include/sys/zio_impl.h

   1 /*
   2  * CDDL HEADER START
   3  *
   4  * The contents of this file are subject to the terms of the
   5  * Common Development and Distribution License (the "License").
   6  * You may not use this file except in compliance with the License.
   7  *
   8  * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
   9  * or http://www.opensolaris.org/os/licensing.
  10  * See the License for the specific language governing permissions
  11  * and limitations under the License.
  12  *
  13  * When distributing Covered Code, include this CDDL HEADER in each
  14  * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
  15  * If applicable, add the following below this CDDL HEADER, with the
  16  * fields enclosed by brackets "[]" replaced with your own identifying
  17  * information: Portions Copyright [yyyy] [name of copyright owner]
  18  *
  19  * CDDL HEADER END
  20  */
  21 /*
  22  * Copyright 2009 Sun Microsystems, Inc.  All rights reserved.
  23  * Use is subject to license terms.
  24  */
  25
  26 /*
  27  * Copyright (c) 2012, 2015 by Delphix. All rights reserved.
  28  */
  29
  30 #ifndef _ZIO_IMPL_H
  31 #define _ZIO_IMPL_H
  32
  33 #ifdef  __cplusplus
  34 extern "C" {
  35 #endif
  36
  37 /*
  38  * XXX -- Describe ZFS I/O pipeline here. Fill in as needed.
  39  *
  40  * The ZFS I/O pipeline is comprised of various stages which are defined
  41  * in the zio_stage enum below. The individual stages are used to construct
  42  * these basic I/O operations: Read, Write, Free, Claim, and Ioctl.
  43  *
  44  * I/O operations: (XXX - provide detail for each of the operations)
  45  *
  46  * Read:
  47  * Write:
  48  * Free:
  49  * Claim:
  50  * Ioctl:
  51  *
  52  * Although the most common pipeline are used by the basic I/O operations
  53  * above, there are some helper pipelines (one could consider them
  54  * sub-pipelines) which are used internally by the ZIO module and are
  55  * explained below:
  56  *
  57  * Interlock Pipeline:
  58  * The interlock pipeline is the most basic pipeline and is used by all
  59  * of the I/O operations. The interlock pipeline does not perform any I/O
  60  * and is used to coordinate the dependencies between I/Os that are being
  61  * issued (i.e. the parent/child relationship).
  62  *
  63  * Vdev child Pipeline:
  64  * The vdev child pipeline is responsible for performing the physical I/O.
  65  * It is in this pipeline where the I/O are queued and possibly cached.
  66  *
  67  * In addition to performing I/O, the pipeline is also responsible for
  68  * data transformations. The transformations performed are based on the
  69  * specific properties that user may have selected and modify the
  70  * behavior of the pipeline. Examples of supported transformations are
  71  * compression, dedup, and nop writes. Transformations will either modify
  72  * the data or the pipeline. This list below further describes each of
  73  * the supported transformations:
  74  *
  75  * Compression:
  76  * ZFS supports five different flavors of compression -- gzip, lzjb, lz4, zle,
  77  * and zstd. Compression occurs as part of the write pipeline and is
  78  * performed in the ZIO_STAGE_WRITE_BP_INIT stage.
  79  *
  80  * Dedup:
  81  * Dedup reads are handled by the ZIO_STAGE_DDT_READ_START and
  82  * ZIO_STAGE_DDT_READ_DONE stages. These stages are added to an existing
  83  * read pipeline if the dedup bit is set on the block pointer.
  84  * Writing a dedup block is performed by the ZIO_STAGE_DDT_WRITE stage
  85  * and added to a write pipeline if a user has enabled dedup on that
  86  * particular dataset.
  87  *
  88  * NOP Write:
  89  * The NOP write feature is performed by the ZIO_STAGE_NOP_WRITE stage
  90  * and is added to an existing write pipeline if a cryptographically
  91  * secure checksum (i.e. SHA256) is enabled and compression is turned on.
  92  * The NOP write stage will compare the checksums of the current data
  93  * on-disk (level-0 blocks only) and the data that is currently being written.
  94  * If the checksum values are identical then the pipeline is converted to
  95  * an interlock pipeline skipping block allocation and bypassing the
  96  * physical I/O.  The nop write feature can handle writes in either
  97  * syncing or open context (i.e. zil writes) and as a result is mutually
  98  * exclusive with dedup.
  99  *
 100  * Encryption:
 101  * Encryption and authentication is handled by the ZIO_STAGE_ENCRYPT stage.
 102  * This stage determines how the encryption metadata is stored in the bp.
 103  * Decryption and MAC verification is performed during zio_decrypt() as a
 104  * transform callback. Encryption is mutually exclusive with nopwrite, because
 105  * blocks with the same plaintext will be encrypted with different salts and
 106  * IV's (if dedup is off), and therefore have different ciphertexts. For dedup
 107  * blocks we deterministically generate the IV and salt by performing an HMAC
 108  * of the plaintext, which is computationally expensive, but allows us to keep
 109  * support for encrypted dedup. See the block comment in zio_crypt.c for
 110  * details.
 111  */
 112
 113 /*
 114  * zio pipeline stage definitions
 115  */
 116 enum zio_stage {
 117         ZIO_STAGE_OPEN                  = 1 << 0,       /* RWFCI */
 118
 119         ZIO_STAGE_READ_BP_INIT          = 1 << 1,       /* R---- */
 120         ZIO_STAGE_WRITE_BP_INIT         = 1 << 2,       /* -W--- */
 121         ZIO_STAGE_FREE_BP_INIT          = 1 << 3,       /* --F-- */
 122         ZIO_STAGE_ISSUE_ASYNC           = 1 << 4,       /* RWF-- */
 123         ZIO_STAGE_WRITE_COMPRESS        = 1 << 5,       /* -W--- */
 124
 125         ZIO_STAGE_ENCRYPT               = 1 << 6,       /* -W--- */
 126         ZIO_STAGE_CHECKSUM_GENERATE     = 1 << 7,       /* -W--- */
 127
 128         ZIO_STAGE_NOP_WRITE             = 1 << 8,       /* -W--- */
 129
 130         ZIO_STAGE_DDT_READ_START        = 1 << 9,       /* R---- */
 131         ZIO_STAGE_DDT_READ_DONE         = 1 << 10,      /* R---- */
 132         ZIO_STAGE_DDT_WRITE             = 1 << 11,      /* -W--- */
 133         ZIO_STAGE_DDT_FREE              = 1 << 12,      /* --F-- */
 134
 135         ZIO_STAGE_GANG_ASSEMBLE         = 1 << 13,      /* RWFC- */
 136         ZIO_STAGE_GANG_ISSUE            = 1 << 14,      /* RWFC- */
 137
 138         ZIO_STAGE_DVA_THROTTLE          = 1 << 15,      /* -W--- */
 139         ZIO_STAGE_DVA_ALLOCATE          = 1 << 16,      /* -W--- */
 140         ZIO_STAGE_DVA_FREE              = 1 << 17,      /* --F-- */
 141         ZIO_STAGE_DVA_CLAIM             = 1 << 18,      /* ---C- */
 142
 143         ZIO_STAGE_READY                 = 1 << 19,      /* RWFCI */
 144
 145         ZIO_STAGE_VDEV_IO_START         = 1 << 20,      /* RW--I */
 146         ZIO_STAGE_VDEV_IO_DONE          = 1 << 21,      /* RW--I */
 147         ZIO_STAGE_VDEV_IO_ASSESS        = 1 << 22,      /* RW--I */
 148
 149         ZIO_STAGE_CHECKSUM_VERIFY       = 1 << 23,      /* R---- */
 150
 151         ZIO_STAGE_DONE                  = 1 << 24       /* RWFCI */
 152 };
 153
 154 #define ZIO_INTERLOCK_STAGES                    \
 155         (ZIO_STAGE_READY |                      \
 156         ZIO_STAGE_DONE)
 157
 158 #define ZIO_INTERLOCK_PIPELINE                  \
 159         ZIO_INTERLOCK_STAGES
 160
 161 #define ZIO_VDEV_IO_STAGES                      \
 162         (ZIO_STAGE_VDEV_IO_START |              \
 163         ZIO_STAGE_VDEV_IO_DONE |                \
 164         ZIO_STAGE_VDEV_IO_ASSESS)
 165
 166 #define ZIO_VDEV_CHILD_PIPELINE                 \
 167         (ZIO_VDEV_IO_STAGES |                   \
 168         ZIO_STAGE_DONE)
 169
 170 #define ZIO_READ_COMMON_STAGES                  \
 171         (ZIO_INTERLOCK_STAGES |                 \
 172         ZIO_VDEV_IO_STAGES |                    \
 173         ZIO_STAGE_CHECKSUM_VERIFY)
 174
 175 #define ZIO_READ_PHYS_PIPELINE                  \
 176         ZIO_READ_COMMON_STAGES
 177
 178 #define ZIO_READ_PIPELINE                       \
 179         (ZIO_READ_COMMON_STAGES |               \
 180         ZIO_STAGE_READ_BP_INIT)
 181
 182 #define ZIO_DDT_CHILD_READ_PIPELINE             \
 183         ZIO_READ_COMMON_STAGES
 184
 185 #define ZIO_DDT_READ_PIPELINE                   \
 186         (ZIO_INTERLOCK_STAGES |                 \
 187         ZIO_STAGE_READ_BP_INIT |                \
 188         ZIO_STAGE_DDT_READ_START |              \
 189         ZIO_STAGE_DDT_READ_DONE)
 190
 191 #define ZIO_WRITE_COMMON_STAGES                 \
 192         (ZIO_INTERLOCK_STAGES |                 \
 193         ZIO_VDEV_IO_STAGES |                    \
 194         ZIO_STAGE_ISSUE_ASYNC |                 \
 195         ZIO_STAGE_CHECKSUM_GENERATE)
 196
 197 #define ZIO_WRITE_PHYS_PIPELINE                 \
 198         ZIO_WRITE_COMMON_STAGES
 199
 200 #define ZIO_REWRITE_PIPELINE                    \
 201         (ZIO_WRITE_COMMON_STAGES |              \
 202         ZIO_STAGE_WRITE_COMPRESS |              \
 203         ZIO_STAGE_ENCRYPT |                     \
 204         ZIO_STAGE_WRITE_BP_INIT)
 205
 206 #define ZIO_WRITE_PIPELINE                      \
 207         (ZIO_WRITE_COMMON_STAGES |              \
 208         ZIO_STAGE_WRITE_BP_INIT |               \
 209         ZIO_STAGE_WRITE_COMPRESS |              \
 210         ZIO_STAGE_ENCRYPT |                     \
 211         ZIO_STAGE_DVA_THROTTLE |                \
 212         ZIO_STAGE_DVA_ALLOCATE)
 213
 214 #define ZIO_DDT_CHILD_WRITE_PIPELINE            \
 215         (ZIO_INTERLOCK_STAGES |                 \
 216         ZIO_VDEV_IO_STAGES |                    \
 217         ZIO_STAGE_DVA_THROTTLE |                \
 218         ZIO_STAGE_DVA_ALLOCATE)
 219
 220 #define ZIO_DDT_WRITE_PIPELINE                  \
 221         (ZIO_INTERLOCK_STAGES |                 \
 222         ZIO_STAGE_WRITE_BP_INIT |               \
 223         ZIO_STAGE_ISSUE_ASYNC |                 \
 224         ZIO_STAGE_WRITE_COMPRESS |              \
 225         ZIO_STAGE_ENCRYPT |                     \
 226         ZIO_STAGE_CHECKSUM_GENERATE |           \
 227         ZIO_STAGE_DDT_WRITE)
 228
 229 #define ZIO_GANG_STAGES                         \
 230         (ZIO_STAGE_GANG_ASSEMBLE |              \
 231         ZIO_STAGE_GANG_ISSUE)
 232
 233 #define ZIO_FREE_PIPELINE                       \
 234         (ZIO_INTERLOCK_STAGES |                 \
 235         ZIO_STAGE_FREE_BP_INIT |                \
 236         ZIO_STAGE_DVA_FREE)
 237
 238 #define ZIO_DDT_FREE_PIPELINE                   \
 239         (ZIO_INTERLOCK_STAGES |                 \
 240         ZIO_STAGE_FREE_BP_INIT |                \
 241         ZIO_STAGE_ISSUE_ASYNC |                 \
 242         ZIO_STAGE_DDT_FREE)
 243
 244 #define ZIO_CLAIM_PIPELINE                      \
 245         (ZIO_INTERLOCK_STAGES |                 \
 246         ZIO_STAGE_DVA_CLAIM)
 247
 248 #define ZIO_IOCTL_PIPELINE                      \
 249         (ZIO_INTERLOCK_STAGES |                 \
 250         ZIO_STAGE_VDEV_IO_START |               \
 251         ZIO_STAGE_VDEV_IO_ASSESS)
 252
 253 #define ZIO_TRIM_PIPELINE                       \
 254         (ZIO_INTERLOCK_STAGES |                 \
 255         ZIO_STAGE_ISSUE_ASYNC |                 \
 256         ZIO_VDEV_IO_STAGES)
 257
 258 #define ZIO_BLOCKING_STAGES                     \
 259         (ZIO_STAGE_DVA_ALLOCATE |               \
 260         ZIO_STAGE_DVA_CLAIM |                   \
 261         ZIO_STAGE_VDEV_IO_START)
 262
 263 extern void zio_inject_init(void);
 264 extern void zio_inject_fini(void);
 265
 266 #ifdef  __cplusplus
 267 }
 268 #endif
 269
 270 #endif  /* _ZIO_IMPL_H */