1
0
mirror of git://sourceware.org/git/lvm2.git synced 2024-12-21 13:34:40 +03:00
lvm2/lib/format_text/export.c

1120 lines
23 KiB
C
Raw Normal View History

/*
2008-01-30 17:00:02 +03:00
* Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.
lvconvert: add infrastructure for RaidLV reshaping support In order to support striped raid5/6/10 LV reshaping (change of LV type, stripesize or number of legs), this patch introduces infrastructure prerequisites to be used by raid_manip.c extensions in followup patches. This base is needed for allocation of out-of-place reshape space required by the MD raid personalities to avoid writing over data in-place when reading off the current RAID layout or number of legs and writing out the new layout or to a different number of legs (i.e. restripe) Changes: - add members reshape_len to 'struct lv_segment' to store out-of-place reshape length per component rimage - add member data_copies to struct lv_segment to support more than 2 raid10 data copies - make alloc_lv_segment() aware of both reshape_len and data_copies - adjust all alloc_lv_segment() callers to the new API - add functions to retrieve the current data offset (needed for out-of-place reshaping space allocation) and the devices count from the kernel - make libdm deptree code aware of reshape_len - add LV flags for disk add/remove reshaping - support import/export of the new 'struct lv_segment' members - enhance lv_extend/_lv_reduce to cope with reshape_len - add seg_is_*/segtype_is_* macros related to reshaping - add target version check for reshaping - grow rebuilds/writemostly bitmaps to 246 bit to support kernel maximal - enhance libdm deptree code to support data_offset (out-of-place reshaping) and delta_disk (legs add/remove reshaping) target arguments Related: rhbz834579 Related: rhbz1191935 Related: rhbz1191978
2017-02-24 02:50:00 +03:00
* Copyright (C) 2004-2017 Red Hat, Inc. All rights reserved.
*
2004-03-30 23:35:44 +04:00
* This file is part of LVM2.
*
* This copyrighted material is made available to anyone wishing to use,
* modify, copy, or redistribute it subject to the terms and conditions
* of the GNU Lesser General Public License v.2.1.
2004-03-30 23:35:44 +04:00
*
* You should have received a copy of the GNU Lesser General Public License
2004-03-30 23:35:44 +04:00
* along with this program; if not, write to the Free Software Foundation,
* Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
#include "base/memory/zalloc.h"
#include "lib/misc/lib.h"
#include "import-export.h"
2018-12-14 15:10:25 +03:00
#include "lvm-version.h"
#include "lib/metadata/metadata.h"
#include "lib/display/display.h"
#include "lib/misc/lvm-string.h"
#include "lib/metadata/segtype.h"
#include "lib/format_text/text_export.h"
#include "lib/commands/toolcontext.h"
device usage based on devices file The LVM devices file lists devices that lvm can use. The default file is /etc/lvm/devices/system.devices, and the lvmdevices(8) command is used to add or remove device entries. If the file does not exist, or if lvm.conf includes use_devicesfile=0, then lvm will not use a devices file. When the devices file is in use, the regex filter is not used, and the filter settings in lvm.conf or on the command line are ignored. LVM records devices in the devices file using hardware-specific IDs, such as the WWID, and attempts to use subsystem-specific IDs for virtual device types. These device IDs are also written in the VG metadata. When no hardware or virtual ID is available, lvm falls back using the unstable device name as the device ID. When devnames are used, lvm performs extra scanning to find devices if their devname changes, e.g. after reboot. When proper device IDs are used, an lvm command will not look at devices outside the devices file, but when devnames are used as a fallback, lvm will scan devices outside the devices file to locate PVs on renamed devices. A config setting search_for_devnames can be used to control the scanning for renamed devname entries. Related to the devices file, the new command option --devices <devnames> allows a list of devices to be specified for the command to use, overriding the devices file. The listed devices act as a sort of devices file in terms of limiting which devices lvm will see and use. Devices that are not listed will appear to be missing to the lvm command. Multiple devices files can be kept in /etc/lvm/devices, which allows lvm to be used with different sets of devices, e.g. system devices do not need to be exposed to a specific application, and the application can use lvm on its own set of devices that are not exposed to the system. The option --devicesfile <filename> is used to select the devices file to use with the command. Without the option set, the default system devices file is used. Setting --devicesfile "" causes lvm to not use a devices file. An existing, empty devices file means lvm will see no devices. The new command vgimportdevices adds PVs from a VG to the devices file and updates the VG metadata to include the device IDs. vgimportdevices -a will import all VGs into the system devices file. LVM commands run by dmeventd not use a devices file by default, and will look at all devices on the system. A devices file can be created for dmeventd (/etc/lvm/devices/dmeventd.devices) If this file exists, lvm commands run by dmeventd will use it. Internal implementaion: - device_ids_read - read the devices file . add struct dev_use (du) to cmd->use_devices for each devices file entry - dev_cache_scan - get /dev entries . add struct device (dev) to dev_cache for each device on the system - device_ids_match - match devices file entries to /dev entries . match each du on cmd->use_devices to a dev in dev_cache, using device ID . on match, set du->dev, dev->id, dev->flags MATCHED_USE_ID - label_scan - read lvm headers and metadata from devices . filters are applied, those that do not need data from the device . filter-deviceid skips devs without MATCHED_USE_ID, i.e. skips /dev entries that are not listed in the devices file . read lvm label from dev . filters are applied, those that use data from the device . read lvm metadata from dev . add info/vginfo structs for PVs/VGs (info is "lvmcache") - device_ids_find_renamed_devs - handle devices with unstable devname ID where devname changed . this step only needed when devs do not have proper device IDs, and their dev names change, e.g. after reboot sdb becomes sdc. . detect incorrect match because PVID in the devices file entry does not match the PVID found when the device was read above . undo incorrect match between du and dev above . search system devices for new location of PVID . update devices file with new devnames for PVIDs on renamed devices . label_scan the renamed devs - continue with command processing
2020-06-23 21:25:41 +03:00
#include "lib/device/device_id.h"
#include "libdaemon/client/config-util.h"
2001-12-20 14:52:54 +03:00
#include <stdarg.h>
#include <time.h>
2002-11-18 17:04:08 +03:00
#include <sys/utsname.h>
2002-11-18 17:04:08 +03:00
struct formatter;
__attribute__((format(printf, 3, 0)))
2002-11-18 17:04:08 +03:00
typedef int (*out_with_comment_fn) (struct formatter * f, const char *comment,
const char *fmt, va_list ap);
typedef int (*nl_fn) (struct formatter * f);
/*
* Macro for formatted output.
* out_with_comment_fn returns -1 if data didn't fit and buffer was expanded.
* Then argument list is reset and out_with_comment_fn is called again.
*/
#define _out_with_comment(f, buffer, fmt, ap) \
2008-01-30 17:00:02 +03:00
do { \
va_start(ap, fmt); \
2017-07-19 17:17:30 +03:00
r = (f)->out_with_comment((f), (buffer), (fmt), ap); \
va_end(ap); \
} while (r == -1)
2002-11-18 17:04:08 +03:00
/*
* The first half of this file deals with
* exporting the vg, ie. writing it to a file.
*/
struct formatter {
struct dm_pool *mem; /* pv names allocated from here */
struct dm_hash_table *pv_names; /* dev_name -> pv_name (eg, pv1) */
2002-11-18 17:04:08 +03:00
union {
FILE *fp; /* where we're writing to */
struct {
char *start;
2002-11-18 17:04:08 +03:00
uint32_t size;
uint32_t used;
} buf;
} data;
out_with_comment_fn out_with_comment;
nl_fn nl;
2002-11-18 17:04:08 +03:00
int indent; /* current level of indentation */
int error;
2002-11-18 17:04:08 +03:00
int header; /* 1 => comments at start; 0 => end */
};
2002-11-18 17:04:08 +03:00
static struct utsname _utsname;
static void _init(void)
{
static int _initialised = 0;
if (_initialised)
return;
if (uname(&_utsname)) {
log_error("uname failed: %s", strerror(errno));
memset(&_utsname, 0, sizeof(_utsname));
}
_initialised = 1;
}
/*
* Formatting functions.
*/
#define MAX_INDENT 5
static void _inc_indent(struct formatter *f)
{
if (++f->indent > MAX_INDENT)
f->indent = MAX_INDENT;
}
static void _dec_indent(struct formatter *f)
{
if (!f->indent--) {
log_error(INTERNAL_ERROR "problem tracking indentation");
f->indent = 0;
}
}
/*
* Newline function for prettier layout.
*/
static int _nl_file(struct formatter *f)
{
2002-11-18 17:04:08 +03:00
fprintf(f->data.fp, "\n");
return 1;
2002-11-18 17:04:08 +03:00
}
static int _extend_buffer(struct formatter *f)
2002-11-18 17:04:08 +03:00
{
char *newbuf;
2002-11-18 17:04:08 +03:00
2017-12-11 18:32:53 +03:00
log_debug_metadata("Doubling metadata output buffer to " FMTu32,
f->data.buf.size * 2);
if (!(newbuf = realloc(f->data.buf.start,
2018-04-20 18:43:50 +03:00
f->data.buf.size * 2))) {
log_error("Buffer reallocation failed.");
return 0;
}
memset(newbuf + f->data.buf.size, 0, f->data.buf.size);
f->data.buf.start = newbuf;
f->data.buf.size *= 2;
return 1;
}
static int _nl_raw(struct formatter *f)
{
/* If metadata doesn't fit, extend buffer */
if ((f->data.buf.used + 2 > f->data.buf.size) &&
2008-01-30 16:19:47 +03:00
(!_extend_buffer(f)))
return_0;
*(f->data.buf.start + f->data.buf.used) = '\n';
2002-11-18 17:04:08 +03:00
f->data.buf.used += 1;
*(f->data.buf.start + f->data.buf.used) = '\0';
return 1;
}
#define COMMENT_TAB 6
__attribute__((format(printf, 3, 0)))
2002-11-18 17:04:08 +03:00
static int _out_with_comment_file(struct formatter *f, const char *comment,
const char *fmt, va_list ap)
{
int i;
char white_space[MAX_INDENT + 1];
2002-11-18 17:04:08 +03:00
if (ferror(f->data.fp))
return 0;
for (i = 0; i < f->indent; i++)
white_space[i] = '\t';
white_space[i] = '\0';
fputs(white_space, f->data.fp);
2002-11-18 17:04:08 +03:00
i = vfprintf(f->data.fp, fmt, ap);
if (comment) {
/*
* line comments up if possible.
*/
i += 8 * f->indent;
i /= 8;
i++;
do
2002-11-18 17:04:08 +03:00
fputc('\t', f->data.fp);
while (++i < COMMENT_TAB);
fputs(comment, f->data.fp);
}
2002-11-18 17:04:08 +03:00
fputc('\n', f->data.fp);
return 1;
}
__attribute__((format(printf, 3, 0)))
static int _out_with_comment_raw(struct formatter *f,
const char *comment __attribute__((unused)),
2002-11-18 17:04:08 +03:00
const char *fmt, va_list ap)
{
int n;
va_list apc;
2002-11-18 17:04:08 +03:00
va_copy(apc, ap);
n = vsnprintf(f->data.buf.start + f->data.buf.used,
f->data.buf.size - f->data.buf.used, fmt, apc);
va_end(apc);
2002-11-18 17:04:08 +03:00
/* If metadata doesn't fit, extend buffer */
if (n < 0 || (n + f->data.buf.used + 2 > f->data.buf.size)) {
2008-01-30 16:19:47 +03:00
if (!_extend_buffer(f))
return_0;
return -1; /* Retry */
}
2002-11-18 17:04:08 +03:00
f->data.buf.used += n;
outnl(f);
2002-11-18 17:04:08 +03:00
return 1;
}
/*
* Formats a string, converting a size specified
* in 512-byte sectors to a more human readable
* form (eg, megabytes). We may want to lift this
* for other code to use.
*/
static int _sectors_to_units(uint64_t sectors, char *buffer, size_t s)
{
static const char *_units[] = {
"Kilobytes",
"Megabytes",
"Gigabytes",
2002-11-18 17:04:08 +03:00
"Terabytes",
"Petabytes",
"Exabytes",
NULL
};
int i;
double d = (double) sectors;
/* to convert to K */
d /= 2.0;
for (i = 0; (d > 1024.0) && _units[i]; i++)
d /= 1024.0;
2006-08-21 16:54:53 +04:00
return dm_snprintf(buffer, s, "# %g %s", d, _units[i]) > 0;
}
/* increment indention level */
void out_inc_indent(struct formatter *f)
{
_inc_indent(f);
}
/* decrement indention level */
void out_dec_indent(struct formatter *f)
{
_dec_indent(f);
}
/* insert new line */
int out_newline(struct formatter *f)
{
return f->nl(f);
}
/*
* Appends a comment giving a size in more easily
* readable form (eg, 4M instead of 8096).
*/
2004-05-05 01:25:57 +04:00
int out_size(struct formatter *f, uint64_t size, const char *fmt, ...)
{
char buffer[64];
va_list ap;
2002-11-18 17:04:08 +03:00
int r;
2002-11-18 17:04:08 +03:00
if (!_sectors_to_units(size, buffer, sizeof(buffer)))
return 0;
_out_with_comment(f, buffer, fmt, ap);
2002-11-18 17:04:08 +03:00
return r;
}
/*
* Appends a comment indicating that the line is
* only a hint.
*/
2004-05-05 01:25:57 +04:00
int out_hint(struct formatter *f, const char *fmt, ...)
{
va_list ap;
2002-11-18 17:04:08 +03:00
int r;
_out_with_comment(f, "# Hint only", fmt, ap);
2002-11-18 17:04:08 +03:00
return r;
}
/*
* The normal output function with comment
*/
int out_text_with_comment(struct formatter *f, const char *comment, const char *fmt, ...)
{
va_list ap;
int r;
_out_with_comment(f, comment, fmt, ap);
return r;
}
/*
* The normal output function.
*/
2004-05-05 01:25:57 +04:00
int out_text(struct formatter *f, const char *fmt, ...)
{
va_list ap;
2002-11-18 17:04:08 +03:00
int r;
_out_with_comment(f, NULL, fmt, ap);
2002-11-18 17:04:08 +03:00
return r;
}
static int _out_line(const char *line, void *_f) {
struct formatter *f = (struct formatter *) _f;
return out_text(f, "%s", line);
}
int out_config_node(struct formatter *f, const struct dm_config_node *cn)
{
return dm_config_write_node(cn, _out_line, f);
}
static int _print_header(struct cmd_context *cmd, struct formatter *f,
const char *desc)
{
char *buf;
2001-12-20 14:52:54 +03:00
time_t t;
t = time(NULL);
outf(f, "# Generated by LVM2 version %s: %s", LVM_VERSION, ctime(&t));
2004-05-05 01:25:57 +04:00
outf(f, CONTENTS_FIELD " = \"" CONTENTS_VALUE "\"");
outf(f, FORMAT_VERSION_FIELD " = %d", FORMAT_VERSION_VALUE);
outnl(f);
buf = alloca(dm_escaped_len(desc));
outf(f, "description = \"%s\"", dm_escape_double_quotes(buf, desc));
outnl(f);
outf(f, "creation_host = \"%s\"\t# %s %s %s %s %s", _utsname.nodename,
_utsname.sysname, _utsname.nodename, _utsname.release,
_utsname.version, _utsname.machine);
if (cmd->system_id && *cmd->system_id)
outf(f, "creation_host_system_id = \"%s\"", cmd->system_id);
outf(f, "creation_time = " FMTu64 "\t# %s", (uint64_t)t, ctime(&t));
return 1;
}
static int _print_flag_config(struct formatter *f, uint64_t status, int type)
{
char buffer[4096];
if (!print_flags(buffer, sizeof(buffer), type, STATUS_FLAG, status))
return_0;
outf(f, "status = %s", buffer);
if (!print_flags(buffer, sizeof(buffer), type, COMPATIBLE_FLAG, status))
return_0;
outf(f, "flags = %s", buffer);
return 1;
}
static char *_alloc_printed_str_list(struct dm_list *list)
{
struct dm_str_list *sl;
int first = 1;
size_t size = 0;
char *buffer, *buf;
dm_list_iterate_items(sl, list)
/* '"' + item + '"' + ',' + ' ' */
size += strlen(sl->str) + 4;
/* '[' + ']' + '\0' */
size += 3;
if (!(buffer = buf = malloc(size))) {
log_error("Could not allocate memory for string list buffer.");
return NULL;
}
if (!emit_to_buffer(&buf, &size, "["))
goto_bad;
dm_list_iterate_items(sl, list) {
if (!first) {
if (!emit_to_buffer(&buf, &size, ", "))
goto_bad;
} else
first = 0;
if (!emit_to_buffer(&buf, &size, "\"%s\"", sl->str))
goto_bad;
}
if (!emit_to_buffer(&buf, &size, "]"))
goto_bad;
return buffer;
bad:
free(buffer);
return_NULL;
}
static int _out_list(struct formatter *f, struct dm_list *list,
const char *list_name)
{
char *buffer;
if (!dm_list_empty(list)) {
if (!(buffer = _alloc_printed_str_list(list)))
return_0;
if (!out_text(f, "%s = %s", list_name, buffer)) {
free(buffer);
return_0;
}
free(buffer);
}
return 1;
}
static int _print_vg(struct formatter *f, struct volume_group *vg)
{
2004-03-08 20:19:15 +03:00
char buffer[4096];
const struct format_type *fmt = NULL;
uint64_t status = vg->status;
2008-01-30 16:19:47 +03:00
if (!id_write_format(&vg->id, buffer, sizeof(buffer)))
return_0;
2004-05-05 01:25:57 +04:00
outf(f, "id = \"%s\"", buffer);
2004-05-05 01:25:57 +04:00
outf(f, "seqno = %u", vg->seqno);
2004-03-08 20:19:15 +03:00
if (vg->original_fmt)
fmt = vg->original_fmt;
else if (vg->fid)
fmt = vg->fid->fmt;
if (fmt)
outfc(f, "# informational", "format = \"%s\"", fmt->name);
/*
2015-03-09 21:53:22 +03:00
* Removing WRITE and adding LVM_WRITE_LOCKED makes it read-only
* to old versions of lvm that only look for LVM_WRITE.
*/
if ((status & LVM_WRITE) && vg_flag_write_locked(vg)) {
status &= ~LVM_WRITE;
status |= LVM_WRITE_LOCKED;
}
if (!_print_flag_config(f, status, VG_FLAGS))
2008-01-30 16:19:47 +03:00
return_0;
2004-03-08 20:19:15 +03:00
if (!_out_list(f, &vg->tags, "tags"))
return_0;
if (vg->system_id && *vg->system_id)
2004-05-05 01:25:57 +04:00
outf(f, "system_id = \"%s\"", vg->system_id);
2004-03-08 20:19:15 +03:00
2015-03-05 23:00:44 +03:00
if (vg->lock_type) {
outf(f, "lock_type = \"%s\"", vg->lock_type);
2015-03-05 23:00:44 +03:00
if (vg->lock_args)
outf(f, "lock_args = \"%s\"", vg->lock_args);
}
outsize(f, (uint64_t) vg->extent_size, "extent_size = %u",
vg->extent_size);
2004-05-05 01:25:57 +04:00
outf(f, "max_lv = %u", vg->max_lv);
outf(f, "max_pv = %u", vg->max_pv);
/* Default policy is NORMAL; INHERIT is meaningless */
if (vg->alloc != ALLOC_NORMAL && vg->alloc != ALLOC_INHERIT) {
outnl(f);
outf(f, "allocation_policy = \"%s\"",
get_alloc_string(vg->alloc));
}
if (vg->profile)
outf(f, "profile = \"%s\"", vg->profile->name);
outf(f, "metadata_copies = %u", vg->mda_copies);
return 1;
}
/*
* Get the pv%d name from the formatters hash
* table.
*/
static const char *_get_pv_name_from_uuid(struct formatter *f, char *uuid)
{
const char *pv_name = dm_hash_lookup(f->pv_names, uuid);
if (!pv_name)
log_error(INTERNAL_ERROR "PV name for uuid %s missing from text metadata export hash table.",
uuid);
return pv_name;
}
static const char *_get_pv_name(struct formatter *f, struct physical_volume *pv)
{
char uuid[64] __attribute__((aligned(8)));
if (!pv || !id_write_format(&pv->id, uuid, sizeof(uuid)))
return_NULL;
return _get_pv_name_from_uuid(f, uuid);
}
static int _print_pvs(struct formatter *f, struct volume_group *vg)
{
2005-06-01 20:51:55 +04:00
struct pv_list *pvl;
struct physical_volume *pv;
char buffer[PATH_MAX * 2];
const char *name;
device usage based on devices file The LVM devices file lists devices that lvm can use. The default file is /etc/lvm/devices/system.devices, and the lvmdevices(8) command is used to add or remove device entries. If the file does not exist, or if lvm.conf includes use_devicesfile=0, then lvm will not use a devices file. When the devices file is in use, the regex filter is not used, and the filter settings in lvm.conf or on the command line are ignored. LVM records devices in the devices file using hardware-specific IDs, such as the WWID, and attempts to use subsystem-specific IDs for virtual device types. These device IDs are also written in the VG metadata. When no hardware or virtual ID is available, lvm falls back using the unstable device name as the device ID. When devnames are used, lvm performs extra scanning to find devices if their devname changes, e.g. after reboot. When proper device IDs are used, an lvm command will not look at devices outside the devices file, but when devnames are used as a fallback, lvm will scan devices outside the devices file to locate PVs on renamed devices. A config setting search_for_devnames can be used to control the scanning for renamed devname entries. Related to the devices file, the new command option --devices <devnames> allows a list of devices to be specified for the command to use, overriding the devices file. The listed devices act as a sort of devices file in terms of limiting which devices lvm will see and use. Devices that are not listed will appear to be missing to the lvm command. Multiple devices files can be kept in /etc/lvm/devices, which allows lvm to be used with different sets of devices, e.g. system devices do not need to be exposed to a specific application, and the application can use lvm on its own set of devices that are not exposed to the system. The option --devicesfile <filename> is used to select the devices file to use with the command. Without the option set, the default system devices file is used. Setting --devicesfile "" causes lvm to not use a devices file. An existing, empty devices file means lvm will see no devices. The new command vgimportdevices adds PVs from a VG to the devices file and updates the VG metadata to include the device IDs. vgimportdevices -a will import all VGs into the system devices file. LVM commands run by dmeventd not use a devices file by default, and will look at all devices on the system. A devices file can be created for dmeventd (/etc/lvm/devices/dmeventd.devices) If this file exists, lvm commands run by dmeventd will use it. Internal implementaion: - device_ids_read - read the devices file . add struct dev_use (du) to cmd->use_devices for each devices file entry - dev_cache_scan - get /dev entries . add struct device (dev) to dev_cache for each device on the system - device_ids_match - match devices file entries to /dev entries . match each du on cmd->use_devices to a dev in dev_cache, using device ID . on match, set du->dev, dev->id, dev->flags MATCHED_USE_ID - label_scan - read lvm headers and metadata from devices . filters are applied, those that do not need data from the device . filter-deviceid skips devs without MATCHED_USE_ID, i.e. skips /dev entries that are not listed in the devices file . read lvm label from dev . filters are applied, those that use data from the device . read lvm metadata from dev . add info/vginfo structs for PVs/VGs (info is "lvmcache") - device_ids_find_renamed_devs - handle devices with unstable devname ID where devname changed . this step only needed when devs do not have proper device IDs, and their dev names change, e.g. after reboot sdb becomes sdc. . detect incorrect match because PVID in the devices file entry does not match the PVID found when the device was read above . undo incorrect match between du and dev above . search system devices for new location of PVID . update devices file with new devnames for PVIDs on renamed devices . label_scan the renamed devs - continue with command processing
2020-06-23 21:25:41 +03:00
const char *idtype, *idname;
2004-05-05 01:25:57 +04:00
outf(f, "physical_volumes {");
_inc_indent(f);
dm_list_iterate_items(pvl, &vg->pvs) {
2005-06-01 20:51:55 +04:00
pv = pvl->pv;
if (!id_write_format(&pv->id, buffer, sizeof(buffer)))
return_0;
if (!(name = _get_pv_name_from_uuid(f, buffer)))
2008-01-30 16:19:47 +03:00
return_0;
outnl(f);
2004-05-05 01:25:57 +04:00
outf(f, "%s {", name);
_inc_indent(f);
2004-05-05 01:25:57 +04:00
outf(f, "id = \"%s\"", buffer);
if (strlen(pv_dev_name(pv)) >= PATH_MAX) {
log_error("pv device name size is out of bounds.");
return 0;
}
outhint(f, "device = \"%s\"",
dm_escape_double_quotes(buffer, pv_dev_name(pv)));
outnl(f);
device usage based on devices file The LVM devices file lists devices that lvm can use. The default file is /etc/lvm/devices/system.devices, and the lvmdevices(8) command is used to add or remove device entries. If the file does not exist, or if lvm.conf includes use_devicesfile=0, then lvm will not use a devices file. When the devices file is in use, the regex filter is not used, and the filter settings in lvm.conf or on the command line are ignored. LVM records devices in the devices file using hardware-specific IDs, such as the WWID, and attempts to use subsystem-specific IDs for virtual device types. These device IDs are also written in the VG metadata. When no hardware or virtual ID is available, lvm falls back using the unstable device name as the device ID. When devnames are used, lvm performs extra scanning to find devices if their devname changes, e.g. after reboot. When proper device IDs are used, an lvm command will not look at devices outside the devices file, but when devnames are used as a fallback, lvm will scan devices outside the devices file to locate PVs on renamed devices. A config setting search_for_devnames can be used to control the scanning for renamed devname entries. Related to the devices file, the new command option --devices <devnames> allows a list of devices to be specified for the command to use, overriding the devices file. The listed devices act as a sort of devices file in terms of limiting which devices lvm will see and use. Devices that are not listed will appear to be missing to the lvm command. Multiple devices files can be kept in /etc/lvm/devices, which allows lvm to be used with different sets of devices, e.g. system devices do not need to be exposed to a specific application, and the application can use lvm on its own set of devices that are not exposed to the system. The option --devicesfile <filename> is used to select the devices file to use with the command. Without the option set, the default system devices file is used. Setting --devicesfile "" causes lvm to not use a devices file. An existing, empty devices file means lvm will see no devices. The new command vgimportdevices adds PVs from a VG to the devices file and updates the VG metadata to include the device IDs. vgimportdevices -a will import all VGs into the system devices file. LVM commands run by dmeventd not use a devices file by default, and will look at all devices on the system. A devices file can be created for dmeventd (/etc/lvm/devices/dmeventd.devices) If this file exists, lvm commands run by dmeventd will use it. Internal implementaion: - device_ids_read - read the devices file . add struct dev_use (du) to cmd->use_devices for each devices file entry - dev_cache_scan - get /dev entries . add struct device (dev) to dev_cache for each device on the system - device_ids_match - match devices file entries to /dev entries . match each du on cmd->use_devices to a dev in dev_cache, using device ID . on match, set du->dev, dev->id, dev->flags MATCHED_USE_ID - label_scan - read lvm headers and metadata from devices . filters are applied, those that do not need data from the device . filter-deviceid skips devs without MATCHED_USE_ID, i.e. skips /dev entries that are not listed in the devices file . read lvm label from dev . filters are applied, those that use data from the device . read lvm metadata from dev . add info/vginfo structs for PVs/VGs (info is "lvmcache") - device_ids_find_renamed_devs - handle devices with unstable devname ID where devname changed . this step only needed when devs do not have proper device IDs, and their dev names change, e.g. after reboot sdb becomes sdc. . detect incorrect match because PVID in the devices file entry does not match the PVID found when the device was read above . undo incorrect match between du and dev above . search system devices for new location of PVID . update devices file with new devnames for PVIDs on renamed devices . label_scan the renamed devs - continue with command processing
2020-06-23 21:25:41 +03:00
idtype = dev_idtype_for_metadata(vg->cmd, pv->dev);
idname = dev_idname_for_metadata(vg->cmd, pv->dev);
if (idtype && idname) {
outf(f, "device_id_type = \"%s\"", idtype);
outf(f, "device_id = \"%s\"", idname);
}
if (!_print_flag_config(f, pv->status, PV_FLAGS))
2008-01-30 16:19:47 +03:00
return_0;
2004-03-08 20:19:15 +03:00
if (!_out_list(f, &pv->tags, "tags"))
return_0;
2004-03-08 20:19:15 +03:00
2017-12-11 18:32:53 +03:00
outsize(f, pv->size, "dev_size = " FMTu64, pv->size);
2017-12-11 18:32:53 +03:00
outf(f, "pe_start = " FMTu64, pv->pe_start);
outsize(f, vg->extent_size * (uint64_t) pv->pe_count,
"pe_count = %u", pv->pe_count);
if (pv->ba_start && pv->ba_size) {
2017-12-11 18:32:53 +03:00
outf(f, "ba_start = " FMTu64, pv->ba_start);
outsize(f, pv->ba_size, "ba_size = " FMTu64, pv->ba_size);
}
_dec_indent(f);
2004-05-05 01:25:57 +04:00
outf(f, "}");
}
_dec_indent(f);
2004-05-05 01:25:57 +04:00
outf(f, "}");
return 1;
}
static int _print_segment(struct formatter *f, struct volume_group *vg,
2002-11-18 17:04:08 +03:00
int count, struct lv_segment *seg)
{
char buffer[2048];
if (!print_segtype_lvflags(buffer, sizeof(buffer), seg->lv->status))
return_0;
2004-05-05 01:25:57 +04:00
outf(f, "segment%u {", count);
_inc_indent(f);
2004-05-05 01:25:57 +04:00
outf(f, "start_extent = %u", seg->le);
outsize(f, (uint64_t) seg->len * vg->extent_size,
"extent_count = %u", seg->len);
outnl(f);
lvconvert: add infrastructure for RaidLV reshaping support In order to support striped raid5/6/10 LV reshaping (change of LV type, stripesize or number of legs), this patch introduces infrastructure prerequisites to be used by raid_manip.c extensions in followup patches. This base is needed for allocation of out-of-place reshape space required by the MD raid personalities to avoid writing over data in-place when reading off the current RAID layout or number of legs and writing out the new layout or to a different number of legs (i.e. restripe) Changes: - add members reshape_len to 'struct lv_segment' to store out-of-place reshape length per component rimage - add member data_copies to struct lv_segment to support more than 2 raid10 data copies - make alloc_lv_segment() aware of both reshape_len and data_copies - adjust all alloc_lv_segment() callers to the new API - add functions to retrieve the current data offset (needed for out-of-place reshaping space allocation) and the devices count from the kernel - make libdm deptree code aware of reshape_len - add LV flags for disk add/remove reshaping - support import/export of the new 'struct lv_segment' members - enhance lv_extend/_lv_reduce to cope with reshape_len - add seg_is_*/segtype_is_* macros related to reshaping - add target version check for reshaping - grow rebuilds/writemostly bitmaps to 246 bit to support kernel maximal - enhance libdm deptree code to support data_offset (out-of-place reshaping) and delta_disk (legs add/remove reshaping) target arguments Related: rhbz834579 Related: rhbz1191935 Related: rhbz1191978
2017-02-24 02:50:00 +03:00
if (seg->reshape_len)
outsize(f, (uint64_t) seg->reshape_len * vg->extent_size,
"reshape_count = %u", seg->reshape_len);
outf(f, "type = \"%s%s\"", seg->segtype->name, buffer);
if (!_out_list(f, &seg->tags, "tags"))
return_0;
2004-03-08 20:19:15 +03:00
if (seg->segtype->ops->text_export &&
2008-01-30 16:19:47 +03:00
!seg->segtype->ops->text_export(seg, f))
return_0;
2004-05-05 01:25:57 +04:00
_dec_indent(f);
outf(f, "}");
2004-05-05 01:25:57 +04:00
return 1;
}
2003-05-06 16:02:36 +04:00
2004-05-05 01:25:57 +04:00
int out_areas(struct formatter *f, const struct lv_segment *seg,
const char *type)
{
const char *name;
unsigned int s;
struct physical_volume *pv;
2002-11-18 17:04:08 +03:00
outnl(f);
2004-05-05 01:25:57 +04:00
outf(f, "%ss = [", type);
_inc_indent(f);
2002-11-18 17:04:08 +03:00
2004-05-05 01:25:57 +04:00
for (s = 0; s < seg->area_count; s++) {
2005-06-01 20:51:55 +04:00
switch (seg_type(seg, s)) {
2004-05-05 01:25:57 +04:00
case AREA_PV:
if (!(pv = seg_pv(seg, s))) {
2017-12-11 18:32:53 +03:00
log_error(INTERNAL_ERROR "Missing PV for area " FMTu32 " of %s segment of LV %s.",
s, type, display_lvname(seg->lv));
return 0;
}
if (!(name = _get_pv_name(f, pv)))
2008-01-30 16:19:47 +03:00
return_0;
2004-05-05 01:25:57 +04:00
outf(f, "\"%s\", %u%s", name,
2005-06-01 20:51:55 +04:00
seg_pe(seg, s),
2004-05-05 01:25:57 +04:00
(s == seg->area_count - 1) ? "" : ",");
break;
case AREA_LV:
/* FIXME This helper code should be target-independent! Check for metadata LV property. */
if (!seg_is_raid(seg)) {
outf(f, "\"%s\", %u%s",
seg_lv(seg, s)->name,
seg_le(seg, s),
(s == seg->area_count - 1) ? "" : ",");
continue;
}
/* RAID devices are laid-out in metadata/data pairs */
/* FIXME Validation should be elsewhere, not here! */
if (!lv_is_raid_image(seg_lv(seg, s)) ||
(seg->meta_areas && seg_metalv(seg, s) && !lv_is_raid_metadata(seg_metalv(seg, s)))) {
log_error("RAID segment has non-RAID areas");
return 0;
}
if (seg->meta_areas && seg_metalv(seg,s))
outf(f, "\"%s\", \"%s\"%s",
(seg->meta_areas && seg_metalv(seg, s)) ? seg_metalv(seg, s)->name : "",
seg_lv(seg, s)->name, (s == seg->area_count - 1) ? "" : ",");
else
outf(f, "\"%s\"%s", seg_lv(seg, s)->name, (s == seg->area_count - 1) ? "" : ",");
break;
case AREA_UNASSIGNED:
2017-12-11 18:32:53 +03:00
log_error(INTERNAL_ERROR "Invalid type for area " FMTu32 " of %s segment of LV %s.",
s, type, display_lvname(seg->lv));
return 0;
2004-05-05 01:25:57 +04:00
}
}
_dec_indent(f);
2004-05-05 01:25:57 +04:00
outf(f, "]");
return 1;
}
static int _print_timestamp(struct formatter *f,
const char *name, time_t ts,
char *buf, size_t buf_size)
{
struct tm *local_tm;
if (ts) {
strncpy(buf, "# ", buf_size);
if (!(local_tm = localtime(&ts)) ||
!strftime(buf + 2, buf_size - 2,
"%Y-%m-%d %T %z", local_tm))
buf[0] = 0;
2017-12-11 18:32:53 +03:00
outfc(f, buf, "%s = " FMTu64, name, (uint64_t) ts);
}
return 1;
}
2005-06-01 20:51:55 +04:00
static int _print_lv(struct formatter *f, struct logical_volume *lv)
2001-12-20 14:52:54 +03:00
{
2005-06-01 20:51:55 +04:00
struct lv_segment *seg;
char buffer[4096];
int seg_count;
uint64_t status = lv->status;
2001-12-20 14:52:54 +03:00
outnl(f);
2005-06-01 20:51:55 +04:00
outf(f, "%s {", lv->name);
_inc_indent(f);
2001-12-20 14:52:54 +03:00
2005-06-01 20:51:55 +04:00
/* FIXME: Write full lvid */
2008-01-30 16:19:47 +03:00
if (!id_write_format(&lv->lvid.id[1], buffer, sizeof(buffer)))
return_0;
2005-06-01 20:51:55 +04:00
outf(f, "id = \"%s\"", buffer);
/*
2015-03-09 21:53:22 +03:00
* Removing WRITE and adding LVM_WRITE_LOCKED makes it read-only
* to old versions of lvm that only look for LVM_WRITE.
*/
if ((status & LVM_WRITE) && vg_flag_write_locked(lv->vg)) {
status &= ~LVM_WRITE;
status |= LVM_WRITE_LOCKED;
}
if (!_print_flag_config(f, status, LV_FLAGS))
2008-01-30 16:19:47 +03:00
return_0;
2005-06-01 20:51:55 +04:00
if (!_out_list(f, &lv->tags, "tags"))
return_0;
2005-06-01 20:51:55 +04:00
if (lv->timestamp) {
if (!_print_timestamp(f, "creation_time", lv->timestamp,
buffer, sizeof(buffer)))
return_0;
outf(f, "creation_host = \"%s\"", lv->hostname);
}
2015-03-05 23:00:44 +03:00
if (lv->lock_args)
outf(f, "lock_args = \"%s\"", lv->lock_args);
2005-06-01 20:51:55 +04:00
if (lv->alloc != ALLOC_INHERIT)
outf(f, "allocation_policy = \"%s\"",
get_alloc_string(lv->alloc));
if (lv->profile)
outf(f, "profile = \"%s\"", lv->profile->name);
switch (lv->read_ahead) {
case DM_READ_AHEAD_NONE:
outfc(f, "# None", "read_ahead = -1");
break;
case DM_READ_AHEAD_AUTO:
/* No output - use default */
break;
default:
2005-06-01 20:51:55 +04:00
outf(f, "read_ahead = %u", lv->read_ahead);
}
2005-06-01 20:51:55 +04:00
if (lv->major >= 0)
outf(f, "major = %d", lv->major);
if (lv->minor >= 0)
outf(f, "minor = %d", lv->minor);
outf(f, "segment_count = %u", dm_list_size(&lv->segments));
outnl(f);
2005-06-01 20:51:55 +04:00
seg_count = 1;
dm_list_iterate_items(seg, &lv->segments) {
2008-01-30 16:19:47 +03:00
if (!_print_segment(f, lv->vg, seg_count++, seg))
return_0;
2005-06-01 20:51:55 +04:00
}
_dec_indent(f);
outf(f, "}");
return 1;
2001-12-20 14:52:54 +03:00
}
static int _print_lvs(struct formatter *f, struct volume_group *vg)
{
2005-06-01 20:51:55 +04:00
struct lv_list *lvl;
/*
* Don't bother with an lv section if there are no lvs.
*/
if (dm_list_empty(&vg->lvs))
return 1;
2004-05-05 01:25:57 +04:00
outf(f, "logical_volumes {");
_inc_indent(f);
2005-06-01 20:51:55 +04:00
/*
* Write visible LVs first
*/
dm_list_iterate_items(lvl, &vg->lvs) {
if (!(lv_is_visible(lvl->lv)))
2005-06-01 20:51:55 +04:00
continue;
2008-01-30 16:19:47 +03:00
if (!_print_lv(f, lvl->lv))
return_0;
2005-06-01 20:51:55 +04:00
}
dm_list_iterate_items(lvl, &vg->lvs) {
if ((lv_is_visible(lvl->lv)))
2005-06-01 20:51:55 +04:00
continue;
2008-01-30 16:19:47 +03:00
if (!_print_lv(f, lvl->lv))
return_0;
}
_dec_indent(f);
2004-05-05 01:25:57 +04:00
outf(f, "}");
return 1;
}
static int _alloc_printed_indirect_descendants(struct dm_list *indirect_glvs, char **buffer)
{
struct glv_list *user_glvl;
size_t buf_size = 0;
int first = 1;
char *buf;
*buffer = NULL;
dm_list_iterate_items(user_glvl, indirect_glvs) {
if (user_glvl->glv->is_historical)
continue;
/* '"' + name + '"' + ',' + ' ' */
buf_size += strlen(user_glvl->glv->live->name) + 4;
}
if (!buf_size)
return 1;
/* '[' + ']' + '\0' */
buf_size += 3;
if (!(*buffer = malloc(buf_size))) {
log_error("Could not allocate memory for ancestor list buffer.");
return 0;
}
buf = *buffer;
if (!emit_to_buffer(&buf, &buf_size, "["))
goto_bad;
dm_list_iterate_items(user_glvl, indirect_glvs) {
if (user_glvl->glv->is_historical)
continue;
if (!first) {
if (!emit_to_buffer(&buf, &buf_size, ", "))
goto_bad;
} else
first = 0;
if (!emit_to_buffer(&buf, &buf_size, "\"%s\"", user_glvl->glv->live->name))
goto_bad;
}
if (!emit_to_buffer(&buf, &buf_size, "]"))
goto_bad;
return 1;
bad:
if (*buffer) {
free(*buffer);
*buffer = NULL;
}
return 0;
}
static int _print_historical_lv(struct formatter *f, struct historical_logical_volume *hlv)
{
char buffer[40];
char *descendants_buffer = NULL;
int r = 0;
if (!id_write_format(&hlv->lvid.id[1], buffer, sizeof(buffer)))
goto_out;
if (!_alloc_printed_indirect_descendants(&hlv->indirect_glvs, &descendants_buffer))
goto_out;
outnlgo(f);
outfgo(f, "%s {", hlv->name);
_inc_indent(f);
outfgo(f, "id = \"%s\"", buffer);
if (!_print_timestamp(f, "creation_time", hlv->timestamp, buffer, sizeof(buffer)))
goto_out;
if (!_print_timestamp(f, "removal_time", hlv->timestamp_removed, buffer, sizeof(buffer)))
goto_out;
if (hlv->indirect_origin) {
if (hlv->indirect_origin->is_historical)
outfgo(f, "origin = \"%s%s\"", HISTORICAL_LV_PREFIX, hlv->indirect_origin->historical->name);
else
outfgo(f, "origin = \"%s\"", hlv->indirect_origin->live->name);
}
if (descendants_buffer)
outfgo(f, "descendants = %s", descendants_buffer);
_dec_indent(f);
outfgo(f, "}");
r = 1;
out:
free(descendants_buffer);
return r;
}
static int _print_historical_lvs(struct formatter *f, struct volume_group *vg)
{
struct glv_list *glvl;
if (dm_list_empty(&vg->historical_lvs))
return 1;
outf(f, "historical_logical_volumes {");
_inc_indent(f);
dm_list_iterate_items(glvl, &vg->historical_lvs) {
if (!_print_historical_lv(f, glvl->glv->historical))
return_0;
}
_dec_indent(f);
outf(f, "}");
return 1;
}
/*
* In the text format we refer to pv's as 'pv1',
* 'pv2' etc. This function builds a hash table
* to enable a quick lookup from device -> name.
*/
static int _build_pv_names(struct formatter *f, struct volume_group *vg)
{
int count = 0;
2005-06-01 20:51:55 +04:00
struct pv_list *pvl;
struct physical_volume *pv;
char buffer[32], *name;
char uuid[64];
if (!(f->mem = dm_pool_create("text pv_names", 512)))
return_0;
if (!(f->pv_names = dm_hash_create(115)))
return_0;
dm_list_iterate_items(pvl, &vg->pvs) {
2005-06-01 20:51:55 +04:00
pv = pvl->pv;
/* FIXME But skip if there's already an LV called pv%d ! */
2006-08-21 16:54:53 +04:00
if (dm_snprintf(buffer, sizeof(buffer), "pv%d", count++) < 0)
return_0;
if (!(name = dm_pool_strdup(f->mem, buffer)))
return_0;
if (!id_write_format(&pv->id, uuid, sizeof(uuid)))
return_0;
if (!dm_hash_insert(f->pv_names, uuid, name))
return_0;
}
return 1;
}
2002-11-18 17:04:08 +03:00
static int _text_vg_export(struct formatter *f,
struct volume_group *vg, const char *desc)
{
int r = 0;
2008-01-30 16:19:47 +03:00
if (!_build_pv_names(f, vg))
goto_out;
if (f->header && !_print_header(vg->cmd, f, desc))
goto_out;
2002-11-18 17:04:08 +03:00
2004-05-05 01:25:57 +04:00
if (!out_text(f, "%s {", vg->name))
goto_out;
2001-12-20 14:52:54 +03:00
_inc_indent(f);
if (!_print_vg(f, vg))
goto_out;
outnl(f);
if (!_print_pvs(f, vg))
goto_out;
outnl(f);
if (!_print_lvs(f, vg))
goto_out;
outnl(f);
if (!_print_historical_lvs(f, vg))
goto_out;
2002-11-18 17:04:08 +03:00
_dec_indent(f);
2004-05-05 01:25:57 +04:00
if (!out_text(f, "}"))
goto_out;
if (!f->header && !_print_header(vg->cmd, f, desc))
goto_out;
2002-11-18 17:04:08 +03:00
r = 1;
out:
if (f->mem) {
dm_pool_destroy(f->mem);
f->mem = NULL;
}
if (f->pv_names) {
dm_hash_destroy(f->pv_names);
f->pv_names = NULL;
}
2002-11-18 17:04:08 +03:00
return r;
}
int text_vg_export_file(struct volume_group *vg, const char *desc, FILE *fp)
{
int r;
struct formatter f = {
.indent = 0,
.header = 1,
.out_with_comment = &_out_with_comment_file,
.nl = &_nl_file,
.data.fp = fp,
};
2002-11-18 17:04:08 +03:00
_init();
if ((r = _text_vg_export(&f, vg, desc)))
r = !ferror(f.data.fp);
2002-11-18 17:04:08 +03:00
return r;
}
2002-11-18 17:04:08 +03:00
/* Returns amount of buffer used incl. terminating NUL */
size_t text_vg_export_raw(struct volume_group *vg, const char *desc, char **buf, uint32_t *buf_size)
2002-11-18 17:04:08 +03:00
{
size_t r;
struct formatter f = {
.indent = 0,
.header = 0,
.out_with_comment = &_out_with_comment_raw,
.nl = &_nl_raw,
.data.buf.size = 65536, /* Initial metadata limit */
};
2002-11-18 17:04:08 +03:00
_init();
if (!(f.data.buf.start = zalloc(f.data.buf.size))) {
log_error("text_export buffer allocation failed");
return 0;
}
if (!_text_vg_export(&f, vg, desc)) {
free(f.data.buf.start);
return 0;
2002-11-18 17:04:08 +03:00
}
r = f.data.buf.used + 1;
*buf = f.data.buf.start;
2002-11-18 17:04:08 +03:00
if (buf_size)
*buf_size = f.data.buf.size;
2002-11-18 17:04:08 +03:00
return r;
}
static size_t _export_vg_to_buffer(struct volume_group *vg, char **buf)
{
return text_vg_export_raw(vg, "", buf, NULL);
}
struct dm_config_tree *export_vg_to_config_tree(struct volume_group *vg)
{
char *buf = NULL;
struct dm_config_tree *vg_cft;
if (!_export_vg_to_buffer(vg, &buf)) {
log_error("Could not format metadata for VG %s.", vg->name);
2018-12-23 01:44:42 +03:00
return NULL;
}
if (!(vg_cft = config_tree_from_string_without_dup_node_check(buf))) {
log_error("Error parsing metadata for VG %s.", vg->name);
free(buf);
2018-12-23 01:44:42 +03:00
return NULL;
}
free(buf);
return vg_cft;
}
2004-05-05 01:25:57 +04:00
#undef outf
#undef outnl