-/* -*- mode: c; c-basic-offset: 8; indent-tabs-mode: nil; -*-
- * vim:expandtab:shiftwidth=8:tabstop=8:
- *
+/*
* GPL HEADER START
*
* DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
* GPL HEADER END
*/
/*
- * Copyright 2008 Sun Microsystems, Inc. All rights reserved
+ * Copyright (c) 2008, 2010, Oracle and/or its affiliates. All rights reserved.
* Use is subject to license terms.
+ *
+ * Copyright (c) 2012, 2015, Intel Corporation.
*/
/*
* This file is part of Lustre, http://www.lustre.org/
* Internal interfaces of LOV layer.
*
* Author: Nikita Danilov <nikita.danilov@sun.com>
+ * Author: Jinshan Xiong <jinshan.xiong@intel.com>
*/
#ifndef LOV_CL_INTERNAL_H
#define LOV_CL_INTERNAL_H
-#ifdef __KERNEL__
-# include <libcfs/libcfs.h>
-#else
-# include <liblustre.h>
-#endif
-
+#include <libcfs/libcfs.h>
#include <obd.h>
#include <cl_object.h>
#include "lov_internal.h"
-/** \addtogroup lov lov @{ */
-
/** \defgroup lov lov
* Logical object volume layer. This layer implements data striping (raid0).
*
* - top-page keeps a reference to its sub-page, and destroys it when it
* is destroyed.
*
- * - sub-lock keep a reference to its top-locks. Top-lock keeps a
- * reference (and a hold, see cl_lock_hold()) on its sub-locks when it
- * actively using them (that is, in cl_lock_state::CLS_QUEUING,
- * cl_lock_state::CLS_ENQUEUED, cl_lock_state::CLS_HELD states). When
- * moving into cl_lock_state::CLS_CACHED state, top-lock releases a
- * hold. From this moment top-lock has only a 'weak' reference to its
- * sub-locks. This reference is protected by top-lock
- * cl_lock::cll_guard, and will be automatically cleared by the sub-lock
- * when the latter is destroyed. When a sub-lock is canceled, a
- * reference to it is removed from the top-lock array, and top-lock is
- * moved into CLS_NEW state. It is guaranteed that all sub-locks exits
- * while their top-lock is in CLS_HELD or CLS_CACHED states.
- *
* - IO's are not reference counted.
*
* To implement a connection between top and sub entities, lov layer is split
* into two pieces: lov ("upper half"), and lovsub ("bottom half"), both
- * implementing full set of cl-interfaces. For example, top-object has clu and
+ * implementing full set of cl-interfaces. For example, top-object has vvp and
* lov layers, and it's sub-object has lovsub and osc layers. lovsub layer is
* used to track child-parent relationship.
*
*
* \see cl_env_get()
*/
- int emrg_refcheck;
+ __u16 emrg_refcheck;
};
struct lov_device {
* Serializes access to lov_device::ld_emrg in low-memory
* conditions.
*/
- struct mutex ld_mutex;
+ struct mutex ld_mutex;
};
/**
* Layout type.
*/
enum lov_layout_type {
- /** empty file without body */
- LLT_EMPTY,
- /** striped file */
- LLT_RAID0,
- /** join file */
- LLT_JOIN,
- LLT_NR
+ LLT_EMPTY, /** empty file without body (mknod + truncate) */
+ LLT_RAID0, /** striped file */
+ LLT_RELEASED, /** file with no objects (data in HSM) */
+ LLT_NR
};
+static inline char *llt2str(enum lov_layout_type llt)
+{
+ switch (llt) {
+ case LLT_EMPTY:
+ return "EMPTY";
+ case LLT_RAID0:
+ return "RAID0";
+ case LLT_RELEASED:
+ return "RELEASED";
+ case LLT_NR:
+ LBUG();
+ }
+ LBUG();
+ return "";
+}
+
/**
* lov-specific file state.
*
* function corresponding to the current layout type.
*/
struct lov_object {
- struct cl_object lo_cl;
- /**
- * Serializes object operations with transitions between layout types.
- *
- * This semaphore is taken in shared mode by all object methods, and
- * is taken in exclusive mode when object type is changed.
- *
- * \see lov_object::lo_type
- */
- struct rw_semaphore lo_type_guard;
- /**
- * Type of an object. Protected by lov_object::lo_type_guard.
- */
- enum lov_layout_type lo_type;
-
- union lov_layout_state {
- struct lov_layout_raid0 {
- unsigned lo_nr;
- struct lov_stripe_md *lo_lsm;
- /**
- * Array of sub-objects. Allocated when top-object is
- * created (lov_init_raid0()).
- *
- * Top-object is a strict master of its sub-objects:
- * it is created before them, and outlives its
- * children (this later is necessary so that basic
- * functions like cl_object_top() always
- * work). Top-object keeps a reference on every
- * sub-object.
- *
- * When top-object is destroyed (lov_delete_raid0())
- * it releases its reference to a sub-object and waits
- * until the latter is finally destroyed.
- */
- struct lovsub_object **lo_sub;
- /**
- * When this is true, lov_object::lo_attr contains
- * valid up to date attributes for a top-level
- * object. This field is reset to 0 when attributes of
- * any sub-object change.
- */
- int lo_attr_valid;
- /**
- * Cached object attribute, built from sub-object
- * attributes.
- */
- struct cl_attr lo_attr;
- } raid0;
- struct lov_layout_state_empty {
- } empty;
- struct lov_layout_state_join {
- } join;
- } u;
- /**
- * Thread that acquired lov_object::lo_type_guard in an exclusive
- * mode.
- */
- cfs_task_t *lo_owner;
-};
-
-/**
- * Flags that top-lock can set on each of its sub-locks.
- */
-enum lov_sub_flags {
- /** Top-lock acquired a hold (cl_lock_hold()) on a sub-lock. */
- LSF_HELD = 1 << 0
+ struct cl_object lo_cl;
+ /**
+ * Serializes object operations with transitions between layout types.
+ *
+ * This semaphore is taken in shared mode by all object methods, and
+ * is taken in exclusive mode when object type is changed.
+ *
+ * \see lov_object::lo_type
+ */
+ struct rw_semaphore lo_type_guard;
+ /**
+ * Type of an object. Protected by lov_object::lo_type_guard.
+ */
+ enum lov_layout_type lo_type;
+ /**
+ * True if layout is invalid. This bit is cleared when layout lock
+ * is lost.
+ */
+ bool lo_layout_invalid;
+ /**
+ * How many IOs are on going on this object. Layout can be changed
+ * only if there is no active IO.
+ */
+ atomic_t lo_active_ios;
+ /**
+ * Waitq - wait for no one else is using lo_lsm
+ */
+ wait_queue_head_t lo_waitq;
+ /**
+ * Layout metadata. NULL if empty layout.
+ */
+ struct lov_stripe_md *lo_lsm;
+
+ union lov_layout_state {
+ struct lov_layout_raid0 {
+ unsigned lo_nr;
+ /**
+ * When this is true, lov_object::lo_attr contains
+ * valid up to date attributes for a top-level
+ * object. This field is reset to 0 when attributes of
+ * any sub-object change.
+ */
+ int lo_attr_valid;
+ /**
+ * Array of sub-objects. Allocated when top-object is
+ * created (lov_init_raid0()).
+ *
+ * Top-object is a strict master of its sub-objects:
+ * it is created before them, and outlives its
+ * children (this later is necessary so that basic
+ * functions like cl_object_top() always
+ * work). Top-object keeps a reference on every
+ * sub-object.
+ *
+ * When top-object is destroyed (lov_delete_raid0())
+ * it releases its reference to a sub-object and waits
+ * until the latter is finally destroyed.
+ *
+ * May be vmalloc'd, must be freed with OBD_FREE_LARGE.
+ */
+ struct lovsub_object **lo_sub;
+ /**
+ * protect lo_sub
+ */
+ spinlock_t lo_sub_lock;
+ /**
+ * Cached object attribute, built from sub-object
+ * attributes.
+ */
+ struct cl_attr lo_attr;
+ } raid0;
+ struct lov_layout_state_empty {
+ } empty;
+ struct lov_layout_state_released {
+ } released;
+ } u;
+ /**
+ * Thread that acquired lov_object::lo_type_guard in an exclusive
+ * mode.
+ */
+ struct task_struct *lo_owner;
};
/**
* State lov_lock keeps for each sub-lock.
*/
struct lov_lock_sub {
- /** sub-lock itself */
- struct lovsub_lock *sub_lock;
- /** An array of per-sub-lock flags, taken from enum lov_sub_flags */
- unsigned sub_flags;
- int sub_stripe;
- struct cl_lock_descr sub_descr;
- struct cl_lock_descr sub_got;
+ /** sub-lock itself */
+ struct cl_lock sub_lock;
+ /** Set if the sublock has ever been enqueued, meaning it may
+ * hold resources of underlying layers */
+ unsigned int sub_is_enqueued:1,
+ sub_initialized:1;
+ int sub_stripe;
};
/**
* lov-specific lock state.
*/
struct lov_lock {
- struct cl_lock_slice lls_cl;
- /** Number of sub-locks in this lock */
- int lls_nr;
- /**
- * Number of existing sub-locks.
- */
- unsigned lls_nr_filled;
- /**
- * Set when sub-lock was canceled, while top-lock was being
- * unlocked.
- */
- int lls_unuse_race;
- /**
- * An array of sub-locks
- *
- * There are two issues with managing sub-locks:
- *
- * - sub-locks are concurrently canceled, and
- *
- * - sub-locks are shared with other top-locks.
- *
- * To manage cancellation, top-lock acquires a hold on a sublock
- * (lov_sublock_adopt()) when the latter is inserted into
- * lov_lock::lls_sub[]. This hold is released (lov_sublock_release())
- * when top-lock is going into CLS_CACHED state or destroyed. Hold
- * prevents sub-lock from cancellation.
- *
- * Sub-lock sharing means, among other things, that top-lock that is
- * in the process of creation (i.e., not yet inserted into lock list)
- * is already accessible to other threads once at least one of its
- * sub-locks is created, see lov_lock_sub_init().
- *
- * Sub-lock can be in one of the following states:
- *
- * - doesn't exist, lov_lock::lls_sub[]::sub_lock == NULL. Such
- * sub-lock was either never created (top-lock is in CLS_NEW
- * state), or it was created, then canceled, then destroyed
- * (lov_lock_unlink() cleared sub-lock pointer in the top-lock).
- *
- * - sub-lock exists and is on
- * hold. (lov_lock::lls_sub[]::sub_flags & LSF_HELD). This is a
- * normal state of a sub-lock in CLS_HELD and CLS_CACHED states
- * of a top-lock.
- *
- * - sub-lock exists, but is not held by the top-lock. This
- * happens after top-lock released a hold on sub-locks before
- * going into cache (lov_lock_unuse()).
- *
- * \todo To support wide-striping, array has to be replaced with a set
- * of queues to avoid scanning.
- */
- struct lov_lock_sub *lls_sub;
- /**
- * Original description with which lock was enqueued.
- */
- struct cl_lock_descr lls_orig;
+ struct cl_lock_slice lls_cl;
+ /** Number of sub-locks in this lock */
+ int lls_nr;
+ /** sublock array */
+ struct lov_lock_sub lls_sub[0];
};
struct lov_page {
- struct cl_page_slice lps_cl;
- int lps_invalid;
+ struct cl_page_slice lps_cl;
+ unsigned int lps_stripe; /* stripe index */
};
/*
* A linkage into per sub-lock list of all corresponding top-locks,
* hanging off lovsub_lock::lss_parents.
*/
- struct list_head lll_list;
+ struct list_head lll_list;
};
/**
* List of top-locks that have given sub-lock as their part. Protected
* by cl_lock::cll_guard mutex.
*/
- struct list_head lss_parents;
+ struct list_head lss_parents;
/**
* Top-lock that initiated current operation on this sub-lock. This is
* only set during top-to-bottom lock operations like enqueue, and is
struct lov_thread_info {
- struct cl_object_conf lti_stripe_conf;
- struct lu_fid lti_fid;
- struct cl_lock_descr lti_ldescr;
- struct ost_lvb lti_lvb;
- struct cl_2queue lti_cl2q;
- union lov_layout_state lti_state;
- struct cl_lock_closure lti_closure;
- cfs_waitlink_t lti_waiter;
+ struct cl_object_conf lti_stripe_conf;
+ struct lu_fid lti_fid;
+ struct cl_lock_descr lti_ldescr;
+ struct ost_lvb lti_lvb;
+ struct cl_2queue lti_cl2q;
+ struct cl_page_list lti_plist;
+ wait_queue_t lti_waiter;
+ struct cl_attr lti_attr;
};
/**
* State that lov_io maintains for every sub-io.
*/
struct lov_io_sub {
- int sub_stripe;
- /**
- * sub-io for a stripe. Ideally sub-io's can be stopped and resumed
- * independently, with lov acting as a scheduler to maximize overall
- * throughput.
- */
- struct cl_io *sub_io;
- /**
- * Linkage into a list (hanging off lov_io::lis_active) of all
- * sub-io's active for the current IO iteration.
- */
- struct list_head sub_linkage;
- /**
- * true, iff cl_io_init() was successfully executed against
- * lov_io_sub::sub_io.
- */
- int sub_io_initialized;
- /**
- * True, iff lov_io_sub::sub_io and lov_io_sub::sub_env weren't
- * allocated, but borrowed from a per-device emergency pool.
- */
- int sub_borrowed;
- /**
- * environment, in which sub-io executes.
- */
- struct lu_env *sub_env;
- /**
- * environment's refcheck.
- *
- * \see cl_env_get()
- */
- int sub_refcheck;
- int sub_refcheck2;
- int sub_reenter;
- void *sub_cookie;
+ __u16 sub_stripe;
+ /**
+ * environment's refcheck.
+ *
+ * \see cl_env_get()
+ */
+ __u16 sub_refcheck;
+ __u16 sub_reenter;
+ /**
+ * true, iff cl_io_init() was successfully executed against
+ * lov_io_sub::sub_io.
+ */
+ __u16 sub_io_initialized:1,
+ /**
+ * True, iff lov_io_sub::sub_io and lov_io_sub::sub_env weren't
+ * allocated, but borrowed from a per-device emergency pool.
+ */
+ sub_borrowed:1;
+ /**
+ * Linkage into a list (hanging off lov_io::lis_active) of all
+ * sub-io's active for the current IO iteration.
+ */
+ struct list_head sub_linkage;
+ /**
+ * sub-io for a stripe. Ideally sub-io's can be stopped and resumed
+ * independently, with lov acting as a scheduler to maximize overall
+ * throughput.
+ */
+ struct cl_io *sub_io;
+ /**
+ * environment, in which sub-io executes.
+ */
+ struct lu_env *sub_env;
};
/**
* starting position within a file, for the current io loop iteration
* (stripe), used by ci_io_loop().
*/
- obd_off lis_pos;
- /**
- * end position with in a file, for the current stripe io. This is
- * exclusive (i.e., next offset after last byte affected by io).
- */
- obd_off lis_endpos;
-
- int lis_mem_frozen;
- int lis_stripe_count;
- int lis_active_subios;
-
- /**
- * the index of ls_single_subio in ls_subios array
- */
- int lis_single_subio_index;
- struct cl_io lis_single_subio;
-
- /**
- * size of ls_subios array, actually the highest stripe #
- */
- int lis_nr_subios;
- struct lov_io_sub *lis_subs;
- /**
- * List of active sub-io's.
- */
- struct list_head lis_active;
+ loff_t lis_pos;
+ /**
+ * end position with in a file, for the current stripe io. This is
+ * exclusive (i.e., next offset after last byte affected by io).
+ */
+ loff_t lis_endpos;
+
+ int lis_mem_frozen;
+ int lis_stripe_count;
+ int lis_active_subios;
+
+ /**
+ * the index of ls_single_subio in ls_subios array
+ */
+ int lis_single_subio_index;
+ struct cl_io lis_single_subio;
+
+ /**
+ * size of ls_subios array, actually the highest stripe #
+ * May be vmalloc'd, must be freed with OBD_FREE_LARGE().
+ */
+ int lis_nr_subios;
+ struct lov_io_sub *lis_subs;
+ /**
+ * List of active sub-io's.
+ */
+ struct list_head lis_active;
};
struct lov_session {
struct lov_sublock_env ls_subenv;
};
-/**
- * State of transfer for lov.
- */
-struct lov_req {
- struct cl_req_slice lr_cl;
-};
-
-/**
- * State of transfer for lovsub.
- */
-struct lovsub_req {
- struct cl_req_slice lsrq_cl;
-};
-
extern struct lu_device_type lov_device_type;
extern struct lu_device_type lovsub_device_type;
extern struct lu_context_key lov_key;
extern struct lu_context_key lov_session_key;
-extern cfs_mem_cache_t *lov_page_kmem;
-extern cfs_mem_cache_t *lov_lock_kmem;
-extern cfs_mem_cache_t *lov_object_kmem;
-extern cfs_mem_cache_t *lov_thread_kmem;
-extern cfs_mem_cache_t *lov_session_kmem;
-extern cfs_mem_cache_t *lov_req_kmem;
+extern struct kmem_cache *lov_lock_kmem;
+extern struct kmem_cache *lov_object_kmem;
+extern struct kmem_cache *lov_thread_kmem;
+extern struct kmem_cache *lov_session_kmem;
-extern cfs_mem_cache_t *lovsub_page_kmem;
-extern cfs_mem_cache_t *lovsub_lock_kmem;
-extern cfs_mem_cache_t *lovsub_object_kmem;
-extern cfs_mem_cache_t *lovsub_req_kmem;
+extern struct kmem_cache *lovsub_lock_kmem;
+extern struct kmem_cache *lovsub_object_kmem;
-extern cfs_mem_cache_t *lov_lock_link_kmem;
+extern struct kmem_cache *lov_lock_link_kmem;
int lov_object_init (const struct lu_env *env, struct lu_object *obj,
const struct lu_object_conf *conf);
int lov_lock_init_raid0 (const struct lu_env *env, struct cl_object *obj,
struct cl_lock *lock, const struct cl_io *io);
+int lov_lock_init_empty (const struct lu_env *env, struct cl_object *obj,
+ struct cl_lock *lock, const struct cl_io *io);
int lov_io_init_raid0 (const struct lu_env *env, struct cl_object *obj,
struct cl_io *io);
int lov_io_init_empty (const struct lu_env *env, struct cl_object *obj,
struct cl_io *io);
+int lov_io_init_released(const struct lu_env *env, struct cl_object *obj,
+ struct cl_io *io);
void lov_lock_unlink (const struct lu_env *env, struct lov_lock_link *link,
struct lovsub_lock *sub);
const struct cl_lock_descr *d, int idx);
-struct cl_page *lov_page_init (const struct lu_env *env, struct cl_object *ob,
- struct cl_page *page, cfs_page_t *vmpage);
-struct cl_page *lovsub_page_init(const struct lu_env *env, struct cl_object *ob,
- struct cl_page *page, cfs_page_t *vmpage);
-
-struct cl_page *lov_page_init_empty(const struct lu_env *env,
- struct cl_object *obj,
- struct cl_page *page, cfs_page_t *vmpage);
-struct cl_page *lov_page_init_raid0(const struct lu_env *env,
- struct cl_object *obj,
- struct cl_page *page, cfs_page_t *vmpage);
+int lov_page_init (const struct lu_env *env, struct cl_object *ob,
+ struct cl_page *page, pgoff_t index);
+int lovsub_page_init (const struct lu_env *env, struct cl_object *ob,
+ struct cl_page *page, pgoff_t index);
+int lov_page_init_empty (const struct lu_env *env, struct cl_object *obj,
+ struct cl_page *page, pgoff_t index);
+int lov_page_init_raid0 (const struct lu_env *env, struct cl_object *obj,
+ struct cl_page *page, pgoff_t index);
struct lu_object *lov_object_alloc (const struct lu_env *env,
const struct lu_object_header *hdr,
struct lu_device *dev);
struct lov_io *lio,
const struct cl_page_slice *slice);
+struct lov_stripe_md *lov_lsm_addref(struct lov_object *lov);
+int lov_page_stripe(const struct cl_page *page);
#define lov_foreach_target(lov, var) \
for (var = 0; var < lov_targets_nr(lov); ++var)
return container_of0(slice, struct lov_page, lps_cl);
}
-static inline struct lov_req *cl2lov_req(const struct cl_req_slice *slice)
-{
- return container_of0(slice, struct lov_req, lr_cl);
-}
-
static inline struct lovsub_page *
cl2lovsub_page(const struct cl_page_slice *slice)
{
return container_of0(slice, struct lovsub_page, lsb_cl);
}
-static inline struct lovsub_req *cl2lovsub_req(const struct cl_req_slice *slice)
-{
- return container_of0(slice, struct lovsub_req, lsrq_cl);
-}
-
-static inline struct cl_page *lov_sub_page(const struct cl_page_slice *slice)
-{
- return slice->cpl_page->cp_child;
-}
-
static inline struct lov_io *cl2lov_io(const struct lu_env *env,
const struct cl_io_slice *ios)
{
static inline struct lov_layout_raid0 *lov_r0(struct lov_object *lov)
{
- struct lov_layout_raid0 *raid0;
-
- LASSERT(lov->lo_type == LLT_RAID0);
- raid0 = &lov->u.raid0;
- LASSERT(raid0->lo_lsm->lsm_wire.lw_magic == LOV_MAGIC);
- return raid0;
+ LASSERT(lov->lo_type == LLT_RAID0);
+ LASSERT(lov->lo_lsm->lsm_magic == LOV_MAGIC ||
+ lov->lo_lsm->lsm_magic == LOV_MAGIC_V3);
+ return &lov->u.raid0;
}
+/* lov_pack.c */
+int lov_getstripe(struct lov_object *obj, struct lov_stripe_md *lsm,
+ struct lov_user_md __user *lump);
+
/** @} lov */
#endif
-