All of lore.kernel.org
 help / color / mirror / Atom feed
diff for duplicates of <20111222155111.GA21432@ca-server1.us.oracle.com>

diff --git a/a/1.txt b/N1/1.txt
index 8b13789..40067ae 100644
--- a/a/1.txt
+++ b/N1/1.txt
@@ -1 +1,3048 @@
+>From 60ce12cd6fb7d81d3bad48cf3766269547b4c241 Mon Sep 17 00:00:00 2001
+From: Dan Magenheimer <dan.magenheimer@oracle.com>
+Date: Wed, 21 Dec 2011 14:01:56 -0700
+Subject: [PATCH V2 2/6] drivers/staging/ramster: local compression + tmem
 
+Copy files from drivers/staging/zcache.  Ramster compresses pages
+locally before transmitting them to another node, so we can
+leverage the zcache and tmem code directly.  Note: there are
+no ramster-specific changes yet to these files.
+
+Signed-off-by: Dan Magenheimer <dan.magenheimer@oracle.com>
+
+---
+
+ drivers/staging/ramster/Kconfig       |   13 +
+ drivers/staging/ramster/Makefile      |    3 +
+ drivers/staging/ramster/tmem.c        |  770 +++++++++++++
+ drivers/staging/ramster/tmem.h        |  206 ++++
+ drivers/staging/ramster/zcache-main.c | 2003 +++++++++++++++++++++++++++++++++
+ 5 files changed, 2995 insertions(+), 0 deletions(-)
+
+diff --git a/drivers/staging/ramster/Kconfig b/drivers/staging/ramster/Kconfig
+new file mode 100644
+index 0000000..7fabcb2
+--- /dev/null
++++ b/drivers/staging/ramster/Kconfig
+@@ -0,0 +1,13 @@
++config ZCACHE
++	tristate "Dynamic compression of swap pages and clean pagecache pages"
++	depends on CLEANCACHE || FRONTSWAP
++	select XVMALLOC
++	select LZO_COMPRESS
++	select LZO_DECOMPRESS
++	default n
++	help
++	  Zcache doubles RAM efficiency while providing a significant
++	  performance boosts on many workloads.  Zcache uses lzo1x
++	  compression and an in-kernel implementation of transcendent
++	  memory to store clean page cache pages and swap in RAM,
++	  providing a noticeable reduction in disk I/O.
+diff --git a/drivers/staging/ramster/Makefile b/drivers/staging/ramster/Makefile
+new file mode 100644
+index 0000000..60daa27
+--- /dev/null
++++ b/drivers/staging/ramster/Makefile
+@@ -0,0 +1,3 @@
++zcache-y	:=	zcache-main.o tmem.o
++
++obj-$(CONFIG_ZCACHE)	+=	zcache.o
+diff --git a/drivers/staging/ramster/tmem.c b/drivers/staging/ramster/tmem.c
+new file mode 100644
+index 0000000..1ca66ea
+--- /dev/null
++++ b/drivers/staging/ramster/tmem.c
+@@ -0,0 +1,770 @@
++/*
++ * In-kernel transcendent memory (generic implementation)
++ *
++ * Copyright (c) 2009-2011, Dan Magenheimer, Oracle Corp.
++ *
++ * The primary purpose of Transcedent Memory ("tmem") is to map object-oriented
++ * "handles" (triples containing a pool id, and object id, and an index), to
++ * pages in a page-accessible memory (PAM).  Tmem references the PAM pages via
++ * an abstract "pampd" (PAM page-descriptor), which can be operated on by a
++ * set of functions (pamops).  Each pampd contains some representation of
++ * PAGE_SIZE bytes worth of data. Tmem must support potentially millions of
++ * pages and must be able to insert, find, and delete these pages at a
++ * potential frequency of thousands per second concurrently across many CPUs,
++ * (and, if used with KVM, across many vcpus across many guests).
++ * Tmem is tracked with a hierarchy of data structures, organized by
++ * the elements in a handle-tuple: pool_id, object_id, and page index.
++ * One or more "clients" (e.g. guests) each provide one or more tmem_pools.
++ * Each pool, contains a hash table of rb_trees of tmem_objs.  Each
++ * tmem_obj contains a radix-tree-like tree of pointers, with intermediate
++ * nodes called tmem_objnodes.  Each leaf pointer in this tree points to
++ * a pampd, which is accessible only through a small set of callbacks
++ * registered by the PAM implementation (see tmem_register_pamops). Tmem
++ * does all memory allocation via a set of callbacks registered by the tmem
++ * host implementation (e.g. see tmem_register_hostops).
++ */
++
++#include <linux/list.h>
++#include <linux/spinlock.h>
++#include <linux/atomic.h>
++
++#include "tmem.h"
++
++/* data structure sentinels used for debugging... see tmem.h */
++#define POOL_SENTINEL 0x87658765
++#define OBJ_SENTINEL 0x12345678
++#define OBJNODE_SENTINEL 0xfedcba09
++
++/*
++ * A tmem host implementation must use this function to register callbacks
++ * for memory allocation.
++ */
++static struct tmem_hostops tmem_hostops;
++
++static void tmem_objnode_tree_init(void);
++
++void tmem_register_hostops(struct tmem_hostops *m)
++{
++	tmem_objnode_tree_init();
++	tmem_hostops = *m;
++}
++
++/*
++ * A tmem host implementation must use this function to register
++ * callbacks for a page-accessible memory (PAM) implementation
++ */
++static struct tmem_pamops tmem_pamops;
++
++void tmem_register_pamops(struct tmem_pamops *m)
++{
++	tmem_pamops = *m;
++}
++
++/*
++ * Oid's are potentially very sparse and tmem_objs may have an indeterminately
++ * short life, being added and deleted at a relatively high frequency.
++ * So an rb_tree is an ideal data structure to manage tmem_objs.  But because
++ * of the potentially huge number of tmem_objs, each pool manages a hashtable
++ * of rb_trees to reduce search, insert, delete, and rebalancing time.
++ * Each hashbucket also has a lock to manage concurrent access.
++ *
++ * The following routines manage tmem_objs.  When any tmem_obj is accessed,
++ * the hashbucket lock must be held.
++ */
++
++/* searches for object==oid in pool, returns locked object if found */
++static struct tmem_obj *tmem_obj_find(struct tmem_hashbucket *hb,
++					struct tmem_oid *oidp)
++{
++	struct rb_node *rbnode;
++	struct tmem_obj *obj;
++
++	rbnode = hb->obj_rb_root.rb_node;
++	while (rbnode) {
++		BUG_ON(RB_EMPTY_NODE(rbnode));
++		obj = rb_entry(rbnode, struct tmem_obj, rb_tree_node);
++		switch (tmem_oid_compare(oidp, &obj->oid)) {
++		case 0: /* equal */
++			goto out;
++		case -1:
++			rbnode = rbnode->rb_left;
++			break;
++		case 1:
++			rbnode = rbnode->rb_right;
++			break;
++		}
++	}
++	obj = NULL;
++out:
++	return obj;
++}
++
++static void tmem_pampd_destroy_all_in_obj(struct tmem_obj *);
++
++/* free an object that has no more pampds in it */
++static void tmem_obj_free(struct tmem_obj *obj, struct tmem_hashbucket *hb)
++{
++	struct tmem_pool *pool;
++
++	BUG_ON(obj == NULL);
++	ASSERT_SENTINEL(obj, OBJ);
++	BUG_ON(obj->pampd_count > 0);
++	pool = obj->pool;
++	BUG_ON(pool == NULL);
++	if (obj->objnode_tree_root != NULL) /* may be "stump" with no leaves */
++		tmem_pampd_destroy_all_in_obj(obj);
++	BUG_ON(obj->objnode_tree_root != NULL);
++	BUG_ON((long)obj->objnode_count != 0);
++	atomic_dec(&pool->obj_count);
++	BUG_ON(atomic_read(&pool->obj_count) < 0);
++	INVERT_SENTINEL(obj, OBJ);
++	obj->pool = NULL;
++	tmem_oid_set_invalid(&obj->oid);
++	rb_erase(&obj->rb_tree_node, &hb->obj_rb_root);
++}
++
++/*
++ * initialize, and insert an tmem_object_root (called only if find failed)
++ */
++static void tmem_obj_init(struct tmem_obj *obj, struct tmem_hashbucket *hb,
++					struct tmem_pool *pool,
++					struct tmem_oid *oidp)
++{
++	struct rb_root *root = &hb->obj_rb_root;
++	struct rb_node **new = &(root->rb_node), *parent = NULL;
++	struct tmem_obj *this;
++
++	BUG_ON(pool == NULL);
++	atomic_inc(&pool->obj_count);
++	obj->objnode_tree_height = 0;
++	obj->objnode_tree_root = NULL;
++	obj->pool = pool;
++	obj->oid = *oidp;
++	obj->objnode_count = 0;
++	obj->pampd_count = 0;
++	(*tmem_pamops.new_obj)(obj);
++	SET_SENTINEL(obj, OBJ);
++	while (*new) {
++		BUG_ON(RB_EMPTY_NODE(*new));
++		this = rb_entry(*new, struct tmem_obj, rb_tree_node);
++		parent = *new;
++		switch (tmem_oid_compare(oidp, &this->oid)) {
++		case 0:
++			BUG(); /* already present; should never happen! */
++			break;
++		case -1:
++			new = &(*new)->rb_left;
++			break;
++		case 1:
++			new = &(*new)->rb_right;
++			break;
++		}
++	}
++	rb_link_node(&obj->rb_tree_node, parent, new);
++	rb_insert_color(&obj->rb_tree_node, root);
++}
++
++/*
++ * Tmem is managed as a set of tmem_pools with certain attributes, such as
++ * "ephemeral" vs "persistent".  These attributes apply to all tmem_objs
++ * and all pampds that belong to a tmem_pool.  A tmem_pool is created
++ * or deleted relatively rarely (for example, when a filesystem is
++ * mounted or unmounted.
++ */
++
++/* flush all data from a pool and, optionally, free it */
++static void tmem_pool_flush(struct tmem_pool *pool, bool destroy)
++{
++	struct rb_node *rbnode;
++	struct tmem_obj *obj;
++	struct tmem_hashbucket *hb = &pool->hashbucket[0];
++	int i;
++
++	BUG_ON(pool == NULL);
++	for (i = 0; i < TMEM_HASH_BUCKETS; i++, hb++) {
++		spin_lock(&hb->lock);
++		rbnode = rb_first(&hb->obj_rb_root);
++		while (rbnode != NULL) {
++			obj = rb_entry(rbnode, struct tmem_obj, rb_tree_node);
++			rbnode = rb_next(rbnode);
++			tmem_pampd_destroy_all_in_obj(obj);
++			tmem_obj_free(obj, hb);
++			(*tmem_hostops.obj_free)(obj, pool);
++		}
++		spin_unlock(&hb->lock);
++	}
++	if (destroy)
++		list_del(&pool->pool_list);
++}
++
++/*
++ * A tmem_obj contains a radix-tree-like tree in which the intermediate
++ * nodes are called tmem_objnodes.  (The kernel lib/radix-tree.c implementation
++ * is very specialized and tuned for specific uses and is not particularly
++ * suited for use from this code, though some code from the core algorithms has
++ * been reused, thus the copyright notices below).  Each tmem_objnode contains
++ * a set of pointers which point to either a set of intermediate tmem_objnodes
++ * or a set of of pampds.
++ *
++ * Portions Copyright (C) 2001 Momchil Velikov
++ * Portions Copyright (C) 2001 Christoph Hellwig
++ * Portions Copyright (C) 2005 SGI, Christoph Lameter <clameter@sgi.com>
++ */
++
++struct tmem_objnode_tree_path {
++	struct tmem_objnode *objnode;
++	int offset;
++};
++
++/* objnode height_to_maxindex translation */
++static unsigned long tmem_objnode_tree_h2max[OBJNODE_TREE_MAX_PATH + 1];
++
++static void tmem_objnode_tree_init(void)
++{
++	unsigned int ht, tmp;
++
++	for (ht = 0; ht < ARRAY_SIZE(tmem_objnode_tree_h2max); ht++) {
++		tmp = ht * OBJNODE_TREE_MAP_SHIFT;
++		if (tmp >= OBJNODE_TREE_INDEX_BITS)
++			tmem_objnode_tree_h2max[ht] = ~0UL;
++		else
++			tmem_objnode_tree_h2max[ht] =
++			    (~0UL >> (OBJNODE_TREE_INDEX_BITS - tmp - 1)) >> 1;
++	}
++}
++
++static struct tmem_objnode *tmem_objnode_alloc(struct tmem_obj *obj)
++{
++	struct tmem_objnode *objnode;
++
++	ASSERT_SENTINEL(obj, OBJ);
++	BUG_ON(obj->pool == NULL);
++	ASSERT_SENTINEL(obj->pool, POOL);
++	objnode = (*tmem_hostops.objnode_alloc)(obj->pool);
++	if (unlikely(objnode == NULL))
++		goto out;
++	objnode->obj = obj;
++	SET_SENTINEL(objnode, OBJNODE);
++	memset(&objnode->slots, 0, sizeof(objnode->slots));
++	objnode->slots_in_use = 0;
++	obj->objnode_count++;
++out:
++	return objnode;
++}
++
++static void tmem_objnode_free(struct tmem_objnode *objnode)
++{
++	struct tmem_pool *pool;
++	int i;
++
++	BUG_ON(objnode == NULL);
++	for (i = 0; i < OBJNODE_TREE_MAP_SIZE; i++)
++		BUG_ON(objnode->slots[i] != NULL);
++	ASSERT_SENTINEL(objnode, OBJNODE);
++	INVERT_SENTINEL(objnode, OBJNODE);
++	BUG_ON(objnode->obj == NULL);
++	ASSERT_SENTINEL(objnode->obj, OBJ);
++	pool = objnode->obj->pool;
++	BUG_ON(pool == NULL);
++	ASSERT_SENTINEL(pool, POOL);
++	objnode->obj->objnode_count--;
++	objnode->obj = NULL;
++	(*tmem_hostops.objnode_free)(objnode, pool);
++}
++
++/*
++ * lookup index in object and return associated pampd (or NULL if not found)
++ */
++static void **__tmem_pampd_lookup_in_obj(struct tmem_obj *obj, uint32_t index)
++{
++	unsigned int height, shift;
++	struct tmem_objnode **slot = NULL;
++
++	BUG_ON(obj == NULL);
++	ASSERT_SENTINEL(obj, OBJ);
++	BUG_ON(obj->pool == NULL);
++	ASSERT_SENTINEL(obj->pool, POOL);
++
++	height = obj->objnode_tree_height;
++	if (index > tmem_objnode_tree_h2max[obj->objnode_tree_height])
++		goto out;
++	if (height == 0 && obj->objnode_tree_root) {
++		slot = &obj->objnode_tree_root;
++		goto out;
++	}
++	shift = (height-1) * OBJNODE_TREE_MAP_SHIFT;
++	slot = &obj->objnode_tree_root;
++	while (height > 0) {
++		if (*slot == NULL)
++			goto out;
++		slot = (struct tmem_objnode **)
++			((*slot)->slots +
++			 ((index >> shift) & OBJNODE_TREE_MAP_MASK));
++		shift -= OBJNODE_TREE_MAP_SHIFT;
++		height--;
++	}
++out:
++	return slot != NULL ? (void **)slot : NULL;
++}
++
++static void *tmem_pampd_lookup_in_obj(struct tmem_obj *obj, uint32_t index)
++{
++	struct tmem_objnode **slot;
++
++	slot = (struct tmem_objnode **)__tmem_pampd_lookup_in_obj(obj, index);
++	return slot != NULL ? *slot : NULL;
++}
++
++static void *tmem_pampd_replace_in_obj(struct tmem_obj *obj, uint32_t index,
++					void *new_pampd)
++{
++	struct tmem_objnode **slot;
++	void *ret = NULL;
++
++	slot = (struct tmem_objnode **)__tmem_pampd_lookup_in_obj(obj, index);
++	if ((slot != NULL) && (*slot != NULL)) {
++		void *old_pampd = *(void **)slot;
++		*(void **)slot = new_pampd;
++		(*tmem_pamops.free)(old_pampd, obj->pool, NULL, 0);
++		ret = new_pampd;
++	}
++	return ret;
++}
++
++static int tmem_pampd_add_to_obj(struct tmem_obj *obj, uint32_t index,
++					void *pampd)
++{
++	int ret = 0;
++	struct tmem_objnode *objnode = NULL, *newnode, *slot;
++	unsigned int height, shift;
++	int offset = 0;
++
++	/* if necessary, extend the tree to be higher  */
++	if (index > tmem_objnode_tree_h2max[obj->objnode_tree_height]) {
++		height = obj->objnode_tree_height + 1;
++		if (index > tmem_objnode_tree_h2max[height])
++			while (index > tmem_objnode_tree_h2max[height])
++				height++;
++		if (obj->objnode_tree_root == NULL) {
++			obj->objnode_tree_height = height;
++			goto insert;
++		}
++		do {
++			newnode = tmem_objnode_alloc(obj);
++			if (!newnode) {
++				ret = -ENOMEM;
++				goto out;
++			}
++			newnode->slots[0] = obj->objnode_tree_root;
++			newnode->slots_in_use = 1;
++			obj->objnode_tree_root = newnode;
++			obj->objnode_tree_height++;
++		} while (height > obj->objnode_tree_height);
++	}
++insert:
++	slot = obj->objnode_tree_root;
++	height = obj->objnode_tree_height;
++	shift = (height-1) * OBJNODE_TREE_MAP_SHIFT;
++	while (height > 0) {
++		if (slot == NULL) {
++			/* add a child objnode.  */
++			slot = tmem_objnode_alloc(obj);
++			if (!slot) {
++				ret = -ENOMEM;
++				goto out;
++			}
++			if (objnode) {
++
++				objnode->slots[offset] = slot;
++				objnode->slots_in_use++;
++			} else
++				obj->objnode_tree_root = slot;
++		}
++		/* go down a level */
++		offset = (index >> shift) & OBJNODE_TREE_MAP_MASK;
++		objnode = slot;
++		slot = objnode->slots[offset];
++		shift -= OBJNODE_TREE_MAP_SHIFT;
++		height--;
++	}
++	BUG_ON(slot != NULL);
++	if (objnode) {
++		objnode->slots_in_use++;
++		objnode->slots[offset] = pampd;
++	} else
++		obj->objnode_tree_root = pampd;
++	obj->pampd_count++;
++out:
++	return ret;
++}
++
++static void *tmem_pampd_delete_from_obj(struct tmem_obj *obj, uint32_t index)
++{
++	struct tmem_objnode_tree_path path[OBJNODE_TREE_MAX_PATH + 1];
++	struct tmem_objnode_tree_path *pathp = path;
++	struct tmem_objnode *slot = NULL;
++	unsigned int height, shift;
++	int offset;
++
++	BUG_ON(obj == NULL);
++	ASSERT_SENTINEL(obj, OBJ);
++	BUG_ON(obj->pool == NULL);
++	ASSERT_SENTINEL(obj->pool, POOL);
++	height = obj->objnode_tree_height;
++	if (index > tmem_objnode_tree_h2max[height])
++		goto out;
++	slot = obj->objnode_tree_root;
++	if (height == 0 && obj->objnode_tree_root) {
++		obj->objnode_tree_root = NULL;
++		goto out;
++	}
++	shift = (height - 1) * OBJNODE_TREE_MAP_SHIFT;
++	pathp->objnode = NULL;
++	do {
++		if (slot == NULL)
++			goto out;
++		pathp++;
++		offset = (index >> shift) & OBJNODE_TREE_MAP_MASK;
++		pathp->offset = offset;
++		pathp->objnode = slot;
++		slot = slot->slots[offset];
++		shift -= OBJNODE_TREE_MAP_SHIFT;
++		height--;
++	} while (height > 0);
++	if (slot == NULL)
++		goto out;
++	while (pathp->objnode) {
++		pathp->objnode->slots[pathp->offset] = NULL;
++		pathp->objnode->slots_in_use--;
++		if (pathp->objnode->slots_in_use) {
++			if (pathp->objnode == obj->objnode_tree_root) {
++				while (obj->objnode_tree_height > 0 &&
++				  obj->objnode_tree_root->slots_in_use == 1 &&
++				  obj->objnode_tree_root->slots[0]) {
++					struct tmem_objnode *to_free =
++						obj->objnode_tree_root;
++
++					obj->objnode_tree_root =
++							to_free->slots[0];
++					obj->objnode_tree_height--;
++					to_free->slots[0] = NULL;
++					to_free->slots_in_use = 0;
++					tmem_objnode_free(to_free);
++				}
++			}
++			goto out;
++		}
++		tmem_objnode_free(pathp->objnode); /* 0 slots used, free it */
++		pathp--;
++	}
++	obj->objnode_tree_height = 0;
++	obj->objnode_tree_root = NULL;
++
++out:
++	if (slot != NULL)
++		obj->pampd_count--;
++	BUG_ON(obj->pampd_count < 0);
++	return slot;
++}
++
++/* recursively walk the objnode_tree destroying pampds and objnodes */
++static void tmem_objnode_node_destroy(struct tmem_obj *obj,
++					struct tmem_objnode *objnode,
++					unsigned int ht)
++{
++	int i;
++
++	if (ht == 0)
++		return;
++	for (i = 0; i < OBJNODE_TREE_MAP_SIZE; i++) {
++		if (objnode->slots[i]) {
++			if (ht == 1) {
++				obj->pampd_count--;
++				(*tmem_pamops.free)(objnode->slots[i],
++						obj->pool, NULL, 0);
++				objnode->slots[i] = NULL;
++				continue;
++			}
++			tmem_objnode_node_destroy(obj, objnode->slots[i], ht-1);
++			tmem_objnode_free(objnode->slots[i]);
++			objnode->slots[i] = NULL;
++		}
++	}
++}
++
++static void tmem_pampd_destroy_all_in_obj(struct tmem_obj *obj)
++{
++	if (obj->objnode_tree_root == NULL)
++		return;
++	if (obj->objnode_tree_height == 0) {
++		obj->pampd_count--;
++		(*tmem_pamops.free)(obj->objnode_tree_root, obj->pool, NULL, 0);
++	} else {
++		tmem_objnode_node_destroy(obj, obj->objnode_tree_root,
++					obj->objnode_tree_height);
++		tmem_objnode_free(obj->objnode_tree_root);
++		obj->objnode_tree_height = 0;
++	}
++	obj->objnode_tree_root = NULL;
++	(*tmem_pamops.free_obj)(obj->pool, obj);
++}
++
++/*
++ * Tmem is operated on by a set of well-defined actions:
++ * "put", "get", "flush", "flush_object", "new pool" and "destroy pool".
++ * (The tmem ABI allows for subpages and exchanges but these operations
++ * are not included in this implementation.)
++ *
++ * These "tmem core" operations are implemented in the following functions.
++ */
++
++/*
++ * "Put" a page, e.g. copy a page from the kernel into newly allocated
++ * PAM space (if such space is available).  Tmem_put is complicated by
++ * a corner case: What if a page with matching handle already exists in
++ * tmem?  To guarantee coherency, one of two actions is necessary: Either
++ * the data for the page must be overwritten, or the page must be
++ * "flushed" so that the data is not accessible to a subsequent "get".
++ * Since these "duplicate puts" are relatively rare, this implementation
++ * always flushes for simplicity.
++ */
++int tmem_put(struct tmem_pool *pool, struct tmem_oid *oidp, uint32_t index,
++		char *data, size_t size, bool raw, bool ephemeral)
++{
++	struct tmem_obj *obj = NULL, *objfound = NULL, *objnew = NULL;
++	void *pampd = NULL, *pampd_del = NULL;
++	int ret = -ENOMEM;
++	struct tmem_hashbucket *hb;
++
++	hb = &pool->hashbucket[tmem_oid_hash(oidp)];
++	spin_lock(&hb->lock);
++	obj = objfound = tmem_obj_find(hb, oidp);
++	if (obj != NULL) {
++		pampd = tmem_pampd_lookup_in_obj(objfound, index);
++		if (pampd != NULL) {
++			/* if found, is a dup put, flush the old one */
++			pampd_del = tmem_pampd_delete_from_obj(obj, index);
++			BUG_ON(pampd_del != pampd);
++			(*tmem_pamops.free)(pampd, pool, oidp, index);
++			if (obj->pampd_count == 0) {
++				objnew = obj;
++				objfound = NULL;
++			}
++			pampd = NULL;
++		}
++	} else {
++		obj = objnew = (*tmem_hostops.obj_alloc)(pool);
++		if (unlikely(obj == NULL)) {
++			ret = -ENOMEM;
++			goto out;
++		}
++		tmem_obj_init(obj, hb, pool, oidp);
++	}
++	BUG_ON(obj == NULL);
++	BUG_ON(((objnew != obj) && (objfound != obj)) || (objnew == objfound));
++	pampd = (*tmem_pamops.create)(data, size, raw, ephemeral,
++					obj->pool, &obj->oid, index);
++	if (unlikely(pampd == NULL))
++		goto free;
++	ret = tmem_pampd_add_to_obj(obj, index, pampd);
++	if (unlikely(ret == -ENOMEM))
++		/* may have partially built objnode tree ("stump") */
++		goto delete_and_free;
++	goto out;
++
++delete_and_free:
++	(void)tmem_pampd_delete_from_obj(obj, index);
++free:
++	if (pampd)
++		(*tmem_pamops.free)(pampd, pool, NULL, 0);
++	if (objnew) {
++		tmem_obj_free(objnew, hb);
++		(*tmem_hostops.obj_free)(objnew, pool);
++	}
++out:
++	spin_unlock(&hb->lock);
++	return ret;
++}
++
++/*
++ * "Get" a page, e.g. if one can be found, copy the tmem page with the
++ * matching handle from PAM space to the kernel.  By tmem definition,
++ * when a "get" is successful on an ephemeral page, the page is "flushed",
++ * and when a "get" is successful on a persistent page, the page is retained
++ * in tmem.  Note that to preserve
++ * coherency, "get" can never be skipped if tmem contains the data.
++ * That is, if a get is done with a certain handle and fails, any
++ * subsequent "get" must also fail (unless of course there is a
++ * "put" done with the same handle).
++
++ */
++int tmem_get(struct tmem_pool *pool, struct tmem_oid *oidp, uint32_t index,
++		char *data, size_t *size, bool raw, int get_and_free)
++{
++	struct tmem_obj *obj;
++	void *pampd;
++	bool ephemeral = is_ephemeral(pool);
++	int ret = -1;
++	struct tmem_hashbucket *hb;
++	bool free = (get_and_free == 1) || ((get_and_free == 0) && ephemeral);
++	bool lock_held = false;
++
++	hb = &pool->hashbucket[tmem_oid_hash(oidp)];
++	spin_lock(&hb->lock);
++	lock_held = true;
++	obj = tmem_obj_find(hb, oidp);
++	if (obj == NULL)
++		goto out;
++	if (free)
++		pampd = tmem_pampd_delete_from_obj(obj, index);
++	else
++		pampd = tmem_pampd_lookup_in_obj(obj, index);
++	if (pampd == NULL)
++		goto out;
++	if (free) {
++		if (obj->pampd_count == 0) {
++			tmem_obj_free(obj, hb);
++			(*tmem_hostops.obj_free)(obj, pool);
++			obj = NULL;
++		}
++	}
++	if (tmem_pamops.is_remote(pampd)) {
++		lock_held = false;
++		spin_unlock(&hb->lock);
++	}
++	if (free)
++		ret = (*tmem_pamops.get_data_and_free)(
++				data, size, raw, pampd, pool, oidp, index);
++	else
++		ret = (*tmem_pamops.get_data)(
++				data, size, raw, pampd, pool, oidp, index);
++	if (ret < 0)
++		goto out;
++	ret = 0;
++out:
++	if (lock_held)
++		spin_unlock(&hb->lock);
++	return ret;
++}
++
++/*
++ * If a page in tmem matches the handle, "flush" this page from tmem such
++ * that any subsequent "get" does not succeed (unless, of course, there
++ * was another "put" with the same handle).
++ */
++int tmem_flush_page(struct tmem_pool *pool,
++				struct tmem_oid *oidp, uint32_t index)
++{
++	struct tmem_obj *obj;
++	void *pampd;
++	int ret = -1;
++	struct tmem_hashbucket *hb;
++
++	hb = &pool->hashbucket[tmem_oid_hash(oidp)];
++	spin_lock(&hb->lock);
++	obj = tmem_obj_find(hb, oidp);
++	if (obj == NULL)
++		goto out;
++	pampd = tmem_pampd_delete_from_obj(obj, index);
++	if (pampd == NULL)
++		goto out;
++	(*tmem_pamops.free)(pampd, pool, oidp, index);
++	if (obj->pampd_count == 0) {
++		tmem_obj_free(obj, hb);
++		(*tmem_hostops.obj_free)(obj, pool);
++	}
++	ret = 0;
++
++out:
++	spin_unlock(&hb->lock);
++	return ret;
++}
++
++/*
++ * If a page in tmem matches the handle, replace the page so that any
++ * subsequent "get" gets the new page.  Returns 0 if
++ * there was a page to replace, else returns -1.
++ */
++int tmem_replace(struct tmem_pool *pool, struct tmem_oid *oidp,
++			uint32_t index, void *new_pampd)
++{
++	struct tmem_obj *obj;
++	int ret = -1;
++	struct tmem_hashbucket *hb;
++
++	hb = &pool->hashbucket[tmem_oid_hash(oidp)];
++	spin_lock(&hb->lock);
++	obj = tmem_obj_find(hb, oidp);
++	if (obj == NULL)
++		goto out;
++	new_pampd = tmem_pampd_replace_in_obj(obj, index, new_pampd);
++	ret = (*tmem_pamops.replace_in_obj)(new_pampd, obj);
++out:
++	spin_unlock(&hb->lock);
++	return ret;
++}
++
++/*
++ * "Flush" all pages in tmem matching this oid.
++ */
++int tmem_flush_object(struct tmem_pool *pool, struct tmem_oid *oidp)
++{
++	struct tmem_obj *obj;
++	struct tmem_hashbucket *hb;
++	int ret = -1;
++
++	hb = &pool->hashbucket[tmem_oid_hash(oidp)];
++	spin_lock(&hb->lock);
++	obj = tmem_obj_find(hb, oidp);
++	if (obj == NULL)
++		goto out;
++	tmem_pampd_destroy_all_in_obj(obj);
++	tmem_obj_free(obj, hb);
++	(*tmem_hostops.obj_free)(obj, pool);
++	ret = 0;
++
++out:
++	spin_unlock(&hb->lock);
++	return ret;
++}
++
++/*
++ * "Flush" all pages (and tmem_objs) from this tmem_pool and disable
++ * all subsequent access to this tmem_pool.
++ */
++int tmem_destroy_pool(struct tmem_pool *pool)
++{
++	int ret = -1;
++
++	if (pool == NULL)
++		goto out;
++	tmem_pool_flush(pool, 1);
++	ret = 0;
++out:
++	return ret;
++}
++
++static LIST_HEAD(tmem_global_pool_list);
++
++/*
++ * Create a new tmem_pool with the provided flag and return
++ * a pool id provided by the tmem host implementation.
++ */
++void tmem_new_pool(struct tmem_pool *pool, uint32_t flags)
++{
++	int persistent = flags & TMEM_POOL_PERSIST;
++	int shared = flags & TMEM_POOL_SHARED;
++	struct tmem_hashbucket *hb = &pool->hashbucket[0];
++	int i;
++
++	for (i = 0; i < TMEM_HASH_BUCKETS; i++, hb++) {
++		hb->obj_rb_root = RB_ROOT;
++		spin_lock_init(&hb->lock);
++	}
++	INIT_LIST_HEAD(&pool->pool_list);
++	atomic_set(&pool->obj_count, 0);
++	SET_SENTINEL(pool, POOL);
++	list_add_tail(&pool->pool_list, &tmem_global_pool_list);
++	pool->persistent = persistent;
++	pool->shared = shared;
++}
+diff --git a/drivers/staging/ramster/tmem.h b/drivers/staging/ramster/tmem.h
+new file mode 100644
+index 0000000..ed147c4
+--- /dev/null
++++ b/drivers/staging/ramster/tmem.h
+@@ -0,0 +1,206 @@
++/*
++ * tmem.h
++ *
++ * Transcendent memory
++ *
++ * Copyright (c) 2009-2011, Dan Magenheimer, Oracle Corp.
++ */
++
++#ifndef _TMEM_H_
++#define _TMEM_H_
++
++#include <linux/types.h>
++#include <linux/highmem.h>
++#include <linux/hash.h>
++#include <linux/atomic.h>
++
++/*
++ * These are pre-defined by the Xen<->Linux ABI
++ */
++#define TMEM_PUT_PAGE			4
++#define TMEM_GET_PAGE			5
++#define TMEM_FLUSH_PAGE			6
++#define TMEM_FLUSH_OBJECT		7
++#define TMEM_POOL_PERSIST		1
++#define TMEM_POOL_SHARED		2
++#define TMEM_POOL_PRECOMPRESSED		4
++#define TMEM_POOL_PAGESIZE_SHIFT	4
++#define TMEM_POOL_PAGESIZE_MASK		0xf
++#define TMEM_POOL_RESERVED_BITS		0x00ffff00
++
++/*
++ * sentinels have proven very useful for debugging but can be removed
++ * or disabled before final merge.
++ */
++#define SENTINELS
++#ifdef SENTINELS
++#define DECL_SENTINEL uint32_t sentinel;
++#define SET_SENTINEL(_x, _y) (_x->sentinel = _y##_SENTINEL)
++#define INVERT_SENTINEL(_x, _y) (_x->sentinel = ~_y##_SENTINEL)
++#define ASSERT_SENTINEL(_x, _y) WARN_ON(_x->sentinel != _y##_SENTINEL)
++#define ASSERT_INVERTED_SENTINEL(_x, _y) WARN_ON(_x->sentinel != ~_y##_SENTINEL)
++#else
++#define DECL_SENTINEL
++#define SET_SENTINEL(_x, _y) do { } while (0)
++#define INVERT_SENTINEL(_x, _y) do { } while (0)
++#define ASSERT_SENTINEL(_x, _y) do { } while (0)
++#define ASSERT_INVERTED_SENTINEL(_x, _y) do { } while (0)
++#endif
++
++#define ASSERT_SPINLOCK(_l)	WARN_ON(!spin_is_locked(_l))
++
++/*
++ * A pool is the highest-level data structure managed by tmem and
++ * usually corresponds to a large independent set of pages such as
++ * a filesystem.  Each pool has an id, and certain attributes and counters.
++ * It also contains a set of hash buckets, each of which contains an rbtree
++ * of objects and a lock to manage concurrency within the pool.
++ */
++
++#define TMEM_HASH_BUCKET_BITS	8
++#define TMEM_HASH_BUCKETS	(1<<TMEM_HASH_BUCKET_BITS)
++
++struct tmem_hashbucket {
++	struct rb_root obj_rb_root;
++	spinlock_t lock;
++};
++
++struct tmem_pool {
++	void *client; /* "up" for some clients, avoids table lookup */
++	struct list_head pool_list;
++	uint32_t pool_id;
++	bool persistent;
++	bool shared;
++	atomic_t obj_count;
++	atomic_t refcount;
++	struct tmem_hashbucket hashbucket[TMEM_HASH_BUCKETS];
++	DECL_SENTINEL
++};
++
++#define is_persistent(_p)  (_p->persistent)
++#define is_ephemeral(_p)   (!(_p->persistent))
++
++/*
++ * An object id ("oid") is large: 192-bits (to ensure, for example, files
++ * in a modern filesystem can be uniquely identified).
++ */
++
++struct tmem_oid {
++	uint64_t oid[3];
++};
++
++static inline void tmem_oid_set_invalid(struct tmem_oid *oidp)
++{
++	oidp->oid[0] = oidp->oid[1] = oidp->oid[2] = -1UL;
++}
++
++static inline bool tmem_oid_valid(struct tmem_oid *oidp)
++{
++	return oidp->oid[0] != -1UL || oidp->oid[1] != -1UL ||
++		oidp->oid[2] != -1UL;
++}
++
++static inline int tmem_oid_compare(struct tmem_oid *left,
++					struct tmem_oid *right)
++{
++	int ret;
++
++	if (left->oid[2] == right->oid[2]) {
++		if (left->oid[1] == right->oid[1]) {
++			if (left->oid[0] == right->oid[0])
++				ret = 0;
++			else if (left->oid[0] < right->oid[0])
++				ret = -1;
++			else
++				return 1;
++		} else if (left->oid[1] < right->oid[1])
++			ret = -1;
++		else
++			ret = 1;
++	} else if (left->oid[2] < right->oid[2])
++		ret = -1;
++	else
++		ret = 1;
++	return ret;
++}
++
++static inline unsigned tmem_oid_hash(struct tmem_oid *oidp)
++{
++	return hash_long(oidp->oid[0] ^ oidp->oid[1] ^ oidp->oid[2],
++				TMEM_HASH_BUCKET_BITS);
++}
++
++/*
++ * A tmem_obj contains an identifier (oid), pointers to the parent
++ * pool and the rb_tree to which it belongs, counters, and an ordered
++ * set of pampds, structured in a radix-tree-like tree.  The intermediate
++ * nodes of the tree are called tmem_objnodes.
++ */
++
++struct tmem_objnode;
++
++struct tmem_obj {
++	struct tmem_oid oid;
++	struct tmem_pool *pool;
++	struct rb_node rb_tree_node;
++	struct tmem_objnode *objnode_tree_root;
++	unsigned int objnode_tree_height;
++	unsigned long objnode_count;
++	long pampd_count;
++	void *extra; /* for private use by pampd implementation */
++	DECL_SENTINEL
++};
++
++#define OBJNODE_TREE_MAP_SHIFT 6
++#define OBJNODE_TREE_MAP_SIZE (1UL << OBJNODE_TREE_MAP_SHIFT)
++#define OBJNODE_TREE_MAP_MASK (OBJNODE_TREE_MAP_SIZE-1)
++#define OBJNODE_TREE_INDEX_BITS (8 /* CHAR_BIT */ * sizeof(unsigned long))
++#define OBJNODE_TREE_MAX_PATH \
++		(OBJNODE_TREE_INDEX_BITS/OBJNODE_TREE_MAP_SHIFT + 2)
++
++struct tmem_objnode {
++	struct tmem_obj *obj;
++	DECL_SENTINEL
++	void *slots[OBJNODE_TREE_MAP_SIZE];
++	unsigned int slots_in_use;
++};
++
++/* pampd abstract datatype methods provided by the PAM implementation */
++struct tmem_pamops {
++	void *(*create)(char *, size_t, bool, int,
++			struct tmem_pool *, struct tmem_oid *, uint32_t);
++	int (*get_data)(char *, size_t *, bool, void *, struct tmem_pool *,
++				struct tmem_oid *, uint32_t);
++	int (*get_data_and_free)(char *, size_t *, bool, void *,
++				struct tmem_pool *, struct tmem_oid *,
++				uint32_t);
++	void (*free)(void *, struct tmem_pool *, struct tmem_oid *, uint32_t);
++	void (*free_obj)(struct tmem_pool *, struct tmem_obj *);
++	bool (*is_remote)(void *);
++	void (*new_obj)(struct tmem_obj *);
++	int (*replace_in_obj)(void *, struct tmem_obj *);
++};
++extern void tmem_register_pamops(struct tmem_pamops *m);
++
++/* memory allocation methods provided by the host implementation */
++struct tmem_hostops {
++	struct tmem_obj *(*obj_alloc)(struct tmem_pool *);
++	void (*obj_free)(struct tmem_obj *, struct tmem_pool *);
++	struct tmem_objnode *(*objnode_alloc)(struct tmem_pool *);
++	void (*objnode_free)(struct tmem_objnode *, struct tmem_pool *);
++};
++extern void tmem_register_hostops(struct tmem_hostops *m);
++
++/* core tmem accessor functions */
++extern int tmem_put(struct tmem_pool *, struct tmem_oid *, uint32_t index,
++			char *, size_t, bool, bool);
++extern int tmem_get(struct tmem_pool *, struct tmem_oid *, uint32_t index,
++			char *, size_t *, bool, int);
++extern int tmem_replace(struct tmem_pool *, struct tmem_oid *, uint32_t index,
++			void *);
++extern int tmem_flush_page(struct tmem_pool *, struct tmem_oid *,
++			uint32_t index);
++extern int tmem_flush_object(struct tmem_pool *, struct tmem_oid *);
++extern int tmem_destroy_pool(struct tmem_pool *);
++extern void tmem_new_pool(struct tmem_pool *, uint32_t);
++#endif /* _TMEM_H */
+diff --git a/drivers/staging/ramster/zcache-main.c b/drivers/staging/ramster/zcache-main.c
+new file mode 100644
+index 0000000..cd0ed84
+--- /dev/null
++++ b/drivers/staging/ramster/zcache-main.c
+@@ -0,0 +1,2003 @@
++/*
++ * zcache.c
++ *
++ * Copyright (c) 2010,2011, Dan Magenheimer, Oracle Corp.
++ * Copyright (c) 2010,2011, Nitin Gupta
++ *
++ * Zcache provides an in-kernel "host implementation" for transcendent memory
++ * and, thus indirectly, for cleancache and frontswap.  Zcache includes two
++ * page-accessible memory [1] interfaces, both utilizing lzo1x compression:
++ * 1) "compression buddies" ("zbud") is used for ephemeral pages
++ * 2) xvmalloc is used for persistent pages.
++ * Xvmalloc (based on the TLSF allocator) has very low fragmentation
++ * so maximizes space efficiency, while zbud allows pairs (and potentially,
++ * in the future, more than a pair of) compressed pages to be closely linked
++ * so that reclaiming can be done via the kernel's physical-page-oriented
++ * "shrinker" interface.
++ *
++ * [1] For a definition of page-accessible memory (aka PAM), see:
++ *   http://marc.info/?l=linux-mm&m=127811271605009
++ */
++
++#include <linux/module.h>
++#include <linux/cpu.h>
++#include <linux/highmem.h>
++#include <linux/list.h>
++#include <linux/lzo.h>
++#include <linux/slab.h>
++#include <linux/spinlock.h>
++#include <linux/types.h>
++#include <linux/atomic.h>
++#include <linux/math64.h>
++#include "tmem.h"
++
++#include "../zram/xvmalloc.h" /* if built in drivers/staging */
++
++#if (!defined(CONFIG_CLEANCACHE) && !defined(CONFIG_FRONTSWAP))
++#error "zcache is useless without CONFIG_CLEANCACHE or CONFIG_FRONTSWAP"
++#endif
++#ifdef CONFIG_CLEANCACHE
++#include <linux/cleancache.h>
++#endif
++#ifdef CONFIG_FRONTSWAP
++#include <linux/frontswap.h>
++#endif
++
++#if 0
++/* this is more aggressive but may cause other problems? */
++#define ZCACHE_GFP_MASK	(GFP_ATOMIC | __GFP_NORETRY | __GFP_NOWARN)
++#else
++#define ZCACHE_GFP_MASK \
++	(__GFP_FS | __GFP_NORETRY | __GFP_NOWARN | __GFP_NOMEMALLOC)
++#endif
++
++#define MAX_POOLS_PER_CLIENT 16
++
++#define MAX_CLIENTS 16
++#define LOCAL_CLIENT ((uint16_t)-1)
++
++MODULE_LICENSE("GPL");
++
++struct zcache_client {
++	struct tmem_pool *tmem_pools[MAX_POOLS_PER_CLIENT];
++	struct xv_pool *xvpool;
++	bool allocated;
++	atomic_t refcount;
++};
++
++static struct zcache_client zcache_host;
++static struct zcache_client zcache_clients[MAX_CLIENTS];
++
++static inline uint16_t get_client_id_from_client(struct zcache_client *cli)
++{
++	BUG_ON(cli == NULL);
++	if (cli == &zcache_host)
++		return LOCAL_CLIENT;
++	return cli - &zcache_clients[0];
++}
++
++static inline bool is_local_client(struct zcache_client *cli)
++{
++	return cli == &zcache_host;
++}
++
++/**********
++ * Compression buddies ("zbud") provides for packing two (or, possibly
++ * in the future, more) compressed ephemeral pages into a single "raw"
++ * (physical) page and tracking them with data structures so that
++ * the raw pages can be easily reclaimed.
++ *
++ * A zbud page ("zbpg") is an aligned page containing a list_head,
++ * a lock, and two "zbud headers".  The remainder of the physical
++ * page is divided up into aligned 64-byte "chunks" which contain
++ * the compressed data for zero, one, or two zbuds.  Each zbpg
++ * resides on: (1) an "unused list" if it has no zbuds; (2) a
++ * "buddied" list if it is fully populated  with two zbuds; or
++ * (3) one of PAGE_SIZE/64 "unbuddied" lists indexed by how many chunks
++ * the one unbuddied zbud uses.  The data inside a zbpg cannot be
++ * read or written unless the zbpg's lock is held.
++ */
++
++#define ZBH_SENTINEL  0x43214321
++#define ZBPG_SENTINEL  0xdeadbeef
++
++#define ZBUD_MAX_BUDS 2
++
++struct zbud_hdr {
++	uint16_t client_id;
++	uint16_t pool_id;
++	struct tmem_oid oid;
++	uint32_t index;
++	uint16_t size; /* compressed size in bytes, zero means unused */
++	DECL_SENTINEL
++};
++
++struct zbud_page {
++	struct list_head bud_list;
++	spinlock_t lock;
++	struct zbud_hdr buddy[ZBUD_MAX_BUDS];
++	DECL_SENTINEL
++	/* followed by NUM_CHUNK aligned CHUNK_SIZE-byte chunks */
++};
++
++#define CHUNK_SHIFT	6
++#define CHUNK_SIZE	(1 << CHUNK_SHIFT)
++#define CHUNK_MASK	(~(CHUNK_SIZE-1))
++#define NCHUNKS		(((PAGE_SIZE - sizeof(struct zbud_page)) & \
++				CHUNK_MASK) >> CHUNK_SHIFT)
++#define MAX_CHUNK	(NCHUNKS-1)
++
++static struct {
++	struct list_head list;
++	unsigned count;
++} zbud_unbuddied[NCHUNKS];
++/* list N contains pages with N chunks USED and NCHUNKS-N unused */
++/* element 0 is never used but optimizing that isn't worth it */
++static unsigned long zbud_cumul_chunk_counts[NCHUNKS];
++
++struct list_head zbud_buddied_list;
++static unsigned long zcache_zbud_buddied_count;
++
++/* protects the buddied list and all unbuddied lists */
++static DEFINE_SPINLOCK(zbud_budlists_spinlock);
++
++static LIST_HEAD(zbpg_unused_list);
++static unsigned long zcache_zbpg_unused_list_count;
++
++/* protects the unused page list */
++static DEFINE_SPINLOCK(zbpg_unused_list_spinlock);
++
++static atomic_t zcache_zbud_curr_raw_pages;
++static atomic_t zcache_zbud_curr_zpages;
++static unsigned long zcache_zbud_curr_zbytes;
++static unsigned long zcache_zbud_cumul_zpages;
++static unsigned long zcache_zbud_cumul_zbytes;
++static unsigned long zcache_compress_poor;
++static unsigned long zcache_mean_compress_poor;
++
++/* forward references */
++static void *zcache_get_free_page(void);
++static void zcache_free_page(void *p);
++
++/*
++ * zbud helper functions
++ */
++
++static inline unsigned zbud_max_buddy_size(void)
++{
++	return MAX_CHUNK << CHUNK_SHIFT;
++}
++
++static inline unsigned zbud_size_to_chunks(unsigned size)
++{
++	BUG_ON(size == 0 || size > zbud_max_buddy_size());
++	return (size + CHUNK_SIZE - 1) >> CHUNK_SHIFT;
++}
++
++static inline int zbud_budnum(struct zbud_hdr *zh)
++{
++	unsigned offset = (unsigned long)zh & (PAGE_SIZE - 1);
++	struct zbud_page *zbpg = NULL;
++	unsigned budnum = -1U;
++	int i;
++
++	for (i = 0; i < ZBUD_MAX_BUDS; i++)
++		if (offset == offsetof(typeof(*zbpg), buddy[i])) {
++			budnum = i;
++			break;
++		}
++	BUG_ON(budnum == -1U);
++	return budnum;
++}
++
++static char *zbud_data(struct zbud_hdr *zh, unsigned size)
++{
++	struct zbud_page *zbpg;
++	char *p;
++	unsigned budnum;
++
++	ASSERT_SENTINEL(zh, ZBH);
++	budnum = zbud_budnum(zh);
++	BUG_ON(size == 0 || size > zbud_max_buddy_size());
++	zbpg = container_of(zh, struct zbud_page, buddy[budnum]);
++	ASSERT_SPINLOCK(&zbpg->lock);
++	p = (char *)zbpg;
++	if (budnum == 0)
++		p += ((sizeof(struct zbud_page) + CHUNK_SIZE - 1) &
++							CHUNK_MASK);
++	else if (budnum == 1)
++		p += PAGE_SIZE - ((size + CHUNK_SIZE - 1) & CHUNK_MASK);
++	return p;
++}
++
++/*
++ * zbud raw page management
++ */
++
++static struct zbud_page *zbud_alloc_raw_page(void)
++{
++	struct zbud_page *zbpg = NULL;
++	struct zbud_hdr *zh0, *zh1;
++	bool recycled = 0;
++
++	/* if any pages on the zbpg list, use one */
++	spin_lock(&zbpg_unused_list_spinlock);
++	if (!list_empty(&zbpg_unused_list)) {
++		zbpg = list_first_entry(&zbpg_unused_list,
++				struct zbud_page, bud_list);
++		list_del_init(&zbpg->bud_list);
++		zcache_zbpg_unused_list_count--;
++		recycled = 1;
++	}
++	spin_unlock(&zbpg_unused_list_spinlock);
++	if (zbpg == NULL)
++		/* none on zbpg list, try to get a kernel page */
++		zbpg = zcache_get_free_page();
++	if (likely(zbpg != NULL)) {
++		INIT_LIST_HEAD(&zbpg->bud_list);
++		zh0 = &zbpg->buddy[0]; zh1 = &zbpg->buddy[1];
++		spin_lock_init(&zbpg->lock);
++		if (recycled) {
++			ASSERT_INVERTED_SENTINEL(zbpg, ZBPG);
++			SET_SENTINEL(zbpg, ZBPG);
++			BUG_ON(zh0->size != 0 || tmem_oid_valid(&zh0->oid));
++			BUG_ON(zh1->size != 0 || tmem_oid_valid(&zh1->oid));
++		} else {
++			atomic_inc(&zcache_zbud_curr_raw_pages);
++			INIT_LIST_HEAD(&zbpg->bud_list);
++			SET_SENTINEL(zbpg, ZBPG);
++			zh0->size = 0; zh1->size = 0;
++			tmem_oid_set_invalid(&zh0->oid);
++			tmem_oid_set_invalid(&zh1->oid);
++		}
++	}
++	return zbpg;
++}
++
++static void zbud_free_raw_page(struct zbud_page *zbpg)
++{
++	struct zbud_hdr *zh0 = &zbpg->buddy[0], *zh1 = &zbpg->buddy[1];
++
++	ASSERT_SENTINEL(zbpg, ZBPG);
++	BUG_ON(!list_empty(&zbpg->bud_list));
++	ASSERT_SPINLOCK(&zbpg->lock);
++	BUG_ON(zh0->size != 0 || tmem_oid_valid(&zh0->oid));
++	BUG_ON(zh1->size != 0 || tmem_oid_valid(&zh1->oid));
++	INVERT_SENTINEL(zbpg, ZBPG);
++	spin_unlock(&zbpg->lock);
++	spin_lock(&zbpg_unused_list_spinlock);
++	list_add(&zbpg->bud_list, &zbpg_unused_list);
++	zcache_zbpg_unused_list_count++;
++	spin_unlock(&zbpg_unused_list_spinlock);
++}
++
++/*
++ * core zbud handling routines
++ */
++
++static unsigned zbud_free(struct zbud_hdr *zh)
++{
++	unsigned size;
++
++	ASSERT_SENTINEL(zh, ZBH);
++	BUG_ON(!tmem_oid_valid(&zh->oid));
++	size = zh->size;
++	BUG_ON(zh->size == 0 || zh->size > zbud_max_buddy_size());
++	zh->size = 0;
++	tmem_oid_set_invalid(&zh->oid);
++	INVERT_SENTINEL(zh, ZBH);
++	zcache_zbud_curr_zbytes -= size;
++	atomic_dec(&zcache_zbud_curr_zpages);
++	return size;
++}
++
++static void zbud_free_and_delist(struct zbud_hdr *zh)
++{
++	unsigned chunks;
++	struct zbud_hdr *zh_other;
++	unsigned budnum = zbud_budnum(zh), size;
++	struct zbud_page *zbpg =
++		container_of(zh, struct zbud_page, buddy[budnum]);
++
++	spin_lock(&zbpg->lock);
++	if (list_empty(&zbpg->bud_list)) {
++		/* ignore zombie page... see zbud_evict_pages() */
++		spin_unlock(&zbpg->lock);
++		return;
++	}
++	size = zbud_free(zh);
++	ASSERT_SPINLOCK(&zbpg->lock);
++	zh_other = &zbpg->buddy[(budnum == 0) ? 1 : 0];
++	if (zh_other->size == 0) { /* was unbuddied: unlist and free */
++		chunks = zbud_size_to_chunks(size) ;
++		spin_lock(&zbud_budlists_spinlock);
++		BUG_ON(list_empty(&zbud_unbuddied[chunks].list));
++		list_del_init(&zbpg->bud_list);
++		zbud_unbuddied[chunks].count--;
++		spin_unlock(&zbud_budlists_spinlock);
++		zbud_free_raw_page(zbpg);
++	} else { /* was buddied: move remaining buddy to unbuddied list */
++		chunks = zbud_size_to_chunks(zh_other->size) ;
++		spin_lock(&zbud_budlists_spinlock);
++		list_del_init(&zbpg->bud_list);
++		zcache_zbud_buddied_count--;
++		list_add_tail(&zbpg->bud_list, &zbud_unbuddied[chunks].list);
++		zbud_unbuddied[chunks].count++;
++		spin_unlock(&zbud_budlists_spinlock);
++		spin_unlock(&zbpg->lock);
++	}
++}
++
++static struct zbud_hdr *zbud_create(uint16_t client_id, uint16_t pool_id,
++					struct tmem_oid *oid,
++					uint32_t index, struct page *page,
++					void *cdata, unsigned size)
++{
++	struct zbud_hdr *zh0, *zh1, *zh = NULL;
++	struct zbud_page *zbpg = NULL, *ztmp;
++	unsigned nchunks;
++	char *to;
++	int i, found_good_buddy = 0;
++
++	nchunks = zbud_size_to_chunks(size) ;
++	for (i = MAX_CHUNK - nchunks + 1; i > 0; i--) {
++		spin_lock(&zbud_budlists_spinlock);
++		if (!list_empty(&zbud_unbuddied[i].list)) {
++			list_for_each_entry_safe(zbpg, ztmp,
++				    &zbud_unbuddied[i].list, bud_list) {
++				if (spin_trylock(&zbpg->lock)) {
++					found_good_buddy = i;
++					goto found_unbuddied;
++				}
++			}
++		}
++		spin_unlock(&zbud_budlists_spinlock);
++	}
++	/* didn't find a good buddy, try allocating a new page */
++	zbpg = zbud_alloc_raw_page();
++	if (unlikely(zbpg == NULL))
++		goto out;
++	/* ok, have a page, now compress the data before taking locks */
++	spin_lock(&zbpg->lock);
++	spin_lock(&zbud_budlists_spinlock);
++	list_add_tail(&zbpg->bud_list, &zbud_unbuddied[nchunks].list);
++	zbud_unbuddied[nchunks].count++;
++	zh = &zbpg->buddy[0];
++	goto init_zh;
++
++found_unbuddied:
++	ASSERT_SPINLOCK(&zbpg->lock);
++	zh0 = &zbpg->buddy[0]; zh1 = &zbpg->buddy[1];
++	BUG_ON(!((zh0->size == 0) ^ (zh1->size == 0)));
++	if (zh0->size != 0) { /* buddy0 in use, buddy1 is vacant */
++		ASSERT_SENTINEL(zh0, ZBH);
++		zh = zh1;
++	} else if (zh1->size != 0) { /* buddy1 in use, buddy0 is vacant */
++		ASSERT_SENTINEL(zh1, ZBH);
++		zh = zh0;
++	} else
++		BUG();
++	list_del_init(&zbpg->bud_list);
++	zbud_unbuddied[found_good_buddy].count--;
++	list_add_tail(&zbpg->bud_list, &zbud_buddied_list);
++	zcache_zbud_buddied_count++;
++
++init_zh:
++	SET_SENTINEL(zh, ZBH);
++	zh->size = size;
++	zh->index = index;
++	zh->oid = *oid;
++	zh->pool_id = pool_id;
++	zh->client_id = client_id;
++	/* can wait to copy the data until the list locks are dropped */
++	spin_unlock(&zbud_budlists_spinlock);
++
++	to = zbud_data(zh, size);
++	memcpy(to, cdata, size);
++	spin_unlock(&zbpg->lock);
++	zbud_cumul_chunk_counts[nchunks]++;
++	atomic_inc(&zcache_zbud_curr_zpages);
++	zcache_zbud_cumul_zpages++;
++	zcache_zbud_curr_zbytes += size;
++	zcache_zbud_cumul_zbytes += size;
++out:
++	return zh;
++}
++
++static int zbud_decompress(struct page *page, struct zbud_hdr *zh)
++{
++	struct zbud_page *zbpg;
++	unsigned budnum = zbud_budnum(zh);
++	size_t out_len = PAGE_SIZE;
++	char *to_va, *from_va;
++	unsigned size;
++	int ret = 0;
++
++	zbpg = container_of(zh, struct zbud_page, buddy[budnum]);
++	spin_lock(&zbpg->lock);
++	if (list_empty(&zbpg->bud_list)) {
++		/* ignore zombie page... see zbud_evict_pages() */
++		ret = -EINVAL;
++		goto out;
++	}
++	ASSERT_SENTINEL(zh, ZBH);
++	BUG_ON(zh->size == 0 || zh->size > zbud_max_buddy_size());
++	to_va = kmap_atomic(page, KM_USER0);
++	size = zh->size;
++	from_va = zbud_data(zh, size);
++	ret = lzo1x_decompress_safe(from_va, size, to_va, &out_len);
++	BUG_ON(ret != LZO_E_OK);
++	BUG_ON(out_len != PAGE_SIZE);
++	kunmap_atomic(to_va, KM_USER0);
++out:
++	spin_unlock(&zbpg->lock);
++	return ret;
++}
++
++/*
++ * The following routines handle shrinking of ephemeral pages by evicting
++ * pages "least valuable" first.
++ */
++
++static unsigned long zcache_evicted_raw_pages;
++static unsigned long zcache_evicted_buddied_pages;
++static unsigned long zcache_evicted_unbuddied_pages;
++
++static struct tmem_pool *zcache_get_pool_by_id(uint16_t cli_id,
++						uint16_t poolid);
++static void zcache_put_pool(struct tmem_pool *pool);
++
++/*
++ * Flush and free all zbuds in a zbpg, then free the pageframe
++ */
++static void zbud_evict_zbpg(struct zbud_page *zbpg)
++{
++	struct zbud_hdr *zh;
++	int i, j;
++	uint32_t pool_id[ZBUD_MAX_BUDS], client_id[ZBUD_MAX_BUDS];
++	uint32_t index[ZBUD_MAX_BUDS];
++	struct tmem_oid oid[ZBUD_MAX_BUDS];
++	struct tmem_pool *pool;
++
++	ASSERT_SPINLOCK(&zbpg->lock);
++	BUG_ON(!list_empty(&zbpg->bud_list));
++	for (i = 0, j = 0; i < ZBUD_MAX_BUDS; i++) {
++		zh = &zbpg->buddy[i];
++		if (zh->size) {
++			client_id[j] = zh->client_id;
++			pool_id[j] = zh->pool_id;
++			oid[j] = zh->oid;
++			index[j] = zh->index;
++			j++;
++			zbud_free(zh);
++		}
++	}
++	spin_unlock(&zbpg->lock);
++	for (i = 0; i < j; i++) {
++		pool = zcache_get_pool_by_id(client_id[i], pool_id[i]);
++		if (pool != NULL) {
++			tmem_flush_page(pool, &oid[i], index[i]);
++			zcache_put_pool(pool);
++		}
++	}
++	ASSERT_SENTINEL(zbpg, ZBPG);
++	spin_lock(&zbpg->lock);
++	zbud_free_raw_page(zbpg);
++}
++
++/*
++ * Free nr pages.  This code is funky because we want to hold the locks
++ * protecting various lists for as short a time as possible, and in some
++ * circumstances the list may change asynchronously when the list lock is
++ * not held.  In some cases we also trylock not only to avoid waiting on a
++ * page in use by another cpu, but also to avoid potential deadlock due to
++ * lock inversion.
++ */
++static void zbud_evict_pages(int nr)
++{
++	struct zbud_page *zbpg;
++	int i;
++
++	/* first try freeing any pages on unused list */
++retry_unused_list:
++	spin_lock_bh(&zbpg_unused_list_spinlock);
++	if (!list_empty(&zbpg_unused_list)) {
++		/* can't walk list here, since it may change when unlocked */
++		zbpg = list_first_entry(&zbpg_unused_list,
++				struct zbud_page, bud_list);
++		list_del_init(&zbpg->bud_list);
++		zcache_zbpg_unused_list_count--;
++		atomic_dec(&zcache_zbud_curr_raw_pages);
++		spin_unlock_bh(&zbpg_unused_list_spinlock);
++		zcache_free_page(zbpg);
++		zcache_evicted_raw_pages++;
++		if (--nr <= 0)
++			goto out;
++		goto retry_unused_list;
++	}
++	spin_unlock_bh(&zbpg_unused_list_spinlock);
++
++	/* now try freeing unbuddied pages, starting with least space avail */
++	for (i = 0; i < MAX_CHUNK; i++) {
++retry_unbud_list_i:
++		spin_lock_bh(&zbud_budlists_spinlock);
++		if (list_empty(&zbud_unbuddied[i].list)) {
++			spin_unlock_bh(&zbud_budlists_spinlock);
++			continue;
++		}
++		list_for_each_entry(zbpg, &zbud_unbuddied[i].list, bud_list) {
++			if (unlikely(!spin_trylock(&zbpg->lock)))
++				continue;
++			list_del_init(&zbpg->bud_list);
++			zbud_unbuddied[i].count--;
++			spin_unlock(&zbud_budlists_spinlock);
++			zcache_evicted_unbuddied_pages++;
++			/* want budlists unlocked when doing zbpg eviction */
++			zbud_evict_zbpg(zbpg);
++			local_bh_enable();
++			if (--nr <= 0)
++				goto out;
++			goto retry_unbud_list_i;
++		}
++		spin_unlock_bh(&zbud_budlists_spinlock);
++	}
++
++	/* as a last resort, free buddied pages */
++retry_bud_list:
++	spin_lock_bh(&zbud_budlists_spinlock);
++	if (list_empty(&zbud_buddied_list)) {
++		spin_unlock_bh(&zbud_budlists_spinlock);
++		goto out;
++	}
++	list_for_each_entry(zbpg, &zbud_buddied_list, bud_list) {
++		if (unlikely(!spin_trylock(&zbpg->lock)))
++			continue;
++		list_del_init(&zbpg->bud_list);
++		zcache_zbud_buddied_count--;
++		spin_unlock(&zbud_budlists_spinlock);
++		zcache_evicted_buddied_pages++;
++		/* want budlists unlocked when doing zbpg eviction */
++		zbud_evict_zbpg(zbpg);
++		local_bh_enable();
++		if (--nr <= 0)
++			goto out;
++		goto retry_bud_list;
++	}
++	spin_unlock_bh(&zbud_budlists_spinlock);
++out:
++	return;
++}
++
++static void zbud_init(void)
++{
++	int i;
++
++	INIT_LIST_HEAD(&zbud_buddied_list);
++	zcache_zbud_buddied_count = 0;
++	for (i = 0; i < NCHUNKS; i++) {
++		INIT_LIST_HEAD(&zbud_unbuddied[i].list);
++		zbud_unbuddied[i].count = 0;
++	}
++}
++
++#ifdef CONFIG_SYSFS
++/*
++ * These sysfs routines show a nice distribution of how many zbpg's are
++ * currently (and have ever been placed) in each unbuddied list.  It's fun
++ * to watch but can probably go away before final merge.
++ */
++static int zbud_show_unbuddied_list_counts(char *buf)
++{
++	int i;
++	char *p = buf;
++
++	for (i = 0; i < NCHUNKS; i++)
++		p += sprintf(p, "%u ", zbud_unbuddied[i].count);
++	return p - buf;
++}
++
++static int zbud_show_cumul_chunk_counts(char *buf)
++{
++	unsigned long i, chunks = 0, total_chunks = 0, sum_total_chunks = 0;
++	unsigned long total_chunks_lte_21 = 0, total_chunks_lte_32 = 0;
++	unsigned long total_chunks_lte_42 = 0;
++	char *p = buf;
++
++	for (i = 0; i < NCHUNKS; i++) {
++		p += sprintf(p, "%lu ", zbud_cumul_chunk_counts[i]);
++		chunks += zbud_cumul_chunk_counts[i];
++		total_chunks += zbud_cumul_chunk_counts[i];
++		sum_total_chunks += i * zbud_cumul_chunk_counts[i];
++		if (i == 21)
++			total_chunks_lte_21 = total_chunks;
++		if (i == 32)
++			total_chunks_lte_32 = total_chunks;
++		if (i == 42)
++			total_chunks_lte_42 = total_chunks;
++	}
++	p += sprintf(p, "<=21:%lu <=32:%lu <=42:%lu, mean:%lu\n",
++		total_chunks_lte_21, total_chunks_lte_32, total_chunks_lte_42,
++		chunks == 0 ? 0 : sum_total_chunks / chunks);
++	return p - buf;
++}
++#endif
++
++/**********
++ * This "zv" PAM implementation combines the TLSF-based xvMalloc
++ * with lzo1x compression to maximize the amount of data that can
++ * be packed into a physical page.
++ *
++ * Zv represents a PAM page with the index and object (plus a "size" value
++ * necessary for decompression) immediately preceding the compressed data.
++ */
++
++#define ZVH_SENTINEL  0x43214321
++
++struct zv_hdr {
++	uint32_t pool_id;
++	struct tmem_oid oid;
++	uint32_t index;
++	DECL_SENTINEL
++};
++
++/* rudimentary policy limits */
++/* total number of persistent pages may not exceed this percentage */
++static unsigned int zv_page_count_policy_percent = 75;
++/*
++ * byte count defining poor compression; pages with greater zsize will be
++ * rejected
++ */
++static unsigned int zv_max_zsize = (PAGE_SIZE / 8) * 7;
++/*
++ * byte count defining poor *mean* compression; pages with greater zsize
++ * will be rejected until sufficient better-compressed pages are accepted
++ * driving the man below this threshold
++ */
++static unsigned int zv_max_mean_zsize = (PAGE_SIZE / 8) * 5;
++
++static unsigned long zv_curr_dist_counts[NCHUNKS];
++static unsigned long zv_cumul_dist_counts[NCHUNKS];
++
++static struct zv_hdr *zv_create(struct xv_pool *xvpool, uint32_t pool_id,
++				struct tmem_oid *oid, uint32_t index,
++				void *cdata, unsigned clen)
++{
++	struct page *page;
++	struct zv_hdr *zv = NULL;
++	uint32_t offset;
++	int alloc_size = clen + sizeof(struct zv_hdr);
++	int chunks = (alloc_size + (CHUNK_SIZE - 1)) >> CHUNK_SHIFT;
++	int ret;
++
++	BUG_ON(!irqs_disabled());
++	BUG_ON(chunks >= NCHUNKS);
++	ret = xv_malloc(xvpool, alloc_size,
++			&page, &offset, ZCACHE_GFP_MASK);
++	if (unlikely(ret))
++		goto out;
++	zv_curr_dist_counts[chunks]++;
++	zv_cumul_dist_counts[chunks]++;
++	zv = kmap_atomic(page, KM_USER0) + offset;
++	zv->index = index;
++	zv->oid = *oid;
++	zv->pool_id = pool_id;
++	SET_SENTINEL(zv, ZVH);
++	memcpy((char *)zv + sizeof(struct zv_hdr), cdata, clen);
++	kunmap_atomic(zv, KM_USER0);
++out:
++	return zv;
++}
++
++static void zv_free(struct xv_pool *xvpool, struct zv_hdr *zv)
++{
++	unsigned long flags;
++	struct page *page;
++	uint32_t offset;
++	uint16_t size = xv_get_object_size(zv);
++	int chunks = (size + (CHUNK_SIZE - 1)) >> CHUNK_SHIFT;
++
++	ASSERT_SENTINEL(zv, ZVH);
++	BUG_ON(chunks >= NCHUNKS);
++	zv_curr_dist_counts[chunks]--;
++	size -= sizeof(*zv);
++	BUG_ON(size == 0);
++	INVERT_SENTINEL(zv, ZVH);
++	page = virt_to_page(zv);
++	offset = (unsigned long)zv & ~PAGE_MASK;
++	local_irq_save(flags);
++	xv_free(xvpool, page, offset);
++	local_irq_restore(flags);
++}
++
++static void zv_decompress(struct page *page, struct zv_hdr *zv)
++{
++	size_t clen = PAGE_SIZE;
++	char *to_va;
++	unsigned size;
++	int ret;
++
++	ASSERT_SENTINEL(zv, ZVH);
++	size = xv_get_object_size(zv) - sizeof(*zv);
++	BUG_ON(size == 0);
++	to_va = kmap_atomic(page, KM_USER0);
++	ret = lzo1x_decompress_safe((char *)zv + sizeof(*zv),
++					size, to_va, &clen);
++	kunmap_atomic(to_va, KM_USER0);
++	BUG_ON(ret != LZO_E_OK);
++	BUG_ON(clen != PAGE_SIZE);
++}
++
++#ifdef CONFIG_SYSFS
++/*
++ * show a distribution of compression stats for zv pages.
++ */
++
++static int zv_curr_dist_counts_show(char *buf)
++{
++	unsigned long i, n, chunks = 0, sum_total_chunks = 0;
++	char *p = buf;
++
++	for (i = 0; i < NCHUNKS; i++) {
++		n = zv_curr_dist_counts[i];
++		p += sprintf(p, "%lu ", n);
++		chunks += n;
++		sum_total_chunks += i * n;
++	}
++	p += sprintf(p, "mean:%lu\n",
++		chunks == 0 ? 0 : sum_total_chunks / chunks);
++	return p - buf;
++}
++
++static int zv_cumul_dist_counts_show(char *buf)
++{
++	unsigned long i, n, chunks = 0, sum_total_chunks = 0;
++	char *p = buf;
++
++	for (i = 0; i < NCHUNKS; i++) {
++		n = zv_cumul_dist_counts[i];
++		p += sprintf(p, "%lu ", n);
++		chunks += n;
++		sum_total_chunks += i * n;
++	}
++	p += sprintf(p, "mean:%lu\n",
++		chunks == 0 ? 0 : sum_total_chunks / chunks);
++	return p - buf;
++}
++
++/*
++ * setting zv_max_zsize via sysfs causes all persistent (e.g. swap)
++ * pages that don't compress to less than this value (including metadata
++ * overhead) to be rejected.  We don't allow the value to get too close
++ * to PAGE_SIZE.
++ */
++static ssize_t zv_max_zsize_show(struct kobject *kobj,
++				    struct kobj_attribute *attr,
++				    char *buf)
++{
++	return sprintf(buf, "%u\n", zv_max_zsize);
++}
++
++static ssize_t zv_max_zsize_store(struct kobject *kobj,
++				    struct kobj_attribute *attr,
++				    const char *buf, size_t count)
++{
++	unsigned long val;
++	int err;
++
++	if (!capable(CAP_SYS_ADMIN))
++		return -EPERM;
++
++	err = strict_strtoul(buf, 10, &val);
++	if (err || (val == 0) || (val > (PAGE_SIZE / 8) * 7))
++		return -EINVAL;
++	zv_max_zsize = val;
++	return count;
++}
++
++/*
++ * setting zv_max_mean_zsize via sysfs causes all persistent (e.g. swap)
++ * pages that don't compress to less than this value (including metadata
++ * overhead) to be rejected UNLESS the mean compression is also smaller
++ * than this value.  In other words, we are load-balancing-by-zsize the
++ * accepted pages.  Again, we don't allow the value to get too close
++ * to PAGE_SIZE.
++ */
++static ssize_t zv_max_mean_zsize_show(struct kobject *kobj,
++				    struct kobj_attribute *attr,
++				    char *buf)
++{
++	return sprintf(buf, "%u\n", zv_max_mean_zsize);
++}
++
++static ssize_t zv_max_mean_zsize_store(struct kobject *kobj,
++				    struct kobj_attribute *attr,
++				    const char *buf, size_t count)
++{
++	unsigned long val;
++	int err;
++
++	if (!capable(CAP_SYS_ADMIN))
++		return -EPERM;
++
++	err = strict_strtoul(buf, 10, &val);
++	if (err || (val == 0) || (val > (PAGE_SIZE / 8) * 7))
++		return -EINVAL;
++	zv_max_mean_zsize = val;
++	return count;
++}
++
++/*
++ * setting zv_page_count_policy_percent via sysfs sets an upper bound of
++ * persistent (e.g. swap) pages that will be retained according to:
++ *     (zv_page_count_policy_percent * totalram_pages) / 100)
++ * when that limit is reached, further puts will be rejected (until
++ * some pages have been flushed).  Note that, due to compression,
++ * this number may exceed 100; it defaults to 75 and we set an
++ * arbitary limit of 150.  A poor choice will almost certainly result
++ * in OOM's, so this value should only be changed prudently.
++ */
++static ssize_t zv_page_count_policy_percent_show(struct kobject *kobj,
++						 struct kobj_attribute *attr,
++						 char *buf)
++{
++	return sprintf(buf, "%u\n", zv_page_count_policy_percent);
++}
++
++static ssize_t zv_page_count_policy_percent_store(struct kobject *kobj,
++						  struct kobj_attribute *attr,
++						  const char *buf, size_t count)
++{
++	unsigned long val;
++	int err;
++
++	if (!capable(CAP_SYS_ADMIN))
++		return -EPERM;
++
++	err = strict_strtoul(buf, 10, &val);
++	if (err || (val == 0) || (val > 150))
++		return -EINVAL;
++	zv_page_count_policy_percent = val;
++	return count;
++}
++
++static struct kobj_attribute zcache_zv_max_zsize_attr = {
++		.attr = { .name = "zv_max_zsize", .mode = 0644 },
++		.show = zv_max_zsize_show,
++		.store = zv_max_zsize_store,
++};
++
++static struct kobj_attribute zcache_zv_max_mean_zsize_attr = {
++		.attr = { .name = "zv_max_mean_zsize", .mode = 0644 },
++		.show = zv_max_mean_zsize_show,
++		.store = zv_max_mean_zsize_store,
++};
++
++static struct kobj_attribute zcache_zv_page_count_policy_percent_attr = {
++		.attr = { .name = "zv_page_count_policy_percent",
++			  .mode = 0644 },
++		.show = zv_page_count_policy_percent_show,
++		.store = zv_page_count_policy_percent_store,
++};
++#endif
++
++/*
++ * zcache core code starts here
++ */
++
++/* useful stats not collected by cleancache or frontswap */
++static unsigned long zcache_flush_total;
++static unsigned long zcache_flush_found;
++static unsigned long zcache_flobj_total;
++static unsigned long zcache_flobj_found;
++static unsigned long zcache_failed_eph_puts;
++static unsigned long zcache_failed_pers_puts;
++
++/*
++ * Tmem operations assume the poolid implies the invoking client.
++ * Zcache only has one client (the kernel itself): LOCAL_CLIENT.
++ * RAMster has each client numbered by cluster node, and a KVM version
++ * of zcache would have one client per guest and each client might
++ * have a poolid==N.
++ */
++static struct tmem_pool *zcache_get_pool_by_id(uint16_t cli_id, uint16_t poolid)
++{
++	struct tmem_pool *pool = NULL;
++	struct zcache_client *cli = NULL;
++
++	if (cli_id == LOCAL_CLIENT)
++		cli = &zcache_host;
++	else {
++		if (cli_id >= MAX_CLIENTS)
++			goto out;
++		cli = &zcache_clients[cli_id];
++		if (cli == NULL)
++			goto out;
++		atomic_inc(&cli->refcount);
++	}
++	if (poolid < MAX_POOLS_PER_CLIENT) {
++		pool = cli->tmem_pools[poolid];
++		if (pool != NULL)
++			atomic_inc(&pool->refcount);
++	}
++out:
++	return pool;
++}
++
++static void zcache_put_pool(struct tmem_pool *pool)
++{
++	struct zcache_client *cli = NULL;
++
++	if (pool == NULL)
++		BUG();
++	cli = pool->client;
++	atomic_dec(&pool->refcount);
++	atomic_dec(&cli->refcount);
++}
++
++int zcache_new_client(uint16_t cli_id)
++{
++	struct zcache_client *cli = NULL;
++	int ret = -1;
++
++	if (cli_id == LOCAL_CLIENT)
++		cli = &zcache_host;
++	else if ((unsigned int)cli_id < MAX_CLIENTS)
++		cli = &zcache_clients[cli_id];
++	if (cli == NULL)
++		goto out;
++	if (cli->allocated)
++		goto out;
++	cli->allocated = 1;
++#ifdef CONFIG_FRONTSWAP
++	cli->xvpool = xv_create_pool();
++	if (cli->xvpool == NULL)
++		goto out;
++#endif
++	ret = 0;
++out:
++	return ret;
++}
++
++/* counters for debugging */
++static unsigned long zcache_failed_get_free_pages;
++static unsigned long zcache_failed_alloc;
++static unsigned long zcache_put_to_flush;
++static unsigned long zcache_aborted_preload;
++static unsigned long zcache_aborted_shrink;
++
++/*
++ * Ensure that memory allocation requests in zcache don't result
++ * in direct reclaim requests via the shrinker, which would cause
++ * an infinite loop.  Maybe a GFP flag would be better?
++ */
++static DEFINE_SPINLOCK(zcache_direct_reclaim_lock);
++
++/*
++ * for now, used named slabs so can easily track usage; later can
++ * either just use kmalloc, or perhaps add a slab-like allocator
++ * to more carefully manage total memory utilization
++ */
++static struct kmem_cache *zcache_objnode_cache;
++static struct kmem_cache *zcache_obj_cache;
++static atomic_t zcache_curr_obj_count = ATOMIC_INIT(0);
++static unsigned long zcache_curr_obj_count_max;
++static atomic_t zcache_curr_objnode_count = ATOMIC_INIT(0);
++static unsigned long zcache_curr_objnode_count_max;
++
++/*
++ * to avoid memory allocation recursion (e.g. due to direct reclaim), we
++ * preload all necessary data structures so the hostops callbacks never
++ * actually do a malloc
++ */
++struct zcache_preload {
++	void *page;
++	struct tmem_obj *obj;
++	int nr;
++	struct tmem_objnode *objnodes[OBJNODE_TREE_MAX_PATH];
++};
++static DEFINE_PER_CPU(struct zcache_preload, zcache_preloads) = { 0, };
++
++static int zcache_do_preload(struct tmem_pool *pool)
++{
++	struct zcache_preload *kp;
++	struct tmem_objnode *objnode;
++	struct tmem_obj *obj;
++	void *page;
++	int ret = -ENOMEM;
++
++	if (unlikely(zcache_objnode_cache == NULL))
++		goto out;
++	if (unlikely(zcache_obj_cache == NULL))
++		goto out;
++	if (!spin_trylock(&zcache_direct_reclaim_lock)) {
++		zcache_aborted_preload++;
++		goto out;
++	}
++	preempt_disable();
++	kp = &__get_cpu_var(zcache_preloads);
++	while (kp->nr < ARRAY_SIZE(kp->objnodes)) {
++		preempt_enable_no_resched();
++		objnode = kmem_cache_alloc(zcache_objnode_cache,
++				ZCACHE_GFP_MASK);
++		if (unlikely(objnode == NULL)) {
++			zcache_failed_alloc++;
++			goto unlock_out;
++		}
++		preempt_disable();
++		kp = &__get_cpu_var(zcache_preloads);
++		if (kp->nr < ARRAY_SIZE(kp->objnodes))
++			kp->objnodes[kp->nr++] = objnode;
++		else
++			kmem_cache_free(zcache_objnode_cache, objnode);
++	}
++	preempt_enable_no_resched();
++	obj = kmem_cache_alloc(zcache_obj_cache, ZCACHE_GFP_MASK);
++	if (unlikely(obj == NULL)) {
++		zcache_failed_alloc++;
++		goto unlock_out;
++	}
++	page = (void *)__get_free_page(ZCACHE_GFP_MASK);
++	if (unlikely(page == NULL)) {
++		zcache_failed_get_free_pages++;
++		kmem_cache_free(zcache_obj_cache, obj);
++		goto unlock_out;
++	}
++	preempt_disable();
++	kp = &__get_cpu_var(zcache_preloads);
++	if (kp->obj == NULL)
++		kp->obj = obj;
++	else
++		kmem_cache_free(zcache_obj_cache, obj);
++	if (kp->page == NULL)
++		kp->page = page;
++	else
++		free_page((unsigned long)page);
++	ret = 0;
++unlock_out:
++	spin_unlock(&zcache_direct_reclaim_lock);
++out:
++	return ret;
++}
++
++static void *zcache_get_free_page(void)
++{
++	struct zcache_preload *kp;
++	void *page;
++
++	kp = &__get_cpu_var(zcache_preloads);
++	page = kp->page;
++	BUG_ON(page == NULL);
++	kp->page = NULL;
++	return page;
++}
++
++static void zcache_free_page(void *p)
++{
++	free_page((unsigned long)p);
++}
++
++/*
++ * zcache implementation for tmem host ops
++ */
++
++static struct tmem_objnode *zcache_objnode_alloc(struct tmem_pool *pool)
++{
++	struct tmem_objnode *objnode = NULL;
++	unsigned long count;
++	struct zcache_preload *kp;
++
++	kp = &__get_cpu_var(zcache_preloads);
++	if (kp->nr <= 0)
++		goto out;
++	objnode = kp->objnodes[kp->nr - 1];
++	BUG_ON(objnode == NULL);
++	kp->objnodes[kp->nr - 1] = NULL;
++	kp->nr--;
++	count = atomic_inc_return(&zcache_curr_objnode_count);
++	if (count > zcache_curr_objnode_count_max)
++		zcache_curr_objnode_count_max = count;
++out:
++	return objnode;
++}
++
++static void zcache_objnode_free(struct tmem_objnode *objnode,
++					struct tmem_pool *pool)
++{
++	atomic_dec(&zcache_curr_objnode_count);
++	BUG_ON(atomic_read(&zcache_curr_objnode_count) < 0);
++	kmem_cache_free(zcache_objnode_cache, objnode);
++}
++
++static struct tmem_obj *zcache_obj_alloc(struct tmem_pool *pool)
++{
++	struct tmem_obj *obj = NULL;
++	unsigned long count;
++	struct zcache_preload *kp;
++
++	kp = &__get_cpu_var(zcache_preloads);
++	obj = kp->obj;
++	BUG_ON(obj == NULL);
++	kp->obj = NULL;
++	count = atomic_inc_return(&zcache_curr_obj_count);
++	if (count > zcache_curr_obj_count_max)
++		zcache_curr_obj_count_max = count;
++	return obj;
++}
++
++static void zcache_obj_free(struct tmem_obj *obj, struct tmem_pool *pool)
++{
++	atomic_dec(&zcache_curr_obj_count);
++	BUG_ON(atomic_read(&zcache_curr_obj_count) < 0);
++	kmem_cache_free(zcache_obj_cache, obj);
++}
++
++static struct tmem_hostops zcache_hostops = {
++	.obj_alloc = zcache_obj_alloc,
++	.obj_free = zcache_obj_free,
++	.objnode_alloc = zcache_objnode_alloc,
++	.objnode_free = zcache_objnode_free,
++};
++
++/*
++ * zcache implementations for PAM page descriptor ops
++ */
++
++static atomic_t zcache_curr_eph_pampd_count = ATOMIC_INIT(0);
++static unsigned long zcache_curr_eph_pampd_count_max;
++static atomic_t zcache_curr_pers_pampd_count = ATOMIC_INIT(0);
++static unsigned long zcache_curr_pers_pampd_count_max;
++
++/* forward reference */
++static int zcache_compress(struct page *from, void **out_va, size_t *out_len);
++
++static void *zcache_pampd_create(char *data, size_t size, bool raw, int eph,
++				struct tmem_pool *pool, struct tmem_oid *oid,
++				 uint32_t index)
++{
++	void *pampd = NULL, *cdata;
++	size_t clen;
++	int ret;
++	unsigned long count;
++	struct page *page = (struct page *)(data);
++	struct zcache_client *cli = pool->client;
++	uint16_t client_id = get_client_id_from_client(cli);
++	unsigned long zv_mean_zsize;
++	unsigned long curr_pers_pampd_count;
++	u64 total_zsize;
++
++	if (eph) {
++		ret = zcache_compress(page, &cdata, &clen);
++		if (ret == 0)
++			goto out;
++		if (clen == 0 || clen > zbud_max_buddy_size()) {
++			zcache_compress_poor++;
++			goto out;
++		}
++		pampd = (void *)zbud_create(client_id, pool->pool_id, oid,
++						index, page, cdata, clen);
++		if (pampd != NULL) {
++			count = atomic_inc_return(&zcache_curr_eph_pampd_count);
++			if (count > zcache_curr_eph_pampd_count_max)
++				zcache_curr_eph_pampd_count_max = count;
++		}
++	} else {
++		curr_pers_pampd_count =
++			atomic_read(&zcache_curr_pers_pampd_count);
++		if (curr_pers_pampd_count >
++		    (zv_page_count_policy_percent * totalram_pages) / 100)
++			goto out;
++		ret = zcache_compress(page, &cdata, &clen);
++		if (ret == 0)
++			goto out;
++		/* reject if compression is too poor */
++		if (clen > zv_max_zsize) {
++			zcache_compress_poor++;
++			goto out;
++		}
++		/* reject if mean compression is too poor */
++		if ((clen > zv_max_mean_zsize) && (curr_pers_pampd_count > 0)) {
++			total_zsize = xv_get_total_size_bytes(cli->xvpool);
++			zv_mean_zsize = div_u64(total_zsize,
++						curr_pers_pampd_count);
++			if (zv_mean_zsize > zv_max_mean_zsize) {
++				zcache_mean_compress_poor++;
++				goto out;
++			}
++		}
++		pampd = (void *)zv_create(cli->xvpool, pool->pool_id,
++						oid, index, cdata, clen);
++		if (pampd == NULL)
++			goto out;
++		count = atomic_inc_return(&zcache_curr_pers_pampd_count);
++		if (count > zcache_curr_pers_pampd_count_max)
++			zcache_curr_pers_pampd_count_max = count;
++	}
++out:
++	return pampd;
++}
++
++/*
++ * fill the pageframe corresponding to the struct page with the data
++ * from the passed pampd
++ */
++static int zcache_pampd_get_data(char *data, size_t *bufsize, bool raw,
++					void *pampd, struct tmem_pool *pool,
++					struct tmem_oid *oid, uint32_t index)
++{
++	int ret = 0;
++
++	BUG_ON(is_ephemeral(pool));
++	zv_decompress((struct page *)(data), pampd);
++	return ret;
++}
++
++/*
++ * fill the pageframe corresponding to the struct page with the data
++ * from the passed pampd
++ */
++static int zcache_pampd_get_data_and_free(char *data, size_t *bufsize, bool raw,
++					void *pampd, struct tmem_pool *pool,
++					struct tmem_oid *oid, uint32_t index)
++{
++	int ret = 0;
++
++	BUG_ON(!is_ephemeral(pool));
++	zbud_decompress((struct page *)(data), pampd);
++	zbud_free_and_delist((struct zbud_hdr *)pampd);
++	atomic_dec(&zcache_curr_eph_pampd_count);
++	return ret;
++}
++
++/*
++ * free the pampd and remove it from any zcache lists
++ * pampd must no longer be pointed to from any tmem data structures!
++ */
++static void zcache_pampd_free(void *pampd, struct tmem_pool *pool,
++				struct tmem_oid *oid, uint32_t index)
++{
++	struct zcache_client *cli = pool->client;
++
++	if (is_ephemeral(pool)) {
++		zbud_free_and_delist((struct zbud_hdr *)pampd);
++		atomic_dec(&zcache_curr_eph_pampd_count);
++		BUG_ON(atomic_read(&zcache_curr_eph_pampd_count) < 0);
++	} else {
++		zv_free(cli->xvpool, (struct zv_hdr *)pampd);
++		atomic_dec(&zcache_curr_pers_pampd_count);
++		BUG_ON(atomic_read(&zcache_curr_pers_pampd_count) < 0);
++	}
++}
++
++static void zcache_pampd_free_obj(struct tmem_pool *pool, struct tmem_obj *obj)
++{
++}
++
++static void zcache_pampd_new_obj(struct tmem_obj *obj)
++{
++}
++
++static int zcache_pampd_replace_in_obj(void *pampd, struct tmem_obj *obj)
++{
++	return -1;
++}
++
++static bool zcache_pampd_is_remote(void *pampd)
++{
++	return 0;
++}
++
++static struct tmem_pamops zcache_pamops = {
++	.create = zcache_pampd_create,
++	.get_data = zcache_pampd_get_data,
++	.get_data_and_free = zcache_pampd_get_data_and_free,
++	.free = zcache_pampd_free,
++	.free_obj = zcache_pampd_free_obj,
++	.new_obj = zcache_pampd_new_obj,
++	.replace_in_obj = zcache_pampd_replace_in_obj,
++	.is_remote = zcache_pampd_is_remote,
++};
++
++/*
++ * zcache compression/decompression and related per-cpu stuff
++ */
++
++#define LZO_WORKMEM_BYTES LZO1X_1_MEM_COMPRESS
++#define LZO_DSTMEM_PAGE_ORDER 1
++static DEFINE_PER_CPU(unsigned char *, zcache_workmem);
++static DEFINE_PER_CPU(unsigned char *, zcache_dstmem);
++
++static int zcache_compress(struct page *from, void **out_va, size_t *out_len)
++{
++	int ret = 0;
++	unsigned char *dmem = __get_cpu_var(zcache_dstmem);
++	unsigned char *wmem = __get_cpu_var(zcache_workmem);
++	char *from_va;
++
++	BUG_ON(!irqs_disabled());
++	if (unlikely(dmem == NULL || wmem == NULL))
++		goto out;  /* no buffer, so can't compress */
++	from_va = kmap_atomic(from, KM_USER0);
++	mb();
++	ret = lzo1x_1_compress(from_va, PAGE_SIZE, dmem, out_len, wmem);
++	BUG_ON(ret != LZO_E_OK);
++	*out_va = dmem;
++	kunmap_atomic(from_va, KM_USER0);
++	ret = 1;
++out:
++	return ret;
++}
++
++
++static int zcache_cpu_notifier(struct notifier_block *nb,
++				unsigned long action, void *pcpu)
++{
++	int cpu = (long)pcpu;
++	struct zcache_preload *kp;
++
++	switch (action) {
++	case CPU_UP_PREPARE:
++		per_cpu(zcache_dstmem, cpu) = (void *)__get_free_pages(
++			GFP_KERNEL | __GFP_REPEAT,
++			LZO_DSTMEM_PAGE_ORDER),
++		per_cpu(zcache_workmem, cpu) =
++			kzalloc(LZO1X_MEM_COMPRESS,
++				GFP_KERNEL | __GFP_REPEAT);
++		break;
++	case CPU_DEAD:
++	case CPU_UP_CANCELED:
++		free_pages((unsigned long)per_cpu(zcache_dstmem, cpu),
++				LZO_DSTMEM_PAGE_ORDER);
++		per_cpu(zcache_dstmem, cpu) = NULL;
++		kfree(per_cpu(zcache_workmem, cpu));
++		per_cpu(zcache_workmem, cpu) = NULL;
++		kp = &per_cpu(zcache_preloads, cpu);
++		while (kp->nr) {
++			kmem_cache_free(zcache_objnode_cache,
++					kp->objnodes[kp->nr - 1]);
++			kp->objnodes[kp->nr - 1] = NULL;
++			kp->nr--;
++		}
++		kmem_cache_free(zcache_obj_cache, kp->obj);
++		free_page((unsigned long)kp->page);
++		break;
++	default:
++		break;
++	}
++	return NOTIFY_OK;
++}
++
++static struct notifier_block zcache_cpu_notifier_block = {
++	.notifier_call = zcache_cpu_notifier
++};
++
++#ifdef CONFIG_SYSFS
++#define ZCACHE_SYSFS_RO(_name) \
++	static ssize_t zcache_##_name##_show(struct kobject *kobj, \
++				struct kobj_attribute *attr, char *buf) \
++	{ \
++		return sprintf(buf, "%lu\n", zcache_##_name); \
++	} \
++	static struct kobj_attribute zcache_##_name##_attr = { \
++		.attr = { .name = __stringify(_name), .mode = 0444 }, \
++		.show = zcache_##_name##_show, \
++	}
++
++#define ZCACHE_SYSFS_RO_ATOMIC(_name) \
++	static ssize_t zcache_##_name##_show(struct kobject *kobj, \
++				struct kobj_attribute *attr, char *buf) \
++	{ \
++	    return sprintf(buf, "%d\n", atomic_read(&zcache_##_name)); \
++	} \
++	static struct kobj_attribute zcache_##_name##_attr = { \
++		.attr = { .name = __stringify(_name), .mode = 0444 }, \
++		.show = zcache_##_name##_show, \
++	}
++
++#define ZCACHE_SYSFS_RO_CUSTOM(_name, _func) \
++	static ssize_t zcache_##_name##_show(struct kobject *kobj, \
++				struct kobj_attribute *attr, char *buf) \
++	{ \
++	    return _func(buf); \
++	} \
++	static struct kobj_attribute zcache_##_name##_attr = { \
++		.attr = { .name = __stringify(_name), .mode = 0444 }, \
++		.show = zcache_##_name##_show, \
++	}
++
++ZCACHE_SYSFS_RO(curr_obj_count_max);
++ZCACHE_SYSFS_RO(curr_objnode_count_max);
++ZCACHE_SYSFS_RO(flush_total);
++ZCACHE_SYSFS_RO(flush_found);
++ZCACHE_SYSFS_RO(flobj_total);
++ZCACHE_SYSFS_RO(flobj_found);
++ZCACHE_SYSFS_RO(failed_eph_puts);
++ZCACHE_SYSFS_RO(failed_pers_puts);
++ZCACHE_SYSFS_RO(zbud_curr_zbytes);
++ZCACHE_SYSFS_RO(zbud_cumul_zpages);
++ZCACHE_SYSFS_RO(zbud_cumul_zbytes);
++ZCACHE_SYSFS_RO(zbud_buddied_count);
++ZCACHE_SYSFS_RO(zbpg_unused_list_count);
++ZCACHE_SYSFS_RO(evicted_raw_pages);
++ZCACHE_SYSFS_RO(evicted_unbuddied_pages);
++ZCACHE_SYSFS_RO(evicted_buddied_pages);
++ZCACHE_SYSFS_RO(failed_get_free_pages);
++ZCACHE_SYSFS_RO(failed_alloc);
++ZCACHE_SYSFS_RO(put_to_flush);
++ZCACHE_SYSFS_RO(aborted_preload);
++ZCACHE_SYSFS_RO(aborted_shrink);
++ZCACHE_SYSFS_RO(compress_poor);
++ZCACHE_SYSFS_RO(mean_compress_poor);
++ZCACHE_SYSFS_RO_ATOMIC(zbud_curr_raw_pages);
++ZCACHE_SYSFS_RO_ATOMIC(zbud_curr_zpages);
++ZCACHE_SYSFS_RO_ATOMIC(curr_obj_count);
++ZCACHE_SYSFS_RO_ATOMIC(curr_objnode_count);
++ZCACHE_SYSFS_RO_CUSTOM(zbud_unbuddied_list_counts,
++			zbud_show_unbuddied_list_counts);
++ZCACHE_SYSFS_RO_CUSTOM(zbud_cumul_chunk_counts,
++			zbud_show_cumul_chunk_counts);
++ZCACHE_SYSFS_RO_CUSTOM(zv_curr_dist_counts,
++			zv_curr_dist_counts_show);
++ZCACHE_SYSFS_RO_CUSTOM(zv_cumul_dist_counts,
++			zv_cumul_dist_counts_show);
++
++static struct attribute *zcache_attrs[] = {
++	&zcache_curr_obj_count_attr.attr,
++	&zcache_curr_obj_count_max_attr.attr,
++	&zcache_curr_objnode_count_attr.attr,
++	&zcache_curr_objnode_count_max_attr.attr,
++	&zcache_flush_total_attr.attr,
++	&zcache_flobj_total_attr.attr,
++	&zcache_flush_found_attr.attr,
++	&zcache_flobj_found_attr.attr,
++	&zcache_failed_eph_puts_attr.attr,
++	&zcache_failed_pers_puts_attr.attr,
++	&zcache_compress_poor_attr.attr,
++	&zcache_mean_compress_poor_attr.attr,
++	&zcache_zbud_curr_raw_pages_attr.attr,
++	&zcache_zbud_curr_zpages_attr.attr,
++	&zcache_zbud_curr_zbytes_attr.attr,
++	&zcache_zbud_cumul_zpages_attr.attr,
++	&zcache_zbud_cumul_zbytes_attr.attr,
++	&zcache_zbud_buddied_count_attr.attr,
++	&zcache_zbpg_unused_list_count_attr.attr,
++	&zcache_evicted_raw_pages_attr.attr,
++	&zcache_evicted_unbuddied_pages_attr.attr,
++	&zcache_evicted_buddied_pages_attr.attr,
++	&zcache_failed_get_free_pages_attr.attr,
++	&zcache_failed_alloc_attr.attr,
++	&zcache_put_to_flush_attr.attr,
++	&zcache_aborted_preload_attr.attr,
++	&zcache_aborted_shrink_attr.attr,
++	&zcache_zbud_unbuddied_list_counts_attr.attr,
++	&zcache_zbud_cumul_chunk_counts_attr.attr,
++	&zcache_zv_curr_dist_counts_attr.attr,
++	&zcache_zv_cumul_dist_counts_attr.attr,
++	&zcache_zv_max_zsize_attr.attr,
++	&zcache_zv_max_mean_zsize_attr.attr,
++	&zcache_zv_page_count_policy_percent_attr.attr,
++	NULL,
++};
++
++static struct attribute_group zcache_attr_group = {
++	.attrs = zcache_attrs,
++	.name = "zcache",
++};
++
++#endif /* CONFIG_SYSFS */
++/*
++ * When zcache is disabled ("frozen"), pools can be created and destroyed,
++ * but all puts (and thus all other operations that require memory allocation)
++ * must fail.  If zcache is unfrozen, accepts puts, then frozen again,
++ * data consistency requires all puts while frozen to be converted into
++ * flushes.
++ */
++static bool zcache_freeze;
++
++/*
++ * zcache shrinker interface (only useful for ephemeral pages, so zbud only)
++ */
++static int shrink_zcache_memory(struct shrinker *shrink,
++				struct shrink_control *sc)
++{
++	int ret = -1;
++	int nr = sc->nr_to_scan;
++	gfp_t gfp_mask = sc->gfp_mask;
++
++	if (nr >= 0) {
++		if (!(gfp_mask & __GFP_FS))
++			/* does this case really need to be skipped? */
++			goto out;
++		if (spin_trylock(&zcache_direct_reclaim_lock)) {
++			zbud_evict_pages(nr);
++			spin_unlock(&zcache_direct_reclaim_lock);
++		} else
++			zcache_aborted_shrink++;
++	}
++	ret = (int)atomic_read(&zcache_zbud_curr_raw_pages);
++out:
++	return ret;
++}
++
++static struct shrinker zcache_shrinker = {
++	.shrink = shrink_zcache_memory,
++	.seeks = DEFAULT_SEEKS,
++};
++
++/*
++ * zcache shims between cleancache/frontswap ops and tmem
++ */
++
++static int zcache_put_page(int cli_id, int pool_id, struct tmem_oid *oidp,
++				uint32_t index, struct page *page)
++{
++	struct tmem_pool *pool;
++	int ret = -1;
++
++	BUG_ON(!irqs_disabled());
++	pool = zcache_get_pool_by_id(cli_id, pool_id);
++	if (unlikely(pool == NULL))
++		goto out;
++	if (!zcache_freeze && zcache_do_preload(pool) == 0) {
++		/* preload does preempt_disable on success */
++		ret = tmem_put(pool, oidp, index, (char *)(page),
++				PAGE_SIZE, 0, is_ephemeral(pool));
++		if (ret < 0) {
++			if (is_ephemeral(pool))
++				zcache_failed_eph_puts++;
++			else
++				zcache_failed_pers_puts++;
++		}
++		zcache_put_pool(pool);
++		preempt_enable_no_resched();
++	} else {
++		zcache_put_to_flush++;
++		if (atomic_read(&pool->obj_count) > 0)
++			/* the put fails whether the flush succeeds or not */
++			(void)tmem_flush_page(pool, oidp, index);
++		zcache_put_pool(pool);
++	}
++out:
++	return ret;
++}
++
++static int zcache_get_page(int cli_id, int pool_id, struct tmem_oid *oidp,
++				uint32_t index, struct page *page)
++{
++	struct tmem_pool *pool;
++	int ret = -1;
++	unsigned long flags;
++	size_t size = PAGE_SIZE;
++
++	local_irq_save(flags);
++	pool = zcache_get_pool_by_id(cli_id, pool_id);
++	if (likely(pool != NULL)) {
++		if (atomic_read(&pool->obj_count) > 0)
++			ret = tmem_get(pool, oidp, index, (char *)(page),
++					&size, 0, is_ephemeral(pool));
++		zcache_put_pool(pool);
++	}
++	local_irq_restore(flags);
++	return ret;
++}
++
++static int zcache_flush_page(int cli_id, int pool_id,
++				struct tmem_oid *oidp, uint32_t index)
++{
++	struct tmem_pool *pool;
++	int ret = -1;
++	unsigned long flags;
++
++	local_irq_save(flags);
++	zcache_flush_total++;
++	pool = zcache_get_pool_by_id(cli_id, pool_id);
++	if (likely(pool != NULL)) {
++		if (atomic_read(&pool->obj_count) > 0)
++			ret = tmem_flush_page(pool, oidp, index);
++		zcache_put_pool(pool);
++	}
++	if (ret >= 0)
++		zcache_flush_found++;
++	local_irq_restore(flags);
++	return ret;
++}
++
++static int zcache_flush_object(int cli_id, int pool_id,
++				struct tmem_oid *oidp)
++{
++	struct tmem_pool *pool;
++	int ret = -1;
++	unsigned long flags;
++
++	local_irq_save(flags);
++	zcache_flobj_total++;
++	pool = zcache_get_pool_by_id(cli_id, pool_id);
++	if (likely(pool != NULL)) {
++		if (atomic_read(&pool->obj_count) > 0)
++			ret = tmem_flush_object(pool, oidp);
++		zcache_put_pool(pool);
++	}
++	if (ret >= 0)
++		zcache_flobj_found++;
++	local_irq_restore(flags);
++	return ret;
++}
++
++static int zcache_destroy_pool(int cli_id, int pool_id)
++{
++	struct tmem_pool *pool = NULL;
++	struct zcache_client *cli = NULL;
++	int ret = -1;
++
++	if (pool_id < 0)
++		goto out;
++	if (cli_id == LOCAL_CLIENT)
++		cli = &zcache_host;
++	else if ((unsigned int)cli_id < MAX_CLIENTS)
++		cli = &zcache_clients[cli_id];
++	if (cli == NULL)
++		goto out;
++	atomic_inc(&cli->refcount);
++	pool = cli->tmem_pools[pool_id];
++	if (pool == NULL)
++		goto out;
++	cli->tmem_pools[pool_id] = NULL;
++	/* wait for pool activity on other cpus to quiesce */
++	while (atomic_read(&pool->refcount) != 0)
++		;
++	atomic_dec(&cli->refcount);
++	local_bh_disable();
++	ret = tmem_destroy_pool(pool);
++	local_bh_enable();
++	kfree(pool);
++	pr_info("zcache: destroyed pool id=%d, cli_id=%d\n",
++			pool_id, cli_id);
++out:
++	return ret;
++}
++
++static int zcache_new_pool(uint16_t cli_id, uint32_t flags)
++{
++	int poolid = -1;
++	struct tmem_pool *pool;
++	struct zcache_client *cli = NULL;
++
++	if (cli_id == LOCAL_CLIENT)
++		cli = &zcache_host;
++	else if ((unsigned int)cli_id < MAX_CLIENTS)
++		cli = &zcache_clients[cli_id];
++	if (cli == NULL)
++		goto out;
++	atomic_inc(&cli->refcount);
++	pool = kmalloc(sizeof(struct tmem_pool), GFP_KERNEL);
++	if (pool == NULL) {
++		pr_info("zcache: pool creation failed: out of memory\n");
++		goto out;
++	}
++
++	for (poolid = 0; poolid < MAX_POOLS_PER_CLIENT; poolid++)
++		if (cli->tmem_pools[poolid] == NULL)
++			break;
++	if (poolid >= MAX_POOLS_PER_CLIENT) {
++		pr_info("zcache: pool creation failed: max exceeded\n");
++		kfree(pool);
++		poolid = -1;
++		goto out;
++	}
++	atomic_set(&pool->refcount, 0);
++	pool->client = cli;
++	pool->pool_id = poolid;
++	tmem_new_pool(pool, flags);
++	cli->tmem_pools[poolid] = pool;
++	pr_info("zcache: created %s tmem pool, id=%d, client=%d\n",
++		flags & TMEM_POOL_PERSIST ? "persistent" : "ephemeral",
++		poolid, cli_id);
++out:
++	if (cli != NULL)
++		atomic_dec(&cli->refcount);
++	return poolid;
++}
++
++/**********
++ * Two kernel functionalities currently can be layered on top of tmem.
++ * These are "cleancache" which is used as a second-chance cache for clean
++ * page cache pages; and "frontswap" which is used for swap pages
++ * to avoid writes to disk.  A generic "shim" is provided here for each
++ * to translate in-kernel semantics to zcache semantics.
++ */
++
++#ifdef CONFIG_CLEANCACHE
++static void zcache_cleancache_put_page(int pool_id,
++					struct cleancache_filekey key,
++					pgoff_t index, struct page *page)
++{
++	u32 ind = (u32) index;
++	struct tmem_oid oid = *(struct tmem_oid *)&key;
++
++	if (likely(ind == index))
++		(void)zcache_put_page(LOCAL_CLIENT, pool_id, &oid, index, page);
++}
++
++static int zcache_cleancache_get_page(int pool_id,
++					struct cleancache_filekey key,
++					pgoff_t index, struct page *page)
++{
++	u32 ind = (u32) index;
++	struct tmem_oid oid = *(struct tmem_oid *)&key;
++	int ret = -1;
++
++	if (likely(ind == index))
++		ret = zcache_get_page(LOCAL_CLIENT, pool_id, &oid, index, page);
++	return ret;
++}
++
++static void zcache_cleancache_flush_page(int pool_id,
++					struct cleancache_filekey key,
++					pgoff_t index)
++{
++	u32 ind = (u32) index;
++	struct tmem_oid oid = *(struct tmem_oid *)&key;
++
++	if (likely(ind == index))
++		(void)zcache_flush_page(LOCAL_CLIENT, pool_id, &oid, ind);
++}
++
++static void zcache_cleancache_flush_inode(int pool_id,
++					struct cleancache_filekey key)
++{
++	struct tmem_oid oid = *(struct tmem_oid *)&key;
++
++	(void)zcache_flush_object(LOCAL_CLIENT, pool_id, &oid);
++}
++
++static void zcache_cleancache_flush_fs(int pool_id)
++{
++	if (pool_id >= 0)
++		(void)zcache_destroy_pool(LOCAL_CLIENT, pool_id);
++}
++
++static int zcache_cleancache_init_fs(size_t pagesize)
++{
++	BUG_ON(sizeof(struct cleancache_filekey) !=
++				sizeof(struct tmem_oid));
++	BUG_ON(pagesize != PAGE_SIZE);
++	return zcache_new_pool(LOCAL_CLIENT, 0);
++}
++
++static int zcache_cleancache_init_shared_fs(char *uuid, size_t pagesize)
++{
++	/* shared pools are unsupported and map to private */
++	BUG_ON(sizeof(struct cleancache_filekey) !=
++				sizeof(struct tmem_oid));
++	BUG_ON(pagesize != PAGE_SIZE);
++	return zcache_new_pool(LOCAL_CLIENT, 0);
++}
++
++static struct cleancache_ops zcache_cleancache_ops = {
++	.put_page = zcache_cleancache_put_page,
++	.get_page = zcache_cleancache_get_page,
++	.invalidate_page = zcache_cleancache_flush_page,
++	.invalidate_inode = zcache_cleancache_flush_inode,
++	.invalidate_fs = zcache_cleancache_flush_fs,
++	.init_shared_fs = zcache_cleancache_init_shared_fs,
++	.init_fs = zcache_cleancache_init_fs
++};
++
++struct cleancache_ops zcache_cleancache_register_ops(void)
++{
++	struct cleancache_ops old_ops =
++		cleancache_register_ops(&zcache_cleancache_ops);
++
++	return old_ops;
++}
++#endif
++
++#ifdef CONFIG_FRONTSWAP
++/* a single tmem poolid is used for all frontswap "types" (swapfiles) */
++static int zcache_frontswap_poolid = -1;
++
++/*
++ * Swizzling increases objects per swaptype, increasing tmem concurrency
++ * for heavy swaploads.  Later, larger nr_cpus -> larger SWIZ_BITS
++ */
++#define SWIZ_BITS		4
++#define SWIZ_MASK		((1 << SWIZ_BITS) - 1)
++#define _oswiz(_type, _ind)	((_type << SWIZ_BITS) | (_ind & SWIZ_MASK))
++#define iswiz(_ind)		(_ind >> SWIZ_BITS)
++
++static inline struct tmem_oid oswiz(unsigned type, u32 ind)
++{
++	struct tmem_oid oid = { .oid = { 0 } };
++	oid.oid[0] = _oswiz(type, ind);
++	return oid;
++}
++
++static int zcache_frontswap_put_page(unsigned type, pgoff_t offset,
++				   struct page *page)
++{
++	u64 ind64 = (u64)offset;
++	u32 ind = (u32)offset;
++	struct tmem_oid oid = oswiz(type, ind);
++	int ret = -1;
++	unsigned long flags;
++
++	BUG_ON(!PageLocked(page));
++	if (likely(ind64 == ind)) {
++		local_irq_save(flags);
++		ret = zcache_put_page(LOCAL_CLIENT, zcache_frontswap_poolid,
++					&oid, iswiz(ind), page);
++		local_irq_restore(flags);
++	}
++	return ret;
++}
++
++/* returns 0 if the page was successfully gotten from frontswap, -1 if
++ * was not present (should never happen!) */
++static int zcache_frontswap_get_page(unsigned type, pgoff_t offset,
++				   struct page *page)
++{
++	u64 ind64 = (u64)offset;
++	u32 ind = (u32)offset;
++	struct tmem_oid oid = oswiz(type, ind);
++	int ret = -1;
++
++	BUG_ON(!PageLocked(page));
++	if (likely(ind64 == ind))
++		ret = zcache_get_page(LOCAL_CLIENT, zcache_frontswap_poolid,
++					&oid, iswiz(ind), page);
++	return ret;
++}
++
++/* flush a single page from frontswap */
++static void zcache_frontswap_flush_page(unsigned type, pgoff_t offset)
++{
++	u64 ind64 = (u64)offset;
++	u32 ind = (u32)offset;
++	struct tmem_oid oid = oswiz(type, ind);
++
++	if (likely(ind64 == ind))
++		(void)zcache_flush_page(LOCAL_CLIENT, zcache_frontswap_poolid,
++					&oid, iswiz(ind));
++}
++
++/* flush all pages from the passed swaptype */
++static void zcache_frontswap_flush_area(unsigned type)
++{
++	struct tmem_oid oid;
++	int ind;
++
++	for (ind = SWIZ_MASK; ind >= 0; ind--) {
++		oid = oswiz(type, ind);
++		(void)zcache_flush_object(LOCAL_CLIENT,
++						zcache_frontswap_poolid, &oid);
++	}
++}
++
++static void zcache_frontswap_init(unsigned ignored)
++{
++	/* a single tmem poolid is used for all frontswap "types" (swapfiles) */
++	if (zcache_frontswap_poolid < 0)
++		zcache_frontswap_poolid =
++			zcache_new_pool(LOCAL_CLIENT, TMEM_POOL_PERSIST);
++}
++
++static struct frontswap_ops zcache_frontswap_ops = {
++	.put_page = zcache_frontswap_put_page,
++	.get_page = zcache_frontswap_get_page,
++	.invalidate_page = zcache_frontswap_flush_page,
++	.invalidate_area = zcache_frontswap_flush_area,
++	.init = zcache_frontswap_init
++};
++
++struct frontswap_ops zcache_frontswap_register_ops(void)
++{
++	struct frontswap_ops old_ops =
++		frontswap_register_ops(&zcache_frontswap_ops);
++
++	return old_ops;
++}
++#endif
++
++/*
++ * zcache initialization
++ * NOTE FOR NOW zcache MUST BE PROVIDED AS A KERNEL BOOT PARAMETER OR
++ * NOTHING HAPPENS!
++ */
++
++static int zcache_enabled;
++
++static int __init enable_zcache(char *s)
++{
++	zcache_enabled = 1;
++	return 1;
++}
++__setup("zcache", enable_zcache);
++
++/* allow independent dynamic disabling of cleancache and frontswap */
++
++static int use_cleancache = 1;
++
++static int __init no_cleancache(char *s)
++{
++	use_cleancache = 0;
++	return 1;
++}
++
++__setup("nocleancache", no_cleancache);
++
++static int use_frontswap = 1;
++
++static int __init no_frontswap(char *s)
++{
++	use_frontswap = 0;
++	return 1;
++}
++
++__setup("nofrontswap", no_frontswap);
++
++static int __init zcache_init(void)
++{
++	int ret = 0;
++
++#ifdef CONFIG_SYSFS
++	ret = sysfs_create_group(mm_kobj, &zcache_attr_group);
++	if (ret) {
++		pr_err("zcache: can't create sysfs\n");
++		goto out;
++	}
++#endif /* CONFIG_SYSFS */
++#if defined(CONFIG_CLEANCACHE) || defined(CONFIG_FRONTSWAP)
++	if (zcache_enabled) {
++		unsigned int cpu;
++
++		tmem_register_hostops(&zcache_hostops);
++		tmem_register_pamops(&zcache_pamops);
++		ret = register_cpu_notifier(&zcache_cpu_notifier_block);
++		if (ret) {
++			pr_err("zcache: can't register cpu notifier\n");
++			goto out;
++		}
++		for_each_online_cpu(cpu) {
++			void *pcpu = (void *)(long)cpu;
++			zcache_cpu_notifier(&zcache_cpu_notifier_block,
++				CPU_UP_PREPARE, pcpu);
++		}
++	}
++	zcache_objnode_cache = kmem_cache_create("zcache_objnode",
++				sizeof(struct tmem_objnode), 0, 0, NULL);
++	zcache_obj_cache = kmem_cache_create("zcache_obj",
++				sizeof(struct tmem_obj), 0, 0, NULL);
++	ret = zcache_new_client(LOCAL_CLIENT);
++	if (ret) {
++		pr_err("zcache: can't create client\n");
++		goto out;
++	}
++#endif
++#ifdef CONFIG_CLEANCACHE
++	if (zcache_enabled && use_cleancache) {
++		struct cleancache_ops old_ops;
++
++		zbud_init();
++		register_shrinker(&zcache_shrinker);
++		old_ops = zcache_cleancache_register_ops();
++		pr_info("zcache: cleancache enabled using kernel "
++			"transcendent memory and compression buddies\n");
++		if (old_ops.init_fs != NULL)
++			pr_warning("zcache: cleancache_ops overridden");
++	}
++#endif
++#ifdef CONFIG_FRONTSWAP
++	if (zcache_enabled && use_frontswap) {
++		struct frontswap_ops old_ops;
++
++		old_ops = zcache_frontswap_register_ops();
++		pr_info("zcache: frontswap enabled using kernel "
++			"transcendent memory and xvmalloc\n");
++		if (old_ops.init != NULL)
++			pr_warning("ktmem: frontswap_ops overridden");
++	}
++#endif
++out:
++	return ret;
++}
++
++module_init(zcache_init)
+-- 
+1.7.1
diff --git a/a/content_digest b/N1/content_digest
index a05ef92..4ff1e48 100644
--- a/a/content_digest
+++ b/N1/content_digest
@@ -13,5 +13,3053 @@
  " dan.magenheimer@oracle.com\0"
  "\00:1\0"
  "b\0"
+ ">From 60ce12cd6fb7d81d3bad48cf3766269547b4c241 Mon Sep 17 00:00:00 2001\n"
+ "From: Dan Magenheimer <dan.magenheimer@oracle.com>\n"
+ "Date: Wed, 21 Dec 2011 14:01:56 -0700\n"
+ "Subject: [PATCH V2 2/6] drivers/staging/ramster: local compression + tmem\n"
+ "\n"
+ "Copy files from drivers/staging/zcache.  Ramster compresses pages\n"
+ "locally before transmitting them to another node, so we can\n"
+ "leverage the zcache and tmem code directly.  Note: there are\n"
+ "no ramster-specific changes yet to these files.\n"
+ "\n"
+ "Signed-off-by: Dan Magenheimer <dan.magenheimer@oracle.com>\n"
+ "\n"
+ "---\n"
+ "\n"
+ " drivers/staging/ramster/Kconfig       |   13 +\n"
+ " drivers/staging/ramster/Makefile      |    3 +\n"
+ " drivers/staging/ramster/tmem.c        |  770 +++++++++++++\n"
+ " drivers/staging/ramster/tmem.h        |  206 ++++\n"
+ " drivers/staging/ramster/zcache-main.c | 2003 +++++++++++++++++++++++++++++++++\n"
+ " 5 files changed, 2995 insertions(+), 0 deletions(-)\n"
+ "\n"
+ "diff --git a/drivers/staging/ramster/Kconfig b/drivers/staging/ramster/Kconfig\n"
+ "new file mode 100644\n"
+ "index 0000000..7fabcb2\n"
+ "--- /dev/null\n"
+ "+++ b/drivers/staging/ramster/Kconfig\n"
+ "@@ -0,0 +1,13 @@\n"
+ "+config ZCACHE\n"
+ "+\ttristate \"Dynamic compression of swap pages and clean pagecache pages\"\n"
+ "+\tdepends on CLEANCACHE || FRONTSWAP\n"
+ "+\tselect XVMALLOC\n"
+ "+\tselect LZO_COMPRESS\n"
+ "+\tselect LZO_DECOMPRESS\n"
+ "+\tdefault n\n"
+ "+\thelp\n"
+ "+\t  Zcache doubles RAM efficiency while providing a significant\n"
+ "+\t  performance boosts on many workloads.  Zcache uses lzo1x\n"
+ "+\t  compression and an in-kernel implementation of transcendent\n"
+ "+\t  memory to store clean page cache pages and swap in RAM,\n"
+ "+\t  providing a noticeable reduction in disk I/O.\n"
+ "diff --git a/drivers/staging/ramster/Makefile b/drivers/staging/ramster/Makefile\n"
+ "new file mode 100644\n"
+ "index 0000000..60daa27\n"
+ "--- /dev/null\n"
+ "+++ b/drivers/staging/ramster/Makefile\n"
+ "@@ -0,0 +1,3 @@\n"
+ "+zcache-y\t:=\tzcache-main.o tmem.o\n"
+ "+\n"
+ "+obj-$(CONFIG_ZCACHE)\t+=\tzcache.o\n"
+ "diff --git a/drivers/staging/ramster/tmem.c b/drivers/staging/ramster/tmem.c\n"
+ "new file mode 100644\n"
+ "index 0000000..1ca66ea\n"
+ "--- /dev/null\n"
+ "+++ b/drivers/staging/ramster/tmem.c\n"
+ "@@ -0,0 +1,770 @@\n"
+ "+/*\n"
+ "+ * In-kernel transcendent memory (generic implementation)\n"
+ "+ *\n"
+ "+ * Copyright (c) 2009-2011, Dan Magenheimer, Oracle Corp.\n"
+ "+ *\n"
+ "+ * The primary purpose of Transcedent Memory (\"tmem\") is to map object-oriented\n"
+ "+ * \"handles\" (triples containing a pool id, and object id, and an index), to\n"
+ "+ * pages in a page-accessible memory (PAM).  Tmem references the PAM pages via\n"
+ "+ * an abstract \"pampd\" (PAM page-descriptor), which can be operated on by a\n"
+ "+ * set of functions (pamops).  Each pampd contains some representation of\n"
+ "+ * PAGE_SIZE bytes worth of data. Tmem must support potentially millions of\n"
+ "+ * pages and must be able to insert, find, and delete these pages at a\n"
+ "+ * potential frequency of thousands per second concurrently across many CPUs,\n"
+ "+ * (and, if used with KVM, across many vcpus across many guests).\n"
+ "+ * Tmem is tracked with a hierarchy of data structures, organized by\n"
+ "+ * the elements in a handle-tuple: pool_id, object_id, and page index.\n"
+ "+ * One or more \"clients\" (e.g. guests) each provide one or more tmem_pools.\n"
+ "+ * Each pool, contains a hash table of rb_trees of tmem_objs.  Each\n"
+ "+ * tmem_obj contains a radix-tree-like tree of pointers, with intermediate\n"
+ "+ * nodes called tmem_objnodes.  Each leaf pointer in this tree points to\n"
+ "+ * a pampd, which is accessible only through a small set of callbacks\n"
+ "+ * registered by the PAM implementation (see tmem_register_pamops). Tmem\n"
+ "+ * does all memory allocation via a set of callbacks registered by the tmem\n"
+ "+ * host implementation (e.g. see tmem_register_hostops).\n"
+ "+ */\n"
+ "+\n"
+ "+#include <linux/list.h>\n"
+ "+#include <linux/spinlock.h>\n"
+ "+#include <linux/atomic.h>\n"
+ "+\n"
+ "+#include \"tmem.h\"\n"
+ "+\n"
+ "+/* data structure sentinels used for debugging... see tmem.h */\n"
+ "+#define POOL_SENTINEL 0x87658765\n"
+ "+#define OBJ_SENTINEL 0x12345678\n"
+ "+#define OBJNODE_SENTINEL 0xfedcba09\n"
+ "+\n"
+ "+/*\n"
+ "+ * A tmem host implementation must use this function to register callbacks\n"
+ "+ * for memory allocation.\n"
+ "+ */\n"
+ "+static struct tmem_hostops tmem_hostops;\n"
+ "+\n"
+ "+static void tmem_objnode_tree_init(void);\n"
+ "+\n"
+ "+void tmem_register_hostops(struct tmem_hostops *m)\n"
+ "+{\n"
+ "+\ttmem_objnode_tree_init();\n"
+ "+\ttmem_hostops = *m;\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * A tmem host implementation must use this function to register\n"
+ "+ * callbacks for a page-accessible memory (PAM) implementation\n"
+ "+ */\n"
+ "+static struct tmem_pamops tmem_pamops;\n"
+ "+\n"
+ "+void tmem_register_pamops(struct tmem_pamops *m)\n"
+ "+{\n"
+ "+\ttmem_pamops = *m;\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * Oid's are potentially very sparse and tmem_objs may have an indeterminately\n"
+ "+ * short life, being added and deleted at a relatively high frequency.\n"
+ "+ * So an rb_tree is an ideal data structure to manage tmem_objs.  But because\n"
+ "+ * of the potentially huge number of tmem_objs, each pool manages a hashtable\n"
+ "+ * of rb_trees to reduce search, insert, delete, and rebalancing time.\n"
+ "+ * Each hashbucket also has a lock to manage concurrent access.\n"
+ "+ *\n"
+ "+ * The following routines manage tmem_objs.  When any tmem_obj is accessed,\n"
+ "+ * the hashbucket lock must be held.\n"
+ "+ */\n"
+ "+\n"
+ "+/* searches for object==oid in pool, returns locked object if found */\n"
+ "+static struct tmem_obj *tmem_obj_find(struct tmem_hashbucket *hb,\n"
+ "+\t\t\t\t\tstruct tmem_oid *oidp)\n"
+ "+{\n"
+ "+\tstruct rb_node *rbnode;\n"
+ "+\tstruct tmem_obj *obj;\n"
+ "+\n"
+ "+\trbnode = hb->obj_rb_root.rb_node;\n"
+ "+\twhile (rbnode) {\n"
+ "+\t\tBUG_ON(RB_EMPTY_NODE(rbnode));\n"
+ "+\t\tobj = rb_entry(rbnode, struct tmem_obj, rb_tree_node);\n"
+ "+\t\tswitch (tmem_oid_compare(oidp, &obj->oid)) {\n"
+ "+\t\tcase 0: /* equal */\n"
+ "+\t\t\tgoto out;\n"
+ "+\t\tcase -1:\n"
+ "+\t\t\trbnode = rbnode->rb_left;\n"
+ "+\t\t\tbreak;\n"
+ "+\t\tcase 1:\n"
+ "+\t\t\trbnode = rbnode->rb_right;\n"
+ "+\t\t\tbreak;\n"
+ "+\t\t}\n"
+ "+\t}\n"
+ "+\tobj = NULL;\n"
+ "+out:\n"
+ "+\treturn obj;\n"
+ "+}\n"
+ "+\n"
+ "+static void tmem_pampd_destroy_all_in_obj(struct tmem_obj *);\n"
+ "+\n"
+ "+/* free an object that has no more pampds in it */\n"
+ "+static void tmem_obj_free(struct tmem_obj *obj, struct tmem_hashbucket *hb)\n"
+ "+{\n"
+ "+\tstruct tmem_pool *pool;\n"
+ "+\n"
+ "+\tBUG_ON(obj == NULL);\n"
+ "+\tASSERT_SENTINEL(obj, OBJ);\n"
+ "+\tBUG_ON(obj->pampd_count > 0);\n"
+ "+\tpool = obj->pool;\n"
+ "+\tBUG_ON(pool == NULL);\n"
+ "+\tif (obj->objnode_tree_root != NULL) /* may be \"stump\" with no leaves */\n"
+ "+\t\ttmem_pampd_destroy_all_in_obj(obj);\n"
+ "+\tBUG_ON(obj->objnode_tree_root != NULL);\n"
+ "+\tBUG_ON((long)obj->objnode_count != 0);\n"
+ "+\tatomic_dec(&pool->obj_count);\n"
+ "+\tBUG_ON(atomic_read(&pool->obj_count) < 0);\n"
+ "+\tINVERT_SENTINEL(obj, OBJ);\n"
+ "+\tobj->pool = NULL;\n"
+ "+\ttmem_oid_set_invalid(&obj->oid);\n"
+ "+\trb_erase(&obj->rb_tree_node, &hb->obj_rb_root);\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * initialize, and insert an tmem_object_root (called only if find failed)\n"
+ "+ */\n"
+ "+static void tmem_obj_init(struct tmem_obj *obj, struct tmem_hashbucket *hb,\n"
+ "+\t\t\t\t\tstruct tmem_pool *pool,\n"
+ "+\t\t\t\t\tstruct tmem_oid *oidp)\n"
+ "+{\n"
+ "+\tstruct rb_root *root = &hb->obj_rb_root;\n"
+ "+\tstruct rb_node **new = &(root->rb_node), *parent = NULL;\n"
+ "+\tstruct tmem_obj *this;\n"
+ "+\n"
+ "+\tBUG_ON(pool == NULL);\n"
+ "+\tatomic_inc(&pool->obj_count);\n"
+ "+\tobj->objnode_tree_height = 0;\n"
+ "+\tobj->objnode_tree_root = NULL;\n"
+ "+\tobj->pool = pool;\n"
+ "+\tobj->oid = *oidp;\n"
+ "+\tobj->objnode_count = 0;\n"
+ "+\tobj->pampd_count = 0;\n"
+ "+\t(*tmem_pamops.new_obj)(obj);\n"
+ "+\tSET_SENTINEL(obj, OBJ);\n"
+ "+\twhile (*new) {\n"
+ "+\t\tBUG_ON(RB_EMPTY_NODE(*new));\n"
+ "+\t\tthis = rb_entry(*new, struct tmem_obj, rb_tree_node);\n"
+ "+\t\tparent = *new;\n"
+ "+\t\tswitch (tmem_oid_compare(oidp, &this->oid)) {\n"
+ "+\t\tcase 0:\n"
+ "+\t\t\tBUG(); /* already present; should never happen! */\n"
+ "+\t\t\tbreak;\n"
+ "+\t\tcase -1:\n"
+ "+\t\t\tnew = &(*new)->rb_left;\n"
+ "+\t\t\tbreak;\n"
+ "+\t\tcase 1:\n"
+ "+\t\t\tnew = &(*new)->rb_right;\n"
+ "+\t\t\tbreak;\n"
+ "+\t\t}\n"
+ "+\t}\n"
+ "+\trb_link_node(&obj->rb_tree_node, parent, new);\n"
+ "+\trb_insert_color(&obj->rb_tree_node, root);\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * Tmem is managed as a set of tmem_pools with certain attributes, such as\n"
+ "+ * \"ephemeral\" vs \"persistent\".  These attributes apply to all tmem_objs\n"
+ "+ * and all pampds that belong to a tmem_pool.  A tmem_pool is created\n"
+ "+ * or deleted relatively rarely (for example, when a filesystem is\n"
+ "+ * mounted or unmounted.\n"
+ "+ */\n"
+ "+\n"
+ "+/* flush all data from a pool and, optionally, free it */\n"
+ "+static void tmem_pool_flush(struct tmem_pool *pool, bool destroy)\n"
+ "+{\n"
+ "+\tstruct rb_node *rbnode;\n"
+ "+\tstruct tmem_obj *obj;\n"
+ "+\tstruct tmem_hashbucket *hb = &pool->hashbucket[0];\n"
+ "+\tint i;\n"
+ "+\n"
+ "+\tBUG_ON(pool == NULL);\n"
+ "+\tfor (i = 0; i < TMEM_HASH_BUCKETS; i++, hb++) {\n"
+ "+\t\tspin_lock(&hb->lock);\n"
+ "+\t\trbnode = rb_first(&hb->obj_rb_root);\n"
+ "+\t\twhile (rbnode != NULL) {\n"
+ "+\t\t\tobj = rb_entry(rbnode, struct tmem_obj, rb_tree_node);\n"
+ "+\t\t\trbnode = rb_next(rbnode);\n"
+ "+\t\t\ttmem_pampd_destroy_all_in_obj(obj);\n"
+ "+\t\t\ttmem_obj_free(obj, hb);\n"
+ "+\t\t\t(*tmem_hostops.obj_free)(obj, pool);\n"
+ "+\t\t}\n"
+ "+\t\tspin_unlock(&hb->lock);\n"
+ "+\t}\n"
+ "+\tif (destroy)\n"
+ "+\t\tlist_del(&pool->pool_list);\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * A tmem_obj contains a radix-tree-like tree in which the intermediate\n"
+ "+ * nodes are called tmem_objnodes.  (The kernel lib/radix-tree.c implementation\n"
+ "+ * is very specialized and tuned for specific uses and is not particularly\n"
+ "+ * suited for use from this code, though some code from the core algorithms has\n"
+ "+ * been reused, thus the copyright notices below).  Each tmem_objnode contains\n"
+ "+ * a set of pointers which point to either a set of intermediate tmem_objnodes\n"
+ "+ * or a set of of pampds.\n"
+ "+ *\n"
+ "+ * Portions Copyright (C) 2001 Momchil Velikov\n"
+ "+ * Portions Copyright (C) 2001 Christoph Hellwig\n"
+ "+ * Portions Copyright (C) 2005 SGI, Christoph Lameter <clameter@sgi.com>\n"
+ "+ */\n"
+ "+\n"
+ "+struct tmem_objnode_tree_path {\n"
+ "+\tstruct tmem_objnode *objnode;\n"
+ "+\tint offset;\n"
+ "+};\n"
+ "+\n"
+ "+/* objnode height_to_maxindex translation */\n"
+ "+static unsigned long tmem_objnode_tree_h2max[OBJNODE_TREE_MAX_PATH + 1];\n"
+ "+\n"
+ "+static void tmem_objnode_tree_init(void)\n"
+ "+{\n"
+ "+\tunsigned int ht, tmp;\n"
+ "+\n"
+ "+\tfor (ht = 0; ht < ARRAY_SIZE(tmem_objnode_tree_h2max); ht++) {\n"
+ "+\t\ttmp = ht * OBJNODE_TREE_MAP_SHIFT;\n"
+ "+\t\tif (tmp >= OBJNODE_TREE_INDEX_BITS)\n"
+ "+\t\t\ttmem_objnode_tree_h2max[ht] = ~0UL;\n"
+ "+\t\telse\n"
+ "+\t\t\ttmem_objnode_tree_h2max[ht] =\n"
+ "+\t\t\t    (~0UL >> (OBJNODE_TREE_INDEX_BITS - tmp - 1)) >> 1;\n"
+ "+\t}\n"
+ "+}\n"
+ "+\n"
+ "+static struct tmem_objnode *tmem_objnode_alloc(struct tmem_obj *obj)\n"
+ "+{\n"
+ "+\tstruct tmem_objnode *objnode;\n"
+ "+\n"
+ "+\tASSERT_SENTINEL(obj, OBJ);\n"
+ "+\tBUG_ON(obj->pool == NULL);\n"
+ "+\tASSERT_SENTINEL(obj->pool, POOL);\n"
+ "+\tobjnode = (*tmem_hostops.objnode_alloc)(obj->pool);\n"
+ "+\tif (unlikely(objnode == NULL))\n"
+ "+\t\tgoto out;\n"
+ "+\tobjnode->obj = obj;\n"
+ "+\tSET_SENTINEL(objnode, OBJNODE);\n"
+ "+\tmemset(&objnode->slots, 0, sizeof(objnode->slots));\n"
+ "+\tobjnode->slots_in_use = 0;\n"
+ "+\tobj->objnode_count++;\n"
+ "+out:\n"
+ "+\treturn objnode;\n"
+ "+}\n"
+ "+\n"
+ "+static void tmem_objnode_free(struct tmem_objnode *objnode)\n"
+ "+{\n"
+ "+\tstruct tmem_pool *pool;\n"
+ "+\tint i;\n"
+ "+\n"
+ "+\tBUG_ON(objnode == NULL);\n"
+ "+\tfor (i = 0; i < OBJNODE_TREE_MAP_SIZE; i++)\n"
+ "+\t\tBUG_ON(objnode->slots[i] != NULL);\n"
+ "+\tASSERT_SENTINEL(objnode, OBJNODE);\n"
+ "+\tINVERT_SENTINEL(objnode, OBJNODE);\n"
+ "+\tBUG_ON(objnode->obj == NULL);\n"
+ "+\tASSERT_SENTINEL(objnode->obj, OBJ);\n"
+ "+\tpool = objnode->obj->pool;\n"
+ "+\tBUG_ON(pool == NULL);\n"
+ "+\tASSERT_SENTINEL(pool, POOL);\n"
+ "+\tobjnode->obj->objnode_count--;\n"
+ "+\tobjnode->obj = NULL;\n"
+ "+\t(*tmem_hostops.objnode_free)(objnode, pool);\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * lookup index in object and return associated pampd (or NULL if not found)\n"
+ "+ */\n"
+ "+static void **__tmem_pampd_lookup_in_obj(struct tmem_obj *obj, uint32_t index)\n"
+ "+{\n"
+ "+\tunsigned int height, shift;\n"
+ "+\tstruct tmem_objnode **slot = NULL;\n"
+ "+\n"
+ "+\tBUG_ON(obj == NULL);\n"
+ "+\tASSERT_SENTINEL(obj, OBJ);\n"
+ "+\tBUG_ON(obj->pool == NULL);\n"
+ "+\tASSERT_SENTINEL(obj->pool, POOL);\n"
+ "+\n"
+ "+\theight = obj->objnode_tree_height;\n"
+ "+\tif (index > tmem_objnode_tree_h2max[obj->objnode_tree_height])\n"
+ "+\t\tgoto out;\n"
+ "+\tif (height == 0 && obj->objnode_tree_root) {\n"
+ "+\t\tslot = &obj->objnode_tree_root;\n"
+ "+\t\tgoto out;\n"
+ "+\t}\n"
+ "+\tshift = (height-1) * OBJNODE_TREE_MAP_SHIFT;\n"
+ "+\tslot = &obj->objnode_tree_root;\n"
+ "+\twhile (height > 0) {\n"
+ "+\t\tif (*slot == NULL)\n"
+ "+\t\t\tgoto out;\n"
+ "+\t\tslot = (struct tmem_objnode **)\n"
+ "+\t\t\t((*slot)->slots +\n"
+ "+\t\t\t ((index >> shift) & OBJNODE_TREE_MAP_MASK));\n"
+ "+\t\tshift -= OBJNODE_TREE_MAP_SHIFT;\n"
+ "+\t\theight--;\n"
+ "+\t}\n"
+ "+out:\n"
+ "+\treturn slot != NULL ? (void **)slot : NULL;\n"
+ "+}\n"
+ "+\n"
+ "+static void *tmem_pampd_lookup_in_obj(struct tmem_obj *obj, uint32_t index)\n"
+ "+{\n"
+ "+\tstruct tmem_objnode **slot;\n"
+ "+\n"
+ "+\tslot = (struct tmem_objnode **)__tmem_pampd_lookup_in_obj(obj, index);\n"
+ "+\treturn slot != NULL ? *slot : NULL;\n"
+ "+}\n"
+ "+\n"
+ "+static void *tmem_pampd_replace_in_obj(struct tmem_obj *obj, uint32_t index,\n"
+ "+\t\t\t\t\tvoid *new_pampd)\n"
+ "+{\n"
+ "+\tstruct tmem_objnode **slot;\n"
+ "+\tvoid *ret = NULL;\n"
+ "+\n"
+ "+\tslot = (struct tmem_objnode **)__tmem_pampd_lookup_in_obj(obj, index);\n"
+ "+\tif ((slot != NULL) && (*slot != NULL)) {\n"
+ "+\t\tvoid *old_pampd = *(void **)slot;\n"
+ "+\t\t*(void **)slot = new_pampd;\n"
+ "+\t\t(*tmem_pamops.free)(old_pampd, obj->pool, NULL, 0);\n"
+ "+\t\tret = new_pampd;\n"
+ "+\t}\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+static int tmem_pampd_add_to_obj(struct tmem_obj *obj, uint32_t index,\n"
+ "+\t\t\t\t\tvoid *pampd)\n"
+ "+{\n"
+ "+\tint ret = 0;\n"
+ "+\tstruct tmem_objnode *objnode = NULL, *newnode, *slot;\n"
+ "+\tunsigned int height, shift;\n"
+ "+\tint offset = 0;\n"
+ "+\n"
+ "+\t/* if necessary, extend the tree to be higher  */\n"
+ "+\tif (index > tmem_objnode_tree_h2max[obj->objnode_tree_height]) {\n"
+ "+\t\theight = obj->objnode_tree_height + 1;\n"
+ "+\t\tif (index > tmem_objnode_tree_h2max[height])\n"
+ "+\t\t\twhile (index > tmem_objnode_tree_h2max[height])\n"
+ "+\t\t\t\theight++;\n"
+ "+\t\tif (obj->objnode_tree_root == NULL) {\n"
+ "+\t\t\tobj->objnode_tree_height = height;\n"
+ "+\t\t\tgoto insert;\n"
+ "+\t\t}\n"
+ "+\t\tdo {\n"
+ "+\t\t\tnewnode = tmem_objnode_alloc(obj);\n"
+ "+\t\t\tif (!newnode) {\n"
+ "+\t\t\t\tret = -ENOMEM;\n"
+ "+\t\t\t\tgoto out;\n"
+ "+\t\t\t}\n"
+ "+\t\t\tnewnode->slots[0] = obj->objnode_tree_root;\n"
+ "+\t\t\tnewnode->slots_in_use = 1;\n"
+ "+\t\t\tobj->objnode_tree_root = newnode;\n"
+ "+\t\t\tobj->objnode_tree_height++;\n"
+ "+\t\t} while (height > obj->objnode_tree_height);\n"
+ "+\t}\n"
+ "+insert:\n"
+ "+\tslot = obj->objnode_tree_root;\n"
+ "+\theight = obj->objnode_tree_height;\n"
+ "+\tshift = (height-1) * OBJNODE_TREE_MAP_SHIFT;\n"
+ "+\twhile (height > 0) {\n"
+ "+\t\tif (slot == NULL) {\n"
+ "+\t\t\t/* add a child objnode.  */\n"
+ "+\t\t\tslot = tmem_objnode_alloc(obj);\n"
+ "+\t\t\tif (!slot) {\n"
+ "+\t\t\t\tret = -ENOMEM;\n"
+ "+\t\t\t\tgoto out;\n"
+ "+\t\t\t}\n"
+ "+\t\t\tif (objnode) {\n"
+ "+\n"
+ "+\t\t\t\tobjnode->slots[offset] = slot;\n"
+ "+\t\t\t\tobjnode->slots_in_use++;\n"
+ "+\t\t\t} else\n"
+ "+\t\t\t\tobj->objnode_tree_root = slot;\n"
+ "+\t\t}\n"
+ "+\t\t/* go down a level */\n"
+ "+\t\toffset = (index >> shift) & OBJNODE_TREE_MAP_MASK;\n"
+ "+\t\tobjnode = slot;\n"
+ "+\t\tslot = objnode->slots[offset];\n"
+ "+\t\tshift -= OBJNODE_TREE_MAP_SHIFT;\n"
+ "+\t\theight--;\n"
+ "+\t}\n"
+ "+\tBUG_ON(slot != NULL);\n"
+ "+\tif (objnode) {\n"
+ "+\t\tobjnode->slots_in_use++;\n"
+ "+\t\tobjnode->slots[offset] = pampd;\n"
+ "+\t} else\n"
+ "+\t\tobj->objnode_tree_root = pampd;\n"
+ "+\tobj->pampd_count++;\n"
+ "+out:\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+static void *tmem_pampd_delete_from_obj(struct tmem_obj *obj, uint32_t index)\n"
+ "+{\n"
+ "+\tstruct tmem_objnode_tree_path path[OBJNODE_TREE_MAX_PATH + 1];\n"
+ "+\tstruct tmem_objnode_tree_path *pathp = path;\n"
+ "+\tstruct tmem_objnode *slot = NULL;\n"
+ "+\tunsigned int height, shift;\n"
+ "+\tint offset;\n"
+ "+\n"
+ "+\tBUG_ON(obj == NULL);\n"
+ "+\tASSERT_SENTINEL(obj, OBJ);\n"
+ "+\tBUG_ON(obj->pool == NULL);\n"
+ "+\tASSERT_SENTINEL(obj->pool, POOL);\n"
+ "+\theight = obj->objnode_tree_height;\n"
+ "+\tif (index > tmem_objnode_tree_h2max[height])\n"
+ "+\t\tgoto out;\n"
+ "+\tslot = obj->objnode_tree_root;\n"
+ "+\tif (height == 0 && obj->objnode_tree_root) {\n"
+ "+\t\tobj->objnode_tree_root = NULL;\n"
+ "+\t\tgoto out;\n"
+ "+\t}\n"
+ "+\tshift = (height - 1) * OBJNODE_TREE_MAP_SHIFT;\n"
+ "+\tpathp->objnode = NULL;\n"
+ "+\tdo {\n"
+ "+\t\tif (slot == NULL)\n"
+ "+\t\t\tgoto out;\n"
+ "+\t\tpathp++;\n"
+ "+\t\toffset = (index >> shift) & OBJNODE_TREE_MAP_MASK;\n"
+ "+\t\tpathp->offset = offset;\n"
+ "+\t\tpathp->objnode = slot;\n"
+ "+\t\tslot = slot->slots[offset];\n"
+ "+\t\tshift -= OBJNODE_TREE_MAP_SHIFT;\n"
+ "+\t\theight--;\n"
+ "+\t} while (height > 0);\n"
+ "+\tif (slot == NULL)\n"
+ "+\t\tgoto out;\n"
+ "+\twhile (pathp->objnode) {\n"
+ "+\t\tpathp->objnode->slots[pathp->offset] = NULL;\n"
+ "+\t\tpathp->objnode->slots_in_use--;\n"
+ "+\t\tif (pathp->objnode->slots_in_use) {\n"
+ "+\t\t\tif (pathp->objnode == obj->objnode_tree_root) {\n"
+ "+\t\t\t\twhile (obj->objnode_tree_height > 0 &&\n"
+ "+\t\t\t\t  obj->objnode_tree_root->slots_in_use == 1 &&\n"
+ "+\t\t\t\t  obj->objnode_tree_root->slots[0]) {\n"
+ "+\t\t\t\t\tstruct tmem_objnode *to_free =\n"
+ "+\t\t\t\t\t\tobj->objnode_tree_root;\n"
+ "+\n"
+ "+\t\t\t\t\tobj->objnode_tree_root =\n"
+ "+\t\t\t\t\t\t\tto_free->slots[0];\n"
+ "+\t\t\t\t\tobj->objnode_tree_height--;\n"
+ "+\t\t\t\t\tto_free->slots[0] = NULL;\n"
+ "+\t\t\t\t\tto_free->slots_in_use = 0;\n"
+ "+\t\t\t\t\ttmem_objnode_free(to_free);\n"
+ "+\t\t\t\t}\n"
+ "+\t\t\t}\n"
+ "+\t\t\tgoto out;\n"
+ "+\t\t}\n"
+ "+\t\ttmem_objnode_free(pathp->objnode); /* 0 slots used, free it */\n"
+ "+\t\tpathp--;\n"
+ "+\t}\n"
+ "+\tobj->objnode_tree_height = 0;\n"
+ "+\tobj->objnode_tree_root = NULL;\n"
+ "+\n"
+ "+out:\n"
+ "+\tif (slot != NULL)\n"
+ "+\t\tobj->pampd_count--;\n"
+ "+\tBUG_ON(obj->pampd_count < 0);\n"
+ "+\treturn slot;\n"
+ "+}\n"
+ "+\n"
+ "+/* recursively walk the objnode_tree destroying pampds and objnodes */\n"
+ "+static void tmem_objnode_node_destroy(struct tmem_obj *obj,\n"
+ "+\t\t\t\t\tstruct tmem_objnode *objnode,\n"
+ "+\t\t\t\t\tunsigned int ht)\n"
+ "+{\n"
+ "+\tint i;\n"
+ "+\n"
+ "+\tif (ht == 0)\n"
+ "+\t\treturn;\n"
+ "+\tfor (i = 0; i < OBJNODE_TREE_MAP_SIZE; i++) {\n"
+ "+\t\tif (objnode->slots[i]) {\n"
+ "+\t\t\tif (ht == 1) {\n"
+ "+\t\t\t\tobj->pampd_count--;\n"
+ "+\t\t\t\t(*tmem_pamops.free)(objnode->slots[i],\n"
+ "+\t\t\t\t\t\tobj->pool, NULL, 0);\n"
+ "+\t\t\t\tobjnode->slots[i] = NULL;\n"
+ "+\t\t\t\tcontinue;\n"
+ "+\t\t\t}\n"
+ "+\t\t\ttmem_objnode_node_destroy(obj, objnode->slots[i], ht-1);\n"
+ "+\t\t\ttmem_objnode_free(objnode->slots[i]);\n"
+ "+\t\t\tobjnode->slots[i] = NULL;\n"
+ "+\t\t}\n"
+ "+\t}\n"
+ "+}\n"
+ "+\n"
+ "+static void tmem_pampd_destroy_all_in_obj(struct tmem_obj *obj)\n"
+ "+{\n"
+ "+\tif (obj->objnode_tree_root == NULL)\n"
+ "+\t\treturn;\n"
+ "+\tif (obj->objnode_tree_height == 0) {\n"
+ "+\t\tobj->pampd_count--;\n"
+ "+\t\t(*tmem_pamops.free)(obj->objnode_tree_root, obj->pool, NULL, 0);\n"
+ "+\t} else {\n"
+ "+\t\ttmem_objnode_node_destroy(obj, obj->objnode_tree_root,\n"
+ "+\t\t\t\t\tobj->objnode_tree_height);\n"
+ "+\t\ttmem_objnode_free(obj->objnode_tree_root);\n"
+ "+\t\tobj->objnode_tree_height = 0;\n"
+ "+\t}\n"
+ "+\tobj->objnode_tree_root = NULL;\n"
+ "+\t(*tmem_pamops.free_obj)(obj->pool, obj);\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * Tmem is operated on by a set of well-defined actions:\n"
+ "+ * \"put\", \"get\", \"flush\", \"flush_object\", \"new pool\" and \"destroy pool\".\n"
+ "+ * (The tmem ABI allows for subpages and exchanges but these operations\n"
+ "+ * are not included in this implementation.)\n"
+ "+ *\n"
+ "+ * These \"tmem core\" operations are implemented in the following functions.\n"
+ "+ */\n"
+ "+\n"
+ "+/*\n"
+ "+ * \"Put\" a page, e.g. copy a page from the kernel into newly allocated\n"
+ "+ * PAM space (if such space is available).  Tmem_put is complicated by\n"
+ "+ * a corner case: What if a page with matching handle already exists in\n"
+ "+ * tmem?  To guarantee coherency, one of two actions is necessary: Either\n"
+ "+ * the data for the page must be overwritten, or the page must be\n"
+ "+ * \"flushed\" so that the data is not accessible to a subsequent \"get\".\n"
+ "+ * Since these \"duplicate puts\" are relatively rare, this implementation\n"
+ "+ * always flushes for simplicity.\n"
+ "+ */\n"
+ "+int tmem_put(struct tmem_pool *pool, struct tmem_oid *oidp, uint32_t index,\n"
+ "+\t\tchar *data, size_t size, bool raw, bool ephemeral)\n"
+ "+{\n"
+ "+\tstruct tmem_obj *obj = NULL, *objfound = NULL, *objnew = NULL;\n"
+ "+\tvoid *pampd = NULL, *pampd_del = NULL;\n"
+ "+\tint ret = -ENOMEM;\n"
+ "+\tstruct tmem_hashbucket *hb;\n"
+ "+\n"
+ "+\thb = &pool->hashbucket[tmem_oid_hash(oidp)];\n"
+ "+\tspin_lock(&hb->lock);\n"
+ "+\tobj = objfound = tmem_obj_find(hb, oidp);\n"
+ "+\tif (obj != NULL) {\n"
+ "+\t\tpampd = tmem_pampd_lookup_in_obj(objfound, index);\n"
+ "+\t\tif (pampd != NULL) {\n"
+ "+\t\t\t/* if found, is a dup put, flush the old one */\n"
+ "+\t\t\tpampd_del = tmem_pampd_delete_from_obj(obj, index);\n"
+ "+\t\t\tBUG_ON(pampd_del != pampd);\n"
+ "+\t\t\t(*tmem_pamops.free)(pampd, pool, oidp, index);\n"
+ "+\t\t\tif (obj->pampd_count == 0) {\n"
+ "+\t\t\t\tobjnew = obj;\n"
+ "+\t\t\t\tobjfound = NULL;\n"
+ "+\t\t\t}\n"
+ "+\t\t\tpampd = NULL;\n"
+ "+\t\t}\n"
+ "+\t} else {\n"
+ "+\t\tobj = objnew = (*tmem_hostops.obj_alloc)(pool);\n"
+ "+\t\tif (unlikely(obj == NULL)) {\n"
+ "+\t\t\tret = -ENOMEM;\n"
+ "+\t\t\tgoto out;\n"
+ "+\t\t}\n"
+ "+\t\ttmem_obj_init(obj, hb, pool, oidp);\n"
+ "+\t}\n"
+ "+\tBUG_ON(obj == NULL);\n"
+ "+\tBUG_ON(((objnew != obj) && (objfound != obj)) || (objnew == objfound));\n"
+ "+\tpampd = (*tmem_pamops.create)(data, size, raw, ephemeral,\n"
+ "+\t\t\t\t\tobj->pool, &obj->oid, index);\n"
+ "+\tif (unlikely(pampd == NULL))\n"
+ "+\t\tgoto free;\n"
+ "+\tret = tmem_pampd_add_to_obj(obj, index, pampd);\n"
+ "+\tif (unlikely(ret == -ENOMEM))\n"
+ "+\t\t/* may have partially built objnode tree (\"stump\") */\n"
+ "+\t\tgoto delete_and_free;\n"
+ "+\tgoto out;\n"
+ "+\n"
+ "+delete_and_free:\n"
+ "+\t(void)tmem_pampd_delete_from_obj(obj, index);\n"
+ "+free:\n"
+ "+\tif (pampd)\n"
+ "+\t\t(*tmem_pamops.free)(pampd, pool, NULL, 0);\n"
+ "+\tif (objnew) {\n"
+ "+\t\ttmem_obj_free(objnew, hb);\n"
+ "+\t\t(*tmem_hostops.obj_free)(objnew, pool);\n"
+ "+\t}\n"
+ "+out:\n"
+ "+\tspin_unlock(&hb->lock);\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * \"Get\" a page, e.g. if one can be found, copy the tmem page with the\n"
+ "+ * matching handle from PAM space to the kernel.  By tmem definition,\n"
+ "+ * when a \"get\" is successful on an ephemeral page, the page is \"flushed\",\n"
+ "+ * and when a \"get\" is successful on a persistent page, the page is retained\n"
+ "+ * in tmem.  Note that to preserve\n"
+ "+ * coherency, \"get\" can never be skipped if tmem contains the data.\n"
+ "+ * That is, if a get is done with a certain handle and fails, any\n"
+ "+ * subsequent \"get\" must also fail (unless of course there is a\n"
+ "+ * \"put\" done with the same handle).\n"
+ "+\n"
+ "+ */\n"
+ "+int tmem_get(struct tmem_pool *pool, struct tmem_oid *oidp, uint32_t index,\n"
+ "+\t\tchar *data, size_t *size, bool raw, int get_and_free)\n"
+ "+{\n"
+ "+\tstruct tmem_obj *obj;\n"
+ "+\tvoid *pampd;\n"
+ "+\tbool ephemeral = is_ephemeral(pool);\n"
+ "+\tint ret = -1;\n"
+ "+\tstruct tmem_hashbucket *hb;\n"
+ "+\tbool free = (get_and_free == 1) || ((get_and_free == 0) && ephemeral);\n"
+ "+\tbool lock_held = false;\n"
+ "+\n"
+ "+\thb = &pool->hashbucket[tmem_oid_hash(oidp)];\n"
+ "+\tspin_lock(&hb->lock);\n"
+ "+\tlock_held = true;\n"
+ "+\tobj = tmem_obj_find(hb, oidp);\n"
+ "+\tif (obj == NULL)\n"
+ "+\t\tgoto out;\n"
+ "+\tif (free)\n"
+ "+\t\tpampd = tmem_pampd_delete_from_obj(obj, index);\n"
+ "+\telse\n"
+ "+\t\tpampd = tmem_pampd_lookup_in_obj(obj, index);\n"
+ "+\tif (pampd == NULL)\n"
+ "+\t\tgoto out;\n"
+ "+\tif (free) {\n"
+ "+\t\tif (obj->pampd_count == 0) {\n"
+ "+\t\t\ttmem_obj_free(obj, hb);\n"
+ "+\t\t\t(*tmem_hostops.obj_free)(obj, pool);\n"
+ "+\t\t\tobj = NULL;\n"
+ "+\t\t}\n"
+ "+\t}\n"
+ "+\tif (tmem_pamops.is_remote(pampd)) {\n"
+ "+\t\tlock_held = false;\n"
+ "+\t\tspin_unlock(&hb->lock);\n"
+ "+\t}\n"
+ "+\tif (free)\n"
+ "+\t\tret = (*tmem_pamops.get_data_and_free)(\n"
+ "+\t\t\t\tdata, size, raw, pampd, pool, oidp, index);\n"
+ "+\telse\n"
+ "+\t\tret = (*tmem_pamops.get_data)(\n"
+ "+\t\t\t\tdata, size, raw, pampd, pool, oidp, index);\n"
+ "+\tif (ret < 0)\n"
+ "+\t\tgoto out;\n"
+ "+\tret = 0;\n"
+ "+out:\n"
+ "+\tif (lock_held)\n"
+ "+\t\tspin_unlock(&hb->lock);\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * If a page in tmem matches the handle, \"flush\" this page from tmem such\n"
+ "+ * that any subsequent \"get\" does not succeed (unless, of course, there\n"
+ "+ * was another \"put\" with the same handle).\n"
+ "+ */\n"
+ "+int tmem_flush_page(struct tmem_pool *pool,\n"
+ "+\t\t\t\tstruct tmem_oid *oidp, uint32_t index)\n"
+ "+{\n"
+ "+\tstruct tmem_obj *obj;\n"
+ "+\tvoid *pampd;\n"
+ "+\tint ret = -1;\n"
+ "+\tstruct tmem_hashbucket *hb;\n"
+ "+\n"
+ "+\thb = &pool->hashbucket[tmem_oid_hash(oidp)];\n"
+ "+\tspin_lock(&hb->lock);\n"
+ "+\tobj = tmem_obj_find(hb, oidp);\n"
+ "+\tif (obj == NULL)\n"
+ "+\t\tgoto out;\n"
+ "+\tpampd = tmem_pampd_delete_from_obj(obj, index);\n"
+ "+\tif (pampd == NULL)\n"
+ "+\t\tgoto out;\n"
+ "+\t(*tmem_pamops.free)(pampd, pool, oidp, index);\n"
+ "+\tif (obj->pampd_count == 0) {\n"
+ "+\t\ttmem_obj_free(obj, hb);\n"
+ "+\t\t(*tmem_hostops.obj_free)(obj, pool);\n"
+ "+\t}\n"
+ "+\tret = 0;\n"
+ "+\n"
+ "+out:\n"
+ "+\tspin_unlock(&hb->lock);\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * If a page in tmem matches the handle, replace the page so that any\n"
+ "+ * subsequent \"get\" gets the new page.  Returns 0 if\n"
+ "+ * there was a page to replace, else returns -1.\n"
+ "+ */\n"
+ "+int tmem_replace(struct tmem_pool *pool, struct tmem_oid *oidp,\n"
+ "+\t\t\tuint32_t index, void *new_pampd)\n"
+ "+{\n"
+ "+\tstruct tmem_obj *obj;\n"
+ "+\tint ret = -1;\n"
+ "+\tstruct tmem_hashbucket *hb;\n"
+ "+\n"
+ "+\thb = &pool->hashbucket[tmem_oid_hash(oidp)];\n"
+ "+\tspin_lock(&hb->lock);\n"
+ "+\tobj = tmem_obj_find(hb, oidp);\n"
+ "+\tif (obj == NULL)\n"
+ "+\t\tgoto out;\n"
+ "+\tnew_pampd = tmem_pampd_replace_in_obj(obj, index, new_pampd);\n"
+ "+\tret = (*tmem_pamops.replace_in_obj)(new_pampd, obj);\n"
+ "+out:\n"
+ "+\tspin_unlock(&hb->lock);\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * \"Flush\" all pages in tmem matching this oid.\n"
+ "+ */\n"
+ "+int tmem_flush_object(struct tmem_pool *pool, struct tmem_oid *oidp)\n"
+ "+{\n"
+ "+\tstruct tmem_obj *obj;\n"
+ "+\tstruct tmem_hashbucket *hb;\n"
+ "+\tint ret = -1;\n"
+ "+\n"
+ "+\thb = &pool->hashbucket[tmem_oid_hash(oidp)];\n"
+ "+\tspin_lock(&hb->lock);\n"
+ "+\tobj = tmem_obj_find(hb, oidp);\n"
+ "+\tif (obj == NULL)\n"
+ "+\t\tgoto out;\n"
+ "+\ttmem_pampd_destroy_all_in_obj(obj);\n"
+ "+\ttmem_obj_free(obj, hb);\n"
+ "+\t(*tmem_hostops.obj_free)(obj, pool);\n"
+ "+\tret = 0;\n"
+ "+\n"
+ "+out:\n"
+ "+\tspin_unlock(&hb->lock);\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * \"Flush\" all pages (and tmem_objs) from this tmem_pool and disable\n"
+ "+ * all subsequent access to this tmem_pool.\n"
+ "+ */\n"
+ "+int tmem_destroy_pool(struct tmem_pool *pool)\n"
+ "+{\n"
+ "+\tint ret = -1;\n"
+ "+\n"
+ "+\tif (pool == NULL)\n"
+ "+\t\tgoto out;\n"
+ "+\ttmem_pool_flush(pool, 1);\n"
+ "+\tret = 0;\n"
+ "+out:\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+static LIST_HEAD(tmem_global_pool_list);\n"
+ "+\n"
+ "+/*\n"
+ "+ * Create a new tmem_pool with the provided flag and return\n"
+ "+ * a pool id provided by the tmem host implementation.\n"
+ "+ */\n"
+ "+void tmem_new_pool(struct tmem_pool *pool, uint32_t flags)\n"
+ "+{\n"
+ "+\tint persistent = flags & TMEM_POOL_PERSIST;\n"
+ "+\tint shared = flags & TMEM_POOL_SHARED;\n"
+ "+\tstruct tmem_hashbucket *hb = &pool->hashbucket[0];\n"
+ "+\tint i;\n"
+ "+\n"
+ "+\tfor (i = 0; i < TMEM_HASH_BUCKETS; i++, hb++) {\n"
+ "+\t\thb->obj_rb_root = RB_ROOT;\n"
+ "+\t\tspin_lock_init(&hb->lock);\n"
+ "+\t}\n"
+ "+\tINIT_LIST_HEAD(&pool->pool_list);\n"
+ "+\tatomic_set(&pool->obj_count, 0);\n"
+ "+\tSET_SENTINEL(pool, POOL);\n"
+ "+\tlist_add_tail(&pool->pool_list, &tmem_global_pool_list);\n"
+ "+\tpool->persistent = persistent;\n"
+ "+\tpool->shared = shared;\n"
+ "+}\n"
+ "diff --git a/drivers/staging/ramster/tmem.h b/drivers/staging/ramster/tmem.h\n"
+ "new file mode 100644\n"
+ "index 0000000..ed147c4\n"
+ "--- /dev/null\n"
+ "+++ b/drivers/staging/ramster/tmem.h\n"
+ "@@ -0,0 +1,206 @@\n"
+ "+/*\n"
+ "+ * tmem.h\n"
+ "+ *\n"
+ "+ * Transcendent memory\n"
+ "+ *\n"
+ "+ * Copyright (c) 2009-2011, Dan Magenheimer, Oracle Corp.\n"
+ "+ */\n"
+ "+\n"
+ "+#ifndef _TMEM_H_\n"
+ "+#define _TMEM_H_\n"
+ "+\n"
+ "+#include <linux/types.h>\n"
+ "+#include <linux/highmem.h>\n"
+ "+#include <linux/hash.h>\n"
+ "+#include <linux/atomic.h>\n"
+ "+\n"
+ "+/*\n"
+ "+ * These are pre-defined by the Xen<->Linux ABI\n"
+ "+ */\n"
+ "+#define TMEM_PUT_PAGE\t\t\t4\n"
+ "+#define TMEM_GET_PAGE\t\t\t5\n"
+ "+#define TMEM_FLUSH_PAGE\t\t\t6\n"
+ "+#define TMEM_FLUSH_OBJECT\t\t7\n"
+ "+#define TMEM_POOL_PERSIST\t\t1\n"
+ "+#define TMEM_POOL_SHARED\t\t2\n"
+ "+#define TMEM_POOL_PRECOMPRESSED\t\t4\n"
+ "+#define TMEM_POOL_PAGESIZE_SHIFT\t4\n"
+ "+#define TMEM_POOL_PAGESIZE_MASK\t\t0xf\n"
+ "+#define TMEM_POOL_RESERVED_BITS\t\t0x00ffff00\n"
+ "+\n"
+ "+/*\n"
+ "+ * sentinels have proven very useful for debugging but can be removed\n"
+ "+ * or disabled before final merge.\n"
+ "+ */\n"
+ "+#define SENTINELS\n"
+ "+#ifdef SENTINELS\n"
+ "+#define DECL_SENTINEL uint32_t sentinel;\n"
+ "+#define SET_SENTINEL(_x, _y) (_x->sentinel = _y##_SENTINEL)\n"
+ "+#define INVERT_SENTINEL(_x, _y) (_x->sentinel = ~_y##_SENTINEL)\n"
+ "+#define ASSERT_SENTINEL(_x, _y) WARN_ON(_x->sentinel != _y##_SENTINEL)\n"
+ "+#define ASSERT_INVERTED_SENTINEL(_x, _y) WARN_ON(_x->sentinel != ~_y##_SENTINEL)\n"
+ "+#else\n"
+ "+#define DECL_SENTINEL\n"
+ "+#define SET_SENTINEL(_x, _y) do { } while (0)\n"
+ "+#define INVERT_SENTINEL(_x, _y) do { } while (0)\n"
+ "+#define ASSERT_SENTINEL(_x, _y) do { } while (0)\n"
+ "+#define ASSERT_INVERTED_SENTINEL(_x, _y) do { } while (0)\n"
+ "+#endif\n"
+ "+\n"
+ "+#define ASSERT_SPINLOCK(_l)\tWARN_ON(!spin_is_locked(_l))\n"
+ "+\n"
+ "+/*\n"
+ "+ * A pool is the highest-level data structure managed by tmem and\n"
+ "+ * usually corresponds to a large independent set of pages such as\n"
+ "+ * a filesystem.  Each pool has an id, and certain attributes and counters.\n"
+ "+ * It also contains a set of hash buckets, each of which contains an rbtree\n"
+ "+ * of objects and a lock to manage concurrency within the pool.\n"
+ "+ */\n"
+ "+\n"
+ "+#define TMEM_HASH_BUCKET_BITS\t8\n"
+ "+#define TMEM_HASH_BUCKETS\t(1<<TMEM_HASH_BUCKET_BITS)\n"
+ "+\n"
+ "+struct tmem_hashbucket {\n"
+ "+\tstruct rb_root obj_rb_root;\n"
+ "+\tspinlock_t lock;\n"
+ "+};\n"
+ "+\n"
+ "+struct tmem_pool {\n"
+ "+\tvoid *client; /* \"up\" for some clients, avoids table lookup */\n"
+ "+\tstruct list_head pool_list;\n"
+ "+\tuint32_t pool_id;\n"
+ "+\tbool persistent;\n"
+ "+\tbool shared;\n"
+ "+\tatomic_t obj_count;\n"
+ "+\tatomic_t refcount;\n"
+ "+\tstruct tmem_hashbucket hashbucket[TMEM_HASH_BUCKETS];\n"
+ "+\tDECL_SENTINEL\n"
+ "+};\n"
+ "+\n"
+ "+#define is_persistent(_p)  (_p->persistent)\n"
+ "+#define is_ephemeral(_p)   (!(_p->persistent))\n"
+ "+\n"
+ "+/*\n"
+ "+ * An object id (\"oid\") is large: 192-bits (to ensure, for example, files\n"
+ "+ * in a modern filesystem can be uniquely identified).\n"
+ "+ */\n"
+ "+\n"
+ "+struct tmem_oid {\n"
+ "+\tuint64_t oid[3];\n"
+ "+};\n"
+ "+\n"
+ "+static inline void tmem_oid_set_invalid(struct tmem_oid *oidp)\n"
+ "+{\n"
+ "+\toidp->oid[0] = oidp->oid[1] = oidp->oid[2] = -1UL;\n"
+ "+}\n"
+ "+\n"
+ "+static inline bool tmem_oid_valid(struct tmem_oid *oidp)\n"
+ "+{\n"
+ "+\treturn oidp->oid[0] != -1UL || oidp->oid[1] != -1UL ||\n"
+ "+\t\toidp->oid[2] != -1UL;\n"
+ "+}\n"
+ "+\n"
+ "+static inline int tmem_oid_compare(struct tmem_oid *left,\n"
+ "+\t\t\t\t\tstruct tmem_oid *right)\n"
+ "+{\n"
+ "+\tint ret;\n"
+ "+\n"
+ "+\tif (left->oid[2] == right->oid[2]) {\n"
+ "+\t\tif (left->oid[1] == right->oid[1]) {\n"
+ "+\t\t\tif (left->oid[0] == right->oid[0])\n"
+ "+\t\t\t\tret = 0;\n"
+ "+\t\t\telse if (left->oid[0] < right->oid[0])\n"
+ "+\t\t\t\tret = -1;\n"
+ "+\t\t\telse\n"
+ "+\t\t\t\treturn 1;\n"
+ "+\t\t} else if (left->oid[1] < right->oid[1])\n"
+ "+\t\t\tret = -1;\n"
+ "+\t\telse\n"
+ "+\t\t\tret = 1;\n"
+ "+\t} else if (left->oid[2] < right->oid[2])\n"
+ "+\t\tret = -1;\n"
+ "+\telse\n"
+ "+\t\tret = 1;\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+static inline unsigned tmem_oid_hash(struct tmem_oid *oidp)\n"
+ "+{\n"
+ "+\treturn hash_long(oidp->oid[0] ^ oidp->oid[1] ^ oidp->oid[2],\n"
+ "+\t\t\t\tTMEM_HASH_BUCKET_BITS);\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * A tmem_obj contains an identifier (oid), pointers to the parent\n"
+ "+ * pool and the rb_tree to which it belongs, counters, and an ordered\n"
+ "+ * set of pampds, structured in a radix-tree-like tree.  The intermediate\n"
+ "+ * nodes of the tree are called tmem_objnodes.\n"
+ "+ */\n"
+ "+\n"
+ "+struct tmem_objnode;\n"
+ "+\n"
+ "+struct tmem_obj {\n"
+ "+\tstruct tmem_oid oid;\n"
+ "+\tstruct tmem_pool *pool;\n"
+ "+\tstruct rb_node rb_tree_node;\n"
+ "+\tstruct tmem_objnode *objnode_tree_root;\n"
+ "+\tunsigned int objnode_tree_height;\n"
+ "+\tunsigned long objnode_count;\n"
+ "+\tlong pampd_count;\n"
+ "+\tvoid *extra; /* for private use by pampd implementation */\n"
+ "+\tDECL_SENTINEL\n"
+ "+};\n"
+ "+\n"
+ "+#define OBJNODE_TREE_MAP_SHIFT 6\n"
+ "+#define OBJNODE_TREE_MAP_SIZE (1UL << OBJNODE_TREE_MAP_SHIFT)\n"
+ "+#define OBJNODE_TREE_MAP_MASK (OBJNODE_TREE_MAP_SIZE-1)\n"
+ "+#define OBJNODE_TREE_INDEX_BITS (8 /* CHAR_BIT */ * sizeof(unsigned long))\n"
+ "+#define OBJNODE_TREE_MAX_PATH \\\n"
+ "+\t\t(OBJNODE_TREE_INDEX_BITS/OBJNODE_TREE_MAP_SHIFT + 2)\n"
+ "+\n"
+ "+struct tmem_objnode {\n"
+ "+\tstruct tmem_obj *obj;\n"
+ "+\tDECL_SENTINEL\n"
+ "+\tvoid *slots[OBJNODE_TREE_MAP_SIZE];\n"
+ "+\tunsigned int slots_in_use;\n"
+ "+};\n"
+ "+\n"
+ "+/* pampd abstract datatype methods provided by the PAM implementation */\n"
+ "+struct tmem_pamops {\n"
+ "+\tvoid *(*create)(char *, size_t, bool, int,\n"
+ "+\t\t\tstruct tmem_pool *, struct tmem_oid *, uint32_t);\n"
+ "+\tint (*get_data)(char *, size_t *, bool, void *, struct tmem_pool *,\n"
+ "+\t\t\t\tstruct tmem_oid *, uint32_t);\n"
+ "+\tint (*get_data_and_free)(char *, size_t *, bool, void *,\n"
+ "+\t\t\t\tstruct tmem_pool *, struct tmem_oid *,\n"
+ "+\t\t\t\tuint32_t);\n"
+ "+\tvoid (*free)(void *, struct tmem_pool *, struct tmem_oid *, uint32_t);\n"
+ "+\tvoid (*free_obj)(struct tmem_pool *, struct tmem_obj *);\n"
+ "+\tbool (*is_remote)(void *);\n"
+ "+\tvoid (*new_obj)(struct tmem_obj *);\n"
+ "+\tint (*replace_in_obj)(void *, struct tmem_obj *);\n"
+ "+};\n"
+ "+extern void tmem_register_pamops(struct tmem_pamops *m);\n"
+ "+\n"
+ "+/* memory allocation methods provided by the host implementation */\n"
+ "+struct tmem_hostops {\n"
+ "+\tstruct tmem_obj *(*obj_alloc)(struct tmem_pool *);\n"
+ "+\tvoid (*obj_free)(struct tmem_obj *, struct tmem_pool *);\n"
+ "+\tstruct tmem_objnode *(*objnode_alloc)(struct tmem_pool *);\n"
+ "+\tvoid (*objnode_free)(struct tmem_objnode *, struct tmem_pool *);\n"
+ "+};\n"
+ "+extern void tmem_register_hostops(struct tmem_hostops *m);\n"
+ "+\n"
+ "+/* core tmem accessor functions */\n"
+ "+extern int tmem_put(struct tmem_pool *, struct tmem_oid *, uint32_t index,\n"
+ "+\t\t\tchar *, size_t, bool, bool);\n"
+ "+extern int tmem_get(struct tmem_pool *, struct tmem_oid *, uint32_t index,\n"
+ "+\t\t\tchar *, size_t *, bool, int);\n"
+ "+extern int tmem_replace(struct tmem_pool *, struct tmem_oid *, uint32_t index,\n"
+ "+\t\t\tvoid *);\n"
+ "+extern int tmem_flush_page(struct tmem_pool *, struct tmem_oid *,\n"
+ "+\t\t\tuint32_t index);\n"
+ "+extern int tmem_flush_object(struct tmem_pool *, struct tmem_oid *);\n"
+ "+extern int tmem_destroy_pool(struct tmem_pool *);\n"
+ "+extern void tmem_new_pool(struct tmem_pool *, uint32_t);\n"
+ "+#endif /* _TMEM_H */\n"
+ "diff --git a/drivers/staging/ramster/zcache-main.c b/drivers/staging/ramster/zcache-main.c\n"
+ "new file mode 100644\n"
+ "index 0000000..cd0ed84\n"
+ "--- /dev/null\n"
+ "+++ b/drivers/staging/ramster/zcache-main.c\n"
+ "@@ -0,0 +1,2003 @@\n"
+ "+/*\n"
+ "+ * zcache.c\n"
+ "+ *\n"
+ "+ * Copyright (c) 2010,2011, Dan Magenheimer, Oracle Corp.\n"
+ "+ * Copyright (c) 2010,2011, Nitin Gupta\n"
+ "+ *\n"
+ "+ * Zcache provides an in-kernel \"host implementation\" for transcendent memory\n"
+ "+ * and, thus indirectly, for cleancache and frontswap.  Zcache includes two\n"
+ "+ * page-accessible memory [1] interfaces, both utilizing lzo1x compression:\n"
+ "+ * 1) \"compression buddies\" (\"zbud\") is used for ephemeral pages\n"
+ "+ * 2) xvmalloc is used for persistent pages.\n"
+ "+ * Xvmalloc (based on the TLSF allocator) has very low fragmentation\n"
+ "+ * so maximizes space efficiency, while zbud allows pairs (and potentially,\n"
+ "+ * in the future, more than a pair of) compressed pages to be closely linked\n"
+ "+ * so that reclaiming can be done via the kernel's physical-page-oriented\n"
+ "+ * \"shrinker\" interface.\n"
+ "+ *\n"
+ "+ * [1] For a definition of page-accessible memory (aka PAM), see:\n"
+ "+ *   http://marc.info/?l=linux-mm&m=127811271605009\n"
+ "+ */\n"
+ "+\n"
+ "+#include <linux/module.h>\n"
+ "+#include <linux/cpu.h>\n"
+ "+#include <linux/highmem.h>\n"
+ "+#include <linux/list.h>\n"
+ "+#include <linux/lzo.h>\n"
+ "+#include <linux/slab.h>\n"
+ "+#include <linux/spinlock.h>\n"
+ "+#include <linux/types.h>\n"
+ "+#include <linux/atomic.h>\n"
+ "+#include <linux/math64.h>\n"
+ "+#include \"tmem.h\"\n"
+ "+\n"
+ "+#include \"../zram/xvmalloc.h\" /* if built in drivers/staging */\n"
+ "+\n"
+ "+#if (!defined(CONFIG_CLEANCACHE) && !defined(CONFIG_FRONTSWAP))\n"
+ "+#error \"zcache is useless without CONFIG_CLEANCACHE or CONFIG_FRONTSWAP\"\n"
+ "+#endif\n"
+ "+#ifdef CONFIG_CLEANCACHE\n"
+ "+#include <linux/cleancache.h>\n"
+ "+#endif\n"
+ "+#ifdef CONFIG_FRONTSWAP\n"
+ "+#include <linux/frontswap.h>\n"
+ "+#endif\n"
+ "+\n"
+ "+#if 0\n"
+ "+/* this is more aggressive but may cause other problems? */\n"
+ "+#define ZCACHE_GFP_MASK\t(GFP_ATOMIC | __GFP_NORETRY | __GFP_NOWARN)\n"
+ "+#else\n"
+ "+#define ZCACHE_GFP_MASK \\\n"
+ "+\t(__GFP_FS | __GFP_NORETRY | __GFP_NOWARN | __GFP_NOMEMALLOC)\n"
+ "+#endif\n"
+ "+\n"
+ "+#define MAX_POOLS_PER_CLIENT 16\n"
+ "+\n"
+ "+#define MAX_CLIENTS 16\n"
+ "+#define LOCAL_CLIENT ((uint16_t)-1)\n"
+ "+\n"
+ "+MODULE_LICENSE(\"GPL\");\n"
+ "+\n"
+ "+struct zcache_client {\n"
+ "+\tstruct tmem_pool *tmem_pools[MAX_POOLS_PER_CLIENT];\n"
+ "+\tstruct xv_pool *xvpool;\n"
+ "+\tbool allocated;\n"
+ "+\tatomic_t refcount;\n"
+ "+};\n"
+ "+\n"
+ "+static struct zcache_client zcache_host;\n"
+ "+static struct zcache_client zcache_clients[MAX_CLIENTS];\n"
+ "+\n"
+ "+static inline uint16_t get_client_id_from_client(struct zcache_client *cli)\n"
+ "+{\n"
+ "+\tBUG_ON(cli == NULL);\n"
+ "+\tif (cli == &zcache_host)\n"
+ "+\t\treturn LOCAL_CLIENT;\n"
+ "+\treturn cli - &zcache_clients[0];\n"
+ "+}\n"
+ "+\n"
+ "+static inline bool is_local_client(struct zcache_client *cli)\n"
+ "+{\n"
+ "+\treturn cli == &zcache_host;\n"
+ "+}\n"
+ "+\n"
+ "+/**********\n"
+ "+ * Compression buddies (\"zbud\") provides for packing two (or, possibly\n"
+ "+ * in the future, more) compressed ephemeral pages into a single \"raw\"\n"
+ "+ * (physical) page and tracking them with data structures so that\n"
+ "+ * the raw pages can be easily reclaimed.\n"
+ "+ *\n"
+ "+ * A zbud page (\"zbpg\") is an aligned page containing a list_head,\n"
+ "+ * a lock, and two \"zbud headers\".  The remainder of the physical\n"
+ "+ * page is divided up into aligned 64-byte \"chunks\" which contain\n"
+ "+ * the compressed data for zero, one, or two zbuds.  Each zbpg\n"
+ "+ * resides on: (1) an \"unused list\" if it has no zbuds; (2) a\n"
+ "+ * \"buddied\" list if it is fully populated  with two zbuds; or\n"
+ "+ * (3) one of PAGE_SIZE/64 \"unbuddied\" lists indexed by how many chunks\n"
+ "+ * the one unbuddied zbud uses.  The data inside a zbpg cannot be\n"
+ "+ * read or written unless the zbpg's lock is held.\n"
+ "+ */\n"
+ "+\n"
+ "+#define ZBH_SENTINEL  0x43214321\n"
+ "+#define ZBPG_SENTINEL  0xdeadbeef\n"
+ "+\n"
+ "+#define ZBUD_MAX_BUDS 2\n"
+ "+\n"
+ "+struct zbud_hdr {\n"
+ "+\tuint16_t client_id;\n"
+ "+\tuint16_t pool_id;\n"
+ "+\tstruct tmem_oid oid;\n"
+ "+\tuint32_t index;\n"
+ "+\tuint16_t size; /* compressed size in bytes, zero means unused */\n"
+ "+\tDECL_SENTINEL\n"
+ "+};\n"
+ "+\n"
+ "+struct zbud_page {\n"
+ "+\tstruct list_head bud_list;\n"
+ "+\tspinlock_t lock;\n"
+ "+\tstruct zbud_hdr buddy[ZBUD_MAX_BUDS];\n"
+ "+\tDECL_SENTINEL\n"
+ "+\t/* followed by NUM_CHUNK aligned CHUNK_SIZE-byte chunks */\n"
+ "+};\n"
+ "+\n"
+ "+#define CHUNK_SHIFT\t6\n"
+ "+#define CHUNK_SIZE\t(1 << CHUNK_SHIFT)\n"
+ "+#define CHUNK_MASK\t(~(CHUNK_SIZE-1))\n"
+ "+#define NCHUNKS\t\t(((PAGE_SIZE - sizeof(struct zbud_page)) & \\\n"
+ "+\t\t\t\tCHUNK_MASK) >> CHUNK_SHIFT)\n"
+ "+#define MAX_CHUNK\t(NCHUNKS-1)\n"
+ "+\n"
+ "+static struct {\n"
+ "+\tstruct list_head list;\n"
+ "+\tunsigned count;\n"
+ "+} zbud_unbuddied[NCHUNKS];\n"
+ "+/* list N contains pages with N chunks USED and NCHUNKS-N unused */\n"
+ "+/* element 0 is never used but optimizing that isn't worth it */\n"
+ "+static unsigned long zbud_cumul_chunk_counts[NCHUNKS];\n"
+ "+\n"
+ "+struct list_head zbud_buddied_list;\n"
+ "+static unsigned long zcache_zbud_buddied_count;\n"
+ "+\n"
+ "+/* protects the buddied list and all unbuddied lists */\n"
+ "+static DEFINE_SPINLOCK(zbud_budlists_spinlock);\n"
+ "+\n"
+ "+static LIST_HEAD(zbpg_unused_list);\n"
+ "+static unsigned long zcache_zbpg_unused_list_count;\n"
+ "+\n"
+ "+/* protects the unused page list */\n"
+ "+static DEFINE_SPINLOCK(zbpg_unused_list_spinlock);\n"
+ "+\n"
+ "+static atomic_t zcache_zbud_curr_raw_pages;\n"
+ "+static atomic_t zcache_zbud_curr_zpages;\n"
+ "+static unsigned long zcache_zbud_curr_zbytes;\n"
+ "+static unsigned long zcache_zbud_cumul_zpages;\n"
+ "+static unsigned long zcache_zbud_cumul_zbytes;\n"
+ "+static unsigned long zcache_compress_poor;\n"
+ "+static unsigned long zcache_mean_compress_poor;\n"
+ "+\n"
+ "+/* forward references */\n"
+ "+static void *zcache_get_free_page(void);\n"
+ "+static void zcache_free_page(void *p);\n"
+ "+\n"
+ "+/*\n"
+ "+ * zbud helper functions\n"
+ "+ */\n"
+ "+\n"
+ "+static inline unsigned zbud_max_buddy_size(void)\n"
+ "+{\n"
+ "+\treturn MAX_CHUNK << CHUNK_SHIFT;\n"
+ "+}\n"
+ "+\n"
+ "+static inline unsigned zbud_size_to_chunks(unsigned size)\n"
+ "+{\n"
+ "+\tBUG_ON(size == 0 || size > zbud_max_buddy_size());\n"
+ "+\treturn (size + CHUNK_SIZE - 1) >> CHUNK_SHIFT;\n"
+ "+}\n"
+ "+\n"
+ "+static inline int zbud_budnum(struct zbud_hdr *zh)\n"
+ "+{\n"
+ "+\tunsigned offset = (unsigned long)zh & (PAGE_SIZE - 1);\n"
+ "+\tstruct zbud_page *zbpg = NULL;\n"
+ "+\tunsigned budnum = -1U;\n"
+ "+\tint i;\n"
+ "+\n"
+ "+\tfor (i = 0; i < ZBUD_MAX_BUDS; i++)\n"
+ "+\t\tif (offset == offsetof(typeof(*zbpg), buddy[i])) {\n"
+ "+\t\t\tbudnum = i;\n"
+ "+\t\t\tbreak;\n"
+ "+\t\t}\n"
+ "+\tBUG_ON(budnum == -1U);\n"
+ "+\treturn budnum;\n"
+ "+}\n"
+ "+\n"
+ "+static char *zbud_data(struct zbud_hdr *zh, unsigned size)\n"
+ "+{\n"
+ "+\tstruct zbud_page *zbpg;\n"
+ "+\tchar *p;\n"
+ "+\tunsigned budnum;\n"
+ "+\n"
+ "+\tASSERT_SENTINEL(zh, ZBH);\n"
+ "+\tbudnum = zbud_budnum(zh);\n"
+ "+\tBUG_ON(size == 0 || size > zbud_max_buddy_size());\n"
+ "+\tzbpg = container_of(zh, struct zbud_page, buddy[budnum]);\n"
+ "+\tASSERT_SPINLOCK(&zbpg->lock);\n"
+ "+\tp = (char *)zbpg;\n"
+ "+\tif (budnum == 0)\n"
+ "+\t\tp += ((sizeof(struct zbud_page) + CHUNK_SIZE - 1) &\n"
+ "+\t\t\t\t\t\t\tCHUNK_MASK);\n"
+ "+\telse if (budnum == 1)\n"
+ "+\t\tp += PAGE_SIZE - ((size + CHUNK_SIZE - 1) & CHUNK_MASK);\n"
+ "+\treturn p;\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * zbud raw page management\n"
+ "+ */\n"
+ "+\n"
+ "+static struct zbud_page *zbud_alloc_raw_page(void)\n"
+ "+{\n"
+ "+\tstruct zbud_page *zbpg = NULL;\n"
+ "+\tstruct zbud_hdr *zh0, *zh1;\n"
+ "+\tbool recycled = 0;\n"
+ "+\n"
+ "+\t/* if any pages on the zbpg list, use one */\n"
+ "+\tspin_lock(&zbpg_unused_list_spinlock);\n"
+ "+\tif (!list_empty(&zbpg_unused_list)) {\n"
+ "+\t\tzbpg = list_first_entry(&zbpg_unused_list,\n"
+ "+\t\t\t\tstruct zbud_page, bud_list);\n"
+ "+\t\tlist_del_init(&zbpg->bud_list);\n"
+ "+\t\tzcache_zbpg_unused_list_count--;\n"
+ "+\t\trecycled = 1;\n"
+ "+\t}\n"
+ "+\tspin_unlock(&zbpg_unused_list_spinlock);\n"
+ "+\tif (zbpg == NULL)\n"
+ "+\t\t/* none on zbpg list, try to get a kernel page */\n"
+ "+\t\tzbpg = zcache_get_free_page();\n"
+ "+\tif (likely(zbpg != NULL)) {\n"
+ "+\t\tINIT_LIST_HEAD(&zbpg->bud_list);\n"
+ "+\t\tzh0 = &zbpg->buddy[0]; zh1 = &zbpg->buddy[1];\n"
+ "+\t\tspin_lock_init(&zbpg->lock);\n"
+ "+\t\tif (recycled) {\n"
+ "+\t\t\tASSERT_INVERTED_SENTINEL(zbpg, ZBPG);\n"
+ "+\t\t\tSET_SENTINEL(zbpg, ZBPG);\n"
+ "+\t\t\tBUG_ON(zh0->size != 0 || tmem_oid_valid(&zh0->oid));\n"
+ "+\t\t\tBUG_ON(zh1->size != 0 || tmem_oid_valid(&zh1->oid));\n"
+ "+\t\t} else {\n"
+ "+\t\t\tatomic_inc(&zcache_zbud_curr_raw_pages);\n"
+ "+\t\t\tINIT_LIST_HEAD(&zbpg->bud_list);\n"
+ "+\t\t\tSET_SENTINEL(zbpg, ZBPG);\n"
+ "+\t\t\tzh0->size = 0; zh1->size = 0;\n"
+ "+\t\t\ttmem_oid_set_invalid(&zh0->oid);\n"
+ "+\t\t\ttmem_oid_set_invalid(&zh1->oid);\n"
+ "+\t\t}\n"
+ "+\t}\n"
+ "+\treturn zbpg;\n"
+ "+}\n"
+ "+\n"
+ "+static void zbud_free_raw_page(struct zbud_page *zbpg)\n"
+ "+{\n"
+ "+\tstruct zbud_hdr *zh0 = &zbpg->buddy[0], *zh1 = &zbpg->buddy[1];\n"
+ "+\n"
+ "+\tASSERT_SENTINEL(zbpg, ZBPG);\n"
+ "+\tBUG_ON(!list_empty(&zbpg->bud_list));\n"
+ "+\tASSERT_SPINLOCK(&zbpg->lock);\n"
+ "+\tBUG_ON(zh0->size != 0 || tmem_oid_valid(&zh0->oid));\n"
+ "+\tBUG_ON(zh1->size != 0 || tmem_oid_valid(&zh1->oid));\n"
+ "+\tINVERT_SENTINEL(zbpg, ZBPG);\n"
+ "+\tspin_unlock(&zbpg->lock);\n"
+ "+\tspin_lock(&zbpg_unused_list_spinlock);\n"
+ "+\tlist_add(&zbpg->bud_list, &zbpg_unused_list);\n"
+ "+\tzcache_zbpg_unused_list_count++;\n"
+ "+\tspin_unlock(&zbpg_unused_list_spinlock);\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * core zbud handling routines\n"
+ "+ */\n"
+ "+\n"
+ "+static unsigned zbud_free(struct zbud_hdr *zh)\n"
+ "+{\n"
+ "+\tunsigned size;\n"
+ "+\n"
+ "+\tASSERT_SENTINEL(zh, ZBH);\n"
+ "+\tBUG_ON(!tmem_oid_valid(&zh->oid));\n"
+ "+\tsize = zh->size;\n"
+ "+\tBUG_ON(zh->size == 0 || zh->size > zbud_max_buddy_size());\n"
+ "+\tzh->size = 0;\n"
+ "+\ttmem_oid_set_invalid(&zh->oid);\n"
+ "+\tINVERT_SENTINEL(zh, ZBH);\n"
+ "+\tzcache_zbud_curr_zbytes -= size;\n"
+ "+\tatomic_dec(&zcache_zbud_curr_zpages);\n"
+ "+\treturn size;\n"
+ "+}\n"
+ "+\n"
+ "+static void zbud_free_and_delist(struct zbud_hdr *zh)\n"
+ "+{\n"
+ "+\tunsigned chunks;\n"
+ "+\tstruct zbud_hdr *zh_other;\n"
+ "+\tunsigned budnum = zbud_budnum(zh), size;\n"
+ "+\tstruct zbud_page *zbpg =\n"
+ "+\t\tcontainer_of(zh, struct zbud_page, buddy[budnum]);\n"
+ "+\n"
+ "+\tspin_lock(&zbpg->lock);\n"
+ "+\tif (list_empty(&zbpg->bud_list)) {\n"
+ "+\t\t/* ignore zombie page... see zbud_evict_pages() */\n"
+ "+\t\tspin_unlock(&zbpg->lock);\n"
+ "+\t\treturn;\n"
+ "+\t}\n"
+ "+\tsize = zbud_free(zh);\n"
+ "+\tASSERT_SPINLOCK(&zbpg->lock);\n"
+ "+\tzh_other = &zbpg->buddy[(budnum == 0) ? 1 : 0];\n"
+ "+\tif (zh_other->size == 0) { /* was unbuddied: unlist and free */\n"
+ "+\t\tchunks = zbud_size_to_chunks(size) ;\n"
+ "+\t\tspin_lock(&zbud_budlists_spinlock);\n"
+ "+\t\tBUG_ON(list_empty(&zbud_unbuddied[chunks].list));\n"
+ "+\t\tlist_del_init(&zbpg->bud_list);\n"
+ "+\t\tzbud_unbuddied[chunks].count--;\n"
+ "+\t\tspin_unlock(&zbud_budlists_spinlock);\n"
+ "+\t\tzbud_free_raw_page(zbpg);\n"
+ "+\t} else { /* was buddied: move remaining buddy to unbuddied list */\n"
+ "+\t\tchunks = zbud_size_to_chunks(zh_other->size) ;\n"
+ "+\t\tspin_lock(&zbud_budlists_spinlock);\n"
+ "+\t\tlist_del_init(&zbpg->bud_list);\n"
+ "+\t\tzcache_zbud_buddied_count--;\n"
+ "+\t\tlist_add_tail(&zbpg->bud_list, &zbud_unbuddied[chunks].list);\n"
+ "+\t\tzbud_unbuddied[chunks].count++;\n"
+ "+\t\tspin_unlock(&zbud_budlists_spinlock);\n"
+ "+\t\tspin_unlock(&zbpg->lock);\n"
+ "+\t}\n"
+ "+}\n"
+ "+\n"
+ "+static struct zbud_hdr *zbud_create(uint16_t client_id, uint16_t pool_id,\n"
+ "+\t\t\t\t\tstruct tmem_oid *oid,\n"
+ "+\t\t\t\t\tuint32_t index, struct page *page,\n"
+ "+\t\t\t\t\tvoid *cdata, unsigned size)\n"
+ "+{\n"
+ "+\tstruct zbud_hdr *zh0, *zh1, *zh = NULL;\n"
+ "+\tstruct zbud_page *zbpg = NULL, *ztmp;\n"
+ "+\tunsigned nchunks;\n"
+ "+\tchar *to;\n"
+ "+\tint i, found_good_buddy = 0;\n"
+ "+\n"
+ "+\tnchunks = zbud_size_to_chunks(size) ;\n"
+ "+\tfor (i = MAX_CHUNK - nchunks + 1; i > 0; i--) {\n"
+ "+\t\tspin_lock(&zbud_budlists_spinlock);\n"
+ "+\t\tif (!list_empty(&zbud_unbuddied[i].list)) {\n"
+ "+\t\t\tlist_for_each_entry_safe(zbpg, ztmp,\n"
+ "+\t\t\t\t    &zbud_unbuddied[i].list, bud_list) {\n"
+ "+\t\t\t\tif (spin_trylock(&zbpg->lock)) {\n"
+ "+\t\t\t\t\tfound_good_buddy = i;\n"
+ "+\t\t\t\t\tgoto found_unbuddied;\n"
+ "+\t\t\t\t}\n"
+ "+\t\t\t}\n"
+ "+\t\t}\n"
+ "+\t\tspin_unlock(&zbud_budlists_spinlock);\n"
+ "+\t}\n"
+ "+\t/* didn't find a good buddy, try allocating a new page */\n"
+ "+\tzbpg = zbud_alloc_raw_page();\n"
+ "+\tif (unlikely(zbpg == NULL))\n"
+ "+\t\tgoto out;\n"
+ "+\t/* ok, have a page, now compress the data before taking locks */\n"
+ "+\tspin_lock(&zbpg->lock);\n"
+ "+\tspin_lock(&zbud_budlists_spinlock);\n"
+ "+\tlist_add_tail(&zbpg->bud_list, &zbud_unbuddied[nchunks].list);\n"
+ "+\tzbud_unbuddied[nchunks].count++;\n"
+ "+\tzh = &zbpg->buddy[0];\n"
+ "+\tgoto init_zh;\n"
+ "+\n"
+ "+found_unbuddied:\n"
+ "+\tASSERT_SPINLOCK(&zbpg->lock);\n"
+ "+\tzh0 = &zbpg->buddy[0]; zh1 = &zbpg->buddy[1];\n"
+ "+\tBUG_ON(!((zh0->size == 0) ^ (zh1->size == 0)));\n"
+ "+\tif (zh0->size != 0) { /* buddy0 in use, buddy1 is vacant */\n"
+ "+\t\tASSERT_SENTINEL(zh0, ZBH);\n"
+ "+\t\tzh = zh1;\n"
+ "+\t} else if (zh1->size != 0) { /* buddy1 in use, buddy0 is vacant */\n"
+ "+\t\tASSERT_SENTINEL(zh1, ZBH);\n"
+ "+\t\tzh = zh0;\n"
+ "+\t} else\n"
+ "+\t\tBUG();\n"
+ "+\tlist_del_init(&zbpg->bud_list);\n"
+ "+\tzbud_unbuddied[found_good_buddy].count--;\n"
+ "+\tlist_add_tail(&zbpg->bud_list, &zbud_buddied_list);\n"
+ "+\tzcache_zbud_buddied_count++;\n"
+ "+\n"
+ "+init_zh:\n"
+ "+\tSET_SENTINEL(zh, ZBH);\n"
+ "+\tzh->size = size;\n"
+ "+\tzh->index = index;\n"
+ "+\tzh->oid = *oid;\n"
+ "+\tzh->pool_id = pool_id;\n"
+ "+\tzh->client_id = client_id;\n"
+ "+\t/* can wait to copy the data until the list locks are dropped */\n"
+ "+\tspin_unlock(&zbud_budlists_spinlock);\n"
+ "+\n"
+ "+\tto = zbud_data(zh, size);\n"
+ "+\tmemcpy(to, cdata, size);\n"
+ "+\tspin_unlock(&zbpg->lock);\n"
+ "+\tzbud_cumul_chunk_counts[nchunks]++;\n"
+ "+\tatomic_inc(&zcache_zbud_curr_zpages);\n"
+ "+\tzcache_zbud_cumul_zpages++;\n"
+ "+\tzcache_zbud_curr_zbytes += size;\n"
+ "+\tzcache_zbud_cumul_zbytes += size;\n"
+ "+out:\n"
+ "+\treturn zh;\n"
+ "+}\n"
+ "+\n"
+ "+static int zbud_decompress(struct page *page, struct zbud_hdr *zh)\n"
+ "+{\n"
+ "+\tstruct zbud_page *zbpg;\n"
+ "+\tunsigned budnum = zbud_budnum(zh);\n"
+ "+\tsize_t out_len = PAGE_SIZE;\n"
+ "+\tchar *to_va, *from_va;\n"
+ "+\tunsigned size;\n"
+ "+\tint ret = 0;\n"
+ "+\n"
+ "+\tzbpg = container_of(zh, struct zbud_page, buddy[budnum]);\n"
+ "+\tspin_lock(&zbpg->lock);\n"
+ "+\tif (list_empty(&zbpg->bud_list)) {\n"
+ "+\t\t/* ignore zombie page... see zbud_evict_pages() */\n"
+ "+\t\tret = -EINVAL;\n"
+ "+\t\tgoto out;\n"
+ "+\t}\n"
+ "+\tASSERT_SENTINEL(zh, ZBH);\n"
+ "+\tBUG_ON(zh->size == 0 || zh->size > zbud_max_buddy_size());\n"
+ "+\tto_va = kmap_atomic(page, KM_USER0);\n"
+ "+\tsize = zh->size;\n"
+ "+\tfrom_va = zbud_data(zh, size);\n"
+ "+\tret = lzo1x_decompress_safe(from_va, size, to_va, &out_len);\n"
+ "+\tBUG_ON(ret != LZO_E_OK);\n"
+ "+\tBUG_ON(out_len != PAGE_SIZE);\n"
+ "+\tkunmap_atomic(to_va, KM_USER0);\n"
+ "+out:\n"
+ "+\tspin_unlock(&zbpg->lock);\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * The following routines handle shrinking of ephemeral pages by evicting\n"
+ "+ * pages \"least valuable\" first.\n"
+ "+ */\n"
+ "+\n"
+ "+static unsigned long zcache_evicted_raw_pages;\n"
+ "+static unsigned long zcache_evicted_buddied_pages;\n"
+ "+static unsigned long zcache_evicted_unbuddied_pages;\n"
+ "+\n"
+ "+static struct tmem_pool *zcache_get_pool_by_id(uint16_t cli_id,\n"
+ "+\t\t\t\t\t\tuint16_t poolid);\n"
+ "+static void zcache_put_pool(struct tmem_pool *pool);\n"
+ "+\n"
+ "+/*\n"
+ "+ * Flush and free all zbuds in a zbpg, then free the pageframe\n"
+ "+ */\n"
+ "+static void zbud_evict_zbpg(struct zbud_page *zbpg)\n"
+ "+{\n"
+ "+\tstruct zbud_hdr *zh;\n"
+ "+\tint i, j;\n"
+ "+\tuint32_t pool_id[ZBUD_MAX_BUDS], client_id[ZBUD_MAX_BUDS];\n"
+ "+\tuint32_t index[ZBUD_MAX_BUDS];\n"
+ "+\tstruct tmem_oid oid[ZBUD_MAX_BUDS];\n"
+ "+\tstruct tmem_pool *pool;\n"
+ "+\n"
+ "+\tASSERT_SPINLOCK(&zbpg->lock);\n"
+ "+\tBUG_ON(!list_empty(&zbpg->bud_list));\n"
+ "+\tfor (i = 0, j = 0; i < ZBUD_MAX_BUDS; i++) {\n"
+ "+\t\tzh = &zbpg->buddy[i];\n"
+ "+\t\tif (zh->size) {\n"
+ "+\t\t\tclient_id[j] = zh->client_id;\n"
+ "+\t\t\tpool_id[j] = zh->pool_id;\n"
+ "+\t\t\toid[j] = zh->oid;\n"
+ "+\t\t\tindex[j] = zh->index;\n"
+ "+\t\t\tj++;\n"
+ "+\t\t\tzbud_free(zh);\n"
+ "+\t\t}\n"
+ "+\t}\n"
+ "+\tspin_unlock(&zbpg->lock);\n"
+ "+\tfor (i = 0; i < j; i++) {\n"
+ "+\t\tpool = zcache_get_pool_by_id(client_id[i], pool_id[i]);\n"
+ "+\t\tif (pool != NULL) {\n"
+ "+\t\t\ttmem_flush_page(pool, &oid[i], index[i]);\n"
+ "+\t\t\tzcache_put_pool(pool);\n"
+ "+\t\t}\n"
+ "+\t}\n"
+ "+\tASSERT_SENTINEL(zbpg, ZBPG);\n"
+ "+\tspin_lock(&zbpg->lock);\n"
+ "+\tzbud_free_raw_page(zbpg);\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * Free nr pages.  This code is funky because we want to hold the locks\n"
+ "+ * protecting various lists for as short a time as possible, and in some\n"
+ "+ * circumstances the list may change asynchronously when the list lock is\n"
+ "+ * not held.  In some cases we also trylock not only to avoid waiting on a\n"
+ "+ * page in use by another cpu, but also to avoid potential deadlock due to\n"
+ "+ * lock inversion.\n"
+ "+ */\n"
+ "+static void zbud_evict_pages(int nr)\n"
+ "+{\n"
+ "+\tstruct zbud_page *zbpg;\n"
+ "+\tint i;\n"
+ "+\n"
+ "+\t/* first try freeing any pages on unused list */\n"
+ "+retry_unused_list:\n"
+ "+\tspin_lock_bh(&zbpg_unused_list_spinlock);\n"
+ "+\tif (!list_empty(&zbpg_unused_list)) {\n"
+ "+\t\t/* can't walk list here, since it may change when unlocked */\n"
+ "+\t\tzbpg = list_first_entry(&zbpg_unused_list,\n"
+ "+\t\t\t\tstruct zbud_page, bud_list);\n"
+ "+\t\tlist_del_init(&zbpg->bud_list);\n"
+ "+\t\tzcache_zbpg_unused_list_count--;\n"
+ "+\t\tatomic_dec(&zcache_zbud_curr_raw_pages);\n"
+ "+\t\tspin_unlock_bh(&zbpg_unused_list_spinlock);\n"
+ "+\t\tzcache_free_page(zbpg);\n"
+ "+\t\tzcache_evicted_raw_pages++;\n"
+ "+\t\tif (--nr <= 0)\n"
+ "+\t\t\tgoto out;\n"
+ "+\t\tgoto retry_unused_list;\n"
+ "+\t}\n"
+ "+\tspin_unlock_bh(&zbpg_unused_list_spinlock);\n"
+ "+\n"
+ "+\t/* now try freeing unbuddied pages, starting with least space avail */\n"
+ "+\tfor (i = 0; i < MAX_CHUNK; i++) {\n"
+ "+retry_unbud_list_i:\n"
+ "+\t\tspin_lock_bh(&zbud_budlists_spinlock);\n"
+ "+\t\tif (list_empty(&zbud_unbuddied[i].list)) {\n"
+ "+\t\t\tspin_unlock_bh(&zbud_budlists_spinlock);\n"
+ "+\t\t\tcontinue;\n"
+ "+\t\t}\n"
+ "+\t\tlist_for_each_entry(zbpg, &zbud_unbuddied[i].list, bud_list) {\n"
+ "+\t\t\tif (unlikely(!spin_trylock(&zbpg->lock)))\n"
+ "+\t\t\t\tcontinue;\n"
+ "+\t\t\tlist_del_init(&zbpg->bud_list);\n"
+ "+\t\t\tzbud_unbuddied[i].count--;\n"
+ "+\t\t\tspin_unlock(&zbud_budlists_spinlock);\n"
+ "+\t\t\tzcache_evicted_unbuddied_pages++;\n"
+ "+\t\t\t/* want budlists unlocked when doing zbpg eviction */\n"
+ "+\t\t\tzbud_evict_zbpg(zbpg);\n"
+ "+\t\t\tlocal_bh_enable();\n"
+ "+\t\t\tif (--nr <= 0)\n"
+ "+\t\t\t\tgoto out;\n"
+ "+\t\t\tgoto retry_unbud_list_i;\n"
+ "+\t\t}\n"
+ "+\t\tspin_unlock_bh(&zbud_budlists_spinlock);\n"
+ "+\t}\n"
+ "+\n"
+ "+\t/* as a last resort, free buddied pages */\n"
+ "+retry_bud_list:\n"
+ "+\tspin_lock_bh(&zbud_budlists_spinlock);\n"
+ "+\tif (list_empty(&zbud_buddied_list)) {\n"
+ "+\t\tspin_unlock_bh(&zbud_budlists_spinlock);\n"
+ "+\t\tgoto out;\n"
+ "+\t}\n"
+ "+\tlist_for_each_entry(zbpg, &zbud_buddied_list, bud_list) {\n"
+ "+\t\tif (unlikely(!spin_trylock(&zbpg->lock)))\n"
+ "+\t\t\tcontinue;\n"
+ "+\t\tlist_del_init(&zbpg->bud_list);\n"
+ "+\t\tzcache_zbud_buddied_count--;\n"
+ "+\t\tspin_unlock(&zbud_budlists_spinlock);\n"
+ "+\t\tzcache_evicted_buddied_pages++;\n"
+ "+\t\t/* want budlists unlocked when doing zbpg eviction */\n"
+ "+\t\tzbud_evict_zbpg(zbpg);\n"
+ "+\t\tlocal_bh_enable();\n"
+ "+\t\tif (--nr <= 0)\n"
+ "+\t\t\tgoto out;\n"
+ "+\t\tgoto retry_bud_list;\n"
+ "+\t}\n"
+ "+\tspin_unlock_bh(&zbud_budlists_spinlock);\n"
+ "+out:\n"
+ "+\treturn;\n"
+ "+}\n"
+ "+\n"
+ "+static void zbud_init(void)\n"
+ "+{\n"
+ "+\tint i;\n"
+ "+\n"
+ "+\tINIT_LIST_HEAD(&zbud_buddied_list);\n"
+ "+\tzcache_zbud_buddied_count = 0;\n"
+ "+\tfor (i = 0; i < NCHUNKS; i++) {\n"
+ "+\t\tINIT_LIST_HEAD(&zbud_unbuddied[i].list);\n"
+ "+\t\tzbud_unbuddied[i].count = 0;\n"
+ "+\t}\n"
+ "+}\n"
+ "+\n"
+ "+#ifdef CONFIG_SYSFS\n"
+ "+/*\n"
+ "+ * These sysfs routines show a nice distribution of how many zbpg's are\n"
+ "+ * currently (and have ever been placed) in each unbuddied list.  It's fun\n"
+ "+ * to watch but can probably go away before final merge.\n"
+ "+ */\n"
+ "+static int zbud_show_unbuddied_list_counts(char *buf)\n"
+ "+{\n"
+ "+\tint i;\n"
+ "+\tchar *p = buf;\n"
+ "+\n"
+ "+\tfor (i = 0; i < NCHUNKS; i++)\n"
+ "+\t\tp += sprintf(p, \"%u \", zbud_unbuddied[i].count);\n"
+ "+\treturn p - buf;\n"
+ "+}\n"
+ "+\n"
+ "+static int zbud_show_cumul_chunk_counts(char *buf)\n"
+ "+{\n"
+ "+\tunsigned long i, chunks = 0, total_chunks = 0, sum_total_chunks = 0;\n"
+ "+\tunsigned long total_chunks_lte_21 = 0, total_chunks_lte_32 = 0;\n"
+ "+\tunsigned long total_chunks_lte_42 = 0;\n"
+ "+\tchar *p = buf;\n"
+ "+\n"
+ "+\tfor (i = 0; i < NCHUNKS; i++) {\n"
+ "+\t\tp += sprintf(p, \"%lu \", zbud_cumul_chunk_counts[i]);\n"
+ "+\t\tchunks += zbud_cumul_chunk_counts[i];\n"
+ "+\t\ttotal_chunks += zbud_cumul_chunk_counts[i];\n"
+ "+\t\tsum_total_chunks += i * zbud_cumul_chunk_counts[i];\n"
+ "+\t\tif (i == 21)\n"
+ "+\t\t\ttotal_chunks_lte_21 = total_chunks;\n"
+ "+\t\tif (i == 32)\n"
+ "+\t\t\ttotal_chunks_lte_32 = total_chunks;\n"
+ "+\t\tif (i == 42)\n"
+ "+\t\t\ttotal_chunks_lte_42 = total_chunks;\n"
+ "+\t}\n"
+ "+\tp += sprintf(p, \"<=21:%lu <=32:%lu <=42:%lu, mean:%lu\\n\",\n"
+ "+\t\ttotal_chunks_lte_21, total_chunks_lte_32, total_chunks_lte_42,\n"
+ "+\t\tchunks == 0 ? 0 : sum_total_chunks / chunks);\n"
+ "+\treturn p - buf;\n"
+ "+}\n"
+ "+#endif\n"
+ "+\n"
+ "+/**********\n"
+ "+ * This \"zv\" PAM implementation combines the TLSF-based xvMalloc\n"
+ "+ * with lzo1x compression to maximize the amount of data that can\n"
+ "+ * be packed into a physical page.\n"
+ "+ *\n"
+ "+ * Zv represents a PAM page with the index and object (plus a \"size\" value\n"
+ "+ * necessary for decompression) immediately preceding the compressed data.\n"
+ "+ */\n"
+ "+\n"
+ "+#define ZVH_SENTINEL  0x43214321\n"
+ "+\n"
+ "+struct zv_hdr {\n"
+ "+\tuint32_t pool_id;\n"
+ "+\tstruct tmem_oid oid;\n"
+ "+\tuint32_t index;\n"
+ "+\tDECL_SENTINEL\n"
+ "+};\n"
+ "+\n"
+ "+/* rudimentary policy limits */\n"
+ "+/* total number of persistent pages may not exceed this percentage */\n"
+ "+static unsigned int zv_page_count_policy_percent = 75;\n"
+ "+/*\n"
+ "+ * byte count defining poor compression; pages with greater zsize will be\n"
+ "+ * rejected\n"
+ "+ */\n"
+ "+static unsigned int zv_max_zsize = (PAGE_SIZE / 8) * 7;\n"
+ "+/*\n"
+ "+ * byte count defining poor *mean* compression; pages with greater zsize\n"
+ "+ * will be rejected until sufficient better-compressed pages are accepted\n"
+ "+ * driving the man below this threshold\n"
+ "+ */\n"
+ "+static unsigned int zv_max_mean_zsize = (PAGE_SIZE / 8) * 5;\n"
+ "+\n"
+ "+static unsigned long zv_curr_dist_counts[NCHUNKS];\n"
+ "+static unsigned long zv_cumul_dist_counts[NCHUNKS];\n"
+ "+\n"
+ "+static struct zv_hdr *zv_create(struct xv_pool *xvpool, uint32_t pool_id,\n"
+ "+\t\t\t\tstruct tmem_oid *oid, uint32_t index,\n"
+ "+\t\t\t\tvoid *cdata, unsigned clen)\n"
+ "+{\n"
+ "+\tstruct page *page;\n"
+ "+\tstruct zv_hdr *zv = NULL;\n"
+ "+\tuint32_t offset;\n"
+ "+\tint alloc_size = clen + sizeof(struct zv_hdr);\n"
+ "+\tint chunks = (alloc_size + (CHUNK_SIZE - 1)) >> CHUNK_SHIFT;\n"
+ "+\tint ret;\n"
+ "+\n"
+ "+\tBUG_ON(!irqs_disabled());\n"
+ "+\tBUG_ON(chunks >= NCHUNKS);\n"
+ "+\tret = xv_malloc(xvpool, alloc_size,\n"
+ "+\t\t\t&page, &offset, ZCACHE_GFP_MASK);\n"
+ "+\tif (unlikely(ret))\n"
+ "+\t\tgoto out;\n"
+ "+\tzv_curr_dist_counts[chunks]++;\n"
+ "+\tzv_cumul_dist_counts[chunks]++;\n"
+ "+\tzv = kmap_atomic(page, KM_USER0) + offset;\n"
+ "+\tzv->index = index;\n"
+ "+\tzv->oid = *oid;\n"
+ "+\tzv->pool_id = pool_id;\n"
+ "+\tSET_SENTINEL(zv, ZVH);\n"
+ "+\tmemcpy((char *)zv + sizeof(struct zv_hdr), cdata, clen);\n"
+ "+\tkunmap_atomic(zv, KM_USER0);\n"
+ "+out:\n"
+ "+\treturn zv;\n"
+ "+}\n"
+ "+\n"
+ "+static void zv_free(struct xv_pool *xvpool, struct zv_hdr *zv)\n"
+ "+{\n"
+ "+\tunsigned long flags;\n"
+ "+\tstruct page *page;\n"
+ "+\tuint32_t offset;\n"
+ "+\tuint16_t size = xv_get_object_size(zv);\n"
+ "+\tint chunks = (size + (CHUNK_SIZE - 1)) >> CHUNK_SHIFT;\n"
+ "+\n"
+ "+\tASSERT_SENTINEL(zv, ZVH);\n"
+ "+\tBUG_ON(chunks >= NCHUNKS);\n"
+ "+\tzv_curr_dist_counts[chunks]--;\n"
+ "+\tsize -= sizeof(*zv);\n"
+ "+\tBUG_ON(size == 0);\n"
+ "+\tINVERT_SENTINEL(zv, ZVH);\n"
+ "+\tpage = virt_to_page(zv);\n"
+ "+\toffset = (unsigned long)zv & ~PAGE_MASK;\n"
+ "+\tlocal_irq_save(flags);\n"
+ "+\txv_free(xvpool, page, offset);\n"
+ "+\tlocal_irq_restore(flags);\n"
+ "+}\n"
+ "+\n"
+ "+static void zv_decompress(struct page *page, struct zv_hdr *zv)\n"
+ "+{\n"
+ "+\tsize_t clen = PAGE_SIZE;\n"
+ "+\tchar *to_va;\n"
+ "+\tunsigned size;\n"
+ "+\tint ret;\n"
+ "+\n"
+ "+\tASSERT_SENTINEL(zv, ZVH);\n"
+ "+\tsize = xv_get_object_size(zv) - sizeof(*zv);\n"
+ "+\tBUG_ON(size == 0);\n"
+ "+\tto_va = kmap_atomic(page, KM_USER0);\n"
+ "+\tret = lzo1x_decompress_safe((char *)zv + sizeof(*zv),\n"
+ "+\t\t\t\t\tsize, to_va, &clen);\n"
+ "+\tkunmap_atomic(to_va, KM_USER0);\n"
+ "+\tBUG_ON(ret != LZO_E_OK);\n"
+ "+\tBUG_ON(clen != PAGE_SIZE);\n"
+ "+}\n"
+ "+\n"
+ "+#ifdef CONFIG_SYSFS\n"
+ "+/*\n"
+ "+ * show a distribution of compression stats for zv pages.\n"
+ "+ */\n"
+ "+\n"
+ "+static int zv_curr_dist_counts_show(char *buf)\n"
+ "+{\n"
+ "+\tunsigned long i, n, chunks = 0, sum_total_chunks = 0;\n"
+ "+\tchar *p = buf;\n"
+ "+\n"
+ "+\tfor (i = 0; i < NCHUNKS; i++) {\n"
+ "+\t\tn = zv_curr_dist_counts[i];\n"
+ "+\t\tp += sprintf(p, \"%lu \", n);\n"
+ "+\t\tchunks += n;\n"
+ "+\t\tsum_total_chunks += i * n;\n"
+ "+\t}\n"
+ "+\tp += sprintf(p, \"mean:%lu\\n\",\n"
+ "+\t\tchunks == 0 ? 0 : sum_total_chunks / chunks);\n"
+ "+\treturn p - buf;\n"
+ "+}\n"
+ "+\n"
+ "+static int zv_cumul_dist_counts_show(char *buf)\n"
+ "+{\n"
+ "+\tunsigned long i, n, chunks = 0, sum_total_chunks = 0;\n"
+ "+\tchar *p = buf;\n"
+ "+\n"
+ "+\tfor (i = 0; i < NCHUNKS; i++) {\n"
+ "+\t\tn = zv_cumul_dist_counts[i];\n"
+ "+\t\tp += sprintf(p, \"%lu \", n);\n"
+ "+\t\tchunks += n;\n"
+ "+\t\tsum_total_chunks += i * n;\n"
+ "+\t}\n"
+ "+\tp += sprintf(p, \"mean:%lu\\n\",\n"
+ "+\t\tchunks == 0 ? 0 : sum_total_chunks / chunks);\n"
+ "+\treturn p - buf;\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * setting zv_max_zsize via sysfs causes all persistent (e.g. swap)\n"
+ "+ * pages that don't compress to less than this value (including metadata\n"
+ "+ * overhead) to be rejected.  We don't allow the value to get too close\n"
+ "+ * to PAGE_SIZE.\n"
+ "+ */\n"
+ "+static ssize_t zv_max_zsize_show(struct kobject *kobj,\n"
+ "+\t\t\t\t    struct kobj_attribute *attr,\n"
+ "+\t\t\t\t    char *buf)\n"
+ "+{\n"
+ "+\treturn sprintf(buf, \"%u\\n\", zv_max_zsize);\n"
+ "+}\n"
+ "+\n"
+ "+static ssize_t zv_max_zsize_store(struct kobject *kobj,\n"
+ "+\t\t\t\t    struct kobj_attribute *attr,\n"
+ "+\t\t\t\t    const char *buf, size_t count)\n"
+ "+{\n"
+ "+\tunsigned long val;\n"
+ "+\tint err;\n"
+ "+\n"
+ "+\tif (!capable(CAP_SYS_ADMIN))\n"
+ "+\t\treturn -EPERM;\n"
+ "+\n"
+ "+\terr = strict_strtoul(buf, 10, &val);\n"
+ "+\tif (err || (val == 0) || (val > (PAGE_SIZE / 8) * 7))\n"
+ "+\t\treturn -EINVAL;\n"
+ "+\tzv_max_zsize = val;\n"
+ "+\treturn count;\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * setting zv_max_mean_zsize via sysfs causes all persistent (e.g. swap)\n"
+ "+ * pages that don't compress to less than this value (including metadata\n"
+ "+ * overhead) to be rejected UNLESS the mean compression is also smaller\n"
+ "+ * than this value.  In other words, we are load-balancing-by-zsize the\n"
+ "+ * accepted pages.  Again, we don't allow the value to get too close\n"
+ "+ * to PAGE_SIZE.\n"
+ "+ */\n"
+ "+static ssize_t zv_max_mean_zsize_show(struct kobject *kobj,\n"
+ "+\t\t\t\t    struct kobj_attribute *attr,\n"
+ "+\t\t\t\t    char *buf)\n"
+ "+{\n"
+ "+\treturn sprintf(buf, \"%u\\n\", zv_max_mean_zsize);\n"
+ "+}\n"
+ "+\n"
+ "+static ssize_t zv_max_mean_zsize_store(struct kobject *kobj,\n"
+ "+\t\t\t\t    struct kobj_attribute *attr,\n"
+ "+\t\t\t\t    const char *buf, size_t count)\n"
+ "+{\n"
+ "+\tunsigned long val;\n"
+ "+\tint err;\n"
+ "+\n"
+ "+\tif (!capable(CAP_SYS_ADMIN))\n"
+ "+\t\treturn -EPERM;\n"
+ "+\n"
+ "+\terr = strict_strtoul(buf, 10, &val);\n"
+ "+\tif (err || (val == 0) || (val > (PAGE_SIZE / 8) * 7))\n"
+ "+\t\treturn -EINVAL;\n"
+ "+\tzv_max_mean_zsize = val;\n"
+ "+\treturn count;\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * setting zv_page_count_policy_percent via sysfs sets an upper bound of\n"
+ "+ * persistent (e.g. swap) pages that will be retained according to:\n"
+ "+ *     (zv_page_count_policy_percent * totalram_pages) / 100)\n"
+ "+ * when that limit is reached, further puts will be rejected (until\n"
+ "+ * some pages have been flushed).  Note that, due to compression,\n"
+ "+ * this number may exceed 100; it defaults to 75 and we set an\n"
+ "+ * arbitary limit of 150.  A poor choice will almost certainly result\n"
+ "+ * in OOM's, so this value should only be changed prudently.\n"
+ "+ */\n"
+ "+static ssize_t zv_page_count_policy_percent_show(struct kobject *kobj,\n"
+ "+\t\t\t\t\t\t struct kobj_attribute *attr,\n"
+ "+\t\t\t\t\t\t char *buf)\n"
+ "+{\n"
+ "+\treturn sprintf(buf, \"%u\\n\", zv_page_count_policy_percent);\n"
+ "+}\n"
+ "+\n"
+ "+static ssize_t zv_page_count_policy_percent_store(struct kobject *kobj,\n"
+ "+\t\t\t\t\t\t  struct kobj_attribute *attr,\n"
+ "+\t\t\t\t\t\t  const char *buf, size_t count)\n"
+ "+{\n"
+ "+\tunsigned long val;\n"
+ "+\tint err;\n"
+ "+\n"
+ "+\tif (!capable(CAP_SYS_ADMIN))\n"
+ "+\t\treturn -EPERM;\n"
+ "+\n"
+ "+\terr = strict_strtoul(buf, 10, &val);\n"
+ "+\tif (err || (val == 0) || (val > 150))\n"
+ "+\t\treturn -EINVAL;\n"
+ "+\tzv_page_count_policy_percent = val;\n"
+ "+\treturn count;\n"
+ "+}\n"
+ "+\n"
+ "+static struct kobj_attribute zcache_zv_max_zsize_attr = {\n"
+ "+\t\t.attr = { .name = \"zv_max_zsize\", .mode = 0644 },\n"
+ "+\t\t.show = zv_max_zsize_show,\n"
+ "+\t\t.store = zv_max_zsize_store,\n"
+ "+};\n"
+ "+\n"
+ "+static struct kobj_attribute zcache_zv_max_mean_zsize_attr = {\n"
+ "+\t\t.attr = { .name = \"zv_max_mean_zsize\", .mode = 0644 },\n"
+ "+\t\t.show = zv_max_mean_zsize_show,\n"
+ "+\t\t.store = zv_max_mean_zsize_store,\n"
+ "+};\n"
+ "+\n"
+ "+static struct kobj_attribute zcache_zv_page_count_policy_percent_attr = {\n"
+ "+\t\t.attr = { .name = \"zv_page_count_policy_percent\",\n"
+ "+\t\t\t  .mode = 0644 },\n"
+ "+\t\t.show = zv_page_count_policy_percent_show,\n"
+ "+\t\t.store = zv_page_count_policy_percent_store,\n"
+ "+};\n"
+ "+#endif\n"
+ "+\n"
+ "+/*\n"
+ "+ * zcache core code starts here\n"
+ "+ */\n"
+ "+\n"
+ "+/* useful stats not collected by cleancache or frontswap */\n"
+ "+static unsigned long zcache_flush_total;\n"
+ "+static unsigned long zcache_flush_found;\n"
+ "+static unsigned long zcache_flobj_total;\n"
+ "+static unsigned long zcache_flobj_found;\n"
+ "+static unsigned long zcache_failed_eph_puts;\n"
+ "+static unsigned long zcache_failed_pers_puts;\n"
+ "+\n"
+ "+/*\n"
+ "+ * Tmem operations assume the poolid implies the invoking client.\n"
+ "+ * Zcache only has one client (the kernel itself): LOCAL_CLIENT.\n"
+ "+ * RAMster has each client numbered by cluster node, and a KVM version\n"
+ "+ * of zcache would have one client per guest and each client might\n"
+ "+ * have a poolid==N.\n"
+ "+ */\n"
+ "+static struct tmem_pool *zcache_get_pool_by_id(uint16_t cli_id, uint16_t poolid)\n"
+ "+{\n"
+ "+\tstruct tmem_pool *pool = NULL;\n"
+ "+\tstruct zcache_client *cli = NULL;\n"
+ "+\n"
+ "+\tif (cli_id == LOCAL_CLIENT)\n"
+ "+\t\tcli = &zcache_host;\n"
+ "+\telse {\n"
+ "+\t\tif (cli_id >= MAX_CLIENTS)\n"
+ "+\t\t\tgoto out;\n"
+ "+\t\tcli = &zcache_clients[cli_id];\n"
+ "+\t\tif (cli == NULL)\n"
+ "+\t\t\tgoto out;\n"
+ "+\t\tatomic_inc(&cli->refcount);\n"
+ "+\t}\n"
+ "+\tif (poolid < MAX_POOLS_PER_CLIENT) {\n"
+ "+\t\tpool = cli->tmem_pools[poolid];\n"
+ "+\t\tif (pool != NULL)\n"
+ "+\t\t\tatomic_inc(&pool->refcount);\n"
+ "+\t}\n"
+ "+out:\n"
+ "+\treturn pool;\n"
+ "+}\n"
+ "+\n"
+ "+static void zcache_put_pool(struct tmem_pool *pool)\n"
+ "+{\n"
+ "+\tstruct zcache_client *cli = NULL;\n"
+ "+\n"
+ "+\tif (pool == NULL)\n"
+ "+\t\tBUG();\n"
+ "+\tcli = pool->client;\n"
+ "+\tatomic_dec(&pool->refcount);\n"
+ "+\tatomic_dec(&cli->refcount);\n"
+ "+}\n"
+ "+\n"
+ "+int zcache_new_client(uint16_t cli_id)\n"
+ "+{\n"
+ "+\tstruct zcache_client *cli = NULL;\n"
+ "+\tint ret = -1;\n"
+ "+\n"
+ "+\tif (cli_id == LOCAL_CLIENT)\n"
+ "+\t\tcli = &zcache_host;\n"
+ "+\telse if ((unsigned int)cli_id < MAX_CLIENTS)\n"
+ "+\t\tcli = &zcache_clients[cli_id];\n"
+ "+\tif (cli == NULL)\n"
+ "+\t\tgoto out;\n"
+ "+\tif (cli->allocated)\n"
+ "+\t\tgoto out;\n"
+ "+\tcli->allocated = 1;\n"
+ "+#ifdef CONFIG_FRONTSWAP\n"
+ "+\tcli->xvpool = xv_create_pool();\n"
+ "+\tif (cli->xvpool == NULL)\n"
+ "+\t\tgoto out;\n"
+ "+#endif\n"
+ "+\tret = 0;\n"
+ "+out:\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+/* counters for debugging */\n"
+ "+static unsigned long zcache_failed_get_free_pages;\n"
+ "+static unsigned long zcache_failed_alloc;\n"
+ "+static unsigned long zcache_put_to_flush;\n"
+ "+static unsigned long zcache_aborted_preload;\n"
+ "+static unsigned long zcache_aborted_shrink;\n"
+ "+\n"
+ "+/*\n"
+ "+ * Ensure that memory allocation requests in zcache don't result\n"
+ "+ * in direct reclaim requests via the shrinker, which would cause\n"
+ "+ * an infinite loop.  Maybe a GFP flag would be better?\n"
+ "+ */\n"
+ "+static DEFINE_SPINLOCK(zcache_direct_reclaim_lock);\n"
+ "+\n"
+ "+/*\n"
+ "+ * for now, used named slabs so can easily track usage; later can\n"
+ "+ * either just use kmalloc, or perhaps add a slab-like allocator\n"
+ "+ * to more carefully manage total memory utilization\n"
+ "+ */\n"
+ "+static struct kmem_cache *zcache_objnode_cache;\n"
+ "+static struct kmem_cache *zcache_obj_cache;\n"
+ "+static atomic_t zcache_curr_obj_count = ATOMIC_INIT(0);\n"
+ "+static unsigned long zcache_curr_obj_count_max;\n"
+ "+static atomic_t zcache_curr_objnode_count = ATOMIC_INIT(0);\n"
+ "+static unsigned long zcache_curr_objnode_count_max;\n"
+ "+\n"
+ "+/*\n"
+ "+ * to avoid memory allocation recursion (e.g. due to direct reclaim), we\n"
+ "+ * preload all necessary data structures so the hostops callbacks never\n"
+ "+ * actually do a malloc\n"
+ "+ */\n"
+ "+struct zcache_preload {\n"
+ "+\tvoid *page;\n"
+ "+\tstruct tmem_obj *obj;\n"
+ "+\tint nr;\n"
+ "+\tstruct tmem_objnode *objnodes[OBJNODE_TREE_MAX_PATH];\n"
+ "+};\n"
+ "+static DEFINE_PER_CPU(struct zcache_preload, zcache_preloads) = { 0, };\n"
+ "+\n"
+ "+static int zcache_do_preload(struct tmem_pool *pool)\n"
+ "+{\n"
+ "+\tstruct zcache_preload *kp;\n"
+ "+\tstruct tmem_objnode *objnode;\n"
+ "+\tstruct tmem_obj *obj;\n"
+ "+\tvoid *page;\n"
+ "+\tint ret = -ENOMEM;\n"
+ "+\n"
+ "+\tif (unlikely(zcache_objnode_cache == NULL))\n"
+ "+\t\tgoto out;\n"
+ "+\tif (unlikely(zcache_obj_cache == NULL))\n"
+ "+\t\tgoto out;\n"
+ "+\tif (!spin_trylock(&zcache_direct_reclaim_lock)) {\n"
+ "+\t\tzcache_aborted_preload++;\n"
+ "+\t\tgoto out;\n"
+ "+\t}\n"
+ "+\tpreempt_disable();\n"
+ "+\tkp = &__get_cpu_var(zcache_preloads);\n"
+ "+\twhile (kp->nr < ARRAY_SIZE(kp->objnodes)) {\n"
+ "+\t\tpreempt_enable_no_resched();\n"
+ "+\t\tobjnode = kmem_cache_alloc(zcache_objnode_cache,\n"
+ "+\t\t\t\tZCACHE_GFP_MASK);\n"
+ "+\t\tif (unlikely(objnode == NULL)) {\n"
+ "+\t\t\tzcache_failed_alloc++;\n"
+ "+\t\t\tgoto unlock_out;\n"
+ "+\t\t}\n"
+ "+\t\tpreempt_disable();\n"
+ "+\t\tkp = &__get_cpu_var(zcache_preloads);\n"
+ "+\t\tif (kp->nr < ARRAY_SIZE(kp->objnodes))\n"
+ "+\t\t\tkp->objnodes[kp->nr++] = objnode;\n"
+ "+\t\telse\n"
+ "+\t\t\tkmem_cache_free(zcache_objnode_cache, objnode);\n"
+ "+\t}\n"
+ "+\tpreempt_enable_no_resched();\n"
+ "+\tobj = kmem_cache_alloc(zcache_obj_cache, ZCACHE_GFP_MASK);\n"
+ "+\tif (unlikely(obj == NULL)) {\n"
+ "+\t\tzcache_failed_alloc++;\n"
+ "+\t\tgoto unlock_out;\n"
+ "+\t}\n"
+ "+\tpage = (void *)__get_free_page(ZCACHE_GFP_MASK);\n"
+ "+\tif (unlikely(page == NULL)) {\n"
+ "+\t\tzcache_failed_get_free_pages++;\n"
+ "+\t\tkmem_cache_free(zcache_obj_cache, obj);\n"
+ "+\t\tgoto unlock_out;\n"
+ "+\t}\n"
+ "+\tpreempt_disable();\n"
+ "+\tkp = &__get_cpu_var(zcache_preloads);\n"
+ "+\tif (kp->obj == NULL)\n"
+ "+\t\tkp->obj = obj;\n"
+ "+\telse\n"
+ "+\t\tkmem_cache_free(zcache_obj_cache, obj);\n"
+ "+\tif (kp->page == NULL)\n"
+ "+\t\tkp->page = page;\n"
+ "+\telse\n"
+ "+\t\tfree_page((unsigned long)page);\n"
+ "+\tret = 0;\n"
+ "+unlock_out:\n"
+ "+\tspin_unlock(&zcache_direct_reclaim_lock);\n"
+ "+out:\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+static void *zcache_get_free_page(void)\n"
+ "+{\n"
+ "+\tstruct zcache_preload *kp;\n"
+ "+\tvoid *page;\n"
+ "+\n"
+ "+\tkp = &__get_cpu_var(zcache_preloads);\n"
+ "+\tpage = kp->page;\n"
+ "+\tBUG_ON(page == NULL);\n"
+ "+\tkp->page = NULL;\n"
+ "+\treturn page;\n"
+ "+}\n"
+ "+\n"
+ "+static void zcache_free_page(void *p)\n"
+ "+{\n"
+ "+\tfree_page((unsigned long)p);\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * zcache implementation for tmem host ops\n"
+ "+ */\n"
+ "+\n"
+ "+static struct tmem_objnode *zcache_objnode_alloc(struct tmem_pool *pool)\n"
+ "+{\n"
+ "+\tstruct tmem_objnode *objnode = NULL;\n"
+ "+\tunsigned long count;\n"
+ "+\tstruct zcache_preload *kp;\n"
+ "+\n"
+ "+\tkp = &__get_cpu_var(zcache_preloads);\n"
+ "+\tif (kp->nr <= 0)\n"
+ "+\t\tgoto out;\n"
+ "+\tobjnode = kp->objnodes[kp->nr - 1];\n"
+ "+\tBUG_ON(objnode == NULL);\n"
+ "+\tkp->objnodes[kp->nr - 1] = NULL;\n"
+ "+\tkp->nr--;\n"
+ "+\tcount = atomic_inc_return(&zcache_curr_objnode_count);\n"
+ "+\tif (count > zcache_curr_objnode_count_max)\n"
+ "+\t\tzcache_curr_objnode_count_max = count;\n"
+ "+out:\n"
+ "+\treturn objnode;\n"
+ "+}\n"
+ "+\n"
+ "+static void zcache_objnode_free(struct tmem_objnode *objnode,\n"
+ "+\t\t\t\t\tstruct tmem_pool *pool)\n"
+ "+{\n"
+ "+\tatomic_dec(&zcache_curr_objnode_count);\n"
+ "+\tBUG_ON(atomic_read(&zcache_curr_objnode_count) < 0);\n"
+ "+\tkmem_cache_free(zcache_objnode_cache, objnode);\n"
+ "+}\n"
+ "+\n"
+ "+static struct tmem_obj *zcache_obj_alloc(struct tmem_pool *pool)\n"
+ "+{\n"
+ "+\tstruct tmem_obj *obj = NULL;\n"
+ "+\tunsigned long count;\n"
+ "+\tstruct zcache_preload *kp;\n"
+ "+\n"
+ "+\tkp = &__get_cpu_var(zcache_preloads);\n"
+ "+\tobj = kp->obj;\n"
+ "+\tBUG_ON(obj == NULL);\n"
+ "+\tkp->obj = NULL;\n"
+ "+\tcount = atomic_inc_return(&zcache_curr_obj_count);\n"
+ "+\tif (count > zcache_curr_obj_count_max)\n"
+ "+\t\tzcache_curr_obj_count_max = count;\n"
+ "+\treturn obj;\n"
+ "+}\n"
+ "+\n"
+ "+static void zcache_obj_free(struct tmem_obj *obj, struct tmem_pool *pool)\n"
+ "+{\n"
+ "+\tatomic_dec(&zcache_curr_obj_count);\n"
+ "+\tBUG_ON(atomic_read(&zcache_curr_obj_count) < 0);\n"
+ "+\tkmem_cache_free(zcache_obj_cache, obj);\n"
+ "+}\n"
+ "+\n"
+ "+static struct tmem_hostops zcache_hostops = {\n"
+ "+\t.obj_alloc = zcache_obj_alloc,\n"
+ "+\t.obj_free = zcache_obj_free,\n"
+ "+\t.objnode_alloc = zcache_objnode_alloc,\n"
+ "+\t.objnode_free = zcache_objnode_free,\n"
+ "+};\n"
+ "+\n"
+ "+/*\n"
+ "+ * zcache implementations for PAM page descriptor ops\n"
+ "+ */\n"
+ "+\n"
+ "+static atomic_t zcache_curr_eph_pampd_count = ATOMIC_INIT(0);\n"
+ "+static unsigned long zcache_curr_eph_pampd_count_max;\n"
+ "+static atomic_t zcache_curr_pers_pampd_count = ATOMIC_INIT(0);\n"
+ "+static unsigned long zcache_curr_pers_pampd_count_max;\n"
+ "+\n"
+ "+/* forward reference */\n"
+ "+static int zcache_compress(struct page *from, void **out_va, size_t *out_len);\n"
+ "+\n"
+ "+static void *zcache_pampd_create(char *data, size_t size, bool raw, int eph,\n"
+ "+\t\t\t\tstruct tmem_pool *pool, struct tmem_oid *oid,\n"
+ "+\t\t\t\t uint32_t index)\n"
+ "+{\n"
+ "+\tvoid *pampd = NULL, *cdata;\n"
+ "+\tsize_t clen;\n"
+ "+\tint ret;\n"
+ "+\tunsigned long count;\n"
+ "+\tstruct page *page = (struct page *)(data);\n"
+ "+\tstruct zcache_client *cli = pool->client;\n"
+ "+\tuint16_t client_id = get_client_id_from_client(cli);\n"
+ "+\tunsigned long zv_mean_zsize;\n"
+ "+\tunsigned long curr_pers_pampd_count;\n"
+ "+\tu64 total_zsize;\n"
+ "+\n"
+ "+\tif (eph) {\n"
+ "+\t\tret = zcache_compress(page, &cdata, &clen);\n"
+ "+\t\tif (ret == 0)\n"
+ "+\t\t\tgoto out;\n"
+ "+\t\tif (clen == 0 || clen > zbud_max_buddy_size()) {\n"
+ "+\t\t\tzcache_compress_poor++;\n"
+ "+\t\t\tgoto out;\n"
+ "+\t\t}\n"
+ "+\t\tpampd = (void *)zbud_create(client_id, pool->pool_id, oid,\n"
+ "+\t\t\t\t\t\tindex, page, cdata, clen);\n"
+ "+\t\tif (pampd != NULL) {\n"
+ "+\t\t\tcount = atomic_inc_return(&zcache_curr_eph_pampd_count);\n"
+ "+\t\t\tif (count > zcache_curr_eph_pampd_count_max)\n"
+ "+\t\t\t\tzcache_curr_eph_pampd_count_max = count;\n"
+ "+\t\t}\n"
+ "+\t} else {\n"
+ "+\t\tcurr_pers_pampd_count =\n"
+ "+\t\t\tatomic_read(&zcache_curr_pers_pampd_count);\n"
+ "+\t\tif (curr_pers_pampd_count >\n"
+ "+\t\t    (zv_page_count_policy_percent * totalram_pages) / 100)\n"
+ "+\t\t\tgoto out;\n"
+ "+\t\tret = zcache_compress(page, &cdata, &clen);\n"
+ "+\t\tif (ret == 0)\n"
+ "+\t\t\tgoto out;\n"
+ "+\t\t/* reject if compression is too poor */\n"
+ "+\t\tif (clen > zv_max_zsize) {\n"
+ "+\t\t\tzcache_compress_poor++;\n"
+ "+\t\t\tgoto out;\n"
+ "+\t\t}\n"
+ "+\t\t/* reject if mean compression is too poor */\n"
+ "+\t\tif ((clen > zv_max_mean_zsize) && (curr_pers_pampd_count > 0)) {\n"
+ "+\t\t\ttotal_zsize = xv_get_total_size_bytes(cli->xvpool);\n"
+ "+\t\t\tzv_mean_zsize = div_u64(total_zsize,\n"
+ "+\t\t\t\t\t\tcurr_pers_pampd_count);\n"
+ "+\t\t\tif (zv_mean_zsize > zv_max_mean_zsize) {\n"
+ "+\t\t\t\tzcache_mean_compress_poor++;\n"
+ "+\t\t\t\tgoto out;\n"
+ "+\t\t\t}\n"
+ "+\t\t}\n"
+ "+\t\tpampd = (void *)zv_create(cli->xvpool, pool->pool_id,\n"
+ "+\t\t\t\t\t\toid, index, cdata, clen);\n"
+ "+\t\tif (pampd == NULL)\n"
+ "+\t\t\tgoto out;\n"
+ "+\t\tcount = atomic_inc_return(&zcache_curr_pers_pampd_count);\n"
+ "+\t\tif (count > zcache_curr_pers_pampd_count_max)\n"
+ "+\t\t\tzcache_curr_pers_pampd_count_max = count;\n"
+ "+\t}\n"
+ "+out:\n"
+ "+\treturn pampd;\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * fill the pageframe corresponding to the struct page with the data\n"
+ "+ * from the passed pampd\n"
+ "+ */\n"
+ "+static int zcache_pampd_get_data(char *data, size_t *bufsize, bool raw,\n"
+ "+\t\t\t\t\tvoid *pampd, struct tmem_pool *pool,\n"
+ "+\t\t\t\t\tstruct tmem_oid *oid, uint32_t index)\n"
+ "+{\n"
+ "+\tint ret = 0;\n"
+ "+\n"
+ "+\tBUG_ON(is_ephemeral(pool));\n"
+ "+\tzv_decompress((struct page *)(data), pampd);\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * fill the pageframe corresponding to the struct page with the data\n"
+ "+ * from the passed pampd\n"
+ "+ */\n"
+ "+static int zcache_pampd_get_data_and_free(char *data, size_t *bufsize, bool raw,\n"
+ "+\t\t\t\t\tvoid *pampd, struct tmem_pool *pool,\n"
+ "+\t\t\t\t\tstruct tmem_oid *oid, uint32_t index)\n"
+ "+{\n"
+ "+\tint ret = 0;\n"
+ "+\n"
+ "+\tBUG_ON(!is_ephemeral(pool));\n"
+ "+\tzbud_decompress((struct page *)(data), pampd);\n"
+ "+\tzbud_free_and_delist((struct zbud_hdr *)pampd);\n"
+ "+\tatomic_dec(&zcache_curr_eph_pampd_count);\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+/*\n"
+ "+ * free the pampd and remove it from any zcache lists\n"
+ "+ * pampd must no longer be pointed to from any tmem data structures!\n"
+ "+ */\n"
+ "+static void zcache_pampd_free(void *pampd, struct tmem_pool *pool,\n"
+ "+\t\t\t\tstruct tmem_oid *oid, uint32_t index)\n"
+ "+{\n"
+ "+\tstruct zcache_client *cli = pool->client;\n"
+ "+\n"
+ "+\tif (is_ephemeral(pool)) {\n"
+ "+\t\tzbud_free_and_delist((struct zbud_hdr *)pampd);\n"
+ "+\t\tatomic_dec(&zcache_curr_eph_pampd_count);\n"
+ "+\t\tBUG_ON(atomic_read(&zcache_curr_eph_pampd_count) < 0);\n"
+ "+\t} else {\n"
+ "+\t\tzv_free(cli->xvpool, (struct zv_hdr *)pampd);\n"
+ "+\t\tatomic_dec(&zcache_curr_pers_pampd_count);\n"
+ "+\t\tBUG_ON(atomic_read(&zcache_curr_pers_pampd_count) < 0);\n"
+ "+\t}\n"
+ "+}\n"
+ "+\n"
+ "+static void zcache_pampd_free_obj(struct tmem_pool *pool, struct tmem_obj *obj)\n"
+ "+{\n"
+ "+}\n"
+ "+\n"
+ "+static void zcache_pampd_new_obj(struct tmem_obj *obj)\n"
+ "+{\n"
+ "+}\n"
+ "+\n"
+ "+static int zcache_pampd_replace_in_obj(void *pampd, struct tmem_obj *obj)\n"
+ "+{\n"
+ "+\treturn -1;\n"
+ "+}\n"
+ "+\n"
+ "+static bool zcache_pampd_is_remote(void *pampd)\n"
+ "+{\n"
+ "+\treturn 0;\n"
+ "+}\n"
+ "+\n"
+ "+static struct tmem_pamops zcache_pamops = {\n"
+ "+\t.create = zcache_pampd_create,\n"
+ "+\t.get_data = zcache_pampd_get_data,\n"
+ "+\t.get_data_and_free = zcache_pampd_get_data_and_free,\n"
+ "+\t.free = zcache_pampd_free,\n"
+ "+\t.free_obj = zcache_pampd_free_obj,\n"
+ "+\t.new_obj = zcache_pampd_new_obj,\n"
+ "+\t.replace_in_obj = zcache_pampd_replace_in_obj,\n"
+ "+\t.is_remote = zcache_pampd_is_remote,\n"
+ "+};\n"
+ "+\n"
+ "+/*\n"
+ "+ * zcache compression/decompression and related per-cpu stuff\n"
+ "+ */\n"
+ "+\n"
+ "+#define LZO_WORKMEM_BYTES LZO1X_1_MEM_COMPRESS\n"
+ "+#define LZO_DSTMEM_PAGE_ORDER 1\n"
+ "+static DEFINE_PER_CPU(unsigned char *, zcache_workmem);\n"
+ "+static DEFINE_PER_CPU(unsigned char *, zcache_dstmem);\n"
+ "+\n"
+ "+static int zcache_compress(struct page *from, void **out_va, size_t *out_len)\n"
+ "+{\n"
+ "+\tint ret = 0;\n"
+ "+\tunsigned char *dmem = __get_cpu_var(zcache_dstmem);\n"
+ "+\tunsigned char *wmem = __get_cpu_var(zcache_workmem);\n"
+ "+\tchar *from_va;\n"
+ "+\n"
+ "+\tBUG_ON(!irqs_disabled());\n"
+ "+\tif (unlikely(dmem == NULL || wmem == NULL))\n"
+ "+\t\tgoto out;  /* no buffer, so can't compress */\n"
+ "+\tfrom_va = kmap_atomic(from, KM_USER0);\n"
+ "+\tmb();\n"
+ "+\tret = lzo1x_1_compress(from_va, PAGE_SIZE, dmem, out_len, wmem);\n"
+ "+\tBUG_ON(ret != LZO_E_OK);\n"
+ "+\t*out_va = dmem;\n"
+ "+\tkunmap_atomic(from_va, KM_USER0);\n"
+ "+\tret = 1;\n"
+ "+out:\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+\n"
+ "+static int zcache_cpu_notifier(struct notifier_block *nb,\n"
+ "+\t\t\t\tunsigned long action, void *pcpu)\n"
+ "+{\n"
+ "+\tint cpu = (long)pcpu;\n"
+ "+\tstruct zcache_preload *kp;\n"
+ "+\n"
+ "+\tswitch (action) {\n"
+ "+\tcase CPU_UP_PREPARE:\n"
+ "+\t\tper_cpu(zcache_dstmem, cpu) = (void *)__get_free_pages(\n"
+ "+\t\t\tGFP_KERNEL | __GFP_REPEAT,\n"
+ "+\t\t\tLZO_DSTMEM_PAGE_ORDER),\n"
+ "+\t\tper_cpu(zcache_workmem, cpu) =\n"
+ "+\t\t\tkzalloc(LZO1X_MEM_COMPRESS,\n"
+ "+\t\t\t\tGFP_KERNEL | __GFP_REPEAT);\n"
+ "+\t\tbreak;\n"
+ "+\tcase CPU_DEAD:\n"
+ "+\tcase CPU_UP_CANCELED:\n"
+ "+\t\tfree_pages((unsigned long)per_cpu(zcache_dstmem, cpu),\n"
+ "+\t\t\t\tLZO_DSTMEM_PAGE_ORDER);\n"
+ "+\t\tper_cpu(zcache_dstmem, cpu) = NULL;\n"
+ "+\t\tkfree(per_cpu(zcache_workmem, cpu));\n"
+ "+\t\tper_cpu(zcache_workmem, cpu) = NULL;\n"
+ "+\t\tkp = &per_cpu(zcache_preloads, cpu);\n"
+ "+\t\twhile (kp->nr) {\n"
+ "+\t\t\tkmem_cache_free(zcache_objnode_cache,\n"
+ "+\t\t\t\t\tkp->objnodes[kp->nr - 1]);\n"
+ "+\t\t\tkp->objnodes[kp->nr - 1] = NULL;\n"
+ "+\t\t\tkp->nr--;\n"
+ "+\t\t}\n"
+ "+\t\tkmem_cache_free(zcache_obj_cache, kp->obj);\n"
+ "+\t\tfree_page((unsigned long)kp->page);\n"
+ "+\t\tbreak;\n"
+ "+\tdefault:\n"
+ "+\t\tbreak;\n"
+ "+\t}\n"
+ "+\treturn NOTIFY_OK;\n"
+ "+}\n"
+ "+\n"
+ "+static struct notifier_block zcache_cpu_notifier_block = {\n"
+ "+\t.notifier_call = zcache_cpu_notifier\n"
+ "+};\n"
+ "+\n"
+ "+#ifdef CONFIG_SYSFS\n"
+ "+#define ZCACHE_SYSFS_RO(_name) \\\n"
+ "+\tstatic ssize_t zcache_##_name##_show(struct kobject *kobj, \\\n"
+ "+\t\t\t\tstruct kobj_attribute *attr, char *buf) \\\n"
+ "+\t{ \\\n"
+ "+\t\treturn sprintf(buf, \"%lu\\n\", zcache_##_name); \\\n"
+ "+\t} \\\n"
+ "+\tstatic struct kobj_attribute zcache_##_name##_attr = { \\\n"
+ "+\t\t.attr = { .name = __stringify(_name), .mode = 0444 }, \\\n"
+ "+\t\t.show = zcache_##_name##_show, \\\n"
+ "+\t}\n"
+ "+\n"
+ "+#define ZCACHE_SYSFS_RO_ATOMIC(_name) \\\n"
+ "+\tstatic ssize_t zcache_##_name##_show(struct kobject *kobj, \\\n"
+ "+\t\t\t\tstruct kobj_attribute *attr, char *buf) \\\n"
+ "+\t{ \\\n"
+ "+\t    return sprintf(buf, \"%d\\n\", atomic_read(&zcache_##_name)); \\\n"
+ "+\t} \\\n"
+ "+\tstatic struct kobj_attribute zcache_##_name##_attr = { \\\n"
+ "+\t\t.attr = { .name = __stringify(_name), .mode = 0444 }, \\\n"
+ "+\t\t.show = zcache_##_name##_show, \\\n"
+ "+\t}\n"
+ "+\n"
+ "+#define ZCACHE_SYSFS_RO_CUSTOM(_name, _func) \\\n"
+ "+\tstatic ssize_t zcache_##_name##_show(struct kobject *kobj, \\\n"
+ "+\t\t\t\tstruct kobj_attribute *attr, char *buf) \\\n"
+ "+\t{ \\\n"
+ "+\t    return _func(buf); \\\n"
+ "+\t} \\\n"
+ "+\tstatic struct kobj_attribute zcache_##_name##_attr = { \\\n"
+ "+\t\t.attr = { .name = __stringify(_name), .mode = 0444 }, \\\n"
+ "+\t\t.show = zcache_##_name##_show, \\\n"
+ "+\t}\n"
+ "+\n"
+ "+ZCACHE_SYSFS_RO(curr_obj_count_max);\n"
+ "+ZCACHE_SYSFS_RO(curr_objnode_count_max);\n"
+ "+ZCACHE_SYSFS_RO(flush_total);\n"
+ "+ZCACHE_SYSFS_RO(flush_found);\n"
+ "+ZCACHE_SYSFS_RO(flobj_total);\n"
+ "+ZCACHE_SYSFS_RO(flobj_found);\n"
+ "+ZCACHE_SYSFS_RO(failed_eph_puts);\n"
+ "+ZCACHE_SYSFS_RO(failed_pers_puts);\n"
+ "+ZCACHE_SYSFS_RO(zbud_curr_zbytes);\n"
+ "+ZCACHE_SYSFS_RO(zbud_cumul_zpages);\n"
+ "+ZCACHE_SYSFS_RO(zbud_cumul_zbytes);\n"
+ "+ZCACHE_SYSFS_RO(zbud_buddied_count);\n"
+ "+ZCACHE_SYSFS_RO(zbpg_unused_list_count);\n"
+ "+ZCACHE_SYSFS_RO(evicted_raw_pages);\n"
+ "+ZCACHE_SYSFS_RO(evicted_unbuddied_pages);\n"
+ "+ZCACHE_SYSFS_RO(evicted_buddied_pages);\n"
+ "+ZCACHE_SYSFS_RO(failed_get_free_pages);\n"
+ "+ZCACHE_SYSFS_RO(failed_alloc);\n"
+ "+ZCACHE_SYSFS_RO(put_to_flush);\n"
+ "+ZCACHE_SYSFS_RO(aborted_preload);\n"
+ "+ZCACHE_SYSFS_RO(aborted_shrink);\n"
+ "+ZCACHE_SYSFS_RO(compress_poor);\n"
+ "+ZCACHE_SYSFS_RO(mean_compress_poor);\n"
+ "+ZCACHE_SYSFS_RO_ATOMIC(zbud_curr_raw_pages);\n"
+ "+ZCACHE_SYSFS_RO_ATOMIC(zbud_curr_zpages);\n"
+ "+ZCACHE_SYSFS_RO_ATOMIC(curr_obj_count);\n"
+ "+ZCACHE_SYSFS_RO_ATOMIC(curr_objnode_count);\n"
+ "+ZCACHE_SYSFS_RO_CUSTOM(zbud_unbuddied_list_counts,\n"
+ "+\t\t\tzbud_show_unbuddied_list_counts);\n"
+ "+ZCACHE_SYSFS_RO_CUSTOM(zbud_cumul_chunk_counts,\n"
+ "+\t\t\tzbud_show_cumul_chunk_counts);\n"
+ "+ZCACHE_SYSFS_RO_CUSTOM(zv_curr_dist_counts,\n"
+ "+\t\t\tzv_curr_dist_counts_show);\n"
+ "+ZCACHE_SYSFS_RO_CUSTOM(zv_cumul_dist_counts,\n"
+ "+\t\t\tzv_cumul_dist_counts_show);\n"
+ "+\n"
+ "+static struct attribute *zcache_attrs[] = {\n"
+ "+\t&zcache_curr_obj_count_attr.attr,\n"
+ "+\t&zcache_curr_obj_count_max_attr.attr,\n"
+ "+\t&zcache_curr_objnode_count_attr.attr,\n"
+ "+\t&zcache_curr_objnode_count_max_attr.attr,\n"
+ "+\t&zcache_flush_total_attr.attr,\n"
+ "+\t&zcache_flobj_total_attr.attr,\n"
+ "+\t&zcache_flush_found_attr.attr,\n"
+ "+\t&zcache_flobj_found_attr.attr,\n"
+ "+\t&zcache_failed_eph_puts_attr.attr,\n"
+ "+\t&zcache_failed_pers_puts_attr.attr,\n"
+ "+\t&zcache_compress_poor_attr.attr,\n"
+ "+\t&zcache_mean_compress_poor_attr.attr,\n"
+ "+\t&zcache_zbud_curr_raw_pages_attr.attr,\n"
+ "+\t&zcache_zbud_curr_zpages_attr.attr,\n"
+ "+\t&zcache_zbud_curr_zbytes_attr.attr,\n"
+ "+\t&zcache_zbud_cumul_zpages_attr.attr,\n"
+ "+\t&zcache_zbud_cumul_zbytes_attr.attr,\n"
+ "+\t&zcache_zbud_buddied_count_attr.attr,\n"
+ "+\t&zcache_zbpg_unused_list_count_attr.attr,\n"
+ "+\t&zcache_evicted_raw_pages_attr.attr,\n"
+ "+\t&zcache_evicted_unbuddied_pages_attr.attr,\n"
+ "+\t&zcache_evicted_buddied_pages_attr.attr,\n"
+ "+\t&zcache_failed_get_free_pages_attr.attr,\n"
+ "+\t&zcache_failed_alloc_attr.attr,\n"
+ "+\t&zcache_put_to_flush_attr.attr,\n"
+ "+\t&zcache_aborted_preload_attr.attr,\n"
+ "+\t&zcache_aborted_shrink_attr.attr,\n"
+ "+\t&zcache_zbud_unbuddied_list_counts_attr.attr,\n"
+ "+\t&zcache_zbud_cumul_chunk_counts_attr.attr,\n"
+ "+\t&zcache_zv_curr_dist_counts_attr.attr,\n"
+ "+\t&zcache_zv_cumul_dist_counts_attr.attr,\n"
+ "+\t&zcache_zv_max_zsize_attr.attr,\n"
+ "+\t&zcache_zv_max_mean_zsize_attr.attr,\n"
+ "+\t&zcache_zv_page_count_policy_percent_attr.attr,\n"
+ "+\tNULL,\n"
+ "+};\n"
+ "+\n"
+ "+static struct attribute_group zcache_attr_group = {\n"
+ "+\t.attrs = zcache_attrs,\n"
+ "+\t.name = \"zcache\",\n"
+ "+};\n"
+ "+\n"
+ "+#endif /* CONFIG_SYSFS */\n"
+ "+/*\n"
+ "+ * When zcache is disabled (\"frozen\"), pools can be created and destroyed,\n"
+ "+ * but all puts (and thus all other operations that require memory allocation)\n"
+ "+ * must fail.  If zcache is unfrozen, accepts puts, then frozen again,\n"
+ "+ * data consistency requires all puts while frozen to be converted into\n"
+ "+ * flushes.\n"
+ "+ */\n"
+ "+static bool zcache_freeze;\n"
+ "+\n"
+ "+/*\n"
+ "+ * zcache shrinker interface (only useful for ephemeral pages, so zbud only)\n"
+ "+ */\n"
+ "+static int shrink_zcache_memory(struct shrinker *shrink,\n"
+ "+\t\t\t\tstruct shrink_control *sc)\n"
+ "+{\n"
+ "+\tint ret = -1;\n"
+ "+\tint nr = sc->nr_to_scan;\n"
+ "+\tgfp_t gfp_mask = sc->gfp_mask;\n"
+ "+\n"
+ "+\tif (nr >= 0) {\n"
+ "+\t\tif (!(gfp_mask & __GFP_FS))\n"
+ "+\t\t\t/* does this case really need to be skipped? */\n"
+ "+\t\t\tgoto out;\n"
+ "+\t\tif (spin_trylock(&zcache_direct_reclaim_lock)) {\n"
+ "+\t\t\tzbud_evict_pages(nr);\n"
+ "+\t\t\tspin_unlock(&zcache_direct_reclaim_lock);\n"
+ "+\t\t} else\n"
+ "+\t\t\tzcache_aborted_shrink++;\n"
+ "+\t}\n"
+ "+\tret = (int)atomic_read(&zcache_zbud_curr_raw_pages);\n"
+ "+out:\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+static struct shrinker zcache_shrinker = {\n"
+ "+\t.shrink = shrink_zcache_memory,\n"
+ "+\t.seeks = DEFAULT_SEEKS,\n"
+ "+};\n"
+ "+\n"
+ "+/*\n"
+ "+ * zcache shims between cleancache/frontswap ops and tmem\n"
+ "+ */\n"
+ "+\n"
+ "+static int zcache_put_page(int cli_id, int pool_id, struct tmem_oid *oidp,\n"
+ "+\t\t\t\tuint32_t index, struct page *page)\n"
+ "+{\n"
+ "+\tstruct tmem_pool *pool;\n"
+ "+\tint ret = -1;\n"
+ "+\n"
+ "+\tBUG_ON(!irqs_disabled());\n"
+ "+\tpool = zcache_get_pool_by_id(cli_id, pool_id);\n"
+ "+\tif (unlikely(pool == NULL))\n"
+ "+\t\tgoto out;\n"
+ "+\tif (!zcache_freeze && zcache_do_preload(pool) == 0) {\n"
+ "+\t\t/* preload does preempt_disable on success */\n"
+ "+\t\tret = tmem_put(pool, oidp, index, (char *)(page),\n"
+ "+\t\t\t\tPAGE_SIZE, 0, is_ephemeral(pool));\n"
+ "+\t\tif (ret < 0) {\n"
+ "+\t\t\tif (is_ephemeral(pool))\n"
+ "+\t\t\t\tzcache_failed_eph_puts++;\n"
+ "+\t\t\telse\n"
+ "+\t\t\t\tzcache_failed_pers_puts++;\n"
+ "+\t\t}\n"
+ "+\t\tzcache_put_pool(pool);\n"
+ "+\t\tpreempt_enable_no_resched();\n"
+ "+\t} else {\n"
+ "+\t\tzcache_put_to_flush++;\n"
+ "+\t\tif (atomic_read(&pool->obj_count) > 0)\n"
+ "+\t\t\t/* the put fails whether the flush succeeds or not */\n"
+ "+\t\t\t(void)tmem_flush_page(pool, oidp, index);\n"
+ "+\t\tzcache_put_pool(pool);\n"
+ "+\t}\n"
+ "+out:\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+static int zcache_get_page(int cli_id, int pool_id, struct tmem_oid *oidp,\n"
+ "+\t\t\t\tuint32_t index, struct page *page)\n"
+ "+{\n"
+ "+\tstruct tmem_pool *pool;\n"
+ "+\tint ret = -1;\n"
+ "+\tunsigned long flags;\n"
+ "+\tsize_t size = PAGE_SIZE;\n"
+ "+\n"
+ "+\tlocal_irq_save(flags);\n"
+ "+\tpool = zcache_get_pool_by_id(cli_id, pool_id);\n"
+ "+\tif (likely(pool != NULL)) {\n"
+ "+\t\tif (atomic_read(&pool->obj_count) > 0)\n"
+ "+\t\t\tret = tmem_get(pool, oidp, index, (char *)(page),\n"
+ "+\t\t\t\t\t&size, 0, is_ephemeral(pool));\n"
+ "+\t\tzcache_put_pool(pool);\n"
+ "+\t}\n"
+ "+\tlocal_irq_restore(flags);\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+static int zcache_flush_page(int cli_id, int pool_id,\n"
+ "+\t\t\t\tstruct tmem_oid *oidp, uint32_t index)\n"
+ "+{\n"
+ "+\tstruct tmem_pool *pool;\n"
+ "+\tint ret = -1;\n"
+ "+\tunsigned long flags;\n"
+ "+\n"
+ "+\tlocal_irq_save(flags);\n"
+ "+\tzcache_flush_total++;\n"
+ "+\tpool = zcache_get_pool_by_id(cli_id, pool_id);\n"
+ "+\tif (likely(pool != NULL)) {\n"
+ "+\t\tif (atomic_read(&pool->obj_count) > 0)\n"
+ "+\t\t\tret = tmem_flush_page(pool, oidp, index);\n"
+ "+\t\tzcache_put_pool(pool);\n"
+ "+\t}\n"
+ "+\tif (ret >= 0)\n"
+ "+\t\tzcache_flush_found++;\n"
+ "+\tlocal_irq_restore(flags);\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+static int zcache_flush_object(int cli_id, int pool_id,\n"
+ "+\t\t\t\tstruct tmem_oid *oidp)\n"
+ "+{\n"
+ "+\tstruct tmem_pool *pool;\n"
+ "+\tint ret = -1;\n"
+ "+\tunsigned long flags;\n"
+ "+\n"
+ "+\tlocal_irq_save(flags);\n"
+ "+\tzcache_flobj_total++;\n"
+ "+\tpool = zcache_get_pool_by_id(cli_id, pool_id);\n"
+ "+\tif (likely(pool != NULL)) {\n"
+ "+\t\tif (atomic_read(&pool->obj_count) > 0)\n"
+ "+\t\t\tret = tmem_flush_object(pool, oidp);\n"
+ "+\t\tzcache_put_pool(pool);\n"
+ "+\t}\n"
+ "+\tif (ret >= 0)\n"
+ "+\t\tzcache_flobj_found++;\n"
+ "+\tlocal_irq_restore(flags);\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+static int zcache_destroy_pool(int cli_id, int pool_id)\n"
+ "+{\n"
+ "+\tstruct tmem_pool *pool = NULL;\n"
+ "+\tstruct zcache_client *cli = NULL;\n"
+ "+\tint ret = -1;\n"
+ "+\n"
+ "+\tif (pool_id < 0)\n"
+ "+\t\tgoto out;\n"
+ "+\tif (cli_id == LOCAL_CLIENT)\n"
+ "+\t\tcli = &zcache_host;\n"
+ "+\telse if ((unsigned int)cli_id < MAX_CLIENTS)\n"
+ "+\t\tcli = &zcache_clients[cli_id];\n"
+ "+\tif (cli == NULL)\n"
+ "+\t\tgoto out;\n"
+ "+\tatomic_inc(&cli->refcount);\n"
+ "+\tpool = cli->tmem_pools[pool_id];\n"
+ "+\tif (pool == NULL)\n"
+ "+\t\tgoto out;\n"
+ "+\tcli->tmem_pools[pool_id] = NULL;\n"
+ "+\t/* wait for pool activity on other cpus to quiesce */\n"
+ "+\twhile (atomic_read(&pool->refcount) != 0)\n"
+ "+\t\t;\n"
+ "+\tatomic_dec(&cli->refcount);\n"
+ "+\tlocal_bh_disable();\n"
+ "+\tret = tmem_destroy_pool(pool);\n"
+ "+\tlocal_bh_enable();\n"
+ "+\tkfree(pool);\n"
+ "+\tpr_info(\"zcache: destroyed pool id=%d, cli_id=%d\\n\",\n"
+ "+\t\t\tpool_id, cli_id);\n"
+ "+out:\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+static int zcache_new_pool(uint16_t cli_id, uint32_t flags)\n"
+ "+{\n"
+ "+\tint poolid = -1;\n"
+ "+\tstruct tmem_pool *pool;\n"
+ "+\tstruct zcache_client *cli = NULL;\n"
+ "+\n"
+ "+\tif (cli_id == LOCAL_CLIENT)\n"
+ "+\t\tcli = &zcache_host;\n"
+ "+\telse if ((unsigned int)cli_id < MAX_CLIENTS)\n"
+ "+\t\tcli = &zcache_clients[cli_id];\n"
+ "+\tif (cli == NULL)\n"
+ "+\t\tgoto out;\n"
+ "+\tatomic_inc(&cli->refcount);\n"
+ "+\tpool = kmalloc(sizeof(struct tmem_pool), GFP_KERNEL);\n"
+ "+\tif (pool == NULL) {\n"
+ "+\t\tpr_info(\"zcache: pool creation failed: out of memory\\n\");\n"
+ "+\t\tgoto out;\n"
+ "+\t}\n"
+ "+\n"
+ "+\tfor (poolid = 0; poolid < MAX_POOLS_PER_CLIENT; poolid++)\n"
+ "+\t\tif (cli->tmem_pools[poolid] == NULL)\n"
+ "+\t\t\tbreak;\n"
+ "+\tif (poolid >= MAX_POOLS_PER_CLIENT) {\n"
+ "+\t\tpr_info(\"zcache: pool creation failed: max exceeded\\n\");\n"
+ "+\t\tkfree(pool);\n"
+ "+\t\tpoolid = -1;\n"
+ "+\t\tgoto out;\n"
+ "+\t}\n"
+ "+\tatomic_set(&pool->refcount, 0);\n"
+ "+\tpool->client = cli;\n"
+ "+\tpool->pool_id = poolid;\n"
+ "+\ttmem_new_pool(pool, flags);\n"
+ "+\tcli->tmem_pools[poolid] = pool;\n"
+ "+\tpr_info(\"zcache: created %s tmem pool, id=%d, client=%d\\n\",\n"
+ "+\t\tflags & TMEM_POOL_PERSIST ? \"persistent\" : \"ephemeral\",\n"
+ "+\t\tpoolid, cli_id);\n"
+ "+out:\n"
+ "+\tif (cli != NULL)\n"
+ "+\t\tatomic_dec(&cli->refcount);\n"
+ "+\treturn poolid;\n"
+ "+}\n"
+ "+\n"
+ "+/**********\n"
+ "+ * Two kernel functionalities currently can be layered on top of tmem.\n"
+ "+ * These are \"cleancache\" which is used as a second-chance cache for clean\n"
+ "+ * page cache pages; and \"frontswap\" which is used for swap pages\n"
+ "+ * to avoid writes to disk.  A generic \"shim\" is provided here for each\n"
+ "+ * to translate in-kernel semantics to zcache semantics.\n"
+ "+ */\n"
+ "+\n"
+ "+#ifdef CONFIG_CLEANCACHE\n"
+ "+static void zcache_cleancache_put_page(int pool_id,\n"
+ "+\t\t\t\t\tstruct cleancache_filekey key,\n"
+ "+\t\t\t\t\tpgoff_t index, struct page *page)\n"
+ "+{\n"
+ "+\tu32 ind = (u32) index;\n"
+ "+\tstruct tmem_oid oid = *(struct tmem_oid *)&key;\n"
+ "+\n"
+ "+\tif (likely(ind == index))\n"
+ "+\t\t(void)zcache_put_page(LOCAL_CLIENT, pool_id, &oid, index, page);\n"
+ "+}\n"
+ "+\n"
+ "+static int zcache_cleancache_get_page(int pool_id,\n"
+ "+\t\t\t\t\tstruct cleancache_filekey key,\n"
+ "+\t\t\t\t\tpgoff_t index, struct page *page)\n"
+ "+{\n"
+ "+\tu32 ind = (u32) index;\n"
+ "+\tstruct tmem_oid oid = *(struct tmem_oid *)&key;\n"
+ "+\tint ret = -1;\n"
+ "+\n"
+ "+\tif (likely(ind == index))\n"
+ "+\t\tret = zcache_get_page(LOCAL_CLIENT, pool_id, &oid, index, page);\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+static void zcache_cleancache_flush_page(int pool_id,\n"
+ "+\t\t\t\t\tstruct cleancache_filekey key,\n"
+ "+\t\t\t\t\tpgoff_t index)\n"
+ "+{\n"
+ "+\tu32 ind = (u32) index;\n"
+ "+\tstruct tmem_oid oid = *(struct tmem_oid *)&key;\n"
+ "+\n"
+ "+\tif (likely(ind == index))\n"
+ "+\t\t(void)zcache_flush_page(LOCAL_CLIENT, pool_id, &oid, ind);\n"
+ "+}\n"
+ "+\n"
+ "+static void zcache_cleancache_flush_inode(int pool_id,\n"
+ "+\t\t\t\t\tstruct cleancache_filekey key)\n"
+ "+{\n"
+ "+\tstruct tmem_oid oid = *(struct tmem_oid *)&key;\n"
+ "+\n"
+ "+\t(void)zcache_flush_object(LOCAL_CLIENT, pool_id, &oid);\n"
+ "+}\n"
+ "+\n"
+ "+static void zcache_cleancache_flush_fs(int pool_id)\n"
+ "+{\n"
+ "+\tif (pool_id >= 0)\n"
+ "+\t\t(void)zcache_destroy_pool(LOCAL_CLIENT, pool_id);\n"
+ "+}\n"
+ "+\n"
+ "+static int zcache_cleancache_init_fs(size_t pagesize)\n"
+ "+{\n"
+ "+\tBUG_ON(sizeof(struct cleancache_filekey) !=\n"
+ "+\t\t\t\tsizeof(struct tmem_oid));\n"
+ "+\tBUG_ON(pagesize != PAGE_SIZE);\n"
+ "+\treturn zcache_new_pool(LOCAL_CLIENT, 0);\n"
+ "+}\n"
+ "+\n"
+ "+static int zcache_cleancache_init_shared_fs(char *uuid, size_t pagesize)\n"
+ "+{\n"
+ "+\t/* shared pools are unsupported and map to private */\n"
+ "+\tBUG_ON(sizeof(struct cleancache_filekey) !=\n"
+ "+\t\t\t\tsizeof(struct tmem_oid));\n"
+ "+\tBUG_ON(pagesize != PAGE_SIZE);\n"
+ "+\treturn zcache_new_pool(LOCAL_CLIENT, 0);\n"
+ "+}\n"
+ "+\n"
+ "+static struct cleancache_ops zcache_cleancache_ops = {\n"
+ "+\t.put_page = zcache_cleancache_put_page,\n"
+ "+\t.get_page = zcache_cleancache_get_page,\n"
+ "+\t.invalidate_page = zcache_cleancache_flush_page,\n"
+ "+\t.invalidate_inode = zcache_cleancache_flush_inode,\n"
+ "+\t.invalidate_fs = zcache_cleancache_flush_fs,\n"
+ "+\t.init_shared_fs = zcache_cleancache_init_shared_fs,\n"
+ "+\t.init_fs = zcache_cleancache_init_fs\n"
+ "+};\n"
+ "+\n"
+ "+struct cleancache_ops zcache_cleancache_register_ops(void)\n"
+ "+{\n"
+ "+\tstruct cleancache_ops old_ops =\n"
+ "+\t\tcleancache_register_ops(&zcache_cleancache_ops);\n"
+ "+\n"
+ "+\treturn old_ops;\n"
+ "+}\n"
+ "+#endif\n"
+ "+\n"
+ "+#ifdef CONFIG_FRONTSWAP\n"
+ "+/* a single tmem poolid is used for all frontswap \"types\" (swapfiles) */\n"
+ "+static int zcache_frontswap_poolid = -1;\n"
+ "+\n"
+ "+/*\n"
+ "+ * Swizzling increases objects per swaptype, increasing tmem concurrency\n"
+ "+ * for heavy swaploads.  Later, larger nr_cpus -> larger SWIZ_BITS\n"
+ "+ */\n"
+ "+#define SWIZ_BITS\t\t4\n"
+ "+#define SWIZ_MASK\t\t((1 << SWIZ_BITS) - 1)\n"
+ "+#define _oswiz(_type, _ind)\t((_type << SWIZ_BITS) | (_ind & SWIZ_MASK))\n"
+ "+#define iswiz(_ind)\t\t(_ind >> SWIZ_BITS)\n"
+ "+\n"
+ "+static inline struct tmem_oid oswiz(unsigned type, u32 ind)\n"
+ "+{\n"
+ "+\tstruct tmem_oid oid = { .oid = { 0 } };\n"
+ "+\toid.oid[0] = _oswiz(type, ind);\n"
+ "+\treturn oid;\n"
+ "+}\n"
+ "+\n"
+ "+static int zcache_frontswap_put_page(unsigned type, pgoff_t offset,\n"
+ "+\t\t\t\t   struct page *page)\n"
+ "+{\n"
+ "+\tu64 ind64 = (u64)offset;\n"
+ "+\tu32 ind = (u32)offset;\n"
+ "+\tstruct tmem_oid oid = oswiz(type, ind);\n"
+ "+\tint ret = -1;\n"
+ "+\tunsigned long flags;\n"
+ "+\n"
+ "+\tBUG_ON(!PageLocked(page));\n"
+ "+\tif (likely(ind64 == ind)) {\n"
+ "+\t\tlocal_irq_save(flags);\n"
+ "+\t\tret = zcache_put_page(LOCAL_CLIENT, zcache_frontswap_poolid,\n"
+ "+\t\t\t\t\t&oid, iswiz(ind), page);\n"
+ "+\t\tlocal_irq_restore(flags);\n"
+ "+\t}\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+/* returns 0 if the page was successfully gotten from frontswap, -1 if\n"
+ "+ * was not present (should never happen!) */\n"
+ "+static int zcache_frontswap_get_page(unsigned type, pgoff_t offset,\n"
+ "+\t\t\t\t   struct page *page)\n"
+ "+{\n"
+ "+\tu64 ind64 = (u64)offset;\n"
+ "+\tu32 ind = (u32)offset;\n"
+ "+\tstruct tmem_oid oid = oswiz(type, ind);\n"
+ "+\tint ret = -1;\n"
+ "+\n"
+ "+\tBUG_ON(!PageLocked(page));\n"
+ "+\tif (likely(ind64 == ind))\n"
+ "+\t\tret = zcache_get_page(LOCAL_CLIENT, zcache_frontswap_poolid,\n"
+ "+\t\t\t\t\t&oid, iswiz(ind), page);\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+/* flush a single page from frontswap */\n"
+ "+static void zcache_frontswap_flush_page(unsigned type, pgoff_t offset)\n"
+ "+{\n"
+ "+\tu64 ind64 = (u64)offset;\n"
+ "+\tu32 ind = (u32)offset;\n"
+ "+\tstruct tmem_oid oid = oswiz(type, ind);\n"
+ "+\n"
+ "+\tif (likely(ind64 == ind))\n"
+ "+\t\t(void)zcache_flush_page(LOCAL_CLIENT, zcache_frontswap_poolid,\n"
+ "+\t\t\t\t\t&oid, iswiz(ind));\n"
+ "+}\n"
+ "+\n"
+ "+/* flush all pages from the passed swaptype */\n"
+ "+static void zcache_frontswap_flush_area(unsigned type)\n"
+ "+{\n"
+ "+\tstruct tmem_oid oid;\n"
+ "+\tint ind;\n"
+ "+\n"
+ "+\tfor (ind = SWIZ_MASK; ind >= 0; ind--) {\n"
+ "+\t\toid = oswiz(type, ind);\n"
+ "+\t\t(void)zcache_flush_object(LOCAL_CLIENT,\n"
+ "+\t\t\t\t\t\tzcache_frontswap_poolid, &oid);\n"
+ "+\t}\n"
+ "+}\n"
+ "+\n"
+ "+static void zcache_frontswap_init(unsigned ignored)\n"
+ "+{\n"
+ "+\t/* a single tmem poolid is used for all frontswap \"types\" (swapfiles) */\n"
+ "+\tif (zcache_frontswap_poolid < 0)\n"
+ "+\t\tzcache_frontswap_poolid =\n"
+ "+\t\t\tzcache_new_pool(LOCAL_CLIENT, TMEM_POOL_PERSIST);\n"
+ "+}\n"
+ "+\n"
+ "+static struct frontswap_ops zcache_frontswap_ops = {\n"
+ "+\t.put_page = zcache_frontswap_put_page,\n"
+ "+\t.get_page = zcache_frontswap_get_page,\n"
+ "+\t.invalidate_page = zcache_frontswap_flush_page,\n"
+ "+\t.invalidate_area = zcache_frontswap_flush_area,\n"
+ "+\t.init = zcache_frontswap_init\n"
+ "+};\n"
+ "+\n"
+ "+struct frontswap_ops zcache_frontswap_register_ops(void)\n"
+ "+{\n"
+ "+\tstruct frontswap_ops old_ops =\n"
+ "+\t\tfrontswap_register_ops(&zcache_frontswap_ops);\n"
+ "+\n"
+ "+\treturn old_ops;\n"
+ "+}\n"
+ "+#endif\n"
+ "+\n"
+ "+/*\n"
+ "+ * zcache initialization\n"
+ "+ * NOTE FOR NOW zcache MUST BE PROVIDED AS A KERNEL BOOT PARAMETER OR\n"
+ "+ * NOTHING HAPPENS!\n"
+ "+ */\n"
+ "+\n"
+ "+static int zcache_enabled;\n"
+ "+\n"
+ "+static int __init enable_zcache(char *s)\n"
+ "+{\n"
+ "+\tzcache_enabled = 1;\n"
+ "+\treturn 1;\n"
+ "+}\n"
+ "+__setup(\"zcache\", enable_zcache);\n"
+ "+\n"
+ "+/* allow independent dynamic disabling of cleancache and frontswap */\n"
+ "+\n"
+ "+static int use_cleancache = 1;\n"
+ "+\n"
+ "+static int __init no_cleancache(char *s)\n"
+ "+{\n"
+ "+\tuse_cleancache = 0;\n"
+ "+\treturn 1;\n"
+ "+}\n"
+ "+\n"
+ "+__setup(\"nocleancache\", no_cleancache);\n"
+ "+\n"
+ "+static int use_frontswap = 1;\n"
+ "+\n"
+ "+static int __init no_frontswap(char *s)\n"
+ "+{\n"
+ "+\tuse_frontswap = 0;\n"
+ "+\treturn 1;\n"
+ "+}\n"
+ "+\n"
+ "+__setup(\"nofrontswap\", no_frontswap);\n"
+ "+\n"
+ "+static int __init zcache_init(void)\n"
+ "+{\n"
+ "+\tint ret = 0;\n"
+ "+\n"
+ "+#ifdef CONFIG_SYSFS\n"
+ "+\tret = sysfs_create_group(mm_kobj, &zcache_attr_group);\n"
+ "+\tif (ret) {\n"
+ "+\t\tpr_err(\"zcache: can't create sysfs\\n\");\n"
+ "+\t\tgoto out;\n"
+ "+\t}\n"
+ "+#endif /* CONFIG_SYSFS */\n"
+ "+#if defined(CONFIG_CLEANCACHE) || defined(CONFIG_FRONTSWAP)\n"
+ "+\tif (zcache_enabled) {\n"
+ "+\t\tunsigned int cpu;\n"
+ "+\n"
+ "+\t\ttmem_register_hostops(&zcache_hostops);\n"
+ "+\t\ttmem_register_pamops(&zcache_pamops);\n"
+ "+\t\tret = register_cpu_notifier(&zcache_cpu_notifier_block);\n"
+ "+\t\tif (ret) {\n"
+ "+\t\t\tpr_err(\"zcache: can't register cpu notifier\\n\");\n"
+ "+\t\t\tgoto out;\n"
+ "+\t\t}\n"
+ "+\t\tfor_each_online_cpu(cpu) {\n"
+ "+\t\t\tvoid *pcpu = (void *)(long)cpu;\n"
+ "+\t\t\tzcache_cpu_notifier(&zcache_cpu_notifier_block,\n"
+ "+\t\t\t\tCPU_UP_PREPARE, pcpu);\n"
+ "+\t\t}\n"
+ "+\t}\n"
+ "+\tzcache_objnode_cache = kmem_cache_create(\"zcache_objnode\",\n"
+ "+\t\t\t\tsizeof(struct tmem_objnode), 0, 0, NULL);\n"
+ "+\tzcache_obj_cache = kmem_cache_create(\"zcache_obj\",\n"
+ "+\t\t\t\tsizeof(struct tmem_obj), 0, 0, NULL);\n"
+ "+\tret = zcache_new_client(LOCAL_CLIENT);\n"
+ "+\tif (ret) {\n"
+ "+\t\tpr_err(\"zcache: can't create client\\n\");\n"
+ "+\t\tgoto out;\n"
+ "+\t}\n"
+ "+#endif\n"
+ "+#ifdef CONFIG_CLEANCACHE\n"
+ "+\tif (zcache_enabled && use_cleancache) {\n"
+ "+\t\tstruct cleancache_ops old_ops;\n"
+ "+\n"
+ "+\t\tzbud_init();\n"
+ "+\t\tregister_shrinker(&zcache_shrinker);\n"
+ "+\t\told_ops = zcache_cleancache_register_ops();\n"
+ "+\t\tpr_info(\"zcache: cleancache enabled using kernel \"\n"
+ "+\t\t\t\"transcendent memory and compression buddies\\n\");\n"
+ "+\t\tif (old_ops.init_fs != NULL)\n"
+ "+\t\t\tpr_warning(\"zcache: cleancache_ops overridden\");\n"
+ "+\t}\n"
+ "+#endif\n"
+ "+#ifdef CONFIG_FRONTSWAP\n"
+ "+\tif (zcache_enabled && use_frontswap) {\n"
+ "+\t\tstruct frontswap_ops old_ops;\n"
+ "+\n"
+ "+\t\told_ops = zcache_frontswap_register_ops();\n"
+ "+\t\tpr_info(\"zcache: frontswap enabled using kernel \"\n"
+ "+\t\t\t\"transcendent memory and xvmalloc\\n\");\n"
+ "+\t\tif (old_ops.init != NULL)\n"
+ "+\t\t\tpr_warning(\"ktmem: frontswap_ops overridden\");\n"
+ "+\t}\n"
+ "+#endif\n"
+ "+out:\n"
+ "+\treturn ret;\n"
+ "+}\n"
+ "+\n"
+ "+module_init(zcache_init)\n"
+ "-- \n"
+ 1.7.1
 
-471fbfc23458b1f7c14742b33a50eb755c0ebae6777c0c1ebcdca7e78379a23a
+b72a4d89ff7c1e300e1aae43cde9c5a4585863d660cb414075641f292ca203e1

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.