*/
/*
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
- * Copyright (c) 2011, 2016 by Delphix. All rights reserved.
+ * Copyright (c) 2011, 2017 by Delphix. All rights reserved.
* Copyright 2011 Nexenta Systems, Inc. All rights reserved.
* Copyright (c) 2012, Joyent, Inc. All rights reserved.
* Copyright 2014 HybridCluster. All rights reserved.
* Copyright (c) 2014 Spectra Logic Corporation, All rights reserved.
+ * Copyright 2013 Saso Kiselkov. All rights reserved.
*/
/* Portions Copyright 2010 Robert Milkowski */
struct arc_buf;
struct zio_prop;
struct sa_handle;
+struct dsl_crypto_params;
typedef struct objset objset_t;
typedef struct dmu_tx dmu_tx_t;
#define DMU_OT_NEWTYPE 0x80
#define DMU_OT_METADATA 0x40
-#define DMU_OT_BYTESWAP_MASK 0x3f
+#define DMU_OT_ENCRYPTED 0x20
+#define DMU_OT_BYTESWAP_MASK 0x1f
/*
* Defines a uint8_t object type. Object types specify if the data
* in the object is metadata (boolean) and how to byteswap the data
* (dmu_object_byteswap_t).
*/
-#define DMU_OT(byteswap, metadata) \
+#define DMU_OT(byteswap, metadata, encrypted) \
(DMU_OT_NEWTYPE | \
((metadata) ? DMU_OT_METADATA : 0) | \
+ ((encrypted) ? DMU_OT_ENCRYPTED : 0) | \
((byteswap) & DMU_OT_BYTESWAP_MASK))
#define DMU_OT_IS_VALID(ot) (((ot) & DMU_OT_NEWTYPE) ? \
((ot) & DMU_OT_METADATA) : \
dmu_ot[(int)(ot)].ot_metadata)
+#define DMU_OT_IS_ENCRYPTED(ot) (((ot) & DMU_OT_NEWTYPE) ? \
+ ((ot) & DMU_OT_ENCRYPTED) : \
+ dmu_ot[(int)(ot)].ot_encrypt)
+
/*
* These object types use bp_fill != 1 for their L0 bp's. Therefore they can't
* have their data embedded (i.e. use a BP_IS_EMBEDDED() bp), because bp_fill
/*
* Names for valid types declared with DMU_OT().
*/
- DMU_OTN_UINT8_DATA = DMU_OT(DMU_BSWAP_UINT8, B_FALSE),
- DMU_OTN_UINT8_METADATA = DMU_OT(DMU_BSWAP_UINT8, B_TRUE),
- DMU_OTN_UINT16_DATA = DMU_OT(DMU_BSWAP_UINT16, B_FALSE),
- DMU_OTN_UINT16_METADATA = DMU_OT(DMU_BSWAP_UINT16, B_TRUE),
- DMU_OTN_UINT32_DATA = DMU_OT(DMU_BSWAP_UINT32, B_FALSE),
- DMU_OTN_UINT32_METADATA = DMU_OT(DMU_BSWAP_UINT32, B_TRUE),
- DMU_OTN_UINT64_DATA = DMU_OT(DMU_BSWAP_UINT64, B_FALSE),
- DMU_OTN_UINT64_METADATA = DMU_OT(DMU_BSWAP_UINT64, B_TRUE),
- DMU_OTN_ZAP_DATA = DMU_OT(DMU_BSWAP_ZAP, B_FALSE),
- DMU_OTN_ZAP_METADATA = DMU_OT(DMU_BSWAP_ZAP, B_TRUE),
+ DMU_OTN_UINT8_DATA = DMU_OT(DMU_BSWAP_UINT8, B_FALSE, B_FALSE),
+ DMU_OTN_UINT8_METADATA = DMU_OT(DMU_BSWAP_UINT8, B_TRUE, B_FALSE),
+ DMU_OTN_UINT16_DATA = DMU_OT(DMU_BSWAP_UINT16, B_FALSE, B_FALSE),
+ DMU_OTN_UINT16_METADATA = DMU_OT(DMU_BSWAP_UINT16, B_TRUE, B_FALSE),
+ DMU_OTN_UINT32_DATA = DMU_OT(DMU_BSWAP_UINT32, B_FALSE, B_FALSE),
+ DMU_OTN_UINT32_METADATA = DMU_OT(DMU_BSWAP_UINT32, B_TRUE, B_FALSE),
+ DMU_OTN_UINT64_DATA = DMU_OT(DMU_BSWAP_UINT64, B_FALSE, B_FALSE),
+ DMU_OTN_UINT64_METADATA = DMU_OT(DMU_BSWAP_UINT64, B_TRUE, B_FALSE),
+ DMU_OTN_ZAP_DATA = DMU_OT(DMU_BSWAP_ZAP, B_FALSE, B_FALSE),
+ DMU_OTN_ZAP_METADATA = DMU_OT(DMU_BSWAP_ZAP, B_TRUE, B_FALSE),
+
+ DMU_OTN_UINT8_ENC_DATA = DMU_OT(DMU_BSWAP_UINT8, B_FALSE, B_TRUE),
+ DMU_OTN_UINT8_ENC_METADATA = DMU_OT(DMU_BSWAP_UINT8, B_TRUE, B_TRUE),
+ DMU_OTN_UINT16_ENC_DATA = DMU_OT(DMU_BSWAP_UINT16, B_FALSE, B_TRUE),
+ DMU_OTN_UINT16_ENC_METADATA = DMU_OT(DMU_BSWAP_UINT16, B_TRUE, B_TRUE),
+ DMU_OTN_UINT32_ENC_DATA = DMU_OT(DMU_BSWAP_UINT32, B_FALSE, B_TRUE),
+ DMU_OTN_UINT32_ENC_METADATA = DMU_OT(DMU_BSWAP_UINT32, B_TRUE, B_TRUE),
+ DMU_OTN_UINT64_ENC_DATA = DMU_OT(DMU_BSWAP_UINT64, B_FALSE, B_TRUE),
+ DMU_OTN_UINT64_ENC_METADATA = DMU_OT(DMU_BSWAP_UINT64, B_TRUE, B_TRUE),
+ DMU_OTN_ZAP_ENC_DATA = DMU_OT(DMU_BSWAP_ZAP, B_FALSE, B_TRUE),
+ DMU_OTN_ZAP_ENC_METADATA = DMU_OT(DMU_BSWAP_ZAP, B_TRUE, B_TRUE),
} dmu_object_type_t;
-typedef enum txg_how {
- TXG_WAIT = 1,
- TXG_NOWAIT,
- TXG_WAITED,
-} txg_how_t;
+/*
+ * These flags are intended to be used to specify the "txg_how"
+ * parameter when calling the dmu_tx_assign() function. See the comment
+ * above dmu_tx_assign() for more details on the meaning of these flags.
+ */
+#define TXG_NOWAIT (0ULL)
+#define TXG_WAIT (1ULL<<0)
+#define TXG_NOTHROTTLE (1ULL<<1)
void byteswap_uint64_array(void *buf, size_t size);
void byteswap_uint32_array(void *buf, size_t size);
#define DMU_USERUSED_OBJECT (-1ULL)
#define DMU_GROUPUSED_OBJECT (-2ULL)
+#define DMU_PROJECTUSED_OBJECT (-3ULL)
+
+/*
+ * Zap prefix for object accounting in DMU_{USER,GROUP,PROJECT}USED_OBJECT.
+ */
+#define DMU_OBJACCT_PREFIX "obj-"
+#define DMU_OBJACCT_PREFIX_LEN 4
/*
* artificial blkids for bonus buffer and spill blocks
*/
#define DMU_BONUS_BLKID (-1ULL)
#define DMU_SPILL_BLKID (-2ULL)
+
/*
* Public routines to create, destroy, open, and close objsets.
*/
+typedef void dmu_objset_create_sync_func_t(objset_t *os, void *arg,
+ cred_t *cr, dmu_tx_t *tx);
+
int dmu_objset_hold(const char *name, void *tag, objset_t **osp);
int dmu_objset_own(const char *name, dmu_objset_type_t type,
- boolean_t readonly, void *tag, objset_t **osp);
+ boolean_t readonly, boolean_t key_required, void *tag, objset_t **osp);
void dmu_objset_rele(objset_t *os, void *tag);
-void dmu_objset_disown(objset_t *os, void *tag);
+void dmu_objset_disown(objset_t *os, boolean_t key_required, void *tag);
int dmu_objset_open_ds(struct dsl_dataset *ds, objset_t **osp);
void dmu_objset_evict_dbufs(objset_t *os);
int dmu_objset_create(const char *name, dmu_objset_type_t type, uint64_t flags,
- void (*func)(objset_t *os, void *arg, cred_t *cr, dmu_tx_t *tx), void *arg);
+ struct dsl_crypto_params *dcp, dmu_objset_create_sync_func_t func,
+ void *arg);
int dmu_objset_clone(const char *name, const char *origin);
int dsl_destroy_snapshots_nvl(struct nvlist *snaps, boolean_t defer,
struct nvlist *errlist);
#define DMU_POOL_FREE_BPOBJ "free_bpobj"
#define DMU_POOL_BPTREE_OBJ "bptree_obj"
#define DMU_POOL_EMPTY_BPOBJ "empty_bpobj"
+#define DMU_POOL_CHECKSUM_SALT "org.illumos:checksum_salt"
#define DMU_POOL_VDEV_ZAP_MAP "com.delphix:vdev_zap_map"
/*
int dmu_object_next(objset_t *os, uint64_t *objectp,
boolean_t hole, uint64_t txg);
+/*
+ * Set the number of levels on a dnode. nlevels must be greater than the
+ * current number of levels or an EINVAL will be returned.
+ */
+int dmu_object_set_nlevels(objset_t *os, uint64_t object, int nlevels,
+ dmu_tx_t *tx);
+
/*
* Set the data blocksize for an object.
*
int dmu_object_set_blocksize(objset_t *os, uint64_t object, uint64_t size,
int ibs, dmu_tx_t *tx);
+/*
+ * Manually set the maxblkid on a dnode. This will adjust nlevels accordingly
+ * to accommodate the change.
+ */
+int dmu_object_set_maxblkid(objset_t *os, uint64_t object, uint64_t maxblkid,
+ dmu_tx_t *tx);
+
/*
* Set the checksum property on a dnode. The new checksum algorithm will
* apply to all newly written blocks; existing blocks will not be affected.
void dmu_object_set_compress(objset_t *os, uint64_t object, uint8_t compress,
dmu_tx_t *tx);
-void
-dmu_write_embedded(objset_t *os, uint64_t object, uint64_t offset,
+int dmu_object_dirty_raw(objset_t *os, uint64_t object, dmu_tx_t *tx);
+
+void dmu_write_embedded(objset_t *os, uint64_t object, uint64_t offset,
void *data, uint8_t etype, uint8_t comp, int uncompressed_size,
int compressed_size, int byteorder, dmu_tx_t *tx);
#define WP_DMU_SYNC 0x2
#define WP_SPILL 0x4
-void dmu_write_policy(objset_t *os, struct dnode *dn, int level, int wp,
- enum zio_compress compress_override, struct zio_prop *zp);
+void dmu_write_policy(objset_t *os, dnode_t *dn, int level, int wp,
+ struct zio_prop *zp);
+
/*
* The bonus data is accessed more or less like a regular buffer.
* You must dmu_bonus_hold() to get the buffer, which will give you a
*
* Returns ENOENT, EIO, or 0.
*/
+int dmu_bonus_hold_impl(objset_t *os, uint64_t object, void *tag,
+ uint32_t flags, dmu_buf_t **dbp);
int dmu_bonus_hold(objset_t *os, uint64_t object, void *tag, dmu_buf_t **);
int dmu_bonus_max(void);
int dmu_set_bonus(dmu_buf_t *, int, dmu_tx_t *);
*/
taskq_ent_t dbu_tqent;
- /* This instance's eviction function pointer. */
- dmu_buf_evict_func_t *dbu_evict_func;
+ /*
+ * This instance's eviction function pointers.
+ *
+ * dbu_evict_func_sync is called synchronously and then
+ * dbu_evict_func_async is executed asynchronously on a taskq.
+ */
+ dmu_buf_evict_func_t *dbu_evict_func_sync;
+ dmu_buf_evict_func_t *dbu_evict_func_async;
#ifdef ZFS_DEBUG
/*
* Pointer to user's dbuf pointer. NULL for clients that do
* NOTE: This function should only be called once on a given dmu_buf_user_t.
* To allow enforcement of this, dbu must already be zeroed on entry.
*/
-#ifdef __lint
-/* Very ugly, but it beats issuing suppression directives in many Makefiles. */
-extern void
-dmu_buf_init_user(dmu_buf_user_t *dbu, dmu_buf_evict_func_t *evict_func,
- dmu_buf_t **clear_on_evict_dbufp);
-#else /* __lint */
+/*ARGSUSED*/
static inline void
-dmu_buf_init_user(dmu_buf_user_t *dbu, dmu_buf_evict_func_t *evict_func,
- dmu_buf_t **clear_on_evict_dbufp)
+dmu_buf_init_user(dmu_buf_user_t *dbu, dmu_buf_evict_func_t *evict_func_sync,
+ dmu_buf_evict_func_t *evict_func_async, dmu_buf_t **clear_on_evict_dbufp)
{
- ASSERT(dbu->dbu_evict_func == NULL);
- ASSERT(evict_func != NULL);
- dbu->dbu_evict_func = evict_func;
+ ASSERT(dbu->dbu_evict_func_sync == NULL);
+ ASSERT(dbu->dbu_evict_func_async == NULL);
+
+ /* must have at least one evict func */
+ IMPLY(evict_func_sync == NULL, evict_func_async != NULL);
+ dbu->dbu_evict_func_sync = evict_func_sync;
+ dbu->dbu_evict_func_async = evict_func_async;
taskq_init_ent(&dbu->dbu_tqent);
#ifdef ZFS_DEBUG
dbu->dbu_clear_on_evict_dbufp = clear_on_evict_dbufp;
#endif
}
-#endif /* __lint */
/*
* Attach user data to a dbuf and mark it for normal (when the dbuf's
* (ie. you've called dmu_tx_hold_object(tx, db->db_object)).
*/
void dmu_buf_will_dirty(dmu_buf_t *db, dmu_tx_t *tx);
-
-/*
- * Tells if the given dbuf is freeable.
- */
-boolean_t dmu_buf_freeable(dmu_buf_t *);
+void dmu_buf_will_change_crypt_params(dmu_buf_t *db, dmu_tx_t *tx);
/*
* You must create a transaction, then hold the objects which you will
dmu_tx_t *dmu_tx_create(objset_t *os);
void dmu_tx_hold_write(dmu_tx_t *tx, uint64_t object, uint64_t off, int len);
+void dmu_tx_hold_write_by_dnode(dmu_tx_t *tx, dnode_t *dn, uint64_t off,
+ int len);
void dmu_tx_hold_free(dmu_tx_t *tx, uint64_t object, uint64_t off,
uint64_t len);
+void dmu_tx_hold_free_by_dnode(dmu_tx_t *tx, dnode_t *dn, uint64_t off,
+ uint64_t len);
void dmu_tx_hold_zap(dmu_tx_t *tx, uint64_t object, int add, const char *name);
+void dmu_tx_hold_zap_by_dnode(dmu_tx_t *tx, dnode_t *dn, int add,
+ const char *name);
void dmu_tx_hold_bonus(dmu_tx_t *tx, uint64_t object);
+void dmu_tx_hold_bonus_by_dnode(dmu_tx_t *tx, dnode_t *dn);
void dmu_tx_hold_spill(dmu_tx_t *tx, uint64_t object);
void dmu_tx_hold_sa(dmu_tx_t *tx, struct sa_handle *hdl, boolean_t may_grow);
void dmu_tx_hold_sa_create(dmu_tx_t *tx, int total_size);
void dmu_tx_abort(dmu_tx_t *tx);
-int dmu_tx_assign(dmu_tx_t *tx, enum txg_how txg_how);
+int dmu_tx_assign(dmu_tx_t *tx, uint64_t txg_how);
void dmu_tx_wait(dmu_tx_t *tx);
void dmu_tx_commit(dmu_tx_t *tx);
void dmu_tx_mark_netfree(dmu_tx_t *tx);
* to stable storage and will also be called if the dmu_tx is aborted.
* If there is any error which prevents the transaction from being committed to
* disk, the callback will be called with a value of error != 0.
+ *
+ * When multiple callbacks are registered to the transaction, the callbacks
+ * will be called in reverse order to let Lustre, the only user of commit
+ * callback currently, take the fast path of its commit callback handling.
*/
typedef void dmu_tx_callback_func_t(void *dcb_data, int error);
void dmu_tx_callback_register(dmu_tx_t *tx, dmu_tx_callback_func_t *dcb_func,
void *dcb_data);
+void dmu_tx_do_callbacks(list_t *cb_list, int error);
/*
* Free up the data blocks for a defined range of a file. If size is
* -1, the range from offset to end-of-file is freed.
*/
int dmu_free_range(objset_t *os, uint64_t object, uint64_t offset,
- uint64_t size, dmu_tx_t *tx);
+ uint64_t size, dmu_tx_t *tx);
int dmu_free_long_range(objset_t *os, uint64_t object, uint64_t offset,
- uint64_t size);
+ uint64_t size);
+int dmu_free_long_range_raw(objset_t *os, uint64_t object, uint64_t offset,
+ uint64_t size);
int dmu_free_long_object(objset_t *os, uint64_t object);
+int dmu_free_long_object_raw(objset_t *os, uint64_t object);
/*
* Convenience functions.
*/
#define DMU_READ_PREFETCH 0 /* prefetch */
#define DMU_READ_NO_PREFETCH 1 /* don't prefetch */
+#define DMU_READ_NO_DECRYPT 2 /* don't decrypt */
int dmu_read(objset_t *os, uint64_t object, uint64_t offset, uint64_t size,
void *buf, uint32_t flags);
+int dmu_read_by_dnode(dnode_t *dn, uint64_t offset, uint64_t size, void *buf,
+ uint32_t flags);
void dmu_write(objset_t *os, uint64_t object, uint64_t offset, uint64_t size,
const void *buf, dmu_tx_t *tx);
+void dmu_write_by_dnode(dnode_t *dn, uint64_t offset, uint64_t size,
+ const void *buf, dmu_tx_t *tx);
void dmu_prealloc(objset_t *os, uint64_t object, uint64_t offset, uint64_t size,
dmu_tx_t *tx);
#ifdef _KERNEL
#include <linux/blkdev_compat.h>
int dmu_read_uio(objset_t *os, uint64_t object, struct uio *uio, uint64_t size);
int dmu_read_uio_dbuf(dmu_buf_t *zdb, struct uio *uio, uint64_t size);
+int dmu_read_uio_dnode(dnode_t *dn, struct uio *uio, uint64_t size);
int dmu_write_uio(objset_t *os, uint64_t object, struct uio *uio, uint64_t size,
dmu_tx_t *tx);
int dmu_write_uio_dbuf(dmu_buf_t *zdb, struct uio *uio, uint64_t size,
dmu_tx_t *tx);
+int dmu_write_uio_dnode(dnode_t *dn, struct uio *uio, uint64_t size,
+ dmu_tx_t *tx);
#endif
struct arc_buf *dmu_request_arcbuf(dmu_buf_t *handle, int size);
void dmu_return_arcbuf(struct arc_buf *buf);
-void dmu_assign_arcbuf(dmu_buf_t *handle, uint64_t offset, struct arc_buf *buf,
- dmu_tx_t *tx);
+void dmu_assign_arcbuf_by_dnode(dnode_t *dn, uint64_t offset,
+ struct arc_buf *buf, dmu_tx_t *tx);
+void dmu_assign_arcbuf_by_dbuf(dmu_buf_t *handle, uint64_t offset,
+ struct arc_buf *buf, dmu_tx_t *tx);
+#define dmu_assign_arcbuf dmu_assign_arcbuf_by_dbuf
+void dmu_convert_to_raw(dmu_buf_t *handle, boolean_t byteorder,
+ const uint8_t *salt, const uint8_t *iv, const uint8_t *mac, dmu_tx_t *tx);
+void dmu_copy_from_buf(objset_t *os, uint64_t object, uint64_t offset,
+ dmu_buf_t *handle, dmu_tx_t *tx);
+#ifdef HAVE_UIO_ZEROCOPY
int dmu_xuio_init(struct xuio *uio, int niov);
void dmu_xuio_fini(struct xuio *uio);
int dmu_xuio_add(struct xuio *uio, struct arc_buf *abuf, offset_t off,
int dmu_xuio_cnt(struct xuio *uio);
struct arc_buf *dmu_xuio_arcbuf(struct xuio *uio, int i);
void dmu_xuio_clear(struct xuio *uio, int i);
+#endif /* HAVE_UIO_ZEROCOPY */
void xuio_stat_wbuf_copied(void);
void xuio_stat_wbuf_nocopy(void);
typedef struct dmu_object_type_info {
dmu_object_byteswap_t ot_byteswap;
boolean_t ot_metadata;
+ boolean_t ot_encrypt;
char *ot_name;
} dmu_object_type_info_t;
uint64_t *idp, uint64_t *offp);
typedef int objset_used_cb_t(dmu_object_type_t bonustype,
- void *bonus, uint64_t *userp, uint64_t *groupp);
+ void *bonus, uint64_t *userp, uint64_t *groupp, uint64_t *projectp);
extern void dmu_objset_register_type(dmu_objset_type_t ost,
objset_used_cb_t *cb);
extern void dmu_objset_set_user(objset_t *os, void *user_ptr);
* {zfs,zvol,ztest}_get_done() args
*/
typedef struct zgd {
- struct zilog *zgd_zilog;
+ struct lwb *zgd_lwb;
struct blkptr *zgd_bp;
dmu_buf_t *zgd_db;
struct rl *zgd_rl;
projects / mirror_zfs.git / blobdiff