* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
- * or http://www.opensolaris.org/os/licensing.
+ * or https://opensource.org/licenses/CDDL-1.0.
* See the License for the specific language governing permissions
* and limitations under the License.
*
*/
/*
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
- * Copyright (c) 2011, 2017 by Delphix. All rights reserved.
+ * Copyright (c) 2011, 2020 by Delphix. All rights reserved.
+ * Copyright (c) 2017, Intel Corporation.
+ * Copyright (c) 2023, Klara Inc.
*/
#ifndef _SYS_VDEV_IMPL_H
#include <sys/uberblock_impl.h>
#include <sys/vdev_indirect_mapping.h>
#include <sys/vdev_indirect_births.h>
+#include <sys/vdev_rebuild.h>
#include <sys/vdev_removal.h>
#include <sys/zfs_ratelimit.h>
* Forward declarations that lots of things need.
*/
typedef struct vdev_queue vdev_queue_t;
-typedef struct vdev_cache vdev_cache_t;
-typedef struct vdev_cache_entry vdev_cache_entry_t;
struct abd;
-extern int zfs_vdev_queue_depth_pct;
-extern uint32_t zfs_vdev_async_write_max_active;
+extern uint_t zfs_vdev_queue_depth_pct;
+extern uint_t zfs_vdev_def_queue_depth;
+extern uint_t zfs_vdev_async_write_max_active;
/*
* Virtual device operations
*/
+typedef int vdev_init_func_t(spa_t *spa, nvlist_t *nv, void **tsd);
+typedef void vdev_kobj_post_evt_func_t(vdev_t *vd);
+typedef void vdev_fini_func_t(vdev_t *vd);
typedef int vdev_open_func_t(vdev_t *vd, uint64_t *size, uint64_t *max_size,
- uint64_t *ashift);
+ uint64_t *ashift, uint64_t *pshift);
typedef void vdev_close_func_t(vdev_t *vd);
-typedef uint64_t vdev_asize_func_t(vdev_t *vd, uint64_t psize);
+typedef uint64_t vdev_asize_func_t(vdev_t *vd, uint64_t psize, uint64_t txg);
+typedef uint64_t vdev_min_asize_func_t(vdev_t *vd);
+typedef uint64_t vdev_min_alloc_func_t(vdev_t *vd);
typedef void vdev_io_start_func_t(zio_t *zio);
typedef void vdev_io_done_func_t(zio_t *zio);
typedef void vdev_state_change_func_t(vdev_t *vd, int, int);
-typedef boolean_t vdev_need_resilver_func_t(vdev_t *vd, uint64_t, size_t);
+typedef boolean_t vdev_need_resilver_func_t(vdev_t *vd, const dva_t *dva,
+ size_t psize, uint64_t phys_birth);
typedef void vdev_hold_func_t(vdev_t *vd);
typedef void vdev_rele_func_t(vdev_t *vd);
uint64_t offset, uint64_t size, void *arg);
typedef void vdev_remap_func_t(vdev_t *vd, uint64_t offset, uint64_t size,
vdev_remap_cb_t callback, void *arg);
+/*
+ * Given a target vdev, translates the logical range "in" to the physical
+ * range "res"
+ */
+typedef void vdev_xlation_func_t(vdev_t *cvd, const range_seg64_t *logical,
+ range_seg64_t *physical, range_seg64_t *remain);
+typedef uint64_t vdev_rebuild_asize_func_t(vdev_t *vd, uint64_t start,
+ uint64_t size, uint64_t max_segment);
+typedef void vdev_metaslab_init_func_t(vdev_t *vd, uint64_t *startp,
+ uint64_t *sizep);
+typedef void vdev_config_generate_func_t(vdev_t *vd, nvlist_t *nv);
+typedef uint64_t vdev_nparity_func_t(vdev_t *vd);
+typedef uint64_t vdev_ndisks_func_t(vdev_t *vd);
typedef const struct vdev_ops {
+ vdev_init_func_t *vdev_op_init;
+ vdev_fini_func_t *vdev_op_fini;
vdev_open_func_t *vdev_op_open;
vdev_close_func_t *vdev_op_close;
vdev_asize_func_t *vdev_op_asize;
+ vdev_min_asize_func_t *vdev_op_min_asize;
+ vdev_min_alloc_func_t *vdev_op_min_alloc;
vdev_io_start_func_t *vdev_op_io_start;
vdev_io_done_func_t *vdev_op_io_done;
vdev_state_change_func_t *vdev_op_state_change;
vdev_hold_func_t *vdev_op_hold;
vdev_rele_func_t *vdev_op_rele;
vdev_remap_func_t *vdev_op_remap;
+ vdev_xlation_func_t *vdev_op_xlate;
+ vdev_rebuild_asize_func_t *vdev_op_rebuild_asize;
+ vdev_metaslab_init_func_t *vdev_op_metaslab_init;
+ vdev_config_generate_func_t *vdev_op_config_generate;
+ vdev_nparity_func_t *vdev_op_nparity;
+ vdev_ndisks_func_t *vdev_op_ndisks;
+ vdev_kobj_post_evt_func_t *vdev_op_kobj_evt_post;
char vdev_op_type[16];
boolean_t vdev_op_leaf;
} vdev_ops_t;
/*
* Virtual device properties
*/
-struct vdev_cache_entry {
- struct abd *ve_abd;
- uint64_t ve_offset;
- clock_t ve_lastused;
- avl_node_t ve_offset_node;
- avl_node_t ve_lastused_node;
- uint32_t ve_hits;
- uint16_t ve_missed_update;
- zio_t *ve_fill_io;
-};
-
-struct vdev_cache {
- avl_tree_t vc_offset_tree;
- avl_tree_t vc_lastused_tree;
- kmutex_t vc_lock;
-};
-
-typedef struct vdev_queue_class {
- uint32_t vqc_active;
-
- /*
- * Sorted by offset or timestamp, depending on if the queue is
- * LBA-ordered vs FIFO.
- */
- avl_tree_t vqc_queued_tree;
+typedef union vdev_queue_class {
+ struct {
+ ulong_t vqc_list_numnodes;
+ list_t vqc_list;
+ };
+ avl_tree_t vqc_tree;
} vdev_queue_class_t;
struct vdev_queue {
vdev_t *vq_vdev;
vdev_queue_class_t vq_class[ZIO_PRIORITY_NUM_QUEUEABLE];
- avl_tree_t vq_active_tree;
avl_tree_t vq_read_offset_tree;
avl_tree_t vq_write_offset_tree;
uint64_t vq_last_offset;
+ zio_priority_t vq_last_prio; /* Last sent I/O priority. */
+ uint32_t vq_cqueued; /* Classes with queued I/Os. */
+ uint32_t vq_cactive[ZIO_PRIORITY_NUM_QUEUEABLE];
+ uint32_t vq_active; /* Number of active I/Os. */
+ uint32_t vq_ia_active; /* Active interactive I/Os. */
+ uint32_t vq_nia_credit; /* Non-interactive I/Os credit. */
+ list_t vq_active_list; /* List of active I/Os. */
hrtime_t vq_io_complete_ts; /* time last i/o completed */
hrtime_t vq_io_delta_ts;
zio_t vq_io_search; /* used as local for stack reduction */
kmutex_t vq_lock;
};
+typedef enum vdev_alloc_bias {
+ VDEV_BIAS_NONE,
+ VDEV_BIAS_LOG, /* dedicated to ZIL data (SLOG) */
+ VDEV_BIAS_SPECIAL, /* dedicated to ddt, metadata, and small blks */
+ VDEV_BIAS_DEDUP /* dedicated to dedup metadata */
+} vdev_alloc_bias_t;
+
+
/*
* On-disk indirect vdev state.
*
uint64_t vdev_min_asize; /* min acceptable asize */
uint64_t vdev_max_asize; /* max acceptable asize */
uint64_t vdev_ashift; /* block alignment shift */
+
+ /*
+ * Logical block alignment shift
+ *
+ * The smallest sized/aligned I/O supported by the device.
+ */
+ uint64_t vdev_logical_ashift;
+ /*
+ * Physical block alignment shift
+ *
+ * The device supports logical I/Os with vdev_logical_ashift
+ * size/alignment, but optimum performance will be achieved by
+ * aligning/sizing requests to vdev_physical_ashift. Smaller
+ * requests may be inflated or incur device level read-modify-write
+ * operations.
+ *
+ * May be 0 to indicate no preference (i.e. use vdev_logical_ashift).
+ */
+ uint64_t vdev_physical_ashift;
uint64_t vdev_state; /* see VDEV_STATE_* #defines */
uint64_t vdev_prevstate; /* used when reopening a vdev */
vdev_ops_t *vdev_ops; /* vdev operations */
spa_t *vdev_spa; /* spa for this vdev */
void *vdev_tsd; /* type-specific data */
- vnode_t *vdev_name_vp; /* vnode for pathname */
- vnode_t *vdev_devid_vp; /* vnode for devid */
vdev_t *vdev_top; /* top-level vdev */
vdev_t *vdev_parent; /* parent vdev */
vdev_t **vdev_child; /* array of children */
boolean_t vdev_expanding; /* expand the vdev? */
boolean_t vdev_reopening; /* reopen in progress? */
boolean_t vdev_nonrot; /* true if solid state */
+ int vdev_load_error; /* error on last load */
int vdev_open_error; /* error on last open */
+ int vdev_validate_error; /* error on last validate */
kthread_t *vdev_open_thread; /* thread opening children */
+ kthread_t *vdev_validate_thread; /* thread validating children */
uint64_t vdev_crtxg; /* txg when top-level was added */
+ uint64_t vdev_root_zap;
/*
* Top-level vdev state.
uint64_t vdev_ms_shift; /* metaslab size shift */
uint64_t vdev_ms_count; /* number of metaslabs */
metaslab_group_t *vdev_mg; /* metaslab group */
+ metaslab_group_t *vdev_log_mg; /* embedded slog metaslab group */
metaslab_t **vdev_ms; /* metaslab array */
- uint64_t vdev_pending_fastwrite; /* allocated fastwrites */
txg_list_t vdev_ms_list; /* per-txg dirty metaslab lists */
txg_list_t vdev_dtl_list; /* per-txg dirty DTL lists */
txg_node_t vdev_txg_node; /* per-txg dirty vdev linkage */
list_node_t vdev_state_dirty_node; /* state dirty list */
uint64_t vdev_deflate_ratio; /* deflation ratio (x512) */
uint64_t vdev_islog; /* is an intent log device */
+ uint64_t vdev_noalloc; /* device is passivated? */
uint64_t vdev_removing; /* device is being removed? */
+ uint64_t vdev_failfast; /* device failfast setting */
+ boolean_t vdev_rz_expanding; /* raidz is being expanded? */
boolean_t vdev_ishole; /* is a hole in the namespace */
- kmutex_t vdev_queue_lock; /* protects vdev_queue_depth */
uint64_t vdev_top_zap;
+ vdev_alloc_bias_t vdev_alloc_bias; /* metaslab allocation bias */
/* pool checkpoint related */
space_map_t *vdev_checkpoint_sm; /* contains reserved blocks */
+ /* Initialize related */
+ boolean_t vdev_initialize_exit_wanted;
+ vdev_initializing_state_t vdev_initialize_state;
+ list_node_t vdev_initialize_node;
+ kthread_t *vdev_initialize_thread;
+ /* Protects vdev_initialize_thread and vdev_initialize_state. */
+ kmutex_t vdev_initialize_lock;
+ kcondvar_t vdev_initialize_cv;
+ uint64_t vdev_initialize_offset[TXG_SIZE];
+ uint64_t vdev_initialize_last_offset;
+ range_tree_t *vdev_initialize_tree; /* valid while initializing */
+ uint64_t vdev_initialize_bytes_est;
+ uint64_t vdev_initialize_bytes_done;
+ uint64_t vdev_initialize_action_time; /* start and end time */
+
+ /* TRIM related */
+ boolean_t vdev_trim_exit_wanted;
+ boolean_t vdev_autotrim_exit_wanted;
+ vdev_trim_state_t vdev_trim_state;
+ list_node_t vdev_trim_node;
+ kmutex_t vdev_autotrim_lock;
+ kcondvar_t vdev_autotrim_cv;
+ kcondvar_t vdev_autotrim_kick_cv;
+ kthread_t *vdev_autotrim_thread;
+ /* Protects vdev_trim_thread and vdev_trim_state. */
+ kmutex_t vdev_trim_lock;
+ kcondvar_t vdev_trim_cv;
+ kthread_t *vdev_trim_thread;
+ uint64_t vdev_trim_offset[TXG_SIZE];
+ uint64_t vdev_trim_last_offset;
+ uint64_t vdev_trim_bytes_est;
+ uint64_t vdev_trim_bytes_done;
+ uint64_t vdev_trim_rate; /* requested rate (bytes/sec) */
+ uint64_t vdev_trim_partial; /* requested partial TRIM */
+ uint64_t vdev_trim_secure; /* requested secure TRIM */
+ uint64_t vdev_trim_action_time; /* start and end time */
+
+ /* Rebuild related */
+ boolean_t vdev_rebuilding;
+ boolean_t vdev_rebuild_exit_wanted;
+ boolean_t vdev_rebuild_cancel_wanted;
+ boolean_t vdev_rebuild_reset_wanted;
+ kmutex_t vdev_rebuild_lock;
+ kcondvar_t vdev_rebuild_cv;
+ kthread_t *vdev_rebuild_thread;
+ vdev_rebuild_t vdev_rebuild_config;
+
+ /* For limiting outstanding I/Os (initialize, TRIM) */
+ kmutex_t vdev_initialize_io_lock;
+ kcondvar_t vdev_initialize_io_cv;
+ uint64_t vdev_initialize_inflight;
+ kmutex_t vdev_trim_io_lock;
+ kcondvar_t vdev_trim_io_cv;
+ uint64_t vdev_trim_inflight[3];
+
/*
* Values stored in the config for an indirect or removing vdev.
*/
range_tree_t *vdev_obsolete_segments;
space_map_t *vdev_obsolete_sm;
- /*
- * The queue depth parameters determine how many async writes are
- * still pending (i.e. allocated by net yet issued to disk) per
- * top-level (vdev_async_write_queue_depth) and the maximum allowed
- * (vdev_max_async_write_queue_depth). These values only apply to
- * top-level vdevs.
- */
- uint64_t vdev_async_write_queue_depth;
- uint64_t vdev_max_async_write_queue_depth;
-
/*
* Protects the vdev_scan_io_queue field itself as well as the
* structure's contents (when present).
uint64_t vdev_degraded; /* persistent degraded state */
uint64_t vdev_removed; /* persistent removed state */
uint64_t vdev_resilver_txg; /* persistent resilvering state */
- uint64_t vdev_nparity; /* number of parity devices for raidz */
+ uint64_t vdev_rebuild_txg; /* persistent rebuilding state */
char *vdev_path; /* vdev path (if any) */
char *vdev_devid; /* vdev devid (if any) */
char *vdev_physpath; /* vdev device path (if any) */
uint64_t vdev_not_present; /* not present during import */
uint64_t vdev_unspare; /* unspare when resilvering done */
boolean_t vdev_nowritecache; /* true if flushwritecache failed */
+ boolean_t vdev_has_trim; /* TRIM is supported */
+ boolean_t vdev_has_securetrim; /* secure TRIM is supported */
boolean_t vdev_checkremove; /* temporary online test */
boolean_t vdev_forcefault; /* force online fault */
boolean_t vdev_splitting; /* split or repair in progress */
boolean_t vdev_isspare; /* was a hot spare */
boolean_t vdev_isl2cache; /* was a l2cache device */
boolean_t vdev_copy_uberblocks; /* post expand copy uberblocks */
+ boolean_t vdev_resilver_deferred; /* resilver deferred */
+ boolean_t vdev_kobj_flag; /* kobj event record */
+ boolean_t vdev_attaching; /* vdev attach ashift handling */
vdev_queue_t vdev_queue; /* I/O deadline schedule queue */
- vdev_cache_t vdev_cache; /* physical block cache */
spa_aux_vdev_t *vdev_aux; /* for l2cache and spares vdevs */
zio_t *vdev_probe_zio; /* root of current probe */
vdev_aux_t vdev_label_aux; /* on-disk aux state */
uint64_t vdev_leaf_zap;
hrtime_t vdev_mmp_pending; /* 0 if write finished */
uint64_t vdev_mmp_kstat_id; /* to find kstat entry */
+ uint64_t vdev_expansion_time; /* vdev's last expansion time */
+ list_node_t vdev_leaf_node; /* leaf vdev list */
/*
* For DTrace to work in userland (libzpool) context, these fields must
kmutex_t vdev_probe_lock; /* protects vdev_probe_zio */
/*
- * We rate limit ZIO delay and ZIO checksum events, since they
+ * We rate limit ZIO delay, deadman, and checksum events, since they
* can flood ZED with tons of events when a drive is acting up.
*/
zfs_ratelimit_t vdev_delay_rl;
+ zfs_ratelimit_t vdev_deadman_rl;
zfs_ratelimit_t vdev_checksum_rl;
-};
-#define VDEV_RAIDZ_MAXPARITY 3
+ /*
+ * Vdev properties for tuning ZED
+ */
+ uint64_t vdev_checksum_n;
+ uint64_t vdev_checksum_t;
+ uint64_t vdev_io_n;
+ uint64_t vdev_io_t;
+ uint64_t vdev_slow_io_n;
+ uint64_t vdev_slow_io_t;
+};
#define VDEV_PAD_SIZE (8 << 10)
-/* 2 padding areas (vl_pad1 and vl_pad2) to skip */
+/* 2 padding areas (vl_pad1 and vl_be) to skip */
#define VDEV_SKIP_SIZE VDEV_PAD_SIZE * 2
#define VDEV_PHYS_SIZE (112 << 10)
#define VDEV_UBERBLOCK_RING (128 << 10)
zio_eck_t vp_zbt;
} vdev_phys_t;
+typedef enum vbe_vers {
+ /*
+ * The bootenv file is stored as ascii text in the envblock.
+ * It is used by the GRUB bootloader used on Linux to store the
+ * contents of the grubenv file. The file is stored as raw ASCII,
+ * and is protected by an embedded checksum. By default, GRUB will
+ * check if the boot filesystem supports storing the environment data
+ * in a special location, and if so, will invoke filesystem specific
+ * logic to retrieve it. This can be overridden by a variable, should
+ * the user so desire.
+ */
+ VB_RAW = 0,
+
+ /*
+ * The bootenv file is converted to an nvlist and then packed into the
+ * envblock.
+ */
+ VB_NVLIST = 1
+} vbe_vers_t;
+
+typedef struct vdev_boot_envblock {
+ uint64_t vbe_version;
+ char vbe_bootenv[VDEV_PAD_SIZE - sizeof (uint64_t) -
+ sizeof (zio_eck_t)];
+ zio_eck_t vbe_zbt;
+} vdev_boot_envblock_t;
+_Static_assert(sizeof (vdev_boot_envblock_t) == VDEV_PAD_SIZE,
+ "vdev_boot_envblock_t wrong size");
+
typedef struct vdev_label {
char vl_pad1[VDEV_PAD_SIZE]; /* 8K */
- char vl_pad2[VDEV_PAD_SIZE]; /* 8K */
+ vdev_boot_envblock_t vl_be; /* 8K */
vdev_phys_t vl_vdev_phys; /* 112K */
char vl_uberblock[VDEV_UBERBLOCK_RING]; /* 128K */
-} vdev_label_t; /* 256K total */
+} vdev_label_t; /* 256K total */
/*
* vdev_dirty() flags
/*
* Size of embedded boot loader region on each label.
* The total size of the first two labels plus the boot area is 4MB.
+ * On RAIDZ, this space is overwritten during RAIDZ expansion.
*/
#define VDEV_BOOT_SIZE (7ULL << 19) /* 3.5M */
#define VDEV_LABEL_END_SIZE (2 * sizeof (vdev_label_t))
#define VDEV_LABELS 4
#define VDEV_BEST_LABEL VDEV_LABELS
+#define VDEV_OFFSET_IS_LABEL(vd, off) \
+ (((off) < VDEV_LABEL_START_SIZE) || \
+ ((off) >= ((vd)->vdev_psize - VDEV_LABEL_END_SIZE)))
#define VDEV_ALLOC_LOAD 0
#define VDEV_ALLOC_ADD 1
extern vdev_ops_t vdev_mirror_ops;
extern vdev_ops_t vdev_replacing_ops;
extern vdev_ops_t vdev_raidz_ops;
+extern vdev_ops_t vdev_draid_ops;
+extern vdev_ops_t vdev_draid_spare_ops;
extern vdev_ops_t vdev_disk_ops;
extern vdev_ops_t vdev_file_ops;
extern vdev_ops_t vdev_missing_ops;
/*
* Common size functions
*/
-extern uint64_t vdev_default_asize(vdev_t *vd, uint64_t psize);
+extern void vdev_default_xlate(vdev_t *vd, const range_seg64_t *logical_rs,
+ range_seg64_t *physical_rs, range_seg64_t *remain_rs);
+extern uint64_t vdev_default_asize(vdev_t *vd, uint64_t psize, uint64_t txg);
+extern uint64_t vdev_default_min_asize(vdev_t *vd);
extern uint64_t vdev_get_min_asize(vdev_t *vd);
extern void vdev_set_min_asize(vdev_t *vd);
+extern uint64_t vdev_get_min_alloc(vdev_t *vd);
+extern uint64_t vdev_get_nparity(vdev_t *vd);
+extern uint64_t vdev_get_ndisks(vdev_t *vd);
/*
* Global variables
*/
-extern int vdev_standard_sm_blksz;
-/* zdb uses this tunable, so it must be declared here to make lint happy. */
-extern int zfs_vdev_cache_size;
+extern int zfs_vdev_standard_sm_blksz;
/*
* Functions from vdev_indirect.c
extern void vdev_indirect_sync_obsolete(vdev_t *vd, dmu_tx_t *tx);
extern boolean_t vdev_indirect_should_condense(vdev_t *vd);
extern void spa_condense_indirect_start_sync(vdev_t *vd, dmu_tx_t *tx);
-extern int vdev_obsolete_sm_object(vdev_t *vd);
-extern boolean_t vdev_obsolete_counts_are_precise(vdev_t *vd);
+extern int vdev_obsolete_sm_object(vdev_t *vd, uint64_t *sm_obj);
+extern int vdev_obsolete_counts_are_precise(vdev_t *vd, boolean_t *are_precise);
/*
* Other miscellaneous functions
*/
-int vdev_checkpoint_sm_object(vdev_t *vd);
+int vdev_checkpoint_sm_object(vdev_t *vd, uint64_t *sm_obj);
+void vdev_metaslab_group_create(vdev_t *vd);
+uint64_t vdev_best_ashift(uint64_t logical, uint64_t a, uint64_t b);
+
+/*
+ * Vdev ashift optimization tunables
+ */
+extern uint_t zfs_vdev_min_auto_ashift;
+extern uint_t zfs_vdev_max_auto_ashift;
+int param_set_min_auto_ashift(ZFS_MODULE_PARAM_ARGS);
+int param_set_max_auto_ashift(ZFS_MODULE_PARAM_ARGS);
#ifdef __cplusplus
}
projects / mirror_zfs.git / blobdiff