*/
/*
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 2011, 2014 by Delphix. All rights reserved.
+ * Copyright 2011 Nexenta Systems, Inc. All rights reserved.
+ * Copyright (c) 2014 Spectra Logic Corporation, All rights reserved.
+ * Copyright 2013 Saso Kiselkov. All rights reserved.
+ * Copyright (c) 2014 Integros [integros.com]
+ * Copyright 2017 Joyent, Inc.
+ * Copyright (c) 2017 Datto Inc.
*/
#ifndef _SYS_SPA_H
#include <sys/avl.h>
#include <sys/zfs_context.h>
+#include <sys/kstat.h>
#include <sys/nvpair.h>
#include <sys/sysmacros.h>
#include <sys/types.h>
#include <sys/fs/zfs.h>
+#include <sys/spa_checksum.h>
+#include <sys/dmu.h>
#ifdef __cplusplus
extern "C" {
typedef struct spa_aux_vdev spa_aux_vdev_t;
typedef struct ddt ddt_t;
typedef struct ddt_entry ddt_entry_t;
+typedef struct zbookmark_phys zbookmark_phys_t;
+
struct dsl_pool;
+struct dsl_dataset;
/*
* General-purpose 32-bit and 64-bit bitfield encodings.
#define BF32_GET(x, low, len) BF32_DECODE(x, low, len)
#define BF64_GET(x, low, len) BF64_DECODE(x, low, len)
-#define BF32_SET(x, low, len, val) \
- ((x) ^= BF32_ENCODE((x >> low) ^ (val), low, len))
-#define BF64_SET(x, low, len, val) \
- ((x) ^= BF64_ENCODE((x >> low) ^ (val), low, len))
+#define BF32_SET(x, low, len, val) do { \
+ ASSERT3U(val, <, 1U << (len)); \
+ ASSERT3U(low + len, <=, 32); \
+ (x) ^= BF32_ENCODE((x >> low) ^ (val), low, len); \
+_NOTE(CONSTCOND) } while (0)
+
+#define BF64_SET(x, low, len, val) do { \
+ ASSERT3U(val, <, 1ULL << (len)); \
+ ASSERT3U(low + len, <=, 64); \
+ ((x) ^= BF64_ENCODE((x >> low) ^ (val), low, len)); \
+_NOTE(CONSTCOND) } while (0)
#define BF32_GET_SB(x, low, len, shift, bias) \
((BF32_GET(x, low, len) + (bias)) << (shift))
#define BF64_GET_SB(x, low, len, shift, bias) \
((BF64_GET(x, low, len) + (bias)) << (shift))
-#define BF32_SET_SB(x, low, len, shift, bias, val) \
- BF32_SET(x, low, len, ((val) >> (shift)) - (bias))
-#define BF64_SET_SB(x, low, len, shift, bias, val) \
- BF64_SET(x, low, len, ((val) >> (shift)) - (bias))
+#define BF32_SET_SB(x, low, len, shift, bias, val) do { \
+ ASSERT(IS_P2ALIGNED(val, 1U << shift)); \
+ ASSERT3S((val) >> (shift), >=, bias); \
+ BF32_SET(x, low, len, ((val) >> (shift)) - (bias)); \
+_NOTE(CONSTCOND) } while (0)
+#define BF64_SET_SB(x, low, len, shift, bias, val) do { \
+ ASSERT(IS_P2ALIGNED(val, 1ULL << shift)); \
+ ASSERT3S((val) >> (shift), >=, bias); \
+ BF64_SET(x, low, len, ((val) >> (shift)) - (bias)); \
+_NOTE(CONSTCOND) } while (0)
/*
- * We currently support nine block sizes, from 512 bytes to 128K.
- * We could go higher, but the benefits are near-zero and the cost
- * of COWing a giant block to modify one byte would become excessive.
+ * We currently support block sizes from 512 bytes to 16MB.
+ * The benefits of larger blocks, and thus larger IO, need to be weighed
+ * against the cost of COWing a giant block to modify one byte, and the
+ * large latency of reading or writing a large block.
+ *
+ * Note that although blocks up to 16MB are supported, the recordsize
+ * property can not be set larger than zfs_max_recordsize (default 1MB).
+ * See the comment near zfs_max_recordsize in dsl_dataset.c for details.
+ *
+ * Note that although the LSIZE field of the blkptr_t can store sizes up
+ * to 32MB, the dnode's dn_datablkszsec can only store sizes up to
+ * 32MB - 512 bytes. Therefore, we limit SPA_MAXBLOCKSIZE to 16MB.
*/
#define SPA_MINBLOCKSHIFT 9
-#define SPA_MAXBLOCKSHIFT 17
+#define SPA_OLD_MAXBLOCKSHIFT 17
+#define SPA_MAXBLOCKSHIFT 24
#define SPA_MINBLOCKSIZE (1ULL << SPA_MINBLOCKSHIFT)
+#define SPA_OLD_MAXBLOCKSIZE (1ULL << SPA_OLD_MAXBLOCKSHIFT)
#define SPA_MAXBLOCKSIZE (1ULL << SPA_MAXBLOCKSHIFT)
-#define SPA_BLOCKSIZES (SPA_MAXBLOCKSHIFT - SPA_MINBLOCKSHIFT + 1)
+/*
+ * Alignment Shift (ashift) is an immutable, internal top-level vdev property
+ * which can only be set at vdev creation time. Physical writes are always done
+ * according to it, which makes 2^ashift the smallest possible IO on a vdev.
+ *
+ * We currently allow values ranging from 512 bytes (2^9 = 512) to 64 KiB
+ * (2^16 = 65,536).
+ */
+#define ASHIFT_MIN 9
+#define ASHIFT_MAX 16
/*
* Size of block to hold the configuration data (a packed nvlist)
*/
-#define SPA_CONFIG_BLOCKSIZE (1 << 14)
+#define SPA_CONFIG_BLOCKSIZE (1ULL << 14)
/*
* The DVA size encodings for LSIZE and PSIZE support blocks up to 32MB.
#define SPA_PSIZEBITS 16 /* PSIZE up to 32M (2^16 * 512) */
#define SPA_ASIZEBITS 24 /* ASIZE up to 64 times larger */
+#define SPA_COMPRESSBITS 7
+
/*
* All SPA data is represented by 128-bit data virtual addresses (DVAs).
* The members of the dva_t should be considered opaque outside the SPA.
uint64_t dva_word[2];
} dva_t;
+
/*
- * Each block has a 256-bit checksum -- strong enough for cryptographic hashes.
+ * Some checksums/hashes need a 256-bit initialization salt. This salt is kept
+ * secret and is suitable for use in MAC algorithms as the key.
*/
-typedef struct zio_cksum {
- uint64_t zc_word[4];
-} zio_cksum_t;
+typedef struct zio_cksum_salt {
+ uint8_t zcs_bytes[32];
+} zio_cksum_salt_t;
/*
* Each block is described by its DVAs, time of birth, checksum, etc.
* +-------+-------+-------+-------+-------+-------+-------+-------+
* 5 |G| offset3 |
* +-------+-------+-------+-------+-------+-------+-------+-------+
- * 6 |BDX|lvl| type | cksum | comp | PSIZE | LSIZE |
+ * 6 |BDX|lvl| type | cksum |E| comp| PSIZE | LSIZE |
* +-------+-------+-------+-------+-------+-------+-------+-------+
* 7 | padding |
* +-------+-------+-------+-------+-------+-------+-------+-------+
* G gang block indicator
* B byteorder (endianness)
* D dedup
- * X unused
+ * X encryption (on version 30, which is not supported)
+ * E blkptr_t contains embedded data (see below)
* lvl level of indirection
* type DMU object type
* phys birth txg of block allocation; zero if same as logical birth txg
* fill count number of non-zero blocks under this bp
* checksum[4] 256-bit checksum of the data this bp describes
*/
+
+/*
+ * "Embedded" blkptr_t's don't actually point to a block, instead they
+ * have a data payload embedded in the blkptr_t itself. See the comment
+ * in blkptr.c for more details.
+ *
+ * The blkptr_t is laid out as follows:
+ *
+ * 64 56 48 40 32 24 16 8 0
+ * +-------+-------+-------+-------+-------+-------+-------+-------+
+ * 0 | payload |
+ * 1 | payload |
+ * 2 | payload |
+ * 3 | payload |
+ * 4 | payload |
+ * 5 | payload |
+ * +-------+-------+-------+-------+-------+-------+-------+-------+
+ * 6 |BDX|lvl| type | etype |E| comp| PSIZE| LSIZE |
+ * +-------+-------+-------+-------+-------+-------+-------+-------+
+ * 7 | payload |
+ * 8 | payload |
+ * 9 | payload |
+ * +-------+-------+-------+-------+-------+-------+-------+-------+
+ * a | logical birth txg |
+ * +-------+-------+-------+-------+-------+-------+-------+-------+
+ * b | payload |
+ * c | payload |
+ * d | payload |
+ * e | payload |
+ * f | payload |
+ * +-------+-------+-------+-------+-------+-------+-------+-------+
+ *
+ * Legend:
+ *
+ * payload contains the embedded data
+ * B (byteorder) byteorder (endianness)
+ * D (dedup) padding (set to zero)
+ * X encryption (set to zero; see above)
+ * E (embedded) set to one
+ * lvl indirection level
+ * type DMU object type
+ * etype how to interpret embedded data (BP_EMBEDDED_TYPE_*)
+ * comp compression function of payload
+ * PSIZE size of payload after compression, in bytes
+ * LSIZE logical size of payload, in bytes
+ * note that 25 bits is enough to store the largest
+ * "normal" BP's LSIZE (2^16 * 2^9) in bytes
+ * log. birth transaction group in which the block was logically born
+ *
+ * Note that LSIZE and PSIZE are stored in bytes, whereas for non-embedded
+ * bp's they are stored in units of SPA_MINBLOCKSHIFT.
+ * Generally, the generic BP_GET_*() macros can be used on embedded BP's.
+ * The B, D, X, lvl, type, and comp fields are stored the same as with normal
+ * BP's so the BP_SET_* macros can be used with them. etype, PSIZE, LSIZE must
+ * be set with the BPE_SET_* macros. BP_SET_EMBEDDED() should be called before
+ * other macros, as they assert that they are only used on BP's of the correct
+ * "embedded-ness".
+ */
+
+#define BPE_GET_ETYPE(bp) \
+ (ASSERT(BP_IS_EMBEDDED(bp)), \
+ BF64_GET((bp)->blk_prop, 40, 8))
+#define BPE_SET_ETYPE(bp, t) do { \
+ ASSERT(BP_IS_EMBEDDED(bp)); \
+ BF64_SET((bp)->blk_prop, 40, 8, t); \
+_NOTE(CONSTCOND) } while (0)
+
+#define BPE_GET_LSIZE(bp) \
+ (ASSERT(BP_IS_EMBEDDED(bp)), \
+ BF64_GET_SB((bp)->blk_prop, 0, 25, 0, 1))
+#define BPE_SET_LSIZE(bp, x) do { \
+ ASSERT(BP_IS_EMBEDDED(bp)); \
+ BF64_SET_SB((bp)->blk_prop, 0, 25, 0, 1, x); \
+_NOTE(CONSTCOND) } while (0)
+
+#define BPE_GET_PSIZE(bp) \
+ (ASSERT(BP_IS_EMBEDDED(bp)), \
+ BF64_GET_SB((bp)->blk_prop, 25, 7, 0, 1))
+#define BPE_SET_PSIZE(bp, x) do { \
+ ASSERT(BP_IS_EMBEDDED(bp)); \
+ BF64_SET_SB((bp)->blk_prop, 25, 7, 0, 1, x); \
+_NOTE(CONSTCOND) } while (0)
+
+typedef enum bp_embedded_type {
+ BP_EMBEDDED_TYPE_DATA,
+ BP_EMBEDDED_TYPE_RESERVED, /* Reserved for an unintegrated feature. */
+ NUM_BP_EMBEDDED_TYPES = BP_EMBEDDED_TYPE_RESERVED
+} bp_embedded_type_t;
+
+#define BPE_NUM_WORDS 14
+#define BPE_PAYLOAD_SIZE (BPE_NUM_WORDS * sizeof (uint64_t))
+#define BPE_IS_PAYLOADWORD(bp, wp) \
+ ((wp) != &(bp)->blk_prop && (wp) != &(bp)->blk_birth)
+
#define SPA_BLKPTRSHIFT 7 /* blkptr_t is 128 bytes */
#define SPA_DVAS_PER_BP 3 /* Number of DVAs in a bp */
+/*
+ * A block is a hole when it has either 1) never been written to, or
+ * 2) is zero-filled. In both cases, ZFS can return all zeroes for all reads
+ * without physically allocating disk space. Holes are represented in the
+ * blkptr_t structure by zeroed blk_dva. Correct checking for holes is
+ * done through the BP_IS_HOLE macro. For holes, the logical size, level,
+ * DMU object type, and birth times are all also stored for holes that
+ * were written to at some point (i.e. were punched after having been filled).
+ */
typedef struct blkptr {
dva_t blk_dva[SPA_DVAS_PER_BP]; /* Data Virtual Addresses */
uint64_t blk_prop; /* size, compression, type, etc */
* Macros to get and set fields in a bp or DVA.
*/
#define DVA_GET_ASIZE(dva) \
- BF64_GET_SB((dva)->dva_word[0], 0, 24, SPA_MINBLOCKSHIFT, 0)
+ BF64_GET_SB((dva)->dva_word[0], 0, SPA_ASIZEBITS, SPA_MINBLOCKSHIFT, 0)
#define DVA_SET_ASIZE(dva, x) \
- BF64_SET_SB((dva)->dva_word[0], 0, 24, SPA_MINBLOCKSHIFT, 0, x)
+ BF64_SET_SB((dva)->dva_word[0], 0, SPA_ASIZEBITS, \
+ SPA_MINBLOCKSHIFT, 0, x)
#define DVA_GET_GRID(dva) BF64_GET((dva)->dva_word[0], 24, 8)
#define DVA_SET_GRID(dva, x) BF64_SET((dva)->dva_word[0], 24, 8, x)
#define DVA_SET_GANG(dva, x) BF64_SET((dva)->dva_word[1], 63, 1, x)
#define BP_GET_LSIZE(bp) \
- BF64_GET_SB((bp)->blk_prop, 0, 16, SPA_MINBLOCKSHIFT, 1)
-#define BP_SET_LSIZE(bp, x) \
- BF64_SET_SB((bp)->blk_prop, 0, 16, SPA_MINBLOCKSHIFT, 1, x)
+ (BP_IS_EMBEDDED(bp) ? \
+ (BPE_GET_ETYPE(bp) == BP_EMBEDDED_TYPE_DATA ? BPE_GET_LSIZE(bp) : 0): \
+ BF64_GET_SB((bp)->blk_prop, 0, SPA_LSIZEBITS, SPA_MINBLOCKSHIFT, 1))
+#define BP_SET_LSIZE(bp, x) do { \
+ ASSERT(!BP_IS_EMBEDDED(bp)); \
+ BF64_SET_SB((bp)->blk_prop, \
+ 0, SPA_LSIZEBITS, SPA_MINBLOCKSHIFT, 1, x); \
+_NOTE(CONSTCOND) } while (0)
#define BP_GET_PSIZE(bp) \
- BF64_GET_SB((bp)->blk_prop, 16, 16, SPA_MINBLOCKSHIFT, 1)
-#define BP_SET_PSIZE(bp, x) \
- BF64_SET_SB((bp)->blk_prop, 16, 16, SPA_MINBLOCKSHIFT, 1, x)
+ (BP_IS_EMBEDDED(bp) ? 0 : \
+ BF64_GET_SB((bp)->blk_prop, 16, SPA_PSIZEBITS, SPA_MINBLOCKSHIFT, 1))
+#define BP_SET_PSIZE(bp, x) do { \
+ ASSERT(!BP_IS_EMBEDDED(bp)); \
+ BF64_SET_SB((bp)->blk_prop, \
+ 16, SPA_PSIZEBITS, SPA_MINBLOCKSHIFT, 1, x); \
+_NOTE(CONSTCOND) } while (0)
-#define BP_GET_COMPRESS(bp) BF64_GET((bp)->blk_prop, 32, 8)
-#define BP_SET_COMPRESS(bp, x) BF64_SET((bp)->blk_prop, 32, 8, x)
+#define BP_GET_COMPRESS(bp) \
+ BF64_GET((bp)->blk_prop, 32, SPA_COMPRESSBITS)
+#define BP_SET_COMPRESS(bp, x) \
+ BF64_SET((bp)->blk_prop, 32, SPA_COMPRESSBITS, x)
-#define BP_GET_CHECKSUM(bp) BF64_GET((bp)->blk_prop, 40, 8)
-#define BP_SET_CHECKSUM(bp, x) BF64_SET((bp)->blk_prop, 40, 8, x)
+#define BP_IS_EMBEDDED(bp) BF64_GET((bp)->blk_prop, 39, 1)
+#define BP_SET_EMBEDDED(bp, x) BF64_SET((bp)->blk_prop, 39, 1, x)
+
+#define BP_GET_CHECKSUM(bp) \
+ (BP_IS_EMBEDDED(bp) ? ZIO_CHECKSUM_OFF : \
+ BF64_GET((bp)->blk_prop, 40, 8))
+#define BP_SET_CHECKSUM(bp, x) do { \
+ ASSERT(!BP_IS_EMBEDDED(bp)); \
+ BF64_SET((bp)->blk_prop, 40, 8, x); \
+_NOTE(CONSTCOND) } while (0)
#define BP_GET_TYPE(bp) BF64_GET((bp)->blk_prop, 48, 8)
#define BP_SET_TYPE(bp, x) BF64_SET((bp)->blk_prop, 48, 8, x)
#define BP_GET_LEVEL(bp) BF64_GET((bp)->blk_prop, 56, 5)
#define BP_SET_LEVEL(bp, x) BF64_SET((bp)->blk_prop, 56, 5, x)
-#define BP_GET_PROP_BIT_61(bp) BF64_GET((bp)->blk_prop, 61, 1)
-#define BP_SET_PROP_BIT_61(bp, x) BF64_SET((bp)->blk_prop, 61, 1, x)
-
#define BP_GET_DEDUP(bp) BF64_GET((bp)->blk_prop, 62, 1)
#define BP_SET_DEDUP(bp, x) BF64_SET((bp)->blk_prop, 62, 1, x)
-#define BP_GET_BYTEORDER(bp) (0 - BF64_GET((bp)->blk_prop, 63, 1))
+#define BP_GET_BYTEORDER(bp) BF64_GET((bp)->blk_prop, 63, 1)
#define BP_SET_BYTEORDER(bp, x) BF64_SET((bp)->blk_prop, 63, 1, x)
#define BP_PHYSICAL_BIRTH(bp) \
- ((bp)->blk_phys_birth ? (bp)->blk_phys_birth : (bp)->blk_birth)
+ (BP_IS_EMBEDDED(bp) ? 0 : \
+ (bp)->blk_phys_birth ? (bp)->blk_phys_birth : (bp)->blk_birth)
#define BP_SET_BIRTH(bp, logical, physical) \
{ \
+ ASSERT(!BP_IS_EMBEDDED(bp)); \
(bp)->blk_birth = (logical); \
(bp)->blk_phys_birth = ((logical) == (physical) ? 0 : (physical)); \
}
+#define BP_GET_FILL(bp) (BP_IS_EMBEDDED(bp) ? 1 : (bp)->blk_fill)
+
+#define BP_IS_METADATA(bp) \
+ (BP_GET_LEVEL(bp) > 0 || DMU_OT_IS_METADATA(BP_GET_TYPE(bp)))
+
#define BP_GET_ASIZE(bp) \
- (DVA_GET_ASIZE(&(bp)->blk_dva[0]) + DVA_GET_ASIZE(&(bp)->blk_dva[1]) + \
- DVA_GET_ASIZE(&(bp)->blk_dva[2]))
+ (BP_IS_EMBEDDED(bp) ? 0 : \
+ DVA_GET_ASIZE(&(bp)->blk_dva[0]) + \
+ DVA_GET_ASIZE(&(bp)->blk_dva[1]) + \
+ DVA_GET_ASIZE(&(bp)->blk_dva[2]))
-#define BP_GET_UCSIZE(bp) \
- ((BP_GET_LEVEL(bp) > 0 || dmu_ot[BP_GET_TYPE(bp)].ot_metadata) ? \
- BP_GET_PSIZE(bp) : BP_GET_LSIZE(bp))
+#define BP_GET_UCSIZE(bp) \
+ (BP_IS_METADATA(bp) ? BP_GET_PSIZE(bp) : BP_GET_LSIZE(bp))
#define BP_GET_NDVAS(bp) \
- (!!DVA_GET_ASIZE(&(bp)->blk_dva[0]) + \
+ (BP_IS_EMBEDDED(bp) ? 0 : \
+ !!DVA_GET_ASIZE(&(bp)->blk_dva[0]) + \
!!DVA_GET_ASIZE(&(bp)->blk_dva[1]) + \
!!DVA_GET_ASIZE(&(bp)->blk_dva[2]))
#define BP_COUNT_GANG(bp) \
+ (BP_IS_EMBEDDED(bp) ? 0 : \
(DVA_GET_GANG(&(bp)->blk_dva[0]) + \
DVA_GET_GANG(&(bp)->blk_dva[1]) + \
- DVA_GET_GANG(&(bp)->blk_dva[2]))
+ DVA_GET_GANG(&(bp)->blk_dva[2])))
#define DVA_EQUAL(dva1, dva2) \
((dva1)->dva_word[1] == (dva2)->dva_word[1] && \
#define BP_EQUAL(bp1, bp2) \
(BP_PHYSICAL_BIRTH(bp1) == BP_PHYSICAL_BIRTH(bp2) && \
+ (bp1)->blk_birth == (bp2)->blk_birth && \
DVA_EQUAL(&(bp1)->blk_dva[0], &(bp2)->blk_dva[0]) && \
DVA_EQUAL(&(bp1)->blk_dva[1], &(bp2)->blk_dva[1]) && \
DVA_EQUAL(&(bp1)->blk_dva[2], &(bp2)->blk_dva[2]))
-#define ZIO_CHECKSUM_EQUAL(zc1, zc2) \
- (0 == (((zc1).zc_word[0] - (zc2).zc_word[0]) | \
- ((zc1).zc_word[1] - (zc2).zc_word[1]) | \
- ((zc1).zc_word[2] - (zc2).zc_word[2]) | \
- ((zc1).zc_word[3] - (zc2).zc_word[3])))
#define DVA_IS_VALID(dva) (DVA_GET_ASIZE(dva) != 0)
-#define ZIO_SET_CHECKSUM(zcp, w0, w1, w2, w3) \
-{ \
- (zcp)->zc_word[0] = w0; \
- (zcp)->zc_word[1] = w1; \
- (zcp)->zc_word[2] = w2; \
- (zcp)->zc_word[3] = w3; \
-}
-
-#define BP_IDENTITY(bp) (&(bp)->blk_dva[0])
-#define BP_IS_GANG(bp) DVA_GET_GANG(BP_IDENTITY(bp))
-#define BP_IS_HOLE(bp) ((bp)->blk_birth == 0)
+#define BP_IDENTITY(bp) (ASSERT(!BP_IS_EMBEDDED(bp)), &(bp)->blk_dva[0])
+#define BP_IS_GANG(bp) \
+ (BP_IS_EMBEDDED(bp) ? B_FALSE : DVA_GET_GANG(BP_IDENTITY(bp)))
+#define DVA_IS_EMPTY(dva) ((dva)->dva_word[0] == 0ULL && \
+ (dva)->dva_word[1] == 0ULL)
+#define BP_IS_HOLE(bp) \
+ (!BP_IS_EMBEDDED(bp) && DVA_IS_EMPTY(BP_IDENTITY(bp)))
/* BP_IS_RAIDZ(bp) assumes no block compression */
#define BP_IS_RAIDZ(bp) (DVA_GET_ASIZE(&(bp)->blk_dva[0]) > \
ZIO_SET_CHECKSUM(&(bp)->blk_cksum, 0, 0, 0, 0); \
}
-/*
- * Note: the byteorder is either 0 or -1, both of which are palindromes.
- * This simplifies the endianness handling a bit.
- */
#ifdef _BIG_ENDIAN
#define ZFS_HOST_BYTEORDER (0ULL)
#else
-#define ZFS_HOST_BYTEORDER (-1ULL)
+#define ZFS_HOST_BYTEORDER (1ULL)
#endif
#define BP_SHOULD_BYTESWAP(bp) (BP_GET_BYTEORDER(bp) != ZFS_HOST_BYTEORDER)
* 'func' is either snprintf() or mdb_snprintf().
* 'ws' (whitespace) can be ' ' for single-line format, '\n' for multi-line.
*/
-#define SPRINTF_BLKPTR(func, ws, buf, bp, type, checksum, compress) \
+#define SNPRINTF_BLKPTR(func, ws, buf, size, bp, type, checksum, compress) \
{ \
static const char *copyname[] = \
{ "zero", "single", "double", "triple" }; \
- int size = BP_SPRINTF_LEN; \
int len = 0; \
int copies = 0; \
int d; \
\
if (bp == NULL) { \
- len = func(buf + len, size - len, "<NULL>"); \
+ len += func(buf + len, size - len, "<NULL>"); \
} else if (BP_IS_HOLE(bp)) { \
- len = func(buf + len, size - len, "<hole>"); \
+ len += func(buf + len, size - len, \
+ "HOLE [L%llu %s] " \
+ "size=%llxL birth=%lluL", \
+ (u_longlong_t)BP_GET_LEVEL(bp), \
+ type, \
+ (u_longlong_t)BP_GET_LSIZE(bp), \
+ (u_longlong_t)bp->blk_birth); \
+ } else if (BP_IS_EMBEDDED(bp)) { \
+ len = func(buf + len, size - len, \
+ "EMBEDDED [L%llu %s] et=%u %s " \
+ "size=%llxL/%llxP birth=%lluL", \
+ (u_longlong_t)BP_GET_LEVEL(bp), \
+ type, \
+ (int)BPE_GET_ETYPE(bp), \
+ compress, \
+ (u_longlong_t)BPE_GET_LSIZE(bp), \
+ (u_longlong_t)BPE_GET_PSIZE(bp), \
+ (u_longlong_t)bp->blk_birth); \
} else { \
for (d = 0; d < BP_GET_NDVAS(bp); d++) { \
const dva_t *dva = &bp->blk_dva[d]; \
(u_longlong_t)BP_GET_PSIZE(bp), \
(u_longlong_t)bp->blk_birth, \
(u_longlong_t)BP_PHYSICAL_BIRTH(bp), \
- (u_longlong_t)bp->blk_fill, \
+ (u_longlong_t)BP_GET_FILL(bp), \
ws, \
(u_longlong_t)bp->blk_cksum.zc_word[0], \
(u_longlong_t)bp->blk_cksum.zc_word[1], \
ASSERT(len < size); \
}
-#include <sys/dmu.h>
-
#define BP_GET_BUFC_TYPE(bp) \
- (((BP_GET_LEVEL(bp) > 0) || (dmu_ot[BP_GET_TYPE(bp)].ot_metadata)) ? \
- ARC_BUFC_METADATA : ARC_BUFC_DATA);
+ (BP_IS_METADATA(bp) ? ARC_BUFC_METADATA : ARC_BUFC_DATA)
typedef enum spa_import_type {
SPA_IMPORT_EXISTING,
extern int spa_open(const char *pool, spa_t **, void *tag);
extern int spa_open_rewind(const char *pool, spa_t **, void *tag,
nvlist_t *policy, nvlist_t **config);
-extern int spa_get_stats(const char *pool, nvlist_t **config,
- char *altroot, size_t buflen);
+extern int spa_get_stats(const char *pool, nvlist_t **config, char *altroot,
+ size_t buflen);
extern int spa_create(const char *pool, nvlist_t *config, nvlist_t *props,
- const char *history_str, nvlist_t *zplprops);
-extern int spa_import_rootpool(char *devpath, char *devid);
-extern int spa_import(const char *pool, nvlist_t *config, nvlist_t *props,
+ nvlist_t *zplprops);
+extern int spa_import(char *pool, nvlist_t *config, nvlist_t *props,
uint64_t flags);
extern nvlist_t *spa_tryimport(nvlist_t *tryconfig);
extern int spa_destroy(char *pool);
/* scanning */
extern int spa_scan(spa_t *spa, pool_scan_func_t func);
extern int spa_scan_stop(spa_t *spa);
+extern int spa_scrub_pause_resume(spa_t *spa, pool_scrub_cmd_t flag);
/* spa syncing */
extern void spa_sync(spa_t *spa, uint64_t txg); /* only for DMU use */
extern void spa_sync_allpools(void);
-/*
- * DEFERRED_FREE must be large enough that regular blocks are not
- * deferred. XXX so can't we change it back to 1?
- */
-#define SYNC_PASS_DEFERRED_FREE 2 /* defer frees after this pass */
-#define SYNC_PASS_DONT_COMPRESS 4 /* don't compress after this pass */
-#define SYNC_PASS_REWRITE 1 /* rewrite new bps after this pass */
+extern int zfs_sync_pass_deferred_free;
/* spa namespace global mutex */
extern kmutex_t spa_namespace_lock;
/* Refcount functions */
extern void spa_open_ref(spa_t *spa, void *tag);
extern void spa_close(spa_t *spa, void *tag);
+extern void spa_async_close(spa_t *spa, void *tag);
extern boolean_t spa_refcount_zero(spa_t *spa);
#define SCL_NONE 0x00
#define SCL_ALL ((1 << SCL_LOCKS) - 1)
#define SCL_STATE_ALL (SCL_STATE | SCL_L2ARC | SCL_ZIO)
+/* Historical pool statistics */
+typedef struct spa_stats_history {
+ kmutex_t lock;
+ uint64_t count;
+ uint64_t size;
+ kstat_t *kstat;
+ void *private;
+ list_t list;
+} spa_stats_history_t;
+
+typedef struct spa_stats {
+ spa_stats_history_t read_history;
+ spa_stats_history_t txg_history;
+ spa_stats_history_t tx_assign_histogram;
+ spa_stats_history_t io_history;
+ spa_stats_history_t mmp_history;
+ spa_stats_history_t state; /* pool state */
+} spa_stats_t;
+
+typedef enum txg_state {
+ TXG_STATE_BIRTH = 0,
+ TXG_STATE_OPEN = 1,
+ TXG_STATE_QUIESCED = 2,
+ TXG_STATE_WAIT_FOR_SYNC = 3,
+ TXG_STATE_SYNCED = 4,
+ TXG_STATE_COMMITTED = 5,
+} txg_state_t;
+
+typedef struct txg_stat {
+ vdev_stat_t vs1;
+ vdev_stat_t vs2;
+ uint64_t txg;
+ uint64_t ndirty;
+} txg_stat_t;
+
+extern void spa_stats_init(spa_t *spa);
+extern void spa_stats_destroy(spa_t *spa);
+extern void spa_read_history_add(spa_t *spa, const zbookmark_phys_t *zb,
+ uint32_t aflags);
+extern void spa_txg_history_add(spa_t *spa, uint64_t txg, hrtime_t birth_time);
+extern int spa_txg_history_set(spa_t *spa, uint64_t txg,
+ txg_state_t completed_state, hrtime_t completed_time);
+extern txg_stat_t *spa_txg_history_init_io(spa_t *, uint64_t,
+ struct dsl_pool *);
+extern void spa_txg_history_fini_io(spa_t *, txg_stat_t *);
+extern void spa_tx_assign_add_nsecs(spa_t *spa, uint64_t nsecs);
+extern int spa_mmp_history_set_skip(spa_t *spa, uint64_t mmp_kstat_id);
+extern int spa_mmp_history_set(spa_t *spa, uint64_t mmp_kstat_id, int io_error,
+ hrtime_t duration);
+extern void *spa_mmp_history_add(spa_t *spa, uint64_t txg, uint64_t timestamp,
+ uint64_t mmp_delay, vdev_t *vd, int label, uint64_t mmp_kstat_id,
+ int error);
+
/* Pool configuration locks */
extern int spa_config_tryenter(spa_t *spa, int locks, void *tag, krw_t rw);
extern void spa_config_enter(spa_t *spa, int locks, void *tag, krw_t rw);
/* Log claim callback */
extern void spa_claim_notify(zio_t *zio);
+extern void spa_deadman(void *);
/* Accessor functions */
extern boolean_t spa_shutting_down(spa_t *spa);
extern struct dsl_pool *spa_get_dsl(spa_t *spa);
+extern boolean_t spa_is_initializing(spa_t *spa);
extern blkptr_t *spa_get_rootblkptr(spa_t *spa);
extern void spa_set_rootblkptr(spa_t *spa, const blkptr_t *bp);
extern void spa_altroot(spa_t *, char *, size_t);
extern int spa_sync_pass(spa_t *spa);
extern char *spa_name(spa_t *spa);
extern uint64_t spa_guid(spa_t *spa);
+extern uint64_t spa_load_guid(spa_t *spa);
extern uint64_t spa_last_synced_txg(spa_t *spa);
extern uint64_t spa_first_txg(spa_t *spa);
extern uint64_t spa_syncing_txg(spa_t *spa);
+extern uint64_t spa_final_dirty_txg(spa_t *spa);
extern uint64_t spa_version(spa_t *spa);
extern pool_state_t spa_state(spa_t *spa);
extern spa_load_state_t spa_load_state(spa_t *spa);
extern uint64_t spa_freeze_txg(spa_t *spa);
-extern uint64_t spa_get_asize(spa_t *spa, uint64_t lsize);
+extern uint64_t spa_get_worst_case_asize(spa_t *spa, uint64_t lsize);
extern uint64_t spa_get_dspace(spa_t *spa);
+extern uint64_t spa_get_slop_space(spa_t *spa);
extern void spa_update_dspace(spa_t *spa);
extern uint64_t spa_version(spa_t *spa);
extern boolean_t spa_deflate(spa_t *spa);
extern metaslab_class_t *spa_normal_class(spa_t *spa);
extern metaslab_class_t *spa_log_class(spa_t *spa);
+extern void spa_evicting_os_register(spa_t *, objset_t *os);
+extern void spa_evicting_os_deregister(spa_t *, objset_t *os);
+extern void spa_evicting_os_wait(spa_t *spa);
extern int spa_max_replication(spa_t *spa);
extern int spa_prev_software_version(spa_t *spa);
-extern int spa_busy(void);
extern uint8_t spa_get_failmode(spa_t *spa);
extern boolean_t spa_suspended(spa_t *spa);
extern uint64_t spa_bootfs(spa_t *spa);
extern uint64_t spa_delegation(spa_t *spa);
extern objset_t *spa_meta_objset(spa_t *spa);
+extern uint64_t spa_deadman_synctime(spa_t *spa);
/* Miscellaneous support routines */
+extern void spa_activate_mos_feature(spa_t *spa, const char *feature,
+ dmu_tx_t *tx);
+extern void spa_deactivate_mos_feature(spa_t *spa, const char *feature);
extern int spa_rename(const char *oldname, const char *newname);
extern spa_t *spa_by_guid(uint64_t pool_guid, uint64_t device_guid);
extern boolean_t spa_guid_exists(uint64_t pool_guid, uint64_t device_guid);
extern void spa_strfree(char *);
extern uint64_t spa_get_random(uint64_t range);
extern uint64_t spa_generate_guid(spa_t *spa);
-extern void sprintf_blkptr(char *buf, const blkptr_t *bp);
+extern void snprintf_blkptr(char *buf, size_t buflen, const blkptr_t *bp);
extern void spa_freeze(spa_t *spa);
+extern int spa_change_guid(spa_t *spa);
extern void spa_upgrade(spa_t *spa, uint64_t version);
extern void spa_evict_all(void);
extern vdev_t *spa_lookup_by_guid(spa_t *spa, uint64_t guid,
extern boolean_t spa_has_slogs(spa_t *spa);
extern boolean_t spa_is_root(spa_t *spa);
extern boolean_t spa_writeable(spa_t *spa);
+extern boolean_t spa_has_pending_synctask(spa_t *spa);
+extern int spa_maxblocksize(spa_t *spa);
+extern int spa_maxdnodesize(spa_t *spa);
+extern void zfs_blkptr_verify(spa_t *spa, const blkptr_t *bp);
+extern boolean_t spa_multihost(spa_t *spa);
+extern unsigned long spa_get_hostid(void);
extern int spa_mode(spa_t *spa);
-extern uint64_t strtonum(const char *str, char **nptr);
-
-/* history logging */
-typedef enum history_log_type {
- LOG_CMD_POOL_CREATE,
- LOG_CMD_NORMAL,
- LOG_INTERNAL
-} history_log_type_t;
-
-typedef struct history_arg {
- char *ha_history_str;
- history_log_type_t ha_log_type;
- history_internal_events_t ha_event;
- char *ha_zone;
- uid_t ha_uid;
-} history_arg_t;
+extern uint64_t zfs_strtonum(const char *str, char **nptr);
extern char *spa_his_ievent_table[];
extern void spa_history_create_obj(spa_t *spa, dmu_tx_t *tx);
extern int spa_history_get(spa_t *spa, uint64_t *offset, uint64_t *len_read,
char *his_buf);
-extern int spa_history_log(spa_t *spa, const char *his_buf,
- history_log_type_t what);
-extern void spa_history_log_internal(history_internal_events_t event,
- spa_t *spa, dmu_tx_t *tx, const char *fmt, ...);
-extern void spa_history_log_version(spa_t *spa, history_internal_events_t evt);
+extern int spa_history_log(spa_t *spa, const char *his_buf);
+extern int spa_history_log_nvl(spa_t *spa, nvlist_t *nvl);
+extern void spa_history_log_version(spa_t *spa, const char *operation,
+ dmu_tx_t *tx);
+extern void spa_history_log_internal(spa_t *spa, const char *operation,
+ dmu_tx_t *tx, const char *fmt, ...);
+extern void spa_history_log_internal_ds(struct dsl_dataset *ds, const char *op,
+ dmu_tx_t *tx, const char *fmt, ...);
+extern void spa_history_log_internal_dd(dsl_dir_t *dd, const char *operation,
+ dmu_tx_t *tx, const char *fmt, ...);
+
+extern const char *spa_state_to_name(spa_t *spa);
/* error handling */
-struct zbookmark;
+struct zbookmark_phys;
extern void spa_log_error(spa_t *spa, zio_t *zio);
extern void zfs_ereport_post(const char *class, spa_t *spa, vdev_t *vd,
zio_t *zio, uint64_t stateoroffset, uint64_t length);
+extern nvlist_t *zfs_event_create(spa_t *spa, vdev_t *vd, const char *type,
+ const char *name, nvlist_t *aux);
extern void zfs_post_remove(spa_t *spa, vdev_t *vd);
-extern void zfs_post_state_change(spa_t *spa, vdev_t *vd);
+extern void zfs_post_state_change(spa_t *spa, vdev_t *vd, uint64_t laststate);
extern void zfs_post_autoreplace(spa_t *spa, vdev_t *vd);
extern uint64_t spa_get_errlog_size(spa_t *spa);
extern int spa_get_errlog(spa_t *spa, void *uaddr, size_t *count);
extern void spa_configfile_set(spa_t *, nvlist_t *, boolean_t);
/* asynchronous event notification */
-extern void spa_event_notify(spa_t *spa, vdev_t *vdev, const char *name);
+extern void spa_event_notify(spa_t *spa, vdev_t *vdev, nvlist_t *hist_nvl,
+ const char *name);
#ifdef ZFS_DEBUG
#define dprintf_bp(bp, fmt, ...) do { \
- if (zfs_flags & ZFS_DEBUG_DPRINTF) { \
+ if (zfs_flags & ZFS_DEBUG_DPRINTF) { \
char *__blkbuf = kmem_alloc(BP_SPRINTF_LEN, KM_SLEEP); \
- sprintf_blkptr(__blkbuf, (bp)); \
+ snprintf_blkptr(__blkbuf, BP_SPRINTF_LEN, (bp)); \
dprintf(fmt " %s\n", __VA_ARGS__, __blkbuf); \
kmem_free(__blkbuf, BP_SPRINTF_LEN); \
} \
#define dprintf_bp(bp, fmt, ...)
#endif
+extern boolean_t spa_debug_enabled(spa_t *spa);
+#define spa_dbgmsg(spa, ...) \
+{ \
+ if (spa_debug_enabled(spa)) \
+ zfs_dbgmsg(__VA_ARGS__); \
+}
+
extern int spa_mode_global; /* mode, e.g. FREAD | FWRITE */
#ifdef __cplusplus
projects / mirror_zfs-debian.git / blobdiff