_layout_seg *seg)
{
	u64 end = seg->offset + seg->length;
	return end >= seg->offset ? end : NFS4_MAX_UINT64;
}

static void
layout_update_len(struct nfsd4_layout_seg *lo, u64 end)
{
	if (end == NFS4_MAX_UINT64)
		lo->length = NFS4_MAX_UINT64;
	else
		lo->length = end - lo->offset;
}

static bool
layouts_overlapping(struct nfs4_layout *lo, struct nfsd4_layout_seg *s)
{
	if (s->iomode != IOMODE_ANY && s->iomode != lo->lo_seg.iomode)
		return false;
	if (layout_end(&lo->lo_seg) <= s->offset)
		return false;
	if (layout_end(s) <= lo->lo_seg.offset)
		return false;
	return true;
}

static bool
layouts_try_merge(struct nfsd4_layout_seg *lo, struct nfsd4_layout_seg *new)
{
	if (lo->iomode != new->iomode)
		return false;
	if (layout_end(new) < lo->offset)
		return false;
	if (layout_end(lo) < new->offset)
		return false;

	lo->offset = min(lo->offset, new->offset);
	layout_update_len(lo, max(layout_end(lo), layout_end(new)));
	return true;
}

static __be32
nfsd4_recall_conflict(struct nfs4_layout_stateid *ls)
{
	struct nfs4_file *fp = ls->ls_stid.sc_file;
	struct nfs4_layout_stateid *l, *n;
	__be32 nfserr = nfs_ok;

	assert_spin_locked(&fp->fi_lock);

	list_for_each_entry_safe(l, n, &fp->fi_lo_states, ls_perfile) {
		if (l != ls) {
			nfsd4_recall_file_layout(l);
			nfserr = nfserr_recallconflict;
		}
	}

	return nfserr;
}

__be32
nfsd4_insert_layout(struct nfsd4_layoutget *lgp, struct nfs4_layout_stateid *ls)
{
	struct nfsd4_layout_seg *seg = &lgp->lg_seg;
	struct nfs4_file *fp = ls->ls_stid.sc_file;
	struct nfs4_layout *lp, *new = NULL;
	__be32 nfserr;

	spin_lock(&fp->fi_lock);
	nfserr = nfsd4_recall_conflict(ls);
	if (nfserr)
		goto out;
	spin_lock(&ls->ls_lock);
	list_for_each_entry(lp, &ls->ls_layouts, lo_perstate) {
		if (layouts_try_merge(&lp->lo_seg, seg))
			goto done;
	}
	spin_unlock(&ls->ls_lock);
	spin_unlock(&fp->fi_lock);

	new = kmem_cache_alloc(nfs4_layout_cache, GFP_KERNEL);
	if (!new)
		return nfserr_jukebox;
	memcpy(&new->lo_seg, seg, sizeof(lp->lo_seg));
	new->lo_state = ls;

	spin_lock(&fp->fi_lock);
	nfserr = nfsd4_recall_conflict(ls);
	if (nfserr)
		goto out;
	spin_lock(&ls->ls_lock);
	list_for_each_entry(lp, &ls->ls_layouts, lo_perstate) {
		if (layouts_try_merge(&lp->lo_seg, seg))
			goto done;
	}

	atomic_inc(&ls->ls_stid.sc_count);
	list_add_tail(&new->lo_perstate, &ls->ls_layouts);
	new = NULL;
done:
	nfs4_inc_and_copy_stateid(&lgp->lg_sid, &ls->ls_stid);
	spin_unlock(&ls->ls_lock);
out:
	spin_unlock(&fp->fi_lock);
	if (new)
		kmem_cache_free(nfs4_layout_cache, new);
	return nfserr;
}

static void
nfsd4_free_layouts(struct list_head *reaplist)
{
	while (!list_empty(reaplist)) {
		struct nfs4_layout *lp = list_first_entry(reaplist,
				struct nfs4_layout, lo_perstate);

		list_del(&lp->lo_perstate);
		nfs4_put_stid(&lp->lo_state->ls_stid);
		kmem_cache_free(nfs4_layout_cache, lp);
	}
}

static void
nfsd4_return_file_layout(struct nfs4_layout *lp, struct nfsd4_layout_seg *seg,
		struct list_head *reaplist)
{
	struct nfsd4_layout_seg *lo = &lp->lo_seg;
	u64 end = layout_end(lo);

	if (seg->offset <= lo->offset) {
		if (layout_end(seg) >= end) {
			list_move_tail(&lp->lo_perstate, reaplist);
			return;
		}
		lo->offset = layout_end(seg);
	} else {
		/* retain the whole layout segment on a split. */
		if (layout_end(seg) < end) {
			dprintk("%s: split not supported\n", __func__);
			return;
		}
		end = seg->offset;
	}

	layout_update_len(lo, end);
}

__be32
nfsd4_return_file_layouts(struct svc_rqst *rqstp,
		struct nfsd4_compound_state *cstate,
		struct nfsd4_layoutreturn *lrp)
{
	struct nfs4_layout_stateid *ls;
	struct nfs4_layout *lp, *n;
	LIST_HEAD(reaplist);
	__be32 nfserr;
	int found = 0;

	nfserr = nfsd4_preprocess_layout_stateid(rqstp, cstate, &lrp->lr_sid,
						false, lrp->lr_layout_type,
						&ls);
	if (nfserr) {
		trace_layout_return_lookup_fail(&lrp->lr_sid);
		return nfserr;
	}

	spin_lock(&ls->ls_lock);
	list_for_each_entry_safe(lp, n, &ls->ls_layouts, lo_perstate) {
		if (layouts_overlapping(lp, &lrp->lr_seg)) {
			nfsd4_return_file_layout(lp, &lrp->lr_seg, &reaplist);
			found++;
		}
	}
	if (!list_empty(&ls->ls_layouts)) {
		if (found)
			nfs4_inc_and_copy_stateid(&lrp->lr_sid, &ls->ls_stid);
		lrp->lrs_present = 1;
	} else {
		trace_layoutstate_unhash(&ls->ls_stid.sc_stateid);
		nfs4_unhash_stid(&ls->ls_stid);
		lrp->lrs_present = 0;
	}
	spin_unlock(&ls->ls_lock);

	mutex_unlock(&ls->ls_mutex);
	nfs4_put_stid(&ls->ls_stid);
	nfsd4_free_layouts(&reaplist);
	return nfs_ok;
}

__be32
nfsd4_return_client_layouts(struct svc_rqst *rqstp,
		struct nfsd4_compound_state *cstate,
		struct nfsd4_layoutreturn *lrp)
{
	struct nfs4_layout_stateid *ls, *n;
	struct nfs4_client *clp = cstate->clp;
	struct nfs4_layout *lp, *t;
	LIST_HEAD(reaplist);

	lrp->lrs_present = 0;

	spin_lock(&clp->cl_lock);
	list_for_each_entry_safe(ls, n, &clp->cl_lo_states, ls_perclnt) {
		if (ls->ls_layout_type != lrp->lr_layout_type)
			continue;

		if (lrp->lr_return_type == RETURN_FSID &&
		    !fh_fsid_match(&ls->ls_stid.sc_file->fi_fhandle,
				   &cstate->current_fh.fh_handle))
			continue;

		spin_lock(&ls->ls_lock);
		list_for_each_entry_safe(lp, t, &ls->ls_layouts, lo_perstate) {
			if (lrp->lr_seg.iomode == IOMODE_ANY ||
			    lrp->lr_seg.iomode == lp->lo_seg.iomode)
				list_move_tail(&lp->lo_perstate, &reaplist);
		}
		spin_unlock(&ls->ls_lock);
	}
	spin_unlock(&clp->cl_lock);

	nfsd4_free_layouts(&reaplist);
	return 0;
}

static void
nfsd4_return_all_layouts(struct nfs4_layout_stateid *ls,
		struct list_head *reaplist)
{
	spin_lock(&ls->ls_lock);
	list_splice_init(&ls->ls_layouts, reaplist);
	spin_unlock(&ls->ls_lock);
}

void
nfsd4_return_all_client_layouts(struct nfs4_client *clp)
{
	struct nfs4_layout_stateid *ls, *n;
	LIST_HEAD(reaplist);

	spin_lock(&clp->cl_lock);
	list_for_each_entry_safe(ls, n, &clp->cl_lo_states, ls_perclnt)
		nfsd4_return_all_layouts(ls, &reaplist);
	spin_unlock(&clp->cl_lock);

	nfsd4_free_layouts(&reaplist);
}

void
nfsd4_return_all_file_layouts(struct nfs4_client *clp, struct nfs4_file *fp)
{
	struct nfs4_layout_stateid *ls, *n;
	LIST_HEAD(reaplist);

	spin_lock(&fp->fi_lock);
	list_for_each_entry_safe(ls, n, &fp->fi_lo_states, ls_perfile) {
		if (ls->ls_stid.sc_client == clp)
			nfsd4_return_all_layouts(ls, &reaplist);
	}
	spin_unlock(&fp->fi_lock);

	nfsd4_free_layouts(&reaplist);
}

static void
nfsd4_cb_layout_fail(struct nfs4_layout_stateid *ls)
{
	struct nfs4_client *clp = ls->ls_stid.sc_client;
	char addr_str[INET6_ADDRSTRLEN];
	static char *envp[] = {
		"HOME=/",
		"TERM=linux",
		"PATH=/sbin:/usr/sbin:/bin:/usr/bin",
		NULL
	};
	char *argv[8];
	int error;

	rpc_ntop((struct sockaddr *)&clp->cl_addr, addr_str, sizeof(addr_str));

	trace_layout_recall_fail(&ls->ls_stid.sc_stateid);

	printk(KERN_WARNING
		"nfsd: client %s failed to respond to layout recall. "
		"  Fencing..\n", addr_str);

	argv[0] = "/sbin/nfsd-recall-failed";
	argv[1] = addr_str;
	argv[2] = ls->ls_file->f_path.mnt->mnt_sb->s_id;
	argv[3] = NULL;

	error = call_usermodehelper(argv[0], argv, envp, UMH_WAIT_PROC);
	if (error) {
		printk(KERN_ERR "nfsd: fence failed for client %s: %d!\n",
			addr_str, error);
	}
}

static void
nfsd4_cb_layout_prepare(struct nfsd4_callback *cb)
{
	struct nfs4_layout_stateid *ls =
		container_of(cb, struct nfs4_layout_stateid, ls_recall);

	mutex_lock(&ls->ls_mutex);
	nfs4_inc_and_copy_stateid(&ls->ls_recall_sid, &ls->ls_stid);
	mutex_unlock(&ls->ls_mutex);
}

static int
nfsd4_cb_layout_done(struct nfsd4_callback *cb, struct rpc_task *task)
{
	struct nfs4_layout_stateid *ls =
		container_of(cb, struct nfs4_layout_stateid, ls_recall);
	struct nfsd_net *nn;
	ktime_t now, cutoff;
	LIST_HEAD(reaplist);


	switch (task->tk_status) {
	case 0:
	case -NFS4ERR_DELAY:
		/*
		 * Anything left? If not, then call it done. Note that we don't
		 * take the spinlock since this is an optimization and nothing
		 * should get added until the cb counter goes to zero.
		 */
		if (list_empty(&ls->ls_layouts))
			return 1;

		/* Poll the client until it's done with the layout */
		now = ktime_get();
		nn = net_generic(ls->ls_stid.sc_client->net, nfsd_net_id);

		/* Client gets 2 lease periods to return it */
		cutoff = ktime_add_ns(task->tk_start,
					 nn->nfsd4_lease * NSEC_PER_SEC * 2);

		if (ktime_before(now, cutoff)) {
			rpc_delay(task, HZ/100); /* 10 mili-seconds */
			return 0;
		}
		/* Fallthrough */
	case -NFS4ERR_NOMATCHING_LAYOUT:
		trace_layout_recall_done(&ls->ls_stid.sc_stateid);
		task->tk_status = 0;
		return 1;
	default:
		/*
		 * Unknown error or non-responding client, we'll need to fence.
		 */
		nfsd4_cb_layout_fail(ls);
		return -1;
	}
}

static void
nfsd4_cb_layout_release(struct nfsd4_callback *cb)
{
	struct nfs4_layout_stateid *ls =
		container_of(cb, struct nfs4_layout_stateid, ls_recall);
	LIST_HEAD(reaplist);

	trace_layout_recall_release(&ls->ls_stid.sc_stateid);

	nfsd4_return_all_layouts(ls, &reaplist);
	nfsd4_free_layouts(&reaplist);
	nfs4_put_stid(&ls->ls_stid);
}

static const struct nfsd4_callback_ops nfsd4_cb_layout_ops = {
	.prepare	= nfsd4_cb_layout_prepare,
	.done		= nfsd4_cb_layout_done,
	.release	= nfsd4_cb_layout_release,
};

static bool
nfsd4_layout_lm_break(struct file_lock *fl)
{
	/*
	 * We don't want the locks code to timeout the lease for us;
	 * we'll remove it ourself if a layout isn't returned
	 * in time:
	 */
	fl->fl_break_time = 0;
	nfsd4_recall_file_layout(fl->fl_owner);
	return false;
}

static int
nfsd4_layout_lm_change(struct file_lock *onlist, int arg,
		struct list_head *dispose)
{
	BUG_ON(!(arg & F_UNLCK));
	return lease_modify(onlist, arg, dispose);
}

static const struct lock_manager_operations nfsd4_layouts_lm_ops = {
	.lm_break	= nfsd4_layout_lm_break,
	.lm_change	= nfsd4_layout_lm_change,
};

int
nfsd4_init_pnfs(void)
{
	int i;

	for (i = 0; i < DEVID_HASH_SIZE; i++)
		INIT_LIST_HEAD(&nfsd_devid_hash[i]);

	nfs4_layout_cache = kmem_cache_create("nfs4_layout",
			sizeof(struct nfs4_layout), 0, 0, NULL);
	if (!nfs4_layout_cache)
		return -ENOMEM;

	nfs4_layout_stateid_cache = kmem_cache_create("nfs4_layout_stateid",
			sizeof(struct nfs4_layout_stateid), 0, 0, NULL);
	if (!nfs4_layout_stateid_cache) {
		kmem_cache_destroy(nfs4_layout_cache);
		return -ENOMEM;
	}
	return 0;
}

void
nfsd4_exit_pnfs(void)
{
	int i;

	kmem_cache_destroy(nfs4_layout_cache);
	kmem_cache_destroy(nfs4_layout_stateid_cache);

	for (i = 0; i < DEVID_HASH_SIZE; i++) {
		struct nfsd4_deviceid_map *map, *n;

		list_for_each_entry_safe(map, n, &nfsd_devid_hash[i], hash)
			kfree(map);
	}
}
Commit	Line	Data
9cf514cc CH	1	/*
	2	* Copyright (c) 2014 Christoph Hellwig.
	3	*/
c5c707f9 CH	4	#include <linux/kmod.h>
c5c707f9 CH	5	#include <linux/file.h>
9cf514cc CH	6	#include <linux/jhash.h>
9cf514cc CH	7	#include <linux/sched.h>
c5c707f9	8	#include <linux/sunrpc/addr.h>
9cf514cc CH	9
	10	#include "pnfs.h"
	11	#include "netns.h"
31ef83dc	12	#include "trace.h"
9cf514cc CH	13
	14	#define NFSDDBG_FACILITY NFSDDBG_PNFS
	15
	16	struct nfs4_layout {
	17	struct list_head lo_perstate;
	18	struct nfs4_layout_stateid *lo_state;
	19	struct nfsd4_layout_seg lo_seg;
	20	};
	21
	22	static struct kmem_cache *nfs4_layout_cache;
	23	static struct kmem_cache *nfs4_layout_stateid_cache;
	24
c4cb8974	25	static const struct nfsd4_callback_ops nfsd4_cb_layout_ops;
c5c707f9 CH	26	static const struct lock_manager_operations nfsd4_layouts_lm_ops;
c5c707f9 CH	27
9cf514cc	28	const struct nfsd4_layout_ops *nfsd4_layout_ops[LAYOUT_TYPE_MAX] = {
81c39329	29	#ifdef CONFIG_NFSD_BLOCKLAYOUT
8650b8a0	30	[LAYOUT_BLOCK_VOLUME] = &bl_layout_ops,
81c39329	31	#endif
9cf514cc CH	32	};
	33
	34	/* pNFS device ID to export fsid mapping */
	35	#define DEVID_HASH_BITS 8
	36	#define DEVID_HASH_SIZE (1 << DEVID_HASH_BITS)
	37	#define DEVID_HASH_MASK (DEVID_HASH_SIZE - 1)
	38	static u64 nfsd_devid_seq = 1;
	39	static struct list_head nfsd_devid_hash[DEVID_HASH_SIZE];
	40	static DEFINE_SPINLOCK(nfsd_devid_lock);
	41
	42	static inline u32 devid_hashfn(u64 idx)
	43	{
	44	return jhash_2words(idx, idx >> 32, 0) & DEVID_HASH_MASK;
	45	}
	46
	47	static void
	48	nfsd4_alloc_devid_map(const struct svc_fh *fhp)
	49	{
	50	const struct knfsd_fh *fh = &fhp->fh_handle;
	51	size_t fsid_len = key_len(fh->fh_fsid_type);
	52	struct nfsd4_deviceid_map map, old;
	53	int i;
	54
	55	map = kzalloc(sizeof(*map) + fsid_len, GFP_KERNEL);
	56	if (!map)
	57	return;
	58
	59	map->fsid_type = fh->fh_fsid_type;
	60	memcpy(&map->fsid, fh->fh_fsid, fsid_len);
	61
	62	spin_lock(&nfsd_devid_lock);
	63	if (fhp->fh_export->ex_devid_map)
	64	goto out_unlock;
	65
	66	for (i = 0; i < DEVID_HASH_SIZE; i++) {
	67	list_for_each_entry(old, &nfsd_devid_hash[i], hash) {
	68	if (old->fsid_type != fh->fh_fsid_type)
	69	continue;
	70	if (memcmp(old->fsid, fh->fh_fsid,
	71	key_len(old->fsid_type)))
	72	continue;
	73
	74	fhp->fh_export->ex_devid_map = old;
	75	goto out_unlock;
	76	}
	77	}
	78
	79	map->idx = nfsd_devid_seq++;
	80	list_add_tail_rcu(&map->hash, &nfsd_devid_hash[devid_hashfn(map->idx)]);
	81	fhp->fh_export->ex_devid_map = map;
	82	map = NULL;
	83
	84	out_unlock:
	85	spin_unlock(&nfsd_devid_lock);
	86	kfree(map);
	87	}
	88
	89	struct nfsd4_deviceid_map *
	90	nfsd4_find_devid_map(int idx)
	91	{
	92	struct nfsd4_deviceid_map map, ret = NULL;
	93
	94	rcu_read_lock();
	95	list_for_each_entry_rcu(map, &nfsd_devid_hash[devid_hashfn(idx)], hash)
96	if (map->idx == idx)
97	ret = map;
98	rcu_read_unlock();
99
100	return ret;
101	}
102
103	int
104	nfsd4_set_deviceid(struct nfsd4_deviceid id, const struct svc_fh fhp,
105	u32 device_generation)
106	{
107	if (!fhp->fh_export->ex_devid_map) {
108	nfsd4_alloc_devid_map(fhp);
109	if (!fhp->fh_export->ex_devid_map)
110	return -ENOMEM;
111	}
112
113	id->fsid_idx = fhp->fh_export->ex_devid_map->idx;
114	id->generation = device_generation;
115	id->pad = 0;
116	return 0;
117	}
118
119	void nfsd4_setup_layout_type(struct svc_export *exp)
120	{
8650b8a0 CH	121	struct super_block *sb = exp->ex_path.mnt->mnt_sb;
8650b8a0 CH	122
f3f03330	123	if (!(exp->ex_flags & NFSEXP_PNFS))
9cf514cc	124	return;
8650b8a0	125
81c39329	126	#ifdef CONFIG_NFSD_BLOCKLAYOUT
8650b8a0 CH	127	if (sb->s_export_op->get_uuid &&
	128	sb->s_export_op->map_blocks &&
	129	sb->s_export_op->commit_blocks)
	130	exp->ex_layout_type = LAYOUT_BLOCK_VOLUME;
81c39329	131	#endif
9cf514cc CH	132	}
	133
	134	static void
	135	nfsd4_free_layout_stateid(struct nfs4_stid *stid)
	136	{
	137	struct nfs4_layout_stateid *ls = layoutstateid(stid);
	138	struct nfs4_client *clp = ls->ls_stid.sc_client;
	139	struct nfs4_file *fp = ls->ls_stid.sc_file;
	140
31ef83dc CH	141	trace_layoutstate_free(&ls->ls_stid.sc_stateid);
31ef83dc CH	142
9cf514cc CH	143	spin_lock(&clp->cl_lock);
	144	list_del_init(&ls->ls_perclnt);
	145	spin_unlock(&clp->cl_lock);
	146
	147	spin_lock(&fp->fi_lock);
	148	list_del_init(&ls->ls_perfile);
	149	spin_unlock(&fp->fi_lock);
	150
c5c707f9 CH	151	vfs_setlease(ls->ls_file, F_UNLCK, NULL, (void **)&ls);
	152	fput(ls->ls_file);
	153
	154	if (ls->ls_recalled)
	155	atomic_dec(&ls->ls_stid.sc_file->fi_lo_recalls);
	156
9cf514cc CH	157	kmem_cache_free(nfs4_layout_stateid_cache, ls);
	158	}
	159
c5c707f9 CH	160	static int
	161	nfsd4_layout_setlease(struct nfs4_layout_stateid *ls)
	162	{
	163	struct file_lock *fl;
	164	int status;
	165
	166	fl = locks_alloc_lock();
	167	if (!fl)
	168	return -ENOMEM;
	169	locks_init_lock(fl);
	170	fl->fl_lmops = &nfsd4_layouts_lm_ops;
	171	fl->fl_flags = FL_LAYOUT;
	172	fl->fl_type = F_RDLCK;
	173	fl->fl_end = OFFSET_MAX;
	174	fl->fl_owner = ls;
	175	fl->fl_pid = current->tgid;
	176	fl->fl_file = ls->ls_file;
	177
	178	status = vfs_setlease(fl->fl_file, fl->fl_type, &fl, NULL);
	179	if (status) {
	180	locks_free_lock(fl);
	181	return status;
	182	}
	183	BUG_ON(fl != NULL);
	184	return 0;
	185	}
	186
9cf514cc CH	187	static struct nfs4_layout_stateid *
	188	nfsd4_alloc_layout_stateid(struct nfsd4_compound_state *cstate,
	189	struct nfs4_stid *parent, u32 layout_type)
	190	{
	191	struct nfs4_client *clp = cstate->clp;
	192	struct nfs4_file *fp = parent->sc_file;
	193	struct nfs4_layout_stateid *ls;
	194	struct nfs4_stid *stp;
	195
	196	stp = nfs4_alloc_stid(cstate->clp, nfs4_layout_stateid_cache);
	197	if (!stp)
	198	return NULL;
	199	stp->sc_free = nfsd4_free_layout_stateid;
	200	get_nfs4_file(fp);
	201	stp->sc_file = fp;
	202
	203	ls = layoutstateid(stp);
	204	INIT_LIST_HEAD(&ls->ls_perclnt);
	205	INIT_LIST_HEAD(&ls->ls_perfile);
	206	spin_lock_init(&ls->ls_lock);
	207	INIT_LIST_HEAD(&ls->ls_layouts);
cc8a5532	208	mutex_init(&ls->ls_mutex);
9cf514cc	209	ls->ls_layout_type = layout_type;
c5c707f9 CH	210	nfsd4_init_cb(&ls->ls_recall, clp, &nfsd4_cb_layout_ops,
	211	NFSPROC4_CLNT_CB_LAYOUT);
	212
	213	if (parent->sc_type == NFS4_DELEG_STID)
	214	ls->ls_file = get_file(fp->fi_deleg_file);
	215	else
	216	ls->ls_file = find_any_file(fp);
	217	BUG_ON(!ls->ls_file);
	218
	219	if (nfsd4_layout_setlease(ls)) {
1ca4b88e	220	fput(ls->ls_file);
c5c707f9 CH	221	put_nfs4_file(fp);
	222	kmem_cache_free(nfs4_layout_stateid_cache, ls);
	223	return NULL;
	224	}
9cf514cc CH	225
	226	spin_lock(&clp->cl_lock);
	227	stp->sc_type = NFS4_LAYOUT_STID;
	228	list_add(&ls->ls_perclnt, &clp->cl_lo_states);
	229	spin_unlock(&clp->cl_lock);
	230
	231	spin_lock(&fp->fi_lock);
	232	list_add(&ls->ls_perfile, &fp->fi_lo_states);
	233	spin_unlock(&fp->fi_lock);
	234
31ef83dc	235	trace_layoutstate_alloc(&ls->ls_stid.sc_stateid);
9cf514cc CH	236	return ls;
	237	}
	238
	239	__be32
	240	nfsd4_preprocess_layout_stateid(struct svc_rqst *rqstp,
	241	struct nfsd4_compound_state cstate, stateid_t stateid,
	242	bool create, u32 layout_type, struct nfs4_layout_stateid **lsp)
	243	{
	244	struct nfs4_layout_stateid *ls;
	245	struct nfs4_stid *stid;
	246	unsigned char typemask = NFS4_LAYOUT_STID;
	247	__be32 status;
	248
	249	if (create)
	250	typemask \|= (NFS4_OPEN_STID \| NFS4_LOCK_STID \| NFS4_DELEG_STID);
	251
	252	status = nfsd4_lookup_stateid(cstate, stateid, typemask, &stid,
	253	net_generic(SVC_NET(rqstp), nfsd_net_id));
	254	if (status)
	255	goto out;
	256
	257	if (!fh_match(&cstate->current_fh.fh_handle,
	258	&stid->sc_file->fi_fhandle)) {
	259	status = nfserr_bad_stateid;
	260	goto out_put_stid;
	261	}
	262
	263	if (stid->sc_type != NFS4_LAYOUT_STID) {
	264	ls = nfsd4_alloc_layout_stateid(cstate, stid, layout_type);
	265	nfs4_put_stid(stid);
	266
	267	status = nfserr_jukebox;
	268	if (!ls)
	269	goto out;
cc8a5532	270	mutex_lock(&ls->ls_mutex);
9cf514cc CH	271	} else {
	272	ls = container_of(stid, struct nfs4_layout_stateid, ls_stid);
	273
	274	status = nfserr_bad_stateid;
cc8a5532	275	mutex_lock(&ls->ls_mutex);
9cf514cc	276	if (stateid->si_generation > stid->sc_stateid.si_generation)
cc8a5532	277	goto out_unlock_stid;
9cf514cc	278	if (layout_type != ls->ls_layout_type)
cc8a5532	279	goto out_unlock_stid;
9cf514cc CH	280	}
	281
	282	*lsp = ls;
	283	return 0;
	284
cc8a5532 JL	285	out_unlock_stid:
cc8a5532 JL	286	mutex_unlock(&ls->ls_mutex);
9cf514cc CH	287	out_put_stid:
	288	nfs4_put_stid(stid);
	289	out:
	290	return status;
	291	}
	292
c5c707f9 CH	293	static void
	294	nfsd4_recall_file_layout(struct nfs4_layout_stateid *ls)
	295	{
	296	spin_lock(&ls->ls_lock);
	297	if (ls->ls_recalled)
	298	goto out_unlock;
	299
	300	ls->ls_recalled = true;
	301	atomic_inc(&ls->ls_stid.sc_file->fi_lo_recalls);
	302	if (list_empty(&ls->ls_layouts))
	303	goto out_unlock;
	304
31ef83dc CH	305	trace_layout_recall(&ls->ls_stid.sc_stateid);
31ef83dc CH	306
c5c707f9	307	atomic_inc(&ls->ls_stid.sc_count);
c5c707f9 CH	308	nfsd4_run_cb(&ls->ls_recall);
	309
	310	out_unlock:
	311	spin_unlock(&ls->ls_lock);
	312	}
	313
9cf514cc CH	314	static inline u64
	315	layout_end(struct nfsd4_layout_seg *seg)
	316	{
	317	u64 end = seg->offset + seg->length;
	318	return end >= seg->offset ? end : NFS4_MAX_UINT64;
	319	}
	320
	321	static void
	322	layout_update_len(struct nfsd4_layout_seg *lo, u64 end)
	323	{
	324	if (end == NFS4_MAX_UINT64)
	325	lo->length = NFS4_MAX_UINT64;
	326	else
	327	lo->length = end - lo->offset;
	328	}
	329
	330	static bool
	331	layouts_overlapping(struct nfs4_layout lo, struct nfsd4_layout_seg s)
	332	{
	333	if (s->iomode != IOMODE_ANY && s->iomode != lo->lo_seg.iomode)
	334	return false;
	335	if (layout_end(&lo->lo_seg) <= s->offset)
	336	return false;
	337	if (layout_end(s) <= lo->lo_seg.offset)
	338	return false;
	339	return true;
	340	}
	341
	342	static bool
	343	layouts_try_merge(struct nfsd4_layout_seg lo, struct nfsd4_layout_seg new)
	344	{
	345	if (lo->iomode != new->iomode)
	346	return false;
	347	if (layout_end(new) < lo->offset)
	348	return false;
	349	if (layout_end(lo) < new->offset)
	350	return false;
	351
	352	lo->offset = min(lo->offset, new->offset);
	353	layout_update_len(lo, max(layout_end(lo), layout_end(new)));
	354	return true;
	355	}
	356
c5c707f9 CH	357	static __be32
	358	nfsd4_recall_conflict(struct nfs4_layout_stateid *ls)
	359	{
	360	struct nfs4_file *fp = ls->ls_stid.sc_file;
	361	struct nfs4_layout_stateid l, n;
	362	__be32 nfserr = nfs_ok;
	363
	364	assert_spin_locked(&fp->fi_lock);
	365
	366	list_for_each_entry_safe(l, n, &fp->fi_lo_states, ls_perfile) {
	367	if (l != ls) {
	368	nfsd4_recall_file_layout(l);
	369	nfserr = nfserr_recallconflict;
	370	}
	371	}
	372
	373	return nfserr;
	374	}
	375
9cf514cc CH	376	__be32
	377	nfsd4_insert_layout(struct nfsd4_layoutget lgp, struct nfs4_layout_stateid ls)
	378	{
	379	struct nfsd4_layout_seg *seg = &lgp->lg_seg;
c5c707f9	380	struct nfs4_file *fp = ls->ls_stid.sc_file;
9cf514cc	381	struct nfs4_layout lp, new = NULL;
c5c707f9	382	__be32 nfserr;
9cf514cc	383
c5c707f9 CH	384	spin_lock(&fp->fi_lock);
	385	nfserr = nfsd4_recall_conflict(ls);
	386	if (nfserr)
	387	goto out;
9cf514cc CH	388	spin_lock(&ls->ls_lock);
	389	list_for_each_entry(lp, &ls->ls_layouts, lo_perstate) {
	390	if (layouts_try_merge(&lp->lo_seg, seg))
	391	goto done;
	392	}
	393	spin_unlock(&ls->ls_lock);
c5c707f9	394	spin_unlock(&fp->fi_lock);
9cf514cc CH	395
	396	new = kmem_cache_alloc(nfs4_layout_cache, GFP_KERNEL);
	397	if (!new)
	398	return nfserr_jukebox;
	399	memcpy(&new->lo_seg, seg, sizeof(lp->lo_seg));
	400	new->lo_state = ls;
	401
c5c707f9 CH	402	spin_lock(&fp->fi_lock);
	403	nfserr = nfsd4_recall_conflict(ls);
	404	if (nfserr)
	405	goto out;
9cf514cc CH	406	spin_lock(&ls->ls_lock);
	407	list_for_each_entry(lp, &ls->ls_layouts, lo_perstate) {
	408	if (layouts_try_merge(&lp->lo_seg, seg))
	409	goto done;
	410	}
	411
	412	atomic_inc(&ls->ls_stid.sc_count);
	413	list_add_tail(&new->lo_perstate, &ls->ls_layouts);
	414	new = NULL;
	415	done:
9767feb2	416	nfs4_inc_and_copy_stateid(&lgp->lg_sid, &ls->ls_stid);
9cf514cc	417	spin_unlock(&ls->ls_lock);
c5c707f9 CH	418	out:
c5c707f9 CH	419	spin_unlock(&fp->fi_lock);
9cf514cc CH	420	if (new)
9cf514cc CH	421	kmem_cache_free(nfs4_layout_cache, new);
c5c707f9	422	return nfserr;
9cf514cc CH	423	}
	424
	425	static void
	426	nfsd4_free_layouts(struct list_head *reaplist)
	427	{
	428	while (!list_empty(reaplist)) {
	429	struct nfs4_layout *lp = list_first_entry(reaplist,
	430	struct nfs4_layout, lo_perstate);
	431
	432	list_del(&lp->lo_perstate);
	433	nfs4_put_stid(&lp->lo_state->ls_stid);
	434	kmem_cache_free(nfs4_layout_cache, lp);
	435	}
	436	}
	437
	438	static void
	439	nfsd4_return_file_layout(struct nfs4_layout lp, struct nfsd4_layout_seg seg,
	440	struct list_head *reaplist)
	441	{
	442	struct nfsd4_layout_seg *lo = &lp->lo_seg;
	443	u64 end = layout_end(lo);
	444
	445	if (seg->offset <= lo->offset) {
	446	if (layout_end(seg) >= end) {
	447	list_move_tail(&lp->lo_perstate, reaplist);
	448	return;
	449	}
7890203d	450	lo->offset = layout_end(seg);
9cf514cc CH	451	} else {
	452	/* retain the whole layout segment on a split. */
	453	if (layout_end(seg) < end) {
	454	dprintk("%s: split not supported\n", __func__);
	455	return;
	456	}
7890203d	457	end = seg->offset;
9cf514cc CH	458	}
	459
	460	layout_update_len(lo, end);
	461	}
	462
	463	__be32
	464	nfsd4_return_file_layouts(struct svc_rqst *rqstp,
	465	struct nfsd4_compound_state *cstate,
	466	struct nfsd4_layoutreturn *lrp)
	467	{
	468	struct nfs4_layout_stateid *ls;
	469	struct nfs4_layout lp, n;
	470	LIST_HEAD(reaplist);
	471	__be32 nfserr;
	472	int found = 0;
	473
	474	nfserr = nfsd4_preprocess_layout_stateid(rqstp, cstate, &lrp->lr_sid,
	475	false, lrp->lr_layout_type,
	476	&ls);
31ef83dc CH	477	if (nfserr) {
31ef83dc CH	478	trace_layout_return_lookup_fail(&lrp->lr_sid);
9cf514cc	479	return nfserr;
31ef83dc	480	}
9cf514cc CH	481
	482	spin_lock(&ls->ls_lock);
	483	list_for_each_entry_safe(lp, n, &ls->ls_layouts, lo_perstate) {
	484	if (layouts_overlapping(lp, &lrp->lr_seg)) {
	485	nfsd4_return_file_layout(lp, &lrp->lr_seg, &reaplist);
	486	found++;
	487	}
	488	}
	489	if (!list_empty(&ls->ls_layouts)) {
9767feb2 JL	490	if (found)
9767feb2 JL	491	nfs4_inc_and_copy_stateid(&lrp->lr_sid, &ls->ls_stid);
9cf514cc CH	492	lrp->lrs_present = 1;
9cf514cc CH	493	} else {
31ef83dc	494	trace_layoutstate_unhash(&ls->ls_stid.sc_stateid);
9cf514cc CH	495	nfs4_unhash_stid(&ls->ls_stid);
	496	lrp->lrs_present = 0;
	497	}
	498	spin_unlock(&ls->ls_lock);
	499
cc8a5532	500	mutex_unlock(&ls->ls_mutex);
9cf514cc CH	501	nfs4_put_stid(&ls->ls_stid);
	502	nfsd4_free_layouts(&reaplist);
	503	return nfs_ok;
	504	}
	505
	506	__be32
	507	nfsd4_return_client_layouts(struct svc_rqst *rqstp,
	508	struct nfsd4_compound_state *cstate,
	509	struct nfsd4_layoutreturn *lrp)
	510	{
	511	struct nfs4_layout_stateid ls, n;
	512	struct nfs4_client *clp = cstate->clp;
	513	struct nfs4_layout lp, t;
	514	LIST_HEAD(reaplist);
	515
	516	lrp->lrs_present = 0;
	517
	518	spin_lock(&clp->cl_lock);
	519	list_for_each_entry_safe(ls, n, &clp->cl_lo_states, ls_perclnt) {
6f8f28ec KM	520	if (ls->ls_layout_type != lrp->lr_layout_type)
	521	continue;
	522
9cf514cc CH	523	if (lrp->lr_return_type == RETURN_FSID &&
	524	!fh_fsid_match(&ls->ls_stid.sc_file->fi_fhandle,
	525	&cstate->current_fh.fh_handle))
	526	continue;
	527
	528	spin_lock(&ls->ls_lock);
	529	list_for_each_entry_safe(lp, t, &ls->ls_layouts, lo_perstate) {
	530	if (lrp->lr_seg.iomode == IOMODE_ANY \|\|
	531	lrp->lr_seg.iomode == lp->lo_seg.iomode)
	532	list_move_tail(&lp->lo_perstate, &reaplist);
	533	}
	534	spin_unlock(&ls->ls_lock);
	535	}
	536	spin_unlock(&clp->cl_lock);
	537
	538	nfsd4_free_layouts(&reaplist);
	539	return 0;
	540	}
	541
	542	static void
	543	nfsd4_return_all_layouts(struct nfs4_layout_stateid *ls,
	544	struct list_head *reaplist)
	545	{
	546	spin_lock(&ls->ls_lock);
	547	list_splice_init(&ls->ls_layouts, reaplist);
	548	spin_unlock(&ls->ls_lock);
	549	}
	550
	551	void
	552	nfsd4_return_all_client_layouts(struct nfs4_client *clp)
	553	{
	554	struct nfs4_layout_stateid ls, n;
	555	LIST_HEAD(reaplist);
	556
	557	spin_lock(&clp->cl_lock);
	558	list_for_each_entry_safe(ls, n, &clp->cl_lo_states, ls_perclnt)
	559	nfsd4_return_all_layouts(ls, &reaplist);
	560	spin_unlock(&clp->cl_lock);
	561
	562	nfsd4_free_layouts(&reaplist);
	563	}
	564
	565	void
	566	nfsd4_return_all_file_layouts(struct nfs4_client clp, struct nfs4_file fp)
	567	{
	568	struct nfs4_layout_stateid ls, n;
	569	LIST_HEAD(reaplist);
	570
	571	spin_lock(&fp->fi_lock);
	572	list_for_each_entry_safe(ls, n, &fp->fi_lo_states, ls_perfile) {
	573	if (ls->ls_stid.sc_client == clp)
	574	nfsd4_return_all_layouts(ls, &reaplist);
	575	}
	576	spin_unlock(&fp->fi_lock);
	577
	578	nfsd4_free_layouts(&reaplist);
	579	}
	580
c5c707f9 CH	581	static void
	582	nfsd4_cb_layout_fail(struct nfs4_layout_stateid *ls)
	583	{
	584	struct nfs4_client *clp = ls->ls_stid.sc_client;
	585	char addr_str[INET6_ADDRSTRLEN];
	586	static char *envp[] = {
	587	"HOME=/",
	588	"TERM=linux",
	589	"PATH=/sbin:/usr/sbin:/bin:/usr/bin",
	590	NULL
	591	};
	592	char *argv[8];
	593	int error;
	594
	595	rpc_ntop((struct sockaddr *)&clp->cl_addr, addr_str, sizeof(addr_str));
	596
715a03d2 KM	597	trace_layout_recall_fail(&ls->ls_stid.sc_stateid);
715a03d2 KM	598
c5c707f9 CH	599	printk(KERN_WARNING
	600	"nfsd: client %s failed to respond to layout recall. "
	601	" Fencing..\n", addr_str);
	602
	603	argv[0] = "/sbin/nfsd-recall-failed";
	604	argv[1] = addr_str;
	605	argv[2] = ls->ls_file->f_path.mnt->mnt_sb->s_id;
	606	argv[3] = NULL;
	607
	608	error = call_usermodehelper(argv[0], argv, envp, UMH_WAIT_PROC);
	609	if (error) {
	610	printk(KERN_ERR "nfsd: fence failed for client %s: %d!\n",
	611	addr_str, error);
	612	}
	613	}
	614
cc8a5532 JL	615	static void
	616	nfsd4_cb_layout_prepare(struct nfsd4_callback *cb)
	617	{
	618	struct nfs4_layout_stateid *ls =
	619	container_of(cb, struct nfs4_layout_stateid, ls_recall);
	620
	621	mutex_lock(&ls->ls_mutex);
9767feb2	622	nfs4_inc_and_copy_stateid(&ls->ls_recall_sid, &ls->ls_stid);
be20aa00	623	mutex_unlock(&ls->ls_mutex);
cc8a5532 JL	624	}
cc8a5532 JL	625
c5c707f9 CH	626	static int
	627	nfsd4_cb_layout_done(struct nfsd4_callback cb, struct rpc_task task)
	628	{
	629	struct nfs4_layout_stateid *ls =
	630	container_of(cb, struct nfs4_layout_stateid, ls_recall);
6b9b2107 JL	631	struct nfsd_net *nn;
6b9b2107 JL	632	ktime_t now, cutoff;
c5c707f9 CH	633	LIST_HEAD(reaplist);
c5c707f9 CH	634
6b9b2107	635
c5c707f9 CH	636	switch (task->tk_status) {
c5c707f9 CH	637	case 0:
6b9b2107 JL	638	case -NFS4ERR_DELAY:
	639	/*
	640	* Anything left? If not, then call it done. Note that we don't
	641	* take the spinlock since this is an optimization and nothing
	642	* should get added until the cb counter goes to zero.
	643	*/
	644	if (list_empty(&ls->ls_layouts))
	645	return 1;
	646
	647	/* Poll the client until it's done with the layout */
	648	now = ktime_get();
	649	nn = net_generic(ls->ls_stid.sc_client->net, nfsd_net_id);
	650
	651	/* Client gets 2 lease periods to return it */
	652	cutoff = ktime_add_ns(task->tk_start,
	653	nn->nfsd4_lease * NSEC_PER_SEC * 2);
	654
	655	if (ktime_before(now, cutoff)) {
	656	rpc_delay(task, HZ/100); /* 10 mili-seconds */
	657	return 0;
	658	}
	659	/* Fallthrough */
c5c707f9	660	case -NFS4ERR_NOMATCHING_LAYOUT:
31ef83dc	661	trace_layout_recall_done(&ls->ls_stid.sc_stateid);
c5c707f9 CH	662	task->tk_status = 0;
c5c707f9 CH	663	return 1;
c5c707f9 CH	664	default:
	665	/*
	666	* Unknown error or non-responding client, we'll need to fence.
	667	*/
	668	nfsd4_cb_layout_fail(ls);
	669	return -1;
	670	}
	671	}
	672
	673	static void
	674	nfsd4_cb_layout_release(struct nfsd4_callback *cb)
	675	{
	676	struct nfs4_layout_stateid *ls =
	677	container_of(cb, struct nfs4_layout_stateid, ls_recall);
	678	LIST_HEAD(reaplist);
	679
31ef83dc CH	680	trace_layout_recall_release(&ls->ls_stid.sc_stateid);
31ef83dc CH	681
c5c707f9 CH	682	nfsd4_return_all_layouts(ls, &reaplist);
	683	nfsd4_free_layouts(&reaplist);
	684	nfs4_put_stid(&ls->ls_stid);
	685	}
	686
c4cb8974	687	static const struct nfsd4_callback_ops nfsd4_cb_layout_ops = {
cc8a5532	688	.prepare = nfsd4_cb_layout_prepare,
c5c707f9 CH	689	.done = nfsd4_cb_layout_done,
	690	.release = nfsd4_cb_layout_release,
	691	};
	692
	693	static bool
	694	nfsd4_layout_lm_break(struct file_lock *fl)
	695	{
	696	/*
	697	* We don't want the locks code to timeout the lease for us;
	698	* we'll remove it ourself if a layout isn't returned
	699	* in time:
	700	*/
	701	fl->fl_break_time = 0;
	702	nfsd4_recall_file_layout(fl->fl_owner);
	703	return false;
	704	}
	705
	706	static int
	707	nfsd4_layout_lm_change(struct file_lock *onlist, int arg,
	708	struct list_head *dispose)
	709	{
	710	BUG_ON(!(arg & F_UNLCK));
	711	return lease_modify(onlist, arg, dispose);
	712	}
	713
	714	static const struct lock_manager_operations nfsd4_layouts_lm_ops = {
	715	.lm_break = nfsd4_layout_lm_break,
	716	.lm_change = nfsd4_layout_lm_change,
	717	};
	718
9cf514cc CH	719	int
	720	nfsd4_init_pnfs(void)
	721	{
	722	int i;
	723
	724	for (i = 0; i < DEVID_HASH_SIZE; i++)
	725	INIT_LIST_HEAD(&nfsd_devid_hash[i]);
	726
	727	nfs4_layout_cache = kmem_cache_create("nfs4_layout",
	728	sizeof(struct nfs4_layout), 0, 0, NULL);
	729	if (!nfs4_layout_cache)
	730	return -ENOMEM;
	731
	732	nfs4_layout_stateid_cache = kmem_cache_create("nfs4_layout_stateid",
	733	sizeof(struct nfs4_layout_stateid), 0, 0, NULL);
	734	if (!nfs4_layout_stateid_cache) {
	735	kmem_cache_destroy(nfs4_layout_cache);
	736	return -ENOMEM;
	737	}
	738	return 0;
	739	}
	740
	741	void
	742	nfsd4_exit_pnfs(void)
	743	{
	744	int i;
	745
	746	kmem_cache_destroy(nfs4_layout_cache);
	747	kmem_cache_destroy(nfs4_layout_stateid_cache);
	748
	749	for (i = 0; i < DEVID_HASH_SIZE; i++) {
	750	struct nfsd4_deviceid_map map, n;
	751
	752	list_for_each_entry_safe(map, n, &nfsd_devid_hash[i], hash)
	753	kfree(map);
	754	}
	755	}