Adding upstream version 2.39.3.upstream/2.39.3

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

1 files changed, 2486 insertions, 0 deletions

diff --git a/misc-utils/lsblk.c b/misc-utils/lsblk.c
new file mode 100644
index 0000000..fae918b
--- /dev/null
+++ b/misc-utils/lsblk.c
@@ -0,0 +1,2486 @@
+/*
+ * lsblk(8) - list block devices
+ *
+ * Copyright (C) 2010-2018 Red Hat, Inc. All rights reserved.
+ * Written by Milan Broz <gmazyland@gmail.com>
+ *            Karel Zak <kzak@redhat.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it would be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
+ */
+#include <stdio.h>
+#include <errno.h>
+#include <getopt.h>
+#include <stdlib.h>
+#include <unistd.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <dirent.h>
+#include <fcntl.h>
+#include <string.h>
+#include <sys/ioctl.h>
+#include <stdarg.h>
+#include <locale.h>
+#include <pwd.h>
+#include <grp.h>
+#include <ctype.h>
+#include <assert.h>
+
+#include <blkid.h>
+
+#include "c.h"
+#include "pathnames.h"
+#include "blkdev.h"
+#include "canonicalize.h"
+#include "nls.h"
+#include "xalloc.h"
+#include "strutils.h"
+#include "sysfs.h"
+#include "closestream.h"
+#include "optutils.h"
+#include "fileutils.h"
+#include "loopdev.h"
+#include "buffer.h"
+
+#include "lsblk.h"
+
+UL_DEBUG_DEFINE_MASK(lsblk);
+UL_DEBUG_DEFINE_MASKNAMES(lsblk) = UL_DEBUG_EMPTY_MASKNAMES;
+
+#define LSBLK_EXIT_SOMEOK 64
+#define LSBLK_EXIT_ALLFAILED 32
+
+static int column_id_to_number(int id);
+
+/* column IDs */
+enum {
+	COL_ALIOFF = 0,
+	COL_IDLINK,
+	COL_ID,
+	COL_DALIGN,
+	COL_DAX,
+	COL_DGRAN,
+	COL_DISKSEQ,
+	COL_DMAX,
+	COL_DZERO,
+	COL_FSAVAIL,
+	COL_FSROOTS,
+	COL_FSSIZE,
+	COL_FSTYPE,
+	COL_FSUSED,
+	COL_FSUSEPERC,
+	COL_FSVERSION,
+	COL_GROUP,
+	COL_HCTL,
+	COL_HOTPLUG,
+	COL_KNAME,
+	COL_LABEL,
+	COL_LOGSEC,
+	COL_MAJMIN,
+	COL_MINIO,
+	COL_MODE,
+	COL_MODEL,
+	COL_MQ,
+	COL_NAME,
+	COL_OPTIO,
+	COL_OWNER,
+	COL_PARTFLAGS,
+	COL_PARTLABEL,
+	COL_PARTN,
+	COL_PARTTYPE,
+	COL_PARTTYPENAME,
+	COL_PARTUUID,
+	COL_PATH,
+	COL_PHYSEC,
+	COL_PKNAME,
+	COL_PTTYPE,
+	COL_PTUUID,
+	COL_RA,
+	COL_RAND,
+	COL_REV,
+	COL_RM,
+	COL_RO,
+	COL_ROTA,
+	COL_RQ_SIZE,
+	COL_SCHED,
+	COL_SERIAL,
+	COL_SIZE,
+	COL_START,
+	COL_STATE,
+	COL_SUBSYS,
+	COL_TARGET,
+	COL_TARGETS,
+	COL_TRANSPORT,
+	COL_TYPE,
+	COL_UUID,
+	COL_VENDOR,
+	COL_WSAME,
+	COL_WWN,
+	COL_ZONED,
+	COL_ZONE_SZ,
+	COL_ZONE_WGRAN,
+	COL_ZONE_APP,
+	COL_ZONE_NR,
+	COL_ZONE_OMAX,
+	COL_ZONE_AMAX,
+};
+
+/* basic table settings */
+enum {
+	LSBLK_ASCII =		(1 << 0),
+	LSBLK_RAW =		(1 << 1),
+	LSBLK_NOHEADINGS =	(1 << 2),
+	LSBLK_EXPORT =		(1 << 3),
+	LSBLK_TREE =		(1 << 4),
+	LSBLK_JSON =		(1 << 5),
+	LSBLK_SHELLVAR =	(1 << 6)
+};
+
+/* Types used for qsort() and JSON */
+enum {
+	COLTYPE_STR	= 0,	/* default */
+	COLTYPE_NUM	= 1,	/* always u64 number */
+	COLTYPE_SORTNUM = 2,	/* string on output, u64 for qsort() */
+	COLTYPE_SIZE	= 3,	/* srring by default, number when --bytes */
+	COLTYPE_BOOL	= 4	/* 0 or 1 */
+};
+
+/* column names */
+struct colinfo {
+	const char	*name;		/* header */
+	double		whint;		/* width hint (N < 1 is in percent of termwidth) */
+	int		flags;		/* SCOLS_FL_* */
+	const char	*help;
+	int		type;		/* COLTYPE_* */
+};
+
+/* columns descriptions */
+static struct colinfo infos[] = {
+	[COL_ALIOFF] = { "ALIGNMENT", 6, SCOLS_FL_RIGHT, N_("alignment offset"), COLTYPE_NUM },
+	[COL_ID] = { "ID", 0.1, SCOLS_FL_NOEXTREMES, N_("udev ID (based on ID-LINK)") },
+	[COL_IDLINK] = { "ID-LINK", 0.1, SCOLS_FL_NOEXTREMES, N_("the shortest udev /dev/disk/by-id link name") },
+	[COL_DALIGN] = { "DISC-ALN", 6, SCOLS_FL_RIGHT, N_("discard alignment offset"), COLTYPE_NUM },
+	[COL_DAX] = { "DAX", 1, SCOLS_FL_RIGHT, N_("dax-capable device"), COLTYPE_BOOL },
+	[COL_DGRAN] = { "DISC-GRAN", 6, SCOLS_FL_RIGHT, N_("discard granularity"), COLTYPE_SIZE },
+	[COL_DISKSEQ] = { "DISK-SEQ", 1, SCOLS_FL_RIGHT, N_("disk sequence number"), COLTYPE_NUM },
+	[COL_DMAX] = { "DISC-MAX", 6, SCOLS_FL_RIGHT, N_("discard max bytes"), COLTYPE_SIZE },
+	[COL_DZERO] = { "DISC-ZERO", 1, SCOLS_FL_RIGHT, N_("discard zeroes data"), COLTYPE_BOOL },
+	[COL_FSAVAIL] = { "FSAVAIL", 5, SCOLS_FL_RIGHT, N_("filesystem size available"), COLTYPE_SIZE },
+	[COL_FSROOTS] = { "FSROOTS", 0.1, SCOLS_FL_WRAP, N_("mounted filesystem roots") },
+	[COL_FSSIZE] = { "FSSIZE", 5, SCOLS_FL_RIGHT, N_("filesystem size"), COLTYPE_SIZE },
+	[COL_FSTYPE] = { "FSTYPE", 0.1, SCOLS_FL_TRUNC, N_("filesystem type") },
+	[COL_FSUSED] = { "FSUSED", 5, SCOLS_FL_RIGHT, N_("filesystem size used"), COLTYPE_SIZE },
+	[COL_FSUSEPERC] = { "FSUSE%", 3, SCOLS_FL_RIGHT, N_("filesystem use percentage") },
+	[COL_FSVERSION] = { "FSVER", 0.1, SCOLS_FL_TRUNC, N_("filesystem version") },
+	[COL_GROUP] = { "GROUP", 0.1, SCOLS_FL_TRUNC, N_("group name") },
+	[COL_HCTL] = { "HCTL", 10, 0, N_("Host:Channel:Target:Lun for SCSI") },
+	[COL_HOTPLUG] = { "HOTPLUG", 1, SCOLS_FL_RIGHT, N_("removable or hotplug device (usb, pcmcia, ...)"), COLTYPE_BOOL },
+	[COL_KNAME] = { "KNAME", 0.3, 0, N_("internal kernel device name") },
+	[COL_LABEL] = { "LABEL", 0.1, 0, N_("filesystem LABEL") },
+	[COL_LOGSEC] = { "LOG-SEC", 7, SCOLS_FL_RIGHT, N_("logical sector size"), COLTYPE_NUM },
+	[COL_MAJMIN] = { "MAJ:MIN", 6, 0, N_("major:minor device number"), COLTYPE_SORTNUM },
+	[COL_MINIO] = { "MIN-IO", 6, SCOLS_FL_RIGHT, N_("minimum I/O size"), COLTYPE_NUM },
+	[COL_MODEL] = { "MODEL", 0.1, SCOLS_FL_TRUNC, N_("device identifier") },
+	[COL_MODE] = { "MODE", 10, 0, N_("device node permissions") },
+	[COL_MQ] = { "MQ", 3, SCOLS_FL_RIGHT, N_("device queues") },
+	[COL_NAME] = { "NAME", 0.25, SCOLS_FL_NOEXTREMES, N_("device name") },
+	[COL_OPTIO] = { "OPT-IO", 6, SCOLS_FL_RIGHT, N_("optimal I/O size"), COLTYPE_NUM },
+	[COL_OWNER] = { "OWNER", 0.1, SCOLS_FL_TRUNC, N_("user name"), },
+	[COL_PARTFLAGS] = { "PARTFLAGS", 36,  0, N_("partition flags") },
+	[COL_PARTLABEL] = { "PARTLABEL", 0.1, 0, N_("partition LABEL") },
+	[COL_PARTN] = { "PARTN", 2, SCOLS_FL_RIGHT, N_("partition number as read from the partition table"), COLTYPE_NUM },
+	[COL_PARTTYPENAME]  = { "PARTTYPENAME",  0.1,  0, N_("partition type name") },
+	[COL_PARTTYPE] = { "PARTTYPE", 36,  0, N_("partition type code or UUID") },
+	[COL_PARTUUID] = { "PARTUUID", 36,  0, N_("partition UUID") },
+	[COL_PATH] = { "PATH", 0.3, 0, N_("path to the device node") },
+	[COL_PHYSEC] = { "PHY-SEC", 7, SCOLS_FL_RIGHT, N_("physical sector size"), COLTYPE_NUM },
+	[COL_PKNAME] = { "PKNAME", 0.3, 0, N_("internal parent kernel device name") },
+	[COL_PTTYPE] = { "PTTYPE", 0.1, 0, N_("partition table type") },
+	[COL_PTUUID] = { "PTUUID", 36,  0, N_("partition table identifier (usually UUID)") },
+	[COL_RAND] = { "RAND", 1, SCOLS_FL_RIGHT, N_("adds randomness"), COLTYPE_BOOL },
+	[COL_RA] = { "RA", 3, SCOLS_FL_RIGHT, N_("read-ahead of the device"), COLTYPE_NUM },
+	[COL_REV] = { "REV", 4, SCOLS_FL_RIGHT, N_("device revision") },
+	[COL_RM] = { "RM", 1, SCOLS_FL_RIGHT, N_("removable device"), COLTYPE_BOOL },
+	[COL_ROTA] = { "ROTA", 1, SCOLS_FL_RIGHT, N_("rotational device"), COLTYPE_BOOL },
+	[COL_RO] = { "RO", 1, SCOLS_FL_RIGHT, N_("read-only device"), COLTYPE_BOOL },
+	[COL_RQ_SIZE]= { "RQ-SIZE", 5, SCOLS_FL_RIGHT, N_("request queue size"), COLTYPE_NUM },
+	[COL_SCHED] = { "SCHED", 0.1, 0, N_("I/O scheduler name") },
+	[COL_SERIAL] = { "SERIAL", 0.1, SCOLS_FL_TRUNC, N_("disk serial number") },
+	[COL_SIZE] = { "SIZE", 5, SCOLS_FL_RIGHT, N_("size of the device"), COLTYPE_SIZE },
+	[COL_START] = { "START", 5, SCOLS_FL_RIGHT, N_("partition start offset"), COLTYPE_NUM },
+	[COL_STATE] = { "STATE", 7, SCOLS_FL_TRUNC, N_("state of the device") },
+	[COL_SUBSYS] = { "SUBSYSTEMS", 0.1, SCOLS_FL_NOEXTREMES, N_("de-duplicated chain of subsystems") },
+	[COL_TARGETS] = { "MOUNTPOINTS", 0.10, SCOLS_FL_WRAP | SCOLS_FL_NOEXTREMES,  N_("all locations where device is mounted") },
+	[COL_TARGET] = { "MOUNTPOINT", 0.10, SCOLS_FL_TRUNC | SCOLS_FL_NOEXTREMES, N_("where the device is mounted") },
+	[COL_TRANSPORT] = { "TRAN", 6, 0, N_("device transport type") },
+	[COL_TYPE] = { "TYPE", 4, 0, N_("device type") },
+	[COL_UUID] = { "UUID", 36,  0, N_("filesystem UUID") },
+	[COL_VENDOR] = { "VENDOR", 0.1, SCOLS_FL_TRUNC, N_("device vendor") },
+	[COL_WSAME] = { "WSAME", 6, SCOLS_FL_RIGHT, N_("write same max bytes"), COLTYPE_SIZE },
+	[COL_WWN] = { "WWN", 18, 0, N_("unique storage identifier") },
+	[COL_ZONED] = { "ZONED", 0.3, 0, N_("zone model") },
+	[COL_ZONE_SZ] = { "ZONE-SZ", 9, SCOLS_FL_RIGHT, N_("zone size"), COLTYPE_SIZE },
+	[COL_ZONE_WGRAN] = { "ZONE-WGRAN", 10, SCOLS_FL_RIGHT, N_("zone write granularity"), COLTYPE_SIZE },
+	[COL_ZONE_APP] = { "ZONE-APP", 11, SCOLS_FL_RIGHT, N_("zone append max bytes"), COLTYPE_SIZE },
+	[COL_ZONE_NR] = { "ZONE-NR", 8, SCOLS_FL_RIGHT, N_("number of zones"), COLTYPE_NUM },
+	[COL_ZONE_OMAX] = { "ZONE-OMAX", 10, SCOLS_FL_RIGHT, N_("maximum number of open zones"), COLTYPE_NUM },
+	[COL_ZONE_AMAX] = { "ZONE-AMAX", 10, SCOLS_FL_RIGHT, N_("maximum number of active zones"), COLTYPE_NUM },
+};
+
+struct lsblk *lsblk;	/* global handler */
+
+/*
+ * columns[] array specifies all currently wanted output column. The columns
+ * are defined by infos[] array and you can specify (on command line) each
+ * column twice. That's enough, dynamically allocated array of the columns is
+ * unnecessary overkill and over-engineering in this case
+ */
+static int columns[ARRAY_SIZE(infos) * 2];
+static size_t ncolumns;
+
+static inline void add_column(int id)
+{
+	if (ncolumns >= ARRAY_SIZE(columns))
+		errx(EXIT_FAILURE, _("too many columns specified, "
+				     "the limit is %zu columns"),
+				ARRAY_SIZE(columns) - 1);
+	columns[ ncolumns++ ] =  id;
+}
+
+static inline void add_uniq_column(int id)
+{
+	if (column_id_to_number(id) < 0)
+		add_column(id);
+}
+
+static void lsblk_init_debug(void)
+{
+	__UL_INIT_DEBUG_FROM_ENV(lsblk, LSBLK_DEBUG_, 0, LSBLK_DEBUG);
+}
+
+/*
+ * exclude/include devices filter based on major device numbers
+ */
+static int excludes[256];
+static size_t nexcludes;
+
+static int includes[256];
+static size_t nincludes;
+
+static int is_maj_excluded(int maj)
+{
+	size_t i;
+
+	assert(ARRAY_SIZE(excludes) > nexcludes);
+
+	if (!nexcludes)
+		return 0;	/* filter not enabled, device not excluded */
+
+	for (i = 0; i < nexcludes; i++) {
+		if (excludes[i] == maj) {
+			DBG(FILTER, ul_debug("exclude: maj=%d", maj));
+			return 1;
+		}
+	}
+	return 0;
+}
+
+static int is_maj_included(int maj)
+{
+	size_t i;
+
+	assert(ARRAY_SIZE(includes) > nincludes);
+
+	if (!nincludes)
+		return 1;	/* filter not enabled, device is included */
+
+	for (i = 0; i < nincludes; i++) {
+		if (includes[i] == maj) {
+			DBG(FILTER, ul_debug("include: maj=%d", maj));
+			return 1;
+		}
+	}
+	return 0;
+}
+
+/* Converts column sequential number to column ID (COL_*) */
+static int get_column_id(int num)
+{
+	assert(num >= 0);
+	assert((size_t) num < ncolumns);
+	assert(columns[num] < (int) ARRAY_SIZE(infos));
+	return columns[num];
+}
+
+/* Returns column description for the column sequential number */
+static struct colinfo *get_column_info(int num)
+{
+	return &infos[ get_column_id(num) ];
+}
+
+/* Converts column name (as defined in the infos[] to the column ID */
+static int column_name_to_id(const char *name, size_t namesz)
+{
+	size_t i;
+
+	for (i = 0; i < ARRAY_SIZE(infos); i++) {
+		const char *cn = infos[i].name;
+
+		if (!strncasecmp(name, cn, namesz) && !*(cn + namesz))
+			return i;
+	}
+	warnx(_("unknown column: %s"), name);
+	return -1;
+}
+
+/* Converts column ID (COL_*) to column sequential number */
+static int column_id_to_number(int id)
+{
+	size_t i;
+
+	for (i = 0; i < ncolumns; i++)
+		if (columns[i] == id)
+			return i;
+	return -1;
+}
+
+/* Checks for DM prefix in the device name */
+static int is_dm(const char *name)
+{
+	return strncmp(name, "dm-", 3) ? 0 : 1;
+}
+
+/* Returns full pat to the device node (TODO: what about sysfs_blkdev_get_path()) */
+static char *get_device_path(struct lsblk_device *dev)
+{
+	char path[PATH_MAX];
+
+	assert(dev);
+	assert(dev->name);
+
+	if (is_dm(dev->name))
+		return __canonicalize_dm_name(lsblk->sysroot, dev->name);
+
+	snprintf(path, sizeof(path), "/dev/%s", dev->name);
+	sysfs_devname_sys_to_dev(path);
+	return xstrdup(path);
+}
+
+static int is_readonly_device(struct lsblk_device *dev)
+{
+	int fd, ro = 0;
+
+	if (ul_path_scanf(dev->sysfs, "ro", "%d", &ro) == 1)
+		return ro;
+
+	/* fallback if "ro" attribute does not exist */
+	fd = open(dev->filename, O_RDONLY);
+	if (fd != -1) {
+		if (ioctl(fd, BLKROGET, &ro) != 0)
+			ro = 0;
+		close(fd);
+	}
+	return ro;
+}
+
+static char *get_scheduler(struct lsblk_device *dev)
+{
+	char buf[128];
+	char *p, *res = NULL;
+
+	if (ul_path_read_buffer(dev->sysfs, buf, sizeof(buf), "queue/scheduler") == 0)
+		return NULL;
+	p = strchr(buf, '[');
+	if (p) {
+		res = p + 1;
+		p = strchr(res, ']');
+		if (p) {
+			*p = '\0';
+			res = xstrdup(res);
+		} else
+			res = NULL;
+	}
+	return res;
+}
+
+static char *get_type(struct lsblk_device *dev)
+{
+	char *res = NULL, *p;
+
+	if (device_is_partition(dev))
+		return xstrdup("part");
+
+	if (is_dm(dev->name)) {
+		char *dm_uuid = NULL;
+
+		/* The DM_UUID prefix should be set to subsystem owning
+		 * the device - LVM, CRYPT, DMRAID, MPATH, PART */
+		if (ul_path_read_string(dev->sysfs, &dm_uuid, "dm/uuid") > 0
+		    && dm_uuid) {
+			char *tmp = dm_uuid;
+			char *dm_uuid_prefix = strsep(&tmp, "-");
+
+			if (dm_uuid_prefix) {
+				/* kpartx hack to remove partition number */
+				if (strncasecmp(dm_uuid_prefix, "part", 4) == 0)
+					dm_uuid_prefix[4] = '\0';
+
+				res = xstrdup(dm_uuid_prefix);
+			}
+		}
+
+		free(dm_uuid);
+		if (!res)
+			/* No UUID or no prefix - just mark it as DM device */
+			res = xstrdup("dm");
+
+	} else if (!strncmp(dev->name, "loop", 4)) {
+		res = xstrdup("loop");
+
+	} else if (!strncmp(dev->name, "md", 2)) {
+		char *md_level = NULL;
+
+		ul_path_read_string(dev->sysfs, &md_level, "md/level");
+		res = md_level ? md_level : xstrdup("md");
+
+	} else {
+		const char *type = NULL;
+		int x = 0;
+
+		if (ul_path_read_s32(dev->sysfs, &x, "device/type") == 0)
+			type = blkdev_scsi_type_to_name(x);
+		if (!type)
+			type = "disk";
+		res = xstrdup(type);
+	}
+
+	for (p = res; p && *p; p++)
+		*p = tolower((unsigned char) *p);
+	return res;
+}
+
+/* Thanks to lsscsi code for idea of detection logic used here */
+static const char *get_transport(struct lsblk_device *dev)
+{
+	struct path_cxt *sysfs = dev->sysfs;
+	char *attr = NULL;
+	const char *trans = NULL;
+
+
+	/* SCSI - Serial Peripheral Interface */
+	if (sysfs_blkdev_scsi_host_is(sysfs, "spi"))
+		trans = "spi";
+
+	/* FC/FCoE - Fibre Channel / Fibre Channel over Ethernet */
+	else if (sysfs_blkdev_scsi_host_is(sysfs, "fc")) {
+		attr = sysfs_blkdev_scsi_host_strdup_attribute(sysfs, "fc", "symbolic_name");
+		if (!attr)
+			return NULL;
+		trans = strstr(attr, " over ") ? "fcoe" : "fc";
+		free(attr);
+	}
+
+	/* SAS - Serial Attached SCSI */
+	else if (sysfs_blkdev_scsi_host_is(sysfs, "sas") ||
+		 sysfs_blkdev_scsi_has_attribute(sysfs, "sas_device"))
+		trans = "sas";
+
+
+	/* SBP - Serial Bus Protocol (FireWire) */
+	else if (sysfs_blkdev_scsi_has_attribute(sysfs, "ieee1394_id"))
+		trans = "sbp";
+
+	/* iSCSI */
+	else if (sysfs_blkdev_scsi_host_is(sysfs, "iscsi"))
+		trans ="iscsi";
+
+	/* USB - Universal Serial Bus */
+	else if (sysfs_blkdev_scsi_path_contains(sysfs, "usb"))
+		trans = "usb";
+
+	/* ATA, SATA */
+	else if (sysfs_blkdev_scsi_host_is(sysfs, "scsi")) {
+		attr = sysfs_blkdev_scsi_host_strdup_attribute(sysfs, "scsi", "proc_name");
+		if (!attr)
+			return NULL;
+		if (!strncmp(attr, "ahci", 4) || !strncmp(attr, "sata", 4))
+			trans = "sata";
+		else if (strstr(attr, "ata"))
+			trans = "ata";
+		free(attr);
+
+	} else if (strncmp(dev->name, "nvme", 4) == 0) {
+		trans = "nvme";
+	} else if (strncmp(dev->name, "vd", 2) == 0)
+		trans = "virtio";
+	else if (strncmp(dev->name, "mmcblk", 6) == 0)
+		trans = "mmc";
+
+	return trans;
+}
+
+static char *get_subsystems(struct lsblk_device *dev)
+{
+	char path[PATH_MAX];
+	char *sub, *chain, *res = NULL;
+	size_t len = 0, last = 0;
+
+	chain = sysfs_blkdev_get_devchain(dev->sysfs, path, sizeof(path));
+	if (!chain)
+		return NULL;
+
+	while (sysfs_blkdev_next_subsystem(dev->sysfs, chain, &sub) == 0) {
+		size_t sz;
+
+		/* don't create "block:scsi:scsi", but "block:scsi" */
+		if (len && strcmp(res + last, sub) == 0) {
+			free(sub);
+			continue;
+		}
+
+		sz = strlen(sub);
+		res = xrealloc(res, len + sz + 2);
+		if (len)
+			res[len++] = ':';
+
+		memcpy(res + len, sub, sz + 1);
+		last = len;
+		len += sz;
+		free(sub);
+	}
+
+	return res;
+}
+
+
+#define is_parsable(_l)	(scols_table_is_raw((_l)->table) || \
+			 scols_table_is_export((_l)->table) || \
+			 scols_table_is_json((_l)->table))
+
+static char *mk_name(const char *name)
+{
+	char *p;
+	if (!name)
+		return NULL;
+	if (lsblk->paths)
+		xasprintf(&p, "/dev/%s", name);
+	else
+		p = xstrdup(name);
+	if (p)
+		sysfs_devname_sys_to_dev(p);
+	return p;
+}
+
+static char *mk_dm_name(const char *name)
+{
+	char *p;
+	if (!name)
+		return NULL;
+	if (lsblk->paths)
+		xasprintf(&p, "/dev/mapper/%s", name);
+	else
+		p = xstrdup(name);
+	return p;
+}
+
+/* stores data to scols cell userdata (invisible and independent on output)
+ * to make the original values accessible for sort functions
+ */
+static void set_sortdata_u64(struct libscols_line *ln, int col, uint64_t x)
+{
+	struct libscols_cell *ce = scols_line_get_cell(ln, col);
+	uint64_t *data;
+
+	if (!ce)
+		return;
+	data = xmalloc(sizeof(uint64_t));
+	*data = x;
+	scols_cell_set_userdata(ce, data);
+}
+
+/* do not modify *data on any error */
+static void str2u64(const char *str, uint64_t *data)
+{
+	uintmax_t num;
+	char *end = NULL;
+
+	errno = 0;
+	if (str == NULL || *str == '\0')
+		return;
+	num = strtoumax(str, &end, 10);
+
+	if (errno || str == end || (end && *end))
+		return;
+	*data = num;
+}
+
+static void unref_sortdata(struct libscols_table *tb)
+{
+	struct libscols_iter *itr;
+	struct libscols_line *ln;
+
+	if (!tb || !lsblk->sort_col)
+		return;
+	itr = scols_new_iter(SCOLS_ITER_FORWARD);
+	if (!itr)
+		return;
+	while (scols_table_next_line(tb, itr, &ln) == 0) {
+		struct libscols_cell *ce = scols_line_get_column_cell(ln,
+								lsblk->sort_col);
+		void *data = scols_cell_get_userdata(ce);
+		free(data);
+	}
+
+	scols_free_iter(itr);
+}
+
+static char *get_vfs_attribute(struct lsblk_device *dev, int id)
+{
+	char *sizestr;
+	uint64_t vfs_attr = 0;
+
+	if (!dev->fsstat.f_blocks) {
+		const char *mnt = lsblk_device_get_mountpoint(dev);
+		if (!mnt || dev->is_swap)
+			return NULL;
+		if (statvfs(mnt, &dev->fsstat) != 0)
+			return NULL;
+	}
+
+	switch(id) {
+	case COL_FSSIZE:
+		vfs_attr = dev->fsstat.f_frsize * dev->fsstat.f_blocks;
+		break;
+	case COL_FSAVAIL:
+		vfs_attr = dev->fsstat.f_frsize * dev->fsstat.f_bavail;
+		break;
+	case COL_FSUSED:
+		vfs_attr = dev->fsstat.f_frsize * (dev->fsstat.f_blocks - dev->fsstat.f_bfree);
+		break;
+	case COL_FSUSEPERC:
+		if (dev->fsstat.f_blocks == 0)
+			return xstrdup("-");
+
+		xasprintf(&sizestr, "%.0f%%",
+				(double)(dev->fsstat.f_blocks - dev->fsstat.f_bfree) /
+				dev->fsstat.f_blocks * 100);
+		return sizestr;
+	}
+
+	if (!vfs_attr)
+		sizestr = xstrdup("0");
+	else if (lsblk->bytes)
+		xasprintf(&sizestr, "%ju", vfs_attr);
+	else
+		sizestr = size_to_human_string(SIZE_SUFFIX_1LETTER, vfs_attr);
+
+	return sizestr;
+}
+
+static struct stat *device_get_stat(struct lsblk_device *dev)
+{
+	if (!dev->st.st_rdev
+	    && stat(dev->filename, &dev->st) != 0)
+		return NULL;
+
+	return &dev->st;
+}
+
+static int is_removable_device(struct lsblk_device *dev, struct lsblk_device *parent)
+{
+	struct path_cxt *pc;
+
+	if (dev->removable != -1)
+		goto done;
+
+	dev->removable = sysfs_blkdev_is_removable(dev->sysfs);
+
+	if (!dev->removable && parent) {
+		pc = sysfs_blkdev_get_parent(dev->sysfs);
+		if (!pc)
+			goto done;
+
+		if (pc == parent->sysfs)
+			/* dev is partition and parent is whole-disk  */
+			dev->removable = is_removable_device(parent, NULL);
+		else
+			/* parent is something else, use sysfs parent */
+			dev->removable = sysfs_blkdev_is_removable(pc);
+	}
+done:
+	if (dev->removable == -1)
+		dev->removable = 0;
+	return dev->removable;
+}
+
+static uint64_t device_get_discard_granularity(struct lsblk_device *dev)
+{
+	if (dev->discard_granularity == (uint64_t) -1
+	    && ul_path_read_u64(dev->sysfs, &dev->discard_granularity,
+			        "queue/discard_granularity") != 0)
+		dev->discard_granularity = 0;
+
+	return dev->discard_granularity;
+}
+
+static void device_read_bytes(struct lsblk_device *dev, char *path, char **str,
+			      uint64_t *sortdata)
+{
+	uint64_t x;
+
+	if (lsblk->bytes) {
+		ul_path_read_string(dev->sysfs, str, path);
+		if (sortdata)
+			str2u64(*str, sortdata);
+		return;
+	}
+
+	if (ul_path_read_u64(dev->sysfs, &x, path) == 0) {
+		*str = size_to_human_string(SIZE_SUFFIX_1LETTER, x);
+		if (sortdata)
+			*sortdata = x;
+	}
+}
+
+static void process_mq(struct lsblk_device *dev, char **str)
+{
+	unsigned int queues = 0;
+
+	DBG(DEV, ul_debugobj(dev, "%s: process mq", dev->name));
+
+	queues = ul_path_count_dirents(dev->sysfs, "mq");
+	if (!queues) {
+		*str = xstrdup("1");
+		DBG(DEV, ul_debugobj(dev, "%s: no mq supported, use a single queue", dev->name));
+		return;
+	}
+
+	DBG(DEV, ul_debugobj(dev, "%s: has %d queues", dev->name, queues));
+	xasprintf(str, "%3u", queues);
+}
+
+/*
+ * Generates data (string) for column specified by column ID for specified device. If sortdata
+ * is not NULL then returns number usable to sort the column if the data are available for the
+ * column.
+ */
+static char *device_get_data(
+		struct lsblk_device *dev,		/* device */
+		struct lsblk_device *parent,		/* device parent as defined in the tree */
+		int id,					/* column ID (COL_*) */
+		uint64_t *sortdata)			/* returns sort data as number */
+{
+	struct lsblk_devprop *prop = NULL;
+	char *str = NULL;
+
+	switch(id) {
+	case COL_NAME:
+		str = dev->dm_name ? mk_dm_name(dev->dm_name) :	mk_name(dev->name);
+		break;
+	case COL_KNAME:
+		str = mk_name(dev->name);
+		break;
+	case COL_PKNAME:
+		if (parent)
+			str = mk_name(parent->name);
+		break;
+	case COL_PATH:
+		if (dev->filename)
+			str = xstrdup(dev->filename);
+		break;
+	case COL_OWNER:
+		if (lsblk->sysroot)
+			prop = lsblk_device_get_properties(dev);
+		if (prop && prop->owner) {
+			str = xstrdup(prop->owner);
+		} else {
+			struct stat *st = device_get_stat(dev);
+			struct passwd *pw = st ? getpwuid(st->st_uid) : NULL;
+			if (pw)
+				str = xstrdup(pw->pw_name);
+		}
+		break;
+	case COL_GROUP:
+		if (lsblk->sysroot)
+			prop = lsblk_device_get_properties(dev);
+		if (prop && prop->group) {
+			str = xstrdup(prop->group);
+		} else {
+			struct stat *st = device_get_stat(dev);
+			struct group *gr = st ? getgrgid(st->st_gid) : NULL;
+			if (gr)
+				str = xstrdup(gr->gr_name);
+		}
+		break;
+	case COL_MODE:
+		if (lsblk->sysroot)
+			prop = lsblk_device_get_properties(dev);
+		if (prop && prop->mode) {
+			str = xstrdup(prop->mode);
+		} else {
+			struct stat *st = device_get_stat(dev);
+			char md[11] = { '\0' };
+
+			if (st)
+				str = xstrdup(xstrmode(st->st_mode, md));
+		}
+		break;
+	case COL_MAJMIN:
+		if (is_parsable(lsblk))
+			xasprintf(&str, "%u:%u", dev->maj, dev->min);
+		else
+			xasprintf(&str, "%3u:%-3u", dev->maj, dev->min);
+		if (sortdata)
+			*sortdata = makedev(dev->maj, dev->min);
+		break;
+	case COL_FSTYPE:
+		prop = lsblk_device_get_properties(dev);
+		if (prop && prop->fstype)
+			str = xstrdup(prop->fstype);
+		break;
+	case COL_FSSIZE:
+	case COL_FSAVAIL:
+	case COL_FSUSED:
+	case COL_FSUSEPERC:
+		str = get_vfs_attribute(dev, id);
+		break;
+	case COL_FSVERSION:
+		prop = lsblk_device_get_properties(dev);
+		if (prop && prop->fsversion)
+			str = xstrdup(prop->fsversion);
+		break;
+	case COL_TARGET:
+	{
+		const char *p = lsblk_device_get_mountpoint(dev);
+		if (p)
+			str = xstrdup(p);
+		break;
+	}
+	case COL_TARGETS:
+	{
+		size_t i, n = 0;
+		struct ul_buffer buf = UL_INIT_BUFFER;
+		struct libmnt_fs **fss = lsblk_device_get_filesystems(dev, &n);
+
+		for (i = 0; i < n; i++) {
+			struct libmnt_fs *fs = fss[i];
+			if (mnt_fs_is_swaparea(fs))
+				ul_buffer_append_string(&buf, "[SWAP]");
+			else
+				ul_buffer_append_string(&buf, mnt_fs_get_target(fs));
+			if (i + 1 < n)
+				ul_buffer_append_data(&buf, "\n", 1);
+		}
+		str = ul_buffer_get_data(&buf, NULL, NULL);
+		break;
+	}
+	case COL_FSROOTS:
+	{
+		size_t i, n = 0;
+		struct ul_buffer buf = UL_INIT_BUFFER;
+		struct libmnt_fs **fss = lsblk_device_get_filesystems(dev, &n);
+
+		for (i = 0; i < n; i++) {
+			struct libmnt_fs *fs = fss[i];
+			const char *root = mnt_fs_get_root(fs);
+			if (mnt_fs_is_swaparea(fs))
+				continue;
+			ul_buffer_append_string(&buf, root ? root : "/");
+			if (i + 1 < n)
+				ul_buffer_append_data(&buf, "\n", 1);
+		}
+		str = ul_buffer_get_data(&buf, NULL, NULL);
+		break;
+	}
+	case COL_LABEL:
+		prop = lsblk_device_get_properties(dev);
+		if (prop && prop->label)
+			str = xstrdup(prop->label);
+		break;
+	case COL_UUID:
+		prop = lsblk_device_get_properties(dev);
+		if (prop && prop->uuid)
+			str = xstrdup(prop->uuid);
+		break;
+	case COL_PTUUID:
+		prop = lsblk_device_get_properties(dev);
+		if (prop && prop->ptuuid)
+			str = xstrdup(prop->ptuuid);
+		break;
+	case COL_PTTYPE:
+		prop = lsblk_device_get_properties(dev);
+		if (prop && prop->pttype)
+			str = xstrdup(prop->pttype);
+		break;
+	case COL_PARTTYPE:
+		prop = lsblk_device_get_properties(dev);
+		if (prop && prop->parttype)
+			str = xstrdup(prop->parttype);
+		break;
+	case COL_PARTTYPENAME:
+		prop = lsblk_device_get_properties(dev);
+		if (prop && prop->parttype && prop->pttype) {
+			const char *x = lsblk_parttype_code_to_string(
+						prop->parttype, prop->pttype);
+			if (x)
+				str = xstrdup(x);
+		}
+		break;
+	case COL_PARTLABEL:
+		prop = lsblk_device_get_properties(dev);
+		if (prop && prop->partlabel)
+			str = xstrdup(prop->partlabel);
+		break;
+	case COL_PARTUUID:
+		prop = lsblk_device_get_properties(dev);
+		if (prop && prop->partuuid)
+			str = xstrdup(prop->partuuid);
+		break;
+	case COL_PARTFLAGS:
+		prop = lsblk_device_get_properties(dev);
+		if (prop && prop->partflags)
+			str = xstrdup(prop->partflags);
+		break;
+	case COL_PARTN:
+		prop = lsblk_device_get_properties(dev);
+		if (prop && prop->partn)
+			str = xstrdup(prop->partn);
+		break;
+	case COL_WWN:
+		prop = lsblk_device_get_properties(dev);
+		if (prop && prop->wwn)
+			str = xstrdup(prop->wwn);
+		break;
+	case COL_IDLINK:
+		prop = lsblk_device_get_properties(dev);
+		if (prop && prop->idlink)
+			str = xstrdup(prop->idlink);
+		break;
+	case COL_ID:
+		prop = lsblk_device_get_properties(dev);
+		if (prop && prop->idlink) {
+			/* skip bus/subsystem prefix */
+			const char *p = strchr(prop->idlink, '-');
+			str = p && *(p + 1) ? xstrdup(p+1) : xstrdup(prop->idlink);
+		}
+		break;
+	case COL_RA:
+		ul_path_read_string(dev->sysfs, &str, "queue/read_ahead_kb");
+		if (sortdata)
+			str2u64(str, sortdata);
+		break;
+	case COL_RO:
+		str = xstrdup(is_readonly_device(dev) ? "1" : "0");
+		break;
+	case COL_RM:
+		str = xstrdup(is_removable_device(dev, parent) ? "1" : "0");
+		break;
+	case COL_HOTPLUG:
+		str = sysfs_blkdev_is_hotpluggable(dev->sysfs) ? xstrdup("1") : xstrdup("0");
+		break;
+	case COL_ROTA:
+		ul_path_read_string(dev->sysfs, &str, "queue/rotational");
+		break;
+	case COL_RAND:
+		ul_path_read_string(dev->sysfs, &str, "queue/add_random");
+		break;
+	case COL_MODEL:
+		if (!device_is_partition(dev) && dev->nslaves == 0) {
+			prop = lsblk_device_get_properties(dev);
+			if (prop && prop->model)
+				str = xstrdup(prop->model);
+			else
+				ul_path_read_string(dev->sysfs, &str, "device/model");
+		}
+		break;
+	case COL_SERIAL:
+		if (!device_is_partition(dev) && dev->nslaves == 0) {
+			prop = lsblk_device_get_properties(dev);
+			if (prop && prop->serial)
+				str = xstrdup(prop->serial);
+			else
+				ul_path_read_string(dev->sysfs, &str, "device/serial");
+		}
+		break;
+	case COL_REV:
+		if (!device_is_partition(dev) && dev->nslaves == 0) {
+			prop = lsblk_device_get_properties(dev);
+			if (prop && prop->revision)
+				str = xstrdup(prop->revision);
+			else
+				ul_path_read_string(dev->sysfs, &str, "device/rev");
+		}
+		break;
+	case COL_VENDOR:
+		if (!device_is_partition(dev) && dev->nslaves == 0)
+			ul_path_read_string(dev->sysfs, &str, "device/vendor");
+		break;
+	case COL_SIZE:
+		if (lsblk->bytes)
+			xasprintf(&str, "%ju", dev->size);
+		else
+			str = size_to_human_string(SIZE_SUFFIX_1LETTER, dev->size);
+		if (sortdata)
+			*sortdata = dev->size;
+		break;
+	case COL_START:
+		ul_path_read_string(dev->sysfs, &str, "start");
+		if (sortdata)
+			str2u64(str, sortdata);
+		break;
+	case COL_STATE:
+		if (!device_is_partition(dev) && !dev->dm_name)
+			ul_path_read_string(dev->sysfs, &str, "device/state");
+		else if (dev->dm_name) {
+			int x = 0;
+			if (ul_path_read_s32(dev->sysfs, &x, "dm/suspended") == 0)
+				str = xstrdup(x ? "suspended" : "running");
+		}
+		break;
+	case COL_ALIOFF:
+		ul_path_read_string(dev->sysfs, &str, "alignment_offset");
+		if (sortdata)
+			str2u64(str, sortdata);
+		break;
+	case COL_MINIO:
+		ul_path_read_string(dev->sysfs, &str, "queue/minimum_io_size");
+		if (sortdata)
+			str2u64(str, sortdata);
+		break;
+	case COL_OPTIO:
+		ul_path_read_string(dev->sysfs, &str, "queue/optimal_io_size");
+		if (sortdata)
+			str2u64(str, sortdata);
+		break;
+	case COL_PHYSEC:
+		ul_path_read_string(dev->sysfs, &str, "queue/physical_block_size");
+		if (sortdata)
+			str2u64(str, sortdata);
+		break;
+	case COL_LOGSEC:
+		ul_path_read_string(dev->sysfs, &str, "queue/logical_block_size");
+		if (sortdata)
+			str2u64(str, sortdata);
+		break;
+	case COL_SCHED:
+		str = get_scheduler(dev);
+		break;
+	case COL_RQ_SIZE:
+		ul_path_read_string(dev->sysfs, &str, "queue/nr_requests");
+		if (sortdata)
+			str2u64(str, sortdata);
+		break;
+	case COL_TYPE:
+		str = get_type(dev);
+		break;
+	case COL_HCTL:
+	{
+		int h, c, t, l;
+		if (sysfs_blkdev_scsi_get_hctl(dev->sysfs, &h, &c, &t, &l) == 0)
+			xasprintf(&str, "%d:%d:%d:%d", h, c, t, l);
+		break;
+	}
+	case COL_TRANSPORT:
+	{
+		const char *trans = get_transport(dev);
+		if (trans)
+			str = xstrdup(trans);
+		break;
+	}
+	case COL_SUBSYS:
+		str = get_subsystems(dev);
+		break;
+	case COL_DALIGN:
+		if (device_get_discard_granularity(dev) > 0)
+			ul_path_read_string(dev->sysfs, &str, "discard_alignment");
+		if (!str)
+			str = xstrdup("0");
+		if (sortdata)
+			str2u64(str, sortdata);
+		break;
+	case COL_DGRAN:
+		if (lsblk->bytes) {
+			ul_path_read_string(dev->sysfs, &str, "queue/discard_granularity");
+			if (sortdata)
+				str2u64(str, sortdata);
+		} else {
+			uint64_t x = device_get_discard_granularity(dev);
+			str = size_to_human_string(SIZE_SUFFIX_1LETTER, x);
+			if (sortdata)
+				*sortdata = x;
+		}
+		break;
+	case COL_DMAX:
+		device_read_bytes(dev, "queue/discard_max_bytes", &str, sortdata);
+		break;
+	case COL_DZERO:
+		if (device_get_discard_granularity(dev) > 0)
+			ul_path_read_string(dev->sysfs, &str, "queue/discard_zeroes_data");
+		if (!str)
+			str = xstrdup("0");
+		break;
+	case COL_WSAME:
+		device_read_bytes(dev, "queue/write_same_max_bytes", &str, sortdata);
+		if (!str)
+			str = xstrdup("0");
+		break;
+	case COL_ZONED:
+		ul_path_read_string(dev->sysfs, &str, "queue/zoned");
+		break;
+	case COL_ZONE_SZ:
+	{
+		uint64_t x;
+
+		if (ul_path_read_u64(dev->sysfs, &x, "queue/chunk_sectors") == 0) {
+			x <<= 9;
+			if (lsblk->bytes)
+				xasprintf(&str, "%ju", x);
+			else
+				str = size_to_human_string(SIZE_SUFFIX_1LETTER, x);
+			if (sortdata)
+				*sortdata = x;
+		}
+		break;
+	}
+	case COL_ZONE_WGRAN:
+		device_read_bytes(dev, "queue/zone_write_granularity", &str, sortdata);
+		break;
+	case COL_ZONE_APP:
+		device_read_bytes(dev, "queue/zone_append_max_bytes", &str, sortdata);
+		break;
+	case COL_ZONE_NR:
+		ul_path_read_string(dev->sysfs, &str, "queue/nr_zones");
+		if (sortdata)
+			str2u64(str, sortdata);
+		break;
+	case COL_ZONE_OMAX:
+		ul_path_read_string(dev->sysfs, &str, "queue/max_open_zones");
+		if (!str)
+			str = xstrdup("0");
+		if (sortdata)
+			str2u64(str, sortdata);
+		break;
+	case COL_ZONE_AMAX:
+		ul_path_read_string(dev->sysfs, &str, "queue/max_active_zones");
+		if (!str)
+			str = xstrdup("0");
+		if (sortdata)
+			str2u64(str, sortdata);
+		break;
+	case COL_DAX:
+		ul_path_read_string(dev->sysfs, &str, "queue/dax");
+		break;
+	case COL_MQ:
+		process_mq(dev, &str);
+		break;
+	case COL_DISKSEQ:
+		ul_path_read_string(dev->sysfs, &str, "diskseq");
+		if (sortdata)
+			str2u64(str, sortdata);
+		break;
+	};
+
+	return str;
+}
+
+/*
+ * Adds data for all wanted columns about the device to the smartcols table
+ */
+static void device_to_scols(
+			struct lsblk_device *dev,
+			struct lsblk_device *parent,
+			struct libscols_table *tab,
+			struct libscols_line *parent_line)
+{
+	size_t i;
+	struct libscols_line *ln;
+	struct lsblk_iter itr;
+	struct lsblk_device *child = NULL;
+	int link_group = 0;
+
+
+	DBG(DEV, ul_debugobj(dev, "add '%s' to scols", dev->name));
+	ON_DBG(DEV, if (ul_path_isopen_dirfd(dev->sysfs)) ul_debugobj(dev, " %s ---> is open!", dev->name));
+
+	if (!parent && dev->wholedisk)
+		parent = dev->wholedisk;
+
+	/* Do not print device more than once on --list if tree order is not requested */
+	if (!(lsblk->flags & LSBLK_TREE) && !lsblk->force_tree_order && dev->is_printed)
+		return;
+
+	if (lsblk->merge && list_count_entries(&dev->parents) > 1) {
+		if (!lsblk_device_is_last_parent(dev, parent))
+			return;
+		link_group = 1;
+	}
+
+	ln = scols_table_new_line(tab, link_group ? NULL : parent_line);
+	if (!ln)
+		err(EXIT_FAILURE, _("failed to allocate output line"));
+
+	dev->is_printed = 1;
+
+	if (link_group) {
+		struct lsblk_device *p;
+		struct libscols_line *gr = parent_line;
+
+		/* Merge all my parents to the one group */
+		DBG(DEV, ul_debugobj(dev, " grouping parents [--merge]"));
+		lsblk_reset_iter(&itr, LSBLK_ITER_FORWARD);
+		while (lsblk_device_next_parent(dev, &itr, &p) == 0) {
+			if (!p->scols_line) {
+				DBG(DEV, ul_debugobj(dev, " *** ignore '%s' no scols line yet", p->name));
+				continue;
+			}
+			DBG(DEV, ul_debugobj(dev, " group '%s'", p->name));
+			scols_table_group_lines(tab, p->scols_line, gr, 0);
+		}
+
+		/* Link the group -- this makes group->child connection */
+		DBG(DEV, ul_debugobj(dev, " linking the group [--merge]"));
+		scols_line_link_group(ln, gr, 0);
+	}
+
+	/* read column specific data and set it to smartcols table line */
+	for (i = 0; i < ncolumns; i++) {
+		char *data;
+		int id = get_column_id(i);
+
+		if (lsblk->sort_id != id)
+			data = device_get_data(dev, parent, id, NULL);
+		else {
+			uint64_t sortdata = (uint64_t) -1;
+
+			data = device_get_data(dev, parent, id, &sortdata);
+			if (data && sortdata != (uint64_t) -1)
+				set_sortdata_u64(ln, i, sortdata);
+		}
+		DBG(DEV, ul_debugobj(dev, " refer data[%zu]=\"%s\"", i, data));
+		if (data && scols_line_refer_data(ln, i, data))
+			err(EXIT_FAILURE, _("failed to add output data"));
+	}
+
+	dev->scols_line = ln;
+
+	if (dev->npartitions == 0)
+		/* For partitions we often read from parental whole-disk sysfs,
+		 * otherwise we can close */
+		ul_path_close_dirfd(dev->sysfs);
+
+	lsblk_reset_iter(&itr, LSBLK_ITER_FORWARD);
+	while (lsblk_device_next_child(dev, &itr, &child) == 0) {
+		DBG(DEV, ul_debugobj(dev, "%s -> continue to child", dev->name));
+		device_to_scols(child, dev, tab, ln);
+		DBG(DEV, ul_debugobj(dev, "%s <- child done", dev->name));
+	}
+
+	/* Let's be careful with number of open files */
+	ul_path_close_dirfd(dev->sysfs);
+}
+
+/*
+ * Walks on tree and adds one line for each device to the smartcols table
+ */
+static void devtree_to_scols(struct lsblk_devtree *tr, struct libscols_table *tab)
+{
+	struct lsblk_iter itr;
+	struct lsblk_device *dev = NULL;
+
+	lsblk_reset_iter(&itr, LSBLK_ITER_FORWARD);
+
+	while (lsblk_devtree_next_root(tr, &itr, &dev) == 0)
+		device_to_scols(dev, NULL, tab, NULL);
+}
+
+static int ignore_empty(struct lsblk_device *dev)
+{
+	/* show all non-empty devices */
+	if (dev->size)
+		return 0;
+
+	if (lsblk->noempty && dev->size == 0)
+		return 1;
+
+	/* ignore empty loop devices without backing file */
+	if (dev->maj == LOOPDEV_MAJOR &&
+	    !loopdev_has_backing_file(dev->filename))
+		return 1;
+
+	return 0;
+}
+
+/*
+ * Reads very basic information about the device from sysfs into the device struct
+ */
+static int initialize_device(struct lsblk_device *dev,
+		    struct lsblk_device *wholedisk,
+		    const char *name)
+{
+	dev_t devno;
+
+	DBG(DEV, ul_debugobj(dev, "initialize %s [wholedisk=%p %s]",
+			name, wholedisk, wholedisk ? wholedisk->name : ""));
+
+	if (sysfs_devname_is_hidden(lsblk->sysroot, name)) {
+		DBG(DEV, ul_debugobj(dev, "%s: hidden, ignore", name));
+		return -1;
+	}
+
+	dev->name = xstrdup(name);
+
+	if (wholedisk) {
+		dev->wholedisk = wholedisk;
+		lsblk_ref_device(wholedisk);
+	}
+
+	dev->filename = get_device_path(dev);
+	if (!dev->filename) {
+		DBG(DEV, ul_debugobj(dev, "%s: failed to get device path", dev->name));
+		return -1;
+	}
+	DBG(DEV, ul_debugobj(dev, "%s: filename=%s", dev->name, dev->filename));
+
+	devno = __sysfs_devname_to_devno(lsblk->sysroot, dev->name, wholedisk ? wholedisk->name : NULL);
+	if (!devno) {
+		DBG(DEV, ul_debugobj(dev, "%s: unknown device name", dev->name));
+		return -1;
+	}
+
+	dev->sysfs = ul_new_sysfs_path(devno, wholedisk ? wholedisk->sysfs : NULL, lsblk->sysroot);
+	if (!dev->sysfs) {
+		DBG(DEV, ul_debugobj(dev, "%s: failed to initialize sysfs handler", dev->name));
+		return -1;
+	}
+
+	dev->maj = major(devno);
+	dev->min = minor(devno);
+	dev->size = 0;
+
+	if (ul_path_read_u64(dev->sysfs, &dev->size, "size") == 0)	/* in sectors */
+		dev->size <<= 9;					/* in bytes */
+
+	/* Ignore devices of zero size */
+	if (!lsblk->all_devices && ignore_empty(dev)) {
+		DBG(DEV, ul_debugobj(dev, "zero size device -- ignore"));
+		return -1;
+	}
+	if (is_dm(dev->name)) {
+		ul_path_read_string(dev->sysfs, &dev->dm_name, "dm/name");
+		if (!dev->dm_name) {
+			DBG(DEV, ul_debugobj(dev, "%s: failed to get dm name", dev->name));
+			return -1;
+		}
+	}
+
+	dev->npartitions = sysfs_blkdev_count_partitions(dev->sysfs, dev->name);
+	dev->nholders = ul_path_count_dirents(dev->sysfs, "holders");
+	dev->nslaves = ul_path_count_dirents(dev->sysfs, "slaves");
+
+	DBG(DEV, ul_debugobj(dev, "%s: npartitions=%d, nholders=%d, nslaves=%d",
+			dev->name, dev->npartitions, dev->nholders, dev->nslaves));
+
+	/* ignore non-SCSI devices */
+	if (lsblk->scsi && sysfs_blkdev_scsi_get_hctl(dev->sysfs, NULL, NULL, NULL, NULL)) {
+		DBG(DEV, ul_debugobj(dev, "non-scsi device -- ignore"));
+		return -1;
+	}
+
+	/* ignore non-NVMe devices */
+	if (lsblk->nvme) {
+		const char *transport = get_transport(dev);
+
+		if (!transport || strcmp(transport, "nvme")) {
+			DBG(DEV, ul_debugobj(dev, "non-nvme device -- ignore"));
+			return -1;
+		}
+	}
+
+	/* ignore non-virtio devices */
+	if (lsblk->virtio) {
+		const char *transport = get_transport(dev);
+
+		if (!transport || strcmp(transport, "virtio")) {
+			DBG(DEV, ul_debugobj(dev, "non-virtio device -- ignore"));
+			return -1;
+		}
+	}
+
+	DBG(DEV, ul_debugobj(dev, "%s: context successfully initialized", dev->name));
+	return 0;
+}
+
+static struct lsblk_device *devtree_get_device_or_new(struct lsblk_devtree *tr,
+					       struct lsblk_device *disk,
+					       const char *name)
+{
+	struct lsblk_device *dev = lsblk_devtree_get_device(tr, name);
+
+	if (!dev) {
+		dev = lsblk_new_device();
+		if (!dev)
+			err(EXIT_FAILURE, _("failed to allocate device"));
+
+		if (initialize_device(dev, disk, name) != 0) {
+			lsblk_unref_device(dev);
+			return NULL;
+		}
+		lsblk_devtree_add_device(tr, dev);
+		lsblk_unref_device(dev);		/* keep it referenced by devtree only */
+	} else
+		DBG(DEV, ul_debugobj(dev, "%s: already processed", name));
+
+	return dev;
+}
+
+static struct lsblk_device *devtree_pktcdvd_get_dep(
+			struct lsblk_devtree *tr,
+			struct lsblk_device *dev,
+			int want_slave)
+{
+	char buf[PATH_MAX], *name;
+	dev_t devno;
+
+	devno = lsblk_devtree_pktcdvd_get_mate(tr,
+			makedev(dev->maj, dev->min), !want_slave);
+	if (!devno)
+		return NULL;
+
+	name = sysfs_devno_to_devname(devno, buf, sizeof(buf));
+	if (!name)
+		return NULL;
+
+	return devtree_get_device_or_new(tr, NULL, name);
+}
+
+static int process_dependencies(
+			struct lsblk_devtree *tr,
+			struct lsblk_device *dev,
+			int do_partitions);
+
+/*
+ * Read devices from whole-disk device into tree
+ */
+static int process_partitions(struct lsblk_devtree *tr, struct lsblk_device *disk)
+{
+	DIR *dir;
+	struct dirent *d;
+
+	assert(disk);
+
+	/*
+	 * Do not process further if there are no partitions for
+	 * this device or the device itself is a partition.
+	 */
+	if (!disk->npartitions || device_is_partition(disk))
+		return -EINVAL;
+
+	DBG(DEV, ul_debugobj(disk, "%s: probe whole-disk for partitions", disk->name));
+
+	dir = ul_path_opendir(disk->sysfs, NULL);
+	if (!dir)
+		err(EXIT_FAILURE, _("failed to open device directory in sysfs"));
+
+	while ((d = xreaddir(dir))) {
+		struct lsblk_device *part;
+
+		if (!(sysfs_blkdev_is_partition_dirent(dir, d, disk->name)))
+			continue;
+
+		DBG(DEV, ul_debugobj(disk, "  checking %s", d->d_name));
+
+		part = devtree_get_device_or_new(tr, disk, d->d_name);
+		if (!part)
+			continue;
+
+		if (lsblk_device_new_dependence(disk, part) == 0)
+			process_dependencies(tr, part, 0);
+
+		ul_path_close_dirfd(part->sysfs);
+	}
+
+	/* For partitions we need parental (whole-disk) sysfs directory pretty
+	 * often, so close it now when all is done */
+	ul_path_close_dirfd(disk->sysfs);
+
+	DBG(DEV, ul_debugobj(disk, "probe whole-disk for partitions -- done"));
+	closedir(dir);
+	return 0;
+}
+
+static char *get_wholedisk_from_partition_dirent(DIR *dir, struct dirent *d, char *buf, size_t bufsz)
+{
+	char *p;
+	int len;
+
+	if ((len = readlinkat(dirfd(dir), d->d_name, buf, bufsz - 1)) < 0)
+		return 0;
+
+	buf[len] = '\0';
+
+	/* The path ends with ".../<device>/<partition>" */
+	p = strrchr(buf, '/');
+	if (!p)
+		return NULL;
+	*p = '\0';
+
+	p = strrchr(buf, '/');
+	if (!p)
+		return NULL;
+	p++;
+
+	return p;
+}
+
+/*
+ * Reads slaves/holders and partitions for specified device into device tree
+ */
+static int process_dependencies(
+			struct lsblk_devtree *tr,
+			struct lsblk_device *dev,
+			int do_partitions)
+{
+	DIR *dir;
+	struct dirent *d;
+	const char *depname;
+	struct lsblk_device *dep = NULL;
+
+	assert(dev);
+
+	if (lsblk->nodeps)
+		return 0;
+
+	/* read all or specified partition */
+	if (do_partitions && dev->npartitions)
+		process_partitions(tr, dev);
+
+	DBG(DEV, ul_debugobj(dev, "%s: reading dependencies", dev->name));
+
+	if (!(lsblk->inverse ? dev->nslaves : dev->nholders)) {
+		DBG(DEV, ul_debugobj(dev, " ignore (no slaves/holders)"));
+		goto done;
+	}
+
+	depname = lsblk->inverse ? "slaves" : "holders";
+	dir = ul_path_opendir(dev->sysfs, depname);
+	if (!dir) {
+		DBG(DEV, ul_debugobj(dev, " ignore (no slaves/holders directory)"));
+		goto done;
+	}
+	ul_path_close_dirfd(dev->sysfs);
+
+	DBG(DEV, ul_debugobj(dev, " %s: checking for '%s' dependence", dev->name, depname));
+
+	while ((d = xreaddir(dir))) {
+		struct lsblk_device *disk = NULL;
+
+		/* Is the dependency a partition? */
+		if (sysfs_blkdev_is_partition_dirent(dir, d, NULL)) {
+
+			char buf[PATH_MAX];
+			char *diskname;
+
+			DBG(DEV, ul_debugobj(dev, " %s: dependence is partition", d->d_name));
+
+			diskname = get_wholedisk_from_partition_dirent(dir, d, buf, sizeof(buf));
+			if (diskname)
+				disk = devtree_get_device_or_new(tr, NULL, diskname);
+			if (!disk) {
+				DBG(DEV, ul_debugobj(dev, "  ignore no wholedisk ???"));
+				goto next;
+			}
+
+			dep = devtree_get_device_or_new(tr, disk, d->d_name);
+			if (!dep)
+				goto next;
+
+			if (lsblk_device_new_dependence(dev, dep) == 0)
+				process_dependencies(tr, dep, 1);
+
+			if (lsblk->inverse
+			    && lsblk_device_new_dependence(dep, disk) == 0)
+				process_dependencies(tr, disk, 0);
+		}
+		/* The dependency is a whole device. */
+		else {
+			DBG(DEV, ul_debugobj(dev, " %s: %s: dependence is whole-disk",
+								dev->name, d->d_name));
+
+			dep = devtree_get_device_or_new(tr, NULL, d->d_name);
+			if (!dep)
+				goto next;
+
+			if (lsblk_device_new_dependence(dev, dep) == 0)
+				/* For inverse tree we don't want to show partitions
+				 * if the dependence is on whole-disk */
+				process_dependencies(tr, dep, lsblk->inverse ? 0 : 1);
+		}
+next:
+		if (dep && dep->sysfs)
+			ul_path_close_dirfd(dep->sysfs);
+		if (disk && disk->sysfs)
+			ul_path_close_dirfd(disk->sysfs);
+	}
+	closedir(dir);
+done:
+	dep = devtree_pktcdvd_get_dep(tr, dev, lsblk->inverse);
+
+	if (dep && lsblk_device_new_dependence(dev, dep) == 0) {
+		lsblk_devtree_remove_root(tr, dep);
+		process_dependencies(tr, dep, lsblk->inverse ? 0 : 1);
+	}
+
+	return 0;
+}
+
+/*
+ * Defines the device as root node in the device tree and walks on all dependencies of the device.
+ */
+static int __process_one_device(struct lsblk_devtree *tr, char *devname, dev_t devno)
+{
+	struct lsblk_device *dev = NULL;
+	struct lsblk_device *disk = NULL;
+	char buf[PATH_MAX + 1], *name = NULL, *diskname = NULL;
+	int real_part = 0, rc = -EINVAL;
+
+	if (devno == 0 && devname) {
+		struct stat st;
+
+		DBG(DEV, ul_debug("%s: reading alone device", devname));
+
+		if (stat(devname, &st) || !S_ISBLK(st.st_mode)) {
+			warnx(_("%s: not a block device"), devname);
+			goto leave;
+		}
+		devno = st.st_rdev;
+	} else if (devno) {
+		DBG(DEV, ul_debug("%d:%d: reading alone device", major(devno), minor(devno)));
+	} else {
+		assert(devno || devname);
+		return -EINVAL;
+	}
+
+	/* TODO: sysfs_devno_to_devname() internally initializes path_cxt, it
+	 * would be better to use ul_new_sysfs_path() + sysfs_blkdev_get_name()
+	 * and reuse path_cxt for initialize_device()
+	 */
+	name = sysfs_devno_to_devname(devno, buf, sizeof(buf));
+	if (!name) {
+		if (devname)
+			warn(_("%s: failed to get sysfs name"), devname);
+		goto leave;
+	}
+	name = xstrdup(name);
+
+	if (!strncmp(name, "dm-", 3)) {
+		/* dm mapping is never a real partition! */
+		real_part = 0;
+	} else {
+		dev_t diskno = 0;
+
+		if (blkid_devno_to_wholedisk(devno, buf, sizeof(buf), &diskno)) {
+			warn(_("%s: failed to get whole-disk device number"), name);
+			goto leave;
+		}
+		diskname = buf;
+		real_part = devno != diskno;
+	}
+
+	if (!real_part) {
+		/*
+		 * Device is not a partition.
+		 */
+		DBG(DEV, ul_debug(" non-partition"));
+
+		dev = devtree_get_device_or_new(tr, NULL, name);
+		if (!dev)
+			goto leave;
+
+		lsblk_devtree_add_root(tr, dev);
+		process_dependencies(tr, dev, !lsblk->inverse);
+	} else {
+		/*
+		 * Partition, read sysfs name of the disk device
+		 */
+		DBG(DEV, ul_debug(" partition"));
+
+		disk = devtree_get_device_or_new(tr, NULL, diskname);
+		if (!disk)
+			goto leave;
+
+		dev = devtree_get_device_or_new(tr, disk, name);
+		if (!dev)
+			goto leave;
+
+		lsblk_devtree_add_root(tr, dev);
+		process_dependencies(tr, dev, 1);
+
+		if (lsblk->inverse
+		    && lsblk_device_new_dependence(dev, disk) == 0)
+			process_dependencies(tr, disk, 0);
+		else
+			ul_path_close_dirfd(disk->sysfs);
+	}
+
+	rc = 0;
+leave:
+	if (dev && dev->sysfs)
+		ul_path_close_dirfd(dev->sysfs);
+	if (disk && disk->sysfs)
+		ul_path_close_dirfd(disk->sysfs);
+	free(name);
+	return rc;
+}
+
+static int process_one_device(struct lsblk_devtree *tr, char *devname)
+{
+	assert(devname);
+	return __process_one_device(tr, devname, 0);
+}
+
+/*
+ * The /sys/block contains only root devices, and no partitions. It seems more
+ * simple to scan /sys/dev/block where are all devices without exceptions to get
+ * top-level devices for the reverse tree.
+ */
+static int process_all_devices_inverse(struct lsblk_devtree *tr)
+{
+	DIR *dir;
+	struct dirent *d;
+	struct path_cxt *pc = ul_new_path(_PATH_SYS_DEVBLOCK);
+
+	assert(lsblk->inverse);
+
+	if (!pc)
+		err(EXIT_FAILURE, _("failed to allocate /sys handler"));
+
+	ul_path_set_prefix(pc, lsblk->sysroot);
+	dir = ul_path_opendir(pc, NULL);
+	if (!dir)
+		goto done;
+
+	DBG(DEV, ul_debug("iterate on " _PATH_SYS_DEVBLOCK));
+
+	while ((d = xreaddir(dir))) {
+		dev_t devno;
+		int maj, min;
+
+		DBG(DEV, ul_debug(" %s dentry", d->d_name));
+
+		if (sscanf(d->d_name, "%d:%d", &maj, &min) != 2)
+			continue;
+		devno = makedev(maj, min);
+
+		if (is_maj_excluded(maj) || !is_maj_included(maj))
+			continue;
+		if (ul_path_countf_dirents(pc, "%s/holders", d->d_name) != 0)
+			continue;
+		if (sysfs_devno_count_partitions(devno) != 0)
+			continue;
+		__process_one_device(tr, NULL, devno);
+	}
+
+	closedir(dir);
+done:
+	ul_unref_path(pc);
+	DBG(DEV, ul_debug("iterate on " _PATH_SYS_DEVBLOCK " -- done"));
+	return 0;
+}
+
+/*
+ * Reads root nodes (devices) from /sys/block into devices tree
+ */
+static int process_all_devices(struct lsblk_devtree *tr)
+{
+	DIR *dir;
+	struct dirent *d;
+	struct path_cxt *pc;
+
+	assert(lsblk->inverse == 0);
+
+	pc = ul_new_path(_PATH_SYS_BLOCK);
+	if (!pc)
+		err(EXIT_FAILURE, _("failed to allocate /sys handler"));
+
+	ul_path_set_prefix(pc, lsblk->sysroot);
+	dir = ul_path_opendir(pc, NULL);
+	if (!dir)
+		goto done;
+
+	DBG(DEV, ul_debug("iterate on " _PATH_SYS_BLOCK));
+
+	while ((d = xreaddir(dir))) {
+		struct lsblk_device *dev = NULL;
+
+		DBG(DEV, ul_debug(" %s dentry", d->d_name));
+		dev = devtree_get_device_or_new(tr, NULL, d->d_name);
+		if (!dev)
+			goto next;
+
+		/* remove unwanted devices */
+		if (is_maj_excluded(dev->maj) || !is_maj_included(dev->maj)) {
+			DBG(DEV, ul_debug(" %s: ignore (by filter)", d->d_name));
+			lsblk_devtree_remove_device(tr, dev);
+			dev = NULL;
+			goto next;
+		}
+
+		if (dev->nslaves) {
+			DBG(DEV, ul_debug(" %s: ignore (in-middle)", d->d_name));
+			goto next;
+		}
+
+		lsblk_devtree_add_root(tr, dev);
+		process_dependencies(tr, dev, 1);
+next:
+		/* Let's be careful with number of open files */
+		if (dev && dev->sysfs)
+			ul_path_close_dirfd(dev->sysfs);
+	}
+
+	closedir(dir);
+done:
+	ul_unref_path(pc);
+	DBG(DEV, ul_debug("iterate on " _PATH_SYS_BLOCK " -- done"));
+	return 0;
+}
+
+/*
+ * Parses major numbers as specified on lsblk command line
+ */
+static void parse_excludes(const char *str0)
+{
+	const char *str = str0;
+
+	while (str && *str) {
+		char *end = NULL;
+		unsigned long n;
+
+		errno = 0;
+		n = strtoul(str, &end, 10);
+
+		if (end == str || (end && *end && *end != ','))
+			errx(EXIT_FAILURE, _("failed to parse list '%s'"), str0);
+		if (errno != 0 && (n == ULONG_MAX || n == 0))
+			err(EXIT_FAILURE, _("failed to parse list '%s'"), str0);
+		excludes[nexcludes++] = n;
+
+		if (nexcludes == ARRAY_SIZE(excludes))
+			/* TRANSLATORS: The standard value for %d is 256. */
+			errx(EXIT_FAILURE, _("the list of excluded devices is "
+					"too large (limit is %d devices)"),
+					(int)ARRAY_SIZE(excludes));
+
+		str = end && *end ? end + 1 : NULL;
+	}
+}
+
+/*
+ * Parses major numbers as specified on lsblk command line
+ * (TODO: what about refactor and merge parse_excludes() and parse_includes().)
+ */
+static void parse_includes(const char *str0)
+{
+	const char *str = str0;
+
+	while (str && *str) {
+		char *end = NULL;
+		unsigned long n;
+
+		errno = 0;
+		n = strtoul(str, &end, 10);
+
+		if (end == str || (end && *end && *end != ','))
+			errx(EXIT_FAILURE, _("failed to parse list '%s'"), str0);
+		if (errno != 0 && (n == ULONG_MAX || n == 0))
+			err(EXIT_FAILURE, _("failed to parse list '%s'"), str0);
+		includes[nincludes++] = n;
+
+		if (nincludes == ARRAY_SIZE(includes))
+			/* TRANSLATORS: The standard value for %d is 256. */
+			errx(EXIT_FAILURE, _("the list of included devices is "
+					"too large (limit is %d devices)"),
+					(int)ARRAY_SIZE(includes));
+		str = end && *end ? end + 1 : NULL;
+	}
+}
+
+/*
+ * see set_sortdata_u64() and columns initialization in main()
+ */
+static int cmp_u64_cells(struct libscols_cell *a,
+			 struct libscols_cell *b,
+			 __attribute__((__unused__)) void *data)
+{
+	uint64_t *adata = (uint64_t *) scols_cell_get_userdata(a),
+		 *bdata = (uint64_t *) scols_cell_get_userdata(b);
+
+	if (adata == NULL && bdata == NULL)
+		return 0;
+	if (adata == NULL)
+		return -1;
+	if (bdata == NULL)
+		return 1;
+	return *adata == *bdata ? 0 : *adata >= *bdata ? 1 : -1;
+}
+
+static void device_set_dedupkey(
+			struct lsblk_device *dev,
+			struct lsblk_device *parent,
+			int id)
+{
+	struct lsblk_iter itr;
+	struct lsblk_device *child = NULL;
+
+	dev->dedupkey = device_get_data(dev, parent, id, NULL);
+	if (dev->dedupkey)
+		DBG(DEV, ul_debugobj(dev, "%s: de-duplication key: %s", dev->name, dev->dedupkey));
+
+	if (dev->npartitions == 0)
+		/* For partitions we often read from parental whole-disk sysfs,
+		 * otherwise we can close */
+		ul_path_close_dirfd(dev->sysfs);
+
+	lsblk_reset_iter(&itr, LSBLK_ITER_FORWARD);
+
+	while (lsblk_device_next_child(dev, &itr, &child) == 0)
+		device_set_dedupkey(child, dev, id);
+
+	/* Let's be careful with number of open files */
+	ul_path_close_dirfd(dev->sysfs);
+}
+
+static void devtree_set_dedupkeys(struct lsblk_devtree *tr, int id)
+{
+	struct lsblk_iter itr;
+	struct lsblk_device *dev = NULL;
+
+	lsblk_reset_iter(&itr, LSBLK_ITER_FORWARD);
+
+	while (lsblk_devtree_next_root(tr, &itr, &dev) == 0)
+		device_set_dedupkey(dev, NULL, id);
+}
+
+static void __attribute__((__noreturn__)) usage(void)
+{
+	FILE *out = stdout;
+	size_t i;
+
+	fputs(USAGE_HEADER, out);
+	fprintf(out, _(" %s [options] [<device> ...]\n"), program_invocation_short_name);
+
+	fputs(USAGE_SEPARATOR, out);
+	fputs(_("List information about block devices.\n"), out);
+
+	fputs(USAGE_OPTIONS, out);
+	fputs(_(" -A, --noempty        don't print empty devices\n"), out);
+	fputs(_(" -D, --discard        print discard capabilities\n"), out);
+	fputs(_(" -E, --dedup <column> de-duplicate output by <column>\n"), out);
+	fputs(_(" -I, --include <list> show only devices with specified major numbers\n"), out);
+	fputs(_(" -J, --json           use JSON output format\n"), out);
+	fputs(_(" -M, --merge          group parents of sub-trees (usable for RAIDs, Multi-path)\n"), out);
+	fputs(_(" -O, --output-all     output all columns\n"), out);
+	fputs(_(" -P, --pairs          use key=\"value\" output format\n"), out);
+	fputs(_(" -S, --scsi           output info about SCSI devices\n"), out);
+	fputs(_(" -N, --nvme           output info about NVMe devices\n"), out);
+	fputs(_(" -v, --virtio         output info about virtio devices\n"), out);
+	fputs(_(" -T, --tree[=<column>] use tree format output\n"), out);
+	fputs(_(" -a, --all            print all devices\n"), out);
+	fputs(_(" -b, --bytes          print SIZE in bytes rather than in human readable format\n"), out);
+	fputs(_(" -d, --nodeps         don't print slaves or holders\n"), out);
+	fputs(_(" -e, --exclude <list> exclude devices by major number (default: RAM disks)\n"), out);
+	fputs(_(" -f, --fs             output info about filesystems\n"), out);
+	fputs(_(" -i, --ascii          use ascii characters only\n"), out);
+	fputs(_(" -l, --list           use list format output\n"), out);
+	fputs(_(" -m, --perms          output info about permissions\n"), out);
+	fputs(_(" -n, --noheadings     don't print headings\n"), out);
+	fputs(_(" -o, --output <list>  output columns\n"), out);
+	fputs(_(" -p, --paths          print complete device path\n"), out);
+	fputs(_(" -r, --raw            use raw output format\n"), out);
+	fputs(_(" -s, --inverse        inverse dependencies\n"), out);
+	fputs(_(" -t, --topology       output info about topology\n"), out);
+	fputs(_(" -w, --width <num>    specifies output width as number of characters\n"), out);
+	fputs(_(" -x, --sort <column>  sort output by <column>\n"), out);
+	fputs(_(" -y, --shell          use column names to be usable as shell variable identifiers\n"), out);
+	fputs(_(" -z, --zoned          print zone related information\n"), out);
+	fputs(_("     --sysroot <dir>  use specified directory as system root\n"), out);
+	fputs(USAGE_SEPARATOR, out);
+	printf(USAGE_HELP_OPTIONS(22));
+
+	fprintf(out, USAGE_COLUMNS);
+
+	for (i = 0; i < ARRAY_SIZE(infos); i++)
+		fprintf(out, " %12s  %s\n", infos[i].name, _(infos[i].help));
+
+	printf(USAGE_MAN_TAIL("lsblk(8)"));
+
+	exit(EXIT_SUCCESS);
+}
+
+static void check_sysdevblock(void)
+{
+	if (access(_PATH_SYS_DEVBLOCK, R_OK) != 0)
+		err(EXIT_FAILURE, _("failed to access sysfs directory: %s"),
+		    _PATH_SYS_DEVBLOCK);
+}
+
+int main(int argc, char *argv[])
+{
+	struct lsblk _ls = {
+		.sort_id = -1,
+		.dedup_id = -1,
+		.flags = LSBLK_TREE,
+		.tree_id = COL_NAME
+	};
+	struct lsblk_devtree *tr = NULL;
+	int c, status = EXIT_FAILURE;
+	char *outarg = NULL;
+	size_t i;
+	unsigned int width = 0;
+	int force_tree = 0, has_tree_col = 0;
+
+	enum {
+		OPT_SYSROOT = CHAR_MAX + 1
+	};
+
+	static const struct option longopts[] = {
+		{ "all",	no_argument,       NULL, 'a' },
+		{ "bytes",      no_argument,       NULL, 'b' },
+		{ "nodeps",     no_argument,       NULL, 'd' },
+		{ "noempty",    no_argument,       NULL, 'A' },
+		{ "discard",    no_argument,       NULL, 'D' },
+		{ "dedup",      required_argument, NULL, 'E' },
+		{ "zoned",      no_argument,       NULL, 'z' },
+		{ "help",	no_argument,       NULL, 'h' },
+		{ "json",       no_argument,       NULL, 'J' },
+		{ "output",     required_argument, NULL, 'o' },
+		{ "output-all", no_argument,       NULL, 'O' },
+		{ "merge",      no_argument,       NULL, 'M' },
+		{ "perms",      no_argument,       NULL, 'm' },
+		{ "noheadings",	no_argument,       NULL, 'n' },
+		{ "list",       no_argument,       NULL, 'l' },
+		{ "ascii",	no_argument,       NULL, 'i' },
+		{ "raw",        no_argument,       NULL, 'r' },
+		{ "inverse",	no_argument,       NULL, 's' },
+		{ "fs",         no_argument,       NULL, 'f' },
+		{ "exclude",    required_argument, NULL, 'e' },
+		{ "include",    required_argument, NULL, 'I' },
+		{ "topology",   no_argument,       NULL, 't' },
+		{ "paths",      no_argument,       NULL, 'p' },
+		{ "pairs",      no_argument,       NULL, 'P' },
+		{ "scsi",       no_argument,       NULL, 'S' },
+		{ "nvme",       no_argument,       NULL, 'N' },
+		{ "virtio",     no_argument,       NULL, 'v' },
+		{ "sort",	required_argument, NULL, 'x' },
+		{ "sysroot",    required_argument, NULL, OPT_SYSROOT },
+		{ "shell",      no_argument,       NULL, 'y' },
+		{ "tree",       optional_argument, NULL, 'T' },
+		{ "version",    no_argument,       NULL, 'V' },
+		{ "width",	required_argument, NULL, 'w' },
+		{ NULL, 0, NULL, 0 },
+	};
+
+	static const ul_excl_t excl[] = {       /* rows and cols in ASCII order */
+		{ 'D','O' },
+		{ 'I','e' },
+		{ 'J', 'P', 'r' },
+		{ 'O','S' },
+		{ 'O','f' },
+		{ 'O','m' },
+		{ 'O','o' },
+		{ 'O','t' },
+		{ 'P','T', 'l','r' },
+		{ 0 }
+	};
+	int excl_st[ARRAY_SIZE(excl)] = UL_EXCL_STATUS_INIT;
+
+	setlocale(LC_ALL, "");
+	bindtextdomain(PACKAGE, LOCALEDIR);
+	textdomain(PACKAGE);
+	close_stdout_atexit();
+
+	lsblk = &_ls;
+
+	lsblk_init_debug();
+
+	while((c = getopt_long(argc, argv,
+				"AabdDzE:e:fhJlNnMmo:OpPiI:rstVvST::w:x:y",
+				longopts, NULL)) != -1) {
+
+		err_exclusive_options(c, longopts, excl, excl_st);
+
+		switch(c) {
+		case 'A':
+			lsblk->noempty = 1;
+			break;
+		case 'a':
+			lsblk->all_devices = 1;
+			break;
+		case 'b':
+			lsblk->bytes = 1;
+			break;
+		case 'd':
+			lsblk->nodeps = 1;
+			break;
+		case 'D':
+			add_uniq_column(COL_NAME);
+			add_uniq_column(COL_DALIGN);
+			add_uniq_column(COL_DGRAN);
+			add_uniq_column(COL_DMAX);
+			add_uniq_column(COL_DZERO);
+			break;
+		case 'z':
+			add_uniq_column(COL_NAME);
+			add_uniq_column(COL_ZONED);
+			add_uniq_column(COL_ZONE_SZ);
+			add_uniq_column(COL_ZONE_NR);
+			add_uniq_column(COL_ZONE_AMAX);
+			add_uniq_column(COL_ZONE_OMAX);
+			add_uniq_column(COL_ZONE_APP);
+			add_uniq_column(COL_ZONE_WGRAN);
+			break;
+		case 'e':
+			parse_excludes(optarg);
+			break;
+		case 'J':
+			lsblk->flags |= LSBLK_JSON;
+			break;
+		case 'l':
+			lsblk->flags &= ~LSBLK_TREE; /* disable the default */
+			break;
+		case 'M':
+			lsblk->merge = 1;
+			break;
+		case 'n':
+			lsblk->flags |= LSBLK_NOHEADINGS;
+			break;
+		case 'o':
+			outarg = optarg;
+			break;
+		case 'O':
+			for (ncolumns = 0 ; ncolumns < ARRAY_SIZE(infos); ncolumns++)
+				columns[ncolumns] = ncolumns;
+			break;
+		case 'p':
+			lsblk->paths = 1;
+			break;
+		case 'P':
+			lsblk->flags |= LSBLK_EXPORT;
+			lsblk->flags &= ~LSBLK_TREE;	/* disable the default */
+			break;
+		case 'y':
+			lsblk->flags |= LSBLK_SHELLVAR;
+			break;
+		case 'i':
+			lsblk->flags |= LSBLK_ASCII;
+			break;
+		case 'I':
+			parse_includes(optarg);
+			break;
+		case 'r':
+			lsblk->flags &= ~LSBLK_TREE;	/* disable the default */
+			lsblk->flags |= LSBLK_RAW;		/* enable raw */
+			break;
+		case 's':
+			lsblk->inverse = 1;
+			break;
+		case 'f':
+			add_uniq_column(COL_NAME);
+			add_uniq_column(COL_FSTYPE);
+			add_uniq_column(COL_FSVERSION);
+			add_uniq_column(COL_LABEL);
+			add_uniq_column(COL_UUID);
+			add_uniq_column(COL_FSAVAIL);
+			add_uniq_column(COL_FSUSEPERC);
+			add_uniq_column(COL_TARGETS);
+			break;
+		case 'm':
+			add_uniq_column(COL_NAME);
+			add_uniq_column(COL_SIZE);
+			add_uniq_column(COL_OWNER);
+			add_uniq_column(COL_GROUP);
+			add_uniq_column(COL_MODE);
+			break;
+		case 't':
+			add_uniq_column(COL_NAME);
+			add_uniq_column(COL_ALIOFF);
+			add_uniq_column(COL_MINIO);
+			add_uniq_column(COL_OPTIO);
+			add_uniq_column(COL_PHYSEC);
+			add_uniq_column(COL_LOGSEC);
+			add_uniq_column(COL_ROTA);
+			add_uniq_column(COL_SCHED);
+			add_uniq_column(COL_RQ_SIZE);
+			add_uniq_column(COL_RA);
+			add_uniq_column(COL_WSAME);
+			break;
+		case 'S':
+			lsblk->nodeps = 1;
+			lsblk->scsi = 1;
+			add_uniq_column(COL_NAME);
+			add_uniq_column(COL_HCTL);
+			add_uniq_column(COL_TYPE);
+			add_uniq_column(COL_VENDOR);
+			add_uniq_column(COL_MODEL);
+			add_uniq_column(COL_REV);
+			add_uniq_column(COL_SERIAL);
+			add_uniq_column(COL_TRANSPORT);
+			break;
+		case 'N':
+			lsblk->nodeps = 1;
+			lsblk->nvme = 1;
+			add_uniq_column(COL_NAME);
+			add_uniq_column(COL_TYPE);
+			add_uniq_column(COL_MODEL);
+			add_uniq_column(COL_SERIAL);
+			add_uniq_column(COL_REV);
+			add_uniq_column(COL_TRANSPORT);
+			add_uniq_column(COL_RQ_SIZE);
+			add_uniq_column(COL_MQ);
+			break;
+		case 'v':
+			lsblk->nodeps = 1;
+			lsblk->virtio = 1;
+			add_uniq_column(COL_NAME);
+			add_uniq_column(COL_TYPE);
+			add_uniq_column(COL_TRANSPORT);
+			add_uniq_column(COL_SIZE);
+			add_uniq_column(COL_RQ_SIZE);
+			add_uniq_column(COL_MQ);
+			break;
+		case 'T':
+			force_tree = 1;
+			if (optarg) {
+				if (*optarg == '=')
+					optarg++;
+				lsblk->tree_id = column_name_to_id(optarg, strlen(optarg));
+			}
+			break;
+		case OPT_SYSROOT:
+			lsblk->sysroot = optarg;
+			break;
+		case 'E':
+			lsblk->dedup_id = column_name_to_id(optarg, strlen(optarg));
+			if (lsblk->dedup_id >= 0)
+				break;
+			errtryhelp(EXIT_FAILURE);
+			break;
+		case 'w':
+			width = strtou32_or_err(optarg, _("invalid output width number argument"));
+			break;
+		case 'x':
+			lsblk->flags &= ~LSBLK_TREE; /* disable the default */
+			lsblk->sort_id = column_name_to_id(optarg, strlen(optarg));
+			if (lsblk->sort_id >= 0)
+				break;
+			errtryhelp(EXIT_FAILURE);
+			break;
+
+		case 'h':
+			usage();
+		case 'V':
+			print_version(EXIT_SUCCESS);
+		default:
+			errtryhelp(EXIT_FAILURE);
+		}
+	}
+
+	if (force_tree)
+		lsblk->flags |= LSBLK_TREE;
+
+	check_sysdevblock();
+
+	if (!ncolumns) {
+		add_column(COL_NAME);
+		add_column(COL_MAJMIN);
+		add_column(COL_RM);
+		add_column(COL_SIZE);
+		add_column(COL_RO);
+		add_column(COL_TYPE);
+		add_column(COL_TARGETS);
+	}
+
+	if (outarg && string_add_to_idarray(outarg, columns, ARRAY_SIZE(columns),
+					 &ncolumns, column_name_to_id) < 0)
+		return EXIT_FAILURE;
+
+	if (lsblk->all_devices == 0 && nexcludes == 0 && nincludes == 0)
+		excludes[nexcludes++] = 1;	/* default: ignore RAM disks */
+
+	if (lsblk->sort_id < 0)
+		/* Since Linux 4.8 we have sort devices by default, because
+		 * /sys is no more sorted */
+		lsblk->sort_id = COL_MAJMIN;
+
+	/* For --{inverse,raw,pairs} --list we still follow parent->child relation */
+	if (!(lsblk->flags & LSBLK_TREE)
+	    && (lsblk->inverse || lsblk->flags & LSBLK_EXPORT || lsblk->flags & LSBLK_RAW))
+		lsblk->force_tree_order = 1;
+
+	if (lsblk->sort_id >= 0 && column_id_to_number(lsblk->sort_id) < 0) {
+		/* the sort column is not between output columns -- add as hidden */
+		add_column(lsblk->sort_id);
+		lsblk->sort_hidden = 1;
+	}
+
+	if (lsblk->dedup_id >= 0 && column_id_to_number(lsblk->dedup_id) < 0) {
+		/* the deduplication column is not between output columns -- add as hidden */
+		add_column(lsblk->dedup_id);
+		lsblk->dedup_hidden = 1;
+	}
+
+	lsblk_mnt_init();
+	scols_init_debug(0);
+	ul_path_init_debug();
+
+	/*
+	 * initialize output columns
+	 */
+	if (!(lsblk->table = scols_new_table()))
+		errx(EXIT_FAILURE, _("failed to allocate output table"));
+	scols_table_enable_raw(lsblk->table, !!(lsblk->flags & LSBLK_RAW));
+	scols_table_enable_export(lsblk->table, !!(lsblk->flags & LSBLK_EXPORT));
+	scols_table_enable_shellvar(lsblk->table, !!(lsblk->flags & LSBLK_SHELLVAR));
+	scols_table_enable_ascii(lsblk->table, !!(lsblk->flags & LSBLK_ASCII));
+	scols_table_enable_json(lsblk->table, !!(lsblk->flags & LSBLK_JSON));
+	scols_table_enable_noheadings(lsblk->table, !!(lsblk->flags & LSBLK_NOHEADINGS));
+
+	if (lsblk->flags & LSBLK_JSON)
+		scols_table_set_name(lsblk->table, "blockdevices");
+	if (width) {
+		scols_table_set_termwidth(lsblk->table, width);
+		scols_table_set_termforce(lsblk->table, SCOLS_TERMFORCE_ALWAYS);
+	}
+
+	for (i = 0; i < ncolumns; i++) {
+		struct colinfo *ci = get_column_info(i);
+		struct libscols_column *cl;
+		int id = get_column_id(i), fl = ci->flags;
+
+		if ((lsblk->flags & LSBLK_TREE)
+		    && has_tree_col == 0
+		    && id == lsblk->tree_id) {
+			fl |= SCOLS_FL_TREE;
+			fl &= ~SCOLS_FL_RIGHT;
+			has_tree_col = 1;
+		}
+
+		if (lsblk->sort_hidden && lsblk->sort_id == id)
+			fl |= SCOLS_FL_HIDDEN;
+		if (lsblk->dedup_hidden && lsblk->dedup_id == id)
+			fl |= SCOLS_FL_HIDDEN;
+
+		if (force_tree
+		    && lsblk->flags & LSBLK_JSON
+		    && has_tree_col == 0
+		    && i + 1 == ncolumns)
+			/* The "--tree --json" specified, but no column with
+			 * SCOLS_FL_TREE yet; force it for the last column
+			 */
+			fl |= SCOLS_FL_TREE;
+
+		cl = scols_table_new_column(lsblk->table, ci->name, ci->whint, fl);
+		if (!cl) {
+			warn(_("failed to allocate output column"));
+			goto leave;
+		}
+		if (!lsblk->sort_col && lsblk->sort_id == id) {
+			lsblk->sort_col = cl;
+			scols_column_set_cmpfunc(cl,
+				ci->type == COLTYPE_NUM     ? cmp_u64_cells :
+				ci->type == COLTYPE_SIZE    ? cmp_u64_cells :
+			        ci->type == COLTYPE_SORTNUM ? cmp_u64_cells : scols_cmpstr_cells,
+				NULL);
+		}
+		/* multi-line cells (now used for MOUNTPOINTS) */
+		if (fl & SCOLS_FL_WRAP) {
+			scols_column_set_wrapfunc(cl,
+						scols_wrapnl_chunksize,
+						scols_wrapnl_nextchunk,
+						NULL);
+			scols_column_set_safechars(cl, "\n");
+		}
+
+		if (lsblk->flags & LSBLK_JSON) {
+			switch (ci->type) {
+			case COLTYPE_SIZE:
+				if (!lsblk->bytes)
+					break;
+				/* fallthrough */
+			case COLTYPE_NUM:
+				scols_column_set_json_type(cl, SCOLS_JSON_NUMBER);
+				break;
+			case COLTYPE_BOOL:
+				scols_column_set_json_type(cl, SCOLS_JSON_BOOLEAN);
+				break;
+			default:
+				if (fl & SCOLS_FL_WRAP)
+					scols_column_set_json_type(cl, SCOLS_JSON_ARRAY_STRING);
+				else
+					scols_column_set_json_type(cl, SCOLS_JSON_STRING);
+				break;
+			}
+		}
+	}
+
+	tr = lsblk_new_devtree();
+	if (!tr)
+		err(EXIT_FAILURE, _("failed to allocate device tree"));
+
+	if (optind == argc) {
+		int rc = lsblk->inverse ?
+			process_all_devices_inverse(tr) :
+			process_all_devices(tr);
+
+		status = rc == 0 ? EXIT_SUCCESS : EXIT_FAILURE;
+	} else {
+		int cnt = 0, cnt_err = 0;
+
+		while (optind < argc) {
+			if (process_one_device(tr, argv[optind++]) != 0)
+				cnt_err++;
+			cnt++;
+		}
+		status = cnt == 0	? EXIT_FAILURE :	/* nothing */
+			 cnt == cnt_err	? LSBLK_EXIT_ALLFAILED :/* all failed */
+			 cnt_err	? LSBLK_EXIT_SOMEOK :	/* some ok */
+					  EXIT_SUCCESS;		/* all success */
+	}
+
+	if (lsblk->dedup_id > -1) {
+		devtree_set_dedupkeys(tr, lsblk->dedup_id);
+		lsblk_devtree_deduplicate_devices(tr);
+	}
+
+	devtree_to_scols(tr, lsblk->table);
+
+	if (lsblk->sort_col)
+		scols_sort_table(lsblk->table, lsblk->sort_col);
+	if (lsblk->force_tree_order)
+		scols_sort_table_by_tree(lsblk->table);
+
+	scols_print_table(lsblk->table);
+
+leave:
+	if (lsblk->sort_col)
+		unref_sortdata(lsblk->table);
+
+	scols_unref_table(lsblk->table);
+
+	lsblk_mnt_deinit();
+	lsblk_properties_deinit();
+	lsblk_unref_devtree(tr);
+
+	return status;
+}