diff --git a/cmd/zstream/zstream_token.c b/cmd/zstream/zstream_token.c index 23cc56dce37c..795a372633ef 100644 --- a/cmd/zstream/zstream_token.c +++ b/cmd/zstream/zstream_token.c @@ -1,78 +1,82 @@ /* * CDDL HEADER START * * The contents of this file are subject to the terms of the * Common Development and Distribution License (the "License"). * 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 https://opensource.org/licenses/CDDL-1.0. * See the License for the specific language governing permissions * and limitations under the License. * * When distributing Covered Code, include this CDDL HEADER in each * file and include the License file at usr/src/OPENSOLARIS.LICENSE. * If applicable, add the following below this CDDL HEADER, with the * fields enclosed by brackets "[]" replaced with your own identifying * information: Portions Copyright [yyyy] [name of copyright owner] * * CDDL HEADER END */ /* * Copyright 2010 Sun Microsystems, Inc. All rights reserved. * Use is subject to license terms. * * Portions Copyright 2012 Martin Matuska */ /* * Copyright (c) 2020 by Datto Inc. All rights reserved. */ #include #include #include #include #include #include #include #include #include #include #include #include "zstream.h" int zstream_do_token(int argc, char *argv[]) { char *resume_token = NULL; + libzfs_handle_t *hdl; if (argc < 2) { (void) fprintf(stderr, "Need to pass the resume token\n"); zstream_usage(); } resume_token = argv[1]; - libzfs_handle_t *hdl = libzfs_init(); + if ((hdl = libzfs_init()) == NULL) { + (void) fprintf(stderr, "%s\n", libzfs_error_init(errno)); + return (1); + } nvlist_t *resume_nvl = zfs_send_resume_token_to_nvlist(hdl, resume_token); if (resume_nvl == NULL) { (void) fprintf(stderr, "Unable to parse resume token: %s\n", libzfs_error_description(hdl)); libzfs_fini(hdl); return (1); } dump_nvlist(resume_nvl, 5); nvlist_free(resume_nvl); libzfs_fini(hdl); return (0); } diff --git a/cmd/ztest.c b/cmd/ztest.c index 033bcfd33576..e630d3353187 100644 --- a/cmd/ztest.c +++ b/cmd/ztest.c @@ -1,8289 +1,8295 @@ /* * CDDL HEADER START * * The contents of this file are subject to the terms of the * Common Development and Distribution License (the "License"). * 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 https://opensource.org/licenses/CDDL-1.0. * See the License for the specific language governing permissions * and limitations under the License. * * When distributing Covered Code, include this CDDL HEADER in each * file and include the License file at usr/src/OPENSOLARIS.LICENSE. * If applicable, add the following below this CDDL HEADER, with the * fields enclosed by brackets "[]" replaced with your own identifying * information: Portions Copyright [yyyy] [name of copyright owner] * * CDDL HEADER END */ /* * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved. * Copyright (c) 2011, 2018 by Delphix. All rights reserved. * Copyright 2011 Nexenta Systems, Inc. All rights reserved. * Copyright (c) 2013 Steven Hartland. All rights reserved. * Copyright (c) 2014 Integros [integros.com] * Copyright 2017 Joyent, Inc. * Copyright (c) 2017, Intel Corporation. */ /* * The objective of this program is to provide a DMU/ZAP/SPA stress test * that runs entirely in userland, is easy to use, and easy to extend. * * The overall design of the ztest program is as follows: * * (1) For each major functional area (e.g. adding vdevs to a pool, * creating and destroying datasets, reading and writing objects, etc) * we have a simple routine to test that functionality. These * individual routines do not have to do anything "stressful". * * (2) We turn these simple functionality tests into a stress test by * running them all in parallel, with as many threads as desired, * and spread across as many datasets, objects, and vdevs as desired. * * (3) While all this is happening, we inject faults into the pool to * verify that self-healing data really works. * * (4) Every time we open a dataset, we change its checksum and compression * functions. Thus even individual objects vary from block to block * in which checksum they use and whether they're compressed. * * (5) To verify that we never lose on-disk consistency after a crash, * we run the entire test in a child of the main process. * At random times, the child self-immolates with a SIGKILL. * This is the software equivalent of pulling the power cord. * The parent then runs the test again, using the existing * storage pool, as many times as desired. If backwards compatibility * testing is enabled ztest will sometimes run the "older" version * of ztest after a SIGKILL. * * (6) To verify that we don't have future leaks or temporal incursions, * many of the functional tests record the transaction group number * as part of their data. When reading old data, they verify that * the transaction group number is less than the current, open txg. * If you add a new test, please do this if applicable. * * (7) Threads are created with a reduced stack size, for sanity checking. * Therefore, it's important not to allocate huge buffers on the stack. * * When run with no arguments, ztest runs for about five minutes and * produces no output if successful. To get a little bit of information, * specify -V. To get more information, specify -VV, and so on. * * To turn this into an overnight stress test, use -T to specify run time. * * You can ask more vdevs [-v], datasets [-d], or threads [-t] * to increase the pool capacity, fanout, and overall stress level. * * Use the -k option to set the desired frequency of kills. * * When ztest invokes itself it passes all relevant information through a * temporary file which is mmap-ed in the child process. This allows shared * memory to survive the exec syscall. The ztest_shared_hdr_t struct is always * stored at offset 0 of this file and contains information on the size and * number of shared structures in the file. The information stored in this file * must remain backwards compatible with older versions of ztest so that * ztest can invoke them during backwards compatibility testing (-B). */ #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #if (__GLIBC__ && !__UCLIBC__) #include /* for backtrace() */ #endif static int ztest_fd_data = -1; static int ztest_fd_rand = -1; typedef struct ztest_shared_hdr { uint64_t zh_hdr_size; uint64_t zh_opts_size; uint64_t zh_size; uint64_t zh_stats_size; uint64_t zh_stats_count; uint64_t zh_ds_size; uint64_t zh_ds_count; } ztest_shared_hdr_t; static ztest_shared_hdr_t *ztest_shared_hdr; enum ztest_class_state { ZTEST_VDEV_CLASS_OFF, ZTEST_VDEV_CLASS_ON, ZTEST_VDEV_CLASS_RND }; #define ZO_GVARS_MAX_ARGLEN ((size_t)64) #define ZO_GVARS_MAX_COUNT ((size_t)10) typedef struct ztest_shared_opts { char zo_pool[ZFS_MAX_DATASET_NAME_LEN]; char zo_dir[ZFS_MAX_DATASET_NAME_LEN]; char zo_alt_ztest[MAXNAMELEN]; char zo_alt_libpath[MAXNAMELEN]; uint64_t zo_vdevs; uint64_t zo_vdevtime; size_t zo_vdev_size; int zo_ashift; int zo_mirrors; int zo_raid_children; int zo_raid_parity; char zo_raid_type[8]; int zo_draid_data; int zo_draid_spares; int zo_datasets; int zo_threads; uint64_t zo_passtime; uint64_t zo_killrate; int zo_verbose; int zo_init; uint64_t zo_time; uint64_t zo_maxloops; uint64_t zo_metaslab_force_ganging; int zo_mmp_test; int zo_special_vdevs; int zo_dump_dbgmsg; int zo_gvars_count; char zo_gvars[ZO_GVARS_MAX_COUNT][ZO_GVARS_MAX_ARGLEN]; } ztest_shared_opts_t; /* Default values for command line options. */ #define DEFAULT_POOL "ztest" #define DEFAULT_VDEV_DIR "/tmp" #define DEFAULT_VDEV_COUNT 5 #define DEFAULT_VDEV_SIZE (SPA_MINDEVSIZE * 4) /* 256m default size */ #define DEFAULT_VDEV_SIZE_STR "256M" #define DEFAULT_ASHIFT SPA_MINBLOCKSHIFT #define DEFAULT_MIRRORS 2 #define DEFAULT_RAID_CHILDREN 4 #define DEFAULT_RAID_PARITY 1 #define DEFAULT_DRAID_DATA 4 #define DEFAULT_DRAID_SPARES 1 #define DEFAULT_DATASETS_COUNT 7 #define DEFAULT_THREADS 23 #define DEFAULT_RUN_TIME 300 /* 300 seconds */ #define DEFAULT_RUN_TIME_STR "300 sec" #define DEFAULT_PASS_TIME 60 /* 60 seconds */ #define DEFAULT_PASS_TIME_STR "60 sec" #define DEFAULT_KILL_RATE 70 /* 70% kill rate */ #define DEFAULT_KILLRATE_STR "70%" #define DEFAULT_INITS 1 #define DEFAULT_MAX_LOOPS 50 /* 5 minutes */ #define DEFAULT_FORCE_GANGING (64 << 10) #define DEFAULT_FORCE_GANGING_STR "64K" /* Simplifying assumption: -1 is not a valid default. */ #define NO_DEFAULT -1 static const ztest_shared_opts_t ztest_opts_defaults = { .zo_pool = DEFAULT_POOL, .zo_dir = DEFAULT_VDEV_DIR, .zo_alt_ztest = { '\0' }, .zo_alt_libpath = { '\0' }, .zo_vdevs = DEFAULT_VDEV_COUNT, .zo_ashift = DEFAULT_ASHIFT, .zo_mirrors = DEFAULT_MIRRORS, .zo_raid_children = DEFAULT_RAID_CHILDREN, .zo_raid_parity = DEFAULT_RAID_PARITY, .zo_raid_type = VDEV_TYPE_RAIDZ, .zo_vdev_size = DEFAULT_VDEV_SIZE, .zo_draid_data = DEFAULT_DRAID_DATA, /* data drives */ .zo_draid_spares = DEFAULT_DRAID_SPARES, /* distributed spares */ .zo_datasets = DEFAULT_DATASETS_COUNT, .zo_threads = DEFAULT_THREADS, .zo_passtime = DEFAULT_PASS_TIME, .zo_killrate = DEFAULT_KILL_RATE, .zo_verbose = 0, .zo_mmp_test = 0, .zo_init = DEFAULT_INITS, .zo_time = DEFAULT_RUN_TIME, .zo_maxloops = DEFAULT_MAX_LOOPS, /* max loops during spa_freeze() */ .zo_metaslab_force_ganging = DEFAULT_FORCE_GANGING, .zo_special_vdevs = ZTEST_VDEV_CLASS_RND, .zo_gvars_count = 0, }; extern uint64_t metaslab_force_ganging; extern uint64_t metaslab_df_alloc_threshold; extern unsigned long zfs_deadman_synctime_ms; extern uint_t metaslab_preload_limit; extern int zfs_compressed_arc_enabled; extern int zfs_abd_scatter_enabled; extern uint_t dmu_object_alloc_chunk_shift; extern boolean_t zfs_force_some_double_word_sm_entries; extern unsigned long zio_decompress_fail_fraction; extern unsigned long zfs_reconstruct_indirect_damage_fraction; static ztest_shared_opts_t *ztest_shared_opts; static ztest_shared_opts_t ztest_opts; static const char *const ztest_wkeydata = "abcdefghijklmnopqrstuvwxyz012345"; typedef struct ztest_shared_ds { uint64_t zd_seq; } ztest_shared_ds_t; static ztest_shared_ds_t *ztest_shared_ds; #define ZTEST_GET_SHARED_DS(d) (&ztest_shared_ds[d]) #define BT_MAGIC 0x123456789abcdefULL #define MAXFAULTS(zs) \ (MAX((zs)->zs_mirrors, 1) * (ztest_opts.zo_raid_parity + 1) - 1) enum ztest_io_type { ZTEST_IO_WRITE_TAG, ZTEST_IO_WRITE_PATTERN, ZTEST_IO_WRITE_ZEROES, ZTEST_IO_TRUNCATE, ZTEST_IO_SETATTR, ZTEST_IO_REWRITE, ZTEST_IO_TYPES }; typedef struct ztest_block_tag { uint64_t bt_magic; uint64_t bt_objset; uint64_t bt_object; uint64_t bt_dnodesize; uint64_t bt_offset; uint64_t bt_gen; uint64_t bt_txg; uint64_t bt_crtxg; } ztest_block_tag_t; typedef struct bufwad { uint64_t bw_index; uint64_t bw_txg; uint64_t bw_data; } bufwad_t; /* * It would be better to use a rangelock_t per object. Unfortunately * the rangelock_t is not a drop-in replacement for rl_t, because we * still need to map from object ID to rangelock_t. */ typedef enum { RL_READER, RL_WRITER, RL_APPEND } rl_type_t; typedef struct rll { void *rll_writer; int rll_readers; kmutex_t rll_lock; kcondvar_t rll_cv; } rll_t; typedef struct rl { uint64_t rl_object; uint64_t rl_offset; uint64_t rl_size; rll_t *rl_lock; } rl_t; #define ZTEST_RANGE_LOCKS 64 #define ZTEST_OBJECT_LOCKS 64 /* * Object descriptor. Used as a template for object lookup/create/remove. */ typedef struct ztest_od { uint64_t od_dir; uint64_t od_object; dmu_object_type_t od_type; dmu_object_type_t od_crtype; uint64_t od_blocksize; uint64_t od_crblocksize; uint64_t od_crdnodesize; uint64_t od_gen; uint64_t od_crgen; char od_name[ZFS_MAX_DATASET_NAME_LEN]; } ztest_od_t; /* * Per-dataset state. */ typedef struct ztest_ds { ztest_shared_ds_t *zd_shared; objset_t *zd_os; pthread_rwlock_t zd_zilog_lock; zilog_t *zd_zilog; ztest_od_t *zd_od; /* debugging aid */ char zd_name[ZFS_MAX_DATASET_NAME_LEN]; kmutex_t zd_dirobj_lock; rll_t zd_object_lock[ZTEST_OBJECT_LOCKS]; rll_t zd_range_lock[ZTEST_RANGE_LOCKS]; } ztest_ds_t; /* * Per-iteration state. */ typedef void ztest_func_t(ztest_ds_t *zd, uint64_t id); typedef struct ztest_info { ztest_func_t *zi_func; /* test function */ uint64_t zi_iters; /* iterations per execution */ uint64_t *zi_interval; /* execute every seconds */ const char *zi_funcname; /* name of test function */ } ztest_info_t; typedef struct ztest_shared_callstate { uint64_t zc_count; /* per-pass count */ uint64_t zc_time; /* per-pass time */ uint64_t zc_next; /* next time to call this function */ } ztest_shared_callstate_t; static ztest_shared_callstate_t *ztest_shared_callstate; #define ZTEST_GET_SHARED_CALLSTATE(c) (&ztest_shared_callstate[c]) ztest_func_t ztest_dmu_read_write; ztest_func_t ztest_dmu_write_parallel; ztest_func_t ztest_dmu_object_alloc_free; ztest_func_t ztest_dmu_object_next_chunk; ztest_func_t ztest_dmu_commit_callbacks; ztest_func_t ztest_zap; ztest_func_t ztest_zap_parallel; ztest_func_t ztest_zil_commit; ztest_func_t ztest_zil_remount; ztest_func_t ztest_dmu_read_write_zcopy; ztest_func_t ztest_dmu_objset_create_destroy; ztest_func_t ztest_dmu_prealloc; ztest_func_t ztest_fzap; ztest_func_t ztest_dmu_snapshot_create_destroy; ztest_func_t ztest_dsl_prop_get_set; ztest_func_t ztest_spa_prop_get_set; ztest_func_t ztest_spa_create_destroy; ztest_func_t ztest_fault_inject; ztest_func_t ztest_dmu_snapshot_hold; ztest_func_t ztest_mmp_enable_disable; ztest_func_t ztest_scrub; ztest_func_t ztest_dsl_dataset_promote_busy; ztest_func_t ztest_vdev_attach_detach; ztest_func_t ztest_vdev_LUN_growth; ztest_func_t ztest_vdev_add_remove; ztest_func_t ztest_vdev_class_add; ztest_func_t ztest_vdev_aux_add_remove; ztest_func_t ztest_split_pool; ztest_func_t ztest_reguid; ztest_func_t ztest_spa_upgrade; ztest_func_t ztest_device_removal; ztest_func_t ztest_spa_checkpoint_create_discard; ztest_func_t ztest_initialize; ztest_func_t ztest_trim; ztest_func_t ztest_blake3; ztest_func_t ztest_fletcher; ztest_func_t ztest_fletcher_incr; ztest_func_t ztest_verify_dnode_bt; uint64_t zopt_always = 0ULL * NANOSEC; /* all the time */ uint64_t zopt_incessant = 1ULL * NANOSEC / 10; /* every 1/10 second */ uint64_t zopt_often = 1ULL * NANOSEC; /* every second */ uint64_t zopt_sometimes = 10ULL * NANOSEC; /* every 10 seconds */ uint64_t zopt_rarely = 60ULL * NANOSEC; /* every 60 seconds */ #define ZTI_INIT(func, iters, interval) \ { .zi_func = (func), \ .zi_iters = (iters), \ .zi_interval = (interval), \ .zi_funcname = # func } ztest_info_t ztest_info[] = { ZTI_INIT(ztest_dmu_read_write, 1, &zopt_always), ZTI_INIT(ztest_dmu_write_parallel, 10, &zopt_always), ZTI_INIT(ztest_dmu_object_alloc_free, 1, &zopt_always), ZTI_INIT(ztest_dmu_object_next_chunk, 1, &zopt_sometimes), ZTI_INIT(ztest_dmu_commit_callbacks, 1, &zopt_always), ZTI_INIT(ztest_zap, 30, &zopt_always), ZTI_INIT(ztest_zap_parallel, 100, &zopt_always), ZTI_INIT(ztest_split_pool, 1, &zopt_always), ZTI_INIT(ztest_zil_commit, 1, &zopt_incessant), ZTI_INIT(ztest_zil_remount, 1, &zopt_sometimes), ZTI_INIT(ztest_dmu_read_write_zcopy, 1, &zopt_often), ZTI_INIT(ztest_dmu_objset_create_destroy, 1, &zopt_often), ZTI_INIT(ztest_dsl_prop_get_set, 1, &zopt_often), ZTI_INIT(ztest_spa_prop_get_set, 1, &zopt_sometimes), #if 0 ZTI_INIT(ztest_dmu_prealloc, 1, &zopt_sometimes), #endif ZTI_INIT(ztest_fzap, 1, &zopt_sometimes), ZTI_INIT(ztest_dmu_snapshot_create_destroy, 1, &zopt_sometimes), ZTI_INIT(ztest_spa_create_destroy, 1, &zopt_sometimes), ZTI_INIT(ztest_fault_inject, 1, &zopt_sometimes), ZTI_INIT(ztest_dmu_snapshot_hold, 1, &zopt_sometimes), ZTI_INIT(ztest_mmp_enable_disable, 1, &zopt_sometimes), ZTI_INIT(ztest_reguid, 1, &zopt_rarely), ZTI_INIT(ztest_scrub, 1, &zopt_rarely), ZTI_INIT(ztest_spa_upgrade, 1, &zopt_rarely), ZTI_INIT(ztest_dsl_dataset_promote_busy, 1, &zopt_rarely), ZTI_INIT(ztest_vdev_attach_detach, 1, &zopt_sometimes), ZTI_INIT(ztest_vdev_LUN_growth, 1, &zopt_rarely), ZTI_INIT(ztest_vdev_add_remove, 1, &ztest_opts.zo_vdevtime), ZTI_INIT(ztest_vdev_class_add, 1, &ztest_opts.zo_vdevtime), ZTI_INIT(ztest_vdev_aux_add_remove, 1, &ztest_opts.zo_vdevtime), ZTI_INIT(ztest_device_removal, 1, &zopt_sometimes), ZTI_INIT(ztest_spa_checkpoint_create_discard, 1, &zopt_rarely), ZTI_INIT(ztest_initialize, 1, &zopt_sometimes), ZTI_INIT(ztest_trim, 1, &zopt_sometimes), ZTI_INIT(ztest_blake3, 1, &zopt_rarely), ZTI_INIT(ztest_fletcher, 1, &zopt_rarely), ZTI_INIT(ztest_fletcher_incr, 1, &zopt_rarely), ZTI_INIT(ztest_verify_dnode_bt, 1, &zopt_sometimes), }; #define ZTEST_FUNCS (sizeof (ztest_info) / sizeof (ztest_info_t)) /* * The following struct is used to hold a list of uncalled commit callbacks. * The callbacks are ordered by txg number. */ typedef struct ztest_cb_list { kmutex_t zcl_callbacks_lock; list_t zcl_callbacks; } ztest_cb_list_t; /* * Stuff we need to share writably between parent and child. */ typedef struct ztest_shared { boolean_t zs_do_init; hrtime_t zs_proc_start; hrtime_t zs_proc_stop; hrtime_t zs_thread_start; hrtime_t zs_thread_stop; hrtime_t zs_thread_kill; uint64_t zs_enospc_count; uint64_t zs_vdev_next_leaf; uint64_t zs_vdev_aux; uint64_t zs_alloc; uint64_t zs_space; uint64_t zs_splits; uint64_t zs_mirrors; uint64_t zs_metaslab_sz; uint64_t zs_metaslab_df_alloc_threshold; uint64_t zs_guid; } ztest_shared_t; #define ID_PARALLEL -1ULL static char ztest_dev_template[] = "%s/%s.%llua"; static char ztest_aux_template[] = "%s/%s.%s.%llu"; ztest_shared_t *ztest_shared; static spa_t *ztest_spa = NULL; static ztest_ds_t *ztest_ds; static kmutex_t ztest_vdev_lock; static boolean_t ztest_device_removal_active = B_FALSE; static boolean_t ztest_pool_scrubbed = B_FALSE; static kmutex_t ztest_checkpoint_lock; /* * The ztest_name_lock protects the pool and dataset namespace used by * the individual tests. To modify the namespace, consumers must grab * this lock as writer. Grabbing the lock as reader will ensure that the * namespace does not change while the lock is held. */ static pthread_rwlock_t ztest_name_lock; static boolean_t ztest_dump_core = B_TRUE; static boolean_t ztest_exiting; /* Global commit callback list */ static ztest_cb_list_t zcl; /* Commit cb delay */ static uint64_t zc_min_txg_delay = UINT64_MAX; static int zc_cb_counter = 0; /* * Minimum number of commit callbacks that need to be registered for us to check * whether the minimum txg delay is acceptable. */ #define ZTEST_COMMIT_CB_MIN_REG 100 /* * If a number of txgs equal to this threshold have been created after a commit * callback has been registered but not called, then we assume there is an * implementation bug. */ #define ZTEST_COMMIT_CB_THRESH (TXG_CONCURRENT_STATES + 1000) enum ztest_object { ZTEST_META_DNODE = 0, ZTEST_DIROBJ, ZTEST_OBJECTS }; static __attribute__((noreturn)) void usage(boolean_t requested); static int ztest_scrub_impl(spa_t *spa); /* * These libumem hooks provide a reasonable set of defaults for the allocator's * debugging facilities. */ const char * _umem_debug_init(void) { return ("default,verbose"); /* $UMEM_DEBUG setting */ } const char * _umem_logging_init(void) { return ("fail,contents"); /* $UMEM_LOGGING setting */ } static void dump_debug_buffer(void) { ssize_t ret __attribute__((unused)); if (!ztest_opts.zo_dump_dbgmsg) return; /* * We use write() instead of printf() so that this function * is safe to call from a signal handler. */ ret = write(STDOUT_FILENO, "\n", 1); zfs_dbgmsg_print("ztest"); } #define BACKTRACE_SZ 100 static void sig_handler(int signo) { struct sigaction action; #if (__GLIBC__ && !__UCLIBC__) /* backtrace() is a GNU extension */ int nptrs; void *buffer[BACKTRACE_SZ]; nptrs = backtrace(buffer, BACKTRACE_SZ); backtrace_symbols_fd(buffer, nptrs, STDERR_FILENO); #endif dump_debug_buffer(); /* * Restore default action and re-raise signal so SIGSEGV and * SIGABRT can trigger a core dump. */ action.sa_handler = SIG_DFL; sigemptyset(&action.sa_mask); action.sa_flags = 0; (void) sigaction(signo, &action, NULL); raise(signo); } #define FATAL_MSG_SZ 1024 static const char *fatal_msg; static __attribute__((format(printf, 2, 3))) __attribute__((noreturn)) void fatal(int do_perror, const char *message, ...) { va_list args; int save_errno = errno; char *buf; (void) fflush(stdout); buf = umem_alloc(FATAL_MSG_SZ, UMEM_NOFAIL); if (buf == NULL) goto out; va_start(args, message); (void) sprintf(buf, "ztest: "); /* LINTED */ (void) vsprintf(buf + strlen(buf), message, args); va_end(args); if (do_perror) { (void) snprintf(buf + strlen(buf), FATAL_MSG_SZ - strlen(buf), ": %s", strerror(save_errno)); } (void) fprintf(stderr, "%s\n", buf); fatal_msg = buf; /* to ease debugging */ out: if (ztest_dump_core) abort(); else dump_debug_buffer(); exit(3); } static int str2shift(const char *buf) { const char *ends = "BKMGTPEZ"; int i; if (buf[0] == '\0') return (0); for (i = 0; i < strlen(ends); i++) { if (toupper(buf[0]) == ends[i]) break; } if (i == strlen(ends)) { (void) fprintf(stderr, "ztest: invalid bytes suffix: %s\n", buf); usage(B_FALSE); } if (buf[1] == '\0' || (toupper(buf[1]) == 'B' && buf[2] == '\0')) { return (10*i); } (void) fprintf(stderr, "ztest: invalid bytes suffix: %s\n", buf); usage(B_FALSE); } static uint64_t nicenumtoull(const char *buf) { char *end; uint64_t val; val = strtoull(buf, &end, 0); if (end == buf) { (void) fprintf(stderr, "ztest: bad numeric value: %s\n", buf); usage(B_FALSE); } else if (end[0] == '.') { double fval = strtod(buf, &end); fval *= pow(2, str2shift(end)); /* * UINT64_MAX is not exactly representable as a double. * The closest representation is UINT64_MAX + 1, so we * use a >= comparison instead of > for the bounds check. */ if (fval >= (double)UINT64_MAX) { (void) fprintf(stderr, "ztest: value too large: %s\n", buf); usage(B_FALSE); } val = (uint64_t)fval; } else { int shift = str2shift(end); if (shift >= 64 || (val << shift) >> shift != val) { (void) fprintf(stderr, "ztest: value too large: %s\n", buf); usage(B_FALSE); } val <<= shift; } return (val); } typedef struct ztest_option { const char short_opt; const char *long_opt; const char *long_opt_param; const char *comment; unsigned int default_int; const char *default_str; } ztest_option_t; /* * The following option_table is used for generating the usage info as well as * the long and short option information for calling getopt_long(). */ static ztest_option_t option_table[] = { { 'v', "vdevs", "INTEGER", "Number of vdevs", DEFAULT_VDEV_COUNT, NULL}, { 's', "vdev-size", "INTEGER", "Size of each vdev", NO_DEFAULT, DEFAULT_VDEV_SIZE_STR}, { 'a', "alignment-shift", "INTEGER", "Alignment shift; use 0 for random", DEFAULT_ASHIFT, NULL}, { 'm', "mirror-copies", "INTEGER", "Number of mirror copies", DEFAULT_MIRRORS, NULL}, { 'r', "raid-disks", "INTEGER", "Number of raidz/draid disks", DEFAULT_RAID_CHILDREN, NULL}, { 'R', "raid-parity", "INTEGER", "Raid parity", DEFAULT_RAID_PARITY, NULL}, { 'K', "raid-kind", "raidz|draid|random", "Raid kind", NO_DEFAULT, "random"}, { 'D', "draid-data", "INTEGER", "Number of draid data drives", DEFAULT_DRAID_DATA, NULL}, { 'S', "draid-spares", "INTEGER", "Number of draid spares", DEFAULT_DRAID_SPARES, NULL}, { 'd', "datasets", "INTEGER", "Number of datasets", DEFAULT_DATASETS_COUNT, NULL}, { 't', "threads", "INTEGER", "Number of ztest threads", DEFAULT_THREADS, NULL}, { 'g', "gang-block-threshold", "INTEGER", "Metaslab gang block threshold", NO_DEFAULT, DEFAULT_FORCE_GANGING_STR}, { 'i', "init-count", "INTEGER", "Number of times to initialize pool", DEFAULT_INITS, NULL}, { 'k', "kill-percentage", "INTEGER", "Kill percentage", NO_DEFAULT, DEFAULT_KILLRATE_STR}, { 'p', "pool-name", "STRING", "Pool name", NO_DEFAULT, DEFAULT_POOL}, { 'f', "vdev-file-directory", "PATH", "File directory for vdev files", NO_DEFAULT, DEFAULT_VDEV_DIR}, { 'M', "multi-host", NULL, "Multi-host; simulate pool imported on remote host", NO_DEFAULT, NULL}, { 'E', "use-existing-pool", NULL, "Use existing pool instead of creating new one", NO_DEFAULT, NULL}, { 'T', "run-time", "INTEGER", "Total run time", NO_DEFAULT, DEFAULT_RUN_TIME_STR}, { 'P', "pass-time", "INTEGER", "Time per pass", NO_DEFAULT, DEFAULT_PASS_TIME_STR}, { 'F', "freeze-loops", "INTEGER", "Max loops in spa_freeze()", DEFAULT_MAX_LOOPS, NULL}, { 'B', "alt-ztest", "PATH", "Alternate ztest path", NO_DEFAULT, NULL}, { 'C', "vdev-class-state", "on|off|random", "vdev class state", NO_DEFAULT, "random"}, { 'o', "option", "\"OPTION=INTEGER\"", "Set global variable to an unsigned 32-bit integer value", NO_DEFAULT, NULL}, { 'G', "dump-debug-msg", NULL, "Dump zfs_dbgmsg buffer before exiting due to an error", NO_DEFAULT, NULL}, { 'V', "verbose", NULL, "Verbose (use multiple times for ever more verbosity)", NO_DEFAULT, NULL}, { 'h', "help", NULL, "Show this help", NO_DEFAULT, NULL}, {0, 0, 0, 0, 0, 0} }; static struct option *long_opts = NULL; static char *short_opts = NULL; static void init_options(void) { ASSERT3P(long_opts, ==, NULL); ASSERT3P(short_opts, ==, NULL); int count = sizeof (option_table) / sizeof (option_table[0]); long_opts = umem_alloc(sizeof (struct option) * count, UMEM_NOFAIL); short_opts = umem_alloc(sizeof (char) * 2 * count, UMEM_NOFAIL); int short_opt_index = 0; for (int i = 0; i < count; i++) { long_opts[i].val = option_table[i].short_opt; long_opts[i].name = option_table[i].long_opt; long_opts[i].has_arg = option_table[i].long_opt_param != NULL ? required_argument : no_argument; long_opts[i].flag = NULL; short_opts[short_opt_index++] = option_table[i].short_opt; if (option_table[i].long_opt_param != NULL) { short_opts[short_opt_index++] = ':'; } } } static void fini_options(void) { int count = sizeof (option_table) / sizeof (option_table[0]); umem_free(long_opts, sizeof (struct option) * count); umem_free(short_opts, sizeof (char) * 2 * count); long_opts = NULL; short_opts = NULL; } static __attribute__((noreturn)) void usage(boolean_t requested) { char option[80]; FILE *fp = requested ? stdout : stderr; (void) fprintf(fp, "Usage: %s [OPTIONS...]\n", DEFAULT_POOL); for (int i = 0; option_table[i].short_opt != 0; i++) { if (option_table[i].long_opt_param != NULL) { (void) sprintf(option, " -%c --%s=%s", option_table[i].short_opt, option_table[i].long_opt, option_table[i].long_opt_param); } else { (void) sprintf(option, " -%c --%s", option_table[i].short_opt, option_table[i].long_opt); } (void) fprintf(fp, " %-40s%s", option, option_table[i].comment); if (option_table[i].long_opt_param != NULL) { if (option_table[i].default_str != NULL) { (void) fprintf(fp, " (default: %s)", option_table[i].default_str); } else if (option_table[i].default_int != NO_DEFAULT) { (void) fprintf(fp, " (default: %u)", option_table[i].default_int); } } (void) fprintf(fp, "\n"); } exit(requested ? 0 : 1); } static uint64_t ztest_random(uint64_t range) { uint64_t r; ASSERT3S(ztest_fd_rand, >=, 0); if (range == 0) return (0); if (read(ztest_fd_rand, &r, sizeof (r)) != sizeof (r)) fatal(B_TRUE, "short read from /dev/urandom"); return (r % range); } static void ztest_parse_name_value(const char *input, ztest_shared_opts_t *zo) { char name[32]; char *value; int state = ZTEST_VDEV_CLASS_RND; (void) strlcpy(name, input, sizeof (name)); value = strchr(name, '='); if (value == NULL) { (void) fprintf(stderr, "missing value in property=value " "'-C' argument (%s)\n", input); usage(B_FALSE); } *(value) = '\0'; value++; if (strcmp(value, "on") == 0) { state = ZTEST_VDEV_CLASS_ON; } else if (strcmp(value, "off") == 0) { state = ZTEST_VDEV_CLASS_OFF; } else if (strcmp(value, "random") == 0) { state = ZTEST_VDEV_CLASS_RND; } else { (void) fprintf(stderr, "invalid property value '%s'\n", value); usage(B_FALSE); } if (strcmp(name, "special") == 0) { zo->zo_special_vdevs = state; } else { (void) fprintf(stderr, "invalid property name '%s'\n", name); usage(B_FALSE); } if (zo->zo_verbose >= 3) (void) printf("%s vdev state is '%s'\n", name, value); } static void process_options(int argc, char **argv) { char *path; ztest_shared_opts_t *zo = &ztest_opts; int opt; uint64_t value; const char *raid_kind = "random"; memcpy(zo, &ztest_opts_defaults, sizeof (*zo)); init_options(); while ((opt = getopt_long(argc, argv, short_opts, long_opts, NULL)) != EOF) { value = 0; switch (opt) { case 'v': case 's': case 'a': case 'm': case 'r': case 'R': case 'D': case 'S': case 'd': case 't': case 'g': case 'i': case 'k': case 'T': case 'P': case 'F': value = nicenumtoull(optarg); } switch (opt) { case 'v': zo->zo_vdevs = value; break; case 's': zo->zo_vdev_size = MAX(SPA_MINDEVSIZE, value); break; case 'a': zo->zo_ashift = value; break; case 'm': zo->zo_mirrors = value; break; case 'r': zo->zo_raid_children = MAX(1, value); break; case 'R': zo->zo_raid_parity = MIN(MAX(value, 1), 3); break; case 'K': raid_kind = optarg; break; case 'D': zo->zo_draid_data = MAX(1, value); break; case 'S': zo->zo_draid_spares = MAX(1, value); break; case 'd': zo->zo_datasets = MAX(1, value); break; case 't': zo->zo_threads = MAX(1, value); break; case 'g': zo->zo_metaslab_force_ganging = MAX(SPA_MINBLOCKSIZE << 1, value); break; case 'i': zo->zo_init = value; break; case 'k': zo->zo_killrate = value; break; case 'p': (void) strlcpy(zo->zo_pool, optarg, sizeof (zo->zo_pool)); break; case 'f': path = realpath(optarg, NULL); if (path == NULL) { (void) fprintf(stderr, "error: %s: %s\n", optarg, strerror(errno)); usage(B_FALSE); } else { (void) strlcpy(zo->zo_dir, path, sizeof (zo->zo_dir)); free(path); } break; case 'M': zo->zo_mmp_test = 1; break; case 'V': zo->zo_verbose++; break; case 'E': zo->zo_init = 0; break; case 'T': zo->zo_time = value; break; case 'P': zo->zo_passtime = MAX(1, value); break; case 'F': zo->zo_maxloops = MAX(1, value); break; case 'B': (void) strlcpy(zo->zo_alt_ztest, optarg, sizeof (zo->zo_alt_ztest)); break; case 'C': ztest_parse_name_value(optarg, zo); break; case 'o': if (zo->zo_gvars_count >= ZO_GVARS_MAX_COUNT) { (void) fprintf(stderr, "max global var count (%zu) exceeded\n", ZO_GVARS_MAX_COUNT); usage(B_FALSE); } char *v = zo->zo_gvars[zo->zo_gvars_count]; if (strlcpy(v, optarg, ZO_GVARS_MAX_ARGLEN) >= ZO_GVARS_MAX_ARGLEN) { (void) fprintf(stderr, "global var option '%s' is too long\n", optarg); usage(B_FALSE); } zo->zo_gvars_count++; break; case 'G': zo->zo_dump_dbgmsg = 1; break; case 'h': usage(B_TRUE); break; case '?': default: usage(B_FALSE); break; } } fini_options(); /* When raid choice is 'random' add a draid pool 50% of the time */ if (strcmp(raid_kind, "random") == 0) { raid_kind = (ztest_random(2) == 0) ? "draid" : "raidz"; if (ztest_opts.zo_verbose >= 3) (void) printf("choosing RAID type '%s'\n", raid_kind); } if (strcmp(raid_kind, "draid") == 0) { uint64_t min_devsize; /* With fewer disk use 256M, otherwise 128M is OK */ min_devsize = (ztest_opts.zo_raid_children < 16) ? (256ULL << 20) : (128ULL << 20); /* No top-level mirrors with dRAID for now */ zo->zo_mirrors = 0; /* Use more appropriate defaults for dRAID */ if (zo->zo_vdevs == ztest_opts_defaults.zo_vdevs) zo->zo_vdevs = 1; if (zo->zo_raid_children == ztest_opts_defaults.zo_raid_children) zo->zo_raid_children = 16; if (zo->zo_ashift < 12) zo->zo_ashift = 12; if (zo->zo_vdev_size < min_devsize) zo->zo_vdev_size = min_devsize; if (zo->zo_draid_data + zo->zo_raid_parity > zo->zo_raid_children - zo->zo_draid_spares) { (void) fprintf(stderr, "error: too few draid " "children (%d) for stripe width (%d)\n", zo->zo_raid_children, zo->zo_draid_data + zo->zo_raid_parity); usage(B_FALSE); } (void) strlcpy(zo->zo_raid_type, VDEV_TYPE_DRAID, sizeof (zo->zo_raid_type)); } else /* using raidz */ { ASSERT0(strcmp(raid_kind, "raidz")); zo->zo_raid_parity = MIN(zo->zo_raid_parity, zo->zo_raid_children - 1); } zo->zo_vdevtime = (zo->zo_vdevs > 0 ? zo->zo_time * NANOSEC / zo->zo_vdevs : UINT64_MAX >> 2); if (*zo->zo_alt_ztest) { const char *invalid_what = "ztest"; char *val = zo->zo_alt_ztest; if (0 != access(val, X_OK) || (strrchr(val, '/') == NULL && (errno = EINVAL))) goto invalid; int dirlen = strrchr(val, '/') - val; strlcpy(zo->zo_alt_libpath, val, MIN(sizeof (zo->zo_alt_libpath), dirlen + 1)); invalid_what = "library path", val = zo->zo_alt_libpath; if (strrchr(val, '/') == NULL && (errno = EINVAL)) goto invalid; *strrchr(val, '/') = '\0'; strlcat(val, "/lib", sizeof (zo->zo_alt_libpath)); if (0 != access(zo->zo_alt_libpath, X_OK)) goto invalid; return; invalid: ztest_dump_core = B_FALSE; fatal(B_TRUE, "invalid alternate %s %s", invalid_what, val); } } static void ztest_kill(ztest_shared_t *zs) { zs->zs_alloc = metaslab_class_get_alloc(spa_normal_class(ztest_spa)); zs->zs_space = metaslab_class_get_space(spa_normal_class(ztest_spa)); /* * Before we kill ourselves, make sure that the config is updated. * See comment above spa_write_cachefile(). */ mutex_enter(&spa_namespace_lock); spa_write_cachefile(ztest_spa, B_FALSE, B_FALSE, B_FALSE); mutex_exit(&spa_namespace_lock); (void) raise(SIGKILL); } static void ztest_record_enospc(const char *s) { (void) s; ztest_shared->zs_enospc_count++; } static uint64_t ztest_get_ashift(void) { if (ztest_opts.zo_ashift == 0) return (SPA_MINBLOCKSHIFT + ztest_random(5)); return (ztest_opts.zo_ashift); } static boolean_t ztest_is_draid_spare(const char *name) { uint64_t spare_id = 0, parity = 0, vdev_id = 0; if (sscanf(name, VDEV_TYPE_DRAID "%"PRIu64"-%"PRIu64"-%"PRIu64"", &parity, &vdev_id, &spare_id) == 3) { return (B_TRUE); } return (B_FALSE); } static nvlist_t * make_vdev_file(const char *path, const char *aux, const char *pool, size_t size, uint64_t ashift) { char *pathbuf = NULL; uint64_t vdev; nvlist_t *file; boolean_t draid_spare = B_FALSE; if (ashift == 0) ashift = ztest_get_ashift(); if (path == NULL) { pathbuf = umem_alloc(MAXPATHLEN, UMEM_NOFAIL); path = pathbuf; if (aux != NULL) { vdev = ztest_shared->zs_vdev_aux; (void) snprintf(pathbuf, MAXPATHLEN, ztest_aux_template, ztest_opts.zo_dir, pool == NULL ? ztest_opts.zo_pool : pool, aux, vdev); } else { vdev = ztest_shared->zs_vdev_next_leaf++; (void) snprintf(pathbuf, MAXPATHLEN, ztest_dev_template, ztest_opts.zo_dir, pool == NULL ? ztest_opts.zo_pool : pool, vdev); } } else { draid_spare = ztest_is_draid_spare(path); } if (size != 0 && !draid_spare) { int fd = open(path, O_RDWR | O_CREAT | O_TRUNC, 0666); if (fd == -1) fatal(B_TRUE, "can't open %s", path); if (ftruncate(fd, size) != 0) fatal(B_TRUE, "can't ftruncate %s", path); (void) close(fd); } file = fnvlist_alloc(); fnvlist_add_string(file, ZPOOL_CONFIG_TYPE, draid_spare ? VDEV_TYPE_DRAID_SPARE : VDEV_TYPE_FILE); fnvlist_add_string(file, ZPOOL_CONFIG_PATH, path); fnvlist_add_uint64(file, ZPOOL_CONFIG_ASHIFT, ashift); umem_free(pathbuf, MAXPATHLEN); return (file); } static nvlist_t * make_vdev_raid(const char *path, const char *aux, const char *pool, size_t size, uint64_t ashift, int r) { nvlist_t *raid, **child; int c; if (r < 2) return (make_vdev_file(path, aux, pool, size, ashift)); child = umem_alloc(r * sizeof (nvlist_t *), UMEM_NOFAIL); for (c = 0; c < r; c++) child[c] = make_vdev_file(path, aux, pool, size, ashift); raid = fnvlist_alloc(); fnvlist_add_string(raid, ZPOOL_CONFIG_TYPE, ztest_opts.zo_raid_type); fnvlist_add_uint64(raid, ZPOOL_CONFIG_NPARITY, ztest_opts.zo_raid_parity); fnvlist_add_nvlist_array(raid, ZPOOL_CONFIG_CHILDREN, (const nvlist_t **)child, r); if (strcmp(ztest_opts.zo_raid_type, VDEV_TYPE_DRAID) == 0) { uint64_t ndata = ztest_opts.zo_draid_data; uint64_t nparity = ztest_opts.zo_raid_parity; uint64_t nspares = ztest_opts.zo_draid_spares; uint64_t children = ztest_opts.zo_raid_children; uint64_t ngroups = 1; /* * Calculate the minimum number of groups required to fill a * slice. This is the LCM of the stripe width (data + parity) * and the number of data drives (children - spares). */ while (ngroups * (ndata + nparity) % (children - nspares) != 0) ngroups++; /* Store the basic dRAID configuration. */ fnvlist_add_uint64(raid, ZPOOL_CONFIG_DRAID_NDATA, ndata); fnvlist_add_uint64(raid, ZPOOL_CONFIG_DRAID_NSPARES, nspares); fnvlist_add_uint64(raid, ZPOOL_CONFIG_DRAID_NGROUPS, ngroups); } for (c = 0; c < r; c++) fnvlist_free(child[c]); umem_free(child, r * sizeof (nvlist_t *)); return (raid); } static nvlist_t * make_vdev_mirror(const char *path, const char *aux, const char *pool, size_t size, uint64_t ashift, int r, int m) { nvlist_t *mirror, **child; int c; if (m < 1) return (make_vdev_raid(path, aux, pool, size, ashift, r)); child = umem_alloc(m * sizeof (nvlist_t *), UMEM_NOFAIL); for (c = 0; c < m; c++) child[c] = make_vdev_raid(path, aux, pool, size, ashift, r); mirror = fnvlist_alloc(); fnvlist_add_string(mirror, ZPOOL_CONFIG_TYPE, VDEV_TYPE_MIRROR); fnvlist_add_nvlist_array(mirror, ZPOOL_CONFIG_CHILDREN, (const nvlist_t **)child, m); for (c = 0; c < m; c++) fnvlist_free(child[c]); umem_free(child, m * sizeof (nvlist_t *)); return (mirror); } static nvlist_t * make_vdev_root(const char *path, const char *aux, const char *pool, size_t size, uint64_t ashift, const char *class, int r, int m, int t) { nvlist_t *root, **child; int c; boolean_t log; ASSERT3S(t, >, 0); log = (class != NULL && strcmp(class, "log") == 0); child = umem_alloc(t * sizeof (nvlist_t *), UMEM_NOFAIL); for (c = 0; c < t; c++) { child[c] = make_vdev_mirror(path, aux, pool, size, ashift, r, m); fnvlist_add_uint64(child[c], ZPOOL_CONFIG_IS_LOG, log); if (class != NULL && class[0] != '\0') { ASSERT(m > 1 || log); /* expecting a mirror */ fnvlist_add_string(child[c], ZPOOL_CONFIG_ALLOCATION_BIAS, class); } } root = fnvlist_alloc(); fnvlist_add_string(root, ZPOOL_CONFIG_TYPE, VDEV_TYPE_ROOT); fnvlist_add_nvlist_array(root, aux ? aux : ZPOOL_CONFIG_CHILDREN, (const nvlist_t **)child, t); for (c = 0; c < t; c++) fnvlist_free(child[c]); umem_free(child, t * sizeof (nvlist_t *)); return (root); } /* * Find a random spa version. Returns back a random spa version in the * range [initial_version, SPA_VERSION_FEATURES]. */ static uint64_t ztest_random_spa_version(uint64_t initial_version) { uint64_t version = initial_version; if (version <= SPA_VERSION_BEFORE_FEATURES) { version = version + ztest_random(SPA_VERSION_BEFORE_FEATURES - version + 1); } if (version > SPA_VERSION_BEFORE_FEATURES) version = SPA_VERSION_FEATURES; ASSERT(SPA_VERSION_IS_SUPPORTED(version)); return (version); } static int ztest_random_blocksize(void) { ASSERT3U(ztest_spa->spa_max_ashift, !=, 0); /* * Choose a block size >= the ashift. * If the SPA supports new MAXBLOCKSIZE, test up to 1MB blocks. */ int maxbs = SPA_OLD_MAXBLOCKSHIFT; if (spa_maxblocksize(ztest_spa) == SPA_MAXBLOCKSIZE) maxbs = 20; uint64_t block_shift = ztest_random(maxbs - ztest_spa->spa_max_ashift + 1); return (1 << (SPA_MINBLOCKSHIFT + block_shift)); } static int ztest_random_dnodesize(void) { int slots; int max_slots = spa_maxdnodesize(ztest_spa) >> DNODE_SHIFT; if (max_slots == DNODE_MIN_SLOTS) return (DNODE_MIN_SIZE); /* * Weight the random distribution more heavily toward smaller * dnode sizes since that is more likely to reflect real-world * usage. */ ASSERT3U(max_slots, >, 4); switch (ztest_random(10)) { case 0: slots = 5 + ztest_random(max_slots - 4); break; case 1 ... 4: slots = 2 + ztest_random(3); break; default: slots = 1; break; } return (slots << DNODE_SHIFT); } static int ztest_random_ibshift(void) { return (DN_MIN_INDBLKSHIFT + ztest_random(DN_MAX_INDBLKSHIFT - DN_MIN_INDBLKSHIFT + 1)); } static uint64_t ztest_random_vdev_top(spa_t *spa, boolean_t log_ok) { uint64_t top; vdev_t *rvd = spa->spa_root_vdev; vdev_t *tvd; ASSERT3U(spa_config_held(spa, SCL_ALL, RW_READER), !=, 0); do { top = ztest_random(rvd->vdev_children); tvd = rvd->vdev_child[top]; } while (!vdev_is_concrete(tvd) || (tvd->vdev_islog && !log_ok) || tvd->vdev_mg == NULL || tvd->vdev_mg->mg_class == NULL); return (top); } static uint64_t ztest_random_dsl_prop(zfs_prop_t prop) { uint64_t value; do { value = zfs_prop_random_value(prop, ztest_random(-1ULL)); } while (prop == ZFS_PROP_CHECKSUM && value == ZIO_CHECKSUM_OFF); return (value); } static int ztest_dsl_prop_set_uint64(char *osname, zfs_prop_t prop, uint64_t value, boolean_t inherit) { const char *propname = zfs_prop_to_name(prop); const char *valname; char *setpoint; uint64_t curval; int error; error = dsl_prop_set_int(osname, propname, (inherit ? ZPROP_SRC_NONE : ZPROP_SRC_LOCAL), value); if (error == ENOSPC) { ztest_record_enospc(FTAG); return (error); } ASSERT0(error); setpoint = umem_alloc(MAXPATHLEN, UMEM_NOFAIL); VERIFY0(dsl_prop_get_integer(osname, propname, &curval, setpoint)); if (ztest_opts.zo_verbose >= 6) { int err; err = zfs_prop_index_to_string(prop, curval, &valname); if (err) (void) printf("%s %s = %llu at '%s'\n", osname, propname, (unsigned long long)curval, setpoint); else (void) printf("%s %s = %s at '%s'\n", osname, propname, valname, setpoint); } umem_free(setpoint, MAXPATHLEN); return (error); } static int ztest_spa_prop_set_uint64(zpool_prop_t prop, uint64_t value) { spa_t *spa = ztest_spa; nvlist_t *props = NULL; int error; props = fnvlist_alloc(); fnvlist_add_uint64(props, zpool_prop_to_name(prop), value); error = spa_prop_set(spa, props); fnvlist_free(props); if (error == ENOSPC) { ztest_record_enospc(FTAG); return (error); } ASSERT0(error); return (error); } static int ztest_dmu_objset_own(const char *name, dmu_objset_type_t type, boolean_t readonly, boolean_t decrypt, const void *tag, objset_t **osp) { int err; char *cp = NULL; char ddname[ZFS_MAX_DATASET_NAME_LEN]; strlcpy(ddname, name, sizeof (ddname)); cp = strchr(ddname, '@'); if (cp != NULL) *cp = '\0'; err = dmu_objset_own(name, type, readonly, decrypt, tag, osp); while (decrypt && err == EACCES) { dsl_crypto_params_t *dcp; nvlist_t *crypto_args = fnvlist_alloc(); fnvlist_add_uint8_array(crypto_args, "wkeydata", (uint8_t *)ztest_wkeydata, WRAPPING_KEY_LEN); VERIFY0(dsl_crypto_params_create_nvlist(DCP_CMD_NONE, NULL, crypto_args, &dcp)); err = spa_keystore_load_wkey(ddname, dcp, B_FALSE); /* * Note: if there was an error loading, the wkey was not * consumed, and needs to be freed. */ dsl_crypto_params_free(dcp, (err != 0)); fnvlist_free(crypto_args); if (err == EINVAL) { /* * We couldn't load a key for this dataset so try * the parent. This loop will eventually hit the * encryption root since ztest only makes clones * as children of their origin datasets. */ cp = strrchr(ddname, '/'); if (cp == NULL) return (err); *cp = '\0'; err = EACCES; continue; } else if (err != 0) { break; } err = dmu_objset_own(name, type, readonly, decrypt, tag, osp); break; } return (err); } static void ztest_rll_init(rll_t *rll) { rll->rll_writer = NULL; rll->rll_readers = 0; mutex_init(&rll->rll_lock, NULL, MUTEX_DEFAULT, NULL); cv_init(&rll->rll_cv, NULL, CV_DEFAULT, NULL); } static void ztest_rll_destroy(rll_t *rll) { ASSERT3P(rll->rll_writer, ==, NULL); ASSERT0(rll->rll_readers); mutex_destroy(&rll->rll_lock); cv_destroy(&rll->rll_cv); } static void ztest_rll_lock(rll_t *rll, rl_type_t type) { mutex_enter(&rll->rll_lock); if (type == RL_READER) { while (rll->rll_writer != NULL) (void) cv_wait(&rll->rll_cv, &rll->rll_lock); rll->rll_readers++; } else { while (rll->rll_writer != NULL || rll->rll_readers) (void) cv_wait(&rll->rll_cv, &rll->rll_lock); rll->rll_writer = curthread; } mutex_exit(&rll->rll_lock); } static void ztest_rll_unlock(rll_t *rll) { mutex_enter(&rll->rll_lock); if (rll->rll_writer) { ASSERT0(rll->rll_readers); rll->rll_writer = NULL; } else { ASSERT3S(rll->rll_readers, >, 0); ASSERT3P(rll->rll_writer, ==, NULL); rll->rll_readers--; } if (rll->rll_writer == NULL && rll->rll_readers == 0) cv_broadcast(&rll->rll_cv); mutex_exit(&rll->rll_lock); } static void ztest_object_lock(ztest_ds_t *zd, uint64_t object, rl_type_t type) { rll_t *rll = &zd->zd_object_lock[object & (ZTEST_OBJECT_LOCKS - 1)]; ztest_rll_lock(rll, type); } static void ztest_object_unlock(ztest_ds_t *zd, uint64_t object) { rll_t *rll = &zd->zd_object_lock[object & (ZTEST_OBJECT_LOCKS - 1)]; ztest_rll_unlock(rll); } static rl_t * ztest_range_lock(ztest_ds_t *zd, uint64_t object, uint64_t offset, uint64_t size, rl_type_t type) { uint64_t hash = object ^ (offset % (ZTEST_RANGE_LOCKS + 1)); rll_t *rll = &zd->zd_range_lock[hash & (ZTEST_RANGE_LOCKS - 1)]; rl_t *rl; rl = umem_alloc(sizeof (*rl), UMEM_NOFAIL); rl->rl_object = object; rl->rl_offset = offset; rl->rl_size = size; rl->rl_lock = rll; ztest_rll_lock(rll, type); return (rl); } static void ztest_range_unlock(rl_t *rl) { rll_t *rll = rl->rl_lock; ztest_rll_unlock(rll); umem_free(rl, sizeof (*rl)); } static void ztest_zd_init(ztest_ds_t *zd, ztest_shared_ds_t *szd, objset_t *os) { zd->zd_os = os; zd->zd_zilog = dmu_objset_zil(os); zd->zd_shared = szd; dmu_objset_name(os, zd->zd_name); int l; if (zd->zd_shared != NULL) zd->zd_shared->zd_seq = 0; VERIFY0(pthread_rwlock_init(&zd->zd_zilog_lock, NULL)); mutex_init(&zd->zd_dirobj_lock, NULL, MUTEX_DEFAULT, NULL); for (l = 0; l < ZTEST_OBJECT_LOCKS; l++) ztest_rll_init(&zd->zd_object_lock[l]); for (l = 0; l < ZTEST_RANGE_LOCKS; l++) ztest_rll_init(&zd->zd_range_lock[l]); } static void ztest_zd_fini(ztest_ds_t *zd) { int l; mutex_destroy(&zd->zd_dirobj_lock); (void) pthread_rwlock_destroy(&zd->zd_zilog_lock); for (l = 0; l < ZTEST_OBJECT_LOCKS; l++) ztest_rll_destroy(&zd->zd_object_lock[l]); for (l = 0; l < ZTEST_RANGE_LOCKS; l++) ztest_rll_destroy(&zd->zd_range_lock[l]); } #define TXG_MIGHTWAIT (ztest_random(10) == 0 ? TXG_NOWAIT : TXG_WAIT) static uint64_t ztest_tx_assign(dmu_tx_t *tx, uint64_t txg_how, const char *tag) { uint64_t txg; int error; /* * Attempt to assign tx to some transaction group. */ error = dmu_tx_assign(tx, txg_how); if (error) { if (error == ERESTART) { ASSERT3U(txg_how, ==, TXG_NOWAIT); dmu_tx_wait(tx); } else { ASSERT3U(error, ==, ENOSPC); ztest_record_enospc(tag); } dmu_tx_abort(tx); return (0); } txg = dmu_tx_get_txg(tx); ASSERT3U(txg, !=, 0); return (txg); } static void ztest_bt_generate(ztest_block_tag_t *bt, objset_t *os, uint64_t object, uint64_t dnodesize, uint64_t offset, uint64_t gen, uint64_t txg, uint64_t crtxg) { bt->bt_magic = BT_MAGIC; bt->bt_objset = dmu_objset_id(os); bt->bt_object = object; bt->bt_dnodesize = dnodesize; bt->bt_offset = offset; bt->bt_gen = gen; bt->bt_txg = txg; bt->bt_crtxg = crtxg; } static void ztest_bt_verify(ztest_block_tag_t *bt, objset_t *os, uint64_t object, uint64_t dnodesize, uint64_t offset, uint64_t gen, uint64_t txg, uint64_t crtxg) { ASSERT3U(bt->bt_magic, ==, BT_MAGIC); ASSERT3U(bt->bt_objset, ==, dmu_objset_id(os)); ASSERT3U(bt->bt_object, ==, object); ASSERT3U(bt->bt_dnodesize, ==, dnodesize); ASSERT3U(bt->bt_offset, ==, offset); ASSERT3U(bt->bt_gen, <=, gen); ASSERT3U(bt->bt_txg, <=, txg); ASSERT3U(bt->bt_crtxg, ==, crtxg); } static ztest_block_tag_t * ztest_bt_bonus(dmu_buf_t *db) { dmu_object_info_t doi; ztest_block_tag_t *bt; dmu_object_info_from_db(db, &doi); ASSERT3U(doi.doi_bonus_size, <=, db->db_size); ASSERT3U(doi.doi_bonus_size, >=, sizeof (*bt)); bt = (void *)((char *)db->db_data + doi.doi_bonus_size - sizeof (*bt)); return (bt); } /* * Generate a token to fill up unused bonus buffer space. Try to make * it unique to the object, generation, and offset to verify that data * is not getting overwritten by data from other dnodes. */ #define ZTEST_BONUS_FILL_TOKEN(obj, ds, gen, offset) \ (((ds) << 48) | ((gen) << 32) | ((obj) << 8) | (offset)) /* * Fill up the unused bonus buffer region before the block tag with a * verifiable pattern. Filling the whole bonus area with non-zero data * helps ensure that all dnode traversal code properly skips the * interior regions of large dnodes. */ static void ztest_fill_unused_bonus(dmu_buf_t *db, void *end, uint64_t obj, objset_t *os, uint64_t gen) { uint64_t *bonusp; ASSERT(IS_P2ALIGNED((char *)end - (char *)db->db_data, 8)); for (bonusp = db->db_data; bonusp < (uint64_t *)end; bonusp++) { uint64_t token = ZTEST_BONUS_FILL_TOKEN(obj, dmu_objset_id(os), gen, bonusp - (uint64_t *)db->db_data); *bonusp = token; } } /* * Verify that the unused area of a bonus buffer is filled with the * expected tokens. */ static void ztest_verify_unused_bonus(dmu_buf_t *db, void *end, uint64_t obj, objset_t *os, uint64_t gen) { uint64_t *bonusp; for (bonusp = db->db_data; bonusp < (uint64_t *)end; bonusp++) { uint64_t token = ZTEST_BONUS_FILL_TOKEN(obj, dmu_objset_id(os), gen, bonusp - (uint64_t *)db->db_data); VERIFY3U(*bonusp, ==, token); } } /* * ZIL logging ops */ #define lrz_type lr_mode #define lrz_blocksize lr_uid #define lrz_ibshift lr_gid #define lrz_bonustype lr_rdev #define lrz_dnodesize lr_crtime[1] static void ztest_log_create(ztest_ds_t *zd, dmu_tx_t *tx, lr_create_t *lr) { char *name = (void *)(lr + 1); /* name follows lr */ size_t namesize = strlen(name) + 1; itx_t *itx; if (zil_replaying(zd->zd_zilog, tx)) return; itx = zil_itx_create(TX_CREATE, sizeof (*lr) + namesize); memcpy(&itx->itx_lr + 1, &lr->lr_common + 1, sizeof (*lr) + namesize - sizeof (lr_t)); zil_itx_assign(zd->zd_zilog, itx, tx); } static void ztest_log_remove(ztest_ds_t *zd, dmu_tx_t *tx, lr_remove_t *lr, uint64_t object) { char *name = (void *)(lr + 1); /* name follows lr */ size_t namesize = strlen(name) + 1; itx_t *itx; if (zil_replaying(zd->zd_zilog, tx)) return; itx = zil_itx_create(TX_REMOVE, sizeof (*lr) + namesize); memcpy(&itx->itx_lr + 1, &lr->lr_common + 1, sizeof (*lr) + namesize - sizeof (lr_t)); itx->itx_oid = object; zil_itx_assign(zd->zd_zilog, itx, tx); } static void ztest_log_write(ztest_ds_t *zd, dmu_tx_t *tx, lr_write_t *lr) { itx_t *itx; itx_wr_state_t write_state = ztest_random(WR_NUM_STATES); if (zil_replaying(zd->zd_zilog, tx)) return; if (lr->lr_length > zil_max_log_data(zd->zd_zilog)) write_state = WR_INDIRECT; itx = zil_itx_create(TX_WRITE, sizeof (*lr) + (write_state == WR_COPIED ? lr->lr_length : 0)); if (write_state == WR_COPIED && dmu_read(zd->zd_os, lr->lr_foid, lr->lr_offset, lr->lr_length, ((lr_write_t *)&itx->itx_lr) + 1, DMU_READ_NO_PREFETCH) != 0) { zil_itx_destroy(itx); itx = zil_itx_create(TX_WRITE, sizeof (*lr)); write_state = WR_NEED_COPY; } itx->itx_private = zd; itx->itx_wr_state = write_state; itx->itx_sync = (ztest_random(8) == 0); memcpy(&itx->itx_lr + 1, &lr->lr_common + 1, sizeof (*lr) - sizeof (lr_t)); zil_itx_assign(zd->zd_zilog, itx, tx); } static void ztest_log_truncate(ztest_ds_t *zd, dmu_tx_t *tx, lr_truncate_t *lr) { itx_t *itx; if (zil_replaying(zd->zd_zilog, tx)) return; itx = zil_itx_create(TX_TRUNCATE, sizeof (*lr)); memcpy(&itx->itx_lr + 1, &lr->lr_common + 1, sizeof (*lr) - sizeof (lr_t)); itx->itx_sync = B_FALSE; zil_itx_assign(zd->zd_zilog, itx, tx); } static void ztest_log_setattr(ztest_ds_t *zd, dmu_tx_t *tx, lr_setattr_t *lr) { itx_t *itx; if (zil_replaying(zd->zd_zilog, tx)) return; itx = zil_itx_create(TX_SETATTR, sizeof (*lr)); memcpy(&itx->itx_lr + 1, &lr->lr_common + 1, sizeof (*lr) - sizeof (lr_t)); itx->itx_sync = B_FALSE; zil_itx_assign(zd->zd_zilog, itx, tx); } /* * ZIL replay ops */ static int ztest_replay_create(void *arg1, void *arg2, boolean_t byteswap) { ztest_ds_t *zd = arg1; lr_create_t *lr = arg2; char *name = (void *)(lr + 1); /* name follows lr */ objset_t *os = zd->zd_os; ztest_block_tag_t *bbt; dmu_buf_t *db; dmu_tx_t *tx; uint64_t txg; int error = 0; int bonuslen; if (byteswap) byteswap_uint64_array(lr, sizeof (*lr)); ASSERT3U(lr->lr_doid, ==, ZTEST_DIROBJ); ASSERT3S(name[0], !=, '\0'); tx = dmu_tx_create(os); dmu_tx_hold_zap(tx, lr->lr_doid, B_TRUE, name); if (lr->lrz_type == DMU_OT_ZAP_OTHER) { dmu_tx_hold_zap(tx, DMU_NEW_OBJECT, B_TRUE, NULL); } else { dmu_tx_hold_bonus(tx, DMU_NEW_OBJECT); } txg = ztest_tx_assign(tx, TXG_WAIT, FTAG); if (txg == 0) return (ENOSPC); ASSERT3U(dmu_objset_zil(os)->zl_replay, ==, !!lr->lr_foid); bonuslen = DN_BONUS_SIZE(lr->lrz_dnodesize); if (lr->lrz_type == DMU_OT_ZAP_OTHER) { if (lr->lr_foid == 0) { lr->lr_foid = zap_create_dnsize(os, lr->lrz_type, lr->lrz_bonustype, bonuslen, lr->lrz_dnodesize, tx); } else { error = zap_create_claim_dnsize(os, lr->lr_foid, lr->lrz_type, lr->lrz_bonustype, bonuslen, lr->lrz_dnodesize, tx); } } else { if (lr->lr_foid == 0) { lr->lr_foid = dmu_object_alloc_dnsize(os, lr->lrz_type, 0, lr->lrz_bonustype, bonuslen, lr->lrz_dnodesize, tx); } else { error = dmu_object_claim_dnsize(os, lr->lr_foid, lr->lrz_type, 0, lr->lrz_bonustype, bonuslen, lr->lrz_dnodesize, tx); } } if (error) { ASSERT3U(error, ==, EEXIST); ASSERT(zd->zd_zilog->zl_replay); dmu_tx_commit(tx); return (error); } ASSERT3U(lr->lr_foid, !=, 0); if (lr->lrz_type != DMU_OT_ZAP_OTHER) VERIFY0(dmu_object_set_blocksize(os, lr->lr_foid, lr->lrz_blocksize, lr->lrz_ibshift, tx)); VERIFY0(dmu_bonus_hold(os, lr->lr_foid, FTAG, &db)); bbt = ztest_bt_bonus(db); dmu_buf_will_dirty(db, tx); ztest_bt_generate(bbt, os, lr->lr_foid, lr->lrz_dnodesize, -1ULL, lr->lr_gen, txg, txg); ztest_fill_unused_bonus(db, bbt, lr->lr_foid, os, lr->lr_gen); dmu_buf_rele(db, FTAG); VERIFY0(zap_add(os, lr->lr_doid, name, sizeof (uint64_t), 1, &lr->lr_foid, tx)); (void) ztest_log_create(zd, tx, lr); dmu_tx_commit(tx); return (0); } static int ztest_replay_remove(void *arg1, void *arg2, boolean_t byteswap) { ztest_ds_t *zd = arg1; lr_remove_t *lr = arg2; char *name = (void *)(lr + 1); /* name follows lr */ objset_t *os = zd->zd_os; dmu_object_info_t doi; dmu_tx_t *tx; uint64_t object, txg; if (byteswap) byteswap_uint64_array(lr, sizeof (*lr)); ASSERT3U(lr->lr_doid, ==, ZTEST_DIROBJ); ASSERT3S(name[0], !=, '\0'); VERIFY0( zap_lookup(os, lr->lr_doid, name, sizeof (object), 1, &object)); ASSERT3U(object, !=, 0); ztest_object_lock(zd, object, RL_WRITER); VERIFY0(dmu_object_info(os, object, &doi)); tx = dmu_tx_create(os); dmu_tx_hold_zap(tx, lr->lr_doid, B_FALSE, name); dmu_tx_hold_free(tx, object, 0, DMU_OBJECT_END); txg = ztest_tx_assign(tx, TXG_WAIT, FTAG); if (txg == 0) { ztest_object_unlock(zd, object); return (ENOSPC); } if (doi.doi_type == DMU_OT_ZAP_OTHER) { VERIFY0(zap_destroy(os, object, tx)); } else { VERIFY0(dmu_object_free(os, object, tx)); } VERIFY0(zap_remove(os, lr->lr_doid, name, tx)); (void) ztest_log_remove(zd, tx, lr, object); dmu_tx_commit(tx); ztest_object_unlock(zd, object); return (0); } static int ztest_replay_write(void *arg1, void *arg2, boolean_t byteswap) { ztest_ds_t *zd = arg1; lr_write_t *lr = arg2; objset_t *os = zd->zd_os; void *data = lr + 1; /* data follows lr */ uint64_t offset, length; ztest_block_tag_t *bt = data; ztest_block_tag_t *bbt; uint64_t gen, txg, lrtxg, crtxg; dmu_object_info_t doi; dmu_tx_t *tx; dmu_buf_t *db; arc_buf_t *abuf = NULL; rl_t *rl; if (byteswap) byteswap_uint64_array(lr, sizeof (*lr)); offset = lr->lr_offset; length = lr->lr_length; /* If it's a dmu_sync() block, write the whole block */ if (lr->lr_common.lrc_reclen == sizeof (lr_write_t)) { uint64_t blocksize = BP_GET_LSIZE(&lr->lr_blkptr); if (length < blocksize) { offset -= offset % blocksize; length = blocksize; } } if (bt->bt_magic == BSWAP_64(BT_MAGIC)) byteswap_uint64_array(bt, sizeof (*bt)); if (bt->bt_magic != BT_MAGIC) bt = NULL; ztest_object_lock(zd, lr->lr_foid, RL_READER); rl = ztest_range_lock(zd, lr->lr_foid, offset, length, RL_WRITER); VERIFY0(dmu_bonus_hold(os, lr->lr_foid, FTAG, &db)); dmu_object_info_from_db(db, &doi); bbt = ztest_bt_bonus(db); ASSERT3U(bbt->bt_magic, ==, BT_MAGIC); gen = bbt->bt_gen; crtxg = bbt->bt_crtxg; lrtxg = lr->lr_common.lrc_txg; tx = dmu_tx_create(os); dmu_tx_hold_write(tx, lr->lr_foid, offset, length); if (ztest_random(8) == 0 && length == doi.doi_data_block_size && P2PHASE(offset, length) == 0) abuf = dmu_request_arcbuf(db, length); txg = ztest_tx_assign(tx, TXG_WAIT, FTAG); if (txg == 0) { if (abuf != NULL) dmu_return_arcbuf(abuf); dmu_buf_rele(db, FTAG); ztest_range_unlock(rl); ztest_object_unlock(zd, lr->lr_foid); return (ENOSPC); } if (bt != NULL) { /* * Usually, verify the old data before writing new data -- * but not always, because we also want to verify correct * behavior when the data was not recently read into cache. */ ASSERT0(offset % doi.doi_data_block_size); if (ztest_random(4) != 0) { int prefetch = ztest_random(2) ? DMU_READ_PREFETCH : DMU_READ_NO_PREFETCH; ztest_block_tag_t rbt; VERIFY(dmu_read(os, lr->lr_foid, offset, sizeof (rbt), &rbt, prefetch) == 0); if (rbt.bt_magic == BT_MAGIC) { ztest_bt_verify(&rbt, os, lr->lr_foid, 0, offset, gen, txg, crtxg); } } /* * Writes can appear to be newer than the bonus buffer because * the ztest_get_data() callback does a dmu_read() of the * open-context data, which may be different than the data * as it was when the write was generated. */ if (zd->zd_zilog->zl_replay) { ztest_bt_verify(bt, os, lr->lr_foid, 0, offset, MAX(gen, bt->bt_gen), MAX(txg, lrtxg), bt->bt_crtxg); } /* * Set the bt's gen/txg to the bonus buffer's gen/txg * so that all of the usual ASSERTs will work. */ ztest_bt_generate(bt, os, lr->lr_foid, 0, offset, gen, txg, crtxg); } if (abuf == NULL) { dmu_write(os, lr->lr_foid, offset, length, data, tx); } else { memcpy(abuf->b_data, data, length); VERIFY0(dmu_assign_arcbuf_by_dbuf(db, offset, abuf, tx)); } (void) ztest_log_write(zd, tx, lr); dmu_buf_rele(db, FTAG); dmu_tx_commit(tx); ztest_range_unlock(rl); ztest_object_unlock(zd, lr->lr_foid); return (0); } static int ztest_replay_truncate(void *arg1, void *arg2, boolean_t byteswap) { ztest_ds_t *zd = arg1; lr_truncate_t *lr = arg2; objset_t *os = zd->zd_os; dmu_tx_t *tx; uint64_t txg; rl_t *rl; if (byteswap) byteswap_uint64_array(lr, sizeof (*lr)); ztest_object_lock(zd, lr->lr_foid, RL_READER); rl = ztest_range_lock(zd, lr->lr_foid, lr->lr_offset, lr->lr_length, RL_WRITER); tx = dmu_tx_create(os); dmu_tx_hold_free(tx, lr->lr_foid, lr->lr_offset, lr->lr_length); txg = ztest_tx_assign(tx, TXG_WAIT, FTAG); if (txg == 0) { ztest_range_unlock(rl); ztest_object_unlock(zd, lr->lr_foid); return (ENOSPC); } VERIFY0(dmu_free_range(os, lr->lr_foid, lr->lr_offset, lr->lr_length, tx)); (void) ztest_log_truncate(zd, tx, lr); dmu_tx_commit(tx); ztest_range_unlock(rl); ztest_object_unlock(zd, lr->lr_foid); return (0); } static int ztest_replay_setattr(void *arg1, void *arg2, boolean_t byteswap) { ztest_ds_t *zd = arg1; lr_setattr_t *lr = arg2; objset_t *os = zd->zd_os; dmu_tx_t *tx; dmu_buf_t *db; ztest_block_tag_t *bbt; uint64_t txg, lrtxg, crtxg, dnodesize; if (byteswap) byteswap_uint64_array(lr, sizeof (*lr)); ztest_object_lock(zd, lr->lr_foid, RL_WRITER); VERIFY0(dmu_bonus_hold(os, lr->lr_foid, FTAG, &db)); tx = dmu_tx_create(os); dmu_tx_hold_bonus(tx, lr->lr_foid); txg = ztest_tx_assign(tx, TXG_WAIT, FTAG); if (txg == 0) { dmu_buf_rele(db, FTAG); ztest_object_unlock(zd, lr->lr_foid); return (ENOSPC); } bbt = ztest_bt_bonus(db); ASSERT3U(bbt->bt_magic, ==, BT_MAGIC); crtxg = bbt->bt_crtxg; lrtxg = lr->lr_common.lrc_txg; dnodesize = bbt->bt_dnodesize; if (zd->zd_zilog->zl_replay) { ASSERT3U(lr->lr_size, !=, 0); ASSERT3U(lr->lr_mode, !=, 0); ASSERT3U(lrtxg, !=, 0); } else { /* * Randomly change the size and increment the generation. */ lr->lr_size = (ztest_random(db->db_size / sizeof (*bbt)) + 1) * sizeof (*bbt); lr->lr_mode = bbt->bt_gen + 1; ASSERT0(lrtxg); } /* * Verify that the current bonus buffer is not newer than our txg. */ ztest_bt_verify(bbt, os, lr->lr_foid, dnodesize, -1ULL, lr->lr_mode, MAX(txg, lrtxg), crtxg); dmu_buf_will_dirty(db, tx); ASSERT3U(lr->lr_size, >=, sizeof (*bbt)); ASSERT3U(lr->lr_size, <=, db->db_size); VERIFY0(dmu_set_bonus(db, lr->lr_size, tx)); bbt = ztest_bt_bonus(db); ztest_bt_generate(bbt, os, lr->lr_foid, dnodesize, -1ULL, lr->lr_mode, txg, crtxg); ztest_fill_unused_bonus(db, bbt, lr->lr_foid, os, bbt->bt_gen); dmu_buf_rele(db, FTAG); (void) ztest_log_setattr(zd, tx, lr); dmu_tx_commit(tx); ztest_object_unlock(zd, lr->lr_foid); return (0); } zil_replay_func_t *ztest_replay_vector[TX_MAX_TYPE] = { NULL, /* 0 no such transaction type */ ztest_replay_create, /* TX_CREATE */ NULL, /* TX_MKDIR */ NULL, /* TX_MKXATTR */ NULL, /* TX_SYMLINK */ ztest_replay_remove, /* TX_REMOVE */ NULL, /* TX_RMDIR */ NULL, /* TX_LINK */ NULL, /* TX_RENAME */ ztest_replay_write, /* TX_WRITE */ ztest_replay_truncate, /* TX_TRUNCATE */ ztest_replay_setattr, /* TX_SETATTR */ NULL, /* TX_ACL */ NULL, /* TX_CREATE_ACL */ NULL, /* TX_CREATE_ATTR */ NULL, /* TX_CREATE_ACL_ATTR */ NULL, /* TX_MKDIR_ACL */ NULL, /* TX_MKDIR_ATTR */ NULL, /* TX_MKDIR_ACL_ATTR */ NULL, /* TX_WRITE2 */ NULL, /* TX_SETSAXATTR */ }; /* * ZIL get_data callbacks */ static void ztest_get_done(zgd_t *zgd, int error) { (void) error; ztest_ds_t *zd = zgd->zgd_private; uint64_t object = ((rl_t *)zgd->zgd_lr)->rl_object; if (zgd->zgd_db) dmu_buf_rele(zgd->zgd_db, zgd); ztest_range_unlock((rl_t *)zgd->zgd_lr); ztest_object_unlock(zd, object); umem_free(zgd, sizeof (*zgd)); } static int ztest_get_data(void *arg, uint64_t arg2, lr_write_t *lr, char *buf, struct lwb *lwb, zio_t *zio) { (void) arg2; ztest_ds_t *zd = arg; objset_t *os = zd->zd_os; uint64_t object = lr->lr_foid; uint64_t offset = lr->lr_offset; uint64_t size = lr->lr_length; uint64_t txg = lr->lr_common.lrc_txg; uint64_t crtxg; dmu_object_info_t doi; dmu_buf_t *db; zgd_t *zgd; int error; ASSERT3P(lwb, !=, NULL); ASSERT3P(zio, !=, NULL); ASSERT3U(size, !=, 0); ztest_object_lock(zd, object, RL_READER); error = dmu_bonus_hold(os, object, FTAG, &db); if (error) { ztest_object_unlock(zd, object); return (error); } crtxg = ztest_bt_bonus(db)->bt_crtxg; if (crtxg == 0 || crtxg > txg) { dmu_buf_rele(db, FTAG); ztest_object_unlock(zd, object); return (ENOENT); } dmu_object_info_from_db(db, &doi); dmu_buf_rele(db, FTAG); db = NULL; zgd = umem_zalloc(sizeof (*zgd), UMEM_NOFAIL); zgd->zgd_lwb = lwb; zgd->zgd_private = zd; if (buf != NULL) { /* immediate write */ zgd->zgd_lr = (struct zfs_locked_range *)ztest_range_lock(zd, object, offset, size, RL_READER); error = dmu_read(os, object, offset, size, buf, DMU_READ_NO_PREFETCH); ASSERT0(error); } else { size = doi.doi_data_block_size; if (ISP2(size)) { offset = P2ALIGN(offset, size); } else { ASSERT3U(offset, <, size); offset = 0; } zgd->zgd_lr = (struct zfs_locked_range *)ztest_range_lock(zd, object, offset, size, RL_READER); error = dmu_buf_hold(os, object, offset, zgd, &db, DMU_READ_NO_PREFETCH); if (error == 0) { blkptr_t *bp = &lr->lr_blkptr; zgd->zgd_db = db; zgd->zgd_bp = bp; ASSERT3U(db->db_offset, ==, offset); ASSERT3U(db->db_size, ==, size); error = dmu_sync(zio, lr->lr_common.lrc_txg, ztest_get_done, zgd); if (error == 0) return (0); } } ztest_get_done(zgd, error); return (error); } static void * ztest_lr_alloc(size_t lrsize, char *name) { char *lr; size_t namesize = name ? strlen(name) + 1 : 0; lr = umem_zalloc(lrsize + namesize, UMEM_NOFAIL); if (name) memcpy(lr + lrsize, name, namesize); return (lr); } static void ztest_lr_free(void *lr, size_t lrsize, char *name) { size_t namesize = name ? strlen(name) + 1 : 0; umem_free(lr, lrsize + namesize); } /* * Lookup a bunch of objects. Returns the number of objects not found. */ static int ztest_lookup(ztest_ds_t *zd, ztest_od_t *od, int count) { int missing = 0; int error; int i; ASSERT(MUTEX_HELD(&zd->zd_dirobj_lock)); for (i = 0; i < count; i++, od++) { od->od_object = 0; error = zap_lookup(zd->zd_os, od->od_dir, od->od_name, sizeof (uint64_t), 1, &od->od_object); if (error) { ASSERT3S(error, ==, ENOENT); ASSERT0(od->od_object); missing++; } else { dmu_buf_t *db; ztest_block_tag_t *bbt; dmu_object_info_t doi; ASSERT3U(od->od_object, !=, 0); ASSERT0(missing); /* there should be no gaps */ ztest_object_lock(zd, od->od_object, RL_READER); VERIFY0(dmu_bonus_hold(zd->zd_os, od->od_object, FTAG, &db)); dmu_object_info_from_db(db, &doi); bbt = ztest_bt_bonus(db); ASSERT3U(bbt->bt_magic, ==, BT_MAGIC); od->od_type = doi.doi_type; od->od_blocksize = doi.doi_data_block_size; od->od_gen = bbt->bt_gen; dmu_buf_rele(db, FTAG); ztest_object_unlock(zd, od->od_object); } } return (missing); } static int ztest_create(ztest_ds_t *zd, ztest_od_t *od, int count) { int missing = 0; int i; ASSERT(MUTEX_HELD(&zd->zd_dirobj_lock)); for (i = 0; i < count; i++, od++) { if (missing) { od->od_object = 0; missing++; continue; } lr_create_t *lr = ztest_lr_alloc(sizeof (*lr), od->od_name); lr->lr_doid = od->od_dir; lr->lr_foid = 0; /* 0 to allocate, > 0 to claim */ lr->lrz_type = od->od_crtype; lr->lrz_blocksize = od->od_crblocksize; lr->lrz_ibshift = ztest_random_ibshift(); lr->lrz_bonustype = DMU_OT_UINT64_OTHER; lr->lrz_dnodesize = od->od_crdnodesize; lr->lr_gen = od->od_crgen; lr->lr_crtime[0] = time(NULL); if (ztest_replay_create(zd, lr, B_FALSE) != 0) { ASSERT0(missing); od->od_object = 0; missing++; } else { od->od_object = lr->lr_foid; od->od_type = od->od_crtype; od->od_blocksize = od->od_crblocksize; od->od_gen = od->od_crgen; ASSERT3U(od->od_object, !=, 0); } ztest_lr_free(lr, sizeof (*lr), od->od_name); } return (missing); } static int ztest_remove(ztest_ds_t *zd, ztest_od_t *od, int count) { int missing = 0; int error; int i; ASSERT(MUTEX_HELD(&zd->zd_dirobj_lock)); od += count - 1; for (i = count - 1; i >= 0; i--, od--) { if (missing) { missing++; continue; } /* * No object was found. */ if (od->od_object == 0) continue; lr_remove_t *lr = ztest_lr_alloc(sizeof (*lr), od->od_name); lr->lr_doid = od->od_dir; if ((error = ztest_replay_remove(zd, lr, B_FALSE)) != 0) { ASSERT3U(error, ==, ENOSPC); missing++; } else { od->od_object = 0; } ztest_lr_free(lr, sizeof (*lr), od->od_name); } return (missing); } static int ztest_write(ztest_ds_t *zd, uint64_t object, uint64_t offset, uint64_t size, void *data) { lr_write_t *lr; int error; lr = ztest_lr_alloc(sizeof (*lr) + size, NULL); lr->lr_foid = object; lr->lr_offset = offset; lr->lr_length = size; lr->lr_blkoff = 0; BP_ZERO(&lr->lr_blkptr); memcpy(lr + 1, data, size); error = ztest_replay_write(zd, lr, B_FALSE); ztest_lr_free(lr, sizeof (*lr) + size, NULL); return (error); } static int ztest_truncate(ztest_ds_t *zd, uint64_t object, uint64_t offset, uint64_t size) { lr_truncate_t *lr; int error; lr = ztest_lr_alloc(sizeof (*lr), NULL); lr->lr_foid = object; lr->lr_offset = offset; lr->lr_length = size; error = ztest_replay_truncate(zd, lr, B_FALSE); ztest_lr_free(lr, sizeof (*lr), NULL); return (error); } static int ztest_setattr(ztest_ds_t *zd, uint64_t object) { lr_setattr_t *lr; int error; lr = ztest_lr_alloc(sizeof (*lr), NULL); lr->lr_foid = object; lr->lr_size = 0; lr->lr_mode = 0; error = ztest_replay_setattr(zd, lr, B_FALSE); ztest_lr_free(lr, sizeof (*lr), NULL); return (error); } static void ztest_prealloc(ztest_ds_t *zd, uint64_t object, uint64_t offset, uint64_t size) { objset_t *os = zd->zd_os; dmu_tx_t *tx; uint64_t txg; rl_t *rl; txg_wait_synced(dmu_objset_pool(os), 0); ztest_object_lock(zd, object, RL_READER); rl = ztest_range_lock(zd, object, offset, size, RL_WRITER); tx = dmu_tx_create(os); dmu_tx_hold_write(tx, object, offset, size); txg = ztest_tx_assign(tx, TXG_WAIT, FTAG); if (txg != 0) { dmu_prealloc(os, object, offset, size, tx); dmu_tx_commit(tx); txg_wait_synced(dmu_objset_pool(os), txg); } else { (void) dmu_free_long_range(os, object, offset, size); } ztest_range_unlock(rl); ztest_object_unlock(zd, object); } static void ztest_io(ztest_ds_t *zd, uint64_t object, uint64_t offset) { int err; ztest_block_tag_t wbt; dmu_object_info_t doi; enum ztest_io_type io_type; uint64_t blocksize; void *data; VERIFY0(dmu_object_info(zd->zd_os, object, &doi)); blocksize = doi.doi_data_block_size; data = umem_alloc(blocksize, UMEM_NOFAIL); /* * Pick an i/o type at random, biased toward writing block tags. */ io_type = ztest_random(ZTEST_IO_TYPES); if (ztest_random(2) == 0) io_type = ZTEST_IO_WRITE_TAG; (void) pthread_rwlock_rdlock(&zd->zd_zilog_lock); switch (io_type) { case ZTEST_IO_WRITE_TAG: ztest_bt_generate(&wbt, zd->zd_os, object, doi.doi_dnodesize, offset, 0, 0, 0); (void) ztest_write(zd, object, offset, sizeof (wbt), &wbt); break; case ZTEST_IO_WRITE_PATTERN: (void) memset(data, 'a' + (object + offset) % 5, blocksize); if (ztest_random(2) == 0) { /* * Induce fletcher2 collisions to ensure that * zio_ddt_collision() detects and resolves them * when using fletcher2-verify for deduplication. */ ((uint64_t *)data)[0] ^= 1ULL << 63; ((uint64_t *)data)[4] ^= 1ULL << 63; } (void) ztest_write(zd, object, offset, blocksize, data); break; case ZTEST_IO_WRITE_ZEROES: memset(data, 0, blocksize); (void) ztest_write(zd, object, offset, blocksize, data); break; case ZTEST_IO_TRUNCATE: (void) ztest_truncate(zd, object, offset, blocksize); break; case ZTEST_IO_SETATTR: (void) ztest_setattr(zd, object); break; default: break; case ZTEST_IO_REWRITE: (void) pthread_rwlock_rdlock(&ztest_name_lock); err = ztest_dsl_prop_set_uint64(zd->zd_name, ZFS_PROP_CHECKSUM, spa_dedup_checksum(ztest_spa), B_FALSE); VERIFY(err == 0 || err == ENOSPC); err = ztest_dsl_prop_set_uint64(zd->zd_name, ZFS_PROP_COMPRESSION, ztest_random_dsl_prop(ZFS_PROP_COMPRESSION), B_FALSE); VERIFY(err == 0 || err == ENOSPC); (void) pthread_rwlock_unlock(&ztest_name_lock); VERIFY0(dmu_read(zd->zd_os, object, offset, blocksize, data, DMU_READ_NO_PREFETCH)); (void) ztest_write(zd, object, offset, blocksize, data); break; } (void) pthread_rwlock_unlock(&zd->zd_zilog_lock); umem_free(data, blocksize); } /* * Initialize an object description template. */ static void ztest_od_init(ztest_od_t *od, uint64_t id, const char *tag, uint64_t index, dmu_object_type_t type, uint64_t blocksize, uint64_t dnodesize, uint64_t gen) { od->od_dir = ZTEST_DIROBJ; od->od_object = 0; od->od_crtype = type; od->od_crblocksize = blocksize ? blocksize : ztest_random_blocksize(); od->od_crdnodesize = dnodesize ? dnodesize : ztest_random_dnodesize(); od->od_crgen = gen; od->od_type = DMU_OT_NONE; od->od_blocksize = 0; od->od_gen = 0; (void) snprintf(od->od_name, sizeof (od->od_name), "%s(%"PRId64")[%"PRIu64"]", tag, id, index); } /* * Lookup or create the objects for a test using the od template. * If the objects do not all exist, or if 'remove' is specified, * remove any existing objects and create new ones. Otherwise, * use the existing objects. */ static int ztest_object_init(ztest_ds_t *zd, ztest_od_t *od, size_t size, boolean_t remove) { int count = size / sizeof (*od); int rv = 0; mutex_enter(&zd->zd_dirobj_lock); if ((ztest_lookup(zd, od, count) != 0 || remove) && (ztest_remove(zd, od, count) != 0 || ztest_create(zd, od, count) != 0)) rv = -1; zd->zd_od = od; mutex_exit(&zd->zd_dirobj_lock); return (rv); } void ztest_zil_commit(ztest_ds_t *zd, uint64_t id) { (void) id; zilog_t *zilog = zd->zd_zilog; (void) pthread_rwlock_rdlock(&zd->zd_zilog_lock); zil_commit(zilog, ztest_random(ZTEST_OBJECTS)); /* * Remember the committed values in zd, which is in parent/child * shared memory. If we die, the next iteration of ztest_run() * will verify that the log really does contain this record. */ mutex_enter(&zilog->zl_lock); ASSERT3P(zd->zd_shared, !=, NULL); ASSERT3U(zd->zd_shared->zd_seq, <=, zilog->zl_commit_lr_seq); zd->zd_shared->zd_seq = zilog->zl_commit_lr_seq; mutex_exit(&zilog->zl_lock); (void) pthread_rwlock_unlock(&zd->zd_zilog_lock); } /* * This function is designed to simulate the operations that occur during a * mount/unmount operation. We hold the dataset across these operations in an * attempt to expose any implicit assumptions about ZIL management. */ void ztest_zil_remount(ztest_ds_t *zd, uint64_t id) { (void) id; objset_t *os = zd->zd_os; /* * We hold the ztest_vdev_lock so we don't cause problems with * other threads that wish to remove a log device, such as * ztest_device_removal(). */ mutex_enter(&ztest_vdev_lock); /* * We grab the zd_dirobj_lock to ensure that no other thread is * updating the zil (i.e. adding in-memory log records) and the * zd_zilog_lock to block any I/O. */ mutex_enter(&zd->zd_dirobj_lock); (void) pthread_rwlock_wrlock(&zd->zd_zilog_lock); /* zfsvfs_teardown() */ zil_close(zd->zd_zilog); /* zfsvfs_setup() */ VERIFY3P(zil_open(os, ztest_get_data, NULL), ==, zd->zd_zilog); zil_replay(os, zd, ztest_replay_vector); (void) pthread_rwlock_unlock(&zd->zd_zilog_lock); mutex_exit(&zd->zd_dirobj_lock); mutex_exit(&ztest_vdev_lock); } /* * Verify that we can't destroy an active pool, create an existing pool, * or create a pool with a bad vdev spec. */ void ztest_spa_create_destroy(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; ztest_shared_opts_t *zo = &ztest_opts; spa_t *spa; nvlist_t *nvroot; if (zo->zo_mmp_test) return; /* * Attempt to create using a bad file. */ nvroot = make_vdev_root("/dev/bogus", NULL, NULL, 0, 0, NULL, 0, 0, 1); VERIFY3U(ENOENT, ==, spa_create("ztest_bad_file", nvroot, NULL, NULL, NULL)); fnvlist_free(nvroot); /* * Attempt to create using a bad mirror. */ nvroot = make_vdev_root("/dev/bogus", NULL, NULL, 0, 0, NULL, 0, 2, 1); VERIFY3U(ENOENT, ==, spa_create("ztest_bad_mirror", nvroot, NULL, NULL, NULL)); fnvlist_free(nvroot); /* * Attempt to create an existing pool. It shouldn't matter * what's in the nvroot; we should fail with EEXIST. */ (void) pthread_rwlock_rdlock(&ztest_name_lock); nvroot = make_vdev_root("/dev/bogus", NULL, NULL, 0, 0, NULL, 0, 0, 1); VERIFY3U(EEXIST, ==, spa_create(zo->zo_pool, nvroot, NULL, NULL, NULL)); fnvlist_free(nvroot); /* * We open a reference to the spa and then we try to export it * expecting one of the following errors: * * EBUSY * Because of the reference we just opened. * * ZFS_ERR_EXPORT_IN_PROGRESS * For the case that there is another ztest thread doing * an export concurrently. */ VERIFY0(spa_open(zo->zo_pool, &spa, FTAG)); int error = spa_destroy(zo->zo_pool); if (error != EBUSY && error != ZFS_ERR_EXPORT_IN_PROGRESS) { fatal(B_FALSE, "spa_destroy(%s) returned unexpected value %d", spa->spa_name, error); } spa_close(spa, FTAG); (void) pthread_rwlock_unlock(&ztest_name_lock); } /* * Start and then stop the MMP threads to ensure the startup and shutdown code * works properly. Actual protection and property-related code tested via ZTS. */ void ztest_mmp_enable_disable(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; ztest_shared_opts_t *zo = &ztest_opts; spa_t *spa = ztest_spa; if (zo->zo_mmp_test) return; /* * Since enabling MMP involves setting a property, it could not be done * while the pool is suspended. */ if (spa_suspended(spa)) return; spa_config_enter(spa, SCL_CONFIG, FTAG, RW_READER); mutex_enter(&spa->spa_props_lock); zfs_multihost_fail_intervals = 0; if (!spa_multihost(spa)) { spa->spa_multihost = B_TRUE; mmp_thread_start(spa); } mutex_exit(&spa->spa_props_lock); spa_config_exit(spa, SCL_CONFIG, FTAG); txg_wait_synced(spa_get_dsl(spa), 0); mmp_signal_all_threads(); txg_wait_synced(spa_get_dsl(spa), 0); spa_config_enter(spa, SCL_CONFIG, FTAG, RW_READER); mutex_enter(&spa->spa_props_lock); if (spa_multihost(spa)) { mmp_thread_stop(spa); spa->spa_multihost = B_FALSE; } mutex_exit(&spa->spa_props_lock); spa_config_exit(spa, SCL_CONFIG, FTAG); } void ztest_spa_upgrade(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; spa_t *spa; uint64_t initial_version = SPA_VERSION_INITIAL; uint64_t version, newversion; nvlist_t *nvroot, *props; char *name; if (ztest_opts.zo_mmp_test) return; /* dRAID added after feature flags, skip upgrade test. */ if (strcmp(ztest_opts.zo_raid_type, VDEV_TYPE_DRAID) == 0) return; mutex_enter(&ztest_vdev_lock); name = kmem_asprintf("%s_upgrade", ztest_opts.zo_pool); /* * Clean up from previous runs. */ (void) spa_destroy(name); nvroot = make_vdev_root(NULL, NULL, name, ztest_opts.zo_vdev_size, 0, NULL, ztest_opts.zo_raid_children, ztest_opts.zo_mirrors, 1); /* * If we're configuring a RAIDZ device then make sure that the * initial version is capable of supporting that feature. */ switch (ztest_opts.zo_raid_parity) { case 0: case 1: initial_version = SPA_VERSION_INITIAL; break; case 2: initial_version = SPA_VERSION_RAIDZ2; break; case 3: initial_version = SPA_VERSION_RAIDZ3; break; } /* * Create a pool with a spa version that can be upgraded. Pick * a value between initial_version and SPA_VERSION_BEFORE_FEATURES. */ do { version = ztest_random_spa_version(initial_version); } while (version > SPA_VERSION_BEFORE_FEATURES); props = fnvlist_alloc(); fnvlist_add_uint64(props, zpool_prop_to_name(ZPOOL_PROP_VERSION), version); VERIFY0(spa_create(name, nvroot, props, NULL, NULL)); fnvlist_free(nvroot); fnvlist_free(props); VERIFY0(spa_open(name, &spa, FTAG)); VERIFY3U(spa_version(spa), ==, version); newversion = ztest_random_spa_version(version + 1); if (ztest_opts.zo_verbose >= 4) { (void) printf("upgrading spa version from " "%"PRIu64" to %"PRIu64"\n", version, newversion); } spa_upgrade(spa, newversion); VERIFY3U(spa_version(spa), >, version); VERIFY3U(spa_version(spa), ==, fnvlist_lookup_uint64(spa->spa_config, zpool_prop_to_name(ZPOOL_PROP_VERSION))); spa_close(spa, FTAG); kmem_strfree(name); mutex_exit(&ztest_vdev_lock); } static void ztest_spa_checkpoint(spa_t *spa) { ASSERT(MUTEX_HELD(&ztest_checkpoint_lock)); int error = spa_checkpoint(spa->spa_name); switch (error) { case 0: case ZFS_ERR_DEVRM_IN_PROGRESS: case ZFS_ERR_DISCARDING_CHECKPOINT: case ZFS_ERR_CHECKPOINT_EXISTS: break; case ENOSPC: ztest_record_enospc(FTAG); break; default: fatal(B_FALSE, "spa_checkpoint(%s) = %d", spa->spa_name, error); } } static void ztest_spa_discard_checkpoint(spa_t *spa) { ASSERT(MUTEX_HELD(&ztest_checkpoint_lock)); int error = spa_checkpoint_discard(spa->spa_name); switch (error) { case 0: case ZFS_ERR_DISCARDING_CHECKPOINT: case ZFS_ERR_NO_CHECKPOINT: break; default: fatal(B_FALSE, "spa_discard_checkpoint(%s) = %d", spa->spa_name, error); } } void ztest_spa_checkpoint_create_discard(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; spa_t *spa = ztest_spa; mutex_enter(&ztest_checkpoint_lock); if (ztest_random(2) == 0) { ztest_spa_checkpoint(spa); } else { ztest_spa_discard_checkpoint(spa); } mutex_exit(&ztest_checkpoint_lock); } static vdev_t * vdev_lookup_by_path(vdev_t *vd, const char *path) { vdev_t *mvd; int c; if (vd->vdev_path != NULL && strcmp(path, vd->vdev_path) == 0) return (vd); for (c = 0; c < vd->vdev_children; c++) if ((mvd = vdev_lookup_by_path(vd->vdev_child[c], path)) != NULL) return (mvd); return (NULL); } static int spa_num_top_vdevs(spa_t *spa) { vdev_t *rvd = spa->spa_root_vdev; ASSERT3U(spa_config_held(spa, SCL_VDEV, RW_READER), ==, SCL_VDEV); return (rvd->vdev_children); } /* * Verify that vdev_add() works as expected. */ void ztest_vdev_add_remove(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; ztest_shared_t *zs = ztest_shared; spa_t *spa = ztest_spa; uint64_t leaves; uint64_t guid; nvlist_t *nvroot; int error; if (ztest_opts.zo_mmp_test) return; mutex_enter(&ztest_vdev_lock); leaves = MAX(zs->zs_mirrors + zs->zs_splits, 1) * ztest_opts.zo_raid_children; spa_config_enter(spa, SCL_VDEV, FTAG, RW_READER); ztest_shared->zs_vdev_next_leaf = spa_num_top_vdevs(spa) * leaves; /* * If we have slogs then remove them 1/4 of the time. */ if (spa_has_slogs(spa) && ztest_random(4) == 0) { metaslab_group_t *mg; /* * find the first real slog in log allocation class */ mg = spa_log_class(spa)->mc_allocator[0].mca_rotor; while (!mg->mg_vd->vdev_islog) mg = mg->mg_next; guid = mg->mg_vd->vdev_guid; spa_config_exit(spa, SCL_VDEV, FTAG); /* * We have to grab the zs_name_lock as writer to * prevent a race between removing a slog (dmu_objset_find) * and destroying a dataset. Removing the slog will * grab a reference on the dataset which may cause * dsl_destroy_head() to fail with EBUSY thus * leaving the dataset in an inconsistent state. */ pthread_rwlock_wrlock(&ztest_name_lock); error = spa_vdev_remove(spa, guid, B_FALSE); pthread_rwlock_unlock(&ztest_name_lock); switch (error) { case 0: case EEXIST: /* Generic zil_reset() error */ case EBUSY: /* Replay required */ case EACCES: /* Crypto key not loaded */ case ZFS_ERR_CHECKPOINT_EXISTS: case ZFS_ERR_DISCARDING_CHECKPOINT: break; default: fatal(B_FALSE, "spa_vdev_remove() = %d", error); } } else { spa_config_exit(spa, SCL_VDEV, FTAG); /* * Make 1/4 of the devices be log devices */ nvroot = make_vdev_root(NULL, NULL, NULL, ztest_opts.zo_vdev_size, 0, (ztest_random(4) == 0) ? "log" : NULL, ztest_opts.zo_raid_children, zs->zs_mirrors, 1); error = spa_vdev_add(spa, nvroot); fnvlist_free(nvroot); switch (error) { case 0: break; case ENOSPC: ztest_record_enospc("spa_vdev_add"); break; default: fatal(B_FALSE, "spa_vdev_add() = %d", error); } } mutex_exit(&ztest_vdev_lock); } void ztest_vdev_class_add(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; ztest_shared_t *zs = ztest_shared; spa_t *spa = ztest_spa; uint64_t leaves; nvlist_t *nvroot; const char *class = (ztest_random(2) == 0) ? VDEV_ALLOC_BIAS_SPECIAL : VDEV_ALLOC_BIAS_DEDUP; int error; /* * By default add a special vdev 50% of the time */ if ((ztest_opts.zo_special_vdevs == ZTEST_VDEV_CLASS_OFF) || (ztest_opts.zo_special_vdevs == ZTEST_VDEV_CLASS_RND && ztest_random(2) == 0)) { return; } mutex_enter(&ztest_vdev_lock); /* Only test with mirrors */ if (zs->zs_mirrors < 2) { mutex_exit(&ztest_vdev_lock); return; } /* requires feature@allocation_classes */ if (!spa_feature_is_enabled(spa, SPA_FEATURE_ALLOCATION_CLASSES)) { mutex_exit(&ztest_vdev_lock); return; } leaves = MAX(zs->zs_mirrors + zs->zs_splits, 1) * ztest_opts.zo_raid_children; spa_config_enter(spa, SCL_VDEV, FTAG, RW_READER); ztest_shared->zs_vdev_next_leaf = spa_num_top_vdevs(spa) * leaves; spa_config_exit(spa, SCL_VDEV, FTAG); nvroot = make_vdev_root(NULL, NULL, NULL, ztest_opts.zo_vdev_size, 0, class, ztest_opts.zo_raid_children, zs->zs_mirrors, 1); error = spa_vdev_add(spa, nvroot); fnvlist_free(nvroot); if (error == ENOSPC) ztest_record_enospc("spa_vdev_add"); else if (error != 0) fatal(B_FALSE, "spa_vdev_add() = %d", error); /* * 50% of the time allow small blocks in the special class */ if (error == 0 && spa_special_class(spa)->mc_groups == 1 && ztest_random(2) == 0) { if (ztest_opts.zo_verbose >= 3) (void) printf("Enabling special VDEV small blocks\n"); (void) ztest_dsl_prop_set_uint64(zd->zd_name, ZFS_PROP_SPECIAL_SMALL_BLOCKS, 32768, B_FALSE); } mutex_exit(&ztest_vdev_lock); if (ztest_opts.zo_verbose >= 3) { metaslab_class_t *mc; if (strcmp(class, VDEV_ALLOC_BIAS_SPECIAL) == 0) mc = spa_special_class(spa); else mc = spa_dedup_class(spa); (void) printf("Added a %s mirrored vdev (of %d)\n", class, (int)mc->mc_groups); } } /* * Verify that adding/removing aux devices (l2arc, hot spare) works as expected. */ void ztest_vdev_aux_add_remove(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; ztest_shared_t *zs = ztest_shared; spa_t *spa = ztest_spa; vdev_t *rvd = spa->spa_root_vdev; spa_aux_vdev_t *sav; const char *aux; char *path; uint64_t guid = 0; int error, ignore_err = 0; if (ztest_opts.zo_mmp_test) return; path = umem_alloc(MAXPATHLEN, UMEM_NOFAIL); if (ztest_random(2) == 0) { sav = &spa->spa_spares; aux = ZPOOL_CONFIG_SPARES; } else { sav = &spa->spa_l2cache; aux = ZPOOL_CONFIG_L2CACHE; } mutex_enter(&ztest_vdev_lock); spa_config_enter(spa, SCL_VDEV, FTAG, RW_READER); if (sav->sav_count != 0 && ztest_random(4) == 0) { /* * Pick a random device to remove. */ vdev_t *svd = sav->sav_vdevs[ztest_random(sav->sav_count)]; /* dRAID spares cannot be removed; try anyways to see ENOTSUP */ if (strstr(svd->vdev_path, VDEV_TYPE_DRAID) != NULL) ignore_err = ENOTSUP; guid = svd->vdev_guid; } else { /* * Find an unused device we can add. */ zs->zs_vdev_aux = 0; for (;;) { int c; (void) snprintf(path, MAXPATHLEN, ztest_aux_template, ztest_opts.zo_dir, ztest_opts.zo_pool, aux, zs->zs_vdev_aux); for (c = 0; c < sav->sav_count; c++) if (strcmp(sav->sav_vdevs[c]->vdev_path, path) == 0) break; if (c == sav->sav_count && vdev_lookup_by_path(rvd, path) == NULL) break; zs->zs_vdev_aux++; } } spa_config_exit(spa, SCL_VDEV, FTAG); if (guid == 0) { /* * Add a new device. */ nvlist_t *nvroot = make_vdev_root(NULL, aux, NULL, (ztest_opts.zo_vdev_size * 5) / 4, 0, NULL, 0, 0, 1); error = spa_vdev_add(spa, nvroot); switch (error) { case 0: break; default: fatal(B_FALSE, "spa_vdev_add(%p) = %d", nvroot, error); } fnvlist_free(nvroot); } else { /* * Remove an existing device. Sometimes, dirty its * vdev state first to make sure we handle removal * of devices that have pending state changes. */ if (ztest_random(2) == 0) (void) vdev_online(spa, guid, 0, NULL); error = spa_vdev_remove(spa, guid, B_FALSE); switch (error) { case 0: case EBUSY: case ZFS_ERR_CHECKPOINT_EXISTS: case ZFS_ERR_DISCARDING_CHECKPOINT: break; default: if (error != ignore_err) fatal(B_FALSE, "spa_vdev_remove(%"PRIu64") = %d", guid, error); } } mutex_exit(&ztest_vdev_lock); umem_free(path, MAXPATHLEN); } /* * split a pool if it has mirror tlvdevs */ void ztest_split_pool(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; ztest_shared_t *zs = ztest_shared; spa_t *spa = ztest_spa; vdev_t *rvd = spa->spa_root_vdev; nvlist_t *tree, **child, *config, *split, **schild; uint_t c, children, schildren = 0, lastlogid = 0; int error = 0; if (ztest_opts.zo_mmp_test) return; mutex_enter(&ztest_vdev_lock); /* ensure we have a usable config; mirrors of raidz aren't supported */ if (zs->zs_mirrors < 3 || ztest_opts.zo_raid_children > 1) { mutex_exit(&ztest_vdev_lock); return; } /* clean up the old pool, if any */ (void) spa_destroy("splitp"); spa_config_enter(spa, SCL_VDEV, FTAG, RW_READER); /* generate a config from the existing config */ mutex_enter(&spa->spa_props_lock); tree = fnvlist_lookup_nvlist(spa->spa_config, ZPOOL_CONFIG_VDEV_TREE); mutex_exit(&spa->spa_props_lock); VERIFY0(nvlist_lookup_nvlist_array(tree, ZPOOL_CONFIG_CHILDREN, &child, &children)); schild = malloc(rvd->vdev_children * sizeof (nvlist_t *)); for (c = 0; c < children; c++) { vdev_t *tvd = rvd->vdev_child[c]; nvlist_t **mchild; uint_t mchildren; if (tvd->vdev_islog || tvd->vdev_ops == &vdev_hole_ops) { schild[schildren] = fnvlist_alloc(); fnvlist_add_string(schild[schildren], ZPOOL_CONFIG_TYPE, VDEV_TYPE_HOLE); fnvlist_add_uint64(schild[schildren], ZPOOL_CONFIG_IS_HOLE, 1); if (lastlogid == 0) lastlogid = schildren; ++schildren; continue; } lastlogid = 0; VERIFY0(nvlist_lookup_nvlist_array(child[c], ZPOOL_CONFIG_CHILDREN, &mchild, &mchildren)); schild[schildren++] = fnvlist_dup(mchild[0]); } /* OK, create a config that can be used to split */ split = fnvlist_alloc(); fnvlist_add_string(split, ZPOOL_CONFIG_TYPE, VDEV_TYPE_ROOT); fnvlist_add_nvlist_array(split, ZPOOL_CONFIG_CHILDREN, (const nvlist_t **)schild, lastlogid != 0 ? lastlogid : schildren); config = fnvlist_alloc(); fnvlist_add_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, split); for (c = 0; c < schildren; c++) fnvlist_free(schild[c]); free(schild); fnvlist_free(split); spa_config_exit(spa, SCL_VDEV, FTAG); (void) pthread_rwlock_wrlock(&ztest_name_lock); error = spa_vdev_split_mirror(spa, "splitp", config, NULL, B_FALSE); (void) pthread_rwlock_unlock(&ztest_name_lock); fnvlist_free(config); if (error == 0) { (void) printf("successful split - results:\n"); mutex_enter(&spa_namespace_lock); show_pool_stats(spa); show_pool_stats(spa_lookup("splitp")); mutex_exit(&spa_namespace_lock); ++zs->zs_splits; --zs->zs_mirrors; } mutex_exit(&ztest_vdev_lock); } /* * Verify that we can attach and detach devices. */ void ztest_vdev_attach_detach(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; ztest_shared_t *zs = ztest_shared; spa_t *spa = ztest_spa; spa_aux_vdev_t *sav = &spa->spa_spares; vdev_t *rvd = spa->spa_root_vdev; vdev_t *oldvd, *newvd, *pvd; nvlist_t *root; uint64_t leaves; uint64_t leaf, top; uint64_t ashift = ztest_get_ashift(); uint64_t oldguid, pguid; uint64_t oldsize, newsize; char *oldpath, *newpath; int replacing; int oldvd_has_siblings = B_FALSE; int newvd_is_spare = B_FALSE; int newvd_is_dspare = B_FALSE; int oldvd_is_log; int error, expected_error; if (ztest_opts.zo_mmp_test) return; oldpath = umem_alloc(MAXPATHLEN, UMEM_NOFAIL); newpath = umem_alloc(MAXPATHLEN, UMEM_NOFAIL); mutex_enter(&ztest_vdev_lock); leaves = MAX(zs->zs_mirrors, 1) * ztest_opts.zo_raid_children; spa_config_enter(spa, SCL_ALL, FTAG, RW_WRITER); /* * If a vdev is in the process of being removed, its removal may * finish while we are in progress, leading to an unexpected error * value. Don't bother trying to attach while we are in the middle * of removal. */ if (ztest_device_removal_active) { spa_config_exit(spa, SCL_ALL, FTAG); goto out; } /* * Decide whether to do an attach or a replace. */ replacing = ztest_random(2); /* * Pick a random top-level vdev. */ top = ztest_random_vdev_top(spa, B_TRUE); /* * Pick a random leaf within it. */ leaf = ztest_random(leaves); /* * Locate this vdev. */ oldvd = rvd->vdev_child[top]; /* pick a child from the mirror */ if (zs->zs_mirrors >= 1) { ASSERT3P(oldvd->vdev_ops, ==, &vdev_mirror_ops); ASSERT3U(oldvd->vdev_children, >=, zs->zs_mirrors); oldvd = oldvd->vdev_child[leaf / ztest_opts.zo_raid_children]; } /* pick a child out of the raidz group */ if (ztest_opts.zo_raid_children > 1) { if (strcmp(oldvd->vdev_ops->vdev_op_type, "raidz") == 0) ASSERT3P(oldvd->vdev_ops, ==, &vdev_raidz_ops); else ASSERT3P(oldvd->vdev_ops, ==, &vdev_draid_ops); ASSERT3U(oldvd->vdev_children, ==, ztest_opts.zo_raid_children); oldvd = oldvd->vdev_child[leaf % ztest_opts.zo_raid_children]; } /* * If we're already doing an attach or replace, oldvd may be a * mirror vdev -- in which case, pick a random child. */ while (oldvd->vdev_children != 0) { oldvd_has_siblings = B_TRUE; ASSERT3U(oldvd->vdev_children, >=, 2); oldvd = oldvd->vdev_child[ztest_random(oldvd->vdev_children)]; } oldguid = oldvd->vdev_guid; oldsize = vdev_get_min_asize(oldvd); oldvd_is_log = oldvd->vdev_top->vdev_islog; (void) strlcpy(oldpath, oldvd->vdev_path, MAXPATHLEN); pvd = oldvd->vdev_parent; pguid = pvd->vdev_guid; /* * If oldvd has siblings, then half of the time, detach it. Prior * to the detach the pool is scrubbed in order to prevent creating * unrepairable blocks as a result of the data corruption injection. */ if (oldvd_has_siblings && ztest_random(2) == 0) { spa_config_exit(spa, SCL_ALL, FTAG); error = ztest_scrub_impl(spa); if (error) goto out; error = spa_vdev_detach(spa, oldguid, pguid, B_FALSE); if (error != 0 && error != ENODEV && error != EBUSY && error != ENOTSUP && error != ZFS_ERR_CHECKPOINT_EXISTS && error != ZFS_ERR_DISCARDING_CHECKPOINT) fatal(B_FALSE, "detach (%s) returned %d", oldpath, error); goto out; } /* * For the new vdev, choose with equal probability between the two * standard paths (ending in either 'a' or 'b') or a random hot spare. */ if (sav->sav_count != 0 && ztest_random(3) == 0) { newvd = sav->sav_vdevs[ztest_random(sav->sav_count)]; newvd_is_spare = B_TRUE; if (newvd->vdev_ops == &vdev_draid_spare_ops) newvd_is_dspare = B_TRUE; (void) strlcpy(newpath, newvd->vdev_path, MAXPATHLEN); } else { (void) snprintf(newpath, MAXPATHLEN, ztest_dev_template, ztest_opts.zo_dir, ztest_opts.zo_pool, top * leaves + leaf); if (ztest_random(2) == 0) newpath[strlen(newpath) - 1] = 'b'; newvd = vdev_lookup_by_path(rvd, newpath); } if (newvd) { /* * Reopen to ensure the vdev's asize field isn't stale. */ vdev_reopen(newvd); newsize = vdev_get_min_asize(newvd); } else { /* * Make newsize a little bigger or smaller than oldsize. * If it's smaller, the attach should fail. * If it's larger, and we're doing a replace, * we should get dynamic LUN growth when we're done. */ newsize = 10 * oldsize / (9 + ztest_random(3)); } /* * If pvd is not a mirror or root, the attach should fail with ENOTSUP, * unless it's a replace; in that case any non-replacing parent is OK. * * If newvd is already part of the pool, it should fail with EBUSY. * * If newvd is too small, it should fail with EOVERFLOW. * * If newvd is a distributed spare and it's being attached to a * dRAID which is not its parent it should fail with EINVAL. */ if (pvd->vdev_ops != &vdev_mirror_ops && pvd->vdev_ops != &vdev_root_ops && (!replacing || pvd->vdev_ops == &vdev_replacing_ops || pvd->vdev_ops == &vdev_spare_ops)) expected_error = ENOTSUP; else if (newvd_is_spare && (!replacing || oldvd_is_log)) expected_error = ENOTSUP; else if (newvd == oldvd) expected_error = replacing ? 0 : EBUSY; else if (vdev_lookup_by_path(rvd, newpath) != NULL) expected_error = EBUSY; else if (!newvd_is_dspare && newsize < oldsize) expected_error = EOVERFLOW; else if (ashift > oldvd->vdev_top->vdev_ashift) expected_error = EDOM; else if (newvd_is_dspare && pvd != vdev_draid_spare_get_parent(newvd)) expected_error = ENOTSUP; else expected_error = 0; spa_config_exit(spa, SCL_ALL, FTAG); /* * Build the nvlist describing newpath. */ root = make_vdev_root(newpath, NULL, NULL, newvd == NULL ? newsize : 0, ashift, NULL, 0, 0, 1); /* * When supported select either a healing or sequential resilver. */ boolean_t rebuilding = B_FALSE; if (pvd->vdev_ops == &vdev_mirror_ops || pvd->vdev_ops == &vdev_root_ops) { rebuilding = !!ztest_random(2); } error = spa_vdev_attach(spa, oldguid, root, replacing, rebuilding); fnvlist_free(root); /* * If our parent was the replacing vdev, but the replace completed, * then instead of failing with ENOTSUP we may either succeed, * fail with ENODEV, or fail with EOVERFLOW. */ if (expected_error == ENOTSUP && (error == 0 || error == ENODEV || error == EOVERFLOW)) expected_error = error; /* * If someone grew the LUN, the replacement may be too small. */ if (error == EOVERFLOW || error == EBUSY) expected_error = error; if (error == ZFS_ERR_CHECKPOINT_EXISTS || error == ZFS_ERR_DISCARDING_CHECKPOINT || error == ZFS_ERR_RESILVER_IN_PROGRESS || error == ZFS_ERR_REBUILD_IN_PROGRESS) expected_error = error; if (error != expected_error && expected_error != EBUSY) { fatal(B_FALSE, "attach (%s %"PRIu64", %s %"PRIu64", %d) " "returned %d, expected %d", oldpath, oldsize, newpath, newsize, replacing, error, expected_error); } out: mutex_exit(&ztest_vdev_lock); umem_free(oldpath, MAXPATHLEN); umem_free(newpath, MAXPATHLEN); } void ztest_device_removal(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; spa_t *spa = ztest_spa; vdev_t *vd; uint64_t guid; int error; mutex_enter(&ztest_vdev_lock); if (ztest_device_removal_active) { mutex_exit(&ztest_vdev_lock); return; } /* * Remove a random top-level vdev and wait for removal to finish. */ spa_config_enter(spa, SCL_VDEV, FTAG, RW_READER); vd = vdev_lookup_top(spa, ztest_random_vdev_top(spa, B_FALSE)); guid = vd->vdev_guid; spa_config_exit(spa, SCL_VDEV, FTAG); error = spa_vdev_remove(spa, guid, B_FALSE); if (error == 0) { ztest_device_removal_active = B_TRUE; mutex_exit(&ztest_vdev_lock); /* * spa->spa_vdev_removal is created in a sync task that * is initiated via dsl_sync_task_nowait(). Since the * task may not run before spa_vdev_remove() returns, we * must wait at least 1 txg to ensure that the removal * struct has been created. */ txg_wait_synced(spa_get_dsl(spa), 0); while (spa->spa_removing_phys.sr_state == DSS_SCANNING) txg_wait_synced(spa_get_dsl(spa), 0); } else { mutex_exit(&ztest_vdev_lock); return; } /* * The pool needs to be scrubbed after completing device removal. * Failure to do so may result in checksum errors due to the * strategy employed by ztest_fault_inject() when selecting which * offset are redundant and can be damaged. */ error = spa_scan(spa, POOL_SCAN_SCRUB); if (error == 0) { while (dsl_scan_scrubbing(spa_get_dsl(spa))) txg_wait_synced(spa_get_dsl(spa), 0); } mutex_enter(&ztest_vdev_lock); ztest_device_removal_active = B_FALSE; mutex_exit(&ztest_vdev_lock); } /* * Callback function which expands the physical size of the vdev. */ static vdev_t * grow_vdev(vdev_t *vd, void *arg) { spa_t *spa __maybe_unused = vd->vdev_spa; size_t *newsize = arg; size_t fsize; int fd; ASSERT3S(spa_config_held(spa, SCL_STATE, RW_READER), ==, SCL_STATE); ASSERT(vd->vdev_ops->vdev_op_leaf); if ((fd = open(vd->vdev_path, O_RDWR)) == -1) return (vd); fsize = lseek(fd, 0, SEEK_END); VERIFY0(ftruncate(fd, *newsize)); if (ztest_opts.zo_verbose >= 6) { (void) printf("%s grew from %lu to %lu bytes\n", vd->vdev_path, (ulong_t)fsize, (ulong_t)*newsize); } (void) close(fd); return (NULL); } /* * Callback function which expands a given vdev by calling vdev_online(). */ static vdev_t * online_vdev(vdev_t *vd, void *arg) { (void) arg; spa_t *spa = vd->vdev_spa; vdev_t *tvd = vd->vdev_top; uint64_t guid = vd->vdev_guid; uint64_t generation = spa->spa_config_generation + 1; vdev_state_t newstate = VDEV_STATE_UNKNOWN; int error; ASSERT3S(spa_config_held(spa, SCL_STATE, RW_READER), ==, SCL_STATE); ASSERT(vd->vdev_ops->vdev_op_leaf); /* Calling vdev_online will initialize the new metaslabs */ spa_config_exit(spa, SCL_STATE, spa); error = vdev_online(spa, guid, ZFS_ONLINE_EXPAND, &newstate); spa_config_enter(spa, SCL_STATE, spa, RW_READER); /* * If vdev_online returned an error or the underlying vdev_open * failed then we abort the expand. The only way to know that * vdev_open fails is by checking the returned newstate. */ if (error || newstate != VDEV_STATE_HEALTHY) { if (ztest_opts.zo_verbose >= 5) { (void) printf("Unable to expand vdev, state %u, " "error %d\n", newstate, error); } return (vd); } ASSERT3U(newstate, ==, VDEV_STATE_HEALTHY); /* * Since we dropped the lock we need to ensure that we're * still talking to the original vdev. It's possible this * vdev may have been detached/replaced while we were * trying to online it. */ if (generation != spa->spa_config_generation) { if (ztest_opts.zo_verbose >= 5) { (void) printf("vdev configuration has changed, " "guid %"PRIu64", state %"PRIu64", " "expected gen %"PRIu64", got gen %"PRIu64"\n", guid, tvd->vdev_state, generation, spa->spa_config_generation); } return (vd); } return (NULL); } /* * Traverse the vdev tree calling the supplied function. * We continue to walk the tree until we either have walked all * children or we receive a non-NULL return from the callback. * If a NULL callback is passed, then we just return back the first * leaf vdev we encounter. */ static vdev_t * vdev_walk_tree(vdev_t *vd, vdev_t *(*func)(vdev_t *, void *), void *arg) { uint_t c; if (vd->vdev_ops->vdev_op_leaf) { if (func == NULL) return (vd); else return (func(vd, arg)); } for (c = 0; c < vd->vdev_children; c++) { vdev_t *cvd = vd->vdev_child[c]; if ((cvd = vdev_walk_tree(cvd, func, arg)) != NULL) return (cvd); } return (NULL); } /* * Verify that dynamic LUN growth works as expected. */ void ztest_vdev_LUN_growth(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; spa_t *spa = ztest_spa; vdev_t *vd, *tvd; metaslab_class_t *mc; metaslab_group_t *mg; size_t psize, newsize; uint64_t top; uint64_t old_class_space, new_class_space, old_ms_count, new_ms_count; mutex_enter(&ztest_checkpoint_lock); mutex_enter(&ztest_vdev_lock); spa_config_enter(spa, SCL_STATE, spa, RW_READER); /* * If there is a vdev removal in progress, it could complete while * we are running, in which case we would not be able to verify * that the metaslab_class space increased (because it decreases * when the device removal completes). */ if (ztest_device_removal_active) { spa_config_exit(spa, SCL_STATE, spa); mutex_exit(&ztest_vdev_lock); mutex_exit(&ztest_checkpoint_lock); return; } top = ztest_random_vdev_top(spa, B_TRUE); tvd = spa->spa_root_vdev->vdev_child[top]; mg = tvd->vdev_mg; mc = mg->mg_class; old_ms_count = tvd->vdev_ms_count; old_class_space = metaslab_class_get_space(mc); /* * Determine the size of the first leaf vdev associated with * our top-level device. */ vd = vdev_walk_tree(tvd, NULL, NULL); ASSERT3P(vd, !=, NULL); ASSERT(vd->vdev_ops->vdev_op_leaf); psize = vd->vdev_psize; /* * We only try to expand the vdev if it's healthy, less than 4x its * original size, and it has a valid psize. */ if (tvd->vdev_state != VDEV_STATE_HEALTHY || psize == 0 || psize >= 4 * ztest_opts.zo_vdev_size) { spa_config_exit(spa, SCL_STATE, spa); mutex_exit(&ztest_vdev_lock); mutex_exit(&ztest_checkpoint_lock); return; } ASSERT3U(psize, >, 0); newsize = psize + MAX(psize / 8, SPA_MAXBLOCKSIZE); ASSERT3U(newsize, >, psize); if (ztest_opts.zo_verbose >= 6) { (void) printf("Expanding LUN %s from %lu to %lu\n", vd->vdev_path, (ulong_t)psize, (ulong_t)newsize); } /* * Growing the vdev is a two step process: * 1). expand the physical size (i.e. relabel) * 2). online the vdev to create the new metaslabs */ if (vdev_walk_tree(tvd, grow_vdev, &newsize) != NULL || vdev_walk_tree(tvd, online_vdev, NULL) != NULL || tvd->vdev_state != VDEV_STATE_HEALTHY) { if (ztest_opts.zo_verbose >= 5) { (void) printf("Could not expand LUN because " "the vdev configuration changed.\n"); } spa_config_exit(spa, SCL_STATE, spa); mutex_exit(&ztest_vdev_lock); mutex_exit(&ztest_checkpoint_lock); return; } spa_config_exit(spa, SCL_STATE, spa); /* * Expanding the LUN will update the config asynchronously, * thus we must wait for the async thread to complete any * pending tasks before proceeding. */ for (;;) { boolean_t done; mutex_enter(&spa->spa_async_lock); done = (spa->spa_async_thread == NULL && !spa->spa_async_tasks); mutex_exit(&spa->spa_async_lock); if (done) break; txg_wait_synced(spa_get_dsl(spa), 0); (void) poll(NULL, 0, 100); } spa_config_enter(spa, SCL_STATE, spa, RW_READER); tvd = spa->spa_root_vdev->vdev_child[top]; new_ms_count = tvd->vdev_ms_count; new_class_space = metaslab_class_get_space(mc); if (tvd->vdev_mg != mg || mg->mg_class != mc) { if (ztest_opts.zo_verbose >= 5) { (void) printf("Could not verify LUN expansion due to " "intervening vdev offline or remove.\n"); } spa_config_exit(spa, SCL_STATE, spa); mutex_exit(&ztest_vdev_lock); mutex_exit(&ztest_checkpoint_lock); return; } /* * Make sure we were able to grow the vdev. */ if (new_ms_count <= old_ms_count) { fatal(B_FALSE, "LUN expansion failed: ms_count %"PRIu64" < %"PRIu64"\n", old_ms_count, new_ms_count); } /* * Make sure we were able to grow the pool. */ if (new_class_space <= old_class_space) { fatal(B_FALSE, "LUN expansion failed: class_space %"PRIu64" < %"PRIu64"\n", old_class_space, new_class_space); } if (ztest_opts.zo_verbose >= 5) { char oldnumbuf[NN_NUMBUF_SZ], newnumbuf[NN_NUMBUF_SZ]; nicenum(old_class_space, oldnumbuf, sizeof (oldnumbuf)); nicenum(new_class_space, newnumbuf, sizeof (newnumbuf)); (void) printf("%s grew from %s to %s\n", spa->spa_name, oldnumbuf, newnumbuf); } spa_config_exit(spa, SCL_STATE, spa); mutex_exit(&ztest_vdev_lock); mutex_exit(&ztest_checkpoint_lock); } /* * Verify that dmu_objset_{create,destroy,open,close} work as expected. */ static void ztest_objset_create_cb(objset_t *os, void *arg, cred_t *cr, dmu_tx_t *tx) { (void) arg, (void) cr; /* * Create the objects common to all ztest datasets. */ VERIFY0(zap_create_claim(os, ZTEST_DIROBJ, DMU_OT_ZAP_OTHER, DMU_OT_NONE, 0, tx)); } static int ztest_dataset_create(char *dsname) { int err; uint64_t rand; dsl_crypto_params_t *dcp = NULL; /* * 50% of the time, we create encrypted datasets * using a random cipher suite and a hard-coded * wrapping key. */ rand = ztest_random(2); if (rand != 0) { nvlist_t *crypto_args = fnvlist_alloc(); nvlist_t *props = fnvlist_alloc(); /* slight bias towards the default cipher suite */ rand = ztest_random(ZIO_CRYPT_FUNCTIONS); if (rand < ZIO_CRYPT_AES_128_CCM) rand = ZIO_CRYPT_ON; fnvlist_add_uint64(props, zfs_prop_to_name(ZFS_PROP_ENCRYPTION), rand); fnvlist_add_uint8_array(crypto_args, "wkeydata", (uint8_t *)ztest_wkeydata, WRAPPING_KEY_LEN); /* * These parameters aren't really used by the kernel. They * are simply stored so that userspace knows how to load * the wrapping key. */ fnvlist_add_uint64(props, zfs_prop_to_name(ZFS_PROP_KEYFORMAT), ZFS_KEYFORMAT_RAW); fnvlist_add_string(props, zfs_prop_to_name(ZFS_PROP_KEYLOCATION), "prompt"); fnvlist_add_uint64(props, zfs_prop_to_name(ZFS_PROP_PBKDF2_SALT), 0ULL); fnvlist_add_uint64(props, zfs_prop_to_name(ZFS_PROP_PBKDF2_ITERS), 0ULL); VERIFY0(dsl_crypto_params_create_nvlist(DCP_CMD_NONE, props, crypto_args, &dcp)); /* * Cycle through all available encryption implementations * to verify interoperability. */ VERIFY0(gcm_impl_set("cycle")); VERIFY0(aes_impl_set("cycle")); fnvlist_free(crypto_args); fnvlist_free(props); } err = dmu_objset_create(dsname, DMU_OST_OTHER, 0, dcp, ztest_objset_create_cb, NULL); dsl_crypto_params_free(dcp, !!err); rand = ztest_random(100); if (err || rand < 80) return (err); if (ztest_opts.zo_verbose >= 5) (void) printf("Setting dataset %s to sync always\n", dsname); return (ztest_dsl_prop_set_uint64(dsname, ZFS_PROP_SYNC, ZFS_SYNC_ALWAYS, B_FALSE)); } static int ztest_objset_destroy_cb(const char *name, void *arg) { (void) arg; objset_t *os; dmu_object_info_t doi; int error; /* * Verify that the dataset contains a directory object. */ VERIFY0(ztest_dmu_objset_own(name, DMU_OST_OTHER, B_TRUE, B_TRUE, FTAG, &os)); error = dmu_object_info(os, ZTEST_DIROBJ, &doi); if (error != ENOENT) { /* We could have crashed in the middle of destroying it */ ASSERT0(error); ASSERT3U(doi.doi_type, ==, DMU_OT_ZAP_OTHER); ASSERT3S(doi.doi_physical_blocks_512, >=, 0); } dmu_objset_disown(os, B_TRUE, FTAG); /* * Destroy the dataset. */ if (strchr(name, '@') != NULL) { error = dsl_destroy_snapshot(name, B_TRUE); if (error != ECHRNG) { /* * The program was executed, but encountered a runtime * error, such as insufficient slop, or a hold on the * dataset. */ ASSERT0(error); } } else { error = dsl_destroy_head(name); if (error == ENOSPC) { /* There could be checkpoint or insufficient slop */ ztest_record_enospc(FTAG); } else if (error != EBUSY) { /* There could be a hold on this dataset */ ASSERT0(error); } } return (0); } static boolean_t ztest_snapshot_create(char *osname, uint64_t id) { char snapname[ZFS_MAX_DATASET_NAME_LEN]; int error; (void) snprintf(snapname, sizeof (snapname), "%"PRIu64"", id); error = dmu_objset_snapshot_one(osname, snapname); if (error == ENOSPC) { ztest_record_enospc(FTAG); return (B_FALSE); } if (error != 0 && error != EEXIST) { fatal(B_FALSE, "ztest_snapshot_create(%s@%s) = %d", osname, snapname, error); } return (B_TRUE); } static boolean_t ztest_snapshot_destroy(char *osname, uint64_t id) { char snapname[ZFS_MAX_DATASET_NAME_LEN]; int error; (void) snprintf(snapname, sizeof (snapname), "%s@%"PRIu64"", osname, id); error = dsl_destroy_snapshot(snapname, B_FALSE); if (error != 0 && error != ENOENT) fatal(B_FALSE, "ztest_snapshot_destroy(%s) = %d", snapname, error); return (B_TRUE); } void ztest_dmu_objset_create_destroy(ztest_ds_t *zd, uint64_t id) { (void) zd; ztest_ds_t *zdtmp; int iters; int error; objset_t *os, *os2; char name[ZFS_MAX_DATASET_NAME_LEN]; zilog_t *zilog; int i; zdtmp = umem_alloc(sizeof (ztest_ds_t), UMEM_NOFAIL); (void) pthread_rwlock_rdlock(&ztest_name_lock); (void) snprintf(name, sizeof (name), "%s/temp_%"PRIu64"", ztest_opts.zo_pool, id); /* * If this dataset exists from a previous run, process its replay log * half of the time. If we don't replay it, then dsl_destroy_head() * (invoked from ztest_objset_destroy_cb()) should just throw it away. */ if (ztest_random(2) == 0 && ztest_dmu_objset_own(name, DMU_OST_OTHER, B_FALSE, B_TRUE, FTAG, &os) == 0) { ztest_zd_init(zdtmp, NULL, os); zil_replay(os, zdtmp, ztest_replay_vector); ztest_zd_fini(zdtmp); dmu_objset_disown(os, B_TRUE, FTAG); } /* * There may be an old instance of the dataset we're about to * create lying around from a previous run. If so, destroy it * and all of its snapshots. */ (void) dmu_objset_find(name, ztest_objset_destroy_cb, NULL, DS_FIND_CHILDREN | DS_FIND_SNAPSHOTS); /* * Verify that the destroyed dataset is no longer in the namespace. */ VERIFY3U(ENOENT, ==, ztest_dmu_objset_own(name, DMU_OST_OTHER, B_TRUE, B_TRUE, FTAG, &os)); /* * Verify that we can create a new dataset. */ error = ztest_dataset_create(name); if (error) { if (error == ENOSPC) { ztest_record_enospc(FTAG); goto out; } fatal(B_FALSE, "dmu_objset_create(%s) = %d", name, error); } VERIFY0(ztest_dmu_objset_own(name, DMU_OST_OTHER, B_FALSE, B_TRUE, FTAG, &os)); ztest_zd_init(zdtmp, NULL, os); /* * Open the intent log for it. */ zilog = zil_open(os, ztest_get_data, NULL); /* * Put some objects in there, do a little I/O to them, * and randomly take a couple of snapshots along the way. */ iters = ztest_random(5); for (i = 0; i < iters; i++) { ztest_dmu_object_alloc_free(zdtmp, id); if (ztest_random(iters) == 0) (void) ztest_snapshot_create(name, i); } /* * Verify that we cannot create an existing dataset. */ VERIFY3U(EEXIST, ==, dmu_objset_create(name, DMU_OST_OTHER, 0, NULL, NULL, NULL)); /* * Verify that we can hold an objset that is also owned. */ VERIFY0(dmu_objset_hold(name, FTAG, &os2)); dmu_objset_rele(os2, FTAG); /* * Verify that we cannot own an objset that is already owned. */ VERIFY3U(EBUSY, ==, ztest_dmu_objset_own(name, DMU_OST_OTHER, B_FALSE, B_TRUE, FTAG, &os2)); zil_close(zilog); dmu_objset_disown(os, B_TRUE, FTAG); ztest_zd_fini(zdtmp); out: (void) pthread_rwlock_unlock(&ztest_name_lock); umem_free(zdtmp, sizeof (ztest_ds_t)); } /* * Verify that dmu_snapshot_{create,destroy,open,close} work as expected. */ void ztest_dmu_snapshot_create_destroy(ztest_ds_t *zd, uint64_t id) { (void) pthread_rwlock_rdlock(&ztest_name_lock); (void) ztest_snapshot_destroy(zd->zd_name, id); (void) ztest_snapshot_create(zd->zd_name, id); (void) pthread_rwlock_unlock(&ztest_name_lock); } /* * Cleanup non-standard snapshots and clones. */ static void ztest_dsl_dataset_cleanup(char *osname, uint64_t id) { char *snap1name; char *clone1name; char *snap2name; char *clone2name; char *snap3name; int error; snap1name = umem_alloc(ZFS_MAX_DATASET_NAME_LEN, UMEM_NOFAIL); clone1name = umem_alloc(ZFS_MAX_DATASET_NAME_LEN, UMEM_NOFAIL); snap2name = umem_alloc(ZFS_MAX_DATASET_NAME_LEN, UMEM_NOFAIL); clone2name = umem_alloc(ZFS_MAX_DATASET_NAME_LEN, UMEM_NOFAIL); snap3name = umem_alloc(ZFS_MAX_DATASET_NAME_LEN, UMEM_NOFAIL); (void) snprintf(snap1name, ZFS_MAX_DATASET_NAME_LEN, "%s@s1_%"PRIu64"", osname, id); (void) snprintf(clone1name, ZFS_MAX_DATASET_NAME_LEN, "%s/c1_%"PRIu64"", osname, id); (void) snprintf(snap2name, ZFS_MAX_DATASET_NAME_LEN, "%s@s2_%"PRIu64"", clone1name, id); (void) snprintf(clone2name, ZFS_MAX_DATASET_NAME_LEN, "%s/c2_%"PRIu64"", osname, id); (void) snprintf(snap3name, ZFS_MAX_DATASET_NAME_LEN, "%s@s3_%"PRIu64"", clone1name, id); error = dsl_destroy_head(clone2name); if (error && error != ENOENT) fatal(B_FALSE, "dsl_destroy_head(%s) = %d", clone2name, error); error = dsl_destroy_snapshot(snap3name, B_FALSE); if (error && error != ENOENT) fatal(B_FALSE, "dsl_destroy_snapshot(%s) = %d", snap3name, error); error = dsl_destroy_snapshot(snap2name, B_FALSE); if (error && error != ENOENT) fatal(B_FALSE, "dsl_destroy_snapshot(%s) = %d", snap2name, error); error = dsl_destroy_head(clone1name); if (error && error != ENOENT) fatal(B_FALSE, "dsl_destroy_head(%s) = %d", clone1name, error); error = dsl_destroy_snapshot(snap1name, B_FALSE); if (error && error != ENOENT) fatal(B_FALSE, "dsl_destroy_snapshot(%s) = %d", snap1name, error); umem_free(snap1name, ZFS_MAX_DATASET_NAME_LEN); umem_free(clone1name, ZFS_MAX_DATASET_NAME_LEN); umem_free(snap2name, ZFS_MAX_DATASET_NAME_LEN); umem_free(clone2name, ZFS_MAX_DATASET_NAME_LEN); umem_free(snap3name, ZFS_MAX_DATASET_NAME_LEN); } /* * Verify dsl_dataset_promote handles EBUSY */ void ztest_dsl_dataset_promote_busy(ztest_ds_t *zd, uint64_t id) { objset_t *os; char *snap1name; char *clone1name; char *snap2name; char *clone2name; char *snap3name; char *osname = zd->zd_name; int error; snap1name = umem_alloc(ZFS_MAX_DATASET_NAME_LEN, UMEM_NOFAIL); clone1name = umem_alloc(ZFS_MAX_DATASET_NAME_LEN, UMEM_NOFAIL); snap2name = umem_alloc(ZFS_MAX_DATASET_NAME_LEN, UMEM_NOFAIL); clone2name = umem_alloc(ZFS_MAX_DATASET_NAME_LEN, UMEM_NOFAIL); snap3name = umem_alloc(ZFS_MAX_DATASET_NAME_LEN, UMEM_NOFAIL); (void) pthread_rwlock_rdlock(&ztest_name_lock); ztest_dsl_dataset_cleanup(osname, id); (void) snprintf(snap1name, ZFS_MAX_DATASET_NAME_LEN, "%s@s1_%"PRIu64"", osname, id); (void) snprintf(clone1name, ZFS_MAX_DATASET_NAME_LEN, "%s/c1_%"PRIu64"", osname, id); (void) snprintf(snap2name, ZFS_MAX_DATASET_NAME_LEN, "%s@s2_%"PRIu64"", clone1name, id); (void) snprintf(clone2name, ZFS_MAX_DATASET_NAME_LEN, "%s/c2_%"PRIu64"", osname, id); (void) snprintf(snap3name, ZFS_MAX_DATASET_NAME_LEN, "%s@s3_%"PRIu64"", clone1name, id); error = dmu_objset_snapshot_one(osname, strchr(snap1name, '@') + 1); if (error && error != EEXIST) { if (error == ENOSPC) { ztest_record_enospc(FTAG); goto out; } fatal(B_FALSE, "dmu_take_snapshot(%s) = %d", snap1name, error); } error = dmu_objset_clone(clone1name, snap1name); if (error) { if (error == ENOSPC) { ztest_record_enospc(FTAG); goto out; } fatal(B_FALSE, "dmu_objset_create(%s) = %d", clone1name, error); } error = dmu_objset_snapshot_one(clone1name, strchr(snap2name, '@') + 1); if (error && error != EEXIST) { if (error == ENOSPC) { ztest_record_enospc(FTAG); goto out; } fatal(B_FALSE, "dmu_open_snapshot(%s) = %d", snap2name, error); } error = dmu_objset_snapshot_one(clone1name, strchr(snap3name, '@') + 1); if (error && error != EEXIST) { if (error == ENOSPC) { ztest_record_enospc(FTAG); goto out; } fatal(B_FALSE, "dmu_open_snapshot(%s) = %d", snap3name, error); } error = dmu_objset_clone(clone2name, snap3name); if (error) { if (error == ENOSPC) { ztest_record_enospc(FTAG); goto out; } fatal(B_FALSE, "dmu_objset_create(%s) = %d", clone2name, error); } error = ztest_dmu_objset_own(snap2name, DMU_OST_ANY, B_TRUE, B_TRUE, FTAG, &os); if (error) fatal(B_FALSE, "dmu_objset_own(%s) = %d", snap2name, error); error = dsl_dataset_promote(clone2name, NULL); if (error == ENOSPC) { dmu_objset_disown(os, B_TRUE, FTAG); ztest_record_enospc(FTAG); goto out; } if (error != EBUSY) fatal(B_FALSE, "dsl_dataset_promote(%s), %d, not EBUSY", clone2name, error); dmu_objset_disown(os, B_TRUE, FTAG); out: ztest_dsl_dataset_cleanup(osname, id); (void) pthread_rwlock_unlock(&ztest_name_lock); umem_free(snap1name, ZFS_MAX_DATASET_NAME_LEN); umem_free(clone1name, ZFS_MAX_DATASET_NAME_LEN); umem_free(snap2name, ZFS_MAX_DATASET_NAME_LEN); umem_free(clone2name, ZFS_MAX_DATASET_NAME_LEN); umem_free(snap3name, ZFS_MAX_DATASET_NAME_LEN); } #undef OD_ARRAY_SIZE #define OD_ARRAY_SIZE 4 /* * Verify that dmu_object_{alloc,free} work as expected. */ void ztest_dmu_object_alloc_free(ztest_ds_t *zd, uint64_t id) { ztest_od_t *od; int batchsize; int size; int b; size = sizeof (ztest_od_t) * OD_ARRAY_SIZE; od = umem_alloc(size, UMEM_NOFAIL); batchsize = OD_ARRAY_SIZE; for (b = 0; b < batchsize; b++) ztest_od_init(od + b, id, FTAG, b, DMU_OT_UINT64_OTHER, 0, 0, 0); /* * Destroy the previous batch of objects, create a new batch, * and do some I/O on the new objects. */ if (ztest_object_init(zd, od, size, B_TRUE) != 0) return; while (ztest_random(4 * batchsize) != 0) ztest_io(zd, od[ztest_random(batchsize)].od_object, ztest_random(ZTEST_RANGE_LOCKS) << SPA_MAXBLOCKSHIFT); umem_free(od, size); } /* * Rewind the global allocator to verify object allocation backfilling. */ void ztest_dmu_object_next_chunk(ztest_ds_t *zd, uint64_t id) { (void) id; objset_t *os = zd->zd_os; uint_t dnodes_per_chunk = 1 << dmu_object_alloc_chunk_shift; uint64_t object; /* * Rewind the global allocator randomly back to a lower object number * to force backfilling and reclamation of recently freed dnodes. */ mutex_enter(&os->os_obj_lock); object = ztest_random(os->os_obj_next_chunk); os->os_obj_next_chunk = P2ALIGN(object, dnodes_per_chunk); mutex_exit(&os->os_obj_lock); } #undef OD_ARRAY_SIZE #define OD_ARRAY_SIZE 2 /* * Verify that dmu_{read,write} work as expected. */ void ztest_dmu_read_write(ztest_ds_t *zd, uint64_t id) { int size; ztest_od_t *od; objset_t *os = zd->zd_os; size = sizeof (ztest_od_t) * OD_ARRAY_SIZE; od = umem_alloc(size, UMEM_NOFAIL); dmu_tx_t *tx; int freeit, error; uint64_t i, n, s, txg; bufwad_t *packbuf, *bigbuf, *pack, *bigH, *bigT; uint64_t packobj, packoff, packsize, bigobj, bigoff, bigsize; uint64_t chunksize = (1000 + ztest_random(1000)) * sizeof (uint64_t); uint64_t regions = 997; uint64_t stride = 123456789ULL; uint64_t width = 40; int free_percent = 5; /* * This test uses two objects, packobj and bigobj, that are always * updated together (i.e. in the same tx) so that their contents are * in sync and can be compared. Their contents relate to each other * in a simple way: packobj is a dense array of 'bufwad' structures, * while bigobj is a sparse array of the same bufwads. Specifically, * for any index n, there are three bufwads that should be identical: * * packobj, at offset n * sizeof (bufwad_t) * bigobj, at the head of the nth chunk * bigobj, at the tail of the nth chunk * * The chunk size is arbitrary. It doesn't have to be a power of two, * and it doesn't have any relation to the object blocksize. * The only requirement is that it can hold at least two bufwads. * * Normally, we write the bufwad to each of these locations. * However, free_percent of the time we instead write zeroes to * packobj and perform a dmu_free_range() on bigobj. By comparing * bigobj to packobj, we can verify that the DMU is correctly * tracking which parts of an object are allocated and free, * and that the contents of the allocated blocks are correct. */ /* * Read the directory info. If it's the first time, set things up. */ ztest_od_init(od, id, FTAG, 0, DMU_OT_UINT64_OTHER, 0, 0, chunksize); ztest_od_init(od + 1, id, FTAG, 1, DMU_OT_UINT64_OTHER, 0, 0, chunksize); if (ztest_object_init(zd, od, size, B_FALSE) != 0) { umem_free(od, size); return; } bigobj = od[0].od_object; packobj = od[1].od_object; chunksize = od[0].od_gen; ASSERT3U(chunksize, ==, od[1].od_gen); /* * Prefetch a random chunk of the big object. * Our aim here is to get some async reads in flight * for blocks that we may free below; the DMU should * handle this race correctly. */ n = ztest_random(regions) * stride + ztest_random(width); s = 1 + ztest_random(2 * width - 1); dmu_prefetch(os, bigobj, 0, n * chunksize, s * chunksize, ZIO_PRIORITY_SYNC_READ); /* * Pick a random index and compute the offsets into packobj and bigobj. */ n = ztest_random(regions) * stride + ztest_random(width); s = 1 + ztest_random(width - 1); packoff = n * sizeof (bufwad_t); packsize = s * sizeof (bufwad_t); bigoff = n * chunksize; bigsize = s * chunksize; packbuf = umem_alloc(packsize, UMEM_NOFAIL); bigbuf = umem_alloc(bigsize, UMEM_NOFAIL); /* * free_percent of the time, free a range of bigobj rather than * overwriting it. */ freeit = (ztest_random(100) < free_percent); /* * Read the current contents of our objects. */ error = dmu_read(os, packobj, packoff, packsize, packbuf, DMU_READ_PREFETCH); ASSERT0(error); error = dmu_read(os, bigobj, bigoff, bigsize, bigbuf, DMU_READ_PREFETCH); ASSERT0(error); /* * Get a tx for the mods to both packobj and bigobj. */ tx = dmu_tx_create(os); dmu_tx_hold_write(tx, packobj, packoff, packsize); if (freeit) dmu_tx_hold_free(tx, bigobj, bigoff, bigsize); else dmu_tx_hold_write(tx, bigobj, bigoff, bigsize); /* This accounts for setting the checksum/compression. */ dmu_tx_hold_bonus(tx, bigobj); txg = ztest_tx_assign(tx, TXG_MIGHTWAIT, FTAG); if (txg == 0) { umem_free(packbuf, packsize); umem_free(bigbuf, bigsize); umem_free(od, size); return; } enum zio_checksum cksum; do { cksum = (enum zio_checksum) ztest_random_dsl_prop(ZFS_PROP_CHECKSUM); } while (cksum >= ZIO_CHECKSUM_LEGACY_FUNCTIONS); dmu_object_set_checksum(os, bigobj, cksum, tx); enum zio_compress comp; do { comp = (enum zio_compress) ztest_random_dsl_prop(ZFS_PROP_COMPRESSION); } while (comp >= ZIO_COMPRESS_LEGACY_FUNCTIONS); dmu_object_set_compress(os, bigobj, comp, tx); /* * For each index from n to n + s, verify that the existing bufwad * in packobj matches the bufwads at the head and tail of the * corresponding chunk in bigobj. Then update all three bufwads * with the new values we want to write out. */ for (i = 0; i < s; i++) { /* LINTED */ pack = (bufwad_t *)((char *)packbuf + i * sizeof (bufwad_t)); /* LINTED */ bigH = (bufwad_t *)((char *)bigbuf + i * chunksize); /* LINTED */ bigT = (bufwad_t *)((char *)bigH + chunksize) - 1; ASSERT3U((uintptr_t)bigH - (uintptr_t)bigbuf, <, bigsize); ASSERT3U((uintptr_t)bigT - (uintptr_t)bigbuf, <, bigsize); if (pack->bw_txg > txg) fatal(B_FALSE, "future leak: got %"PRIx64", open txg is %"PRIx64"", pack->bw_txg, txg); if (pack->bw_data != 0 && pack->bw_index != n + i) fatal(B_FALSE, "wrong index: " "got %"PRIx64", wanted %"PRIx64"+%"PRIx64"", pack->bw_index, n, i); if (memcmp(pack, bigH, sizeof (bufwad_t)) != 0) fatal(B_FALSE, "pack/bigH mismatch in %p/%p", pack, bigH); if (memcmp(pack, bigT, sizeof (bufwad_t)) != 0) fatal(B_FALSE, "pack/bigT mismatch in %p/%p", pack, bigT); if (freeit) { memset(pack, 0, sizeof (bufwad_t)); } else { pack->bw_index = n + i; pack->bw_txg = txg; pack->bw_data = 1 + ztest_random(-2ULL); } *bigH = *pack; *bigT = *pack; } /* * We've verified all the old bufwads, and made new ones. * Now write them out. */ dmu_write(os, packobj, packoff, packsize, packbuf, tx); if (freeit) { if (ztest_opts.zo_verbose >= 7) { (void) printf("freeing offset %"PRIx64" size %"PRIx64"" " txg %"PRIx64"\n", bigoff, bigsize, txg); } VERIFY0(dmu_free_range(os, bigobj, bigoff, bigsize, tx)); } else { if (ztest_opts.zo_verbose >= 7) { (void) printf("writing offset %"PRIx64" size %"PRIx64"" " txg %"PRIx64"\n", bigoff, bigsize, txg); } dmu_write(os, bigobj, bigoff, bigsize, bigbuf, tx); } dmu_tx_commit(tx); /* * Sanity check the stuff we just wrote. */ { void *packcheck = umem_alloc(packsize, UMEM_NOFAIL); void *bigcheck = umem_alloc(bigsize, UMEM_NOFAIL); VERIFY0(dmu_read(os, packobj, packoff, packsize, packcheck, DMU_READ_PREFETCH)); VERIFY0(dmu_read(os, bigobj, bigoff, bigsize, bigcheck, DMU_READ_PREFETCH)); ASSERT0(memcmp(packbuf, packcheck, packsize)); ASSERT0(memcmp(bigbuf, bigcheck, bigsize)); umem_free(packcheck, packsize); umem_free(bigcheck, bigsize); } umem_free(packbuf, packsize); umem_free(bigbuf, bigsize); umem_free(od, size); } static void compare_and_update_pbbufs(uint64_t s, bufwad_t *packbuf, bufwad_t *bigbuf, uint64_t bigsize, uint64_t n, uint64_t chunksize, uint64_t txg) { uint64_t i; bufwad_t *pack; bufwad_t *bigH; bufwad_t *bigT; /* * For each index from n to n + s, verify that the existing bufwad * in packobj matches the bufwads at the head and tail of the * corresponding chunk in bigobj. Then update all three bufwads * with the new values we want to write out. */ for (i = 0; i < s; i++) { /* LINTED */ pack = (bufwad_t *)((char *)packbuf + i * sizeof (bufwad_t)); /* LINTED */ bigH = (bufwad_t *)((char *)bigbuf + i * chunksize); /* LINTED */ bigT = (bufwad_t *)((char *)bigH + chunksize) - 1; ASSERT3U((uintptr_t)bigH - (uintptr_t)bigbuf, <, bigsize); ASSERT3U((uintptr_t)bigT - (uintptr_t)bigbuf, <, bigsize); if (pack->bw_txg > txg) fatal(B_FALSE, "future leak: got %"PRIx64", open txg is %"PRIx64"", pack->bw_txg, txg); if (pack->bw_data != 0 && pack->bw_index != n + i) fatal(B_FALSE, "wrong index: " "got %"PRIx64", wanted %"PRIx64"+%"PRIx64"", pack->bw_index, n, i); if (memcmp(pack, bigH, sizeof (bufwad_t)) != 0) fatal(B_FALSE, "pack/bigH mismatch in %p/%p", pack, bigH); if (memcmp(pack, bigT, sizeof (bufwad_t)) != 0) fatal(B_FALSE, "pack/bigT mismatch in %p/%p", pack, bigT); pack->bw_index = n + i; pack->bw_txg = txg; pack->bw_data = 1 + ztest_random(-2ULL); *bigH = *pack; *bigT = *pack; } } #undef OD_ARRAY_SIZE #define OD_ARRAY_SIZE 2 void ztest_dmu_read_write_zcopy(ztest_ds_t *zd, uint64_t id) { objset_t *os = zd->zd_os; ztest_od_t *od; dmu_tx_t *tx; uint64_t i; int error; int size; uint64_t n, s, txg; bufwad_t *packbuf, *bigbuf; uint64_t packobj, packoff, packsize, bigobj, bigoff, bigsize; uint64_t blocksize = ztest_random_blocksize(); uint64_t chunksize = blocksize; uint64_t regions = 997; uint64_t stride = 123456789ULL; uint64_t width = 9; dmu_buf_t *bonus_db; arc_buf_t **bigbuf_arcbufs; dmu_object_info_t doi; size = sizeof (ztest_od_t) * OD_ARRAY_SIZE; od = umem_alloc(size, UMEM_NOFAIL); /* * This test uses two objects, packobj and bigobj, that are always * updated together (i.e. in the same tx) so that their contents are * in sync and can be compared. Their contents relate to each other * in a simple way: packobj is a dense array of 'bufwad' structures, * while bigobj is a sparse array of the same bufwads. Specifically, * for any index n, there are three bufwads that should be identical: * * packobj, at offset n * sizeof (bufwad_t) * bigobj, at the head of the nth chunk * bigobj, at the tail of the nth chunk * * The chunk size is set equal to bigobj block size so that * dmu_assign_arcbuf_by_dbuf() can be tested for object updates. */ /* * Read the directory info. If it's the first time, set things up. */ ztest_od_init(od, id, FTAG, 0, DMU_OT_UINT64_OTHER, blocksize, 0, 0); ztest_od_init(od + 1, id, FTAG, 1, DMU_OT_UINT64_OTHER, 0, 0, chunksize); if (ztest_object_init(zd, od, size, B_FALSE) != 0) { umem_free(od, size); return; } bigobj = od[0].od_object; packobj = od[1].od_object; blocksize = od[0].od_blocksize; chunksize = blocksize; ASSERT3U(chunksize, ==, od[1].od_gen); VERIFY0(dmu_object_info(os, bigobj, &doi)); VERIFY(ISP2(doi.doi_data_block_size)); VERIFY3U(chunksize, ==, doi.doi_data_block_size); VERIFY3U(chunksize, >=, 2 * sizeof (bufwad_t)); /* * Pick a random index and compute the offsets into packobj and bigobj. */ n = ztest_random(regions) * stride + ztest_random(width); s = 1 + ztest_random(width - 1); packoff = n * sizeof (bufwad_t); packsize = s * sizeof (bufwad_t); bigoff = n * chunksize; bigsize = s * chunksize; packbuf = umem_zalloc(packsize, UMEM_NOFAIL); bigbuf = umem_zalloc(bigsize, UMEM_NOFAIL); VERIFY0(dmu_bonus_hold(os, bigobj, FTAG, &bonus_db)); bigbuf_arcbufs = umem_zalloc(2 * s * sizeof (arc_buf_t *), UMEM_NOFAIL); /* * Iteration 0 test zcopy for DB_UNCACHED dbufs. * Iteration 1 test zcopy to already referenced dbufs. * Iteration 2 test zcopy to dirty dbuf in the same txg. * Iteration 3 test zcopy to dbuf dirty in previous txg. * Iteration 4 test zcopy when dbuf is no longer dirty. * Iteration 5 test zcopy when it can't be done. * Iteration 6 one more zcopy write. */ for (i = 0; i < 7; i++) { uint64_t j; uint64_t off; /* * In iteration 5 (i == 5) use arcbufs * that don't match bigobj blksz to test * dmu_assign_arcbuf_by_dbuf() when it can't directly * assign an arcbuf to a dbuf. */ for (j = 0; j < s; j++) { if (i != 5 || chunksize < (SPA_MINBLOCKSIZE * 2)) { bigbuf_arcbufs[j] = dmu_request_arcbuf(bonus_db, chunksize); } else { bigbuf_arcbufs[2 * j] = dmu_request_arcbuf(bonus_db, chunksize / 2); bigbuf_arcbufs[2 * j + 1] = dmu_request_arcbuf(bonus_db, chunksize / 2); } } /* * Get a tx for the mods to both packobj and bigobj. */ tx = dmu_tx_create(os); dmu_tx_hold_write(tx, packobj, packoff, packsize); dmu_tx_hold_write(tx, bigobj, bigoff, bigsize); txg = ztest_tx_assign(tx, TXG_MIGHTWAIT, FTAG); if (txg == 0) { umem_free(packbuf, packsize); umem_free(bigbuf, bigsize); for (j = 0; j < s; j++) { if (i != 5 || chunksize < (SPA_MINBLOCKSIZE * 2)) { dmu_return_arcbuf(bigbuf_arcbufs[j]); } else { dmu_return_arcbuf( bigbuf_arcbufs[2 * j]); dmu_return_arcbuf( bigbuf_arcbufs[2 * j + 1]); } } umem_free(bigbuf_arcbufs, 2 * s * sizeof (arc_buf_t *)); umem_free(od, size); dmu_buf_rele(bonus_db, FTAG); return; } /* * 50% of the time don't read objects in the 1st iteration to * test dmu_assign_arcbuf_by_dbuf() for the case when there are * no existing dbufs for the specified offsets. */ if (i != 0 || ztest_random(2) != 0) { error = dmu_read(os, packobj, packoff, packsize, packbuf, DMU_READ_PREFETCH); ASSERT0(error); error = dmu_read(os, bigobj, bigoff, bigsize, bigbuf, DMU_READ_PREFETCH); ASSERT0(error); } compare_and_update_pbbufs(s, packbuf, bigbuf, bigsize, n, chunksize, txg); /* * We've verified all the old bufwads, and made new ones. * Now write them out. */ dmu_write(os, packobj, packoff, packsize, packbuf, tx); if (ztest_opts.zo_verbose >= 7) { (void) printf("writing offset %"PRIx64" size %"PRIx64"" " txg %"PRIx64"\n", bigoff, bigsize, txg); } for (off = bigoff, j = 0; j < s; j++, off += chunksize) { dmu_buf_t *dbt; if (i != 5 || chunksize < (SPA_MINBLOCKSIZE * 2)) { memcpy(bigbuf_arcbufs[j]->b_data, (caddr_t)bigbuf + (off - bigoff), chunksize); } else { memcpy(bigbuf_arcbufs[2 * j]->b_data, (caddr_t)bigbuf + (off - bigoff), chunksize / 2); memcpy(bigbuf_arcbufs[2 * j + 1]->b_data, (caddr_t)bigbuf + (off - bigoff) + chunksize / 2, chunksize / 2); } if (i == 1) { VERIFY(dmu_buf_hold(os, bigobj, off, FTAG, &dbt, DMU_READ_NO_PREFETCH) == 0); } if (i != 5 || chunksize < (SPA_MINBLOCKSIZE * 2)) { VERIFY0(dmu_assign_arcbuf_by_dbuf(bonus_db, off, bigbuf_arcbufs[j], tx)); } else { VERIFY0(dmu_assign_arcbuf_by_dbuf(bonus_db, off, bigbuf_arcbufs[2 * j], tx)); VERIFY0(dmu_assign_arcbuf_by_dbuf(bonus_db, off + chunksize / 2, bigbuf_arcbufs[2 * j + 1], tx)); } if (i == 1) { dmu_buf_rele(dbt, FTAG); } } dmu_tx_commit(tx); /* * Sanity check the stuff we just wrote. */ { void *packcheck = umem_alloc(packsize, UMEM_NOFAIL); void *bigcheck = umem_alloc(bigsize, UMEM_NOFAIL); VERIFY0(dmu_read(os, packobj, packoff, packsize, packcheck, DMU_READ_PREFETCH)); VERIFY0(dmu_read(os, bigobj, bigoff, bigsize, bigcheck, DMU_READ_PREFETCH)); ASSERT0(memcmp(packbuf, packcheck, packsize)); ASSERT0(memcmp(bigbuf, bigcheck, bigsize)); umem_free(packcheck, packsize); umem_free(bigcheck, bigsize); } if (i == 2) { txg_wait_open(dmu_objset_pool(os), 0, B_TRUE); } else if (i == 3) { txg_wait_synced(dmu_objset_pool(os), 0); } } dmu_buf_rele(bonus_db, FTAG); umem_free(packbuf, packsize); umem_free(bigbuf, bigsize); umem_free(bigbuf_arcbufs, 2 * s * sizeof (arc_buf_t *)); umem_free(od, size); } void ztest_dmu_write_parallel(ztest_ds_t *zd, uint64_t id) { (void) id; ztest_od_t *od; od = umem_alloc(sizeof (ztest_od_t), UMEM_NOFAIL); uint64_t offset = (1ULL << (ztest_random(20) + 43)) + (ztest_random(ZTEST_RANGE_LOCKS) << SPA_MAXBLOCKSHIFT); /* * Have multiple threads write to large offsets in an object * to verify that parallel writes to an object -- even to the * same blocks within the object -- doesn't cause any trouble. */ ztest_od_init(od, ID_PARALLEL, FTAG, 0, DMU_OT_UINT64_OTHER, 0, 0, 0); if (ztest_object_init(zd, od, sizeof (ztest_od_t), B_FALSE) != 0) return; while (ztest_random(10) != 0) ztest_io(zd, od->od_object, offset); umem_free(od, sizeof (ztest_od_t)); } void ztest_dmu_prealloc(ztest_ds_t *zd, uint64_t id) { ztest_od_t *od; uint64_t offset = (1ULL << (ztest_random(4) + SPA_MAXBLOCKSHIFT)) + (ztest_random(ZTEST_RANGE_LOCKS) << SPA_MAXBLOCKSHIFT); uint64_t count = ztest_random(20) + 1; uint64_t blocksize = ztest_random_blocksize(); void *data; od = umem_alloc(sizeof (ztest_od_t), UMEM_NOFAIL); ztest_od_init(od, id, FTAG, 0, DMU_OT_UINT64_OTHER, blocksize, 0, 0); if (ztest_object_init(zd, od, sizeof (ztest_od_t), !ztest_random(2)) != 0) { umem_free(od, sizeof (ztest_od_t)); return; } if (ztest_truncate(zd, od->od_object, offset, count * blocksize) != 0) { umem_free(od, sizeof (ztest_od_t)); return; } ztest_prealloc(zd, od->od_object, offset, count * blocksize); data = umem_zalloc(blocksize, UMEM_NOFAIL); while (ztest_random(count) != 0) { uint64_t randoff = offset + (ztest_random(count) * blocksize); if (ztest_write(zd, od->od_object, randoff, blocksize, data) != 0) break; while (ztest_random(4) != 0) ztest_io(zd, od->od_object, randoff); } umem_free(data, blocksize); umem_free(od, sizeof (ztest_od_t)); } /* * Verify that zap_{create,destroy,add,remove,update} work as expected. */ #define ZTEST_ZAP_MIN_INTS 1 #define ZTEST_ZAP_MAX_INTS 4 #define ZTEST_ZAP_MAX_PROPS 1000 void ztest_zap(ztest_ds_t *zd, uint64_t id) { objset_t *os = zd->zd_os; ztest_od_t *od; uint64_t object; uint64_t txg, last_txg; uint64_t value[ZTEST_ZAP_MAX_INTS]; uint64_t zl_ints, zl_intsize, prop; int i, ints; dmu_tx_t *tx; char propname[100], txgname[100]; int error; const char *const hc[2] = { "s.acl.h", ".s.open.h.hyLZlg" }; od = umem_alloc(sizeof (ztest_od_t), UMEM_NOFAIL); ztest_od_init(od, id, FTAG, 0, DMU_OT_ZAP_OTHER, 0, 0, 0); if (ztest_object_init(zd, od, sizeof (ztest_od_t), !ztest_random(2)) != 0) goto out; object = od->od_object; /* * Generate a known hash collision, and verify that * we can lookup and remove both entries. */ tx = dmu_tx_create(os); dmu_tx_hold_zap(tx, object, B_TRUE, NULL); txg = ztest_tx_assign(tx, TXG_MIGHTWAIT, FTAG); if (txg == 0) goto out; for (i = 0; i < 2; i++) { value[i] = i; VERIFY0(zap_add(os, object, hc[i], sizeof (uint64_t), 1, &value[i], tx)); } for (i = 0; i < 2; i++) { VERIFY3U(EEXIST, ==, zap_add(os, object, hc[i], sizeof (uint64_t), 1, &value[i], tx)); VERIFY0( zap_length(os, object, hc[i], &zl_intsize, &zl_ints)); ASSERT3U(zl_intsize, ==, sizeof (uint64_t)); ASSERT3U(zl_ints, ==, 1); } for (i = 0; i < 2; i++) { VERIFY0(zap_remove(os, object, hc[i], tx)); } dmu_tx_commit(tx); /* * Generate a bunch of random entries. */ ints = MAX(ZTEST_ZAP_MIN_INTS, object % ZTEST_ZAP_MAX_INTS); prop = ztest_random(ZTEST_ZAP_MAX_PROPS); (void) sprintf(propname, "prop_%"PRIu64"", prop); (void) sprintf(txgname, "txg_%"PRIu64"", prop); memset(value, 0, sizeof (value)); last_txg = 0; /* * If these zap entries already exist, validate their contents. */ error = zap_length(os, object, txgname, &zl_intsize, &zl_ints); if (error == 0) { ASSERT3U(zl_intsize, ==, sizeof (uint64_t)); ASSERT3U(zl_ints, ==, 1); VERIFY0(zap_lookup(os, object, txgname, zl_intsize, zl_ints, &last_txg)); VERIFY0(zap_length(os, object, propname, &zl_intsize, &zl_ints)); ASSERT3U(zl_intsize, ==, sizeof (uint64_t)); ASSERT3U(zl_ints, ==, ints); VERIFY0(zap_lookup(os, object, propname, zl_intsize, zl_ints, value)); for (i = 0; i < ints; i++) { ASSERT3U(value[i], ==, last_txg + object + i); } } else { ASSERT3U(error, ==, ENOENT); } /* * Atomically update two entries in our zap object. * The first is named txg_%llu, and contains the txg * in which the property was last updated. The second * is named prop_%llu, and the nth element of its value * should be txg + object + n. */ tx = dmu_tx_create(os); dmu_tx_hold_zap(tx, object, B_TRUE, NULL); txg = ztest_tx_assign(tx, TXG_MIGHTWAIT, FTAG); if (txg == 0) goto out; if (last_txg > txg) fatal(B_FALSE, "zap future leak: old %"PRIu64" new %"PRIu64"", last_txg, txg); for (i = 0; i < ints; i++) value[i] = txg + object + i; VERIFY0(zap_update(os, object, txgname, sizeof (uint64_t), 1, &txg, tx)); VERIFY0(zap_update(os, object, propname, sizeof (uint64_t), ints, value, tx)); dmu_tx_commit(tx); /* * Remove a random pair of entries. */ prop = ztest_random(ZTEST_ZAP_MAX_PROPS); (void) sprintf(propname, "prop_%"PRIu64"", prop); (void) sprintf(txgname, "txg_%"PRIu64"", prop); error = zap_length(os, object, txgname, &zl_intsize, &zl_ints); if (error == ENOENT) goto out; ASSERT0(error); tx = dmu_tx_create(os); dmu_tx_hold_zap(tx, object, B_TRUE, NULL); txg = ztest_tx_assign(tx, TXG_MIGHTWAIT, FTAG); if (txg == 0) goto out; VERIFY0(zap_remove(os, object, txgname, tx)); VERIFY0(zap_remove(os, object, propname, tx)); dmu_tx_commit(tx); out: umem_free(od, sizeof (ztest_od_t)); } /* * Test case to test the upgrading of a microzap to fatzap. */ void ztest_fzap(ztest_ds_t *zd, uint64_t id) { objset_t *os = zd->zd_os; ztest_od_t *od; uint64_t object, txg, value; od = umem_alloc(sizeof (ztest_od_t), UMEM_NOFAIL); ztest_od_init(od, id, FTAG, 0, DMU_OT_ZAP_OTHER, 0, 0, 0); if (ztest_object_init(zd, od, sizeof (ztest_od_t), !ztest_random(2)) != 0) goto out; object = od->od_object; /* * Add entries to this ZAP and make sure it spills over * and gets upgraded to a fatzap. Also, since we are adding * 2050 entries we should see ptrtbl growth and leaf-block split. */ for (value = 0; value < 2050; value++) { char name[ZFS_MAX_DATASET_NAME_LEN]; dmu_tx_t *tx; int error; (void) snprintf(name, sizeof (name), "fzap-%"PRIu64"-%"PRIu64"", id, value); tx = dmu_tx_create(os); dmu_tx_hold_zap(tx, object, B_TRUE, name); txg = ztest_tx_assign(tx, TXG_MIGHTWAIT, FTAG); if (txg == 0) goto out; error = zap_add(os, object, name, sizeof (uint64_t), 1, &value, tx); ASSERT(error == 0 || error == EEXIST); dmu_tx_commit(tx); } out: umem_free(od, sizeof (ztest_od_t)); } void ztest_zap_parallel(ztest_ds_t *zd, uint64_t id) { (void) id; objset_t *os = zd->zd_os; ztest_od_t *od; uint64_t txg, object, count, wsize, wc, zl_wsize, zl_wc; dmu_tx_t *tx; int i, namelen, error; int micro = ztest_random(2); char name[20], string_value[20]; void *data; od = umem_alloc(sizeof (ztest_od_t), UMEM_NOFAIL); ztest_od_init(od, ID_PARALLEL, FTAG, micro, DMU_OT_ZAP_OTHER, 0, 0, 0); if (ztest_object_init(zd, od, sizeof (ztest_od_t), B_FALSE) != 0) { umem_free(od, sizeof (ztest_od_t)); return; } object = od->od_object; /* * Generate a random name of the form 'xxx.....' where each * x is a random printable character and the dots are dots. * There are 94 such characters, and the name length goes from * 6 to 20, so there are 94^3 * 15 = 12,458,760 possible names. */ namelen = ztest_random(sizeof (name) - 5) + 5 + 1; for (i = 0; i < 3; i++) name[i] = '!' + ztest_random('~' - '!' + 1); for (; i < namelen - 1; i++) name[i] = '.'; name[i] = '\0'; if ((namelen & 1) || micro) { wsize = sizeof (txg); wc = 1; data = &txg; } else { wsize = 1; wc = namelen; data = string_value; } count = -1ULL; VERIFY0(zap_count(os, object, &count)); ASSERT3S(count, !=, -1ULL); /* * Select an operation: length, lookup, add, update, remove. */ i = ztest_random(5); if (i >= 2) { tx = dmu_tx_create(os); dmu_tx_hold_zap(tx, object, B_TRUE, NULL); txg = ztest_tx_assign(tx, TXG_MIGHTWAIT, FTAG); if (txg == 0) { umem_free(od, sizeof (ztest_od_t)); return; } memcpy(string_value, name, namelen); } else { tx = NULL; txg = 0; memset(string_value, 0, namelen); } switch (i) { case 0: error = zap_length(os, object, name, &zl_wsize, &zl_wc); if (error == 0) { ASSERT3U(wsize, ==, zl_wsize); ASSERT3U(wc, ==, zl_wc); } else { ASSERT3U(error, ==, ENOENT); } break; case 1: error = zap_lookup(os, object, name, wsize, wc, data); if (error == 0) { if (data == string_value && memcmp(name, data, namelen) != 0) fatal(B_FALSE, "name '%s' != val '%s' len %d", name, (char *)data, namelen); } else { ASSERT3U(error, ==, ENOENT); } break; case 2: error = zap_add(os, object, name, wsize, wc, data, tx); ASSERT(error == 0 || error == EEXIST); break; case 3: VERIFY0(zap_update(os, object, name, wsize, wc, data, tx)); break; case 4: error = zap_remove(os, object, name, tx); ASSERT(error == 0 || error == ENOENT); break; } if (tx != NULL) dmu_tx_commit(tx); umem_free(od, sizeof (ztest_od_t)); } /* * Commit callback data. */ typedef struct ztest_cb_data { list_node_t zcd_node; uint64_t zcd_txg; int zcd_expected_err; boolean_t zcd_added; boolean_t zcd_called; spa_t *zcd_spa; } ztest_cb_data_t; /* This is the actual commit callback function */ static void ztest_commit_callback(void *arg, int error) { ztest_cb_data_t *data = arg; uint64_t synced_txg; VERIFY3P(data, !=, NULL); VERIFY3S(data->zcd_expected_err, ==, error); VERIFY(!data->zcd_called); synced_txg = spa_last_synced_txg(data->zcd_spa); if (data->zcd_txg > synced_txg) fatal(B_FALSE, "commit callback of txg %"PRIu64" called prematurely, " "last synced txg = %"PRIu64"\n", data->zcd_txg, synced_txg); data->zcd_called = B_TRUE; if (error == ECANCELED) { ASSERT0(data->zcd_txg); ASSERT(!data->zcd_added); /* * The private callback data should be destroyed here, but * since we are going to check the zcd_called field after * dmu_tx_abort(), we will destroy it there. */ return; } ASSERT(data->zcd_added); ASSERT3U(data->zcd_txg, !=, 0); (void) mutex_enter(&zcl.zcl_callbacks_lock); /* See if this cb was called more quickly */ if ((synced_txg - data->zcd_txg) < zc_min_txg_delay) zc_min_txg_delay = synced_txg - data->zcd_txg; /* Remove our callback from the list */ list_remove(&zcl.zcl_callbacks, data); (void) mutex_exit(&zcl.zcl_callbacks_lock); umem_free(data, sizeof (ztest_cb_data_t)); } /* Allocate and initialize callback data structure */ static ztest_cb_data_t * ztest_create_cb_data(objset_t *os, uint64_t txg) { ztest_cb_data_t *cb_data; cb_data = umem_zalloc(sizeof (ztest_cb_data_t), UMEM_NOFAIL); cb_data->zcd_txg = txg; cb_data->zcd_spa = dmu_objset_spa(os); list_link_init(&cb_data->zcd_node); return (cb_data); } /* * Commit callback test. */ void ztest_dmu_commit_callbacks(ztest_ds_t *zd, uint64_t id) { objset_t *os = zd->zd_os; ztest_od_t *od; dmu_tx_t *tx; ztest_cb_data_t *cb_data[3], *tmp_cb; uint64_t old_txg, txg; int i, error = 0; od = umem_alloc(sizeof (ztest_od_t), UMEM_NOFAIL); ztest_od_init(od, id, FTAG, 0, DMU_OT_UINT64_OTHER, 0, 0, 0); if (ztest_object_init(zd, od, sizeof (ztest_od_t), B_FALSE) != 0) { umem_free(od, sizeof (ztest_od_t)); return; } tx = dmu_tx_create(os); cb_data[0] = ztest_create_cb_data(os, 0); dmu_tx_callback_register(tx, ztest_commit_callback, cb_data[0]); dmu_tx_hold_write(tx, od->od_object, 0, sizeof (uint64_t)); /* Every once in a while, abort the transaction on purpose */ if (ztest_random(100) == 0) error = -1; if (!error) error = dmu_tx_assign(tx, TXG_NOWAIT); txg = error ? 0 : dmu_tx_get_txg(tx); cb_data[0]->zcd_txg = txg; cb_data[1] = ztest_create_cb_data(os, txg); dmu_tx_callback_register(tx, ztest_commit_callback, cb_data[1]); if (error) { /* * It's not a strict requirement to call the registered * callbacks from inside dmu_tx_abort(), but that's what * it's supposed to happen in the current implementation * so we will check for that. */ for (i = 0; i < 2; i++) { cb_data[i]->zcd_expected_err = ECANCELED; VERIFY(!cb_data[i]->zcd_called); } dmu_tx_abort(tx); for (i = 0; i < 2; i++) { VERIFY(cb_data[i]->zcd_called); umem_free(cb_data[i], sizeof (ztest_cb_data_t)); } umem_free(od, sizeof (ztest_od_t)); return; } cb_data[2] = ztest_create_cb_data(os, txg); dmu_tx_callback_register(tx, ztest_commit_callback, cb_data[2]); /* * Read existing data to make sure there isn't a future leak. */ VERIFY0(dmu_read(os, od->od_object, 0, sizeof (uint64_t), &old_txg, DMU_READ_PREFETCH)); if (old_txg > txg) fatal(B_FALSE, "future leak: got %"PRIu64", open txg is %"PRIu64"", old_txg, txg); dmu_write(os, od->od_object, 0, sizeof (uint64_t), &txg, tx); (void) mutex_enter(&zcl.zcl_callbacks_lock); /* * Since commit callbacks don't have any ordering requirement and since * it is theoretically possible for a commit callback to be called * after an arbitrary amount of time has elapsed since its txg has been * synced, it is difficult to reliably determine whether a commit * callback hasn't been called due to high load or due to a flawed * implementation. * * In practice, we will assume that if after a certain number of txgs a * commit callback hasn't been called, then most likely there's an * implementation bug.. */ tmp_cb = list_head(&zcl.zcl_callbacks); if (tmp_cb != NULL && tmp_cb->zcd_txg + ZTEST_COMMIT_CB_THRESH < txg) { fatal(B_FALSE, "Commit callback threshold exceeded, " "oldest txg: %"PRIu64", open txg: %"PRIu64"\n", tmp_cb->zcd_txg, txg); } /* * Let's find the place to insert our callbacks. * * Even though the list is ordered by txg, it is possible for the * insertion point to not be the end because our txg may already be * quiescing at this point and other callbacks in the open txg * (from other objsets) may have sneaked in. */ tmp_cb = list_tail(&zcl.zcl_callbacks); while (tmp_cb != NULL && tmp_cb->zcd_txg > txg) tmp_cb = list_prev(&zcl.zcl_callbacks, tmp_cb); /* Add the 3 callbacks to the list */ for (i = 0; i < 3; i++) { if (tmp_cb == NULL) list_insert_head(&zcl.zcl_callbacks, cb_data[i]); else list_insert_after(&zcl.zcl_callbacks, tmp_cb, cb_data[i]); cb_data[i]->zcd_added = B_TRUE; VERIFY(!cb_data[i]->zcd_called); tmp_cb = cb_data[i]; } zc_cb_counter += 3; (void) mutex_exit(&zcl.zcl_callbacks_lock); dmu_tx_commit(tx); umem_free(od, sizeof (ztest_od_t)); } /* * Visit each object in the dataset. Verify that its properties * are consistent what was stored in the block tag when it was created, * and that its unused bonus buffer space has not been overwritten. */ void ztest_verify_dnode_bt(ztest_ds_t *zd, uint64_t id) { (void) id; objset_t *os = zd->zd_os; uint64_t obj; int err = 0; for (obj = 0; err == 0; err = dmu_object_next(os, &obj, FALSE, 0)) { ztest_block_tag_t *bt = NULL; dmu_object_info_t doi; dmu_buf_t *db; ztest_object_lock(zd, obj, RL_READER); if (dmu_bonus_hold(os, obj, FTAG, &db) != 0) { ztest_object_unlock(zd, obj); continue; } dmu_object_info_from_db(db, &doi); if (doi.doi_bonus_size >= sizeof (*bt)) bt = ztest_bt_bonus(db); if (bt && bt->bt_magic == BT_MAGIC) { ztest_bt_verify(bt, os, obj, doi.doi_dnodesize, bt->bt_offset, bt->bt_gen, bt->bt_txg, bt->bt_crtxg); ztest_verify_unused_bonus(db, bt, obj, os, bt->bt_gen); } dmu_buf_rele(db, FTAG); ztest_object_unlock(zd, obj); } } void ztest_dsl_prop_get_set(ztest_ds_t *zd, uint64_t id) { (void) id; zfs_prop_t proplist[] = { ZFS_PROP_CHECKSUM, ZFS_PROP_COMPRESSION, ZFS_PROP_COPIES, ZFS_PROP_DEDUP }; (void) pthread_rwlock_rdlock(&ztest_name_lock); for (int p = 0; p < sizeof (proplist) / sizeof (proplist[0]); p++) (void) ztest_dsl_prop_set_uint64(zd->zd_name, proplist[p], ztest_random_dsl_prop(proplist[p]), (int)ztest_random(2)); VERIFY0(ztest_dsl_prop_set_uint64(zd->zd_name, ZFS_PROP_RECORDSIZE, ztest_random_blocksize(), (int)ztest_random(2))); (void) pthread_rwlock_unlock(&ztest_name_lock); } void ztest_spa_prop_get_set(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; nvlist_t *props = NULL; (void) pthread_rwlock_rdlock(&ztest_name_lock); (void) ztest_spa_prop_set_uint64(ZPOOL_PROP_AUTOTRIM, ztest_random(2)); VERIFY0(spa_prop_get(ztest_spa, &props)); if (ztest_opts.zo_verbose >= 6) dump_nvlist(props, 4); fnvlist_free(props); (void) pthread_rwlock_unlock(&ztest_name_lock); } static int user_release_one(const char *snapname, const char *holdname) { nvlist_t *snaps, *holds; int error; snaps = fnvlist_alloc(); holds = fnvlist_alloc(); fnvlist_add_boolean(holds, holdname); fnvlist_add_nvlist(snaps, snapname, holds); fnvlist_free(holds); error = dsl_dataset_user_release(snaps, NULL); fnvlist_free(snaps); return (error); } /* * Test snapshot hold/release and deferred destroy. */ void ztest_dmu_snapshot_hold(ztest_ds_t *zd, uint64_t id) { int error; objset_t *os = zd->zd_os; objset_t *origin; char snapname[100]; char fullname[100]; char clonename[100]; char tag[100]; char osname[ZFS_MAX_DATASET_NAME_LEN]; nvlist_t *holds; (void) pthread_rwlock_rdlock(&ztest_name_lock); dmu_objset_name(os, osname); (void) snprintf(snapname, sizeof (snapname), "sh1_%"PRIu64"", id); (void) snprintf(fullname, sizeof (fullname), "%s@%s", osname, snapname); (void) snprintf(clonename, sizeof (clonename), "%s/ch1_%"PRIu64"", osname, id); (void) snprintf(tag, sizeof (tag), "tag_%"PRIu64"", id); /* * Clean up from any previous run. */ error = dsl_destroy_head(clonename); if (error != ENOENT) ASSERT0(error); error = user_release_one(fullname, tag); if (error != ESRCH && error != ENOENT) ASSERT0(error); error = dsl_destroy_snapshot(fullname, B_FALSE); if (error != ENOENT) ASSERT0(error); /* * Create snapshot, clone it, mark snap for deferred destroy, * destroy clone, verify snap was also destroyed. */ error = dmu_objset_snapshot_one(osname, snapname); if (error) { if (error == ENOSPC) { ztest_record_enospc("dmu_objset_snapshot"); goto out; } fatal(B_FALSE, "dmu_objset_snapshot(%s) = %d", fullname, error); } error = dmu_objset_clone(clonename, fullname); if (error) { if (error == ENOSPC) { ztest_record_enospc("dmu_objset_clone"); goto out; } fatal(B_FALSE, "dmu_objset_clone(%s) = %d", clonename, error); } error = dsl_destroy_snapshot(fullname, B_TRUE); if (error) { fatal(B_FALSE, "dsl_destroy_snapshot(%s, B_TRUE) = %d", fullname, error); } error = dsl_destroy_head(clonename); if (error) fatal(B_FALSE, "dsl_destroy_head(%s) = %d", clonename, error); error = dmu_objset_hold(fullname, FTAG, &origin); if (error != ENOENT) fatal(B_FALSE, "dmu_objset_hold(%s) = %d", fullname, error); /* * Create snapshot, add temporary hold, verify that we can't * destroy a held snapshot, mark for deferred destroy, * release hold, verify snapshot was destroyed. */ error = dmu_objset_snapshot_one(osname, snapname); if (error) { if (error == ENOSPC) { ztest_record_enospc("dmu_objset_snapshot"); goto out; } fatal(B_FALSE, "dmu_objset_snapshot(%s) = %d", fullname, error); } holds = fnvlist_alloc(); fnvlist_add_string(holds, fullname, tag); error = dsl_dataset_user_hold(holds, 0, NULL); fnvlist_free(holds); if (error == ENOSPC) { ztest_record_enospc("dsl_dataset_user_hold"); goto out; } else if (error) { fatal(B_FALSE, "dsl_dataset_user_hold(%s, %s) = %u", fullname, tag, error); } error = dsl_destroy_snapshot(fullname, B_FALSE); if (error != EBUSY) { fatal(B_FALSE, "dsl_destroy_snapshot(%s, B_FALSE) = %d", fullname, error); } error = dsl_destroy_snapshot(fullname, B_TRUE); if (error) { fatal(B_FALSE, "dsl_destroy_snapshot(%s, B_TRUE) = %d", fullname, error); } error = user_release_one(fullname, tag); if (error) fatal(B_FALSE, "user_release_one(%s, %s) = %d", fullname, tag, error); VERIFY3U(dmu_objset_hold(fullname, FTAG, &origin), ==, ENOENT); out: (void) pthread_rwlock_unlock(&ztest_name_lock); } /* * Inject random faults into the on-disk data. */ void ztest_fault_inject(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; ztest_shared_t *zs = ztest_shared; spa_t *spa = ztest_spa; int fd; uint64_t offset; uint64_t leaves; uint64_t bad = 0x1990c0ffeedecadeull; uint64_t top, leaf; char *path0; char *pathrand; size_t fsize; int bshift = SPA_MAXBLOCKSHIFT + 2; int iters = 1000; int maxfaults; int mirror_save; vdev_t *vd0 = NULL; uint64_t guid0 = 0; boolean_t islog = B_FALSE; path0 = umem_alloc(MAXPATHLEN, UMEM_NOFAIL); pathrand = umem_alloc(MAXPATHLEN, UMEM_NOFAIL); mutex_enter(&ztest_vdev_lock); /* * Device removal is in progress, fault injection must be disabled * until it completes and the pool is scrubbed. The fault injection * strategy for damaging blocks does not take in to account evacuated * blocks which may have already been damaged. */ if (ztest_device_removal_active) { mutex_exit(&ztest_vdev_lock); goto out; } maxfaults = MAXFAULTS(zs); leaves = MAX(zs->zs_mirrors, 1) * ztest_opts.zo_raid_children; mirror_save = zs->zs_mirrors; mutex_exit(&ztest_vdev_lock); ASSERT3U(leaves, >=, 1); /* * While ztest is running the number of leaves will not change. This * is critical for the fault injection logic as it determines where * errors can be safely injected such that they are always repairable. * * When restarting ztest a different number of leaves may be requested * which will shift the regions to be damaged. This is fine as long * as the pool has been scrubbed prior to using the new mapping. * Failure to do can result in non-repairable damage being injected. */ if (ztest_pool_scrubbed == B_FALSE) goto out; /* * Grab the name lock as reader. There are some operations * which don't like to have their vdevs changed while * they are in progress (i.e. spa_change_guid). Those * operations will have grabbed the name lock as writer. */ (void) pthread_rwlock_rdlock(&ztest_name_lock); /* * We need SCL_STATE here because we're going to look at vd0->vdev_tsd. */ spa_config_enter(spa, SCL_STATE, FTAG, RW_READER); if (ztest_random(2) == 0) { /* * Inject errors on a normal data device or slog device. */ top = ztest_random_vdev_top(spa, B_TRUE); leaf = ztest_random(leaves) + zs->zs_splits; /* * Generate paths to the first leaf in this top-level vdev, * and to the random leaf we selected. We'll induce transient * write failures and random online/offline activity on leaf 0, * and we'll write random garbage to the randomly chosen leaf. */ (void) snprintf(path0, MAXPATHLEN, ztest_dev_template, ztest_opts.zo_dir, ztest_opts.zo_pool, top * leaves + zs->zs_splits); (void) snprintf(pathrand, MAXPATHLEN, ztest_dev_template, ztest_opts.zo_dir, ztest_opts.zo_pool, top * leaves + leaf); vd0 = vdev_lookup_by_path(spa->spa_root_vdev, path0); if (vd0 != NULL && vd0->vdev_top->vdev_islog) islog = B_TRUE; /* * If the top-level vdev needs to be resilvered * then we only allow faults on the device that is * resilvering. */ if (vd0 != NULL && maxfaults != 1 && (!vdev_resilver_needed(vd0->vdev_top, NULL, NULL) || vd0->vdev_resilver_txg != 0)) { /* * Make vd0 explicitly claim to be unreadable, * or unwritable, or reach behind its back * and close the underlying fd. We can do this if * maxfaults == 0 because we'll fail and reexecute, * and we can do it if maxfaults >= 2 because we'll * have enough redundancy. If maxfaults == 1, the * combination of this with injection of random data * corruption below exceeds the pool's fault tolerance. */ vdev_file_t *vf = vd0->vdev_tsd; zfs_dbgmsg("injecting fault to vdev %llu; maxfaults=%d", (long long)vd0->vdev_id, (int)maxfaults); if (vf != NULL && ztest_random(3) == 0) { (void) close(vf->vf_file->f_fd); vf->vf_file->f_fd = -1; } else if (ztest_random(2) == 0) { vd0->vdev_cant_read = B_TRUE; } else { vd0->vdev_cant_write = B_TRUE; } guid0 = vd0->vdev_guid; } } else { /* * Inject errors on an l2cache device. */ spa_aux_vdev_t *sav = &spa->spa_l2cache; if (sav->sav_count == 0) { spa_config_exit(spa, SCL_STATE, FTAG); (void) pthread_rwlock_unlock(&ztest_name_lock); goto out; } vd0 = sav->sav_vdevs[ztest_random(sav->sav_count)]; guid0 = vd0->vdev_guid; (void) strlcpy(path0, vd0->vdev_path, MAXPATHLEN); (void) strlcpy(pathrand, vd0->vdev_path, MAXPATHLEN); leaf = 0; leaves = 1; maxfaults = INT_MAX; /* no limit on cache devices */ } spa_config_exit(spa, SCL_STATE, FTAG); (void) pthread_rwlock_unlock(&ztest_name_lock); /* * If we can tolerate two or more faults, or we're dealing * with a slog, randomly online/offline vd0. */ if ((maxfaults >= 2 || islog) && guid0 != 0) { if (ztest_random(10) < 6) { int flags = (ztest_random(2) == 0 ? ZFS_OFFLINE_TEMPORARY : 0); /* * We have to grab the zs_name_lock as writer to * prevent a race between offlining a slog and * destroying a dataset. Offlining the slog will * grab a reference on the dataset which may cause * dsl_destroy_head() to fail with EBUSY thus * leaving the dataset in an inconsistent state. */ if (islog) (void) pthread_rwlock_wrlock(&ztest_name_lock); VERIFY3U(vdev_offline(spa, guid0, flags), !=, EBUSY); if (islog) (void) pthread_rwlock_unlock(&ztest_name_lock); } else { /* * Ideally we would like to be able to randomly * call vdev_[on|off]line without holding locks * to force unpredictable failures but the side * effects of vdev_[on|off]line prevent us from * doing so. We grab the ztest_vdev_lock here to * prevent a race between injection testing and * aux_vdev removal. */ mutex_enter(&ztest_vdev_lock); (void) vdev_online(spa, guid0, 0, NULL); mutex_exit(&ztest_vdev_lock); } } if (maxfaults == 0) goto out; /* * We have at least single-fault tolerance, so inject data corruption. */ fd = open(pathrand, O_RDWR); if (fd == -1) /* we hit a gap in the device namespace */ goto out; fsize = lseek(fd, 0, SEEK_END); while (--iters != 0) { /* * The offset must be chosen carefully to ensure that * we do not inject a given logical block with errors * on two different leaf devices, because ZFS can not * tolerate that (if maxfaults==1). * * To achieve this we divide each leaf device into * chunks of size (# leaves * SPA_MAXBLOCKSIZE * 4). * Each chunk is further divided into error-injection * ranges (can accept errors) and clear ranges (we do * not inject errors in those). Each error-injection * range can accept errors only for a single leaf vdev. * Error-injection ranges are separated by clear ranges. * * For example, with 3 leaves, each chunk looks like: * 0 to 32M: injection range for leaf 0 * 32M to 64M: clear range - no injection allowed * 64M to 96M: injection range for leaf 1 * 96M to 128M: clear range - no injection allowed * 128M to 160M: injection range for leaf 2 * 160M to 192M: clear range - no injection allowed * * Each clear range must be large enough such that a * single block cannot straddle it. This way a block * can't be a target in two different injection ranges * (on different leaf vdevs). */ offset = ztest_random(fsize / (leaves << bshift)) * (leaves << bshift) + (leaf << bshift) + (ztest_random(1ULL << (bshift - 1)) & -8ULL); /* * Only allow damage to the labels at one end of the vdev. * * If all labels are damaged, the device will be totally * inaccessible, which will result in loss of data, * because we also damage (parts of) the other side of * the mirror/raidz. * * Additionally, we will always have both an even and an * odd label, so that we can handle crashes in the * middle of vdev_config_sync(). */ if ((leaf & 1) == 0 && offset < VDEV_LABEL_START_SIZE) continue; /* * The two end labels are stored at the "end" of the disk, but * the end of the disk (vdev_psize) is aligned to * sizeof (vdev_label_t). */ uint64_t psize = P2ALIGN(fsize, sizeof (vdev_label_t)); if ((leaf & 1) == 1 && offset + sizeof (bad) > psize - VDEV_LABEL_END_SIZE) continue; mutex_enter(&ztest_vdev_lock); if (mirror_save != zs->zs_mirrors) { mutex_exit(&ztest_vdev_lock); (void) close(fd); goto out; } if (pwrite(fd, &bad, sizeof (bad), offset) != sizeof (bad)) fatal(B_TRUE, "can't inject bad word at 0x%"PRIx64" in %s", offset, pathrand); mutex_exit(&ztest_vdev_lock); if (ztest_opts.zo_verbose >= 7) (void) printf("injected bad word into %s," " offset 0x%"PRIx64"\n", pathrand, offset); } (void) close(fd); out: umem_free(path0, MAXPATHLEN); umem_free(pathrand, MAXPATHLEN); } /* * By design ztest will never inject uncorrectable damage in to the pool. * Issue a scrub, wait for it to complete, and verify there is never any * persistent damage. * * Only after a full scrub has been completed is it safe to start injecting * data corruption. See the comment in zfs_fault_inject(). */ static int ztest_scrub_impl(spa_t *spa) { int error = spa_scan(spa, POOL_SCAN_SCRUB); if (error) return (error); while (dsl_scan_scrubbing(spa_get_dsl(spa))) txg_wait_synced(spa_get_dsl(spa), 0); if (spa_get_errlog_size(spa) > 0) return (ECKSUM); ztest_pool_scrubbed = B_TRUE; return (0); } /* * Scrub the pool. */ void ztest_scrub(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; spa_t *spa = ztest_spa; int error; /* * Scrub in progress by device removal. */ if (ztest_device_removal_active) return; /* * Start a scrub, wait a moment, then force a restart. */ (void) spa_scan(spa, POOL_SCAN_SCRUB); (void) poll(NULL, 0, 100); error = ztest_scrub_impl(spa); if (error == EBUSY) error = 0; ASSERT0(error); } /* * Change the guid for the pool. */ void ztest_reguid(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; spa_t *spa = ztest_spa; uint64_t orig, load; int error; if (ztest_opts.zo_mmp_test) return; orig = spa_guid(spa); load = spa_load_guid(spa); (void) pthread_rwlock_wrlock(&ztest_name_lock); error = spa_change_guid(spa); (void) pthread_rwlock_unlock(&ztest_name_lock); if (error != 0) return; if (ztest_opts.zo_verbose >= 4) { (void) printf("Changed guid old %"PRIu64" -> %"PRIu64"\n", orig, spa_guid(spa)); } VERIFY3U(orig, !=, spa_guid(spa)); VERIFY3U(load, ==, spa_load_guid(spa)); } void ztest_blake3(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; hrtime_t end = gethrtime() + NANOSEC; zio_cksum_salt_t salt; void *salt_ptr = &salt.zcs_bytes; struct abd *abd_data, *abd_meta; void *buf, *templ; int i, *ptr; uint32_t size; BLAKE3_CTX ctx; size = ztest_random_blocksize(); buf = umem_alloc(size, UMEM_NOFAIL); abd_data = abd_alloc(size, B_FALSE); abd_meta = abd_alloc(size, B_TRUE); for (i = 0, ptr = buf; i < size / sizeof (*ptr); i++, ptr++) *ptr = ztest_random(UINT_MAX); memset(salt_ptr, 'A', 32); abd_copy_from_buf_off(abd_data, buf, 0, size); abd_copy_from_buf_off(abd_meta, buf, 0, size); while (gethrtime() <= end) { int run_count = 100; zio_cksum_t zc_ref1, zc_ref2; zio_cksum_t zc_res1, zc_res2; void *ref1 = &zc_ref1; void *ref2 = &zc_ref2; void *res1 = &zc_res1; void *res2 = &zc_res2; /* BLAKE3_KEY_LEN = 32 */ VERIFY0(blake3_impl_setname("generic")); templ = abd_checksum_blake3_tmpl_init(&salt); Blake3_InitKeyed(&ctx, salt_ptr); Blake3_Update(&ctx, buf, size); Blake3_Final(&ctx, ref1); zc_ref2 = zc_ref1; ZIO_CHECKSUM_BSWAP(&zc_ref2); abd_checksum_blake3_tmpl_free(templ); VERIFY0(blake3_impl_setname("cycle")); while (run_count-- > 0) { /* Test current implementation */ Blake3_InitKeyed(&ctx, salt_ptr); Blake3_Update(&ctx, buf, size); Blake3_Final(&ctx, res1); zc_res2 = zc_res1; ZIO_CHECKSUM_BSWAP(&zc_res2); VERIFY0(memcmp(ref1, res1, 32)); VERIFY0(memcmp(ref2, res2, 32)); /* Test ABD - data */ templ = abd_checksum_blake3_tmpl_init(&salt); abd_checksum_blake3_native(abd_data, size, templ, &zc_res1); abd_checksum_blake3_byteswap(abd_data, size, templ, &zc_res2); VERIFY0(memcmp(ref1, res1, 32)); VERIFY0(memcmp(ref2, res2, 32)); /* Test ABD - metadata */ abd_checksum_blake3_native(abd_meta, size, templ, &zc_res1); abd_checksum_blake3_byteswap(abd_meta, size, templ, &zc_res2); abd_checksum_blake3_tmpl_free(templ); VERIFY0(memcmp(ref1, res1, 32)); VERIFY0(memcmp(ref2, res2, 32)); } } abd_free(abd_data); abd_free(abd_meta); umem_free(buf, size); } void ztest_fletcher(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; hrtime_t end = gethrtime() + NANOSEC; while (gethrtime() <= end) { int run_count = 100; void *buf; struct abd *abd_data, *abd_meta; uint32_t size; int *ptr; int i; zio_cksum_t zc_ref; zio_cksum_t zc_ref_byteswap; size = ztest_random_blocksize(); buf = umem_alloc(size, UMEM_NOFAIL); abd_data = abd_alloc(size, B_FALSE); abd_meta = abd_alloc(size, B_TRUE); for (i = 0, ptr = buf; i < size / sizeof (*ptr); i++, ptr++) *ptr = ztest_random(UINT_MAX); abd_copy_from_buf_off(abd_data, buf, 0, size); abd_copy_from_buf_off(abd_meta, buf, 0, size); VERIFY0(fletcher_4_impl_set("scalar")); fletcher_4_native(buf, size, NULL, &zc_ref); fletcher_4_byteswap(buf, size, NULL, &zc_ref_byteswap); VERIFY0(fletcher_4_impl_set("cycle")); while (run_count-- > 0) { zio_cksum_t zc; zio_cksum_t zc_byteswap; fletcher_4_byteswap(buf, size, NULL, &zc_byteswap); fletcher_4_native(buf, size, NULL, &zc); VERIFY0(memcmp(&zc, &zc_ref, sizeof (zc))); VERIFY0(memcmp(&zc_byteswap, &zc_ref_byteswap, sizeof (zc_byteswap))); /* Test ABD - data */ abd_fletcher_4_byteswap(abd_data, size, NULL, &zc_byteswap); abd_fletcher_4_native(abd_data, size, NULL, &zc); VERIFY0(memcmp(&zc, &zc_ref, sizeof (zc))); VERIFY0(memcmp(&zc_byteswap, &zc_ref_byteswap, sizeof (zc_byteswap))); /* Test ABD - metadata */ abd_fletcher_4_byteswap(abd_meta, size, NULL, &zc_byteswap); abd_fletcher_4_native(abd_meta, size, NULL, &zc); VERIFY0(memcmp(&zc, &zc_ref, sizeof (zc))); VERIFY0(memcmp(&zc_byteswap, &zc_ref_byteswap, sizeof (zc_byteswap))); } umem_free(buf, size); abd_free(abd_data); abd_free(abd_meta); } } void ztest_fletcher_incr(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; void *buf; size_t size; int *ptr; int i; zio_cksum_t zc_ref; zio_cksum_t zc_ref_bswap; hrtime_t end = gethrtime() + NANOSEC; while (gethrtime() <= end) { int run_count = 100; size = ztest_random_blocksize(); buf = umem_alloc(size, UMEM_NOFAIL); for (i = 0, ptr = buf; i < size / sizeof (*ptr); i++, ptr++) *ptr = ztest_random(UINT_MAX); VERIFY0(fletcher_4_impl_set("scalar")); fletcher_4_native(buf, size, NULL, &zc_ref); fletcher_4_byteswap(buf, size, NULL, &zc_ref_bswap); VERIFY0(fletcher_4_impl_set("cycle")); while (run_count-- > 0) { zio_cksum_t zc; zio_cksum_t zc_bswap; size_t pos = 0; ZIO_SET_CHECKSUM(&zc, 0, 0, 0, 0); ZIO_SET_CHECKSUM(&zc_bswap, 0, 0, 0, 0); while (pos < size) { size_t inc = 64 * ztest_random(size / 67); /* sometimes add few bytes to test non-simd */ if (ztest_random(100) < 10) inc += P2ALIGN(ztest_random(64), sizeof (uint32_t)); if (inc > (size - pos)) inc = size - pos; fletcher_4_incremental_native(buf + pos, inc, &zc); fletcher_4_incremental_byteswap(buf + pos, inc, &zc_bswap); pos += inc; } VERIFY3U(pos, ==, size); VERIFY(ZIO_CHECKSUM_EQUAL(zc, zc_ref)); VERIFY(ZIO_CHECKSUM_EQUAL(zc_bswap, zc_ref_bswap)); /* * verify if incremental on the whole buffer is * equivalent to non-incremental version */ ZIO_SET_CHECKSUM(&zc, 0, 0, 0, 0); ZIO_SET_CHECKSUM(&zc_bswap, 0, 0, 0, 0); fletcher_4_incremental_native(buf, size, &zc); fletcher_4_incremental_byteswap(buf, size, &zc_bswap); VERIFY(ZIO_CHECKSUM_EQUAL(zc, zc_ref)); VERIFY(ZIO_CHECKSUM_EQUAL(zc_bswap, zc_ref_bswap)); } umem_free(buf, size); } } static int ztest_set_global_vars(void) { for (size_t i = 0; i < ztest_opts.zo_gvars_count; i++) { char *kv = ztest_opts.zo_gvars[i]; VERIFY3U(strlen(kv), <=, ZO_GVARS_MAX_ARGLEN); VERIFY3U(strlen(kv), >, 0); int err = set_global_var(kv); if (ztest_opts.zo_verbose > 0) { (void) printf("setting global var %s ... %s\n", kv, err ? "failed" : "ok"); } if (err != 0) { (void) fprintf(stderr, "failed to set global var '%s'\n", kv); return (err); } } return (0); } static char ** ztest_global_vars_to_zdb_args(void) { char **args = calloc(2*ztest_opts.zo_gvars_count + 1, sizeof (char *)); char **cur = args; + if (args == NULL) + return (NULL); for (size_t i = 0; i < ztest_opts.zo_gvars_count; i++) { *cur++ = (char *)"-o"; *cur++ = ztest_opts.zo_gvars[i]; } ASSERT3P(cur, ==, &args[2*ztest_opts.zo_gvars_count]); *cur = NULL; return (args); } /* The end of strings is indicated by a NULL element */ static char * join_strings(char **strings, const char *sep) { size_t totallen = 0; for (char **sp = strings; *sp != NULL; sp++) { totallen += strlen(*sp); totallen += strlen(sep); } if (totallen > 0) { ASSERT(totallen >= strlen(sep)); totallen -= strlen(sep); } size_t buflen = totallen + 1; char *o = malloc(buflen); /* trailing 0 byte */ o[0] = '\0'; for (char **sp = strings; *sp != NULL; sp++) { size_t would; would = strlcat(o, *sp, buflen); VERIFY3U(would, <, buflen); if (*(sp+1) == NULL) { break; } would = strlcat(o, sep, buflen); VERIFY3U(would, <, buflen); } ASSERT3S(strlen(o), ==, totallen); return (o); } static int ztest_check_path(char *path) { struct stat s; /* return true on success */ return (!stat(path, &s)); } static void ztest_get_zdb_bin(char *bin, int len) { char *zdb_path; /* * Try to use $ZDB and in-tree zdb path. If not successful, just * let popen to search through PATH. */ if ((zdb_path = getenv("ZDB"))) { strlcpy(bin, zdb_path, len); /* In env */ if (!ztest_check_path(bin)) { ztest_dump_core = 0; fatal(B_TRUE, "invalid ZDB '%s'", bin); } return; } VERIFY3P(realpath(getexecname(), bin), !=, NULL); if (strstr(bin, ".libs/ztest")) { strstr(bin, ".libs/ztest")[0] = '\0'; /* In-tree */ strcat(bin, "zdb"); if (ztest_check_path(bin)) return; } strcpy(bin, "zdb"); } static vdev_t * ztest_random_concrete_vdev_leaf(vdev_t *vd) { if (vd == NULL) return (NULL); if (vd->vdev_children == 0) return (vd); vdev_t *eligible[vd->vdev_children]; int eligible_idx = 0, i; for (i = 0; i < vd->vdev_children; i++) { vdev_t *cvd = vd->vdev_child[i]; if (cvd->vdev_top->vdev_removing) continue; if (cvd->vdev_children > 0 || (vdev_is_concrete(cvd) && !cvd->vdev_detached)) { eligible[eligible_idx++] = cvd; } } VERIFY3S(eligible_idx, >, 0); uint64_t child_no = ztest_random(eligible_idx); return (ztest_random_concrete_vdev_leaf(eligible[child_no])); } void ztest_initialize(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; spa_t *spa = ztest_spa; int error = 0; mutex_enter(&ztest_vdev_lock); spa_config_enter(spa, SCL_VDEV, FTAG, RW_READER); /* Random leaf vdev */ vdev_t *rand_vd = ztest_random_concrete_vdev_leaf(spa->spa_root_vdev); if (rand_vd == NULL) { spa_config_exit(spa, SCL_VDEV, FTAG); mutex_exit(&ztest_vdev_lock); return; } /* * The random vdev we've selected may change as soon as we * drop the spa_config_lock. We create local copies of things * we're interested in. */ uint64_t guid = rand_vd->vdev_guid; char *path = strdup(rand_vd->vdev_path); boolean_t active = rand_vd->vdev_initialize_thread != NULL; zfs_dbgmsg("vd %px, guid %llu", rand_vd, (u_longlong_t)guid); spa_config_exit(spa, SCL_VDEV, FTAG); uint64_t cmd = ztest_random(POOL_INITIALIZE_FUNCS); nvlist_t *vdev_guids = fnvlist_alloc(); nvlist_t *vdev_errlist = fnvlist_alloc(); fnvlist_add_uint64(vdev_guids, path, guid); error = spa_vdev_initialize(spa, vdev_guids, cmd, vdev_errlist); fnvlist_free(vdev_guids); fnvlist_free(vdev_errlist); switch (cmd) { case POOL_INITIALIZE_CANCEL: if (ztest_opts.zo_verbose >= 4) { (void) printf("Cancel initialize %s", path); if (!active) (void) printf(" failed (no initialize active)"); (void) printf("\n"); } break; case POOL_INITIALIZE_START: if (ztest_opts.zo_verbose >= 4) { (void) printf("Start initialize %s", path); if (active && error == 0) (void) printf(" failed (already active)"); else if (error != 0) (void) printf(" failed (error %d)", error); (void) printf("\n"); } break; case POOL_INITIALIZE_SUSPEND: if (ztest_opts.zo_verbose >= 4) { (void) printf("Suspend initialize %s", path); if (!active) (void) printf(" failed (no initialize active)"); (void) printf("\n"); } break; } free(path); mutex_exit(&ztest_vdev_lock); } void ztest_trim(ztest_ds_t *zd, uint64_t id) { (void) zd, (void) id; spa_t *spa = ztest_spa; int error = 0; mutex_enter(&ztest_vdev_lock); spa_config_enter(spa, SCL_VDEV, FTAG, RW_READER); /* Random leaf vdev */ vdev_t *rand_vd = ztest_random_concrete_vdev_leaf(spa->spa_root_vdev); if (rand_vd == NULL) { spa_config_exit(spa, SCL_VDEV, FTAG); mutex_exit(&ztest_vdev_lock); return; } /* * The random vdev we've selected may change as soon as we * drop the spa_config_lock. We create local copies of things * we're interested in. */ uint64_t guid = rand_vd->vdev_guid; char *path = strdup(rand_vd->vdev_path); boolean_t active = rand_vd->vdev_trim_thread != NULL; zfs_dbgmsg("vd %p, guid %llu", rand_vd, (u_longlong_t)guid); spa_config_exit(spa, SCL_VDEV, FTAG); uint64_t cmd = ztest_random(POOL_TRIM_FUNCS); uint64_t rate = 1 << ztest_random(30); boolean_t partial = (ztest_random(5) > 0); boolean_t secure = (ztest_random(5) > 0); nvlist_t *vdev_guids = fnvlist_alloc(); nvlist_t *vdev_errlist = fnvlist_alloc(); fnvlist_add_uint64(vdev_guids, path, guid); error = spa_vdev_trim(spa, vdev_guids, cmd, rate, partial, secure, vdev_errlist); fnvlist_free(vdev_guids); fnvlist_free(vdev_errlist); switch (cmd) { case POOL_TRIM_CANCEL: if (ztest_opts.zo_verbose >= 4) { (void) printf("Cancel TRIM %s", path); if (!active) (void) printf(" failed (no TRIM active)"); (void) printf("\n"); } break; case POOL_TRIM_START: if (ztest_opts.zo_verbose >= 4) { (void) printf("Start TRIM %s", path); if (active && error == 0) (void) printf(" failed (already active)"); else if (error != 0) (void) printf(" failed (error %d)", error); (void) printf("\n"); } break; case POOL_TRIM_SUSPEND: if (ztest_opts.zo_verbose >= 4) { (void) printf("Suspend TRIM %s", path); if (!active) (void) printf(" failed (no TRIM active)"); (void) printf("\n"); } break; } free(path); mutex_exit(&ztest_vdev_lock); } /* * Verify pool integrity by running zdb. */ static void ztest_run_zdb(const char *pool) { int status; char *bin; char *zdb; char *zbuf; const int len = MAXPATHLEN + MAXNAMELEN + 20; FILE *fp; bin = umem_alloc(len, UMEM_NOFAIL); zdb = umem_alloc(len, UMEM_NOFAIL); zbuf = umem_alloc(1024, UMEM_NOFAIL); ztest_get_zdb_bin(bin, len); char **set_gvars_args = ztest_global_vars_to_zdb_args(); + if (set_gvars_args == NULL) { + fatal(B_FALSE, "Failed to allocate memory in " + "ztest_global_vars_to_zdb_args(). Cannot run zdb.\n"); + } char *set_gvars_args_joined = join_strings(set_gvars_args, " "); free(set_gvars_args); size_t would = snprintf(zdb, len, "%s -bcc%s%s -G -d -Y -e -y %s -p %s %s", bin, ztest_opts.zo_verbose >= 3 ? "s" : "", ztest_opts.zo_verbose >= 4 ? "v" : "", set_gvars_args_joined, ztest_opts.zo_dir, pool); ASSERT3U(would, <, len); free(set_gvars_args_joined); if (ztest_opts.zo_verbose >= 5) (void) printf("Executing %s\n", zdb); fp = popen(zdb, "r"); while (fgets(zbuf, 1024, fp) != NULL) if (ztest_opts.zo_verbose >= 3) (void) printf("%s", zbuf); status = pclose(fp); if (status == 0) goto out; ztest_dump_core = 0; if (WIFEXITED(status)) fatal(B_FALSE, "'%s' exit code %d", zdb, WEXITSTATUS(status)); else fatal(B_FALSE, "'%s' died with signal %d", zdb, WTERMSIG(status)); out: umem_free(bin, len); umem_free(zdb, len); umem_free(zbuf, 1024); } static void ztest_walk_pool_directory(const char *header) { spa_t *spa = NULL; if (ztest_opts.zo_verbose >= 6) (void) puts(header); mutex_enter(&spa_namespace_lock); while ((spa = spa_next(spa)) != NULL) if (ztest_opts.zo_verbose >= 6) (void) printf("\t%s\n", spa_name(spa)); mutex_exit(&spa_namespace_lock); } static void ztest_spa_import_export(char *oldname, char *newname) { nvlist_t *config, *newconfig; uint64_t pool_guid; spa_t *spa; int error; if (ztest_opts.zo_verbose >= 4) { (void) printf("import/export: old = %s, new = %s\n", oldname, newname); } /* * Clean up from previous runs. */ (void) spa_destroy(newname); /* * Get the pool's configuration and guid. */ VERIFY0(spa_open(oldname, &spa, FTAG)); /* * Kick off a scrub to tickle scrub/export races. */ if (ztest_random(2) == 0) (void) spa_scan(spa, POOL_SCAN_SCRUB); pool_guid = spa_guid(spa); spa_close(spa, FTAG); ztest_walk_pool_directory("pools before export"); /* * Export it. */ VERIFY0(spa_export(oldname, &config, B_FALSE, B_FALSE)); ztest_walk_pool_directory("pools after export"); /* * Try to import it. */ newconfig = spa_tryimport(config); ASSERT3P(newconfig, !=, NULL); fnvlist_free(newconfig); /* * Import it under the new name. */ error = spa_import(newname, config, NULL, 0); if (error != 0) { dump_nvlist(config, 0); fatal(B_FALSE, "couldn't import pool %s as %s: error %u", oldname, newname, error); } ztest_walk_pool_directory("pools after import"); /* * Try to import it again -- should fail with EEXIST. */ VERIFY3U(EEXIST, ==, spa_import(newname, config, NULL, 0)); /* * Try to import it under a different name -- should fail with EEXIST. */ VERIFY3U(EEXIST, ==, spa_import(oldname, config, NULL, 0)); /* * Verify that the pool is no longer visible under the old name. */ VERIFY3U(ENOENT, ==, spa_open(oldname, &spa, FTAG)); /* * Verify that we can open and close the pool using the new name. */ VERIFY0(spa_open(newname, &spa, FTAG)); ASSERT3U(pool_guid, ==, spa_guid(spa)); spa_close(spa, FTAG); fnvlist_free(config); } static void ztest_resume(spa_t *spa) { if (spa_suspended(spa) && ztest_opts.zo_verbose >= 6) (void) printf("resuming from suspended state\n"); spa_vdev_state_enter(spa, SCL_NONE); vdev_clear(spa, NULL); (void) spa_vdev_state_exit(spa, NULL, 0); (void) zio_resume(spa); } static __attribute__((noreturn)) void ztest_resume_thread(void *arg) { spa_t *spa = arg; while (!ztest_exiting) { if (spa_suspended(spa)) ztest_resume(spa); (void) poll(NULL, 0, 100); /* * Periodically change the zfs_compressed_arc_enabled setting. */ if (ztest_random(10) == 0) zfs_compressed_arc_enabled = ztest_random(2); /* * Periodically change the zfs_abd_scatter_enabled setting. */ if (ztest_random(10) == 0) zfs_abd_scatter_enabled = ztest_random(2); } thread_exit(); } static __attribute__((noreturn)) void ztest_deadman_thread(void *arg) { ztest_shared_t *zs = arg; spa_t *spa = ztest_spa; hrtime_t delay, overdue, last_run = gethrtime(); delay = (zs->zs_thread_stop - zs->zs_thread_start) + MSEC2NSEC(zfs_deadman_synctime_ms); while (!ztest_exiting) { /* * Wait for the delay timer while checking occasionally * if we should stop. */ if (gethrtime() < last_run + delay) { (void) poll(NULL, 0, 1000); continue; } /* * If the pool is suspended then fail immediately. Otherwise, * check to see if the pool is making any progress. If * vdev_deadman() discovers that there hasn't been any recent * I/Os then it will end up aborting the tests. */ if (spa_suspended(spa) || spa->spa_root_vdev == NULL) { fatal(B_FALSE, "aborting test after %lu seconds because " "pool has transitioned to a suspended state.", zfs_deadman_synctime_ms / 1000); } vdev_deadman(spa->spa_root_vdev, FTAG); /* * If the process doesn't complete within a grace period of * zfs_deadman_synctime_ms over the expected finish time, * then it may be hung and is terminated. */ overdue = zs->zs_proc_stop + MSEC2NSEC(zfs_deadman_synctime_ms); if (gethrtime() > overdue) { fatal(B_FALSE, "aborting test after %llu seconds because " "the process is overdue for termination.", (gethrtime() - zs->zs_proc_start) / NANOSEC); } (void) printf("ztest has been running for %lld seconds\n", (gethrtime() - zs->zs_proc_start) / NANOSEC); last_run = gethrtime(); delay = MSEC2NSEC(zfs_deadman_checktime_ms); } thread_exit(); } static void ztest_execute(int test, ztest_info_t *zi, uint64_t id) { ztest_ds_t *zd = &ztest_ds[id % ztest_opts.zo_datasets]; ztest_shared_callstate_t *zc = ZTEST_GET_SHARED_CALLSTATE(test); hrtime_t functime = gethrtime(); int i; for (i = 0; i < zi->zi_iters; i++) zi->zi_func(zd, id); functime = gethrtime() - functime; atomic_add_64(&zc->zc_count, 1); atomic_add_64(&zc->zc_time, functime); if (ztest_opts.zo_verbose >= 4) (void) printf("%6.2f sec in %s\n", (double)functime / NANOSEC, zi->zi_funcname); } static __attribute__((noreturn)) void ztest_thread(void *arg) { int rand; uint64_t id = (uintptr_t)arg; ztest_shared_t *zs = ztest_shared; uint64_t call_next; hrtime_t now; ztest_info_t *zi; ztest_shared_callstate_t *zc; while ((now = gethrtime()) < zs->zs_thread_stop) { /* * See if it's time to force a crash. */ if (now > zs->zs_thread_kill) ztest_kill(zs); /* * If we're getting ENOSPC with some regularity, stop. */ if (zs->zs_enospc_count > 10) break; /* * Pick a random function to execute. */ rand = ztest_random(ZTEST_FUNCS); zi = &ztest_info[rand]; zc = ZTEST_GET_SHARED_CALLSTATE(rand); call_next = zc->zc_next; if (now >= call_next && atomic_cas_64(&zc->zc_next, call_next, call_next + ztest_random(2 * zi->zi_interval[0] + 1)) == call_next) { ztest_execute(rand, zi, id); } } thread_exit(); } static void ztest_dataset_name(char *dsname, const char *pool, int d) { (void) snprintf(dsname, ZFS_MAX_DATASET_NAME_LEN, "%s/ds_%d", pool, d); } static void ztest_dataset_destroy(int d) { char name[ZFS_MAX_DATASET_NAME_LEN]; int t; ztest_dataset_name(name, ztest_opts.zo_pool, d); if (ztest_opts.zo_verbose >= 3) (void) printf("Destroying %s to free up space\n", name); /* * Cleanup any non-standard clones and snapshots. In general, * ztest thread t operates on dataset (t % zopt_datasets), * so there may be more than one thing to clean up. */ for (t = d; t < ztest_opts.zo_threads; t += ztest_opts.zo_datasets) ztest_dsl_dataset_cleanup(name, t); (void) dmu_objset_find(name, ztest_objset_destroy_cb, NULL, DS_FIND_SNAPSHOTS | DS_FIND_CHILDREN); } static void ztest_dataset_dirobj_verify(ztest_ds_t *zd) { uint64_t usedobjs, dirobjs, scratch; /* * ZTEST_DIROBJ is the object directory for the entire dataset. * Therefore, the number of objects in use should equal the * number of ZTEST_DIROBJ entries, +1 for ZTEST_DIROBJ itself. * If not, we have an object leak. * * Note that we can only check this in ztest_dataset_open(), * when the open-context and syncing-context values agree. * That's because zap_count() returns the open-context value, * while dmu_objset_space() returns the rootbp fill count. */ VERIFY0(zap_count(zd->zd_os, ZTEST_DIROBJ, &dirobjs)); dmu_objset_space(zd->zd_os, &scratch, &scratch, &usedobjs, &scratch); ASSERT3U(dirobjs + 1, ==, usedobjs); } static int ztest_dataset_open(int d) { ztest_ds_t *zd = &ztest_ds[d]; uint64_t committed_seq = ZTEST_GET_SHARED_DS(d)->zd_seq; objset_t *os; zilog_t *zilog; char name[ZFS_MAX_DATASET_NAME_LEN]; int error; ztest_dataset_name(name, ztest_opts.zo_pool, d); (void) pthread_rwlock_rdlock(&ztest_name_lock); error = ztest_dataset_create(name); if (error == ENOSPC) { (void) pthread_rwlock_unlock(&ztest_name_lock); ztest_record_enospc(FTAG); return (error); } ASSERT(error == 0 || error == EEXIST); VERIFY0(ztest_dmu_objset_own(name, DMU_OST_OTHER, B_FALSE, B_TRUE, zd, &os)); (void) pthread_rwlock_unlock(&ztest_name_lock); ztest_zd_init(zd, ZTEST_GET_SHARED_DS(d), os); zilog = zd->zd_zilog; if (zilog->zl_header->zh_claim_lr_seq != 0 && zilog->zl_header->zh_claim_lr_seq < committed_seq) fatal(B_FALSE, "missing log records: " "claimed %"PRIu64" < committed %"PRIu64"", zilog->zl_header->zh_claim_lr_seq, committed_seq); ztest_dataset_dirobj_verify(zd); zil_replay(os, zd, ztest_replay_vector); ztest_dataset_dirobj_verify(zd); if (ztest_opts.zo_verbose >= 6) (void) printf("%s replay %"PRIu64" blocks, " "%"PRIu64" records, seq %"PRIu64"\n", zd->zd_name, zilog->zl_parse_blk_count, zilog->zl_parse_lr_count, zilog->zl_replaying_seq); zilog = zil_open(os, ztest_get_data, NULL); if (zilog->zl_replaying_seq != 0 && zilog->zl_replaying_seq < committed_seq) fatal(B_FALSE, "missing log records: " "replayed %"PRIu64" < committed %"PRIu64"", zilog->zl_replaying_seq, committed_seq); return (0); } static void ztest_dataset_close(int d) { ztest_ds_t *zd = &ztest_ds[d]; zil_close(zd->zd_zilog); dmu_objset_disown(zd->zd_os, B_TRUE, zd); ztest_zd_fini(zd); } static int ztest_replay_zil_cb(const char *name, void *arg) { (void) arg; objset_t *os; ztest_ds_t *zdtmp; VERIFY0(ztest_dmu_objset_own(name, DMU_OST_ANY, B_TRUE, B_TRUE, FTAG, &os)); zdtmp = umem_alloc(sizeof (ztest_ds_t), UMEM_NOFAIL); ztest_zd_init(zdtmp, NULL, os); zil_replay(os, zdtmp, ztest_replay_vector); ztest_zd_fini(zdtmp); if (dmu_objset_zil(os)->zl_parse_lr_count != 0 && ztest_opts.zo_verbose >= 6) { zilog_t *zilog = dmu_objset_zil(os); (void) printf("%s replay %"PRIu64" blocks, " "%"PRIu64" records, seq %"PRIu64"\n", name, zilog->zl_parse_blk_count, zilog->zl_parse_lr_count, zilog->zl_replaying_seq); } umem_free(zdtmp, sizeof (ztest_ds_t)); dmu_objset_disown(os, B_TRUE, FTAG); return (0); } static void ztest_freeze(void) { ztest_ds_t *zd = &ztest_ds[0]; spa_t *spa; int numloops = 0; if (ztest_opts.zo_verbose >= 3) (void) printf("testing spa_freeze()...\n"); kernel_init(SPA_MODE_READ | SPA_MODE_WRITE); VERIFY0(spa_open(ztest_opts.zo_pool, &spa, FTAG)); VERIFY0(ztest_dataset_open(0)); ztest_spa = spa; /* * Force the first log block to be transactionally allocated. * We have to do this before we freeze the pool -- otherwise * the log chain won't be anchored. */ while (BP_IS_HOLE(&zd->zd_zilog->zl_header->zh_log)) { ztest_dmu_object_alloc_free(zd, 0); zil_commit(zd->zd_zilog, 0); } txg_wait_synced(spa_get_dsl(spa), 0); /* * Freeze the pool. This stops spa_sync() from doing anything, * so that the only way to record changes from now on is the ZIL. */ spa_freeze(spa); /* * Because it is hard to predict how much space a write will actually * require beforehand, we leave ourselves some fudge space to write over * capacity. */ uint64_t capacity = metaslab_class_get_space(spa_normal_class(spa)) / 2; /* * Run tests that generate log records but don't alter the pool config * or depend on DSL sync tasks (snapshots, objset create/destroy, etc). * We do a txg_wait_synced() after each iteration to force the txg * to increase well beyond the last synced value in the uberblock. * The ZIL should be OK with that. * * Run a random number of times less than zo_maxloops and ensure we do * not run out of space on the pool. */ while (ztest_random(10) != 0 && numloops++ < ztest_opts.zo_maxloops && metaslab_class_get_alloc(spa_normal_class(spa)) < capacity) { ztest_od_t od; ztest_od_init(&od, 0, FTAG, 0, DMU_OT_UINT64_OTHER, 0, 0, 0); VERIFY0(ztest_object_init(zd, &od, sizeof (od), B_FALSE)); ztest_io(zd, od.od_object, ztest_random(ZTEST_RANGE_LOCKS) << SPA_MAXBLOCKSHIFT); txg_wait_synced(spa_get_dsl(spa), 0); } /* * Commit all of the changes we just generated. */ zil_commit(zd->zd_zilog, 0); txg_wait_synced(spa_get_dsl(spa), 0); /* * Close our dataset and close the pool. */ ztest_dataset_close(0); spa_close(spa, FTAG); kernel_fini(); /* * Open and close the pool and dataset to induce log replay. */ kernel_init(SPA_MODE_READ | SPA_MODE_WRITE); VERIFY0(spa_open(ztest_opts.zo_pool, &spa, FTAG)); ASSERT3U(spa_freeze_txg(spa), ==, UINT64_MAX); VERIFY0(ztest_dataset_open(0)); ztest_spa = spa; txg_wait_synced(spa_get_dsl(spa), 0); ztest_dataset_close(0); ztest_reguid(NULL, 0); spa_close(spa, FTAG); kernel_fini(); } static void ztest_import_impl(void) { importargs_t args = { 0 }; nvlist_t *cfg = NULL; int nsearch = 1; char *searchdirs[nsearch]; int flags = ZFS_IMPORT_MISSING_LOG; searchdirs[0] = ztest_opts.zo_dir; args.paths = nsearch; args.path = searchdirs; args.can_be_active = B_FALSE; VERIFY0(zpool_find_config(NULL, ztest_opts.zo_pool, &cfg, &args, &libzpool_config_ops)); VERIFY0(spa_import(ztest_opts.zo_pool, cfg, NULL, flags)); fnvlist_free(cfg); } /* * Import a storage pool with the given name. */ static void ztest_import(ztest_shared_t *zs) { spa_t *spa; mutex_init(&ztest_vdev_lock, NULL, MUTEX_DEFAULT, NULL); mutex_init(&ztest_checkpoint_lock, NULL, MUTEX_DEFAULT, NULL); VERIFY0(pthread_rwlock_init(&ztest_name_lock, NULL)); kernel_init(SPA_MODE_READ | SPA_MODE_WRITE); ztest_import_impl(); VERIFY0(spa_open(ztest_opts.zo_pool, &spa, FTAG)); zs->zs_metaslab_sz = 1ULL << spa->spa_root_vdev->vdev_child[0]->vdev_ms_shift; spa_close(spa, FTAG); kernel_fini(); if (!ztest_opts.zo_mmp_test) { ztest_run_zdb(ztest_opts.zo_pool); ztest_freeze(); ztest_run_zdb(ztest_opts.zo_pool); } (void) pthread_rwlock_destroy(&ztest_name_lock); mutex_destroy(&ztest_vdev_lock); mutex_destroy(&ztest_checkpoint_lock); } /* * Kick off threads to run tests on all datasets in parallel. */ static void ztest_run(ztest_shared_t *zs) { spa_t *spa; objset_t *os; kthread_t *resume_thread, *deadman_thread; kthread_t **run_threads; uint64_t object; int error; int t, d; ztest_exiting = B_FALSE; /* * Initialize parent/child shared state. */ mutex_init(&ztest_vdev_lock, NULL, MUTEX_DEFAULT, NULL); mutex_init(&ztest_checkpoint_lock, NULL, MUTEX_DEFAULT, NULL); VERIFY0(pthread_rwlock_init(&ztest_name_lock, NULL)); zs->zs_thread_start = gethrtime(); zs->zs_thread_stop = zs->zs_thread_start + ztest_opts.zo_passtime * NANOSEC; zs->zs_thread_stop = MIN(zs->zs_thread_stop, zs->zs_proc_stop); zs->zs_thread_kill = zs->zs_thread_stop; if (ztest_random(100) < ztest_opts.zo_killrate) { zs->zs_thread_kill -= ztest_random(ztest_opts.zo_passtime * NANOSEC); } mutex_init(&zcl.zcl_callbacks_lock, NULL, MUTEX_DEFAULT, NULL); list_create(&zcl.zcl_callbacks, sizeof (ztest_cb_data_t), offsetof(ztest_cb_data_t, zcd_node)); /* * Open our pool. It may need to be imported first depending on * what tests were running when the previous pass was terminated. */ kernel_init(SPA_MODE_READ | SPA_MODE_WRITE); error = spa_open(ztest_opts.zo_pool, &spa, FTAG); if (error) { VERIFY3S(error, ==, ENOENT); ztest_import_impl(); VERIFY0(spa_open(ztest_opts.zo_pool, &spa, FTAG)); zs->zs_metaslab_sz = 1ULL << spa->spa_root_vdev->vdev_child[0]->vdev_ms_shift; } metaslab_preload_limit = ztest_random(20) + 1; ztest_spa = spa; VERIFY0(vdev_raidz_impl_set("cycle")); dmu_objset_stats_t dds; VERIFY0(ztest_dmu_objset_own(ztest_opts.zo_pool, DMU_OST_ANY, B_TRUE, B_TRUE, FTAG, &os)); dsl_pool_config_enter(dmu_objset_pool(os), FTAG); dmu_objset_fast_stat(os, &dds); dsl_pool_config_exit(dmu_objset_pool(os), FTAG); zs->zs_guid = dds.dds_guid; dmu_objset_disown(os, B_TRUE, FTAG); /* * Create a thread to periodically resume suspended I/O. */ resume_thread = thread_create(NULL, 0, ztest_resume_thread, spa, 0, NULL, TS_RUN | TS_JOINABLE, defclsyspri); /* * Create a deadman thread and set to panic if we hang. */ deadman_thread = thread_create(NULL, 0, ztest_deadman_thread, zs, 0, NULL, TS_RUN | TS_JOINABLE, defclsyspri); spa->spa_deadman_failmode = ZIO_FAILURE_MODE_PANIC; /* * Verify that we can safely inquire about any object, * whether it's allocated or not. To make it interesting, * we probe a 5-wide window around each power of two. * This hits all edge cases, including zero and the max. */ for (t = 0; t < 64; t++) { for (d = -5; d <= 5; d++) { error = dmu_object_info(spa->spa_meta_objset, (1ULL << t) + d, NULL); ASSERT(error == 0 || error == ENOENT || error == EINVAL); } } /* * If we got any ENOSPC errors on the previous run, destroy something. */ if (zs->zs_enospc_count != 0) { int d = ztest_random(ztest_opts.zo_datasets); ztest_dataset_destroy(d); } zs->zs_enospc_count = 0; /* * If we were in the middle of ztest_device_removal() and were killed * we need to ensure the removal and scrub complete before running * any tests that check ztest_device_removal_active. The removal will * be restarted automatically when the spa is opened, but we need to * initiate the scrub manually if it is not already in progress. Note * that we always run the scrub whenever an indirect vdev exists * because we have no way of knowing for sure if ztest_device_removal() * fully completed its scrub before the pool was reimported. */ if (spa->spa_removing_phys.sr_state == DSS_SCANNING || spa->spa_removing_phys.sr_prev_indirect_vdev != -1) { while (spa->spa_removing_phys.sr_state == DSS_SCANNING) txg_wait_synced(spa_get_dsl(spa), 0); error = ztest_scrub_impl(spa); if (error == EBUSY) error = 0; ASSERT0(error); } run_threads = umem_zalloc(ztest_opts.zo_threads * sizeof (kthread_t *), UMEM_NOFAIL); if (ztest_opts.zo_verbose >= 4) (void) printf("starting main threads...\n"); /* * Replay all logs of all datasets in the pool. This is primarily for * temporary datasets which wouldn't otherwise get replayed, which * can trigger failures when attempting to offline a SLOG in * ztest_fault_inject(). */ (void) dmu_objset_find(ztest_opts.zo_pool, ztest_replay_zil_cb, NULL, DS_FIND_CHILDREN); /* * Kick off all the tests that run in parallel. */ for (t = 0; t < ztest_opts.zo_threads; t++) { if (t < ztest_opts.zo_datasets && ztest_dataset_open(t) != 0) { umem_free(run_threads, ztest_opts.zo_threads * sizeof (kthread_t *)); return; } run_threads[t] = thread_create(NULL, 0, ztest_thread, (void *)(uintptr_t)t, 0, NULL, TS_RUN | TS_JOINABLE, defclsyspri); } /* * Wait for all of the tests to complete. */ for (t = 0; t < ztest_opts.zo_threads; t++) VERIFY0(thread_join(run_threads[t])); /* * Close all datasets. This must be done after all the threads * are joined so we can be sure none of the datasets are in-use * by any of the threads. */ for (t = 0; t < ztest_opts.zo_threads; t++) { if (t < ztest_opts.zo_datasets) ztest_dataset_close(t); } txg_wait_synced(spa_get_dsl(spa), 0); zs->zs_alloc = metaslab_class_get_alloc(spa_normal_class(spa)); zs->zs_space = metaslab_class_get_space(spa_normal_class(spa)); umem_free(run_threads, ztest_opts.zo_threads * sizeof (kthread_t *)); /* Kill the resume and deadman threads */ ztest_exiting = B_TRUE; VERIFY0(thread_join(resume_thread)); VERIFY0(thread_join(deadman_thread)); ztest_resume(spa); /* * Right before closing the pool, kick off a bunch of async I/O; * spa_close() should wait for it to complete. */ for (object = 1; object < 50; object++) { dmu_prefetch(spa->spa_meta_objset, object, 0, 0, 1ULL << 20, ZIO_PRIORITY_SYNC_READ); } /* Verify that at least one commit cb was called in a timely fashion */ if (zc_cb_counter >= ZTEST_COMMIT_CB_MIN_REG) VERIFY0(zc_min_txg_delay); spa_close(spa, FTAG); /* * Verify that we can loop over all pools. */ mutex_enter(&spa_namespace_lock); for (spa = spa_next(NULL); spa != NULL; spa = spa_next(spa)) if (ztest_opts.zo_verbose > 3) (void) printf("spa_next: found %s\n", spa_name(spa)); mutex_exit(&spa_namespace_lock); /* * Verify that we can export the pool and reimport it under a * different name. */ if ((ztest_random(2) == 0) && !ztest_opts.zo_mmp_test) { char name[ZFS_MAX_DATASET_NAME_LEN]; (void) snprintf(name, sizeof (name), "%s_import", ztest_opts.zo_pool); ztest_spa_import_export(ztest_opts.zo_pool, name); ztest_spa_import_export(name, ztest_opts.zo_pool); } kernel_fini(); list_destroy(&zcl.zcl_callbacks); mutex_destroy(&zcl.zcl_callbacks_lock); (void) pthread_rwlock_destroy(&ztest_name_lock); mutex_destroy(&ztest_vdev_lock); mutex_destroy(&ztest_checkpoint_lock); } static void print_time(hrtime_t t, char *timebuf) { hrtime_t s = t / NANOSEC; hrtime_t m = s / 60; hrtime_t h = m / 60; hrtime_t d = h / 24; s -= m * 60; m -= h * 60; h -= d * 24; timebuf[0] = '\0'; if (d) (void) sprintf(timebuf, "%llud%02lluh%02llum%02llus", d, h, m, s); else if (h) (void) sprintf(timebuf, "%lluh%02llum%02llus", h, m, s); else if (m) (void) sprintf(timebuf, "%llum%02llus", m, s); else (void) sprintf(timebuf, "%llus", s); } static nvlist_t * make_random_props(void) { nvlist_t *props; props = fnvlist_alloc(); if (ztest_random(2) == 0) return (props); fnvlist_add_uint64(props, zpool_prop_to_name(ZPOOL_PROP_AUTOREPLACE), 1); return (props); } /* * Create a storage pool with the given name and initial vdev size. * Then test spa_freeze() functionality. */ static void ztest_init(ztest_shared_t *zs) { spa_t *spa; nvlist_t *nvroot, *props; int i; mutex_init(&ztest_vdev_lock, NULL, MUTEX_DEFAULT, NULL); mutex_init(&ztest_checkpoint_lock, NULL, MUTEX_DEFAULT, NULL); VERIFY0(pthread_rwlock_init(&ztest_name_lock, NULL)); kernel_init(SPA_MODE_READ | SPA_MODE_WRITE); /* * Create the storage pool. */ (void) spa_destroy(ztest_opts.zo_pool); ztest_shared->zs_vdev_next_leaf = 0; zs->zs_splits = 0; zs->zs_mirrors = ztest_opts.zo_mirrors; nvroot = make_vdev_root(NULL, NULL, NULL, ztest_opts.zo_vdev_size, 0, NULL, ztest_opts.zo_raid_children, zs->zs_mirrors, 1); props = make_random_props(); /* * We don't expect the pool to suspend unless maxfaults == 0, * in which case ztest_fault_inject() temporarily takes away * the only valid replica. */ fnvlist_add_uint64(props, zpool_prop_to_name(ZPOOL_PROP_FAILUREMODE), MAXFAULTS(zs) ? ZIO_FAILURE_MODE_PANIC : ZIO_FAILURE_MODE_WAIT); for (i = 0; i < SPA_FEATURES; i++) { char *buf; if (!spa_feature_table[i].fi_zfs_mod_supported) continue; /* * 75% chance of using the log space map feature. We want ztest * to exercise both the code paths that use the log space map * feature and the ones that don't. */ if (i == SPA_FEATURE_LOG_SPACEMAP && ztest_random(4) == 0) continue; VERIFY3S(-1, !=, asprintf(&buf, "feature@%s", spa_feature_table[i].fi_uname)); fnvlist_add_uint64(props, buf, 0); free(buf); } VERIFY0(spa_create(ztest_opts.zo_pool, nvroot, props, NULL, NULL)); fnvlist_free(nvroot); fnvlist_free(props); VERIFY0(spa_open(ztest_opts.zo_pool, &spa, FTAG)); zs->zs_metaslab_sz = 1ULL << spa->spa_root_vdev->vdev_child[0]->vdev_ms_shift; spa_close(spa, FTAG); kernel_fini(); if (!ztest_opts.zo_mmp_test) { ztest_run_zdb(ztest_opts.zo_pool); ztest_freeze(); ztest_run_zdb(ztest_opts.zo_pool); } (void) pthread_rwlock_destroy(&ztest_name_lock); mutex_destroy(&ztest_vdev_lock); mutex_destroy(&ztest_checkpoint_lock); } static void setup_data_fd(void) { static char ztest_name_data[] = "/tmp/ztest.data.XXXXXX"; ztest_fd_data = mkstemp(ztest_name_data); ASSERT3S(ztest_fd_data, >=, 0); (void) unlink(ztest_name_data); } static int shared_data_size(ztest_shared_hdr_t *hdr) { int size; size = hdr->zh_hdr_size; size += hdr->zh_opts_size; size += hdr->zh_size; size += hdr->zh_stats_size * hdr->zh_stats_count; size += hdr->zh_ds_size * hdr->zh_ds_count; return (size); } static void setup_hdr(void) { int size; ztest_shared_hdr_t *hdr; hdr = (void *)mmap(0, P2ROUNDUP(sizeof (*hdr), getpagesize()), PROT_READ | PROT_WRITE, MAP_SHARED, ztest_fd_data, 0); ASSERT3P(hdr, !=, MAP_FAILED); VERIFY0(ftruncate(ztest_fd_data, sizeof (ztest_shared_hdr_t))); hdr->zh_hdr_size = sizeof (ztest_shared_hdr_t); hdr->zh_opts_size = sizeof (ztest_shared_opts_t); hdr->zh_size = sizeof (ztest_shared_t); hdr->zh_stats_size = sizeof (ztest_shared_callstate_t); hdr->zh_stats_count = ZTEST_FUNCS; hdr->zh_ds_size = sizeof (ztest_shared_ds_t); hdr->zh_ds_count = ztest_opts.zo_datasets; size = shared_data_size(hdr); VERIFY0(ftruncate(ztest_fd_data, size)); (void) munmap((caddr_t)hdr, P2ROUNDUP(sizeof (*hdr), getpagesize())); } static void setup_data(void) { int size, offset; ztest_shared_hdr_t *hdr; uint8_t *buf; hdr = (void *)mmap(0, P2ROUNDUP(sizeof (*hdr), getpagesize()), PROT_READ, MAP_SHARED, ztest_fd_data, 0); ASSERT3P(hdr, !=, MAP_FAILED); size = shared_data_size(hdr); (void) munmap((caddr_t)hdr, P2ROUNDUP(sizeof (*hdr), getpagesize())); hdr = ztest_shared_hdr = (void *)mmap(0, P2ROUNDUP(size, getpagesize()), PROT_READ | PROT_WRITE, MAP_SHARED, ztest_fd_data, 0); ASSERT3P(hdr, !=, MAP_FAILED); buf = (uint8_t *)hdr; offset = hdr->zh_hdr_size; ztest_shared_opts = (void *)&buf[offset]; offset += hdr->zh_opts_size; ztest_shared = (void *)&buf[offset]; offset += hdr->zh_size; ztest_shared_callstate = (void *)&buf[offset]; offset += hdr->zh_stats_size * hdr->zh_stats_count; ztest_shared_ds = (void *)&buf[offset]; } static boolean_t exec_child(char *cmd, char *libpath, boolean_t ignorekill, int *statusp) { pid_t pid; int status; char *cmdbuf = NULL; pid = fork(); if (cmd == NULL) { cmdbuf = umem_alloc(MAXPATHLEN, UMEM_NOFAIL); (void) strlcpy(cmdbuf, getexecname(), MAXPATHLEN); cmd = cmdbuf; } if (pid == -1) fatal(B_TRUE, "fork failed"); if (pid == 0) { /* child */ char fd_data_str[12]; VERIFY3S(11, >=, snprintf(fd_data_str, 12, "%d", ztest_fd_data)); VERIFY0(setenv("ZTEST_FD_DATA", fd_data_str, 1)); if (libpath != NULL) { const char *curlp = getenv("LD_LIBRARY_PATH"); if (curlp == NULL) VERIFY0(setenv("LD_LIBRARY_PATH", libpath, 1)); else { char *newlp = NULL; VERIFY3S(-1, !=, asprintf(&newlp, "%s:%s", libpath, curlp)); VERIFY0(setenv("LD_LIBRARY_PATH", newlp, 1)); free(newlp); } } (void) execl(cmd, cmd, (char *)NULL); ztest_dump_core = B_FALSE; fatal(B_TRUE, "exec failed: %s", cmd); } if (cmdbuf != NULL) { umem_free(cmdbuf, MAXPATHLEN); cmd = NULL; } while (waitpid(pid, &status, 0) != pid) continue; if (statusp != NULL) *statusp = status; if (WIFEXITED(status)) { if (WEXITSTATUS(status) != 0) { (void) fprintf(stderr, "child exited with code %d\n", WEXITSTATUS(status)); exit(2); } return (B_FALSE); } else if (WIFSIGNALED(status)) { if (!ignorekill || WTERMSIG(status) != SIGKILL) { (void) fprintf(stderr, "child died with signal %d\n", WTERMSIG(status)); exit(3); } return (B_TRUE); } else { (void) fprintf(stderr, "something strange happened to child\n"); exit(4); } } static void ztest_run_init(void) { int i; ztest_shared_t *zs = ztest_shared; /* * Blow away any existing copy of zpool.cache */ (void) remove(spa_config_path); if (ztest_opts.zo_init == 0) { if (ztest_opts.zo_verbose >= 1) (void) printf("Importing pool %s\n", ztest_opts.zo_pool); ztest_import(zs); return; } /* * Create and initialize our storage pool. */ for (i = 1; i <= ztest_opts.zo_init; i++) { memset(zs, 0, sizeof (*zs)); if (ztest_opts.zo_verbose >= 3 && ztest_opts.zo_init != 1) { (void) printf("ztest_init(), pass %d\n", i); } ztest_init(zs); } } int main(int argc, char **argv) { int kills = 0; int iters = 0; int older = 0; int newer = 0; ztest_shared_t *zs; ztest_info_t *zi; ztest_shared_callstate_t *zc; char timebuf[100]; char numbuf[NN_NUMBUF_SZ]; char *cmd; boolean_t hasalt; int f, err; char *fd_data_str = getenv("ZTEST_FD_DATA"); struct sigaction action; (void) setvbuf(stdout, NULL, _IOLBF, 0); dprintf_setup(&argc, argv); zfs_deadman_synctime_ms = 300000; zfs_deadman_checktime_ms = 30000; /* * As two-word space map entries may not come up often (especially * if pool and vdev sizes are small) we want to force at least some * of them so the feature get tested. */ zfs_force_some_double_word_sm_entries = B_TRUE; /* * Verify that even extensively damaged split blocks with many * segments can be reconstructed in a reasonable amount of time * when reconstruction is known to be possible. * * Note: the lower this value is, the more damage we inflict, and * the more time ztest spends in recovering that damage. We chose * to induce damage 1/100th of the time so recovery is tested but * not so frequently that ztest doesn't get to test other code paths. */ zfs_reconstruct_indirect_damage_fraction = 100; action.sa_handler = sig_handler; sigemptyset(&action.sa_mask); action.sa_flags = 0; if (sigaction(SIGSEGV, &action, NULL) < 0) { (void) fprintf(stderr, "ztest: cannot catch SIGSEGV: %s.\n", strerror(errno)); exit(EXIT_FAILURE); } if (sigaction(SIGABRT, &action, NULL) < 0) { (void) fprintf(stderr, "ztest: cannot catch SIGABRT: %s.\n", strerror(errno)); exit(EXIT_FAILURE); } /* * Force random_get_bytes() to use /dev/urandom in order to prevent * ztest from needlessly depleting the system entropy pool. */ random_path = "/dev/urandom"; ztest_fd_rand = open(random_path, O_RDONLY | O_CLOEXEC); ASSERT3S(ztest_fd_rand, >=, 0); if (!fd_data_str) { process_options(argc, argv); setup_data_fd(); setup_hdr(); setup_data(); memcpy(ztest_shared_opts, &ztest_opts, sizeof (*ztest_shared_opts)); } else { ztest_fd_data = atoi(fd_data_str); setup_data(); memcpy(&ztest_opts, ztest_shared_opts, sizeof (ztest_opts)); } ASSERT3U(ztest_opts.zo_datasets, ==, ztest_shared_hdr->zh_ds_count); err = ztest_set_global_vars(); if (err != 0 && !fd_data_str) { /* error message done by ztest_set_global_vars */ exit(EXIT_FAILURE); } else { /* children should not be spawned if setting gvars fails */ VERIFY3S(err, ==, 0); } /* Override location of zpool.cache */ VERIFY3S(asprintf((char **)&spa_config_path, "%s/zpool.cache", ztest_opts.zo_dir), !=, -1); ztest_ds = umem_alloc(ztest_opts.zo_datasets * sizeof (ztest_ds_t), UMEM_NOFAIL); zs = ztest_shared; if (fd_data_str) { metaslab_force_ganging = ztest_opts.zo_metaslab_force_ganging; metaslab_df_alloc_threshold = zs->zs_metaslab_df_alloc_threshold; if (zs->zs_do_init) ztest_run_init(); else ztest_run(zs); exit(0); } hasalt = (strlen(ztest_opts.zo_alt_ztest) != 0); if (ztest_opts.zo_verbose >= 1) { (void) printf("%"PRIu64" vdevs, %d datasets, %d threads," "%d %s disks, %"PRIu64" seconds...\n\n", ztest_opts.zo_vdevs, ztest_opts.zo_datasets, ztest_opts.zo_threads, ztest_opts.zo_raid_children, ztest_opts.zo_raid_type, ztest_opts.zo_time); } cmd = umem_alloc(MAXNAMELEN, UMEM_NOFAIL); (void) strlcpy(cmd, getexecname(), MAXNAMELEN); zs->zs_do_init = B_TRUE; if (strlen(ztest_opts.zo_alt_ztest) != 0) { if (ztest_opts.zo_verbose >= 1) { (void) printf("Executing older ztest for " "initialization: %s\n", ztest_opts.zo_alt_ztest); } VERIFY(!exec_child(ztest_opts.zo_alt_ztest, ztest_opts.zo_alt_libpath, B_FALSE, NULL)); } else { VERIFY(!exec_child(NULL, NULL, B_FALSE, NULL)); } zs->zs_do_init = B_FALSE; zs->zs_proc_start = gethrtime(); zs->zs_proc_stop = zs->zs_proc_start + ztest_opts.zo_time * NANOSEC; for (f = 0; f < ZTEST_FUNCS; f++) { zi = &ztest_info[f]; zc = ZTEST_GET_SHARED_CALLSTATE(f); if (zs->zs_proc_start + zi->zi_interval[0] > zs->zs_proc_stop) zc->zc_next = UINT64_MAX; else zc->zc_next = zs->zs_proc_start + ztest_random(2 * zi->zi_interval[0] + 1); } /* * Run the tests in a loop. These tests include fault injection * to verify that self-healing data works, and forced crashes * to verify that we never lose on-disk consistency. */ while (gethrtime() < zs->zs_proc_stop) { int status; boolean_t killed; /* * Initialize the workload counters for each function. */ for (f = 0; f < ZTEST_FUNCS; f++) { zc = ZTEST_GET_SHARED_CALLSTATE(f); zc->zc_count = 0; zc->zc_time = 0; } /* Set the allocation switch size */ zs->zs_metaslab_df_alloc_threshold = ztest_random(zs->zs_metaslab_sz / 4) + 1; if (!hasalt || ztest_random(2) == 0) { if (hasalt && ztest_opts.zo_verbose >= 1) { (void) printf("Executing newer ztest: %s\n", cmd); } newer++; killed = exec_child(cmd, NULL, B_TRUE, &status); } else { if (hasalt && ztest_opts.zo_verbose >= 1) { (void) printf("Executing older ztest: %s\n", ztest_opts.zo_alt_ztest); } older++; killed = exec_child(ztest_opts.zo_alt_ztest, ztest_opts.zo_alt_libpath, B_TRUE, &status); } if (killed) kills++; iters++; if (ztest_opts.zo_verbose >= 1) { hrtime_t now = gethrtime(); now = MIN(now, zs->zs_proc_stop); print_time(zs->zs_proc_stop - now, timebuf); nicenum(zs->zs_space, numbuf, sizeof (numbuf)); (void) printf("Pass %3d, %8s, %3"PRIu64" ENOSPC, " "%4.1f%% of %5s used, %3.0f%% done, %8s to go\n", iters, WIFEXITED(status) ? "Complete" : "SIGKILL", zs->zs_enospc_count, 100.0 * zs->zs_alloc / zs->zs_space, numbuf, 100.0 * (now - zs->zs_proc_start) / (ztest_opts.zo_time * NANOSEC), timebuf); } if (ztest_opts.zo_verbose >= 2) { (void) printf("\nWorkload summary:\n\n"); (void) printf("%7s %9s %s\n", "Calls", "Time", "Function"); (void) printf("%7s %9s %s\n", "-----", "----", "--------"); for (f = 0; f < ZTEST_FUNCS; f++) { zi = &ztest_info[f]; zc = ZTEST_GET_SHARED_CALLSTATE(f); print_time(zc->zc_time, timebuf); (void) printf("%7"PRIu64" %9s %s\n", zc->zc_count, timebuf, zi->zi_funcname); } (void) printf("\n"); } if (!ztest_opts.zo_mmp_test) ztest_run_zdb(ztest_opts.zo_pool); } if (ztest_opts.zo_verbose >= 1) { if (hasalt) { (void) printf("%d runs of older ztest: %s\n", older, ztest_opts.zo_alt_ztest); (void) printf("%d runs of newer ztest: %s\n", newer, cmd); } (void) printf("%d killed, %d completed, %.0f%% kill rate\n", kills, iters - kills, (100.0 * kills) / MAX(1, iters)); } umem_free(cmd, MAXNAMELEN); return (0); } diff --git a/lib/libzfs/libzfs_dataset.c b/lib/libzfs/libzfs_dataset.c index 3288268fb258..133b3b358831 100644 --- a/lib/libzfs/libzfs_dataset.c +++ b/lib/libzfs/libzfs_dataset.c @@ -1,5590 +1,5592 @@ /* * CDDL HEADER START * * The contents of this file are subject to the terms of the * Common Development and Distribution License (the "License"). * 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 https://opensource.org/licenses/CDDL-1.0. * See the License for the specific language governing permissions * and limitations under the License. * * When distributing Covered Code, include this CDDL HEADER in each * file and include the License file at usr/src/OPENSOLARIS.LICENSE. * If applicable, add the following below this CDDL HEADER, with the * fields enclosed by brackets "[]" replaced with your own identifying * information: Portions Copyright [yyyy] [name of copyright owner] * * CDDL HEADER END */ /* * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved. * Copyright 2019 Joyent, Inc. * Copyright (c) 2011, 2020 by Delphix. All rights reserved. * Copyright (c) 2012 DEY Storage Systems, Inc. All rights reserved. * Copyright (c) 2012 Pawel Jakub Dawidek . * Copyright (c) 2013 Martin Matuska. All rights reserved. * Copyright (c) 2013 Steven Hartland. All rights reserved. * Copyright 2017 Nexenta Systems, Inc. * Copyright 2016 Igor Kozhukhov * Copyright 2017-2018 RackTop Systems. * Copyright (c) 2019 Datto Inc. * Copyright (c) 2019, loli10K * Copyright (c) 2021 Matt Fiddaman */ #include #include #include #include #include #include #include #include #include #include #include #include #include #include #ifdef HAVE_IDMAP #include #include #include #endif /* HAVE_IDMAP */ #include #include #include #include #include #include #include "zfs_namecheck.h" #include "zfs_prop.h" #include "libzfs_impl.h" #include "zfs_deleg.h" static int userquota_propname_decode(const char *propname, boolean_t zoned, zfs_userquota_prop_t *typep, char *domain, int domainlen, uint64_t *ridp); /* * Given a single type (not a mask of types), return the type in a human * readable form. */ const char * zfs_type_to_name(zfs_type_t type) { switch (type) { case ZFS_TYPE_FILESYSTEM: return (dgettext(TEXT_DOMAIN, "filesystem")); case ZFS_TYPE_SNAPSHOT: return (dgettext(TEXT_DOMAIN, "snapshot")); case ZFS_TYPE_VOLUME: return (dgettext(TEXT_DOMAIN, "volume")); case ZFS_TYPE_POOL: return (dgettext(TEXT_DOMAIN, "pool")); case ZFS_TYPE_BOOKMARK: return (dgettext(TEXT_DOMAIN, "bookmark")); default: assert(!"unhandled zfs_type_t"); } return (NULL); } /* * Validate a ZFS path. This is used even before trying to open the dataset, to * provide a more meaningful error message. We call zfs_error_aux() to * explain exactly why the name was not valid. */ int zfs_validate_name(libzfs_handle_t *hdl, const char *path, int type, boolean_t modifying) { namecheck_err_t why; char what; if (!(type & ZFS_TYPE_SNAPSHOT) && strchr(path, '@') != NULL) { if (hdl != NULL) zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "snapshot delimiter '@' is not expected here")); return (0); } if (type == ZFS_TYPE_SNAPSHOT && strchr(path, '@') == NULL) { if (hdl != NULL) zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "missing '@' delimiter in snapshot name")); return (0); } if (!(type & ZFS_TYPE_BOOKMARK) && strchr(path, '#') != NULL) { if (hdl != NULL) zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "bookmark delimiter '#' is not expected here")); return (0); } if (type == ZFS_TYPE_BOOKMARK && strchr(path, '#') == NULL) { if (hdl != NULL) zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "missing '#' delimiter in bookmark name")); return (0); } if (modifying && strchr(path, '%') != NULL) { if (hdl != NULL) zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "invalid character %c in name"), '%'); return (0); } if (entity_namecheck(path, &why, &what) != 0) { if (hdl != NULL) { switch (why) { case NAME_ERR_TOOLONG: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "name is too long")); break; case NAME_ERR_LEADING_SLASH: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "leading slash in name")); break; case NAME_ERR_EMPTY_COMPONENT: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "empty component or misplaced '@'" " or '#' delimiter in name")); break; case NAME_ERR_TRAILING_SLASH: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "trailing slash in name")); break; case NAME_ERR_INVALCHAR: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "invalid character " "'%c' in name"), what); break; case NAME_ERR_MULTIPLE_DELIMITERS: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "multiple '@' and/or '#' delimiters in " "name")); break; case NAME_ERR_NOLETTER: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "pool doesn't begin with a letter")); break; case NAME_ERR_RESERVED: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "name is reserved")); break; case NAME_ERR_DISKLIKE: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "reserved disk name")); break; case NAME_ERR_SELF_REF: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "self reference, '.' is found in name")); break; case NAME_ERR_PARENT_REF: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "parent reference, '..' is found in name")); break; default: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "(%d) not defined"), why); break; } } return (0); } return (-1); } int zfs_name_valid(const char *name, zfs_type_t type) { if (type == ZFS_TYPE_POOL) return (zpool_name_valid(NULL, B_FALSE, name)); return (zfs_validate_name(NULL, name, type, B_FALSE)); } /* * This function takes the raw DSL properties, and filters out the user-defined * properties into a separate nvlist. */ static nvlist_t * process_user_props(zfs_handle_t *zhp, nvlist_t *props) { libzfs_handle_t *hdl = zhp->zfs_hdl; nvpair_t *elem; nvlist_t *nvl; if (nvlist_alloc(&nvl, NV_UNIQUE_NAME, 0) != 0) { (void) no_memory(hdl); return (NULL); } elem = NULL; while ((elem = nvlist_next_nvpair(props, elem)) != NULL) { if (!zfs_prop_user(nvpair_name(elem))) continue; nvlist_t *propval = fnvpair_value_nvlist(elem); if (nvlist_add_nvlist(nvl, nvpair_name(elem), propval) != 0) { nvlist_free(nvl); (void) no_memory(hdl); return (NULL); } } return (nvl); } static zpool_handle_t * zpool_add_handle(zfs_handle_t *zhp, const char *pool_name) { libzfs_handle_t *hdl = zhp->zfs_hdl; zpool_handle_t *zph; if ((zph = zpool_open_canfail(hdl, pool_name)) != NULL) { if (hdl->libzfs_pool_handles != NULL) zph->zpool_next = hdl->libzfs_pool_handles; hdl->libzfs_pool_handles = zph; } return (zph); } static zpool_handle_t * zpool_find_handle(zfs_handle_t *zhp, const char *pool_name, int len) { libzfs_handle_t *hdl = zhp->zfs_hdl; zpool_handle_t *zph = hdl->libzfs_pool_handles; while ((zph != NULL) && (strncmp(pool_name, zpool_get_name(zph), len) != 0)) zph = zph->zpool_next; return (zph); } /* * Returns a handle to the pool that contains the provided dataset. * If a handle to that pool already exists then that handle is returned. * Otherwise, a new handle is created and added to the list of handles. */ static zpool_handle_t * zpool_handle(zfs_handle_t *zhp) { char *pool_name; int len; zpool_handle_t *zph; len = strcspn(zhp->zfs_name, "/@#") + 1; pool_name = zfs_alloc(zhp->zfs_hdl, len); (void) strlcpy(pool_name, zhp->zfs_name, len); zph = zpool_find_handle(zhp, pool_name, len); if (zph == NULL) zph = zpool_add_handle(zhp, pool_name); free(pool_name); return (zph); } void zpool_free_handles(libzfs_handle_t *hdl) { zpool_handle_t *next, *zph = hdl->libzfs_pool_handles; while (zph != NULL) { next = zph->zpool_next; zpool_close(zph); zph = next; } hdl->libzfs_pool_handles = NULL; } /* * Utility function to gather stats (objset and zpl) for the given object. */ static int get_stats_ioctl(zfs_handle_t *zhp, zfs_cmd_t *zc) { libzfs_handle_t *hdl = zhp->zfs_hdl; (void) strlcpy(zc->zc_name, zhp->zfs_name, sizeof (zc->zc_name)); while (zfs_ioctl(hdl, ZFS_IOC_OBJSET_STATS, zc) != 0) { if (errno == ENOMEM) zcmd_expand_dst_nvlist(hdl, zc); else return (-1); } return (0); } /* * Utility function to get the received properties of the given object. */ static int get_recvd_props_ioctl(zfs_handle_t *zhp) { libzfs_handle_t *hdl = zhp->zfs_hdl; nvlist_t *recvdprops; zfs_cmd_t zc = {"\0"}; int err; zcmd_alloc_dst_nvlist(hdl, &zc, 0); (void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name)); while (zfs_ioctl(hdl, ZFS_IOC_OBJSET_RECVD_PROPS, &zc) != 0) { if (errno == ENOMEM) zcmd_expand_dst_nvlist(hdl, &zc); else { zcmd_free_nvlists(&zc); return (-1); } } err = zcmd_read_dst_nvlist(zhp->zfs_hdl, &zc, &recvdprops); zcmd_free_nvlists(&zc); if (err != 0) return (-1); nvlist_free(zhp->zfs_recvd_props); zhp->zfs_recvd_props = recvdprops; return (0); } static int put_stats_zhdl(zfs_handle_t *zhp, zfs_cmd_t *zc) { nvlist_t *allprops, *userprops; zhp->zfs_dmustats = zc->zc_objset_stats; /* structure assignment */ if (zcmd_read_dst_nvlist(zhp->zfs_hdl, zc, &allprops) != 0) { return (-1); } /* * XXX Why do we store the user props separately, in addition to * storing them in zfs_props? */ if ((userprops = process_user_props(zhp, allprops)) == NULL) { nvlist_free(allprops); return (-1); } nvlist_free(zhp->zfs_props); nvlist_free(zhp->zfs_user_props); zhp->zfs_props = allprops; zhp->zfs_user_props = userprops; return (0); } static int get_stats(zfs_handle_t *zhp) { int rc = 0; zfs_cmd_t zc = {"\0"}; zcmd_alloc_dst_nvlist(zhp->zfs_hdl, &zc, 0); if (get_stats_ioctl(zhp, &zc) != 0) rc = -1; else if (put_stats_zhdl(zhp, &zc) != 0) rc = -1; zcmd_free_nvlists(&zc); return (rc); } /* * Refresh the properties currently stored in the handle. */ void zfs_refresh_properties(zfs_handle_t *zhp) { (void) get_stats(zhp); } /* * Makes a handle from the given dataset name. Used by zfs_open() and * zfs_iter_* to create child handles on the fly. */ static int make_dataset_handle_common(zfs_handle_t *zhp, zfs_cmd_t *zc) { if (put_stats_zhdl(zhp, zc) != 0) return (-1); /* * We've managed to open the dataset and gather statistics. Determine * the high-level type. */ if (zhp->zfs_dmustats.dds_type == DMU_OST_ZVOL) { zhp->zfs_head_type = ZFS_TYPE_VOLUME; } else if (zhp->zfs_dmustats.dds_type == DMU_OST_ZFS) { zhp->zfs_head_type = ZFS_TYPE_FILESYSTEM; } else if (zhp->zfs_dmustats.dds_type == DMU_OST_OTHER) { errno = EINVAL; return (-1); } else if (zhp->zfs_dmustats.dds_inconsistent) { errno = EBUSY; return (-1); } else { abort(); } if (zhp->zfs_dmustats.dds_is_snapshot) zhp->zfs_type = ZFS_TYPE_SNAPSHOT; else if (zhp->zfs_dmustats.dds_type == DMU_OST_ZVOL) zhp->zfs_type = ZFS_TYPE_VOLUME; else if (zhp->zfs_dmustats.dds_type == DMU_OST_ZFS) zhp->zfs_type = ZFS_TYPE_FILESYSTEM; else abort(); /* we should never see any other types */ if ((zhp->zpool_hdl = zpool_handle(zhp)) == NULL) return (-1); return (0); } zfs_handle_t * make_dataset_handle(libzfs_handle_t *hdl, const char *path) { zfs_cmd_t zc = {"\0"}; zfs_handle_t *zhp = calloc(1, sizeof (zfs_handle_t)); if (zhp == NULL) return (NULL); zhp->zfs_hdl = hdl; (void) strlcpy(zhp->zfs_name, path, sizeof (zhp->zfs_name)); zcmd_alloc_dst_nvlist(hdl, &zc, 0); if (get_stats_ioctl(zhp, &zc) == -1) { zcmd_free_nvlists(&zc); free(zhp); return (NULL); } if (make_dataset_handle_common(zhp, &zc) == -1) { free(zhp); zhp = NULL; } zcmd_free_nvlists(&zc); return (zhp); } zfs_handle_t * make_dataset_handle_zc(libzfs_handle_t *hdl, zfs_cmd_t *zc) { zfs_handle_t *zhp = calloc(1, sizeof (zfs_handle_t)); if (zhp == NULL) return (NULL); zhp->zfs_hdl = hdl; (void) strlcpy(zhp->zfs_name, zc->zc_name, sizeof (zhp->zfs_name)); if (make_dataset_handle_common(zhp, zc) == -1) { free(zhp); return (NULL); } return (zhp); } zfs_handle_t * make_dataset_simple_handle_zc(zfs_handle_t *pzhp, zfs_cmd_t *zc) { zfs_handle_t *zhp = calloc(1, sizeof (zfs_handle_t)); if (zhp == NULL) return (NULL); zhp->zfs_hdl = pzhp->zfs_hdl; (void) strlcpy(zhp->zfs_name, zc->zc_name, sizeof (zhp->zfs_name)); zhp->zfs_head_type = pzhp->zfs_type; zhp->zfs_type = ZFS_TYPE_SNAPSHOT; zhp->zpool_hdl = zpool_handle(zhp); zhp->zfs_dmustats = zc->zc_objset_stats; return (zhp); } zfs_handle_t * zfs_handle_dup(zfs_handle_t *zhp_orig) { zfs_handle_t *zhp = calloc(1, sizeof (zfs_handle_t)); if (zhp == NULL) return (NULL); zhp->zfs_hdl = zhp_orig->zfs_hdl; zhp->zpool_hdl = zhp_orig->zpool_hdl; (void) strlcpy(zhp->zfs_name, zhp_orig->zfs_name, sizeof (zhp->zfs_name)); zhp->zfs_type = zhp_orig->zfs_type; zhp->zfs_head_type = zhp_orig->zfs_head_type; zhp->zfs_dmustats = zhp_orig->zfs_dmustats; if (zhp_orig->zfs_props != NULL) { if (nvlist_dup(zhp_orig->zfs_props, &zhp->zfs_props, 0) != 0) { (void) no_memory(zhp->zfs_hdl); zfs_close(zhp); return (NULL); } } if (zhp_orig->zfs_user_props != NULL) { if (nvlist_dup(zhp_orig->zfs_user_props, &zhp->zfs_user_props, 0) != 0) { (void) no_memory(zhp->zfs_hdl); zfs_close(zhp); return (NULL); } } if (zhp_orig->zfs_recvd_props != NULL) { if (nvlist_dup(zhp_orig->zfs_recvd_props, &zhp->zfs_recvd_props, 0)) { (void) no_memory(zhp->zfs_hdl); zfs_close(zhp); return (NULL); } } zhp->zfs_mntcheck = zhp_orig->zfs_mntcheck; if (zhp_orig->zfs_mntopts != NULL) { zhp->zfs_mntopts = zfs_strdup(zhp_orig->zfs_hdl, zhp_orig->zfs_mntopts); } zhp->zfs_props_table = zhp_orig->zfs_props_table; return (zhp); } boolean_t zfs_bookmark_exists(const char *path) { nvlist_t *bmarks; nvlist_t *props; char fsname[ZFS_MAX_DATASET_NAME_LEN]; char *bmark_name; char *pound; int err; boolean_t rv; (void) strlcpy(fsname, path, sizeof (fsname)); pound = strchr(fsname, '#'); if (pound == NULL) return (B_FALSE); *pound = '\0'; bmark_name = pound + 1; props = fnvlist_alloc(); err = lzc_get_bookmarks(fsname, props, &bmarks); nvlist_free(props); if (err != 0) { nvlist_free(bmarks); return (B_FALSE); } rv = nvlist_exists(bmarks, bmark_name); nvlist_free(bmarks); return (rv); } zfs_handle_t * make_bookmark_handle(zfs_handle_t *parent, const char *path, nvlist_t *bmark_props) { zfs_handle_t *zhp = calloc(1, sizeof (zfs_handle_t)); if (zhp == NULL) return (NULL); /* Fill in the name. */ zhp->zfs_hdl = parent->zfs_hdl; (void) strlcpy(zhp->zfs_name, path, sizeof (zhp->zfs_name)); /* Set the property lists. */ if (nvlist_dup(bmark_props, &zhp->zfs_props, 0) != 0) { free(zhp); return (NULL); } /* Set the types. */ zhp->zfs_head_type = parent->zfs_head_type; zhp->zfs_type = ZFS_TYPE_BOOKMARK; if ((zhp->zpool_hdl = zpool_handle(zhp)) == NULL) { nvlist_free(zhp->zfs_props); free(zhp); return (NULL); } return (zhp); } struct zfs_open_bookmarks_cb_data { const char *path; zfs_handle_t *zhp; }; static int zfs_open_bookmarks_cb(zfs_handle_t *zhp, void *data) { struct zfs_open_bookmarks_cb_data *dp = data; /* * Is it the one we are looking for? */ if (strcmp(dp->path, zfs_get_name(zhp)) == 0) { /* * We found it. Save it and let the caller know we are done. */ dp->zhp = zhp; return (EEXIST); } /* * Not found. Close the handle and ask for another one. */ zfs_close(zhp); return (0); } /* * Opens the given snapshot, bookmark, filesystem, or volume. The 'types' * argument is a mask of acceptable types. The function will print an * appropriate error message and return NULL if it can't be opened. */ zfs_handle_t * zfs_open(libzfs_handle_t *hdl, const char *path, int types) { zfs_handle_t *zhp; char errbuf[ERRBUFLEN]; char *bookp; (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot open '%s'"), path); /* * Validate the name before we even try to open it. */ if (!zfs_validate_name(hdl, path, types, B_FALSE)) { (void) zfs_error(hdl, EZFS_INVALIDNAME, errbuf); return (NULL); } /* * Bookmarks needs to be handled separately. */ bookp = strchr(path, '#'); if (bookp == NULL) { /* * Try to get stats for the dataset, which will tell us if it * exists. */ errno = 0; if ((zhp = make_dataset_handle(hdl, path)) == NULL) { (void) zfs_standard_error(hdl, errno, errbuf); return (NULL); } } else { char dsname[ZFS_MAX_DATASET_NAME_LEN]; zfs_handle_t *pzhp; struct zfs_open_bookmarks_cb_data cb_data = {path, NULL}; /* * We need to cut out '#' and everything after '#' * to get the parent dataset name only. */ assert(bookp - path < sizeof (dsname)); (void) strlcpy(dsname, path, MIN(sizeof (dsname), bookp - path + 1)); /* * Create handle for the parent dataset. */ errno = 0; if ((pzhp = make_dataset_handle(hdl, dsname)) == NULL) { (void) zfs_standard_error(hdl, errno, errbuf); return (NULL); } /* * Iterate bookmarks to find the right one. */ errno = 0; if ((zfs_iter_bookmarks(pzhp, zfs_open_bookmarks_cb, &cb_data) == 0) && (cb_data.zhp == NULL)) { (void) zfs_error(hdl, EZFS_NOENT, errbuf); zfs_close(pzhp); return (NULL); } if (cb_data.zhp == NULL) { (void) zfs_standard_error(hdl, errno, errbuf); zfs_close(pzhp); return (NULL); } zhp = cb_data.zhp; /* * Cleanup. */ zfs_close(pzhp); } if (!(types & zhp->zfs_type)) { (void) zfs_error(hdl, EZFS_BADTYPE, errbuf); zfs_close(zhp); return (NULL); } return (zhp); } /* * Release a ZFS handle. Nothing to do but free the associated memory. */ void zfs_close(zfs_handle_t *zhp) { if (zhp->zfs_mntopts) free(zhp->zfs_mntopts); nvlist_free(zhp->zfs_props); nvlist_free(zhp->zfs_user_props); nvlist_free(zhp->zfs_recvd_props); free(zhp); } typedef struct mnttab_node { struct mnttab mtn_mt; avl_node_t mtn_node; } mnttab_node_t; static int libzfs_mnttab_cache_compare(const void *arg1, const void *arg2) { const mnttab_node_t *mtn1 = (const mnttab_node_t *)arg1; const mnttab_node_t *mtn2 = (const mnttab_node_t *)arg2; int rv; rv = strcmp(mtn1->mtn_mt.mnt_special, mtn2->mtn_mt.mnt_special); return (TREE_ISIGN(rv)); } void libzfs_mnttab_init(libzfs_handle_t *hdl) { pthread_mutex_init(&hdl->libzfs_mnttab_cache_lock, NULL); assert(avl_numnodes(&hdl->libzfs_mnttab_cache) == 0); avl_create(&hdl->libzfs_mnttab_cache, libzfs_mnttab_cache_compare, sizeof (mnttab_node_t), offsetof(mnttab_node_t, mtn_node)); } static int libzfs_mnttab_update(libzfs_handle_t *hdl) { FILE *mnttab; struct mnttab entry; if ((mnttab = fopen(MNTTAB, "re")) == NULL) return (ENOENT); while (getmntent(mnttab, &entry) == 0) { mnttab_node_t *mtn; avl_index_t where; if (strcmp(entry.mnt_fstype, MNTTYPE_ZFS) != 0) continue; mtn = zfs_alloc(hdl, sizeof (mnttab_node_t)); mtn->mtn_mt.mnt_special = zfs_strdup(hdl, entry.mnt_special); mtn->mtn_mt.mnt_mountp = zfs_strdup(hdl, entry.mnt_mountp); mtn->mtn_mt.mnt_fstype = zfs_strdup(hdl, entry.mnt_fstype); mtn->mtn_mt.mnt_mntopts = zfs_strdup(hdl, entry.mnt_mntopts); /* Exclude duplicate mounts */ if (avl_find(&hdl->libzfs_mnttab_cache, mtn, &where) != NULL) { free(mtn->mtn_mt.mnt_special); free(mtn->mtn_mt.mnt_mountp); free(mtn->mtn_mt.mnt_fstype); free(mtn->mtn_mt.mnt_mntopts); free(mtn); continue; } avl_add(&hdl->libzfs_mnttab_cache, mtn); } (void) fclose(mnttab); return (0); } void libzfs_mnttab_fini(libzfs_handle_t *hdl) { void *cookie = NULL; mnttab_node_t *mtn; while ((mtn = avl_destroy_nodes(&hdl->libzfs_mnttab_cache, &cookie)) != NULL) { free(mtn->mtn_mt.mnt_special); free(mtn->mtn_mt.mnt_mountp); free(mtn->mtn_mt.mnt_fstype); free(mtn->mtn_mt.mnt_mntopts); free(mtn); } avl_destroy(&hdl->libzfs_mnttab_cache); (void) pthread_mutex_destroy(&hdl->libzfs_mnttab_cache_lock); } void libzfs_mnttab_cache(libzfs_handle_t *hdl, boolean_t enable) { hdl->libzfs_mnttab_enable = enable; } int libzfs_mnttab_find(libzfs_handle_t *hdl, const char *fsname, struct mnttab *entry) { FILE *mnttab; mnttab_node_t find; mnttab_node_t *mtn; int ret = ENOENT; if (!hdl->libzfs_mnttab_enable) { struct mnttab srch = { 0 }; if (avl_numnodes(&hdl->libzfs_mnttab_cache)) libzfs_mnttab_fini(hdl); if ((mnttab = fopen(MNTTAB, "re")) == NULL) return (ENOENT); srch.mnt_special = (char *)fsname; srch.mnt_fstype = (char *)MNTTYPE_ZFS; ret = getmntany(mnttab, entry, &srch) ? ENOENT : 0; (void) fclose(mnttab); return (ret); } pthread_mutex_lock(&hdl->libzfs_mnttab_cache_lock); if (avl_numnodes(&hdl->libzfs_mnttab_cache) == 0) { int error; if ((error = libzfs_mnttab_update(hdl)) != 0) { pthread_mutex_unlock(&hdl->libzfs_mnttab_cache_lock); return (error); } } find.mtn_mt.mnt_special = (char *)fsname; mtn = avl_find(&hdl->libzfs_mnttab_cache, &find, NULL); if (mtn) { *entry = mtn->mtn_mt; ret = 0; } pthread_mutex_unlock(&hdl->libzfs_mnttab_cache_lock); return (ret); } void libzfs_mnttab_add(libzfs_handle_t *hdl, const char *special, const char *mountp, const char *mntopts) { mnttab_node_t *mtn; pthread_mutex_lock(&hdl->libzfs_mnttab_cache_lock); if (avl_numnodes(&hdl->libzfs_mnttab_cache) != 0) { mtn = zfs_alloc(hdl, sizeof (mnttab_node_t)); mtn->mtn_mt.mnt_special = zfs_strdup(hdl, special); mtn->mtn_mt.mnt_mountp = zfs_strdup(hdl, mountp); mtn->mtn_mt.mnt_fstype = zfs_strdup(hdl, MNTTYPE_ZFS); mtn->mtn_mt.mnt_mntopts = zfs_strdup(hdl, mntopts); /* * Another thread may have already added this entry * via libzfs_mnttab_update. If so we should skip it. */ if (avl_find(&hdl->libzfs_mnttab_cache, mtn, NULL) != NULL) { free(mtn->mtn_mt.mnt_special); free(mtn->mtn_mt.mnt_mountp); free(mtn->mtn_mt.mnt_fstype); free(mtn->mtn_mt.mnt_mntopts); free(mtn); } else { avl_add(&hdl->libzfs_mnttab_cache, mtn); } } pthread_mutex_unlock(&hdl->libzfs_mnttab_cache_lock); } void libzfs_mnttab_remove(libzfs_handle_t *hdl, const char *fsname) { mnttab_node_t find; mnttab_node_t *ret; pthread_mutex_lock(&hdl->libzfs_mnttab_cache_lock); find.mtn_mt.mnt_special = (char *)fsname; if ((ret = avl_find(&hdl->libzfs_mnttab_cache, (void *)&find, NULL)) != NULL) { avl_remove(&hdl->libzfs_mnttab_cache, ret); free(ret->mtn_mt.mnt_special); free(ret->mtn_mt.mnt_mountp); free(ret->mtn_mt.mnt_fstype); free(ret->mtn_mt.mnt_mntopts); free(ret); } pthread_mutex_unlock(&hdl->libzfs_mnttab_cache_lock); } int zfs_spa_version(zfs_handle_t *zhp, int *spa_version) { zpool_handle_t *zpool_handle = zhp->zpool_hdl; if (zpool_handle == NULL) return (-1); *spa_version = zpool_get_prop_int(zpool_handle, ZPOOL_PROP_VERSION, NULL); return (0); } /* * The choice of reservation property depends on the SPA version. */ static int zfs_which_resv_prop(zfs_handle_t *zhp, zfs_prop_t *resv_prop) { int spa_version; if (zfs_spa_version(zhp, &spa_version) < 0) return (-1); if (spa_version >= SPA_VERSION_REFRESERVATION) *resv_prop = ZFS_PROP_REFRESERVATION; else *resv_prop = ZFS_PROP_RESERVATION; return (0); } /* * Given an nvlist of properties to set, validates that they are correct, and * parses any numeric properties (index, boolean, etc) if they are specified as * strings. */ nvlist_t * zfs_valid_proplist(libzfs_handle_t *hdl, zfs_type_t type, nvlist_t *nvl, uint64_t zoned, zfs_handle_t *zhp, zpool_handle_t *zpool_hdl, boolean_t key_params_ok, const char *errbuf) { nvpair_t *elem; uint64_t intval; char *strval; zfs_prop_t prop; nvlist_t *ret; int chosen_normal = -1; int chosen_utf = -1; if (nvlist_alloc(&ret, NV_UNIQUE_NAME, 0) != 0) { (void) no_memory(hdl); return (NULL); } /* * Make sure this property is valid and applies to this type. */ elem = NULL; while ((elem = nvlist_next_nvpair(nvl, elem)) != NULL) { const char *propname = nvpair_name(elem); prop = zfs_name_to_prop(propname); if (prop == ZPROP_USERPROP && zfs_prop_user(propname)) { /* * This is a user property: make sure it's a * string, and that it's less than ZAP_MAXNAMELEN. */ if (nvpair_type(elem) != DATA_TYPE_STRING) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "'%s' must be a string"), propname); (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } if (strlen(nvpair_name(elem)) >= ZAP_MAXNAMELEN) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "property name '%s' is too long"), propname); (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } (void) nvpair_value_string(elem, &strval); if (nvlist_add_string(ret, propname, strval) != 0) { (void) no_memory(hdl); goto error; } continue; } /* * Currently, only user properties can be modified on * snapshots. */ if (type == ZFS_TYPE_SNAPSHOT) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "this property can not be modified for snapshots")); (void) zfs_error(hdl, EZFS_PROPTYPE, errbuf); goto error; } if (prop == ZPROP_USERPROP && zfs_prop_userquota(propname)) { zfs_userquota_prop_t uqtype; char *newpropname = NULL; char domain[128]; uint64_t rid; uint64_t valary[3]; int rc; if (userquota_propname_decode(propname, zoned, &uqtype, domain, sizeof (domain), &rid) != 0) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "'%s' has an invalid user/group name"), propname); (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } if (uqtype != ZFS_PROP_USERQUOTA && uqtype != ZFS_PROP_GROUPQUOTA && uqtype != ZFS_PROP_USEROBJQUOTA && uqtype != ZFS_PROP_GROUPOBJQUOTA && uqtype != ZFS_PROP_PROJECTQUOTA && uqtype != ZFS_PROP_PROJECTOBJQUOTA) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "'%s' is readonly"), propname); (void) zfs_error(hdl, EZFS_PROPREADONLY, errbuf); goto error; } if (nvpair_type(elem) == DATA_TYPE_STRING) { (void) nvpair_value_string(elem, &strval); if (strcmp(strval, "none") == 0) { intval = 0; } else if (zfs_nicestrtonum(hdl, strval, &intval) != 0) { (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } } else if (nvpair_type(elem) == DATA_TYPE_UINT64) { (void) nvpair_value_uint64(elem, &intval); if (intval == 0) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "use 'none' to disable " "{user|group|project}quota")); goto error; } } else { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "'%s' must be a number"), propname); (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } /* * Encode the prop name as * userquota@-domain, to make it easy * for the kernel to decode. */ rc = asprintf(&newpropname, "%s%llx-%s", zfs_userquota_prop_prefixes[uqtype], (longlong_t)rid, domain); if (rc == -1 || newpropname == NULL) { (void) no_memory(hdl); goto error; } valary[0] = uqtype; valary[1] = rid; valary[2] = intval; if (nvlist_add_uint64_array(ret, newpropname, valary, 3) != 0) { free(newpropname); (void) no_memory(hdl); goto error; } free(newpropname); continue; } else if (prop == ZPROP_USERPROP && zfs_prop_written(propname)) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "'%s' is readonly"), propname); (void) zfs_error(hdl, EZFS_PROPREADONLY, errbuf); goto error; } if (prop == ZPROP_INVAL) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "invalid property '%s'"), propname); (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } if (!zfs_prop_valid_for_type(prop, type, B_FALSE)) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "'%s' does not " "apply to datasets of this type"), propname); (void) zfs_error(hdl, EZFS_PROPTYPE, errbuf); goto error; } if (zfs_prop_readonly(prop) && !(zfs_prop_setonce(prop) && zhp == NULL) && !(zfs_prop_encryption_key_param(prop) && key_params_ok)) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "'%s' is readonly"), propname); (void) zfs_error(hdl, EZFS_PROPREADONLY, errbuf); goto error; } if (zprop_parse_value(hdl, elem, prop, type, ret, &strval, &intval, errbuf) != 0) goto error; /* * Perform some additional checks for specific properties. */ switch (prop) { case ZFS_PROP_VERSION: { int version; if (zhp == NULL) break; version = zfs_prop_get_int(zhp, ZFS_PROP_VERSION); if (intval < version) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "Can not downgrade; already at version %u"), version); (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } break; } case ZFS_PROP_VOLBLOCKSIZE: case ZFS_PROP_RECORDSIZE: { int maxbs = SPA_MAXBLOCKSIZE; char buf[64]; if (zpool_hdl != NULL) { maxbs = zpool_get_prop_int(zpool_hdl, ZPOOL_PROP_MAXBLOCKSIZE, NULL); } /* * The value must be a power of two between * SPA_MINBLOCKSIZE and maxbs. */ if (intval < SPA_MINBLOCKSIZE || intval > maxbs || !ISP2(intval)) { zfs_nicebytes(maxbs, buf, sizeof (buf)); zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "'%s' must be power of 2 from 512B " "to %s"), propname, buf); (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } break; } case ZFS_PROP_SPECIAL_SMALL_BLOCKS: { int maxbs = SPA_OLD_MAXBLOCKSIZE; char buf[64]; if (zpool_hdl != NULL) { char state[64] = ""; maxbs = zpool_get_prop_int(zpool_hdl, ZPOOL_PROP_MAXBLOCKSIZE, NULL); /* * Issue a warning but do not fail so that * tests for settable properties succeed. */ if (zpool_prop_get_feature(zpool_hdl, "feature@allocation_classes", state, sizeof (state)) != 0 || strcmp(state, ZFS_FEATURE_ACTIVE) != 0) { (void) fprintf(stderr, gettext( "%s: property requires a special " "device in the pool\n"), propname); } } if (intval != 0 && (intval < SPA_MINBLOCKSIZE || intval > maxbs || !ISP2(intval))) { zfs_nicebytes(maxbs, buf, sizeof (buf)); zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "invalid '%s=%llu' property: must be zero " "or a power of 2 from 512B to %s"), propname, (unsigned long long)intval, buf); (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } break; } case ZFS_PROP_MLSLABEL: { #ifdef HAVE_MLSLABEL /* * Verify the mlslabel string and convert to * internal hex label string. */ m_label_t *new_sl; char *hex = NULL; /* internal label string */ /* Default value is already OK. */ if (strcasecmp(strval, ZFS_MLSLABEL_DEFAULT) == 0) break; /* Verify the label can be converted to binary form */ if (((new_sl = m_label_alloc(MAC_LABEL)) == NULL) || (str_to_label(strval, &new_sl, MAC_LABEL, L_NO_CORRECTION, NULL) == -1)) { goto badlabel; } /* Now translate to hex internal label string */ if (label_to_str(new_sl, &hex, M_INTERNAL, DEF_NAMES) != 0) { if (hex) free(hex); goto badlabel; } m_label_free(new_sl); /* If string is already in internal form, we're done. */ if (strcmp(strval, hex) == 0) { free(hex); break; } /* Replace the label string with the internal form. */ (void) nvlist_remove(ret, zfs_prop_to_name(prop), DATA_TYPE_STRING); fnvlist_add_string(ret, zfs_prop_to_name(prop), hex); free(hex); break; badlabel: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "invalid mlslabel '%s'"), strval); (void) zfs_error(hdl, EZFS_BADPROP, errbuf); m_label_free(new_sl); /* OK if null */ goto error; #else zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "mlslabels are unsupported")); (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; #endif /* HAVE_MLSLABEL */ } case ZFS_PROP_MOUNTPOINT: { namecheck_err_t why; if (strcmp(strval, ZFS_MOUNTPOINT_NONE) == 0 || strcmp(strval, ZFS_MOUNTPOINT_LEGACY) == 0) break; if (mountpoint_namecheck(strval, &why)) { switch (why) { case NAME_ERR_LEADING_SLASH: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "'%s' must be an absolute path, " "'none', or 'legacy'"), propname); break; case NAME_ERR_TOOLONG: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "component of '%s' is too long"), propname); break; default: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "(%d) not defined"), why); break; } (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } zfs_fallthrough; } case ZFS_PROP_SHARESMB: case ZFS_PROP_SHARENFS: /* * For the mountpoint and sharenfs or sharesmb * properties, check if it can be set in a * global/non-global zone based on * the zoned property value: * * global zone non-global zone * -------------------------------------------------- * zoned=on mountpoint (no) mountpoint (yes) * sharenfs (no) sharenfs (no) * sharesmb (no) sharesmb (no) * * zoned=off mountpoint (yes) N/A * sharenfs (yes) * sharesmb (yes) */ if (zoned) { if (getzoneid() == GLOBAL_ZONEID) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "'%s' cannot be set on " "dataset in a non-global zone"), propname); (void) zfs_error(hdl, EZFS_ZONED, errbuf); goto error; } else if (prop == ZFS_PROP_SHARENFS || prop == ZFS_PROP_SHARESMB) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "'%s' cannot be set in " "a non-global zone"), propname); (void) zfs_error(hdl, EZFS_ZONED, errbuf); goto error; } } else if (getzoneid() != GLOBAL_ZONEID) { /* * If zoned property is 'off', this must be in * a global zone. If not, something is wrong. */ zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "'%s' cannot be set while dataset " "'zoned' property is set"), propname); (void) zfs_error(hdl, EZFS_ZONED, errbuf); goto error; } /* * At this point, it is legitimate to set the * property. Now we want to make sure that the * property value is valid if it is sharenfs. */ if ((prop == ZFS_PROP_SHARENFS || prop == ZFS_PROP_SHARESMB) && strcmp(strval, "on") != 0 && strcmp(strval, "off") != 0) { enum sa_protocol proto; if (prop == ZFS_PROP_SHARESMB) proto = SA_PROTOCOL_SMB; else proto = SA_PROTOCOL_NFS; if (sa_validate_shareopts(strval, proto) != SA_OK) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "'%s' cannot be set to invalid " "options"), propname); (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } } break; case ZFS_PROP_KEYLOCATION: if (!zfs_prop_valid_keylocation(strval, B_FALSE)) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "invalid keylocation")); (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } if (zhp != NULL) { uint64_t crypt = zfs_prop_get_int(zhp, ZFS_PROP_ENCRYPTION); if (crypt == ZIO_CRYPT_OFF && strcmp(strval, "none") != 0) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "keylocation must be 'none' " "for unencrypted datasets")); (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } else if (crypt != ZIO_CRYPT_OFF && strcmp(strval, "none") == 0) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "keylocation must not be 'none' " "for encrypted datasets")); (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } } break; case ZFS_PROP_PBKDF2_ITERS: if (intval < MIN_PBKDF2_ITERATIONS) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "minimum pbkdf2 iterations is %u"), MIN_PBKDF2_ITERATIONS); (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } break; case ZFS_PROP_UTF8ONLY: chosen_utf = (int)intval; break; case ZFS_PROP_NORMALIZE: chosen_normal = (int)intval; break; default: break; } /* * For changes to existing volumes, we have some additional * checks to enforce. */ if (type == ZFS_TYPE_VOLUME && zhp != NULL) { uint64_t blocksize = zfs_prop_get_int(zhp, ZFS_PROP_VOLBLOCKSIZE); char buf[64]; switch (prop) { case ZFS_PROP_VOLSIZE: if (intval % blocksize != 0) { zfs_nicebytes(blocksize, buf, sizeof (buf)); zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "'%s' must be a multiple of " "volume block size (%s)"), propname, buf); (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } if (intval == 0) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "'%s' cannot be zero"), propname); (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } break; default: break; } } /* check encryption properties */ if (zhp != NULL) { int64_t crypt = zfs_prop_get_int(zhp, ZFS_PROP_ENCRYPTION); switch (prop) { case ZFS_PROP_COPIES: if (crypt != ZIO_CRYPT_OFF && intval > 2) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "encrypted datasets cannot have " "3 copies")); (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } break; default: break; } } } /* * If normalization was chosen, but no UTF8 choice was made, * enforce rejection of non-UTF8 names. * * If normalization was chosen, but rejecting non-UTF8 names * was explicitly not chosen, it is an error. * * If utf8only was turned off, but the parent has normalization, * turn off normalization. */ if (chosen_normal > 0 && chosen_utf < 0) { if (nvlist_add_uint64(ret, zfs_prop_to_name(ZFS_PROP_UTF8ONLY), 1) != 0) { (void) no_memory(hdl); goto error; } } else if (chosen_normal > 0 && chosen_utf == 0) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "'%s' must be set 'on' if normalization chosen"), zfs_prop_to_name(ZFS_PROP_UTF8ONLY)); (void) zfs_error(hdl, EZFS_BADPROP, errbuf); goto error; } else if (chosen_normal < 0 && chosen_utf == 0) { if (nvlist_add_uint64(ret, zfs_prop_to_name(ZFS_PROP_NORMALIZE), 0) != 0) { (void) no_memory(hdl); goto error; } } return (ret); error: nvlist_free(ret); return (NULL); } static int zfs_add_synthetic_resv(zfs_handle_t *zhp, nvlist_t *nvl) { uint64_t old_volsize; uint64_t new_volsize; uint64_t old_reservation; uint64_t new_reservation; zfs_prop_t resv_prop; nvlist_t *props; zpool_handle_t *zph = zpool_handle(zhp); /* * If this is an existing volume, and someone is setting the volsize, * make sure that it matches the reservation, or add it if necessary. */ old_volsize = zfs_prop_get_int(zhp, ZFS_PROP_VOLSIZE); if (zfs_which_resv_prop(zhp, &resv_prop) < 0) return (-1); old_reservation = zfs_prop_get_int(zhp, resv_prop); props = fnvlist_alloc(); fnvlist_add_uint64(props, zfs_prop_to_name(ZFS_PROP_VOLBLOCKSIZE), zfs_prop_get_int(zhp, ZFS_PROP_VOLBLOCKSIZE)); if ((zvol_volsize_to_reservation(zph, old_volsize, props) != old_reservation) || nvlist_exists(nvl, zfs_prop_to_name(resv_prop))) { fnvlist_free(props); return (0); } if (nvlist_lookup_uint64(nvl, zfs_prop_to_name(ZFS_PROP_VOLSIZE), &new_volsize) != 0) { fnvlist_free(props); return (-1); } new_reservation = zvol_volsize_to_reservation(zph, new_volsize, props); fnvlist_free(props); if (nvlist_add_uint64(nvl, zfs_prop_to_name(resv_prop), new_reservation) != 0) { (void) no_memory(zhp->zfs_hdl); return (-1); } return (1); } /* * Helper for 'zfs {set|clone} refreservation=auto'. Must be called after * zfs_valid_proplist(), as it is what sets the UINT64_MAX sentinel value. * Return codes must match zfs_add_synthetic_resv(). */ static int zfs_fix_auto_resv(zfs_handle_t *zhp, nvlist_t *nvl) { uint64_t volsize; uint64_t resvsize; zfs_prop_t prop; nvlist_t *props; if (!ZFS_IS_VOLUME(zhp)) { return (0); } if (zfs_which_resv_prop(zhp, &prop) != 0) { return (-1); } if (prop != ZFS_PROP_REFRESERVATION) { return (0); } if (nvlist_lookup_uint64(nvl, zfs_prop_to_name(prop), &resvsize) != 0) { /* No value being set, so it can't be "auto" */ return (0); } if (resvsize != UINT64_MAX) { /* Being set to a value other than "auto" */ return (0); } props = fnvlist_alloc(); fnvlist_add_uint64(props, zfs_prop_to_name(ZFS_PROP_VOLBLOCKSIZE), zfs_prop_get_int(zhp, ZFS_PROP_VOLBLOCKSIZE)); if (nvlist_lookup_uint64(nvl, zfs_prop_to_name(ZFS_PROP_VOLSIZE), &volsize) != 0) { volsize = zfs_prop_get_int(zhp, ZFS_PROP_VOLSIZE); } resvsize = zvol_volsize_to_reservation(zpool_handle(zhp), volsize, props); fnvlist_free(props); (void) nvlist_remove_all(nvl, zfs_prop_to_name(prop)); if (nvlist_add_uint64(nvl, zfs_prop_to_name(prop), resvsize) != 0) { (void) no_memory(zhp->zfs_hdl); return (-1); } return (1); } static boolean_t zfs_is_namespace_prop(zfs_prop_t prop) { switch (prop) { case ZFS_PROP_ATIME: case ZFS_PROP_RELATIME: case ZFS_PROP_DEVICES: case ZFS_PROP_EXEC: case ZFS_PROP_SETUID: case ZFS_PROP_READONLY: case ZFS_PROP_XATTR: case ZFS_PROP_NBMAND: return (B_TRUE); default: return (B_FALSE); } } /* * Given a property name and value, set the property for the given dataset. */ int zfs_prop_set(zfs_handle_t *zhp, const char *propname, const char *propval) { int ret = -1; char errbuf[ERRBUFLEN]; libzfs_handle_t *hdl = zhp->zfs_hdl; nvlist_t *nvl = NULL; (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot set property for '%s'"), zhp->zfs_name); if (nvlist_alloc(&nvl, NV_UNIQUE_NAME, 0) != 0 || nvlist_add_string(nvl, propname, propval) != 0) { (void) no_memory(hdl); goto error; } ret = zfs_prop_set_list(zhp, nvl); error: nvlist_free(nvl); return (ret); } /* * Given an nvlist of property names and values, set the properties for the * given dataset. */ int zfs_prop_set_list(zfs_handle_t *zhp, nvlist_t *props) { zfs_cmd_t zc = {"\0"}; int ret = -1; prop_changelist_t **cls = NULL; int cl_idx; char errbuf[ERRBUFLEN]; libzfs_handle_t *hdl = zhp->zfs_hdl; nvlist_t *nvl; int nvl_len = 0; int added_resv = 0; zfs_prop_t prop = 0; nvpair_t *elem; (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot set property for '%s'"), zhp->zfs_name); if ((nvl = zfs_valid_proplist(hdl, zhp->zfs_type, props, zfs_prop_get_int(zhp, ZFS_PROP_ZONED), zhp, zhp->zpool_hdl, B_FALSE, errbuf)) == NULL) goto error; /* * We have to check for any extra properties which need to be added * before computing the length of the nvlist. */ for (elem = nvlist_next_nvpair(nvl, NULL); elem != NULL; elem = nvlist_next_nvpair(nvl, elem)) { if (zfs_name_to_prop(nvpair_name(elem)) == ZFS_PROP_VOLSIZE && (added_resv = zfs_add_synthetic_resv(zhp, nvl)) == -1) { goto error; } } if (added_resv != 1 && (added_resv = zfs_fix_auto_resv(zhp, nvl)) == -1) { goto error; } /* * Check how many properties we're setting and allocate an array to * store changelist pointers for postfix(). */ for (elem = nvlist_next_nvpair(nvl, NULL); elem != NULL; elem = nvlist_next_nvpair(nvl, elem)) nvl_len++; if ((cls = calloc(nvl_len, sizeof (prop_changelist_t *))) == NULL) goto error; cl_idx = 0; for (elem = nvlist_next_nvpair(nvl, NULL); elem != NULL; elem = nvlist_next_nvpair(nvl, elem)) { prop = zfs_name_to_prop(nvpair_name(elem)); assert(cl_idx < nvl_len); /* * We don't want to unmount & remount the dataset when changing * its canmount property to 'on' or 'noauto'. We only use * the changelist logic to unmount when setting canmount=off. */ if (prop != ZFS_PROP_CANMOUNT || (fnvpair_value_uint64(elem) == ZFS_CANMOUNT_OFF && zfs_is_mounted(zhp, NULL))) { cls[cl_idx] = changelist_gather(zhp, prop, 0, 0); if (cls[cl_idx] == NULL) goto error; } if (prop == ZFS_PROP_MOUNTPOINT && changelist_haszonedchild(cls[cl_idx])) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "child dataset with inherited mountpoint is used " "in a non-global zone")); ret = zfs_error(hdl, EZFS_ZONED, errbuf); goto error; } if (cls[cl_idx] != NULL && (ret = changelist_prefix(cls[cl_idx])) != 0) goto error; cl_idx++; } assert(cl_idx == nvl_len); /* * Execute the corresponding ioctl() to set this list of properties. */ (void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name)); zcmd_write_src_nvlist(hdl, &zc, nvl); zcmd_alloc_dst_nvlist(hdl, &zc, 0); ret = zfs_ioctl(hdl, ZFS_IOC_SET_PROP, &zc); if (ret != 0) { if (zc.zc_nvlist_dst_filled == B_FALSE) { (void) zfs_standard_error(hdl, errno, errbuf); goto error; } /* Get the list of unset properties back and report them. */ nvlist_t *errorprops = NULL; if (zcmd_read_dst_nvlist(hdl, &zc, &errorprops) != 0) goto error; for (nvpair_t *elem = nvlist_next_nvpair(errorprops, NULL); elem != NULL; elem = nvlist_next_nvpair(errorprops, elem)) { prop = zfs_name_to_prop(nvpair_name(elem)); zfs_setprop_error(hdl, prop, errno, errbuf); } nvlist_free(errorprops); if (added_resv && errno == ENOSPC) { /* clean up the volsize property we tried to set */ uint64_t old_volsize = zfs_prop_get_int(zhp, ZFS_PROP_VOLSIZE); nvlist_free(nvl); nvl = NULL; zcmd_free_nvlists(&zc); if (nvlist_alloc(&nvl, NV_UNIQUE_NAME, 0) != 0) goto error; if (nvlist_add_uint64(nvl, zfs_prop_to_name(ZFS_PROP_VOLSIZE), old_volsize) != 0) goto error; zcmd_write_src_nvlist(hdl, &zc, nvl); (void) zfs_ioctl(hdl, ZFS_IOC_SET_PROP, &zc); } } else { for (cl_idx = 0; cl_idx < nvl_len; cl_idx++) { if (cls[cl_idx] != NULL) { int clp_err = changelist_postfix(cls[cl_idx]); if (clp_err != 0) ret = clp_err; } } if (ret == 0) { /* * Refresh the statistics so the new property * value is reflected. */ (void) get_stats(zhp); /* * Remount the filesystem to propagate the change * if one of the options handled by the generic * Linux namespace layer has been modified. */ if (zfs_is_namespace_prop(prop) && zfs_is_mounted(zhp, NULL)) ret = zfs_mount(zhp, MNTOPT_REMOUNT, 0); } } error: nvlist_free(nvl); zcmd_free_nvlists(&zc); if (cls != NULL) { for (cl_idx = 0; cl_idx < nvl_len; cl_idx++) { if (cls[cl_idx] != NULL) changelist_free(cls[cl_idx]); } free(cls); } return (ret); } /* * Given a property, inherit the value from the parent dataset, or if received * is TRUE, revert to the received value, if any. */ int zfs_prop_inherit(zfs_handle_t *zhp, const char *propname, boolean_t received) { zfs_cmd_t zc = {"\0"}; int ret; prop_changelist_t *cl; libzfs_handle_t *hdl = zhp->zfs_hdl; char errbuf[ERRBUFLEN]; zfs_prop_t prop; (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot inherit %s for '%s'"), propname, zhp->zfs_name); zc.zc_cookie = received; if ((prop = zfs_name_to_prop(propname)) == ZPROP_USERPROP) { /* * For user properties, the amount of work we have to do is very * small, so just do it here. */ if (!zfs_prop_user(propname)) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "invalid property")); return (zfs_error(hdl, EZFS_BADPROP, errbuf)); } (void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name)); (void) strlcpy(zc.zc_value, propname, sizeof (zc.zc_value)); if (zfs_ioctl(zhp->zfs_hdl, ZFS_IOC_INHERIT_PROP, &zc) != 0) return (zfs_standard_error(hdl, errno, errbuf)); (void) get_stats(zhp); return (0); } /* * Verify that this property is inheritable. */ if (zfs_prop_readonly(prop)) return (zfs_error(hdl, EZFS_PROPREADONLY, errbuf)); if (!zfs_prop_inheritable(prop) && !received) return (zfs_error(hdl, EZFS_PROPNONINHERIT, errbuf)); /* * Check to see if the value applies to this type */ if (!zfs_prop_valid_for_type(prop, zhp->zfs_type, B_FALSE)) return (zfs_error(hdl, EZFS_PROPTYPE, errbuf)); /* * Normalize the name, to get rid of shorthand abbreviations. */ propname = zfs_prop_to_name(prop); (void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name)); (void) strlcpy(zc.zc_value, propname, sizeof (zc.zc_value)); if (prop == ZFS_PROP_MOUNTPOINT && getzoneid() == GLOBAL_ZONEID && zfs_prop_get_int(zhp, ZFS_PROP_ZONED)) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "dataset is used in a non-global zone")); return (zfs_error(hdl, EZFS_ZONED, errbuf)); } /* * Determine datasets which will be affected by this change, if any. */ if ((cl = changelist_gather(zhp, prop, 0, 0)) == NULL) return (-1); if (prop == ZFS_PROP_MOUNTPOINT && changelist_haszonedchild(cl)) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "child dataset with inherited mountpoint is used " "in a non-global zone")); ret = zfs_error(hdl, EZFS_ZONED, errbuf); goto error; } if ((ret = changelist_prefix(cl)) != 0) goto error; if ((ret = zfs_ioctl(zhp->zfs_hdl, ZFS_IOC_INHERIT_PROP, &zc)) != 0) { changelist_free(cl); return (zfs_standard_error(hdl, errno, errbuf)); } else { if ((ret = changelist_postfix(cl)) != 0) goto error; /* * Refresh the statistics so the new property is reflected. */ (void) get_stats(zhp); /* * Remount the filesystem to propagate the change * if one of the options handled by the generic * Linux namespace layer has been modified. */ if (zfs_is_namespace_prop(prop) && zfs_is_mounted(zhp, NULL)) ret = zfs_mount(zhp, MNTOPT_REMOUNT, 0); } error: changelist_free(cl); return (ret); } /* * True DSL properties are stored in an nvlist. The following two functions * extract them appropriately. */ uint64_t getprop_uint64(zfs_handle_t *zhp, zfs_prop_t prop, char **source) { nvlist_t *nv; uint64_t value; *source = NULL; if (nvlist_lookup_nvlist(zhp->zfs_props, zfs_prop_to_name(prop), &nv) == 0) { value = fnvlist_lookup_uint64(nv, ZPROP_VALUE); (void) nvlist_lookup_string(nv, ZPROP_SOURCE, source); } else { verify(!zhp->zfs_props_table || zhp->zfs_props_table[prop] == B_TRUE); value = zfs_prop_default_numeric(prop); *source = (char *)""; } return (value); } static const char * getprop_string(zfs_handle_t *zhp, zfs_prop_t prop, char **source) { nvlist_t *nv; const char *value; *source = NULL; if (nvlist_lookup_nvlist(zhp->zfs_props, zfs_prop_to_name(prop), &nv) == 0) { value = fnvlist_lookup_string(nv, ZPROP_VALUE); (void) nvlist_lookup_string(nv, ZPROP_SOURCE, source); } else { verify(!zhp->zfs_props_table || zhp->zfs_props_table[prop] == B_TRUE); value = zfs_prop_default_string(prop); *source = (char *)""; } return (value); } static boolean_t zfs_is_recvd_props_mode(zfs_handle_t *zhp) { return (zhp->zfs_props == zhp->zfs_recvd_props); } static void zfs_set_recvd_props_mode(zfs_handle_t *zhp, uint64_t *cookie) { *cookie = (uint64_t)(uintptr_t)zhp->zfs_props; zhp->zfs_props = zhp->zfs_recvd_props; } static void zfs_unset_recvd_props_mode(zfs_handle_t *zhp, uint64_t *cookie) { zhp->zfs_props = (nvlist_t *)(uintptr_t)*cookie; *cookie = 0; } /* * Internal function for getting a numeric property. Both zfs_prop_get() and * zfs_prop_get_int() are built using this interface. * * Certain properties can be overridden using 'mount -o'. In this case, scan * the contents of the /proc/self/mounts entry, searching for the * appropriate options. If they differ from the on-disk values, report the * current values and mark the source "temporary". */ static int get_numeric_property(zfs_handle_t *zhp, zfs_prop_t prop, zprop_source_t *src, char **source, uint64_t *val) { zfs_cmd_t zc = {"\0"}; nvlist_t *zplprops = NULL; struct mnttab mnt; const char *mntopt_on = NULL; const char *mntopt_off = NULL; boolean_t received = zfs_is_recvd_props_mode(zhp); *source = NULL; /* * If the property is being fetched for a snapshot, check whether * the property is valid for the snapshot's head dataset type. */ if (zhp->zfs_type == ZFS_TYPE_SNAPSHOT && !zfs_prop_valid_for_type(prop, zhp->zfs_head_type, B_TRUE)) { *val = zfs_prop_default_numeric(prop); return (-1); } switch (prop) { case ZFS_PROP_ATIME: mntopt_on = MNTOPT_ATIME; mntopt_off = MNTOPT_NOATIME; break; case ZFS_PROP_RELATIME: mntopt_on = MNTOPT_RELATIME; mntopt_off = MNTOPT_NORELATIME; break; case ZFS_PROP_DEVICES: mntopt_on = MNTOPT_DEVICES; mntopt_off = MNTOPT_NODEVICES; break; case ZFS_PROP_EXEC: mntopt_on = MNTOPT_EXEC; mntopt_off = MNTOPT_NOEXEC; break; case ZFS_PROP_READONLY: mntopt_on = MNTOPT_RO; mntopt_off = MNTOPT_RW; break; case ZFS_PROP_SETUID: mntopt_on = MNTOPT_SETUID; mntopt_off = MNTOPT_NOSETUID; break; case ZFS_PROP_XATTR: mntopt_on = MNTOPT_XATTR; mntopt_off = MNTOPT_NOXATTR; break; case ZFS_PROP_NBMAND: mntopt_on = MNTOPT_NBMAND; mntopt_off = MNTOPT_NONBMAND; break; default: break; } /* * Because looking up the mount options is potentially expensive * (iterating over all of /proc/self/mounts), we defer its * calculation until we're looking up a property which requires * its presence. */ if (!zhp->zfs_mntcheck && (mntopt_on != NULL || prop == ZFS_PROP_MOUNTED)) { libzfs_handle_t *hdl = zhp->zfs_hdl; struct mnttab entry; if (libzfs_mnttab_find(hdl, zhp->zfs_name, &entry) == 0) zhp->zfs_mntopts = zfs_strdup(hdl, entry.mnt_mntopts); zhp->zfs_mntcheck = B_TRUE; } if (zhp->zfs_mntopts == NULL) mnt.mnt_mntopts = (char *)""; else mnt.mnt_mntopts = zhp->zfs_mntopts; switch (prop) { case ZFS_PROP_ATIME: case ZFS_PROP_RELATIME: case ZFS_PROP_DEVICES: case ZFS_PROP_EXEC: case ZFS_PROP_READONLY: case ZFS_PROP_SETUID: #ifndef __FreeBSD__ case ZFS_PROP_XATTR: #endif case ZFS_PROP_NBMAND: *val = getprop_uint64(zhp, prop, source); if (received) break; if (hasmntopt(&mnt, mntopt_on) && !*val) { *val = B_TRUE; if (src) *src = ZPROP_SRC_TEMPORARY; } else if (hasmntopt(&mnt, mntopt_off) && *val) { *val = B_FALSE; if (src) *src = ZPROP_SRC_TEMPORARY; } break; case ZFS_PROP_CANMOUNT: case ZFS_PROP_VOLSIZE: case ZFS_PROP_QUOTA: case ZFS_PROP_REFQUOTA: case ZFS_PROP_RESERVATION: case ZFS_PROP_REFRESERVATION: case ZFS_PROP_FILESYSTEM_LIMIT: case ZFS_PROP_SNAPSHOT_LIMIT: case ZFS_PROP_FILESYSTEM_COUNT: case ZFS_PROP_SNAPSHOT_COUNT: *val = getprop_uint64(zhp, prop, source); if (*source == NULL) { /* not default, must be local */ *source = zhp->zfs_name; } break; case ZFS_PROP_MOUNTED: *val = (zhp->zfs_mntopts != NULL); break; case ZFS_PROP_NUMCLONES: *val = zhp->zfs_dmustats.dds_num_clones; break; case ZFS_PROP_VERSION: case ZFS_PROP_NORMALIZE: case ZFS_PROP_UTF8ONLY: case ZFS_PROP_CASE: zcmd_alloc_dst_nvlist(zhp->zfs_hdl, &zc, 0); (void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name)); if (zfs_ioctl(zhp->zfs_hdl, ZFS_IOC_OBJSET_ZPLPROPS, &zc)) { zcmd_free_nvlists(&zc); if (prop == ZFS_PROP_VERSION && zhp->zfs_type == ZFS_TYPE_VOLUME) *val = zfs_prop_default_numeric(prop); return (-1); } if (zcmd_read_dst_nvlist(zhp->zfs_hdl, &zc, &zplprops) != 0 || nvlist_lookup_uint64(zplprops, zfs_prop_to_name(prop), val) != 0) { zcmd_free_nvlists(&zc); return (-1); } nvlist_free(zplprops); zcmd_free_nvlists(&zc); break; case ZFS_PROP_INCONSISTENT: *val = zhp->zfs_dmustats.dds_inconsistent; break; case ZFS_PROP_REDACTED: *val = zhp->zfs_dmustats.dds_redacted; break; case ZFS_PROP_CREATETXG: /* * We can directly read createtxg property from zfs * handle for Filesystem, Snapshot and ZVOL types. */ if ((zhp->zfs_type == ZFS_TYPE_FILESYSTEM) || (zhp->zfs_type == ZFS_TYPE_SNAPSHOT) || (zhp->zfs_type == ZFS_TYPE_VOLUME)) { *val = zhp->zfs_dmustats.dds_creation_txg; break; } zfs_fallthrough; default: switch (zfs_prop_get_type(prop)) { case PROP_TYPE_NUMBER: case PROP_TYPE_INDEX: *val = getprop_uint64(zhp, prop, source); /* * If we tried to use a default value for a * readonly property, it means that it was not * present. Note this only applies to "truly" * readonly properties, not set-once properties * like volblocksize. */ if (zfs_prop_readonly(prop) && !zfs_prop_setonce(prop) && *source != NULL && (*source)[0] == '\0') { *source = NULL; return (-1); } break; case PROP_TYPE_STRING: default: zfs_error_aux(zhp->zfs_hdl, dgettext(TEXT_DOMAIN, "cannot get non-numeric property")); return (zfs_error(zhp->zfs_hdl, EZFS_BADPROP, dgettext(TEXT_DOMAIN, "internal error"))); } } return (0); } /* * Calculate the source type, given the raw source string. */ static void get_source(zfs_handle_t *zhp, zprop_source_t *srctype, char *source, char *statbuf, size_t statlen) { if (statbuf == NULL || srctype == NULL || *srctype == ZPROP_SRC_TEMPORARY) { return; } if (source == NULL) { *srctype = ZPROP_SRC_NONE; } else if (source[0] == '\0') { *srctype = ZPROP_SRC_DEFAULT; } else if (strstr(source, ZPROP_SOURCE_VAL_RECVD) != NULL) { *srctype = ZPROP_SRC_RECEIVED; } else { if (strcmp(source, zhp->zfs_name) == 0) { *srctype = ZPROP_SRC_LOCAL; } else { (void) strlcpy(statbuf, source, statlen); *srctype = ZPROP_SRC_INHERITED; } } } int zfs_prop_get_recvd(zfs_handle_t *zhp, const char *propname, char *propbuf, size_t proplen, boolean_t literal) { zfs_prop_t prop; int err = 0; if (zhp->zfs_recvd_props == NULL) if (get_recvd_props_ioctl(zhp) != 0) return (-1); prop = zfs_name_to_prop(propname); if (prop != ZPROP_USERPROP) { uint64_t cookie; if (!nvlist_exists(zhp->zfs_recvd_props, propname)) return (-1); zfs_set_recvd_props_mode(zhp, &cookie); err = zfs_prop_get(zhp, prop, propbuf, proplen, NULL, NULL, 0, literal); zfs_unset_recvd_props_mode(zhp, &cookie); } else { nvlist_t *propval; char *recvdval; if (nvlist_lookup_nvlist(zhp->zfs_recvd_props, propname, &propval) != 0) return (-1); recvdval = fnvlist_lookup_string(propval, ZPROP_VALUE); (void) strlcpy(propbuf, recvdval, proplen); } return (err == 0 ? 0 : -1); } static int get_clones_string(zfs_handle_t *zhp, char *propbuf, size_t proplen) { nvlist_t *value; nvpair_t *pair; value = zfs_get_clones_nvl(zhp); if (value == NULL || nvlist_empty(value)) return (-1); propbuf[0] = '\0'; for (pair = nvlist_next_nvpair(value, NULL); pair != NULL; pair = nvlist_next_nvpair(value, pair)) { if (propbuf[0] != '\0') (void) strlcat(propbuf, ",", proplen); (void) strlcat(propbuf, nvpair_name(pair), proplen); } return (0); } struct get_clones_arg { uint64_t numclones; nvlist_t *value; const char *origin; char buf[ZFS_MAX_DATASET_NAME_LEN]; }; static int get_clones_cb(zfs_handle_t *zhp, void *arg) { struct get_clones_arg *gca = arg; if (gca->numclones == 0) { zfs_close(zhp); return (0); } if (zfs_prop_get(zhp, ZFS_PROP_ORIGIN, gca->buf, sizeof (gca->buf), NULL, NULL, 0, B_TRUE) != 0) goto out; if (strcmp(gca->buf, gca->origin) == 0) { fnvlist_add_boolean(gca->value, zfs_get_name(zhp)); gca->numclones--; } out: (void) zfs_iter_children(zhp, get_clones_cb, gca); zfs_close(zhp); return (0); } nvlist_t * zfs_get_clones_nvl(zfs_handle_t *zhp) { nvlist_t *nv, *value; if (nvlist_lookup_nvlist(zhp->zfs_props, zfs_prop_to_name(ZFS_PROP_CLONES), &nv) != 0) { struct get_clones_arg gca; /* * if this is a snapshot, then the kernel wasn't able * to get the clones. Do it by slowly iterating. */ if (zhp->zfs_type != ZFS_TYPE_SNAPSHOT) return (NULL); if (nvlist_alloc(&nv, NV_UNIQUE_NAME, 0) != 0) return (NULL); if (nvlist_alloc(&value, NV_UNIQUE_NAME, 0) != 0) { nvlist_free(nv); return (NULL); } gca.numclones = zfs_prop_get_int(zhp, ZFS_PROP_NUMCLONES); gca.value = value; gca.origin = zhp->zfs_name; if (gca.numclones != 0) { zfs_handle_t *root; char pool[ZFS_MAX_DATASET_NAME_LEN]; char *cp = pool; /* get the pool name */ (void) strlcpy(pool, zhp->zfs_name, sizeof (pool)); (void) strsep(&cp, "/@"); root = zfs_open(zhp->zfs_hdl, pool, ZFS_TYPE_FILESYSTEM); if (root == NULL) { nvlist_free(nv); nvlist_free(value); return (NULL); } (void) get_clones_cb(root, &gca); } if (gca.numclones != 0 || nvlist_add_nvlist(nv, ZPROP_VALUE, value) != 0 || nvlist_add_nvlist(zhp->zfs_props, zfs_prop_to_name(ZFS_PROP_CLONES), nv) != 0) { nvlist_free(nv); nvlist_free(value); return (NULL); } nvlist_free(nv); nvlist_free(value); nv = fnvlist_lookup_nvlist(zhp->zfs_props, zfs_prop_to_name(ZFS_PROP_CLONES)); } return (fnvlist_lookup_nvlist(nv, ZPROP_VALUE)); } static int get_rsnaps_string(zfs_handle_t *zhp, char *propbuf, size_t proplen) { nvlist_t *value; uint64_t *snaps; uint_t nsnaps; if (nvlist_lookup_nvlist(zhp->zfs_props, zfs_prop_to_name(ZFS_PROP_REDACT_SNAPS), &value) != 0) return (-1); if (nvlist_lookup_uint64_array(value, ZPROP_VALUE, &snaps, &nsnaps) != 0) return (-1); if (nsnaps == 0) { /* There's no redaction snapshots; pass a special value back */ (void) snprintf(propbuf, proplen, "none"); return (0); } propbuf[0] = '\0'; for (int i = 0; i < nsnaps; i++) { char buf[128]; if (propbuf[0] != '\0') (void) strlcat(propbuf, ",", proplen); (void) snprintf(buf, sizeof (buf), "%llu", (u_longlong_t)snaps[i]); (void) strlcat(propbuf, buf, proplen); } return (0); } /* * Accepts a property and value and checks that the value * matches the one found by the channel program. If they are * not equal, print both of them. */ static void zcp_check(zfs_handle_t *zhp, zfs_prop_t prop, uint64_t intval, const char *strval) { if (!zhp->zfs_hdl->libzfs_prop_debug) return; int error; char *poolname = zhp->zpool_hdl->zpool_name; const char *prop_name = zfs_prop_to_name(prop); const char *program = "args = ...\n" "ds = args['dataset']\n" "prop = args['property']\n" "value, setpoint = zfs.get_prop(ds, prop)\n" "return {value=value, setpoint=setpoint}\n"; nvlist_t *outnvl; nvlist_t *retnvl; nvlist_t *argnvl = fnvlist_alloc(); fnvlist_add_string(argnvl, "dataset", zhp->zfs_name); fnvlist_add_string(argnvl, "property", zfs_prop_to_name(prop)); error = lzc_channel_program_nosync(poolname, program, 10 * 1000 * 1000, 10 * 1024 * 1024, argnvl, &outnvl); if (error == 0) { retnvl = fnvlist_lookup_nvlist(outnvl, "return"); if (zfs_prop_get_type(prop) == PROP_TYPE_NUMBER) { int64_t ans; error = nvlist_lookup_int64(retnvl, "value", &ans); if (error != 0) { (void) fprintf(stderr, "%s: zcp check error: " "%u\n", prop_name, error); return; } if (ans != intval) { (void) fprintf(stderr, "%s: zfs found %llu, " "but zcp found %llu\n", prop_name, (u_longlong_t)intval, (u_longlong_t)ans); } } else { char *str_ans; error = nvlist_lookup_string(retnvl, "value", &str_ans); if (error != 0) { (void) fprintf(stderr, "%s: zcp check error: " "%u\n", prop_name, error); return; } if (strcmp(strval, str_ans) != 0) { (void) fprintf(stderr, "%s: zfs found '%s', but zcp found '%s'\n", prop_name, strval, str_ans); } } } else { (void) fprintf(stderr, "%s: zcp check failed, channel program " "error: %u\n", prop_name, error); } nvlist_free(argnvl); nvlist_free(outnvl); } /* * Retrieve a property from the given object. If 'literal' is specified, then * numbers are left as exact values. Otherwise, numbers are converted to a * human-readable form. * * Returns 0 on success, or -1 on error. */ int zfs_prop_get(zfs_handle_t *zhp, zfs_prop_t prop, char *propbuf, size_t proplen, zprop_source_t *src, char *statbuf, size_t statlen, boolean_t literal) { char *source = NULL; uint64_t val; const char *str; const char *strval; boolean_t received = zfs_is_recvd_props_mode(zhp); /* * Check to see if this property applies to our object */ if (!zfs_prop_valid_for_type(prop, zhp->zfs_type, B_FALSE)) return (-1); if (received && zfs_prop_readonly(prop)) return (-1); if (src) *src = ZPROP_SRC_NONE; switch (prop) { case ZFS_PROP_CREATION: /* * 'creation' is a time_t stored in the statistics. We convert * this into a string unless 'literal' is specified. */ { val = getprop_uint64(zhp, prop, &source); time_t time = (time_t)val; struct tm t; if (literal || localtime_r(&time, &t) == NULL || strftime(propbuf, proplen, "%a %b %e %k:%M %Y", &t) == 0) (void) snprintf(propbuf, proplen, "%llu", (u_longlong_t)val); } zcp_check(zhp, prop, val, NULL); break; case ZFS_PROP_MOUNTPOINT: /* * Getting the precise mountpoint can be tricky. * * - for 'none' or 'legacy', return those values. * - for inherited mountpoints, we want to take everything * after our ancestor and append it to the inherited value. * * If the pool has an alternate root, we want to prepend that * root to any values we return. */ str = getprop_string(zhp, prop, &source); if (str[0] == '/') { char buf[MAXPATHLEN]; char *root = buf; const char *relpath; /* * If we inherit the mountpoint, even from a dataset * with a received value, the source will be the path of * the dataset we inherit from. If source is * ZPROP_SOURCE_VAL_RECVD, the received value is not * inherited. */ if (strcmp(source, ZPROP_SOURCE_VAL_RECVD) == 0) { relpath = ""; } else { relpath = zhp->zfs_name + strlen(source); if (relpath[0] == '/') relpath++; } if ((zpool_get_prop(zhp->zpool_hdl, ZPOOL_PROP_ALTROOT, buf, MAXPATHLEN, NULL, B_FALSE)) || (strcmp(root, "-") == 0)) root[0] = '\0'; /* * Special case an alternate root of '/'. This will * avoid having multiple leading slashes in the * mountpoint path. */ if (strcmp(root, "/") == 0) root++; /* * If the mountpoint is '/' then skip over this * if we are obtaining either an alternate root or * an inherited mountpoint. */ if (str[1] == '\0' && (root[0] != '\0' || relpath[0] != '\0')) str++; if (relpath[0] == '\0') (void) snprintf(propbuf, proplen, "%s%s", root, str); else (void) snprintf(propbuf, proplen, "%s%s%s%s", root, str, relpath[0] == '@' ? "" : "/", relpath); } else { /* 'legacy' or 'none' */ (void) strlcpy(propbuf, str, proplen); } zcp_check(zhp, prop, 0, propbuf); break; case ZFS_PROP_ORIGIN: str = getprop_string(zhp, prop, &source); if (str == NULL) return (-1); (void) strlcpy(propbuf, str, proplen); zcp_check(zhp, prop, 0, str); break; case ZFS_PROP_REDACT_SNAPS: if (get_rsnaps_string(zhp, propbuf, proplen) != 0) return (-1); break; case ZFS_PROP_CLONES: if (get_clones_string(zhp, propbuf, proplen) != 0) return (-1); break; case ZFS_PROP_QUOTA: case ZFS_PROP_REFQUOTA: case ZFS_PROP_RESERVATION: case ZFS_PROP_REFRESERVATION: if (get_numeric_property(zhp, prop, src, &source, &val) != 0) return (-1); /* * If quota or reservation is 0, we translate this into 'none' * (unless literal is set), and indicate that it's the default * value. Otherwise, we print the number nicely and indicate * that its set locally. */ if (val == 0) { if (literal) (void) strlcpy(propbuf, "0", proplen); else (void) strlcpy(propbuf, "none", proplen); } else { if (literal) (void) snprintf(propbuf, proplen, "%llu", (u_longlong_t)val); else zfs_nicebytes(val, propbuf, proplen); } zcp_check(zhp, prop, val, NULL); break; case ZFS_PROP_FILESYSTEM_LIMIT: case ZFS_PROP_SNAPSHOT_LIMIT: case ZFS_PROP_FILESYSTEM_COUNT: case ZFS_PROP_SNAPSHOT_COUNT: if (get_numeric_property(zhp, prop, src, &source, &val) != 0) return (-1); /* * If limit is UINT64_MAX, we translate this into 'none', and * indicate that it's the default value. Otherwise, we print * the number nicely and indicate that it's set locally. */ if (val == UINT64_MAX) { (void) strlcpy(propbuf, "none", proplen); } else if (literal) { (void) snprintf(propbuf, proplen, "%llu", (u_longlong_t)val); } else { zfs_nicenum(val, propbuf, proplen); } zcp_check(zhp, prop, val, NULL); break; case ZFS_PROP_REFRATIO: case ZFS_PROP_COMPRESSRATIO: if (get_numeric_property(zhp, prop, src, &source, &val) != 0) return (-1); if (literal) (void) snprintf(propbuf, proplen, "%llu.%02llu", (u_longlong_t)(val / 100), (u_longlong_t)(val % 100)); else (void) snprintf(propbuf, proplen, "%llu.%02llux", (u_longlong_t)(val / 100), (u_longlong_t)(val % 100)); zcp_check(zhp, prop, val, NULL); break; case ZFS_PROP_TYPE: switch (zhp->zfs_type) { case ZFS_TYPE_FILESYSTEM: str = "filesystem"; break; case ZFS_TYPE_VOLUME: str = "volume"; break; case ZFS_TYPE_SNAPSHOT: str = "snapshot"; break; case ZFS_TYPE_BOOKMARK: str = "bookmark"; break; default: abort(); } (void) snprintf(propbuf, proplen, "%s", str); zcp_check(zhp, prop, 0, propbuf); break; case ZFS_PROP_MOUNTED: /* * The 'mounted' property is a pseudo-property that described * whether the filesystem is currently mounted. Even though * it's a boolean value, the typical values of "on" and "off" * don't make sense, so we translate to "yes" and "no". */ if (get_numeric_property(zhp, ZFS_PROP_MOUNTED, src, &source, &val) != 0) return (-1); if (val) (void) strlcpy(propbuf, "yes", proplen); else (void) strlcpy(propbuf, "no", proplen); break; case ZFS_PROP_NAME: /* * The 'name' property is a pseudo-property derived from the * dataset name. It is presented as a real property to simplify * consumers. */ (void) strlcpy(propbuf, zhp->zfs_name, proplen); zcp_check(zhp, prop, 0, propbuf); break; case ZFS_PROP_MLSLABEL: { #ifdef HAVE_MLSLABEL m_label_t *new_sl = NULL; char *ascii = NULL; /* human readable label */ (void) strlcpy(propbuf, getprop_string(zhp, prop, &source), proplen); if (literal || (strcasecmp(propbuf, ZFS_MLSLABEL_DEFAULT) == 0)) break; /* * Try to translate the internal hex string to * human-readable output. If there are any * problems just use the hex string. */ if (str_to_label(propbuf, &new_sl, MAC_LABEL, L_NO_CORRECTION, NULL) == -1) { m_label_free(new_sl); break; } if (label_to_str(new_sl, &ascii, M_LABEL, DEF_NAMES) != 0) { if (ascii) free(ascii); m_label_free(new_sl); break; } m_label_free(new_sl); (void) strlcpy(propbuf, ascii, proplen); free(ascii); #else (void) strlcpy(propbuf, getprop_string(zhp, prop, &source), proplen); #endif /* HAVE_MLSLABEL */ } break; case ZFS_PROP_GUID: case ZFS_PROP_KEY_GUID: case ZFS_PROP_IVSET_GUID: case ZFS_PROP_CREATETXG: case ZFS_PROP_OBJSETID: case ZFS_PROP_PBKDF2_ITERS: /* * These properties are stored as numbers, but they are * identifiers or counters. * We don't want them to be pretty printed, because pretty * printing truncates their values making them useless. */ if (get_numeric_property(zhp, prop, src, &source, &val) != 0) return (-1); (void) snprintf(propbuf, proplen, "%llu", (u_longlong_t)val); zcp_check(zhp, prop, val, NULL); break; case ZFS_PROP_REFERENCED: case ZFS_PROP_AVAILABLE: case ZFS_PROP_USED: case ZFS_PROP_USEDSNAP: case ZFS_PROP_USEDDS: case ZFS_PROP_USEDREFRESERV: case ZFS_PROP_USEDCHILD: if (get_numeric_property(zhp, prop, src, &source, &val) != 0) return (-1); if (literal) { (void) snprintf(propbuf, proplen, "%llu", (u_longlong_t)val); } else { zfs_nicebytes(val, propbuf, proplen); } zcp_check(zhp, prop, val, NULL); break; case ZFS_PROP_SNAPSHOTS_CHANGED: { if ((get_numeric_property(zhp, prop, src, &source, &val) != 0) || val == 0) { return (-1); } time_t time = (time_t)val; struct tm t; if (literal || localtime_r(&time, &t) == NULL || strftime(propbuf, proplen, "%a %b %e %k:%M:%S %Y", &t) == 0) (void) snprintf(propbuf, proplen, "%llu", (u_longlong_t)val); } zcp_check(zhp, prop, val, NULL); break; default: switch (zfs_prop_get_type(prop)) { case PROP_TYPE_NUMBER: if (get_numeric_property(zhp, prop, src, &source, &val) != 0) { return (-1); } if (literal) { (void) snprintf(propbuf, proplen, "%llu", (u_longlong_t)val); } else { zfs_nicenum(val, propbuf, proplen); } zcp_check(zhp, prop, val, NULL); break; case PROP_TYPE_STRING: str = getprop_string(zhp, prop, &source); if (str == NULL) return (-1); (void) strlcpy(propbuf, str, proplen); zcp_check(zhp, prop, 0, str); break; case PROP_TYPE_INDEX: if (get_numeric_property(zhp, prop, src, &source, &val) != 0) return (-1); if (zfs_prop_index_to_string(prop, val, &strval) != 0) return (-1); (void) strlcpy(propbuf, strval, proplen); zcp_check(zhp, prop, 0, strval); break; default: abort(); } } get_source(zhp, src, source, statbuf, statlen); return (0); } /* * Utility function to get the given numeric property. Does no validation that * the given property is the appropriate type; should only be used with * hard-coded property types. */ uint64_t zfs_prop_get_int(zfs_handle_t *zhp, zfs_prop_t prop) { char *source; uint64_t val = 0; (void) get_numeric_property(zhp, prop, NULL, &source, &val); return (val); } static int zfs_prop_set_int(zfs_handle_t *zhp, zfs_prop_t prop, uint64_t val) { char buf[64]; (void) snprintf(buf, sizeof (buf), "%llu", (longlong_t)val); return (zfs_prop_set(zhp, zfs_prop_to_name(prop), buf)); } /* * Similar to zfs_prop_get(), but returns the value as an integer. */ int zfs_prop_get_numeric(zfs_handle_t *zhp, zfs_prop_t prop, uint64_t *value, zprop_source_t *src, char *statbuf, size_t statlen) { char *source; /* * Check to see if this property applies to our object */ if (!zfs_prop_valid_for_type(prop, zhp->zfs_type, B_FALSE)) { return (zfs_error_fmt(zhp->zfs_hdl, EZFS_PROPTYPE, dgettext(TEXT_DOMAIN, "cannot get property '%s'"), zfs_prop_to_name(prop))); } if (src) *src = ZPROP_SRC_NONE; if (get_numeric_property(zhp, prop, src, &source, value) != 0) return (-1); get_source(zhp, src, source, statbuf, statlen); return (0); } #ifdef HAVE_IDMAP static int idmap_id_to_numeric_domain_rid(uid_t id, boolean_t isuser, char **domainp, idmap_rid_t *ridp) { idmap_get_handle_t *get_hdl = NULL; idmap_stat status; int err = EINVAL; if (idmap_get_create(&get_hdl) != IDMAP_SUCCESS) goto out; if (isuser) { err = idmap_get_sidbyuid(get_hdl, id, IDMAP_REQ_FLG_USE_CACHE, domainp, ridp, &status); } else { err = idmap_get_sidbygid(get_hdl, id, IDMAP_REQ_FLG_USE_CACHE, domainp, ridp, &status); } if (err == IDMAP_SUCCESS && idmap_get_mappings(get_hdl) == IDMAP_SUCCESS && status == IDMAP_SUCCESS) err = 0; else err = EINVAL; out: if (get_hdl) idmap_get_destroy(get_hdl); return (err); } #endif /* HAVE_IDMAP */ /* * convert the propname into parameters needed by kernel * Eg: userquota@ahrens -> ZFS_PROP_USERQUOTA, "", 126829 * Eg: userused@matt@domain -> ZFS_PROP_USERUSED, "S-1-123-456", 789 * Eg: groupquota@staff -> ZFS_PROP_GROUPQUOTA, "", 1234 * Eg: groupused@staff -> ZFS_PROP_GROUPUSED, "", 1234 * Eg: projectquota@123 -> ZFS_PROP_PROJECTQUOTA, "", 123 * Eg: projectused@789 -> ZFS_PROP_PROJECTUSED, "", 789 */ static int userquota_propname_decode(const char *propname, boolean_t zoned, zfs_userquota_prop_t *typep, char *domain, int domainlen, uint64_t *ridp) { zfs_userquota_prop_t type; char *cp; boolean_t isuser; boolean_t isgroup; boolean_t isproject; struct passwd *pw; struct group *gr; domain[0] = '\0'; /* Figure out the property type ({user|group|project}{quota|space}) */ for (type = 0; type < ZFS_NUM_USERQUOTA_PROPS; type++) { if (strncmp(propname, zfs_userquota_prop_prefixes[type], strlen(zfs_userquota_prop_prefixes[type])) == 0) break; } if (type == ZFS_NUM_USERQUOTA_PROPS) return (EINVAL); *typep = type; isuser = (type == ZFS_PROP_USERQUOTA || type == ZFS_PROP_USERUSED || type == ZFS_PROP_USEROBJQUOTA || type == ZFS_PROP_USEROBJUSED); isgroup = (type == ZFS_PROP_GROUPQUOTA || type == ZFS_PROP_GROUPUSED || type == ZFS_PROP_GROUPOBJQUOTA || type == ZFS_PROP_GROUPOBJUSED); isproject = (type == ZFS_PROP_PROJECTQUOTA || type == ZFS_PROP_PROJECTUSED || type == ZFS_PROP_PROJECTOBJQUOTA || type == ZFS_PROP_PROJECTOBJUSED); cp = strchr(propname, '@') + 1; if (isuser && (pw = getpwnam(cp)) != NULL) { if (zoned && getzoneid() == GLOBAL_ZONEID) return (ENOENT); *ridp = pw->pw_uid; } else if (isgroup && (gr = getgrnam(cp)) != NULL) { if (zoned && getzoneid() == GLOBAL_ZONEID) return (ENOENT); *ridp = gr->gr_gid; } else if (!isproject && strchr(cp, '@')) { #ifdef HAVE_IDMAP /* * It's a SID name (eg "user@domain") that needs to be * turned into S-1-domainID-RID. */ directory_error_t e; char *numericsid = NULL; char *end; if (zoned && getzoneid() == GLOBAL_ZONEID) return (ENOENT); if (isuser) { e = directory_sid_from_user_name(NULL, cp, &numericsid); } else { e = directory_sid_from_group_name(NULL, cp, &numericsid); } if (e != NULL) { directory_error_free(e); return (ENOENT); } if (numericsid == NULL) return (ENOENT); cp = numericsid; (void) strlcpy(domain, cp, domainlen); cp = strrchr(domain, '-'); *cp = '\0'; cp++; errno = 0; *ridp = strtoull(cp, &end, 10); free(numericsid); if (errno != 0 || *end != '\0') return (EINVAL); #else (void) domainlen; return (ENOSYS); #endif /* HAVE_IDMAP */ } else { /* It's a user/group/project ID (eg "12345"). */ uid_t id; char *end; id = strtoul(cp, &end, 10); if (*end != '\0') return (EINVAL); if (id > MAXUID && !isproject) { #ifdef HAVE_IDMAP /* It's an ephemeral ID. */ idmap_rid_t rid; char *mapdomain; if (idmap_id_to_numeric_domain_rid(id, isuser, &mapdomain, &rid) != 0) return (ENOENT); (void) strlcpy(domain, mapdomain, domainlen); *ridp = rid; #else return (ENOSYS); #endif /* HAVE_IDMAP */ } else { *ridp = id; } } return (0); } static int zfs_prop_get_userquota_common(zfs_handle_t *zhp, const char *propname, uint64_t *propvalue, zfs_userquota_prop_t *typep) { int err; zfs_cmd_t zc = {"\0"}; (void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name)); err = userquota_propname_decode(propname, zfs_prop_get_int(zhp, ZFS_PROP_ZONED), typep, zc.zc_value, sizeof (zc.zc_value), &zc.zc_guid); zc.zc_objset_type = *typep; if (err) return (err); err = zfs_ioctl(zhp->zfs_hdl, ZFS_IOC_USERSPACE_ONE, &zc); if (err) return (err); *propvalue = zc.zc_cookie; return (0); } int zfs_prop_get_userquota_int(zfs_handle_t *zhp, const char *propname, uint64_t *propvalue) { zfs_userquota_prop_t type; return (zfs_prop_get_userquota_common(zhp, propname, propvalue, &type)); } int zfs_prop_get_userquota(zfs_handle_t *zhp, const char *propname, char *propbuf, int proplen, boolean_t literal) { int err; uint64_t propvalue; zfs_userquota_prop_t type; err = zfs_prop_get_userquota_common(zhp, propname, &propvalue, &type); if (err) return (err); if (literal) { (void) snprintf(propbuf, proplen, "%llu", (u_longlong_t)propvalue); } else if (propvalue == 0 && (type == ZFS_PROP_USERQUOTA || type == ZFS_PROP_GROUPQUOTA || type == ZFS_PROP_USEROBJQUOTA || type == ZFS_PROP_GROUPOBJQUOTA || type == ZFS_PROP_PROJECTQUOTA || type == ZFS_PROP_PROJECTOBJQUOTA)) { (void) strlcpy(propbuf, "none", proplen); } else if (type == ZFS_PROP_USERQUOTA || type == ZFS_PROP_GROUPQUOTA || type == ZFS_PROP_USERUSED || type == ZFS_PROP_GROUPUSED || type == ZFS_PROP_PROJECTUSED || type == ZFS_PROP_PROJECTQUOTA) { zfs_nicebytes(propvalue, propbuf, proplen); } else { zfs_nicenum(propvalue, propbuf, proplen); } return (0); } /* * propname must start with "written@" or "written#". */ int zfs_prop_get_written_int(zfs_handle_t *zhp, const char *propname, uint64_t *propvalue) { int err; zfs_cmd_t zc = {"\0"}; const char *snapname; (void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name)); assert(zfs_prop_written(propname)); snapname = propname + strlen("written@"); if (strchr(snapname, '@') != NULL || strchr(snapname, '#') != NULL) { /* full snapshot or bookmark name specified */ (void) strlcpy(zc.zc_value, snapname, sizeof (zc.zc_value)); } else { /* snapname is the short name, append it to zhp's fsname */ char *cp; (void) strlcpy(zc.zc_value, zhp->zfs_name, sizeof (zc.zc_value)); cp = strchr(zc.zc_value, '@'); if (cp != NULL) *cp = '\0'; (void) strlcat(zc.zc_value, snapname - 1, sizeof (zc.zc_value)); } err = zfs_ioctl(zhp->zfs_hdl, ZFS_IOC_SPACE_WRITTEN, &zc); if (err) return (err); *propvalue = zc.zc_cookie; return (0); } int zfs_prop_get_written(zfs_handle_t *zhp, const char *propname, char *propbuf, int proplen, boolean_t literal) { int err; uint64_t propvalue; err = zfs_prop_get_written_int(zhp, propname, &propvalue); if (err) return (err); if (literal) { (void) snprintf(propbuf, proplen, "%llu", (u_longlong_t)propvalue); } else { zfs_nicebytes(propvalue, propbuf, proplen); } return (0); } /* * Returns the name of the given zfs handle. */ const char * zfs_get_name(const zfs_handle_t *zhp) { return (zhp->zfs_name); } /* * Returns the name of the parent pool for the given zfs handle. */ const char * zfs_get_pool_name(const zfs_handle_t *zhp) { return (zhp->zpool_hdl->zpool_name); } /* * Returns the type of the given zfs handle. */ zfs_type_t zfs_get_type(const zfs_handle_t *zhp) { return (zhp->zfs_type); } /* * Returns the type of the given zfs handle, * or, if a snapshot, the type of the snapshotted dataset. */ zfs_type_t zfs_get_underlying_type(const zfs_handle_t *zhp) { return (zhp->zfs_head_type); } /* * Is one dataset name a child dataset of another? * * Needs to handle these cases: * Dataset 1 "a/foo" "a/foo" "a/foo" "a/foo" * Dataset 2 "a/fo" "a/foobar" "a/bar/baz" "a/foo/bar" * Descendant? No. No. No. Yes. */ static boolean_t is_descendant(const char *ds1, const char *ds2) { size_t d1len = strlen(ds1); /* ds2 can't be a descendant if it's smaller */ if (strlen(ds2) < d1len) return (B_FALSE); /* otherwise, compare strings and verify that there's a '/' char */ return (ds2[d1len] == '/' && (strncmp(ds1, ds2, d1len) == 0)); } /* * Given a complete name, return just the portion that refers to the parent. * Will return -1 if there is no parent (path is just the name of the * pool). */ static int parent_name(const char *path, char *buf, size_t buflen) { char *slashp; (void) strlcpy(buf, path, buflen); if ((slashp = strrchr(buf, '/')) == NULL) return (-1); *slashp = '\0'; return (0); } int zfs_parent_name(zfs_handle_t *zhp, char *buf, size_t buflen) { return (parent_name(zfs_get_name(zhp), buf, buflen)); } /* * If accept_ancestor is false, then check to make sure that the given path has * a parent, and that it exists. If accept_ancestor is true, then find the * closest existing ancestor for the given path. In prefixlen return the * length of already existing prefix of the given path. We also fetch the * 'zoned' property, which is used to validate property settings when creating * new datasets. */ static int check_parents(libzfs_handle_t *hdl, const char *path, uint64_t *zoned, boolean_t accept_ancestor, int *prefixlen) { zfs_cmd_t zc = {"\0"}; char parent[ZFS_MAX_DATASET_NAME_LEN]; char *slash; zfs_handle_t *zhp; char errbuf[ERRBUFLEN]; uint64_t is_zoned; (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot create '%s'"), path); /* get parent, and check to see if this is just a pool */ if (parent_name(path, parent, sizeof (parent)) != 0) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "missing dataset name")); return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf)); } /* check to see if the pool exists */ if ((slash = strchr(parent, '/')) == NULL) slash = parent + strlen(parent); (void) strlcpy(zc.zc_name, parent, MIN(sizeof (zc.zc_name), slash - parent + 1)); if (zfs_ioctl(hdl, ZFS_IOC_OBJSET_STATS, &zc) != 0 && errno == ENOENT) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "no such pool '%s'"), zc.zc_name); return (zfs_error(hdl, EZFS_NOENT, errbuf)); } /* check to see if the parent dataset exists */ while ((zhp = make_dataset_handle(hdl, parent)) == NULL) { if (errno == ENOENT && accept_ancestor) { /* * Go deeper to find an ancestor, give up on top level. */ if (parent_name(parent, parent, sizeof (parent)) != 0) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "no such pool '%s'"), zc.zc_name); return (zfs_error(hdl, EZFS_NOENT, errbuf)); } } else if (errno == ENOENT) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "parent does not exist")); return (zfs_error(hdl, EZFS_NOENT, errbuf)); } else return (zfs_standard_error(hdl, errno, errbuf)); } is_zoned = zfs_prop_get_int(zhp, ZFS_PROP_ZONED); if (zoned != NULL) *zoned = is_zoned; /* we are in a non-global zone, but parent is in the global zone */ if (getzoneid() != GLOBAL_ZONEID && !is_zoned) { (void) zfs_standard_error(hdl, EPERM, errbuf); zfs_close(zhp); return (-1); } /* make sure parent is a filesystem */ if (zfs_get_type(zhp) != ZFS_TYPE_FILESYSTEM) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "parent is not a filesystem")); (void) zfs_error(hdl, EZFS_BADTYPE, errbuf); zfs_close(zhp); return (-1); } zfs_close(zhp); if (prefixlen != NULL) *prefixlen = strlen(parent); return (0); } /* * Finds whether the dataset of the given type(s) exists. */ boolean_t zfs_dataset_exists(libzfs_handle_t *hdl, const char *path, zfs_type_t types) { zfs_handle_t *zhp; if (!zfs_validate_name(hdl, path, types, B_FALSE)) return (B_FALSE); /* * Try to get stats for the dataset, which will tell us if it exists. */ if ((zhp = make_dataset_handle(hdl, path)) != NULL) { int ds_type = zhp->zfs_type; zfs_close(zhp); if (types & ds_type) return (B_TRUE); } return (B_FALSE); } /* * Given a path to 'target', create all the ancestors between * the prefixlen portion of the path, and the target itself. * Fail if the initial prefixlen-ancestor does not already exist. */ int create_parents(libzfs_handle_t *hdl, char *target, int prefixlen) { zfs_handle_t *h; char *cp; const char *opname; /* make sure prefix exists */ cp = target + prefixlen; if (*cp != '/') { assert(strchr(cp, '/') == NULL); h = zfs_open(hdl, target, ZFS_TYPE_FILESYSTEM); } else { *cp = '\0'; h = zfs_open(hdl, target, ZFS_TYPE_FILESYSTEM); *cp = '/'; } if (h == NULL) return (-1); zfs_close(h); /* * Attempt to create, mount, and share any ancestor filesystems, * up to the prefixlen-long one. */ for (cp = target + prefixlen + 1; (cp = strchr(cp, '/')) != NULL; *cp = '/', cp++) { *cp = '\0'; h = make_dataset_handle(hdl, target); if (h) { /* it already exists, nothing to do here */ zfs_close(h); continue; } if (zfs_create(hdl, target, ZFS_TYPE_FILESYSTEM, NULL) != 0) { opname = dgettext(TEXT_DOMAIN, "create"); goto ancestorerr; } h = zfs_open(hdl, target, ZFS_TYPE_FILESYSTEM); if (h == NULL) { opname = dgettext(TEXT_DOMAIN, "open"); goto ancestorerr; } if (zfs_mount(h, NULL, 0) != 0) { opname = dgettext(TEXT_DOMAIN, "mount"); goto ancestorerr; } if (zfs_share(h, NULL) != 0) { opname = dgettext(TEXT_DOMAIN, "share"); goto ancestorerr; } zfs_close(h); } zfs_commit_shares(NULL); return (0); ancestorerr: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "failed to %s ancestor '%s'"), opname, target); return (-1); } /* * Creates non-existing ancestors of the given path. */ int zfs_create_ancestors(libzfs_handle_t *hdl, const char *path) { int prefix; char *path_copy; char errbuf[ERRBUFLEN]; int rc = 0; (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot create '%s'"), path); /* * Check that we are not passing the nesting limit * before we start creating any ancestors. */ if (dataset_nestcheck(path) != 0) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "maximum name nesting depth exceeded")); return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf)); } if (check_parents(hdl, path, NULL, B_TRUE, &prefix) != 0) return (-1); if ((path_copy = strdup(path)) != NULL) { rc = create_parents(hdl, path_copy, prefix); free(path_copy); } if (path_copy == NULL || rc != 0) return (-1); return (0); } /* * Create a new filesystem or volume. */ int zfs_create(libzfs_handle_t *hdl, const char *path, zfs_type_t type, nvlist_t *props) { int ret; uint64_t size = 0; uint64_t blocksize = zfs_prop_default_numeric(ZFS_PROP_VOLBLOCKSIZE); uint64_t zoned; enum lzc_dataset_type ost; zpool_handle_t *zpool_handle; uint8_t *wkeydata = NULL; uint_t wkeylen = 0; char errbuf[ERRBUFLEN]; char parent[ZFS_MAX_DATASET_NAME_LEN]; (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot create '%s'"), path); /* validate the path, taking care to note the extended error message */ if (!zfs_validate_name(hdl, path, type, B_TRUE)) return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf)); if (dataset_nestcheck(path) != 0) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "maximum name nesting depth exceeded")); return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf)); } /* validate parents exist */ if (check_parents(hdl, path, &zoned, B_FALSE, NULL) != 0) return (-1); /* * The failure modes when creating a dataset of a different type over * one that already exists is a little strange. In particular, if you * try to create a dataset on top of an existing dataset, the ioctl() * will return ENOENT, not EEXIST. To prevent this from happening, we * first try to see if the dataset exists. */ if (zfs_dataset_exists(hdl, path, ZFS_TYPE_DATASET)) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "dataset already exists")); return (zfs_error(hdl, EZFS_EXISTS, errbuf)); } if (type == ZFS_TYPE_VOLUME) ost = LZC_DATSET_TYPE_ZVOL; else ost = LZC_DATSET_TYPE_ZFS; /* open zpool handle for prop validation */ char pool_path[ZFS_MAX_DATASET_NAME_LEN]; (void) strlcpy(pool_path, path, sizeof (pool_path)); /* truncate pool_path at first slash */ char *p = strchr(pool_path, '/'); if (p != NULL) *p = '\0'; if ((zpool_handle = zpool_open(hdl, pool_path)) == NULL) return (-1); if (props && (props = zfs_valid_proplist(hdl, type, props, zoned, NULL, zpool_handle, B_TRUE, errbuf)) == 0) { zpool_close(zpool_handle); return (-1); } zpool_close(zpool_handle); if (type == ZFS_TYPE_VOLUME) { /* * If we are creating a volume, the size and block size must * satisfy a few restraints. First, the blocksize must be a * valid block size between SPA_{MIN,MAX}BLOCKSIZE. Second, the * volsize must be a multiple of the block size, and cannot be * zero. */ if (props == NULL || nvlist_lookup_uint64(props, zfs_prop_to_name(ZFS_PROP_VOLSIZE), &size) != 0) { nvlist_free(props); zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "missing volume size")); return (zfs_error(hdl, EZFS_BADPROP, errbuf)); } if ((ret = nvlist_lookup_uint64(props, zfs_prop_to_name(ZFS_PROP_VOLBLOCKSIZE), &blocksize)) != 0) { if (ret == ENOENT) { blocksize = zfs_prop_default_numeric( ZFS_PROP_VOLBLOCKSIZE); } else { nvlist_free(props); zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "missing volume block size")); return (zfs_error(hdl, EZFS_BADPROP, errbuf)); } } if (size == 0) { nvlist_free(props); zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "volume size cannot be zero")); return (zfs_error(hdl, EZFS_BADPROP, errbuf)); } if (size % blocksize != 0) { nvlist_free(props); zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "volume size must be a multiple of volume block " "size")); return (zfs_error(hdl, EZFS_BADPROP, errbuf)); } } (void) parent_name(path, parent, sizeof (parent)); if (zfs_crypto_create(hdl, parent, props, NULL, B_TRUE, &wkeydata, &wkeylen) != 0) { nvlist_free(props); return (zfs_error(hdl, EZFS_CRYPTOFAILED, errbuf)); } /* create the dataset */ ret = lzc_create(path, ost, props, wkeydata, wkeylen); nvlist_free(props); if (wkeydata != NULL) free(wkeydata); /* check for failure */ if (ret != 0) { switch (errno) { case ENOENT: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "no such parent '%s'"), parent); return (zfs_error(hdl, EZFS_NOENT, errbuf)); case ENOTSUP: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "pool must be upgraded to set this " "property or value")); return (zfs_error(hdl, EZFS_BADVERSION, errbuf)); case EACCES: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "encryption root's key is not loaded " "or provided")); return (zfs_error(hdl, EZFS_CRYPTOFAILED, errbuf)); case ERANGE: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "invalid property value(s) specified")); return (zfs_error(hdl, EZFS_BADPROP, errbuf)); #ifdef _ILP32 case EOVERFLOW: /* * This platform can't address a volume this big. */ if (type == ZFS_TYPE_VOLUME) return (zfs_error(hdl, EZFS_VOLTOOBIG, errbuf)); zfs_fallthrough; #endif default: return (zfs_standard_error(hdl, errno, errbuf)); } } return (0); } /* * Destroys the given dataset. The caller must make sure that the filesystem * isn't mounted, and that there are no active dependents. If the file system * does not exist this function does nothing. */ int zfs_destroy(zfs_handle_t *zhp, boolean_t defer) { int error; if (zhp->zfs_type != ZFS_TYPE_SNAPSHOT && defer) return (EINVAL); if (zhp->zfs_type == ZFS_TYPE_BOOKMARK) { nvlist_t *nv = fnvlist_alloc(); fnvlist_add_boolean(nv, zhp->zfs_name); error = lzc_destroy_bookmarks(nv, NULL); fnvlist_free(nv); if (error != 0) { return (zfs_standard_error_fmt(zhp->zfs_hdl, error, dgettext(TEXT_DOMAIN, "cannot destroy '%s'"), zhp->zfs_name)); } return (0); } if (zhp->zfs_type == ZFS_TYPE_SNAPSHOT) { nvlist_t *nv = fnvlist_alloc(); fnvlist_add_boolean(nv, zhp->zfs_name); error = lzc_destroy_snaps(nv, defer, NULL); fnvlist_free(nv); } else { error = lzc_destroy(zhp->zfs_name); } if (error != 0 && error != ENOENT) { return (zfs_standard_error_fmt(zhp->zfs_hdl, errno, dgettext(TEXT_DOMAIN, "cannot destroy '%s'"), zhp->zfs_name)); } remove_mountpoint(zhp); return (0); } struct destroydata { nvlist_t *nvl; const char *snapname; }; static int zfs_check_snap_cb(zfs_handle_t *zhp, void *arg) { struct destroydata *dd = arg; char name[ZFS_MAX_DATASET_NAME_LEN]; int rv = 0; if (snprintf(name, sizeof (name), "%s@%s", zhp->zfs_name, dd->snapname) >= sizeof (name)) return (EINVAL); if (lzc_exists(name)) fnvlist_add_boolean(dd->nvl, name); rv = zfs_iter_filesystems(zhp, zfs_check_snap_cb, dd); zfs_close(zhp); return (rv); } /* * Destroys all snapshots with the given name in zhp & descendants. */ int zfs_destroy_snaps(zfs_handle_t *zhp, char *snapname, boolean_t defer) { int ret; struct destroydata dd = { 0 }; dd.snapname = snapname; dd.nvl = fnvlist_alloc(); (void) zfs_check_snap_cb(zfs_handle_dup(zhp), &dd); if (nvlist_empty(dd.nvl)) { ret = zfs_standard_error_fmt(zhp->zfs_hdl, ENOENT, dgettext(TEXT_DOMAIN, "cannot destroy '%s@%s'"), zhp->zfs_name, snapname); } else { ret = zfs_destroy_snaps_nvl(zhp->zfs_hdl, dd.nvl, defer); } fnvlist_free(dd.nvl); return (ret); } /* * Destroys all the snapshots named in the nvlist. */ int zfs_destroy_snaps_nvl(libzfs_handle_t *hdl, nvlist_t *snaps, boolean_t defer) { nvlist_t *errlist = NULL; nvpair_t *pair; int ret = zfs_destroy_snaps_nvl_os(hdl, snaps); if (ret != 0) return (ret); ret = lzc_destroy_snaps(snaps, defer, &errlist); if (ret == 0) { nvlist_free(errlist); return (0); } if (nvlist_empty(errlist)) { char errbuf[ERRBUFLEN]; (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot destroy snapshots")); ret = zfs_standard_error(hdl, ret, errbuf); } for (pair = nvlist_next_nvpair(errlist, NULL); pair != NULL; pair = nvlist_next_nvpair(errlist, pair)) { char errbuf[ERRBUFLEN]; (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot destroy snapshot %s"), nvpair_name(pair)); switch (fnvpair_value_int32(pair)) { case EEXIST: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "snapshot is cloned")); ret = zfs_error(hdl, EZFS_EXISTS, errbuf); break; default: ret = zfs_standard_error(hdl, errno, errbuf); break; } } nvlist_free(errlist); return (ret); } /* * Clones the given dataset. The target must be of the same type as the source. */ int zfs_clone(zfs_handle_t *zhp, const char *target, nvlist_t *props) { char parent[ZFS_MAX_DATASET_NAME_LEN]; int ret; char errbuf[ERRBUFLEN]; libzfs_handle_t *hdl = zhp->zfs_hdl; uint64_t zoned; assert(zhp->zfs_type == ZFS_TYPE_SNAPSHOT); (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot create '%s'"), target); /* validate the target/clone name */ if (!zfs_validate_name(hdl, target, ZFS_TYPE_FILESYSTEM, B_TRUE)) return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf)); /* validate parents exist */ if (check_parents(hdl, target, &zoned, B_FALSE, NULL) != 0) return (-1); (void) parent_name(target, parent, sizeof (parent)); /* do the clone */ if (props) { zfs_type_t type = ZFS_TYPE_FILESYSTEM; if (ZFS_IS_VOLUME(zhp)) type = ZFS_TYPE_VOLUME; if ((props = zfs_valid_proplist(hdl, type, props, zoned, zhp, zhp->zpool_hdl, B_TRUE, errbuf)) == NULL) return (-1); if (zfs_fix_auto_resv(zhp, props) == -1) { nvlist_free(props); return (-1); } } if (zfs_crypto_clone_check(hdl, zhp, parent, props) != 0) { nvlist_free(props); return (zfs_error(hdl, EZFS_CRYPTOFAILED, errbuf)); } ret = lzc_clone(target, zhp->zfs_name, props); nvlist_free(props); if (ret != 0) { switch (errno) { case ENOENT: /* * The parent doesn't exist. We should have caught this * above, but there may a race condition that has since * destroyed the parent. * * At this point, we don't know whether it's the source * that doesn't exist anymore, or whether the target * dataset doesn't exist. */ zfs_error_aux(zhp->zfs_hdl, dgettext(TEXT_DOMAIN, "no such parent '%s'"), parent); return (zfs_error(zhp->zfs_hdl, EZFS_NOENT, errbuf)); case EXDEV: zfs_error_aux(zhp->zfs_hdl, dgettext(TEXT_DOMAIN, "source and target pools differ")); return (zfs_error(zhp->zfs_hdl, EZFS_CROSSTARGET, errbuf)); default: return (zfs_standard_error(zhp->zfs_hdl, errno, errbuf)); } } return (ret); } /* * Promotes the given clone fs to be the clone parent. */ int zfs_promote(zfs_handle_t *zhp) { libzfs_handle_t *hdl = zhp->zfs_hdl; char snapname[ZFS_MAX_DATASET_NAME_LEN]; int ret; char errbuf[ERRBUFLEN]; (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot promote '%s'"), zhp->zfs_name); if (zhp->zfs_type == ZFS_TYPE_SNAPSHOT) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "snapshots can not be promoted")); return (zfs_error(hdl, EZFS_BADTYPE, errbuf)); } if (zhp->zfs_dmustats.dds_origin[0] == '\0') { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "not a cloned filesystem")); return (zfs_error(hdl, EZFS_BADTYPE, errbuf)); } if (!zfs_validate_name(hdl, zhp->zfs_name, zhp->zfs_type, B_TRUE)) return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf)); ret = lzc_promote(zhp->zfs_name, snapname, sizeof (snapname)); if (ret != 0) { switch (ret) { case EACCES: /* * Promoting encrypted dataset outside its * encryption root. */ zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "cannot promote dataset outside its " "encryption root")); return (zfs_error(hdl, EZFS_EXISTS, errbuf)); case EEXIST: /* There is a conflicting snapshot name. */ zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "conflicting snapshot '%s' from parent '%s'"), snapname, zhp->zfs_dmustats.dds_origin); return (zfs_error(hdl, EZFS_EXISTS, errbuf)); default: return (zfs_standard_error(hdl, ret, errbuf)); } } return (ret); } typedef struct snapdata { nvlist_t *sd_nvl; const char *sd_snapname; } snapdata_t; static int zfs_snapshot_cb(zfs_handle_t *zhp, void *arg) { snapdata_t *sd = arg; char name[ZFS_MAX_DATASET_NAME_LEN]; int rv = 0; if (zfs_prop_get_int(zhp, ZFS_PROP_INCONSISTENT) == 0) { if (snprintf(name, sizeof (name), "%s@%s", zfs_get_name(zhp), sd->sd_snapname) >= sizeof (name)) return (EINVAL); fnvlist_add_boolean(sd->sd_nvl, name); rv = zfs_iter_filesystems(zhp, zfs_snapshot_cb, sd); } zfs_close(zhp); return (rv); } /* * Creates snapshots. The keys in the snaps nvlist are the snapshots to be * created. */ int zfs_snapshot_nvl(libzfs_handle_t *hdl, nvlist_t *snaps, nvlist_t *props) { int ret; char errbuf[ERRBUFLEN]; nvpair_t *elem; nvlist_t *errors; zpool_handle_t *zpool_hdl; char pool[ZFS_MAX_DATASET_NAME_LEN]; (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot create snapshots ")); elem = NULL; while ((elem = nvlist_next_nvpair(snaps, elem)) != NULL) { const char *snapname = nvpair_name(elem); /* validate the target name */ if (!zfs_validate_name(hdl, snapname, ZFS_TYPE_SNAPSHOT, B_TRUE)) { (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot create snapshot '%s'"), snapname); return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf)); } } /* * get pool handle for prop validation. assumes all snaps are in the * same pool, as does lzc_snapshot (below). */ elem = nvlist_next_nvpair(snaps, NULL); + if (elem == NULL) + return (-1); (void) strlcpy(pool, nvpair_name(elem), sizeof (pool)); pool[strcspn(pool, "/@")] = '\0'; zpool_hdl = zpool_open(hdl, pool); if (zpool_hdl == NULL) return (-1); if (props != NULL && (props = zfs_valid_proplist(hdl, ZFS_TYPE_SNAPSHOT, props, B_FALSE, NULL, zpool_hdl, B_FALSE, errbuf)) == NULL) { zpool_close(zpool_hdl); return (-1); } zpool_close(zpool_hdl); ret = lzc_snapshot(snaps, props, &errors); if (ret != 0) { boolean_t printed = B_FALSE; for (elem = nvlist_next_nvpair(errors, NULL); elem != NULL; elem = nvlist_next_nvpair(errors, elem)) { (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot create snapshot '%s'"), nvpair_name(elem)); (void) zfs_standard_error(hdl, fnvpair_value_int32(elem), errbuf); printed = B_TRUE; } if (!printed) { switch (ret) { case EXDEV: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "multiple snapshots of same " "fs not allowed")); (void) zfs_error(hdl, EZFS_EXISTS, errbuf); break; default: (void) zfs_standard_error(hdl, ret, errbuf); } } } nvlist_free(props); nvlist_free(errors); return (ret); } int zfs_snapshot(libzfs_handle_t *hdl, const char *path, boolean_t recursive, nvlist_t *props) { int ret; snapdata_t sd = { 0 }; char fsname[ZFS_MAX_DATASET_NAME_LEN]; char *cp; zfs_handle_t *zhp; char errbuf[ERRBUFLEN]; (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot snapshot %s"), path); if (!zfs_validate_name(hdl, path, ZFS_TYPE_SNAPSHOT, B_TRUE)) return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf)); (void) strlcpy(fsname, path, sizeof (fsname)); cp = strchr(fsname, '@'); *cp = '\0'; sd.sd_snapname = cp + 1; if ((zhp = zfs_open(hdl, fsname, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME)) == NULL) { return (-1); } sd.sd_nvl = fnvlist_alloc(); if (recursive) { (void) zfs_snapshot_cb(zfs_handle_dup(zhp), &sd); } else { fnvlist_add_boolean(sd.sd_nvl, path); } ret = zfs_snapshot_nvl(hdl, sd.sd_nvl, props); fnvlist_free(sd.sd_nvl); zfs_close(zhp); return (ret); } /* * Destroy any more recent snapshots. We invoke this callback on any dependents * of the snapshot first. If the 'cb_dependent' member is non-zero, then this * is a dependent and we should just destroy it without checking the transaction * group. */ typedef struct rollback_data { const char *cb_target; /* the snapshot */ uint64_t cb_create; /* creation time reference */ boolean_t cb_error; boolean_t cb_force; } rollback_data_t; static int rollback_destroy_dependent(zfs_handle_t *zhp, void *data) { rollback_data_t *cbp = data; prop_changelist_t *clp; /* We must destroy this clone; first unmount it */ clp = changelist_gather(zhp, ZFS_PROP_NAME, 0, cbp->cb_force ? MS_FORCE: 0); if (clp == NULL || changelist_prefix(clp) != 0) { cbp->cb_error = B_TRUE; zfs_close(zhp); return (0); } if (zfs_destroy(zhp, B_FALSE) != 0) cbp->cb_error = B_TRUE; else changelist_remove(clp, zhp->zfs_name); (void) changelist_postfix(clp); changelist_free(clp); zfs_close(zhp); return (0); } static int rollback_destroy(zfs_handle_t *zhp, void *data) { rollback_data_t *cbp = data; if (zfs_prop_get_int(zhp, ZFS_PROP_CREATETXG) > cbp->cb_create) { cbp->cb_error |= zfs_iter_dependents(zhp, B_FALSE, rollback_destroy_dependent, cbp); cbp->cb_error |= zfs_destroy(zhp, B_FALSE); } zfs_close(zhp); return (0); } /* * Given a dataset, rollback to a specific snapshot, discarding any * data changes since then and making it the active dataset. * * Any snapshots and bookmarks more recent than the target are * destroyed, along with their dependents (i.e. clones). */ int zfs_rollback(zfs_handle_t *zhp, zfs_handle_t *snap, boolean_t force) { rollback_data_t cb = { 0 }; int err; boolean_t restore_resv = 0; uint64_t old_volsize = 0, new_volsize; zfs_prop_t resv_prop = { 0 }; uint64_t min_txg = 0; assert(zhp->zfs_type == ZFS_TYPE_FILESYSTEM || zhp->zfs_type == ZFS_TYPE_VOLUME); /* * Destroy all recent snapshots and their dependents. */ cb.cb_force = force; cb.cb_target = snap->zfs_name; cb.cb_create = zfs_prop_get_int(snap, ZFS_PROP_CREATETXG); if (cb.cb_create > 0) min_txg = cb.cb_create; (void) zfs_iter_snapshots(zhp, B_FALSE, rollback_destroy, &cb, min_txg, 0); (void) zfs_iter_bookmarks(zhp, rollback_destroy, &cb); if (cb.cb_error) return (-1); /* * Now that we have verified that the snapshot is the latest, * rollback to the given snapshot. */ if (zhp->zfs_type == ZFS_TYPE_VOLUME) { if (zfs_which_resv_prop(zhp, &resv_prop) < 0) return (-1); old_volsize = zfs_prop_get_int(zhp, ZFS_PROP_VOLSIZE); restore_resv = (old_volsize == zfs_prop_get_int(zhp, resv_prop)); } /* * Pass both the filesystem and the wanted snapshot names, * we would get an error back if the snapshot is destroyed or * a new snapshot is created before this request is processed. */ err = lzc_rollback_to(zhp->zfs_name, snap->zfs_name); if (err != 0) { char errbuf[ERRBUFLEN]; (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot rollback '%s'"), zhp->zfs_name); switch (err) { case EEXIST: zfs_error_aux(zhp->zfs_hdl, dgettext(TEXT_DOMAIN, "there is a snapshot or bookmark more recent " "than '%s'"), snap->zfs_name); (void) zfs_error(zhp->zfs_hdl, EZFS_EXISTS, errbuf); break; case ESRCH: zfs_error_aux(zhp->zfs_hdl, dgettext(TEXT_DOMAIN, "'%s' is not found among snapshots of '%s'"), snap->zfs_name, zhp->zfs_name); (void) zfs_error(zhp->zfs_hdl, EZFS_NOENT, errbuf); break; case EINVAL: (void) zfs_error(zhp->zfs_hdl, EZFS_BADTYPE, errbuf); break; default: (void) zfs_standard_error(zhp->zfs_hdl, err, errbuf); } return (err); } /* * For volumes, if the pre-rollback volsize matched the pre- * rollback reservation and the volsize has changed then set * the reservation property to the post-rollback volsize. * Make a new handle since the rollback closed the dataset. */ if ((zhp->zfs_type == ZFS_TYPE_VOLUME) && (zhp = make_dataset_handle(zhp->zfs_hdl, zhp->zfs_name))) { if (restore_resv) { new_volsize = zfs_prop_get_int(zhp, ZFS_PROP_VOLSIZE); if (old_volsize != new_volsize) err = zfs_prop_set_int(zhp, resv_prop, new_volsize); } zfs_close(zhp); } return (err); } /* * Renames the given dataset. */ int zfs_rename(zfs_handle_t *zhp, const char *target, renameflags_t flags) { int ret = 0; zfs_cmd_t zc = {"\0"}; char *delim; prop_changelist_t *cl = NULL; char parent[ZFS_MAX_DATASET_NAME_LEN]; char property[ZFS_MAXPROPLEN]; libzfs_handle_t *hdl = zhp->zfs_hdl; char errbuf[ERRBUFLEN]; /* if we have the same exact name, just return success */ if (strcmp(zhp->zfs_name, target) == 0) return (0); (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot rename to '%s'"), target); /* make sure source name is valid */ if (!zfs_validate_name(hdl, zhp->zfs_name, zhp->zfs_type, B_TRUE)) return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf)); /* * Make sure the target name is valid */ if (zhp->zfs_type == ZFS_TYPE_SNAPSHOT) { if ((strchr(target, '@') == NULL) || *target == '@') { /* * Snapshot target name is abbreviated, * reconstruct full dataset name */ (void) strlcpy(parent, zhp->zfs_name, sizeof (parent)); delim = strchr(parent, '@'); if (strchr(target, '@') == NULL) *(++delim) = '\0'; else *delim = '\0'; (void) strlcat(parent, target, sizeof (parent)); target = parent; } else { /* * Make sure we're renaming within the same dataset. */ delim = strchr(target, '@'); if (strncmp(zhp->zfs_name, target, delim - target) != 0 || zhp->zfs_name[delim - target] != '@') { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "snapshots must be part of same " "dataset")); return (zfs_error(hdl, EZFS_CROSSTARGET, errbuf)); } } if (!zfs_validate_name(hdl, target, zhp->zfs_type, B_TRUE)) return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf)); } else { if (flags.recursive) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "recursive rename must be a snapshot")); return (zfs_error(hdl, EZFS_BADTYPE, errbuf)); } if (!zfs_validate_name(hdl, target, zhp->zfs_type, B_TRUE)) return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf)); /* validate parents */ if (check_parents(hdl, target, NULL, B_FALSE, NULL) != 0) return (-1); /* make sure we're in the same pool */ verify((delim = strchr(target, '/')) != NULL); if (strncmp(zhp->zfs_name, target, delim - target) != 0 || zhp->zfs_name[delim - target] != '/') { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "datasets must be within same pool")); return (zfs_error(hdl, EZFS_CROSSTARGET, errbuf)); } /* new name cannot be a child of the current dataset name */ if (is_descendant(zhp->zfs_name, target)) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "New dataset name cannot be a descendant of " "current dataset name")); return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf)); } } (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot rename '%s'"), zhp->zfs_name); if (getzoneid() == GLOBAL_ZONEID && zfs_prop_get_int(zhp, ZFS_PROP_ZONED)) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "dataset is used in a non-global zone")); return (zfs_error(hdl, EZFS_ZONED, errbuf)); } /* * Avoid unmounting file systems with mountpoint property set to * 'legacy' or 'none' even if -u option is not given. */ if (zhp->zfs_type == ZFS_TYPE_FILESYSTEM && !flags.recursive && !flags.nounmount && zfs_prop_get(zhp, ZFS_PROP_MOUNTPOINT, property, sizeof (property), NULL, NULL, 0, B_FALSE) == 0 && (strcmp(property, "legacy") == 0 || strcmp(property, "none") == 0)) { flags.nounmount = B_TRUE; } if (flags.recursive) { char *parentname = zfs_strdup(zhp->zfs_hdl, zhp->zfs_name); delim = strchr(parentname, '@'); *delim = '\0'; zfs_handle_t *zhrp = zfs_open(zhp->zfs_hdl, parentname, ZFS_TYPE_DATASET); free(parentname); if (zhrp == NULL) { ret = -1; goto error; } zfs_close(zhrp); } else if (zhp->zfs_type != ZFS_TYPE_SNAPSHOT) { if ((cl = changelist_gather(zhp, ZFS_PROP_NAME, flags.nounmount ? CL_GATHER_DONT_UNMOUNT : CL_GATHER_ITER_MOUNTED, flags.forceunmount ? MS_FORCE : 0)) == NULL) return (-1); if (changelist_haszonedchild(cl)) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "child dataset with inherited mountpoint is used " "in a non-global zone")); (void) zfs_error(hdl, EZFS_ZONED, errbuf); ret = -1; goto error; } if ((ret = changelist_prefix(cl)) != 0) goto error; } if (ZFS_IS_VOLUME(zhp)) zc.zc_objset_type = DMU_OST_ZVOL; else zc.zc_objset_type = DMU_OST_ZFS; (void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name)); (void) strlcpy(zc.zc_value, target, sizeof (zc.zc_value)); zc.zc_cookie = !!flags.recursive; zc.zc_cookie |= (!!flags.nounmount) << 1; if ((ret = zfs_ioctl(zhp->zfs_hdl, ZFS_IOC_RENAME, &zc)) != 0) { /* * if it was recursive, the one that actually failed will * be in zc.zc_name */ (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot rename '%s'"), zc.zc_name); if (flags.recursive && errno == EEXIST) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "a child dataset already has a snapshot " "with the new name")); (void) zfs_error(hdl, EZFS_EXISTS, errbuf); } else if (errno == EACCES) { zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "cannot move encrypted child outside of " "its encryption root")); (void) zfs_error(hdl, EZFS_CRYPTOFAILED, errbuf); } else { (void) zfs_standard_error(zhp->zfs_hdl, errno, errbuf); } /* * On failure, we still want to remount any filesystems that * were previously mounted, so we don't alter the system state. */ if (cl != NULL) (void) changelist_postfix(cl); } else { if (cl != NULL) { changelist_rename(cl, zfs_get_name(zhp), target); ret = changelist_postfix(cl); } } error: if (cl != NULL) { changelist_free(cl); } return (ret); } nvlist_t * zfs_get_all_props(zfs_handle_t *zhp) { return (zhp->zfs_props); } nvlist_t * zfs_get_recvd_props(zfs_handle_t *zhp) { if (zhp->zfs_recvd_props == NULL) if (get_recvd_props_ioctl(zhp) != 0) return (NULL); return (zhp->zfs_recvd_props); } nvlist_t * zfs_get_user_props(zfs_handle_t *zhp) { return (zhp->zfs_user_props); } /* * This function is used by 'zfs list' to determine the exact set of columns to * display, and their maximum widths. This does two main things: * * - If this is a list of all properties, then expand the list to include * all native properties, and set a flag so that for each dataset we look * for new unique user properties and add them to the list. * * - For non fixed-width properties, keep track of the maximum width seen * so that we can size the column appropriately. If the user has * requested received property values, we also need to compute the width * of the RECEIVED column. */ int zfs_expand_proplist(zfs_handle_t *zhp, zprop_list_t **plp, boolean_t received, boolean_t literal) { libzfs_handle_t *hdl = zhp->zfs_hdl; zprop_list_t *entry; zprop_list_t **last, **start; nvlist_t *userprops, *propval; nvpair_t *elem; char *strval; char buf[ZFS_MAXPROPLEN]; if (zprop_expand_list(hdl, plp, ZFS_TYPE_DATASET) != 0) return (-1); userprops = zfs_get_user_props(zhp); entry = *plp; if (entry->pl_all && nvlist_next_nvpair(userprops, NULL) != NULL) { /* * Go through and add any user properties as necessary. We * start by incrementing our list pointer to the first * non-native property. */ start = plp; while (*start != NULL) { if ((*start)->pl_prop == ZPROP_USERPROP) break; start = &(*start)->pl_next; } elem = NULL; while ((elem = nvlist_next_nvpair(userprops, elem)) != NULL) { /* * See if we've already found this property in our list. */ for (last = start; *last != NULL; last = &(*last)->pl_next) { if (strcmp((*last)->pl_user_prop, nvpair_name(elem)) == 0) break; } if (*last == NULL) { entry = zfs_alloc(hdl, sizeof (zprop_list_t)); entry->pl_user_prop = zfs_strdup(hdl, nvpair_name(elem)); entry->pl_prop = ZPROP_USERPROP; entry->pl_width = strlen(nvpair_name(elem)); entry->pl_all = B_TRUE; *last = entry; } } } /* * Now go through and check the width of any non-fixed columns */ for (entry = *plp; entry != NULL; entry = entry->pl_next) { if (entry->pl_fixed && !literal) continue; if (entry->pl_prop != ZPROP_USERPROP) { if (zfs_prop_get(zhp, entry->pl_prop, buf, sizeof (buf), NULL, NULL, 0, literal) == 0) { if (strlen(buf) > entry->pl_width) entry->pl_width = strlen(buf); } if (received && zfs_prop_get_recvd(zhp, zfs_prop_to_name(entry->pl_prop), buf, sizeof (buf), literal) == 0) if (strlen(buf) > entry->pl_recvd_width) entry->pl_recvd_width = strlen(buf); } else { if (nvlist_lookup_nvlist(userprops, entry->pl_user_prop, &propval) == 0) { strval = fnvlist_lookup_string(propval, ZPROP_VALUE); if (strlen(strval) > entry->pl_width) entry->pl_width = strlen(strval); } if (received && zfs_prop_get_recvd(zhp, entry->pl_user_prop, buf, sizeof (buf), literal) == 0) if (strlen(buf) > entry->pl_recvd_width) entry->pl_recvd_width = strlen(buf); } } return (0); } void zfs_prune_proplist(zfs_handle_t *zhp, uint8_t *props) { nvpair_t *curr; nvpair_t *next; /* * Keep a reference to the props-table against which we prune the * properties. */ zhp->zfs_props_table = props; curr = nvlist_next_nvpair(zhp->zfs_props, NULL); while (curr) { zfs_prop_t zfs_prop = zfs_name_to_prop(nvpair_name(curr)); next = nvlist_next_nvpair(zhp->zfs_props, curr); /* * User properties will result in ZPROP_USERPROP (an alias * for ZPROP_INVAL), and since we * only know how to prune standard ZFS properties, we always * leave these in the list. This can also happen if we * encounter an unknown DSL property (when running older * software, for example). */ if (zfs_prop != ZPROP_USERPROP && props[zfs_prop] == B_FALSE) (void) nvlist_remove(zhp->zfs_props, nvpair_name(curr), nvpair_type(curr)); curr = next; } } static int zfs_smb_acl_mgmt(libzfs_handle_t *hdl, char *dataset, char *path, zfs_smb_acl_op_t cmd, char *resource1, char *resource2) { zfs_cmd_t zc = {"\0"}; nvlist_t *nvlist = NULL; int error; (void) strlcpy(zc.zc_name, dataset, sizeof (zc.zc_name)); (void) strlcpy(zc.zc_value, path, sizeof (zc.zc_value)); zc.zc_cookie = (uint64_t)cmd; if (cmd == ZFS_SMB_ACL_RENAME) { if (nvlist_alloc(&nvlist, NV_UNIQUE_NAME, 0) != 0) { (void) no_memory(hdl); return (0); } } switch (cmd) { case ZFS_SMB_ACL_ADD: case ZFS_SMB_ACL_REMOVE: (void) strlcpy(zc.zc_string, resource1, sizeof (zc.zc_string)); break; case ZFS_SMB_ACL_RENAME: if (nvlist_add_string(nvlist, ZFS_SMB_ACL_SRC, resource1) != 0) { (void) no_memory(hdl); return (-1); } if (nvlist_add_string(nvlist, ZFS_SMB_ACL_TARGET, resource2) != 0) { (void) no_memory(hdl); return (-1); } zcmd_write_src_nvlist(hdl, &zc, nvlist); break; case ZFS_SMB_ACL_PURGE: break; default: return (-1); } error = ioctl(hdl->libzfs_fd, ZFS_IOC_SMB_ACL, &zc); nvlist_free(nvlist); return (error); } int zfs_smb_acl_add(libzfs_handle_t *hdl, char *dataset, char *path, char *resource) { return (zfs_smb_acl_mgmt(hdl, dataset, path, ZFS_SMB_ACL_ADD, resource, NULL)); } int zfs_smb_acl_remove(libzfs_handle_t *hdl, char *dataset, char *path, char *resource) { return (zfs_smb_acl_mgmt(hdl, dataset, path, ZFS_SMB_ACL_REMOVE, resource, NULL)); } int zfs_smb_acl_purge(libzfs_handle_t *hdl, char *dataset, char *path) { return (zfs_smb_acl_mgmt(hdl, dataset, path, ZFS_SMB_ACL_PURGE, NULL, NULL)); } int zfs_smb_acl_rename(libzfs_handle_t *hdl, char *dataset, char *path, char *oldname, char *newname) { return (zfs_smb_acl_mgmt(hdl, dataset, path, ZFS_SMB_ACL_RENAME, oldname, newname)); } int zfs_userspace(zfs_handle_t *zhp, zfs_userquota_prop_t type, zfs_userspace_cb_t func, void *arg) { zfs_cmd_t zc = {"\0"}; zfs_useracct_t buf[100]; libzfs_handle_t *hdl = zhp->zfs_hdl; int ret; (void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name)); zc.zc_objset_type = type; zc.zc_nvlist_dst = (uintptr_t)buf; for (;;) { zfs_useracct_t *zua = buf; zc.zc_nvlist_dst_size = sizeof (buf); if (zfs_ioctl(hdl, ZFS_IOC_USERSPACE_MANY, &zc) != 0) { if ((errno == ENOTSUP && (type == ZFS_PROP_USEROBJUSED || type == ZFS_PROP_GROUPOBJUSED || type == ZFS_PROP_USEROBJQUOTA || type == ZFS_PROP_GROUPOBJQUOTA || type == ZFS_PROP_PROJECTOBJUSED || type == ZFS_PROP_PROJECTOBJQUOTA || type == ZFS_PROP_PROJECTUSED || type == ZFS_PROP_PROJECTQUOTA))) break; return (zfs_standard_error_fmt(hdl, errno, dgettext(TEXT_DOMAIN, "cannot get used/quota for %s"), zc.zc_name)); } if (zc.zc_nvlist_dst_size == 0) break; while (zc.zc_nvlist_dst_size > 0) { if ((ret = func(arg, zua->zu_domain, zua->zu_rid, zua->zu_space)) != 0) return (ret); zua++; zc.zc_nvlist_dst_size -= sizeof (zfs_useracct_t); } } return (0); } struct holdarg { nvlist_t *nvl; const char *snapname; const char *tag; boolean_t recursive; int error; }; static int zfs_hold_one(zfs_handle_t *zhp, void *arg) { struct holdarg *ha = arg; char name[ZFS_MAX_DATASET_NAME_LEN]; int rv = 0; if (snprintf(name, sizeof (name), "%s@%s", zhp->zfs_name, ha->snapname) >= sizeof (name)) return (EINVAL); if (lzc_exists(name)) fnvlist_add_string(ha->nvl, name, ha->tag); if (ha->recursive) rv = zfs_iter_filesystems(zhp, zfs_hold_one, ha); zfs_close(zhp); return (rv); } int zfs_hold(zfs_handle_t *zhp, const char *snapname, const char *tag, boolean_t recursive, int cleanup_fd) { int ret; struct holdarg ha; ha.nvl = fnvlist_alloc(); ha.snapname = snapname; ha.tag = tag; ha.recursive = recursive; (void) zfs_hold_one(zfs_handle_dup(zhp), &ha); if (nvlist_empty(ha.nvl)) { char errbuf[ERRBUFLEN]; fnvlist_free(ha.nvl); ret = ENOENT; (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot hold snapshot '%s@%s'"), zhp->zfs_name, snapname); (void) zfs_standard_error(zhp->zfs_hdl, ret, errbuf); return (ret); } ret = zfs_hold_nvl(zhp, cleanup_fd, ha.nvl); fnvlist_free(ha.nvl); return (ret); } int zfs_hold_nvl(zfs_handle_t *zhp, int cleanup_fd, nvlist_t *holds) { int ret; nvlist_t *errors; libzfs_handle_t *hdl = zhp->zfs_hdl; char errbuf[ERRBUFLEN]; nvpair_t *elem; errors = NULL; ret = lzc_hold(holds, cleanup_fd, &errors); if (ret == 0) { /* There may be errors even in the success case. */ fnvlist_free(errors); return (0); } if (nvlist_empty(errors)) { /* no hold-specific errors */ (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot hold")); switch (ret) { case ENOTSUP: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "pool must be upgraded")); (void) zfs_error(hdl, EZFS_BADVERSION, errbuf); break; case EINVAL: (void) zfs_error(hdl, EZFS_BADTYPE, errbuf); break; default: (void) zfs_standard_error(hdl, ret, errbuf); } } for (elem = nvlist_next_nvpair(errors, NULL); elem != NULL; elem = nvlist_next_nvpair(errors, elem)) { (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot hold snapshot '%s'"), nvpair_name(elem)); switch (fnvpair_value_int32(elem)) { case E2BIG: /* * Temporary tags wind up having the ds object id * prepended. So even if we passed the length check * above, it's still possible for the tag to wind * up being slightly too long. */ (void) zfs_error(hdl, EZFS_TAGTOOLONG, errbuf); break; case EINVAL: (void) zfs_error(hdl, EZFS_BADTYPE, errbuf); break; case EEXIST: (void) zfs_error(hdl, EZFS_REFTAG_HOLD, errbuf); break; default: (void) zfs_standard_error(hdl, fnvpair_value_int32(elem), errbuf); } } fnvlist_free(errors); return (ret); } static int zfs_release_one(zfs_handle_t *zhp, void *arg) { struct holdarg *ha = arg; char name[ZFS_MAX_DATASET_NAME_LEN]; int rv = 0; nvlist_t *existing_holds; if (snprintf(name, sizeof (name), "%s@%s", zhp->zfs_name, ha->snapname) >= sizeof (name)) { ha->error = EINVAL; rv = EINVAL; } if (lzc_get_holds(name, &existing_holds) != 0) { ha->error = ENOENT; } else if (!nvlist_exists(existing_holds, ha->tag)) { ha->error = ESRCH; } else { nvlist_t *torelease = fnvlist_alloc(); fnvlist_add_boolean(torelease, ha->tag); fnvlist_add_nvlist(ha->nvl, name, torelease); fnvlist_free(torelease); } if (ha->recursive) rv = zfs_iter_filesystems(zhp, zfs_release_one, ha); zfs_close(zhp); return (rv); } int zfs_release(zfs_handle_t *zhp, const char *snapname, const char *tag, boolean_t recursive) { int ret; struct holdarg ha; nvlist_t *errors = NULL; nvpair_t *elem; libzfs_handle_t *hdl = zhp->zfs_hdl; char errbuf[ERRBUFLEN]; ha.nvl = fnvlist_alloc(); ha.snapname = snapname; ha.tag = tag; ha.recursive = recursive; ha.error = 0; (void) zfs_release_one(zfs_handle_dup(zhp), &ha); if (nvlist_empty(ha.nvl)) { fnvlist_free(ha.nvl); ret = ha.error; (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot release hold from snapshot '%s@%s'"), zhp->zfs_name, snapname); if (ret == ESRCH) { (void) zfs_error(hdl, EZFS_REFTAG_RELE, errbuf); } else { (void) zfs_standard_error(hdl, ret, errbuf); } return (ret); } ret = lzc_release(ha.nvl, &errors); fnvlist_free(ha.nvl); if (ret == 0) { /* There may be errors even in the success case. */ fnvlist_free(errors); return (0); } if (nvlist_empty(errors)) { /* no hold-specific errors */ (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot release")); switch (errno) { case ENOTSUP: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "pool must be upgraded")); (void) zfs_error(hdl, EZFS_BADVERSION, errbuf); break; default: (void) zfs_standard_error(hdl, errno, errbuf); } } for (elem = nvlist_next_nvpair(errors, NULL); elem != NULL; elem = nvlist_next_nvpair(errors, elem)) { (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot release hold from snapshot '%s'"), nvpair_name(elem)); switch (fnvpair_value_int32(elem)) { case ESRCH: (void) zfs_error(hdl, EZFS_REFTAG_RELE, errbuf); break; case EINVAL: (void) zfs_error(hdl, EZFS_BADTYPE, errbuf); break; default: (void) zfs_standard_error(hdl, fnvpair_value_int32(elem), errbuf); } } fnvlist_free(errors); return (ret); } int zfs_get_fsacl(zfs_handle_t *zhp, nvlist_t **nvl) { zfs_cmd_t zc = {"\0"}; libzfs_handle_t *hdl = zhp->zfs_hdl; int nvsz = 2048; void *nvbuf; int err = 0; char errbuf[ERRBUFLEN]; assert(zhp->zfs_type == ZFS_TYPE_VOLUME || zhp->zfs_type == ZFS_TYPE_FILESYSTEM); tryagain: nvbuf = malloc(nvsz); if (nvbuf == NULL) { err = (zfs_error(hdl, EZFS_NOMEM, strerror(errno))); goto out; } zc.zc_nvlist_dst_size = nvsz; zc.zc_nvlist_dst = (uintptr_t)nvbuf; (void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name)); if (zfs_ioctl(hdl, ZFS_IOC_GET_FSACL, &zc) != 0) { (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot get permissions on '%s'"), zc.zc_name); switch (errno) { case ENOMEM: free(nvbuf); nvsz = zc.zc_nvlist_dst_size; goto tryagain; case ENOTSUP: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "pool must be upgraded")); err = zfs_error(hdl, EZFS_BADVERSION, errbuf); break; case EINVAL: err = zfs_error(hdl, EZFS_BADTYPE, errbuf); break; case ENOENT: err = zfs_error(hdl, EZFS_NOENT, errbuf); break; default: err = zfs_standard_error(hdl, errno, errbuf); break; } } else { /* success */ int rc = nvlist_unpack(nvbuf, zc.zc_nvlist_dst_size, nvl, 0); if (rc) { err = zfs_standard_error_fmt(hdl, rc, dgettext( TEXT_DOMAIN, "cannot get permissions on '%s'"), zc.zc_name); } } free(nvbuf); out: return (err); } int zfs_set_fsacl(zfs_handle_t *zhp, boolean_t un, nvlist_t *nvl) { zfs_cmd_t zc = {"\0"}; libzfs_handle_t *hdl = zhp->zfs_hdl; char *nvbuf; char errbuf[ERRBUFLEN]; size_t nvsz; int err; assert(zhp->zfs_type == ZFS_TYPE_VOLUME || zhp->zfs_type == ZFS_TYPE_FILESYSTEM); err = nvlist_size(nvl, &nvsz, NV_ENCODE_NATIVE); assert(err == 0); nvbuf = malloc(nvsz); err = nvlist_pack(nvl, &nvbuf, &nvsz, NV_ENCODE_NATIVE, 0); assert(err == 0); zc.zc_nvlist_src_size = nvsz; zc.zc_nvlist_src = (uintptr_t)nvbuf; zc.zc_perm_action = un; (void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name)); if (zfs_ioctl(hdl, ZFS_IOC_SET_FSACL, &zc) != 0) { (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot set permissions on '%s'"), zc.zc_name); switch (errno) { case ENOTSUP: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "pool must be upgraded")); err = zfs_error(hdl, EZFS_BADVERSION, errbuf); break; case EINVAL: err = zfs_error(hdl, EZFS_BADTYPE, errbuf); break; case ENOENT: err = zfs_error(hdl, EZFS_NOENT, errbuf); break; default: err = zfs_standard_error(hdl, errno, errbuf); break; } } free(nvbuf); return (err); } int zfs_get_holds(zfs_handle_t *zhp, nvlist_t **nvl) { int err; char errbuf[ERRBUFLEN]; err = lzc_get_holds(zhp->zfs_name, nvl); if (err != 0) { libzfs_handle_t *hdl = zhp->zfs_hdl; (void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN, "cannot get holds for '%s'"), zhp->zfs_name); switch (err) { case ENOTSUP: zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "pool must be upgraded")); err = zfs_error(hdl, EZFS_BADVERSION, errbuf); break; case EINVAL: err = zfs_error(hdl, EZFS_BADTYPE, errbuf); break; case ENOENT: err = zfs_error(hdl, EZFS_NOENT, errbuf); break; default: err = zfs_standard_error(hdl, errno, errbuf); break; } } return (err); } /* * The theory of raidz space accounting * * The "referenced" property of RAIDZ vdevs is scaled such that a 128KB block * will "reference" 128KB, even though it allocates more than that, to store the * parity information (and perhaps skip sectors). This concept of the * "referenced" (and other DMU space accounting) being lower than the allocated * space by a constant factor is called "raidz deflation." * * As mentioned above, the constant factor for raidz deflation assumes a 128KB * block size. However, zvols typically have a much smaller block size (default * 8KB). These smaller blocks may require proportionally much more parity * information (and perhaps skip sectors). In this case, the change to the * "referenced" property may be much more than the logical block size. * * Suppose a raidz vdev has 5 disks with ashift=12. A 128k block may be written * as follows. * * +-------+-------+-------+-------+-------+ * | disk1 | disk2 | disk3 | disk4 | disk5 | * +-------+-------+-------+-------+-------+ * | P0 | D0 | D8 | D16 | D24 | * | P1 | D1 | D9 | D17 | D25 | * | P2 | D2 | D10 | D18 | D26 | * | P3 | D3 | D11 | D19 | D27 | * | P4 | D4 | D12 | D20 | D28 | * | P5 | D5 | D13 | D21 | D29 | * | P6 | D6 | D14 | D22 | D30 | * | P7 | D7 | D15 | D23 | D31 | * +-------+-------+-------+-------+-------+ * * Above, notice that 160k was allocated: 8 x 4k parity sectors + 32 x 4k data * sectors. The dataset's referenced will increase by 128k and the pool's * allocated and free properties will be adjusted by 160k. * * A 4k block written to the same raidz vdev will require two 4k sectors. The * blank cells represent unallocated space. * * +-------+-------+-------+-------+-------+ * | disk1 | disk2 | disk3 | disk4 | disk5 | * +-------+-------+-------+-------+-------+ * | P0 | D0 | | | | * +-------+-------+-------+-------+-------+ * * Above, notice that the 4k block required one sector for parity and another * for data. vdev_raidz_asize() will return 8k and as such the pool's allocated * and free properties will be adjusted by 8k. The dataset will not be charged * 8k. Rather, it will be charged a value that is scaled according to the * overhead of the 128k block on the same vdev. This 8k allocation will be * charged 8k * 128k / 160k. 128k is from SPA_OLD_MAXBLOCKSIZE and 160k is as * calculated in the 128k block example above. * * Every raidz allocation is sized to be a multiple of nparity+1 sectors. That * is, every raidz1 allocation will be a multiple of 2 sectors, raidz2 * allocations are a multiple of 3 sectors, and raidz3 allocations are a * multiple of of 4 sectors. When a block does not fill the required number of * sectors, skip blocks (sectors) are used. * * An 8k block being written to a raidz vdev may be written as follows: * * +-------+-------+-------+-------+-------+ * | disk1 | disk2 | disk3 | disk4 | disk5 | * +-------+-------+-------+-------+-------+ * | P0 | D0 | D1 | S0 | | * +-------+-------+-------+-------+-------+ * * In order to maintain the nparity+1 allocation size, a skip block (S0) was * added. For this 8k block, the pool's allocated and free properties are * adjusted by 16k and the dataset's referenced is increased by 16k * 128k / * 160k. Again, 128k is from SPA_OLD_MAXBLOCKSIZE and 160k is as calculated in * the 128k block example above. * * The situation is slightly different for dRAID since the minimum allocation * size is the full group width. The same 8K block above would be written as * follows in a dRAID group: * * +-------+-------+-------+-------+-------+ * | disk1 | disk2 | disk3 | disk4 | disk5 | * +-------+-------+-------+-------+-------+ * | P0 | D0 | D1 | S0 | S1 | * +-------+-------+-------+-------+-------+ * * Compression may lead to a variety of block sizes being written for the same * volume or file. There is no clear way to reserve just the amount of space * that will be required, so the worst case (no compression) is assumed. * Note that metadata blocks will typically be compressed, so the reservation * size returned by zvol_volsize_to_reservation() will generally be slightly * larger than the maximum that the volume can reference. */ /* * Derived from function of same name in module/zfs/vdev_raidz.c. Returns the * amount of space (in bytes) that will be allocated for the specified block * size. Note that the "referenced" space accounted will be less than this, but * not necessarily equal to "blksize", due to RAIDZ deflation. */ static uint64_t vdev_raidz_asize(uint64_t ndisks, uint64_t nparity, uint64_t ashift, uint64_t blksize) { uint64_t asize, ndata; ASSERT3U(ndisks, >, nparity); ndata = ndisks - nparity; asize = ((blksize - 1) >> ashift) + 1; asize += nparity * ((asize + ndata - 1) / ndata); asize = roundup(asize, nparity + 1) << ashift; return (asize); } /* * Derived from function of same name in module/zfs/vdev_draid.c. Returns the * amount of space (in bytes) that will be allocated for the specified block * size. */ static uint64_t vdev_draid_asize(uint64_t ndisks, uint64_t nparity, uint64_t ashift, uint64_t blksize) { ASSERT3U(ndisks, >, nparity); uint64_t ndata = ndisks - nparity; uint64_t rows = ((blksize - 1) / (ndata << ashift)) + 1; uint64_t asize = (rows * ndisks) << ashift; return (asize); } /* * Determine how much space will be allocated if it lands on the most space- * inefficient top-level vdev. Returns the size in bytes required to store one * copy of the volume data. See theory comment above. */ static uint64_t volsize_from_vdevs(zpool_handle_t *zhp, uint64_t nblocks, uint64_t blksize) { nvlist_t *config, *tree, **vdevs; uint_t nvdevs; uint64_t ret = 0; config = zpool_get_config(zhp, NULL); if (nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, &tree) != 0 || nvlist_lookup_nvlist_array(tree, ZPOOL_CONFIG_CHILDREN, &vdevs, &nvdevs) != 0) { return (nblocks * blksize); } for (int v = 0; v < nvdevs; v++) { char *type; uint64_t nparity, ashift, asize, tsize; uint64_t volsize; if (nvlist_lookup_string(vdevs[v], ZPOOL_CONFIG_TYPE, &type) != 0) continue; if (strcmp(type, VDEV_TYPE_RAIDZ) != 0 && strcmp(type, VDEV_TYPE_DRAID) != 0) continue; if (nvlist_lookup_uint64(vdevs[v], ZPOOL_CONFIG_NPARITY, &nparity) != 0) continue; if (nvlist_lookup_uint64(vdevs[v], ZPOOL_CONFIG_ASHIFT, &ashift) != 0) continue; if (strcmp(type, VDEV_TYPE_RAIDZ) == 0) { nvlist_t **disks; uint_t ndisks; if (nvlist_lookup_nvlist_array(vdevs[v], ZPOOL_CONFIG_CHILDREN, &disks, &ndisks) != 0) continue; /* allocation size for the "typical" 128k block */ tsize = vdev_raidz_asize(ndisks, nparity, ashift, SPA_OLD_MAXBLOCKSIZE); /* allocation size for the blksize block */ asize = vdev_raidz_asize(ndisks, nparity, ashift, blksize); } else { uint64_t ndata; if (nvlist_lookup_uint64(vdevs[v], ZPOOL_CONFIG_DRAID_NDATA, &ndata) != 0) continue; /* allocation size for the "typical" 128k block */ tsize = vdev_draid_asize(ndata + nparity, nparity, ashift, SPA_OLD_MAXBLOCKSIZE); /* allocation size for the blksize block */ asize = vdev_draid_asize(ndata + nparity, nparity, ashift, blksize); } /* * Scale this size down as a ratio of 128k / tsize. * See theory statement above. */ volsize = nblocks * asize * SPA_OLD_MAXBLOCKSIZE / tsize; if (volsize > ret) { ret = volsize; } } if (ret == 0) { ret = nblocks * blksize; } return (ret); } /* * Convert the zvol's volume size to an appropriate reservation. See theory * comment above. * * Note: If this routine is updated, it is necessary to update the ZFS test * suite's shell version in reservation.shlib. */ uint64_t zvol_volsize_to_reservation(zpool_handle_t *zph, uint64_t volsize, nvlist_t *props) { uint64_t numdb; uint64_t nblocks, volblocksize; int ncopies; char *strval; if (nvlist_lookup_string(props, zfs_prop_to_name(ZFS_PROP_COPIES), &strval) == 0) ncopies = atoi(strval); else ncopies = 1; if (nvlist_lookup_uint64(props, zfs_prop_to_name(ZFS_PROP_VOLBLOCKSIZE), &volblocksize) != 0) volblocksize = ZVOL_DEFAULT_BLOCKSIZE; nblocks = volsize / volblocksize; /* * Metadata defaults to using 128k blocks, not volblocksize blocks. For * this reason, only the data blocks are scaled based on vdev config. */ volsize = volsize_from_vdevs(zph, nblocks, volblocksize); /* start with metadnode L0-L6 */ numdb = 7; /* calculate number of indirects */ while (nblocks > 1) { nblocks += DNODES_PER_LEVEL - 1; nblocks /= DNODES_PER_LEVEL; numdb += nblocks; } numdb *= MIN(SPA_DVAS_PER_BP, ncopies + 1); volsize *= ncopies; /* * this is exactly DN_MAX_INDBLKSHIFT when metadata isn't * compressed, but in practice they compress down to about * 1100 bytes */ numdb *= 1ULL << DN_MAX_INDBLKSHIFT; volsize += numdb; return (volsize); } /* * Wait for the given activity and return the status of the wait (whether or not * any waiting was done) in the 'waited' parameter. Non-existent fses are * reported via the 'missing' parameter, rather than by printing an error * message. This is convenient when this function is called in a loop over a * long period of time (as it is, for example, by zfs's wait cmd). In that * scenario, a fs being exported or destroyed should be considered a normal * event, so we don't want to print an error when we find that the fs doesn't * exist. */ int zfs_wait_status(zfs_handle_t *zhp, zfs_wait_activity_t activity, boolean_t *missing, boolean_t *waited) { int error = lzc_wait_fs(zhp->zfs_name, activity, waited); *missing = (error == ENOENT); if (*missing) return (0); if (error != 0) { (void) zfs_standard_error_fmt(zhp->zfs_hdl, error, dgettext(TEXT_DOMAIN, "error waiting in fs '%s'"), zhp->zfs_name); } return (error); }