diff options
Diffstat (limited to 'cmd')
-rw-r--r-- | cmd/pyzfs/pyzfs.py | 82 | ||||
-rw-r--r-- | cmd/sgs/include/alist.h | 24 | ||||
-rw-r--r-- | cmd/sgs/include/debug.h | 477 | ||||
-rw-r--r-- | cmd/sgs/include/sgs.h | 126 | ||||
-rw-r--r-- | cmd/sgs/messages/sgs.ident | 8 | ||||
-rw-r--r-- | cmd/sgs/tools/common/sgsmsg.c | 57 | ||||
-rw-r--r-- | cmd/sgs/tools/common/string_table.c | 31 | ||||
-rw-r--r-- | cmd/stat/common/statcommon.h | 328 | ||||
-rw-r--r-- | cmd/stat/common/timestamp.c | 54 | ||||
-rw-r--r-- | cmd/zdb/zdb.c | 3158 | ||||
-rw-r--r-- | cmd/zdb/zdb_il.c | 384 | ||||
-rw-r--r-- | cmd/zfs/zfs_iter.c | 464 | ||||
-rw-r--r-- | cmd/zfs/zfs_iter.h | 56 | ||||
-rw-r--r-- | cmd/zfs/zfs_main.c | 4160 | ||||
-rw-r--r-- | cmd/zfs/zfs_util.h | 42 | ||||
-rw-r--r-- | cmd/zinject/translate.c | 494 | ||||
-rw-r--r-- | cmd/zinject/zinject.c | 972 | ||||
-rw-r--r-- | cmd/zinject/zinject.h | 70 | ||||
-rw-r--r-- | cmd/zlook/zlook.c | 411 | ||||
-rw-r--r-- | cmd/zpool/zpool_iter.c | 252 | ||||
-rw-r--r-- | cmd/zpool/zpool_main.c | 4467 | ||||
-rw-r--r-- | cmd/zpool/zpool_util.c | 86 | ||||
-rw-r--r-- | cmd/zpool/zpool_util.h | 72 | ||||
-rw-r--r-- | cmd/zpool/zpool_vdev.c | 1469 | ||||
-rw-r--r-- | cmd/zstreamdump/zstreamdump.c | 429 | ||||
-rw-r--r-- | cmd/ztest/ztest.c | 5604 |
26 files changed, 23523 insertions, 254 deletions
diff --git a/cmd/pyzfs/pyzfs.py b/cmd/pyzfs/pyzfs.py new file mode 100644 index 000000000000..2088993d6d88 --- /dev/null +++ b/cmd/pyzfs/pyzfs.py @@ -0,0 +1,82 @@ +#! /usr/bin/python2.6 -S +# +# 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 http://www.opensolaris.org/os/licensing. +# 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) 2009, 2010, Oracle and/or its affiliates. All rights reserved. +# + +# Note, we want SIGINT (control-c) to exit the process quietly, to mimic +# the standard behavior of C programs. The best we can do with pure +# Python is to run with -S (to disable "import site"), and start our +# program with a "try" statement. Hopefully nobody hits ^C before our +# try statement is executed. + +try: + import site + import gettext + import zfs.util + import zfs.ioctl + import sys + import errno + import solaris.misc + + """This is the main script for doing zfs subcommands. It doesn't know + what subcommands there are, it just looks for a module zfs.<subcommand> + that implements that subcommand.""" + + try: + _ = gettext.translation("SUNW_OST_OSCMD", "/usr/lib/locale", + fallback=True).gettext + except: + _ = solaris.misc.gettext + + if len(sys.argv) < 2: + sys.exit(_("missing subcommand argument")) + + zfs.ioctl.set_cmdstr(" ".join(["zfs"] + sys.argv[1:])) + + try: + # import zfs.<subcommand> + # subfunc = zfs.<subcommand>.do_<subcommand> + + subcmd = sys.argv[1] + __import__("zfs." + subcmd) + submod = getattr(zfs, subcmd) + subfunc = getattr(submod, "do_" + subcmd) + except (ImportError, AttributeError): + sys.exit(_("invalid subcommand")) + + try: + subfunc() + except zfs.util.ZFSError, e: + print(e) + sys.exit(1) + +except IOError, e: + import errno + import sys + + if e.errno == errno.EPIPE: + sys.exit(1) + raise +except KeyboardInterrupt: + import sys + + sys.exit(1) diff --git a/cmd/sgs/include/alist.h b/cmd/sgs/include/alist.h index e92811f2ee5c..c5071a8cd289 100644 --- a/cmd/sgs/include/alist.h +++ b/cmd/sgs/include/alist.h @@ -20,7 +20,7 @@ */ /* - * Copyright 2008 Sun Microsystems, Inc. All rights reserved. + * Copyright 2009 Sun Microsystems, Inc. All rights reserved. * Use is subject to license terms. * * Define an Alist, a list maintained as a reallocable array, and a for() loop @@ -32,8 +32,6 @@ #ifndef _ALIST_H #define _ALIST_H -#pragma ident "%Z%%M% %I% %E% SMI" - #ifdef __cplusplus extern "C" { #endif @@ -138,6 +136,13 @@ typedef struct { void *apl_data[1]; /* data area: (arrcnt * size) bytes */ } APlist; +#ifdef _SYSCALL32 /* required by librtld_db */ +typedef struct { + Elf32_Word apl_arritems; + Elf32_Word apl_nitems; + Elf32_Addr apl_data[1]; +} APlist32; +#endif /* _SYSCALL32 */ /* * The ALIST_OFF_DATA and APLIST_OFF_DATA macros give the byte offset @@ -224,7 +229,7 @@ typedef struct { * Possible values returned by aplist_test() */ typedef enum { - ALE_ALLOCFAIL = 0, /* Memory allocation error */ + ALE_ALLOCFAIL = 0, /* memory allocation error */ ALE_EXISTS = 1, /* alist entry already exists */ ALE_NOTFND = 2, /* item not found and insert not required */ ALE_CREATE = 3 /* alist entry created */ @@ -244,11 +249,14 @@ typedef enum { ((void *)((_off) + (char *)(_lp))) /* - * # of items currently found in a list. These macros handle the case - * where the list has not been allocated yet. + * The number of items currently found in a list (nitems), and the total number + * of slots in the current data allocation (arritems). These macros handle the + * case where the list has not been allocated yet. */ -#define alist_nitems(_lp) (((_lp) == NULL) ? 0 : (_lp)->al_nitems) -#define aplist_nitems(_lp) (((_lp) == NULL) ? 0 : (_lp)->apl_nitems) +#define alist_nitems(_lp) (((_lp) == NULL) ? 0 : (_lp)->al_nitems) +#define aplist_nitems(_lp) (((_lp) == NULL) ? 0 : (_lp)->apl_nitems) +#define alist_arritems(_lp) (((_lp) == NULL) ? 0 : (_lp)->al_arritems) +#define aplist_arritems(_lp) (((_lp) == NULL) ? 0 : (_lp)->apl_arritems) extern void *alist_append(Alist **, const void *, size_t, Aliste); diff --git a/cmd/sgs/include/debug.h b/cmd/sgs/include/debug.h index 0a42f8dd9e9b..9db202574d0f 100644 --- a/cmd/sgs/include/debug.h +++ b/cmd/sgs/include/debug.h @@ -20,15 +20,12 @@ */ /* - * Copyright 2008 Sun Microsystems, Inc. All rights reserved. - * Use is subject to license terms. + * Copyright (c) 1991, 2010, Oracle and/or its affiliates. All rights reserved. */ #ifndef _DEBUG_H #define _DEBUG_H -#pragma ident "%Z%%M% %I% %E% SMI" - /* * Global include file for lddbg debugging. * @@ -45,6 +42,7 @@ * start with the `Elf_' prefix. These latter routines are the only * routines used by the elfdump(1) utility. */ +#include <sys/times.h> #include <sgs.h> #include <libld.h> #include <rtld.h> @@ -55,13 +53,12 @@ extern "C" { #endif /* - * Define Dbg_*() interface flags. These flags direct the debugging routine to - * generate different diagnostics, thus the strings themselves are maintained + * Define Dbg_*() interface values. These values direct the debugging routine + * to generate different diagnostics, thus the strings themselves are maintained * in the debugging library. */ #define DBG_SUP_ENVIRON 1 #define DBG_SUP_CMDLINE 2 -#define DBG_SUP_DEFAULT 3 #define DBG_CONF_IGNORE 1 /* configuration processing errors */ #define DBG_CONF_VERSION 2 @@ -87,18 +84,22 @@ extern "C" { #define DBG_DLSYM_SELF 3 #define DBG_DLSYM_PROBE 4 #define DBG_DLSYM_SINGLETON 5 +#define DBG_DLSYM_NUM DBG_DLSYM_SINGLETON + 1 #define DBG_DLCLOSE_NULL 0 #define DBG_DLCLOSE_IGNORE 1 #define DBG_DLCLOSE_RESCAN 2 -#define DBG_WAIT_INIT 1 -#define DBG_WAIT_FINI 2 -#define DBG_WAIT_SYMBOL 3 - #define DBG_SYM_REDUCE_GLOBAL 1 /* reporting global symbols to local */ #define DBG_SYM_REDUCE_RETAIN 2 /* reporting non reduced local syms */ +#define DBG_AUD_CALL 1 /* original call to auditor */ +#define DBG_AUD_RET 2 /* return from auditor diagnostic */ + +#define DBG_AUD_LOCAL 0 /* auditor is local */ +#define DBG_AUD_GLOBAL 1 /* auditor is global */ +#define DBG_AUD_PRELOAD 2 /* auditor is preloaded */ + /* * Group handle operations - passed to Dbg_file_hdl_title(). Indicate why * handle dependencies are being manipulated. @@ -139,12 +140,18 @@ extern "C" { #define DBG_BINFO_REF_PARENT 0x2000 /* reference to PARENT */ #define DBG_BINFO_REF_MSK 0xf000 - -#define DBG_CAP_INITIAL 0 -#define DBG_CAP_IGNORE 1 -#define DBG_CAP_OLD 2 -#define DBG_CAP_NEW 3 -#define DBG_CAP_RESOLVED 4 +/* + * ld.so.1(1) symbol capabilities processing. + */ +#define DBG_CAP_DEFAULT 0 +#define DBG_CAP_USED 1 +#define DBG_CAP_CANDIDATE 2 +#define DBG_CAP_REJECTED 3 +#define DBG_CAP_HW_1 4 +#define DBG_CAP_SF_1 5 +#define DBG_CAP_HW_2 6 +#define DBG_CAP_PLAT 7 +#define DBG_CAP_MACH 8 #define DBG_REL_START 1 #define DBG_REL_FINISH 2 @@ -153,23 +160,53 @@ extern "C" { #define DBG_NL_STD 0 /* newline controllers - standard and */ #define DBG_NL_FRC 2 /* forced. */ -#define DBG_BNDREJ_NODIR 0 /* bind rejected, direct to nodirect */ -#define DBG_BNDREJ_SINGLE 1 /* bind rejected, singleton without */ +#define DBG_BNDREJ_DIRECT 0 /* bind rejected, direct to nodirect */ +#define DBG_BNDREJ_GROUP 1 /* bind rejected, group to nodirect */ +#define DBG_BNDREJ_SINGLE 2 /* bind rejected, singleton without */ /* default search model */ #define DBG_BNDREJ_NUM DBG_BNDREJ_SINGLE /* + * Dbg_state_str() is used to obtain commonly used "state transition" + * strings used in various debugging output. + */ +#define DBG_STATE_ADD 0 /* add */ +#define DBG_STATE_CURRENT 1 /* current */ +#define DBG_STATE_EXCLUDE 2 /* exclude */ +#define DBG_STATE_IGNORED 3 /* ignored */ +#define DBG_STATE_MOD_BEFORE 4 /* modify (before) */ +#define DBG_STATE_MOD_AFTER 5 /* modify (after) */ +#define DBG_STATE_NEW 6 /* new */ +#define DBG_STATE_NEW_IMPLICIT 7 /* new (implicit) */ +#define DBG_STATE_RESET 8 /* reset */ +#define DBG_STATE_ORIGINAL 9 /* original */ +#define DBG_STATE_RESOLVED 10 /* resolved */ + +#define DBG_STATE_NUM 11 +typedef uint_t dbg_state_t; +extern const char *Dbg_state_str(dbg_state_t); + +/* * Define a debug descriptor, and a user macro that inspects the descriptor as * a means of triggering a class of diagnostic output. */ typedef struct { uint_t d_class; /* debugging classes */ - uint_t d_extra; /* extra information for classes */ - APlist *d_list; /* associated strings */ + uint_t d_extra; /* extra public information */ + APlist *d_list; /* accepted link-map list names */ + struct timeval d_totaltime; /* total time since entry - */ + /* gettimeofday(3c) */ + struct timeval d_deltatime; /* delta time since last diagnostic - */ + /* gettimeofday(3c) */ } Dbg_desc; extern Dbg_desc *dbg_desc; +/* + * Macros used to avoid calls to liblddbg unless debugging is enabled. + * liblddbg is lazy loaded --- this prevents it from happening unless + * it will actually be used. + */ #define DBG_ENABLED (dbg_desc->d_class) #define DBG_CALL(func) if (DBG_ENABLED) func @@ -180,26 +217,51 @@ extern Dbg_desc *dbg_desc; * may be interpreted by the debugging library itself or from the callers * dbg_print() routine. */ -#define DBG_E_DETAIL 0x0001 /* add detail to a class */ -#define DBG_E_LONG 0x0002 /* use long names (ie. no truncation) */ - -#define DBG_E_STDNL 0x0010 /* standard newline indicator */ +#define DBG_E_DETAIL 0x00000001 /* add detail to a class */ +#define DBG_E_LONG 0x00000002 /* use long names (ie. no truncation) */ +#define DBG_E_DEMANGLE 0x00000004 /* demangle symbol names */ +#define DBG_E_STDNL 0x00000008 /* standard newline indicator */ +#define DBG_E_HELP 0x00000010 /* help requested */ +#define DBG_E_HELP_EXIT 0x00000020 /* hint: user should exit after help */ +#define DBG_E_TTIME 0x00000040 /* prepend total time */ +#define DBG_E_DTIME 0x00000080 /* prepend delta time */ +#define DBG_E_RESET 0x00000100 /* reset times */ + +/* ld only */ +#define DBG_E_SNAME 0x00001000 /* prepend simple name */ +#define DBG_E_FNAME 0x00002000 /* prepend full name */ +#define DBG_E_CLASS 0x00004000 /* prepend ELF class */ + +/* ld.so.1 only */ +#define DBG_E_LMID 0x00100000 /* prepend link-map id */ +#define DBG_E_LMID_LDSO 0x00200000 /* show ldso link-map list */ +#define DBG_E_LMID_ALL 0x00400000 /* show all non-ldso link-map lists */ +#define DBG_E_LMID_ALT 0x00800000 /* show all ALT link-map lists */ +#define DBG_E_LMID_BASE 0x01000000 /* show BASE link-map list */ -#define DBG_E_SNAME 0x0100 /* prepend simple name (ld only) */ -#define DBG_E_FNAME 0x0200 /* prepend full name (ld only) */ -#define DBG_E_CLASS 0x0400 /* prepend ELF class (ld only) */ -#define DBG_E_LMID 0x0800 /* prepend link-map id (ld.so.1 only) */ -#define DBG_E_DEMANGLE 0x1000 /* demangle symbol names */ #define DBG_NOTDETAIL() !(dbg_desc->d_extra & DBG_E_DETAIL) #define DBG_NOTLONG() !(dbg_desc->d_extra & DBG_E_LONG) +#define DBG_ISDEMANGLE() \ + (dbg_desc->d_extra & DBG_E_DEMANGLE) + +#define DBG_TOTALTIME (dbg_desc->d_totaltime) +#define DBG_DELTATIME (dbg_desc->d_deltatime) + +#define DBG_ISTTIME() (dbg_desc->d_extra & DBG_E_TTIME) +#define DBG_ISDTIME() (dbg_desc->d_extra & DBG_E_DTIME) +#define DBG_ISTIME() (dbg_desc->d_extra & (DBG_E_TTIME | DBG_E_DTIME)) +#define DBG_NOTTIME() !(dbg_desc->d_extra & (DBG_E_TTIME | DBG_E_DTIME)) + +#define DBG_ISRESET() (dbg_desc->d_extra & DBG_E_RESET) +#define DBG_ONRESET() (dbg_desc->d_extra |= DBG_E_RESET) +#define DBG_OFFRESET() (dbg_desc->d_extra &= ~DBG_E_RESET) + #define DBG_ISSNAME() (dbg_desc->d_extra & DBG_E_SNAME) #define DBG_ISFNAME() (dbg_desc->d_extra & DBG_E_FNAME) #define DBG_ISCLASS() (dbg_desc->d_extra & DBG_E_CLASS) #define DBG_ISLMID() (dbg_desc->d_extra & DBG_E_LMID) -#define DBG_ISDEMANGLE() \ - (dbg_desc->d_extra & DBG_E_DEMANGLE) /* * Print routine, this must be supplied by the application. The initial @@ -209,7 +271,19 @@ extern Dbg_desc *dbg_desc; /* PRINTFLIKE2 */ extern void dbg_print(Lm_list *, const char *, ...); -extern uintptr_t Dbg_setup(const char *, Dbg_desc *); +/* + * Initialization routine, called before any other Dbg routines to + * establish the necessary state. + */ +typedef enum { DBG_CALLER_LD, DBG_CALLER_RTLD } dbg_setup_caller_t; +extern int Dbg_setup(dbg_setup_caller_t, const char *, + Dbg_desc *, const char **); + +/* Call dbg_print() to produce linker version output */ +extern void Dbg_version(void); + +/* Call dbg_print() to produce help output */ +extern void Dbg_help(void); /* * Establish ELF32 and ELF64 class Dbg_*() interfaces. @@ -225,12 +299,28 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_bind_reject Dbg64_bind_reject #define Dbg_bind_weak Dbg64_bind_weak -#define Dbg_cap_val_hw1 Dbg64_cap_val_hw1 -#define Dbg_cap_hw_candidate Dbg64_cap_hw_candidate -#define Dbg_cap_hw_filter Dbg64_cap_hw_filter -#define Dbg_cap_mapfile Dbg64_cap_mapfile -#define Dbg_cap_sec_entry Dbg64_cap_sec_entry +#define Dbg_cap_candidate Dbg64_cap_candidate +#define Dbg_cap_filter Dbg64_cap_filter +#define Dbg_cap_id Dbg64_cap_id +#define Dbg_cap_identical Dbg64_cap_identical +#define Dbg_cap_mapfile_title Dbg64_cap_mapfile_title +#define Dbg_cap_post_title Dbg64_cap_post_title #define Dbg_cap_sec_title Dbg64_cap_sec_title +#define Dbg_cap_val Dbg64_cap_val +#define Dbg_cap_ptr_entry Dbg64_cap_ptr_entry +#define Dbg_cap_val_entry Dbg64_cap_val_entry + +#define Dbg_dl_dladdr Dbg64_dl_dladdr +#define Dbg_dl_dlclose Dbg64_dl_dlclose +#define Dbg_dl_dldump Dbg64_dl_dldump +#define Dbg_dl_dlerror Dbg64_dl_dlerror +#define Dbg_dl_dlinfo Dbg64_dl_dlinfo +#define Dbg_dl_dlopen Dbg64_dl_dlopen +#define Dbg_dl_dlsym Dbg64_dl_dlsym +#define Dbg_dl_iphdr_enter Dbg64_dl_iphdr_enter +#define Dbg_dl_iphdr_callback Dbg64_dl_iphdr_callback +#define Dbg_dl_iphdr_mapchange Dbg64_dl_iphdr_mapchange +#define Dbg_dl_iphdr_unmap_ret Dbg64_dl_iphdr_unmap_ret #define Dbg_ent_entry Dbg64_ent_entry #define Dbg_ent_print Dbg64_ent_print @@ -241,15 +331,14 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_file_ar_rescan Dbg64_file_ar_rescan #define Dbg_file_bind_entry Dbg64_file_bind_entry #define Dbg_file_bindings Dbg64_file_bindings +#define Dbg_file_bindings_done Dbg64_file_bindings_done #define Dbg_file_cleanup Dbg64_file_cleanup #define Dbg_file_cntl Dbg64_file_cntl #define Dbg_file_config_dis Dbg64_file_config_dis #define Dbg_file_config_obj Dbg64_file_config_obj #define Dbg_file_del_rescan Dbg64_file_del_rescan #define Dbg_file_delete Dbg64_file_delete -#define Dbg_file_dlclose Dbg64_file_dlclose -#define Dbg_file_dldump Dbg64_file_dldump -#define Dbg_file_dlopen Dbg64_file_dlopen +#define Dbg_file_deferred Dbg64_file_deferred #define Dbg_file_elf Dbg64_file_elf #define Dbg_file_filtee Dbg64_file_filtee #define Dbg_file_filter Dbg64_file_filter @@ -260,6 +349,7 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_file_hdl_title Dbg64_file_hdl_title #define Dbg_file_lazyload Dbg64_file_lazyload #define Dbg_file_ldso Dbg64_file_ldso +#define Dbg_file_mmapobj Dbg64_file_mmapobj #define Dbg_file_mode_promote Dbg64_file_mode_promote #define Dbg_file_modified Dbg64_file_modified #define Dbg_file_needed Dbg64_file_needed @@ -275,7 +365,7 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_libs_audit Dbg64_libs_audit #define Dbg_libs_find Dbg64_libs_find #define Dbg_libs_found Dbg64_libs_found -#define Dbg_libs_ignore Dbg64_libs_ignore +#define Dbg_libs_insecure Dbg64_libs_insecure #define Dbg_libs_init Dbg64_libs_init #define Dbg_libs_l Dbg64_libs_l #define Dbg_libs_path Dbg64_libs_path @@ -284,22 +374,28 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_libs_yp Dbg64_libs_yp #define Dbg_libs_ylu Dbg64_libs_ylu -#define Dbg_map_dash Dbg64_map_dash +#define Dbg_map_cexp_id Dbg64_map_cexp_id +#define Dbg_map_dv Dbg64_map_dv +#define Dbg_map_dv_entry Dbg64_map_dv_entry #define Dbg_map_ent Dbg64_map_ent +#define Dbg_map_ent_ord_title Dbg64_map_ent_ord_title +#define Dbg_map_hdr_noalloc Dbg64_map_hdr_noalloc #define Dbg_map_parse Dbg64_map_parse -#define Dbg_map_pipe Dbg64_map_pipe +#define Dbg_map_pass Dbg64_map_pass +#define Dbg_map_post_title Dbg64_map_post_title #define Dbg_map_seg Dbg64_map_seg -#define Dbg_map_set_atsign Dbg64_map_set_atsign -#define Dbg_map_set_equal Dbg64_map_set_equal +#define Dbg_map_seg_order Dbg64_map_seg_order +#define Dbg_map_seg_os_order Dbg64_map_seg_os_order #define Dbg_map_size_new Dbg64_map_size_new #define Dbg_map_size_old Dbg64_map_size_old -#define Dbg_map_sort_fini Dbg64_map_sort_fini -#define Dbg_map_sort_orig Dbg64_map_sort_orig +#define Dbg_map_sort_seg Dbg64_map_sort_seg +#define Dbg_map_sort_title Dbg64_map_sort_title #define Dbg_map_symbol Dbg64_map_symbol #define Dbg_map_version Dbg64_map_version #define Dbg_move_adjexpandreloc Dbg64_move_adjexpandreloc #define Dbg_move_adjmovereloc Dbg64_move_adjmovereloc +#define Dbg_move_bad Dbg64_move_bad #define Dbg_move_data Dbg64_move_data #define Dbg_move_entry1 Dbg64_move_entry1 #define Dbg_move_entry2 Dbg64_move_entry2 @@ -329,13 +425,16 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_reloc_sloppycomdat Dbg64_reloc_sloppycomdat #define Dbg_sec_added Dbg64_sec_added +#define Dbg_sec_backing Dbg64_sec_backing #define Dbg_sec_created Dbg64_sec_created #define Dbg_sec_discarded Dbg64_sec_discarded #define Dbg_sec_genstr_compress Dbg64_sec_genstr_compress #define Dbg_sec_group Dbg64_sec_group +#define Dbg_sec_gnu_comdat Dbg64_sec_gnu_comdat #define Dbg_sec_in Dbg64_sec_in #define Dbg_sec_order_error Dbg64_sec_order_error #define Dbg_sec_order_list Dbg64_sec_order_list +#define Dbg_sec_redirected Dbg64_sec_redirected #define Dbg_sec_strtab Dbg64_sec_strtab #define Dbg_sec_unsup_strmerge Dbg64_sec_unsup_strmerge @@ -353,17 +452,24 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_support_action Dbg64_support_action #define Dbg_support_load Dbg64_support_load #define Dbg_support_req Dbg64_support_req +#define Dbg_support_vnone Dbg64_support_vnone #define Dbg_syminfo_entry Dbg64_syminfo_entry #define Dbg_syminfo_title Dbg64_syminfo_title #define Dbg_syms_ar_checking Dbg64_syms_ar_checking -#define Dbg_syms_ar_entry Dbg64_syms_ar_entry +#define Dbg_syms_ar_force Dbg64_syms_ar_force #define Dbg_syms_ar_resolve Dbg64_syms_ar_resolve +#define Dbg_syms_ar_skip Dbg64_syms_ar_skip #define Dbg_syms_ar_title Dbg64_syms_ar_title +#define Dbg_syms_cap_convert Dbg64_syms_cap_convert +#define Dbg_syms_cap_local Dbg64_syms_cap_local +#define Dbg_syms_cap_lookup Dbg64_syms_cap_lookup +#define Dbg_syms_cap_title Dbg64_syms_cap_title +#define Dbg_syms_copy_reloc Dbg64_syms_copy_reloc #define Dbg_syms_created Dbg64_syms_created #define Dbg_syms_discarded Dbg64_syms_discarded -#define Dbg_syms_dlsym Dbg64_syms_dlsym +#define Dbg_syms_dup_discarded Dbg64_syms_dup_discarded #define Dbg_syms_dup_sort_addr Dbg64_syms_dup_sort_addr #define Dbg_syms_entered Dbg64_syms_entered #define Dbg_syms_entry Dbg64_syms_entry @@ -376,7 +482,6 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_syms_old Dbg64_syms_old #define Dbg_syms_process Dbg64_syms_process #define Dbg_syms_reduce Dbg64_syms_reduce -#define Dbg_syms_reloc Dbg64_syms_reloc #define Dbg_syms_resolved Dbg64_syms_resolved #define Dbg_syms_resolving Dbg64_syms_resolving #define Dbg_syms_sec_entry Dbg64_syms_sec_entry @@ -384,8 +489,8 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_syms_spec_title Dbg64_syms_spec_title #define Dbg_syms_updated Dbg64_syms_updated #define Dbg_syms_up_title Dbg64_syms_up_title +#define Dbg_syms_wrap Dbg64_syms_wrap -#define Dbg_util_broadcast Dbg64_util_broadcast #define Dbg_util_call_array Dbg64_util_call_array #define Dbg_util_call_fini Dbg64_util_call_fini #define Dbg_util_call_init Dbg64_util_call_init @@ -401,7 +506,6 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_util_scc_entry Dbg64_util_scc_entry #define Dbg_util_scc_title Dbg64_util_scc_title #define Dbg_util_str Dbg64_util_str -#define Dbg_util_wait Dbg64_util_wait #define Dbg_unused_file Dbg64_unused_file #define Dbg_unused_lcinterface Dbg64_unused_lcinterface @@ -413,6 +517,7 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_ver_avail_title Dbg64_ver_avail_title #define Dbg_ver_def_title Dbg64_ver_def_title #define Dbg_ver_desc_entry Dbg64_ver_desc_entry +#define Dbg_ver_need_done Dbg64_ver_need_done #define Dbg_ver_need_entry Dbg64_ver_need_entry #define Dbg_ver_need_title Dbg64_ver_need_title #define Dbg_ver_nointerface Dbg64_ver_nointerface @@ -427,12 +532,28 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_bind_reject Dbg32_bind_reject #define Dbg_bind_weak Dbg32_bind_weak -#define Dbg_cap_val_hw1 Dbg32_cap_val_hw1 -#define Dbg_cap_hw_candidate Dbg32_cap_hw_candidate -#define Dbg_cap_hw_filter Dbg32_cap_hw_filter -#define Dbg_cap_mapfile Dbg32_cap_mapfile -#define Dbg_cap_sec_entry Dbg32_cap_sec_entry +#define Dbg_cap_candidate Dbg32_cap_candidate +#define Dbg_cap_filter Dbg32_cap_filter +#define Dbg_cap_id Dbg32_cap_id +#define Dbg_cap_identical Dbg32_cap_identical +#define Dbg_cap_mapfile_title Dbg32_cap_mapfile_title +#define Dbg_cap_post_title Dbg32_cap_post_title #define Dbg_cap_sec_title Dbg32_cap_sec_title +#define Dbg_cap_val Dbg32_cap_val +#define Dbg_cap_ptr_entry Dbg32_cap_ptr_entry +#define Dbg_cap_val_entry Dbg32_cap_val_entry + +#define Dbg_dl_dladdr Dbg32_dl_dladdr +#define Dbg_dl_dlclose Dbg32_dl_dlclose +#define Dbg_dl_dldump Dbg32_dl_dldump +#define Dbg_dl_dlerror Dbg32_dl_dlerror +#define Dbg_dl_dlinfo Dbg32_dl_dlinfo +#define Dbg_dl_dlopen Dbg32_dl_dlopen +#define Dbg_dl_dlsym Dbg32_dl_dlsym +#define Dbg_dl_iphdr_enter Dbg32_dl_iphdr_enter +#define Dbg_dl_iphdr_callback Dbg32_dl_iphdr_callback +#define Dbg_dl_iphdr_mapchange Dbg32_dl_iphdr_mapchange +#define Dbg_dl_iphdr_unmap_ret Dbg32_dl_iphdr_unmap_ret #define Dbg_ent_entry Dbg32_ent_entry #define Dbg_ent_print Dbg32_ent_print @@ -443,15 +564,14 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_file_ar_rescan Dbg32_file_ar_rescan #define Dbg_file_bind_entry Dbg32_file_bind_entry #define Dbg_file_bindings Dbg32_file_bindings +#define Dbg_file_bindings_done Dbg32_file_bindings_done #define Dbg_file_cleanup Dbg32_file_cleanup #define Dbg_file_cntl Dbg32_file_cntl #define Dbg_file_config_dis Dbg32_file_config_dis #define Dbg_file_config_obj Dbg32_file_config_obj #define Dbg_file_del_rescan Dbg32_file_del_rescan #define Dbg_file_delete Dbg32_file_delete -#define Dbg_file_dlclose Dbg32_file_dlclose -#define Dbg_file_dldump Dbg32_file_dldump -#define Dbg_file_dlopen Dbg32_file_dlopen +#define Dbg_file_deferred Dbg32_file_deferred #define Dbg_file_elf Dbg32_file_elf #define Dbg_file_filtee Dbg32_file_filtee #define Dbg_file_filter Dbg32_file_filter @@ -462,6 +582,7 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_file_hdl_title Dbg32_file_hdl_title #define Dbg_file_lazyload Dbg32_file_lazyload #define Dbg_file_ldso Dbg32_file_ldso +#define Dbg_file_mmapobj Dbg32_file_mmapobj #define Dbg_file_mode_promote Dbg32_file_mode_promote #define Dbg_file_modified Dbg32_file_modified #define Dbg_file_needed Dbg32_file_needed @@ -477,7 +598,7 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_libs_audit Dbg32_libs_audit #define Dbg_libs_find Dbg32_libs_find #define Dbg_libs_found Dbg32_libs_found -#define Dbg_libs_ignore Dbg32_libs_ignore +#define Dbg_libs_insecure Dbg32_libs_insecure #define Dbg_libs_init Dbg32_libs_init #define Dbg_libs_l Dbg32_libs_l #define Dbg_libs_path Dbg32_libs_path @@ -486,22 +607,28 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_libs_yp Dbg32_libs_yp #define Dbg_libs_ylu Dbg32_libs_ylu -#define Dbg_map_dash Dbg32_map_dash +#define Dbg_map_cexp_id Dbg32_map_cexp_id +#define Dbg_map_dv Dbg32_map_dv +#define Dbg_map_dv_entry Dbg32_map_dv_entry #define Dbg_map_ent Dbg32_map_ent +#define Dbg_map_ent_ord_title Dbg32_map_ent_ord_title +#define Dbg_map_hdr_noalloc Dbg32_map_hdr_noalloc #define Dbg_map_parse Dbg32_map_parse -#define Dbg_map_pipe Dbg32_map_pipe +#define Dbg_map_pass Dbg32_map_pass +#define Dbg_map_post_title Dbg32_map_post_title #define Dbg_map_seg Dbg32_map_seg -#define Dbg_map_set_atsign Dbg32_map_set_atsign -#define Dbg_map_set_equal Dbg32_map_set_equal +#define Dbg_map_seg_order Dbg32_map_seg_order +#define Dbg_map_seg_os_order Dbg32_map_seg_os_order #define Dbg_map_size_new Dbg32_map_size_new #define Dbg_map_size_old Dbg32_map_size_old -#define Dbg_map_sort_fini Dbg32_map_sort_fini -#define Dbg_map_sort_orig Dbg32_map_sort_orig +#define Dbg_map_sort_seg Dbg32_map_sort_seg +#define Dbg_map_sort_title Dbg32_map_sort_title #define Dbg_map_symbol Dbg32_map_symbol #define Dbg_map_version Dbg32_map_version #define Dbg_move_adjexpandreloc Dbg32_move_adjexpandreloc #define Dbg_move_adjmovereloc Dbg32_move_adjmovereloc +#define Dbg_move_bad Dbg32_move_bad #define Dbg_move_data Dbg32_move_data #define Dbg_move_entry1 Dbg32_move_entry1 #define Dbg_move_entry2 Dbg32_move_entry2 @@ -531,13 +658,16 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_reloc_sloppycomdat Dbg32_reloc_sloppycomdat #define Dbg_sec_added Dbg32_sec_added +#define Dbg_sec_backing Dbg32_sec_backing #define Dbg_sec_created Dbg32_sec_created #define Dbg_sec_discarded Dbg32_sec_discarded #define Dbg_sec_genstr_compress Dbg32_sec_genstr_compress #define Dbg_sec_group Dbg32_sec_group +#define Dbg_sec_gnu_comdat Dbg32_sec_gnu_comdat #define Dbg_sec_in Dbg32_sec_in #define Dbg_sec_order_error Dbg32_sec_order_error #define Dbg_sec_order_list Dbg32_sec_order_list +#define Dbg_sec_redirected Dbg32_sec_redirected #define Dbg_sec_strtab Dbg32_sec_strtab #define Dbg_sec_unsup_strmerge Dbg32_sec_unsup_strmerge @@ -555,17 +685,24 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_support_action Dbg32_support_action #define Dbg_support_load Dbg32_support_load #define Dbg_support_req Dbg32_support_req +#define Dbg_support_vnone Dbg32_support_vnone #define Dbg_syminfo_entry Dbg32_syminfo_entry #define Dbg_syminfo_title Dbg32_syminfo_title #define Dbg_syms_ar_checking Dbg32_syms_ar_checking -#define Dbg_syms_ar_entry Dbg32_syms_ar_entry +#define Dbg_syms_ar_force Dbg32_syms_ar_force #define Dbg_syms_ar_resolve Dbg32_syms_ar_resolve +#define Dbg_syms_ar_skip Dbg32_syms_ar_skip #define Dbg_syms_ar_title Dbg32_syms_ar_title +#define Dbg_syms_cap_convert Dbg32_syms_cap_convert +#define Dbg_syms_cap_local Dbg32_syms_cap_local +#define Dbg_syms_cap_lookup Dbg32_syms_cap_lookup +#define Dbg_syms_cap_title Dbg32_syms_cap_title +#define Dbg_syms_copy_reloc Dbg32_syms_copy_reloc #define Dbg_syms_created Dbg32_syms_created #define Dbg_syms_discarded Dbg32_syms_discarded -#define Dbg_syms_dlsym Dbg32_syms_dlsym +#define Dbg_syms_dup_discarded Dbg32_syms_dup_discarded #define Dbg_syms_dup_sort_addr Dbg32_syms_dup_sort_addr #define Dbg_syms_entered Dbg32_syms_entered #define Dbg_syms_entry Dbg32_syms_entry @@ -579,7 +716,6 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_syms_old Dbg32_syms_old #define Dbg_syms_process Dbg32_syms_process #define Dbg_syms_reduce Dbg32_syms_reduce -#define Dbg_syms_reloc Dbg32_syms_reloc #define Dbg_syms_resolved Dbg32_syms_resolved #define Dbg_syms_resolving Dbg32_syms_resolving #define Dbg_syms_sec_entry Dbg32_syms_sec_entry @@ -587,8 +723,8 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_syms_spec_title Dbg32_syms_spec_title #define Dbg_syms_updated Dbg32_syms_updated #define Dbg_syms_up_title Dbg32_syms_up_title +#define Dbg_syms_wrap Dbg32_syms_wrap -#define Dbg_util_broadcast Dbg32_util_broadcast #define Dbg_util_call_array Dbg32_util_call_array #define Dbg_util_call_fini Dbg32_util_call_fini #define Dbg_util_call_init Dbg32_util_call_init @@ -604,7 +740,6 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_util_scc_entry Dbg32_util_scc_entry #define Dbg_util_scc_title Dbg32_util_scc_title #define Dbg_util_str Dbg32_util_str -#define Dbg_util_wait Dbg32_util_wait #define Dbg_unused_file Dbg32_unused_file #define Dbg_unused_lcinterface Dbg32_unused_lcinterface @@ -616,6 +751,7 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); #define Dbg_ver_avail_title Dbg32_ver_avail_title #define Dbg_ver_def_title Dbg32_ver_def_title #define Dbg_ver_desc_entry Dbg32_ver_desc_entry +#define Dbg_ver_need_done Dbg32_ver_need_done #define Dbg_ver_need_entry Dbg32_ver_need_entry #define Dbg_ver_need_title Dbg32_ver_need_title #define Dbg_ver_nointerface Dbg32_ver_nointerface @@ -626,17 +762,40 @@ extern uintptr_t Dbg_setup(const char *, Dbg_desc *); /* * External Dbg_*() interface routines. */ -extern void Dbg_args_files(Lm_list *, int, char *); -extern void Dbg_args_flags(Lm_list *, int, int); +extern void Dbg_args_file(Lm_list *, int, char *); +extern void Dbg_args_guidance_unknown(Lm_list *, const char *); +extern void Dbg_args_option(Lm_list *, int, int, char *); +extern void Dbg_args_str2chr(Lm_list *, int, const char *, int); +extern void Dbg_args_Wldel(Lm_list *, int, const char *); +extern void Dbg_audit_activity(Lm_list *, const char *, const char *, + uint_t); extern void Dbg_audit_ignore(Rt_map *); extern void Dbg_audit_interface(Lm_list *, const char *, const char *); -extern void Dbg_audit_lib(Lm_list *, const char *); -extern void Dbg_audit_object(Lm_list *, const char *, const char *); -extern void Dbg_audit_symval(Lm_list *, const char *, const char *, - const char *, Addr, Addr); +extern void Dbg_audit_lib(Rt_map *, const char *, int); +extern void Dbg_audit_objclose(Lm_list *, const char *, const char *); +extern void Dbg_audit_objopen(Lm_list *, int, const char *, const char *, + uint_t, Boolean); +extern void Dbg_audit_objfilter(Lm_list *, int, const char *, const char *, + const char *, const char *); +extern void Dbg_audit_objsearch(Lm_list *, int, const char *, const char *, + uint_t, const char *); +extern void Dbg_audit_pltenter(Lm_list *, int, const char *, const char *, + Addr); +extern void Dbg_audit_pltexit(Lm_list *, const char *, const char *); +extern void Dbg_audit_preinit(Lm_list *, const char *, const char *); +extern void Dbg_audit_symbind(Lm_list *, int, const char *, const char *, + Addr, uint_t); extern void Dbg_audit_skip(Lm_list *, const char *, const char *); extern void Dbg_audit_terminate(Lm_list *, const char *); -extern void Dbg_audit_version(Lm_list *, const char *, ulong_t); +extern void Dbg_audit_version(Lm_list *, const char *, uint_t, uint_t); + +extern void Dbg_basic_collect(Lm_list *); +extern void Dbg_basic_create(Lm_list *); +extern void Dbg_basic_finish(Lm_list *); +extern void Dbg_basic_files(Lm_list *); +extern void Dbg_basic_options(Lm_list *); +extern void Dbg_basic_relocate(Lm_list *); +extern void Dbg_basic_validate(Lm_list *); extern void Dbg_bind_global(Rt_map *, Addr, Off, Xword, Pltbindtype, Rt_map *, Addr, Off, const char *, uint_t); @@ -649,26 +808,43 @@ extern void Dbg_bind_pltpad_to(Rt_map *, Addr, const char *, const char *); extern void Dbg_bind_reject(Rt_map *, Rt_map *, const char *, int); extern void Dbg_bind_weak(Rt_map *, Addr, Addr, const char *); -extern void Dbg_cap_hw_candidate(Lm_list *, const char *); -extern void Dbg_cap_hw_filter(Lm_list *, const char *, Rt_map *); -extern void Dbg_cap_mapfile(Lm_list *, Xword, Xword, Half); -extern void Dbg_cap_sec_entry(Lm_list *, uint_t, Xword, Xword, Half); -extern void Dbg_cap_sec_title(Ofl_desc *); -extern void Dbg_cap_val_hw1(Lm_list *, Xword, Half); +extern void Dbg_cap_candidate(Lm_list *, const char *); +extern void Dbg_cap_filter(Lm_list *, const char *, Rt_map *); +extern void Dbg_cap_id(Lm_list *, Lineno, const char *, const char *); +extern void Dbg_cap_identical(Lm_list *, const char *, const char *); +extern void Dbg_cap_mapfile_title(Lm_list *, Lineno); +extern void Dbg_cap_post_title(Lm_list *, int *); +extern void Dbg_cap_sec_title(Lm_list *, const char *); +extern void Dbg_cap_val(Lm_list *, Syscapset *, Syscapset *, Half); +extern void Dbg_cap_ptr_entry(Lm_list *, dbg_state_t, Xword, const char *); +extern void Dbg_cap_val_entry(Lm_list *, dbg_state_t, Xword, Xword, Half); + +extern void Dbg_dl_dladdr(Rt_map *, void *); +extern void Dbg_dl_dlclose(Rt_map *, const char *, int); +extern void Dbg_dl_dldump(Rt_map *, const char *, const char *, int); +extern void Dbg_dl_dlerror(Rt_map *, const char *); +extern void Dbg_dl_dlinfo(Rt_map *, const char *, int, void *); +extern void Dbg_dl_dlopen(Rt_map *, const char *, int *, int); +extern void Dbg_dl_dlsym(Rt_map *, const char *, int *, const char *, int); +extern void Dbg_dl_iphdr_enter(Rt_map *, u_longlong_t, u_longlong_t); +extern void Dbg_dl_iphdr_callback(Lm_list *, struct dl_phdr_info *); +extern void Dbg_dl_iphdr_mapchange(Lm_list *, u_longlong_t, u_longlong_t); +extern void Dbg_dl_iphdr_unmap_ret(Lm_list *); extern const char * Dbg_demangle_name(const char *); -extern void Dbg_ent_entry(Lm_list *, Half, Ent_desc *); -extern void Dbg_ent_print(Lm_list *, Half, List *, Boolean); +extern void Dbg_ent_entry(Lm_list *, uchar_t, Half, Ent_desc *); +extern void Dbg_ent_print(Lm_list *, uchar_t, Half, APlist *); extern void Dbg_file_analyze(Rt_map *); -extern void Dbg_file_aout(Lm_list *, const char *, ulong_t, ulong_t, - ulong_t, const char *, Aliste); -extern void Dbg_file_ar(Lm_list *, const char *, int); -extern void Dbg_file_ar_rescan(Lm_list *); +extern void Dbg_file_aout(Lm_list *, const char *, Addr, size_t, + const char *, Aliste); +extern void Dbg_file_ar(Lm_list *, const char *, Boolean); +extern void Dbg_file_ar_rescan(Lm_list *, int, int); extern void Dbg_file_bind_entry(Lm_list *, Bnd_desc *); extern void Dbg_file_bindings(Rt_map *, int); +extern void Dbg_file_bindings_done(Lm_list *); extern void Dbg_file_cleanup(Lm_list *, const char *, Aliste); extern void Dbg_file_cntl(Lm_list *, Aliste, Aliste); extern void Dbg_file_config_dis(Lm_list *, const char *, int); @@ -676,11 +852,9 @@ extern void Dbg_file_config_obj(Lm_list *, const char *, const char *, const char *); extern void Dbg_file_del_rescan(Lm_list *); extern void Dbg_file_delete(Rt_map *); -extern void Dbg_file_dlclose(Lm_list *, const char *, int); -extern void Dbg_file_dldump(Rt_map *, const char *, int); -extern void Dbg_file_dlopen(Rt_map *, const char *, int *, int); -extern void Dbg_file_elf(Lm_list *, const char *, ulong_t, ulong_t, - ulong_t, ulong_t, const char *, Aliste); +extern void Dbg_file_deferred(Lm_list *, const char *, const char *); +extern void Dbg_file_elf(Lm_list *, const char *, Addr, size_t, + const char *, Aliste); extern void Dbg_file_filtee(Lm_list *, const char *, const char *, int); extern void Dbg_file_filter(Lm_list *, const char *, const char *, int); extern void Dbg_file_fixname(Lm_list *, const char *, const char *); @@ -691,6 +865,8 @@ extern void Dbg_file_hdl_title(int); extern void Dbg_file_lazyload(Rt_map *, const char *, const char *); extern void Dbg_file_ldso(Rt_map *, char **, auxv_t *, const char *, Aliste); +extern void Dbg_file_mmapobj(Lm_list *, const char *, mmapobj_result_t *, + uint_t); extern void Dbg_file_mode_promote(Rt_map *, int); extern void Dbg_file_modified(Lm_list *, const char *, const char *, const char *, int, int, Elf *, Elf *); @@ -707,32 +883,42 @@ extern void Dbg_got_display(Ofl_desc *, Off, int, Word, size_t); extern void Dbg_libs_audit(Lm_list *, const char *, const char *); extern void Dbg_libs_find(Lm_list *, const char *); extern void Dbg_libs_found(Lm_list *, const char *, int); -extern void Dbg_libs_ignore(Lm_list *, const char *); -extern void Dbg_libs_init(Lm_list *, List *, List *); +extern void Dbg_libs_insecure(Lm_list *, const char *, int); +extern void Dbg_libs_init(Lm_list *, APlist *, APlist *); extern void Dbg_libs_l(Lm_list *, const char *, const char *); extern void Dbg_libs_path(Lm_list *, const char *, uint_t, const char *); extern void Dbg_libs_req(Lm_list *, const char *, const char *, const char *); -extern void Dbg_libs_update(Lm_list *, List *, List *); +extern void Dbg_libs_update(Lm_list *, APlist *, APlist *); extern void Dbg_libs_yp(Lm_list *, const char *); extern void Dbg_libs_ylu(Lm_list *, const char *, const char *, int); -extern void Dbg_map_dash(Lm_list *, const char *, Sdf_desc *); -extern void Dbg_map_ent(Lm_list *, Boolean, Ent_desc *, Ofl_desc *); -extern void Dbg_map_parse(Lm_list *, const char *); -extern void Dbg_map_pipe(Lm_list *, Sg_desc *, const char *, const Word); -extern void Dbg_map_seg(Ofl_desc *, int, Sg_desc *); -extern void Dbg_map_set_atsign(Boolean); -extern void Dbg_map_set_equal(Boolean); -extern void Dbg_map_size_new(Lm_list *, const char *); -extern void Dbg_map_size_old(Ofl_desc *, Sym_desc *); -extern void Dbg_map_sort_fini(Lm_list *, Sg_desc *); -extern void Dbg_map_sort_orig(Lm_list *, Sg_desc *); +extern void Dbg_map_cexp_id(Lm_list *, Boolean, const char *, Lineno, + const char *); +extern void Dbg_map_dv(Lm_list *, const char *, Lineno); +extern void Dbg_map_dv_entry(Lm_list *, Lineno, int, const char *); +extern void Dbg_map_ent(Lm_list *, Ent_desc *, Ofl_desc *, Lineno); +extern void Dbg_map_ent_ord_title(Lm_list *, const char *); +extern void Dbg_map_hdr_noalloc(Lm_list *, Lineno); +extern void Dbg_map_parse(Lm_list *, const char *, int); +extern void Dbg_map_pass(Lm_list *, Boolean, const char *, Lineno, + const char *); +extern void Dbg_map_post_title(Lm_list *); +extern void Dbg_map_seg(Ofl_desc *, dbg_state_t, int, Sg_desc *, Lineno); +extern void Dbg_map_seg_order(Ofl_desc *, uchar_t, Half, dbg_state_t, + Lineno); +extern void Dbg_map_seg_os_order(Lm_list *, Sg_desc *, const char *, + Word, Lineno); +extern void Dbg_map_size_new(Lm_list *, const char *, const char *, Lineno); +extern void Dbg_map_size_old(Ofl_desc *, Sym_desc *, const char *, Lineno); +extern void Dbg_map_sort_title(Lm_list *, Boolean); +extern void Dbg_map_sort_seg(Lm_list *, uchar_t, Half, Sg_desc *); extern void Dbg_map_symbol(Ofl_desc *, Sym_desc *); extern void Dbg_map_version(Lm_list *, const char *, const char *, int); extern void Dbg_move_adjexpandreloc(Lm_list *, Xword, const char *); extern void Dbg_move_adjmovereloc(Lm_list *, Xword, Xword, const char *); +extern void Dbg_move_bad(Lm_list *, ulong_t, const char *, Addr); extern void Dbg_move_data(Rt_map *); extern void Dbg_move_entry1(Lm_list *, int, Move *, Sym_desc *); extern void Dbg_move_entry2(Lm_list *, Move *, Word, const char *); @@ -747,8 +933,8 @@ extern void Dbg_reloc_apply_val(Lm_list *, int, Xword, Xword); extern void Dbg_reloc_ars_entry(Lm_list *, int, Word, Half, Rel_desc *); extern void Dbg_reloc_copy(Rt_map *, Rt_map *, const char *, int); extern void Dbg_reloc_discard(Lm_list *, Half, Rel_desc *); -extern void Dbg_reloc_doact(Lm_list *, int, Half, Word, Word, Xword, Xword, - const char *, Os_desc *); +extern void Dbg_reloc_doact(Lm_list *, int, Half, Word, Rel_desc *, + Xword, Xword, rel_desc_sname_func_t); extern void Dbg_reloc_doact_title(Lm_list *); extern void Dbg_reloc_dooutrel(Lm_list *, Word); extern void Dbg_reloc_entry(Lm_list *, const char *, Half, Word, void *, @@ -757,35 +943,40 @@ extern void Dbg_reloc_error(Lm_list *, int, Half, Word, void *, const char *); extern void Dbg_reloc_generate(Lm_list *, Os_desc *, Word); extern void Dbg_reloc_in(Lm_list *, int, Half, Word, void *, const char *, - const char *); + Word, const char *); extern void Dbg_reloc_ors_entry(Lm_list *, int, Word, Half, Rel_desc *); extern void Dbg_reloc_out(Ofl_desc *, int, Word, void *, const char *, const char *); extern void Dbg_reloc_proc(Lm_list *, Os_desc *, Is_desc *, Is_desc *); extern void Dbg_reloc_run(Rt_map *, uint_t, int, int); -extern void Dbg_reloc_transition(Lm_list *, Half, Word, Rel_desc *); -extern void Dbg_reloc_sloppycomdat(Lm_list *, const char *, Sym_desc *); +extern void Dbg_reloc_transition(Lm_list *, Half, Word, Rel_desc *, + rel_desc_sname_func_t); +extern void Dbg_reloc_sloppycomdat(Lm_list *, Sym_desc *); extern void Dbg_sec_added(Lm_list *, Os_desc *, Sg_desc *); +extern void Dbg_sec_backing(Lm_list *); extern void Dbg_sec_created(Lm_list *, Os_desc *, Sg_desc *); extern void Dbg_sec_discarded(Lm_list *, Is_desc *, Is_desc *); extern void Dbg_sec_genstr_compress(Lm_list *, const char *, Xword, Xword); extern void Dbg_sec_group(Lm_list *, Is_desc *, Group_desc *); +extern void Dbg_sec_gnu_comdat(Lm_list *, Is_desc *, Boolean, Boolean); extern void Dbg_sec_in(Lm_list *, Is_desc *); extern void Dbg_sec_order_error(Lm_list *, Ifl_desc *, Word, int); extern void Dbg_sec_order_list(Ofl_desc *, int); +extern void Dbg_sec_redirected(Lm_list *, Is_desc *, const char *); extern void Dbg_sec_strtab(Lm_list *, Os_desc *, Str_tbl *); extern void Dbg_sec_unsup_strmerge(Lm_list *, Is_desc *); -extern void Dbg_seg_desc_entry(Lm_list *, Half, int, Sg_desc *); +extern void Dbg_seg_desc_entry(Lm_list *, uchar_t, Half, int, Sg_desc *, + Boolean); extern void Dbg_seg_entry(Ofl_desc *, int, Sg_desc *); -extern void Dbg_seg_list(Lm_list *, Half, List *); +extern void Dbg_seg_list(Lm_list *, uchar_t, Half, APlist *); extern void Dbg_seg_os(Ofl_desc *, Os_desc *, int); extern void Dbg_seg_title(Lm_list *); -extern void Dbg_shdr_modified(Lm_list *, const char *, Half, Shdr *, Shdr *, - const char *); +extern void Dbg_shdr_modified(Lm_list *, const char *, uchar_t, Half, + Word, Shdr *, Shdr *, const char *); extern void Dbg_statistics_ar(Ofl_desc *); extern void Dbg_statistics_ld(Ofl_desc *); @@ -794,21 +985,29 @@ extern void Dbg_support_action(Lm_list *, const char *, const char *, Support_ndx, const char *); extern void Dbg_support_load(Lm_list *, const char *, const char *); extern void Dbg_support_req(Lm_list *, const char *, int); +extern void Dbg_support_vnone(Lm_list *, const char *); extern void Dbg_syminfo_entry(Lm_list *, Word, Syminfo *, Sym *, const char *, Dyn *); extern void Dbg_syminfo_title(Lm_list *); -extern void Dbg_syms_ar_checking(Lm_list *, Xword, Elf_Arsym *, - const char *); -extern void Dbg_syms_ar_entry(Lm_list *, Xword, Elf_Arsym *); -extern void Dbg_syms_ar_resolve(Lm_list *, Xword, Elf_Arsym *, - const char *, int); -extern void Dbg_syms_ar_title(Lm_list *, const char *, int); +extern void Dbg_syms_ar_checking(Lm_list *, const char *, const char *, + Elf_Arsym *); +extern void Dbg_syms_ar_force(Lm_list *, const char *, const char *); +extern void Dbg_syms_ar_resolve(Lm_list *, const char *, const char *, + Elf_Arsym *); +extern void Dbg_syms_ar_skip(Lm_list *, const char *, Elf_Arsym *); +extern void Dbg_syms_ar_title(Lm_list *, const char *, Boolean); +extern void Dbg_syms_cap_convert(Ofl_desc *, Word, const char *, Sym *); +extern void Dbg_syms_cap_local(Ofl_desc *, Word, const char *, Sym *, + Sym_desc *); +extern void Dbg_syms_cap_lookup(Rt_map *, uint_t, const char *, uint_t, + Half, Syscapset *); +extern void Dbg_syms_cap_title(Ofl_desc *); +extern void Dbg_syms_copy_reloc(Ofl_desc *, Sym_desc *, Word); extern void Dbg_syms_created(Lm_list *, const char *); extern void Dbg_syms_discarded(Lm_list *, Sym_desc *); -extern void Dbg_syms_dlsym(Rt_map *, const char *, int *, const char *, - int); +extern void Dbg_syms_dup_discarded(Lm_list *, Word ndx, Sym_desc *); extern void Dbg_syms_dup_sort_addr(Lm_list *, const char *, const char *, const char *, Addr); extern void Dbg_syms_entered(Ofl_desc *, Sym *, Sym_desc *); @@ -826,7 +1025,6 @@ extern void Dbg_syms_old(Ofl_desc *, Sym_desc *); extern void Dbg_syms_process(Lm_list *, Ifl_desc *); extern void Dbg_syms_reduce(Ofl_desc *, int, Sym_desc *, int, const char *); -extern void Dbg_syms_reloc(Ofl_desc *, Sym_desc *); extern void Dbg_syms_resolved(Ofl_desc *, Sym_desc *); extern void Dbg_syms_resolving(Ofl_desc *, Word, const char *, int, int, Sym *, Sym *, Sym_desc *, Ifl_desc *); @@ -835,12 +1033,12 @@ extern void Dbg_syms_sec_title(Lm_list *); extern void Dbg_syms_spec_title(Lm_list *); extern void Dbg_syms_updated(Ofl_desc *, Sym_desc *, const char *); extern void Dbg_syms_up_title(Lm_list *); +extern void Dbg_syms_wrap(Lm_list *, Word, const char *, const char *); extern void Dbg_tls_modactivity(Lm_list *, void *, uint_t); extern void Dbg_tls_static_block(Lm_list *, void *, ulong_t, ulong_t); extern void Dbg_tls_static_resv(Rt_map *, ulong_t, ulong_t); -extern void Dbg_util_broadcast(Rt_map *); extern void Dbg_util_call_array(Rt_map *, void *, int, Word); extern void Dbg_util_call_fini(Rt_map *); extern void Dbg_util_call_init(Rt_map *, int); @@ -854,10 +1052,9 @@ extern void Dbg_util_intoolate(Rt_map *); extern void Dbg_util_lcinterface(Rt_map *, int, char *); extern void Dbg_util_nl(Lm_list *, int); extern void Dbg_util_no_init(Rt_map *); -extern void Dbg_util_str(Lm_list *, const char *); extern void Dbg_util_scc_entry(Rt_map *, uint_t); extern void Dbg_util_scc_title(Lm_list *, int); -extern void Dbg_util_wait(Rt_map *, Rt_map *, int); +extern void Dbg_util_str(Lm_list *, const char *); extern void Dbg_unused_file(Lm_list *, const char *, int, uint_t); extern void Dbg_unused_lcinterface(Rt_map *, Rt_map *, int); @@ -870,6 +1067,7 @@ extern void Dbg_ver_avail_entry(Lm_list *, Ver_index *, const char *); extern void Dbg_ver_avail_title(Lm_list *, const char *); extern void Dbg_ver_def_title(Lm_list *, const char *); extern void Dbg_ver_desc_entry(Lm_list *, Ver_desc *); +extern void Dbg_ver_need_done(Lm_list *); extern void Dbg_ver_need_entry(Lm_list *, Half, const char *, const char *); extern void Dbg_ver_need_title(Lm_list *, const char *); @@ -882,6 +1080,7 @@ extern void Dbg_ver_symbol(Lm_list *, const char *); #define ELF_DBG_ELFDUMP 1 #define ELF_DBG_RTLD 2 #define ELF_DBG_LD 3 +#define ELF_DBG_LD_ACT 4 /* * Define generic Elf_*() interfaces. @@ -967,12 +1166,14 @@ extern void Elf_syminfo_title(Lm_list *); #endif -extern void Elf_cap_entry(Lm_list *, Cap *, int, Half); +extern void Elf_cap_entry(Lm_list *, Cap *, int, const char *, size_t, + Half); extern void Elf_cap_title(Lm_list *); extern const char \ *Elf_demangle_name(const char *); -extern void Elf_dyn_entry(Lm_list *, Dyn *, int, const char *, Half); +extern void Elf_dyn_entry(Lm_list *, Dyn *, int, const char *, + uchar_t, Half); extern void Elf_dyn_null_entry(Lm_list *, Dyn *, int, int); extern void Elf_dyn_title(Lm_list *); @@ -982,7 +1183,7 @@ extern void Elf_got_entry(Lm_list *, Sword, Addr, Xword, Half, uchar_t, uchar_t, Word, void *, const char *); extern void Elf_got_title(Lm_list *); -extern void Elf_phdr(Lm_list *, Half, Phdr *); +extern void Elf_phdr(Lm_list *, uchar_t, Half, Phdr *); extern void Elf_reloc_apply_val(Lm_list *, int, Xword, Xword); extern void Elf_reloc_apply_reg(Lm_list *, int, Half, Xword, Xword); @@ -993,10 +1194,10 @@ extern void Elf_reloc_entry_2(Lm_list *, int, const char *, Word, const char *); extern void Elf_reloc_title(Lm_list *, int, Word); -extern void Elf_shdr(Lm_list *, Half, Shdr *); +extern void Elf_shdr(Lm_list *, uchar_t, Half, Shdr *); -extern void Elf_syms_table_entry(Lm_list *, int, const char *, Half, Sym *, - Versym, int, const char *, const char *); +extern void Elf_syms_table_entry(Lm_list *, int, const char *, uchar_t, + Half, Sym *, Versym, int, const char *, const char *); extern void Elf_syms_table_title(Lm_list *, int); extern void Elf_ver_def_title(Lm_list *); diff --git a/cmd/sgs/include/sgs.h b/cmd/sgs/include/sgs.h index 388ec978ec02..fc6248107a06 100644 --- a/cmd/sgs/include/sgs.h +++ b/cmd/sgs/include/sgs.h @@ -24,8 +24,7 @@ * All Rights Reserved * * - * Copyright 2008 Sun Microsystems, Inc. All rights reserved. - * Use is subject to license terms. + * Copyright (c) 1989, 2010, Oracle and/or its affiliates. All rights reserved. * * Global include file for all sgs. */ @@ -33,9 +32,6 @@ #ifndef _SGS_H #define _SGS_H -#pragma ident "%Z%%M% %I% %E% SMI" - - #ifdef __cplusplus extern "C" { #endif @@ -51,6 +47,7 @@ extern "C" { #include <sys/types.h> #include <sys/machelf.h> #include <stdlib.h> +#include <stdarg.h> #include <libelf.h> #include <assert.h> #include <alist.h> @@ -122,17 +119,27 @@ typedef enum { } Boolean; /* - * Types of errors (used by eprintf()), together with a generic error return + * Types of errors (used by veprintf()), together with a generic error return * value. */ typedef enum { - ERR_NONE, - ERR_WARNING, - ERR_FATAL, - ERR_ELF, - ERR_NUM /* Must be last */ + ERR_NONE, /* plain message */ + ERR_WARNING_NF, /* warning that cannot be promoted to fatal */ + ERR_WARNING, /* warning that can be promoted to fatal */ + ERR_GUIDANCE, /* guidance warning that can be promoted */ + ERR_FATAL, /* fatal error */ + ERR_ELF, /* fatal libelf error */ + ERR_NUM /* # of Error codes. Must be last */ } Error; +/* + * Type used to represent line numbers within files, and a corresponding + * printing macro for it. + */ +typedef ulong_t Lineno; +#define EC_LINENO(_x) EC_XWORD(_x) /* "llu" */ + + #if defined(_LP64) && !defined(_ELF64) #define S_ERROR (~(uint_t)0) #else @@ -140,48 +147,16 @@ typedef enum { #endif /* - * LIST_TRAVERSE() is used as the only "argument" of a "for" loop to - * traverse a linked list. The node pointer `node' is set to each node in - * turn and the corresponding data pointer is copied to `data'. The macro - * is used as in - * for (LIST_TRAVERSE(List *list, Listnode *node, void *data)) { - * process(data); - * } + * CTF currently does not handle automatic array variables sized via function + * arguments (VLA arrays) properly, when the code is compiled with gcc. + * Adding 1 to the size is a workaround. VLA_SIZE, and its use, should be + * pulled when CTF is fixed or replaced. */ -#define LIST_TRAVERSE(L, N, D) \ - (void) (((N) = (L)->head) != NULL && ((D) = (N)->data) != NULL); \ - (N) != NULL; \ - (void) (((N) = (N)->next) != NULL && ((D) = (N)->data) != NULL) - -typedef struct listnode Listnode; -typedef struct list List; - -struct listnode { /* a node on a linked list */ - void *data; /* the data item */ - Listnode *next; /* the next element */ -}; - -struct list { /* a linked list */ - Listnode *head; /* the first element */ - Listnode *tail; /* the last element */ -}; - - -#ifdef _SYSCALL32 -typedef struct listnode32 Listnode32; -typedef struct list32 List32; - -struct listnode32 { /* a node on a linked list */ - Elf32_Addr data; /* the data item */ - Elf32_Addr next; /* the next element */ -}; - -struct list32 { /* a linked list */ - Elf32_Addr head; /* the first element */ - Elf32_Addr tail; /* the last element */ -}; -#endif /* _SYSCALL32 */ - +#ifdef __GNUC__ +#define VLA_SIZE(_arg) ((_arg) + 1) +#else +#define VLA_SIZE(_arg) (_arg) +#endif /* * Structure to maintain rejected files elf information. Files that are not @@ -192,7 +167,7 @@ struct list32 { /* a linked list */ */ typedef struct { ushort_t rej_type; /* SGS_REJ_ value */ - ushort_t rej_flag; /* additional information */ + ushort_t rej_flags; /* additional information */ uint_t rej_info; /* numeric and string information */ const char *rej_str; /* associated with error */ const char *rej_name; /* object name - expanded library */ @@ -212,7 +187,16 @@ typedef struct { /* required */ #define SGS_REJ_STR 10 /* generic error - info is a string */ #define SGS_REJ_UNKFILE 11 /* unknown file type */ -#define SGS_REJ_HWCAP_1 12 /* hardware capabilities mismatch */ +#define SGS_REJ_UNKCAP 12 /* unknown capabilities */ +#define SGS_REJ_HWCAP_1 13 /* hardware capabilities mismatch */ +#define SGS_REJ_SFCAP_1 14 /* software capabilities mismatch */ +#define SGS_REJ_MACHCAP 15 /* machine capability mismatch */ +#define SGS_REJ_PLATCAP 16 /* platform capability mismatch */ +#define SGS_REJ_HWCAP_2 17 /* hardware capabilities mismatch */ +#define SGS_REJ_ARCHIVE 18 /* archive used in invalid context */ +#define SGS_REJ_NUM 19 + +#define FLG_REJ_ALTER 0x01 /* object name is an alternative */ /* * For those source files used both inside and outside of the @@ -227,44 +211,42 @@ typedef struct { #define realloc libld_realloc #define libld_calloc(x, a) libld_malloc(((size_t)x) * ((size_t)a)) -extern void libld_free(void *); -extern void *libld_malloc(size_t); -extern void *libld_realloc(void *, size_t); +extern void libld_free(void *); +extern void *libld_malloc(size_t); +extern void *libld_realloc(void *, size_t); #endif - /* * Data structures (defined in libld.h). */ +typedef struct audit_desc Audit_desc; +typedef struct audit_info Audit_info; +typedef struct audit_list Audit_list; +typedef struct cap_desc Cap_desc; typedef struct ent_desc Ent_desc; typedef struct group_desc Group_desc; typedef struct ifl_desc Ifl_desc; typedef struct is_desc Is_desc; typedef struct isa_desc Isa_desc; typedef struct isa_opt Isa_opt; -typedef struct mv_desc Mv_desc; -typedef struct ofl_desc Ofl_desc; typedef struct os_desc Os_desc; +typedef struct ofl_desc Ofl_desc; typedef struct rel_cache Rel_cache; +typedef struct rel_cachebuf Rel_cachebuf; +typedef struct rel_aux_cachebuf Rel_aux_cachebuf; +typedef struct rel_aux Rel_aux; +typedef struct rel_desc Rel_desc; typedef struct sdf_desc Sdf_desc; typedef struct sdv_desc Sdv_desc; +typedef struct sec_order Sec_order; typedef struct sg_desc Sg_desc; typedef struct sort_desc Sort_desc; -typedef struct sec_order Sec_order; -typedef struct sym_desc Sym_desc; -typedef struct sym_aux Sym_aux; typedef struct sym_avlnode Sym_avlnode; +typedef struct sym_aux Sym_aux; +typedef struct sym_desc Sym_desc; typedef struct uts_desc Uts_desc; typedef struct ver_desc Ver_desc; typedef struct ver_index Ver_index; -typedef struct audit_desc Audit_desc; -typedef struct audit_info Audit_info; -typedef struct audit_list Audit_list; - -/* - * Data structures defined in machrel.h. - */ -typedef struct rel_desc Rel_desc; /* * Data structures defined in rtld.h. @@ -279,7 +261,7 @@ typedef struct lm_list32 Lm_list32; */ extern int assfail(const char *, const char *, int); extern void eprintf(Lm_list *, Error, const char *, ...); -extern char *sgs_demangle(char *); +extern void veprintf(Lm_list *, Error, const char *, va_list); extern uint_t sgs_str_hash(const char *); extern uint_t findprime(uint_t); diff --git a/cmd/sgs/messages/sgs.ident b/cmd/sgs/messages/sgs.ident index 6afbf5ff5e26..9aa369693f66 100644 --- a/cmd/sgs/messages/sgs.ident +++ b/cmd/sgs/messages/sgs.ident @@ -1,6 +1,5 @@ # -# Copyright 2008 Sun Microsystems, Inc. All rights reserved. -# Use is subject to license terms. +# Copyright (c) 1995, 2010, Oracle and/or its affiliates. All rights reserved. # # CDDL HEADER START # @@ -21,8 +20,6 @@ # # CDDL HEADER END # -# ident "%Z%%M% %I% %E% SMI" -# # # Global message identifiers for the sgs utilities. This information is read # by sgsmsg(1l) using the -i option. @@ -37,7 +34,7 @@ MSG_ID_RTLD 1 SUNW_OST_SGS /* sgs/rtld */ MSG_ID_LIBRTLD 2 SUNW_OST_SGS /* sgs/librtld */ MSG_ID_LIBLD 3 SUNW_OST_SGS /* sgs/libld */ MSG_ID_LIBLDDBG 4 SUNW_OST_SGS /* sgs/liblddbg */ -MSG_ID_LIBLDSTAB 5 SUNW_OST_SGS /* sgs/libldstab */ + MSG_ID_LIBRTLD_DB 6 SUNW_OST_SGS /* sgs/librtld_db */ MSG_ID_LIBPROF 7 SUNW_OST_SGS /* sgs/libprof */ MSG_ID_LIBCRLE 8 SUNW_OST_SGS /* sgs/libcrle */ @@ -60,3 +57,4 @@ MSG_ID_ELFEDIT_STR 27 SUNW_OST_SGS /* str: */ MSG_ID_ELFEDIT_SYM 27 SUNW_OST_SGS /* sym: */ MSG_ID_ELFEDIT_SYMINFO 27 SUNW_OST_SGS /* syminfo: */ MSG_ID_ELFWRAP 28 SUNW_OST_SGS /* sgs/elfwrap */ +MSG_ID_AR 29 SUNW_OST_SGS /* sgs/ar */ diff --git a/cmd/sgs/tools/common/sgsmsg.c b/cmd/sgs/tools/common/sgsmsg.c index 9b2e37b507d6..1cdd584c7427 100644 --- a/cmd/sgs/tools/common/sgsmsg.c +++ b/cmd/sgs/tools/common/sgsmsg.c @@ -19,8 +19,7 @@ * CDDL HEADER END */ /* - * Copyright 2008 Sun Microsystems, Inc. All rights reserved. - * Use is subject to license terms. + * Copyright (c) 1995, 2010, Oracle and/or its affiliates. All rights reserved. * * sgsmsg generates several message files from an input template file. Messages * are constructed for use with gettext(3i) - the default - or catgets(3c). The @@ -66,7 +65,6 @@ * the data array being built in msg.c. The index into this array * becomes the `message' identifier created in the msg.h file. */ -#pragma ident "%Z%%M% %I% %E% SMI" #include <fcntl.h> #include <stdlib.h> @@ -393,15 +391,22 @@ init_defs(void) return (1); } + if (fprintf(fddefs, "#include <sgsmsg.h>\t/* Msg typedef */\n\n") < 0) { + (void) fprintf(stderr, Errmsg_wrte, fldefs, strerror(errno)); + return (1); + } + if (fprintf(fddefs, "#ifndef\t__lint\n\n") < 0) { (void) fprintf(stderr, Errmsg_wrte, fldefs, strerror(errno)); return (1); } /* - * add "typedef int Msg;" + * The MSG_SGS_ARRAY_NAME macro supplies a generic way to + * reference the string table regardless of its name. */ - if (fprintf(fddefs, "typedef int\tMsg;\n\n") < 0) { + if (fprintf(fddefs, "#define\tMSG_SGS_LOCAL_ARRAY\t__%s\n\n", + interface) < 0) { (void) fprintf(stderr, Errmsg_wrte, fldefs, strerror(errno)); return (1); } @@ -418,7 +423,13 @@ init_defs(void) return (1); } } - if (fprintf(fddefs, "#define\tMSG_ORIG(x)\t&__%s[x]\n\n", + if (fprintf(fddefs, + "#define\tMSG_ORIG_STRTAB(_x, _s)\t&_s[_x]\n\n") < 0) { + (void) fprintf(stderr, Errmsg_wrte, fldefs, strerror(errno)); + return (1); + } + if (fprintf(fddefs, + "#define\tMSG_ORIG(x)\tMSG_ORIG_STRTAB(x, __%s)\n\n", interface) < 0) { (void) fprintf(stderr, Errmsg_wrte, fldefs, strerror(errno)); return (1); @@ -453,17 +464,14 @@ fini_defs(void) return (1); } - /* - * When __lint is defined, Msg is a char *. This allows lint to - * check our format strings against it's arguments. - */ - if (fprintf(fddefs, "\ntypedef char *\tMsg;\n\n") < 0) { + if (fprintf(fddefs, "extern\tconst char *\t_%s(Msg);\n\n", + interface) < 0) { (void) fprintf(stderr, Errmsg_wrte, fldefs, strerror(errno)); return (1); } - if (fprintf(fddefs, "extern\tconst char *\t_%s(Msg);\n\n", - interface) < 0) { + if (fprintf(fddefs, "#ifndef MSG_SGS_LOCAL_ARRAY\n" + "#define\tMSG_SGS_LOCAL_ARRAY\t\"\"\n#endif\n\n") < 0) { (void) fprintf(stderr, Errmsg_wrte, fldefs, strerror(errno)); return (1); } @@ -478,12 +486,25 @@ fini_defs(void) } if (fprintf(fddefs, + "#define MSG_ORIG_STRTAB(_x, _s)\t_x\n" "#define MSG_ORIG(x)\tx\n#define MSG_INTL(x)\tx\n") < 0) { (void) fprintf(stderr, Errmsg_wrte, fldefs, strerror(errno)); return (1); } /* + * Provide a way to get the array and function declarations above + * without also getting the actual messages. This is useful in + * our lintsup.c files that include more than one message header. + * lintsup doesn't need the actual messages, and this prevents + * macro name collisions. + */ + if (fprintf(fddefs, "\n#ifndef LINTSUP_SUPPRESS_STRINGS\n") < 0) { + (void) fprintf(stderr, Errmsg_wrte, fldefs, strerror(errno)); + return (1); + } + + /* * Copy the temporary lint defs file into the new header. */ if (fdlint) { @@ -510,6 +531,11 @@ fini_defs(void) (void) free(buf); } + if (fprintf(fddefs, "\n#endif\t/* LINTSUP_SUPPRESS_STRINGS */\n") < 0) { + (void) fprintf(stderr, Errmsg_wrte, fldefs, strerror(errno)); + return (1); + } + if (fprintf(fddefs, "\n#endif\t/* __lint */\n") < 0) { (void) fprintf(stderr, Errmsg_wrte, fldefs, strerror(errno)); return (1); @@ -1102,8 +1128,9 @@ main(int argc, char ** argv) } } if (fddefs && fddata) { - (void) sprintf(fllint, "%s.%d", nmlint, (int)getpid()); - if ((fdlint = fopen(fllint, "w+")) == NULL) { + (void) sprintf(fllint, "%s.%d.XXXXXX", nmlint, (int)getpid()); + if ((mkstemp(fllint) == -1) || + ((fdlint = fopen(fllint, "w+")) == NULL)) { (void) fprintf(stderr, Errmsg_opne, fllint, strerror(errno)); return (1); diff --git a/cmd/sgs/tools/common/string_table.c b/cmd/sgs/tools/common/string_table.c index e174acaf0419..c15473150e4c 100644 --- a/cmd/sgs/tools/common/string_table.c +++ b/cmd/sgs/tools/common/string_table.c @@ -20,12 +20,10 @@ */ /* - * Copyright 2008 Sun Microsystems, Inc. All rights reserved. + * Copyright 2009 Sun Microsystems, Inc. All rights reserved. * Use is subject to license terms. */ -#pragma ident "%Z%%M% %I% %E% SMI" - #include <_string_table.h> #include <strings.h> #include <sgs.h> @@ -160,7 +158,7 @@ st_new(uint_t flags) { Str_tbl *stp; - if ((stp = calloc(sizeof (Str_tbl), 1)) == NULL) + if ((stp = calloc(sizeof (*stp), 1)) == NULL) return (NULL); /* @@ -175,7 +173,7 @@ st_new(uint_t flags) if ((stp->st_flags & FLG_STTAB_COMPRESS) == 0) return (stp); - if ((stp->st_lentree = calloc(sizeof (avl_tree_t), 1)) == NULL) + if ((stp->st_lentree = calloc(sizeof (*stp->st_lentree), 1)) == NULL) return (NULL); avl_create(stp->st_lentree, &avl_len_compare, sizeof (LenNode), @@ -187,9 +185,9 @@ st_new(uint_t flags) /* * Insert a new string into the Str_tbl. There are two AVL trees used. * - * . The first LenNode AVL tree maintains a tree of nodes based on string + * - The first LenNode AVL tree maintains a tree of nodes based on string * sizes. - * . Each LenNode maintains a StrNode AVL tree for each string. Large + * - Each LenNode maintains a StrNode AVL tree for each string. Large * applications have been known to contribute thousands of strings of * the same size. Should strings need to be removed (-z ignore), then * the string AVL tree makes this removal efficient and scalable. @@ -227,12 +225,13 @@ st_insert(Str_tbl *stp, const char *str) */ ln.ln_strlen = len; if ((lnp = avl_find(stp->st_lentree, &ln, &where)) == NULL) { - if ((lnp = calloc(sizeof (LenNode), 1)) == NULL) + if ((lnp = calloc(sizeof (*lnp), 1)) == NULL) return (-1); lnp->ln_strlen = len; avl_insert(stp->st_lentree, lnp, where); - if ((lnp->ln_strtree = calloc(sizeof (avl_tree_t), 1)) == NULL) + if ((lnp->ln_strtree = calloc(sizeof (*lnp->ln_strtree), 1)) == + NULL) return (0); avl_create(lnp->ln_strtree, &avl_str_compare, sizeof (StrNode), @@ -246,7 +245,7 @@ st_insert(Str_tbl *stp, const char *str) */ sn.sn_str = str; if ((snp = avl_find(lnp->ln_strtree, &sn, &where)) == NULL) { - if ((snp = calloc(sizeof (StrNode), 1)) == NULL) + if ((snp = calloc(sizeof (*snp), 1)) == NULL) return (-1); snp->sn_str = str; avl_insert(lnp->ln_strtree, snp, where); @@ -513,7 +512,7 @@ st_hash_insert(Str_tbl *stp, const char *str, size_t len) /* * allocate a new master string */ - if ((mstr = calloc(sizeof (Str_hash), 1)) == 0) + if ((mstr = calloc(sizeof (*mstr), 1)) == 0) return (-1); mstr->sm_next = stp->st_mstrlist; stp->st_mstrlist = mstr; @@ -528,7 +527,7 @@ st_hash_insert(Str_tbl *stp, const char *str, size_t len) stp->st_strsize += len - mstr->sm_strlen; } - if ((sthash = calloc(sizeof (Str_hash), 1)) == 0) + if ((sthash = calloc(sizeof (*sthash), 1)) == 0) return (-1); mstr->sm_hashval = sthash->hi_hashval = hashval; @@ -569,8 +568,8 @@ st_getstrtab_sz(Str_tbl *stp) * strings input. */ stp->st_hbckcnt = findprime(stp->st_strcnt); - if ((stp->st_hashbcks = - calloc(sizeof (Str_hash), stp->st_hbckcnt)) == NULL) + if ((stp->st_hashbcks = calloc(sizeof (*stp->st_hashbcks), + stp->st_hbckcnt)) == NULL) return (0); /* @@ -673,8 +672,8 @@ st_setstrbuf(Str_tbl *stp, char *stbuf, size_t bufsize) #ifdef DEBUG /* * for debug builds - start with a stringtable filled in - * with '0xff'. This makes it very easy to find wholes - * which we failed to fill in - in the strtab. + * with '0xff'. This makes it very easy to spot unfilled + * holes in the strtab. */ memset(stbuf, 0xff, bufsize); stbuf[0] = '\0'; diff --git a/cmd/stat/common/statcommon.h b/cmd/stat/common/statcommon.h new file mode 100644 index 000000000000..9ee077431ddb --- /dev/null +++ b/cmd/stat/common/statcommon.h @@ -0,0 +1,328 @@ +/* + * 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 http://www.opensolaris.org/os/licensing. + * 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) 2004, 2010, Oracle and/or its affiliates. All rights reserved. + * + * Common routines for acquiring snapshots of kstats for + * iostat, mpstat, and vmstat. + */ + +#ifndef _STATCOMMON_H +#define _STATCOMMON_H + +#ifdef __cplusplus +extern "C" { +#endif + +#include <stdio.h> +#include <kstat.h> +#include <sys/time.h> +#include <sys/types.h> +#include <sys/buf.h> +#include <sys/dnlc.h> +#include <sys/sysinfo.h> +#include <sys/processor.h> +#include <sys/pset.h> +#include <sys/avl.h> + +/* No CPU present at this CPU position */ +#define ID_NO_CPU -1 +/* CPU belongs to no pset (we number this as "pset 0") */ +#define ID_NO_PSET 0 +/* CPU is usable */ +#define CPU_ONLINE(s) ((s) == P_ONLINE || (s) == P_NOINTR) +/* will the CPU have kstats */ +#define CPU_ACTIVE(c) (CPU_ONLINE((c)->cs_state) && (c)->cs_id != ID_NO_CPU) +/* IO device has no identified ID */ +#define IODEV_NO_ID -1 +/* no limit to iodevs to collect */ +#define UNLIMITED_IODEVS ((size_t)-1) + +#define NODATE 0 /* Default: No time stamp */ +#define DDATE 1 /* Standard date format */ +#define UDATE 2 /* Internal representation of Unix time */ + + +enum snapshot_types { + /* All CPUs separately */ + SNAP_CPUS = 1 << 0, + /* Aggregated processor sets */ + SNAP_PSETS = 1 << 1, + /* sys-wide stats including aggregated CPU stats */ + SNAP_SYSTEM = 1 << 2, + /* interrupt sources and counts */ + SNAP_INTERRUPTS = 1 << 3, + /* disk etc. stats */ + SNAP_IODEVS = 1 << 4, + /* disk controller aggregates */ + SNAP_CONTROLLERS = 1 << 5, + /* mpxio L I (multipath) paths: -X: Lun,LunInitiator */ + SNAP_IOPATHS_LI = 1 << 6, + /* mpxio LTI (multipath) paths: -Y: Lun,LunTarget,LunTargetInitiator */ + SNAP_IOPATHS_LTI = 1 << 7, + /* disk error stats */ + SNAP_IODEV_ERRORS = 1 << 8, + /* pretty names for iodevs */ + SNAP_IODEV_PRETTY = 1 << 9, + /* devid for iodevs */ + SNAP_IODEV_DEVID = 1 << 10 +}; + +struct cpu_snapshot { + /* may be ID_NO_CPU if no CPU present */ + processorid_t cs_id; + /* may be ID_NO_PSET if no pset */ + psetid_t cs_pset_id; + /* as in p_online(2) */ + int cs_state; + /* stats for this CPU */ + kstat_t cs_vm; + kstat_t cs_sys; +}; + +struct pset_snapshot { + /* ID may be zero to indicate the "none set" */ + psetid_t ps_id; + /* number of CPUs in set */ + size_t ps_nr_cpus; + /* the CPUs in this set */ + struct cpu_snapshot **ps_cpus; +}; + +struct intr_snapshot { + /* name of interrupt source */ + char is_name[KSTAT_STRLEN]; + /* total number of interrupts from this source */ + ulong_t is_total; +}; + +struct sys_snapshot { + sysinfo_t ss_sysinfo; + vminfo_t ss_vminfo; + struct nc_stats ss_nc; + /* vm/sys stats aggregated across all CPUs */ + kstat_t ss_agg_vm; + kstat_t ss_agg_sys; + /* ticks since boot */ + ulong_t ss_ticks; + long ss_deficit; +}; + +/* order is significant (see sort_before()) */ +enum iodev_type { + IODEV_CONTROLLER = 1 << 0, + IODEV_DISK = 1 << 1, + IODEV_PARTITION = 1 << 2, + IODEV_TAPE = 1 << 3, + IODEV_NFS = 1 << 4, + IODEV_IOPATH_LT = 1 << 5, /* synthetic LunTarget */ + IODEV_IOPATH_LI = 1 << 6, /* synthetic LunInitiator */ + IODEV_IOPATH_LTI = 1 << 7, /* LunTgtInitiator (pathinfo) */ + IODEV_UNKNOWN = 1 << 8 +}; + +/* identify a disk, partition, etc. */ +struct iodev_id { + int id; + /* target id (for disks) */ + char tid[KSTAT_STRLEN]; +}; + +/* + * Used for disks, partitions, tapes, nfs, controllers, iopaths + * Each entry can be a branch of a tree; for example, the disks + * of a controller constitute the children of the controller + * iodev_snapshot. This relationship is not strictly maintained + * if is_pretty can't be found. + */ +struct iodev_snapshot { + /* original kstat name */ + char is_name[KSTAT_STRLEN]; + /* type of kstat */ + enum iodev_type is_type; + /* ID if meaningful */ + struct iodev_id is_id; + /* parent ID if meaningful */ + struct iodev_id is_parent_id; + /* user-friendly name if found */ + char *is_pretty; + /* device ID if applicable */ + char *is_devid; + /* mount-point if applicable */ + char *is_dname; + /* number of direct children */ + int is_nr_children; + /* children of this I/O device */ + struct iodev_snapshot *is_children; + /* standard I/O stats */ + kstat_io_t is_stats; + /* iodev error stats */ + kstat_t is_errors; + /* creation time of the stats */ + hrtime_t is_crtime; + /* time at which iodev snapshot was taken */ + hrtime_t is_snaptime; + /* kstat module */ + char is_module[KSTAT_STRLEN]; + /* kstat instance */ + int is_instance; + /* kstat (only used temporarily) */ + kstat_t *is_ksp; + struct iodev_snapshot *is_prev; + struct iodev_snapshot *is_next; + /* AVL structures to speedup insertion */ + avl_tree_t *avl_list; /* list this element belongs to */ + avl_node_t avl_link; +}; + +/* which iodevs to show. */ +struct iodev_filter { + /* nr. of iodevs to choose */ + size_t if_max_iodevs; + /* bit mask of enum io_types to allow */ + int if_allowed_types; + /* should we show floppy ? if_names can override this */ + int if_skip_floppy; + /* nr. of named iodevs */ + size_t if_nr_names; + char **if_names; +}; + +/* The primary structure of a system snapshot. */ +struct snapshot { + /* what types were *requested* */ + enum snapshot_types s_types; + size_t s_nr_cpus; + struct cpu_snapshot *s_cpus; + size_t s_nr_psets; + struct pset_snapshot *s_psets; + size_t s_nr_intrs; + struct intr_snapshot *s_intrs; + size_t s_nr_iodevs; + struct iodev_snapshot *s_iodevs; + size_t s_iodevs_is_name_maxlen; + struct sys_snapshot s_sys; + struct biostats s_biostats; +}; + +/* print a message and exit with failure */ +void fail(int do_perror, char *message, ...); + +/* strdup str, or exit with failure */ +char *safe_strdup(char *str); + +/* malloc successfully, or exit with failure */ +void *safe_alloc(size_t size); + +/* + * Copy a kstat from src to dst. If the source kstat contains no data, + * then set the destination kstat data to NULL and size to zero. + * Returns 0 on success. + */ +int kstat_copy(const kstat_t *src, kstat_t *dst); + +/* + * Look up the named kstat, and give the ui64 difference i.e. + * new - old, or if old is NULL, return new. + */ +uint64_t kstat_delta(kstat_t *old, kstat_t *new, char *name); + +/* Return the number of ticks delta between two hrtime_t values. */ +uint64_t hrtime_delta(hrtime_t old, hrtime_t new); + +/* + * Add the integer-valued stats from "src" to the + * existing ones in "dst". If "dst" does not contain + * stats, then a kstat_copy() is performed. + */ +int kstat_add(const kstat_t *src, kstat_t *dst); + +/* return the number of CPUs with kstats (i.e. present and online) */ +int nr_active_cpus(struct snapshot *ss); + +/* + * Return the difference in CPU ticks between the two sys + * kstats. + */ +uint64_t cpu_ticks_delta(kstat_t *old, kstat_t *new); + +/* + * Open the kstat chain. Cannot fail. + */ +kstat_ctl_t *open_kstat(void); + +/* + * Return a struct snapshot based on the snapshot_types parameter + * passed in. iodev_filter may be NULL in which case all iodevs + * are selected if SNAP_IODEVS is passed. + */ +struct snapshot *acquire_snapshot(kstat_ctl_t *, int, struct iodev_filter *); + +/* free a snapshot */ +void free_snapshot(struct snapshot *ss); + +typedef void (*snapshot_cb)(void *old, void *new, void *data); + +/* + * Call the call back for each pair of data items of the given type, + * passing the data pointer passed in as well. If an item has been + * added, the first pointer will be NULL; if removed, the second pointer + * will be NULL. + * + * A non-zero return value indicates configuration has changed. + */ +int snapshot_walk(enum snapshot_types type, struct snapshot *old, + struct snapshot *new, snapshot_cb cb, void *data); + +/* + * Output a line detailing any configuration changes such as a CPU + * brought online, etc, bracketed by << >>. + */ +void snapshot_report_changes(struct snapshot *old, struct snapshot *new); + +/* Return non-zero if configuration has changed. */ +int snapshot_has_changed(struct snapshot *old, struct snapshot *new); + +/* free the given iodev */ +void free_iodev(struct iodev_snapshot *iodev); + +/* acquire the I/O devices */ +int acquire_iodevs(struct snapshot *ss, kstat_ctl_t *kc, + struct iodev_filter *df); + +/* strcmp-style I/O device comparator */ +int iodev_cmp(struct iodev_snapshot *io1, struct iodev_snapshot *io2); + +/* sleep until *wakeup + interval, keeping cadence where desired */ +void sleep_until(hrtime_t *wakeup, hrtime_t interval, int forever, + int *caught_cont); + +/* signal handler - so we can be aware of SIGCONT */ +void cont_handler(int sig_number); + +/* Print a timestamp in either Unix or standard format. */ +void print_timestamp(uint_t); + +#ifdef __cplusplus +} +#endif + +#endif /* _STATCOMMON_H */ diff --git a/cmd/stat/common/timestamp.c b/cmd/stat/common/timestamp.c new file mode 100644 index 000000000000..e0592f3806f7 --- /dev/null +++ b/cmd/stat/common/timestamp.c @@ -0,0 +1,54 @@ +/* + * 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 http://www.opensolaris.org/os/licensing. + * 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 2009 Sun Microsystems, Inc. All rights reserved. + * Use is subject to license terms. + */ + +#include "statcommon.h" + +#include <langinfo.h> + +/* + * Print timestamp as decimal reprentation of time_t value (-T u was specified) + * or in date(1) format (-T d was specified). + */ +void +print_timestamp(uint_t timestamp_fmt) +{ + time_t t = time(NULL); + static char *fmt = NULL; + + /* We only need to retrieve this once per invocation */ + if (fmt == NULL) + fmt = nl_langinfo(_DATE_FMT); + + if (timestamp_fmt == UDATE) { + (void) printf("%ld\n", t); + } else if (timestamp_fmt == DDATE) { + char dstr[64]; + int len; + + len = strftime(dstr, sizeof (dstr), fmt, localtime(&t)); + if (len > 0) + (void) printf("%s\n", dstr); + } +} diff --git a/cmd/zdb/zdb.c b/cmd/zdb/zdb.c new file mode 100644 index 000000000000..c6e219df9e1d --- /dev/null +++ b/cmd/zdb/zdb.c @@ -0,0 +1,3158 @@ +/* + * 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 http://www.opensolaris.org/os/licensing. + * 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. + */ + +#include <stdio.h> +#include <stdio_ext.h> +#include <stdlib.h> +#include <ctype.h> +#include <sys/zfs_context.h> +#include <sys/spa.h> +#include <sys/spa_impl.h> +#include <sys/dmu.h> +#include <sys/zap.h> +#include <sys/fs/zfs.h> +#include <sys/zfs_znode.h> +#include <sys/zfs_sa.h> +#include <sys/sa.h> +#include <sys/sa_impl.h> +#include <sys/vdev.h> +#include <sys/vdev_impl.h> +#include <sys/metaslab_impl.h> +#include <sys/dmu_objset.h> +#include <sys/dsl_dir.h> +#include <sys/dsl_dataset.h> +#include <sys/dsl_pool.h> +#include <sys/dbuf.h> +#include <sys/zil.h> +#include <sys/zil_impl.h> +#include <sys/stat.h> +#include <sys/resource.h> +#include <sys/dmu_traverse.h> +#include <sys/zio_checksum.h> +#include <sys/zio_compress.h> +#include <sys/zfs_fuid.h> +#include <sys/arc.h> +#include <sys/ddt.h> +#undef ZFS_MAXNAMELEN +#undef verify +#include <libzfs.h> + +#define ZDB_COMPRESS_NAME(idx) ((idx) < ZIO_COMPRESS_FUNCTIONS ? \ + zio_compress_table[(idx)].ci_name : "UNKNOWN") +#define ZDB_CHECKSUM_NAME(idx) ((idx) < ZIO_CHECKSUM_FUNCTIONS ? \ + zio_checksum_table[(idx)].ci_name : "UNKNOWN") +#define ZDB_OT_NAME(idx) ((idx) < DMU_OT_NUMTYPES ? \ + dmu_ot[(idx)].ot_name : "UNKNOWN") +#define ZDB_OT_TYPE(idx) ((idx) < DMU_OT_NUMTYPES ? (idx) : DMU_OT_NUMTYPES) + +#ifndef lint +extern int zfs_recover; +#else +int zfs_recover; +#endif + +const char cmdname[] = "zdb"; +uint8_t dump_opt[256]; + +typedef void object_viewer_t(objset_t *, uint64_t, void *data, size_t size); + +extern void dump_intent_log(zilog_t *); +uint64_t *zopt_object = NULL; +int zopt_objects = 0; +libzfs_handle_t *g_zfs; + +/* + * These libumem hooks provide a reasonable set of defaults for the allocator's + * debugging facilities. + */ +const char * +_umem_debug_init() +{ + return ("default,verbose"); /* $UMEM_DEBUG setting */ +} + +const char * +_umem_logging_init(void) +{ + return ("fail,contents"); /* $UMEM_LOGGING setting */ +} + +static void +usage(void) +{ + (void) fprintf(stderr, + "Usage: %s [-CumdibcsDvhL] poolname [object...]\n" + " %s [-div] dataset [object...]\n" + " %s -m [-L] poolname [vdev [metaslab...]]\n" + " %s -R poolname vdev:offset:size[:flags]\n" + " %s -S poolname\n" + " %s -l [-u] device\n" + " %s -C\n\n", + cmdname, cmdname, cmdname, cmdname, cmdname, cmdname, cmdname); + + (void) fprintf(stderr, " Dataset name must include at least one " + "separator character '/' or '@'\n"); + (void) fprintf(stderr, " If dataset name is specified, only that " + "dataset is dumped\n"); + (void) fprintf(stderr, " If object numbers are specified, only " + "those objects are dumped\n\n"); + (void) fprintf(stderr, " Options to control amount of output:\n"); + (void) fprintf(stderr, " -u uberblock\n"); + (void) fprintf(stderr, " -d dataset(s)\n"); + (void) fprintf(stderr, " -i intent logs\n"); + (void) fprintf(stderr, " -C config (or cachefile if alone)\n"); + (void) fprintf(stderr, " -h pool history\n"); + (void) fprintf(stderr, " -b block statistics\n"); + (void) fprintf(stderr, " -m metaslabs\n"); + (void) fprintf(stderr, " -c checksum all metadata (twice for " + "all data) blocks\n"); + (void) fprintf(stderr, " -s report stats on zdb's I/O\n"); + (void) fprintf(stderr, " -D dedup statistics\n"); + (void) fprintf(stderr, " -S simulate dedup to measure effect\n"); + (void) fprintf(stderr, " -v verbose (applies to all others)\n"); + (void) fprintf(stderr, " -l dump label contents\n"); + (void) fprintf(stderr, " -L disable leak tracking (do not " + "load spacemaps)\n"); + (void) fprintf(stderr, " -R read and display block from a " + "device\n\n"); + (void) fprintf(stderr, " Below options are intended for use " + "with other options (except -l):\n"); + (void) fprintf(stderr, " -A ignore assertions (-A), enable " + "panic recovery (-AA) or both (-AAA)\n"); + (void) fprintf(stderr, " -F attempt automatic rewind within " + "safe range of transaction groups\n"); + (void) fprintf(stderr, " -U <cachefile_path> -- use alternate " + "cachefile\n"); + (void) fprintf(stderr, " -X attempt extreme rewind (does not " + "work with dataset)\n"); + (void) fprintf(stderr, " -e pool is exported/destroyed/" + "has altroot/not in a cachefile\n"); + (void) fprintf(stderr, " -p <path> -- use one or more with " + "-e to specify path to vdev dir\n"); + (void) fprintf(stderr, " -P print numbers parsable\n"); + (void) fprintf(stderr, " -t <txg> -- highest txg to use when " + "searching for uberblocks\n"); + (void) fprintf(stderr, "Specify an option more than once (e.g. -bb) " + "to make only that option verbose\n"); + (void) fprintf(stderr, "Default is to dump everything non-verbosely\n"); + exit(1); +} + +/* + * Called for usage errors that are discovered after a call to spa_open(), + * dmu_bonus_hold(), or pool_match(). abort() is called for other errors. + */ + +static void +fatal(const char *fmt, ...) +{ + va_list ap; + + va_start(ap, fmt); + (void) fprintf(stderr, "%s: ", cmdname); + (void) vfprintf(stderr, fmt, ap); + va_end(ap); + (void) fprintf(stderr, "\n"); + + exit(1); +} + +/* ARGSUSED */ +static void +dump_packed_nvlist(objset_t *os, uint64_t object, void *data, size_t size) +{ + nvlist_t *nv; + size_t nvsize = *(uint64_t *)data; + char *packed = umem_alloc(nvsize, UMEM_NOFAIL); + + VERIFY(0 == dmu_read(os, object, 0, nvsize, packed, DMU_READ_PREFETCH)); + + VERIFY(nvlist_unpack(packed, nvsize, &nv, 0) == 0); + + umem_free(packed, nvsize); + + dump_nvlist(nv, 8); + + nvlist_free(nv); +} + +static void +zdb_nicenum(uint64_t num, char *buf) +{ + if (dump_opt['P']) + (void) sprintf(buf, "%llu", (longlong_t)num); + else + nicenum(num, buf); +} + +const char dump_zap_stars[] = "****************************************"; +const int dump_zap_width = sizeof (dump_zap_stars) - 1; + +static void +dump_zap_histogram(uint64_t histo[ZAP_HISTOGRAM_SIZE]) +{ + int i; + int minidx = ZAP_HISTOGRAM_SIZE - 1; + int maxidx = 0; + uint64_t max = 0; + + for (i = 0; i < ZAP_HISTOGRAM_SIZE; i++) { + if (histo[i] > max) + max = histo[i]; + if (histo[i] > 0 && i > maxidx) + maxidx = i; + if (histo[i] > 0 && i < minidx) + minidx = i; + } + + if (max < dump_zap_width) + max = dump_zap_width; + + for (i = minidx; i <= maxidx; i++) + (void) printf("\t\t\t%u: %6llu %s\n", i, (u_longlong_t)histo[i], + &dump_zap_stars[(max - histo[i]) * dump_zap_width / max]); +} + +static void +dump_zap_stats(objset_t *os, uint64_t object) +{ + int error; + zap_stats_t zs; + + error = zap_get_stats(os, object, &zs); + if (error) + return; + + if (zs.zs_ptrtbl_len == 0) { + ASSERT(zs.zs_num_blocks == 1); + (void) printf("\tmicrozap: %llu bytes, %llu entries\n", + (u_longlong_t)zs.zs_blocksize, + (u_longlong_t)zs.zs_num_entries); + return; + } + + (void) printf("\tFat ZAP stats:\n"); + + (void) printf("\t\tPointer table:\n"); + (void) printf("\t\t\t%llu elements\n", + (u_longlong_t)zs.zs_ptrtbl_len); + (void) printf("\t\t\tzt_blk: %llu\n", + (u_longlong_t)zs.zs_ptrtbl_zt_blk); + (void) printf("\t\t\tzt_numblks: %llu\n", + (u_longlong_t)zs.zs_ptrtbl_zt_numblks); + (void) printf("\t\t\tzt_shift: %llu\n", + (u_longlong_t)zs.zs_ptrtbl_zt_shift); + (void) printf("\t\t\tzt_blks_copied: %llu\n", + (u_longlong_t)zs.zs_ptrtbl_blks_copied); + (void) printf("\t\t\tzt_nextblk: %llu\n", + (u_longlong_t)zs.zs_ptrtbl_nextblk); + + (void) printf("\t\tZAP entries: %llu\n", + (u_longlong_t)zs.zs_num_entries); + (void) printf("\t\tLeaf blocks: %llu\n", + (u_longlong_t)zs.zs_num_leafs); + (void) printf("\t\tTotal blocks: %llu\n", + (u_longlong_t)zs.zs_num_blocks); + (void) printf("\t\tzap_block_type: 0x%llx\n", + (u_longlong_t)zs.zs_block_type); + (void) printf("\t\tzap_magic: 0x%llx\n", + (u_longlong_t)zs.zs_magic); + (void) printf("\t\tzap_salt: 0x%llx\n", + (u_longlong_t)zs.zs_salt); + + (void) printf("\t\tLeafs with 2^n pointers:\n"); + dump_zap_histogram(zs.zs_leafs_with_2n_pointers); + + (void) printf("\t\tBlocks with n*5 entries:\n"); + dump_zap_histogram(zs.zs_blocks_with_n5_entries); + + (void) printf("\t\tBlocks n/10 full:\n"); + dump_zap_histogram(zs.zs_blocks_n_tenths_full); + + (void) printf("\t\tEntries with n chunks:\n"); + dump_zap_histogram(zs.zs_entries_using_n_chunks); + + (void) printf("\t\tBuckets with n entries:\n"); + dump_zap_histogram(zs.zs_buckets_with_n_entries); +} + +/*ARGSUSED*/ +static void +dump_none(objset_t *os, uint64_t object, void *data, size_t size) +{ +} + +/*ARGSUSED*/ +static void +dump_unknown(objset_t *os, uint64_t object, void *data, size_t size) +{ + (void) printf("\tUNKNOWN OBJECT TYPE\n"); +} + +/*ARGSUSED*/ +void +dump_uint8(objset_t *os, uint64_t object, void *data, size_t size) +{ +} + +/*ARGSUSED*/ +static void +dump_uint64(objset_t *os, uint64_t object, void *data, size_t size) +{ +} + +/*ARGSUSED*/ +static void +dump_zap(objset_t *os, uint64_t object, void *data, size_t size) +{ + zap_cursor_t zc; + zap_attribute_t attr; + void *prop; + int i; + + dump_zap_stats(os, object); + (void) printf("\n"); + + for (zap_cursor_init(&zc, os, object); + zap_cursor_retrieve(&zc, &attr) == 0; + zap_cursor_advance(&zc)) { + (void) printf("\t\t%s = ", attr.za_name); + if (attr.za_num_integers == 0) { + (void) printf("\n"); + continue; + } + prop = umem_zalloc(attr.za_num_integers * + attr.za_integer_length, UMEM_NOFAIL); + (void) zap_lookup(os, object, attr.za_name, + attr.za_integer_length, attr.za_num_integers, prop); + if (attr.za_integer_length == 1) { + (void) printf("%s", (char *)prop); + } else { + for (i = 0; i < attr.za_num_integers; i++) { + switch (attr.za_integer_length) { + case 2: + (void) printf("%u ", + ((uint16_t *)prop)[i]); + break; + case 4: + (void) printf("%u ", + ((uint32_t *)prop)[i]); + break; + case 8: + (void) printf("%lld ", + (u_longlong_t)((int64_t *)prop)[i]); + break; + } + } + } + (void) printf("\n"); + umem_free(prop, attr.za_num_integers * attr.za_integer_length); + } + zap_cursor_fini(&zc); +} + +/*ARGSUSED*/ +static void +dump_ddt_zap(objset_t *os, uint64_t object, void *data, size_t size) +{ + dump_zap_stats(os, object); + /* contents are printed elsewhere, properly decoded */ +} + +/*ARGSUSED*/ +static void +dump_sa_attrs(objset_t *os, uint64_t object, void *data, size_t size) +{ + zap_cursor_t zc; + zap_attribute_t attr; + + dump_zap_stats(os, object); + (void) printf("\n"); + + for (zap_cursor_init(&zc, os, object); + zap_cursor_retrieve(&zc, &attr) == 0; + zap_cursor_advance(&zc)) { + (void) printf("\t\t%s = ", attr.za_name); + if (attr.za_num_integers == 0) { + (void) printf("\n"); + continue; + } + (void) printf(" %llx : [%d:%d:%d]\n", + (u_longlong_t)attr.za_first_integer, + (int)ATTR_LENGTH(attr.za_first_integer), + (int)ATTR_BSWAP(attr.za_first_integer), + (int)ATTR_NUM(attr.za_first_integer)); + } + zap_cursor_fini(&zc); +} + +/*ARGSUSED*/ +static void +dump_sa_layouts(objset_t *os, uint64_t object, void *data, size_t size) +{ + zap_cursor_t zc; + zap_attribute_t attr; + uint16_t *layout_attrs; + int i; + + dump_zap_stats(os, object); + (void) printf("\n"); + + for (zap_cursor_init(&zc, os, object); + zap_cursor_retrieve(&zc, &attr) == 0; + zap_cursor_advance(&zc)) { + (void) printf("\t\t%s = [", attr.za_name); + if (attr.za_num_integers == 0) { + (void) printf("\n"); + continue; + } + + VERIFY(attr.za_integer_length == 2); + layout_attrs = umem_zalloc(attr.za_num_integers * + attr.za_integer_length, UMEM_NOFAIL); + + VERIFY(zap_lookup(os, object, attr.za_name, + attr.za_integer_length, + attr.za_num_integers, layout_attrs) == 0); + + for (i = 0; i != attr.za_num_integers; i++) + (void) printf(" %d ", (int)layout_attrs[i]); + (void) printf("]\n"); + umem_free(layout_attrs, + attr.za_num_integers * attr.za_integer_length); + } + zap_cursor_fini(&zc); +} + +/*ARGSUSED*/ +static void +dump_zpldir(objset_t *os, uint64_t object, void *data, size_t size) +{ + zap_cursor_t zc; + zap_attribute_t attr; + const char *typenames[] = { + /* 0 */ "not specified", + /* 1 */ "FIFO", + /* 2 */ "Character Device", + /* 3 */ "3 (invalid)", + /* 4 */ "Directory", + /* 5 */ "5 (invalid)", + /* 6 */ "Block Device", + /* 7 */ "7 (invalid)", + /* 8 */ "Regular File", + /* 9 */ "9 (invalid)", + /* 10 */ "Symbolic Link", + /* 11 */ "11 (invalid)", + /* 12 */ "Socket", + /* 13 */ "Door", + /* 14 */ "Event Port", + /* 15 */ "15 (invalid)", + }; + + dump_zap_stats(os, object); + (void) printf("\n"); + + for (zap_cursor_init(&zc, os, object); + zap_cursor_retrieve(&zc, &attr) == 0; + zap_cursor_advance(&zc)) { + (void) printf("\t\t%s = %lld (type: %s)\n", + attr.za_name, ZFS_DIRENT_OBJ(attr.za_first_integer), + typenames[ZFS_DIRENT_TYPE(attr.za_first_integer)]); + } + zap_cursor_fini(&zc); +} + +static void +dump_spacemap(objset_t *os, space_map_obj_t *smo, space_map_t *sm) +{ + uint64_t alloc, offset, entry; + uint8_t mapshift = sm->sm_shift; + uint64_t mapstart = sm->sm_start; + char *ddata[] = { "ALLOC", "FREE", "CONDENSE", "INVALID", + "INVALID", "INVALID", "INVALID", "INVALID" }; + + if (smo->smo_object == 0) + return; + + /* + * Print out the freelist entries in both encoded and decoded form. + */ + alloc = 0; + for (offset = 0; offset < smo->smo_objsize; offset += sizeof (entry)) { + VERIFY3U(0, ==, dmu_read(os, smo->smo_object, offset, + sizeof (entry), &entry, DMU_READ_PREFETCH)); + if (SM_DEBUG_DECODE(entry)) { + (void) printf("\t [%6llu] %s: txg %llu, pass %llu\n", + (u_longlong_t)(offset / sizeof (entry)), + ddata[SM_DEBUG_ACTION_DECODE(entry)], + (u_longlong_t)SM_DEBUG_TXG_DECODE(entry), + (u_longlong_t)SM_DEBUG_SYNCPASS_DECODE(entry)); + } else { + (void) printf("\t [%6llu] %c range:" + " %010llx-%010llx size: %06llx\n", + (u_longlong_t)(offset / sizeof (entry)), + SM_TYPE_DECODE(entry) == SM_ALLOC ? 'A' : 'F', + (u_longlong_t)((SM_OFFSET_DECODE(entry) << + mapshift) + mapstart), + (u_longlong_t)((SM_OFFSET_DECODE(entry) << + mapshift) + mapstart + (SM_RUN_DECODE(entry) << + mapshift)), + (u_longlong_t)(SM_RUN_DECODE(entry) << mapshift)); + if (SM_TYPE_DECODE(entry) == SM_ALLOC) + alloc += SM_RUN_DECODE(entry) << mapshift; + else + alloc -= SM_RUN_DECODE(entry) << mapshift; + } + } + if (alloc != smo->smo_alloc) { + (void) printf("space_map_object alloc (%llu) INCONSISTENT " + "with space map summary (%llu)\n", + (u_longlong_t)smo->smo_alloc, (u_longlong_t)alloc); + } +} + +static void +dump_metaslab_stats(metaslab_t *msp) +{ + char maxbuf[32]; + space_map_t *sm = &msp->ms_map; + avl_tree_t *t = sm->sm_pp_root; + int free_pct = sm->sm_space * 100 / sm->sm_size; + + zdb_nicenum(space_map_maxsize(sm), maxbuf); + + (void) printf("\t %25s %10lu %7s %6s %4s %4d%%\n", + "segments", avl_numnodes(t), "maxsize", maxbuf, + "freepct", free_pct); +} + +static void +dump_metaslab(metaslab_t *msp) +{ + vdev_t *vd = msp->ms_group->mg_vd; + spa_t *spa = vd->vdev_spa; + space_map_t *sm = &msp->ms_map; + space_map_obj_t *smo = &msp->ms_smo; + char freebuf[32]; + + zdb_nicenum(sm->sm_size - smo->smo_alloc, freebuf); + + (void) printf( + "\tmetaslab %6llu offset %12llx spacemap %6llu free %5s\n", + (u_longlong_t)(sm->sm_start / sm->sm_size), + (u_longlong_t)sm->sm_start, (u_longlong_t)smo->smo_object, freebuf); + + if (dump_opt['m'] > 1 && !dump_opt['L']) { + mutex_enter(&msp->ms_lock); + space_map_load_wait(sm); + if (!sm->sm_loaded) + VERIFY(space_map_load(sm, zfs_metaslab_ops, + SM_FREE, smo, spa->spa_meta_objset) == 0); + dump_metaslab_stats(msp); + space_map_unload(sm); + mutex_exit(&msp->ms_lock); + } + + if (dump_opt['d'] > 5 || dump_opt['m'] > 2) { + ASSERT(sm->sm_size == (1ULL << vd->vdev_ms_shift)); + + mutex_enter(&msp->ms_lock); + dump_spacemap(spa->spa_meta_objset, smo, sm); + mutex_exit(&msp->ms_lock); + } +} + +static void +print_vdev_metaslab_header(vdev_t *vd) +{ + (void) printf("\tvdev %10llu\n\t%-10s%5llu %-19s %-15s %-10s\n", + (u_longlong_t)vd->vdev_id, + "metaslabs", (u_longlong_t)vd->vdev_ms_count, + "offset", "spacemap", "free"); + (void) printf("\t%15s %19s %15s %10s\n", + "---------------", "-------------------", + "---------------", "-------------"); +} + +static void +dump_metaslabs(spa_t *spa) +{ + vdev_t *vd, *rvd = spa->spa_root_vdev; + uint64_t m, c = 0, children = rvd->vdev_children; + + (void) printf("\nMetaslabs:\n"); + + if (!dump_opt['d'] && zopt_objects > 0) { + c = zopt_object[0]; + + if (c >= children) + (void) fatal("bad vdev id: %llu", (u_longlong_t)c); + + if (zopt_objects > 1) { + vd = rvd->vdev_child[c]; + print_vdev_metaslab_header(vd); + + for (m = 1; m < zopt_objects; m++) { + if (zopt_object[m] < vd->vdev_ms_count) + dump_metaslab( + vd->vdev_ms[zopt_object[m]]); + else + (void) fprintf(stderr, "bad metaslab " + "number %llu\n", + (u_longlong_t)zopt_object[m]); + } + (void) printf("\n"); + return; + } + children = c + 1; + } + for (; c < children; c++) { + vd = rvd->vdev_child[c]; + print_vdev_metaslab_header(vd); + + for (m = 0; m < vd->vdev_ms_count; m++) + dump_metaslab(vd->vdev_ms[m]); + (void) printf("\n"); + } +} + +static void +dump_dde(const ddt_t *ddt, const ddt_entry_t *dde, uint64_t index) +{ + const ddt_phys_t *ddp = dde->dde_phys; + const ddt_key_t *ddk = &dde->dde_key; + char *types[4] = { "ditto", "single", "double", "triple" }; + char blkbuf[BP_SPRINTF_LEN]; + blkptr_t blk; + + for (int p = 0; p < DDT_PHYS_TYPES; p++, ddp++) { + if (ddp->ddp_phys_birth == 0) + continue; + ddt_bp_create(ddt->ddt_checksum, ddk, ddp, &blk); + sprintf_blkptr(blkbuf, &blk); + (void) printf("index %llx refcnt %llu %s %s\n", + (u_longlong_t)index, (u_longlong_t)ddp->ddp_refcnt, + types[p], blkbuf); + } +} + +static void +dump_dedup_ratio(const ddt_stat_t *dds) +{ + double rL, rP, rD, D, dedup, compress, copies; + + if (dds->dds_blocks == 0) + return; + + rL = (double)dds->dds_ref_lsize; + rP = (double)dds->dds_ref_psize; + rD = (double)dds->dds_ref_dsize; + D = (double)dds->dds_dsize; + + dedup = rD / D; + compress = rL / rP; + copies = rD / rP; + + (void) printf("dedup = %.2f, compress = %.2f, copies = %.2f, " + "dedup * compress / copies = %.2f\n\n", + dedup, compress, copies, dedup * compress / copies); +} + +static void +dump_ddt(ddt_t *ddt, enum ddt_type type, enum ddt_class class) +{ + char name[DDT_NAMELEN]; + ddt_entry_t dde; + uint64_t walk = 0; + dmu_object_info_t doi; + uint64_t count, dspace, mspace; + int error; + + error = ddt_object_info(ddt, type, class, &doi); + + if (error == ENOENT) + return; + ASSERT(error == 0); + + if ((count = ddt_object_count(ddt, type, class)) == 0) + return; + + dspace = doi.doi_physical_blocks_512 << 9; + mspace = doi.doi_fill_count * doi.doi_data_block_size; + + ddt_object_name(ddt, type, class, name); + + (void) printf("%s: %llu entries, size %llu on disk, %llu in core\n", + name, + (u_longlong_t)count, + (u_longlong_t)(dspace / count), + (u_longlong_t)(mspace / count)); + + if (dump_opt['D'] < 3) + return; + + zpool_dump_ddt(NULL, &ddt->ddt_histogram[type][class]); + + if (dump_opt['D'] < 4) + return; + + if (dump_opt['D'] < 5 && class == DDT_CLASS_UNIQUE) + return; + + (void) printf("%s contents:\n\n", name); + + while ((error = ddt_object_walk(ddt, type, class, &walk, &dde)) == 0) + dump_dde(ddt, &dde, walk); + + ASSERT(error == ENOENT); + + (void) printf("\n"); +} + +static void +dump_all_ddts(spa_t *spa) +{ + ddt_histogram_t ddh_total = { 0 }; + ddt_stat_t dds_total = { 0 }; + + for (enum zio_checksum c = 0; c < ZIO_CHECKSUM_FUNCTIONS; c++) { + ddt_t *ddt = spa->spa_ddt[c]; + for (enum ddt_type type = 0; type < DDT_TYPES; type++) { + for (enum ddt_class class = 0; class < DDT_CLASSES; + class++) { + dump_ddt(ddt, type, class); + } + } + } + + ddt_get_dedup_stats(spa, &dds_total); + + if (dds_total.dds_blocks == 0) { + (void) printf("All DDTs are empty\n"); + return; + } + + (void) printf("\n"); + + if (dump_opt['D'] > 1) { + (void) printf("DDT histogram (aggregated over all DDTs):\n"); + ddt_get_dedup_histogram(spa, &ddh_total); + zpool_dump_ddt(&dds_total, &ddh_total); + } + + dump_dedup_ratio(&dds_total); +} + +static void +dump_dtl_seg(space_map_t *sm, uint64_t start, uint64_t size) +{ + char *prefix = (void *)sm; + + (void) printf("%s [%llu,%llu) length %llu\n", + prefix, + (u_longlong_t)start, + (u_longlong_t)(start + size), + (u_longlong_t)(size)); +} + +static void +dump_dtl(vdev_t *vd, int indent) +{ + spa_t *spa = vd->vdev_spa; + boolean_t required; + char *name[DTL_TYPES] = { "missing", "partial", "scrub", "outage" }; + char prefix[256]; + + spa_vdev_state_enter(spa, SCL_NONE); + required = vdev_dtl_required(vd); + (void) spa_vdev_state_exit(spa, NULL, 0); + + if (indent == 0) + (void) printf("\nDirty time logs:\n\n"); + + (void) printf("\t%*s%s [%s]\n", indent, "", + vd->vdev_path ? vd->vdev_path : + vd->vdev_parent ? vd->vdev_ops->vdev_op_type : spa_name(spa), + required ? "DTL-required" : "DTL-expendable"); + + for (int t = 0; t < DTL_TYPES; t++) { + space_map_t *sm = &vd->vdev_dtl[t]; + if (sm->sm_space == 0) + continue; + (void) snprintf(prefix, sizeof (prefix), "\t%*s%s", + indent + 2, "", name[t]); + mutex_enter(sm->sm_lock); + space_map_walk(sm, dump_dtl_seg, (void *)prefix); + mutex_exit(sm->sm_lock); + if (dump_opt['d'] > 5 && vd->vdev_children == 0) + dump_spacemap(spa->spa_meta_objset, + &vd->vdev_dtl_smo, sm); + } + + for (int c = 0; c < vd->vdev_children; c++) + dump_dtl(vd->vdev_child[c], indent + 4); +} + +static void +dump_history(spa_t *spa) +{ + nvlist_t **events = NULL; + char buf[SPA_MAXBLOCKSIZE]; + uint64_t resid, len, off = 0; + uint_t num = 0; + int error; + time_t tsec; + struct tm t; + char tbuf[30]; + char internalstr[MAXPATHLEN]; + + do { + len = sizeof (buf); + + if ((error = spa_history_get(spa, &off, &len, buf)) != 0) { + (void) fprintf(stderr, "Unable to read history: " + "error %d\n", error); + return; + } + + if (zpool_history_unpack(buf, len, &resid, &events, &num) != 0) + break; + + off -= resid; + } while (len != 0); + + (void) printf("\nHistory:\n"); + for (int i = 0; i < num; i++) { + uint64_t time, txg, ievent; + char *cmd, *intstr; + + if (nvlist_lookup_uint64(events[i], ZPOOL_HIST_TIME, + &time) != 0) + continue; + if (nvlist_lookup_string(events[i], ZPOOL_HIST_CMD, + &cmd) != 0) { + if (nvlist_lookup_uint64(events[i], + ZPOOL_HIST_INT_EVENT, &ievent) != 0) + continue; + verify(nvlist_lookup_uint64(events[i], + ZPOOL_HIST_TXG, &txg) == 0); + verify(nvlist_lookup_string(events[i], + ZPOOL_HIST_INT_STR, &intstr) == 0); + if (ievent >= LOG_END) + continue; + + (void) snprintf(internalstr, + sizeof (internalstr), + "[internal %s txg:%lld] %s", + zfs_history_event_names[ievent], txg, + intstr); + cmd = internalstr; + } + tsec = time; + (void) localtime_r(&tsec, &t); + (void) strftime(tbuf, sizeof (tbuf), "%F.%T", &t); + (void) printf("%s %s\n", tbuf, cmd); + } +} + +/*ARGSUSED*/ +static void +dump_dnode(objset_t *os, uint64_t object, void *data, size_t size) +{ +} + +static uint64_t +blkid2offset(const dnode_phys_t *dnp, const blkptr_t *bp, const zbookmark_t *zb) +{ + if (dnp == NULL) { + ASSERT(zb->zb_level < 0); + if (zb->zb_object == 0) + return (zb->zb_blkid); + return (zb->zb_blkid * BP_GET_LSIZE(bp)); + } + + ASSERT(zb->zb_level >= 0); + + return ((zb->zb_blkid << + (zb->zb_level * (dnp->dn_indblkshift - SPA_BLKPTRSHIFT))) * + dnp->dn_datablkszsec << SPA_MINBLOCKSHIFT); +} + +static void +sprintf_blkptr_compact(char *blkbuf, const blkptr_t *bp) +{ + const dva_t *dva = bp->blk_dva; + int ndvas = dump_opt['d'] > 5 ? BP_GET_NDVAS(bp) : 1; + + if (dump_opt['b'] >= 5) { + sprintf_blkptr(blkbuf, bp); + return; + } + + blkbuf[0] = '\0'; + + for (int i = 0; i < ndvas; i++) + (void) sprintf(blkbuf + strlen(blkbuf), "%llu:%llx:%llx ", + (u_longlong_t)DVA_GET_VDEV(&dva[i]), + (u_longlong_t)DVA_GET_OFFSET(&dva[i]), + (u_longlong_t)DVA_GET_ASIZE(&dva[i])); + + (void) sprintf(blkbuf + strlen(blkbuf), + "%llxL/%llxP F=%llu B=%llu/%llu", + (u_longlong_t)BP_GET_LSIZE(bp), + (u_longlong_t)BP_GET_PSIZE(bp), + (u_longlong_t)bp->blk_fill, + (u_longlong_t)bp->blk_birth, + (u_longlong_t)BP_PHYSICAL_BIRTH(bp)); +} + +static void +print_indirect(blkptr_t *bp, const zbookmark_t *zb, + const dnode_phys_t *dnp) +{ + char blkbuf[BP_SPRINTF_LEN]; + int l; + + ASSERT3U(BP_GET_TYPE(bp), ==, dnp->dn_type); + ASSERT3U(BP_GET_LEVEL(bp), ==, zb->zb_level); + + (void) printf("%16llx ", (u_longlong_t)blkid2offset(dnp, bp, zb)); + + ASSERT(zb->zb_level >= 0); + + for (l = dnp->dn_nlevels - 1; l >= -1; l--) { + if (l == zb->zb_level) { + (void) printf("L%llx", (u_longlong_t)zb->zb_level); + } else { + (void) printf(" "); + } + } + + sprintf_blkptr_compact(blkbuf, bp); + (void) printf("%s\n", blkbuf); +} + +static int +visit_indirect(spa_t *spa, const dnode_phys_t *dnp, + blkptr_t *bp, const zbookmark_t *zb) +{ + int err = 0; + + if (bp->blk_birth == 0) + return (0); + + print_indirect(bp, zb, dnp); + + if (BP_GET_LEVEL(bp) > 0) { + uint32_t flags = ARC_WAIT; + int i; + blkptr_t *cbp; + int epb = BP_GET_LSIZE(bp) >> SPA_BLKPTRSHIFT; + arc_buf_t *buf; + uint64_t fill = 0; + + err = arc_read_nolock(NULL, spa, bp, arc_getbuf_func, &buf, + ZIO_PRIORITY_ASYNC_READ, ZIO_FLAG_CANFAIL, &flags, zb); + if (err) + return (err); + ASSERT(buf->b_data); + + /* recursively visit blocks below this */ + cbp = buf->b_data; + for (i = 0; i < epb; i++, cbp++) { + zbookmark_t czb; + + SET_BOOKMARK(&czb, zb->zb_objset, zb->zb_object, + zb->zb_level - 1, + zb->zb_blkid * epb + i); + err = visit_indirect(spa, dnp, cbp, &czb); + if (err) + break; + fill += cbp->blk_fill; + } + if (!err) + ASSERT3U(fill, ==, bp->blk_fill); + (void) arc_buf_remove_ref(buf, &buf); + } + + return (err); +} + +/*ARGSUSED*/ +static void +dump_indirect(dnode_t *dn) +{ + dnode_phys_t *dnp = dn->dn_phys; + int j; + zbookmark_t czb; + + (void) printf("Indirect blocks:\n"); + + SET_BOOKMARK(&czb, dmu_objset_id(dn->dn_objset), + dn->dn_object, dnp->dn_nlevels - 1, 0); + for (j = 0; j < dnp->dn_nblkptr; j++) { + czb.zb_blkid = j; + (void) visit_indirect(dmu_objset_spa(dn->dn_objset), dnp, + &dnp->dn_blkptr[j], &czb); + } + + (void) printf("\n"); +} + +/*ARGSUSED*/ +static void +dump_dsl_dir(objset_t *os, uint64_t object, void *data, size_t size) +{ + dsl_dir_phys_t *dd = data; + time_t crtime; + char nice[32]; + + if (dd == NULL) + return; + + ASSERT3U(size, >=, sizeof (dsl_dir_phys_t)); + + crtime = dd->dd_creation_time; + (void) printf("\t\tcreation_time = %s", ctime(&crtime)); + (void) printf("\t\thead_dataset_obj = %llu\n", + (u_longlong_t)dd->dd_head_dataset_obj); + (void) printf("\t\tparent_dir_obj = %llu\n", + (u_longlong_t)dd->dd_parent_obj); + (void) printf("\t\torigin_obj = %llu\n", + (u_longlong_t)dd->dd_origin_obj); + (void) printf("\t\tchild_dir_zapobj = %llu\n", + (u_longlong_t)dd->dd_child_dir_zapobj); + zdb_nicenum(dd->dd_used_bytes, nice); + (void) printf("\t\tused_bytes = %s\n", nice); + zdb_nicenum(dd->dd_compressed_bytes, nice); + (void) printf("\t\tcompressed_bytes = %s\n", nice); + zdb_nicenum(dd->dd_uncompressed_bytes, nice); + (void) printf("\t\tuncompressed_bytes = %s\n", nice); + zdb_nicenum(dd->dd_quota, nice); + (void) printf("\t\tquota = %s\n", nice); + zdb_nicenum(dd->dd_reserved, nice); + (void) printf("\t\treserved = %s\n", nice); + (void) printf("\t\tprops_zapobj = %llu\n", + (u_longlong_t)dd->dd_props_zapobj); + (void) printf("\t\tdeleg_zapobj = %llu\n", + (u_longlong_t)dd->dd_deleg_zapobj); + (void) printf("\t\tflags = %llx\n", + (u_longlong_t)dd->dd_flags); + +#define DO(which) \ + zdb_nicenum(dd->dd_used_breakdown[DD_USED_ ## which], nice); \ + (void) printf("\t\tused_breakdown[" #which "] = %s\n", nice) + DO(HEAD); + DO(SNAP); + DO(CHILD); + DO(CHILD_RSRV); + DO(REFRSRV); +#undef DO +} + +/*ARGSUSED*/ +static void +dump_dsl_dataset(objset_t *os, uint64_t object, void *data, size_t size) +{ + dsl_dataset_phys_t *ds = data; + time_t crtime; + char used[32], compressed[32], uncompressed[32], unique[32]; + char blkbuf[BP_SPRINTF_LEN]; + + if (ds == NULL) + return; + + ASSERT(size == sizeof (*ds)); + crtime = ds->ds_creation_time; + zdb_nicenum(ds->ds_used_bytes, used); + zdb_nicenum(ds->ds_compressed_bytes, compressed); + zdb_nicenum(ds->ds_uncompressed_bytes, uncompressed); + zdb_nicenum(ds->ds_unique_bytes, unique); + sprintf_blkptr(blkbuf, &ds->ds_bp); + + (void) printf("\t\tdir_obj = %llu\n", + (u_longlong_t)ds->ds_dir_obj); + (void) printf("\t\tprev_snap_obj = %llu\n", + (u_longlong_t)ds->ds_prev_snap_obj); + (void) printf("\t\tprev_snap_txg = %llu\n", + (u_longlong_t)ds->ds_prev_snap_txg); + (void) printf("\t\tnext_snap_obj = %llu\n", + (u_longlong_t)ds->ds_next_snap_obj); + (void) printf("\t\tsnapnames_zapobj = %llu\n", + (u_longlong_t)ds->ds_snapnames_zapobj); + (void) printf("\t\tnum_children = %llu\n", + (u_longlong_t)ds->ds_num_children); + (void) printf("\t\tuserrefs_obj = %llu\n", + (u_longlong_t)ds->ds_userrefs_obj); + (void) printf("\t\tcreation_time = %s", ctime(&crtime)); + (void) printf("\t\tcreation_txg = %llu\n", + (u_longlong_t)ds->ds_creation_txg); + (void) printf("\t\tdeadlist_obj = %llu\n", + (u_longlong_t)ds->ds_deadlist_obj); + (void) printf("\t\tused_bytes = %s\n", used); + (void) printf("\t\tcompressed_bytes = %s\n", compressed); + (void) printf("\t\tuncompressed_bytes = %s\n", uncompressed); + (void) printf("\t\tunique = %s\n", unique); + (void) printf("\t\tfsid_guid = %llu\n", + (u_longlong_t)ds->ds_fsid_guid); + (void) printf("\t\tguid = %llu\n", + (u_longlong_t)ds->ds_guid); + (void) printf("\t\tflags = %llx\n", + (u_longlong_t)ds->ds_flags); + (void) printf("\t\tnext_clones_obj = %llu\n", + (u_longlong_t)ds->ds_next_clones_obj); + (void) printf("\t\tprops_obj = %llu\n", + (u_longlong_t)ds->ds_props_obj); + (void) printf("\t\tbp = %s\n", blkbuf); +} + +/* ARGSUSED */ +static int +dump_bpobj_cb(void *arg, const blkptr_t *bp, dmu_tx_t *tx) +{ + char blkbuf[BP_SPRINTF_LEN]; + + ASSERT(bp->blk_birth != 0); + sprintf_blkptr_compact(blkbuf, bp); + (void) printf("\t%s\n", blkbuf); + return (0); +} + +static void +dump_bpobj(bpobj_t *bpo, char *name) +{ + char bytes[32]; + char comp[32]; + char uncomp[32]; + + if (dump_opt['d'] < 3) + return; + + zdb_nicenum(bpo->bpo_phys->bpo_bytes, bytes); + if (bpo->bpo_havesubobj) { + zdb_nicenum(bpo->bpo_phys->bpo_comp, comp); + zdb_nicenum(bpo->bpo_phys->bpo_uncomp, uncomp); + (void) printf("\n %s: %llu local blkptrs, %llu subobjs, " + "%s (%s/%s comp)\n", + name, (u_longlong_t)bpo->bpo_phys->bpo_num_blkptrs, + (u_longlong_t)bpo->bpo_phys->bpo_num_subobjs, + bytes, comp, uncomp); + } else { + (void) printf("\n %s: %llu blkptrs, %s\n", + name, (u_longlong_t)bpo->bpo_phys->bpo_num_blkptrs, bytes); + } + + if (dump_opt['d'] < 5) + return; + + (void) printf("\n"); + + (void) bpobj_iterate_nofree(bpo, dump_bpobj_cb, NULL, NULL); +} + +static void +dump_deadlist(dsl_deadlist_t *dl) +{ + dsl_deadlist_entry_t *dle; + char bytes[32]; + char comp[32]; + char uncomp[32]; + + if (dump_opt['d'] < 3) + return; + + zdb_nicenum(dl->dl_phys->dl_used, bytes); + zdb_nicenum(dl->dl_phys->dl_comp, comp); + zdb_nicenum(dl->dl_phys->dl_uncomp, uncomp); + (void) printf("\n Deadlist: %s (%s/%s comp)\n", + bytes, comp, uncomp); + + if (dump_opt['d'] < 4) + return; + + (void) printf("\n"); + + for (dle = avl_first(&dl->dl_tree); dle; + dle = AVL_NEXT(&dl->dl_tree, dle)) { + (void) printf(" mintxg %llu -> obj %llu\n", + (longlong_t)dle->dle_mintxg, + (longlong_t)dle->dle_bpobj.bpo_object); + + if (dump_opt['d'] >= 5) + dump_bpobj(&dle->dle_bpobj, ""); + } +} + +static avl_tree_t idx_tree; +static avl_tree_t domain_tree; +static boolean_t fuid_table_loaded; +static boolean_t sa_loaded; +sa_attr_type_t *sa_attr_table; + +static void +fuid_table_destroy() +{ + if (fuid_table_loaded) { + zfs_fuid_table_destroy(&idx_tree, &domain_tree); + fuid_table_loaded = B_FALSE; + } +} + +/* + * print uid or gid information. + * For normal POSIX id just the id is printed in decimal format. + * For CIFS files with FUID the fuid is printed in hex followed by + * the doman-rid string. + */ +static void +print_idstr(uint64_t id, const char *id_type) +{ + if (FUID_INDEX(id)) { + char *domain; + + domain = zfs_fuid_idx_domain(&idx_tree, FUID_INDEX(id)); + (void) printf("\t%s %llx [%s-%d]\n", id_type, + (u_longlong_t)id, domain, (int)FUID_RID(id)); + } else { + (void) printf("\t%s %llu\n", id_type, (u_longlong_t)id); + } + +} + +static void +dump_uidgid(objset_t *os, uint64_t uid, uint64_t gid) +{ + uint32_t uid_idx, gid_idx; + + uid_idx = FUID_INDEX(uid); + gid_idx = FUID_INDEX(gid); + + /* Load domain table, if not already loaded */ + if (!fuid_table_loaded && (uid_idx || gid_idx)) { + uint64_t fuid_obj; + + /* first find the fuid object. It lives in the master node */ + VERIFY(zap_lookup(os, MASTER_NODE_OBJ, ZFS_FUID_TABLES, + 8, 1, &fuid_obj) == 0); + zfs_fuid_avl_tree_create(&idx_tree, &domain_tree); + (void) zfs_fuid_table_load(os, fuid_obj, + &idx_tree, &domain_tree); + fuid_table_loaded = B_TRUE; + } + + print_idstr(uid, "uid"); + print_idstr(gid, "gid"); +} + +/*ARGSUSED*/ +static void +dump_znode(objset_t *os, uint64_t object, void *data, size_t size) +{ + char path[MAXPATHLEN * 2]; /* allow for xattr and failure prefix */ + sa_handle_t *hdl; + uint64_t xattr, rdev, gen; + uint64_t uid, gid, mode, fsize, parent, links; + uint64_t pflags; + uint64_t acctm[2], modtm[2], chgtm[2], crtm[2]; + time_t z_crtime, z_atime, z_mtime, z_ctime; + sa_bulk_attr_t bulk[12]; + int idx = 0; + int error; + + if (!sa_loaded) { + uint64_t sa_attrs = 0; + uint64_t version; + + VERIFY(zap_lookup(os, MASTER_NODE_OBJ, ZPL_VERSION_STR, + 8, 1, &version) == 0); + if (version >= ZPL_VERSION_SA) { + VERIFY(zap_lookup(os, MASTER_NODE_OBJ, ZFS_SA_ATTRS, + 8, 1, &sa_attrs) == 0); + } + if ((error = sa_setup(os, sa_attrs, zfs_attr_table, + ZPL_END, &sa_attr_table)) != 0) { + (void) printf("sa_setup failed errno %d, can't " + "display znode contents\n", error); + return; + } + sa_loaded = B_TRUE; + } + + if (sa_handle_get(os, object, NULL, SA_HDL_PRIVATE, &hdl)) { + (void) printf("Failed to get handle for SA znode\n"); + return; + } + + SA_ADD_BULK_ATTR(bulk, idx, sa_attr_table[ZPL_UID], NULL, &uid, 8); + SA_ADD_BULK_ATTR(bulk, idx, sa_attr_table[ZPL_GID], NULL, &gid, 8); + SA_ADD_BULK_ATTR(bulk, idx, sa_attr_table[ZPL_LINKS], NULL, + &links, 8); + SA_ADD_BULK_ATTR(bulk, idx, sa_attr_table[ZPL_GEN], NULL, &gen, 8); + SA_ADD_BULK_ATTR(bulk, idx, sa_attr_table[ZPL_MODE], NULL, + &mode, 8); + SA_ADD_BULK_ATTR(bulk, idx, sa_attr_table[ZPL_PARENT], + NULL, &parent, 8); + SA_ADD_BULK_ATTR(bulk, idx, sa_attr_table[ZPL_SIZE], NULL, + &fsize, 8); + SA_ADD_BULK_ATTR(bulk, idx, sa_attr_table[ZPL_ATIME], NULL, + acctm, 16); + SA_ADD_BULK_ATTR(bulk, idx, sa_attr_table[ZPL_MTIME], NULL, + modtm, 16); + SA_ADD_BULK_ATTR(bulk, idx, sa_attr_table[ZPL_CRTIME], NULL, + crtm, 16); + SA_ADD_BULK_ATTR(bulk, idx, sa_attr_table[ZPL_CTIME], NULL, + chgtm, 16); + SA_ADD_BULK_ATTR(bulk, idx, sa_attr_table[ZPL_FLAGS], NULL, + &pflags, 8); + + if (sa_bulk_lookup(hdl, bulk, idx)) { + (void) sa_handle_destroy(hdl); + return; + } + + error = zfs_obj_to_path(os, object, path, sizeof (path)); + if (error != 0) { + (void) snprintf(path, sizeof (path), "\?\?\?<object#%llu>", + (u_longlong_t)object); + } + if (dump_opt['d'] < 3) { + (void) printf("\t%s\n", path); + (void) sa_handle_destroy(hdl); + return; + } + + z_crtime = (time_t)crtm[0]; + z_atime = (time_t)acctm[0]; + z_mtime = (time_t)modtm[0]; + z_ctime = (time_t)chgtm[0]; + + (void) printf("\tpath %s\n", path); + dump_uidgid(os, uid, gid); + (void) printf("\tatime %s", ctime(&z_atime)); + (void) printf("\tmtime %s", ctime(&z_mtime)); + (void) printf("\tctime %s", ctime(&z_ctime)); + (void) printf("\tcrtime %s", ctime(&z_crtime)); + (void) printf("\tgen %llu\n", (u_longlong_t)gen); + (void) printf("\tmode %llo\n", (u_longlong_t)mode); + (void) printf("\tsize %llu\n", (u_longlong_t)fsize); + (void) printf("\tparent %llu\n", (u_longlong_t)parent); + (void) printf("\tlinks %llu\n", (u_longlong_t)links); + (void) printf("\tpflags %llx\n", (u_longlong_t)pflags); + if (sa_lookup(hdl, sa_attr_table[ZPL_XATTR], &xattr, + sizeof (uint64_t)) == 0) + (void) printf("\txattr %llu\n", (u_longlong_t)xattr); + if (sa_lookup(hdl, sa_attr_table[ZPL_RDEV], &rdev, + sizeof (uint64_t)) == 0) + (void) printf("\trdev 0x%016llx\n", (u_longlong_t)rdev); + sa_handle_destroy(hdl); +} + +/*ARGSUSED*/ +static void +dump_acl(objset_t *os, uint64_t object, void *data, size_t size) +{ +} + +/*ARGSUSED*/ +static void +dump_dmu_objset(objset_t *os, uint64_t object, void *data, size_t size) +{ +} + +static object_viewer_t *object_viewer[DMU_OT_NUMTYPES + 1] = { + dump_none, /* unallocated */ + dump_zap, /* object directory */ + dump_uint64, /* object array */ + dump_none, /* packed nvlist */ + dump_packed_nvlist, /* packed nvlist size */ + dump_none, /* bplist */ + dump_none, /* bplist header */ + dump_none, /* SPA space map header */ + dump_none, /* SPA space map */ + dump_none, /* ZIL intent log */ + dump_dnode, /* DMU dnode */ + dump_dmu_objset, /* DMU objset */ + dump_dsl_dir, /* DSL directory */ + dump_zap, /* DSL directory child map */ + dump_zap, /* DSL dataset snap map */ + dump_zap, /* DSL props */ + dump_dsl_dataset, /* DSL dataset */ + dump_znode, /* ZFS znode */ + dump_acl, /* ZFS V0 ACL */ + dump_uint8, /* ZFS plain file */ + dump_zpldir, /* ZFS directory */ + dump_zap, /* ZFS master node */ + dump_zap, /* ZFS delete queue */ + dump_uint8, /* zvol object */ + dump_zap, /* zvol prop */ + dump_uint8, /* other uint8[] */ + dump_uint64, /* other uint64[] */ + dump_zap, /* other ZAP */ + dump_zap, /* persistent error log */ + dump_uint8, /* SPA history */ + dump_uint64, /* SPA history offsets */ + dump_zap, /* Pool properties */ + dump_zap, /* DSL permissions */ + dump_acl, /* ZFS ACL */ + dump_uint8, /* ZFS SYSACL */ + dump_none, /* FUID nvlist */ + dump_packed_nvlist, /* FUID nvlist size */ + dump_zap, /* DSL dataset next clones */ + dump_zap, /* DSL scrub queue */ + dump_zap, /* ZFS user/group used */ + dump_zap, /* ZFS user/group quota */ + dump_zap, /* snapshot refcount tags */ + dump_ddt_zap, /* DDT ZAP object */ + dump_zap, /* DDT statistics */ + dump_znode, /* SA object */ + dump_zap, /* SA Master Node */ + dump_sa_attrs, /* SA attribute registration */ + dump_sa_layouts, /* SA attribute layouts */ + dump_zap, /* DSL scrub translations */ + dump_none, /* fake dedup BP */ + dump_zap, /* deadlist */ + dump_none, /* deadlist hdr */ + dump_zap, /* dsl clones */ + dump_none, /* bpobj subobjs */ + dump_unknown, /* Unknown type, must be last */ +}; + +static void +dump_object(objset_t *os, uint64_t object, int verbosity, int *print_header) +{ + dmu_buf_t *db = NULL; + dmu_object_info_t doi; + dnode_t *dn; + void *bonus = NULL; + size_t bsize = 0; + char iblk[32], dblk[32], lsize[32], asize[32], fill[32]; + char bonus_size[32]; + char aux[50]; + int error; + + if (*print_header) { + (void) printf("\n%10s %3s %5s %5s %5s %5s %6s %s\n", + "Object", "lvl", "iblk", "dblk", "dsize", "lsize", + "%full", "type"); + *print_header = 0; + } + + if (object == 0) { + dn = DMU_META_DNODE(os); + } else { + error = dmu_bonus_hold(os, object, FTAG, &db); + if (error) + fatal("dmu_bonus_hold(%llu) failed, errno %u", + object, error); + bonus = db->db_data; + bsize = db->db_size; + dn = DB_DNODE((dmu_buf_impl_t *)db); + } + dmu_object_info_from_dnode(dn, &doi); + + zdb_nicenum(doi.doi_metadata_block_size, iblk); + zdb_nicenum(doi.doi_data_block_size, dblk); + zdb_nicenum(doi.doi_max_offset, lsize); + zdb_nicenum(doi.doi_physical_blocks_512 << 9, asize); + zdb_nicenum(doi.doi_bonus_size, bonus_size); + (void) sprintf(fill, "%6.2f", 100.0 * doi.doi_fill_count * + doi.doi_data_block_size / (object == 0 ? DNODES_PER_BLOCK : 1) / + doi.doi_max_offset); + + aux[0] = '\0'; + + if (doi.doi_checksum != ZIO_CHECKSUM_INHERIT || verbosity >= 6) { + (void) snprintf(aux + strlen(aux), sizeof (aux), " (K=%s)", + ZDB_CHECKSUM_NAME(doi.doi_checksum)); + } + + if (doi.doi_compress != ZIO_COMPRESS_INHERIT || verbosity >= 6) { + (void) snprintf(aux + strlen(aux), sizeof (aux), " (Z=%s)", + ZDB_COMPRESS_NAME(doi.doi_compress)); + } + + (void) printf("%10lld %3u %5s %5s %5s %5s %6s %s%s\n", + (u_longlong_t)object, doi.doi_indirection, iblk, dblk, + asize, lsize, fill, ZDB_OT_NAME(doi.doi_type), aux); + + if (doi.doi_bonus_type != DMU_OT_NONE && verbosity > 3) { + (void) printf("%10s %3s %5s %5s %5s %5s %6s %s\n", + "", "", "", "", "", bonus_size, "bonus", + ZDB_OT_NAME(doi.doi_bonus_type)); + } + + if (verbosity >= 4) { + (void) printf("\tdnode flags: %s%s%s\n", + (dn->dn_phys->dn_flags & DNODE_FLAG_USED_BYTES) ? + "USED_BYTES " : "", + (dn->dn_phys->dn_flags & DNODE_FLAG_USERUSED_ACCOUNTED) ? + "USERUSED_ACCOUNTED " : "", + (dn->dn_phys->dn_flags & DNODE_FLAG_SPILL_BLKPTR) ? + "SPILL_BLKPTR" : ""); + (void) printf("\tdnode maxblkid: %llu\n", + (longlong_t)dn->dn_phys->dn_maxblkid); + + object_viewer[ZDB_OT_TYPE(doi.doi_bonus_type)](os, object, + bonus, bsize); + object_viewer[ZDB_OT_TYPE(doi.doi_type)](os, object, NULL, 0); + *print_header = 1; + } + + if (verbosity >= 5) + dump_indirect(dn); + + if (verbosity >= 5) { + /* + * Report the list of segments that comprise the object. + */ + uint64_t start = 0; + uint64_t end; + uint64_t blkfill = 1; + int minlvl = 1; + + if (dn->dn_type == DMU_OT_DNODE) { + minlvl = 0; + blkfill = DNODES_PER_BLOCK; + } + + for (;;) { + char segsize[32]; + error = dnode_next_offset(dn, + 0, &start, minlvl, blkfill, 0); + if (error) + break; + end = start; + error = dnode_next_offset(dn, + DNODE_FIND_HOLE, &end, minlvl, blkfill, 0); + zdb_nicenum(end - start, segsize); + (void) printf("\t\tsegment [%016llx, %016llx)" + " size %5s\n", (u_longlong_t)start, + (u_longlong_t)end, segsize); + if (error) + break; + start = end; + } + } + + if (db != NULL) + dmu_buf_rele(db, FTAG); +} + +static char *objset_types[DMU_OST_NUMTYPES] = { + "NONE", "META", "ZPL", "ZVOL", "OTHER", "ANY" }; + +static void +dump_dir(objset_t *os) +{ + dmu_objset_stats_t dds; + uint64_t object, object_count; + uint64_t refdbytes, usedobjs, scratch; + char numbuf[32]; + char blkbuf[BP_SPRINTF_LEN + 20]; + char osname[MAXNAMELEN]; + char *type = "UNKNOWN"; + int verbosity = dump_opt['d']; + int print_header = 1; + int i, error; + + dmu_objset_fast_stat(os, &dds); + + if (dds.dds_type < DMU_OST_NUMTYPES) + type = objset_types[dds.dds_type]; + + if (dds.dds_type == DMU_OST_META) { + dds.dds_creation_txg = TXG_INITIAL; + usedobjs = os->os_rootbp->blk_fill; + refdbytes = os->os_spa->spa_dsl_pool-> + dp_mos_dir->dd_phys->dd_used_bytes; + } else { + dmu_objset_space(os, &refdbytes, &scratch, &usedobjs, &scratch); + } + + ASSERT3U(usedobjs, ==, os->os_rootbp->blk_fill); + + zdb_nicenum(refdbytes, numbuf); + + if (verbosity >= 4) { + (void) sprintf(blkbuf, ", rootbp "); + (void) sprintf_blkptr(blkbuf + strlen(blkbuf), os->os_rootbp); + } else { + blkbuf[0] = '\0'; + } + + dmu_objset_name(os, osname); + + (void) printf("Dataset %s [%s], ID %llu, cr_txg %llu, " + "%s, %llu objects%s\n", + osname, type, (u_longlong_t)dmu_objset_id(os), + (u_longlong_t)dds.dds_creation_txg, + numbuf, (u_longlong_t)usedobjs, blkbuf); + + if (zopt_objects != 0) { + for (i = 0; i < zopt_objects; i++) + dump_object(os, zopt_object[i], verbosity, + &print_header); + (void) printf("\n"); + return; + } + + if (dump_opt['i'] != 0 || verbosity >= 2) + dump_intent_log(dmu_objset_zil(os)); + + if (dmu_objset_ds(os) != NULL) + dump_deadlist(&dmu_objset_ds(os)->ds_deadlist); + + if (verbosity < 2) + return; + + if (os->os_rootbp->blk_birth == 0) + return; + + dump_object(os, 0, verbosity, &print_header); + object_count = 0; + if (DMU_USERUSED_DNODE(os) != NULL && + DMU_USERUSED_DNODE(os)->dn_type != 0) { + dump_object(os, DMU_USERUSED_OBJECT, verbosity, &print_header); + dump_object(os, DMU_GROUPUSED_OBJECT, verbosity, &print_header); + } + + object = 0; + while ((error = dmu_object_next(os, &object, B_FALSE, 0)) == 0) { + dump_object(os, object, verbosity, &print_header); + object_count++; + } + + ASSERT3U(object_count, ==, usedobjs); + + (void) printf("\n"); + + if (error != ESRCH) { + (void) fprintf(stderr, "dmu_object_next() = %d\n", error); + abort(); + } +} + +static void +dump_uberblock(uberblock_t *ub, const char *header, const char *footer) +{ + time_t timestamp = ub->ub_timestamp; + + (void) printf(header ? header : ""); + (void) printf("\tmagic = %016llx\n", (u_longlong_t)ub->ub_magic); + (void) printf("\tversion = %llu\n", (u_longlong_t)ub->ub_version); + (void) printf("\ttxg = %llu\n", (u_longlong_t)ub->ub_txg); + (void) printf("\tguid_sum = %llu\n", (u_longlong_t)ub->ub_guid_sum); + (void) printf("\ttimestamp = %llu UTC = %s", + (u_longlong_t)ub->ub_timestamp, asctime(localtime(×tamp))); + if (dump_opt['u'] >= 3) { + char blkbuf[BP_SPRINTF_LEN]; + sprintf_blkptr(blkbuf, &ub->ub_rootbp); + (void) printf("\trootbp = %s\n", blkbuf); + } + (void) printf(footer ? footer : ""); +} + +static void +dump_config(spa_t *spa) +{ + dmu_buf_t *db; + size_t nvsize = 0; + int error = 0; + + + error = dmu_bonus_hold(spa->spa_meta_objset, + spa->spa_config_object, FTAG, &db); + + if (error == 0) { + nvsize = *(uint64_t *)db->db_data; + dmu_buf_rele(db, FTAG); + + (void) printf("\nMOS Configuration:\n"); + dump_packed_nvlist(spa->spa_meta_objset, + spa->spa_config_object, (void *)&nvsize, 1); + } else { + (void) fprintf(stderr, "dmu_bonus_hold(%llu) failed, errno %d", + (u_longlong_t)spa->spa_config_object, error); + } +} + +static void +dump_cachefile(const char *cachefile) +{ + int fd; + struct stat64 statbuf; + char *buf; + nvlist_t *config; + + if ((fd = open64(cachefile, O_RDONLY)) < 0) { + (void) printf("cannot open '%s': %s\n", cachefile, + strerror(errno)); + exit(1); + } + + if (fstat64(fd, &statbuf) != 0) { + (void) printf("failed to stat '%s': %s\n", cachefile, + strerror(errno)); + exit(1); + } + + if ((buf = malloc(statbuf.st_size)) == NULL) { + (void) fprintf(stderr, "failed to allocate %llu bytes\n", + (u_longlong_t)statbuf.st_size); + exit(1); + } + + if (read(fd, buf, statbuf.st_size) != statbuf.st_size) { + (void) fprintf(stderr, "failed to read %llu bytes\n", + (u_longlong_t)statbuf.st_size); + exit(1); + } + + (void) close(fd); + + if (nvlist_unpack(buf, statbuf.st_size, &config, 0) != 0) { + (void) fprintf(stderr, "failed to unpack nvlist\n"); + exit(1); + } + + free(buf); + + dump_nvlist(config, 0); + + nvlist_free(config); +} + +#define ZDB_MAX_UB_HEADER_SIZE 32 + +static void +dump_label_uberblocks(vdev_label_t *lbl, uint64_t ashift) +{ + vdev_t vd; + vdev_t *vdp = &vd; + char header[ZDB_MAX_UB_HEADER_SIZE]; + + vd.vdev_ashift = ashift; + vdp->vdev_top = vdp; + + for (int i = 0; i < VDEV_UBERBLOCK_COUNT(vdp); i++) { + uint64_t uoff = VDEV_UBERBLOCK_OFFSET(vdp, i); + uberblock_t *ub = (void *)((char *)lbl + uoff); + + if (uberblock_verify(ub)) + continue; + (void) snprintf(header, ZDB_MAX_UB_HEADER_SIZE, + "Uberblock[%d]\n", i); + dump_uberblock(ub, header, ""); + } +} + +static void +dump_label(const char *dev) +{ + int fd; + vdev_label_t label; + char *path, *buf = label.vl_vdev_phys.vp_nvlist; + size_t buflen = sizeof (label.vl_vdev_phys.vp_nvlist); + struct stat64 statbuf; + uint64_t psize, ashift; + int len = strlen(dev) + 1; + + if (strncmp(dev, "/dev/dsk/", 9) == 0) { + len++; + path = malloc(len); + (void) snprintf(path, len, "%s%s", "/dev/rdsk/", dev + 9); + } else { + path = strdup(dev); + } + + if ((fd = open64(path, O_RDONLY)) < 0) { + (void) printf("cannot open '%s': %s\n", path, strerror(errno)); + free(path); + exit(1); + } + + if (fstat64(fd, &statbuf) != 0) { + (void) printf("failed to stat '%s': %s\n", path, + strerror(errno)); + free(path); + (void) close(fd); + exit(1); + } + + if (S_ISBLK(statbuf.st_mode)) { + (void) printf("cannot use '%s': character device required\n", + path); + free(path); + (void) close(fd); + exit(1); + } + + psize = statbuf.st_size; + psize = P2ALIGN(psize, (uint64_t)sizeof (vdev_label_t)); + + for (int l = 0; l < VDEV_LABELS; l++) { + nvlist_t *config = NULL; + + (void) printf("--------------------------------------------\n"); + (void) printf("LABEL %d\n", l); + (void) printf("--------------------------------------------\n"); + + if (pread64(fd, &label, sizeof (label), + vdev_label_offset(psize, l, 0)) != sizeof (label)) { + (void) printf("failed to read label %d\n", l); + continue; + } + + if (nvlist_unpack(buf, buflen, &config, 0) != 0) { + (void) printf("failed to unpack label %d\n", l); + ashift = SPA_MINBLOCKSHIFT; + } else { + nvlist_t *vdev_tree = NULL; + + dump_nvlist(config, 4); + if ((nvlist_lookup_nvlist(config, + ZPOOL_CONFIG_VDEV_TREE, &vdev_tree) != 0) || + (nvlist_lookup_uint64(vdev_tree, + ZPOOL_CONFIG_ASHIFT, &ashift) != 0)) + ashift = SPA_MINBLOCKSHIFT; + nvlist_free(config); + } + if (dump_opt['u']) + dump_label_uberblocks(&label, ashift); + } + + free(path); + (void) close(fd); +} + +/*ARGSUSED*/ +static int +dump_one_dir(const char *dsname, void *arg) +{ + int error; + objset_t *os; + + error = dmu_objset_own(dsname, DMU_OST_ANY, B_TRUE, FTAG, &os); + if (error) { + (void) printf("Could not open %s, error %d\n", dsname, error); + return (0); + } + dump_dir(os); + dmu_objset_disown(os, FTAG); + fuid_table_destroy(); + sa_loaded = B_FALSE; + return (0); +} + +/* + * Block statistics. + */ +typedef struct zdb_blkstats { + uint64_t zb_asize; + uint64_t zb_lsize; + uint64_t zb_psize; + uint64_t zb_count; +} zdb_blkstats_t; + +/* + * Extended object types to report deferred frees and dedup auto-ditto blocks. + */ +#define ZDB_OT_DEFERRED (DMU_OT_NUMTYPES + 0) +#define ZDB_OT_DITTO (DMU_OT_NUMTYPES + 1) +#define ZDB_OT_TOTAL (DMU_OT_NUMTYPES + 2) + +static char *zdb_ot_extname[] = { + "deferred free", + "dedup ditto", + "Total", +}; + +#define ZB_TOTAL DN_MAX_LEVELS + +typedef struct zdb_cb { + zdb_blkstats_t zcb_type[ZB_TOTAL + 1][ZDB_OT_TOTAL + 1]; + uint64_t zcb_dedup_asize; + uint64_t zcb_dedup_blocks; + uint64_t zcb_errors[256]; + int zcb_readfails; + int zcb_haderrors; + spa_t *zcb_spa; +} zdb_cb_t; + +static void +zdb_count_block(zdb_cb_t *zcb, zilog_t *zilog, const blkptr_t *bp, + dmu_object_type_t type) +{ + uint64_t refcnt = 0; + + ASSERT(type < ZDB_OT_TOTAL); + + if (zilog && zil_bp_tree_add(zilog, bp) != 0) + return; + + for (int i = 0; i < 4; i++) { + int l = (i < 2) ? BP_GET_LEVEL(bp) : ZB_TOTAL; + int t = (i & 1) ? type : ZDB_OT_TOTAL; + zdb_blkstats_t *zb = &zcb->zcb_type[l][t]; + + zb->zb_asize += BP_GET_ASIZE(bp); + zb->zb_lsize += BP_GET_LSIZE(bp); + zb->zb_psize += BP_GET_PSIZE(bp); + zb->zb_count++; + } + + if (dump_opt['L']) + return; + + if (BP_GET_DEDUP(bp)) { + ddt_t *ddt; + ddt_entry_t *dde; + + ddt = ddt_select(zcb->zcb_spa, bp); + ddt_enter(ddt); + dde = ddt_lookup(ddt, bp, B_FALSE); + + if (dde == NULL) { + refcnt = 0; + } else { + ddt_phys_t *ddp = ddt_phys_select(dde, bp); + ddt_phys_decref(ddp); + refcnt = ddp->ddp_refcnt; + if (ddt_phys_total_refcnt(dde) == 0) + ddt_remove(ddt, dde); + } + ddt_exit(ddt); + } + + VERIFY3U(zio_wait(zio_claim(NULL, zcb->zcb_spa, + refcnt ? 0 : spa_first_txg(zcb->zcb_spa), + bp, NULL, NULL, ZIO_FLAG_CANFAIL)), ==, 0); +} + +/* ARGSUSED */ +static int +zdb_blkptr_cb(spa_t *spa, zilog_t *zilog, const blkptr_t *bp, arc_buf_t *pbuf, + const zbookmark_t *zb, const dnode_phys_t *dnp, void *arg) +{ + zdb_cb_t *zcb = arg; + char blkbuf[BP_SPRINTF_LEN]; + dmu_object_type_t type; + boolean_t is_metadata; + + if (bp == NULL) + return (0); + + type = BP_GET_TYPE(bp); + + zdb_count_block(zcb, zilog, bp, type); + + is_metadata = (BP_GET_LEVEL(bp) != 0 || dmu_ot[type].ot_metadata); + + if (dump_opt['c'] > 1 || (dump_opt['c'] && is_metadata)) { + int ioerr; + size_t size = BP_GET_PSIZE(bp); + void *data = malloc(size); + int flags = ZIO_FLAG_CANFAIL | ZIO_FLAG_SCRUB | ZIO_FLAG_RAW; + + /* If it's an intent log block, failure is expected. */ + if (zb->zb_level == ZB_ZIL_LEVEL) + flags |= ZIO_FLAG_SPECULATIVE; + + ioerr = zio_wait(zio_read(NULL, spa, bp, data, size, + NULL, NULL, ZIO_PRIORITY_ASYNC_READ, flags, zb)); + + free(data); + + if (ioerr && !(flags & ZIO_FLAG_SPECULATIVE)) { + zcb->zcb_haderrors = 1; + zcb->zcb_errors[ioerr]++; + + if (dump_opt['b'] >= 2) + sprintf_blkptr(blkbuf, bp); + else + blkbuf[0] = '\0'; + + (void) printf("zdb_blkptr_cb: " + "Got error %d reading " + "<%llu, %llu, %lld, %llx> %s -- skipping\n", + ioerr, + (u_longlong_t)zb->zb_objset, + (u_longlong_t)zb->zb_object, + (u_longlong_t)zb->zb_level, + (u_longlong_t)zb->zb_blkid, + blkbuf); + } + } + + zcb->zcb_readfails = 0; + + if (dump_opt['b'] >= 4) { + sprintf_blkptr(blkbuf, bp); + (void) printf("objset %llu object %llu " + "level %lld offset 0x%llx %s\n", + (u_longlong_t)zb->zb_objset, + (u_longlong_t)zb->zb_object, + (longlong_t)zb->zb_level, + (u_longlong_t)blkid2offset(dnp, bp, zb), + blkbuf); + } + + return (0); +} + +static void +zdb_leak(space_map_t *sm, uint64_t start, uint64_t size) +{ + vdev_t *vd = sm->sm_ppd; + + (void) printf("leaked space: vdev %llu, offset 0x%llx, size %llu\n", + (u_longlong_t)vd->vdev_id, (u_longlong_t)start, (u_longlong_t)size); +} + +/* ARGSUSED */ +static void +zdb_space_map_load(space_map_t *sm) +{ +} + +static void +zdb_space_map_unload(space_map_t *sm) +{ + space_map_vacate(sm, zdb_leak, sm); +} + +/* ARGSUSED */ +static void +zdb_space_map_claim(space_map_t *sm, uint64_t start, uint64_t size) +{ +} + +static space_map_ops_t zdb_space_map_ops = { + zdb_space_map_load, + zdb_space_map_unload, + NULL, /* alloc */ + zdb_space_map_claim, + NULL, /* free */ + NULL /* maxsize */ +}; + +static void +zdb_ddt_leak_init(spa_t *spa, zdb_cb_t *zcb) +{ + ddt_bookmark_t ddb = { 0 }; + ddt_entry_t dde; + int error; + + while ((error = ddt_walk(spa, &ddb, &dde)) == 0) { + blkptr_t blk; + ddt_phys_t *ddp = dde.dde_phys; + + if (ddb.ddb_class == DDT_CLASS_UNIQUE) + return; + + ASSERT(ddt_phys_total_refcnt(&dde) > 1); + + for (int p = 0; p < DDT_PHYS_TYPES; p++, ddp++) { + if (ddp->ddp_phys_birth == 0) + continue; + ddt_bp_create(ddb.ddb_checksum, + &dde.dde_key, ddp, &blk); + if (p == DDT_PHYS_DITTO) { + zdb_count_block(zcb, NULL, &blk, ZDB_OT_DITTO); + } else { + zcb->zcb_dedup_asize += + BP_GET_ASIZE(&blk) * (ddp->ddp_refcnt - 1); + zcb->zcb_dedup_blocks++; + } + } + if (!dump_opt['L']) { + ddt_t *ddt = spa->spa_ddt[ddb.ddb_checksum]; + ddt_enter(ddt); + VERIFY(ddt_lookup(ddt, &blk, B_TRUE) != NULL); + ddt_exit(ddt); + } + } + + ASSERT(error == ENOENT); +} + +static void +zdb_leak_init(spa_t *spa, zdb_cb_t *zcb) +{ + zcb->zcb_spa = spa; + + if (!dump_opt['L']) { + vdev_t *rvd = spa->spa_root_vdev; + for (int c = 0; c < rvd->vdev_children; c++) { + vdev_t *vd = rvd->vdev_child[c]; + for (int m = 0; m < vd->vdev_ms_count; m++) { + metaslab_t *msp = vd->vdev_ms[m]; + mutex_enter(&msp->ms_lock); + space_map_unload(&msp->ms_map); + VERIFY(space_map_load(&msp->ms_map, + &zdb_space_map_ops, SM_ALLOC, &msp->ms_smo, + spa->spa_meta_objset) == 0); + msp->ms_map.sm_ppd = vd; + mutex_exit(&msp->ms_lock); + } + } + } + + spa_config_enter(spa, SCL_CONFIG, FTAG, RW_READER); + + zdb_ddt_leak_init(spa, zcb); + + spa_config_exit(spa, SCL_CONFIG, FTAG); +} + +static void +zdb_leak_fini(spa_t *spa) +{ + if (!dump_opt['L']) { + vdev_t *rvd = spa->spa_root_vdev; + for (int c = 0; c < rvd->vdev_children; c++) { + vdev_t *vd = rvd->vdev_child[c]; + for (int m = 0; m < vd->vdev_ms_count; m++) { + metaslab_t *msp = vd->vdev_ms[m]; + mutex_enter(&msp->ms_lock); + space_map_unload(&msp->ms_map); + mutex_exit(&msp->ms_lock); + } + } + } +} + +/* ARGSUSED */ +static int +count_block_cb(void *arg, const blkptr_t *bp, dmu_tx_t *tx) +{ + zdb_cb_t *zcb = arg; + + if (dump_opt['b'] >= 4) { + char blkbuf[BP_SPRINTF_LEN]; + sprintf_blkptr(blkbuf, bp); + (void) printf("[%s] %s\n", + "deferred free", blkbuf); + } + zdb_count_block(zcb, NULL, bp, ZDB_OT_DEFERRED); + return (0); +} + +static int +dump_block_stats(spa_t *spa) +{ + zdb_cb_t zcb = { 0 }; + zdb_blkstats_t *zb, *tzb; + uint64_t norm_alloc, norm_space, total_alloc, total_found; + int flags = TRAVERSE_PRE | TRAVERSE_PREFETCH_METADATA | TRAVERSE_HARD; + int leaks = 0; + + (void) printf("\nTraversing all blocks %s%s%s%s%s...\n", + (dump_opt['c'] || !dump_opt['L']) ? "to verify " : "", + (dump_opt['c'] == 1) ? "metadata " : "", + dump_opt['c'] ? "checksums " : "", + (dump_opt['c'] && !dump_opt['L']) ? "and verify " : "", + !dump_opt['L'] ? "nothing leaked " : ""); + + /* + * Load all space maps as SM_ALLOC maps, then traverse the pool + * claiming each block we discover. If the pool is perfectly + * consistent, the space maps will be empty when we're done. + * Anything left over is a leak; any block we can't claim (because + * it's not part of any space map) is a double allocation, + * reference to a freed block, or an unclaimed log block. + */ + zdb_leak_init(spa, &zcb); + + /* + * If there's a deferred-free bplist, process that first. + */ + (void) bpobj_iterate_nofree(&spa->spa_deferred_bpobj, + count_block_cb, &zcb, NULL); + (void) bpobj_iterate_nofree(&spa->spa_dsl_pool->dp_free_bpobj, + count_block_cb, &zcb, NULL); + + if (dump_opt['c'] > 1) + flags |= TRAVERSE_PREFETCH_DATA; + + zcb.zcb_haderrors |= traverse_pool(spa, 0, flags, zdb_blkptr_cb, &zcb); + + if (zcb.zcb_haderrors) { + (void) printf("\nError counts:\n\n"); + (void) printf("\t%5s %s\n", "errno", "count"); + for (int e = 0; e < 256; e++) { + if (zcb.zcb_errors[e] != 0) { + (void) printf("\t%5d %llu\n", + e, (u_longlong_t)zcb.zcb_errors[e]); + } + } + } + + /* + * Report any leaked segments. + */ + zdb_leak_fini(spa); + + tzb = &zcb.zcb_type[ZB_TOTAL][ZDB_OT_TOTAL]; + + norm_alloc = metaslab_class_get_alloc(spa_normal_class(spa)); + norm_space = metaslab_class_get_space(spa_normal_class(spa)); + + total_alloc = norm_alloc + metaslab_class_get_alloc(spa_log_class(spa)); + total_found = tzb->zb_asize - zcb.zcb_dedup_asize; + + if (total_found == total_alloc) { + if (!dump_opt['L']) + (void) printf("\n\tNo leaks (block sum matches space" + " maps exactly)\n"); + } else { + (void) printf("block traversal size %llu != alloc %llu " + "(%s %lld)\n", + (u_longlong_t)total_found, + (u_longlong_t)total_alloc, + (dump_opt['L']) ? "unreachable" : "leaked", + (longlong_t)(total_alloc - total_found)); + leaks = 1; + } + + if (tzb->zb_count == 0) + return (2); + + (void) printf("\n"); + (void) printf("\tbp count: %10llu\n", + (u_longlong_t)tzb->zb_count); + (void) printf("\tbp logical: %10llu avg: %6llu\n", + (u_longlong_t)tzb->zb_lsize, + (u_longlong_t)(tzb->zb_lsize / tzb->zb_count)); + (void) printf("\tbp physical: %10llu avg:" + " %6llu compression: %6.2f\n", + (u_longlong_t)tzb->zb_psize, + (u_longlong_t)(tzb->zb_psize / tzb->zb_count), + (double)tzb->zb_lsize / tzb->zb_psize); + (void) printf("\tbp allocated: %10llu avg:" + " %6llu compression: %6.2f\n", + (u_longlong_t)tzb->zb_asize, + (u_longlong_t)(tzb->zb_asize / tzb->zb_count), + (double)tzb->zb_lsize / tzb->zb_asize); + (void) printf("\tbp deduped: %10llu ref>1:" + " %6llu deduplication: %6.2f\n", + (u_longlong_t)zcb.zcb_dedup_asize, + (u_longlong_t)zcb.zcb_dedup_blocks, + (double)zcb.zcb_dedup_asize / tzb->zb_asize + 1.0); + (void) printf("\tSPA allocated: %10llu used: %5.2f%%\n", + (u_longlong_t)norm_alloc, 100.0 * norm_alloc / norm_space); + + if (dump_opt['b'] >= 2) { + int l, t, level; + (void) printf("\nBlocks\tLSIZE\tPSIZE\tASIZE" + "\t avg\t comp\t%%Total\tType\n"); + + for (t = 0; t <= ZDB_OT_TOTAL; t++) { + char csize[32], lsize[32], psize[32], asize[32]; + char avg[32]; + char *typename; + + if (t < DMU_OT_NUMTYPES) + typename = dmu_ot[t].ot_name; + else + typename = zdb_ot_extname[t - DMU_OT_NUMTYPES]; + + if (zcb.zcb_type[ZB_TOTAL][t].zb_asize == 0) { + (void) printf("%6s\t%5s\t%5s\t%5s" + "\t%5s\t%5s\t%6s\t%s\n", + "-", + "-", + "-", + "-", + "-", + "-", + "-", + typename); + continue; + } + + for (l = ZB_TOTAL - 1; l >= -1; l--) { + level = (l == -1 ? ZB_TOTAL : l); + zb = &zcb.zcb_type[level][t]; + + if (zb->zb_asize == 0) + continue; + + if (dump_opt['b'] < 3 && level != ZB_TOTAL) + continue; + + if (level == 0 && zb->zb_asize == + zcb.zcb_type[ZB_TOTAL][t].zb_asize) + continue; + + zdb_nicenum(zb->zb_count, csize); + zdb_nicenum(zb->zb_lsize, lsize); + zdb_nicenum(zb->zb_psize, psize); + zdb_nicenum(zb->zb_asize, asize); + zdb_nicenum(zb->zb_asize / zb->zb_count, avg); + + (void) printf("%6s\t%5s\t%5s\t%5s\t%5s" + "\t%5.2f\t%6.2f\t", + csize, lsize, psize, asize, avg, + (double)zb->zb_lsize / zb->zb_psize, + 100.0 * zb->zb_asize / tzb->zb_asize); + + if (level == ZB_TOTAL) + (void) printf("%s\n", typename); + else + (void) printf(" L%d %s\n", + level, typename); + } + } + } + + (void) printf("\n"); + + if (leaks) + return (2); + + if (zcb.zcb_haderrors) + return (3); + + return (0); +} + +typedef struct zdb_ddt_entry { + ddt_key_t zdde_key; + uint64_t zdde_ref_blocks; + uint64_t zdde_ref_lsize; + uint64_t zdde_ref_psize; + uint64_t zdde_ref_dsize; + avl_node_t zdde_node; +} zdb_ddt_entry_t; + +/* ARGSUSED */ +static int +zdb_ddt_add_cb(spa_t *spa, zilog_t *zilog, const blkptr_t *bp, + arc_buf_t *pbuf, const zbookmark_t *zb, const dnode_phys_t *dnp, void *arg) +{ + avl_tree_t *t = arg; + avl_index_t where; + zdb_ddt_entry_t *zdde, zdde_search; + + if (bp == NULL) + return (0); + + if (dump_opt['S'] > 1 && zb->zb_level == ZB_ROOT_LEVEL) { + (void) printf("traversing objset %llu, %llu objects, " + "%lu blocks so far\n", + (u_longlong_t)zb->zb_objset, + (u_longlong_t)bp->blk_fill, + avl_numnodes(t)); + } + + if (BP_IS_HOLE(bp) || BP_GET_CHECKSUM(bp) == ZIO_CHECKSUM_OFF || + BP_GET_LEVEL(bp) > 0 || dmu_ot[BP_GET_TYPE(bp)].ot_metadata) + return (0); + + ddt_key_fill(&zdde_search.zdde_key, bp); + + zdde = avl_find(t, &zdde_search, &where); + + if (zdde == NULL) { + zdde = umem_zalloc(sizeof (*zdde), UMEM_NOFAIL); + zdde->zdde_key = zdde_search.zdde_key; + avl_insert(t, zdde, where); + } + + zdde->zdde_ref_blocks += 1; + zdde->zdde_ref_lsize += BP_GET_LSIZE(bp); + zdde->zdde_ref_psize += BP_GET_PSIZE(bp); + zdde->zdde_ref_dsize += bp_get_dsize_sync(spa, bp); + + return (0); +} + +static void +dump_simulated_ddt(spa_t *spa) +{ + avl_tree_t t; + void *cookie = NULL; + zdb_ddt_entry_t *zdde; + ddt_histogram_t ddh_total = { 0 }; + ddt_stat_t dds_total = { 0 }; + + avl_create(&t, ddt_entry_compare, + sizeof (zdb_ddt_entry_t), offsetof(zdb_ddt_entry_t, zdde_node)); + + spa_config_enter(spa, SCL_CONFIG, FTAG, RW_READER); + + (void) traverse_pool(spa, 0, TRAVERSE_PRE | TRAVERSE_PREFETCH_METADATA, + zdb_ddt_add_cb, &t); + + spa_config_exit(spa, SCL_CONFIG, FTAG); + + while ((zdde = avl_destroy_nodes(&t, &cookie)) != NULL) { + ddt_stat_t dds; + uint64_t refcnt = zdde->zdde_ref_blocks; + ASSERT(refcnt != 0); + + dds.dds_blocks = zdde->zdde_ref_blocks / refcnt; + dds.dds_lsize = zdde->zdde_ref_lsize / refcnt; + dds.dds_psize = zdde->zdde_ref_psize / refcnt; + dds.dds_dsize = zdde->zdde_ref_dsize / refcnt; + + dds.dds_ref_blocks = zdde->zdde_ref_blocks; + dds.dds_ref_lsize = zdde->zdde_ref_lsize; + dds.dds_ref_psize = zdde->zdde_ref_psize; + dds.dds_ref_dsize = zdde->zdde_ref_dsize; + + ddt_stat_add(&ddh_total.ddh_stat[highbit(refcnt) - 1], &dds, 0); + + umem_free(zdde, sizeof (*zdde)); + } + + avl_destroy(&t); + + ddt_histogram_stat(&dds_total, &ddh_total); + + (void) printf("Simulated DDT histogram:\n"); + + zpool_dump_ddt(&dds_total, &ddh_total); + + dump_dedup_ratio(&dds_total); +} + +static void +dump_zpool(spa_t *spa) +{ + dsl_pool_t *dp = spa_get_dsl(spa); + int rc = 0; + + if (dump_opt['S']) { + dump_simulated_ddt(spa); + return; + } + + if (!dump_opt['e'] && dump_opt['C'] > 1) { + (void) printf("\nCached configuration:\n"); + dump_nvlist(spa->spa_config, 8); + } + + if (dump_opt['C']) + dump_config(spa); + + if (dump_opt['u']) + dump_uberblock(&spa->spa_uberblock, "\nUberblock:\n", "\n"); + + if (dump_opt['D']) + dump_all_ddts(spa); + + if (dump_opt['d'] > 2 || dump_opt['m']) + dump_metaslabs(spa); + + if (dump_opt['d'] || dump_opt['i']) { + dump_dir(dp->dp_meta_objset); + if (dump_opt['d'] >= 3) { + dump_bpobj(&spa->spa_deferred_bpobj, "Deferred frees"); + if (spa_version(spa) >= SPA_VERSION_DEADLISTS) { + dump_bpobj(&spa->spa_dsl_pool->dp_free_bpobj, + "Pool frees"); + } + dump_dtl(spa->spa_root_vdev, 0); + } + (void) dmu_objset_find(spa_name(spa), dump_one_dir, + NULL, DS_FIND_SNAPSHOTS | DS_FIND_CHILDREN); + } + if (dump_opt['b'] || dump_opt['c']) + rc = dump_block_stats(spa); + + if (dump_opt['s']) + show_pool_stats(spa); + + if (dump_opt['h']) + dump_history(spa); + + if (rc != 0) + exit(rc); +} + +#define ZDB_FLAG_CHECKSUM 0x0001 +#define ZDB_FLAG_DECOMPRESS 0x0002 +#define ZDB_FLAG_BSWAP 0x0004 +#define ZDB_FLAG_GBH 0x0008 +#define ZDB_FLAG_INDIRECT 0x0010 +#define ZDB_FLAG_PHYS 0x0020 +#define ZDB_FLAG_RAW 0x0040 +#define ZDB_FLAG_PRINT_BLKPTR 0x0080 + +int flagbits[256]; + +static void +zdb_print_blkptr(blkptr_t *bp, int flags) +{ + char blkbuf[BP_SPRINTF_LEN]; + + if (flags & ZDB_FLAG_BSWAP) + byteswap_uint64_array((void *)bp, sizeof (blkptr_t)); + + sprintf_blkptr(blkbuf, bp); + (void) printf("%s\n", blkbuf); +} + +static void +zdb_dump_indirect(blkptr_t *bp, int nbps, int flags) +{ + int i; + + for (i = 0; i < nbps; i++) + zdb_print_blkptr(&bp[i], flags); +} + +static void +zdb_dump_gbh(void *buf, int flags) +{ + zdb_dump_indirect((blkptr_t *)buf, SPA_GBH_NBLKPTRS, flags); +} + +static void +zdb_dump_block_raw(void *buf, uint64_t size, int flags) +{ + if (flags & ZDB_FLAG_BSWAP) + byteswap_uint64_array(buf, size); + (void) write(1, buf, size); +} + +static void +zdb_dump_block(char *label, void *buf, uint64_t size, int flags) +{ + uint64_t *d = (uint64_t *)buf; + int nwords = size / sizeof (uint64_t); + int do_bswap = !!(flags & ZDB_FLAG_BSWAP); + int i, j; + char *hdr, *c; + + + if (do_bswap) + hdr = " 7 6 5 4 3 2 1 0 f e d c b a 9 8"; + else + hdr = " 0 1 2 3 4 5 6 7 8 9 a b c d e f"; + + (void) printf("\n%s\n%6s %s 0123456789abcdef\n", label, "", hdr); + + for (i = 0; i < nwords; i += 2) { + (void) printf("%06llx: %016llx %016llx ", + (u_longlong_t)(i * sizeof (uint64_t)), + (u_longlong_t)(do_bswap ? BSWAP_64(d[i]) : d[i]), + (u_longlong_t)(do_bswap ? BSWAP_64(d[i + 1]) : d[i + 1])); + + c = (char *)&d[i]; + for (j = 0; j < 2 * sizeof (uint64_t); j++) + (void) printf("%c", isprint(c[j]) ? c[j] : '.'); + (void) printf("\n"); + } +} + +/* + * There are two acceptable formats: + * leaf_name - For example: c1t0d0 or /tmp/ztest.0a + * child[.child]* - For example: 0.1.1 + * + * The second form can be used to specify arbitrary vdevs anywhere + * in the heirarchy. For example, in a pool with a mirror of + * RAID-Zs, you can specify either RAID-Z vdev with 0.0 or 0.1 . + */ +static vdev_t * +zdb_vdev_lookup(vdev_t *vdev, char *path) +{ + char *s, *p, *q; + int i; + + if (vdev == NULL) + return (NULL); + + /* First, assume the x.x.x.x format */ + i = (int)strtoul(path, &s, 10); + if (s == path || (s && *s != '.' && *s != '\0')) + goto name; + if (i < 0 || i >= vdev->vdev_children) + return (NULL); + + vdev = vdev->vdev_child[i]; + if (*s == '\0') + return (vdev); + return (zdb_vdev_lookup(vdev, s+1)); + +name: + for (i = 0; i < vdev->vdev_children; i++) { + vdev_t *vc = vdev->vdev_child[i]; + + if (vc->vdev_path == NULL) { + vc = zdb_vdev_lookup(vc, path); + if (vc == NULL) + continue; + else + return (vc); + } + + p = strrchr(vc->vdev_path, '/'); + p = p ? p + 1 : vc->vdev_path; + q = &vc->vdev_path[strlen(vc->vdev_path) - 2]; + + if (strcmp(vc->vdev_path, path) == 0) + return (vc); + if (strcmp(p, path) == 0) + return (vc); + if (strcmp(q, "s0") == 0 && strncmp(p, path, q - p) == 0) + return (vc); + } + + return (NULL); +} + +/* + * Read a block from a pool and print it out. The syntax of the + * block descriptor is: + * + * pool:vdev_specifier:offset:size[:flags] + * + * pool - The name of the pool you wish to read from + * vdev_specifier - Which vdev (see comment for zdb_vdev_lookup) + * offset - offset, in hex, in bytes + * size - Amount of data to read, in hex, in bytes + * flags - A string of characters specifying options + * b: Decode a blkptr at given offset within block + * *c: Calculate and display checksums + * d: Decompress data before dumping + * e: Byteswap data before dumping + * g: Display data as a gang block header + * i: Display as an indirect block + * p: Do I/O to physical offset + * r: Dump raw data to stdout + * + * * = not yet implemented + */ +static void +zdb_read_block(char *thing, spa_t *spa) +{ + blkptr_t blk, *bp = &blk; + dva_t *dva = bp->blk_dva; + int flags = 0; + uint64_t offset = 0, size = 0, psize = 0, lsize = 0, blkptr_offset = 0; + zio_t *zio; + vdev_t *vd; + void *pbuf, *lbuf, *buf; + char *s, *p, *dup, *vdev, *flagstr; + int i, error; + + dup = strdup(thing); + s = strtok(dup, ":"); + vdev = s ? s : ""; + s = strtok(NULL, ":"); + offset = strtoull(s ? s : "", NULL, 16); + s = strtok(NULL, ":"); + size = strtoull(s ? s : "", NULL, 16); + s = strtok(NULL, ":"); + flagstr = s ? s : ""; + + s = NULL; + if (size == 0) + s = "size must not be zero"; + if (!IS_P2ALIGNED(size, DEV_BSIZE)) + s = "size must be a multiple of sector size"; + if (!IS_P2ALIGNED(offset, DEV_BSIZE)) + s = "offset must be a multiple of sector size"; + if (s) { + (void) printf("Invalid block specifier: %s - %s\n", thing, s); + free(dup); + return; + } + + for (s = strtok(flagstr, ":"); s; s = strtok(NULL, ":")) { + for (i = 0; flagstr[i]; i++) { + int bit = flagbits[(uchar_t)flagstr[i]]; + + if (bit == 0) { + (void) printf("***Invalid flag: %c\n", + flagstr[i]); + continue; + } + flags |= bit; + + /* If it's not something with an argument, keep going */ + if ((bit & (ZDB_FLAG_CHECKSUM | + ZDB_FLAG_PRINT_BLKPTR)) == 0) + continue; + + p = &flagstr[i + 1]; + if (bit == ZDB_FLAG_PRINT_BLKPTR) + blkptr_offset = strtoull(p, &p, 16); + if (*p != ':' && *p != '\0') { + (void) printf("***Invalid flag arg: '%s'\n", s); + free(dup); + return; + } + } + } + + vd = zdb_vdev_lookup(spa->spa_root_vdev, vdev); + if (vd == NULL) { + (void) printf("***Invalid vdev: %s\n", vdev); + free(dup); + return; + } else { + if (vd->vdev_path) + (void) fprintf(stderr, "Found vdev: %s\n", + vd->vdev_path); + else + (void) fprintf(stderr, "Found vdev type: %s\n", + vd->vdev_ops->vdev_op_type); + } + + psize = size; + lsize = size; + + pbuf = umem_alloc(SPA_MAXBLOCKSIZE, UMEM_NOFAIL); + lbuf = umem_alloc(SPA_MAXBLOCKSIZE, UMEM_NOFAIL); + + BP_ZERO(bp); + + DVA_SET_VDEV(&dva[0], vd->vdev_id); + DVA_SET_OFFSET(&dva[0], offset); + DVA_SET_GANG(&dva[0], !!(flags & ZDB_FLAG_GBH)); + DVA_SET_ASIZE(&dva[0], vdev_psize_to_asize(vd, psize)); + + BP_SET_BIRTH(bp, TXG_INITIAL, TXG_INITIAL); + + BP_SET_LSIZE(bp, lsize); + BP_SET_PSIZE(bp, psize); + BP_SET_COMPRESS(bp, ZIO_COMPRESS_OFF); + BP_SET_CHECKSUM(bp, ZIO_CHECKSUM_OFF); + BP_SET_TYPE(bp, DMU_OT_NONE); + BP_SET_LEVEL(bp, 0); + BP_SET_DEDUP(bp, 0); + BP_SET_BYTEORDER(bp, ZFS_HOST_BYTEORDER); + + spa_config_enter(spa, SCL_STATE, FTAG, RW_READER); + zio = zio_root(spa, NULL, NULL, 0); + + if (vd == vd->vdev_top) { + /* + * Treat this as a normal block read. + */ + zio_nowait(zio_read(zio, spa, bp, pbuf, psize, NULL, NULL, + ZIO_PRIORITY_SYNC_READ, + ZIO_FLAG_CANFAIL | ZIO_FLAG_RAW, NULL)); + } else { + /* + * Treat this as a vdev child I/O. + */ + zio_nowait(zio_vdev_child_io(zio, bp, vd, offset, pbuf, psize, + ZIO_TYPE_READ, ZIO_PRIORITY_SYNC_READ, + ZIO_FLAG_DONT_CACHE | ZIO_FLAG_DONT_QUEUE | + ZIO_FLAG_DONT_PROPAGATE | ZIO_FLAG_DONT_RETRY | + ZIO_FLAG_CANFAIL | ZIO_FLAG_RAW, NULL, NULL)); + } + + error = zio_wait(zio); + spa_config_exit(spa, SCL_STATE, FTAG); + + if (error) { + (void) printf("Read of %s failed, error: %d\n", thing, error); + goto out; + } + + if (flags & ZDB_FLAG_DECOMPRESS) { + /* + * We don't know how the data was compressed, so just try + * every decompress function at every inflated blocksize. + */ + enum zio_compress c; + void *pbuf2 = umem_alloc(SPA_MAXBLOCKSIZE, UMEM_NOFAIL); + void *lbuf2 = umem_alloc(SPA_MAXBLOCKSIZE, UMEM_NOFAIL); + + bcopy(pbuf, pbuf2, psize); + + VERIFY(random_get_pseudo_bytes((uint8_t *)pbuf + psize, + SPA_MAXBLOCKSIZE - psize) == 0); + + VERIFY(random_get_pseudo_bytes((uint8_t *)pbuf2 + psize, + SPA_MAXBLOCKSIZE - psize) == 0); + + for (lsize = SPA_MAXBLOCKSIZE; lsize > psize; + lsize -= SPA_MINBLOCKSIZE) { + for (c = 0; c < ZIO_COMPRESS_FUNCTIONS; c++) { + if (zio_decompress_data(c, pbuf, lbuf, + psize, lsize) == 0 && + zio_decompress_data(c, pbuf2, lbuf2, + psize, lsize) == 0 && + bcmp(lbuf, lbuf2, lsize) == 0) + break; + } + if (c != ZIO_COMPRESS_FUNCTIONS) + break; + lsize -= SPA_MINBLOCKSIZE; + } + + umem_free(pbuf2, SPA_MAXBLOCKSIZE); + umem_free(lbuf2, SPA_MAXBLOCKSIZE); + + if (lsize <= psize) { + (void) printf("Decompress of %s failed\n", thing); + goto out; + } + buf = lbuf; + size = lsize; + } else { + buf = pbuf; + size = psize; + } + + if (flags & ZDB_FLAG_PRINT_BLKPTR) + zdb_print_blkptr((blkptr_t *)(void *) + ((uintptr_t)buf + (uintptr_t)blkptr_offset), flags); + else if (flags & ZDB_FLAG_RAW) + zdb_dump_block_raw(buf, size, flags); + else if (flags & ZDB_FLAG_INDIRECT) + zdb_dump_indirect((blkptr_t *)buf, size / sizeof (blkptr_t), + flags); + else if (flags & ZDB_FLAG_GBH) + zdb_dump_gbh(buf, flags); + else + zdb_dump_block(thing, buf, size, flags); + +out: + umem_free(pbuf, SPA_MAXBLOCKSIZE); + umem_free(lbuf, SPA_MAXBLOCKSIZE); + free(dup); +} + +static boolean_t +pool_match(nvlist_t *cfg, char *tgt) +{ + uint64_t v, guid = strtoull(tgt, NULL, 0); + char *s; + + if (guid != 0) { + if (nvlist_lookup_uint64(cfg, ZPOOL_CONFIG_POOL_GUID, &v) == 0) + return (v == guid); + } else { + if (nvlist_lookup_string(cfg, ZPOOL_CONFIG_POOL_NAME, &s) == 0) + return (strcmp(s, tgt) == 0); + } + return (B_FALSE); +} + +static char * +find_zpool(char **target, nvlist_t **configp, int dirc, char **dirv) +{ + nvlist_t *pools; + nvlist_t *match = NULL; + char *name = NULL; + char *sepp = NULL; + char sep; + int count = 0; + importargs_t args = { 0 }; + + args.paths = dirc; + args.path = dirv; + args.can_be_active = B_TRUE; + + if ((sepp = strpbrk(*target, "/@")) != NULL) { + sep = *sepp; + *sepp = '\0'; + } + + pools = zpool_search_import(g_zfs, &args); + + if (pools != NULL) { + nvpair_t *elem = NULL; + while ((elem = nvlist_next_nvpair(pools, elem)) != NULL) { + verify(nvpair_value_nvlist(elem, configp) == 0); + if (pool_match(*configp, *target)) { + count++; + if (match != NULL) { + /* print previously found config */ + if (name != NULL) { + (void) printf("%s\n", name); + dump_nvlist(match, 8); + name = NULL; + } + (void) printf("%s\n", + nvpair_name(elem)); + dump_nvlist(*configp, 8); + } else { + match = *configp; + name = nvpair_name(elem); + } + } + } + } + if (count > 1) + (void) fatal("\tMatched %d pools - use pool GUID " + "instead of pool name or \n" + "\tpool name part of a dataset name to select pool", count); + + if (sepp) + *sepp = sep; + /* + * If pool GUID was specified for pool id, replace it with pool name + */ + if (name && (strstr(*target, name) != *target)) { + int sz = 1 + strlen(name) + ((sepp) ? strlen(sepp) : 0); + + *target = umem_alloc(sz, UMEM_NOFAIL); + (void) snprintf(*target, sz, "%s%s", name, sepp ? sepp : ""); + } + + *configp = name ? match : NULL; + + return (name); +} + +int +main(int argc, char **argv) +{ + int i, c; + struct rlimit rl = { 1024, 1024 }; + spa_t *spa = NULL; + objset_t *os = NULL; + int dump_all = 1; + int verbose = 0; + int error = 0; + char **searchdirs = NULL; + int nsearch = 0; + char *target; + nvlist_t *policy = NULL; + uint64_t max_txg = UINT64_MAX; + int rewind = ZPOOL_NEVER_REWIND; + + (void) setrlimit(RLIMIT_NOFILE, &rl); + (void) enable_extended_FILE_stdio(-1, -1); + + dprintf_setup(&argc, argv); + + while ((c = getopt(argc, argv, "bcdhilmsuCDRSAFLXevp:t:U:P")) != -1) { + switch (c) { + case 'b': + case 'c': + case 'd': + case 'h': + case 'i': + case 'l': + case 'm': + case 's': + case 'u': + case 'C': + case 'D': + case 'R': + case 'S': + dump_opt[c]++; + dump_all = 0; + break; + case 'A': + case 'F': + case 'L': + case 'X': + case 'e': + case 'P': + dump_opt[c]++; + break; + case 'v': + verbose++; + break; + case 'p': + if (searchdirs == NULL) { + searchdirs = umem_alloc(sizeof (char *), + UMEM_NOFAIL); + } else { + char **tmp = umem_alloc((nsearch + 1) * + sizeof (char *), UMEM_NOFAIL); + bcopy(searchdirs, tmp, nsearch * + sizeof (char *)); + umem_free(searchdirs, + nsearch * sizeof (char *)); + searchdirs = tmp; + } + searchdirs[nsearch++] = optarg; + break; + case 't': + max_txg = strtoull(optarg, NULL, 0); + if (max_txg < TXG_INITIAL) { + (void) fprintf(stderr, "incorrect txg " + "specified: %s\n", optarg); + usage(); + } + break; + case 'U': + spa_config_path = optarg; + break; + default: + usage(); + break; + } + } + + if (!dump_opt['e'] && searchdirs != NULL) { + (void) fprintf(stderr, "-p option requires use of -e\n"); + usage(); + } + + kernel_init(FREAD); + g_zfs = libzfs_init(); + ASSERT(g_zfs != NULL); + + if (dump_all) + verbose = MAX(verbose, 1); + + for (c = 0; c < 256; c++) { + if (dump_all && !strchr("elAFLRSXP", c)) + dump_opt[c] = 1; + if (dump_opt[c]) + dump_opt[c] += verbose; + } + + aok = (dump_opt['A'] == 1) || (dump_opt['A'] > 2); + zfs_recover = (dump_opt['A'] > 1); + + argc -= optind; + argv += optind; + + if (argc < 2 && dump_opt['R']) + usage(); + if (argc < 1) { + if (!dump_opt['e'] && dump_opt['C']) { + dump_cachefile(spa_config_path); + return (0); + } + usage(); + } + + if (dump_opt['l']) { + dump_label(argv[0]); + return (0); + } + + if (dump_opt['X'] || dump_opt['F']) + rewind = ZPOOL_DO_REWIND | + (dump_opt['X'] ? ZPOOL_EXTREME_REWIND : 0); + + if (nvlist_alloc(&policy, NV_UNIQUE_NAME_TYPE, 0) != 0 || + nvlist_add_uint64(policy, ZPOOL_REWIND_REQUEST_TXG, max_txg) != 0 || + nvlist_add_uint32(policy, ZPOOL_REWIND_REQUEST, rewind) != 0) + fatal("internal error: %s", strerror(ENOMEM)); + + error = 0; + target = argv[0]; + + if (dump_opt['e']) { + nvlist_t *cfg = NULL; + char *name = find_zpool(&target, &cfg, nsearch, searchdirs); + + error = ENOENT; + if (name) { + if (dump_opt['C'] > 1) { + (void) printf("\nConfiguration for import:\n"); + dump_nvlist(cfg, 8); + } + if (nvlist_add_nvlist(cfg, + ZPOOL_REWIND_POLICY, policy) != 0) { + fatal("can't open '%s': %s", + target, strerror(ENOMEM)); + } + if ((error = spa_import(name, cfg, NULL, + ZFS_IMPORT_MISSING_LOG)) != 0) { + error = spa_import(name, cfg, NULL, + ZFS_IMPORT_VERBATIM); + } + } + } + + if (error == 0) { + if (strpbrk(target, "/@") == NULL || dump_opt['R']) { + error = spa_open_rewind(target, &spa, FTAG, policy, + NULL); + if (error) { + /* + * If we're missing the log device then + * try opening the pool after clearing the + * log state. + */ + mutex_enter(&spa_namespace_lock); + if ((spa = spa_lookup(target)) != NULL && + spa->spa_log_state == SPA_LOG_MISSING) { + spa->spa_log_state = SPA_LOG_CLEAR; + error = 0; + } + mutex_exit(&spa_namespace_lock); + + if (!error) { + error = spa_open_rewind(target, &spa, + FTAG, policy, NULL); + } + } + } else { + error = dmu_objset_own(target, DMU_OST_ANY, + B_TRUE, FTAG, &os); + } + } + nvlist_free(policy); + + if (error) + fatal("can't open '%s': %s", target, strerror(error)); + + argv++; + argc--; + if (!dump_opt['R']) { + if (argc > 0) { + zopt_objects = argc; + zopt_object = calloc(zopt_objects, sizeof (uint64_t)); + for (i = 0; i < zopt_objects; i++) { + errno = 0; + zopt_object[i] = strtoull(argv[i], NULL, 0); + if (zopt_object[i] == 0 && errno != 0) + fatal("bad number %s: %s", + argv[i], strerror(errno)); + } + } + (os != NULL) ? dump_dir(os) : dump_zpool(spa); + } else { + flagbits['b'] = ZDB_FLAG_PRINT_BLKPTR; + flagbits['c'] = ZDB_FLAG_CHECKSUM; + flagbits['d'] = ZDB_FLAG_DECOMPRESS; + flagbits['e'] = ZDB_FLAG_BSWAP; + flagbits['g'] = ZDB_FLAG_GBH; + flagbits['i'] = ZDB_FLAG_INDIRECT; + flagbits['p'] = ZDB_FLAG_PHYS; + flagbits['r'] = ZDB_FLAG_RAW; + + for (i = 0; i < argc; i++) + zdb_read_block(argv[i], spa); + } + + (os != NULL) ? dmu_objset_disown(os, FTAG) : spa_close(spa, FTAG); + + fuid_table_destroy(); + sa_loaded = B_FALSE; + + libzfs_fini(g_zfs); + kernel_fini(); + + return (0); +} diff --git a/cmd/zdb/zdb_il.c b/cmd/zdb/zdb_il.c new file mode 100644 index 000000000000..a0ed985f52b7 --- /dev/null +++ b/cmd/zdb/zdb_il.c @@ -0,0 +1,384 @@ +/* + * 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 http://www.opensolaris.org/os/licensing. + * 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 2009 Sun Microsystems, Inc. All rights reserved. + * Use is subject to license terms. + */ + +/* + * Print intent log header and statistics. + */ + +#include <stdio.h> +#include <stdlib.h> +#include <ctype.h> +#include <sys/zfs_context.h> +#include <sys/spa.h> +#include <sys/dmu.h> +#include <sys/stat.h> +#include <sys/resource.h> +#include <sys/zil.h> +#include <sys/zil_impl.h> + +extern uint8_t dump_opt[256]; + +static char prefix[4] = "\t\t\t"; + +static void +print_log_bp(const blkptr_t *bp, const char *prefix) +{ + char blkbuf[BP_SPRINTF_LEN]; + + sprintf_blkptr(blkbuf, bp); + (void) printf("%s%s\n", prefix, blkbuf); +} + +/* ARGSUSED */ +static void +zil_prt_rec_create(zilog_t *zilog, int txtype, lr_create_t *lr) +{ + time_t crtime = lr->lr_crtime[0]; + char *name, *link; + lr_attr_t *lrattr; + + name = (char *)(lr + 1); + + if (lr->lr_common.lrc_txtype == TX_CREATE_ATTR || + lr->lr_common.lrc_txtype == TX_MKDIR_ATTR) { + lrattr = (lr_attr_t *)(lr + 1); + name += ZIL_XVAT_SIZE(lrattr->lr_attr_masksize); + } + + if (txtype == TX_SYMLINK) { + link = name + strlen(name) + 1; + (void) printf("%s%s -> %s\n", prefix, name, link); + } else if (txtype != TX_MKXATTR) { + (void) printf("%s%s\n", prefix, name); + } + + (void) printf("%s%s", prefix, ctime(&crtime)); + (void) printf("%sdoid %llu, foid %llu, mode %llo\n", prefix, + (u_longlong_t)lr->lr_doid, (u_longlong_t)lr->lr_foid, + (longlong_t)lr->lr_mode); + (void) printf("%suid %llu, gid %llu, gen %llu, rdev 0x%llx\n", prefix, + (u_longlong_t)lr->lr_uid, (u_longlong_t)lr->lr_gid, + (u_longlong_t)lr->lr_gen, (u_longlong_t)lr->lr_rdev); +} + +/* ARGSUSED */ +static void +zil_prt_rec_remove(zilog_t *zilog, int txtype, lr_remove_t *lr) +{ + (void) printf("%sdoid %llu, name %s\n", prefix, + (u_longlong_t)lr->lr_doid, (char *)(lr + 1)); +} + +/* ARGSUSED */ +static void +zil_prt_rec_link(zilog_t *zilog, int txtype, lr_link_t *lr) +{ + (void) printf("%sdoid %llu, link_obj %llu, name %s\n", prefix, + (u_longlong_t)lr->lr_doid, (u_longlong_t)lr->lr_link_obj, + (char *)(lr + 1)); +} + +/* ARGSUSED */ +static void +zil_prt_rec_rename(zilog_t *zilog, int txtype, lr_rename_t *lr) +{ + char *snm = (char *)(lr + 1); + char *tnm = snm + strlen(snm) + 1; + + (void) printf("%ssdoid %llu, tdoid %llu\n", prefix, + (u_longlong_t)lr->lr_sdoid, (u_longlong_t)lr->lr_tdoid); + (void) printf("%ssrc %s tgt %s\n", prefix, snm, tnm); +} + +/* ARGSUSED */ +static void +zil_prt_rec_write(zilog_t *zilog, int txtype, lr_write_t *lr) +{ + char *data, *dlimit; + blkptr_t *bp = &lr->lr_blkptr; + zbookmark_t zb; + char buf[SPA_MAXBLOCKSIZE]; + int verbose = MAX(dump_opt['d'], dump_opt['i']); + int error; + + (void) printf("%sfoid %llu, offset %llx, length %llx\n", prefix, + (u_longlong_t)lr->lr_foid, (u_longlong_t)lr->lr_offset, + (u_longlong_t)lr->lr_length); + + if (txtype == TX_WRITE2 || verbose < 5) + return; + + if (lr->lr_common.lrc_reclen == sizeof (lr_write_t)) { + (void) printf("%shas blkptr, %s\n", prefix, + bp->blk_birth >= spa_first_txg(zilog->zl_spa) ? + "will claim" : "won't claim"); + print_log_bp(bp, prefix); + + if (BP_IS_HOLE(bp)) { + (void) printf("\t\t\tLSIZE 0x%llx\n", + (u_longlong_t)BP_GET_LSIZE(bp)); + } + if (bp->blk_birth == 0) { + bzero(buf, sizeof (buf)); + (void) printf("%s<hole>\n", prefix); + return; + } + if (bp->blk_birth < zilog->zl_header->zh_claim_txg) { + (void) printf("%s<block already committed>\n", prefix); + return; + } + + SET_BOOKMARK(&zb, dmu_objset_id(zilog->zl_os), + lr->lr_foid, ZB_ZIL_LEVEL, + lr->lr_offset / BP_GET_LSIZE(bp)); + + error = zio_wait(zio_read(NULL, zilog->zl_spa, + bp, buf, BP_GET_LSIZE(bp), NULL, NULL, + ZIO_PRIORITY_SYNC_READ, ZIO_FLAG_CANFAIL, &zb)); + if (error) + return; + data = buf; + } else { + data = (char *)(lr + 1); + } + + dlimit = data + MIN(lr->lr_length, + (verbose < 6 ? 20 : SPA_MAXBLOCKSIZE)); + + (void) printf("%s", prefix); + while (data < dlimit) { + if (isprint(*data)) + (void) printf("%c ", *data); + else + (void) printf("%2X", *data); + data++; + } + (void) printf("\n"); +} + +/* ARGSUSED */ +static void +zil_prt_rec_truncate(zilog_t *zilog, int txtype, lr_truncate_t *lr) +{ + (void) printf("%sfoid %llu, offset 0x%llx, length 0x%llx\n", prefix, + (u_longlong_t)lr->lr_foid, (longlong_t)lr->lr_offset, + (u_longlong_t)lr->lr_length); +} + +/* ARGSUSED */ +static void +zil_prt_rec_setattr(zilog_t *zilog, int txtype, lr_setattr_t *lr) +{ + time_t atime = (time_t)lr->lr_atime[0]; + time_t mtime = (time_t)lr->lr_mtime[0]; + + (void) printf("%sfoid %llu, mask 0x%llx\n", prefix, + (u_longlong_t)lr->lr_foid, (u_longlong_t)lr->lr_mask); + + if (lr->lr_mask & AT_MODE) { + (void) printf("%sAT_MODE %llo\n", prefix, + (longlong_t)lr->lr_mode); + } + + if (lr->lr_mask & AT_UID) { + (void) printf("%sAT_UID %llu\n", prefix, + (u_longlong_t)lr->lr_uid); + } + + if (lr->lr_mask & AT_GID) { + (void) printf("%sAT_GID %llu\n", prefix, + (u_longlong_t)lr->lr_gid); + } + + if (lr->lr_mask & AT_SIZE) { + (void) printf("%sAT_SIZE %llu\n", prefix, + (u_longlong_t)lr->lr_size); + } + + if (lr->lr_mask & AT_ATIME) { + (void) printf("%sAT_ATIME %llu.%09llu %s", prefix, + (u_longlong_t)lr->lr_atime[0], + (u_longlong_t)lr->lr_atime[1], + ctime(&atime)); + } + + if (lr->lr_mask & AT_MTIME) { + (void) printf("%sAT_MTIME %llu.%09llu %s", prefix, + (u_longlong_t)lr->lr_mtime[0], + (u_longlong_t)lr->lr_mtime[1], + ctime(&mtime)); + } +} + +/* ARGSUSED */ +static void +zil_prt_rec_acl(zilog_t *zilog, int txtype, lr_acl_t *lr) +{ + (void) printf("%sfoid %llu, aclcnt %llu\n", prefix, + (u_longlong_t)lr->lr_foid, (u_longlong_t)lr->lr_aclcnt); +} + +typedef void (*zil_prt_rec_func_t)(); +typedef struct zil_rec_info { + zil_prt_rec_func_t zri_print; + char *zri_name; + uint64_t zri_count; +} zil_rec_info_t; + +static zil_rec_info_t zil_rec_info[TX_MAX_TYPE] = { + { NULL, "Total " }, + { zil_prt_rec_create, "TX_CREATE " }, + { zil_prt_rec_create, "TX_MKDIR " }, + { zil_prt_rec_create, "TX_MKXATTR " }, + { zil_prt_rec_create, "TX_SYMLINK " }, + { zil_prt_rec_remove, "TX_REMOVE " }, + { zil_prt_rec_remove, "TX_RMDIR " }, + { zil_prt_rec_link, "TX_LINK " }, + { zil_prt_rec_rename, "TX_RENAME " }, + { zil_prt_rec_write, "TX_WRITE " }, + { zil_prt_rec_truncate, "TX_TRUNCATE " }, + { zil_prt_rec_setattr, "TX_SETATTR " }, + { zil_prt_rec_acl, "TX_ACL_V0 " }, + { zil_prt_rec_acl, "TX_ACL_ACL " }, + { zil_prt_rec_create, "TX_CREATE_ACL " }, + { zil_prt_rec_create, "TX_CREATE_ATTR " }, + { zil_prt_rec_create, "TX_CREATE_ACL_ATTR " }, + { zil_prt_rec_create, "TX_MKDIR_ACL " }, + { zil_prt_rec_create, "TX_MKDIR_ATTR " }, + { zil_prt_rec_create, "TX_MKDIR_ACL_ATTR " }, + { zil_prt_rec_write, "TX_WRITE2 " }, +}; + +/* ARGSUSED */ +static int +print_log_record(zilog_t *zilog, lr_t *lr, void *arg, uint64_t claim_txg) +{ + int txtype; + int verbose = MAX(dump_opt['d'], dump_opt['i']); + + /* reduce size of txtype to strip off TX_CI bit */ + txtype = lr->lrc_txtype; + + ASSERT(txtype != 0 && (uint_t)txtype < TX_MAX_TYPE); + ASSERT(lr->lrc_txg); + + (void) printf("\t\t%s%s len %6llu, txg %llu, seq %llu\n", + (lr->lrc_txtype & TX_CI) ? "CI-" : "", + zil_rec_info[txtype].zri_name, + (u_longlong_t)lr->lrc_reclen, + (u_longlong_t)lr->lrc_txg, + (u_longlong_t)lr->lrc_seq); + + if (txtype && verbose >= 3) + zil_rec_info[txtype].zri_print(zilog, txtype, lr); + + zil_rec_info[txtype].zri_count++; + zil_rec_info[0].zri_count++; + + return (0); +} + +/* ARGSUSED */ +static int +print_log_block(zilog_t *zilog, blkptr_t *bp, void *arg, uint64_t claim_txg) +{ + char blkbuf[BP_SPRINTF_LEN + 10]; + int verbose = MAX(dump_opt['d'], dump_opt['i']); + char *claim; + + if (verbose <= 3) + return (0); + + if (verbose >= 5) { + (void) strcpy(blkbuf, ", "); + sprintf_blkptr(blkbuf + strlen(blkbuf), bp); + } else { + blkbuf[0] = '\0'; + } + + if (claim_txg != 0) + claim = "already claimed"; + else if (bp->blk_birth >= spa_first_txg(zilog->zl_spa)) + claim = "will claim"; + else + claim = "won't claim"; + + (void) printf("\tBlock seqno %llu, %s%s\n", + (u_longlong_t)bp->blk_cksum.zc_word[ZIL_ZC_SEQ], claim, blkbuf); + + return (0); +} + +static void +print_log_stats(int verbose) +{ + int i, w, p10; + + if (verbose > 3) + (void) printf("\n"); + + if (zil_rec_info[0].zri_count == 0) + return; + + for (w = 1, p10 = 10; zil_rec_info[0].zri_count >= p10; p10 *= 10) + w++; + + for (i = 0; i < TX_MAX_TYPE; i++) + if (zil_rec_info[i].zri_count || verbose >= 3) + (void) printf("\t\t%s %*llu\n", + zil_rec_info[i].zri_name, w, + (u_longlong_t)zil_rec_info[i].zri_count); + (void) printf("\n"); +} + +/* ARGSUSED */ +void +dump_intent_log(zilog_t *zilog) +{ + const zil_header_t *zh = zilog->zl_header; + int verbose = MAX(dump_opt['d'], dump_opt['i']); + int i; + + if (zh->zh_log.blk_birth == 0 || verbose < 1) + return; + + (void) printf("\n ZIL header: claim_txg %llu, " + "claim_blk_seq %llu, claim_lr_seq %llu", + (u_longlong_t)zh->zh_claim_txg, + (u_longlong_t)zh->zh_claim_blk_seq, + (u_longlong_t)zh->zh_claim_lr_seq); + (void) printf(" replay_seq %llu, flags 0x%llx\n", + (u_longlong_t)zh->zh_replay_seq, (u_longlong_t)zh->zh_flags); + + for (i = 0; i < TX_MAX_TYPE; i++) + zil_rec_info[i].zri_count = 0; + + if (verbose >= 2) { + (void) printf("\n"); + (void) zil_parse(zilog, print_log_block, print_log_record, NULL, + zh->zh_claim_txg); + print_log_stats(verbose); + } +} diff --git a/cmd/zfs/zfs_iter.c b/cmd/zfs/zfs_iter.c new file mode 100644 index 000000000000..e2ab90eaf14f --- /dev/null +++ b/cmd/zfs/zfs_iter.c @@ -0,0 +1,464 @@ +/* + * 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 http://www.opensolaris.org/os/licensing. + * 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. + */ + +#include <libintl.h> +#include <libuutil.h> +#include <stddef.h> +#include <stdio.h> +#include <stdlib.h> +#include <strings.h> + +#include <libzfs.h> + +#include "zfs_util.h" +#include "zfs_iter.h" + +/* + * This is a private interface used to gather up all the datasets specified on + * the command line so that we can iterate over them in order. + * + * First, we iterate over all filesystems, gathering them together into an + * AVL tree. We report errors for any explicitly specified datasets + * that we couldn't open. + * + * When finished, we have an AVL tree of ZFS handles. We go through and execute + * the provided callback for each one, passing whatever data the user supplied. + */ + +typedef struct zfs_node { + zfs_handle_t *zn_handle; + uu_avl_node_t zn_avlnode; +} zfs_node_t; + +typedef struct callback_data { + uu_avl_t *cb_avl; + int cb_flags; + zfs_type_t cb_types; + zfs_sort_column_t *cb_sortcol; + zprop_list_t **cb_proplist; + int cb_depth_limit; + int cb_depth; + uint8_t cb_props_table[ZFS_NUM_PROPS]; +} callback_data_t; + +uu_avl_pool_t *avl_pool; + +/* + * Include snaps if they were requested or if this a zfs list where types + * were not specified and the "listsnapshots" property is set on this pool. + */ +static int +zfs_include_snapshots(zfs_handle_t *zhp, callback_data_t *cb) +{ + zpool_handle_t *zph; + + if ((cb->cb_flags & ZFS_ITER_PROP_LISTSNAPS) == 0) + return (cb->cb_types & ZFS_TYPE_SNAPSHOT); + + zph = zfs_get_pool_handle(zhp); + return (zpool_get_prop_int(zph, ZPOOL_PROP_LISTSNAPS, NULL)); +} + +/* + * Called for each dataset. If the object is of an appropriate type, + * add it to the avl tree and recurse over any children as necessary. + */ +static int +zfs_callback(zfs_handle_t *zhp, void *data) +{ + callback_data_t *cb = data; + int dontclose = 0; + int include_snaps = zfs_include_snapshots(zhp, cb); + + if ((zfs_get_type(zhp) & cb->cb_types) || + ((zfs_get_type(zhp) == ZFS_TYPE_SNAPSHOT) && include_snaps)) { + uu_avl_index_t idx; + zfs_node_t *node = safe_malloc(sizeof (zfs_node_t)); + + node->zn_handle = zhp; + uu_avl_node_init(node, &node->zn_avlnode, avl_pool); + if (uu_avl_find(cb->cb_avl, node, cb->cb_sortcol, + &idx) == NULL) { + if (cb->cb_proplist) { + if ((*cb->cb_proplist) && + !(*cb->cb_proplist)->pl_all) + zfs_prune_proplist(zhp, + cb->cb_props_table); + + if (zfs_expand_proplist(zhp, cb->cb_proplist, + (cb->cb_flags & ZFS_ITER_RECVD_PROPS)) + != 0) { + free(node); + return (-1); + } + } + uu_avl_insert(cb->cb_avl, node, idx); + dontclose = 1; + } else { + free(node); + } + } + + /* + * Recurse if necessary. + */ + if (cb->cb_flags & ZFS_ITER_RECURSE && + ((cb->cb_flags & ZFS_ITER_DEPTH_LIMIT) == 0 || + cb->cb_depth < cb->cb_depth_limit)) { + cb->cb_depth++; + if (zfs_get_type(zhp) == ZFS_TYPE_FILESYSTEM) + (void) zfs_iter_filesystems(zhp, zfs_callback, data); + if ((zfs_get_type(zhp) != ZFS_TYPE_SNAPSHOT) && include_snaps) + (void) zfs_iter_snapshots(zhp, zfs_callback, data); + cb->cb_depth--; + } + + if (!dontclose) + zfs_close(zhp); + + return (0); +} + +int +zfs_add_sort_column(zfs_sort_column_t **sc, const char *name, + boolean_t reverse) +{ + zfs_sort_column_t *col; + zfs_prop_t prop; + + if ((prop = zfs_name_to_prop(name)) == ZPROP_INVAL && + !zfs_prop_user(name)) + return (-1); + + col = safe_malloc(sizeof (zfs_sort_column_t)); + + col->sc_prop = prop; + col->sc_reverse = reverse; + if (prop == ZPROP_INVAL) { + col->sc_user_prop = safe_malloc(strlen(name) + 1); + (void) strcpy(col->sc_user_prop, name); + } + + if (*sc == NULL) { + col->sc_last = col; + *sc = col; + } else { + (*sc)->sc_last->sc_next = col; + (*sc)->sc_last = col; + } + + return (0); +} + +void +zfs_free_sort_columns(zfs_sort_column_t *sc) +{ + zfs_sort_column_t *col; + + while (sc != NULL) { + col = sc->sc_next; + free(sc->sc_user_prop); + free(sc); + sc = col; + } +} + +/* ARGSUSED */ +static int +zfs_compare(const void *larg, const void *rarg, void *unused) +{ + zfs_handle_t *l = ((zfs_node_t *)larg)->zn_handle; + zfs_handle_t *r = ((zfs_node_t *)rarg)->zn_handle; + const char *lname = zfs_get_name(l); + const char *rname = zfs_get_name(r); + char *lat, *rat; + uint64_t lcreate, rcreate; + int ret; + + lat = (char *)strchr(lname, '@'); + rat = (char *)strchr(rname, '@'); + + if (lat != NULL) + *lat = '\0'; + if (rat != NULL) + *rat = '\0'; + + ret = strcmp(lname, rname); + if (ret == 0) { + /* + * If we're comparing a dataset to one of its snapshots, we + * always make the full dataset first. + */ + if (lat == NULL) { + ret = -1; + } else if (rat == NULL) { + ret = 1; + } else { + /* + * If we have two snapshots from the same dataset, then + * we want to sort them according to creation time. We + * use the hidden CREATETXG property to get an absolute + * ordering of snapshots. + */ + lcreate = zfs_prop_get_int(l, ZFS_PROP_CREATETXG); + rcreate = zfs_prop_get_int(r, ZFS_PROP_CREATETXG); + + if (lcreate < rcreate) + ret = -1; + else if (lcreate > rcreate) + ret = 1; + } + } + + if (lat != NULL) + *lat = '@'; + if (rat != NULL) + *rat = '@'; + + return (ret); +} + +/* + * Sort datasets by specified columns. + * + * o Numeric types sort in ascending order. + * o String types sort in alphabetical order. + * o Types inappropriate for a row sort that row to the literal + * bottom, regardless of the specified ordering. + * + * If no sort columns are specified, or two datasets compare equally + * across all specified columns, they are sorted alphabetically by name + * with snapshots grouped under their parents. + */ +static int +zfs_sort(const void *larg, const void *rarg, void *data) +{ + zfs_handle_t *l = ((zfs_node_t *)larg)->zn_handle; + zfs_handle_t *r = ((zfs_node_t *)rarg)->zn_handle; + zfs_sort_column_t *sc = (zfs_sort_column_t *)data; + zfs_sort_column_t *psc; + + for (psc = sc; psc != NULL; psc = psc->sc_next) { + char lbuf[ZFS_MAXPROPLEN], rbuf[ZFS_MAXPROPLEN]; + char *lstr, *rstr; + uint64_t lnum, rnum; + boolean_t lvalid, rvalid; + int ret = 0; + + /* + * We group the checks below the generic code. If 'lstr' and + * 'rstr' are non-NULL, then we do a string based comparison. + * Otherwise, we compare 'lnum' and 'rnum'. + */ + lstr = rstr = NULL; + if (psc->sc_prop == ZPROP_INVAL) { + nvlist_t *luser, *ruser; + nvlist_t *lval, *rval; + + luser = zfs_get_user_props(l); + ruser = zfs_get_user_props(r); + + lvalid = (nvlist_lookup_nvlist(luser, + psc->sc_user_prop, &lval) == 0); + rvalid = (nvlist_lookup_nvlist(ruser, + psc->sc_user_prop, &rval) == 0); + + if (lvalid) + verify(nvlist_lookup_string(lval, + ZPROP_VALUE, &lstr) == 0); + if (rvalid) + verify(nvlist_lookup_string(rval, + ZPROP_VALUE, &rstr) == 0); + + } else if (zfs_prop_is_string(psc->sc_prop)) { + lvalid = (zfs_prop_get(l, psc->sc_prop, lbuf, + sizeof (lbuf), NULL, NULL, 0, B_TRUE) == 0); + rvalid = (zfs_prop_get(r, psc->sc_prop, rbuf, + sizeof (rbuf), NULL, NULL, 0, B_TRUE) == 0); + + lstr = lbuf; + rstr = rbuf; + } else { + lvalid = zfs_prop_valid_for_type(psc->sc_prop, + zfs_get_type(l)); + rvalid = zfs_prop_valid_for_type(psc->sc_prop, + zfs_get_type(r)); + + if (lvalid) + (void) zfs_prop_get_numeric(l, psc->sc_prop, + &lnum, NULL, NULL, 0); + if (rvalid) + (void) zfs_prop_get_numeric(r, psc->sc_prop, + &rnum, NULL, NULL, 0); + } + + if (!lvalid && !rvalid) + continue; + else if (!lvalid) + return (1); + else if (!rvalid) + return (-1); + + if (lstr) + ret = strcmp(lstr, rstr); + else if (lnum < rnum) + ret = -1; + else if (lnum > rnum) + ret = 1; + + if (ret != 0) { + if (psc->sc_reverse == B_TRUE) + ret = (ret < 0) ? 1 : -1; + return (ret); + } + } + + return (zfs_compare(larg, rarg, NULL)); +} + +int +zfs_for_each(int argc, char **argv, int flags, zfs_type_t types, + zfs_sort_column_t *sortcol, zprop_list_t **proplist, int limit, + zfs_iter_f callback, void *data) +{ + callback_data_t cb = {0}; + int ret = 0; + zfs_node_t *node; + uu_avl_walk_t *walk; + + avl_pool = uu_avl_pool_create("zfs_pool", sizeof (zfs_node_t), + offsetof(zfs_node_t, zn_avlnode), zfs_sort, UU_DEFAULT); + + if (avl_pool == NULL) + nomem(); + + cb.cb_sortcol = sortcol; + cb.cb_flags = flags; + cb.cb_proplist = proplist; + cb.cb_types = types; + cb.cb_depth_limit = limit; + /* + * If cb_proplist is provided then in the zfs_handles created we + * retain only those properties listed in cb_proplist and sortcol. + * The rest are pruned. So, the caller should make sure that no other + * properties other than those listed in cb_proplist/sortcol are + * accessed. + * + * If cb_proplist is NULL then we retain all the properties. We + * always retain the zoned property, which some other properties + * need (userquota & friends), and the createtxg property, which + * we need to sort snapshots. + */ + if (cb.cb_proplist && *cb.cb_proplist) { + zprop_list_t *p = *cb.cb_proplist; + + while (p) { + if (p->pl_prop >= ZFS_PROP_TYPE && + p->pl_prop < ZFS_NUM_PROPS) { + cb.cb_props_table[p->pl_prop] = B_TRUE; + } + p = p->pl_next; + } + + while (sortcol) { + if (sortcol->sc_prop >= ZFS_PROP_TYPE && + sortcol->sc_prop < ZFS_NUM_PROPS) { + cb.cb_props_table[sortcol->sc_prop] = B_TRUE; + } + sortcol = sortcol->sc_next; + } + + cb.cb_props_table[ZFS_PROP_ZONED] = B_TRUE; + cb.cb_props_table[ZFS_PROP_CREATETXG] = B_TRUE; + } else { + (void) memset(cb.cb_props_table, B_TRUE, + sizeof (cb.cb_props_table)); + } + + if ((cb.cb_avl = uu_avl_create(avl_pool, NULL, UU_DEFAULT)) == NULL) + nomem(); + + if (argc == 0) { + /* + * If given no arguments, iterate over all datasets. + */ + cb.cb_flags |= ZFS_ITER_RECURSE; + ret = zfs_iter_root(g_zfs, zfs_callback, &cb); + } else { + int i; + zfs_handle_t *zhp; + zfs_type_t argtype; + + /* + * If we're recursive, then we always allow filesystems as + * arguments. If we also are interested in snapshots, then we + * can take volumes as well. + */ + argtype = types; + if (flags & ZFS_ITER_RECURSE) { + argtype |= ZFS_TYPE_FILESYSTEM; + if (types & ZFS_TYPE_SNAPSHOT) + argtype |= ZFS_TYPE_VOLUME; + } + + for (i = 0; i < argc; i++) { + if (flags & ZFS_ITER_ARGS_CAN_BE_PATHS) { + zhp = zfs_path_to_zhandle(g_zfs, argv[i], + argtype); + } else { + zhp = zfs_open(g_zfs, argv[i], argtype); + } + if (zhp != NULL) + ret |= zfs_callback(zhp, &cb); + else + ret = 1; + } + } + + /* + * At this point we've got our AVL tree full of zfs handles, so iterate + * over each one and execute the real user callback. + */ + for (node = uu_avl_first(cb.cb_avl); node != NULL; + node = uu_avl_next(cb.cb_avl, node)) + ret |= callback(node->zn_handle, data); + + /* + * Finally, clean up the AVL tree. + */ + if ((walk = uu_avl_walk_start(cb.cb_avl, UU_WALK_ROBUST)) == NULL) + nomem(); + + while ((node = uu_avl_walk_next(walk)) != NULL) { + uu_avl_remove(cb.cb_avl, node); + zfs_close(node->zn_handle); + free(node); + } + + uu_avl_walk_end(walk); + uu_avl_destroy(cb.cb_avl); + uu_avl_pool_destroy(avl_pool); + + return (ret); +} diff --git a/cmd/zfs/zfs_iter.h b/cmd/zfs/zfs_iter.h new file mode 100644 index 000000000000..8c6b9fdef54f --- /dev/null +++ b/cmd/zfs/zfs_iter.h @@ -0,0 +1,56 @@ +/* + * 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 http://www.opensolaris.org/os/licensing. + * 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 2009 Sun Microsystems, Inc. All rights reserved. + * Use is subject to license terms. + */ + +#ifndef ZFS_ITER_H +#define ZFS_ITER_H + +#ifdef __cplusplus +extern "C" { +#endif + +typedef struct zfs_sort_column { + struct zfs_sort_column *sc_next; + struct zfs_sort_column *sc_last; + zfs_prop_t sc_prop; + char *sc_user_prop; + boolean_t sc_reverse; +} zfs_sort_column_t; + +#define ZFS_ITER_RECURSE (1 << 0) +#define ZFS_ITER_ARGS_CAN_BE_PATHS (1 << 1) +#define ZFS_ITER_PROP_LISTSNAPS (1 << 2) +#define ZFS_ITER_DEPTH_LIMIT (1 << 3) +#define ZFS_ITER_RECVD_PROPS (1 << 4) + +int zfs_for_each(int, char **, int options, zfs_type_t, + zfs_sort_column_t *, zprop_list_t **, int, zfs_iter_f, void *); +int zfs_add_sort_column(zfs_sort_column_t **, const char *, boolean_t); +void zfs_free_sort_columns(zfs_sort_column_t *); + +#ifdef __cplusplus +} +#endif + +#endif /* ZFS_ITER_H */ diff --git a/cmd/zfs/zfs_main.c b/cmd/zfs/zfs_main.c new file mode 100644 index 000000000000..9516697390ef --- /dev/null +++ b/cmd/zfs/zfs_main.c @@ -0,0 +1,4160 @@ +/* + * 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 http://www.opensolaris.org/os/licensing. + * 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. + */ + +#include <assert.h> +#include <ctype.h> +#include <errno.h> +#include <libgen.h> +#include <libintl.h> +#include <libuutil.h> +#include <libnvpair.h> +#include <locale.h> +#include <stddef.h> +#include <stdio.h> +#include <stdlib.h> +#include <strings.h> +#include <unistd.h> +#include <fcntl.h> +#include <zone.h> +#include <grp.h> +#include <pwd.h> +#include <signal.h> +#include <sys/mkdev.h> +#include <sys/mntent.h> +#include <sys/mnttab.h> +#include <sys/mount.h> +#include <sys/stat.h> +#include <sys/fs/zfs.h> +#include <sys/types.h> +#include <time.h> + +#include <libzfs.h> +#include <libuutil.h> + +#include "zfs_iter.h" +#include "zfs_util.h" +#include "zfs_comutil.h" + +libzfs_handle_t *g_zfs; + +static FILE *mnttab_file; +static char history_str[HIS_MAX_RECORD_LEN]; +const char *pypath = "/usr/lib/zfs/pyzfs.py"; + +static int zfs_do_clone(int argc, char **argv); +static int zfs_do_create(int argc, char **argv); +static int zfs_do_destroy(int argc, char **argv); +static int zfs_do_get(int argc, char **argv); +static int zfs_do_inherit(int argc, char **argv); +static int zfs_do_list(int argc, char **argv); +static int zfs_do_mount(int argc, char **argv); +static int zfs_do_rename(int argc, char **argv); +static int zfs_do_rollback(int argc, char **argv); +static int zfs_do_set(int argc, char **argv); +static int zfs_do_upgrade(int argc, char **argv); +static int zfs_do_snapshot(int argc, char **argv); +static int zfs_do_unmount(int argc, char **argv); +static int zfs_do_share(int argc, char **argv); +static int zfs_do_unshare(int argc, char **argv); +static int zfs_do_send(int argc, char **argv); +static int zfs_do_receive(int argc, char **argv); +static int zfs_do_promote(int argc, char **argv); +static int zfs_do_userspace(int argc, char **argv); +static int zfs_do_python(int argc, char **argv); +static int zfs_do_hold(int argc, char **argv); +static int zfs_do_release(int argc, char **argv); +static int zfs_do_diff(int argc, char **argv); + +/* + * Enable a reasonable set of defaults for libumem debugging on DEBUG builds. + */ + +#ifdef DEBUG +const char * +_umem_debug_init(void) +{ + return ("default,verbose"); /* $UMEM_DEBUG setting */ +} + +const char * +_umem_logging_init(void) +{ + return ("fail,contents"); /* $UMEM_LOGGING setting */ +} +#endif + +typedef enum { + HELP_CLONE, + HELP_CREATE, + HELP_DESTROY, + HELP_GET, + HELP_INHERIT, + HELP_UPGRADE, + HELP_LIST, + HELP_MOUNT, + HELP_PROMOTE, + HELP_RECEIVE, + HELP_RENAME, + HELP_ROLLBACK, + HELP_SEND, + HELP_SET, + HELP_SHARE, + HELP_SNAPSHOT, + HELP_UNMOUNT, + HELP_UNSHARE, + HELP_ALLOW, + HELP_UNALLOW, + HELP_USERSPACE, + HELP_GROUPSPACE, + HELP_HOLD, + HELP_HOLDS, + HELP_RELEASE, + HELP_DIFF +} zfs_help_t; + +typedef struct zfs_command { + const char *name; + int (*func)(int argc, char **argv); + zfs_help_t usage; +} zfs_command_t; + +/* + * Master command table. Each ZFS command has a name, associated function, and + * usage message. The usage messages need to be internationalized, so we have + * to have a function to return the usage message based on a command index. + * + * These commands are organized according to how they are displayed in the usage + * message. An empty command (one with a NULL name) indicates an empty line in + * the generic usage message. + */ +static zfs_command_t command_table[] = { + { "create", zfs_do_create, HELP_CREATE }, + { "destroy", zfs_do_destroy, HELP_DESTROY }, + { NULL }, + { "snapshot", zfs_do_snapshot, HELP_SNAPSHOT }, + { "rollback", zfs_do_rollback, HELP_ROLLBACK }, + { "clone", zfs_do_clone, HELP_CLONE }, + { "promote", zfs_do_promote, HELP_PROMOTE }, + { "rename", zfs_do_rename, HELP_RENAME }, + { NULL }, + { "list", zfs_do_list, HELP_LIST }, + { NULL }, + { "set", zfs_do_set, HELP_SET }, + { "get", zfs_do_get, HELP_GET }, + { "inherit", zfs_do_inherit, HELP_INHERIT }, + { "upgrade", zfs_do_upgrade, HELP_UPGRADE }, + { "userspace", zfs_do_userspace, HELP_USERSPACE }, + { "groupspace", zfs_do_userspace, HELP_GROUPSPACE }, + { NULL }, + { "mount", zfs_do_mount, HELP_MOUNT }, + { "unmount", zfs_do_unmount, HELP_UNMOUNT }, + { "share", zfs_do_share, HELP_SHARE }, + { "unshare", zfs_do_unshare, HELP_UNSHARE }, + { NULL }, + { "send", zfs_do_send, HELP_SEND }, + { "receive", zfs_do_receive, HELP_RECEIVE }, + { NULL }, + { "allow", zfs_do_python, HELP_ALLOW }, + { NULL }, + { "unallow", zfs_do_python, HELP_UNALLOW }, + { NULL }, + { "hold", zfs_do_hold, HELP_HOLD }, + { "holds", zfs_do_python, HELP_HOLDS }, + { "release", zfs_do_release, HELP_RELEASE }, + { "diff", zfs_do_diff, HELP_DIFF }, +}; + +#define NCOMMAND (sizeof (command_table) / sizeof (command_table[0])) + +zfs_command_t *current_command; + +static const char * +get_usage(zfs_help_t idx) +{ + switch (idx) { + case HELP_CLONE: + return (gettext("\tclone [-p] [-o property=value] ... " + "<snapshot> <filesystem|volume>\n")); + case HELP_CREATE: + return (gettext("\tcreate [-p] [-o property=value] ... " + "<filesystem>\n" + "\tcreate [-ps] [-b blocksize] [-o property=value] ... " + "-V <size> <volume>\n")); + case HELP_DESTROY: + return (gettext("\tdestroy [-rRf] <filesystem|volume>\n" + "\tdestroy [-rRd] <snapshot>\n")); + case HELP_GET: + return (gettext("\tget [-rHp] [-d max] " + "[-o \"all\" | field[,...]] [-s source[,...]]\n" + "\t <\"all\" | property[,...]> " + "[filesystem|volume|snapshot] ...\n")); + case HELP_INHERIT: + return (gettext("\tinherit [-rS] <property> " + "<filesystem|volume|snapshot> ...\n")); + case HELP_UPGRADE: + return (gettext("\tupgrade [-v]\n" + "\tupgrade [-r] [-V version] <-a | filesystem ...>\n")); + case HELP_LIST: + return (gettext("\tlist [-rH][-d max] " + "[-o property[,...]] [-t type[,...]] [-s property] ...\n" + "\t [-S property] ... " + "[filesystem|volume|snapshot] ...\n")); + case HELP_MOUNT: + return (gettext("\tmount\n" + "\tmount [-vO] [-o opts] <-a | filesystem>\n")); + case HELP_PROMOTE: + return (gettext("\tpromote <clone-filesystem>\n")); + case HELP_RECEIVE: + return (gettext("\treceive [-vnFu] <filesystem|volume|" + "snapshot>\n" + "\treceive [-vnFu] [-d | -e] <filesystem>\n")); + case HELP_RENAME: + return (gettext("\trename <filesystem|volume|snapshot> " + "<filesystem|volume|snapshot>\n" + "\trename -p <filesystem|volume> <filesystem|volume>\n" + "\trename -r <snapshot> <snapshot>")); + case HELP_ROLLBACK: + return (gettext("\trollback [-rRf] <snapshot>\n")); + case HELP_SEND: + return (gettext("\tsend [-RDp] [-[iI] snapshot] <snapshot>\n")); + case HELP_SET: + return (gettext("\tset <property=value> " + "<filesystem|volume|snapshot> ...\n")); + case HELP_SHARE: + return (gettext("\tshare <-a | filesystem>\n")); + case HELP_SNAPSHOT: + return (gettext("\tsnapshot [-r] [-o property=value] ... " + "<filesystem@snapname|volume@snapname>\n")); + case HELP_UNMOUNT: + return (gettext("\tunmount [-f] " + "<-a | filesystem|mountpoint>\n")); + case HELP_UNSHARE: + return (gettext("\tunshare " + "<-a | filesystem|mountpoint>\n")); + case HELP_ALLOW: + return (gettext("\tallow <filesystem|volume>\n" + "\tallow [-ldug] " + "<\"everyone\"|user|group>[,...] <perm|@setname>[,...]\n" + "\t <filesystem|volume>\n" + "\tallow [-ld] -e <perm|@setname>[,...] " + "<filesystem|volume>\n" + "\tallow -c <perm|@setname>[,...] <filesystem|volume>\n" + "\tallow -s @setname <perm|@setname>[,...] " + "<filesystem|volume>\n")); + case HELP_UNALLOW: + return (gettext("\tunallow [-rldug] " + "<\"everyone\"|user|group>[,...]\n" + "\t [<perm|@setname>[,...]] <filesystem|volume>\n" + "\tunallow [-rld] -e [<perm|@setname>[,...]] " + "<filesystem|volume>\n" + "\tunallow [-r] -c [<perm|@setname>[,...]] " + "<filesystem|volume>\n" + "\tunallow [-r] -s @setname [<perm|@setname>[,...]] " + "<filesystem|volume>\n")); + case HELP_USERSPACE: + return (gettext("\tuserspace [-hniHp] [-o field[,...]] " + "[-sS field] ... [-t type[,...]]\n" + "\t <filesystem|snapshot>\n")); + case HELP_GROUPSPACE: + return (gettext("\tgroupspace [-hniHpU] [-o field[,...]] " + "[-sS field] ... [-t type[,...]]\n" + "\t <filesystem|snapshot>\n")); + case HELP_HOLD: + return (gettext("\thold [-r] <tag> <snapshot> ...\n")); + case HELP_HOLDS: + return (gettext("\tholds [-r] <snapshot> ...\n")); + case HELP_RELEASE: + return (gettext("\trelease [-r] <tag> <snapshot> ...\n")); + case HELP_DIFF: + return (gettext("\tdiff [-FHt] <snapshot> " + "[snapshot|filesystem]\n")); + } + + abort(); + /* NOTREACHED */ +} + +void +nomem(void) +{ + (void) fprintf(stderr, gettext("internal error: out of memory\n")); + exit(1); +} + +/* + * Utility function to guarantee malloc() success. + */ + +void * +safe_malloc(size_t size) +{ + void *data; + + if ((data = calloc(1, size)) == NULL) + nomem(); + + return (data); +} + +static char * +safe_strdup(char *str) +{ + char *dupstr = strdup(str); + + if (dupstr == NULL) + nomem(); + + return (dupstr); +} + +/* + * Callback routine that will print out information for each of + * the properties. + */ +static int +usage_prop_cb(int prop, void *cb) +{ + FILE *fp = cb; + + (void) fprintf(fp, "\t%-15s ", zfs_prop_to_name(prop)); + + if (zfs_prop_readonly(prop)) + (void) fprintf(fp, " NO "); + else + (void) fprintf(fp, "YES "); + + if (zfs_prop_inheritable(prop)) + (void) fprintf(fp, " YES "); + else + (void) fprintf(fp, " NO "); + + if (zfs_prop_values(prop) == NULL) + (void) fprintf(fp, "-\n"); + else + (void) fprintf(fp, "%s\n", zfs_prop_values(prop)); + + return (ZPROP_CONT); +} + +/* + * Display usage message. If we're inside a command, display only the usage for + * that command. Otherwise, iterate over the entire command table and display + * a complete usage message. + */ +static void +usage(boolean_t requested) +{ + int i; + boolean_t show_properties = B_FALSE; + FILE *fp = requested ? stdout : stderr; + + if (current_command == NULL) { + + (void) fprintf(fp, gettext("usage: zfs command args ...\n")); + (void) fprintf(fp, + gettext("where 'command' is one of the following:\n\n")); + + for (i = 0; i < NCOMMAND; i++) { + if (command_table[i].name == NULL) + (void) fprintf(fp, "\n"); + else + (void) fprintf(fp, "%s", + get_usage(command_table[i].usage)); + } + + (void) fprintf(fp, gettext("\nEach dataset is of the form: " + "pool/[dataset/]*dataset[@name]\n")); + } else { + (void) fprintf(fp, gettext("usage:\n")); + (void) fprintf(fp, "%s", get_usage(current_command->usage)); + } + + if (current_command != NULL && + (strcmp(current_command->name, "set") == 0 || + strcmp(current_command->name, "get") == 0 || + strcmp(current_command->name, "inherit") == 0 || + strcmp(current_command->name, "list") == 0)) + show_properties = B_TRUE; + + if (show_properties) { + (void) fprintf(fp, + gettext("\nThe following properties are supported:\n")); + + (void) fprintf(fp, "\n\t%-14s %s %s %s\n\n", + "PROPERTY", "EDIT", "INHERIT", "VALUES"); + + /* Iterate over all properties */ + (void) zprop_iter(usage_prop_cb, fp, B_FALSE, B_TRUE, + ZFS_TYPE_DATASET); + + (void) fprintf(fp, "\t%-15s ", "userused@..."); + (void) fprintf(fp, " NO NO <size>\n"); + (void) fprintf(fp, "\t%-15s ", "groupused@..."); + (void) fprintf(fp, " NO NO <size>\n"); + (void) fprintf(fp, "\t%-15s ", "userquota@..."); + (void) fprintf(fp, "YES NO <size> | none\n"); + (void) fprintf(fp, "\t%-15s ", "groupquota@..."); + (void) fprintf(fp, "YES NO <size> | none\n"); + + (void) fprintf(fp, gettext("\nSizes are specified in bytes " + "with standard units such as K, M, G, etc.\n")); + (void) fprintf(fp, gettext("\nUser-defined properties can " + "be specified by using a name containing a colon (:).\n")); + (void) fprintf(fp, gettext("\nThe {user|group}{used|quota}@ " + "properties must be appended with\n" + "a user or group specifier of one of these forms:\n" + " POSIX name (eg: \"matt\")\n" + " POSIX id (eg: \"126829\")\n" + " SMB name@domain (eg: \"matt@sun\")\n" + " SMB SID (eg: \"S-1-234-567-89\")\n")); + } else { + (void) fprintf(fp, + gettext("\nFor the property list, run: %s\n"), + "zfs set|get"); + (void) fprintf(fp, + gettext("\nFor the delegated permission list, run: %s\n"), + "zfs allow|unallow"); + } + + /* + * See comments at end of main(). + */ + if (getenv("ZFS_ABORT") != NULL) { + (void) printf("dumping core by request\n"); + abort(); + } + + exit(requested ? 0 : 2); +} + +static int +parseprop(nvlist_t *props) +{ + char *propname = optarg; + char *propval, *strval; + + if ((propval = strchr(propname, '=')) == NULL) { + (void) fprintf(stderr, gettext("missing " + "'=' for -o option\n")); + return (-1); + } + *propval = '\0'; + propval++; + if (nvlist_lookup_string(props, propname, &strval) == 0) { + (void) fprintf(stderr, gettext("property '%s' " + "specified multiple times\n"), propname); + return (-1); + } + if (nvlist_add_string(props, propname, propval) != 0) + nomem(); + return (0); +} + +static int +parse_depth(char *opt, int *flags) +{ + char *tmp; + int depth; + + depth = (int)strtol(opt, &tmp, 0); + if (*tmp) { + (void) fprintf(stderr, + gettext("%s is not an integer\n"), optarg); + usage(B_FALSE); + } + if (depth < 0) { + (void) fprintf(stderr, + gettext("Depth can not be negative.\n")); + usage(B_FALSE); + } + *flags |= (ZFS_ITER_DEPTH_LIMIT|ZFS_ITER_RECURSE); + return (depth); +} + +#define PROGRESS_DELAY 2 /* seconds */ + +static char *pt_reverse = "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b"; +static time_t pt_begin; +static char *pt_header = NULL; +static boolean_t pt_shown; + +static void +start_progress_timer(void) +{ + pt_begin = time(NULL) + PROGRESS_DELAY; + pt_shown = B_FALSE; +} + +static void +set_progress_header(char *header) +{ + assert(pt_header == NULL); + pt_header = safe_strdup(header); + if (pt_shown) { + (void) printf("%s: ", header); + (void) fflush(stdout); + } +} + +static void +update_progress(char *update) +{ + if (!pt_shown && time(NULL) > pt_begin) { + int len = strlen(update); + + (void) printf("%s: %s%*.*s", pt_header, update, len, len, + pt_reverse); + (void) fflush(stdout); + pt_shown = B_TRUE; + } else if (pt_shown) { + int len = strlen(update); + + (void) printf("%s%*.*s", update, len, len, pt_reverse); + (void) fflush(stdout); + } +} + +static void +finish_progress(char *done) +{ + if (pt_shown) { + (void) printf("%s\n", done); + (void) fflush(stdout); + } + free(pt_header); + pt_header = NULL; +} +/* + * zfs clone [-p] [-o prop=value] ... <snap> <fs | vol> + * + * Given an existing dataset, create a writable copy whose initial contents + * are the same as the source. The newly created dataset maintains a + * dependency on the original; the original cannot be destroyed so long as + * the clone exists. + * + * The '-p' flag creates all the non-existing ancestors of the target first. + */ +static int +zfs_do_clone(int argc, char **argv) +{ + zfs_handle_t *zhp = NULL; + boolean_t parents = B_FALSE; + nvlist_t *props; + int ret; + int c; + + if (nvlist_alloc(&props, NV_UNIQUE_NAME, 0) != 0) + nomem(); + + /* check options */ + while ((c = getopt(argc, argv, "o:p")) != -1) { + switch (c) { + case 'o': + if (parseprop(props)) + return (1); + break; + case 'p': + parents = B_TRUE; + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + goto usage; + } + } + + argc -= optind; + argv += optind; + + /* check number of arguments */ + if (argc < 1) { + (void) fprintf(stderr, gettext("missing source dataset " + "argument\n")); + goto usage; + } + if (argc < 2) { + (void) fprintf(stderr, gettext("missing target dataset " + "argument\n")); + goto usage; + } + if (argc > 2) { + (void) fprintf(stderr, gettext("too many arguments\n")); + goto usage; + } + + /* open the source dataset */ + if ((zhp = zfs_open(g_zfs, argv[0], ZFS_TYPE_SNAPSHOT)) == NULL) + return (1); + + if (parents && zfs_name_valid(argv[1], ZFS_TYPE_FILESYSTEM | + ZFS_TYPE_VOLUME)) { + /* + * Now create the ancestors of the target dataset. If the + * target already exists and '-p' option was used we should not + * complain. + */ + if (zfs_dataset_exists(g_zfs, argv[1], ZFS_TYPE_FILESYSTEM | + ZFS_TYPE_VOLUME)) + return (0); + if (zfs_create_ancestors(g_zfs, argv[1]) != 0) + return (1); + } + + /* pass to libzfs */ + ret = zfs_clone(zhp, argv[1], props); + + /* create the mountpoint if necessary */ + if (ret == 0) { + zfs_handle_t *clone; + + clone = zfs_open(g_zfs, argv[1], ZFS_TYPE_DATASET); + if (clone != NULL) { + if (zfs_get_type(clone) != ZFS_TYPE_VOLUME) + if ((ret = zfs_mount(clone, NULL, 0)) == 0) + ret = zfs_share(clone); + zfs_close(clone); + } + } + + zfs_close(zhp); + nvlist_free(props); + + return (!!ret); + +usage: + if (zhp) + zfs_close(zhp); + nvlist_free(props); + usage(B_FALSE); + return (-1); +} + +/* + * zfs create [-p] [-o prop=value] ... fs + * zfs create [-ps] [-b blocksize] [-o prop=value] ... -V vol size + * + * Create a new dataset. This command can be used to create filesystems + * and volumes. Snapshot creation is handled by 'zfs snapshot'. + * For volumes, the user must specify a size to be used. + * + * The '-s' flag applies only to volumes, and indicates that we should not try + * to set the reservation for this volume. By default we set a reservation + * equal to the size for any volume. For pools with SPA_VERSION >= + * SPA_VERSION_REFRESERVATION, we set a refreservation instead. + * + * The '-p' flag creates all the non-existing ancestors of the target first. + */ +static int +zfs_do_create(int argc, char **argv) +{ + zfs_type_t type = ZFS_TYPE_FILESYSTEM; + zfs_handle_t *zhp = NULL; + uint64_t volsize; + int c; + boolean_t noreserve = B_FALSE; + boolean_t bflag = B_FALSE; + boolean_t parents = B_FALSE; + int ret = 1; + nvlist_t *props; + uint64_t intval; + int canmount = ZFS_CANMOUNT_OFF; + + if (nvlist_alloc(&props, NV_UNIQUE_NAME, 0) != 0) + nomem(); + + /* check options */ + while ((c = getopt(argc, argv, ":V:b:so:p")) != -1) { + switch (c) { + case 'V': + type = ZFS_TYPE_VOLUME; + if (zfs_nicestrtonum(g_zfs, optarg, &intval) != 0) { + (void) fprintf(stderr, gettext("bad volume " + "size '%s': %s\n"), optarg, + libzfs_error_description(g_zfs)); + goto error; + } + + if (nvlist_add_uint64(props, + zfs_prop_to_name(ZFS_PROP_VOLSIZE), intval) != 0) + nomem(); + volsize = intval; + break; + case 'p': + parents = B_TRUE; + break; + case 'b': + bflag = B_TRUE; + if (zfs_nicestrtonum(g_zfs, optarg, &intval) != 0) { + (void) fprintf(stderr, gettext("bad volume " + "block size '%s': %s\n"), optarg, + libzfs_error_description(g_zfs)); + goto error; + } + + if (nvlist_add_uint64(props, + zfs_prop_to_name(ZFS_PROP_VOLBLOCKSIZE), + intval) != 0) + nomem(); + break; + case 'o': + if (parseprop(props)) + goto error; + break; + case 's': + noreserve = B_TRUE; + break; + case ':': + (void) fprintf(stderr, gettext("missing size " + "argument\n")); + goto badusage; + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + goto badusage; + } + } + + if ((bflag || noreserve) && type != ZFS_TYPE_VOLUME) { + (void) fprintf(stderr, gettext("'-s' and '-b' can only be " + "used when creating a volume\n")); + goto badusage; + } + + argc -= optind; + argv += optind; + + /* check number of arguments */ + if (argc == 0) { + (void) fprintf(stderr, gettext("missing %s argument\n"), + zfs_type_to_name(type)); + goto badusage; + } + if (argc > 1) { + (void) fprintf(stderr, gettext("too many arguments\n")); + goto badusage; + } + + if (type == ZFS_TYPE_VOLUME && !noreserve) { + zpool_handle_t *zpool_handle; + uint64_t spa_version; + char *p; + zfs_prop_t resv_prop; + char *strval; + + if (p = strchr(argv[0], '/')) + *p = '\0'; + zpool_handle = zpool_open(g_zfs, argv[0]); + if (p != NULL) + *p = '/'; + if (zpool_handle == NULL) + goto error; + spa_version = zpool_get_prop_int(zpool_handle, + ZPOOL_PROP_VERSION, NULL); + zpool_close(zpool_handle); + if (spa_version >= SPA_VERSION_REFRESERVATION) + resv_prop = ZFS_PROP_REFRESERVATION; + else + resv_prop = ZFS_PROP_RESERVATION; + volsize = zvol_volsize_to_reservation(volsize, props); + + if (nvlist_lookup_string(props, zfs_prop_to_name(resv_prop), + &strval) != 0) { + if (nvlist_add_uint64(props, + zfs_prop_to_name(resv_prop), volsize) != 0) { + nvlist_free(props); + nomem(); + } + } + } + + if (parents && zfs_name_valid(argv[0], type)) { + /* + * Now create the ancestors of target dataset. If the target + * already exists and '-p' option was used we should not + * complain. + */ + if (zfs_dataset_exists(g_zfs, argv[0], type)) { + ret = 0; + goto error; + } + if (zfs_create_ancestors(g_zfs, argv[0]) != 0) + goto error; + } + + /* pass to libzfs */ + if (zfs_create(g_zfs, argv[0], type, props) != 0) + goto error; + + if ((zhp = zfs_open(g_zfs, argv[0], ZFS_TYPE_DATASET)) == NULL) + goto error; + + ret = 0; + /* + * if the user doesn't want the dataset automatically mounted, + * then skip the mount/share step + */ + if (zfs_prop_valid_for_type(ZFS_PROP_CANMOUNT, type)) + canmount = zfs_prop_get_int(zhp, ZFS_PROP_CANMOUNT); + + /* + * Mount and/or share the new filesystem as appropriate. We provide a + * verbose error message to let the user know that their filesystem was + * in fact created, even if we failed to mount or share it. + */ + if (canmount == ZFS_CANMOUNT_ON) { + if (zfs_mount(zhp, NULL, 0) != 0) { + (void) fprintf(stderr, gettext("filesystem " + "successfully created, but not mounted\n")); + ret = 1; + } else if (zfs_share(zhp) != 0) { + (void) fprintf(stderr, gettext("filesystem " + "successfully created, but not shared\n")); + ret = 1; + } + } + +error: + if (zhp) + zfs_close(zhp); + nvlist_free(props); + return (ret); +badusage: + nvlist_free(props); + usage(B_FALSE); + return (2); +} + +/* + * zfs destroy [-rRf] <fs, vol> + * zfs destroy [-rRd] <snap> + * + * -r Recursively destroy all children + * -R Recursively destroy all dependents, including clones + * -f Force unmounting of any dependents + * -d If we can't destroy now, mark for deferred destruction + * + * Destroys the given dataset. By default, it will unmount any filesystems, + * and refuse to destroy a dataset that has any dependents. A dependent can + * either be a child, or a clone of a child. + */ +typedef struct destroy_cbdata { + boolean_t cb_first; + int cb_force; + int cb_recurse; + int cb_error; + int cb_needforce; + int cb_doclones; + boolean_t cb_closezhp; + zfs_handle_t *cb_target; + char *cb_snapname; + boolean_t cb_defer_destroy; +} destroy_cbdata_t; + +/* + * Check for any dependents based on the '-r' or '-R' flags. + */ +static int +destroy_check_dependent(zfs_handle_t *zhp, void *data) +{ + destroy_cbdata_t *cbp = data; + const char *tname = zfs_get_name(cbp->cb_target); + const char *name = zfs_get_name(zhp); + + if (strncmp(tname, name, strlen(tname)) == 0 && + (name[strlen(tname)] == '/' || name[strlen(tname)] == '@')) { + /* + * This is a direct descendant, not a clone somewhere else in + * the hierarchy. + */ + if (cbp->cb_recurse) + goto out; + + if (cbp->cb_first) { + (void) fprintf(stderr, gettext("cannot destroy '%s': " + "%s has children\n"), + zfs_get_name(cbp->cb_target), + zfs_type_to_name(zfs_get_type(cbp->cb_target))); + (void) fprintf(stderr, gettext("use '-r' to destroy " + "the following datasets:\n")); + cbp->cb_first = B_FALSE; + cbp->cb_error = 1; + } + + (void) fprintf(stderr, "%s\n", zfs_get_name(zhp)); + } else { + /* + * This is a clone. We only want to report this if the '-r' + * wasn't specified, or the target is a snapshot. + */ + if (!cbp->cb_recurse && + zfs_get_type(cbp->cb_target) != ZFS_TYPE_SNAPSHOT) + goto out; + + if (cbp->cb_first) { + (void) fprintf(stderr, gettext("cannot destroy '%s': " + "%s has dependent clones\n"), + zfs_get_name(cbp->cb_target), + zfs_type_to_name(zfs_get_type(cbp->cb_target))); + (void) fprintf(stderr, gettext("use '-R' to destroy " + "the following datasets:\n")); + cbp->cb_first = B_FALSE; + cbp->cb_error = 1; + } + + (void) fprintf(stderr, "%s\n", zfs_get_name(zhp)); + } + +out: + zfs_close(zhp); + return (0); +} + +static int +destroy_callback(zfs_handle_t *zhp, void *data) +{ + destroy_cbdata_t *cbp = data; + + /* + * Ignore pools (which we've already flagged as an error before getting + * here). + */ + if (strchr(zfs_get_name(zhp), '/') == NULL && + zfs_get_type(zhp) == ZFS_TYPE_FILESYSTEM) { + zfs_close(zhp); + return (0); + } + + /* + * Bail out on the first error. + */ + if (zfs_unmount(zhp, NULL, cbp->cb_force ? MS_FORCE : 0) != 0 || + zfs_destroy(zhp, cbp->cb_defer_destroy) != 0) { + zfs_close(zhp); + return (-1); + } + + zfs_close(zhp); + return (0); +} + +static int +destroy_snap_clones(zfs_handle_t *zhp, void *arg) +{ + destroy_cbdata_t *cbp = arg; + char thissnap[MAXPATHLEN]; + zfs_handle_t *szhp; + boolean_t closezhp = cbp->cb_closezhp; + int rv; + + (void) snprintf(thissnap, sizeof (thissnap), + "%s@%s", zfs_get_name(zhp), cbp->cb_snapname); + + libzfs_print_on_error(g_zfs, B_FALSE); + szhp = zfs_open(g_zfs, thissnap, ZFS_TYPE_SNAPSHOT); + libzfs_print_on_error(g_zfs, B_TRUE); + if (szhp) { + /* + * Destroy any clones of this snapshot + */ + if (zfs_iter_dependents(szhp, B_FALSE, destroy_callback, + cbp) != 0) { + zfs_close(szhp); + if (closezhp) + zfs_close(zhp); + return (-1); + } + zfs_close(szhp); + } + + cbp->cb_closezhp = B_TRUE; + rv = zfs_iter_filesystems(zhp, destroy_snap_clones, arg); + if (closezhp) + zfs_close(zhp); + return (rv); +} + +static int +zfs_do_destroy(int argc, char **argv) +{ + destroy_cbdata_t cb = { 0 }; + int c; + zfs_handle_t *zhp; + char *cp; + zfs_type_t type = ZFS_TYPE_DATASET; + + /* check options */ + while ((c = getopt(argc, argv, "dfrR")) != -1) { + switch (c) { + case 'd': + cb.cb_defer_destroy = B_TRUE; + type = ZFS_TYPE_SNAPSHOT; + break; + case 'f': + cb.cb_force = 1; + break; + case 'r': + cb.cb_recurse = 1; + break; + case 'R': + cb.cb_recurse = 1; + cb.cb_doclones = 1; + break; + case '?': + default: + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + /* check number of arguments */ + if (argc == 0) { + (void) fprintf(stderr, gettext("missing path argument\n")); + usage(B_FALSE); + } + if (argc > 1) { + (void) fprintf(stderr, gettext("too many arguments\n")); + usage(B_FALSE); + } + + /* + * If we are doing recursive destroy of a snapshot, then the + * named snapshot may not exist. Go straight to libzfs. + */ + if (cb.cb_recurse && (cp = strchr(argv[0], '@'))) { + int ret; + + *cp = '\0'; + if ((zhp = zfs_open(g_zfs, argv[0], ZFS_TYPE_DATASET)) == NULL) + return (1); + *cp = '@'; + cp++; + + if (cb.cb_doclones) { + boolean_t defer = cb.cb_defer_destroy; + + /* + * Temporarily ignore the defer_destroy setting since + * it's not supported for clones. + */ + cb.cb_defer_destroy = B_FALSE; + cb.cb_snapname = cp; + if (destroy_snap_clones(zhp, &cb) != 0) { + zfs_close(zhp); + return (1); + } + cb.cb_defer_destroy = defer; + } + + ret = zfs_destroy_snaps(zhp, cp, cb.cb_defer_destroy); + zfs_close(zhp); + if (ret) { + (void) fprintf(stderr, + gettext("no snapshots destroyed\n")); + } + return (ret != 0); + } + + /* Open the given dataset */ + if ((zhp = zfs_open(g_zfs, argv[0], type)) == NULL) + return (1); + + cb.cb_target = zhp; + + /* + * Perform an explicit check for pools before going any further. + */ + if (!cb.cb_recurse && strchr(zfs_get_name(zhp), '/') == NULL && + zfs_get_type(zhp) == ZFS_TYPE_FILESYSTEM) { + (void) fprintf(stderr, gettext("cannot destroy '%s': " + "operation does not apply to pools\n"), + zfs_get_name(zhp)); + (void) fprintf(stderr, gettext("use 'zfs destroy -r " + "%s' to destroy all datasets in the pool\n"), + zfs_get_name(zhp)); + (void) fprintf(stderr, gettext("use 'zpool destroy %s' " + "to destroy the pool itself\n"), zfs_get_name(zhp)); + zfs_close(zhp); + return (1); + } + + /* + * Check for any dependents and/or clones. + */ + cb.cb_first = B_TRUE; + if (!cb.cb_doclones && !cb.cb_defer_destroy && + zfs_iter_dependents(zhp, B_TRUE, destroy_check_dependent, + &cb) != 0) { + zfs_close(zhp); + return (1); + } + + if (cb.cb_error || (!cb.cb_defer_destroy && + (zfs_iter_dependents(zhp, B_FALSE, destroy_callback, &cb) != 0))) { + zfs_close(zhp); + return (1); + } + + /* + * Do the real thing. The callback will close the handle regardless of + * whether it succeeds or not. + */ + + if (destroy_callback(zhp, &cb) != 0) + return (1); + + return (0); +} + +static boolean_t +is_recvd_column(zprop_get_cbdata_t *cbp) +{ + int i; + zfs_get_column_t col; + + for (i = 0; i < ZFS_GET_NCOLS && + (col = cbp->cb_columns[i]) != GET_COL_NONE; i++) + if (col == GET_COL_RECVD) + return (B_TRUE); + return (B_FALSE); +} + +/* + * zfs get [-rHp] [-o all | field[,field]...] [-s source[,source]...] + * < all | property[,property]... > < fs | snap | vol > ... + * + * -r recurse over any child datasets + * -H scripted mode. Headers are stripped, and fields are separated + * by tabs instead of spaces. + * -o Set of fields to display. One of "name,property,value, + * received,source". Default is "name,property,value,source". + * "all" is an alias for all five. + * -s Set of sources to allow. One of + * "local,default,inherited,received,temporary,none". Default is + * all six. + * -p Display values in parsable (literal) format. + * + * Prints properties for the given datasets. The user can control which + * columns to display as well as which property types to allow. + */ + +/* + * Invoked to display the properties for a single dataset. + */ +static int +get_callback(zfs_handle_t *zhp, void *data) +{ + char buf[ZFS_MAXPROPLEN]; + char rbuf[ZFS_MAXPROPLEN]; + zprop_source_t sourcetype; + char source[ZFS_MAXNAMELEN]; + zprop_get_cbdata_t *cbp = data; + nvlist_t *user_props = zfs_get_user_props(zhp); + zprop_list_t *pl = cbp->cb_proplist; + nvlist_t *propval; + char *strval; + char *sourceval; + boolean_t received = is_recvd_column(cbp); + + for (; pl != NULL; pl = pl->pl_next) { + char *recvdval = NULL; + /* + * Skip the special fake placeholder. This will also skip over + * the name property when 'all' is specified. + */ + if (pl->pl_prop == ZFS_PROP_NAME && + pl == cbp->cb_proplist) + continue; + + if (pl->pl_prop != ZPROP_INVAL) { + if (zfs_prop_get(zhp, pl->pl_prop, buf, + sizeof (buf), &sourcetype, source, + sizeof (source), + cbp->cb_literal) != 0) { + if (pl->pl_all) + continue; + if (!zfs_prop_valid_for_type(pl->pl_prop, + ZFS_TYPE_DATASET)) { + (void) fprintf(stderr, + gettext("No such property '%s'\n"), + zfs_prop_to_name(pl->pl_prop)); + continue; + } + sourcetype = ZPROP_SRC_NONE; + (void) strlcpy(buf, "-", sizeof (buf)); + } + + if (received && (zfs_prop_get_recvd(zhp, + zfs_prop_to_name(pl->pl_prop), rbuf, sizeof (rbuf), + cbp->cb_literal) == 0)) + recvdval = rbuf; + + zprop_print_one_property(zfs_get_name(zhp), cbp, + zfs_prop_to_name(pl->pl_prop), + buf, sourcetype, source, recvdval); + } else if (zfs_prop_userquota(pl->pl_user_prop)) { + sourcetype = ZPROP_SRC_LOCAL; + + if (zfs_prop_get_userquota(zhp, pl->pl_user_prop, + buf, sizeof (buf), cbp->cb_literal) != 0) { + sourcetype = ZPROP_SRC_NONE; + (void) strlcpy(buf, "-", sizeof (buf)); + } + + zprop_print_one_property(zfs_get_name(zhp), cbp, + pl->pl_user_prop, buf, sourcetype, source, NULL); + } else { + if (nvlist_lookup_nvlist(user_props, + pl->pl_user_prop, &propval) != 0) { + if (pl->pl_all) + continue; + sourcetype = ZPROP_SRC_NONE; + strval = "-"; + } else { + verify(nvlist_lookup_string(propval, + ZPROP_VALUE, &strval) == 0); + verify(nvlist_lookup_string(propval, + ZPROP_SOURCE, &sourceval) == 0); + + if (strcmp(sourceval, + zfs_get_name(zhp)) == 0) { + sourcetype = ZPROP_SRC_LOCAL; + } else if (strcmp(sourceval, + ZPROP_SOURCE_VAL_RECVD) == 0) { + sourcetype = ZPROP_SRC_RECEIVED; + } else { + sourcetype = ZPROP_SRC_INHERITED; + (void) strlcpy(source, + sourceval, sizeof (source)); + } + } + + if (received && (zfs_prop_get_recvd(zhp, + pl->pl_user_prop, rbuf, sizeof (rbuf), + cbp->cb_literal) == 0)) + recvdval = rbuf; + + zprop_print_one_property(zfs_get_name(zhp), cbp, + pl->pl_user_prop, strval, sourcetype, + source, recvdval); + } + } + + return (0); +} + +static int +zfs_do_get(int argc, char **argv) +{ + zprop_get_cbdata_t cb = { 0 }; + int i, c, flags = 0; + char *value, *fields; + int ret; + int limit = 0; + zprop_list_t fake_name = { 0 }; + + /* + * Set up default columns and sources. + */ + cb.cb_sources = ZPROP_SRC_ALL; + cb.cb_columns[0] = GET_COL_NAME; + cb.cb_columns[1] = GET_COL_PROPERTY; + cb.cb_columns[2] = GET_COL_VALUE; + cb.cb_columns[3] = GET_COL_SOURCE; + cb.cb_type = ZFS_TYPE_DATASET; + + /* check options */ + while ((c = getopt(argc, argv, ":d:o:s:rHp")) != -1) { + switch (c) { + case 'p': + cb.cb_literal = B_TRUE; + break; + case 'd': + limit = parse_depth(optarg, &flags); + break; + case 'r': + flags |= ZFS_ITER_RECURSE; + break; + case 'H': + cb.cb_scripted = B_TRUE; + break; + case ':': + (void) fprintf(stderr, gettext("missing argument for " + "'%c' option\n"), optopt); + usage(B_FALSE); + break; + case 'o': + /* + * Process the set of columns to display. We zero out + * the structure to give us a blank slate. + */ + bzero(&cb.cb_columns, sizeof (cb.cb_columns)); + i = 0; + while (*optarg != '\0') { + static char *col_subopts[] = + { "name", "property", "value", "received", + "source", "all", NULL }; + + if (i == ZFS_GET_NCOLS) { + (void) fprintf(stderr, gettext("too " + "many fields given to -o " + "option\n")); + usage(B_FALSE); + } + + switch (getsubopt(&optarg, col_subopts, + &value)) { + case 0: + cb.cb_columns[i++] = GET_COL_NAME; + break; + case 1: + cb.cb_columns[i++] = GET_COL_PROPERTY; + break; + case 2: + cb.cb_columns[i++] = GET_COL_VALUE; + break; + case 3: + cb.cb_columns[i++] = GET_COL_RECVD; + flags |= ZFS_ITER_RECVD_PROPS; + break; + case 4: + cb.cb_columns[i++] = GET_COL_SOURCE; + break; + case 5: + if (i > 0) { + (void) fprintf(stderr, + gettext("\"all\" conflicts " + "with specific fields " + "given to -o option\n")); + usage(B_FALSE); + } + cb.cb_columns[0] = GET_COL_NAME; + cb.cb_columns[1] = GET_COL_PROPERTY; + cb.cb_columns[2] = GET_COL_VALUE; + cb.cb_columns[3] = GET_COL_RECVD; + cb.cb_columns[4] = GET_COL_SOURCE; + flags |= ZFS_ITER_RECVD_PROPS; + i = ZFS_GET_NCOLS; + break; + default: + (void) fprintf(stderr, + gettext("invalid column name " + "'%s'\n"), value); + usage(B_FALSE); + } + } + break; + + case 's': + cb.cb_sources = 0; + while (*optarg != '\0') { + static char *source_subopts[] = { + "local", "default", "inherited", + "received", "temporary", "none", + NULL }; + + switch (getsubopt(&optarg, source_subopts, + &value)) { + case 0: + cb.cb_sources |= ZPROP_SRC_LOCAL; + break; + case 1: + cb.cb_sources |= ZPROP_SRC_DEFAULT; + break; + case 2: + cb.cb_sources |= ZPROP_SRC_INHERITED; + break; + case 3: + cb.cb_sources |= ZPROP_SRC_RECEIVED; + break; + case 4: + cb.cb_sources |= ZPROP_SRC_TEMPORARY; + break; + case 5: + cb.cb_sources |= ZPROP_SRC_NONE; + break; + default: + (void) fprintf(stderr, + gettext("invalid source " + "'%s'\n"), value); + usage(B_FALSE); + } + } + break; + + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + if (argc < 1) { + (void) fprintf(stderr, gettext("missing property " + "argument\n")); + usage(B_FALSE); + } + + fields = argv[0]; + + if (zprop_get_list(g_zfs, fields, &cb.cb_proplist, ZFS_TYPE_DATASET) + != 0) + usage(B_FALSE); + + argc--; + argv++; + + /* + * As part of zfs_expand_proplist(), we keep track of the maximum column + * width for each property. For the 'NAME' (and 'SOURCE') columns, we + * need to know the maximum name length. However, the user likely did + * not specify 'name' as one of the properties to fetch, so we need to + * make sure we always include at least this property for + * print_get_headers() to work properly. + */ + if (cb.cb_proplist != NULL) { + fake_name.pl_prop = ZFS_PROP_NAME; + fake_name.pl_width = strlen(gettext("NAME")); + fake_name.pl_next = cb.cb_proplist; + cb.cb_proplist = &fake_name; + } + + cb.cb_first = B_TRUE; + + /* run for each object */ + ret = zfs_for_each(argc, argv, flags, ZFS_TYPE_DATASET, NULL, + &cb.cb_proplist, limit, get_callback, &cb); + + if (cb.cb_proplist == &fake_name) + zprop_free_list(fake_name.pl_next); + else + zprop_free_list(cb.cb_proplist); + + return (ret); +} + +/* + * inherit [-rS] <property> <fs|vol> ... + * + * -r Recurse over all children + * -S Revert to received value, if any + * + * For each dataset specified on the command line, inherit the given property + * from its parent. Inheriting a property at the pool level will cause it to + * use the default value. The '-r' flag will recurse over all children, and is + * useful for setting a property on a hierarchy-wide basis, regardless of any + * local modifications for each dataset. + */ + +typedef struct inherit_cbdata { + const char *cb_propname; + boolean_t cb_received; +} inherit_cbdata_t; + +static int +inherit_recurse_cb(zfs_handle_t *zhp, void *data) +{ + inherit_cbdata_t *cb = data; + zfs_prop_t prop = zfs_name_to_prop(cb->cb_propname); + + /* + * If we're doing it recursively, then ignore properties that + * are not valid for this type of dataset. + */ + if (prop != ZPROP_INVAL && + !zfs_prop_valid_for_type(prop, zfs_get_type(zhp))) + return (0); + + return (zfs_prop_inherit(zhp, cb->cb_propname, cb->cb_received) != 0); +} + +static int +inherit_cb(zfs_handle_t *zhp, void *data) +{ + inherit_cbdata_t *cb = data; + + return (zfs_prop_inherit(zhp, cb->cb_propname, cb->cb_received) != 0); +} + +static int +zfs_do_inherit(int argc, char **argv) +{ + int c; + zfs_prop_t prop; + inherit_cbdata_t cb = { 0 }; + char *propname; + int ret; + int flags = 0; + boolean_t received = B_FALSE; + + /* check options */ + while ((c = getopt(argc, argv, "rS")) != -1) { + switch (c) { + case 'r': + flags |= ZFS_ITER_RECURSE; + break; + case 'S': + received = B_TRUE; + break; + case '?': + default: + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + /* check number of arguments */ + if (argc < 1) { + (void) fprintf(stderr, gettext("missing property argument\n")); + usage(B_FALSE); + } + if (argc < 2) { + (void) fprintf(stderr, gettext("missing dataset argument\n")); + usage(B_FALSE); + } + + propname = argv[0]; + argc--; + argv++; + + if ((prop = zfs_name_to_prop(propname)) != ZPROP_INVAL) { + if (zfs_prop_readonly(prop)) { + (void) fprintf(stderr, gettext( + "%s property is read-only\n"), + propname); + return (1); + } + if (!zfs_prop_inheritable(prop) && !received) { + (void) fprintf(stderr, gettext("'%s' property cannot " + "be inherited\n"), propname); + if (prop == ZFS_PROP_QUOTA || + prop == ZFS_PROP_RESERVATION || + prop == ZFS_PROP_REFQUOTA || + prop == ZFS_PROP_REFRESERVATION) + (void) fprintf(stderr, gettext("use 'zfs set " + "%s=none' to clear\n"), propname); + return (1); + } + if (received && (prop == ZFS_PROP_VOLSIZE || + prop == ZFS_PROP_VERSION)) { + (void) fprintf(stderr, gettext("'%s' property cannot " + "be reverted to a received value\n"), propname); + return (1); + } + } else if (!zfs_prop_user(propname)) { + (void) fprintf(stderr, gettext("invalid property '%s'\n"), + propname); + usage(B_FALSE); + } + + cb.cb_propname = propname; + cb.cb_received = received; + + if (flags & ZFS_ITER_RECURSE) { + ret = zfs_for_each(argc, argv, flags, ZFS_TYPE_DATASET, + NULL, NULL, 0, inherit_recurse_cb, &cb); + } else { + ret = zfs_for_each(argc, argv, flags, ZFS_TYPE_DATASET, + NULL, NULL, 0, inherit_cb, &cb); + } + + return (ret); +} + +typedef struct upgrade_cbdata { + uint64_t cb_numupgraded; + uint64_t cb_numsamegraded; + uint64_t cb_numfailed; + uint64_t cb_version; + boolean_t cb_newer; + boolean_t cb_foundone; + char cb_lastfs[ZFS_MAXNAMELEN]; +} upgrade_cbdata_t; + +static int +same_pool(zfs_handle_t *zhp, const char *name) +{ + int len1 = strcspn(name, "/@"); + const char *zhname = zfs_get_name(zhp); + int len2 = strcspn(zhname, "/@"); + + if (len1 != len2) + return (B_FALSE); + return (strncmp(name, zhname, len1) == 0); +} + +static int +upgrade_list_callback(zfs_handle_t *zhp, void *data) +{ + upgrade_cbdata_t *cb = data; + int version = zfs_prop_get_int(zhp, ZFS_PROP_VERSION); + + /* list if it's old/new */ + if ((!cb->cb_newer && version < ZPL_VERSION) || + (cb->cb_newer && version > ZPL_VERSION)) { + char *str; + if (cb->cb_newer) { + str = gettext("The following filesystems are " + "formatted using a newer software version and\n" + "cannot be accessed on the current system.\n\n"); + } else { + str = gettext("The following filesystems are " + "out of date, and can be upgraded. After being\n" + "upgraded, these filesystems (and any 'zfs send' " + "streams generated from\n" + "subsequent snapshots) will no longer be " + "accessible by older software versions.\n\n"); + } + + if (!cb->cb_foundone) { + (void) puts(str); + (void) printf(gettext("VER FILESYSTEM\n")); + (void) printf(gettext("--- ------------\n")); + cb->cb_foundone = B_TRUE; + } + + (void) printf("%2u %s\n", version, zfs_get_name(zhp)); + } + + return (0); +} + +static int +upgrade_set_callback(zfs_handle_t *zhp, void *data) +{ + upgrade_cbdata_t *cb = data; + int version = zfs_prop_get_int(zhp, ZFS_PROP_VERSION); + int needed_spa_version; + int spa_version; + + if (zfs_spa_version(zhp, &spa_version) < 0) + return (-1); + + needed_spa_version = zfs_spa_version_map(cb->cb_version); + + if (needed_spa_version < 0) + return (-1); + + if (spa_version < needed_spa_version) { + /* can't upgrade */ + (void) printf(gettext("%s: can not be " + "upgraded; the pool version needs to first " + "be upgraded\nto version %d\n\n"), + zfs_get_name(zhp), needed_spa_version); + cb->cb_numfailed++; + return (0); + } + + /* upgrade */ + if (version < cb->cb_version) { + char verstr[16]; + (void) snprintf(verstr, sizeof (verstr), + "%llu", cb->cb_version); + if (cb->cb_lastfs[0] && !same_pool(zhp, cb->cb_lastfs)) { + /* + * If they did "zfs upgrade -a", then we could + * be doing ioctls to different pools. We need + * to log this history once to each pool. + */ + verify(zpool_stage_history(g_zfs, history_str) == 0); + } + if (zfs_prop_set(zhp, "version", verstr) == 0) + cb->cb_numupgraded++; + else + cb->cb_numfailed++; + (void) strcpy(cb->cb_lastfs, zfs_get_name(zhp)); + } else if (version > cb->cb_version) { + /* can't downgrade */ + (void) printf(gettext("%s: can not be downgraded; " + "it is already at version %u\n"), + zfs_get_name(zhp), version); + cb->cb_numfailed++; + } else { + cb->cb_numsamegraded++; + } + return (0); +} + +/* + * zfs upgrade + * zfs upgrade -v + * zfs upgrade [-r] [-V <version>] <-a | filesystem> + */ +static int +zfs_do_upgrade(int argc, char **argv) +{ + boolean_t all = B_FALSE; + boolean_t showversions = B_FALSE; + int ret; + upgrade_cbdata_t cb = { 0 }; + char c; + int flags = ZFS_ITER_ARGS_CAN_BE_PATHS; + + /* check options */ + while ((c = getopt(argc, argv, "rvV:a")) != -1) { + switch (c) { + case 'r': + flags |= ZFS_ITER_RECURSE; + break; + case 'v': + showversions = B_TRUE; + break; + case 'V': + if (zfs_prop_string_to_index(ZFS_PROP_VERSION, + optarg, &cb.cb_version) != 0) { + (void) fprintf(stderr, + gettext("invalid version %s\n"), optarg); + usage(B_FALSE); + } + break; + case 'a': + all = B_TRUE; + break; + case '?': + default: + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + if ((!all && !argc) && ((flags & ZFS_ITER_RECURSE) | cb.cb_version)) + usage(B_FALSE); + if (showversions && (flags & ZFS_ITER_RECURSE || all || + cb.cb_version || argc)) + usage(B_FALSE); + if ((all || argc) && (showversions)) + usage(B_FALSE); + if (all && argc) + usage(B_FALSE); + + if (showversions) { + /* Show info on available versions. */ + (void) printf(gettext("The following filesystem versions are " + "supported:\n\n")); + (void) printf(gettext("VER DESCRIPTION\n")); + (void) printf("--- -----------------------------------------" + "---------------\n"); + (void) printf(gettext(" 1 Initial ZFS filesystem version\n")); + (void) printf(gettext(" 2 Enhanced directory entries\n")); + (void) printf(gettext(" 3 Case insensitive and File system " + "unique identifier (FUID)\n")); + (void) printf(gettext(" 4 userquota, groupquota " + "properties\n")); + (void) printf(gettext(" 5 System attributes\n")); + (void) printf(gettext("\nFor more information on a particular " + "version, including supported releases,\n")); + (void) printf("see the ZFS Administration Guide.\n\n"); + ret = 0; + } else if (argc || all) { + /* Upgrade filesystems */ + if (cb.cb_version == 0) + cb.cb_version = ZPL_VERSION; + ret = zfs_for_each(argc, argv, flags, ZFS_TYPE_FILESYSTEM, + NULL, NULL, 0, upgrade_set_callback, &cb); + (void) printf(gettext("%llu filesystems upgraded\n"), + cb.cb_numupgraded); + if (cb.cb_numsamegraded) { + (void) printf(gettext("%llu filesystems already at " + "this version\n"), + cb.cb_numsamegraded); + } + if (cb.cb_numfailed != 0) + ret = 1; + } else { + /* List old-version filesytems */ + boolean_t found; + (void) printf(gettext("This system is currently running " + "ZFS filesystem version %llu.\n\n"), ZPL_VERSION); + + flags |= ZFS_ITER_RECURSE; + ret = zfs_for_each(0, NULL, flags, ZFS_TYPE_FILESYSTEM, + NULL, NULL, 0, upgrade_list_callback, &cb); + + found = cb.cb_foundone; + cb.cb_foundone = B_FALSE; + cb.cb_newer = B_TRUE; + + ret = zfs_for_each(0, NULL, flags, ZFS_TYPE_FILESYSTEM, + NULL, NULL, 0, upgrade_list_callback, &cb); + + if (!cb.cb_foundone && !found) { + (void) printf(gettext("All filesystems are " + "formatted with the current version.\n")); + } + } + + return (ret); +} + +/* + * zfs userspace + */ +static int +userspace_cb(void *arg, const char *domain, uid_t rid, uint64_t space) +{ + zfs_userquota_prop_t *typep = arg; + zfs_userquota_prop_t p = *typep; + char *name = NULL; + char *ug, *propname; + char namebuf[32]; + char sizebuf[32]; + + if (domain == NULL || domain[0] == '\0') { + if (p == ZFS_PROP_GROUPUSED || p == ZFS_PROP_GROUPQUOTA) { + struct group *g = getgrgid(rid); + if (g) + name = g->gr_name; + } else { + struct passwd *p = getpwuid(rid); + if (p) + name = p->pw_name; + } + } + + if (p == ZFS_PROP_GROUPUSED || p == ZFS_PROP_GROUPQUOTA) + ug = "group"; + else + ug = "user"; + + if (p == ZFS_PROP_USERUSED || p == ZFS_PROP_GROUPUSED) + propname = "used"; + else + propname = "quota"; + + if (name == NULL) { + (void) snprintf(namebuf, sizeof (namebuf), + "%llu", (longlong_t)rid); + name = namebuf; + } + zfs_nicenum(space, sizebuf, sizeof (sizebuf)); + + (void) printf("%s %s %s%c%s %s\n", propname, ug, domain, + domain[0] ? '-' : ' ', name, sizebuf); + + return (0); +} + +static int +zfs_do_userspace(int argc, char **argv) +{ + zfs_handle_t *zhp; + zfs_userquota_prop_t p; + int error; + + /* + * Try the python version. If the execv fails, we'll continue + * and do a simplistic implementation. + */ + (void) execv(pypath, argv-1); + + (void) printf("internal error: %s not found\n" + "falling back on built-in implementation, " + "some features will not work\n", pypath); + + if ((zhp = zfs_open(g_zfs, argv[argc-1], ZFS_TYPE_DATASET)) == NULL) + return (1); + + (void) printf("PROP TYPE NAME VALUE\n"); + + for (p = 0; p < ZFS_NUM_USERQUOTA_PROPS; p++) { + error = zfs_userspace(zhp, p, userspace_cb, &p); + if (error) + break; + } + return (error); +} + +/* + * list [-r][-d max] [-H] [-o property[,property]...] [-t type[,type]...] + * [-s property [-s property]...] [-S property [-S property]...] + * <dataset> ... + * + * -r Recurse over all children + * -d Limit recursion by depth. + * -H Scripted mode; elide headers and separate columns by tabs + * -o Control which fields to display. + * -t Control which object types to display. + * -s Specify sort columns, descending order. + * -S Specify sort columns, ascending order. + * + * When given no arguments, lists all filesystems in the system. + * Otherwise, list the specified datasets, optionally recursing down them if + * '-r' is specified. + */ +typedef struct list_cbdata { + boolean_t cb_first; + boolean_t cb_scripted; + zprop_list_t *cb_proplist; +} list_cbdata_t; + +/* + * Given a list of columns to display, output appropriate headers for each one. + */ +static void +print_header(zprop_list_t *pl) +{ + char headerbuf[ZFS_MAXPROPLEN]; + const char *header; + int i; + boolean_t first = B_TRUE; + boolean_t right_justify; + + for (; pl != NULL; pl = pl->pl_next) { + if (!first) { + (void) printf(" "); + } else { + first = B_FALSE; + } + + right_justify = B_FALSE; + if (pl->pl_prop != ZPROP_INVAL) { + header = zfs_prop_column_name(pl->pl_prop); + right_justify = zfs_prop_align_right(pl->pl_prop); + } else { + for (i = 0; pl->pl_user_prop[i] != '\0'; i++) + headerbuf[i] = toupper(pl->pl_user_prop[i]); + headerbuf[i] = '\0'; + header = headerbuf; + } + + if (pl->pl_next == NULL && !right_justify) + (void) printf("%s", header); + else if (right_justify) + (void) printf("%*s", pl->pl_width, header); + else + (void) printf("%-*s", pl->pl_width, header); + } + + (void) printf("\n"); +} + +/* + * Given a dataset and a list of fields, print out all the properties according + * to the described layout. + */ +static void +print_dataset(zfs_handle_t *zhp, zprop_list_t *pl, boolean_t scripted) +{ + boolean_t first = B_TRUE; + char property[ZFS_MAXPROPLEN]; + nvlist_t *userprops = zfs_get_user_props(zhp); + nvlist_t *propval; + char *propstr; + boolean_t right_justify; + int width; + + for (; pl != NULL; pl = pl->pl_next) { + if (!first) { + if (scripted) + (void) printf("\t"); + else + (void) printf(" "); + } else { + first = B_FALSE; + } + + if (pl->pl_prop != ZPROP_INVAL) { + if (zfs_prop_get(zhp, pl->pl_prop, property, + sizeof (property), NULL, NULL, 0, B_FALSE) != 0) + propstr = "-"; + else + propstr = property; + + right_justify = zfs_prop_align_right(pl->pl_prop); + } else if (zfs_prop_userquota(pl->pl_user_prop)) { + if (zfs_prop_get_userquota(zhp, pl->pl_user_prop, + property, sizeof (property), B_FALSE) != 0) + propstr = "-"; + else + propstr = property; + right_justify = B_TRUE; + } else { + if (nvlist_lookup_nvlist(userprops, + pl->pl_user_prop, &propval) != 0) + propstr = "-"; + else + verify(nvlist_lookup_string(propval, + ZPROP_VALUE, &propstr) == 0); + right_justify = B_FALSE; + } + + width = pl->pl_width; + + /* + * If this is being called in scripted mode, or if this is the + * last column and it is left-justified, don't include a width + * format specifier. + */ + if (scripted || (pl->pl_next == NULL && !right_justify)) + (void) printf("%s", propstr); + else if (right_justify) + (void) printf("%*s", width, propstr); + else + (void) printf("%-*s", width, propstr); + } + + (void) printf("\n"); +} + +/* + * Generic callback function to list a dataset or snapshot. + */ +static int +list_callback(zfs_handle_t *zhp, void *data) +{ + list_cbdata_t *cbp = data; + + if (cbp->cb_first) { + if (!cbp->cb_scripted) + print_header(cbp->cb_proplist); + cbp->cb_first = B_FALSE; + } + + print_dataset(zhp, cbp->cb_proplist, cbp->cb_scripted); + + return (0); +} + +static int +zfs_do_list(int argc, char **argv) +{ + int c; + boolean_t scripted = B_FALSE; + static char default_fields[] = + "name,used,available,referenced,mountpoint"; + int types = ZFS_TYPE_DATASET; + boolean_t types_specified = B_FALSE; + char *fields = NULL; + list_cbdata_t cb = { 0 }; + char *value; + int limit = 0; + int ret; + zfs_sort_column_t *sortcol = NULL; + int flags = ZFS_ITER_PROP_LISTSNAPS | ZFS_ITER_ARGS_CAN_BE_PATHS; + + /* check options */ + while ((c = getopt(argc, argv, ":d:o:rt:Hs:S:")) != -1) { + switch (c) { + case 'o': + fields = optarg; + break; + case 'd': + limit = parse_depth(optarg, &flags); + break; + case 'r': + flags |= ZFS_ITER_RECURSE; + break; + case 'H': + scripted = B_TRUE; + break; + case 's': + if (zfs_add_sort_column(&sortcol, optarg, + B_FALSE) != 0) { + (void) fprintf(stderr, + gettext("invalid property '%s'\n"), optarg); + usage(B_FALSE); + } + break; + case 'S': + if (zfs_add_sort_column(&sortcol, optarg, + B_TRUE) != 0) { + (void) fprintf(stderr, + gettext("invalid property '%s'\n"), optarg); + usage(B_FALSE); + } + break; + case 't': + types = 0; + types_specified = B_TRUE; + flags &= ~ZFS_ITER_PROP_LISTSNAPS; + while (*optarg != '\0') { + static char *type_subopts[] = { "filesystem", + "volume", "snapshot", "all", NULL }; + + switch (getsubopt(&optarg, type_subopts, + &value)) { + case 0: + types |= ZFS_TYPE_FILESYSTEM; + break; + case 1: + types |= ZFS_TYPE_VOLUME; + break; + case 2: + types |= ZFS_TYPE_SNAPSHOT; + break; + case 3: + types = ZFS_TYPE_DATASET; + break; + + default: + (void) fprintf(stderr, + gettext("invalid type '%s'\n"), + value); + usage(B_FALSE); + } + } + break; + case ':': + (void) fprintf(stderr, gettext("missing argument for " + "'%c' option\n"), optopt); + usage(B_FALSE); + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + if (fields == NULL) + fields = default_fields; + + /* + * If "-o space" and no types were specified, don't display snapshots. + */ + if (strcmp(fields, "space") == 0 && types_specified == B_FALSE) + types &= ~ZFS_TYPE_SNAPSHOT; + + /* + * If the user specifies '-o all', the zprop_get_list() doesn't + * normally include the name of the dataset. For 'zfs list', we always + * want this property to be first. + */ + if (zprop_get_list(g_zfs, fields, &cb.cb_proplist, ZFS_TYPE_DATASET) + != 0) + usage(B_FALSE); + + cb.cb_scripted = scripted; + cb.cb_first = B_TRUE; + + ret = zfs_for_each(argc, argv, flags, types, sortcol, &cb.cb_proplist, + limit, list_callback, &cb); + + zprop_free_list(cb.cb_proplist); + zfs_free_sort_columns(sortcol); + + if (ret == 0 && cb.cb_first && !cb.cb_scripted) + (void) printf(gettext("no datasets available\n")); + + return (ret); +} + +/* + * zfs rename <fs | snap | vol> <fs | snap | vol> + * zfs rename -p <fs | vol> <fs | vol> + * zfs rename -r <snap> <snap> + * + * Renames the given dataset to another of the same type. + * + * The '-p' flag creates all the non-existing ancestors of the target first. + */ +/* ARGSUSED */ +static int +zfs_do_rename(int argc, char **argv) +{ + zfs_handle_t *zhp; + int c; + int ret; + boolean_t recurse = B_FALSE; + boolean_t parents = B_FALSE; + + /* check options */ + while ((c = getopt(argc, argv, "pr")) != -1) { + switch (c) { + case 'p': + parents = B_TRUE; + break; + case 'r': + recurse = B_TRUE; + break; + case '?': + default: + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + /* check number of arguments */ + if (argc < 1) { + (void) fprintf(stderr, gettext("missing source dataset " + "argument\n")); + usage(B_FALSE); + } + if (argc < 2) { + (void) fprintf(stderr, gettext("missing target dataset " + "argument\n")); + usage(B_FALSE); + } + if (argc > 2) { + (void) fprintf(stderr, gettext("too many arguments\n")); + usage(B_FALSE); + } + + if (recurse && parents) { + (void) fprintf(stderr, gettext("-p and -r options are mutually " + "exclusive\n")); + usage(B_FALSE); + } + + if (recurse && strchr(argv[0], '@') == 0) { + (void) fprintf(stderr, gettext("source dataset for recursive " + "rename must be a snapshot\n")); + usage(B_FALSE); + } + + if ((zhp = zfs_open(g_zfs, argv[0], parents ? ZFS_TYPE_FILESYSTEM | + ZFS_TYPE_VOLUME : ZFS_TYPE_DATASET)) == NULL) + return (1); + + /* If we were asked and the name looks good, try to create ancestors. */ + if (parents && zfs_name_valid(argv[1], zfs_get_type(zhp)) && + zfs_create_ancestors(g_zfs, argv[1]) != 0) { + zfs_close(zhp); + return (1); + } + + ret = (zfs_rename(zhp, argv[1], recurse) != 0); + + zfs_close(zhp); + return (ret); +} + +/* + * zfs promote <fs> + * + * Promotes the given clone fs to be the parent + */ +/* ARGSUSED */ +static int +zfs_do_promote(int argc, char **argv) +{ + zfs_handle_t *zhp; + int ret; + + /* check options */ + if (argc > 1 && argv[1][0] == '-') { + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + argv[1][1]); + usage(B_FALSE); + } + + /* check number of arguments */ + if (argc < 2) { + (void) fprintf(stderr, gettext("missing clone filesystem" + " argument\n")); + usage(B_FALSE); + } + if (argc > 2) { + (void) fprintf(stderr, gettext("too many arguments\n")); + usage(B_FALSE); + } + + zhp = zfs_open(g_zfs, argv[1], ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME); + if (zhp == NULL) + return (1); + + ret = (zfs_promote(zhp) != 0); + + + zfs_close(zhp); + return (ret); +} + +/* + * zfs rollback [-rRf] <snapshot> + * + * -r Delete any intervening snapshots before doing rollback + * -R Delete any snapshots and their clones + * -f ignored for backwards compatability + * + * Given a filesystem, rollback to a specific snapshot, discarding any changes + * since then and making it the active dataset. If more recent snapshots exist, + * the command will complain unless the '-r' flag is given. + */ +typedef struct rollback_cbdata { + uint64_t cb_create; + boolean_t cb_first; + int cb_doclones; + char *cb_target; + int cb_error; + boolean_t cb_recurse; + boolean_t cb_dependent; +} rollback_cbdata_t; + +/* + * Report any snapshots more recent than the one specified. Used when '-r' is + * not specified. We reuse this same callback for the snapshot dependents - if + * 'cb_dependent' is set, then this is a dependent and we should report it + * without checking the transaction group. + */ +static int +rollback_check(zfs_handle_t *zhp, void *data) +{ + rollback_cbdata_t *cbp = data; + + if (cbp->cb_doclones) { + zfs_close(zhp); + return (0); + } + + if (!cbp->cb_dependent) { + if (strcmp(zfs_get_name(zhp), cbp->cb_target) != 0 && + zfs_get_type(zhp) == ZFS_TYPE_SNAPSHOT && + zfs_prop_get_int(zhp, ZFS_PROP_CREATETXG) > + cbp->cb_create) { + + if (cbp->cb_first && !cbp->cb_recurse) { + (void) fprintf(stderr, gettext("cannot " + "rollback to '%s': more recent snapshots " + "exist\n"), + cbp->cb_target); + (void) fprintf(stderr, gettext("use '-r' to " + "force deletion of the following " + "snapshots:\n")); + cbp->cb_first = 0; + cbp->cb_error = 1; + } + + if (cbp->cb_recurse) { + cbp->cb_dependent = B_TRUE; + if (zfs_iter_dependents(zhp, B_TRUE, + rollback_check, cbp) != 0) { + zfs_close(zhp); + return (-1); + } + cbp->cb_dependent = B_FALSE; + } else { + (void) fprintf(stderr, "%s\n", + zfs_get_name(zhp)); + } + } + } else { + if (cbp->cb_first && cbp->cb_recurse) { + (void) fprintf(stderr, gettext("cannot rollback to " + "'%s': clones of previous snapshots exist\n"), + cbp->cb_target); + (void) fprintf(stderr, gettext("use '-R' to " + "force deletion of the following clones and " + "dependents:\n")); + cbp->cb_first = 0; + cbp->cb_error = 1; + } + + (void) fprintf(stderr, "%s\n", zfs_get_name(zhp)); + } + + zfs_close(zhp); + return (0); +} + +static int +zfs_do_rollback(int argc, char **argv) +{ + int ret; + int c; + boolean_t force = B_FALSE; + rollback_cbdata_t cb = { 0 }; + zfs_handle_t *zhp, *snap; + char parentname[ZFS_MAXNAMELEN]; + char *delim; + + /* check options */ + while ((c = getopt(argc, argv, "rRf")) != -1) { + switch (c) { + case 'r': + cb.cb_recurse = 1; + break; + case 'R': + cb.cb_recurse = 1; + cb.cb_doclones = 1; + break; + case 'f': + force = B_TRUE; + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + /* check number of arguments */ + if (argc < 1) { + (void) fprintf(stderr, gettext("missing dataset argument\n")); + usage(B_FALSE); + } + if (argc > 1) { + (void) fprintf(stderr, gettext("too many arguments\n")); + usage(B_FALSE); + } + + /* open the snapshot */ + if ((snap = zfs_open(g_zfs, argv[0], ZFS_TYPE_SNAPSHOT)) == NULL) + return (1); + + /* open the parent dataset */ + (void) strlcpy(parentname, argv[0], sizeof (parentname)); + verify((delim = strrchr(parentname, '@')) != NULL); + *delim = '\0'; + if ((zhp = zfs_open(g_zfs, parentname, ZFS_TYPE_DATASET)) == NULL) { + zfs_close(snap); + return (1); + } + + /* + * Check for more recent snapshots and/or clones based on the presence + * of '-r' and '-R'. + */ + cb.cb_target = argv[0]; + cb.cb_create = zfs_prop_get_int(snap, ZFS_PROP_CREATETXG); + cb.cb_first = B_TRUE; + cb.cb_error = 0; + if ((ret = zfs_iter_children(zhp, rollback_check, &cb)) != 0) + goto out; + + if ((ret = cb.cb_error) != 0) + goto out; + + /* + * Rollback parent to the given snapshot. + */ + ret = zfs_rollback(zhp, snap, force); + +out: + zfs_close(snap); + zfs_close(zhp); + + if (ret == 0) + return (0); + else + return (1); +} + +/* + * zfs set property=value { fs | snap | vol } ... + * + * Sets the given property for all datasets specified on the command line. + */ +typedef struct set_cbdata { + char *cb_propname; + char *cb_value; +} set_cbdata_t; + +static int +set_callback(zfs_handle_t *zhp, void *data) +{ + set_cbdata_t *cbp = data; + + if (zfs_prop_set(zhp, cbp->cb_propname, cbp->cb_value) != 0) { + switch (libzfs_errno(g_zfs)) { + case EZFS_MOUNTFAILED: + (void) fprintf(stderr, gettext("property may be set " + "but unable to remount filesystem\n")); + break; + case EZFS_SHARENFSFAILED: + (void) fprintf(stderr, gettext("property may be set " + "but unable to reshare filesystem\n")); + break; + } + return (1); + } + return (0); +} + +static int +zfs_do_set(int argc, char **argv) +{ + set_cbdata_t cb; + int ret; + + /* check for options */ + if (argc > 1 && argv[1][0] == '-') { + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + argv[1][1]); + usage(B_FALSE); + } + + /* check number of arguments */ + if (argc < 2) { + (void) fprintf(stderr, gettext("missing property=value " + "argument\n")); + usage(B_FALSE); + } + if (argc < 3) { + (void) fprintf(stderr, gettext("missing dataset name\n")); + usage(B_FALSE); + } + + /* validate property=value argument */ + cb.cb_propname = argv[1]; + if (((cb.cb_value = strchr(cb.cb_propname, '=')) == NULL) || + (cb.cb_value[1] == '\0')) { + (void) fprintf(stderr, gettext("missing value in " + "property=value argument\n")); + usage(B_FALSE); + } + + *cb.cb_value = '\0'; + cb.cb_value++; + + if (*cb.cb_propname == '\0') { + (void) fprintf(stderr, + gettext("missing property in property=value argument\n")); + usage(B_FALSE); + } + + ret = zfs_for_each(argc - 2, argv + 2, NULL, + ZFS_TYPE_DATASET, NULL, NULL, 0, set_callback, &cb); + + return (ret); +} + +/* + * zfs snapshot [-r] [-o prop=value] ... <fs@snap> + * + * Creates a snapshot with the given name. While functionally equivalent to + * 'zfs create', it is a separate command to differentiate intent. + */ +static int +zfs_do_snapshot(int argc, char **argv) +{ + boolean_t recursive = B_FALSE; + int ret; + char c; + nvlist_t *props; + + if (nvlist_alloc(&props, NV_UNIQUE_NAME, 0) != 0) + nomem(); + + /* check options */ + while ((c = getopt(argc, argv, "ro:")) != -1) { + switch (c) { + case 'o': + if (parseprop(props)) + return (1); + break; + case 'r': + recursive = B_TRUE; + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + goto usage; + } + } + + argc -= optind; + argv += optind; + + /* check number of arguments */ + if (argc < 1) { + (void) fprintf(stderr, gettext("missing snapshot argument\n")); + goto usage; + } + if (argc > 1) { + (void) fprintf(stderr, gettext("too many arguments\n")); + goto usage; + } + + ret = zfs_snapshot(g_zfs, argv[0], recursive, props); + nvlist_free(props); + if (ret && recursive) + (void) fprintf(stderr, gettext("no snapshots were created\n")); + return (ret != 0); + +usage: + nvlist_free(props); + usage(B_FALSE); + return (-1); +} + +/* + * zfs send [-vDp] -R [-i|-I <@snap>] <fs@snap> + * zfs send [-vDp] [-i|-I <@snap>] <fs@snap> + * + * Send a backup stream to stdout. + */ +static int +zfs_do_send(int argc, char **argv) +{ + char *fromname = NULL; + char *toname = NULL; + char *cp; + zfs_handle_t *zhp; + sendflags_t flags = { 0 }; + int c, err; + nvlist_t *dbgnv; + boolean_t extraverbose = B_FALSE; + + /* check options */ + while ((c = getopt(argc, argv, ":i:I:RDpv")) != -1) { + switch (c) { + case 'i': + if (fromname) + usage(B_FALSE); + fromname = optarg; + break; + case 'I': + if (fromname) + usage(B_FALSE); + fromname = optarg; + flags.doall = B_TRUE; + break; + case 'R': + flags.replicate = B_TRUE; + break; + case 'p': + flags.props = B_TRUE; + break; + case 'v': + if (flags.verbose) + extraverbose = B_TRUE; + flags.verbose = B_TRUE; + break; + case 'D': + flags.dedup = B_TRUE; + break; + case ':': + (void) fprintf(stderr, gettext("missing argument for " + "'%c' option\n"), optopt); + usage(B_FALSE); + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + /* check number of arguments */ + if (argc < 1) { + (void) fprintf(stderr, gettext("missing snapshot argument\n")); + usage(B_FALSE); + } + if (argc > 1) { + (void) fprintf(stderr, gettext("too many arguments\n")); + usage(B_FALSE); + } + + if (isatty(STDOUT_FILENO)) { + (void) fprintf(stderr, + gettext("Error: Stream can not be written to a terminal.\n" + "You must redirect standard output.\n")); + return (1); + } + + cp = strchr(argv[0], '@'); + if (cp == NULL) { + (void) fprintf(stderr, + gettext("argument must be a snapshot\n")); + usage(B_FALSE); + } + *cp = '\0'; + toname = cp + 1; + zhp = zfs_open(g_zfs, argv[0], ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME); + if (zhp == NULL) + return (1); + + /* + * If they specified the full path to the snapshot, chop off + * everything except the short name of the snapshot, but special + * case if they specify the origin. + */ + if (fromname && (cp = strchr(fromname, '@')) != NULL) { + char origin[ZFS_MAXNAMELEN]; + zprop_source_t src; + + (void) zfs_prop_get(zhp, ZFS_PROP_ORIGIN, + origin, sizeof (origin), &src, NULL, 0, B_FALSE); + + if (strcmp(origin, fromname) == 0) { + fromname = NULL; + flags.fromorigin = B_TRUE; + } else { + *cp = '\0'; + if (cp != fromname && strcmp(argv[0], fromname)) { + (void) fprintf(stderr, + gettext("incremental source must be " + "in same filesystem\n")); + usage(B_FALSE); + } + fromname = cp + 1; + if (strchr(fromname, '@') || strchr(fromname, '/')) { + (void) fprintf(stderr, + gettext("invalid incremental source\n")); + usage(B_FALSE); + } + } + } + + if (flags.replicate && fromname == NULL) + flags.doall = B_TRUE; + + err = zfs_send(zhp, fromname, toname, flags, STDOUT_FILENO, NULL, 0, + extraverbose ? &dbgnv : NULL); + + if (extraverbose) { + /* + * dump_nvlist prints to stdout, but that's been + * redirected to a file. Make it print to stderr + * instead. + */ + (void) dup2(STDERR_FILENO, STDOUT_FILENO); + dump_nvlist(dbgnv, 0); + nvlist_free(dbgnv); + } + zfs_close(zhp); + + return (err != 0); +} + +/* + * zfs receive [-vnFu] [-d | -e] <fs@snap> + * + * Restore a backup stream from stdin. + */ +static int +zfs_do_receive(int argc, char **argv) +{ + int c, err; + recvflags_t flags = { 0 }; + + /* check options */ + while ((c = getopt(argc, argv, ":denuvF")) != -1) { + switch (c) { + case 'd': + flags.isprefix = B_TRUE; + break; + case 'e': + flags.isprefix = B_TRUE; + flags.istail = B_TRUE; + break; + case 'n': + flags.dryrun = B_TRUE; + break; + case 'u': + flags.nomount = B_TRUE; + break; + case 'v': + flags.verbose = B_TRUE; + break; + case 'F': + flags.force = B_TRUE; + break; + case ':': + (void) fprintf(stderr, gettext("missing argument for " + "'%c' option\n"), optopt); + usage(B_FALSE); + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + /* check number of arguments */ + if (argc < 1) { + (void) fprintf(stderr, gettext("missing snapshot argument\n")); + usage(B_FALSE); + } + if (argc > 1) { + (void) fprintf(stderr, gettext("too many arguments\n")); + usage(B_FALSE); + } + + if (isatty(STDIN_FILENO)) { + (void) fprintf(stderr, + gettext("Error: Backup stream can not be read " + "from a terminal.\n" + "You must redirect standard input.\n")); + return (1); + } + + err = zfs_receive(g_zfs, argv[0], flags, STDIN_FILENO, NULL); + + return (err != 0); +} + +static int +zfs_do_hold_rele_impl(int argc, char **argv, boolean_t holding) +{ + int errors = 0; + int i; + const char *tag; + boolean_t recursive = B_FALSE; + boolean_t temphold = B_FALSE; + const char *opts = holding ? "rt" : "r"; + int c; + + /* check options */ + while ((c = getopt(argc, argv, opts)) != -1) { + switch (c) { + case 'r': + recursive = B_TRUE; + break; + case 't': + temphold = B_TRUE; + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + /* check number of arguments */ + if (argc < 2) + usage(B_FALSE); + + tag = argv[0]; + --argc; + ++argv; + + if (holding && tag[0] == '.') { + /* tags starting with '.' are reserved for libzfs */ + (void) fprintf(stderr, gettext("tag may not start with '.'\n")); + usage(B_FALSE); + } + + for (i = 0; i < argc; ++i) { + zfs_handle_t *zhp; + char parent[ZFS_MAXNAMELEN]; + const char *delim; + char *path = argv[i]; + + delim = strchr(path, '@'); + if (delim == NULL) { + (void) fprintf(stderr, + gettext("'%s' is not a snapshot\n"), path); + ++errors; + continue; + } + (void) strncpy(parent, path, delim - path); + parent[delim - path] = '\0'; + + zhp = zfs_open(g_zfs, parent, + ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME); + if (zhp == NULL) { + ++errors; + continue; + } + if (holding) { + if (zfs_hold(zhp, delim+1, tag, recursive, + temphold, B_FALSE, -1, 0, 0) != 0) + ++errors; + } else { + if (zfs_release(zhp, delim+1, tag, recursive) != 0) + ++errors; + } + zfs_close(zhp); + } + + return (errors != 0); +} + +/* + * zfs hold [-r] [-t] <tag> <snap> ... + * + * -r Recursively hold + * -t Temporary hold (hidden option) + * + * Apply a user-hold with the given tag to the list of snapshots. + */ +static int +zfs_do_hold(int argc, char **argv) +{ + return (zfs_do_hold_rele_impl(argc, argv, B_TRUE)); +} + +/* + * zfs release [-r] <tag> <snap> ... + * + * -r Recursively release + * + * Release a user-hold with the given tag from the list of snapshots. + */ +static int +zfs_do_release(int argc, char **argv) +{ + return (zfs_do_hold_rele_impl(argc, argv, B_FALSE)); +} + +#define CHECK_SPINNER 30 +#define SPINNER_TIME 3 /* seconds */ +#define MOUNT_TIME 5 /* seconds */ + +static int +get_one_dataset(zfs_handle_t *zhp, void *data) +{ + static char *spin[] = { "-", "\\", "|", "/" }; + static int spinval = 0; + static int spincheck = 0; + static time_t last_spin_time = (time_t)0; + get_all_cb_t *cbp = data; + zfs_type_t type = zfs_get_type(zhp); + + if (cbp->cb_verbose) { + if (--spincheck < 0) { + time_t now = time(NULL); + if (last_spin_time + SPINNER_TIME < now) { + update_progress(spin[spinval++ % 4]); + last_spin_time = now; + } + spincheck = CHECK_SPINNER; + } + } + + /* + * Interate over any nested datasets. + */ + if (zfs_iter_filesystems(zhp, get_one_dataset, data) != 0) { + zfs_close(zhp); + return (1); + } + + /* + * Skip any datasets whose type does not match. + */ + if ((type & ZFS_TYPE_FILESYSTEM) == 0) { + zfs_close(zhp); + return (0); + } + libzfs_add_handle(cbp, zhp); + assert(cbp->cb_used <= cbp->cb_alloc); + + return (0); +} + +static void +get_all_datasets(zfs_handle_t ***dslist, size_t *count, boolean_t verbose) +{ + get_all_cb_t cb = { 0 }; + cb.cb_verbose = verbose; + cb.cb_getone = get_one_dataset; + + if (verbose) + set_progress_header(gettext("Reading ZFS config")); + (void) zfs_iter_root(g_zfs, get_one_dataset, &cb); + + *dslist = cb.cb_handles; + *count = cb.cb_used; + + if (verbose) + finish_progress(gettext("done.")); +} + +/* + * Generic callback for sharing or mounting filesystems. Because the code is so + * similar, we have a common function with an extra parameter to determine which + * mode we are using. + */ +#define OP_SHARE 0x1 +#define OP_MOUNT 0x2 + +/* + * Share or mount a dataset. + */ +static int +share_mount_one(zfs_handle_t *zhp, int op, int flags, char *protocol, + boolean_t explicit, const char *options) +{ + char mountpoint[ZFS_MAXPROPLEN]; + char shareopts[ZFS_MAXPROPLEN]; + char smbshareopts[ZFS_MAXPROPLEN]; + const char *cmdname = op == OP_SHARE ? "share" : "mount"; + struct mnttab mnt; + uint64_t zoned, canmount; + boolean_t shared_nfs, shared_smb; + + assert(zfs_get_type(zhp) & ZFS_TYPE_FILESYSTEM); + + /* + * Check to make sure we can mount/share this dataset. If we + * are in the global zone and the filesystem is exported to a + * local zone, or if we are in a local zone and the + * filesystem is not exported, then it is an error. + */ + zoned = zfs_prop_get_int(zhp, ZFS_PROP_ZONED); + + if (zoned && getzoneid() == GLOBAL_ZONEID) { + if (!explicit) + return (0); + + (void) fprintf(stderr, gettext("cannot %s '%s': " + "dataset is exported to a local zone\n"), cmdname, + zfs_get_name(zhp)); + return (1); + + } else if (!zoned && getzoneid() != GLOBAL_ZONEID) { + if (!explicit) + return (0); + + (void) fprintf(stderr, gettext("cannot %s '%s': " + "permission denied\n"), cmdname, + zfs_get_name(zhp)); + return (1); + } + + /* + * Ignore any filesystems which don't apply to us. This + * includes those with a legacy mountpoint, or those with + * legacy share options. + */ + verify(zfs_prop_get(zhp, ZFS_PROP_MOUNTPOINT, mountpoint, + sizeof (mountpoint), NULL, NULL, 0, B_FALSE) == 0); + verify(zfs_prop_get(zhp, ZFS_PROP_SHARENFS, shareopts, + sizeof (shareopts), NULL, NULL, 0, B_FALSE) == 0); + verify(zfs_prop_get(zhp, ZFS_PROP_SHARESMB, smbshareopts, + sizeof (smbshareopts), NULL, NULL, 0, B_FALSE) == 0); + + if (op == OP_SHARE && strcmp(shareopts, "off") == 0 && + strcmp(smbshareopts, "off") == 0) { + if (!explicit) + return (0); + + (void) fprintf(stderr, gettext("cannot share '%s': " + "legacy share\n"), zfs_get_name(zhp)); + (void) fprintf(stderr, gettext("use share(1M) to " + "share this filesystem, or set " + "sharenfs property on\n")); + return (1); + } + + /* + * We cannot share or mount legacy filesystems. If the + * shareopts is non-legacy but the mountpoint is legacy, we + * treat it as a legacy share. + */ + if (strcmp(mountpoint, "legacy") == 0) { + if (!explicit) + return (0); + + (void) fprintf(stderr, gettext("cannot %s '%s': " + "legacy mountpoint\n"), cmdname, zfs_get_name(zhp)); + (void) fprintf(stderr, gettext("use %s(1M) to " + "%s this filesystem\n"), cmdname, cmdname); + return (1); + } + + if (strcmp(mountpoint, "none") == 0) { + if (!explicit) + return (0); + + (void) fprintf(stderr, gettext("cannot %s '%s': no " + "mountpoint set\n"), cmdname, zfs_get_name(zhp)); + return (1); + } + + /* + * canmount explicit outcome + * on no pass through + * on yes pass through + * off no return 0 + * off yes display error, return 1 + * noauto no return 0 + * noauto yes pass through + */ + canmount = zfs_prop_get_int(zhp, ZFS_PROP_CANMOUNT); + if (canmount == ZFS_CANMOUNT_OFF) { + if (!explicit) + return (0); + + (void) fprintf(stderr, gettext("cannot %s '%s': " + "'canmount' property is set to 'off'\n"), cmdname, + zfs_get_name(zhp)); + return (1); + } else if (canmount == ZFS_CANMOUNT_NOAUTO && !explicit) { + return (0); + } + + /* + * At this point, we have verified that the mountpoint and/or + * shareopts are appropriate for auto management. If the + * filesystem is already mounted or shared, return (failing + * for explicit requests); otherwise mount or share the + * filesystem. + */ + switch (op) { + case OP_SHARE: + + shared_nfs = zfs_is_shared_nfs(zhp, NULL); + shared_smb = zfs_is_shared_smb(zhp, NULL); + + if (shared_nfs && shared_smb || + (shared_nfs && strcmp(shareopts, "on") == 0 && + strcmp(smbshareopts, "off") == 0) || + (shared_smb && strcmp(smbshareopts, "on") == 0 && + strcmp(shareopts, "off") == 0)) { + if (!explicit) + return (0); + + (void) fprintf(stderr, gettext("cannot share " + "'%s': filesystem already shared\n"), + zfs_get_name(zhp)); + return (1); + } + + if (!zfs_is_mounted(zhp, NULL) && + zfs_mount(zhp, NULL, 0) != 0) + return (1); + + if (protocol == NULL) { + if (zfs_shareall(zhp) != 0) + return (1); + } else if (strcmp(protocol, "nfs") == 0) { + if (zfs_share_nfs(zhp)) + return (1); + } else if (strcmp(protocol, "smb") == 0) { + if (zfs_share_smb(zhp)) + return (1); + } else { + (void) fprintf(stderr, gettext("cannot share " + "'%s': invalid share type '%s' " + "specified\n"), + zfs_get_name(zhp), protocol); + return (1); + } + + break; + + case OP_MOUNT: + if (options == NULL) + mnt.mnt_mntopts = ""; + else + mnt.mnt_mntopts = (char *)options; + + if (!hasmntopt(&mnt, MNTOPT_REMOUNT) && + zfs_is_mounted(zhp, NULL)) { + if (!explicit) + return (0); + + (void) fprintf(stderr, gettext("cannot mount " + "'%s': filesystem already mounted\n"), + zfs_get_name(zhp)); + return (1); + } + + if (zfs_mount(zhp, options, flags) != 0) + return (1); + break; + } + + return (0); +} + +/* + * Reports progress in the form "(current/total)". Not thread-safe. + */ +static void +report_mount_progress(int current, int total) +{ + static time_t last_progress_time = 0; + time_t now = time(NULL); + char info[32]; + + /* report 1..n instead of 0..n-1 */ + ++current; + + /* display header if we're here for the first time */ + if (current == 1) { + set_progress_header(gettext("Mounting ZFS filesystems")); + } else if (current != total && last_progress_time + MOUNT_TIME >= now) { + /* too soon to report again */ + return; + } + + last_progress_time = now; + + (void) sprintf(info, "(%d/%d)", current, total); + + if (current == total) + finish_progress(info); + else + update_progress(info); +} + +static void +append_options(char *mntopts, char *newopts) +{ + int len = strlen(mntopts); + + /* original length plus new string to append plus 1 for the comma */ + if (len + 1 + strlen(newopts) >= MNT_LINE_MAX) { + (void) fprintf(stderr, gettext("the opts argument for " + "'%c' option is too long (more than %d chars)\n"), + "-o", MNT_LINE_MAX); + usage(B_FALSE); + } + + if (*mntopts) + mntopts[len++] = ','; + + (void) strcpy(&mntopts[len], newopts); +} + +static int +share_mount(int op, int argc, char **argv) +{ + int do_all = 0; + boolean_t verbose = B_FALSE; + int c, ret = 0; + char *options = NULL; + int flags = 0; + + /* check options */ + while ((c = getopt(argc, argv, op == OP_MOUNT ? ":avo:O" : "a")) + != -1) { + switch (c) { + case 'a': + do_all = 1; + break; + case 'v': + verbose = B_TRUE; + break; + case 'o': + if (*optarg == '\0') { + (void) fprintf(stderr, gettext("empty mount " + "options (-o) specified\n")); + usage(B_FALSE); + } + + if (options == NULL) + options = safe_malloc(MNT_LINE_MAX + 1); + + /* option validation is done later */ + append_options(options, optarg); + break; + + case 'O': + flags |= MS_OVERLAY; + break; + case ':': + (void) fprintf(stderr, gettext("missing argument for " + "'%c' option\n"), optopt); + usage(B_FALSE); + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + /* check number of arguments */ + if (do_all) { + zfs_handle_t **dslist = NULL; + size_t i, count = 0; + char *protocol = NULL; + + if (op == OP_SHARE && argc > 0) { + if (strcmp(argv[0], "nfs") != 0 && + strcmp(argv[0], "smb") != 0) { + (void) fprintf(stderr, gettext("share type " + "must be 'nfs' or 'smb'\n")); + usage(B_FALSE); + } + protocol = argv[0]; + argc--; + argv++; + } + + if (argc != 0) { + (void) fprintf(stderr, gettext("too many arguments\n")); + usage(B_FALSE); + } + + start_progress_timer(); + get_all_datasets(&dslist, &count, verbose); + + if (count == 0) + return (0); + + qsort(dslist, count, sizeof (void *), libzfs_dataset_cmp); + + for (i = 0; i < count; i++) { + if (verbose) + report_mount_progress(i, count); + + if (share_mount_one(dslist[i], op, flags, protocol, + B_FALSE, options) != 0) + ret = 1; + zfs_close(dslist[i]); + } + + free(dslist); + } else if (argc == 0) { + struct mnttab entry; + + if ((op == OP_SHARE) || (options != NULL)) { + (void) fprintf(stderr, gettext("missing filesystem " + "argument (specify -a for all)\n")); + usage(B_FALSE); + } + + /* + * When mount is given no arguments, go through /etc/mnttab and + * display any active ZFS mounts. We hide any snapshots, since + * they are controlled automatically. + */ + rewind(mnttab_file); + while (getmntent(mnttab_file, &entry) == 0) { + if (strcmp(entry.mnt_fstype, MNTTYPE_ZFS) != 0 || + strchr(entry.mnt_special, '@') != NULL) + continue; + + (void) printf("%-30s %s\n", entry.mnt_special, + entry.mnt_mountp); + } + + } else { + zfs_handle_t *zhp; + + if (argc > 1) { + (void) fprintf(stderr, + gettext("too many arguments\n")); + usage(B_FALSE); + } + + if ((zhp = zfs_open(g_zfs, argv[0], + ZFS_TYPE_FILESYSTEM)) == NULL) { + ret = 1; + } else { + ret = share_mount_one(zhp, op, flags, NULL, B_TRUE, + options); + zfs_close(zhp); + } + } + + return (ret); +} + +/* + * zfs mount -a [nfs] + * zfs mount filesystem + * + * Mount all filesystems, or mount the given filesystem. + */ +static int +zfs_do_mount(int argc, char **argv) +{ + return (share_mount(OP_MOUNT, argc, argv)); +} + +/* + * zfs share -a [nfs | smb] + * zfs share filesystem + * + * Share all filesystems, or share the given filesystem. + */ +static int +zfs_do_share(int argc, char **argv) +{ + return (share_mount(OP_SHARE, argc, argv)); +} + +typedef struct unshare_unmount_node { + zfs_handle_t *un_zhp; + char *un_mountp; + uu_avl_node_t un_avlnode; +} unshare_unmount_node_t; + +/* ARGSUSED */ +static int +unshare_unmount_compare(const void *larg, const void *rarg, void *unused) +{ + const unshare_unmount_node_t *l = larg; + const unshare_unmount_node_t *r = rarg; + + return (strcmp(l->un_mountp, r->un_mountp)); +} + +/* + * Convenience routine used by zfs_do_umount() and manual_unmount(). Given an + * absolute path, find the entry /etc/mnttab, verify that its a ZFS filesystem, + * and unmount it appropriately. + */ +static int +unshare_unmount_path(int op, char *path, int flags, boolean_t is_manual) +{ + zfs_handle_t *zhp; + int ret; + struct stat64 statbuf; + struct extmnttab entry; + const char *cmdname = (op == OP_SHARE) ? "unshare" : "unmount"; + ino_t path_inode; + + /* + * Search for the path in /etc/mnttab. Rather than looking for the + * specific path, which can be fooled by non-standard paths (i.e. ".." + * or "//"), we stat() the path and search for the corresponding + * (major,minor) device pair. + */ + if (stat64(path, &statbuf) != 0) { + (void) fprintf(stderr, gettext("cannot %s '%s': %s\n"), + cmdname, path, strerror(errno)); + return (1); + } + path_inode = statbuf.st_ino; + + /* + * Search for the given (major,minor) pair in the mount table. + */ + rewind(mnttab_file); + while ((ret = getextmntent(mnttab_file, &entry, 0)) == 0) { + if (entry.mnt_major == major(statbuf.st_dev) && + entry.mnt_minor == minor(statbuf.st_dev)) + break; + } + if (ret != 0) { + if (op == OP_SHARE) { + (void) fprintf(stderr, gettext("cannot %s '%s': not " + "currently mounted\n"), cmdname, path); + return (1); + } + (void) fprintf(stderr, gettext("warning: %s not in mnttab\n"), + path); + if ((ret = umount2(path, flags)) != 0) + (void) fprintf(stderr, gettext("%s: %s\n"), path, + strerror(errno)); + return (ret != 0); + } + + if (strcmp(entry.mnt_fstype, MNTTYPE_ZFS) != 0) { + (void) fprintf(stderr, gettext("cannot %s '%s': not a ZFS " + "filesystem\n"), cmdname, path); + return (1); + } + + if ((zhp = zfs_open(g_zfs, entry.mnt_special, + ZFS_TYPE_FILESYSTEM)) == NULL) + return (1); + + ret = 1; + if (stat64(entry.mnt_mountp, &statbuf) != 0) { + (void) fprintf(stderr, gettext("cannot %s '%s': %s\n"), + cmdname, path, strerror(errno)); + goto out; + } else if (statbuf.st_ino != path_inode) { + (void) fprintf(stderr, gettext("cannot " + "%s '%s': not a mountpoint\n"), cmdname, path); + goto out; + } + + if (op == OP_SHARE) { + char nfs_mnt_prop[ZFS_MAXPROPLEN]; + char smbshare_prop[ZFS_MAXPROPLEN]; + + verify(zfs_prop_get(zhp, ZFS_PROP_SHARENFS, nfs_mnt_prop, + sizeof (nfs_mnt_prop), NULL, NULL, 0, B_FALSE) == 0); + verify(zfs_prop_get(zhp, ZFS_PROP_SHARESMB, smbshare_prop, + sizeof (smbshare_prop), NULL, NULL, 0, B_FALSE) == 0); + + if (strcmp(nfs_mnt_prop, "off") == 0 && + strcmp(smbshare_prop, "off") == 0) { + (void) fprintf(stderr, gettext("cannot unshare " + "'%s': legacy share\n"), path); + (void) fprintf(stderr, gettext("use " + "unshare(1M) to unshare this filesystem\n")); + } else if (!zfs_is_shared(zhp)) { + (void) fprintf(stderr, gettext("cannot unshare '%s': " + "not currently shared\n"), path); + } else { + ret = zfs_unshareall_bypath(zhp, path); + } + } else { + char mtpt_prop[ZFS_MAXPROPLEN]; + + verify(zfs_prop_get(zhp, ZFS_PROP_MOUNTPOINT, mtpt_prop, + sizeof (mtpt_prop), NULL, NULL, 0, B_FALSE) == 0); + + if (is_manual) { + ret = zfs_unmount(zhp, NULL, flags); + } else if (strcmp(mtpt_prop, "legacy") == 0) { + (void) fprintf(stderr, gettext("cannot unmount " + "'%s': legacy mountpoint\n"), + zfs_get_name(zhp)); + (void) fprintf(stderr, gettext("use umount(1M) " + "to unmount this filesystem\n")); + } else { + ret = zfs_unmountall(zhp, flags); + } + } + +out: + zfs_close(zhp); + + return (ret != 0); +} + +/* + * Generic callback for unsharing or unmounting a filesystem. + */ +static int +unshare_unmount(int op, int argc, char **argv) +{ + int do_all = 0; + int flags = 0; + int ret = 0; + int c; + zfs_handle_t *zhp; + char nfs_mnt_prop[ZFS_MAXPROPLEN]; + char sharesmb[ZFS_MAXPROPLEN]; + + /* check options */ + while ((c = getopt(argc, argv, op == OP_SHARE ? "a" : "af")) != -1) { + switch (c) { + case 'a': + do_all = 1; + break; + case 'f': + flags = MS_FORCE; + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + if (do_all) { + /* + * We could make use of zfs_for_each() to walk all datasets in + * the system, but this would be very inefficient, especially + * since we would have to linearly search /etc/mnttab for each + * one. Instead, do one pass through /etc/mnttab looking for + * zfs entries and call zfs_unmount() for each one. + * + * Things get a little tricky if the administrator has created + * mountpoints beneath other ZFS filesystems. In this case, we + * have to unmount the deepest filesystems first. To accomplish + * this, we place all the mountpoints in an AVL tree sorted by + * the special type (dataset name), and walk the result in + * reverse to make sure to get any snapshots first. + */ + struct mnttab entry; + uu_avl_pool_t *pool; + uu_avl_t *tree; + unshare_unmount_node_t *node; + uu_avl_index_t idx; + uu_avl_walk_t *walk; + + if (argc != 0) { + (void) fprintf(stderr, gettext("too many arguments\n")); + usage(B_FALSE); + } + + if (((pool = uu_avl_pool_create("unmount_pool", + sizeof (unshare_unmount_node_t), + offsetof(unshare_unmount_node_t, un_avlnode), + unshare_unmount_compare, UU_DEFAULT)) == NULL) || + ((tree = uu_avl_create(pool, NULL, UU_DEFAULT)) == NULL)) + nomem(); + + rewind(mnttab_file); + while (getmntent(mnttab_file, &entry) == 0) { + + /* ignore non-ZFS entries */ + if (strcmp(entry.mnt_fstype, MNTTYPE_ZFS) != 0) + continue; + + /* ignore snapshots */ + if (strchr(entry.mnt_special, '@') != NULL) + continue; + + if ((zhp = zfs_open(g_zfs, entry.mnt_special, + ZFS_TYPE_FILESYSTEM)) == NULL) { + ret = 1; + continue; + } + + switch (op) { + case OP_SHARE: + verify(zfs_prop_get(zhp, ZFS_PROP_SHARENFS, + nfs_mnt_prop, + sizeof (nfs_mnt_prop), + NULL, NULL, 0, B_FALSE) == 0); + if (strcmp(nfs_mnt_prop, "off") != 0) + break; + verify(zfs_prop_get(zhp, ZFS_PROP_SHARESMB, + nfs_mnt_prop, + sizeof (nfs_mnt_prop), + NULL, NULL, 0, B_FALSE) == 0); + if (strcmp(nfs_mnt_prop, "off") == 0) + continue; + break; + case OP_MOUNT: + /* Ignore legacy mounts */ + verify(zfs_prop_get(zhp, ZFS_PROP_MOUNTPOINT, + nfs_mnt_prop, + sizeof (nfs_mnt_prop), + NULL, NULL, 0, B_FALSE) == 0); + if (strcmp(nfs_mnt_prop, "legacy") == 0) + continue; + /* Ignore canmount=noauto mounts */ + if (zfs_prop_get_int(zhp, ZFS_PROP_CANMOUNT) == + ZFS_CANMOUNT_NOAUTO) + continue; + default: + break; + } + + node = safe_malloc(sizeof (unshare_unmount_node_t)); + node->un_zhp = zhp; + node->un_mountp = safe_strdup(entry.mnt_mountp); + + uu_avl_node_init(node, &node->un_avlnode, pool); + + if (uu_avl_find(tree, node, NULL, &idx) == NULL) { + uu_avl_insert(tree, node, idx); + } else { + zfs_close(node->un_zhp); + free(node->un_mountp); + free(node); + } + } + + /* + * Walk the AVL tree in reverse, unmounting each filesystem and + * removing it from the AVL tree in the process. + */ + if ((walk = uu_avl_walk_start(tree, + UU_WALK_REVERSE | UU_WALK_ROBUST)) == NULL) + nomem(); + + while ((node = uu_avl_walk_next(walk)) != NULL) { + uu_avl_remove(tree, node); + + switch (op) { + case OP_SHARE: + if (zfs_unshareall_bypath(node->un_zhp, + node->un_mountp) != 0) + ret = 1; + break; + + case OP_MOUNT: + if (zfs_unmount(node->un_zhp, + node->un_mountp, flags) != 0) + ret = 1; + break; + } + + zfs_close(node->un_zhp); + free(node->un_mountp); + free(node); + } + + uu_avl_walk_end(walk); + uu_avl_destroy(tree); + uu_avl_pool_destroy(pool); + + } else { + if (argc != 1) { + if (argc == 0) + (void) fprintf(stderr, + gettext("missing filesystem argument\n")); + else + (void) fprintf(stderr, + gettext("too many arguments\n")); + usage(B_FALSE); + } + + /* + * We have an argument, but it may be a full path or a ZFS + * filesystem. Pass full paths off to unmount_path() (shared by + * manual_unmount), otherwise open the filesystem and pass to + * zfs_unmount(). + */ + if (argv[0][0] == '/') + return (unshare_unmount_path(op, argv[0], + flags, B_FALSE)); + + if ((zhp = zfs_open(g_zfs, argv[0], + ZFS_TYPE_FILESYSTEM)) == NULL) + return (1); + + verify(zfs_prop_get(zhp, op == OP_SHARE ? + ZFS_PROP_SHARENFS : ZFS_PROP_MOUNTPOINT, + nfs_mnt_prop, sizeof (nfs_mnt_prop), NULL, + NULL, 0, B_FALSE) == 0); + + switch (op) { + case OP_SHARE: + verify(zfs_prop_get(zhp, ZFS_PROP_SHARENFS, + nfs_mnt_prop, + sizeof (nfs_mnt_prop), + NULL, NULL, 0, B_FALSE) == 0); + verify(zfs_prop_get(zhp, ZFS_PROP_SHARESMB, + sharesmb, sizeof (sharesmb), NULL, NULL, + 0, B_FALSE) == 0); + + if (strcmp(nfs_mnt_prop, "off") == 0 && + strcmp(sharesmb, "off") == 0) { + (void) fprintf(stderr, gettext("cannot " + "unshare '%s': legacy share\n"), + zfs_get_name(zhp)); + (void) fprintf(stderr, gettext("use " + "unshare(1M) to unshare this " + "filesystem\n")); + ret = 1; + } else if (!zfs_is_shared(zhp)) { + (void) fprintf(stderr, gettext("cannot " + "unshare '%s': not currently " + "shared\n"), zfs_get_name(zhp)); + ret = 1; + } else if (zfs_unshareall(zhp) != 0) { + ret = 1; + } + break; + + case OP_MOUNT: + if (strcmp(nfs_mnt_prop, "legacy") == 0) { + (void) fprintf(stderr, gettext("cannot " + "unmount '%s': legacy " + "mountpoint\n"), zfs_get_name(zhp)); + (void) fprintf(stderr, gettext("use " + "umount(1M) to unmount this " + "filesystem\n")); + ret = 1; + } else if (!zfs_is_mounted(zhp, NULL)) { + (void) fprintf(stderr, gettext("cannot " + "unmount '%s': not currently " + "mounted\n"), + zfs_get_name(zhp)); + ret = 1; + } else if (zfs_unmountall(zhp, flags) != 0) { + ret = 1; + } + break; + } + + zfs_close(zhp); + } + + return (ret); +} + +/* + * zfs unmount -a + * zfs unmount filesystem + * + * Unmount all filesystems, or a specific ZFS filesystem. + */ +static int +zfs_do_unmount(int argc, char **argv) +{ + return (unshare_unmount(OP_MOUNT, argc, argv)); +} + +/* + * zfs unshare -a + * zfs unshare filesystem + * + * Unshare all filesystems, or a specific ZFS filesystem. + */ +static int +zfs_do_unshare(int argc, char **argv) +{ + return (unshare_unmount(OP_SHARE, argc, argv)); +} + +/* ARGSUSED */ +static int +zfs_do_python(int argc, char **argv) +{ + (void) execv(pypath, argv-1); + (void) printf("internal error: %s not found\n", pypath); + return (-1); +} + +/* + * Called when invoked as /etc/fs/zfs/mount. Do the mount if the mountpoint is + * 'legacy'. Otherwise, complain that use should be using 'zfs mount'. + */ +static int +manual_mount(int argc, char **argv) +{ + zfs_handle_t *zhp; + char mountpoint[ZFS_MAXPROPLEN]; + char mntopts[MNT_LINE_MAX] = { '\0' }; + int ret; + int c; + int flags = 0; + char *dataset, *path; + + /* check options */ + while ((c = getopt(argc, argv, ":mo:O")) != -1) { + switch (c) { + case 'o': + (void) strlcpy(mntopts, optarg, sizeof (mntopts)); + break; + case 'O': + flags |= MS_OVERLAY; + break; + case 'm': + flags |= MS_NOMNTTAB; + break; + case ':': + (void) fprintf(stderr, gettext("missing argument for " + "'%c' option\n"), optopt); + usage(B_FALSE); + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + (void) fprintf(stderr, gettext("usage: mount [-o opts] " + "<path>\n")); + return (2); + } + } + + argc -= optind; + argv += optind; + + /* check that we only have two arguments */ + if (argc != 2) { + if (argc == 0) + (void) fprintf(stderr, gettext("missing dataset " + "argument\n")); + else if (argc == 1) + (void) fprintf(stderr, + gettext("missing mountpoint argument\n")); + else + (void) fprintf(stderr, gettext("too many arguments\n")); + (void) fprintf(stderr, "usage: mount <dataset> <mountpoint>\n"); + return (2); + } + + dataset = argv[0]; + path = argv[1]; + + /* try to open the dataset */ + if ((zhp = zfs_open(g_zfs, dataset, ZFS_TYPE_FILESYSTEM)) == NULL) + return (1); + + (void) zfs_prop_get(zhp, ZFS_PROP_MOUNTPOINT, mountpoint, + sizeof (mountpoint), NULL, NULL, 0, B_FALSE); + + /* check for legacy mountpoint and complain appropriately */ + ret = 0; + if (strcmp(mountpoint, ZFS_MOUNTPOINT_LEGACY) == 0) { + if (mount(dataset, path, MS_OPTIONSTR | flags, MNTTYPE_ZFS, + NULL, 0, mntopts, sizeof (mntopts)) != 0) { + (void) fprintf(stderr, gettext("mount failed: %s\n"), + strerror(errno)); + ret = 1; + } + } else { + (void) fprintf(stderr, gettext("filesystem '%s' cannot be " + "mounted using 'mount -F zfs'\n"), dataset); + (void) fprintf(stderr, gettext("Use 'zfs set mountpoint=%s' " + "instead.\n"), path); + (void) fprintf(stderr, gettext("If you must use 'mount -F zfs' " + "or /etc/vfstab, use 'zfs set mountpoint=legacy'.\n")); + (void) fprintf(stderr, gettext("See zfs(1M) for more " + "information.\n")); + ret = 1; + } + + return (ret); +} + +/* + * Called when invoked as /etc/fs/zfs/umount. Unlike a manual mount, we allow + * unmounts of non-legacy filesystems, as this is the dominant administrative + * interface. + */ +static int +manual_unmount(int argc, char **argv) +{ + int flags = 0; + int c; + + /* check options */ + while ((c = getopt(argc, argv, "f")) != -1) { + switch (c) { + case 'f': + flags = MS_FORCE; + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + (void) fprintf(stderr, gettext("usage: unmount [-f] " + "<path>\n")); + return (2); + } + } + + argc -= optind; + argv += optind; + + /* check arguments */ + if (argc != 1) { + if (argc == 0) + (void) fprintf(stderr, gettext("missing path " + "argument\n")); + else + (void) fprintf(stderr, gettext("too many arguments\n")); + (void) fprintf(stderr, gettext("usage: unmount [-f] <path>\n")); + return (2); + } + + return (unshare_unmount_path(OP_MOUNT, argv[0], flags, B_TRUE)); +} + +static int +find_command_idx(char *command, int *idx) +{ + int i; + + for (i = 0; i < NCOMMAND; i++) { + if (command_table[i].name == NULL) + continue; + + if (strcmp(command, command_table[i].name) == 0) { + *idx = i; + return (0); + } + } + return (1); +} + +static int +zfs_do_diff(int argc, char **argv) +{ + zfs_handle_t *zhp; + int flags = 0; + char *tosnap = NULL; + char *fromsnap = NULL; + char *atp, *copy; + int err; + int c; + + while ((c = getopt(argc, argv, "FHt")) != -1) { + switch (c) { + case 'F': + flags |= ZFS_DIFF_CLASSIFY; + break; + case 'H': + flags |= ZFS_DIFF_PARSEABLE; + break; + case 't': + flags |= ZFS_DIFF_TIMESTAMP; + break; + default: + (void) fprintf(stderr, + gettext("invalid option '%c'\n"), optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + if (argc < 1) { + (void) fprintf(stderr, + gettext("must provide at least one snapshot name\n")); + usage(B_FALSE); + } + + if (argc > 2) { + (void) fprintf(stderr, gettext("too many arguments\n")); + usage(B_FALSE); + } + + fromsnap = argv[0]; + tosnap = (argc == 2) ? argv[1] : NULL; + + copy = NULL; + if (*fromsnap != '@') + copy = strdup(fromsnap); + else if (tosnap) + copy = strdup(tosnap); + if (copy == NULL) + usage(B_FALSE); + + if (atp = strchr(copy, '@')) + *atp = '\0'; + + if ((zhp = zfs_open(g_zfs, copy, ZFS_TYPE_FILESYSTEM)) == NULL) + return (1); + + free(copy); + + /* + * Ignore SIGPIPE so that the library can give us + * information on any failure + */ + (void) sigignore(SIGPIPE); + + err = zfs_show_diffs(zhp, STDOUT_FILENO, fromsnap, tosnap, flags); + + zfs_close(zhp); + + return (err != 0); +} + +int +main(int argc, char **argv) +{ + int ret; + int i; + char *progname; + char *cmdname; + + (void) setlocale(LC_ALL, ""); + (void) textdomain(TEXT_DOMAIN); + + opterr = 0; + + if ((g_zfs = libzfs_init()) == NULL) { + (void) fprintf(stderr, gettext("internal error: failed to " + "initialize ZFS library\n")); + return (1); + } + + zpool_set_history_str("zfs", argc, argv, history_str); + verify(zpool_stage_history(g_zfs, history_str) == 0); + + libzfs_print_on_error(g_zfs, B_TRUE); + + if ((mnttab_file = fopen(MNTTAB, "r")) == NULL) { + (void) fprintf(stderr, gettext("internal error: unable to " + "open %s\n"), MNTTAB); + return (1); + } + + /* + * This command also doubles as the /etc/fs mount and unmount program. + * Determine if we should take this behavior based on argv[0]. + */ + progname = basename(argv[0]); + if (strcmp(progname, "mount") == 0) { + ret = manual_mount(argc, argv); + } else if (strcmp(progname, "umount") == 0) { + ret = manual_unmount(argc, argv); + } else { + /* + * Make sure the user has specified some command. + */ + if (argc < 2) { + (void) fprintf(stderr, gettext("missing command\n")); + usage(B_FALSE); + } + + cmdname = argv[1]; + + /* + * The 'umount' command is an alias for 'unmount' + */ + if (strcmp(cmdname, "umount") == 0) + cmdname = "unmount"; + + /* + * The 'recv' command is an alias for 'receive' + */ + if (strcmp(cmdname, "recv") == 0) + cmdname = "receive"; + + /* + * Special case '-?' + */ + if (strcmp(cmdname, "-?") == 0) + usage(B_TRUE); + + /* + * Run the appropriate command. + */ + libzfs_mnttab_cache(g_zfs, B_TRUE); + if (find_command_idx(cmdname, &i) == 0) { + current_command = &command_table[i]; + ret = command_table[i].func(argc - 1, argv + 1); + } else if (strchr(cmdname, '=') != NULL) { + verify(find_command_idx("set", &i) == 0); + current_command = &command_table[i]; + ret = command_table[i].func(argc, argv); + } else { + (void) fprintf(stderr, gettext("unrecognized " + "command '%s'\n"), cmdname); + usage(B_FALSE); + } + libzfs_mnttab_cache(g_zfs, B_FALSE); + } + + (void) fclose(mnttab_file); + + libzfs_fini(g_zfs); + + /* + * The 'ZFS_ABORT' environment variable causes us to dump core on exit + * for the purposes of running ::findleaks. + */ + if (getenv("ZFS_ABORT") != NULL) { + (void) printf("dumping core by request\n"); + abort(); + } + + return (ret); +} diff --git a/cmd/zfs/zfs_util.h b/cmd/zfs/zfs_util.h new file mode 100644 index 000000000000..3ddff9e22d7d --- /dev/null +++ b/cmd/zfs/zfs_util.h @@ -0,0 +1,42 @@ +/* + * 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 http://www.opensolaris.org/os/licensing. + * 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. + */ + +#ifndef _ZFS_UTIL_H +#define _ZFS_UTIL_H + +#include <libzfs.h> + +#ifdef __cplusplus +extern "C" { +#endif + +void * safe_malloc(size_t size); +void nomem(void); +libzfs_handle_t *g_zfs; + +#ifdef __cplusplus +} +#endif + +#endif /* _ZFS_UTIL_H */ diff --git a/cmd/zinject/translate.c b/cmd/zinject/translate.c new file mode 100644 index 000000000000..87751e315e89 --- /dev/null +++ b/cmd/zinject/translate.c @@ -0,0 +1,494 @@ +/* + * 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 http://www.opensolaris.org/os/licensing. + * 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) 2006, 2010, Oracle and/or its affiliates. All rights reserved. + */ + +#include <libzfs.h> + +#include <sys/zfs_context.h> + +#include <errno.h> +#include <fcntl.h> +#include <stdarg.h> +#include <stddef.h> +#include <stdio.h> +#include <stdlib.h> +#include <strings.h> +#include <sys/file.h> +#include <sys/mntent.h> +#include <sys/mnttab.h> +#include <sys/param.h> +#include <sys/stat.h> + +#include <sys/dmu.h> +#include <sys/dmu_objset.h> +#include <sys/dnode.h> +#include <sys/vdev_impl.h> + +#include <sys/mkdev.h> + +#include "zinject.h" + +extern void kernel_init(int); +extern void kernel_fini(void); + +static int debug; + +static void +ziprintf(const char *fmt, ...) +{ + va_list ap; + + if (!debug) + return; + + va_start(ap, fmt); + (void) vprintf(fmt, ap); + va_end(ap); +} + +static void +compress_slashes(const char *src, char *dest) +{ + while (*src != '\0') { + *dest = *src++; + while (*dest == '/' && *src == '/') + ++src; + ++dest; + } + *dest = '\0'; +} + +/* + * Given a full path to a file, translate into a dataset name and a relative + * path within the dataset. 'dataset' must be at least MAXNAMELEN characters, + * and 'relpath' must be at least MAXPATHLEN characters. We also pass a stat64 + * buffer, which we need later to get the object ID. + */ +static int +parse_pathname(const char *inpath, char *dataset, char *relpath, + struct stat64 *statbuf) +{ + struct extmnttab mp; + FILE *fp; + int match; + const char *rel; + char fullpath[MAXPATHLEN]; + + compress_slashes(inpath, fullpath); + + if (fullpath[0] != '/') { + (void) fprintf(stderr, "invalid object '%s': must be full " + "path\n", fullpath); + usage(); + return (-1); + } + + if (strlen(fullpath) >= MAXPATHLEN) { + (void) fprintf(stderr, "invalid object; pathname too long\n"); + return (-1); + } + + if (stat64(fullpath, statbuf) != 0) { + (void) fprintf(stderr, "cannot open '%s': %s\n", + fullpath, strerror(errno)); + return (-1); + } + + if ((fp = fopen(MNTTAB, "r")) == NULL) { + (void) fprintf(stderr, "cannot open /etc/mnttab\n"); + return (-1); + } + + match = 0; + while (getextmntent(fp, &mp, sizeof (mp)) == 0) { + if (makedev(mp.mnt_major, mp.mnt_minor) == statbuf->st_dev) { + match = 1; + break; + } + } + + if (!match) { + (void) fprintf(stderr, "cannot find mountpoint for '%s'\n", + fullpath); + return (-1); + } + + if (strcmp(mp.mnt_fstype, MNTTYPE_ZFS) != 0) { + (void) fprintf(stderr, "invalid path '%s': not a ZFS " + "filesystem\n", fullpath); + return (-1); + } + + if (strncmp(fullpath, mp.mnt_mountp, strlen(mp.mnt_mountp)) != 0) { + (void) fprintf(stderr, "invalid path '%s': mountpoint " + "doesn't match path\n", fullpath); + return (-1); + } + + (void) strcpy(dataset, mp.mnt_special); + + rel = fullpath + strlen(mp.mnt_mountp); + if (rel[0] == '/') + rel++; + (void) strcpy(relpath, rel); + + return (0); +} + +/* + * Convert from a (dataset, path) pair into a (objset, object) pair. Note that + * we grab the object number from the inode number, since looking this up via + * libzpool is a real pain. + */ +/* ARGSUSED */ +static int +object_from_path(const char *dataset, const char *path, struct stat64 *statbuf, + zinject_record_t *record) +{ + objset_t *os; + int err; + + /* + * Before doing any libzpool operations, call sync() to ensure that the + * on-disk state is consistent with the in-core state. + */ + sync(); + + err = dmu_objset_own(dataset, DMU_OST_ZFS, B_TRUE, FTAG, &os); + if (err != 0) { + (void) fprintf(stderr, "cannot open dataset '%s': %s\n", + dataset, strerror(err)); + return (-1); + } + + record->zi_objset = dmu_objset_id(os); + record->zi_object = statbuf->st_ino; + + dmu_objset_disown(os, FTAG); + + return (0); +} + +/* + * Calculate the real range based on the type, level, and range given. + */ +static int +calculate_range(const char *dataset, err_type_t type, int level, char *range, + zinject_record_t *record) +{ + objset_t *os = NULL; + dnode_t *dn = NULL; + int err; + int ret = -1; + + /* + * Determine the numeric range from the string. + */ + if (range == NULL) { + /* + * If range is unspecified, set the range to [0,-1], which + * indicates that the whole object should be treated as an + * error. + */ + record->zi_start = 0; + record->zi_end = -1ULL; + } else { + char *end; + + /* XXX add support for suffixes */ + record->zi_start = strtoull(range, &end, 10); + + + if (*end == '\0') + record->zi_end = record->zi_start + 1; + else if (*end == ',') + record->zi_end = strtoull(end + 1, &end, 10); + + if (*end != '\0') { + (void) fprintf(stderr, "invalid range '%s': must be " + "a numeric range of the form 'start[,end]'\n", + range); + goto out; + } + } + + switch (type) { + case TYPE_DATA: + break; + + case TYPE_DNODE: + /* + * If this is a request to inject faults into the dnode, then we + * must translate the current (objset,object) pair into an + * offset within the metadnode for the objset. Specifying any + * kind of range with type 'dnode' is illegal. + */ + if (range != NULL) { + (void) fprintf(stderr, "range cannot be specified when " + "type is 'dnode'\n"); + goto out; + } + + record->zi_start = record->zi_object * sizeof (dnode_phys_t); + record->zi_end = record->zi_start + sizeof (dnode_phys_t); + record->zi_object = 0; + break; + } + + /* + * Get the dnode associated with object, so we can calculate the block + * size. + */ + if ((err = dmu_objset_own(dataset, DMU_OST_ANY, + B_TRUE, FTAG, &os)) != 0) { + (void) fprintf(stderr, "cannot open dataset '%s': %s\n", + dataset, strerror(err)); + goto out; + } + + if (record->zi_object == 0) { + dn = DMU_META_DNODE(os); + } else { + err = dnode_hold(os, record->zi_object, FTAG, &dn); + if (err != 0) { + (void) fprintf(stderr, "failed to hold dnode " + "for object %llu\n", + (u_longlong_t)record->zi_object); + goto out; + } + } + + + ziprintf("data shift: %d\n", (int)dn->dn_datablkshift); + ziprintf(" ind shift: %d\n", (int)dn->dn_indblkshift); + + /* + * Translate range into block IDs. + */ + if (record->zi_start != 0 || record->zi_end != -1ULL) { + record->zi_start >>= dn->dn_datablkshift; + record->zi_end >>= dn->dn_datablkshift; + } + + /* + * Check level, and then translate level 0 blkids into ranges + * appropriate for level of indirection. + */ + record->zi_level = level; + if (level > 0) { + ziprintf("level 0 blkid range: [%llu, %llu]\n", + record->zi_start, record->zi_end); + + if (level >= dn->dn_nlevels) { + (void) fprintf(stderr, "level %d exceeds max level " + "of object (%d)\n", level, dn->dn_nlevels - 1); + goto out; + } + + if (record->zi_start != 0 || record->zi_end != 0) { + int shift = dn->dn_indblkshift - SPA_BLKPTRSHIFT; + + for (; level > 0; level--) { + record->zi_start >>= shift; + record->zi_end >>= shift; + } + } + } + + ret = 0; +out: + if (dn) { + if (dn != DMU_META_DNODE(os)) + dnode_rele(dn, FTAG); + } + if (os) + dmu_objset_disown(os, FTAG); + + return (ret); +} + +int +translate_record(err_type_t type, const char *object, const char *range, + int level, zinject_record_t *record, char *poolname, char *dataset) +{ + char path[MAXPATHLEN]; + char *slash; + struct stat64 statbuf; + int ret = -1; + + kernel_init(FREAD); + + debug = (getenv("ZINJECT_DEBUG") != NULL); + + ziprintf("translating: %s\n", object); + + if (MOS_TYPE(type)) { + /* + * MOS objects are treated specially. + */ + switch (type) { + case TYPE_MOS: + record->zi_type = 0; + break; + case TYPE_MOSDIR: + record->zi_type = DMU_OT_OBJECT_DIRECTORY; + break; + case TYPE_METASLAB: + record->zi_type = DMU_OT_OBJECT_ARRAY; + break; + case TYPE_CONFIG: + record->zi_type = DMU_OT_PACKED_NVLIST; + break; + case TYPE_BPOBJ: + record->zi_type = DMU_OT_BPOBJ; + break; + case TYPE_SPACEMAP: + record->zi_type = DMU_OT_SPACE_MAP; + break; + case TYPE_ERRLOG: + record->zi_type = DMU_OT_ERROR_LOG; + break; + } + + dataset[0] = '\0'; + (void) strcpy(poolname, object); + return (0); + } + + /* + * Convert a full path into a (dataset, file) pair. + */ + if (parse_pathname(object, dataset, path, &statbuf) != 0) + goto err; + + ziprintf(" dataset: %s\n", dataset); + ziprintf(" path: %s\n", path); + + /* + * Convert (dataset, file) into (objset, object) + */ + if (object_from_path(dataset, path, &statbuf, record) != 0) + goto err; + + ziprintf("raw objset: %llu\n", record->zi_objset); + ziprintf("raw object: %llu\n", record->zi_object); + + /* + * For the given object, calculate the real (type, level, range) + */ + if (calculate_range(dataset, type, level, (char *)range, record) != 0) + goto err; + + ziprintf(" objset: %llu\n", record->zi_objset); + ziprintf(" object: %llu\n", record->zi_object); + if (record->zi_start == 0 && + record->zi_end == -1ULL) + ziprintf(" range: all\n"); + else + ziprintf(" range: [%llu, %llu]\n", record->zi_start, + record->zi_end); + + /* + * Copy the pool name + */ + (void) strcpy(poolname, dataset); + if ((slash = strchr(poolname, '/')) != NULL) + *slash = '\0'; + + ret = 0; + +err: + kernel_fini(); + return (ret); +} + +int +translate_raw(const char *str, zinject_record_t *record) +{ + /* + * A raw bookmark of the form objset:object:level:blkid, where each + * number is a hexidecimal value. + */ + if (sscanf(str, "%llx:%llx:%x:%llx", (u_longlong_t *)&record->zi_objset, + (u_longlong_t *)&record->zi_object, &record->zi_level, + (u_longlong_t *)&record->zi_start) != 4) { + (void) fprintf(stderr, "bad raw spec '%s': must be of the form " + "'objset:object:level:blkid'\n", str); + return (-1); + } + + record->zi_end = record->zi_start; + + return (0); +} + +int +translate_device(const char *pool, const char *device, err_type_t label_type, + zinject_record_t *record) +{ + char *end; + zpool_handle_t *zhp; + nvlist_t *tgt; + boolean_t isspare, iscache; + + /* + * Given a device name or GUID, create an appropriate injection record + * with zi_guid set. + */ + if ((zhp = zpool_open(g_zfs, pool)) == NULL) + return (-1); + + record->zi_guid = strtoull(device, &end, 16); + if (record->zi_guid == 0 || *end != '\0') { + tgt = zpool_find_vdev(zhp, device, &isspare, &iscache, NULL); + + if (tgt == NULL) { + (void) fprintf(stderr, "cannot find device '%s' in " + "pool '%s'\n", device, pool); + return (-1); + } + + verify(nvlist_lookup_uint64(tgt, ZPOOL_CONFIG_GUID, + &record->zi_guid) == 0); + } + + switch (label_type) { + case TYPE_LABEL_UBERBLOCK: + record->zi_start = offsetof(vdev_label_t, vl_uberblock[0]); + record->zi_end = record->zi_start + VDEV_UBERBLOCK_RING - 1; + break; + case TYPE_LABEL_NVLIST: + record->zi_start = offsetof(vdev_label_t, vl_vdev_phys); + record->zi_end = record->zi_start + VDEV_PHYS_SIZE - 1; + break; + case TYPE_LABEL_PAD1: + record->zi_start = offsetof(vdev_label_t, vl_pad1); + record->zi_end = record->zi_start + VDEV_PAD_SIZE - 1; + break; + case TYPE_LABEL_PAD2: + record->zi_start = offsetof(vdev_label_t, vl_pad2); + record->zi_end = record->zi_start + VDEV_PAD_SIZE - 1; + break; + } + return (0); +} diff --git a/cmd/zinject/zinject.c b/cmd/zinject/zinject.c new file mode 100644 index 000000000000..60c53ceb3fce --- /dev/null +++ b/cmd/zinject/zinject.c @@ -0,0 +1,972 @@ +/* + * 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 http://www.opensolaris.org/os/licensing. + * 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. + */ + +/* + * ZFS Fault Injector + * + * This userland component takes a set of options and uses libzpool to translate + * from a user-visible object type and name to an internal representation. + * There are two basic types of faults: device faults and data faults. + * + * + * DEVICE FAULTS + * + * Errors can be injected into a particular vdev using the '-d' option. This + * option takes a path or vdev GUID to uniquely identify the device within a + * pool. There are two types of errors that can be injected, EIO and ENXIO, + * that can be controlled through the '-e' option. The default is ENXIO. For + * EIO failures, any attempt to read data from the device will return EIO, but + * subsequent attempt to reopen the device will succeed. For ENXIO failures, + * any attempt to read from the device will return EIO, but any attempt to + * reopen the device will also return ENXIO. + * For label faults, the -L option must be specified. This allows faults + * to be injected into either the nvlist, uberblock, pad1, or pad2 region + * of all the labels for the specified device. + * + * This form of the command looks like: + * + * zinject -d device [-e errno] [-L <uber | nvlist | pad1 | pad2>] pool + * + * + * DATA FAULTS + * + * We begin with a tuple of the form: + * + * <type,level,range,object> + * + * type A string describing the type of data to target. Each type + * implicitly describes how to interpret 'object'. Currently, + * the following values are supported: + * + * data User data for a file + * dnode Dnode for a file or directory + * + * The following MOS objects are special. Instead of injecting + * errors on a particular object or blkid, we inject errors across + * all objects of the given type. + * + * mos Any data in the MOS + * mosdir object directory + * config pool configuration + * bpobj blkptr list + * spacemap spacemap + * metaslab metaslab + * errlog persistent error log + * + * level Object level. Defaults to '0', not applicable to all types. If + * a range is given, this corresponds to the indirect block + * corresponding to the specific range. + * + * range A numerical range [start,end) within the object. Defaults to + * the full size of the file. + * + * object A string describing the logical location of the object. For + * files and directories (currently the only supported types), + * this is the path of the object on disk. + * + * This is translated, via libzpool, into the following internal representation: + * + * <type,objset,object,level,range> + * + * These types should be self-explanatory. This tuple is then passed to the + * kernel via a special ioctl() to initiate fault injection for the given + * object. Note that 'type' is not strictly necessary for fault injection, but + * is used when translating existing faults into a human-readable string. + * + * + * The command itself takes one of the forms: + * + * zinject + * zinject <-a | -u pool> + * zinject -c <id|all> + * zinject [-q] <-t type> [-f freq] [-u] [-a] [-m] [-e errno] [-l level] + * [-r range] <object> + * zinject [-f freq] [-a] [-m] [-u] -b objset:object:level:start:end pool + * + * With no arguments, the command prints all currently registered injection + * handlers, with their numeric identifiers. + * + * The '-c' option will clear the given handler, or all handlers if 'all' is + * specified. + * + * The '-e' option takes a string describing the errno to simulate. This must + * be either 'io' or 'checksum'. In most cases this will result in the same + * behavior, but RAID-Z will produce a different set of ereports for this + * situation. + * + * The '-a', '-u', and '-m' flags toggle internal flush behavior. If '-a' is + * specified, then the ARC cache is flushed appropriately. If '-u' is + * specified, then the underlying SPA is unloaded. Either of these flags can be + * specified independently of any other handlers. The '-m' flag automatically + * does an unmount and remount of the underlying dataset to aid in flushing the + * cache. + * + * The '-f' flag controls the frequency of errors injected, expressed as a + * integer percentage between 1 and 100. The default is 100. + * + * The this form is responsible for actually injecting the handler into the + * framework. It takes the arguments described above, translates them to the + * internal tuple using libzpool, and then issues an ioctl() to register the + * handler. + * + * The final form can target a specific bookmark, regardless of whether a + * human-readable interface has been designed. It allows developers to specify + * a particular block by number. + */ + +#include <errno.h> +#include <fcntl.h> +#include <stdio.h> +#include <stdlib.h> +#include <strings.h> +#include <unistd.h> + +#include <sys/fs/zfs.h> +#include <sys/mount.h> + +#include <libzfs.h> + +#undef verify /* both libzfs.h and zfs_context.h want to define this */ + +#include "zinject.h" + +libzfs_handle_t *g_zfs; +int zfs_fd; + +#define ECKSUM EBADE + +static const char *errtable[TYPE_INVAL] = { + "data", + "dnode", + "mos", + "mosdir", + "metaslab", + "config", + "bpobj", + "spacemap", + "errlog", + "uber", + "nvlist", + "pad1", + "pad2" +}; + +static err_type_t +name_to_type(const char *arg) +{ + int i; + for (i = 0; i < TYPE_INVAL; i++) + if (strcmp(errtable[i], arg) == 0) + return (i); + + return (TYPE_INVAL); +} + +static const char * +type_to_name(uint64_t type) +{ + switch (type) { + case DMU_OT_OBJECT_DIRECTORY: + return ("mosdir"); + case DMU_OT_OBJECT_ARRAY: + return ("metaslab"); + case DMU_OT_PACKED_NVLIST: + return ("config"); + case DMU_OT_BPOBJ: + return ("bpobj"); + case DMU_OT_SPACE_MAP: + return ("spacemap"); + case DMU_OT_ERROR_LOG: + return ("errlog"); + default: + return ("-"); + } +} + + +/* + * Print usage message. + */ +void +usage(void) +{ + (void) printf( + "usage:\n" + "\n" + "\tzinject\n" + "\n" + "\t\tList all active injection records.\n" + "\n" + "\tzinject -c <id|all>\n" + "\n" + "\t\tClear the particular record (if given a numeric ID), or\n" + "\t\tall records if 'all' is specificed.\n" + "\n" + "\tzinject -p <function name> pool\n" + "\t\tInject a panic fault at the specified function. Only \n" + "\t\tfunctions which call spa_vdev_config_exit(), or \n" + "\t\tspa_vdev_exit() will trigger a panic.\n" + "\n" + "\tzinject -d device [-e errno] [-L <nvlist|uber|pad1|pad2>] [-F]\n" + "\t [-T <read|write|free|claim|all> pool\n" + "\t\tInject a fault into a particular device or the device's\n" + "\t\tlabel. Label injection can either be 'nvlist', 'uber',\n " + "\t\t'pad1', or 'pad2'.\n" + "\t\t'errno' can be 'nxio' (the default), 'io', or 'dtl'.\n" + "\n" + "\tzinject -d device -A <degrade|fault> pool\n" + "\t\tPerform a specific action on a particular device\n" + "\n" + "\tzinject -I [-s <seconds> | -g <txgs>] pool\n" + "\t\tCause the pool to stop writing blocks yet not\n" + "\t\treport errors for a duration. Simulates buggy hardware\n" + "\t\tthat fails to honor cache flush requests.\n" + "\t\tDefault duration is 30 seconds. The machine is panicked\n" + "\t\tat the end of the duration.\n" + "\n" + "\tzinject -b objset:object:level:blkid pool\n" + "\n" + "\t\tInject an error into pool 'pool' with the numeric bookmark\n" + "\t\tspecified by the remaining tuple. Each number is in\n" + "\t\thexidecimal, and only one block can be specified.\n" + "\n" + "\tzinject [-q] <-t type> [-e errno] [-l level] [-r range]\n" + "\t [-a] [-m] [-u] [-f freq] <object>\n" + "\n" + "\t\tInject an error into the object specified by the '-t' option\n" + "\t\tand the object descriptor. The 'object' parameter is\n" + "\t\tinterperted depending on the '-t' option.\n" + "\n" + "\t\t-q\tQuiet mode. Only print out the handler number added.\n" + "\t\t-e\tInject a specific error. Must be either 'io' or\n" + "\t\t\t'checksum'. Default is 'io'.\n" + "\t\t-l\tInject error at a particular block level. Default is " + "0.\n" + "\t\t-m\tAutomatically remount underlying filesystem.\n" + "\t\t-r\tInject error over a particular logical range of an\n" + "\t\t\tobject. Will be translated to the appropriate blkid\n" + "\t\t\trange according to the object's properties.\n" + "\t\t-a\tFlush the ARC cache. Can be specified without any\n" + "\t\t\tassociated object.\n" + "\t\t-u\tUnload the associated pool. Can be specified with only\n" + "\t\t\ta pool object.\n" + "\t\t-f\tOnly inject errors a fraction of the time. Expressed as\n" + "\t\t\ta percentage between 1 and 100.\n" + "\n" + "\t-t data\t\tInject an error into the plain file contents of a\n" + "\t\t\tfile. The object must be specified as a complete path\n" + "\t\t\tto a file on a ZFS filesystem.\n" + "\n" + "\t-t dnode\tInject an error into the metadnode in the block\n" + "\t\t\tcorresponding to the dnode for a file or directory. The\n" + "\t\t\t'-r' option is incompatible with this mode. The object\n" + "\t\t\tis specified as a complete path to a file or directory\n" + "\t\t\ton a ZFS filesystem.\n" + "\n" + "\t-t <mos>\tInject errors into the MOS for objects of the given\n" + "\t\t\ttype. Valid types are: mos, mosdir, config, bpobj,\n" + "\t\t\tspacemap, metaslab, errlog. The only valid <object> is\n" + "\t\t\tthe poolname.\n"); +} + +static int +iter_handlers(int (*func)(int, const char *, zinject_record_t *, void *), + void *data) +{ + zfs_cmd_t zc; + int ret; + + zc.zc_guid = 0; + + while (ioctl(zfs_fd, ZFS_IOC_INJECT_LIST_NEXT, &zc) == 0) + if ((ret = func((int)zc.zc_guid, zc.zc_name, + &zc.zc_inject_record, data)) != 0) + return (ret); + + if (errno != ENOENT) { + (void) fprintf(stderr, "Unable to list handlers: %s\n", + strerror(errno)); + return (-1); + } + + return (0); +} + +static int +print_data_handler(int id, const char *pool, zinject_record_t *record, + void *data) +{ + int *count = data; + + if (record->zi_guid != 0 || record->zi_func[0] != '\0') + return (0); + + if (*count == 0) { + (void) printf("%3s %-15s %-6s %-6s %-8s %3s %-15s\n", + "ID", "POOL", "OBJSET", "OBJECT", "TYPE", "LVL", "RANGE"); + (void) printf("--- --------------- ------ " + "------ -------- --- ---------------\n"); + } + + *count += 1; + + (void) printf("%3d %-15s %-6llu %-6llu %-8s %3d ", id, pool, + (u_longlong_t)record->zi_objset, (u_longlong_t)record->zi_object, + type_to_name(record->zi_type), record->zi_level); + + if (record->zi_start == 0 && + record->zi_end == -1ULL) + (void) printf("all\n"); + else + (void) printf("[%llu, %llu]\n", (u_longlong_t)record->zi_start, + (u_longlong_t)record->zi_end); + + return (0); +} + +static int +print_device_handler(int id, const char *pool, zinject_record_t *record, + void *data) +{ + int *count = data; + + if (record->zi_guid == 0 || record->zi_func[0] != '\0') + return (0); + + if (*count == 0) { + (void) printf("%3s %-15s %s\n", "ID", "POOL", "GUID"); + (void) printf("--- --------------- ----------------\n"); + } + + *count += 1; + + (void) printf("%3d %-15s %llx\n", id, pool, + (u_longlong_t)record->zi_guid); + + return (0); +} + +static int +print_panic_handler(int id, const char *pool, zinject_record_t *record, + void *data) +{ + int *count = data; + + if (record->zi_func[0] == '\0') + return (0); + + if (*count == 0) { + (void) printf("%3s %-15s %s\n", "ID", "POOL", "FUNCTION"); + (void) printf("--- --------------- ----------------\n"); + } + + *count += 1; + + (void) printf("%3d %-15s %s\n", id, pool, record->zi_func); + + return (0); +} + +/* + * Print all registered error handlers. Returns the number of handlers + * registered. + */ +static int +print_all_handlers(void) +{ + int count = 0, total = 0; + + (void) iter_handlers(print_device_handler, &count); + if (count > 0) { + total += count; + (void) printf("\n"); + count = 0; + } + + (void) iter_handlers(print_data_handler, &count); + if (count > 0) { + total += count; + (void) printf("\n"); + count = 0; + } + + (void) iter_handlers(print_panic_handler, &count); + + return (count + total); +} + +/* ARGSUSED */ +static int +cancel_one_handler(int id, const char *pool, zinject_record_t *record, + void *data) +{ + zfs_cmd_t zc; + + zc.zc_guid = (uint64_t)id; + + if (ioctl(zfs_fd, ZFS_IOC_CLEAR_FAULT, &zc) != 0) { + (void) fprintf(stderr, "failed to remove handler %d: %s\n", + id, strerror(errno)); + return (1); + } + + return (0); +} + +/* + * Remove all fault injection handlers. + */ +static int +cancel_all_handlers(void) +{ + int ret = iter_handlers(cancel_one_handler, NULL); + + if (ret == 0) + (void) printf("removed all registered handlers\n"); + + return (ret); +} + +/* + * Remove a specific fault injection handler. + */ +static int +cancel_handler(int id) +{ + zfs_cmd_t zc; + + zc.zc_guid = (uint64_t)id; + + if (ioctl(zfs_fd, ZFS_IOC_CLEAR_FAULT, &zc) != 0) { + (void) fprintf(stderr, "failed to remove handler %d: %s\n", + id, strerror(errno)); + return (1); + } + + (void) printf("removed handler %d\n", id); + + return (0); +} + +/* + * Register a new fault injection handler. + */ +static int +register_handler(const char *pool, int flags, zinject_record_t *record, + int quiet) +{ + zfs_cmd_t zc; + + (void) strcpy(zc.zc_name, pool); + zc.zc_inject_record = *record; + zc.zc_guid = flags; + + if (ioctl(zfs_fd, ZFS_IOC_INJECT_FAULT, &zc) != 0) { + (void) fprintf(stderr, "failed to add handler: %s\n", + strerror(errno)); + return (1); + } + + if (flags & ZINJECT_NULL) + return (0); + + if (quiet) { + (void) printf("%llu\n", (u_longlong_t)zc.zc_guid); + } else { + (void) printf("Added handler %llu with the following " + "properties:\n", (u_longlong_t)zc.zc_guid); + (void) printf(" pool: %s\n", pool); + if (record->zi_guid) { + (void) printf(" vdev: %llx\n", + (u_longlong_t)record->zi_guid); + } else if (record->zi_func[0] != '\0') { + (void) printf(" panic function: %s\n", + record->zi_func); + } else if (record->zi_duration > 0) { + (void) printf(" time: %lld seconds\n", + (u_longlong_t)record->zi_duration); + } else if (record->zi_duration < 0) { + (void) printf(" txgs: %lld \n", + (u_longlong_t)-record->zi_duration); + } else { + (void) printf("objset: %llu\n", + (u_longlong_t)record->zi_objset); + (void) printf("object: %llu\n", + (u_longlong_t)record->zi_object); + (void) printf(" type: %llu\n", + (u_longlong_t)record->zi_type); + (void) printf(" level: %d\n", record->zi_level); + if (record->zi_start == 0 && + record->zi_end == -1ULL) + (void) printf(" range: all\n"); + else + (void) printf(" range: [%llu, %llu)\n", + (u_longlong_t)record->zi_start, + (u_longlong_t)record->zi_end); + } + } + + return (0); +} + +int +perform_action(const char *pool, zinject_record_t *record, int cmd) +{ + zfs_cmd_t zc; + + ASSERT(cmd == VDEV_STATE_DEGRADED || cmd == VDEV_STATE_FAULTED); + (void) strlcpy(zc.zc_name, pool, sizeof (zc.zc_name)); + zc.zc_guid = record->zi_guid; + zc.zc_cookie = cmd; + + if (ioctl(zfs_fd, ZFS_IOC_VDEV_SET_STATE, &zc) == 0) + return (0); + + return (1); +} + +int +main(int argc, char **argv) +{ + int c; + char *range = NULL; + char *cancel = NULL; + char *end; + char *raw = NULL; + char *device = NULL; + int level = 0; + int quiet = 0; + int error = 0; + int domount = 0; + int io_type = ZIO_TYPES; + int action = VDEV_STATE_UNKNOWN; + err_type_t type = TYPE_INVAL; + err_type_t label = TYPE_INVAL; + zinject_record_t record = { 0 }; + char pool[MAXNAMELEN]; + char dataset[MAXNAMELEN]; + zfs_handle_t *zhp; + int nowrites = 0; + int dur_txg = 0; + int dur_secs = 0; + int ret; + int flags = 0; + + if ((g_zfs = libzfs_init()) == NULL) { + (void) fprintf(stderr, "internal error: failed to " + "initialize ZFS library\n"); + return (1); + } + + libzfs_print_on_error(g_zfs, B_TRUE); + + if ((zfs_fd = open(ZFS_DEV, O_RDWR)) < 0) { + (void) fprintf(stderr, "failed to open ZFS device\n"); + return (1); + } + + if (argc == 1) { + /* + * No arguments. Print the available handlers. If there are no + * available handlers, direct the user to '-h' for help + * information. + */ + if (print_all_handlers() == 0) { + (void) printf("No handlers registered.\n"); + (void) printf("Run 'zinject -h' for usage " + "information.\n"); + } + + return (0); + } + + while ((c = getopt(argc, argv, + ":aA:b:d:f:Fg:qhIc:t:T:l:mr:s:e:uL:p:")) != -1) { + switch (c) { + case 'a': + flags |= ZINJECT_FLUSH_ARC; + break; + case 'A': + if (strcasecmp(optarg, "degrade") == 0) { + action = VDEV_STATE_DEGRADED; + } else if (strcasecmp(optarg, "fault") == 0) { + action = VDEV_STATE_FAULTED; + } else { + (void) fprintf(stderr, "invalid action '%s': " + "must be 'degrade' or 'fault'\n", optarg); + usage(); + return (1); + } + break; + case 'b': + raw = optarg; + break; + case 'c': + cancel = optarg; + break; + case 'd': + device = optarg; + break; + case 'e': + if (strcasecmp(optarg, "io") == 0) { + error = EIO; + } else if (strcasecmp(optarg, "checksum") == 0) { + error = ECKSUM; + } else if (strcasecmp(optarg, "nxio") == 0) { + error = ENXIO; + } else if (strcasecmp(optarg, "dtl") == 0) { + error = ECHILD; + } else { + (void) fprintf(stderr, "invalid error type " + "'%s': must be 'io', 'checksum' or " + "'nxio'\n", optarg); + usage(); + return (1); + } + break; + case 'f': + record.zi_freq = atoi(optarg); + if (record.zi_freq < 1 || record.zi_freq > 100) { + (void) fprintf(stderr, "frequency range must " + "be in the range (0, 100]\n"); + return (1); + } + break; + case 'F': + record.zi_failfast = B_TRUE; + break; + case 'g': + dur_txg = 1; + record.zi_duration = (int)strtol(optarg, &end, 10); + if (record.zi_duration <= 0 || *end != '\0') { + (void) fprintf(stderr, "invalid duration '%s': " + "must be a positive integer\n", optarg); + usage(); + return (1); + } + /* store duration of txgs as its negative */ + record.zi_duration *= -1; + break; + case 'h': + usage(); + return (0); + case 'I': + /* default duration, if one hasn't yet been defined */ + nowrites = 1; + if (dur_secs == 0 && dur_txg == 0) + record.zi_duration = 30; + break; + case 'l': + level = (int)strtol(optarg, &end, 10); + if (*end != '\0') { + (void) fprintf(stderr, "invalid level '%s': " + "must be an integer\n", optarg); + usage(); + return (1); + } + break; + case 'm': + domount = 1; + break; + case 'p': + (void) strlcpy(record.zi_func, optarg, + sizeof (record.zi_func)); + break; + case 'q': + quiet = 1; + break; + case 'r': + range = optarg; + break; + case 's': + dur_secs = 1; + record.zi_duration = (int)strtol(optarg, &end, 10); + if (record.zi_duration <= 0 || *end != '\0') { + (void) fprintf(stderr, "invalid duration '%s': " + "must be a positive integer\n", optarg); + usage(); + return (1); + } + break; + case 'T': + if (strcasecmp(optarg, "read") == 0) { + io_type = ZIO_TYPE_READ; + } else if (strcasecmp(optarg, "write") == 0) { + io_type = ZIO_TYPE_WRITE; + } else if (strcasecmp(optarg, "free") == 0) { + io_type = ZIO_TYPE_FREE; + } else if (strcasecmp(optarg, "claim") == 0) { + io_type = ZIO_TYPE_CLAIM; + } else if (strcasecmp(optarg, "all") == 0) { + io_type = ZIO_TYPES; + } else { + (void) fprintf(stderr, "invalid I/O type " + "'%s': must be 'read', 'write', 'free', " + "'claim' or 'all'\n", optarg); + usage(); + return (1); + } + break; + case 't': + if ((type = name_to_type(optarg)) == TYPE_INVAL && + !MOS_TYPE(type)) { + (void) fprintf(stderr, "invalid type '%s'\n", + optarg); + usage(); + return (1); + } + break; + case 'u': + flags |= ZINJECT_UNLOAD_SPA; + break; + case 'L': + if ((label = name_to_type(optarg)) == TYPE_INVAL && + !LABEL_TYPE(type)) { + (void) fprintf(stderr, "invalid label type " + "'%s'\n", optarg); + usage(); + return (1); + } + break; + case ':': + (void) fprintf(stderr, "option -%c requires an " + "operand\n", optopt); + usage(); + return (1); + case '?': + (void) fprintf(stderr, "invalid option '%c'\n", + optopt); + usage(); + return (2); + } + } + + argc -= optind; + argv += optind; + + if (cancel != NULL) { + /* + * '-c' is invalid with any other options. + */ + if (raw != NULL || range != NULL || type != TYPE_INVAL || + level != 0 || record.zi_func[0] != '\0' || + record.zi_duration != 0) { + (void) fprintf(stderr, "cancel (-c) incompatible with " + "any other options\n"); + usage(); + return (2); + } + if (argc != 0) { + (void) fprintf(stderr, "extraneous argument to '-c'\n"); + usage(); + return (2); + } + + if (strcmp(cancel, "all") == 0) { + return (cancel_all_handlers()); + } else { + int id = (int)strtol(cancel, &end, 10); + if (*end != '\0') { + (void) fprintf(stderr, "invalid handle id '%s':" + " must be an integer or 'all'\n", cancel); + usage(); + return (1); + } + return (cancel_handler(id)); + } + } + + if (device != NULL) { + /* + * Device (-d) injection uses a completely different mechanism + * for doing injection, so handle it separately here. + */ + if (raw != NULL || range != NULL || type != TYPE_INVAL || + level != 0 || record.zi_func[0] != '\0' || + record.zi_duration != 0) { + (void) fprintf(stderr, "device (-d) incompatible with " + "data error injection\n"); + usage(); + return (2); + } + + if (argc != 1) { + (void) fprintf(stderr, "device (-d) injection requires " + "a single pool name\n"); + usage(); + return (2); + } + + (void) strcpy(pool, argv[0]); + dataset[0] = '\0'; + + if (error == ECKSUM) { + (void) fprintf(stderr, "device error type must be " + "'io' or 'nxio'\n"); + return (1); + } + + record.zi_iotype = io_type; + if (translate_device(pool, device, label, &record) != 0) + return (1); + if (!error) + error = ENXIO; + + if (action != VDEV_STATE_UNKNOWN) + return (perform_action(pool, &record, action)); + + } else if (raw != NULL) { + if (range != NULL || type != TYPE_INVAL || level != 0 || + record.zi_func[0] != '\0' || record.zi_duration != 0) { + (void) fprintf(stderr, "raw (-b) format with " + "any other options\n"); + usage(); + return (2); + } + + if (argc != 1) { + (void) fprintf(stderr, "raw (-b) format expects a " + "single pool name\n"); + usage(); + return (2); + } + + (void) strcpy(pool, argv[0]); + dataset[0] = '\0'; + + if (error == ENXIO) { + (void) fprintf(stderr, "data error type must be " + "'checksum' or 'io'\n"); + return (1); + } + + if (translate_raw(raw, &record) != 0) + return (1); + if (!error) + error = EIO; + } else if (record.zi_func[0] != '\0') { + if (raw != NULL || range != NULL || type != TYPE_INVAL || + level != 0 || device != NULL || record.zi_duration != 0) { + (void) fprintf(stderr, "panic (-p) incompatible with " + "other options\n"); + usage(); + return (2); + } + + if (argc < 1 || argc > 2) { + (void) fprintf(stderr, "panic (-p) injection requires " + "a single pool name and an optional id\n"); + usage(); + return (2); + } + + (void) strcpy(pool, argv[0]); + if (argv[1] != NULL) + record.zi_type = atoi(argv[1]); + dataset[0] = '\0'; + } else if (record.zi_duration != 0) { + if (nowrites == 0) { + (void) fprintf(stderr, "-s or -g meaningless " + "without -I (ignore writes)\n"); + usage(); + return (2); + } else if (dur_secs && dur_txg) { + (void) fprintf(stderr, "choose a duration either " + "in seconds (-s) or a number of txgs (-g) " + "but not both\n"); + usage(); + return (2); + } else if (argc != 1) { + (void) fprintf(stderr, "ignore writes (-I) " + "injection requires a single pool name\n"); + usage(); + return (2); + } + + (void) strcpy(pool, argv[0]); + dataset[0] = '\0'; + } else if (type == TYPE_INVAL) { + if (flags == 0) { + (void) fprintf(stderr, "at least one of '-b', '-d', " + "'-t', '-a', '-p', '-I' or '-u' " + "must be specified\n"); + usage(); + return (2); + } + + if (argc == 1 && (flags & ZINJECT_UNLOAD_SPA)) { + (void) strcpy(pool, argv[0]); + dataset[0] = '\0'; + } else if (argc != 0) { + (void) fprintf(stderr, "extraneous argument for " + "'-f'\n"); + usage(); + return (2); + } + + flags |= ZINJECT_NULL; + } else { + if (argc != 1) { + (void) fprintf(stderr, "missing object\n"); + usage(); + return (2); + } + + if (error == ENXIO) { + (void) fprintf(stderr, "data error type must be " + "'checksum' or 'io'\n"); + return (1); + } + + if (translate_record(type, argv[0], range, level, &record, pool, + dataset) != 0) + return (1); + if (!error) + error = EIO; + } + + /* + * If this is pool-wide metadata, unmount everything. The ioctl() will + * unload the pool, so that we trigger spa-wide reopen of metadata next + * time we access the pool. + */ + if (dataset[0] != '\0' && domount) { + if ((zhp = zfs_open(g_zfs, dataset, ZFS_TYPE_DATASET)) == NULL) + return (1); + + if (zfs_unmount(zhp, NULL, 0) != 0) + return (1); + } + + record.zi_error = error; + + ret = register_handler(pool, flags, &record, quiet); + + if (dataset[0] != '\0' && domount) + ret = (zfs_mount(zhp, NULL, 0) != 0); + + libzfs_fini(g_zfs); + + return (ret); +} diff --git a/cmd/zinject/zinject.h b/cmd/zinject/zinject.h new file mode 100644 index 000000000000..46fdcad8b31f --- /dev/null +++ b/cmd/zinject/zinject.h @@ -0,0 +1,70 @@ +/* + * 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 http://www.opensolaris.org/os/licensing. + * 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. + */ + +#ifndef _ZINJECT_H +#define _ZINJECT_H + +#include <sys/zfs_ioctl.h> + +#ifdef __cplusplus +extern "C" { +#endif + +typedef enum { + TYPE_DATA, /* plain file contents */ + TYPE_DNODE, /* metadnode contents */ + TYPE_MOS, /* all MOS data */ + TYPE_MOSDIR, /* MOS object directory */ + TYPE_METASLAB, /* metaslab objects */ + TYPE_CONFIG, /* MOS config */ + TYPE_BPOBJ, /* block pointer list */ + TYPE_SPACEMAP, /* space map objects */ + TYPE_ERRLOG, /* persistent error log */ + TYPE_LABEL_UBERBLOCK, /* label specific uberblock */ + TYPE_LABEL_NVLIST, /* label specific nvlist */ + TYPE_LABEL_PAD1, /* label specific 8K pad1 area */ + TYPE_LABEL_PAD2, /* label specific 8K pad2 area */ + TYPE_INVAL +} err_type_t; + +#define MOS_TYPE(t) \ + ((t) >= TYPE_MOS && (t) < TYPE_LABEL_UBERBLOCK) + +#define LABEL_TYPE(t) \ + ((t) >= TYPE_LABEL_UBERBLOCK && (t) < TYPE_INVAL) + +int translate_record(err_type_t type, const char *object, const char *range, + int level, zinject_record_t *record, char *poolname, char *dataset); +int translate_raw(const char *raw, zinject_record_t *record); +int translate_device(const char *pool, const char *device, + err_type_t label_type, zinject_record_t *record); +void usage(void); + +extern libzfs_handle_t *g_zfs; + +#ifdef __cplusplus +} +#endif + +#endif /* _ZINJECT_H */ diff --git a/cmd/zlook/zlook.c b/cmd/zlook/zlook.c new file mode 100644 index 000000000000..29a6559f9023 --- /dev/null +++ b/cmd/zlook/zlook.c @@ -0,0 +1,411 @@ +/* + * 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 http://www.opensolaris.org/os/licensing. + * 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) 2009, 2010, Oracle and/or its affiliates. All rights reserved. + */ + +/* + * This is a test program that uses ioctls to the ZFS Unit Test driver + * to perform readdirs or lookups using flags not normally available + * to user-land programs. This allows testing of the flags' + * behavior outside of a complicated consumer, such as the SMB driver. + */ + +#include <stdio.h> +#include <stdlib.h> +#include <unistd.h> +#include <stropts.h> +#include <errno.h> +#include <sys/stat.h> +#include <sys/types.h> +#include <sys/dirent.h> +#include <sys/attr.h> +#include <stddef.h> +#include <fcntl.h> +#include <string.h> +#include <time.h> + +#define _KERNEL + +#include <sys/fs/zut.h> +#include <sys/extdirent.h> + +#undef _KERNEL + +#define MAXBUF (64 * 1024) +#define BIGBUF 4096 +#define LILBUF (sizeof (dirent_t)) + +#define DIRENT_NAMELEN(reclen) \ + ((reclen) - (offsetof(dirent_t, d_name[0]))) + +static void +usage(char *pnam) +{ + (void) fprintf(stderr, "Usage:\n %s -l [-is] dir-to-look-in " + "file-in-dir [xfile-on-file]\n", pnam); + (void) fprintf(stderr, " %s -i [-ls] dir-to-look-in " + "file-in-dir [xfile-on-file]\n", pnam); + (void) fprintf(stderr, " %s -s [-il] dir-to-look-in " + "file-in-dir [xfile-on-file]\n", pnam); + (void) fprintf(stderr, "\t Perform a lookup\n"); + (void) fprintf(stderr, "\t -l == lookup\n"); + (void) fprintf(stderr, "\t -i == request FIGNORECASE\n"); + (void) fprintf(stderr, "\t -s == request stat(2) and xvattr info\n"); + (void) fprintf(stderr, " %s -r [-ea] [-b buffer-size-in-bytes] " + "dir-to-look-in [file-in-dir]\n", pnam); + (void) fprintf(stderr, " %s -e [-ra] [-b buffer-size-in-bytes] " + "dir-to-look-in [file-in-dir]\n", pnam); + (void) fprintf(stderr, " %s -a [-re] [-b buffer-size-in-bytes] " + "dir-to-look-in [file-in-dir]\n", pnam); + (void) fprintf(stderr, "\t Perform a readdir\n"); + (void) fprintf(stderr, "\t -r == readdir\n"); + (void) fprintf(stderr, "\t -e == request extended entries\n"); + (void) fprintf(stderr, "\t -a == request access filtering\n"); + (void) fprintf(stderr, "\t -b == buffer size (default 4K)\n"); + (void) fprintf(stderr, " %s -A path\n", pnam); + (void) fprintf(stderr, "\t Look up _PC_ACCESS_FILTERING " + "for path with pathconf(2)\n"); + (void) fprintf(stderr, " %s -E path\n", pnam); + (void) fprintf(stderr, "\t Look up _PC_SATTR_EXISTS " + "for path with pathconf(2)\n"); + (void) fprintf(stderr, " %s -S path\n", pnam); + (void) fprintf(stderr, "\t Look up _PC_SATTR_EXISTS " + "for path with pathconf(2)\n"); + exit(EINVAL); +} + +static void +print_extd_entries(zut_readdir_t *r) +{ + struct edirent *eodp; + char *bufstart; + + eodp = (edirent_t *)(uintptr_t)r->zr_buf; + bufstart = (char *)eodp; + while ((char *)eodp < bufstart + r->zr_bytes) { + char *blanks = " "; + int i = 0; + while (i < EDIRENT_NAMELEN(eodp->ed_reclen)) { + if (!eodp->ed_name[i]) + break; + (void) printf("%c", eodp->ed_name[i++]); + } + if (i < 16) + (void) printf("%.*s", 16 - i, blanks); + (void) printf("\t%x\n", eodp->ed_eflags); + eodp = (edirent_t *)((intptr_t)eodp + eodp->ed_reclen); + } +} + +static void +print_entries(zut_readdir_t *r) +{ + dirent64_t *dp; + char *bufstart; + + dp = (dirent64_t *)(intptr_t)r->zr_buf; + bufstart = (char *)dp; + while ((char *)dp < bufstart + r->zr_bytes) { + int i = 0; + while (i < DIRENT_NAMELEN(dp->d_reclen)) { + if (!dp->d_name[i]) + break; + (void) printf("%c", dp->d_name[i++]); + } + (void) printf("\n"); + dp = (dirent64_t *)((intptr_t)dp + dp->d_reclen); + } +} + +static void +print_stats(struct stat64 *sb) +{ + char timebuf[512]; + + (void) printf("st_mode\t\t\t%04lo\n", (unsigned long)sb->st_mode); + (void) printf("st_ino\t\t\t%llu\n", (unsigned long long)sb->st_ino); + (void) printf("st_nlink\t\t%lu\n", (unsigned long)sb->st_nlink); + (void) printf("st_uid\t\t\t%d\n", sb->st_uid); + (void) printf("st_gid\t\t\t%d\n", sb->st_gid); + (void) printf("st_size\t\t\t%lld\n", (long long)sb->st_size); + (void) printf("st_blksize\t\t%ld\n", (long)sb->st_blksize); + (void) printf("st_blocks\t\t%lld\n", (long long)sb->st_blocks); + + timebuf[0] = 0; + if (ctime_r(&sb->st_atime, timebuf, 512)) { + (void) printf("st_atime\t\t"); + (void) printf("%s", timebuf); + } + timebuf[0] = 0; + if (ctime_r(&sb->st_mtime, timebuf, 512)) { + (void) printf("st_mtime\t\t"); + (void) printf("%s", timebuf); + } + timebuf[0] = 0; + if (ctime_r(&sb->st_ctime, timebuf, 512)) { + (void) printf("st_ctime\t\t"); + (void) printf("%s", timebuf); + } +} + +static void +print_xvs(uint64_t xvs) +{ + uint_t bits; + int idx = 0; + + if (xvs == 0) + return; + + (void) printf("-------------------\n"); + (void) printf("Attribute bit(s) set:\n"); + (void) printf("-------------------\n"); + + bits = xvs & ((1 << F_ATTR_ALL) - 1); + while (bits) { + uint_t rest = bits >> 1; + if (bits & 1) { + (void) printf("%s", attr_to_name((f_attr_t)idx)); + if (rest) + (void) printf(", "); + } + idx++; + bits = rest; + } + (void) printf("\n"); +} + +int +main(int argc, char **argv) +{ + zut_lookup_t lk = {0}; + zut_readdir_t rd = {0}; + boolean_t checking = B_FALSE; + boolean_t looking = B_FALSE; + boolean_t reading = B_FALSE; + boolean_t bflag = B_FALSE; + long rddir_bufsize = BIGBUF; + int error = 0; + int check; + int fd; + int c; + + while ((c = getopt(argc, argv, "lisaerb:ASE")) != -1) { + switch (c) { + case 'l': + looking = B_TRUE; + break; + case 'i': + lk.zl_reqflags |= ZUT_IGNORECASE; + looking = B_TRUE; + break; + case 's': + lk.zl_reqflags |= ZUT_GETSTAT; + looking = B_TRUE; + break; + case 'a': + rd.zr_reqflags |= ZUT_ACCFILTER; + reading = B_TRUE; + break; + case 'e': + rd.zr_reqflags |= ZUT_EXTRDDIR; + reading = B_TRUE; + break; + case 'r': + reading = B_TRUE; + break; + case 'b': + reading = B_TRUE; + bflag = B_TRUE; + rddir_bufsize = strtol(optarg, NULL, 0); + break; + case 'A': + checking = B_TRUE; + check = _PC_ACCESS_FILTERING; + break; + case 'S': + checking = B_TRUE; + check = _PC_SATTR_ENABLED; + break; + case 'E': + checking = B_TRUE; + check = _PC_SATTR_EXISTS; + break; + case '?': + default: + usage(argv[0]); /* no return */ + } + } + + if ((checking && looking) || (checking && reading) || + (looking && reading) || (!reading && bflag) || + (!checking && !reading && !looking)) + usage(argv[0]); /* no return */ + + if (rddir_bufsize < LILBUF || rddir_bufsize > MAXBUF) { + (void) fprintf(stderr, "Sorry, buffer size " + "must be >= %d and less than or equal to %d bytes.\n", + (int)LILBUF, MAXBUF); + exit(EINVAL); + } + + if (checking) { + char pathbuf[MAXPATHLEN]; + long result; + + if (argc - optind < 1) + usage(argv[0]); /* no return */ + (void) strlcpy(pathbuf, argv[optind], MAXPATHLEN); + result = pathconf(pathbuf, check); + (void) printf("pathconf(2) check for %s\n", pathbuf); + switch (check) { + case _PC_SATTR_ENABLED: + (void) printf("System attributes "); + if (result != 0) + (void) printf("Enabled\n"); + else + (void) printf("Not enabled\n"); + break; + case _PC_SATTR_EXISTS: + (void) printf("System attributes "); + if (result != 0) + (void) printf("Exist\n"); + else + (void) printf("Do not exist\n"); + break; + case _PC_ACCESS_FILTERING: + (void) printf("Access filtering "); + if (result != 0) + (void) printf("Available\n"); + else + (void) printf("Not available\n"); + break; + } + return (result); + } + + if ((fd = open(ZUT_DEV, O_RDONLY)) < 0) { + perror(ZUT_DEV); + return (ENXIO); + } + + if (reading) { + char *buf; + + if (argc - optind < 1) + usage(argv[0]); /* no return */ + + (void) strlcpy(rd.zr_dir, argv[optind], MAXPATHLEN); + if (argc - optind > 1) { + (void) strlcpy(rd.zr_file, argv[optind + 1], + MAXNAMELEN); + rd.zr_reqflags |= ZUT_XATTR; + } + + if ((buf = malloc(rddir_bufsize)) == NULL) { + error = errno; + perror("malloc"); + (void) close(fd); + return (error); + } + + rd.zr_buf = (uint64_t)(uintptr_t)buf; + rd.zr_buflen = rddir_bufsize; + + while (!rd.zr_eof) { + int ierr; + + if ((ierr = ioctl(fd, ZUT_IOC_READDIR, &rd)) != 0) { + (void) fprintf(stderr, + "IOCTL error: %s (%d)\n", + strerror(ierr), ierr); + free(buf); + (void) close(fd); + return (ierr); + } + if (rd.zr_retcode) { + (void) fprintf(stderr, + "readdir result: %s (%d)\n", + strerror(rd.zr_retcode), rd.zr_retcode); + free(buf); + (void) close(fd); + return (rd.zr_retcode); + } + if (rd.zr_reqflags & ZUT_EXTRDDIR) + print_extd_entries(&rd); + else + print_entries(&rd); + } + free(buf); + } else { + int ierr; + + if (argc - optind < 2) + usage(argv[0]); /* no return */ + + (void) strlcpy(lk.zl_dir, argv[optind], MAXPATHLEN); + (void) strlcpy(lk.zl_file, argv[optind + 1], MAXNAMELEN); + if (argc - optind > 2) { + (void) strlcpy(lk.zl_xfile, + argv[optind + 2], MAXNAMELEN); + lk.zl_reqflags |= ZUT_XATTR; + } + + if ((ierr = ioctl(fd, ZUT_IOC_LOOKUP, &lk)) != 0) { + (void) fprintf(stderr, + "IOCTL error: %s (%d)\n", + strerror(ierr), ierr); + (void) close(fd); + return (ierr); + } + + (void) printf("\nLookup of "); + if (lk.zl_reqflags & ZUT_XATTR) { + (void) printf("extended attribute \"%s\" of ", + lk.zl_xfile); + } + (void) printf("file \"%s\" ", lk.zl_file); + (void) printf("in directory \"%s\" ", lk.zl_dir); + if (lk.zl_retcode) { + (void) printf("failed: %s (%d)\n", + strerror(lk.zl_retcode), lk.zl_retcode); + (void) close(fd); + return (lk.zl_retcode); + } + + (void) printf("succeeded.\n"); + if (lk.zl_reqflags & ZUT_IGNORECASE) { + (void) printf("----------------------------\n"); + (void) printf("dirent flags: 0x%0x\n", lk.zl_deflags); + (void) printf("real name: %s\n", lk.zl_real); + } + if (lk.zl_reqflags & ZUT_GETSTAT) { + (void) printf("----------------------------\n"); + print_stats(&lk.zl_statbuf); + print_xvs(lk.zl_xvattrs); + } + } + + (void) close(fd); + return (0); +} diff --git a/cmd/zpool/zpool_iter.c b/cmd/zpool/zpool_iter.c new file mode 100644 index 000000000000..2f0daefd5572 --- /dev/null +++ b/cmd/zpool/zpool_iter.c @@ -0,0 +1,252 @@ +/* + * 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 http://www.opensolaris.org/os/licensing. + * 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 2007 Sun Microsystems, Inc. All rights reserved. + * Use is subject to license terms. + */ + +#pragma ident "%Z%%M% %I% %E% SMI" + +#include <libintl.h> +#include <libuutil.h> +#include <stddef.h> +#include <stdio.h> +#include <stdlib.h> +#include <strings.h> + +#include <libzfs.h> + +#include "zpool_util.h" + +/* + * Private interface for iterating over pools specified on the command line. + * Most consumers will call for_each_pool, but in order to support iostat, we + * allow fined grained control through the zpool_list_t interface. + */ + +typedef struct zpool_node { + zpool_handle_t *zn_handle; + uu_avl_node_t zn_avlnode; + int zn_mark; +} zpool_node_t; + +struct zpool_list { + boolean_t zl_findall; + uu_avl_t *zl_avl; + uu_avl_pool_t *zl_pool; + zprop_list_t **zl_proplist; +}; + +/* ARGSUSED */ +static int +zpool_compare(const void *larg, const void *rarg, void *unused) +{ + zpool_handle_t *l = ((zpool_node_t *)larg)->zn_handle; + zpool_handle_t *r = ((zpool_node_t *)rarg)->zn_handle; + const char *lname = zpool_get_name(l); + const char *rname = zpool_get_name(r); + + return (strcmp(lname, rname)); +} + +/* + * Callback function for pool_list_get(). Adds the given pool to the AVL tree + * of known pools. + */ +static int +add_pool(zpool_handle_t *zhp, void *data) +{ + zpool_list_t *zlp = data; + zpool_node_t *node = safe_malloc(sizeof (zpool_node_t)); + uu_avl_index_t idx; + + node->zn_handle = zhp; + uu_avl_node_init(node, &node->zn_avlnode, zlp->zl_pool); + if (uu_avl_find(zlp->zl_avl, node, NULL, &idx) == NULL) { + if (zlp->zl_proplist && + zpool_expand_proplist(zhp, zlp->zl_proplist) != 0) { + zpool_close(zhp); + free(node); + return (-1); + } + uu_avl_insert(zlp->zl_avl, node, idx); + } else { + zpool_close(zhp); + free(node); + return (-1); + } + + return (0); +} + +/* + * Create a list of pools based on the given arguments. If we're given no + * arguments, then iterate over all pools in the system and add them to the AVL + * tree. Otherwise, add only those pool explicitly specified on the command + * line. + */ +zpool_list_t * +pool_list_get(int argc, char **argv, zprop_list_t **proplist, int *err) +{ + zpool_list_t *zlp; + + zlp = safe_malloc(sizeof (zpool_list_t)); + + zlp->zl_pool = uu_avl_pool_create("zfs_pool", sizeof (zpool_node_t), + offsetof(zpool_node_t, zn_avlnode), zpool_compare, UU_DEFAULT); + + if (zlp->zl_pool == NULL) + zpool_no_memory(); + + if ((zlp->zl_avl = uu_avl_create(zlp->zl_pool, NULL, + UU_DEFAULT)) == NULL) + zpool_no_memory(); + + zlp->zl_proplist = proplist; + + if (argc == 0) { + (void) zpool_iter(g_zfs, add_pool, zlp); + zlp->zl_findall = B_TRUE; + } else { + int i; + + for (i = 0; i < argc; i++) { + zpool_handle_t *zhp; + + if (zhp = zpool_open_canfail(g_zfs, argv[i])) { + if (add_pool(zhp, zlp) != 0) + *err = B_TRUE; + } else { + *err = B_TRUE; + } + } + } + + return (zlp); +} + +/* + * Search for any new pools, adding them to the list. We only add pools when no + * options were given on the command line. Otherwise, we keep the list fixed as + * those that were explicitly specified. + */ +void +pool_list_update(zpool_list_t *zlp) +{ + if (zlp->zl_findall) + (void) zpool_iter(g_zfs, add_pool, zlp); +} + +/* + * Iterate over all pools in the list, executing the callback for each + */ +int +pool_list_iter(zpool_list_t *zlp, int unavail, zpool_iter_f func, + void *data) +{ + zpool_node_t *node, *next_node; + int ret = 0; + + for (node = uu_avl_first(zlp->zl_avl); node != NULL; node = next_node) { + next_node = uu_avl_next(zlp->zl_avl, node); + if (zpool_get_state(node->zn_handle) != POOL_STATE_UNAVAIL || + unavail) + ret |= func(node->zn_handle, data); + } + + return (ret); +} + +/* + * Remove the given pool from the list. When running iostat, we want to remove + * those pools that no longer exist. + */ +void +pool_list_remove(zpool_list_t *zlp, zpool_handle_t *zhp) +{ + zpool_node_t search, *node; + + search.zn_handle = zhp; + if ((node = uu_avl_find(zlp->zl_avl, &search, NULL, NULL)) != NULL) { + uu_avl_remove(zlp->zl_avl, node); + zpool_close(node->zn_handle); + free(node); + } +} + +/* + * Free all the handles associated with this list. + */ +void +pool_list_free(zpool_list_t *zlp) +{ + uu_avl_walk_t *walk; + zpool_node_t *node; + + if ((walk = uu_avl_walk_start(zlp->zl_avl, UU_WALK_ROBUST)) == NULL) { + (void) fprintf(stderr, + gettext("internal error: out of memory")); + exit(1); + } + + while ((node = uu_avl_walk_next(walk)) != NULL) { + uu_avl_remove(zlp->zl_avl, node); + zpool_close(node->zn_handle); + free(node); + } + + uu_avl_walk_end(walk); + uu_avl_destroy(zlp->zl_avl); + uu_avl_pool_destroy(zlp->zl_pool); + + free(zlp); +} + +/* + * Returns the number of elements in the pool list. + */ +int +pool_list_count(zpool_list_t *zlp) +{ + return (uu_avl_numnodes(zlp->zl_avl)); +} + +/* + * High level function which iterates over all pools given on the command line, + * using the pool_list_* interfaces. + */ +int +for_each_pool(int argc, char **argv, boolean_t unavail, + zprop_list_t **proplist, zpool_iter_f func, void *data) +{ + zpool_list_t *list; + int ret = 0; + + if ((list = pool_list_get(argc, argv, proplist, &ret)) == NULL) + return (1); + + if (pool_list_iter(list, unavail, func, data) != 0) + ret = 1; + + pool_list_free(list); + + return (ret); +} diff --git a/cmd/zpool/zpool_main.c b/cmd/zpool/zpool_main.c new file mode 100644 index 000000000000..8aa985b1a552 --- /dev/null +++ b/cmd/zpool/zpool_main.c @@ -0,0 +1,4467 @@ +/* + * 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 http://www.opensolaris.org/os/licensing. + * 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. + */ + +#include <assert.h> +#include <ctype.h> +#include <dirent.h> +#include <errno.h> +#include <fcntl.h> +#include <libgen.h> +#include <libintl.h> +#include <libuutil.h> +#include <locale.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <strings.h> +#include <unistd.h> +#include <priv.h> +#include <pwd.h> +#include <zone.h> +#include <sys/fs/zfs.h> +#include <sys/stat.h> + +#include <libzfs.h> + +#include "zpool_util.h" +#include "zfs_comutil.h" + +#include "statcommon.h" + +static int zpool_do_create(int, char **); +static int zpool_do_destroy(int, char **); + +static int zpool_do_add(int, char **); +static int zpool_do_remove(int, char **); + +static int zpool_do_list(int, char **); +static int zpool_do_iostat(int, char **); +static int zpool_do_status(int, char **); + +static int zpool_do_online(int, char **); +static int zpool_do_offline(int, char **); +static int zpool_do_clear(int, char **); + +static int zpool_do_attach(int, char **); +static int zpool_do_detach(int, char **); +static int zpool_do_replace(int, char **); +static int zpool_do_split(int, char **); + +static int zpool_do_scrub(int, char **); + +static int zpool_do_import(int, char **); +static int zpool_do_export(int, char **); + +static int zpool_do_upgrade(int, char **); + +static int zpool_do_history(int, char **); + +static int zpool_do_get(int, char **); +static int zpool_do_set(int, char **); + +/* + * These libumem hooks provide a reasonable set of defaults for the allocator's + * debugging facilities. + */ + +#ifdef DEBUG +const char * +_umem_debug_init(void) +{ + return ("default,verbose"); /* $UMEM_DEBUG setting */ +} + +const char * +_umem_logging_init(void) +{ + return ("fail,contents"); /* $UMEM_LOGGING setting */ +} +#endif + +typedef enum { + HELP_ADD, + HELP_ATTACH, + HELP_CLEAR, + HELP_CREATE, + HELP_DESTROY, + HELP_DETACH, + HELP_EXPORT, + HELP_HISTORY, + HELP_IMPORT, + HELP_IOSTAT, + HELP_LIST, + HELP_OFFLINE, + HELP_ONLINE, + HELP_REPLACE, + HELP_REMOVE, + HELP_SCRUB, + HELP_STATUS, + HELP_UPGRADE, + HELP_GET, + HELP_SET, + HELP_SPLIT +} zpool_help_t; + + +typedef struct zpool_command { + const char *name; + int (*func)(int, char **); + zpool_help_t usage; +} zpool_command_t; + +/* + * Master command table. Each ZFS command has a name, associated function, and + * usage message. The usage messages need to be internationalized, so we have + * to have a function to return the usage message based on a command index. + * + * These commands are organized according to how they are displayed in the usage + * message. An empty command (one with a NULL name) indicates an empty line in + * the generic usage message. + */ +static zpool_command_t command_table[] = { + { "create", zpool_do_create, HELP_CREATE }, + { "destroy", zpool_do_destroy, HELP_DESTROY }, + { NULL }, + { "add", zpool_do_add, HELP_ADD }, + { "remove", zpool_do_remove, HELP_REMOVE }, + { NULL }, + { "list", zpool_do_list, HELP_LIST }, + { "iostat", zpool_do_iostat, HELP_IOSTAT }, + { "status", zpool_do_status, HELP_STATUS }, + { NULL }, + { "online", zpool_do_online, HELP_ONLINE }, + { "offline", zpool_do_offline, HELP_OFFLINE }, + { "clear", zpool_do_clear, HELP_CLEAR }, + { NULL }, + { "attach", zpool_do_attach, HELP_ATTACH }, + { "detach", zpool_do_detach, HELP_DETACH }, + { "replace", zpool_do_replace, HELP_REPLACE }, + { "split", zpool_do_split, HELP_SPLIT }, + { NULL }, + { "scrub", zpool_do_scrub, HELP_SCRUB }, + { NULL }, + { "import", zpool_do_import, HELP_IMPORT }, + { "export", zpool_do_export, HELP_EXPORT }, + { "upgrade", zpool_do_upgrade, HELP_UPGRADE }, + { NULL }, + { "history", zpool_do_history, HELP_HISTORY }, + { "get", zpool_do_get, HELP_GET }, + { "set", zpool_do_set, HELP_SET }, +}; + +#define NCOMMAND (sizeof (command_table) / sizeof (command_table[0])) + +zpool_command_t *current_command; +static char history_str[HIS_MAX_RECORD_LEN]; + +static uint_t timestamp_fmt = NODATE; + +static const char * +get_usage(zpool_help_t idx) { + switch (idx) { + case HELP_ADD: + return (gettext("\tadd [-fn] <pool> <vdev> ...\n")); + case HELP_ATTACH: + return (gettext("\tattach [-f] <pool> <device> " + "<new-device>\n")); + case HELP_CLEAR: + return (gettext("\tclear [-nF] <pool> [device]\n")); + case HELP_CREATE: + return (gettext("\tcreate [-fn] [-o property=value] ... \n" + "\t [-O file-system-property=value] ... \n" + "\t [-m mountpoint] [-R root] <pool> <vdev> ...\n")); + case HELP_DESTROY: + return (gettext("\tdestroy [-f] <pool>\n")); + case HELP_DETACH: + return (gettext("\tdetach <pool> <device>\n")); + case HELP_EXPORT: + return (gettext("\texport [-f] <pool> ...\n")); + case HELP_HISTORY: + return (gettext("\thistory [-il] [<pool>] ...\n")); + case HELP_IMPORT: + return (gettext("\timport [-d dir] [-D]\n" + "\timport [-d dir | -c cachefile] [-F [-n]] <pool | id>\n" + "\timport [-o mntopts] [-o property=value] ... \n" + "\t [-d dir | -c cachefile] [-D] [-f] [-m] [-N] " + "[-R root] [-F [-n]] -a\n" + "\timport [-o mntopts] [-o property=value] ... \n" + "\t [-d dir | -c cachefile] [-D] [-f] [-m] [-N] " + "[-R root] [-F [-n]]\n" + "\t <pool | id> [newpool]\n")); + case HELP_IOSTAT: + return (gettext("\tiostat [-v] [-T d|u] [pool] ... [interval " + "[count]]\n")); + case HELP_LIST: + return (gettext("\tlist [-H] [-o property[,...]] " + "[-T d|u] [pool] ... [interval [count]]\n")); + case HELP_OFFLINE: + return (gettext("\toffline [-t] <pool> <device> ...\n")); + case HELP_ONLINE: + return (gettext("\tonline <pool> <device> ...\n")); + case HELP_REPLACE: + return (gettext("\treplace [-f] <pool> <device> " + "[new-device]\n")); + case HELP_REMOVE: + return (gettext("\tremove <pool> <device> ...\n")); + case HELP_SCRUB: + return (gettext("\tscrub [-s] <pool> ...\n")); + case HELP_STATUS: + return (gettext("\tstatus [-vx] [-T d|u] [pool] ... [interval " + "[count]]\n")); + case HELP_UPGRADE: + return (gettext("\tupgrade\n" + "\tupgrade -v\n" + "\tupgrade [-V version] <-a | pool ...>\n")); + case HELP_GET: + return (gettext("\tget <\"all\" | property[,...]> " + "<pool> ...\n")); + case HELP_SET: + return (gettext("\tset <property=value> <pool> \n")); + case HELP_SPLIT: + return (gettext("\tsplit [-n] [-R altroot] [-o mntopts]\n" + "\t [-o property=value] <pool> <newpool> " + "[<device> ...]\n")); + } + + abort(); + /* NOTREACHED */ +} + + +/* + * Callback routine that will print out a pool property value. + */ +static int +print_prop_cb(int prop, void *cb) +{ + FILE *fp = cb; + + (void) fprintf(fp, "\t%-15s ", zpool_prop_to_name(prop)); + + if (zpool_prop_readonly(prop)) + (void) fprintf(fp, " NO "); + else + (void) fprintf(fp, " YES "); + + if (zpool_prop_values(prop) == NULL) + (void) fprintf(fp, "-\n"); + else + (void) fprintf(fp, "%s\n", zpool_prop_values(prop)); + + return (ZPROP_CONT); +} + +/* + * Display usage message. If we're inside a command, display only the usage for + * that command. Otherwise, iterate over the entire command table and display + * a complete usage message. + */ +void +usage(boolean_t requested) +{ + FILE *fp = requested ? stdout : stderr; + + if (current_command == NULL) { + int i; + + (void) fprintf(fp, gettext("usage: zpool command args ...\n")); + (void) fprintf(fp, + gettext("where 'command' is one of the following:\n\n")); + + for (i = 0; i < NCOMMAND; i++) { + if (command_table[i].name == NULL) + (void) fprintf(fp, "\n"); + else + (void) fprintf(fp, "%s", + get_usage(command_table[i].usage)); + } + } else { + (void) fprintf(fp, gettext("usage:\n")); + (void) fprintf(fp, "%s", get_usage(current_command->usage)); + } + + if (current_command != NULL && + ((strcmp(current_command->name, "set") == 0) || + (strcmp(current_command->name, "get") == 0) || + (strcmp(current_command->name, "list") == 0))) { + + (void) fprintf(fp, + gettext("\nthe following properties are supported:\n")); + + (void) fprintf(fp, "\n\t%-15s %s %s\n\n", + "PROPERTY", "EDIT", "VALUES"); + + /* Iterate over all properties */ + (void) zprop_iter(print_prop_cb, fp, B_FALSE, B_TRUE, + ZFS_TYPE_POOL); + } + + /* + * See comments at end of main(). + */ + if (getenv("ZFS_ABORT") != NULL) { + (void) printf("dumping core by request\n"); + abort(); + } + + exit(requested ? 0 : 2); +} + +void +print_vdev_tree(zpool_handle_t *zhp, const char *name, nvlist_t *nv, int indent, + boolean_t print_logs) +{ + nvlist_t **child; + uint_t c, children; + char *vname; + + if (name != NULL) + (void) printf("\t%*s%s\n", indent, "", name); + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_CHILDREN, + &child, &children) != 0) + return; + + for (c = 0; c < children; c++) { + uint64_t is_log = B_FALSE; + + (void) nvlist_lookup_uint64(child[c], ZPOOL_CONFIG_IS_LOG, + &is_log); + if ((is_log && !print_logs) || (!is_log && print_logs)) + continue; + + vname = zpool_vdev_name(g_zfs, zhp, child[c], B_FALSE); + print_vdev_tree(zhp, vname, child[c], indent + 2, + B_FALSE); + free(vname); + } +} + +/* + * Add a property pair (name, string-value) into a property nvlist. + */ +static int +add_prop_list(const char *propname, char *propval, nvlist_t **props, + boolean_t poolprop) +{ + zpool_prop_t prop = ZPROP_INVAL; + zfs_prop_t fprop; + nvlist_t *proplist; + const char *normnm; + char *strval; + + if (*props == NULL && + nvlist_alloc(props, NV_UNIQUE_NAME, 0) != 0) { + (void) fprintf(stderr, + gettext("internal error: out of memory\n")); + return (1); + } + + proplist = *props; + + if (poolprop) { + if ((prop = zpool_name_to_prop(propname)) == ZPROP_INVAL) { + (void) fprintf(stderr, gettext("property '%s' is " + "not a valid pool property\n"), propname); + return (2); + } + normnm = zpool_prop_to_name(prop); + } else { + if ((fprop = zfs_name_to_prop(propname)) != ZPROP_INVAL) { + normnm = zfs_prop_to_name(fprop); + } else { + normnm = propname; + } + } + + if (nvlist_lookup_string(proplist, normnm, &strval) == 0 && + prop != ZPOOL_PROP_CACHEFILE) { + (void) fprintf(stderr, gettext("property '%s' " + "specified multiple times\n"), propname); + return (2); + } + + if (nvlist_add_string(proplist, normnm, propval) != 0) { + (void) fprintf(stderr, gettext("internal " + "error: out of memory\n")); + return (1); + } + + return (0); +} + +/* + * zpool add [-fn] <pool> <vdev> ... + * + * -f Force addition of devices, even if they appear in use + * -n Do not add the devices, but display the resulting layout if + * they were to be added. + * + * Adds the given vdevs to 'pool'. As with create, the bulk of this work is + * handled by get_vdev_spec(), which constructs the nvlist needed to pass to + * libzfs. + */ +int +zpool_do_add(int argc, char **argv) +{ + boolean_t force = B_FALSE; + boolean_t dryrun = B_FALSE; + int c; + nvlist_t *nvroot; + char *poolname; + int ret; + zpool_handle_t *zhp; + nvlist_t *config; + + /* check options */ + while ((c = getopt(argc, argv, "fn")) != -1) { + switch (c) { + case 'f': + force = B_TRUE; + break; + case 'n': + dryrun = B_TRUE; + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + /* get pool name and check number of arguments */ + if (argc < 1) { + (void) fprintf(stderr, gettext("missing pool name argument\n")); + usage(B_FALSE); + } + if (argc < 2) { + (void) fprintf(stderr, gettext("missing vdev specification\n")); + usage(B_FALSE); + } + + poolname = argv[0]; + + argc--; + argv++; + + if ((zhp = zpool_open(g_zfs, poolname)) == NULL) + return (1); + + if ((config = zpool_get_config(zhp, NULL)) == NULL) { + (void) fprintf(stderr, gettext("pool '%s' is unavailable\n"), + poolname); + zpool_close(zhp); + return (1); + } + + /* pass off to get_vdev_spec for processing */ + nvroot = make_root_vdev(zhp, force, !force, B_FALSE, dryrun, + argc, argv); + if (nvroot == NULL) { + zpool_close(zhp); + return (1); + } + + if (dryrun) { + nvlist_t *poolnvroot; + + verify(nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, + &poolnvroot) == 0); + + (void) printf(gettext("would update '%s' to the following " + "configuration:\n"), zpool_get_name(zhp)); + + /* print original main pool and new tree */ + print_vdev_tree(zhp, poolname, poolnvroot, 0, B_FALSE); + print_vdev_tree(zhp, NULL, nvroot, 0, B_FALSE); + + /* Do the same for the logs */ + if (num_logs(poolnvroot) > 0) { + print_vdev_tree(zhp, "logs", poolnvroot, 0, B_TRUE); + print_vdev_tree(zhp, NULL, nvroot, 0, B_TRUE); + } else if (num_logs(nvroot) > 0) { + print_vdev_tree(zhp, "logs", nvroot, 0, B_TRUE); + } + + ret = 0; + } else { + ret = (zpool_add(zhp, nvroot) != 0); + } + + nvlist_free(nvroot); + zpool_close(zhp); + + return (ret); +} + +/* + * zpool remove <pool> <vdev> ... + * + * Removes the given vdev from the pool. Currently, this supports removing + * spares, cache, and log devices from the pool. + */ +int +zpool_do_remove(int argc, char **argv) +{ + char *poolname; + int i, ret = 0; + zpool_handle_t *zhp; + + argc--; + argv++; + + /* get pool name and check number of arguments */ + if (argc < 1) { + (void) fprintf(stderr, gettext("missing pool name argument\n")); + usage(B_FALSE); + } + if (argc < 2) { + (void) fprintf(stderr, gettext("missing device\n")); + usage(B_FALSE); + } + + poolname = argv[0]; + + if ((zhp = zpool_open(g_zfs, poolname)) == NULL) + return (1); + + for (i = 1; i < argc; i++) { + if (zpool_vdev_remove(zhp, argv[i]) != 0) + ret = 1; + } + + return (ret); +} + +/* + * zpool create [-fn] [-o property=value] ... + * [-O file-system-property=value] ... + * [-R root] [-m mountpoint] <pool> <dev> ... + * + * -f Force creation, even if devices appear in use + * -n Do not create the pool, but display the resulting layout if it + * were to be created. + * -R Create a pool under an alternate root + * -m Set default mountpoint for the root dataset. By default it's + * '/<pool>' + * -o Set property=value. + * -O Set fsproperty=value in the pool's root file system + * + * Creates the named pool according to the given vdev specification. The + * bulk of the vdev processing is done in get_vdev_spec() in zpool_vdev.c. Once + * we get the nvlist back from get_vdev_spec(), we either print out the contents + * (if '-n' was specified), or pass it to libzfs to do the creation. + */ +int +zpool_do_create(int argc, char **argv) +{ + boolean_t force = B_FALSE; + boolean_t dryrun = B_FALSE; + int c; + nvlist_t *nvroot = NULL; + char *poolname; + int ret = 1; + char *altroot = NULL; + char *mountpoint = NULL; + nvlist_t *fsprops = NULL; + nvlist_t *props = NULL; + char *propval; + + /* check options */ + while ((c = getopt(argc, argv, ":fnR:m:o:O:")) != -1) { + switch (c) { + case 'f': + force = B_TRUE; + break; + case 'n': + dryrun = B_TRUE; + break; + case 'R': + altroot = optarg; + if (add_prop_list(zpool_prop_to_name( + ZPOOL_PROP_ALTROOT), optarg, &props, B_TRUE)) + goto errout; + if (nvlist_lookup_string(props, + zpool_prop_to_name(ZPOOL_PROP_CACHEFILE), + &propval) == 0) + break; + if (add_prop_list(zpool_prop_to_name( + ZPOOL_PROP_CACHEFILE), "none", &props, B_TRUE)) + goto errout; + break; + case 'm': + mountpoint = optarg; + break; + case 'o': + if ((propval = strchr(optarg, '=')) == NULL) { + (void) fprintf(stderr, gettext("missing " + "'=' for -o option\n")); + goto errout; + } + *propval = '\0'; + propval++; + + if (add_prop_list(optarg, propval, &props, B_TRUE)) + goto errout; + break; + case 'O': + if ((propval = strchr(optarg, '=')) == NULL) { + (void) fprintf(stderr, gettext("missing " + "'=' for -O option\n")); + goto errout; + } + *propval = '\0'; + propval++; + + if (add_prop_list(optarg, propval, &fsprops, B_FALSE)) + goto errout; + break; + case ':': + (void) fprintf(stderr, gettext("missing argument for " + "'%c' option\n"), optopt); + goto badusage; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + goto badusage; + } + } + + argc -= optind; + argv += optind; + + /* get pool name and check number of arguments */ + if (argc < 1) { + (void) fprintf(stderr, gettext("missing pool name argument\n")); + goto badusage; + } + if (argc < 2) { + (void) fprintf(stderr, gettext("missing vdev specification\n")); + goto badusage; + } + + poolname = argv[0]; + + /* + * As a special case, check for use of '/' in the name, and direct the + * user to use 'zfs create' instead. + */ + if (strchr(poolname, '/') != NULL) { + (void) fprintf(stderr, gettext("cannot create '%s': invalid " + "character '/' in pool name\n"), poolname); + (void) fprintf(stderr, gettext("use 'zfs create' to " + "create a dataset\n")); + goto errout; + } + + /* pass off to get_vdev_spec for bulk processing */ + nvroot = make_root_vdev(NULL, force, !force, B_FALSE, dryrun, + argc - 1, argv + 1); + if (nvroot == NULL) + goto errout; + + /* make_root_vdev() allows 0 toplevel children if there are spares */ + if (!zfs_allocatable_devs(nvroot)) { + (void) fprintf(stderr, gettext("invalid vdev " + "specification: at least one toplevel vdev must be " + "specified\n")); + goto errout; + } + + + if (altroot != NULL && altroot[0] != '/') { + (void) fprintf(stderr, gettext("invalid alternate root '%s': " + "must be an absolute path\n"), altroot); + goto errout; + } + + /* + * Check the validity of the mountpoint and direct the user to use the + * '-m' mountpoint option if it looks like its in use. + */ + if (mountpoint == NULL || + (strcmp(mountpoint, ZFS_MOUNTPOINT_LEGACY) != 0 && + strcmp(mountpoint, ZFS_MOUNTPOINT_NONE) != 0)) { + char buf[MAXPATHLEN]; + DIR *dirp; + + if (mountpoint && mountpoint[0] != '/') { + (void) fprintf(stderr, gettext("invalid mountpoint " + "'%s': must be an absolute path, 'legacy', or " + "'none'\n"), mountpoint); + goto errout; + } + + if (mountpoint == NULL) { + if (altroot != NULL) + (void) snprintf(buf, sizeof (buf), "%s/%s", + altroot, poolname); + else + (void) snprintf(buf, sizeof (buf), "/%s", + poolname); + } else { + if (altroot != NULL) + (void) snprintf(buf, sizeof (buf), "%s%s", + altroot, mountpoint); + else + (void) snprintf(buf, sizeof (buf), "%s", + mountpoint); + } + + if ((dirp = opendir(buf)) == NULL && errno != ENOENT) { + (void) fprintf(stderr, gettext("mountpoint '%s' : " + "%s\n"), buf, strerror(errno)); + (void) fprintf(stderr, gettext("use '-m' " + "option to provide a different default\n")); + goto errout; + } else if (dirp) { + int count = 0; + + while (count < 3 && readdir(dirp) != NULL) + count++; + (void) closedir(dirp); + + if (count > 2) { + (void) fprintf(stderr, gettext("mountpoint " + "'%s' exists and is not empty\n"), buf); + (void) fprintf(stderr, gettext("use '-m' " + "option to provide a " + "different default\n")); + goto errout; + } + } + } + + if (dryrun) { + /* + * For a dry run invocation, print out a basic message and run + * through all the vdevs in the list and print out in an + * appropriate hierarchy. + */ + (void) printf(gettext("would create '%s' with the " + "following layout:\n\n"), poolname); + + print_vdev_tree(NULL, poolname, nvroot, 0, B_FALSE); + if (num_logs(nvroot) > 0) + print_vdev_tree(NULL, "logs", nvroot, 0, B_TRUE); + + ret = 0; + } else { + /* + * Hand off to libzfs. + */ + if (zpool_create(g_zfs, poolname, + nvroot, props, fsprops) == 0) { + zfs_handle_t *pool = zfs_open(g_zfs, poolname, + ZFS_TYPE_FILESYSTEM); + if (pool != NULL) { + if (mountpoint != NULL) + verify(zfs_prop_set(pool, + zfs_prop_to_name( + ZFS_PROP_MOUNTPOINT), + mountpoint) == 0); + if (zfs_mount(pool, NULL, 0) == 0) + ret = zfs_shareall(pool); + zfs_close(pool); + } + } else if (libzfs_errno(g_zfs) == EZFS_INVALIDNAME) { + (void) fprintf(stderr, gettext("pool name may have " + "been omitted\n")); + } + } + +errout: + nvlist_free(nvroot); + nvlist_free(fsprops); + nvlist_free(props); + return (ret); +badusage: + nvlist_free(fsprops); + nvlist_free(props); + usage(B_FALSE); + return (2); +} + +/* + * zpool destroy <pool> + * + * -f Forcefully unmount any datasets + * + * Destroy the given pool. Automatically unmounts any datasets in the pool. + */ +int +zpool_do_destroy(int argc, char **argv) +{ + boolean_t force = B_FALSE; + int c; + char *pool; + zpool_handle_t *zhp; + int ret; + + /* check options */ + while ((c = getopt(argc, argv, "f")) != -1) { + switch (c) { + case 'f': + force = B_TRUE; + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + /* check arguments */ + if (argc < 1) { + (void) fprintf(stderr, gettext("missing pool argument\n")); + usage(B_FALSE); + } + if (argc > 1) { + (void) fprintf(stderr, gettext("too many arguments\n")); + usage(B_FALSE); + } + + pool = argv[0]; + + if ((zhp = zpool_open_canfail(g_zfs, pool)) == NULL) { + /* + * As a special case, check for use of '/' in the name, and + * direct the user to use 'zfs destroy' instead. + */ + if (strchr(pool, '/') != NULL) + (void) fprintf(stderr, gettext("use 'zfs destroy' to " + "destroy a dataset\n")); + return (1); + } + + if (zpool_disable_datasets(zhp, force) != 0) { + (void) fprintf(stderr, gettext("could not destroy '%s': " + "could not unmount datasets\n"), zpool_get_name(zhp)); + return (1); + } + + ret = (zpool_destroy(zhp) != 0); + + zpool_close(zhp); + + return (ret); +} + +/* + * zpool export [-f] <pool> ... + * + * -f Forcefully unmount datasets + * + * Export the given pools. By default, the command will attempt to cleanly + * unmount any active datasets within the pool. If the '-f' flag is specified, + * then the datasets will be forcefully unmounted. + */ +int +zpool_do_export(int argc, char **argv) +{ + boolean_t force = B_FALSE; + boolean_t hardforce = B_FALSE; + int c; + zpool_handle_t *zhp; + int ret; + int i; + + /* check options */ + while ((c = getopt(argc, argv, "fF")) != -1) { + switch (c) { + case 'f': + force = B_TRUE; + break; + case 'F': + hardforce = B_TRUE; + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + /* check arguments */ + if (argc < 1) { + (void) fprintf(stderr, gettext("missing pool argument\n")); + usage(B_FALSE); + } + + ret = 0; + for (i = 0; i < argc; i++) { + if ((zhp = zpool_open_canfail(g_zfs, argv[i])) == NULL) { + ret = 1; + continue; + } + + if (zpool_disable_datasets(zhp, force) != 0) { + ret = 1; + zpool_close(zhp); + continue; + } + + if (hardforce) { + if (zpool_export_force(zhp) != 0) + ret = 1; + } else if (zpool_export(zhp, force) != 0) { + ret = 1; + } + + zpool_close(zhp); + } + + return (ret); +} + +/* + * Given a vdev configuration, determine the maximum width needed for the device + * name column. + */ +static int +max_width(zpool_handle_t *zhp, nvlist_t *nv, int depth, int max) +{ + char *name = zpool_vdev_name(g_zfs, zhp, nv, B_TRUE); + nvlist_t **child; + uint_t c, children; + int ret; + + if (strlen(name) + depth > max) + max = strlen(name) + depth; + + free(name); + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_SPARES, + &child, &children) == 0) { + for (c = 0; c < children; c++) + if ((ret = max_width(zhp, child[c], depth + 2, + max)) > max) + max = ret; + } + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_L2CACHE, + &child, &children) == 0) { + for (c = 0; c < children; c++) + if ((ret = max_width(zhp, child[c], depth + 2, + max)) > max) + max = ret; + } + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_CHILDREN, + &child, &children) == 0) { + for (c = 0; c < children; c++) + if ((ret = max_width(zhp, child[c], depth + 2, + max)) > max) + max = ret; + } + + + return (max); +} + +typedef struct spare_cbdata { + uint64_t cb_guid; + zpool_handle_t *cb_zhp; +} spare_cbdata_t; + +static boolean_t +find_vdev(nvlist_t *nv, uint64_t search) +{ + uint64_t guid; + nvlist_t **child; + uint_t c, children; + + if (nvlist_lookup_uint64(nv, ZPOOL_CONFIG_GUID, &guid) == 0 && + search == guid) + return (B_TRUE); + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_CHILDREN, + &child, &children) == 0) { + for (c = 0; c < children; c++) + if (find_vdev(child[c], search)) + return (B_TRUE); + } + + return (B_FALSE); +} + +static int +find_spare(zpool_handle_t *zhp, void *data) +{ + spare_cbdata_t *cbp = data; + nvlist_t *config, *nvroot; + + config = zpool_get_config(zhp, NULL); + verify(nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, + &nvroot) == 0); + + if (find_vdev(nvroot, cbp->cb_guid)) { + cbp->cb_zhp = zhp; + return (1); + } + + zpool_close(zhp); + return (0); +} + +/* + * Print out configuration state as requested by status_callback. + */ +void +print_status_config(zpool_handle_t *zhp, const char *name, nvlist_t *nv, + int namewidth, int depth, boolean_t isspare) +{ + nvlist_t **child; + uint_t c, children; + pool_scan_stat_t *ps = NULL; + vdev_stat_t *vs; + char rbuf[6], wbuf[6], cbuf[6]; + char *vname; + uint64_t notpresent; + spare_cbdata_t cb; + char *state; + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_CHILDREN, + &child, &children) != 0) + children = 0; + + verify(nvlist_lookup_uint64_array(nv, ZPOOL_CONFIG_VDEV_STATS, + (uint64_t **)&vs, &c) == 0); + + state = zpool_state_to_name(vs->vs_state, vs->vs_aux); + if (isspare) { + /* + * For hot spares, we use the terms 'INUSE' and 'AVAILABLE' for + * online drives. + */ + if (vs->vs_aux == VDEV_AUX_SPARED) + state = "INUSE"; + else if (vs->vs_state == VDEV_STATE_HEALTHY) + state = "AVAIL"; + } + + (void) printf("\t%*s%-*s %-8s", depth, "", namewidth - depth, + name, state); + + if (!isspare) { + zfs_nicenum(vs->vs_read_errors, rbuf, sizeof (rbuf)); + zfs_nicenum(vs->vs_write_errors, wbuf, sizeof (wbuf)); + zfs_nicenum(vs->vs_checksum_errors, cbuf, sizeof (cbuf)); + (void) printf(" %5s %5s %5s", rbuf, wbuf, cbuf); + } + + if (nvlist_lookup_uint64(nv, ZPOOL_CONFIG_NOT_PRESENT, + ¬present) == 0) { + char *path; + verify(nvlist_lookup_string(nv, ZPOOL_CONFIG_PATH, &path) == 0); + (void) printf(" was %s", path); + } else if (vs->vs_aux != 0) { + (void) printf(" "); + + switch (vs->vs_aux) { + case VDEV_AUX_OPEN_FAILED: + (void) printf(gettext("cannot open")); + break; + + case VDEV_AUX_BAD_GUID_SUM: + (void) printf(gettext("missing device")); + break; + + case VDEV_AUX_NO_REPLICAS: + (void) printf(gettext("insufficient replicas")); + break; + + case VDEV_AUX_VERSION_NEWER: + (void) printf(gettext("newer version")); + break; + + case VDEV_AUX_SPARED: + verify(nvlist_lookup_uint64(nv, ZPOOL_CONFIG_GUID, + &cb.cb_guid) == 0); + if (zpool_iter(g_zfs, find_spare, &cb) == 1) { + if (strcmp(zpool_get_name(cb.cb_zhp), + zpool_get_name(zhp)) == 0) + (void) printf(gettext("currently in " + "use")); + else + (void) printf(gettext("in use by " + "pool '%s'"), + zpool_get_name(cb.cb_zhp)); + zpool_close(cb.cb_zhp); + } else { + (void) printf(gettext("currently in use")); + } + break; + + case VDEV_AUX_ERR_EXCEEDED: + (void) printf(gettext("too many errors")); + break; + + case VDEV_AUX_IO_FAILURE: + (void) printf(gettext("experienced I/O failures")); + break; + + case VDEV_AUX_BAD_LOG: + (void) printf(gettext("bad intent log")); + break; + + case VDEV_AUX_EXTERNAL: + (void) printf(gettext("external device fault")); + break; + + case VDEV_AUX_SPLIT_POOL: + (void) printf(gettext("split into new pool")); + break; + + default: + (void) printf(gettext("corrupted data")); + break; + } + } + + (void) nvlist_lookup_uint64_array(nv, ZPOOL_CONFIG_SCAN_STATS, + (uint64_t **)&ps, &c); + + if (ps && ps->pss_state == DSS_SCANNING && + vs->vs_scan_processed != 0 && children == 0) { + (void) printf(gettext(" (%s)"), + (ps->pss_func == POOL_SCAN_RESILVER) ? + "resilvering" : "repairing"); + } + + (void) printf("\n"); + + for (c = 0; c < children; c++) { + uint64_t islog = B_FALSE, ishole = B_FALSE; + + /* Don't print logs or holes here */ + (void) nvlist_lookup_uint64(child[c], ZPOOL_CONFIG_IS_LOG, + &islog); + (void) nvlist_lookup_uint64(child[c], ZPOOL_CONFIG_IS_HOLE, + &ishole); + if (islog || ishole) + continue; + vname = zpool_vdev_name(g_zfs, zhp, child[c], B_TRUE); + print_status_config(zhp, vname, child[c], + namewidth, depth + 2, isspare); + free(vname); + } +} + + +/* + * Print the configuration of an exported pool. Iterate over all vdevs in the + * pool, printing out the name and status for each one. + */ +void +print_import_config(const char *name, nvlist_t *nv, int namewidth, int depth) +{ + nvlist_t **child; + uint_t c, children; + vdev_stat_t *vs; + char *type, *vname; + + verify(nvlist_lookup_string(nv, ZPOOL_CONFIG_TYPE, &type) == 0); + if (strcmp(type, VDEV_TYPE_MISSING) == 0 || + strcmp(type, VDEV_TYPE_HOLE) == 0) + return; + + verify(nvlist_lookup_uint64_array(nv, ZPOOL_CONFIG_VDEV_STATS, + (uint64_t **)&vs, &c) == 0); + + (void) printf("\t%*s%-*s", depth, "", namewidth - depth, name); + (void) printf(" %s", zpool_state_to_name(vs->vs_state, vs->vs_aux)); + + if (vs->vs_aux != 0) { + (void) printf(" "); + + switch (vs->vs_aux) { + case VDEV_AUX_OPEN_FAILED: + (void) printf(gettext("cannot open")); + break; + + case VDEV_AUX_BAD_GUID_SUM: + (void) printf(gettext("missing device")); + break; + + case VDEV_AUX_NO_REPLICAS: + (void) printf(gettext("insufficient replicas")); + break; + + case VDEV_AUX_VERSION_NEWER: + (void) printf(gettext("newer version")); + break; + + case VDEV_AUX_ERR_EXCEEDED: + (void) printf(gettext("too many errors")); + break; + + default: + (void) printf(gettext("corrupted data")); + break; + } + } + (void) printf("\n"); + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_CHILDREN, + &child, &children) != 0) + return; + + for (c = 0; c < children; c++) { + uint64_t is_log = B_FALSE; + + (void) nvlist_lookup_uint64(child[c], ZPOOL_CONFIG_IS_LOG, + &is_log); + if (is_log) + continue; + + vname = zpool_vdev_name(g_zfs, NULL, child[c], B_TRUE); + print_import_config(vname, child[c], namewidth, depth + 2); + free(vname); + } + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_L2CACHE, + &child, &children) == 0) { + (void) printf(gettext("\tcache\n")); + for (c = 0; c < children; c++) { + vname = zpool_vdev_name(g_zfs, NULL, child[c], B_FALSE); + (void) printf("\t %s\n", vname); + free(vname); + } + } + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_SPARES, + &child, &children) == 0) { + (void) printf(gettext("\tspares\n")); + for (c = 0; c < children; c++) { + vname = zpool_vdev_name(g_zfs, NULL, child[c], B_FALSE); + (void) printf("\t %s\n", vname); + free(vname); + } + } +} + +/* + * Print log vdevs. + * Logs are recorded as top level vdevs in the main pool child array + * but with "is_log" set to 1. We use either print_status_config() or + * print_import_config() to print the top level logs then any log + * children (eg mirrored slogs) are printed recursively - which + * works because only the top level vdev is marked "is_log" + */ +static void +print_logs(zpool_handle_t *zhp, nvlist_t *nv, int namewidth, boolean_t verbose) +{ + uint_t c, children; + nvlist_t **child; + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_CHILDREN, &child, + &children) != 0) + return; + + (void) printf(gettext("\tlogs\n")); + + for (c = 0; c < children; c++) { + uint64_t is_log = B_FALSE; + char *name; + + (void) nvlist_lookup_uint64(child[c], ZPOOL_CONFIG_IS_LOG, + &is_log); + if (!is_log) + continue; + name = zpool_vdev_name(g_zfs, zhp, child[c], B_TRUE); + if (verbose) + print_status_config(zhp, name, child[c], namewidth, + 2, B_FALSE); + else + print_import_config(name, child[c], namewidth, 2); + free(name); + } +} + +/* + * Display the status for the given pool. + */ +static void +show_import(nvlist_t *config) +{ + uint64_t pool_state; + vdev_stat_t *vs; + char *name; + uint64_t guid; + char *msgid; + nvlist_t *nvroot; + int reason; + const char *health; + uint_t vsc; + int namewidth; + + verify(nvlist_lookup_string(config, ZPOOL_CONFIG_POOL_NAME, + &name) == 0); + verify(nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_GUID, + &guid) == 0); + verify(nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_STATE, + &pool_state) == 0); + verify(nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, + &nvroot) == 0); + + verify(nvlist_lookup_uint64_array(nvroot, ZPOOL_CONFIG_VDEV_STATS, + (uint64_t **)&vs, &vsc) == 0); + health = zpool_state_to_name(vs->vs_state, vs->vs_aux); + + reason = zpool_import_status(config, &msgid); + + (void) printf(gettext(" pool: %s\n"), name); + (void) printf(gettext(" id: %llu\n"), (u_longlong_t)guid); + (void) printf(gettext(" state: %s"), health); + if (pool_state == POOL_STATE_DESTROYED) + (void) printf(gettext(" (DESTROYED)")); + (void) printf("\n"); + + switch (reason) { + case ZPOOL_STATUS_MISSING_DEV_R: + case ZPOOL_STATUS_MISSING_DEV_NR: + case ZPOOL_STATUS_BAD_GUID_SUM: + (void) printf(gettext("status: One or more devices are missing " + "from the system.\n")); + break; + + case ZPOOL_STATUS_CORRUPT_LABEL_R: + case ZPOOL_STATUS_CORRUPT_LABEL_NR: + (void) printf(gettext("status: One or more devices contains " + "corrupted data.\n")); + break; + + case ZPOOL_STATUS_CORRUPT_DATA: + (void) printf(gettext("status: The pool data is corrupted.\n")); + break; + + case ZPOOL_STATUS_OFFLINE_DEV: + (void) printf(gettext("status: One or more devices " + "are offlined.\n")); + break; + + case ZPOOL_STATUS_CORRUPT_POOL: + (void) printf(gettext("status: The pool metadata is " + "corrupted.\n")); + break; + + case ZPOOL_STATUS_VERSION_OLDER: + (void) printf(gettext("status: The pool is formatted using an " + "older on-disk version.\n")); + break; + + case ZPOOL_STATUS_VERSION_NEWER: + (void) printf(gettext("status: The pool is formatted using an " + "incompatible version.\n")); + break; + + case ZPOOL_STATUS_HOSTID_MISMATCH: + (void) printf(gettext("status: The pool was last accessed by " + "another system.\n")); + break; + + case ZPOOL_STATUS_FAULTED_DEV_R: + case ZPOOL_STATUS_FAULTED_DEV_NR: + (void) printf(gettext("status: One or more devices are " + "faulted.\n")); + break; + + case ZPOOL_STATUS_BAD_LOG: + (void) printf(gettext("status: An intent log record cannot be " + "read.\n")); + break; + + case ZPOOL_STATUS_RESILVERING: + (void) printf(gettext("status: One or more devices were being " + "resilvered.\n")); + break; + + default: + /* + * No other status can be seen when importing pools. + */ + assert(reason == ZPOOL_STATUS_OK); + } + + /* + * Print out an action according to the overall state of the pool. + */ + if (vs->vs_state == VDEV_STATE_HEALTHY) { + if (reason == ZPOOL_STATUS_VERSION_OLDER) + (void) printf(gettext("action: The pool can be " + "imported using its name or numeric identifier, " + "though\n\tsome features will not be available " + "without an explicit 'zpool upgrade'.\n")); + else if (reason == ZPOOL_STATUS_HOSTID_MISMATCH) + (void) printf(gettext("action: The pool can be " + "imported using its name or numeric " + "identifier and\n\tthe '-f' flag.\n")); + else + (void) printf(gettext("action: The pool can be " + "imported using its name or numeric " + "identifier.\n")); + } else if (vs->vs_state == VDEV_STATE_DEGRADED) { + (void) printf(gettext("action: The pool can be imported " + "despite missing or damaged devices. The\n\tfault " + "tolerance of the pool may be compromised if imported.\n")); + } else { + switch (reason) { + case ZPOOL_STATUS_VERSION_NEWER: + (void) printf(gettext("action: The pool cannot be " + "imported. Access the pool on a system running " + "newer\n\tsoftware, or recreate the pool from " + "backup.\n")); + break; + case ZPOOL_STATUS_MISSING_DEV_R: + case ZPOOL_STATUS_MISSING_DEV_NR: + case ZPOOL_STATUS_BAD_GUID_SUM: + (void) printf(gettext("action: The pool cannot be " + "imported. Attach the missing\n\tdevices and try " + "again.\n")); + break; + default: + (void) printf(gettext("action: The pool cannot be " + "imported due to damaged devices or data.\n")); + } + } + + /* + * If the state is "closed" or "can't open", and the aux state + * is "corrupt data": + */ + if (((vs->vs_state == VDEV_STATE_CLOSED) || + (vs->vs_state == VDEV_STATE_CANT_OPEN)) && + (vs->vs_aux == VDEV_AUX_CORRUPT_DATA)) { + if (pool_state == POOL_STATE_DESTROYED) + (void) printf(gettext("\tThe pool was destroyed, " + "but can be imported using the '-Df' flags.\n")); + else if (pool_state != POOL_STATE_EXPORTED) + (void) printf(gettext("\tThe pool may be active on " + "another system, but can be imported using\n\t" + "the '-f' flag.\n")); + } + + if (msgid != NULL) + (void) printf(gettext(" see: http://www.sun.com/msg/%s\n"), + msgid); + + (void) printf(gettext("config:\n\n")); + + namewidth = max_width(NULL, nvroot, 0, 0); + if (namewidth < 10) + namewidth = 10; + + print_import_config(name, nvroot, namewidth, 0); + if (num_logs(nvroot) > 0) + print_logs(NULL, nvroot, namewidth, B_FALSE); + + if (reason == ZPOOL_STATUS_BAD_GUID_SUM) { + (void) printf(gettext("\n\tAdditional devices are known to " + "be part of this pool, though their\n\texact " + "configuration cannot be determined.\n")); + } +} + +/* + * Perform the import for the given configuration. This passes the heavy + * lifting off to zpool_import_props(), and then mounts the datasets contained + * within the pool. + */ +static int +do_import(nvlist_t *config, const char *newname, const char *mntopts, + nvlist_t *props, int flags) +{ + zpool_handle_t *zhp; + char *name; + uint64_t state; + uint64_t version; + + verify(nvlist_lookup_string(config, ZPOOL_CONFIG_POOL_NAME, + &name) == 0); + + verify(nvlist_lookup_uint64(config, + ZPOOL_CONFIG_POOL_STATE, &state) == 0); + verify(nvlist_lookup_uint64(config, + ZPOOL_CONFIG_VERSION, &version) == 0); + if (version > SPA_VERSION) { + (void) fprintf(stderr, gettext("cannot import '%s': pool " + "is formatted using a newer ZFS version\n"), name); + return (1); + } else if (state != POOL_STATE_EXPORTED && + !(flags & ZFS_IMPORT_ANY_HOST)) { + uint64_t hostid; + + if (nvlist_lookup_uint64(config, ZPOOL_CONFIG_HOSTID, + &hostid) == 0) { + if ((unsigned long)hostid != gethostid()) { + char *hostname; + uint64_t timestamp; + time_t t; + + verify(nvlist_lookup_string(config, + ZPOOL_CONFIG_HOSTNAME, &hostname) == 0); + verify(nvlist_lookup_uint64(config, + ZPOOL_CONFIG_TIMESTAMP, ×tamp) == 0); + t = timestamp; + (void) fprintf(stderr, gettext("cannot import " + "'%s': pool may be in use from other " + "system, it was last accessed by %s " + "(hostid: 0x%lx) on %s"), name, hostname, + (unsigned long)hostid, + asctime(localtime(&t))); + (void) fprintf(stderr, gettext("use '-f' to " + "import anyway\n")); + return (1); + } + } else { + (void) fprintf(stderr, gettext("cannot import '%s': " + "pool may be in use from other system\n"), name); + (void) fprintf(stderr, gettext("use '-f' to import " + "anyway\n")); + return (1); + } + } + + if (zpool_import_props(g_zfs, config, newname, props, flags) != 0) + return (1); + + if (newname != NULL) + name = (char *)newname; + + if ((zhp = zpool_open_canfail(g_zfs, name)) == NULL) + return (1); + + if (zpool_get_state(zhp) != POOL_STATE_UNAVAIL && + !(flags & ZFS_IMPORT_ONLY) && + zpool_enable_datasets(zhp, mntopts, 0) != 0) { + zpool_close(zhp); + return (1); + } + + zpool_close(zhp); + return (0); +} + +/* + * zpool import [-d dir] [-D] + * import [-o mntopts] [-o prop=value] ... [-R root] [-D] + * [-d dir | -c cachefile] [-f] -a + * import [-o mntopts] [-o prop=value] ... [-R root] [-D] + * [-d dir | -c cachefile] [-f] [-n] [-F] <pool | id> [newpool] + * + * -c Read pool information from a cachefile instead of searching + * devices. + * + * -d Scan in a specific directory, other than /dev/dsk. More than + * one directory can be specified using multiple '-d' options. + * + * -D Scan for previously destroyed pools or import all or only + * specified destroyed pools. + * + * -R Temporarily import the pool, with all mountpoints relative to + * the given root. The pool will remain exported when the machine + * is rebooted. + * + * -V Import even in the presence of faulted vdevs. This is an + * intentionally undocumented option for testing purposes, and + * treats the pool configuration as complete, leaving any bad + * vdevs in the FAULTED state. In other words, it does verbatim + * import. + * + * -f Force import, even if it appears that the pool is active. + * + * -F Attempt rewind if necessary. + * + * -n See if rewind would work, but don't actually rewind. + * + * -N Import the pool but don't mount datasets. + * + * -T Specify a starting txg to use for import. This option is + * intentionally undocumented option for testing purposes. + * + * -a Import all pools found. + * + * -o Set property=value and/or temporary mount options (without '='). + * + * The import command scans for pools to import, and import pools based on pool + * name and GUID. The pool can also be renamed as part of the import process. + */ +int +zpool_do_import(int argc, char **argv) +{ + char **searchdirs = NULL; + int nsearch = 0; + int c; + int err = 0; + nvlist_t *pools = NULL; + boolean_t do_all = B_FALSE; + boolean_t do_destroyed = B_FALSE; + char *mntopts = NULL; + nvpair_t *elem; + nvlist_t *config; + uint64_t searchguid = 0; + char *searchname = NULL; + char *propval; + nvlist_t *found_config; + nvlist_t *policy = NULL; + nvlist_t *props = NULL; + boolean_t first; + int flags = ZFS_IMPORT_NORMAL; + uint32_t rewind_policy = ZPOOL_NO_REWIND; + boolean_t dryrun = B_FALSE; + boolean_t do_rewind = B_FALSE; + boolean_t xtreme_rewind = B_FALSE; + uint64_t pool_state, txg = -1ULL; + char *cachefile = NULL; + importargs_t idata = { 0 }; + char *endptr; + + /* check options */ + while ((c = getopt(argc, argv, ":aCc:d:DEfFmnNo:rR:T:VX")) != -1) { + switch (c) { + case 'a': + do_all = B_TRUE; + break; + case 'c': + cachefile = optarg; + break; + case 'd': + if (searchdirs == NULL) { + searchdirs = safe_malloc(sizeof (char *)); + } else { + char **tmp = safe_malloc((nsearch + 1) * + sizeof (char *)); + bcopy(searchdirs, tmp, nsearch * + sizeof (char *)); + free(searchdirs); + searchdirs = tmp; + } + searchdirs[nsearch++] = optarg; + break; + case 'D': + do_destroyed = B_TRUE; + break; + case 'f': + flags |= ZFS_IMPORT_ANY_HOST; + break; + case 'F': + do_rewind = B_TRUE; + break; + case 'm': + flags |= ZFS_IMPORT_MISSING_LOG; + break; + case 'n': + dryrun = B_TRUE; + break; + case 'N': + flags |= ZFS_IMPORT_ONLY; + break; + case 'o': + if ((propval = strchr(optarg, '=')) != NULL) { + *propval = '\0'; + propval++; + if (add_prop_list(optarg, propval, + &props, B_TRUE)) + goto error; + } else { + mntopts = optarg; + } + break; + case 'R': + if (add_prop_list(zpool_prop_to_name( + ZPOOL_PROP_ALTROOT), optarg, &props, B_TRUE)) + goto error; + if (nvlist_lookup_string(props, + zpool_prop_to_name(ZPOOL_PROP_CACHEFILE), + &propval) == 0) + break; + if (add_prop_list(zpool_prop_to_name( + ZPOOL_PROP_CACHEFILE), "none", &props, B_TRUE)) + goto error; + break; + case 'T': + errno = 0; + txg = strtoull(optarg, &endptr, 10); + if (errno != 0 || *endptr != '\0') { + (void) fprintf(stderr, + gettext("invalid txg value\n")); + usage(B_FALSE); + } + rewind_policy = ZPOOL_DO_REWIND | ZPOOL_EXTREME_REWIND; + break; + case 'V': + flags |= ZFS_IMPORT_VERBATIM; + break; + case 'X': + xtreme_rewind = B_TRUE; + break; + case ':': + (void) fprintf(stderr, gettext("missing argument for " + "'%c' option\n"), optopt); + usage(B_FALSE); + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + if (cachefile && nsearch != 0) { + (void) fprintf(stderr, gettext("-c is incompatible with -d\n")); + usage(B_FALSE); + } + + if ((dryrun || xtreme_rewind) && !do_rewind) { + (void) fprintf(stderr, + gettext("-n or -X only meaningful with -F\n")); + usage(B_FALSE); + } + if (dryrun) + rewind_policy = ZPOOL_TRY_REWIND; + else if (do_rewind) + rewind_policy = ZPOOL_DO_REWIND; + if (xtreme_rewind) + rewind_policy |= ZPOOL_EXTREME_REWIND; + + /* In the future, we can capture further policy and include it here */ + if (nvlist_alloc(&policy, NV_UNIQUE_NAME, 0) != 0 || + nvlist_add_uint64(policy, ZPOOL_REWIND_REQUEST_TXG, txg) != 0 || + nvlist_add_uint32(policy, ZPOOL_REWIND_REQUEST, rewind_policy) != 0) + goto error; + + if (searchdirs == NULL) { + searchdirs = safe_malloc(sizeof (char *)); + searchdirs[0] = "/dev/dsk"; + nsearch = 1; + } + + /* check argument count */ + if (do_all) { + if (argc != 0) { + (void) fprintf(stderr, gettext("too many arguments\n")); + usage(B_FALSE); + } + } else { + if (argc > 2) { + (void) fprintf(stderr, gettext("too many arguments\n")); + usage(B_FALSE); + } + + /* + * Check for the SYS_CONFIG privilege. We do this explicitly + * here because otherwise any attempt to discover pools will + * silently fail. + */ + if (argc == 0 && !priv_ineffect(PRIV_SYS_CONFIG)) { + (void) fprintf(stderr, gettext("cannot " + "discover pools: permission denied\n")); + free(searchdirs); + nvlist_free(policy); + return (1); + } + } + + /* + * Depending on the arguments given, we do one of the following: + * + * <none> Iterate through all pools and display information about + * each one. + * + * -a Iterate through all pools and try to import each one. + * + * <id> Find the pool that corresponds to the given GUID/pool + * name and import that one. + * + * -D Above options applies only to destroyed pools. + */ + if (argc != 0) { + char *endptr; + + errno = 0; + searchguid = strtoull(argv[0], &endptr, 10); + if (errno != 0 || *endptr != '\0') + searchname = argv[0]; + found_config = NULL; + + /* + * User specified a name or guid. Ensure it's unique. + */ + idata.unique = B_TRUE; + } + + + idata.path = searchdirs; + idata.paths = nsearch; + idata.poolname = searchname; + idata.guid = searchguid; + idata.cachefile = cachefile; + + pools = zpool_search_import(g_zfs, &idata); + + if (pools != NULL && idata.exists && + (argc == 1 || strcmp(argv[0], argv[1]) == 0)) { + (void) fprintf(stderr, gettext("cannot import '%s': " + "a pool with that name already exists\n"), + argv[0]); + (void) fprintf(stderr, gettext("use the form '%s " + "<pool | id> <newpool>' to give it a new name\n"), + "zpool import"); + err = 1; + } else if (pools == NULL && idata.exists) { + (void) fprintf(stderr, gettext("cannot import '%s': " + "a pool with that name is already created/imported,\n"), + argv[0]); + (void) fprintf(stderr, gettext("and no additional pools " + "with that name were found\n")); + err = 1; + } else if (pools == NULL) { + if (argc != 0) { + (void) fprintf(stderr, gettext("cannot import '%s': " + "no such pool available\n"), argv[0]); + } + err = 1; + } + + if (err == 1) { + free(searchdirs); + nvlist_free(policy); + return (1); + } + + /* + * At this point we have a list of import candidate configs. Even if + * we were searching by pool name or guid, we still need to + * post-process the list to deal with pool state and possible + * duplicate names. + */ + err = 0; + elem = NULL; + first = B_TRUE; + while ((elem = nvlist_next_nvpair(pools, elem)) != NULL) { + + verify(nvpair_value_nvlist(elem, &config) == 0); + + verify(nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_STATE, + &pool_state) == 0); + if (!do_destroyed && pool_state == POOL_STATE_DESTROYED) + continue; + if (do_destroyed && pool_state != POOL_STATE_DESTROYED) + continue; + + verify(nvlist_add_nvlist(config, ZPOOL_REWIND_POLICY, + policy) == 0); + + if (argc == 0) { + if (first) + first = B_FALSE; + else if (!do_all) + (void) printf("\n"); + + if (do_all) { + err |= do_import(config, NULL, mntopts, + props, flags); + } else { + show_import(config); + } + } else if (searchname != NULL) { + char *name; + + /* + * We are searching for a pool based on name. + */ + verify(nvlist_lookup_string(config, + ZPOOL_CONFIG_POOL_NAME, &name) == 0); + + if (strcmp(name, searchname) == 0) { + if (found_config != NULL) { + (void) fprintf(stderr, gettext( + "cannot import '%s': more than " + "one matching pool\n"), searchname); + (void) fprintf(stderr, gettext( + "import by numeric ID instead\n")); + err = B_TRUE; + } + found_config = config; + } + } else { + uint64_t guid; + + /* + * Search for a pool by guid. + */ + verify(nvlist_lookup_uint64(config, + ZPOOL_CONFIG_POOL_GUID, &guid) == 0); + + if (guid == searchguid) + found_config = config; + } + } + + /* + * If we were searching for a specific pool, verify that we found a + * pool, and then do the import. + */ + if (argc != 0 && err == 0) { + if (found_config == NULL) { + (void) fprintf(stderr, gettext("cannot import '%s': " + "no such pool available\n"), argv[0]); + err = B_TRUE; + } else { + err |= do_import(found_config, argc == 1 ? NULL : + argv[1], mntopts, props, flags); + } + } + + /* + * If we were just looking for pools, report an error if none were + * found. + */ + if (argc == 0 && first) + (void) fprintf(stderr, + gettext("no pools available to import\n")); + +error: + nvlist_free(props); + nvlist_free(pools); + nvlist_free(policy); + free(searchdirs); + + return (err ? 1 : 0); +} + +typedef struct iostat_cbdata { + zpool_list_t *cb_list; + int cb_verbose; + int cb_iteration; + int cb_namewidth; +} iostat_cbdata_t; + +static void +print_iostat_separator(iostat_cbdata_t *cb) +{ + int i = 0; + + for (i = 0; i < cb->cb_namewidth; i++) + (void) printf("-"); + (void) printf(" ----- ----- ----- ----- ----- -----\n"); +} + +static void +print_iostat_header(iostat_cbdata_t *cb) +{ + (void) printf("%*s capacity operations bandwidth\n", + cb->cb_namewidth, ""); + (void) printf("%-*s alloc free read write read write\n", + cb->cb_namewidth, "pool"); + print_iostat_separator(cb); +} + +/* + * Display a single statistic. + */ +static void +print_one_stat(uint64_t value) +{ + char buf[64]; + + zfs_nicenum(value, buf, sizeof (buf)); + (void) printf(" %5s", buf); +} + +/* + * Print out all the statistics for the given vdev. This can either be the + * toplevel configuration, or called recursively. If 'name' is NULL, then this + * is a verbose output, and we don't want to display the toplevel pool stats. + */ +void +print_vdev_stats(zpool_handle_t *zhp, const char *name, nvlist_t *oldnv, + nvlist_t *newnv, iostat_cbdata_t *cb, int depth) +{ + nvlist_t **oldchild, **newchild; + uint_t c, children; + vdev_stat_t *oldvs, *newvs; + vdev_stat_t zerovs = { 0 }; + uint64_t tdelta; + double scale; + char *vname; + + if (oldnv != NULL) { + verify(nvlist_lookup_uint64_array(oldnv, + ZPOOL_CONFIG_VDEV_STATS, (uint64_t **)&oldvs, &c) == 0); + } else { + oldvs = &zerovs; + } + + verify(nvlist_lookup_uint64_array(newnv, ZPOOL_CONFIG_VDEV_STATS, + (uint64_t **)&newvs, &c) == 0); + + if (strlen(name) + depth > cb->cb_namewidth) + (void) printf("%*s%s", depth, "", name); + else + (void) printf("%*s%s%*s", depth, "", name, + (int)(cb->cb_namewidth - strlen(name) - depth), ""); + + tdelta = newvs->vs_timestamp - oldvs->vs_timestamp; + + if (tdelta == 0) + scale = 1.0; + else + scale = (double)NANOSEC / tdelta; + + /* only toplevel vdevs have capacity stats */ + if (newvs->vs_space == 0) { + (void) printf(" - -"); + } else { + print_one_stat(newvs->vs_alloc); + print_one_stat(newvs->vs_space - newvs->vs_alloc); + } + + print_one_stat((uint64_t)(scale * (newvs->vs_ops[ZIO_TYPE_READ] - + oldvs->vs_ops[ZIO_TYPE_READ]))); + + print_one_stat((uint64_t)(scale * (newvs->vs_ops[ZIO_TYPE_WRITE] - + oldvs->vs_ops[ZIO_TYPE_WRITE]))); + + print_one_stat((uint64_t)(scale * (newvs->vs_bytes[ZIO_TYPE_READ] - + oldvs->vs_bytes[ZIO_TYPE_READ]))); + + print_one_stat((uint64_t)(scale * (newvs->vs_bytes[ZIO_TYPE_WRITE] - + oldvs->vs_bytes[ZIO_TYPE_WRITE]))); + + (void) printf("\n"); + + if (!cb->cb_verbose) + return; + + if (nvlist_lookup_nvlist_array(newnv, ZPOOL_CONFIG_CHILDREN, + &newchild, &children) != 0) + return; + + if (oldnv && nvlist_lookup_nvlist_array(oldnv, ZPOOL_CONFIG_CHILDREN, + &oldchild, &c) != 0) + return; + + for (c = 0; c < children; c++) { + uint64_t ishole = B_FALSE; + + if (nvlist_lookup_uint64(newchild[c], + ZPOOL_CONFIG_IS_HOLE, &ishole) == 0 && ishole) + continue; + + vname = zpool_vdev_name(g_zfs, zhp, newchild[c], B_FALSE); + print_vdev_stats(zhp, vname, oldnv ? oldchild[c] : NULL, + newchild[c], cb, depth + 2); + free(vname); + } + + /* + * Include level 2 ARC devices in iostat output + */ + if (nvlist_lookup_nvlist_array(newnv, ZPOOL_CONFIG_L2CACHE, + &newchild, &children) != 0) + return; + + if (oldnv && nvlist_lookup_nvlist_array(oldnv, ZPOOL_CONFIG_L2CACHE, + &oldchild, &c) != 0) + return; + + if (children > 0) { + (void) printf("%-*s - - - - - " + "-\n", cb->cb_namewidth, "cache"); + for (c = 0; c < children; c++) { + vname = zpool_vdev_name(g_zfs, zhp, newchild[c], + B_FALSE); + print_vdev_stats(zhp, vname, oldnv ? oldchild[c] : NULL, + newchild[c], cb, depth + 2); + free(vname); + } + } +} + +static int +refresh_iostat(zpool_handle_t *zhp, void *data) +{ + iostat_cbdata_t *cb = data; + boolean_t missing; + + /* + * If the pool has disappeared, remove it from the list and continue. + */ + if (zpool_refresh_stats(zhp, &missing) != 0) + return (-1); + + if (missing) + pool_list_remove(cb->cb_list, zhp); + + return (0); +} + +/* + * Callback to print out the iostats for the given pool. + */ +int +print_iostat(zpool_handle_t *zhp, void *data) +{ + iostat_cbdata_t *cb = data; + nvlist_t *oldconfig, *newconfig; + nvlist_t *oldnvroot, *newnvroot; + + newconfig = zpool_get_config(zhp, &oldconfig); + + if (cb->cb_iteration == 1) + oldconfig = NULL; + + verify(nvlist_lookup_nvlist(newconfig, ZPOOL_CONFIG_VDEV_TREE, + &newnvroot) == 0); + + if (oldconfig == NULL) + oldnvroot = NULL; + else + verify(nvlist_lookup_nvlist(oldconfig, ZPOOL_CONFIG_VDEV_TREE, + &oldnvroot) == 0); + + /* + * Print out the statistics for the pool. + */ + print_vdev_stats(zhp, zpool_get_name(zhp), oldnvroot, newnvroot, cb, 0); + + if (cb->cb_verbose) + print_iostat_separator(cb); + + return (0); +} + +int +get_namewidth(zpool_handle_t *zhp, void *data) +{ + iostat_cbdata_t *cb = data; + nvlist_t *config, *nvroot; + + if ((config = zpool_get_config(zhp, NULL)) != NULL) { + verify(nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, + &nvroot) == 0); + if (!cb->cb_verbose) + cb->cb_namewidth = strlen(zpool_get_name(zhp)); + else + cb->cb_namewidth = max_width(zhp, nvroot, 0, 0); + } + + /* + * The width must fall into the range [10,38]. The upper limit is the + * maximum we can have and still fit in 80 columns. + */ + if (cb->cb_namewidth < 10) + cb->cb_namewidth = 10; + if (cb->cb_namewidth > 38) + cb->cb_namewidth = 38; + + return (0); +} + +/* + * Parse the input string, get the 'interval' and 'count' value if there is one. + */ +static void +get_interval_count(int *argcp, char **argv, unsigned long *iv, + unsigned long *cnt) +{ + unsigned long interval = 0, count = 0; + int argc = *argcp, errno; + + /* + * Determine if the last argument is an integer or a pool name + */ + if (argc > 0 && isdigit(argv[argc - 1][0])) { + char *end; + + errno = 0; + interval = strtoul(argv[argc - 1], &end, 10); + + if (*end == '\0' && errno == 0) { + if (interval == 0) { + (void) fprintf(stderr, gettext("interval " + "cannot be zero\n")); + usage(B_FALSE); + } + /* + * Ignore the last parameter + */ + argc--; + } else { + /* + * If this is not a valid number, just plow on. The + * user will get a more informative error message later + * on. + */ + interval = 0; + } + } + + /* + * If the last argument is also an integer, then we have both a count + * and an interval. + */ + if (argc > 0 && isdigit(argv[argc - 1][0])) { + char *end; + + errno = 0; + count = interval; + interval = strtoul(argv[argc - 1], &end, 10); + + if (*end == '\0' && errno == 0) { + if (interval == 0) { + (void) fprintf(stderr, gettext("interval " + "cannot be zero\n")); + usage(B_FALSE); + } + + /* + * Ignore the last parameter + */ + argc--; + } else { + interval = 0; + } + } + + *iv = interval; + *cnt = count; + *argcp = argc; +} + +static void +get_timestamp_arg(char c) +{ + if (c == 'u') + timestamp_fmt = UDATE; + else if (c == 'd') + timestamp_fmt = DDATE; + else + usage(B_FALSE); +} + +/* + * zpool iostat [-v] [-T d|u] [pool] ... [interval [count]] + * + * -v Display statistics for individual vdevs + * -T Display a timestamp in date(1) or Unix format + * + * This command can be tricky because we want to be able to deal with pool + * creation/destruction as well as vdev configuration changes. The bulk of this + * processing is handled by the pool_list_* routines in zpool_iter.c. We rely + * on pool_list_update() to detect the addition of new pools. Configuration + * changes are all handled within libzfs. + */ +int +zpool_do_iostat(int argc, char **argv) +{ + int c; + int ret; + int npools; + unsigned long interval = 0, count = 0; + zpool_list_t *list; + boolean_t verbose = B_FALSE; + iostat_cbdata_t cb; + + /* check options */ + while ((c = getopt(argc, argv, "T:v")) != -1) { + switch (c) { + case 'T': + get_timestamp_arg(*optarg); + break; + case 'v': + verbose = B_TRUE; + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + get_interval_count(&argc, argv, &interval, &count); + + /* + * Construct the list of all interesting pools. + */ + ret = 0; + if ((list = pool_list_get(argc, argv, NULL, &ret)) == NULL) + return (1); + + if (pool_list_count(list) == 0 && argc != 0) { + pool_list_free(list); + return (1); + } + + if (pool_list_count(list) == 0 && interval == 0) { + pool_list_free(list); + (void) fprintf(stderr, gettext("no pools available\n")); + return (1); + } + + /* + * Enter the main iostat loop. + */ + cb.cb_list = list; + cb.cb_verbose = verbose; + cb.cb_iteration = 0; + cb.cb_namewidth = 0; + + for (;;) { + pool_list_update(list); + + if ((npools = pool_list_count(list)) == 0) + break; + + /* + * Refresh all statistics. This is done as an explicit step + * before calculating the maximum name width, so that any + * configuration changes are properly accounted for. + */ + (void) pool_list_iter(list, B_FALSE, refresh_iostat, &cb); + + /* + * Iterate over all pools to determine the maximum width + * for the pool / device name column across all pools. + */ + cb.cb_namewidth = 0; + (void) pool_list_iter(list, B_FALSE, get_namewidth, &cb); + + if (timestamp_fmt != NODATE) + print_timestamp(timestamp_fmt); + + /* + * If it's the first time, or verbose mode, print the header. + */ + if (++cb.cb_iteration == 1 || verbose) + print_iostat_header(&cb); + + (void) pool_list_iter(list, B_FALSE, print_iostat, &cb); + + /* + * If there's more than one pool, and we're not in verbose mode + * (which prints a separator for us), then print a separator. + */ + if (npools > 1 && !verbose) + print_iostat_separator(&cb); + + if (verbose) + (void) printf("\n"); + + /* + * Flush the output so that redirection to a file isn't buffered + * indefinitely. + */ + (void) fflush(stdout); + + if (interval == 0) + break; + + if (count != 0 && --count == 0) + break; + + (void) sleep(interval); + } + + pool_list_free(list); + + return (ret); +} + +typedef struct list_cbdata { + boolean_t cb_scripted; + boolean_t cb_first; + zprop_list_t *cb_proplist; +} list_cbdata_t; + +/* + * Given a list of columns to display, output appropriate headers for each one. + */ +static void +print_header(zprop_list_t *pl) +{ + const char *header; + boolean_t first = B_TRUE; + boolean_t right_justify; + + for (; pl != NULL; pl = pl->pl_next) { + if (pl->pl_prop == ZPROP_INVAL) + continue; + + if (!first) + (void) printf(" "); + else + first = B_FALSE; + + header = zpool_prop_column_name(pl->pl_prop); + right_justify = zpool_prop_align_right(pl->pl_prop); + + if (pl->pl_next == NULL && !right_justify) + (void) printf("%s", header); + else if (right_justify) + (void) printf("%*s", pl->pl_width, header); + else + (void) printf("%-*s", pl->pl_width, header); + } + + (void) printf("\n"); +} + +/* + * Given a pool and a list of properties, print out all the properties according + * to the described layout. + */ +static void +print_pool(zpool_handle_t *zhp, zprop_list_t *pl, int scripted) +{ + boolean_t first = B_TRUE; + char property[ZPOOL_MAXPROPLEN]; + char *propstr; + boolean_t right_justify; + int width; + + for (; pl != NULL; pl = pl->pl_next) { + if (!first) { + if (scripted) + (void) printf("\t"); + else + (void) printf(" "); + } else { + first = B_FALSE; + } + + right_justify = B_FALSE; + if (pl->pl_prop != ZPROP_INVAL) { + if (zpool_get_prop(zhp, pl->pl_prop, property, + sizeof (property), NULL) != 0) + propstr = "-"; + else + propstr = property; + + right_justify = zpool_prop_align_right(pl->pl_prop); + } else { + propstr = "-"; + } + + width = pl->pl_width; + + /* + * If this is being called in scripted mode, or if this is the + * last column and it is left-justified, don't include a width + * format specifier. + */ + if (scripted || (pl->pl_next == NULL && !right_justify)) + (void) printf("%s", propstr); + else if (right_justify) + (void) printf("%*s", width, propstr); + else + (void) printf("%-*s", width, propstr); + } + + (void) printf("\n"); +} + +/* + * Generic callback function to list a pool. + */ +int +list_callback(zpool_handle_t *zhp, void *data) +{ + list_cbdata_t *cbp = data; + + if (cbp->cb_first) { + if (!cbp->cb_scripted) + print_header(cbp->cb_proplist); + cbp->cb_first = B_FALSE; + } + + print_pool(zhp, cbp->cb_proplist, cbp->cb_scripted); + + return (0); +} + +/* + * zpool list [-H] [-o prop[,prop]*] [-T d|u] [pool] ... [interval [count]] + * + * -H Scripted mode. Don't display headers, and separate properties + * by a single tab. + * -o List of properties to display. Defaults to + * "name,size,allocated,free,capacity,health,altroot" + * -T Display a timestamp in date(1) or Unix format + * + * List all pools in the system, whether or not they're healthy. Output space + * statistics for each one, as well as health status summary. + */ +int +zpool_do_list(int argc, char **argv) +{ + int c; + int ret; + list_cbdata_t cb = { 0 }; + static char default_props[] = + "name,size,allocated,free,capacity,dedupratio,health,altroot"; + char *props = default_props; + unsigned long interval = 0, count = 0; + + /* check options */ + while ((c = getopt(argc, argv, ":Ho:T:")) != -1) { + switch (c) { + case 'H': + cb.cb_scripted = B_TRUE; + break; + case 'o': + props = optarg; + break; + case 'T': + get_timestamp_arg(*optarg); + break; + case ':': + (void) fprintf(stderr, gettext("missing argument for " + "'%c' option\n"), optopt); + usage(B_FALSE); + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + get_interval_count(&argc, argv, &interval, &count); + + if (zprop_get_list(g_zfs, props, &cb.cb_proplist, ZFS_TYPE_POOL) != 0) + usage(B_FALSE); + + cb.cb_first = B_TRUE; + + for (;;) { + + if (timestamp_fmt != NODATE) + print_timestamp(timestamp_fmt); + + ret = for_each_pool(argc, argv, B_TRUE, &cb.cb_proplist, + list_callback, &cb); + + if (argc == 0 && cb.cb_first && !cb.cb_scripted) { + (void) printf(gettext("no pools available\n")); + zprop_free_list(cb.cb_proplist); + return (0); + } + + if (interval == 0) + break; + + if (count != 0 && --count == 0) + break; + + (void) sleep(interval); + } + + zprop_free_list(cb.cb_proplist); + return (ret); +} + +static nvlist_t * +zpool_get_vdev_by_name(nvlist_t *nv, char *name) +{ + nvlist_t **child; + uint_t c, children; + nvlist_t *match; + char *path; + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_CHILDREN, + &child, &children) != 0) { + verify(nvlist_lookup_string(nv, ZPOOL_CONFIG_PATH, &path) == 0); + if (strncmp(name, "/dev/dsk/", 9) == 0) + name += 9; + if (strncmp(path, "/dev/dsk/", 9) == 0) + path += 9; + if (strcmp(name, path) == 0) + return (nv); + return (NULL); + } + + for (c = 0; c < children; c++) + if ((match = zpool_get_vdev_by_name(child[c], name)) != NULL) + return (match); + + return (NULL); +} + +static int +zpool_do_attach_or_replace(int argc, char **argv, int replacing) +{ + boolean_t force = B_FALSE; + int c; + nvlist_t *nvroot; + char *poolname, *old_disk, *new_disk; + zpool_handle_t *zhp; + int ret; + + /* check options */ + while ((c = getopt(argc, argv, "f")) != -1) { + switch (c) { + case 'f': + force = B_TRUE; + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + /* get pool name and check number of arguments */ + if (argc < 1) { + (void) fprintf(stderr, gettext("missing pool name argument\n")); + usage(B_FALSE); + } + + poolname = argv[0]; + + if (argc < 2) { + (void) fprintf(stderr, + gettext("missing <device> specification\n")); + usage(B_FALSE); + } + + old_disk = argv[1]; + + if (argc < 3) { + if (!replacing) { + (void) fprintf(stderr, + gettext("missing <new_device> specification\n")); + usage(B_FALSE); + } + new_disk = old_disk; + argc -= 1; + argv += 1; + } else { + new_disk = argv[2]; + argc -= 2; + argv += 2; + } + + if (argc > 1) { + (void) fprintf(stderr, gettext("too many arguments\n")); + usage(B_FALSE); + } + + if ((zhp = zpool_open(g_zfs, poolname)) == NULL) + return (1); + + if (zpool_get_config(zhp, NULL) == NULL) { + (void) fprintf(stderr, gettext("pool '%s' is unavailable\n"), + poolname); + zpool_close(zhp); + return (1); + } + + nvroot = make_root_vdev(zhp, force, B_FALSE, replacing, B_FALSE, + argc, argv); + if (nvroot == NULL) { + zpool_close(zhp); + return (1); + } + + ret = zpool_vdev_attach(zhp, old_disk, new_disk, nvroot, replacing); + + nvlist_free(nvroot); + zpool_close(zhp); + + return (ret); +} + +/* + * zpool replace [-f] <pool> <device> <new_device> + * + * -f Force attach, even if <new_device> appears to be in use. + * + * Replace <device> with <new_device>. + */ +/* ARGSUSED */ +int +zpool_do_replace(int argc, char **argv) +{ + return (zpool_do_attach_or_replace(argc, argv, B_TRUE)); +} + +/* + * zpool attach [-f] <pool> <device> <new_device> + * + * -f Force attach, even if <new_device> appears to be in use. + * + * Attach <new_device> to the mirror containing <device>. If <device> is not + * part of a mirror, then <device> will be transformed into a mirror of + * <device> and <new_device>. In either case, <new_device> will begin life + * with a DTL of [0, now], and will immediately begin to resilver itself. + */ +int +zpool_do_attach(int argc, char **argv) +{ + return (zpool_do_attach_or_replace(argc, argv, B_FALSE)); +} + +/* + * zpool detach [-f] <pool> <device> + * + * -f Force detach of <device>, even if DTLs argue against it + * (not supported yet) + * + * Detach a device from a mirror. The operation will be refused if <device> + * is the last device in the mirror, or if the DTLs indicate that this device + * has the only valid copy of some data. + */ +/* ARGSUSED */ +int +zpool_do_detach(int argc, char **argv) +{ + int c; + char *poolname, *path; + zpool_handle_t *zhp; + int ret; + + /* check options */ + while ((c = getopt(argc, argv, "f")) != -1) { + switch (c) { + case 'f': + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + /* get pool name and check number of arguments */ + if (argc < 1) { + (void) fprintf(stderr, gettext("missing pool name argument\n")); + usage(B_FALSE); + } + + if (argc < 2) { + (void) fprintf(stderr, + gettext("missing <device> specification\n")); + usage(B_FALSE); + } + + poolname = argv[0]; + path = argv[1]; + + if ((zhp = zpool_open(g_zfs, poolname)) == NULL) + return (1); + + ret = zpool_vdev_detach(zhp, path); + + zpool_close(zhp); + + return (ret); +} + +/* + * zpool split [-n] [-o prop=val] ... + * [-o mntopt] ... + * [-R altroot] <pool> <newpool> [<device> ...] + * + * -n Do not split the pool, but display the resulting layout if + * it were to be split. + * -o Set property=value, or set mount options. + * -R Mount the split-off pool under an alternate root. + * + * Splits the named pool and gives it the new pool name. Devices to be split + * off may be listed, provided that no more than one device is specified + * per top-level vdev mirror. The newly split pool is left in an exported + * state unless -R is specified. + * + * Restrictions: the top-level of the pool pool must only be made up of + * mirrors; all devices in the pool must be healthy; no device may be + * undergoing a resilvering operation. + */ +int +zpool_do_split(int argc, char **argv) +{ + char *srcpool, *newpool, *propval; + char *mntopts = NULL; + splitflags_t flags; + int c, ret = 0; + zpool_handle_t *zhp; + nvlist_t *config, *props = NULL; + + flags.dryrun = B_FALSE; + flags.import = B_FALSE; + + /* check options */ + while ((c = getopt(argc, argv, ":R:no:")) != -1) { + switch (c) { + case 'R': + flags.import = B_TRUE; + if (add_prop_list( + zpool_prop_to_name(ZPOOL_PROP_ALTROOT), optarg, + &props, B_TRUE) != 0) { + if (props) + nvlist_free(props); + usage(B_FALSE); + } + break; + case 'n': + flags.dryrun = B_TRUE; + break; + case 'o': + if ((propval = strchr(optarg, '=')) != NULL) { + *propval = '\0'; + propval++; + if (add_prop_list(optarg, propval, + &props, B_TRUE) != 0) { + if (props) + nvlist_free(props); + usage(B_FALSE); + } + } else { + mntopts = optarg; + } + break; + case ':': + (void) fprintf(stderr, gettext("missing argument for " + "'%c' option\n"), optopt); + usage(B_FALSE); + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + break; + } + } + + if (!flags.import && mntopts != NULL) { + (void) fprintf(stderr, gettext("setting mntopts is only " + "valid when importing the pool\n")); + usage(B_FALSE); + } + + argc -= optind; + argv += optind; + + if (argc < 1) { + (void) fprintf(stderr, gettext("Missing pool name\n")); + usage(B_FALSE); + } + if (argc < 2) { + (void) fprintf(stderr, gettext("Missing new pool name\n")); + usage(B_FALSE); + } + + srcpool = argv[0]; + newpool = argv[1]; + + argc -= 2; + argv += 2; + + if ((zhp = zpool_open(g_zfs, srcpool)) == NULL) + return (1); + + config = split_mirror_vdev(zhp, newpool, props, flags, argc, argv); + if (config == NULL) { + ret = 1; + } else { + if (flags.dryrun) { + (void) printf(gettext("would create '%s' with the " + "following layout:\n\n"), newpool); + print_vdev_tree(NULL, newpool, config, 0, B_FALSE); + } + nvlist_free(config); + } + + zpool_close(zhp); + + if (ret != 0 || flags.dryrun || !flags.import) + return (ret); + + /* + * The split was successful. Now we need to open the new + * pool and import it. + */ + if ((zhp = zpool_open_canfail(g_zfs, newpool)) == NULL) + return (1); + if (zpool_get_state(zhp) != POOL_STATE_UNAVAIL && + zpool_enable_datasets(zhp, mntopts, 0) != 0) { + ret = 1; + (void) fprintf(stderr, gettext("Split was succssful, but " + "the datasets could not all be mounted\n")); + (void) fprintf(stderr, gettext("Try doing '%s' with a " + "different altroot\n"), "zpool import"); + } + zpool_close(zhp); + + return (ret); +} + + + +/* + * zpool online <pool> <device> ... + */ +int +zpool_do_online(int argc, char **argv) +{ + int c, i; + char *poolname; + zpool_handle_t *zhp; + int ret = 0; + vdev_state_t newstate; + int flags = 0; + + /* check options */ + while ((c = getopt(argc, argv, "et")) != -1) { + switch (c) { + case 'e': + flags |= ZFS_ONLINE_EXPAND; + break; + case 't': + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + /* get pool name and check number of arguments */ + if (argc < 1) { + (void) fprintf(stderr, gettext("missing pool name\n")); + usage(B_FALSE); + } + if (argc < 2) { + (void) fprintf(stderr, gettext("missing device name\n")); + usage(B_FALSE); + } + + poolname = argv[0]; + + if ((zhp = zpool_open(g_zfs, poolname)) == NULL) + return (1); + + for (i = 1; i < argc; i++) { + if (zpool_vdev_online(zhp, argv[i], flags, &newstate) == 0) { + if (newstate != VDEV_STATE_HEALTHY) { + (void) printf(gettext("warning: device '%s' " + "onlined, but remains in faulted state\n"), + argv[i]); + if (newstate == VDEV_STATE_FAULTED) + (void) printf(gettext("use 'zpool " + "clear' to restore a faulted " + "device\n")); + else + (void) printf(gettext("use 'zpool " + "replace' to replace devices " + "that are no longer present\n")); + } + } else { + ret = 1; + } + } + + zpool_close(zhp); + + return (ret); +} + +/* + * zpool offline [-ft] <pool> <device> ... + * + * -f Force the device into the offline state, even if doing + * so would appear to compromise pool availability. + * (not supported yet) + * + * -t Only take the device off-line temporarily. The offline + * state will not be persistent across reboots. + */ +/* ARGSUSED */ +int +zpool_do_offline(int argc, char **argv) +{ + int c, i; + char *poolname; + zpool_handle_t *zhp; + int ret = 0; + boolean_t istmp = B_FALSE; + + /* check options */ + while ((c = getopt(argc, argv, "ft")) != -1) { + switch (c) { + case 't': + istmp = B_TRUE; + break; + case 'f': + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + /* get pool name and check number of arguments */ + if (argc < 1) { + (void) fprintf(stderr, gettext("missing pool name\n")); + usage(B_FALSE); + } + if (argc < 2) { + (void) fprintf(stderr, gettext("missing device name\n")); + usage(B_FALSE); + } + + poolname = argv[0]; + + if ((zhp = zpool_open(g_zfs, poolname)) == NULL) + return (1); + + for (i = 1; i < argc; i++) { + if (zpool_vdev_offline(zhp, argv[i], istmp) != 0) + ret = 1; + } + + zpool_close(zhp); + + return (ret); +} + +/* + * zpool clear <pool> [device] + * + * Clear all errors associated with a pool or a particular device. + */ +int +zpool_do_clear(int argc, char **argv) +{ + int c; + int ret = 0; + boolean_t dryrun = B_FALSE; + boolean_t do_rewind = B_FALSE; + boolean_t xtreme_rewind = B_FALSE; + uint32_t rewind_policy = ZPOOL_NO_REWIND; + nvlist_t *policy = NULL; + zpool_handle_t *zhp; + char *pool, *device; + + /* check options */ + while ((c = getopt(argc, argv, "FnX")) != -1) { + switch (c) { + case 'F': + do_rewind = B_TRUE; + break; + case 'n': + dryrun = B_TRUE; + break; + case 'X': + xtreme_rewind = B_TRUE; + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + if (argc < 1) { + (void) fprintf(stderr, gettext("missing pool name\n")); + usage(B_FALSE); + } + + if (argc > 2) { + (void) fprintf(stderr, gettext("too many arguments\n")); + usage(B_FALSE); + } + + if ((dryrun || xtreme_rewind) && !do_rewind) { + (void) fprintf(stderr, + gettext("-n or -X only meaningful with -F\n")); + usage(B_FALSE); + } + if (dryrun) + rewind_policy = ZPOOL_TRY_REWIND; + else if (do_rewind) + rewind_policy = ZPOOL_DO_REWIND; + if (xtreme_rewind) + rewind_policy |= ZPOOL_EXTREME_REWIND; + + /* In future, further rewind policy choices can be passed along here */ + if (nvlist_alloc(&policy, NV_UNIQUE_NAME, 0) != 0 || + nvlist_add_uint32(policy, ZPOOL_REWIND_REQUEST, rewind_policy) != 0) + return (1); + + pool = argv[0]; + device = argc == 2 ? argv[1] : NULL; + + if ((zhp = zpool_open_canfail(g_zfs, pool)) == NULL) { + nvlist_free(policy); + return (1); + } + + if (zpool_clear(zhp, device, policy) != 0) + ret = 1; + + zpool_close(zhp); + + nvlist_free(policy); + + return (ret); +} + +typedef struct scrub_cbdata { + int cb_type; + int cb_argc; + char **cb_argv; +} scrub_cbdata_t; + +int +scrub_callback(zpool_handle_t *zhp, void *data) +{ + scrub_cbdata_t *cb = data; + int err; + + /* + * Ignore faulted pools. + */ + if (zpool_get_state(zhp) == POOL_STATE_UNAVAIL) { + (void) fprintf(stderr, gettext("cannot scrub '%s': pool is " + "currently unavailable\n"), zpool_get_name(zhp)); + return (1); + } + + err = zpool_scan(zhp, cb->cb_type); + + return (err != 0); +} + +/* + * zpool scrub [-s] <pool> ... + * + * -s Stop. Stops any in-progress scrub. + */ +int +zpool_do_scrub(int argc, char **argv) +{ + int c; + scrub_cbdata_t cb; + + cb.cb_type = POOL_SCAN_SCRUB; + + /* check options */ + while ((c = getopt(argc, argv, "s")) != -1) { + switch (c) { + case 's': + cb.cb_type = POOL_SCAN_NONE; + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + cb.cb_argc = argc; + cb.cb_argv = argv; + argc -= optind; + argv += optind; + + if (argc < 1) { + (void) fprintf(stderr, gettext("missing pool name argument\n")); + usage(B_FALSE); + } + + return (for_each_pool(argc, argv, B_TRUE, NULL, scrub_callback, &cb)); +} + +typedef struct status_cbdata { + int cb_count; + boolean_t cb_allpools; + boolean_t cb_verbose; + boolean_t cb_explain; + boolean_t cb_first; + boolean_t cb_dedup_stats; +} status_cbdata_t; + +/* + * Print out detailed scrub status. + */ +void +print_scan_status(pool_scan_stat_t *ps) +{ + time_t start, end; + uint64_t elapsed, mins_left, hours_left; + uint64_t pass_exam, examined, total; + uint_t rate; + double fraction_done; + char processed_buf[7], examined_buf[7], total_buf[7], rate_buf[7]; + + (void) printf(gettext(" scan: ")); + + /* If there's never been a scan, there's not much to say. */ + if (ps == NULL || ps->pss_func == POOL_SCAN_NONE || + ps->pss_func >= POOL_SCAN_FUNCS) { + (void) printf(gettext("none requested\n")); + return; + } + + start = ps->pss_start_time; + end = ps->pss_end_time; + zfs_nicenum(ps->pss_processed, processed_buf, sizeof (processed_buf)); + + assert(ps->pss_func == POOL_SCAN_SCRUB || + ps->pss_func == POOL_SCAN_RESILVER); + /* + * Scan is finished or canceled. + */ + if (ps->pss_state == DSS_FINISHED) { + uint64_t minutes_taken = (end - start) / 60; + char *fmt; + + if (ps->pss_func == POOL_SCAN_SCRUB) { + fmt = gettext("scrub repaired %s in %lluh%um with " + "%llu errors on %s"); + } else if (ps->pss_func == POOL_SCAN_RESILVER) { + fmt = gettext("resilvered %s in %lluh%um with " + "%llu errors on %s"); + } + /* LINTED */ + (void) printf(fmt, processed_buf, + (u_longlong_t)(minutes_taken / 60), + (uint_t)(minutes_taken % 60), + (u_longlong_t)ps->pss_errors, + ctime((time_t *)&end)); + return; + } else if (ps->pss_state == DSS_CANCELED) { + if (ps->pss_func == POOL_SCAN_SCRUB) { + (void) printf(gettext("scrub canceled on %s"), + ctime(&end)); + } else if (ps->pss_func == POOL_SCAN_RESILVER) { + (void) printf(gettext("resilver canceled on %s"), + ctime(&end)); + } + return; + } + + assert(ps->pss_state == DSS_SCANNING); + + /* + * Scan is in progress. + */ + if (ps->pss_func == POOL_SCAN_SCRUB) { + (void) printf(gettext("scrub in progress since %s"), + ctime(&start)); + } else if (ps->pss_func == POOL_SCAN_RESILVER) { + (void) printf(gettext("resilver in progress since %s"), + ctime(&start)); + } + + examined = ps->pss_examined ? ps->pss_examined : 1; + total = ps->pss_to_examine; + fraction_done = (double)examined / total; + + /* elapsed time for this pass */ + elapsed = time(NULL) - ps->pss_pass_start; + elapsed = elapsed ? elapsed : 1; + pass_exam = ps->pss_pass_exam ? ps->pss_pass_exam : 1; + rate = pass_exam / elapsed; + rate = rate ? rate : 1; + mins_left = ((total - examined) / rate) / 60; + hours_left = mins_left / 60; + + zfs_nicenum(examined, examined_buf, sizeof (examined_buf)); + zfs_nicenum(total, total_buf, sizeof (total_buf)); + zfs_nicenum(rate, rate_buf, sizeof (rate_buf)); + + /* + * do not print estimated time if hours_left is more than 30 days + */ + (void) printf(gettext(" %s scanned out of %s at %s/s"), + examined_buf, total_buf, rate_buf); + if (hours_left < (30 * 24)) { + (void) printf(gettext(", %lluh%um to go\n"), + (u_longlong_t)hours_left, (uint_t)(mins_left % 60)); + } else { + (void) printf(gettext( + ", (scan is slow, no estimated time)\n")); + } + + if (ps->pss_func == POOL_SCAN_RESILVER) { + (void) printf(gettext(" %s resilvered, %.2f%% done\n"), + processed_buf, 100 * fraction_done); + } else if (ps->pss_func == POOL_SCAN_SCRUB) { + (void) printf(gettext(" %s repaired, %.2f%% done\n"), + processed_buf, 100 * fraction_done); + } +} + +static void +print_error_log(zpool_handle_t *zhp) +{ + nvlist_t *nverrlist = NULL; + nvpair_t *elem; + char *pathname; + size_t len = MAXPATHLEN * 2; + + if (zpool_get_errlog(zhp, &nverrlist) != 0) { + (void) printf("errors: List of errors unavailable " + "(insufficient privileges)\n"); + return; + } + + (void) printf("errors: Permanent errors have been " + "detected in the following files:\n\n"); + + pathname = safe_malloc(len); + elem = NULL; + while ((elem = nvlist_next_nvpair(nverrlist, elem)) != NULL) { + nvlist_t *nv; + uint64_t dsobj, obj; + + verify(nvpair_value_nvlist(elem, &nv) == 0); + verify(nvlist_lookup_uint64(nv, ZPOOL_ERR_DATASET, + &dsobj) == 0); + verify(nvlist_lookup_uint64(nv, ZPOOL_ERR_OBJECT, + &obj) == 0); + zpool_obj_to_path(zhp, dsobj, obj, pathname, len); + (void) printf("%7s %s\n", "", pathname); + } + free(pathname); + nvlist_free(nverrlist); +} + +static void +print_spares(zpool_handle_t *zhp, nvlist_t **spares, uint_t nspares, + int namewidth) +{ + uint_t i; + char *name; + + if (nspares == 0) + return; + + (void) printf(gettext("\tspares\n")); + + for (i = 0; i < nspares; i++) { + name = zpool_vdev_name(g_zfs, zhp, spares[i], B_FALSE); + print_status_config(zhp, name, spares[i], + namewidth, 2, B_TRUE); + free(name); + } +} + +static void +print_l2cache(zpool_handle_t *zhp, nvlist_t **l2cache, uint_t nl2cache, + int namewidth) +{ + uint_t i; + char *name; + + if (nl2cache == 0) + return; + + (void) printf(gettext("\tcache\n")); + + for (i = 0; i < nl2cache; i++) { + name = zpool_vdev_name(g_zfs, zhp, l2cache[i], B_FALSE); + print_status_config(zhp, name, l2cache[i], + namewidth, 2, B_FALSE); + free(name); + } +} + +static void +print_dedup_stats(nvlist_t *config) +{ + ddt_histogram_t *ddh; + ddt_stat_t *dds; + ddt_object_t *ddo; + uint_t c; + + /* + * If the pool was faulted then we may not have been able to + * obtain the config. Otherwise, if have anything in the dedup + * table continue processing the stats. + */ + if (nvlist_lookup_uint64_array(config, ZPOOL_CONFIG_DDT_OBJ_STATS, + (uint64_t **)&ddo, &c) != 0 || ddo->ddo_count == 0) + return; + + (void) printf("\n"); + (void) printf("DDT entries %llu, size %llu on disk, %llu in core\n", + (u_longlong_t)ddo->ddo_count, + (u_longlong_t)ddo->ddo_dspace, + (u_longlong_t)ddo->ddo_mspace); + + verify(nvlist_lookup_uint64_array(config, ZPOOL_CONFIG_DDT_STATS, + (uint64_t **)&dds, &c) == 0); + verify(nvlist_lookup_uint64_array(config, ZPOOL_CONFIG_DDT_HISTOGRAM, + (uint64_t **)&ddh, &c) == 0); + zpool_dump_ddt(dds, ddh); +} + +/* + * Display a summary of pool status. Displays a summary such as: + * + * pool: tank + * status: DEGRADED + * reason: One or more devices ... + * see: http://www.sun.com/msg/ZFS-xxxx-01 + * config: + * mirror DEGRADED + * c1t0d0 OK + * c2t0d0 UNAVAIL + * + * When given the '-v' option, we print out the complete config. If the '-e' + * option is specified, then we print out error rate information as well. + */ +int +status_callback(zpool_handle_t *zhp, void *data) +{ + status_cbdata_t *cbp = data; + nvlist_t *config, *nvroot; + char *msgid; + int reason; + const char *health; + uint_t c; + vdev_stat_t *vs; + + config = zpool_get_config(zhp, NULL); + reason = zpool_get_status(zhp, &msgid); + + cbp->cb_count++; + + /* + * If we were given 'zpool status -x', only report those pools with + * problems. + */ + if (reason == ZPOOL_STATUS_OK && cbp->cb_explain) { + if (!cbp->cb_allpools) { + (void) printf(gettext("pool '%s' is healthy\n"), + zpool_get_name(zhp)); + if (cbp->cb_first) + cbp->cb_first = B_FALSE; + } + return (0); + } + + if (cbp->cb_first) + cbp->cb_first = B_FALSE; + else + (void) printf("\n"); + + verify(nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, + &nvroot) == 0); + verify(nvlist_lookup_uint64_array(nvroot, ZPOOL_CONFIG_VDEV_STATS, + (uint64_t **)&vs, &c) == 0); + health = zpool_state_to_name(vs->vs_state, vs->vs_aux); + + (void) printf(gettext(" pool: %s\n"), zpool_get_name(zhp)); + (void) printf(gettext(" state: %s\n"), health); + + switch (reason) { + case ZPOOL_STATUS_MISSING_DEV_R: + (void) printf(gettext("status: One or more devices could not " + "be opened. Sufficient replicas exist for\n\tthe pool to " + "continue functioning in a degraded state.\n")); + (void) printf(gettext("action: Attach the missing device and " + "online it using 'zpool online'.\n")); + break; + + case ZPOOL_STATUS_MISSING_DEV_NR: + (void) printf(gettext("status: One or more devices could not " + "be opened. There are insufficient\n\treplicas for the " + "pool to continue functioning.\n")); + (void) printf(gettext("action: Attach the missing device and " + "online it using 'zpool online'.\n")); + break; + + case ZPOOL_STATUS_CORRUPT_LABEL_R: + (void) printf(gettext("status: One or more devices could not " + "be used because the label is missing or\n\tinvalid. " + "Sufficient replicas exist for the pool to continue\n\t" + "functioning in a degraded state.\n")); + (void) printf(gettext("action: Replace the device using " + "'zpool replace'.\n")); + break; + + case ZPOOL_STATUS_CORRUPT_LABEL_NR: + (void) printf(gettext("status: One or more devices could not " + "be used because the label is missing \n\tor invalid. " + "There are insufficient replicas for the pool to " + "continue\n\tfunctioning.\n")); + zpool_explain_recover(zpool_get_handle(zhp), + zpool_get_name(zhp), reason, config); + break; + + case ZPOOL_STATUS_FAILING_DEV: + (void) printf(gettext("status: One or more devices has " + "experienced an unrecoverable error. An\n\tattempt was " + "made to correct the error. Applications are " + "unaffected.\n")); + (void) printf(gettext("action: Determine if the device needs " + "to be replaced, and clear the errors\n\tusing " + "'zpool clear' or replace the device with 'zpool " + "replace'.\n")); + break; + + case ZPOOL_STATUS_OFFLINE_DEV: + (void) printf(gettext("status: One or more devices has " + "been taken offline by the administrator.\n\tSufficient " + "replicas exist for the pool to continue functioning in " + "a\n\tdegraded state.\n")); + (void) printf(gettext("action: Online the device using " + "'zpool online' or replace the device with\n\t'zpool " + "replace'.\n")); + break; + + case ZPOOL_STATUS_REMOVED_DEV: + (void) printf(gettext("status: One or more devices has " + "been removed by the administrator.\n\tSufficient " + "replicas exist for the pool to continue functioning in " + "a\n\tdegraded state.\n")); + (void) printf(gettext("action: Online the device using " + "'zpool online' or replace the device with\n\t'zpool " + "replace'.\n")); + break; + + case ZPOOL_STATUS_RESILVERING: + (void) printf(gettext("status: One or more devices is " + "currently being resilvered. The pool will\n\tcontinue " + "to function, possibly in a degraded state.\n")); + (void) printf(gettext("action: Wait for the resilver to " + "complete.\n")); + break; + + case ZPOOL_STATUS_CORRUPT_DATA: + (void) printf(gettext("status: One or more devices has " + "experienced an error resulting in data\n\tcorruption. " + "Applications may be affected.\n")); + (void) printf(gettext("action: Restore the file in question " + "if possible. Otherwise restore the\n\tentire pool from " + "backup.\n")); + break; + + case ZPOOL_STATUS_CORRUPT_POOL: + (void) printf(gettext("status: The pool metadata is corrupted " + "and the pool cannot be opened.\n")); + zpool_explain_recover(zpool_get_handle(zhp), + zpool_get_name(zhp), reason, config); + break; + + case ZPOOL_STATUS_VERSION_OLDER: + (void) printf(gettext("status: The pool is formatted using an " + "older on-disk format. The pool can\n\tstill be used, but " + "some features are unavailable.\n")); + (void) printf(gettext("action: Upgrade the pool using 'zpool " + "upgrade'. Once this is done, the\n\tpool will no longer " + "be accessible on older software versions.\n")); + break; + + case ZPOOL_STATUS_VERSION_NEWER: + (void) printf(gettext("status: The pool has been upgraded to a " + "newer, incompatible on-disk version.\n\tThe pool cannot " + "be accessed on this system.\n")); + (void) printf(gettext("action: Access the pool from a system " + "running more recent software, or\n\trestore the pool from " + "backup.\n")); + break; + + case ZPOOL_STATUS_FAULTED_DEV_R: + (void) printf(gettext("status: One or more devices are " + "faulted in response to persistent errors.\n\tSufficient " + "replicas exist for the pool to continue functioning " + "in a\n\tdegraded state.\n")); + (void) printf(gettext("action: Replace the faulted device, " + "or use 'zpool clear' to mark the device\n\trepaired.\n")); + break; + + case ZPOOL_STATUS_FAULTED_DEV_NR: + (void) printf(gettext("status: One or more devices are " + "faulted in response to persistent errors. There are " + "insufficient replicas for the pool to\n\tcontinue " + "functioning.\n")); + (void) printf(gettext("action: Destroy and re-create the pool " + "from a backup source. Manually marking the device\n" + "\trepaired using 'zpool clear' may allow some data " + "to be recovered.\n")); + break; + + case ZPOOL_STATUS_IO_FAILURE_WAIT: + case ZPOOL_STATUS_IO_FAILURE_CONTINUE: + (void) printf(gettext("status: One or more devices are " + "faulted in response to IO failures.\n")); + (void) printf(gettext("action: Make sure the affected devices " + "are connected, then run 'zpool clear'.\n")); + break; + + case ZPOOL_STATUS_BAD_LOG: + (void) printf(gettext("status: An intent log record " + "could not be read.\n" + "\tWaiting for adminstrator intervention to fix the " + "faulted pool.\n")); + (void) printf(gettext("action: Either restore the affected " + "device(s) and run 'zpool online',\n" + "\tor ignore the intent log records by running " + "'zpool clear'.\n")); + break; + + default: + /* + * The remaining errors can't actually be generated, yet. + */ + assert(reason == ZPOOL_STATUS_OK); + } + + if (msgid != NULL) + (void) printf(gettext(" see: http://www.sun.com/msg/%s\n"), + msgid); + + if (config != NULL) { + int namewidth; + uint64_t nerr; + nvlist_t **spares, **l2cache; + uint_t nspares, nl2cache; + pool_scan_stat_t *ps = NULL; + + (void) nvlist_lookup_uint64_array(nvroot, + ZPOOL_CONFIG_SCAN_STATS, (uint64_t **)&ps, &c); + print_scan_status(ps); + + namewidth = max_width(zhp, nvroot, 0, 0); + if (namewidth < 10) + namewidth = 10; + + (void) printf(gettext("config:\n\n")); + (void) printf(gettext("\t%-*s %-8s %5s %5s %5s\n"), namewidth, + "NAME", "STATE", "READ", "WRITE", "CKSUM"); + print_status_config(zhp, zpool_get_name(zhp), nvroot, + namewidth, 0, B_FALSE); + + if (num_logs(nvroot) > 0) + print_logs(zhp, nvroot, namewidth, B_TRUE); + if (nvlist_lookup_nvlist_array(nvroot, ZPOOL_CONFIG_L2CACHE, + &l2cache, &nl2cache) == 0) + print_l2cache(zhp, l2cache, nl2cache, namewidth); + + if (nvlist_lookup_nvlist_array(nvroot, ZPOOL_CONFIG_SPARES, + &spares, &nspares) == 0) + print_spares(zhp, spares, nspares, namewidth); + + if (nvlist_lookup_uint64(config, ZPOOL_CONFIG_ERRCOUNT, + &nerr) == 0) { + nvlist_t *nverrlist = NULL; + + /* + * If the approximate error count is small, get a + * precise count by fetching the entire log and + * uniquifying the results. + */ + if (nerr > 0 && nerr < 100 && !cbp->cb_verbose && + zpool_get_errlog(zhp, &nverrlist) == 0) { + nvpair_t *elem; + + elem = NULL; + nerr = 0; + while ((elem = nvlist_next_nvpair(nverrlist, + elem)) != NULL) { + nerr++; + } + } + nvlist_free(nverrlist); + + (void) printf("\n"); + + if (nerr == 0) + (void) printf(gettext("errors: No known data " + "errors\n")); + else if (!cbp->cb_verbose) + (void) printf(gettext("errors: %llu data " + "errors, use '-v' for a list\n"), + (u_longlong_t)nerr); + else + print_error_log(zhp); + } + + if (cbp->cb_dedup_stats) + print_dedup_stats(config); + } else { + (void) printf(gettext("config: The configuration cannot be " + "determined.\n")); + } + + return (0); +} + +/* + * zpool status [-vx] [-T d|u] [pool] ... [interval [count]] + * + * -v Display complete error logs + * -x Display only pools with potential problems + * -D Display dedup status (undocumented) + * -T Display a timestamp in date(1) or Unix format + * + * Describes the health status of all pools or some subset. + */ +int +zpool_do_status(int argc, char **argv) +{ + int c; + int ret; + unsigned long interval = 0, count = 0; + status_cbdata_t cb = { 0 }; + + /* check options */ + while ((c = getopt(argc, argv, "vxDT:")) != -1) { + switch (c) { + case 'v': + cb.cb_verbose = B_TRUE; + break; + case 'x': + cb.cb_explain = B_TRUE; + break; + case 'D': + cb.cb_dedup_stats = B_TRUE; + break; + case 'T': + get_timestamp_arg(*optarg); + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + argc -= optind; + argv += optind; + + get_interval_count(&argc, argv, &interval, &count); + + if (argc == 0) + cb.cb_allpools = B_TRUE; + + cb.cb_first = B_TRUE; + + for (;;) { + if (timestamp_fmt != NODATE) + print_timestamp(timestamp_fmt); + + ret = for_each_pool(argc, argv, B_TRUE, NULL, + status_callback, &cb); + + if (argc == 0 && cb.cb_count == 0) + (void) printf(gettext("no pools available\n")); + else if (cb.cb_explain && cb.cb_first && cb.cb_allpools) + (void) printf(gettext("all pools are healthy\n")); + + if (ret != 0) + return (ret); + + if (interval == 0) + break; + + if (count != 0 && --count == 0) + break; + + (void) sleep(interval); + } + + return (0); +} + +typedef struct upgrade_cbdata { + int cb_all; + int cb_first; + int cb_newer; + int cb_argc; + uint64_t cb_version; + char **cb_argv; +} upgrade_cbdata_t; + +static int +upgrade_cb(zpool_handle_t *zhp, void *arg) +{ + upgrade_cbdata_t *cbp = arg; + nvlist_t *config; + uint64_t version; + int ret = 0; + + config = zpool_get_config(zhp, NULL); + verify(nvlist_lookup_uint64(config, ZPOOL_CONFIG_VERSION, + &version) == 0); + + if (!cbp->cb_newer && version < SPA_VERSION) { + if (!cbp->cb_all) { + if (cbp->cb_first) { + (void) printf(gettext("The following pools are " + "out of date, and can be upgraded. After " + "being\nupgraded, these pools will no " + "longer be accessible by older software " + "versions.\n\n")); + (void) printf(gettext("VER POOL\n")); + (void) printf(gettext("--- ------------\n")); + cbp->cb_first = B_FALSE; + } + + (void) printf("%2llu %s\n", (u_longlong_t)version, + zpool_get_name(zhp)); + } else { + cbp->cb_first = B_FALSE; + ret = zpool_upgrade(zhp, cbp->cb_version); + if (!ret) { + (void) printf(gettext("Successfully upgraded " + "'%s'\n\n"), zpool_get_name(zhp)); + } + } + } else if (cbp->cb_newer && version > SPA_VERSION) { + assert(!cbp->cb_all); + + if (cbp->cb_first) { + (void) printf(gettext("The following pools are " + "formatted using a newer software version and\n" + "cannot be accessed on the current system.\n\n")); + (void) printf(gettext("VER POOL\n")); + (void) printf(gettext("--- ------------\n")); + cbp->cb_first = B_FALSE; + } + + (void) printf("%2llu %s\n", (u_longlong_t)version, + zpool_get_name(zhp)); + } + + zpool_close(zhp); + return (ret); +} + +/* ARGSUSED */ +static int +upgrade_one(zpool_handle_t *zhp, void *data) +{ + upgrade_cbdata_t *cbp = data; + uint64_t cur_version; + int ret; + + if (strcmp("log", zpool_get_name(zhp)) == 0) { + (void) printf(gettext("'log' is now a reserved word\n" + "Pool 'log' must be renamed using export and import" + " to upgrade.\n")); + return (1); + } + + cur_version = zpool_get_prop_int(zhp, ZPOOL_PROP_VERSION, NULL); + if (cur_version > cbp->cb_version) { + (void) printf(gettext("Pool '%s' is already formatted " + "using more current version '%llu'.\n"), + zpool_get_name(zhp), cur_version); + return (0); + } + if (cur_version == cbp->cb_version) { + (void) printf(gettext("Pool '%s' is already formatted " + "using the current version.\n"), zpool_get_name(zhp)); + return (0); + } + + ret = zpool_upgrade(zhp, cbp->cb_version); + + if (!ret) { + (void) printf(gettext("Successfully upgraded '%s' " + "from version %llu to version %llu\n\n"), + zpool_get_name(zhp), (u_longlong_t)cur_version, + (u_longlong_t)cbp->cb_version); + } + + return (ret != 0); +} + +/* + * zpool upgrade + * zpool upgrade -v + * zpool upgrade [-V version] <-a | pool ...> + * + * With no arguments, display downrev'd ZFS pool available for upgrade. + * Individual pools can be upgraded by specifying the pool, and '-a' will + * upgrade all pools. + */ +int +zpool_do_upgrade(int argc, char **argv) +{ + int c; + upgrade_cbdata_t cb = { 0 }; + int ret = 0; + boolean_t showversions = B_FALSE; + char *end; + + + /* check options */ + while ((c = getopt(argc, argv, ":avV:")) != -1) { + switch (c) { + case 'a': + cb.cb_all = B_TRUE; + break; + case 'v': + showversions = B_TRUE; + break; + case 'V': + cb.cb_version = strtoll(optarg, &end, 10); + if (*end != '\0' || cb.cb_version > SPA_VERSION || + cb.cb_version < SPA_VERSION_1) { + (void) fprintf(stderr, + gettext("invalid version '%s'\n"), optarg); + usage(B_FALSE); + } + break; + case ':': + (void) fprintf(stderr, gettext("missing argument for " + "'%c' option\n"), optopt); + usage(B_FALSE); + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + + cb.cb_argc = argc; + cb.cb_argv = argv; + argc -= optind; + argv += optind; + + if (cb.cb_version == 0) { + cb.cb_version = SPA_VERSION; + } else if (!cb.cb_all && argc == 0) { + (void) fprintf(stderr, gettext("-V option is " + "incompatible with other arguments\n")); + usage(B_FALSE); + } + + if (showversions) { + if (cb.cb_all || argc != 0) { + (void) fprintf(stderr, gettext("-v option is " + "incompatible with other arguments\n")); + usage(B_FALSE); + } + } else if (cb.cb_all) { + if (argc != 0) { + (void) fprintf(stderr, gettext("-a option should not " + "be used along with a pool name\n")); + usage(B_FALSE); + } + } + + (void) printf(gettext("This system is currently running " + "ZFS pool version %llu.\n\n"), SPA_VERSION); + cb.cb_first = B_TRUE; + if (showversions) { + (void) printf(gettext("The following versions are " + "supported:\n\n")); + (void) printf(gettext("VER DESCRIPTION\n")); + (void) printf("--- -----------------------------------------" + "---------------\n"); + (void) printf(gettext(" 1 Initial ZFS version\n")); + (void) printf(gettext(" 2 Ditto blocks " + "(replicated metadata)\n")); + (void) printf(gettext(" 3 Hot spares and double parity " + "RAID-Z\n")); + (void) printf(gettext(" 4 zpool history\n")); + (void) printf(gettext(" 5 Compression using the gzip " + "algorithm\n")); + (void) printf(gettext(" 6 bootfs pool property\n")); + (void) printf(gettext(" 7 Separate intent log devices\n")); + (void) printf(gettext(" 8 Delegated administration\n")); + (void) printf(gettext(" 9 refquota and refreservation " + "properties\n")); + (void) printf(gettext(" 10 Cache devices\n")); + (void) printf(gettext(" 11 Improved scrub performance\n")); + (void) printf(gettext(" 12 Snapshot properties\n")); + (void) printf(gettext(" 13 snapused property\n")); + (void) printf(gettext(" 14 passthrough-x aclinherit\n")); + (void) printf(gettext(" 15 user/group space accounting\n")); + (void) printf(gettext(" 16 stmf property support\n")); + (void) printf(gettext(" 17 Triple-parity RAID-Z\n")); + (void) printf(gettext(" 18 Snapshot user holds\n")); + (void) printf(gettext(" 19 Log device removal\n")); + (void) printf(gettext(" 20 Compression using zle " + "(zero-length encoding)\n")); + (void) printf(gettext(" 21 Deduplication\n")); + (void) printf(gettext(" 22 Received properties\n")); + (void) printf(gettext(" 23 Slim ZIL\n")); + (void) printf(gettext(" 24 System attributes\n")); + (void) printf(gettext(" 25 Improved scrub stats\n")); + (void) printf(gettext(" 26 Improved snapshot deletion " + "performance\n")); + (void) printf(gettext(" 27 Improved snapshot creation " + "performance\n")); + (void) printf(gettext(" 28 Multiple vdev replacements\n")); + (void) printf(gettext("\nFor more information on a particular " + "version, including supported releases,\n")); + (void) printf(gettext("see the ZFS Administration Guide.\n\n")); + } else if (argc == 0) { + int notfound; + + ret = zpool_iter(g_zfs, upgrade_cb, &cb); + notfound = cb.cb_first; + + if (!cb.cb_all && ret == 0) { + if (!cb.cb_first) + (void) printf("\n"); + cb.cb_first = B_TRUE; + cb.cb_newer = B_TRUE; + ret = zpool_iter(g_zfs, upgrade_cb, &cb); + if (!cb.cb_first) { + notfound = B_FALSE; + (void) printf("\n"); + } + } + + if (ret == 0) { + if (notfound) + (void) printf(gettext("All pools are formatted " + "using this version.\n")); + else if (!cb.cb_all) + (void) printf(gettext("Use 'zpool upgrade -v' " + "for a list of available versions and " + "their associated\nfeatures.\n")); + } + } else { + ret = for_each_pool(argc, argv, B_FALSE, NULL, + upgrade_one, &cb); + } + + return (ret); +} + +typedef struct hist_cbdata { + boolean_t first; + int longfmt; + int internal; +} hist_cbdata_t; + +/* + * Print out the command history for a specific pool. + */ +static int +get_history_one(zpool_handle_t *zhp, void *data) +{ + nvlist_t *nvhis; + nvlist_t **records; + uint_t numrecords; + char *cmdstr; + char *pathstr; + uint64_t dst_time; + time_t tsec; + struct tm t; + char tbuf[30]; + int ret, i; + uint64_t who; + struct passwd *pwd; + char *hostname; + char *zonename; + char internalstr[MAXPATHLEN]; + hist_cbdata_t *cb = (hist_cbdata_t *)data; + uint64_t txg; + uint64_t ievent; + + cb->first = B_FALSE; + + (void) printf(gettext("History for '%s':\n"), zpool_get_name(zhp)); + + if ((ret = zpool_get_history(zhp, &nvhis)) != 0) + return (ret); + + verify(nvlist_lookup_nvlist_array(nvhis, ZPOOL_HIST_RECORD, + &records, &numrecords) == 0); + for (i = 0; i < numrecords; i++) { + if (nvlist_lookup_uint64(records[i], ZPOOL_HIST_TIME, + &dst_time) != 0) + continue; + + /* is it an internal event or a standard event? */ + if (nvlist_lookup_string(records[i], ZPOOL_HIST_CMD, + &cmdstr) != 0) { + if (cb->internal == 0) + continue; + + if (nvlist_lookup_uint64(records[i], + ZPOOL_HIST_INT_EVENT, &ievent) != 0) + continue; + verify(nvlist_lookup_uint64(records[i], + ZPOOL_HIST_TXG, &txg) == 0); + verify(nvlist_lookup_string(records[i], + ZPOOL_HIST_INT_STR, &pathstr) == 0); + if (ievent >= LOG_END) + continue; + (void) snprintf(internalstr, + sizeof (internalstr), + "[internal %s txg:%lld] %s", + zfs_history_event_names[ievent], txg, + pathstr); + cmdstr = internalstr; + } + tsec = dst_time; + (void) localtime_r(&tsec, &t); + (void) strftime(tbuf, sizeof (tbuf), "%F.%T", &t); + (void) printf("%s %s", tbuf, cmdstr); + + if (!cb->longfmt) { + (void) printf("\n"); + continue; + } + (void) printf(" ["); + if (nvlist_lookup_uint64(records[i], + ZPOOL_HIST_WHO, &who) == 0) { + pwd = getpwuid((uid_t)who); + if (pwd) + (void) printf("user %s on", + pwd->pw_name); + else + (void) printf("user %d on", + (int)who); + } else { + (void) printf(gettext("no info]\n")); + continue; + } + if (nvlist_lookup_string(records[i], + ZPOOL_HIST_HOST, &hostname) == 0) { + (void) printf(" %s", hostname); + } + if (nvlist_lookup_string(records[i], + ZPOOL_HIST_ZONE, &zonename) == 0) { + (void) printf(":%s", zonename); + } + + (void) printf("]"); + (void) printf("\n"); + } + (void) printf("\n"); + nvlist_free(nvhis); + + return (ret); +} + +/* + * zpool history <pool> + * + * Displays the history of commands that modified pools. + */ + + +int +zpool_do_history(int argc, char **argv) +{ + hist_cbdata_t cbdata = { 0 }; + int ret; + int c; + + cbdata.first = B_TRUE; + /* check options */ + while ((c = getopt(argc, argv, "li")) != -1) { + switch (c) { + case 'l': + cbdata.longfmt = 1; + break; + case 'i': + cbdata.internal = 1; + break; + case '?': + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + optopt); + usage(B_FALSE); + } + } + argc -= optind; + argv += optind; + + ret = for_each_pool(argc, argv, B_FALSE, NULL, get_history_one, + &cbdata); + + if (argc == 0 && cbdata.first == B_TRUE) { + (void) printf(gettext("no pools available\n")); + return (0); + } + + return (ret); +} + +static int +get_callback(zpool_handle_t *zhp, void *data) +{ + zprop_get_cbdata_t *cbp = (zprop_get_cbdata_t *)data; + char value[MAXNAMELEN]; + zprop_source_t srctype; + zprop_list_t *pl; + + for (pl = cbp->cb_proplist; pl != NULL; pl = pl->pl_next) { + + /* + * Skip the special fake placeholder. This will also skip + * over the name property when 'all' is specified. + */ + if (pl->pl_prop == ZPOOL_PROP_NAME && + pl == cbp->cb_proplist) + continue; + + if (zpool_get_prop(zhp, pl->pl_prop, + value, sizeof (value), &srctype) != 0) + continue; + + zprop_print_one_property(zpool_get_name(zhp), cbp, + zpool_prop_to_name(pl->pl_prop), value, srctype, NULL, + NULL); + } + return (0); +} + +int +zpool_do_get(int argc, char **argv) +{ + zprop_get_cbdata_t cb = { 0 }; + zprop_list_t fake_name = { 0 }; + int ret; + + if (argc < 3) + usage(B_FALSE); + + cb.cb_first = B_TRUE; + cb.cb_sources = ZPROP_SRC_ALL; + cb.cb_columns[0] = GET_COL_NAME; + cb.cb_columns[1] = GET_COL_PROPERTY; + cb.cb_columns[2] = GET_COL_VALUE; + cb.cb_columns[3] = GET_COL_SOURCE; + cb.cb_type = ZFS_TYPE_POOL; + + if (zprop_get_list(g_zfs, argv[1], &cb.cb_proplist, + ZFS_TYPE_POOL) != 0) + usage(B_FALSE); + + if (cb.cb_proplist != NULL) { + fake_name.pl_prop = ZPOOL_PROP_NAME; + fake_name.pl_width = strlen(gettext("NAME")); + fake_name.pl_next = cb.cb_proplist; + cb.cb_proplist = &fake_name; + } + + ret = for_each_pool(argc - 2, argv + 2, B_TRUE, &cb.cb_proplist, + get_callback, &cb); + + if (cb.cb_proplist == &fake_name) + zprop_free_list(fake_name.pl_next); + else + zprop_free_list(cb.cb_proplist); + + return (ret); +} + +typedef struct set_cbdata { + char *cb_propname; + char *cb_value; + boolean_t cb_any_successful; +} set_cbdata_t; + +int +set_callback(zpool_handle_t *zhp, void *data) +{ + int error; + set_cbdata_t *cb = (set_cbdata_t *)data; + + error = zpool_set_prop(zhp, cb->cb_propname, cb->cb_value); + + if (!error) + cb->cb_any_successful = B_TRUE; + + return (error); +} + +int +zpool_do_set(int argc, char **argv) +{ + set_cbdata_t cb = { 0 }; + int error; + + if (argc > 1 && argv[1][0] == '-') { + (void) fprintf(stderr, gettext("invalid option '%c'\n"), + argv[1][1]); + usage(B_FALSE); + } + + if (argc < 2) { + (void) fprintf(stderr, gettext("missing property=value " + "argument\n")); + usage(B_FALSE); + } + + if (argc < 3) { + (void) fprintf(stderr, gettext("missing pool name\n")); + usage(B_FALSE); + } + + if (argc > 3) { + (void) fprintf(stderr, gettext("too many pool names\n")); + usage(B_FALSE); + } + + cb.cb_propname = argv[1]; + cb.cb_value = strchr(cb.cb_propname, '='); + if (cb.cb_value == NULL) { + (void) fprintf(stderr, gettext("missing value in " + "property=value argument\n")); + usage(B_FALSE); + } + + *(cb.cb_value) = '\0'; + cb.cb_value++; + + error = for_each_pool(argc - 2, argv + 2, B_TRUE, NULL, + set_callback, &cb); + + return (error); +} + +static int +find_command_idx(char *command, int *idx) +{ + int i; + + for (i = 0; i < NCOMMAND; i++) { + if (command_table[i].name == NULL) + continue; + + if (strcmp(command, command_table[i].name) == 0) { + *idx = i; + return (0); + } + } + return (1); +} + +int +main(int argc, char **argv) +{ + int ret; + int i; + char *cmdname; + + (void) setlocale(LC_ALL, ""); + (void) textdomain(TEXT_DOMAIN); + + if ((g_zfs = libzfs_init()) == NULL) { + (void) fprintf(stderr, gettext("internal error: failed to " + "initialize ZFS library\n")); + return (1); + } + + libzfs_print_on_error(g_zfs, B_TRUE); + + opterr = 0; + + /* + * Make sure the user has specified some command. + */ + if (argc < 2) { + (void) fprintf(stderr, gettext("missing command\n")); + usage(B_FALSE); + } + + cmdname = argv[1]; + + /* + * Special case '-?' + */ + if (strcmp(cmdname, "-?") == 0) + usage(B_TRUE); + + zpool_set_history_str("zpool", argc, argv, history_str); + verify(zpool_stage_history(g_zfs, history_str) == 0); + + /* + * Run the appropriate command. + */ + if (find_command_idx(cmdname, &i) == 0) { + current_command = &command_table[i]; + ret = command_table[i].func(argc - 1, argv + 1); + } else if (strchr(cmdname, '=')) { + verify(find_command_idx("set", &i) == 0); + current_command = &command_table[i]; + ret = command_table[i].func(argc, argv); + } else if (strcmp(cmdname, "freeze") == 0 && argc == 3) { + /* + * 'freeze' is a vile debugging abomination, so we treat + * it as such. + */ + char buf[16384]; + int fd = open(ZFS_DEV, O_RDWR); + (void) strcpy((void *)buf, argv[2]); + return (!!ioctl(fd, ZFS_IOC_POOL_FREEZE, buf)); + } else { + (void) fprintf(stderr, gettext("unrecognized " + "command '%s'\n"), cmdname); + usage(B_FALSE); + } + + libzfs_fini(g_zfs); + + /* + * The 'ZFS_ABORT' environment variable causes us to dump core on exit + * for the purposes of running ::findleaks. + */ + if (getenv("ZFS_ABORT") != NULL) { + (void) printf("dumping core by request\n"); + abort(); + } + + return (ret); +} diff --git a/cmd/zpool/zpool_util.c b/cmd/zpool/zpool_util.c new file mode 100644 index 000000000000..c7a002efb17c --- /dev/null +++ b/cmd/zpool/zpool_util.c @@ -0,0 +1,86 @@ +/* + * 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 http://www.opensolaris.org/os/licensing. + * 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 2009 Sun Microsystems, Inc. All rights reserved. + * Use is subject to license terms. + */ + +#include <errno.h> +#include <libgen.h> +#include <libintl.h> +#include <stdio.h> +#include <stdlib.h> +#include <strings.h> + +#include "zpool_util.h" + +/* + * Utility function to guarantee malloc() success. + */ +void * +safe_malloc(size_t size) +{ + void *data; + + if ((data = calloc(1, size)) == NULL) { + (void) fprintf(stderr, "internal error: out of memory\n"); + exit(1); + } + + return (data); +} + +/* + * Display an out of memory error message and abort the current program. + */ +void +zpool_no_memory(void) +{ + assert(errno == ENOMEM); + (void) fprintf(stderr, + gettext("internal error: out of memory\n")); + exit(1); +} + +/* + * Return the number of logs in supplied nvlist + */ +uint_t +num_logs(nvlist_t *nv) +{ + uint_t nlogs = 0; + uint_t c, children; + nvlist_t **child; + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_CHILDREN, + &child, &children) != 0) + return (0); + + for (c = 0; c < children; c++) { + uint64_t is_log = B_FALSE; + + (void) nvlist_lookup_uint64(child[c], ZPOOL_CONFIG_IS_LOG, + &is_log); + if (is_log) + nlogs++; + } + return (nlogs); +} diff --git a/cmd/zpool/zpool_util.h b/cmd/zpool/zpool_util.h new file mode 100644 index 000000000000..134c730fcf8e --- /dev/null +++ b/cmd/zpool/zpool_util.h @@ -0,0 +1,72 @@ +/* + * 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 http://www.opensolaris.org/os/licensing. + * 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. + */ + +#ifndef ZPOOL_UTIL_H +#define ZPOOL_UTIL_H + +#include <libnvpair.h> +#include <libzfs.h> + +#ifdef __cplusplus +extern "C" { +#endif + +/* + * Basic utility functions + */ +void *safe_malloc(size_t); +void zpool_no_memory(void); +uint_t num_logs(nvlist_t *nv); + +/* + * Virtual device functions + */ + +nvlist_t *make_root_vdev(zpool_handle_t *zhp, int force, int check_rep, + boolean_t replacing, boolean_t dryrun, int argc, char **argv); +nvlist_t *split_mirror_vdev(zpool_handle_t *zhp, char *newname, + nvlist_t *props, splitflags_t flags, int argc, char **argv); + +/* + * Pool list functions + */ +int for_each_pool(int, char **, boolean_t unavail, zprop_list_t **, + zpool_iter_f, void *); + +typedef struct zpool_list zpool_list_t; + +zpool_list_t *pool_list_get(int, char **, zprop_list_t **, int *); +void pool_list_update(zpool_list_t *); +int pool_list_iter(zpool_list_t *, int unavail, zpool_iter_f, void *); +void pool_list_free(zpool_list_t *); +int pool_list_count(zpool_list_t *); +void pool_list_remove(zpool_list_t *, zpool_handle_t *); + +libzfs_handle_t *g_zfs; + +#ifdef __cplusplus +} +#endif + +#endif /* ZPOOL_UTIL_H */ diff --git a/cmd/zpool/zpool_vdev.c b/cmd/zpool/zpool_vdev.c new file mode 100644 index 000000000000..53c2e60b7dd5 --- /dev/null +++ b/cmd/zpool/zpool_vdev.c @@ -0,0 +1,1469 @@ +/* + * 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 http://www.opensolaris.org/os/licensing. + * 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. + */ + +/* + * Functions to convert between a list of vdevs and an nvlist representing the + * configuration. Each entry in the list can be one of: + * + * Device vdevs + * disk=(path=..., devid=...) + * file=(path=...) + * + * Group vdevs + * raidz[1|2]=(...) + * mirror=(...) + * + * Hot spares + * + * While the underlying implementation supports it, group vdevs cannot contain + * other group vdevs. All userland verification of devices is contained within + * this file. If successful, the nvlist returned can be passed directly to the + * kernel; we've done as much verification as possible in userland. + * + * Hot spares are a special case, and passed down as an array of disk vdevs, at + * the same level as the root of the vdev tree. + * + * The only function exported by this file is 'make_root_vdev'. The + * function performs several passes: + * + * 1. Construct the vdev specification. Performs syntax validation and + * makes sure each device is valid. + * 2. Check for devices in use. Using libdiskmgt, makes sure that no + * devices are also in use. Some can be overridden using the 'force' + * flag, others cannot. + * 3. Check for replication errors if the 'force' flag is not specified. + * validates that the replication level is consistent across the + * entire pool. + * 4. Call libzfs to label any whole disks with an EFI label. + */ + +#include <assert.h> +#include <devid.h> +#include <errno.h> +#include <fcntl.h> +#include <libdiskmgt.h> +#include <libintl.h> +#include <libnvpair.h> +#include <limits.h> +#include <stdio.h> +#include <string.h> +#include <unistd.h> +#include <sys/efi_partition.h> +#include <sys/stat.h> +#include <sys/vtoc.h> +#include <sys/mntent.h> + +#include "zpool_util.h" + +#define DISK_ROOT "/dev/dsk" +#define RDISK_ROOT "/dev/rdsk" +#define BACKUP_SLICE "s2" + +/* + * For any given vdev specification, we can have multiple errors. The + * vdev_error() function keeps track of whether we have seen an error yet, and + * prints out a header if its the first error we've seen. + */ +boolean_t error_seen; +boolean_t is_force; + +/*PRINTFLIKE1*/ +static void +vdev_error(const char *fmt, ...) +{ + va_list ap; + + if (!error_seen) { + (void) fprintf(stderr, gettext("invalid vdev specification\n")); + if (!is_force) + (void) fprintf(stderr, gettext("use '-f' to override " + "the following errors:\n")); + else + (void) fprintf(stderr, gettext("the following errors " + "must be manually repaired:\n")); + error_seen = B_TRUE; + } + + va_start(ap, fmt); + (void) vfprintf(stderr, fmt, ap); + va_end(ap); +} + +static void +libdiskmgt_error(int error) +{ + /* + * ENXIO/ENODEV is a valid error message if the device doesn't live in + * /dev/dsk. Don't bother printing an error message in this case. + */ + if (error == ENXIO || error == ENODEV) + return; + + (void) fprintf(stderr, gettext("warning: device in use checking " + "failed: %s\n"), strerror(error)); +} + +/* + * Validate a device, passing the bulk of the work off to libdiskmgt. + */ +static int +check_slice(const char *path, int force, boolean_t wholedisk, boolean_t isspare) +{ + char *msg; + int error = 0; + dm_who_type_t who; + + if (force) + who = DM_WHO_ZPOOL_FORCE; + else if (isspare) + who = DM_WHO_ZPOOL_SPARE; + else + who = DM_WHO_ZPOOL; + + if (dm_inuse((char *)path, &msg, who, &error) || error) { + if (error != 0) { + libdiskmgt_error(error); + return (0); + } else { + vdev_error("%s", msg); + free(msg); + return (-1); + } + } + + /* + * If we're given a whole disk, ignore overlapping slices since we're + * about to label it anyway. + */ + error = 0; + if (!wholedisk && !force && + (dm_isoverlapping((char *)path, &msg, &error) || error)) { + if (error == 0) { + /* dm_isoverlapping returned -1 */ + vdev_error(gettext("%s overlaps with %s\n"), path, msg); + free(msg); + return (-1); + } else if (error != ENODEV) { + /* libdiskmgt's devcache only handles physical drives */ + libdiskmgt_error(error); + return (0); + } + } + + return (0); +} + + +/* + * Validate a whole disk. Iterate over all slices on the disk and make sure + * that none is in use by calling check_slice(). + */ +static int +check_disk(const char *name, dm_descriptor_t disk, int force, int isspare) +{ + dm_descriptor_t *drive, *media, *slice; + int err = 0; + int i; + int ret; + + /* + * Get the drive associated with this disk. This should never fail, + * because we already have an alias handle open for the device. + */ + if ((drive = dm_get_associated_descriptors(disk, DM_DRIVE, + &err)) == NULL || *drive == NULL) { + if (err) + libdiskmgt_error(err); + return (0); + } + + if ((media = dm_get_associated_descriptors(*drive, DM_MEDIA, + &err)) == NULL) { + dm_free_descriptors(drive); + if (err) + libdiskmgt_error(err); + return (0); + } + + dm_free_descriptors(drive); + + /* + * It is possible that the user has specified a removable media drive, + * and the media is not present. + */ + if (*media == NULL) { + dm_free_descriptors(media); + vdev_error(gettext("'%s' has no media in drive\n"), name); + return (-1); + } + + if ((slice = dm_get_associated_descriptors(*media, DM_SLICE, + &err)) == NULL) { + dm_free_descriptors(media); + if (err) + libdiskmgt_error(err); + return (0); + } + + dm_free_descriptors(media); + + ret = 0; + + /* + * Iterate over all slices and report any errors. We don't care about + * overlapping slices because we are using the whole disk. + */ + for (i = 0; slice[i] != NULL; i++) { + char *name = dm_get_name(slice[i], &err); + + if (check_slice(name, force, B_TRUE, isspare) != 0) + ret = -1; + + dm_free_name(name); + } + + dm_free_descriptors(slice); + return (ret); +} + +/* + * Validate a device. + */ +static int +check_device(const char *path, boolean_t force, boolean_t isspare) +{ + dm_descriptor_t desc; + int err; + char *dev; + + /* + * For whole disks, libdiskmgt does not include the leading dev path. + */ + dev = strrchr(path, '/'); + assert(dev != NULL); + dev++; + if ((desc = dm_get_descriptor_by_name(DM_ALIAS, dev, &err)) != NULL) { + err = check_disk(path, desc, force, isspare); + dm_free_descriptor(desc); + return (err); + } + + return (check_slice(path, force, B_FALSE, isspare)); +} + +/* + * Check that a file is valid. All we can do in this case is check that it's + * not in use by another pool, and not in use by swap. + */ +static int +check_file(const char *file, boolean_t force, boolean_t isspare) +{ + char *name; + int fd; + int ret = 0; + int err; + pool_state_t state; + boolean_t inuse; + + if (dm_inuse_swap(file, &err)) { + if (err) + libdiskmgt_error(err); + else + vdev_error(gettext("%s is currently used by swap. " + "Please see swap(1M).\n"), file); + return (-1); + } + + if ((fd = open(file, O_RDONLY)) < 0) + return (0); + + if (zpool_in_use(g_zfs, fd, &state, &name, &inuse) == 0 && inuse) { + const char *desc; + + switch (state) { + case POOL_STATE_ACTIVE: + desc = gettext("active"); + break; + + case POOL_STATE_EXPORTED: + desc = gettext("exported"); + break; + + case POOL_STATE_POTENTIALLY_ACTIVE: + desc = gettext("potentially active"); + break; + + default: + desc = gettext("unknown"); + break; + } + + /* + * Allow hot spares to be shared between pools. + */ + if (state == POOL_STATE_SPARE && isspare) + return (0); + + if (state == POOL_STATE_ACTIVE || + state == POOL_STATE_SPARE || !force) { + switch (state) { + case POOL_STATE_SPARE: + vdev_error(gettext("%s is reserved as a hot " + "spare for pool %s\n"), file, name); + break; + default: + vdev_error(gettext("%s is part of %s pool " + "'%s'\n"), file, desc, name); + break; + } + ret = -1; + } + + free(name); + } + + (void) close(fd); + return (ret); +} + + +/* + * By "whole disk" we mean an entire physical disk (something we can + * label, toggle the write cache on, etc.) as opposed to the full + * capacity of a pseudo-device such as lofi or did. We act as if we + * are labeling the disk, which should be a pretty good test of whether + * it's a viable device or not. Returns B_TRUE if it is and B_FALSE if + * it isn't. + */ +static boolean_t +is_whole_disk(const char *arg) +{ + struct dk_gpt *label; + int fd; + char path[MAXPATHLEN]; + + (void) snprintf(path, sizeof (path), "%s%s%s", + RDISK_ROOT, strrchr(arg, '/'), BACKUP_SLICE); + if ((fd = open(path, O_RDWR | O_NDELAY)) < 0) + return (B_FALSE); + if (efi_alloc_and_init(fd, EFI_NUMPAR, &label) != 0) { + (void) close(fd); + return (B_FALSE); + } + efi_free(label); + (void) close(fd); + return (B_TRUE); +} + +/* + * Create a leaf vdev. Determine if this is a file or a device. If it's a + * device, fill in the device id to make a complete nvlist. Valid forms for a + * leaf vdev are: + * + * /dev/dsk/xxx Complete disk path + * /xxx Full path to file + * xxx Shorthand for /dev/dsk/xxx + */ +static nvlist_t * +make_leaf_vdev(const char *arg, uint64_t is_log) +{ + char path[MAXPATHLEN]; + struct stat64 statbuf; + nvlist_t *vdev = NULL; + char *type = NULL; + boolean_t wholedisk = B_FALSE; + + /* + * Determine what type of vdev this is, and put the full path into + * 'path'. We detect whether this is a device of file afterwards by + * checking the st_mode of the file. + */ + if (arg[0] == '/') { + /* + * Complete device or file path. Exact type is determined by + * examining the file descriptor afterwards. + */ + wholedisk = is_whole_disk(arg); + if (!wholedisk && (stat64(arg, &statbuf) != 0)) { + (void) fprintf(stderr, + gettext("cannot open '%s': %s\n"), + arg, strerror(errno)); + return (NULL); + } + + (void) strlcpy(path, arg, sizeof (path)); + } else { + /* + * This may be a short path for a device, or it could be total + * gibberish. Check to see if it's a known device in + * /dev/dsk/. As part of this check, see if we've been given a + * an entire disk (minus the slice number). + */ + (void) snprintf(path, sizeof (path), "%s/%s", DISK_ROOT, + arg); + wholedisk = is_whole_disk(path); + if (!wholedisk && (stat64(path, &statbuf) != 0)) { + /* + * If we got ENOENT, then the user gave us + * gibberish, so try to direct them with a + * reasonable error message. Otherwise, + * regurgitate strerror() since it's the best we + * can do. + */ + if (errno == ENOENT) { + (void) fprintf(stderr, + gettext("cannot open '%s': no such " + "device in %s\n"), arg, DISK_ROOT); + (void) fprintf(stderr, + gettext("must be a full path or " + "shorthand device name\n")); + return (NULL); + } else { + (void) fprintf(stderr, + gettext("cannot open '%s': %s\n"), + path, strerror(errno)); + return (NULL); + } + } + } + + /* + * Determine whether this is a device or a file. + */ + if (wholedisk || S_ISBLK(statbuf.st_mode)) { + type = VDEV_TYPE_DISK; + } else if (S_ISREG(statbuf.st_mode)) { + type = VDEV_TYPE_FILE; + } else { + (void) fprintf(stderr, gettext("cannot use '%s': must be a " + "block device or regular file\n"), path); + return (NULL); + } + + /* + * Finally, we have the complete device or file, and we know that it is + * acceptable to use. Construct the nvlist to describe this vdev. All + * vdevs have a 'path' element, and devices also have a 'devid' element. + */ + verify(nvlist_alloc(&vdev, NV_UNIQUE_NAME, 0) == 0); + verify(nvlist_add_string(vdev, ZPOOL_CONFIG_PATH, path) == 0); + verify(nvlist_add_string(vdev, ZPOOL_CONFIG_TYPE, type) == 0); + verify(nvlist_add_uint64(vdev, ZPOOL_CONFIG_IS_LOG, is_log) == 0); + if (strcmp(type, VDEV_TYPE_DISK) == 0) + verify(nvlist_add_uint64(vdev, ZPOOL_CONFIG_WHOLE_DISK, + (uint64_t)wholedisk) == 0); + + /* + * For a whole disk, defer getting its devid until after labeling it. + */ + if (S_ISBLK(statbuf.st_mode) && !wholedisk) { + /* + * Get the devid for the device. + */ + int fd; + ddi_devid_t devid; + char *minor = NULL, *devid_str = NULL; + + if ((fd = open(path, O_RDONLY)) < 0) { + (void) fprintf(stderr, gettext("cannot open '%s': " + "%s\n"), path, strerror(errno)); + nvlist_free(vdev); + return (NULL); + } + + if (devid_get(fd, &devid) == 0) { + if (devid_get_minor_name(fd, &minor) == 0 && + (devid_str = devid_str_encode(devid, minor)) != + NULL) { + verify(nvlist_add_string(vdev, + ZPOOL_CONFIG_DEVID, devid_str) == 0); + } + if (devid_str != NULL) + devid_str_free(devid_str); + if (minor != NULL) + devid_str_free(minor); + devid_free(devid); + } + + (void) close(fd); + } + + return (vdev); +} + +/* + * Go through and verify the replication level of the pool is consistent. + * Performs the following checks: + * + * For the new spec, verifies that devices in mirrors and raidz are the + * same size. + * + * If the current configuration already has inconsistent replication + * levels, ignore any other potential problems in the new spec. + * + * Otherwise, make sure that the current spec (if there is one) and the new + * spec have consistent replication levels. + */ +typedef struct replication_level { + char *zprl_type; + uint64_t zprl_children; + uint64_t zprl_parity; +} replication_level_t; + +#define ZPOOL_FUZZ (16 * 1024 * 1024) + +/* + * Given a list of toplevel vdevs, return the current replication level. If + * the config is inconsistent, then NULL is returned. If 'fatal' is set, then + * an error message will be displayed for each self-inconsistent vdev. + */ +static replication_level_t * +get_replication(nvlist_t *nvroot, boolean_t fatal) +{ + nvlist_t **top; + uint_t t, toplevels; + nvlist_t **child; + uint_t c, children; + nvlist_t *nv; + char *type; + replication_level_t lastrep, rep, *ret; + boolean_t dontreport; + + ret = safe_malloc(sizeof (replication_level_t)); + + verify(nvlist_lookup_nvlist_array(nvroot, ZPOOL_CONFIG_CHILDREN, + &top, &toplevels) == 0); + + lastrep.zprl_type = NULL; + for (t = 0; t < toplevels; t++) { + uint64_t is_log = B_FALSE; + + nv = top[t]; + + /* + * For separate logs we ignore the top level vdev replication + * constraints. + */ + (void) nvlist_lookup_uint64(nv, ZPOOL_CONFIG_IS_LOG, &is_log); + if (is_log) + continue; + + verify(nvlist_lookup_string(nv, ZPOOL_CONFIG_TYPE, + &type) == 0); + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_CHILDREN, + &child, &children) != 0) { + /* + * This is a 'file' or 'disk' vdev. + */ + rep.zprl_type = type; + rep.zprl_children = 1; + rep.zprl_parity = 0; + } else { + uint64_t vdev_size; + + /* + * This is a mirror or RAID-Z vdev. Go through and make + * sure the contents are all the same (files vs. disks), + * keeping track of the number of elements in the + * process. + * + * We also check that the size of each vdev (if it can + * be determined) is the same. + */ + rep.zprl_type = type; + rep.zprl_children = 0; + + if (strcmp(type, VDEV_TYPE_RAIDZ) == 0) { + verify(nvlist_lookup_uint64(nv, + ZPOOL_CONFIG_NPARITY, + &rep.zprl_parity) == 0); + assert(rep.zprl_parity != 0); + } else { + rep.zprl_parity = 0; + } + + /* + * The 'dontreport' variable indicates that we've + * already reported an error for this spec, so don't + * bother doing it again. + */ + type = NULL; + dontreport = 0; + vdev_size = -1ULL; + for (c = 0; c < children; c++) { + nvlist_t *cnv = child[c]; + char *path; + struct stat64 statbuf; + uint64_t size = -1ULL; + char *childtype; + int fd, err; + + rep.zprl_children++; + + verify(nvlist_lookup_string(cnv, + ZPOOL_CONFIG_TYPE, &childtype) == 0); + + /* + * If this is a replacing or spare vdev, then + * get the real first child of the vdev. + */ + if (strcmp(childtype, + VDEV_TYPE_REPLACING) == 0 || + strcmp(childtype, VDEV_TYPE_SPARE) == 0) { + nvlist_t **rchild; + uint_t rchildren; + + verify(nvlist_lookup_nvlist_array(cnv, + ZPOOL_CONFIG_CHILDREN, &rchild, + &rchildren) == 0); + assert(rchildren == 2); + cnv = rchild[0]; + + verify(nvlist_lookup_string(cnv, + ZPOOL_CONFIG_TYPE, + &childtype) == 0); + } + + verify(nvlist_lookup_string(cnv, + ZPOOL_CONFIG_PATH, &path) == 0); + + /* + * If we have a raidz/mirror that combines disks + * with files, report it as an error. + */ + if (!dontreport && type != NULL && + strcmp(type, childtype) != 0) { + if (ret != NULL) + free(ret); + ret = NULL; + if (fatal) + vdev_error(gettext( + "mismatched replication " + "level: %s contains both " + "files and devices\n"), + rep.zprl_type); + else + return (NULL); + dontreport = B_TRUE; + } + + /* + * According to stat(2), the value of 'st_size' + * is undefined for block devices and character + * devices. But there is no effective way to + * determine the real size in userland. + * + * Instead, we'll take advantage of an + * implementation detail of spec_size(). If the + * device is currently open, then we (should) + * return a valid size. + * + * If we still don't get a valid size (indicated + * by a size of 0 or MAXOFFSET_T), then ignore + * this device altogether. + */ + if ((fd = open(path, O_RDONLY)) >= 0) { + err = fstat64(fd, &statbuf); + (void) close(fd); + } else { + err = stat64(path, &statbuf); + } + + if (err != 0 || + statbuf.st_size == 0 || + statbuf.st_size == MAXOFFSET_T) + continue; + + size = statbuf.st_size; + + /* + * Also make sure that devices and + * slices have a consistent size. If + * they differ by a significant amount + * (~16MB) then report an error. + */ + if (!dontreport && + (vdev_size != -1ULL && + (labs(size - vdev_size) > + ZPOOL_FUZZ))) { + if (ret != NULL) + free(ret); + ret = NULL; + if (fatal) + vdev_error(gettext( + "%s contains devices of " + "different sizes\n"), + rep.zprl_type); + else + return (NULL); + dontreport = B_TRUE; + } + + type = childtype; + vdev_size = size; + } + } + + /* + * At this point, we have the replication of the last toplevel + * vdev in 'rep'. Compare it to 'lastrep' to see if its + * different. + */ + if (lastrep.zprl_type != NULL) { + if (strcmp(lastrep.zprl_type, rep.zprl_type) != 0) { + if (ret != NULL) + free(ret); + ret = NULL; + if (fatal) + vdev_error(gettext( + "mismatched replication level: " + "both %s and %s vdevs are " + "present\n"), + lastrep.zprl_type, rep.zprl_type); + else + return (NULL); + } else if (lastrep.zprl_parity != rep.zprl_parity) { + if (ret) + free(ret); + ret = NULL; + if (fatal) + vdev_error(gettext( + "mismatched replication level: " + "both %llu and %llu device parity " + "%s vdevs are present\n"), + lastrep.zprl_parity, + rep.zprl_parity, + rep.zprl_type); + else + return (NULL); + } else if (lastrep.zprl_children != rep.zprl_children) { + if (ret) + free(ret); + ret = NULL; + if (fatal) + vdev_error(gettext( + "mismatched replication level: " + "both %llu-way and %llu-way %s " + "vdevs are present\n"), + lastrep.zprl_children, + rep.zprl_children, + rep.zprl_type); + else + return (NULL); + } + } + lastrep = rep; + } + + if (ret != NULL) + *ret = rep; + + return (ret); +} + +/* + * Check the replication level of the vdev spec against the current pool. Calls + * get_replication() to make sure the new spec is self-consistent. If the pool + * has a consistent replication level, then we ignore any errors. Otherwise, + * report any difference between the two. + */ +static int +check_replication(nvlist_t *config, nvlist_t *newroot) +{ + nvlist_t **child; + uint_t children; + replication_level_t *current = NULL, *new; + int ret; + + /* + * If we have a current pool configuration, check to see if it's + * self-consistent. If not, simply return success. + */ + if (config != NULL) { + nvlist_t *nvroot; + + verify(nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, + &nvroot) == 0); + if ((current = get_replication(nvroot, B_FALSE)) == NULL) + return (0); + } + /* + * for spares there may be no children, and therefore no + * replication level to check + */ + if ((nvlist_lookup_nvlist_array(newroot, ZPOOL_CONFIG_CHILDREN, + &child, &children) != 0) || (children == 0)) { + free(current); + return (0); + } + + /* + * If all we have is logs then there's no replication level to check. + */ + if (num_logs(newroot) == children) { + free(current); + return (0); + } + + /* + * Get the replication level of the new vdev spec, reporting any + * inconsistencies found. + */ + if ((new = get_replication(newroot, B_TRUE)) == NULL) { + free(current); + return (-1); + } + + /* + * Check to see if the new vdev spec matches the replication level of + * the current pool. + */ + ret = 0; + if (current != NULL) { + if (strcmp(current->zprl_type, new->zprl_type) != 0) { + vdev_error(gettext( + "mismatched replication level: pool uses %s " + "and new vdev is %s\n"), + current->zprl_type, new->zprl_type); + ret = -1; + } else if (current->zprl_parity != new->zprl_parity) { + vdev_error(gettext( + "mismatched replication level: pool uses %llu " + "device parity and new vdev uses %llu\n"), + current->zprl_parity, new->zprl_parity); + ret = -1; + } else if (current->zprl_children != new->zprl_children) { + vdev_error(gettext( + "mismatched replication level: pool uses %llu-way " + "%s and new vdev uses %llu-way %s\n"), + current->zprl_children, current->zprl_type, + new->zprl_children, new->zprl_type); + ret = -1; + } + } + + free(new); + if (current != NULL) + free(current); + + return (ret); +} + +/* + * Go through and find any whole disks in the vdev specification, labelling them + * as appropriate. When constructing the vdev spec, we were unable to open this + * device in order to provide a devid. Now that we have labelled the disk and + * know that slice 0 is valid, we can construct the devid now. + * + * If the disk was already labeled with an EFI label, we will have gotten the + * devid already (because we were able to open the whole disk). Otherwise, we + * need to get the devid after we label the disk. + */ +static int +make_disks(zpool_handle_t *zhp, nvlist_t *nv) +{ + nvlist_t **child; + uint_t c, children; + char *type, *path, *diskname; + char buf[MAXPATHLEN]; + uint64_t wholedisk; + int fd; + int ret; + ddi_devid_t devid; + char *minor = NULL, *devid_str = NULL; + + verify(nvlist_lookup_string(nv, ZPOOL_CONFIG_TYPE, &type) == 0); + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_CHILDREN, + &child, &children) != 0) { + + if (strcmp(type, VDEV_TYPE_DISK) != 0) + return (0); + + /* + * We have a disk device. Get the path to the device + * and see if it's a whole disk by appending the backup + * slice and stat()ing the device. + */ + verify(nvlist_lookup_string(nv, ZPOOL_CONFIG_PATH, &path) == 0); + if (nvlist_lookup_uint64(nv, ZPOOL_CONFIG_WHOLE_DISK, + &wholedisk) != 0 || !wholedisk) + return (0); + + diskname = strrchr(path, '/'); + assert(diskname != NULL); + diskname++; + if (zpool_label_disk(g_zfs, zhp, diskname) == -1) + return (-1); + + /* + * Fill in the devid, now that we've labeled the disk. + */ + (void) snprintf(buf, sizeof (buf), "%ss0", path); + if ((fd = open(buf, O_RDONLY)) < 0) { + (void) fprintf(stderr, + gettext("cannot open '%s': %s\n"), + buf, strerror(errno)); + return (-1); + } + + if (devid_get(fd, &devid) == 0) { + if (devid_get_minor_name(fd, &minor) == 0 && + (devid_str = devid_str_encode(devid, minor)) != + NULL) { + verify(nvlist_add_string(nv, + ZPOOL_CONFIG_DEVID, devid_str) == 0); + } + if (devid_str != NULL) + devid_str_free(devid_str); + if (minor != NULL) + devid_str_free(minor); + devid_free(devid); + } + + /* + * Update the path to refer to the 's0' slice. The presence of + * the 'whole_disk' field indicates to the CLI that we should + * chop off the slice number when displaying the device in + * future output. + */ + verify(nvlist_add_string(nv, ZPOOL_CONFIG_PATH, buf) == 0); + + (void) close(fd); + + return (0); + } + + for (c = 0; c < children; c++) + if ((ret = make_disks(zhp, child[c])) != 0) + return (ret); + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_SPARES, + &child, &children) == 0) + for (c = 0; c < children; c++) + if ((ret = make_disks(zhp, child[c])) != 0) + return (ret); + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_L2CACHE, + &child, &children) == 0) + for (c = 0; c < children; c++) + if ((ret = make_disks(zhp, child[c])) != 0) + return (ret); + + return (0); +} + +/* + * Determine if the given path is a hot spare within the given configuration. + */ +static boolean_t +is_spare(nvlist_t *config, const char *path) +{ + int fd; + pool_state_t state; + char *name = NULL; + nvlist_t *label; + uint64_t guid, spareguid; + nvlist_t *nvroot; + nvlist_t **spares; + uint_t i, nspares; + boolean_t inuse; + + if ((fd = open(path, O_RDONLY)) < 0) + return (B_FALSE); + + if (zpool_in_use(g_zfs, fd, &state, &name, &inuse) != 0 || + !inuse || + state != POOL_STATE_SPARE || + zpool_read_label(fd, &label) != 0) { + free(name); + (void) close(fd); + return (B_FALSE); + } + free(name); + (void) close(fd); + + verify(nvlist_lookup_uint64(label, ZPOOL_CONFIG_GUID, &guid) == 0); + nvlist_free(label); + + verify(nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, + &nvroot) == 0); + if (nvlist_lookup_nvlist_array(nvroot, ZPOOL_CONFIG_SPARES, + &spares, &nspares) == 0) { + for (i = 0; i < nspares; i++) { + verify(nvlist_lookup_uint64(spares[i], + ZPOOL_CONFIG_GUID, &spareguid) == 0); + if (spareguid == guid) + return (B_TRUE); + } + } + + return (B_FALSE); +} + +/* + * Go through and find any devices that are in use. We rely on libdiskmgt for + * the majority of this task. + */ +static int +check_in_use(nvlist_t *config, nvlist_t *nv, boolean_t force, + boolean_t replacing, boolean_t isspare) +{ + nvlist_t **child; + uint_t c, children; + char *type, *path; + int ret; + char buf[MAXPATHLEN]; + uint64_t wholedisk; + + verify(nvlist_lookup_string(nv, ZPOOL_CONFIG_TYPE, &type) == 0); + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_CHILDREN, + &child, &children) != 0) { + + verify(nvlist_lookup_string(nv, ZPOOL_CONFIG_PATH, &path) == 0); + + /* + * As a generic check, we look to see if this is a replace of a + * hot spare within the same pool. If so, we allow it + * regardless of what libdiskmgt or zpool_in_use() says. + */ + if (replacing) { + if (nvlist_lookup_uint64(nv, ZPOOL_CONFIG_WHOLE_DISK, + &wholedisk) == 0 && wholedisk) + (void) snprintf(buf, sizeof (buf), "%ss0", + path); + else + (void) strlcpy(buf, path, sizeof (buf)); + + if (is_spare(config, buf)) + return (0); + } + + if (strcmp(type, VDEV_TYPE_DISK) == 0) + ret = check_device(path, force, isspare); + + if (strcmp(type, VDEV_TYPE_FILE) == 0) + ret = check_file(path, force, isspare); + + return (ret); + } + + for (c = 0; c < children; c++) + if ((ret = check_in_use(config, child[c], force, + replacing, B_FALSE)) != 0) + return (ret); + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_SPARES, + &child, &children) == 0) + for (c = 0; c < children; c++) + if ((ret = check_in_use(config, child[c], force, + replacing, B_TRUE)) != 0) + return (ret); + + if (nvlist_lookup_nvlist_array(nv, ZPOOL_CONFIG_L2CACHE, + &child, &children) == 0) + for (c = 0; c < children; c++) + if ((ret = check_in_use(config, child[c], force, + replacing, B_FALSE)) != 0) + return (ret); + + return (0); +} + +static const char * +is_grouping(const char *type, int *mindev, int *maxdev) +{ + if (strncmp(type, "raidz", 5) == 0) { + const char *p = type + 5; + char *end; + long nparity; + + if (*p == '\0') { + nparity = 1; + } else if (*p == '0') { + return (NULL); /* no zero prefixes allowed */ + } else { + errno = 0; + nparity = strtol(p, &end, 10); + if (errno != 0 || nparity < 1 || nparity >= 255 || + *end != '\0') + return (NULL); + } + + if (mindev != NULL) + *mindev = nparity + 1; + if (maxdev != NULL) + *maxdev = 255; + return (VDEV_TYPE_RAIDZ); + } + + if (maxdev != NULL) + *maxdev = INT_MAX; + + if (strcmp(type, "mirror") == 0) { + if (mindev != NULL) + *mindev = 2; + return (VDEV_TYPE_MIRROR); + } + + if (strcmp(type, "spare") == 0) { + if (mindev != NULL) + *mindev = 1; + return (VDEV_TYPE_SPARE); + } + + if (strcmp(type, "log") == 0) { + if (mindev != NULL) + *mindev = 1; + return (VDEV_TYPE_LOG); + } + + if (strcmp(type, "cache") == 0) { + if (mindev != NULL) + *mindev = 1; + return (VDEV_TYPE_L2CACHE); + } + + return (NULL); +} + +/* + * Construct a syntactically valid vdev specification, + * and ensure that all devices and files exist and can be opened. + * Note: we don't bother freeing anything in the error paths + * because the program is just going to exit anyway. + */ +nvlist_t * +construct_spec(int argc, char **argv) +{ + nvlist_t *nvroot, *nv, **top, **spares, **l2cache; + int t, toplevels, mindev, maxdev, nspares, nlogs, nl2cache; + const char *type; + uint64_t is_log; + boolean_t seen_logs; + + top = NULL; + toplevels = 0; + spares = NULL; + l2cache = NULL; + nspares = 0; + nlogs = 0; + nl2cache = 0; + is_log = B_FALSE; + seen_logs = B_FALSE; + + while (argc > 0) { + nv = NULL; + + /* + * If it's a mirror or raidz, the subsequent arguments are + * its leaves -- until we encounter the next mirror or raidz. + */ + if ((type = is_grouping(argv[0], &mindev, &maxdev)) != NULL) { + nvlist_t **child = NULL; + int c, children = 0; + + if (strcmp(type, VDEV_TYPE_SPARE) == 0) { + if (spares != NULL) { + (void) fprintf(stderr, + gettext("invalid vdev " + "specification: 'spare' can be " + "specified only once\n")); + return (NULL); + } + is_log = B_FALSE; + } + + if (strcmp(type, VDEV_TYPE_LOG) == 0) { + if (seen_logs) { + (void) fprintf(stderr, + gettext("invalid vdev " + "specification: 'log' can be " + "specified only once\n")); + return (NULL); + } + seen_logs = B_TRUE; + is_log = B_TRUE; + argc--; + argv++; + /* + * A log is not a real grouping device. + * We just set is_log and continue. + */ + continue; + } + + if (strcmp(type, VDEV_TYPE_L2CACHE) == 0) { + if (l2cache != NULL) { + (void) fprintf(stderr, + gettext("invalid vdev " + "specification: 'cache' can be " + "specified only once\n")); + return (NULL); + } + is_log = B_FALSE; + } + + if (is_log) { + if (strcmp(type, VDEV_TYPE_MIRROR) != 0) { + (void) fprintf(stderr, + gettext("invalid vdev " + "specification: unsupported 'log' " + "device: %s\n"), type); + return (NULL); + } + nlogs++; + } + + for (c = 1; c < argc; c++) { + if (is_grouping(argv[c], NULL, NULL) != NULL) + break; + children++; + child = realloc(child, + children * sizeof (nvlist_t *)); + if (child == NULL) + zpool_no_memory(); + if ((nv = make_leaf_vdev(argv[c], B_FALSE)) + == NULL) + return (NULL); + child[children - 1] = nv; + } + + if (children < mindev) { + (void) fprintf(stderr, gettext("invalid vdev " + "specification: %s requires at least %d " + "devices\n"), argv[0], mindev); + return (NULL); + } + + if (children > maxdev) { + (void) fprintf(stderr, gettext("invalid vdev " + "specification: %s supports no more than " + "%d devices\n"), argv[0], maxdev); + return (NULL); + } + + argc -= c; + argv += c; + + if (strcmp(type, VDEV_TYPE_SPARE) == 0) { + spares = child; + nspares = children; + continue; + } else if (strcmp(type, VDEV_TYPE_L2CACHE) == 0) { + l2cache = child; + nl2cache = children; + continue; + } else { + verify(nvlist_alloc(&nv, NV_UNIQUE_NAME, + 0) == 0); + verify(nvlist_add_string(nv, ZPOOL_CONFIG_TYPE, + type) == 0); + verify(nvlist_add_uint64(nv, + ZPOOL_CONFIG_IS_LOG, is_log) == 0); + if (strcmp(type, VDEV_TYPE_RAIDZ) == 0) { + verify(nvlist_add_uint64(nv, + ZPOOL_CONFIG_NPARITY, + mindev - 1) == 0); + } + verify(nvlist_add_nvlist_array(nv, + ZPOOL_CONFIG_CHILDREN, child, + children) == 0); + + for (c = 0; c < children; c++) + nvlist_free(child[c]); + free(child); + } + } else { + /* + * We have a device. Pass off to make_leaf_vdev() to + * construct the appropriate nvlist describing the vdev. + */ + if ((nv = make_leaf_vdev(argv[0], is_log)) == NULL) + return (NULL); + if (is_log) + nlogs++; + argc--; + argv++; + } + + toplevels++; + top = realloc(top, toplevels * sizeof (nvlist_t *)); + if (top == NULL) + zpool_no_memory(); + top[toplevels - 1] = nv; + } + + if (toplevels == 0 && nspares == 0 && nl2cache == 0) { + (void) fprintf(stderr, gettext("invalid vdev " + "specification: at least one toplevel vdev must be " + "specified\n")); + return (NULL); + } + + if (seen_logs && nlogs == 0) { + (void) fprintf(stderr, gettext("invalid vdev specification: " + "log requires at least 1 device\n")); + return (NULL); + } + + /* + * Finally, create nvroot and add all top-level vdevs to it. + */ + verify(nvlist_alloc(&nvroot, NV_UNIQUE_NAME, 0) == 0); + verify(nvlist_add_string(nvroot, ZPOOL_CONFIG_TYPE, + VDEV_TYPE_ROOT) == 0); + verify(nvlist_add_nvlist_array(nvroot, ZPOOL_CONFIG_CHILDREN, + top, toplevels) == 0); + if (nspares != 0) + verify(nvlist_add_nvlist_array(nvroot, ZPOOL_CONFIG_SPARES, + spares, nspares) == 0); + if (nl2cache != 0) + verify(nvlist_add_nvlist_array(nvroot, ZPOOL_CONFIG_L2CACHE, + l2cache, nl2cache) == 0); + + for (t = 0; t < toplevels; t++) + nvlist_free(top[t]); + for (t = 0; t < nspares; t++) + nvlist_free(spares[t]); + for (t = 0; t < nl2cache; t++) + nvlist_free(l2cache[t]); + if (spares) + free(spares); + if (l2cache) + free(l2cache); + free(top); + + return (nvroot); +} + +nvlist_t * +split_mirror_vdev(zpool_handle_t *zhp, char *newname, nvlist_t *props, + splitflags_t flags, int argc, char **argv) +{ + nvlist_t *newroot = NULL, **child; + uint_t c, children; + + if (argc > 0) { + if ((newroot = construct_spec(argc, argv)) == NULL) { + (void) fprintf(stderr, gettext("Unable to build a " + "pool from the specified devices\n")); + return (NULL); + } + + if (!flags.dryrun && make_disks(zhp, newroot) != 0) { + nvlist_free(newroot); + return (NULL); + } + + /* avoid any tricks in the spec */ + verify(nvlist_lookup_nvlist_array(newroot, + ZPOOL_CONFIG_CHILDREN, &child, &children) == 0); + for (c = 0; c < children; c++) { + char *path; + const char *type; + int min, max; + + verify(nvlist_lookup_string(child[c], + ZPOOL_CONFIG_PATH, &path) == 0); + if ((type = is_grouping(path, &min, &max)) != NULL) { + (void) fprintf(stderr, gettext("Cannot use " + "'%s' as a device for splitting\n"), type); + nvlist_free(newroot); + return (NULL); + } + } + } + + if (zpool_vdev_split(zhp, newname, &newroot, props, flags) != 0) { + if (newroot != NULL) + nvlist_free(newroot); + return (NULL); + } + + return (newroot); +} + +/* + * Get and validate the contents of the given vdev specification. This ensures + * that the nvlist returned is well-formed, that all the devices exist, and that + * they are not currently in use by any other known consumer. The 'poolconfig' + * parameter is the current configuration of the pool when adding devices + * existing pool, and is used to perform additional checks, such as changing the + * replication level of the pool. It can be 'NULL' to indicate that this is a + * new pool. The 'force' flag controls whether devices should be forcefully + * added, even if they appear in use. + */ +nvlist_t * +make_root_vdev(zpool_handle_t *zhp, int force, int check_rep, + boolean_t replacing, boolean_t dryrun, int argc, char **argv) +{ + nvlist_t *newroot; + nvlist_t *poolconfig = NULL; + is_force = force; + + /* + * Construct the vdev specification. If this is successful, we know + * that we have a valid specification, and that all devices can be + * opened. + */ + if ((newroot = construct_spec(argc, argv)) == NULL) + return (NULL); + + if (zhp && ((poolconfig = zpool_get_config(zhp, NULL)) == NULL)) + return (NULL); + + /* + * Validate each device to make sure that its not shared with another + * subsystem. We do this even if 'force' is set, because there are some + * uses (such as a dedicated dump device) that even '-f' cannot + * override. + */ + if (check_in_use(poolconfig, newroot, force, replacing, B_FALSE) != 0) { + nvlist_free(newroot); + return (NULL); + } + + /* + * Check the replication level of the given vdevs and report any errors + * found. We include the existing pool spec, if any, as we need to + * catch changes against the existing replication level. + */ + if (check_rep && check_replication(poolconfig, newroot) != 0) { + nvlist_free(newroot); + return (NULL); + } + + /* + * Run through the vdev specification and label any whole disks found. + */ + if (!dryrun && make_disks(zhp, newroot) != 0) { + nvlist_free(newroot); + return (NULL); + } + + return (newroot); +} diff --git a/cmd/zstreamdump/zstreamdump.c b/cmd/zstreamdump/zstreamdump.c new file mode 100644 index 000000000000..df23cc1e5a38 --- /dev/null +++ b/cmd/zstreamdump/zstreamdump.c @@ -0,0 +1,429 @@ +/* + * 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 http://www.opensolaris.org/os/licensing. + * 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. + */ + +#include <libnvpair.h> +#include <stdio.h> +#include <stdlib.h> +#include <strings.h> +#include <unistd.h> + +#include <sys/dmu.h> +#include <sys/zfs_ioctl.h> +#include <zfs_fletcher.h> + +uint64_t drr_record_count[DRR_NUMTYPES]; +uint64_t total_write_size = 0; +uint64_t total_stream_len = 0; +FILE *send_stream = 0; +boolean_t do_byteswap = B_FALSE; +boolean_t do_cksum = B_TRUE; +#define INITIAL_BUFLEN (1<<20) + +static void +usage(void) +{ + (void) fprintf(stderr, "usage: zstreamdump [-v] [-C] < file\n"); + (void) fprintf(stderr, "\t -v -- verbose\n"); + (void) fprintf(stderr, "\t -C -- suppress checksum verification\n"); + exit(1); +} + +/* + * ssread - send stream read. + * + * Read while computing incremental checksum + */ + +static size_t +ssread(void *buf, size_t len, zio_cksum_t *cksum) +{ + size_t outlen; + + if ((outlen = fread(buf, len, 1, send_stream)) == 0) + return (0); + + if (do_cksum && cksum) { + if (do_byteswap) + fletcher_4_incremental_byteswap(buf, len, cksum); + else + fletcher_4_incremental_native(buf, len, cksum); + } + total_stream_len += len; + return (outlen); +} + +int +main(int argc, char *argv[]) +{ + char *buf = malloc(INITIAL_BUFLEN); + dmu_replay_record_t thedrr; + dmu_replay_record_t *drr = &thedrr; + struct drr_begin *drrb = &thedrr.drr_u.drr_begin; + struct drr_end *drre = &thedrr.drr_u.drr_end; + struct drr_object *drro = &thedrr.drr_u.drr_object; + struct drr_freeobjects *drrfo = &thedrr.drr_u.drr_freeobjects; + struct drr_write *drrw = &thedrr.drr_u.drr_write; + struct drr_write_byref *drrwbr = &thedrr.drr_u.drr_write_byref; + struct drr_free *drrf = &thedrr.drr_u.drr_free; + struct drr_spill *drrs = &thedrr.drr_u.drr_spill; + char c; + boolean_t verbose = B_FALSE; + boolean_t first = B_TRUE; + int err; + zio_cksum_t zc = { 0 }; + zio_cksum_t pcksum = { 0 }; + + while ((c = getopt(argc, argv, ":vC")) != -1) { + switch (c) { + case 'C': + do_cksum = B_FALSE; + break; + case 'v': + verbose = B_TRUE; + break; + case ':': + (void) fprintf(stderr, + "missing argument for '%c' option\n", optopt); + usage(); + break; + case '?': + (void) fprintf(stderr, "invalid option '%c'\n", + optopt); + usage(); + } + } + + if (isatty(STDIN_FILENO)) { + (void) fprintf(stderr, + "Error: Backup stream can not be read " + "from a terminal.\n" + "You must redirect standard input.\n"); + exit(1); + } + + send_stream = stdin; + pcksum = zc; + while (ssread(drr, sizeof (dmu_replay_record_t), &zc)) { + + if (first) { + if (drrb->drr_magic == BSWAP_64(DMU_BACKUP_MAGIC)) { + do_byteswap = B_TRUE; + if (do_cksum) { + ZIO_SET_CHECKSUM(&zc, 0, 0, 0, 0); + /* + * recalculate header checksum now + * that we know it needs to be + * byteswapped. + */ + fletcher_4_incremental_byteswap(drr, + sizeof (dmu_replay_record_t), &zc); + } + } else if (drrb->drr_magic != DMU_BACKUP_MAGIC) { + (void) fprintf(stderr, "Invalid stream " + "(bad magic number)\n"); + exit(1); + } + first = B_FALSE; + } + if (do_byteswap) { + drr->drr_type = BSWAP_32(drr->drr_type); + drr->drr_payloadlen = + BSWAP_32(drr->drr_payloadlen); + } + + /* + * At this point, the leading fields of the replay record + * (drr_type and drr_payloadlen) have been byte-swapped if + * necessary, but the rest of the data structure (the + * union of type-specific structures) is still in its + * original state. + */ + if (drr->drr_type >= DRR_NUMTYPES) { + (void) printf("INVALID record found: type 0x%x\n", + drr->drr_type); + (void) printf("Aborting.\n"); + exit(1); + } + + drr_record_count[drr->drr_type]++; + + switch (drr->drr_type) { + case DRR_BEGIN: + if (do_byteswap) { + drrb->drr_magic = BSWAP_64(drrb->drr_magic); + drrb->drr_versioninfo = + BSWAP_64(drrb->drr_versioninfo); + drrb->drr_creation_time = + BSWAP_64(drrb->drr_creation_time); + drrb->drr_type = BSWAP_32(drrb->drr_type); + drrb->drr_flags = BSWAP_32(drrb->drr_flags); + drrb->drr_toguid = BSWAP_64(drrb->drr_toguid); + drrb->drr_fromguid = + BSWAP_64(drrb->drr_fromguid); + } + + (void) printf("BEGIN record\n"); + (void) printf("\thdrtype = %lld\n", + DMU_GET_STREAM_HDRTYPE(drrb->drr_versioninfo)); + (void) printf("\tfeatures = %llx\n", + DMU_GET_FEATUREFLAGS(drrb->drr_versioninfo)); + (void) printf("\tmagic = %llx\n", + (u_longlong_t)drrb->drr_magic); + (void) printf("\tcreation_time = %llx\n", + (u_longlong_t)drrb->drr_creation_time); + (void) printf("\ttype = %u\n", drrb->drr_type); + (void) printf("\tflags = 0x%x\n", drrb->drr_flags); + (void) printf("\ttoguid = %llx\n", + (u_longlong_t)drrb->drr_toguid); + (void) printf("\tfromguid = %llx\n", + (u_longlong_t)drrb->drr_fromguid); + (void) printf("\ttoname = %s\n", drrb->drr_toname); + if (verbose) + (void) printf("\n"); + + if ((DMU_GET_STREAM_HDRTYPE(drrb->drr_versioninfo) == + DMU_COMPOUNDSTREAM) && drr->drr_payloadlen != 0) { + nvlist_t *nv; + int sz = drr->drr_payloadlen; + + if (sz > 1<<20) { + free(buf); + buf = malloc(sz); + } + (void) ssread(buf, sz, &zc); + if (ferror(send_stream)) + perror("fread"); + err = nvlist_unpack(buf, sz, &nv, 0); + if (err) + perror(strerror(err)); + nvlist_print(stdout, nv); + nvlist_free(nv); + } + break; + + case DRR_END: + if (do_byteswap) { + drre->drr_checksum.zc_word[0] = + BSWAP_64(drre->drr_checksum.zc_word[0]); + drre->drr_checksum.zc_word[1] = + BSWAP_64(drre->drr_checksum.zc_word[1]); + drre->drr_checksum.zc_word[2] = + BSWAP_64(drre->drr_checksum.zc_word[2]); + drre->drr_checksum.zc_word[3] = + BSWAP_64(drre->drr_checksum.zc_word[3]); + } + /* + * We compare against the *previous* checksum + * value, because the stored checksum is of + * everything before the DRR_END record. + */ + if (do_cksum && !ZIO_CHECKSUM_EQUAL(drre->drr_checksum, + pcksum)) { + (void) printf("Expected checksum differs from " + "checksum in stream.\n"); + (void) printf("Expected checksum = " + "%llx/%llx/%llx/%llx\n", + pcksum.zc_word[0], + pcksum.zc_word[1], + pcksum.zc_word[2], + pcksum.zc_word[3]); + } + (void) printf("END checksum = %llx/%llx/%llx/%llx\n", + drre->drr_checksum.zc_word[0], + drre->drr_checksum.zc_word[1], + drre->drr_checksum.zc_word[2], + drre->drr_checksum.zc_word[3]); + + ZIO_SET_CHECKSUM(&zc, 0, 0, 0, 0); + break; + + case DRR_OBJECT: + if (do_byteswap) { + drro->drr_object = BSWAP_64(drro->drr_object); + drro->drr_type = BSWAP_32(drro->drr_type); + drro->drr_bonustype = + BSWAP_32(drro->drr_bonustype); + drro->drr_blksz = BSWAP_32(drro->drr_blksz); + drro->drr_bonuslen = + BSWAP_32(drro->drr_bonuslen); + drro->drr_toguid = BSWAP_64(drro->drr_toguid); + } + if (verbose) { + (void) printf("OBJECT object = %llu type = %u " + "bonustype = %u blksz = %u bonuslen = %u\n", + (u_longlong_t)drro->drr_object, + drro->drr_type, + drro->drr_bonustype, + drro->drr_blksz, + drro->drr_bonuslen); + } + if (drro->drr_bonuslen > 0) { + (void) ssread(buf, P2ROUNDUP(drro->drr_bonuslen, + 8), &zc); + } + break; + + case DRR_FREEOBJECTS: + if (do_byteswap) { + drrfo->drr_firstobj = + BSWAP_64(drrfo->drr_firstobj); + drrfo->drr_numobjs = + BSWAP_64(drrfo->drr_numobjs); + drrfo->drr_toguid = BSWAP_64(drrfo->drr_toguid); + } + if (verbose) { + (void) printf("FREEOBJECTS firstobj = %llu " + "numobjs = %llu\n", + (u_longlong_t)drrfo->drr_firstobj, + (u_longlong_t)drrfo->drr_numobjs); + } + break; + + case DRR_WRITE: + if (do_byteswap) { + drrw->drr_object = BSWAP_64(drrw->drr_object); + drrw->drr_type = BSWAP_32(drrw->drr_type); + drrw->drr_offset = BSWAP_64(drrw->drr_offset); + drrw->drr_length = BSWAP_64(drrw->drr_length); + drrw->drr_toguid = BSWAP_64(drrw->drr_toguid); + drrw->drr_key.ddk_prop = + BSWAP_64(drrw->drr_key.ddk_prop); + } + if (verbose) { + (void) printf("WRITE object = %llu type = %u " + "checksum type = %u\n" + "offset = %llu length = %llu " + "props = %llx\n", + (u_longlong_t)drrw->drr_object, + drrw->drr_type, + drrw->drr_checksumtype, + (u_longlong_t)drrw->drr_offset, + (u_longlong_t)drrw->drr_length, + (u_longlong_t)drrw->drr_key.ddk_prop); + } + (void) ssread(buf, drrw->drr_length, &zc); + total_write_size += drrw->drr_length; + break; + + case DRR_WRITE_BYREF: + if (do_byteswap) { + drrwbr->drr_object = + BSWAP_64(drrwbr->drr_object); + drrwbr->drr_offset = + BSWAP_64(drrwbr->drr_offset); + drrwbr->drr_length = + BSWAP_64(drrwbr->drr_length); + drrwbr->drr_toguid = + BSWAP_64(drrwbr->drr_toguid); + drrwbr->drr_refguid = + BSWAP_64(drrwbr->drr_refguid); + drrwbr->drr_refobject = + BSWAP_64(drrwbr->drr_refobject); + drrwbr->drr_refoffset = + BSWAP_64(drrwbr->drr_refoffset); + drrwbr->drr_key.ddk_prop = + BSWAP_64(drrwbr->drr_key.ddk_prop); + } + if (verbose) { + (void) printf("WRITE_BYREF object = %llu " + "checksum type = %u props = %llx\n" + "offset = %llu length = %llu\n" + "toguid = %llx refguid = %llx\n" + "refobject = %llu refoffset = %llu\n", + (u_longlong_t)drrwbr->drr_object, + drrwbr->drr_checksumtype, + (u_longlong_t)drrwbr->drr_key.ddk_prop, + (u_longlong_t)drrwbr->drr_offset, + (u_longlong_t)drrwbr->drr_length, + (u_longlong_t)drrwbr->drr_toguid, + (u_longlong_t)drrwbr->drr_refguid, + (u_longlong_t)drrwbr->drr_refobject, + (u_longlong_t)drrwbr->drr_refoffset); + } + break; + + case DRR_FREE: + if (do_byteswap) { + drrf->drr_object = BSWAP_64(drrf->drr_object); + drrf->drr_offset = BSWAP_64(drrf->drr_offset); + drrf->drr_length = BSWAP_64(drrf->drr_length); + } + if (verbose) { + (void) printf("FREE object = %llu " + "offset = %llu length = %lld\n", + (u_longlong_t)drrf->drr_object, + (u_longlong_t)drrf->drr_offset, + (longlong_t)drrf->drr_length); + } + break; + case DRR_SPILL: + if (do_byteswap) { + drrs->drr_object = BSWAP_64(drrs->drr_object); + drrs->drr_length = BSWAP_64(drrs->drr_length); + } + if (verbose) { + (void) printf("SPILL block for object = %llu " + "length = %llu\n", drrs->drr_object, + drrs->drr_length); + } + (void) ssread(buf, drrs->drr_length, &zc); + break; + } + pcksum = zc; + } + free(buf); + + /* Print final summary */ + + (void) printf("SUMMARY:\n"); + (void) printf("\tTotal DRR_BEGIN records = %lld\n", + (u_longlong_t)drr_record_count[DRR_BEGIN]); + (void) printf("\tTotal DRR_END records = %lld\n", + (u_longlong_t)drr_record_count[DRR_END]); + (void) printf("\tTotal DRR_OBJECT records = %lld\n", + (u_longlong_t)drr_record_count[DRR_OBJECT]); + (void) printf("\tTotal DRR_FREEOBJECTS records = %lld\n", + (u_longlong_t)drr_record_count[DRR_FREEOBJECTS]); + (void) printf("\tTotal DRR_WRITE records = %lld\n", + (u_longlong_t)drr_record_count[DRR_WRITE]); + (void) printf("\tTotal DRR_FREE records = %lld\n", + (u_longlong_t)drr_record_count[DRR_FREE]); + (void) printf("\tTotal DRR_SPILL records = %lld\n", + (u_longlong_t)drr_record_count[DRR_SPILL]); + (void) printf("\tTotal records = %lld\n", + (u_longlong_t)(drr_record_count[DRR_BEGIN] + + drr_record_count[DRR_OBJECT] + + drr_record_count[DRR_FREEOBJECTS] + + drr_record_count[DRR_WRITE] + + drr_record_count[DRR_FREE] + + drr_record_count[DRR_SPILL] + + drr_record_count[DRR_END])); + (void) printf("\tTotal write size = %lld (0x%llx)\n", + (u_longlong_t)total_write_size, (u_longlong_t)total_write_size); + (void) printf("\tTotal stream length = %lld (0x%llx)\n", + (u_longlong_t)total_stream_len, (u_longlong_t)total_stream_len); + return (0); +} diff --git a/cmd/ztest/ztest.c b/cmd/ztest/ztest.c new file mode 100644 index 000000000000..b2d81b558882 --- /dev/null +++ b/cmd/ztest/ztest.c @@ -0,0 +1,5604 @@ +/* + * 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 http://www.opensolaris.org/os/licensing. + * 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. + */ + +/* + * 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. + * + * (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. + * + * 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 more vdevs [-v], datasets [-d], or threads [-t] + * to increase the pool capacity, fanout, and overall stress level. + * + * The -N(okill) option will suppress kills, so each child runs to completion. + * This can be useful when you're trying to distinguish temporal incursions + * from plain old race conditions. + */ + +#include <sys/zfs_context.h> +#include <sys/spa.h> +#include <sys/dmu.h> +#include <sys/txg.h> +#include <sys/dbuf.h> +#include <sys/zap.h> +#include <sys/dmu_objset.h> +#include <sys/poll.h> +#include <sys/stat.h> +#include <sys/time.h> +#include <sys/wait.h> +#include <sys/mman.h> +#include <sys/resource.h> +#include <sys/zio.h> +#include <sys/zil.h> +#include <sys/zil_impl.h> +#include <sys/vdev_impl.h> +#include <sys/vdev_file.h> +#include <sys/spa_impl.h> +#include <sys/metaslab_impl.h> +#include <sys/dsl_prop.h> +#include <sys/dsl_dataset.h> +#include <sys/dsl_scan.h> +#include <sys/zio_checksum.h> +#include <sys/refcount.h> +#include <stdio.h> +#include <stdio_ext.h> +#include <stdlib.h> +#include <unistd.h> +#include <signal.h> +#include <umem.h> +#include <dlfcn.h> +#include <ctype.h> +#include <math.h> +#include <sys/fs/zfs.h> +#include <libnvpair.h> + +static char cmdname[] = "ztest"; +static char *zopt_pool = cmdname; + +static uint64_t zopt_vdevs = 5; +static uint64_t zopt_vdevtime; +static int zopt_ashift = SPA_MINBLOCKSHIFT; +static int zopt_mirrors = 2; +static int zopt_raidz = 4; +static int zopt_raidz_parity = 1; +static size_t zopt_vdev_size = SPA_MINDEVSIZE; +static int zopt_datasets = 7; +static int zopt_threads = 23; +static uint64_t zopt_passtime = 60; /* 60 seconds */ +static uint64_t zopt_killrate = 70; /* 70% kill rate */ +static int zopt_verbose = 0; +static int zopt_init = 1; +static char *zopt_dir = "/tmp"; +static uint64_t zopt_time = 300; /* 5 minutes */ +static uint64_t zopt_maxloops = 50; /* max loops during spa_freeze() */ + +#define BT_MAGIC 0x123456789abcdefULL +#define MAXFAULTS() (MAX(zs->zs_mirrors, 1) * (zopt_raidz_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_TYPES +}; + +typedef struct ztest_block_tag { + uint64_t bt_magic; + uint64_t bt_objset; + uint64_t bt_object; + 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; + +/* + * XXX -- fix zfs range locks to be generic so we can use them here. + */ +typedef enum { + RL_READER, + RL_WRITER, + RL_APPEND +} rl_type_t; + +typedef struct rll { + void *rll_writer; + int rll_readers; + mutex_t rll_lock; + cond_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_gen; + uint64_t od_crgen; + char od_name[MAXNAMELEN]; +} ztest_od_t; + +/* + * Per-dataset state. + */ +typedef struct ztest_ds { + objset_t *zd_os; + zilog_t *zd_zilog; + uint64_t zd_seq; + ztest_od_t *zd_od; /* debugging aid */ + char zd_name[MAXNAMELEN]; + mutex_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 <interval> seconds */ + uint64_t zi_call_count; /* per-pass count */ + uint64_t zi_call_time; /* per-pass time */ + uint64_t zi_call_next; /* next time to call this function */ +} ztest_info_t; + +/* + * Note: these aren't static because we want dladdr() to work. + */ +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_commit_callbacks; +ztest_func_t ztest_zap; +ztest_func_t ztest_zap_parallel; +ztest_func_t ztest_zil_commit; +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_ddt_repair; +ztest_func_t ztest_dmu_snapshot_hold; +ztest_func_t ztest_spa_rename; +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_aux_add_remove; +ztest_func_t ztest_split_pool; + +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 */ + +ztest_info_t ztest_info[] = { + { ztest_dmu_read_write, 1, &zopt_always }, + { ztest_dmu_write_parallel, 10, &zopt_always }, + { ztest_dmu_object_alloc_free, 1, &zopt_always }, + { ztest_dmu_commit_callbacks, 1, &zopt_always }, + { ztest_zap, 30, &zopt_always }, + { ztest_zap_parallel, 100, &zopt_always }, + { ztest_split_pool, 1, &zopt_always }, + { ztest_zil_commit, 1, &zopt_incessant }, + { ztest_dmu_read_write_zcopy, 1, &zopt_often }, + { ztest_dmu_objset_create_destroy, 1, &zopt_often }, + { ztest_dsl_prop_get_set, 1, &zopt_often }, + { ztest_spa_prop_get_set, 1, &zopt_sometimes }, +#if 0 + { ztest_dmu_prealloc, 1, &zopt_sometimes }, +#endif + { ztest_fzap, 1, &zopt_sometimes }, + { ztest_dmu_snapshot_create_destroy, 1, &zopt_sometimes }, + { ztest_spa_create_destroy, 1, &zopt_sometimes }, + { ztest_fault_inject, 1, &zopt_sometimes }, + { ztest_ddt_repair, 1, &zopt_sometimes }, + { ztest_dmu_snapshot_hold, 1, &zopt_sometimes }, + { ztest_spa_rename, 1, &zopt_rarely }, + { ztest_scrub, 1, &zopt_rarely }, + { ztest_dsl_dataset_promote_busy, 1, &zopt_rarely }, + { ztest_vdev_attach_detach, 1, &zopt_rarely }, + { ztest_vdev_LUN_growth, 1, &zopt_rarely }, + { ztest_vdev_add_remove, 1, &zopt_vdevtime }, + { ztest_vdev_aux_add_remove, 1, &zopt_vdevtime }, +}; + +#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 { + mutex_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 { + char *zs_pool; + spa_t *zs_spa; + 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; + mutex_t zs_vdev_lock; + rwlock_t zs_name_lock; + ztest_info_t zs_info[ZTEST_FUNCS]; + uint64_t zs_splits; + uint64_t zs_mirrors; + ztest_ds_t zs_zd[]; +} 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; +uint64_t *ztest_seq; + +static int ztest_random_fd; +static int ztest_dump_core = 1; + +static boolean_t ztest_exiting; + +/* Global commit callback list */ +static ztest_cb_list_t zcl; + +extern uint64_t metaslab_gang_bang; +extern uint64_t metaslab_df_alloc_threshold; +static uint64_t metaslab_sz; + +enum ztest_object { + ZTEST_META_DNODE = 0, + ZTEST_DIROBJ, + ZTEST_OBJECTS +}; + +static void usage(boolean_t) __NORETURN; + +/* + * These libumem hooks provide a reasonable set of defaults for the allocator's + * debugging facilities. + */ +const char * +_umem_debug_init() +{ + return ("default,verbose"); /* $UMEM_DEBUG setting */ +} + +const char * +_umem_logging_init(void) +{ + return ("fail,contents"); /* $UMEM_LOGGING setting */ +} + +#define FATAL_MSG_SZ 1024 + +char *fatal_msg; + +static void +fatal(int do_perror, char *message, ...) +{ + va_list args; + int save_errno = errno; + char buf[FATAL_MSG_SZ]; + + (void) fflush(stdout); + + 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 */ + if (ztest_dump_core) + abort(); + 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); + /* NOTREACHED */ +} + +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)); + if (fval > 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); +} + +static void +usage(boolean_t requested) +{ + char nice_vdev_size[10]; + char nice_gang_bang[10]; + FILE *fp = requested ? stdout : stderr; + + nicenum(zopt_vdev_size, nice_vdev_size); + nicenum(metaslab_gang_bang, nice_gang_bang); + + (void) fprintf(fp, "Usage: %s\n" + "\t[-v vdevs (default: %llu)]\n" + "\t[-s size_of_each_vdev (default: %s)]\n" + "\t[-a alignment_shift (default: %d)] use 0 for random\n" + "\t[-m mirror_copies (default: %d)]\n" + "\t[-r raidz_disks (default: %d)]\n" + "\t[-R raidz_parity (default: %d)]\n" + "\t[-d datasets (default: %d)]\n" + "\t[-t threads (default: %d)]\n" + "\t[-g gang_block_threshold (default: %s)]\n" + "\t[-i init_count (default: %d)] initialize pool i times\n" + "\t[-k kill_percentage (default: %llu%%)]\n" + "\t[-p pool_name (default: %s)]\n" + "\t[-f dir (default: %s)] file directory for vdev files\n" + "\t[-V] verbose (use multiple times for ever more blather)\n" + "\t[-E] use existing pool instead of creating new one\n" + "\t[-T time (default: %llu sec)] total run time\n" + "\t[-F freezeloops (default: %llu)] max loops in spa_freeze()\n" + "\t[-P passtime (default: %llu sec)] time per pass\n" + "\t[-h] (print help)\n" + "", + cmdname, + (u_longlong_t)zopt_vdevs, /* -v */ + nice_vdev_size, /* -s */ + zopt_ashift, /* -a */ + zopt_mirrors, /* -m */ + zopt_raidz, /* -r */ + zopt_raidz_parity, /* -R */ + zopt_datasets, /* -d */ + zopt_threads, /* -t */ + nice_gang_bang, /* -g */ + zopt_init, /* -i */ + (u_longlong_t)zopt_killrate, /* -k */ + zopt_pool, /* -p */ + zopt_dir, /* -f */ + (u_longlong_t)zopt_time, /* -T */ + (u_longlong_t)zopt_maxloops, /* -F */ + (u_longlong_t)zopt_passtime); /* -P */ + exit(requested ? 0 : 1); +} + +static void +process_options(int argc, char **argv) +{ + int opt; + uint64_t value; + + /* By default, test gang blocks for blocks 32K and greater */ + metaslab_gang_bang = 32 << 10; + + while ((opt = getopt(argc, argv, + "v:s:a:m:r:R:d:t:g:i:k:p:f:VET:P:hF:")) != EOF) { + value = 0; + switch (opt) { + case 'v': + case 's': + case 'a': + case 'm': + case 'r': + case 'R': + case 'd': + case 't': + case 'g': + case 'i': + case 'k': + case 'T': + case 'P': + case 'F': + value = nicenumtoull(optarg); + } + switch (opt) { + case 'v': + zopt_vdevs = value; + break; + case 's': + zopt_vdev_size = MAX(SPA_MINDEVSIZE, value); + break; + case 'a': + zopt_ashift = value; + break; + case 'm': + zopt_mirrors = value; + break; + case 'r': + zopt_raidz = MAX(1, value); + break; + case 'R': + zopt_raidz_parity = MIN(MAX(value, 1), 3); + break; + case 'd': + zopt_datasets = MAX(1, value); + break; + case 't': + zopt_threads = MAX(1, value); + break; + case 'g': + metaslab_gang_bang = MAX(SPA_MINBLOCKSIZE << 1, value); + break; + case 'i': + zopt_init = value; + break; + case 'k': + zopt_killrate = value; + break; + case 'p': + zopt_pool = strdup(optarg); + break; + case 'f': + zopt_dir = strdup(optarg); + break; + case 'V': + zopt_verbose++; + break; + case 'E': + zopt_init = 0; + break; + case 'T': + zopt_time = value; + break; + case 'P': + zopt_passtime = MAX(1, value); + break; + case 'F': + zopt_maxloops = MAX(1, value); + break; + case 'h': + usage(B_TRUE); + break; + case '?': + default: + usage(B_FALSE); + break; + } + } + + zopt_raidz_parity = MIN(zopt_raidz_parity, zopt_raidz - 1); + + zopt_vdevtime = (zopt_vdevs > 0 ? zopt_time * NANOSEC / zopt_vdevs : + UINT64_MAX >> 2); +} + +static void +ztest_kill(ztest_shared_t *zs) +{ + zs->zs_alloc = metaslab_class_get_alloc(spa_normal_class(zs->zs_spa)); + zs->zs_space = metaslab_class_get_space(spa_normal_class(zs->zs_spa)); + (void) kill(getpid(), SIGKILL); +} + +static uint64_t +ztest_random(uint64_t range) +{ + uint64_t r; + + if (range == 0) + return (0); + + if (read(ztest_random_fd, &r, sizeof (r)) != sizeof (r)) + fatal(1, "short read from /dev/urandom"); + + return (r % range); +} + +/* ARGSUSED */ +static void +ztest_record_enospc(const char *s) +{ + ztest_shared->zs_enospc_count++; +} + +static uint64_t +ztest_get_ashift(void) +{ + if (zopt_ashift == 0) + return (SPA_MINBLOCKSHIFT + ztest_random(3)); + return (zopt_ashift); +} + +static nvlist_t * +make_vdev_file(char *path, char *aux, size_t size, uint64_t ashift) +{ + char pathbuf[MAXPATHLEN]; + uint64_t vdev; + nvlist_t *file; + + if (ashift == 0) + ashift = ztest_get_ashift(); + + if (path == NULL) { + path = pathbuf; + + if (aux != NULL) { + vdev = ztest_shared->zs_vdev_aux; + (void) sprintf(path, ztest_aux_template, + zopt_dir, zopt_pool, aux, vdev); + } else { + vdev = ztest_shared->zs_vdev_next_leaf++; + (void) sprintf(path, ztest_dev_template, + zopt_dir, zopt_pool, vdev); + } + } + + if (size != 0) { + int fd = open(path, O_RDWR | O_CREAT | O_TRUNC, 0666); + if (fd == -1) + fatal(1, "can't open %s", path); + if (ftruncate(fd, size) != 0) + fatal(1, "can't ftruncate %s", path); + (void) close(fd); + } + + VERIFY(nvlist_alloc(&file, NV_UNIQUE_NAME, 0) == 0); + VERIFY(nvlist_add_string(file, ZPOOL_CONFIG_TYPE, VDEV_TYPE_FILE) == 0); + VERIFY(nvlist_add_string(file, ZPOOL_CONFIG_PATH, path) == 0); + VERIFY(nvlist_add_uint64(file, ZPOOL_CONFIG_ASHIFT, ashift) == 0); + + return (file); +} + +static nvlist_t * +make_vdev_raidz(char *path, char *aux, size_t size, uint64_t ashift, int r) +{ + nvlist_t *raidz, **child; + int c; + + if (r < 2) + return (make_vdev_file(path, aux, size, ashift)); + child = umem_alloc(r * sizeof (nvlist_t *), UMEM_NOFAIL); + + for (c = 0; c < r; c++) + child[c] = make_vdev_file(path, aux, size, ashift); + + VERIFY(nvlist_alloc(&raidz, NV_UNIQUE_NAME, 0) == 0); + VERIFY(nvlist_add_string(raidz, ZPOOL_CONFIG_TYPE, + VDEV_TYPE_RAIDZ) == 0); + VERIFY(nvlist_add_uint64(raidz, ZPOOL_CONFIG_NPARITY, + zopt_raidz_parity) == 0); + VERIFY(nvlist_add_nvlist_array(raidz, ZPOOL_CONFIG_CHILDREN, + child, r) == 0); + + for (c = 0; c < r; c++) + nvlist_free(child[c]); + + umem_free(child, r * sizeof (nvlist_t *)); + + return (raidz); +} + +static nvlist_t * +make_vdev_mirror(char *path, char *aux, size_t size, uint64_t ashift, + int r, int m) +{ + nvlist_t *mirror, **child; + int c; + + if (m < 1) + return (make_vdev_raidz(path, aux, size, ashift, r)); + + child = umem_alloc(m * sizeof (nvlist_t *), UMEM_NOFAIL); + + for (c = 0; c < m; c++) + child[c] = make_vdev_raidz(path, aux, size, ashift, r); + + VERIFY(nvlist_alloc(&mirror, NV_UNIQUE_NAME, 0) == 0); + VERIFY(nvlist_add_string(mirror, ZPOOL_CONFIG_TYPE, + VDEV_TYPE_MIRROR) == 0); + VERIFY(nvlist_add_nvlist_array(mirror, ZPOOL_CONFIG_CHILDREN, + child, m) == 0); + + for (c = 0; c < m; c++) + nvlist_free(child[c]); + + umem_free(child, m * sizeof (nvlist_t *)); + + return (mirror); +} + +static nvlist_t * +make_vdev_root(char *path, char *aux, size_t size, uint64_t ashift, + int log, int r, int m, int t) +{ + nvlist_t *root, **child; + int c; + + ASSERT(t > 0); + + child = umem_alloc(t * sizeof (nvlist_t *), UMEM_NOFAIL); + + for (c = 0; c < t; c++) { + child[c] = make_vdev_mirror(path, aux, size, ashift, r, m); + VERIFY(nvlist_add_uint64(child[c], ZPOOL_CONFIG_IS_LOG, + log) == 0); + } + + VERIFY(nvlist_alloc(&root, NV_UNIQUE_NAME, 0) == 0); + VERIFY(nvlist_add_string(root, ZPOOL_CONFIG_TYPE, VDEV_TYPE_ROOT) == 0); + VERIFY(nvlist_add_nvlist_array(root, aux ? aux : ZPOOL_CONFIG_CHILDREN, + child, t) == 0); + + for (c = 0; c < t; c++) + nvlist_free(child[c]); + + umem_free(child, t * sizeof (nvlist_t *)); + + return (root); +} + +static int +ztest_random_blocksize(void) +{ + return (1 << (SPA_MINBLOCKSHIFT + + ztest_random(SPA_MAXBLOCKSHIFT - SPA_MINBLOCKSHIFT + 1))); +} + +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; + + ASSERT(spa_config_held(spa, SCL_ALL, RW_READER) != 0); + + do { + top = ztest_random(rvd->vdev_children); + tvd = rvd->vdev_child[top]; + } while (tvd->vdev_ishole || (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[MAXPATHLEN]; + uint64_t curval; + int error; + + error = dsl_prop_set(osname, propname, + (inherit ? ZPROP_SRC_NONE : ZPROP_SRC_LOCAL), + sizeof (value), 1, &value); + + if (error == ENOSPC) { + ztest_record_enospc(FTAG); + return (error); + } + ASSERT3U(error, ==, 0); + + VERIFY3U(dsl_prop_get(osname, propname, sizeof (curval), + 1, &curval, setpoint), ==, 0); + + if (zopt_verbose >= 6) { + VERIFY(zfs_prop_index_to_string(prop, curval, &valname) == 0); + (void) printf("%s %s = %s at '%s'\n", + osname, propname, valname, setpoint); + } + + return (error); +} + +static int +ztest_spa_prop_set_uint64(ztest_shared_t *zs, zpool_prop_t prop, uint64_t value) +{ + spa_t *spa = zs->zs_spa; + nvlist_t *props = NULL; + int error; + + VERIFY(nvlist_alloc(&props, NV_UNIQUE_NAME, 0) == 0); + VERIFY(nvlist_add_uint64(props, zpool_prop_to_name(prop), value) == 0); + + error = spa_prop_set(spa, props); + + nvlist_free(props); + + if (error == ENOSPC) { + ztest_record_enospc(FTAG); + return (error); + } + ASSERT3U(error, ==, 0); + + return (error); +} + +static void +ztest_rll_init(rll_t *rll) +{ + rll->rll_writer = NULL; + rll->rll_readers = 0; + VERIFY(_mutex_init(&rll->rll_lock, USYNC_THREAD, NULL) == 0); + VERIFY(cond_init(&rll->rll_cv, USYNC_THREAD, NULL) == 0); +} + +static void +ztest_rll_destroy(rll_t *rll) +{ + ASSERT(rll->rll_writer == NULL); + ASSERT(rll->rll_readers == 0); + VERIFY(_mutex_destroy(&rll->rll_lock) == 0); + VERIFY(cond_destroy(&rll->rll_cv) == 0); +} + +static void +ztest_rll_lock(rll_t *rll, rl_type_t type) +{ + VERIFY(mutex_lock(&rll->rll_lock) == 0); + + if (type == RL_READER) { + while (rll->rll_writer != NULL) + (void) cond_wait(&rll->rll_cv, &rll->rll_lock); + rll->rll_readers++; + } else { + while (rll->rll_writer != NULL || rll->rll_readers) + (void) cond_wait(&rll->rll_cv, &rll->rll_lock); + rll->rll_writer = curthread; + } + + VERIFY(mutex_unlock(&rll->rll_lock) == 0); +} + +static void +ztest_rll_unlock(rll_t *rll) +{ + VERIFY(mutex_lock(&rll->rll_lock) == 0); + + if (rll->rll_writer) { + ASSERT(rll->rll_readers == 0); + rll->rll_writer = NULL; + } else { + ASSERT(rll->rll_readers != 0); + ASSERT(rll->rll_writer == NULL); + rll->rll_readers--; + } + + if (rll->rll_writer == NULL && rll->rll_readers == 0) + VERIFY(cond_broadcast(&rll->rll_cv) == 0); + + VERIFY(mutex_unlock(&rll->rll_lock) == 0); +} + +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, objset_t *os) +{ + zd->zd_os = os; + zd->zd_zilog = dmu_objset_zil(os); + zd->zd_seq = 0; + dmu_objset_name(os, zd->zd_name); + + VERIFY(_mutex_init(&zd->zd_dirobj_lock, USYNC_THREAD, NULL) == 0); + + for (int l = 0; l < ZTEST_OBJECT_LOCKS; l++) + ztest_rll_init(&zd->zd_object_lock[l]); + + for (int l = 0; l < ZTEST_RANGE_LOCKS; l++) + ztest_rll_init(&zd->zd_range_lock[l]); +} + +static void +ztest_zd_fini(ztest_ds_t *zd) +{ + VERIFY(_mutex_destroy(&zd->zd_dirobj_lock) == 0); + + for (int l = 0; l < ZTEST_OBJECT_LOCKS; l++) + ztest_rll_destroy(&zd->zd_object_lock[l]); + + for (int 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) { + ASSERT(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); + ASSERT(txg != 0); + return (txg); +} + +static void +ztest_pattern_set(void *buf, uint64_t size, uint64_t value) +{ + uint64_t *ip = buf; + uint64_t *ip_end = (uint64_t *)((uintptr_t)buf + (uintptr_t)size); + + while (ip < ip_end) + *ip++ = value; +} + +static boolean_t +ztest_pattern_match(void *buf, uint64_t size, uint64_t value) +{ + uint64_t *ip = buf; + uint64_t *ip_end = (uint64_t *)((uintptr_t)buf + (uintptr_t)size); + uint64_t diff = 0; + + while (ip < ip_end) + diff |= (value - *ip++); + + return (diff == 0); +} + +static void +ztest_bt_generate(ztest_block_tag_t *bt, objset_t *os, uint64_t object, + 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_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 offset, uint64_t gen, uint64_t txg, uint64_t crtxg) +{ + ASSERT(bt->bt_magic == BT_MAGIC); + ASSERT(bt->bt_objset == dmu_objset_id(os)); + ASSERT(bt->bt_object == object); + ASSERT(bt->bt_offset == offset); + ASSERT(bt->bt_gen <= gen); + ASSERT(bt->bt_txg <= txg); + ASSERT(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); +} + +/* + * 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_bonuslen 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); + bcopy(&lr->lr_common + 1, &itx->itx_lr + 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); + bcopy(&lr->lr_common + 1, &itx->itx_lr + 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) + 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); + itx->itx_sod += (write_state == WR_NEED_COPY ? lr->lr_length : 0); + + bcopy(&lr->lr_common + 1, &itx->itx_lr + 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)); + bcopy(&lr->lr_common + 1, &itx->itx_lr + 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)); + bcopy(&lr->lr_common + 1, &itx->itx_lr + 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(ztest_ds_t *zd, lr_create_t *lr, boolean_t byteswap) +{ + 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; + + if (byteswap) + byteswap_uint64_array(lr, sizeof (*lr)); + + ASSERT(lr->lr_doid == ZTEST_DIROBJ); + ASSERT(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); + + ASSERT(dmu_objset_zil(os)->zl_replay == !!lr->lr_foid); + + if (lr->lrz_type == DMU_OT_ZAP_OTHER) { + if (lr->lr_foid == 0) { + lr->lr_foid = zap_create(os, + lr->lrz_type, lr->lrz_bonustype, + lr->lrz_bonuslen, tx); + } else { + error = zap_create_claim(os, lr->lr_foid, + lr->lrz_type, lr->lrz_bonustype, + lr->lrz_bonuslen, tx); + } + } else { + if (lr->lr_foid == 0) { + lr->lr_foid = dmu_object_alloc(os, + lr->lrz_type, 0, lr->lrz_bonustype, + lr->lrz_bonuslen, tx); + } else { + error = dmu_object_claim(os, lr->lr_foid, + lr->lrz_type, 0, lr->lrz_bonustype, + lr->lrz_bonuslen, tx); + } + } + + if (error) { + ASSERT3U(error, ==, EEXIST); + ASSERT(zd->zd_zilog->zl_replay); + dmu_tx_commit(tx); + return (error); + } + + ASSERT(lr->lr_foid != 0); + + if (lr->lrz_type != DMU_OT_ZAP_OTHER) + VERIFY3U(0, ==, dmu_object_set_blocksize(os, lr->lr_foid, + lr->lrz_blocksize, lr->lrz_ibshift, tx)); + + VERIFY3U(0, ==, 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, -1ULL, lr->lr_gen, txg, txg); + dmu_buf_rele(db, FTAG); + + VERIFY3U(0, ==, 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(ztest_ds_t *zd, lr_remove_t *lr, boolean_t byteswap) +{ + 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)); + + ASSERT(lr->lr_doid == ZTEST_DIROBJ); + ASSERT(name[0] != '\0'); + + VERIFY3U(0, ==, + zap_lookup(os, lr->lr_doid, name, sizeof (object), 1, &object)); + ASSERT(object != 0); + + ztest_object_lock(zd, object, RL_WRITER); + + VERIFY3U(0, ==, 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) { + VERIFY3U(0, ==, zap_destroy(os, object, tx)); + } else { + VERIFY3U(0, ==, dmu_object_free(os, object, tx)); + } + + VERIFY3U(0, ==, 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(ztest_ds_t *zd, lr_write_t *lr, boolean_t byteswap) +{ + 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); + + VERIFY3U(0, ==, 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. + */ + ASSERT(offset % doi.doi_data_block_size == 0); + 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, + 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, 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, offset, gen, txg, crtxg); + } + + if (abuf == NULL) { + dmu_write(os, lr->lr_foid, offset, length, data, tx); + } else { + bcopy(data, abuf->b_data, length); + dmu_assign_arcbuf(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(ztest_ds_t *zd, lr_truncate_t *lr, boolean_t byteswap) +{ + 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); + } + + VERIFY(dmu_free_range(os, lr->lr_foid, lr->lr_offset, + lr->lr_length, tx) == 0); + + (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(ztest_ds_t *zd, lr_setattr_t *lr, boolean_t byteswap) +{ + objset_t *os = zd->zd_os; + dmu_tx_t *tx; + dmu_buf_t *db; + ztest_block_tag_t *bbt; + uint64_t txg, lrtxg, crtxg; + + if (byteswap) + byteswap_uint64_array(lr, sizeof (*lr)); + + ztest_object_lock(zd, lr->lr_foid, RL_WRITER); + + VERIFY3U(0, ==, 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; + + if (zd->zd_zilog->zl_replay) { + ASSERT(lr->lr_size != 0); + ASSERT(lr->lr_mode != 0); + ASSERT(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; + ASSERT(lrtxg == 0); + } + + /* + * Verify that the current bonus buffer is not newer than our txg. + */ + ztest_bt_verify(bbt, os, lr->lr_foid, -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); + VERIFY3U(dmu_set_bonus(db, lr->lr_size, tx), ==, 0); + bbt = ztest_bt_bonus(db); + + ztest_bt_generate(bbt, os, lr->lr_foid, -1ULL, lr->lr_mode, txg, crtxg); + + 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 */ +}; + +/* + * ZIL get_data callbacks + */ + +static void +ztest_get_done(zgd_t *zgd, int error) +{ + ztest_ds_t *zd = zgd->zgd_private; + uint64_t object = zgd->zgd_rl->rl_object; + + if (zgd->zgd_db) + dmu_buf_rele(zgd->zgd_db, zgd); + + ztest_range_unlock(zgd->zgd_rl); + ztest_object_unlock(zd, object); + + if (error == 0 && zgd->zgd_bp) + zil_add_block(zgd->zgd_zilog, zgd->zgd_bp); + + umem_free(zgd, sizeof (*zgd)); +} + +static int +ztest_get_data(void *arg, lr_write_t *lr, char *buf, zio_t *zio) +{ + 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; + blkptr_t *bp = &lr->lr_blkptr; + 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; + + 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_zilog = zd->zd_zilog; + zgd->zgd_private = zd; + + if (buf != NULL) { /* immediate write */ + zgd->zgd_rl = ztest_range_lock(zd, object, offset, size, + RL_READER); + + error = dmu_read(os, object, offset, size, buf, + DMU_READ_NO_PREFETCH); + ASSERT(error == 0); + } else { + size = doi.doi_data_block_size; + if (ISP2(size)) { + offset = P2ALIGN(offset, size); + } else { + ASSERT(offset < size); + offset = 0; + } + + zgd->zgd_rl = 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) { + zgd->zgd_db = db; + zgd->zgd_bp = bp; + + ASSERT(db->db_offset == offset); + ASSERT(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) + bcopy(name, lr + lrsize, namesize); + + return (lr); +} + +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; + + ASSERT(_mutex_held(&zd->zd_dirobj_lock)); + + for (int 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) { + ASSERT(error == ENOENT); + ASSERT(od->od_object == 0); + missing++; + } else { + dmu_buf_t *db; + ztest_block_tag_t *bbt; + dmu_object_info_t doi; + + ASSERT(od->od_object != 0); + ASSERT(missing == 0); /* there should be no gaps */ + + ztest_object_lock(zd, od->od_object, RL_READER); + VERIFY3U(0, ==, 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; + + ASSERT(_mutex_held(&zd->zd_dirobj_lock)); + + for (int 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_bonuslen = dmu_bonus_max(); + lr->lr_gen = od->od_crgen; + lr->lr_crtime[0] = time(NULL); + + if (ztest_replay_create(zd, lr, B_FALSE) != 0) { + ASSERT(missing == 0); + 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; + ASSERT(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; + + ASSERT(_mutex_held(&zd->zd_dirobj_lock)); + + od += count - 1; + + for (int i = count - 1; i >= 0; i--, od--) { + if (missing) { + missing++; + continue; + } + + 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); + + bcopy(data, lr + 1, 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) +{ + ztest_block_tag_t wbt; + dmu_object_info_t doi; + enum ztest_io_type io_type; + uint64_t blocksize; + void *data; + + VERIFY(dmu_object_info(zd->zd_os, object, &doi) == 0); + 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; + + switch (io_type) { + + case ZTEST_IO_WRITE_TAG: + ztest_bt_generate(&wbt, zd->zd_os, object, 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: + bzero(data, 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; + } + + umem_free(data, blocksize); +} + +/* + * Initialize an object description template. + */ +static void +ztest_od_init(ztest_od_t *od, uint64_t id, char *tag, uint64_t index, + dmu_object_type_t type, uint64_t blocksize, 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_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(%lld)[%llu]", + tag, (int64_t)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; + + VERIFY(mutex_lock(&zd->zd_dirobj_lock) == 0); + 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; + VERIFY(mutex_unlock(&zd->zd_dirobj_lock) == 0); + + return (rv); +} + +/* ARGSUSED */ +void +ztest_zil_commit(ztest_ds_t *zd, uint64_t id) +{ + zilog_t *zilog = zd->zd_zilog; + + 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); + ASSERT(zd->zd_seq <= zilog->zl_commit_lr_seq); + zd->zd_seq = zilog->zl_commit_lr_seq; + mutex_exit(&zilog->zl_lock); +} + +/* + * Verify that we can't destroy an active pool, create an existing pool, + * or create a pool with a bad vdev spec. + */ +/* ARGSUSED */ +void +ztest_spa_create_destroy(ztest_ds_t *zd, uint64_t id) +{ + ztest_shared_t *zs = ztest_shared; + spa_t *spa; + nvlist_t *nvroot; + + /* + * Attempt to create using a bad file. + */ + nvroot = make_vdev_root("/dev/bogus", NULL, 0, 0, 0, 0, 0, 1); + VERIFY3U(ENOENT, ==, + spa_create("ztest_bad_file", nvroot, NULL, NULL, NULL)); + nvlist_free(nvroot); + + /* + * Attempt to create using a bad mirror. + */ + nvroot = make_vdev_root("/dev/bogus", NULL, 0, 0, 0, 0, 2, 1); + VERIFY3U(ENOENT, ==, + spa_create("ztest_bad_mirror", nvroot, NULL, NULL, NULL)); + nvlist_free(nvroot); + + /* + * Attempt to create an existing pool. It shouldn't matter + * what's in the nvroot; we should fail with EEXIST. + */ + (void) rw_rdlock(&zs->zs_name_lock); + nvroot = make_vdev_root("/dev/bogus", NULL, 0, 0, 0, 0, 0, 1); + VERIFY3U(EEXIST, ==, spa_create(zs->zs_pool, nvroot, NULL, NULL, NULL)); + nvlist_free(nvroot); + VERIFY3U(0, ==, spa_open(zs->zs_pool, &spa, FTAG)); + VERIFY3U(EBUSY, ==, spa_destroy(zs->zs_pool)); + spa_close(spa, FTAG); + + (void) rw_unlock(&zs->zs_name_lock); +} + +static vdev_t * +vdev_lookup_by_path(vdev_t *vd, const char *path) +{ + vdev_t *mvd; + + if (vd->vdev_path != NULL && strcmp(path, vd->vdev_path) == 0) + return (vd); + + for (int c = 0; c < vd->vdev_children; c++) + if ((mvd = vdev_lookup_by_path(vd->vdev_child[c], path)) != + NULL) + return (mvd); + + return (NULL); +} + +/* + * Find the first available hole which can be used as a top-level. + */ +int +find_vdev_hole(spa_t *spa) +{ + vdev_t *rvd = spa->spa_root_vdev; + int c; + + ASSERT(spa_config_held(spa, SCL_VDEV, RW_READER) == SCL_VDEV); + + for (c = 0; c < rvd->vdev_children; c++) { + vdev_t *cvd = rvd->vdev_child[c]; + + if (cvd->vdev_ishole) + break; + } + return (c); +} + +/* + * Verify that vdev_add() works as expected. + */ +/* ARGSUSED */ +void +ztest_vdev_add_remove(ztest_ds_t *zd, uint64_t id) +{ + ztest_shared_t *zs = ztest_shared; + spa_t *spa = zs->zs_spa; + uint64_t leaves; + uint64_t guid; + nvlist_t *nvroot; + int error; + + VERIFY(mutex_lock(&zs->zs_vdev_lock) == 0); + leaves = MAX(zs->zs_mirrors + zs->zs_splits, 1) * zopt_raidz; + + spa_config_enter(spa, SCL_VDEV, FTAG, RW_READER); + + ztest_shared->zs_vdev_next_leaf = find_vdev_hole(spa) * leaves; + + /* + * If we have slogs then remove them 1/4 of the time. + */ + if (spa_has_slogs(spa) && ztest_random(4) == 0) { + /* + * Grab the guid from the head of the log class rotor. + */ + guid = spa_log_class(spa)->mc_rotor->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 + * dmu_objset_destroy() to fail with EBUSY thus + * leaving the dataset in an inconsistent state. + */ + VERIFY(rw_wrlock(&ztest_shared->zs_name_lock) == 0); + error = spa_vdev_remove(spa, guid, B_FALSE); + VERIFY(rw_unlock(&ztest_shared->zs_name_lock) == 0); + + if (error && error != EEXIST) + fatal(0, "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, zopt_vdev_size, 0, + ztest_random(4) == 0, zopt_raidz, zs->zs_mirrors, 1); + + error = spa_vdev_add(spa, nvroot); + nvlist_free(nvroot); + + if (error == ENOSPC) + ztest_record_enospc("spa_vdev_add"); + else if (error != 0) + fatal(0, "spa_vdev_add() = %d", error); + } + + VERIFY(mutex_unlock(&ztest_shared->zs_vdev_lock) == 0); +} + +/* + * Verify that adding/removing aux devices (l2arc, hot spare) works as expected. + */ +/* ARGSUSED */ +void +ztest_vdev_aux_add_remove(ztest_ds_t *zd, uint64_t id) +{ + ztest_shared_t *zs = ztest_shared; + spa_t *spa = zs->zs_spa; + vdev_t *rvd = spa->spa_root_vdev; + spa_aux_vdev_t *sav; + char *aux; + uint64_t guid = 0; + int error; + + if (ztest_random(2) == 0) { + sav = &spa->spa_spares; + aux = ZPOOL_CONFIG_SPARES; + } else { + sav = &spa->spa_l2cache; + aux = ZPOOL_CONFIG_L2CACHE; + } + + VERIFY(mutex_lock(&zs->zs_vdev_lock) == 0); + + spa_config_enter(spa, SCL_VDEV, FTAG, RW_READER); + + if (sav->sav_count != 0 && ztest_random(4) == 0) { + /* + * Pick a random device to remove. + */ + guid = sav->sav_vdevs[ztest_random(sav->sav_count)]->vdev_guid; + } else { + /* + * Find an unused device we can add. + */ + zs->zs_vdev_aux = 0; + for (;;) { + char path[MAXPATHLEN]; + int c; + (void) sprintf(path, ztest_aux_template, zopt_dir, + zopt_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, + (zopt_vdev_size * 5) / 4, 0, 0, 0, 0, 1); + error = spa_vdev_add(spa, nvroot); + if (error != 0) + fatal(0, "spa_vdev_add(%p) = %d", nvroot, error); + nvlist_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); + if (error != 0 && error != EBUSY) + fatal(0, "spa_vdev_remove(%llu) = %d", guid, error); + } + + VERIFY(mutex_unlock(&zs->zs_vdev_lock) == 0); +} + +/* + * split a pool if it has mirror tlvdevs + */ +/* ARGSUSED */ +void +ztest_split_pool(ztest_ds_t *zd, uint64_t id) +{ + ztest_shared_t *zs = ztest_shared; + spa_t *spa = zs->zs_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; + + VERIFY(mutex_lock(&zs->zs_vdev_lock) == 0); + + /* ensure we have a useable config; mirrors of raidz aren't supported */ + if (zs->zs_mirrors < 3 || zopt_raidz > 1) { + VERIFY(mutex_unlock(&zs->zs_vdev_lock) == 0); + 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); + VERIFY(nvlist_lookup_nvlist(spa->spa_config, ZPOOL_CONFIG_VDEV_TREE, + &tree) == 0); + mutex_exit(&spa->spa_props_lock); + + VERIFY(nvlist_lookup_nvlist_array(tree, ZPOOL_CONFIG_CHILDREN, &child, + &children) == 0); + + 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) { + VERIFY(nvlist_alloc(&schild[schildren], NV_UNIQUE_NAME, + 0) == 0); + VERIFY(nvlist_add_string(schild[schildren], + ZPOOL_CONFIG_TYPE, VDEV_TYPE_HOLE) == 0); + VERIFY(nvlist_add_uint64(schild[schildren], + ZPOOL_CONFIG_IS_HOLE, 1) == 0); + if (lastlogid == 0) + lastlogid = schildren; + ++schildren; + continue; + } + lastlogid = 0; + VERIFY(nvlist_lookup_nvlist_array(child[c], + ZPOOL_CONFIG_CHILDREN, &mchild, &mchildren) == 0); + VERIFY(nvlist_dup(mchild[0], &schild[schildren++], 0) == 0); + } + + /* OK, create a config that can be used to split */ + VERIFY(nvlist_alloc(&split, NV_UNIQUE_NAME, 0) == 0); + VERIFY(nvlist_add_string(split, ZPOOL_CONFIG_TYPE, + VDEV_TYPE_ROOT) == 0); + VERIFY(nvlist_add_nvlist_array(split, ZPOOL_CONFIG_CHILDREN, schild, + lastlogid != 0 ? lastlogid : schildren) == 0); + + VERIFY(nvlist_alloc(&config, NV_UNIQUE_NAME, 0) == 0); + VERIFY(nvlist_add_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, split) == 0); + + for (c = 0; c < schildren; c++) + nvlist_free(schild[c]); + free(schild); + nvlist_free(split); + + spa_config_exit(spa, SCL_VDEV, FTAG); + + (void) rw_wrlock(&zs->zs_name_lock); + error = spa_vdev_split_mirror(spa, "splitp", config, NULL, B_FALSE); + (void) rw_unlock(&zs->zs_name_lock); + + nvlist_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; + } + VERIFY(mutex_unlock(&zs->zs_vdev_lock) == 0); + +} + +/* + * Verify that we can attach and detach devices. + */ +/* ARGSUSED */ +void +ztest_vdev_attach_detach(ztest_ds_t *zd, uint64_t id) +{ + ztest_shared_t *zs = ztest_shared; + spa_t *spa = zs->zs_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; + size_t oldsize, newsize; + char oldpath[MAXPATHLEN], newpath[MAXPATHLEN]; + int replacing; + int oldvd_has_siblings = B_FALSE; + int newvd_is_spare = B_FALSE; + int oldvd_is_log; + int error, expected_error; + + VERIFY(mutex_lock(&zs->zs_vdev_lock) == 0); + leaves = MAX(zs->zs_mirrors, 1) * zopt_raidz; + + spa_config_enter(spa, SCL_VDEV, FTAG, RW_READER); + + /* + * 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]; + if (zs->zs_mirrors >= 1) { + ASSERT(oldvd->vdev_ops == &vdev_mirror_ops); + ASSERT(oldvd->vdev_children >= zs->zs_mirrors); + oldvd = oldvd->vdev_child[leaf / zopt_raidz]; + } + if (zopt_raidz > 1) { + ASSERT(oldvd->vdev_ops == &vdev_raidz_ops); + ASSERT(oldvd->vdev_children == zopt_raidz); + oldvd = oldvd->vdev_child[leaf % zopt_raidz]; + } + + /* + * 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; + ASSERT(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) strcpy(oldpath, oldvd->vdev_path); + pvd = oldvd->vdev_parent; + pguid = pvd->vdev_guid; + + /* + * If oldvd has siblings, then half of the time, detach it. + */ + if (oldvd_has_siblings && ztest_random(2) == 0) { + spa_config_exit(spa, SCL_VDEV, FTAG); + error = spa_vdev_detach(spa, oldguid, pguid, B_FALSE); + if (error != 0 && error != ENODEV && error != EBUSY && + error != ENOTSUP) + fatal(0, "detach (%s) returned %d", oldpath, error); + VERIFY(mutex_unlock(&zs->zs_vdev_lock) == 0); + return; + } + + /* + * 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; + (void) strcpy(newpath, newvd->vdev_path); + } else { + (void) snprintf(newpath, sizeof (newpath), ztest_dev_template, + zopt_dir, zopt_pool, top * leaves + leaf); + if (ztest_random(2) == 0) + newpath[strlen(newpath) - 1] = 'b'; + newvd = vdev_lookup_by_path(rvd, newpath); + } + + if (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 (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 (newsize < oldsize) + expected_error = EOVERFLOW; + else if (ashift > oldvd->vdev_top->vdev_ashift) + expected_error = EDOM; + else + expected_error = 0; + + spa_config_exit(spa, SCL_VDEV, FTAG); + + /* + * Build the nvlist describing newpath. + */ + root = make_vdev_root(newpath, NULL, newvd == NULL ? newsize : 0, + ashift, 0, 0, 0, 1); + + error = spa_vdev_attach(spa, oldguid, root, replacing); + + nvlist_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; + + /* XXX workaround 6690467 */ + if (error != expected_error && expected_error != EBUSY) { + fatal(0, "attach (%s %llu, %s %llu, %d) " + "returned %d, expected %d", + oldpath, (longlong_t)oldsize, newpath, + (longlong_t)newsize, replacing, error, expected_error); + } + + VERIFY(mutex_unlock(&zs->zs_vdev_lock) == 0); +} + +/* + * Callback function which expands the physical size of the vdev. + */ +vdev_t * +grow_vdev(vdev_t *vd, void *arg) +{ + spa_t *spa = vd->vdev_spa; + size_t *newsize = arg; + size_t fsize; + int fd; + + ASSERT(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); + (void) ftruncate(fd, *newsize); + + if (zopt_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(). + */ +/* ARGSUSED */ +vdev_t * +online_vdev(vdev_t *vd, 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; + + ASSERT(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 (zopt_verbose >= 5) { + (void) printf("Unable to expand vdev, state %llu, " + "error %d\n", (u_longlong_t)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 (zopt_verbose >= 5) { + (void) printf("vdev configuration has changed, " + "guid %llu, state %llu, expected gen %llu, " + "got gen %llu\n", + (u_longlong_t)guid, + (u_longlong_t)tvd->vdev_state, + (u_longlong_t)generation, + (u_longlong_t)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. + */ +vdev_t * +vdev_walk_tree(vdev_t *vd, vdev_t *(*func)(vdev_t *, void *), void *arg) +{ + if (vd->vdev_ops->vdev_op_leaf) { + if (func == NULL) + return (vd); + else + return (func(vd, arg)); + } + + for (uint_t 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. + */ +/* ARGSUSED */ +void +ztest_vdev_LUN_growth(ztest_ds_t *zd, uint64_t id) +{ + ztest_shared_t *zs = ztest_shared; + spa_t *spa = zs->zs_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; + + VERIFY(mutex_lock(&zs->zs_vdev_lock) == 0); + spa_config_enter(spa, SCL_STATE, spa, RW_READER); + + 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 * zopt_vdev_size) { + spa_config_exit(spa, SCL_STATE, spa); + VERIFY(mutex_unlock(&zs->zs_vdev_lock) == 0); + return; + } + ASSERT(psize > 0); + newsize = psize + psize / 8; + ASSERT3U(newsize, >, psize); + + if (zopt_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 (zopt_verbose >= 5) { + (void) printf("Could not expand LUN because " + "the vdev configuration changed.\n"); + } + spa_config_exit(spa, SCL_STATE, spa); + VERIFY(mutex_unlock(&zs->zs_vdev_lock) == 0); + 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 (zopt_verbose >= 5) { + (void) printf("Could not verify LUN expansion due to " + "intervening vdev offline or remove.\n"); + } + spa_config_exit(spa, SCL_STATE, spa); + VERIFY(mutex_unlock(&zs->zs_vdev_lock) == 0); + return; + } + + /* + * Make sure we were able to grow the vdev. + */ + if (new_ms_count <= old_ms_count) + fatal(0, "LUN expansion failed: ms_count %llu <= %llu\n", + old_ms_count, new_ms_count); + + /* + * Make sure we were able to grow the pool. + */ + if (new_class_space <= old_class_space) + fatal(0, "LUN expansion failed: class_space %llu <= %llu\n", + old_class_space, new_class_space); + + if (zopt_verbose >= 5) { + char oldnumbuf[6], newnumbuf[6]; + + nicenum(old_class_space, oldnumbuf); + nicenum(new_class_space, newnumbuf); + (void) printf("%s grew from %s to %s\n", + spa->spa_name, oldnumbuf, newnumbuf); + } + + spa_config_exit(spa, SCL_STATE, spa); + VERIFY(mutex_unlock(&zs->zs_vdev_lock) == 0); +} + +/* + * Verify that dmu_objset_{create,destroy,open,close} work as expected. + */ +/* ARGSUSED */ +static void +ztest_objset_create_cb(objset_t *os, void *arg, cred_t *cr, dmu_tx_t *tx) +{ + /* + * Create the objects common to all ztest datasets. + */ + VERIFY(zap_create_claim(os, ZTEST_DIROBJ, + DMU_OT_ZAP_OTHER, DMU_OT_NONE, 0, tx) == 0); +} + +static int +ztest_dataset_create(char *dsname) +{ + uint64_t zilset = ztest_random(100); + int err = dmu_objset_create(dsname, DMU_OST_OTHER, 0, + ztest_objset_create_cb, NULL); + + if (err || zilset < 80) + return (err); + + (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)); +} + +/* ARGSUSED */ +static int +ztest_objset_destroy_cb(const char *name, void *arg) +{ + objset_t *os; + dmu_object_info_t doi; + int error; + + /* + * Verify that the dataset contains a directory object. + */ + VERIFY3U(0, ==, dmu_objset_hold(name, FTAG, &os)); + error = dmu_object_info(os, ZTEST_DIROBJ, &doi); + if (error != ENOENT) { + /* We could have crashed in the middle of destroying it */ + ASSERT3U(error, ==, 0); + ASSERT3U(doi.doi_type, ==, DMU_OT_ZAP_OTHER); + ASSERT3S(doi.doi_physical_blocks_512, >=, 0); + } + dmu_objset_rele(os, FTAG); + + /* + * Destroy the dataset. + */ + VERIFY3U(0, ==, dmu_objset_destroy(name, B_FALSE)); + return (0); +} + +static boolean_t +ztest_snapshot_create(char *osname, uint64_t id) +{ + char snapname[MAXNAMELEN]; + int error; + + (void) snprintf(snapname, MAXNAMELEN, "%s@%llu", osname, + (u_longlong_t)id); + + error = dmu_objset_snapshot(osname, strchr(snapname, '@') + 1, + NULL, NULL, B_FALSE, B_FALSE, -1); + if (error == ENOSPC) { + ztest_record_enospc(FTAG); + return (B_FALSE); + } + if (error != 0 && error != EEXIST) + fatal(0, "ztest_snapshot_create(%s) = %d", snapname, error); + return (B_TRUE); +} + +static boolean_t +ztest_snapshot_destroy(char *osname, uint64_t id) +{ + char snapname[MAXNAMELEN]; + int error; + + (void) snprintf(snapname, MAXNAMELEN, "%s@%llu", osname, + (u_longlong_t)id); + + error = dmu_objset_destroy(snapname, B_FALSE); + if (error != 0 && error != ENOENT) + fatal(0, "ztest_snapshot_destroy(%s) = %d", snapname, error); + return (B_TRUE); +} + +/* ARGSUSED */ +void +ztest_dmu_objset_create_destroy(ztest_ds_t *zd, uint64_t id) +{ + ztest_shared_t *zs = ztest_shared; + ztest_ds_t zdtmp; + int iters; + int error; + objset_t *os, *os2; + char name[MAXNAMELEN]; + zilog_t *zilog; + + (void) rw_rdlock(&zs->zs_name_lock); + + (void) snprintf(name, MAXNAMELEN, "%s/temp_%llu", + zs->zs_pool, (u_longlong_t)id); + + /* + * If this dataset exists from a previous run, process its replay log + * half of the time. If we don't replay it, then dmu_objset_destroy() + * (invoked from ztest_objset_destroy_cb()) should just throw it away. + */ + if (ztest_random(2) == 0 && + dmu_objset_own(name, DMU_OST_OTHER, B_FALSE, FTAG, &os) == 0) { + ztest_zd_init(&zdtmp, os); + zil_replay(os, &zdtmp, ztest_replay_vector); + ztest_zd_fini(&zdtmp); + dmu_objset_disown(os, 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, ==, dmu_objset_hold(name, FTAG, &os)); + + /* + * Verify that we can create a new dataset. + */ + error = ztest_dataset_create(name); + if (error) { + if (error == ENOSPC) { + ztest_record_enospc(FTAG); + (void) rw_unlock(&zs->zs_name_lock); + return; + } + fatal(0, "dmu_objset_create(%s) = %d", name, error); + } + + VERIFY3U(0, ==, + dmu_objset_own(name, DMU_OST_OTHER, B_FALSE, FTAG, &os)); + + ztest_zd_init(&zdtmp, os); + + /* + * Open the intent log for it. + */ + zilog = zil_open(os, ztest_get_data); + + /* + * 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 (int 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)); + + /* + * Verify that we can hold an objset that is also owned. + */ + VERIFY3U(0, ==, dmu_objset_hold(name, FTAG, &os2)); + dmu_objset_rele(os2, FTAG); + + /* + * Verify that we cannot own an objset that is already owned. + */ + VERIFY3U(EBUSY, ==, + dmu_objset_own(name, DMU_OST_OTHER, B_FALSE, FTAG, &os2)); + + zil_close(zilog); + dmu_objset_disown(os, FTAG); + ztest_zd_fini(&zdtmp); + + (void) rw_unlock(&zs->zs_name_lock); +} + +/* + * Verify that dmu_snapshot_{create,destroy,open,close} work as expected. + */ +void +ztest_dmu_snapshot_create_destroy(ztest_ds_t *zd, uint64_t id) +{ + ztest_shared_t *zs = ztest_shared; + + (void) rw_rdlock(&zs->zs_name_lock); + (void) ztest_snapshot_destroy(zd->zd_name, id); + (void) ztest_snapshot_create(zd->zd_name, id); + (void) rw_unlock(&zs->zs_name_lock); +} + +/* + * Cleanup non-standard snapshots and clones. + */ +void +ztest_dsl_dataset_cleanup(char *osname, uint64_t id) +{ + char snap1name[MAXNAMELEN]; + char clone1name[MAXNAMELEN]; + char snap2name[MAXNAMELEN]; + char clone2name[MAXNAMELEN]; + char snap3name[MAXNAMELEN]; + int error; + + (void) snprintf(snap1name, MAXNAMELEN, "%s@s1_%llu", osname, id); + (void) snprintf(clone1name, MAXNAMELEN, "%s/c1_%llu", osname, id); + (void) snprintf(snap2name, MAXNAMELEN, "%s@s2_%llu", clone1name, id); + (void) snprintf(clone2name, MAXNAMELEN, "%s/c2_%llu", osname, id); + (void) snprintf(snap3name, MAXNAMELEN, "%s@s3_%llu", clone1name, id); + + error = dmu_objset_destroy(clone2name, B_FALSE); + if (error && error != ENOENT) + fatal(0, "dmu_objset_destroy(%s) = %d", clone2name, error); + error = dmu_objset_destroy(snap3name, B_FALSE); + if (error && error != ENOENT) + fatal(0, "dmu_objset_destroy(%s) = %d", snap3name, error); + error = dmu_objset_destroy(snap2name, B_FALSE); + if (error && error != ENOENT) + fatal(0, "dmu_objset_destroy(%s) = %d", snap2name, error); + error = dmu_objset_destroy(clone1name, B_FALSE); + if (error && error != ENOENT) + fatal(0, "dmu_objset_destroy(%s) = %d", clone1name, error); + error = dmu_objset_destroy(snap1name, B_FALSE); + if (error && error != ENOENT) + fatal(0, "dmu_objset_destroy(%s) = %d", snap1name, error); +} + +/* + * Verify dsl_dataset_promote handles EBUSY + */ +void +ztest_dsl_dataset_promote_busy(ztest_ds_t *zd, uint64_t id) +{ + ztest_shared_t *zs = ztest_shared; + objset_t *clone; + dsl_dataset_t *ds; + char snap1name[MAXNAMELEN]; + char clone1name[MAXNAMELEN]; + char snap2name[MAXNAMELEN]; + char clone2name[MAXNAMELEN]; + char snap3name[MAXNAMELEN]; + char *osname = zd->zd_name; + int error; + + (void) rw_rdlock(&zs->zs_name_lock); + + ztest_dsl_dataset_cleanup(osname, id); + + (void) snprintf(snap1name, MAXNAMELEN, "%s@s1_%llu", osname, id); + (void) snprintf(clone1name, MAXNAMELEN, "%s/c1_%llu", osname, id); + (void) snprintf(snap2name, MAXNAMELEN, "%s@s2_%llu", clone1name, id); + (void) snprintf(clone2name, MAXNAMELEN, "%s/c2_%llu", osname, id); + (void) snprintf(snap3name, MAXNAMELEN, "%s@s3_%llu", clone1name, id); + + error = dmu_objset_snapshot(osname, strchr(snap1name, '@')+1, + NULL, NULL, B_FALSE, B_FALSE, -1); + if (error && error != EEXIST) { + if (error == ENOSPC) { + ztest_record_enospc(FTAG); + goto out; + } + fatal(0, "dmu_take_snapshot(%s) = %d", snap1name, error); + } + + error = dmu_objset_hold(snap1name, FTAG, &clone); + if (error) + fatal(0, "dmu_open_snapshot(%s) = %d", snap1name, error); + + error = dmu_objset_clone(clone1name, dmu_objset_ds(clone), 0); + dmu_objset_rele(clone, FTAG); + if (error) { + if (error == ENOSPC) { + ztest_record_enospc(FTAG); + goto out; + } + fatal(0, "dmu_objset_create(%s) = %d", clone1name, error); + } + + error = dmu_objset_snapshot(clone1name, strchr(snap2name, '@')+1, + NULL, NULL, B_FALSE, B_FALSE, -1); + if (error && error != EEXIST) { + if (error == ENOSPC) { + ztest_record_enospc(FTAG); + goto out; + } + fatal(0, "dmu_open_snapshot(%s) = %d", snap2name, error); + } + + error = dmu_objset_snapshot(clone1name, strchr(snap3name, '@')+1, + NULL, NULL, B_FALSE, B_FALSE, -1); + if (error && error != EEXIST) { + if (error == ENOSPC) { + ztest_record_enospc(FTAG); + goto out; + } + fatal(0, "dmu_open_snapshot(%s) = %d", snap3name, error); + } + + error = dmu_objset_hold(snap3name, FTAG, &clone); + if (error) + fatal(0, "dmu_open_snapshot(%s) = %d", snap3name, error); + + error = dmu_objset_clone(clone2name, dmu_objset_ds(clone), 0); + dmu_objset_rele(clone, FTAG); + if (error) { + if (error == ENOSPC) { + ztest_record_enospc(FTAG); + goto out; + } + fatal(0, "dmu_objset_create(%s) = %d", clone2name, error); + } + + error = dsl_dataset_own(snap2name, B_FALSE, FTAG, &ds); + if (error) + fatal(0, "dsl_dataset_own(%s) = %d", snap2name, error); + error = dsl_dataset_promote(clone2name, NULL); + if (error != EBUSY) + fatal(0, "dsl_dataset_promote(%s), %d, not EBUSY", clone2name, + error); + dsl_dataset_disown(ds, FTAG); + +out: + ztest_dsl_dataset_cleanup(osname, id); + + (void) rw_unlock(&zs->zs_name_lock); +} + +/* + * 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[4]; + int batchsize = sizeof (od) / sizeof (od[0]); + + for (int b = 0; b < batchsize; b++) + ztest_od_init(&od[b], id, FTAG, b, DMU_OT_UINT64_OTHER, 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, sizeof (od), 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); +} + +/* + * Verify that dmu_{read,write} work as expected. + */ +void +ztest_dmu_read_write(ztest_ds_t *zd, uint64_t id) +{ + objset_t *os = zd->zd_os; + ztest_od_t od[2]; + dmu_tx_t *tx; + int i, freeit, error; + uint64_t 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[0], id, FTAG, 0, DMU_OT_UINT64_OTHER, 0, chunksize); + ztest_od_init(&od[1], id, FTAG, 1, DMU_OT_UINT64_OTHER, 0, chunksize); + + if (ztest_object_init(zd, od, sizeof (od), B_FALSE) != 0) + return; + + bigobj = od[0].od_object; + packobj = od[1].od_object; + chunksize = od[0].od_gen; + ASSERT(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, n * chunksize, s * chunksize); + + /* + * 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); + ASSERT3U(error, ==, 0); + error = dmu_read(os, bigobj, bigoff, bigsize, bigbuf, + DMU_READ_PREFETCH); + ASSERT3U(error, ==, 0); + + /* + * 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); + + txg = ztest_tx_assign(tx, TXG_MIGHTWAIT, FTAG); + if (txg == 0) { + umem_free(packbuf, packsize); + umem_free(bigbuf, bigsize); + return; + } + + dmu_object_set_checksum(os, bigobj, + (enum zio_checksum)ztest_random_dsl_prop(ZFS_PROP_CHECKSUM), tx); + + dmu_object_set_compress(os, bigobj, + (enum zio_compress)ztest_random_dsl_prop(ZFS_PROP_COMPRESSION), 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; + + ASSERT((uintptr_t)bigH - (uintptr_t)bigbuf < bigsize); + ASSERT((uintptr_t)bigT - (uintptr_t)bigbuf < bigsize); + + if (pack->bw_txg > txg) + fatal(0, "future leak: got %llx, open txg is %llx", + pack->bw_txg, txg); + + if (pack->bw_data != 0 && pack->bw_index != n + i) + fatal(0, "wrong index: got %llx, wanted %llx+%llx", + pack->bw_index, n, i); + + if (bcmp(pack, bigH, sizeof (bufwad_t)) != 0) + fatal(0, "pack/bigH mismatch in %p/%p", pack, bigH); + + if (bcmp(pack, bigT, sizeof (bufwad_t)) != 0) + fatal(0, "pack/bigT mismatch in %p/%p", pack, bigT); + + if (freeit) { + bzero(pack, 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 (zopt_verbose >= 7) { + (void) printf("freeing offset %llx size %llx" + " txg %llx\n", + (u_longlong_t)bigoff, + (u_longlong_t)bigsize, + (u_longlong_t)txg); + } + VERIFY(0 == dmu_free_range(os, bigobj, bigoff, bigsize, tx)); + } else { + if (zopt_verbose >= 7) { + (void) printf("writing offset %llx size %llx" + " txg %llx\n", + (u_longlong_t)bigoff, + (u_longlong_t)bigsize, + (u_longlong_t)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); + + VERIFY(0 == dmu_read(os, packobj, packoff, + packsize, packcheck, DMU_READ_PREFETCH)); + VERIFY(0 == dmu_read(os, bigobj, bigoff, + bigsize, bigcheck, DMU_READ_PREFETCH)); + + ASSERT(bcmp(packbuf, packcheck, packsize) == 0); + ASSERT(bcmp(bigbuf, bigcheck, bigsize) == 0); + + umem_free(packcheck, packsize); + umem_free(bigcheck, bigsize); + } + + umem_free(packbuf, packsize); + umem_free(bigbuf, bigsize); +} + +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; + + ASSERT((uintptr_t)bigH - (uintptr_t)bigbuf < bigsize); + ASSERT((uintptr_t)bigT - (uintptr_t)bigbuf < bigsize); + + if (pack->bw_txg > txg) + fatal(0, "future leak: got %llx, open txg is %llx", + pack->bw_txg, txg); + + if (pack->bw_data != 0 && pack->bw_index != n + i) + fatal(0, "wrong index: got %llx, wanted %llx+%llx", + pack->bw_index, n, i); + + if (bcmp(pack, bigH, sizeof (bufwad_t)) != 0) + fatal(0, "pack/bigH mismatch in %p/%p", pack, bigH); + + if (bcmp(pack, bigT, sizeof (bufwad_t)) != 0) + fatal(0, "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; + } +} + +void +ztest_dmu_read_write_zcopy(ztest_ds_t *zd, uint64_t id) +{ + objset_t *os = zd->zd_os; + ztest_od_t od[2]; + dmu_tx_t *tx; + uint64_t i; + int error; + 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; + + /* + * 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() can be tested for object updates. + */ + + /* + * Read the directory info. If it's the first time, set things up. + */ + ztest_od_init(&od[0], id, FTAG, 0, DMU_OT_UINT64_OTHER, blocksize, 0); + ztest_od_init(&od[1], id, FTAG, 1, DMU_OT_UINT64_OTHER, 0, chunksize); + + if (ztest_object_init(zd, od, sizeof (od), B_FALSE) != 0) + return; + + bigobj = od[0].od_object; + packobj = od[1].od_object; + blocksize = od[0].od_blocksize; + chunksize = blocksize; + ASSERT(chunksize == od[1].od_gen); + + VERIFY(dmu_object_info(os, bigobj, &doi) == 0); + VERIFY(ISP2(doi.doi_data_block_size)); + VERIFY(chunksize == doi.doi_data_block_size); + VERIFY(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); + + VERIFY3U(0, ==, 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() when it can't directly + * assign an arcbuf to a dbuf. + */ + for (j = 0; j < s; j++) { + if (i != 5) { + 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) { + 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 *)); + dmu_buf_rele(bonus_db, FTAG); + return; + } + + /* + * 50% of the time don't read objects in the 1st iteration to + * test dmu_assign_arcbuf() for the case when there're 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); + ASSERT3U(error, ==, 0); + error = dmu_read(os, bigobj, bigoff, bigsize, + bigbuf, DMU_READ_PREFETCH); + ASSERT3U(error, ==, 0); + } + 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 (zopt_verbose >= 7) { + (void) printf("writing offset %llx size %llx" + " txg %llx\n", + (u_longlong_t)bigoff, + (u_longlong_t)bigsize, + (u_longlong_t)txg); + } + for (off = bigoff, j = 0; j < s; j++, off += chunksize) { + dmu_buf_t *dbt; + if (i != 5) { + bcopy((caddr_t)bigbuf + (off - bigoff), + bigbuf_arcbufs[j]->b_data, chunksize); + } else { + bcopy((caddr_t)bigbuf + (off - bigoff), + bigbuf_arcbufs[2 * j]->b_data, + chunksize / 2); + bcopy((caddr_t)bigbuf + (off - bigoff) + + chunksize / 2, + bigbuf_arcbufs[2 * j + 1]->b_data, + chunksize / 2); + } + + if (i == 1) { + VERIFY(dmu_buf_hold(os, bigobj, off, + FTAG, &dbt, DMU_READ_NO_PREFETCH) == 0); + } + if (i != 5) { + dmu_assign_arcbuf(bonus_db, off, + bigbuf_arcbufs[j], tx); + } else { + dmu_assign_arcbuf(bonus_db, off, + bigbuf_arcbufs[2 * j], tx); + dmu_assign_arcbuf(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); + + VERIFY(0 == dmu_read(os, packobj, packoff, + packsize, packcheck, DMU_READ_PREFETCH)); + VERIFY(0 == dmu_read(os, bigobj, bigoff, + bigsize, bigcheck, DMU_READ_PREFETCH)); + + ASSERT(bcmp(packbuf, packcheck, packsize) == 0); + ASSERT(bcmp(bigbuf, bigcheck, bigsize) == 0); + + umem_free(packcheck, packsize); + umem_free(bigcheck, bigsize); + } + if (i == 2) { + txg_wait_open(dmu_objset_pool(os), 0); + } 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 *)); +} + +/* ARGSUSED */ +void +ztest_dmu_write_parallel(ztest_ds_t *zd, uint64_t id) +{ + ztest_od_t od[1]; + 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[0], ID_PARALLEL, FTAG, 0, DMU_OT_UINT64_OTHER, 0, 0); + + if (ztest_object_init(zd, od, sizeof (od), B_FALSE) != 0) + return; + + while (ztest_random(10) != 0) + ztest_io(zd, od[0].od_object, offset); +} + +void +ztest_dmu_prealloc(ztest_ds_t *zd, uint64_t id) +{ + ztest_od_t od[1]; + 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; + + ztest_od_init(&od[0], id, FTAG, 0, DMU_OT_UINT64_OTHER, blocksize, 0); + + if (ztest_object_init(zd, od, sizeof (od), !ztest_random(2)) != 0) + return; + + if (ztest_truncate(zd, od[0].od_object, offset, count * blocksize) != 0) + return; + + ztest_prealloc(zd, od[0].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[0].od_object, randoff, blocksize, + data) != 0) + break; + while (ztest_random(4) != 0) + ztest_io(zd, od[0].od_object, randoff); + } + + umem_free(data, blocksize); +} + +/* + * 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[1]; + 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; + char *hc[2] = { "s.acl.h", ".s.open.h.hyLZlg" }; + + ztest_od_init(&od[0], id, FTAG, 0, DMU_OT_ZAP_OTHER, 0, 0); + + if (ztest_object_init(zd, od, sizeof (od), !ztest_random(2)) != 0) + return; + + object = od[0].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) + return; + for (i = 0; i < 2; i++) { + value[i] = i; + VERIFY3U(0, ==, 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)); + VERIFY3U(0, ==, + 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++) { + VERIFY3U(0, ==, zap_remove(os, object, hc[i], tx)); + } + dmu_tx_commit(tx); + + /* + * Generate a buch 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_%llu", (u_longlong_t)prop); + (void) sprintf(txgname, "txg_%llu", (u_longlong_t)prop); + bzero(value, 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); + + VERIFY(zap_lookup(os, object, txgname, zl_intsize, + zl_ints, &last_txg) == 0); + + VERIFY(zap_length(os, object, propname, &zl_intsize, + &zl_ints) == 0); + + ASSERT3U(zl_intsize, ==, sizeof (uint64_t)); + ASSERT3U(zl_ints, ==, ints); + + VERIFY(zap_lookup(os, object, propname, zl_intsize, + zl_ints, value) == 0); + + 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) + return; + + if (last_txg > txg) + fatal(0, "zap future leak: old %llu new %llu", last_txg, txg); + + for (i = 0; i < ints; i++) + value[i] = txg + object + i; + + VERIFY3U(0, ==, zap_update(os, object, txgname, sizeof (uint64_t), + 1, &txg, tx)); + VERIFY3U(0, ==, 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_%llu", (u_longlong_t)prop); + (void) sprintf(txgname, "txg_%llu", (u_longlong_t)prop); + + error = zap_length(os, object, txgname, &zl_intsize, &zl_ints); + + if (error == ENOENT) + return; + + ASSERT3U(error, ==, 0); + + 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) + return; + VERIFY3U(0, ==, zap_remove(os, object, txgname, tx)); + VERIFY3U(0, ==, zap_remove(os, object, propname, tx)); + dmu_tx_commit(tx); +} + +/* + * Testcase 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[1]; + uint64_t object, txg; + + ztest_od_init(&od[0], id, FTAG, 0, DMU_OT_ZAP_OTHER, 0, 0); + + if (ztest_object_init(zd, od, sizeof (od), !ztest_random(2)) != 0) + return; + + object = od[0].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 (int i = 0; i < 2050; i++) { + char name[MAXNAMELEN]; + uint64_t value = i; + dmu_tx_t *tx; + int error; + + (void) snprintf(name, sizeof (name), "fzap-%llu-%llu", + 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) + return; + error = zap_add(os, object, name, sizeof (uint64_t), 1, + &value, tx); + ASSERT(error == 0 || error == EEXIST); + dmu_tx_commit(tx); + } +} + +/* ARGSUSED */ +void +ztest_zap_parallel(ztest_ds_t *zd, uint64_t id) +{ + objset_t *os = zd->zd_os; + ztest_od_t od[1]; + 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; + + ztest_od_init(&od[0], ID_PARALLEL, FTAG, micro, DMU_OT_ZAP_OTHER, 0, 0); + + if (ztest_object_init(zd, od, sizeof (od), B_FALSE) != 0) + return; + + object = od[0].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; + VERIFY(zap_count(os, object, &count) == 0); + ASSERT(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) + return; + bcopy(name, string_value, namelen); + } else { + tx = NULL; + txg = 0; + bzero(string_value, 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 && + bcmp(name, data, namelen) != 0) + fatal(0, "name '%s' != val '%s' len %d", + name, 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: + VERIFY(zap_update(os, object, name, wsize, wc, data, tx) == 0); + break; + + case 4: + error = zap_remove(os, object, name, tx); + ASSERT(error == 0 || error == ENOENT); + break; + } + + if (tx != NULL) + dmu_tx_commit(tx); +} + +/* + * 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; + + VERIFY(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(0, "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) { + ASSERT3U(data->zcd_txg, ==, 0); + 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; + } + + /* Was this callback added to the global callback list? */ + if (!data->zcd_added) + goto out; + + ASSERT3U(data->zcd_txg, !=, 0); + + /* Remove our callback from the list */ + (void) mutex_lock(&zcl.zcl_callbacks_lock); + list_remove(&zcl.zcl_callbacks, data); + (void) mutex_unlock(&zcl.zcl_callbacks_lock); + +out: + 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); + + return (cb_data); +} + +/* + * 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_CALLBACK_THRESH (TXG_CONCURRENT_STATES + 2) + +/* + * 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[1]; + dmu_tx_t *tx; + ztest_cb_data_t *cb_data[3], *tmp_cb; + uint64_t old_txg, txg; + int i, error; + + ztest_od_init(&od[0], id, FTAG, 0, DMU_OT_UINT64_OTHER, 0, 0); + + if (ztest_object_init(zd, od, sizeof (od), B_FALSE) != 0) + 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[0].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)); + } + + 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. + */ + VERIFY(0 == dmu_read(os, od[0].od_object, 0, sizeof (uint64_t), + &old_txg, DMU_READ_PREFETCH)); + + if (old_txg > txg) + fatal(0, "future leak: got %" PRIu64 ", open txg is %" PRIu64, + old_txg, txg); + + dmu_write(os, od[0].od_object, 0, sizeof (uint64_t), &txg, tx); + + (void) mutex_lock(&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 > txg - ZTEST_COMMIT_CALLBACK_THRESH) { + fatal(0, "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]; + } + + (void) mutex_unlock(&zcl.zcl_callbacks_lock); + + dmu_tx_commit(tx); +} + +/* ARGSUSED */ +void +ztest_dsl_prop_get_set(ztest_ds_t *zd, uint64_t id) +{ + zfs_prop_t proplist[] = { + ZFS_PROP_CHECKSUM, + ZFS_PROP_COMPRESSION, + ZFS_PROP_COPIES, + ZFS_PROP_DEDUP + }; + ztest_shared_t *zs = ztest_shared; + + (void) rw_rdlock(&zs->zs_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)); + + (void) rw_unlock(&zs->zs_name_lock); +} + +/* ARGSUSED */ +void +ztest_spa_prop_get_set(ztest_ds_t *zd, uint64_t id) +{ + ztest_shared_t *zs = ztest_shared; + nvlist_t *props = NULL; + + (void) rw_rdlock(&zs->zs_name_lock); + + (void) ztest_spa_prop_set_uint64(zs, ZPOOL_PROP_DEDUPDITTO, + ZIO_DEDUPDITTO_MIN + ztest_random(ZIO_DEDUPDITTO_MIN)); + + VERIFY3U(spa_prop_get(zs->zs_spa, &props), ==, 0); + + if (zopt_verbose >= 6) + dump_nvlist(props, 4); + + nvlist_free(props); + + (void) rw_unlock(&zs->zs_name_lock); +} + +/* + * 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[MAXNAMELEN]; + + (void) rw_rdlock(&ztest_shared->zs_name_lock); + + dmu_objset_name(os, osname); + + (void) snprintf(snapname, 100, "sh1_%llu", id); + (void) snprintf(fullname, 100, "%s@%s", osname, snapname); + (void) snprintf(clonename, 100, "%s/ch1_%llu", osname, id); + (void) snprintf(tag, 100, "%tag_%llu", id); + + /* + * Clean up from any previous run. + */ + (void) dmu_objset_destroy(clonename, B_FALSE); + (void) dsl_dataset_user_release(osname, snapname, tag, B_FALSE); + (void) dmu_objset_destroy(fullname, B_FALSE); + + /* + * Create snapshot, clone it, mark snap for deferred destroy, + * destroy clone, verify snap was also destroyed. + */ + error = dmu_objset_snapshot(osname, snapname, NULL, NULL, FALSE, + FALSE, -1); + if (error) { + if (error == ENOSPC) { + ztest_record_enospc("dmu_objset_snapshot"); + goto out; + } + fatal(0, "dmu_objset_snapshot(%s) = %d", fullname, error); + } + + error = dmu_objset_hold(fullname, FTAG, &origin); + if (error) + fatal(0, "dmu_objset_hold(%s) = %d", fullname, error); + + error = dmu_objset_clone(clonename, dmu_objset_ds(origin), 0); + dmu_objset_rele(origin, FTAG); + if (error) { + if (error == ENOSPC) { + ztest_record_enospc("dmu_objset_clone"); + goto out; + } + fatal(0, "dmu_objset_clone(%s) = %d", clonename, error); + } + + error = dmu_objset_destroy(fullname, B_TRUE); + if (error) { + fatal(0, "dmu_objset_destroy(%s, B_TRUE) = %d", + fullname, error); + } + + error = dmu_objset_destroy(clonename, B_FALSE); + if (error) + fatal(0, "dmu_objset_destroy(%s) = %d", clonename, error); + + error = dmu_objset_hold(fullname, FTAG, &origin); + if (error != ENOENT) + fatal(0, "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(osname, snapname, NULL, NULL, FALSE, + FALSE, -1); + if (error) { + if (error == ENOSPC) { + ztest_record_enospc("dmu_objset_snapshot"); + goto out; + } + fatal(0, "dmu_objset_snapshot(%s) = %d", fullname, error); + } + + error = dsl_dataset_user_hold(osname, snapname, tag, B_FALSE, + B_TRUE, -1); + if (error) + fatal(0, "dsl_dataset_user_hold(%s)", fullname, tag); + + error = dmu_objset_destroy(fullname, B_FALSE); + if (error != EBUSY) { + fatal(0, "dmu_objset_destroy(%s, B_FALSE) = %d", + fullname, error); + } + + error = dmu_objset_destroy(fullname, B_TRUE); + if (error) { + fatal(0, "dmu_objset_destroy(%s, B_TRUE) = %d", + fullname, error); + } + + error = dsl_dataset_user_release(osname, snapname, tag, B_FALSE); + if (error) + fatal(0, "dsl_dataset_user_release(%s)", fullname, tag); + + VERIFY(dmu_objset_hold(fullname, FTAG, &origin) == ENOENT); + +out: + (void) rw_unlock(&ztest_shared->zs_name_lock); +} + +/* + * Inject random faults into the on-disk data. + */ +/* ARGSUSED */ +void +ztest_fault_inject(ztest_ds_t *zd, uint64_t id) +{ + ztest_shared_t *zs = ztest_shared; + spa_t *spa = zs->zs_spa; + int fd; + uint64_t offset; + uint64_t leaves; + uint64_t bad = 0x1990c0ffeedecade; + uint64_t top, leaf; + char path0[MAXPATHLEN]; + char pathrand[MAXPATHLEN]; + size_t fsize; + int bshift = SPA_MAXBLOCKSHIFT + 2; /* don't scrog all labels */ + int iters = 1000; + int maxfaults; + int mirror_save; + vdev_t *vd0 = NULL; + uint64_t guid0 = 0; + boolean_t islog = B_FALSE; + + VERIFY(mutex_lock(&zs->zs_vdev_lock) == 0); + maxfaults = MAXFAULTS(); + leaves = MAX(zs->zs_mirrors, 1) * zopt_raidz; + mirror_save = zs->zs_mirrors; + VERIFY(mutex_unlock(&zs->zs_vdev_lock) == 0); + + ASSERT(leaves >= 1); + + /* + * 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, sizeof (path0), ztest_dev_template, + zopt_dir, zopt_pool, top * leaves + zs->zs_splits); + (void) snprintf(pathrand, sizeof (pathrand), ztest_dev_template, + zopt_dir, zopt_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 (vd0 != NULL && maxfaults != 1) { + /* + * Make vd0 explicitly claim to be unreadable, + * or unwriteable, 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; + + if (vf != NULL && ztest_random(3) == 0) { + (void) close(vf->vf_vnode->v_fd); + vf->vf_vnode->v_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); + return; + } + vd0 = sav->sav_vdevs[ztest_random(sav->sav_count)]; + guid0 = vd0->vdev_guid; + (void) strcpy(path0, vd0->vdev_path); + (void) strcpy(pathrand, vd0->vdev_path); + + leaf = 0; + leaves = 1; + maxfaults = INT_MAX; /* no limit on cache devices */ + } + + spa_config_exit(spa, SCL_STATE, FTAG); + + /* + * 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 + * dmu_objset_destroy() to fail with EBUSY thus + * leaving the dataset in an inconsistent state. + */ + if (islog) + (void) rw_wrlock(&ztest_shared->zs_name_lock); + + VERIFY(vdev_offline(spa, guid0, flags) != EBUSY); + + if (islog) + (void) rw_unlock(&ztest_shared->zs_name_lock); + } else { + (void) vdev_online(spa, guid0, 0, NULL); + } + } + + if (maxfaults == 0) + return; + + /* + * 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 */ + return; + + fsize = lseek(fd, 0, SEEK_END); + + while (--iters != 0) { + offset = ztest_random(fsize / (leaves << bshift)) * + (leaves << bshift) + (leaf << bshift) + + (ztest_random(1ULL << (bshift - 1)) & -8ULL); + + if (offset >= fsize) + continue; + + VERIFY(mutex_lock(&zs->zs_vdev_lock) == 0); + if (mirror_save != zs->zs_mirrors) { + VERIFY(mutex_unlock(&zs->zs_vdev_lock) == 0); + (void) close(fd); + return; + } + + if (pwrite(fd, &bad, sizeof (bad), offset) != sizeof (bad)) + fatal(1, "can't inject bad word at 0x%llx in %s", + offset, pathrand); + + VERIFY(mutex_unlock(&zs->zs_vdev_lock) == 0); + + if (zopt_verbose >= 7) + (void) printf("injected bad word into %s," + " offset 0x%llx\n", pathrand, (u_longlong_t)offset); + } + + (void) close(fd); +} + +/* + * Verify that DDT repair works as expected. + */ +void +ztest_ddt_repair(ztest_ds_t *zd, uint64_t id) +{ + ztest_shared_t *zs = ztest_shared; + spa_t *spa = zs->zs_spa; + objset_t *os = zd->zd_os; + ztest_od_t od[1]; + uint64_t object, blocksize, txg, pattern, psize; + enum zio_checksum checksum = spa_dedup_checksum(spa); + dmu_buf_t *db; + dmu_tx_t *tx; + void *buf; + blkptr_t blk; + int copies = 2 * ZIO_DEDUPDITTO_MIN; + + blocksize = ztest_random_blocksize(); + blocksize = MIN(blocksize, 2048); /* because we write so many */ + + ztest_od_init(&od[0], id, FTAG, 0, DMU_OT_UINT64_OTHER, blocksize, 0); + + if (ztest_object_init(zd, od, sizeof (od), B_FALSE) != 0) + return; + + /* + * Take the name lock as writer to prevent anyone else from changing + * the pool and dataset properies we need to maintain during this test. + */ + (void) rw_wrlock(&zs->zs_name_lock); + + if (ztest_dsl_prop_set_uint64(zd->zd_name, ZFS_PROP_DEDUP, checksum, + B_FALSE) != 0 || + ztest_dsl_prop_set_uint64(zd->zd_name, ZFS_PROP_COPIES, 1, + B_FALSE) != 0) { + (void) rw_unlock(&zs->zs_name_lock); + return; + } + + object = od[0].od_object; + blocksize = od[0].od_blocksize; + pattern = spa_guid(spa) ^ dmu_objset_fsid_guid(os); + + ASSERT(object != 0); + + tx = dmu_tx_create(os); + dmu_tx_hold_write(tx, object, 0, copies * blocksize); + txg = ztest_tx_assign(tx, TXG_WAIT, FTAG); + if (txg == 0) { + (void) rw_unlock(&zs->zs_name_lock); + return; + } + + /* + * Write all the copies of our block. + */ + for (int i = 0; i < copies; i++) { + uint64_t offset = i * blocksize; + VERIFY(dmu_buf_hold(os, object, offset, FTAG, &db, + DMU_READ_NO_PREFETCH) == 0); + ASSERT(db->db_offset == offset); + ASSERT(db->db_size == blocksize); + ASSERT(ztest_pattern_match(db->db_data, db->db_size, pattern) || + ztest_pattern_match(db->db_data, db->db_size, 0ULL)); + dmu_buf_will_fill(db, tx); + ztest_pattern_set(db->db_data, db->db_size, pattern); + dmu_buf_rele(db, FTAG); + } + + dmu_tx_commit(tx); + txg_wait_synced(spa_get_dsl(spa), txg); + + /* + * Find out what block we got. + */ + VERIFY(dmu_buf_hold(os, object, 0, FTAG, &db, + DMU_READ_NO_PREFETCH) == 0); + blk = *((dmu_buf_impl_t *)db)->db_blkptr; + dmu_buf_rele(db, FTAG); + + /* + * Damage the block. Dedup-ditto will save us when we read it later. + */ + psize = BP_GET_PSIZE(&blk); + buf = zio_buf_alloc(psize); + ztest_pattern_set(buf, psize, ~pattern); + + (void) zio_wait(zio_rewrite(NULL, spa, 0, &blk, + buf, psize, NULL, NULL, ZIO_PRIORITY_SYNC_WRITE, + ZIO_FLAG_CANFAIL | ZIO_FLAG_INDUCE_DAMAGE, NULL)); + + zio_buf_free(buf, psize); + + (void) rw_unlock(&zs->zs_name_lock); +} + +/* + * Scrub the pool. + */ +/* ARGSUSED */ +void +ztest_scrub(ztest_ds_t *zd, uint64_t id) +{ + ztest_shared_t *zs = ztest_shared; + spa_t *spa = zs->zs_spa; + + (void) spa_scan(spa, POOL_SCAN_SCRUB); + (void) poll(NULL, 0, 100); /* wait a moment, then force a restart */ + (void) spa_scan(spa, POOL_SCAN_SCRUB); +} + +/* + * Rename the pool to a different name and then rename it back. + */ +/* ARGSUSED */ +void +ztest_spa_rename(ztest_ds_t *zd, uint64_t id) +{ + ztest_shared_t *zs = ztest_shared; + char *oldname, *newname; + spa_t *spa; + + (void) rw_wrlock(&zs->zs_name_lock); + + oldname = zs->zs_pool; + newname = umem_alloc(strlen(oldname) + 5, UMEM_NOFAIL); + (void) strcpy(newname, oldname); + (void) strcat(newname, "_tmp"); + + /* + * Do the rename + */ + VERIFY3U(0, ==, spa_rename(oldname, newname)); + + /* + * Try to open it under the old name, which shouldn't exist + */ + VERIFY3U(ENOENT, ==, spa_open(oldname, &spa, FTAG)); + + /* + * Open it under the new name and make sure it's still the same spa_t. + */ + VERIFY3U(0, ==, spa_open(newname, &spa, FTAG)); + + ASSERT(spa == zs->zs_spa); + spa_close(spa, FTAG); + + /* + * Rename it back to the original + */ + VERIFY3U(0, ==, spa_rename(newname, oldname)); + + /* + * Make sure it can still be opened + */ + VERIFY3U(0, ==, spa_open(oldname, &spa, FTAG)); + + ASSERT(spa == zs->zs_spa); + spa_close(spa, FTAG); + + umem_free(newname, strlen(newname) + 1); + + (void) rw_unlock(&zs->zs_name_lock); +} + +/* + * Verify pool integrity by running zdb. + */ +static void +ztest_run_zdb(char *pool) +{ + int status; + char zdb[MAXPATHLEN + MAXNAMELEN + 20]; + char zbuf[1024]; + char *bin; + char *ztest; + char *isa; + int isalen; + FILE *fp; + + (void) realpath(getexecname(), zdb); + + /* zdb lives in /usr/sbin, while ztest lives in /usr/bin */ + bin = strstr(zdb, "/usr/bin/"); + ztest = strstr(bin, "/ztest"); + isa = bin + 8; + isalen = ztest - isa; + isa = strdup(isa); + /* LINTED */ + (void) sprintf(bin, + "/usr/sbin%.*s/zdb -bcc%s%s -U %s %s", + isalen, + isa, + zopt_verbose >= 3 ? "s" : "", + zopt_verbose >= 4 ? "v" : "", + spa_config_path, + pool); + free(isa); + + if (zopt_verbose >= 5) + (void) printf("Executing %s\n", strstr(zdb, "zdb ")); + + fp = popen(zdb, "r"); + + while (fgets(zbuf, sizeof (zbuf), fp) != NULL) + if (zopt_verbose >= 3) + (void) printf("%s", zbuf); + + status = pclose(fp); + + if (status == 0) + return; + + ztest_dump_core = 0; + if (WIFEXITED(status)) + fatal(0, "'%s' exit code %d", zdb, WEXITSTATUS(status)); + else + fatal(0, "'%s' died with signal %d", zdb, WTERMSIG(status)); +} + +static void +ztest_walk_pool_directory(char *header) +{ + spa_t *spa = NULL; + + if (zopt_verbose >= 6) + (void) printf("%s\n", header); + + mutex_enter(&spa_namespace_lock); + while ((spa = spa_next(spa)) != NULL) + if (zopt_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; + + if (zopt_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. + */ + VERIFY3U(0, ==, 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. + */ + VERIFY3U(0, ==, spa_export(oldname, &config, B_FALSE, B_FALSE)); + + ztest_walk_pool_directory("pools after export"); + + /* + * Try to import it. + */ + newconfig = spa_tryimport(config); + ASSERT(newconfig != NULL); + nvlist_free(newconfig); + + /* + * Import it under the new name. + */ + VERIFY3U(0, ==, spa_import(newname, config, NULL, 0)); + + 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. + */ + VERIFY3U(0, ==, spa_open(newname, &spa, FTAG)); + ASSERT(pool_guid == spa_guid(spa)); + spa_close(spa, FTAG); + + nvlist_free(config); +} + +static void +ztest_resume(spa_t *spa) +{ + if (spa_suspended(spa) && zopt_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 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); + } + return (NULL); +} + +static void * +ztest_deadman_thread(void *arg) +{ + ztest_shared_t *zs = arg; + int grace = 300; + hrtime_t delta; + + delta = (zs->zs_thread_stop - zs->zs_thread_start) / NANOSEC + grace; + + (void) poll(NULL, 0, (int)(1000 * delta)); + + fatal(0, "failed to complete within %d seconds of deadline", grace); + + return (NULL); +} + +static void +ztest_execute(ztest_info_t *zi, uint64_t id) +{ + ztest_shared_t *zs = ztest_shared; + ztest_ds_t *zd = &zs->zs_zd[id % zopt_datasets]; + hrtime_t functime = gethrtime(); + + for (int i = 0; i < zi->zi_iters; i++) + zi->zi_func(zd, id); + + functime = gethrtime() - functime; + + atomic_add_64(&zi->zi_call_count, 1); + atomic_add_64(&zi->zi_call_time, functime); + + if (zopt_verbose >= 4) { + Dl_info dli; + (void) dladdr((void *)zi->zi_func, &dli); + (void) printf("%6.2f sec in %s\n", + (double)functime / NANOSEC, dli.dli_sname); + } +} + +static void * +ztest_thread(void *arg) +{ + uint64_t id = (uintptr_t)arg; + ztest_shared_t *zs = ztest_shared; + uint64_t call_next; + hrtime_t now; + ztest_info_t *zi; + + 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. + */ + zi = &zs->zs_info[ztest_random(ZTEST_FUNCS)]; + call_next = zi->zi_call_next; + + if (now >= call_next && + atomic_cas_64(&zi->zi_call_next, call_next, call_next + + ztest_random(2 * zi->zi_interval[0] + 1)) == call_next) + ztest_execute(zi, id); + } + + return (NULL); +} + +static void +ztest_dataset_name(char *dsname, char *pool, int d) +{ + (void) snprintf(dsname, MAXNAMELEN, "%s/ds_%d", pool, d); +} + +static void +ztest_dataset_destroy(ztest_shared_t *zs, int d) +{ + char name[MAXNAMELEN]; + + ztest_dataset_name(name, zs->zs_pool, d); + + if (zopt_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 (int t = d; t < zopt_threads; t += zopt_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. + */ + VERIFY3U(0, ==, 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(ztest_shared_t *zs, int d) +{ + ztest_ds_t *zd = &zs->zs_zd[d]; + uint64_t committed_seq = zd->zd_seq; + objset_t *os; + zilog_t *zilog; + char name[MAXNAMELEN]; + int error; + + ztest_dataset_name(name, zs->zs_pool, d); + + (void) rw_rdlock(&zs->zs_name_lock); + + error = ztest_dataset_create(name); + if (error == ENOSPC) { + (void) rw_unlock(&zs->zs_name_lock); + ztest_record_enospc(FTAG); + return (error); + } + ASSERT(error == 0 || error == EEXIST); + + VERIFY3U(dmu_objset_hold(name, zd, &os), ==, 0); + (void) rw_unlock(&zs->zs_name_lock); + + ztest_zd_init(zd, os); + + zilog = zd->zd_zilog; + + if (zilog->zl_header->zh_claim_lr_seq != 0 && + zilog->zl_header->zh_claim_lr_seq < committed_seq) + fatal(0, "missing log records: claimed %llu < committed %llu", + 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 (zopt_verbose >= 6) + (void) printf("%s replay %llu blocks, %llu records, seq %llu\n", + zd->zd_name, + (u_longlong_t)zilog->zl_parse_blk_count, + (u_longlong_t)zilog->zl_parse_lr_count, + (u_longlong_t)zilog->zl_replaying_seq); + + zilog = zil_open(os, ztest_get_data); + + if (zilog->zl_replaying_seq != 0 && + zilog->zl_replaying_seq < committed_seq) + fatal(0, "missing log records: replayed %llu < committed %llu", + zilog->zl_replaying_seq, committed_seq); + + return (0); +} + +static void +ztest_dataset_close(ztest_shared_t *zs, int d) +{ + ztest_ds_t *zd = &zs->zs_zd[d]; + + zil_close(zd->zd_zilog); + dmu_objset_rele(zd->zd_os, zd); + + ztest_zd_fini(zd); +} + +/* + * Kick off threads to run tests on all datasets in parallel. + */ +static void +ztest_run(ztest_shared_t *zs) +{ + thread_t *tid; + spa_t *spa; + thread_t resume_tid; + int error; + + ztest_exiting = B_FALSE; + + /* + * Initialize parent/child shared state. + */ + VERIFY(_mutex_init(&zs->zs_vdev_lock, USYNC_THREAD, NULL) == 0); + VERIFY(rwlock_init(&zs->zs_name_lock, USYNC_THREAD, NULL) == 0); + + zs->zs_thread_start = gethrtime(); + zs->zs_thread_stop = zs->zs_thread_start + zopt_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) < zopt_killrate) + zs->zs_thread_kill -= ztest_random(zopt_passtime * NANOSEC); + + (void) _mutex_init(&zcl.zcl_callbacks_lock, USYNC_THREAD, NULL); + + list_create(&zcl.zcl_callbacks, sizeof (ztest_cb_data_t), + offsetof(ztest_cb_data_t, zcd_node)); + + /* + * Open our pool. + */ + kernel_init(FREAD | FWRITE); + VERIFY(spa_open(zs->zs_pool, &spa, FTAG) == 0); + zs->zs_spa = spa; + + spa->spa_dedup_ditto = 2 * ZIO_DEDUPDITTO_MIN; + + /* + * We don't expect the pool to suspend unless maxfaults == 0, + * in which case ztest_fault_inject() temporarily takes away + * the only valid replica. + */ + if (MAXFAULTS() == 0) + spa->spa_failmode = ZIO_FAILURE_MODE_WAIT; + else + spa->spa_failmode = ZIO_FAILURE_MODE_PANIC; + + /* + * Create a thread to periodically resume suspended I/O. + */ + VERIFY(thr_create(0, 0, ztest_resume_thread, spa, THR_BOUND, + &resume_tid) == 0); + + /* + * Create a deadman thread to abort() if we hang. + */ + VERIFY(thr_create(0, 0, ztest_deadman_thread, zs, THR_BOUND, + NULL) == 0); + + /* + * Verify that we can safely inquire about 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 (int t = 0; t < 64; t++) { + for (int 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(zopt_datasets); + ztest_dataset_destroy(zs, d); + } + zs->zs_enospc_count = 0; + + tid = umem_zalloc(zopt_threads * sizeof (thread_t), UMEM_NOFAIL); + + if (zopt_verbose >= 4) + (void) printf("starting main threads...\n"); + + /* + * Kick off all the tests that run in parallel. + */ + for (int t = 0; t < zopt_threads; t++) { + if (t < zopt_datasets && ztest_dataset_open(zs, t) != 0) + return; + VERIFY(thr_create(0, 0, ztest_thread, (void *)(uintptr_t)t, + THR_BOUND, &tid[t]) == 0); + } + + /* + * Wait for all of the tests to complete. We go in reverse order + * so we don't close datasets while threads are still using them. + */ + for (int t = zopt_threads - 1; t >= 0; t--) { + VERIFY(thr_join(tid[t], NULL, NULL) == 0); + if (t < zopt_datasets) + ztest_dataset_close(zs, 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(tid, zopt_threads * sizeof (thread_t)); + + /* Kill the resume thread */ + ztest_exiting = B_TRUE; + VERIFY(thr_join(resume_tid, NULL, NULL) == 0); + 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 (uint64_t object = 1; object < 50; object++) + dmu_prefetch(spa->spa_meta_objset, object, 0, 1ULL << 20); + + 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 (zopt_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) { + char name[MAXNAMELEN]; + (void) snprintf(name, MAXNAMELEN, "%s_import", zs->zs_pool); + ztest_spa_import_export(zs->zs_pool, name); + ztest_spa_import_export(name, zs->zs_pool); + } + + kernel_fini(); + + list_destroy(&zcl.zcl_callbacks); + + (void) _mutex_destroy(&zcl.zcl_callbacks_lock); + + (void) rwlock_destroy(&zs->zs_name_lock); + (void) _mutex_destroy(&zs->zs_vdev_lock); +} + +static void +ztest_freeze(ztest_shared_t *zs) +{ + ztest_ds_t *zd = &zs->zs_zd[0]; + spa_t *spa; + int numloops = 0; + + if (zopt_verbose >= 3) + (void) printf("testing spa_freeze()...\n"); + + kernel_init(FREAD | FWRITE); + VERIFY3U(0, ==, spa_open(zs->zs_pool, &spa, FTAG)); + VERIFY3U(0, ==, ztest_dataset_open(zs, 0)); + + /* + * 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); + + /* + * 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. + */ + while (ztest_random(10) != 0 && numloops++ < zopt_maxloops) { + ztest_dmu_write_parallel(zd, 0); + ztest_dmu_object_alloc_free(zd, 0); + 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(zs, 0); + spa_close(spa, FTAG); + kernel_fini(); + + /* + * Open and close the pool and dataset to induce log replay. + */ + kernel_init(FREAD | FWRITE); + VERIFY3U(0, ==, spa_open(zs->zs_pool, &spa, FTAG)); + VERIFY3U(0, ==, ztest_dataset_open(zs, 0)); + ztest_dataset_close(zs, 0); + spa_close(spa, FTAG); + kernel_fini(); +} + +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() +{ + nvlist_t *props; + + if (ztest_random(2) == 0) + return (NULL); + + VERIFY(nvlist_alloc(&props, NV_UNIQUE_NAME, 0) == 0); + VERIFY(nvlist_add_uint64(props, "autoreplace", 1) == 0); + + (void) printf("props:\n"); + dump_nvlist(props, 4); + + 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; + + VERIFY(_mutex_init(&zs->zs_vdev_lock, USYNC_THREAD, NULL) == 0); + VERIFY(rwlock_init(&zs->zs_name_lock, USYNC_THREAD, NULL) == 0); + + kernel_init(FREAD | FWRITE); + + /* + * Create the storage pool. + */ + (void) spa_destroy(zs->zs_pool); + ztest_shared->zs_vdev_next_leaf = 0; + zs->zs_splits = 0; + zs->zs_mirrors = zopt_mirrors; + nvroot = make_vdev_root(NULL, NULL, zopt_vdev_size, 0, + 0, zopt_raidz, zs->zs_mirrors, 1); + props = make_random_props(); + VERIFY3U(0, ==, spa_create(zs->zs_pool, nvroot, props, NULL, NULL)); + nvlist_free(nvroot); + + VERIFY3U(0, ==, spa_open(zs->zs_pool, &spa, FTAG)); + metaslab_sz = 1ULL << spa->spa_root_vdev->vdev_child[0]->vdev_ms_shift; + spa_close(spa, FTAG); + + kernel_fini(); + + ztest_run_zdb(zs->zs_pool); + + ztest_freeze(zs); + + ztest_run_zdb(zs->zs_pool); + + (void) rwlock_destroy(&zs->zs_name_lock); + (void) _mutex_destroy(&zs->zs_vdev_lock); +} + +int +main(int argc, char **argv) +{ + int kills = 0; + int iters = 0; + ztest_shared_t *zs; + size_t shared_size; + ztest_info_t *zi; + char timebuf[100]; + char numbuf[6]; + spa_t *spa; + + (void) setvbuf(stdout, NULL, _IOLBF, 0); + + ztest_random_fd = open("/dev/urandom", O_RDONLY); + + process_options(argc, argv); + + /* Override location of zpool.cache */ + (void) asprintf((char **)&spa_config_path, "%s/zpool.cache", zopt_dir); + + /* + * Blow away any existing copy of zpool.cache + */ + if (zopt_init != 0) + (void) remove(spa_config_path); + + shared_size = sizeof (*zs) + zopt_datasets * sizeof (ztest_ds_t); + + zs = ztest_shared = (void *)mmap(0, + P2ROUNDUP(shared_size, getpagesize()), + PROT_READ | PROT_WRITE, MAP_SHARED | MAP_ANON, -1, 0); + + if (zopt_verbose >= 1) { + (void) printf("%llu vdevs, %d datasets, %d threads," + " %llu seconds...\n", + (u_longlong_t)zopt_vdevs, zopt_datasets, zopt_threads, + (u_longlong_t)zopt_time); + } + + /* + * Create and initialize our storage pool. + */ + for (int i = 1; i <= zopt_init; i++) { + bzero(zs, sizeof (ztest_shared_t)); + if (zopt_verbose >= 3 && zopt_init != 1) + (void) printf("ztest_init(), pass %d\n", i); + zs->zs_pool = zopt_pool; + ztest_init(zs); + } + + zs->zs_pool = zopt_pool; + zs->zs_proc_start = gethrtime(); + zs->zs_proc_stop = zs->zs_proc_start + zopt_time * NANOSEC; + + for (int f = 0; f < ZTEST_FUNCS; f++) { + zi = &zs->zs_info[f]; + *zi = ztest_info[f]; + if (zs->zs_proc_start + zi->zi_interval[0] > zs->zs_proc_stop) + zi->zi_call_next = UINT64_MAX; + else + zi->zi_call_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; + pid_t pid; + + /* + * Initialize the workload counters for each function. + */ + for (int f = 0; f < ZTEST_FUNCS; f++) { + zi = &zs->zs_info[f]; + zi->zi_call_count = 0; + zi->zi_call_time = 0; + } + + /* Set the allocation switch size */ + metaslab_df_alloc_threshold = ztest_random(metaslab_sz / 4) + 1; + + pid = fork(); + + if (pid == -1) + fatal(1, "fork failed"); + + if (pid == 0) { /* child */ + struct rlimit rl = { 1024, 1024 }; + (void) setrlimit(RLIMIT_NOFILE, &rl); + (void) enable_extended_FILE_stdio(-1, -1); + ztest_run(zs); + exit(0); + } + + while (waitpid(pid, &status, 0) != pid) + continue; + + if (WIFEXITED(status)) { + if (WEXITSTATUS(status) != 0) { + (void) fprintf(stderr, + "child exited with code %d\n", + WEXITSTATUS(status)); + exit(2); + } + } else if (WIFSIGNALED(status)) { + if (WTERMSIG(status) != SIGKILL) { + (void) fprintf(stderr, + "child died with signal %d\n", + WTERMSIG(status)); + exit(3); + } + kills++; + } else { + (void) fprintf(stderr, "something strange happened " + "to child\n"); + exit(4); + } + + iters++; + + if (zopt_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); + + (void) printf("Pass %3d, %8s, %3llu ENOSPC, " + "%4.1f%% of %5s used, %3.0f%% done, %8s to go\n", + iters, + WIFEXITED(status) ? "Complete" : "SIGKILL", + (u_longlong_t)zs->zs_enospc_count, + 100.0 * zs->zs_alloc / zs->zs_space, + numbuf, + 100.0 * (now - zs->zs_proc_start) / + (zopt_time * NANOSEC), timebuf); + } + + if (zopt_verbose >= 2) { + (void) printf("\nWorkload summary:\n\n"); + (void) printf("%7s %9s %s\n", + "Calls", "Time", "Function"); + (void) printf("%7s %9s %s\n", + "-----", "----", "--------"); + for (int f = 0; f < ZTEST_FUNCS; f++) { + Dl_info dli; + + zi = &zs->zs_info[f]; + print_time(zi->zi_call_time, timebuf); + (void) dladdr((void *)zi->zi_func, &dli); + (void) printf("%7llu %9s %s\n", + (u_longlong_t)zi->zi_call_count, timebuf, + dli.dli_sname); + } + (void) printf("\n"); + } + + /* + * It's possible that we killed a child during a rename test, + * in which case we'll have a 'ztest_tmp' pool lying around + * instead of 'ztest'. Do a blind rename in case this happened. + */ + kernel_init(FREAD); + if (spa_open(zopt_pool, &spa, FTAG) == 0) { + spa_close(spa, FTAG); + } else { + char tmpname[MAXNAMELEN]; + kernel_fini(); + kernel_init(FREAD | FWRITE); + (void) snprintf(tmpname, sizeof (tmpname), "%s_tmp", + zopt_pool); + (void) spa_rename(tmpname, zopt_pool); + } + kernel_fini(); + + ztest_run_zdb(zopt_pool); + } + + if (zopt_verbose >= 1) { + (void) printf("%d killed, %d completed, %.0f%% kill rate\n", + kills, iters - kills, (100.0 * kills) / MAX(1, iters)); + } + + return (0); +} |