On starting a system upgraded from Fedora 37 (working system) to 38 with the latest updates the numad.service fails to start with buffer overflow detected. This is preventing most of the KVMs from starting up. $ sudo systemctl status numad -l × numad.service - numad - The NUMA daemon that manages application locality. Loaded: loaded (/usr/lib/systemd/system/numad.service; enabled; preset: disabled) Drop-In: /usr/lib/systemd/system/service.d └─10-timeout-abort.conf Active: failed (Result: core-dump) since Sat 2023-05-27 17:59:18 BST; 37min ago Docs: man:numad Process: 3921 ExecStart=/usr/bin/numad -i 15 (code=dumped, signal=ABRT) CPU: 5ms May 27 17:59:18 kvm-125 systemd[1]: Starting numad.service - numad - The NUMA daemon that manages application locality.... May 27 17:59:18 kvm-125 numad[3921]: *** buffer overflow detected ***: terminated May 27 17:59:18 kvm-125 systemd[1]: numad.service: Control process exited, code=dumped, status=6/ABRT May 27 17:59:18 kvm-125 systemd[1]: numad.service: Failed with result 'core-dump'. May 27 17:59:18 kvm-125 systemd[1]: Failed to start numad.service - numad - The NUMA daemon that manages application locality.. $ $ uname -r 6.2.15-300.fc38.x86_64 $ rpm -qiv numad Name : numad Version : 0.5 Release : 38.20150602git.fc38 Architecture: x86_64 Install Date: Sat 20 May 2023 13:49:21 BST Group : Unspecified Size : 61824 License : LGPLv2 Signature : RSA/SHA256, Fri 20 Jan 2023 07:10:14 GMT, Key ID 809a8d7ceb10b464 Source RPM : numad-0.5-38.20150602git.fc38.src.rpm Build Date : Fri 20 Jan 2023 00:36:52 GMT Build Host : buildvm-x86-20.iad2.fedoraproject.org Packager : Fedora Project Vendor : Fedora Project URL : https://pagure.io/numad Bug URL : https://bugz.fedoraproject.org/numad Summary : NUMA user daemon Description : Numad, a daemon for NUMA (Non-Uniform Memory Architecture) systems, that monitors NUMA characteristics and manages placement of processes and memory to minimize memory latency and thus provide optimum performance. $ $ sudo numactl -H available: 1 nodes (0) node 0 cpus: 0 1 2 3 node 0 size: 3901 MB node 0 free: 2852 MB node distances: node 0 0: 10 $ I tried to disable SELinux just in case that was interferring but same result: $ getenforce Enforcing $ sudo setenforce 0 $ getenforce Permissive $ $ sudo numad -i 15 *** buffer overflow detected ***: terminated Aborted $ Installed numad debuginfo: $ sudo dnf debuginfo-install numad enabling fedora-debuginfo repository enabling fedora-cisco-openh264-debuginfo repository enabling fedora-modular-debuginfo repository enabling updates-debuginfo repository enabling updates-modular-debuginfo repository enabling rpmfusion-free-debuginfo repository enabling rpmfusion-free-updates-debuginfo repository enabling rpmfusion-nonfree-debuginfo repository enabling rpmfusion-nonfree-updates-debuginfo repository Fedora 38 - x86_64 - Debug 2.8 MB/s | 38 MB 00:13 Fedora 38 openh264 (From Cisco) - x86_64 - Debug 4.2 kB/s | 3.7 kB 00:00 Fedora Modular 38 - x86_64 - Debug 940 kB/s | 1.0 MB 00:01 Fedora 38 - x86_64 - Updates - Debug 4.3 MB/s | 11 MB 00:02 Fedora Modular 38 - x86_64 - Updates - Debug 439 B/s | 257 B 00:00 RPM Fusion for Fedora 38 - Free - Debug 1.3 MB/s | 760 kB 00:00 RPM Fusion for Fedora 38 - Free - Updates Debug 580 kB/s | 321 kB 00:00 RPM Fusion for Fedora 38 - Nonfree - Debug 304 kB/s | 148 kB 00:00 RPM Fusion for Fedora 38 - Nonfree - Updates Debug 23 kB/s | 5.2 kB 00:00 Dependencies resolved. ============================================================================================================================================ Package Architecture Version Repository Size ============================================================================================================================================ Installing: numad-debuginfo x86_64 0.5-38.20150602git.fc38 fedora-debuginfo 62 k numad-debugsource x86_64 0.5-38.20150602git.fc38 fedora-debuginfo 29 k Transaction Summary ============================================================================================================================================ Install 2 Packages Total download size: 92 k Installed size: 238 k Is this ok [y/N]: y Downloading Packages: (1/2): numad-debuginfo-0.5-38.20150602git.fc38.x86_64.rpm 138 kB/s | 62 kB 00:00 (2/2): numad-debugsource-0.5-38.20150602git.fc38.x86_64.rpm 64 kB/s | 29 kB 00:00 -------------------------------------------------------------------------------------------------------------------------------------------- Total 103 kB/s | 92 kB 00:00 Running transaction check Transaction check succeeded. Running transaction test Transaction test succeeded. Running transaction Preparing : 1/1 Installing : numad-debugsource-0.5-38.20150602git.fc38.x86_64 1/2 Installing : numad-debuginfo-0.5-38.20150602git.fc38.x86_64 2/2 Running scriptlet: numad-debuginfo-0.5-38.20150602git.fc38.x86_64 2/2 Verifying : numad-debuginfo-0.5-38.20150602git.fc38.x86_64 1/2 Verifying : numad-debugsource-0.5-38.20150602git.fc38.x86_64 2/2 Installed: numad-debuginfo-0.5-38.20150602git.fc38.x86_64 numad-debugsource-0.5-38.20150602git.fc38.x86_64 Complete! $ Run gdb against numad: $ sudo gdb numad GNU gdb (GDB) Fedora Linux 13.1-4.fc38 Copyright (C) 2023 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html> This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type "show copying" and "show warranty" for details. This GDB was configured as "x86_64-redhat-linux-gnu". Type "show configuration" for configuration details. For bug reporting instructions, please see: <https://www.gnu.org/software/gdb/bugs/>. Find the GDB manual and other documentation resources online at: <http://www.gnu.org/software/gdb/documentation/>. For help, type "help". Type "apropos word" to search for commands related to "word"... Reading symbols from numad... Reading symbols from /usr/lib/debug/usr/bin/numad-0.5-38.20150602git.fc38.x86_64.debug... (gdb) run -i 15 Starting program: /usr/bin/numad -i 15 [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib64/libthread_db.so.1". *** buffer overflow detected ***: terminated Program received signal SIGABRT, Aborted. 0x00007ffff7d68844 in __pthread_kill_implementation () from /lib64/libc.so.6 Missing separate debuginfos, use: dnf debuginfo-install glibc-2.37-4.fc38.x86_64 (gdb) quit A debugging session is active. Inferior 1 [process 6004] will be killed. Quit anyway? (y or n) y $ Install glibc debuginfo: $ sudo dnf debuginfo-install glibc enabling fedora-debuginfo repository enabling fedora-cisco-openh264-debuginfo repository enabling fedora-modular-debuginfo repository enabling updates-debuginfo repository enabling updates-modular-debuginfo repository enabling rpmfusion-free-debuginfo repository enabling rpmfusion-free-updates-debuginfo repository enabling rpmfusion-nonfree-debuginfo repository enabling rpmfusion-nonfree-updates-debuginfo repository Last metadata expiration check: 0:02:12 ago on Sat 27 May 2023 18:55:17 BST. Dependencies resolved. ============================================================================================================================================ Package Architecture Version Repository Size ============================================================================================================================================ Installing: glibc-debuginfo i686 2.37-4.fc38 updates-debuginfo 4.3 M glibc-debuginfo x86_64 2.37-4.fc38 updates-debuginfo 4.6 M glibc-debugsource i686 2.37-4.fc38 updates-debuginfo 5.5 M glibc-debugsource x86_64 2.37-4.fc38 updates-debuginfo 6.1 M Transaction Summary ============================================================================================================================================ Install 4 Packages Total download size: 21 M Installed size: 106 M Is this ok [y/N]: y Downloading Packages: (1/4): glibc-debuginfo-2.37-4.fc38.i686.rpm (2/4): glibc-debugsource-2.37-4.fc38.i686.rpm 2.3 MB/s | 5.5 MB 00:02 (3/4): glibc-debuginfo-2.37-4.fc38.x86_64.rpm 1.6 MB/s | 4.6 MB 00:02 (4/4): glibc-debugsource-2.37-4.fc38.x86_64.rpm 4.1 MB/s | 6.1 MB 00:01 -------------------------------------------------------------------------------------------------------------------------------------------- Total 5.3 MB/s | 21 MB 00:03 Running transaction check Transaction check succeeded. Running transaction test Transaction test succeeded. Running transaction Preparing : 1/1 Installing : glibc-debugsource-2.37-4.fc38.x86_64 1/4 Installing : glibc-debugsource-2.37-4.fc38.i686 2/4 Installing : glibc-debuginfo-2.37-4.fc38.i686 3/4 Installing : glibc-debuginfo-2.37-4.fc38.x86_64 4/4 Running scriptlet: glibc-debuginfo-2.37-4.fc38.x86_64 4/4 Verifying : glibc-debuginfo-2.37-4.fc38.i686 1/4 Verifying : glibc-debuginfo-2.37-4.fc38.x86_64 2/4 Verifying : glibc-debugsource-2.37-4.fc38.i686 3/4 Verifying : glibc-debugsource-2.37-4.fc38.x86_64 4/4 Installed: glibc-debuginfo-2.37-4.fc38.i686 glibc-debuginfo-2.37-4.fc38.x86_64 glibc-debugsource-2.37-4.fc38.i686 glibc-debugsource-2.37-4.fc38.x86_64 Complete! $ Run gdb against numad: $ sudo gdb numad GNU gdb (GDB) Fedora Linux 13.1-4.fc38 Copyright (C) 2023 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html> This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type "show copying" and "show warranty" for details. This GDB was configured as "x86_64-redhat-linux-gnu". Type "show configuration" for configuration details. For bug reporting instructions, please see: <https://www.gnu.org/software/gdb/bugs/>. Find the GDB manual and other documentation resources online at: <http://www.gnu.org/software/gdb/documentation/>. For help, type "help". Type "apropos word" to search for commands related to "word"... Reading symbols from numad... Reading symbols from /usr/lib/debug/usr/bin/numad-0.5-38.20150602git.fc38.x86_64.debug... (gdb) run -i 15 Starting program: /usr/bin/numad -i 15 [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib64/libthread_db.so.1". *** buffer overflow detected ***: terminated Program received signal SIGABRT, Aborted. __pthread_kill_implementation (threadid=<optimized out>, signo=signo@entry=6, no_tid=no_tid@entry=0) at pthread_kill.c:44 44 return INTERNAL_SYSCALL_ERROR_P (ret) ? INTERNAL_SYSCALL_ERRNO (ret) : 0; (gdb) thread apply all bt full Thread 1 (Thread 0x7ffff7cd7740 (LWP 6232) "numad"): #0 __pthread_kill_implementation (threadid=<optimized out>, signo=signo@entry=6, no_tid=no_tid@entry=0) at pthread_kill.c:44 tid = <optimized out> ret = 0 pd = <optimized out> old_mask = {__val = {0}} ret = <optimized out> #1 0x00007ffff7d688b3 in __pthread_kill_internal (signo=6, threadid=<optimized out>) at pthread_kill.c:78 No locals. #2 0x00007ffff7d17abe in __GI_raise (sig=sig@entry=6) at ../sysdeps/posix/raise.c:26 ret = <optimized out> #3 0x00007ffff7d0087f in __GI_abort () at abort.c:79 save_stage = 1 act = {__sigaction_handler = {sa_handler = 0x20, sa_sigaction = 0x20}, sa_mask = {__val = {140737488342080, 10000, 4, 140737488340480, 140737351364001, 140737352783296, 140737352783321, 140737352783410, 0, 34359738370, 8, 8, 17, 0, 0, 93824992297868}}, sa_flags = 1431716748, sa_restorer = 0x3266695a54} #4 0x00007ffff7d0160f in __libc_message (fmt=fmt@entry=0x7ffff7e7b2e6 "*** %s ***: terminated\n") at ../sysdeps/posix/libc_fatal.c:150 ap = {{gp_offset = 16, fp_offset = 0, overflow_arg_area = 0x7fffffffc700, reg_save_area = 0x7fffffffc690}} fd = 2 list = <optimized out> nlist = <optimized out> cp = <optimized out> #5 0x00007ffff7dfc979 in __GI___fortify_fail (msg=msg@entry=0x7ffff7e7b28c "buffer overflow detected") at fortify_fail.c:24 No locals. #6 0x00007ffff7dfb1b4 in __GI___chk_fail () at chk_fail.c:28 No locals. #7 0x00007ffff7dfadd1 in ___vsnprintf_chk (s=<optimized out>, maxlen=<optimized out>, flag=<optimized out>, slen=<optimized out>, format=<optimized out>, ap=<optimized out>) at vsnprintf_chk.c:28 mode = <optimized out> #8 0x0000555555559071 in vsnprintf (__ap=0x7fffffffc740, __fmt=0x55555555ef68 "Changing THP scan time in %s from %d to %d ms.\n", __n=1024, __s=<optimized out>) at /usr/include/bits/stdio2.h:68 No locals. #9 numad_log (level=level@entry=5, fmt=fmt@entry=0x55555555ef68 "Changing THP scan time in %s from %d to %d ms.\n") at /usr/src/debug/numad-0.5-38.20150602git.fc38.x86_64/numad.c:156 buf = "Sat May 27 18:58:59 2023: ", '\000' <repeats 997 times> ts = 1685210339 p = <optimized out> ap = {{gp_offset = 16, fp_offset = 48, overflow_arg_area = 0x7fffffffcc40, reg_save_area = 0x7fffffffcb70}} #10 0x000055555555cd68 in set_thp_scan_sleep_ms (new_ms=1000) at /usr/src/debug/numad-0.5-38.20150602git.fc38.x86_64/numad.c:764 cur_ms = 10000 p = <optimized out> buf = "10000\n", '\000' <repeats 418 times>... bytes = <optimized out> thp_scan_fname = 0x55555555ef20 "/sys/kernel/mm/transparent_hugepage/khugepaged/scan_sleep_millisecs" fd = 4 #11 0x000055555555723d in set_thp_scan_sleep_ms (new_ms=<optimized out>) at /usr/src/debug/numad-0.5-38.20150602git.fc38.x86_64/numad.c:748 thp_scan_fname = <optimized out> fd = <optimized out> buf = <optimized out> bytes = <optimized out> cur_ms = <optimized out> p = <optimized out> #12 check_prereqs (prog_name=<optimized out>) at /usr/src/debug/numad-0.5-38.20150602git.fc38.x86_64/numad.c:775 No locals. #13 main (argc=<optimized out>, argv=<optimized out>) at /usr/src/debug/numad-0.5-38.20150602git.fc38.x86_64/numad.c:2592 sid = <optimized out> sa = {__sigaction_handler = {sa_handler = 0x0, sa_sigaction = 0x0}, sa_mask = {__val = {0, 0, 0, 0, 140737353963383, 0, 0, 140737354125312, 140737353717008, 140737488344736, 140737353948969, 17179869190, 64, 64, 64, 784}}, sa_flags = 784, sa_restorer = 0x8} attr = {__size = '\000' <repeats 55 times>, __align = 0} tid = 0 opt = <optimized out> C_flag = <optimized out> d_flag = <optimized out> H_flag = <optimized out> i_flag = <optimized out> K_flag = <optimized out> l_flag = <optimized out> m_flag = <optimized out> p_flag = <optimized out> r_flag = <optimized out> --Type <RET> for more, q to quit, c to continue without paging--c S_flag = <optimized out> t_flag = <optimized out> u_flag = <optimized out> v_flag = <optimized out> w_flag = <optimized out> x_flag = <optimized out> tmp_int = <optimized out> list_pid = <optimized out> daemon_pid = <optimized out> (gdb) So looks like an issue with vsnprintf function in glibc??? Reproducible: Always Steps to Reproduce: 1.See above. 2. 3. I also compared the source code for numad between Fedora 37 against 38 and confirmed there were no changes between these versions besides the version number. I have since restored the system back to Fedora 37 and numad is working as before the upgrade (the above debug trace info is from a Fedora 38 KVM but the result is the same as the physical machine).
Found same issue on fc39 build: # rpm -qiv numad Name : numad Version : 0.5 Release : 39.20150602git.fc39 Architecture: x86_64 Install Date: Sun 18 Jun 2023 11:27:48 PM EDT Group : Unspecified Size : 61328 License : LGPL-2.1-only Signature : RSA/SHA256, Fri 16 Jun 2023 04:39:48 AM EDT, Key ID 75cf5ac418b8e74c Source RPM : numad-0.5-39.20150602git.fc39.src.rpm Build Date : Fri 16 Jun 2023 04:39:31 AM EDT Build Host : buildvm-x86-12.iad2.fedoraproject.org Packager : Fedora Project Vendor : Fedora Project URL : https://pagure.io/numad Bug URL : https://bugz.fedoraproject.org/numad Summary : NUMA user daemon Description : Numad, a daemon for NUMA (Non-Uniform Memory Architecture) systems, that monitors NUMA characteristics and manages placement of processes and memory to minimize memory latency and thus provide optimum performance. # numad *** buffer overflow detected ***: terminated Aborted (core dumped)
If I look in /usr/src/debug/glibc-2.36-9.fc37.x86_64/libio/vsnprintf.c we see this code: ... int __vsnprintf_internal (char *string, size_t maxlen, const char *format, va_list args, unsigned int mode_flags) { _IO_strnfile sf; int ret; #ifdef _IO_MTSAFE_IO sf.f._sbf._f._lock = NULL; #endif /* We need to handle the special case where MAXLEN is 0. Use the overflow buffer right from the start. */ if (maxlen == 0) { string = sf.overflow_buf; maxlen = sizeof (sf.overflow_buf); } ... But it's been re-written in /usr/src/debug/glibc-2.37-4.fc38.x86_64/libio/vsnprintf.c: ... void __printf_buffer_snprintf_init (struct __printf_buffer_snprintf *buf, char *buffer, size_t length) { __printf_buffer_init (&buf->base, buffer, length, __printf_buffer_mode_snprintf); if (length > 0) /* Historic behavior for trivially overlapping buffers (checked by the test suite). */ *buffer = '\0'; } ... int __vsnprintf_internal (char *string, size_t maxlen, const char *format, va_list args, unsigned int mode_flags) { struct __printf_buffer_snprintf buf; __printf_buffer_snprintf_init (&buf, string, maxlen); __printf_buffer (&buf.base, format, args, mode_flags); return __printf_buffer_snprintf_done (&buf); } So should the statement "if (length > 0)" be "if (length == 0)"??? You can see in the previous GDB output that the buf[fer] in numad_log function is repeating 997.
Reproduced on my PC (numad-0.5-38.20150602git.fc38). numad_log() in numad-0.5git/numad.c: ``` char buf[BUF_SIZE]; time_t ts = time(NULL); strncpy(buf, ctime(&ts), sizeof(buf)); char *p = &buf[strlen(buf) - 1]; *p++ = ':'; *p++ = ' '; va_list ap; va_start(ap, fmt); vsnprintf(p, BUF_SIZE, fmt, ap); ``` The length of buf is BUF_SIZE. p may point a position between buf and buf + BUF_SIZE. So the length of the buffer pointed by p may be smaller than BUF_SIZE. I think the following change is needed: ``` diff -uN /tmp/numad.c.orig /tmp/numad.c --- /tmp/numad.c.orig 2023-06-25 02:39:54.642694621 +0900 +++ /tmp/numad.c 2023-06-25 02:39:43.414453349 +0900 @@ -19,7 +19,7 @@ *p++ = ' '; va_list ap; va_start(ap, fmt); - vsnprintf(p, BUF_SIZE, fmt, ap); + vsnprintf(p, BUF_SIZE - (p - buf), fmt, ap); va_end(ap); fprintf(log_fs, "%s", buf); fflush(log_fs); ```
systemd-coredump captures the coredump: [root@dev192 yamato]# coredumpctl debug 1605 PID: 1605 (numad) UID: 0 (root) GID: 0 (root) Signal: 6 (ABRT) Timestamp: Fri 2023-06-23 05:13:20 JST (1 day 21h ago) Command Line: /usr/bin/numad -i 15 Executable: /usr/bin/numad Control Group: /system.slice/numad.service Unit: numad.service Slice: system.slice Boot ID: 10a20ef42fc3437286cfd15f17e0896b Machine ID: ab2726b4e6744a0d98c7f57073398301 Hostname: dev192 Storage: /var/lib/systemd/coredump/core.numad.0.10a20ef42fc3437286cfd15f17e0896b.1605.1687464800000000.zst (present) Size on Disk: 23.6K Package: numad/0.5-38.20150602git.fc38 build-id: b02ab1f2d69771b9076deef117c2f53b7d6addc3 Message: Process 1605 (numad) of user 0 dumped core. Module numad from rpm numad-0.5-38.20150602git.fc38.x86_64 Stack trace of thread 1605: #0 0x00007f5584eb9844 __pthread_kill_implementation (libc.so.6 + 0x8e844) #1 0x00007f5584e68abe raise (libc.so.6 + 0x3dabe) #2 0x00007f5584e5187f abort (libc.so.6 + 0x2687f) #3 0x00007f5584e5260f __libc_message.cold (libc.so.6 + 0x2760f) #4 0x00007f5584f4d979 __fortify_fail (libc.so.6 + 0x122979) #5 0x00007f5584f4c1b4 __chk_fail (libc.so.6 + 0x1211b4) #6 0x00007f5584f4bdd1 __vsnprintf_chk (libc.so.6 + 0x120dd1) #7 0x0000557f76583071 numad_log (numad + 0x5071) #8 0x0000557f76586d68 set_thp_scan_sleep_ms.part.0 (numad + 0x8d68) #9 0x0000557f7658123d main (numad + 0x323d) #10 0x00007f5584e52b4a __libc_start_call_main (libc.so.6 + 0x27b4a) #11 0x00007f5584e52c0b __libc_start_main@@GLIBC_2.34 (libc.so.6 + 0x27c0b) #12 0x0000557f76582e25 _start (numad + 0x4e25) ELF object binary architecture: AMD x86-64 GNU gdb (GDB) Fedora Linux 13.2-1.fc38
I built numad 0.7.1 from https://github.com/yhaenggi/numad and replaced /usr/bin/numad and the affected KVM VMs now start. I also found that editing the domain xml and changing vcpu placement from 'auto' to 'static', and deleting the numatune element allowed the VMs to start with numad 0.5, although I assume that is because numad is simply not called.
Interestingly if you compile the Pagure version of numad as listed in the package info, i.e. $ rpm -qiv numad Name : numad Version : 0.5 Release : 38.20150602git.fc38 Architecture: x86_64 Install Date: Sat 20 May 2023 13:49:21 BST Group : Unspecified Size : 61824 License : LGPLv2 Signature : RSA/SHA256, Fri 20 Jan 2023 07:10:14 GMT, Key ID 809a8d7ceb10b464 Source RPM : numad-0.5-38.20150602git.fc38.src.rpm Build Date : Fri 20 Jan 2023 00:36:52 GMT Build Host : buildvm-x86-20.iad2.fedoraproject.org Packager : Fedora Project Vendor : Fedora Project URL : https://pagure.io/numad Bug URL : https://bugz.fedoraproject.org/numad Summary : NUMA user daemon Description : Numad, a daemon for NUMA (Non-Uniform Memory Architecture) systems, that monitors NUMA characteristics and manages placement of processes and memory to minimize memory latency and thus provide optimum performance. It's the same version as the original buggy version but works (but also note it's the same size as the yhaengi version, e.g. $ ll -Z /usr/bin/numad* lrwxrwxrwx. 1 root root unconfined_u:object_r:bin_t:s0 21 Aug 17 12:30 /usr/bin/numad -> /usr/bin/numad.pagure -rwxr-xr-x. 1 root root system_u:object_r:bin_t:s0 41000 Jan 19 2023 /usr/bin/numademo -rwxr-xr-x. 1 root root system_u:object_r:numad_exec_t:s0 57848 Jan 19 2023 /usr/bin/numad.orig -rwxr-xr-x. 1 root root system_u:object_r:numad_exec_t:s0 82400 Aug 17 12:28 /usr/bin/numad.pagure -rwxr-xr-x. 1 root root system_u:object_r:numad_exec_t:s0 82400 Aug 17 12:16 /usr/bin/numad.yhaenggi $ /usr/bin/numad.pagure -V /usr/bin/numad.pagure version: 20150602: compiled Aug 17 2023 But now it works (also confirmed yhaenggi version works as well but not shown): $ sudo numad -i 15 $ systemctl status numad -l ● numad.service - numad - The NUMA daemon that manages application locality. Loaded: loaded (/usr/lib/systemd/system/numad.service; enabled; preset: disabled) Drop-In: /usr/lib/systemd/system/service.d └─10-timeout-abort.conf Active: active (running) since Wed 2023-08-23 13:44:37 BST; 52min ago Docs: man:numad Process: 946 ExecStart=/usr/bin/numad -i 15 (code=exited, status=0/SUCCESS) Main PID: 951 (numad) Tasks: 2 (limit: 4631) Memory: 380.0K CPU: 36ms CGroup: /system.slice/numad.service └─951 /usr/bin/numad -i 15 Aug 23 13:44:37 kvm-125 systemd[1]: Starting numad.service - numad - The NUMA daemon that manages application locality.... Aug 23 13:44:37 kvm-125 systemd[1]: Started numad.service - numad - The NUMA daemon that manages application locality.. $ So do the F38+ versions of numad need recompiling from the Pagure Git source?
FEDORA-2023-e098610c97 has been submitted as an update to Fedora 39. https://bodhi.fedoraproject.org/updates/FEDORA-2023-e098610c97
FEDORA-2023-4ebdc3f086 has been submitted as an update to Fedora 38. https://bodhi.fedoraproject.org/updates/FEDORA-2023-4ebdc3f086
FEDORA-2023-e098610c97 has been pushed to the Fedora 39 testing repository. Soon you'll be able to install the update with the following command: `sudo dnf upgrade --enablerepo=updates-testing --refresh --advisory=FEDORA-2023-e098610c97` You can provide feedback for this update here: https://bodhi.fedoraproject.org/updates/FEDORA-2023-e098610c97 See also https://fedoraproject.org/wiki/QA:Updates_Testing for more information on how to test updates.
FEDORA-2023-4ebdc3f086 has been pushed to the Fedora 38 testing repository. Soon you'll be able to install the update with the following command: `sudo dnf upgrade --enablerepo=updates-testing --refresh --advisory=FEDORA-2023-4ebdc3f086` You can provide feedback for this update here: https://bodhi.fedoraproject.org/updates/FEDORA-2023-4ebdc3f086 See also https://fedoraproject.org/wiki/QA:Updates_Testing for more information on how to test updates.
FEDORA-2023-4ebdc3f086 has been pushed to the Fedora 38 stable repository. If problem still persists, please make note of it in this bug report.
FEDORA-2023-e098610c97 has been pushed to the Fedora 39 stable repository. If problem still persists, please make note of it in this bug report.