util-linux is a random collection of Linux utilities

A CD (DVD+R), formatted as udf file system. Burn the disc under the windows system for the first time, and the disc can be mounted normally under linux.

Then burn the disc for the second time under windows (additional burn). At this time, the disc cannot be automatically mounted, and it can be used after manual mounting.

Note that by changing the version of libblkid1, it is found that this problem does not exist in versions prior to 2.31.1-0.5 (including this version).

You can test it through the Ubuntu system.

The following is the blkid command information of the CD (Disc burned twice in the windos system) in the linux system.

libblkid1_2.31.1-0.5：
uos-PC:~$ blkid 
……
/dev/sr0: UUID="0930121120554446" LABEL="kelu2" TYPE="udf"
/dev/sdb1: LABEL_FATBOOT="UKNOWN" LABEL="UKNOWN" TYPE="vfat"

uos-PC:~$ blkid -o list
device        fs_type       label      mount                              point UUID
……   
/dev/sr0   udf               kelu2    /media/uos/disk               0930121120554446

libblkid1_2.32-0.1：
uos-PC:~$ blkid
……
/dev/sr0: UUID="0930121120554446" LABEL="kelu2" TYPE="udf"

uos@uos-PC:~$ blkid -o list
device          fs_type      label              mount point                           UUID
……
/dev/sr0       udf            kelu2           (not mounted)               0930121120554446

As you noticed at https://github.com/lsof-org/lsof/issues/89, I have developed lsfd. lsfd is a command like lsof. The critical differences are (1) lsfd is specialized to Linux, and (2) lsfd is based on libsmartcols.

This pull request is for getting comments. Socket-related information is not extracted yet. The filtering feature that lsof is not implemented yet. Test cases are not available yet. The man page is not available yet. They are parts of future work.

A quirk in this pull request is about POSIX threads. I assumed using multiple threads may improve the performance of listing. After implementing and measuring the execution time, I found the performance is not so improved as I expected. So I deleted the code for multi-threading.

Anyway, I made so many commits. I would like to get comments before going ahead more.

$ sudo ./lsfd | grep -v mem | head -30
COMMAND                                PID            USER  ASSOC MODE TYPE              DEVNAME MNTID      INODE NAME
systemd                                  1            root    exe  ---  REG                 dm-0     0  201398717 /usr/lib/systemd/systemd
systemd                                  1            root    cwd  ---  DIR                 dm-0     0         96 /
systemd                                  1            root    rtd  ---  DIR                 dm-0     0         96 /
systemd                                  1            root cgroup  ---  REG                  0:4     0 4026531835 cgroup:[4026531835]
systemd                                  1            root    ipc  ---  REG                  0:4     0 4026531839 ipc:[4026531839]
systemd                                  1            root    mnt  ---  REG                  0:4     0 4026531840 mnt:[4026531840]
systemd                                  1            root    net  ---  REG                  0:4     0 4026532000 net:[4026532000]
systemd                                  1            root    pid  ---  REG                  0:4     0 4026531836 pid:[4026531836]
systemd                                  1            root  pid4c  ---  REG                  0:4     0 4026531836 pid:[4026531836]
systemd                                  1            root   time  ---  REG                  0:4     0 4026531834 time:[4026531834]
systemd                                  1            root time4c  ---  REG                  0:4     0 4026531834 time:[4026531834]
systemd                                  1            root   user  ---  REG                  0:4     0 4026531837 user:[4026531837]
systemd                                  1            root    uts  ---  REG                  0:4     0 4026531838 uts:[4026531838]
...

I got the following error when I run make on Mac OS X.

Does anybody know what is wrong?

$ make sys-utils/flock
gcc -g -O2    sys-utils/flock.c   -o sys-utils/flock
sys-utils/flock.c:42:10: fatal error: 'c.h' file not found
#include "c.h"
         ^~~~~
1 error generated.
make: *** [sys-utils/flock] Error 1

$ make
...
  CC       sys-utils/ipcrm.o
sys-utils/ipcrm.c:246:17: error: variable has incomplete type 'struct seminfo'
        struct seminfo seminfo;
                       ^
sys-utils/ipcrm.c:246:9: note: forward declaration of 'struct seminfo'
        struct seminfo seminfo;
               ^
sys-utils/ipcrm.c:253:21: error: use of undeclared identifier 'SHM_INFO'
                maxid = shmctl(0, SHM_INFO, &shmseg);
                                  ^
sys-utils/ipcrm.c:258:23: error: use of undeclared identifier 'SHM_STAT'
                        rm_me = shmctl(id, SHM_STAT, &shmseg);
                                           ^
sys-utils/ipcrm.c:266:24: error: use of undeclared identifier 'SEM_INFO'
                maxid = semctl(0, 0, SEM_INFO, arg);
                                     ^
sys-utils/ipcrm.c:272:26: error: use of undeclared identifier 'SEM_STAT'
                        rm_me = semctl(id, 0, SEM_STAT, arg);
                                              ^
5 errors generated.
make[2]: *** [sys-utils/ipcrm.o] Error 1
make[1]: *** [all-recursive] Error 1
make: *** [all] Error 2

Until recently, libmount was a low-level library with few dependencies, that could be safely depended on by things like GLib.

The addition of the optional libcryptsetup dependency changes that somewhat. There seem to be two main problem areas causing crashes in distributions that have enabled it, such as recent development versions of Debian and Ubuntu:

• The json-c library contains symbols that collide with other JSON libraries such as libjansson (akheron/jansson#523, json-c/json-c#621) and json-glib (json-glib#33), because it seems every implementation of JSON in C has chosen to use the json_ namespace. This is particularly troublesome for anything that depends on json-glib, which depends on GLib, which depends on libmount, which depends on libcryptsetup, which depends on json-c, almost guaranteeing symbol conflicts.

• libcryptsetup's default/recommended OpenSSL backend for crypto has symbol conflicts with anything that statically links a different version of OpenSSL, which seems to be popular in the proprietary software world (Steam and Minecraft are known to be affected). ValveSoftware/steam-for-linux#7223 and ValveSoftware/steam-for-linux#6861 are examples of this.

I don't know which project(s) should take action to solve this, but I'm trying to make sure all the upstreams involved in this collision are at least aware that there's a problem, so that whatever action is taken can be somewhat coordinated.

One way that this might be solved within libmount (if that is considered to be a desirable place to solve it) would be to move the verity feature into a helper executable, or perhaps into a dlopened module that is loaded with RTLD_LOCAL to avoid it interfering with symbol resolution in the flat default ELF namespace.

Another possible solution, if the verity feature is mostly of interest to embedded or special-purpose distributions, would be to recommend that general-purpose distributions disable that feature.

This addresses some feedback on https://github.com/karelzak/util-linux/pull/872 by ensuring (rather than assuming) that a partition table entry corresponds to the ISO9660 filesystem, and by only applying the logic to restrict the metadata to a partition device in cases where there is actually a device (which does not happen for CD/DVD /dev/sr0).

As noted, the udev behaviour here is also questionable and indeed if it behaved differently I might not have put attention into modifying blkid. However, I think the information presented by blkid in the case of a combined ISO & partition setup (where a partition points at the ISO data too) is questionable, where it effectively reports the same filesystem on two devices, so I think this approach still makes sense for blkid. Separately I'll take another look at the udev side, but probably not until next week.

Let's run at least the unit-tests on Fedora Rawhide via Packit, as it supports a couple of alternative architectures (aarch64, ppc64le, s390x) as well as i*86, which were source of issues in the past.

Good morning,

I have installed an RTC module with and when trying to read the time it gives me this error.

hwclock -r -v --rtc /dev/rtc1 hwclock from util-linux 2.37.154-c2ca2 System Time: 1625749015.843650 Using the rtc interface to the clock. Last drift adjustment done at 1625749005 seconds after 1969 Last calibration done at 1625749005 seconds after 1969 Hardware clock is on UTC time Assuming hardware clock is kept in UTC time. Waiting for clock tick... ioctl(3, RTC_UIE_ON, 0): Invalid argument Waiting in loop for time from /dev/rtc1 to change hwclock: Timed out waiting for time change. ...synchronization failed

Any idea?

Thanks!

It seems you have merged the Fedora version of hardlink into util-linux. This is unfortunate, as Debian, Ubuntu, Arch, and others are shipping a different hardlink, developed my me, that has a lot more features:

Usage: ./hardlink [options] directory|file ...
Options:
  -V, --version         show program's version number and exit
  -h, --help            show this help message and exit
  -v, --verbose         Increase verbosity (repeat for more verbosity)
  -n, --dry-run         Modify nothing, just print what would happen
  -f, --respect-name    Filenames have to be identical
  -p, --ignore-mode     Ignore changes of file mode
  -o, --ignore-owner    Ignore owner changes
  -t, --ignore-time     Ignore timestamps (when testing for equality)
  -X, --respect-xattrs  Respect extended attributes
  -m, --maximize        Maximize the hardlink count, remove the file with
                        lowest hardlink cout
  -M, --minimize        Reverse the meaning of -m
  -O, --keep-oldest     Keep the oldest file of multiple equal files
                        (lower precedence than minimize/maximize)
  -x REGEXP, --exclude=REGEXP
                        Regular expression to exclude files
  -i REGEXP, --include=REGEXP
                        Regular expression to include files/dirs
  -s <num>[K,M,G], --minimum-size=<num>[K,M,G]
                        Minimum size for files. Optional suffix
                        allows for using KiB, MiB, or GiB

Compatibility options to Jakub Jelinek's hardlink:
  -c                    Compare only file contents, same as -pot

Switching to this version of hardlink would break things for a lot of users. Apart from the meaning of -f, mine is a superset option-wise.

I'd argue given that my version is a superset of this version, we should probably consolidate on this one. I think I mentioned this to Jakub back in the day, but I'm not sure.

Recent OpenWrt install seg faults with all the parameter combinations I've tested (except -h and -V, they work fine). Reported on https://github.com/openwrt/openwrt/issues/10737, where it's felt to be an upstream issue.

$ cat /etc/os-release
NAME="OpenWrt"
VERSION="SNAPSHOT"
ID="openwrt"
ID_LIKE="lede openwrt"
PRETTY_NAME="OpenWrt SNAPSHOT"
VERSION_ID="snapshot"
HOME_URL="https://openwrt.org/"
BUG_URL="https://bugs.openwrt.org/"
SUPPORT_URL="https://forum.openwrt.org/"
BUILD_ID="r20614-c27279dc26"
OPENWRT_BOARD="mediatek/mt7622"
OPENWRT_ARCH="aarch64_cortex-a53"
OPENWRT_TAINTS=""
OPENWRT_DEVICE_MANUFACTURER="OpenWrt"
OPENWRT_DEVICE_MANUFACTURER_URL="https://openwrt.org/"
OPENWRT_DEVICE_PRODUCT="Generic"
OPENWRT_DEVICE_REVISION="v0"
OPENWRT_RELEASE="OpenWrt SNAPSHOT r20614-c27279dc26"

$ opkg install lscpu
Package lscpu (2.38-2) installed in root is up to date.

$ lscpu -V
lscpu from util-linux 2.38

$ lscpu
Segmentation fault

$ strace lscpu
execve("/usr/bin/lscpu", ["lscpu"], 0x7fe684d110 /* 15 vars */) = 0
set_tid_address(0x7f8f939248)           = 6160
brk(NULL)                               = 0x2980f000
brk(0x29811000)                         = 0x29811000
mmap(0x2980f000, 4096, PROT_NONE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x2980f000
openat(AT_FDCWD, "/etc/ld-musl-aarch64.path", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD, "/lib/libsmartcols.so.1", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD, "/usr/local/lib/libsmartcols.so.1", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD, "/usr/lib/libsmartcols.so.1", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 3
fcntl(3, F_SETFD, FD_CLOEXEC)           = 0
fstat(3, {st_mode=S_IFREG|0755, st_size=98307, ...}) = 0
read(3, "\177ELF\2\1\1\0\0\0\0\0\0\0\0\0\3\0\267\0\1\0\0\0`R\0\0\0\0\0\0"..., 960) = 960
mmap(NULL, 167936, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x7f8f879000
mmap(0x7f8f8a0000, 8192, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3, 0x17000) = 0x7f8f8a0000
close(3)                                = 0
openat(AT_FDCWD, "/lib/libgcc_s.so.1", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 3
fcntl(3, F_SETFD, FD_CLOEXEC)           = 0
fstat(3, {st_mode=S_IFREG|0644, st_size=73744, ...}) = 0
read(3, "\177ELF\2\1\1\0\0\0\0\0\0\0\0\0\3\0\267\0\1\0\0\0\0-\0\0\0\0\0\0"..., 960) = 960
mmap(NULL, 143360, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x7f8f856000
mmap(0x7f8f877000, 8192, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3, 0x11000) = 0x7f8f877000
close(3)                                = 0
mprotect(0x7f8f8a0000, 4096, PROT_READ) = 0
mprotect(0x7f8f877000, 4096, PROT_READ) = 0
mprotect(0x428000, 4096, PROT_READ)     = 0
ioctl(1, TIOCGWINSZ, {ws_row=60, ws_col=200, ws_xpixel=2200, ws_ypixel=1200}) = 0
mmap(NULL, 16384, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f8f92e000
openat(AT_FDCWD, "/sys/devices/system/cpu", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 3
fcntl(3, F_SETFD, FD_CLOEXEC)           = 0
openat(3, "kernel_max", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 4
mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f8f92d000
fcntl(4, F_SETFD, FD_CLOEXEC)           = 0
read(4, "1\n", 1024)                    = 2
lseek(4, -1, SEEK_CUR)                  = 1
close(4)                                = 0
munmap(0x7f8f92d000, 4096)              = 0
openat(3, "possible", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 4
mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f8f92d000
fcntl(4, F_SETFD, FD_CLOEXEC)           = 0
read(4, "0-1\n", 1024)                  = 4
close(4)                                = 0
munmap(0x7f8f92d000, 4096)              = 0
openat(3, "present", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 4
mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f8f92d000
fcntl(4, F_SETFD, FD_CLOEXEC)           = 0
read(4, "0-1\n", 1024)                  = 4
close(4)                                = 0
munmap(0x7f8f92d000, 4096)              = 0
openat(3, "online", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 4
mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f8f92d000
fcntl(4, F_SETFD, FD_CLOEXEC)           = 0
read(4, "0-1\n", 1024)                  = 4
close(4)                                = 0
munmap(0x7f8f92d000, 4096)              = 0
openat(AT_FDCWD, "/proc", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 4
fcntl(4, F_SETFD, FD_CLOEXEC)           = 0
openat(4, "cpuinfo", O_RDONLY|O_LARGEFILE) = 5
mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f8f92d000
read(5, "processor\t: 0\nBogoMIPS\t: 25.00\nF"..., 1024) = 374
read(5, "", 1024)                       = 0
close(5)                                = 0
munmap(0x7f8f92d000, 4096)              = 0
uname({sysname="Linux", nodename="panther", ...}) = 0
personality(PER_LINUX32)                = 0 (PER_LINUX)
personality(PER_LINUX)                  = 0x8 (PER_LINUX32)
openat(3, "dispatching", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(3, "cpufreq/boost", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(4, "sysinfo", O_RDONLY|O_LARGEFILE) = -1 ENOENT (No such file or directory)
openat(3, "vulnerabilities", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 5
fstat(5, {st_mode=S_IFDIR|0755, st_size=0, ...}) = 0
fcntl(5, F_GETFL)                       = 0x20000 (flags O_RDONLY|O_LARGEFILE)
mmap(NULL, 8192, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f8f92c000
fcntl(5, F_SETFD, FD_CLOEXEC)           = 0
getdents64(5, 0x7f8f92c098 /* 13 entries */, 2048) = 416
getdents64(5, 0x7f8f92c098 /* 0 entries */, 2048) = 0
lseek(5, 0, SEEK_SET)                   = 0
getdents64(5, 0x7f8f92c098 /* 13 entries */, 2048) = 416
openat(3, "vulnerabilities/spectre_v2", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 6
read(6, "Not affected\n", 1023)         = 13
read(6, "", 1010)                       = 0
close(6)                                = 0
openat(3, "vulnerabilities/itlb_multihit", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 6
read(6, "Not affected\n", 1023)         = 13
read(6, "", 1010)                       = 0
close(6)                                = 0
openat(3, "vulnerabilities/mmio_stale_data", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 6
read(6, "Not affected\n", 1023)         = 13
read(6, "", 1010)                       = 0
close(6)                                = 0
openat(3, "vulnerabilities/mds", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 6
read(6, "Not affected\n", 1023)         = 13
read(6, "", 1010)                       = 0
close(6)                                = 0
openat(3, "vulnerabilities/l1tf", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 6
read(6, "Not affected\n", 1023)         = 13
read(6, "", 1010)                       = 0
close(6)                                = 0
openat(3, "vulnerabilities/spec_store_bypass", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 6
read(6, "Not affected\n", 1023)         = 13
read(6, "", 1010)                       = 0
close(6)                                = 0
openat(3, "vulnerabilities/tsx_async_abort", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 6
read(6, "Not affected\n", 1023)         = 13
read(6, "", 1010)                       = 0
close(6)                                = 0
openat(3, "vulnerabilities/spectre_v1", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 6
read(6, "Mitigation: __user pointer sanit"..., 1023) = 40
read(6, "", 983)                        = 0
close(6)                                = 0
openat(3, "vulnerabilities/retbleed", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 6
read(6, "Not affected\n", 1023)         = 13
read(6, "", 1010)                       = 0
close(6)                                = 0
openat(3, "vulnerabilities/srbds", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 6
read(6, "Not affected\n", 1023)         = 13
read(6, "", 1010)                       = 0
close(6)                                = 0
openat(3, "vulnerabilities/meltdown", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 6
read(6, "Not affected\n", 1023)         = 13
read(6, "", 1010)                       = 0
close(6)                                = 0
close(5)                                = 0
munmap(0x7f8f92c000, 8192)              = 0
openat(AT_FDCWD, "/sys/devices/system/node", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = -1 ENOENT (No such file or directory)
faccessat(3, "cpu0/topology/thread_siblings", F_OK) = 0
openat(3, "cpu0/topology/thread_siblings", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 5
mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f8f92d000
fcntl(5, F_SETFD, FD_CLOEXEC)           = 0
read(5, "", 1024)                       = 0
close(5)                                = 0
munmap(0x7f8f92d000, 4096)              = 0
openat(3, "cpu0/topology/core_siblings", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 5
mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f8f92d000
fcntl(5, F_SETFD, FD_CLOEXEC)           = 0
read(5, "", 1024)                       = 0
close(5)                                = 0
munmap(0x7f8f92d000, 4096)              = 0
openat(3, "cpu0/topology/book_siblings", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(3, "cpu0/topology/drawer_siblings", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = -1 ENOENT (No such file or directory)
faccessat(3, "cpu1/topology/thread_siblings", F_OK) = 0
openat(3, "cpu1/topology/thread_siblings", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 5
mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f8f92d000
fcntl(5, F_SETFD, FD_CLOEXEC)           = 0
read(5, "", 1024)                       = 0
close(5)                                = 0
munmap(0x7f8f92d000, 4096)              = 0
openat(3, "cpu1/topology/core_siblings", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = 5
mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f8f92d000
fcntl(5, F_SETFD, FD_CLOEXEC)           = 0
read(5, "", 1024)                       = 0
close(5)                                = 0
munmap(0x7f8f92d000, 4096)              = 0
openat(3, "cpu1/topology/book_siblings", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(3, "cpu1/topology/drawer_siblings", O_RDONLY|O_LARGEFILE|O_CLOEXEC) = -1 ENOENT (No such file or directory)
--- SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_MAPERR, si_addr=0x10} ---
+++ killed by SIGSEGV +++
Segmentation fault

lscpu version 2.32 reports the CPU has 8 MiB cache, while the correct value is 16 MiB:

$ lscpu
Model name:          AMD Ryzen 5 1600 Six-Core Processor
L3 cache:            8192K

/sys (Linux 4.17.0) provides the correct cache IDs:

$ cat /sys/devices/system/cpu/cpu0/cache/index3/id
0
$ cat /sys/devices/system/cpu/cpu6/cache/index3/id
1
$ cat /sys/devices/system/cpu/cpu0/cache/index3/size
8192K

It would be nice if lscpu printed the number of L1/L2/L3 caches, such as:

L1d cache: 6*32K
L1i cache: 6*64K
L2 cache:  6*512K
L3 cache:  2*8192K

For comparison, lstopo version 1.11.2 (http://www.open-mpi.org/projects/hwloc/) works OK.

Hello!

@masatake

sparc64 host:

tests$ git describe
v2.37-648-g1b44d0923

tests$ ./run.sh --show-diff lsfd

-------------------- util-linux regression tests --------------------

                    For development purpose only.
                 Don't execute on production system!

       kernel: 5.16.0-rc2

      options: --show-diff \
               --srcdir=/1/mator/util-linux/tests/.. \
               --builddir=/1/mator/util-linux/tests/..

         lsfd: directory                      ... SKIPPED (no root permissions)
         lsfd: directory                      ... OK
         lsfd: pipe, no fork                  ... OK
         lsfd: read-only regular file         ...
diff-{{{
--- /1/mator/util-linux/tests/expected/lsfd/mkfds-ro-regular-file       2021-11-26 16:29:39.397686937 +0300
+++ /1/mator/util-linux/tests/output/lsfd/mkfds-ro-regular-file 2021-11-26 17:57:41.055617042 +0300
@@ -13,4 +13,4 @@
 MNTID[RUN]: 0
 DEV[RUN]: 0
 FINDMNT[RUN]: 0
-DEV[STR]: 0
+DEV[STR]: 1
}}}-diff

 FAILED (lsfd/mkfds-ro-regular-file)
         lsfd: character device with O_RDWR   ... OK
         lsfd: AF_UNIX socket pair created with socketpair(2) ...
diff-{{{
--- /1/mator/util-linux/tests/expected/lsfd/mkfds-socketpair    2021-11-26 16:29:39.397686937 +0300
+++ /1/mator/util-linux/tests/output/lsfd/mkfds-socketpair      2021-11-26 17:57:41.831650266 +0300
@@ -1,3 +1,3 @@
-    3  rw- SOCK sockfs      UNIX
-    4  rw- SOCK sockfs      UNIX
+    3  rw- SOCK sockfs UNIX-STREAM
+    4  rw- SOCK sockfs UNIX-STREAM
 ASSOC,MODE,TYPE,SOURCE,PROTONAME: 0
}}}-diff

 FAILED (lsfd/mkfds-socketpair)
         lsfd: symbolic link itself opened with O_PATH ... OK
         lsfd: directory                      ... SKIPPED (no root permissions)
         lsfd: directory                      ... OK

---------------------------------------------------------------------
  2 tests of 9 FAILED
---------------------------------------------------------------------

ppc64 host:

tests$ ./run.sh --show-diff lsfd

-------------------- util-linux regression tests --------------------

                    For development purpose only.
                 Don't execute on production system!

       kernel: 5.15.0-1-powerpc64

      options: --show-diff \
               --srcdir=/home/mator/util-linux.git/tests/.. \
               --builddir=/home/mator/util-linux.git/tests/..

         lsfd: directory                      ... SKIPPED (no root permissions)
         lsfd: directory                      ... OK
         lsfd: pipe, no fork                  ... OK
         lsfd: read-only regular file         ... OK
         lsfd: character device with O_RDWR   ...
diff-{{{
--- /home/mator/util-linux.git/tests/expected/lsfd/mkfds-rw-character-device    2021-11-26 18:08:03.405921213 +0300
+++ /home/mator/util-linux.git/tests/output/lsfd/mkfds-rw-character-device      2021-11-26 19:08:30.724078893 +0300
@@ -3,4 +3,4 @@
 MNTID[RUN]: 0
 DEV[RUN]: 0
 FINDMNT[RUN]: 0
-DEV[STR]: 0
+DEV[STR]: 1
}}}-diff

 FAILED (lsfd/mkfds-rw-character-device)
         lsfd: AF_UNIX socket pair created with socketpair(2) ...
diff-{{{
--- /home/mator/util-linux.git/tests/expected/lsfd/mkfds-socketpair     2021-11-26 18:08:03.405921213 +0300
+++ /home/mator/util-linux.git/tests/output/lsfd/mkfds-socketpair       2021-11-26 19:08:30.868079286 +0300
@@ -1,3 +1,3 @@
-    3  rw- SOCK sockfs      UNIX
-    4  rw- SOCK sockfs      UNIX
+    3  rw- SOCK sockfs UNIX-STREAM
+    4  rw- SOCK sockfs UNIX-STREAM
 ASSOC,MODE,TYPE,SOURCE,PROTONAME: 0
}}}-diff

 FAILED (lsfd/mkfds-socketpair)
         lsfd: symbolic link itself opened with O_PATH ... OK
         lsfd: directory                      ... SKIPPED (no root permissions)
         lsfd: directory                      ... OK

---------------------------------------------------------------------
  2 tests of 9 FAILED
---------------------------------------------------------------------

When we create a new namespace in a child process to persist it we need to ensure that the child hangs around. During exit the child will drop all references to its namespaces and so by the time we call open we might already fail to open the namespace. Fix this.

Signed-off-by: Christian Brauner (Microsoft) [email protected]

Provide some minor improvements to the existing taskset man page, namely:

-- Illustrate the ability to specify cpumasks without a leading "0x" with an example that is clearly hexadecimal.

-- Clarify that a successful return code from the command does not mean that a process has actually migrated.

-- Provide a couple of examples to further explain the significance of the taskset return code.

I see this link from " man umont ". I am not reporting any bugs. I tried only ask.

I've been a Linux user for many years and I don't always see a change for the better . Maybe it's just a problem with my Linux distribution, so I tried ask. For now I tried test Ubuntu 22.04.1 LTS (Jammy Jellyfish) MATE 1.26.0

1. I don't see GUI which will work from Desktop or file manager.

• For to change disk partition permissions.

Fedora created gnome-disks , but it's a separate app and when I used it , the path was changed every time, when I booted the system. I need constant path, so for now I'm using /etc/fstab ( and path /media )

3. I see snap in mount point where there should only be hard drive partitions.

$ lsblk
NAME                         MAJ:MIN RM   SIZE RO TYPE MOUNTPOINTS
loop0                          7:0    0     4K  1 loop /snap/bare/5
loop1                          7:1    0 146,6M  1 loop /snap/chromium/2238
loop2                          7:2    0 146,9M  1 loop /snap/chromium/2254
loop3                          7:3    0  63,2M  1 loop /snap/core20/1738
loop4                          7:4    0  63,3M  1 loop /snap/core20/1778
loop5                          7:5    0  55,1M  1 loop /snap/cups/836
loop6                          7:6    0    55M  1 loop /snap/cups/872

4. I can't uninstall all partitions with one command. I used before something like umount -all

Today, after this command is only partially functional.

• I can't start the terminal
• the partition is still unmounted
• when I tried to restart the system, the system showed red errors on a black background about snap.

5. Today, that may not be a problem. But I don't remember how to check which process is blocking mount access. Maybe there is some command that shows the processes using the lock. For example when I'm using GParted mount is locked / busy.

I wanted to ask how it looks in your Linux distribution ? Is there any chance of an idea of the specification in https://www.freedesktop.org/wiki/Specifications/

Do you see a chance for a better future now?

Python libmount fails if libmount was built with mountfd api.

    ctx = libmount.Context()
    ctx.fstype="overlay"
    ctx.source="overlay"
    lower = ":".join("{:x}".format(x) for x in range(int(yamldata["maxidx"])))
    ctx.options="lowerdir={},upperdir={},workdir={}".format(lower, yamldata["upper"], yamldata["work"])
    ctx.target=yamldata["target"]
    ctx.mount()

Traceback (most recent call last):
  File "/home/asavah/kross/kbs/overmount.py", line 61, in <module>
    main()
  File "/home/asavah/kross/kbs/overmount.py", line 51, in main
    __do_mount(options.mount)
  File "/home/asavah/kross/kbs/overmount.py", line 27, in __do_mount
    ctx.mount()
TypeError: Invalid argument

Building util-linux with --disable-libmount-mountfd-support works as intended.