canonical-ubuntu-qa team mailing list archive

Thread
Date
[Bug 2119369] [NEW] ubuntu_ltp_controllers:cpuacct_100_100 cpuacct.sh: Cannot fork failure

To: canonical-ubuntu-qa@xxxxxxxxxxxxxxxxxxx
From: Mehmet Basaran <2119369@xxxxxxxxxxxxxxxxxx>
Date: Sat, 02 Aug 2025 11:53:43 -0000
Reply-to: Bug 2119369 <2119369@xxxxxxxxxxxxxxxxxx>
Sender: noreply@xxxxxxxxxxxxx
Public bug reported:

I have seen this failure for bionic:linux-aws-5.4
(5.4.0-1149.159~18.04.1) on aws:m5a.large instance. However, failure is
due to incorrect setup and can happen on any instance if PAM (pluggable
authentication modules) is misconfigured. I was not able to reproduce
the issue because when I reserved m5a.large instance from aws, it was
setup properly.

dmesg states "cgroup: fork rejected by pids controller in
/system.slice/ssh.service"

log states "tst_pid.c:84: TINFO: Cannot read session user limits from
'/sys/fs/cgroup/pids/user.slice/user-1000.slice/pids.max'"

Since PAM is misconfigured, user processes cannot transition from
system.slice to user.slice and thus limited by the number of processes
in "/sys/fs/cgroup/pids/system.slice/ssh.service/pids.max" which is only
4915.

Probably installing pam_systemd package will resolve this issue. I created this
bug to hint test results. This is not a kernel regression.

Some links: https://www.suse.com/support/kb/doc/?id=000019044

Full log:

09:48:46 INFO | START ubuntu_ltp_controllers.cpuacct_100_100 ubuntu_ltp_controllers.cpuacct_100_100 timestamp=1754041726 timeout=4500 localtime=Aug 01 09:48:46
09:48:46 DEBUG| Persistent state client._record_indent now set to 2
09:48:46 DEBUG| Persistent state client.unexpected_reboot now set to ('ubuntu_ltp_controllers.cpuacct_100_100', 'ubuntu_ltp_controllers.cpuacct_100_100')
09:48:46 DEBUG| Waiting for pid 8115 for 4500 seconds
09:48:46 ERROR| [stderr] -------------------------------------------
09:48:46 ERROR| [stderr] INFO: runltp script is deprecated, try kirk
09:48:46 ERROR| [stderr] https://github.com/linux-test-project/kirk
09:48:46 ERROR| [stderr] -------------------------------------------
09:48:46 DEBUG| [stdout] Checking for required user/group ids
09:48:46 DEBUG| [stdout]
09:48:46 DEBUG| [stdout] 'root' user id and group found.
09:48:46 DEBUG| [stdout] 'nobody' user id and group found.
09:48:46 DEBUG| [stdout] 'bin' user id and group found.
09:48:46 DEBUG| [stdout] 'daemon' user id and group found.
09:48:46 DEBUG| [stdout] Users group found.
09:48:46 DEBUG| [stdout] Sys group found.
09:48:46 DEBUG| [stdout] Required users/groups exist.
09:48:46 DEBUG| [stdout] no big block device was specified on commandline.
09:48:46 DEBUG| [stdout] Tests which require a big block device are disabled.
09:48:46 DEBUG| [stdout] You can specify it with option -z
09:48:46 DEBUG| [stdout] INFO: Test start time: Fri Aug 1 09:48:46 UTC 2025
09:48:46 DEBUG| [stdout] COMMAND: /opt/ltp/bin/ltp-pan -q -e -S -a 8119 -n 8119 -f /tmp/ltp-LwNIp0yl3q/alltests -l /dev/null -C /dev/null -T /dev/null
09:48:46 DEBUG| [stdout] LOG File: /dev/null
09:48:46 DEBUG| [stdout] FAILED COMMAND File: /dev/null
09:48:46 DEBUG| [stdout] TCONF COMMAND File: /dev/null
09:48:46 DEBUG| [stdout] Running tests.......
09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: Running: cpuacct.sh 100 100
09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: Tested kernel: Linux ip-172-31-95-118 5.4.0-1149-aws #159~18.04.1-Ubuntu SMP Wed Jul 16 07:37:15 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux
09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: Using /tmp/ltp-LwNIp0yl3q/LTP_cpuacct.vTpsrKgADv as tmpdir (ext2/ext3 filesystem)
09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: timeout per run is 0h 5m 0s
09:48:46 DEBUG| [stdout] tst_pid.c:84: TINFO: Cannot read session user limits from '/sys/fs/cgroup/user.slice/user-1000.slice/pids.max'
09:48:46 DEBUG| [stdout] tst_pid.c:84: TINFO: Cannot read session user limits from '/sys/fs/cgroup/pids/user.slice/user-1000.slice/pids.max'
09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: task limit fulfilled (approximate need 10000, limit 32523)
09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: memory requirements fulfilled (approximate need 1638400 kB, free 7431092 kB)
09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: cpuacct: /sys/fs/cgroup/cpu,cpuacct
09:48:47 DEBUG| [stdout] cpuacct 1 TINFO: Creating 100 subgroups each with 100 processes
09:48:51 DEBUG| [stdout] /opt/ltp/testcases/bin/cpuacct.sh: 0: /opt/ltp/testcases/bin/cpuacct.sh: Cannot fork
09:48:56 DEBUG| [stdout] INFO: ltp-pan reported some tests FAIL
09:48:56 DEBUG| [stdout] LTP Version: 20250130
09:48:56 DEBUG| [stdout] INFO: Test end time: Fri Aug 1 09:48:56 UTC 2025
09:48:56 ERROR| [stderr] -------------------------------------------
09:48:56 ERROR| [stderr] INFO: runltp script is deprecated, try kirk
09:48:56 ERROR| [stderr] https://github.com/linux-test-project/kirk
09:48:56 ERROR| [stderr] -------------------------------------------

** Affects: ubuntu-kernel-tests
     Importance: Undecided
         Status: New

** Description changed:

  I have seen this failure for bionic:linux-aws-5.4 on aws:m5a.large
  instance. However, failure is due to incorrect setup and can happen on
  any instance if PAM (pluggable authentication modules) is misconfigured.
  I was not able to reproduce the issue because when I reserved m5a.large
  instance from aws, it was setup properly.
  
  dmesg states "cgroup: fork rejected by pids controller in
  /system.slice/ssh.service"
  
  log states "tst_pid.c:84: TINFO: Cannot read session user limits from
  '/sys/fs/cgroup/pids/user.slice/user-1000.slice/pids.max'"
  
  Since PAM is misconfigured, user processes cannot transition from
  system.slice to user.slice and thus limited by the number of processes
  in "/sys/fs/cgroup/pids/system.slice/ssh.service/pids.max" which is only
  4915.
  
- Probably installing pam_systemd package will resolve this issue.
+ Probably installing pam_systemd package will resolve this issue. I created this
+ bug to hint test results. This is not a kernel regression.
  
  Some links: https://www.suse.com/support/kb/doc/?id=000019044
- 
  
  Full log:
  
  09:48:46 INFO | START ubuntu_ltp_controllers.cpuacct_100_100 ubuntu_ltp_controllers.cpuacct_100_100 timestamp=1754041726 timeout=4500 localtime=Aug 01 09:48:46
  09:48:46 DEBUG| Persistent state client._record_indent now set to 2
  09:48:46 DEBUG| Persistent state client.unexpected_reboot now set to ('ubuntu_ltp_controllers.cpuacct_100_100', 'ubuntu_ltp_controllers.cpuacct_100_100')
  09:48:46 DEBUG| Waiting for pid 8115 for 4500 seconds
  09:48:46 ERROR| [stderr] -------------------------------------------
  09:48:46 ERROR| [stderr] INFO: runltp script is deprecated, try kirk
  09:48:46 ERROR| [stderr] https://github.com/linux-test-project/kirk
  09:48:46 ERROR| [stderr] -------------------------------------------
  09:48:46 DEBUG| [stdout] Checking for required user/group ids
  09:48:46 DEBUG| [stdout]
  09:48:46 DEBUG| [stdout] 'root' user id and group found.
  09:48:46 DEBUG| [stdout] 'nobody' user id and group found.
  09:48:46 DEBUG| [stdout] 'bin' user id and group found.
  09:48:46 DEBUG| [stdout] 'daemon' user id and group found.
  09:48:46 DEBUG| [stdout] Users group found.
  09:48:46 DEBUG| [stdout] Sys group found.
  09:48:46 DEBUG| [stdout] Required users/groups exist.
  09:48:46 DEBUG| [stdout] no big block device was specified on commandline.
  09:48:46 DEBUG| [stdout] Tests which require a big block device are disabled.
  09:48:46 DEBUG| [stdout] You can specify it with option -z
  09:48:46 DEBUG| [stdout] INFO: Test start time: Fri Aug 1 09:48:46 UTC 2025
  09:48:46 DEBUG| [stdout] COMMAND: /opt/ltp/bin/ltp-pan -q -e -S -a 8119 -n 8119 -f /tmp/ltp-LwNIp0yl3q/alltests -l /dev/null -C /dev/null -T /dev/null
  09:48:46 DEBUG| [stdout] LOG File: /dev/null
  09:48:46 DEBUG| [stdout] FAILED COMMAND File: /dev/null
  09:48:46 DEBUG| [stdout] TCONF COMMAND File: /dev/null
  09:48:46 DEBUG| [stdout] Running tests.......
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: Running: cpuacct.sh 100 100
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: Tested kernel: Linux ip-172-31-95-118 5.4.0-1149-aws #159~18.04.1-Ubuntu SMP Wed Jul 16 07:37:15 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: Using /tmp/ltp-LwNIp0yl3q/LTP_cpuacct.vTpsrKgADv as tmpdir (ext2/ext3 filesystem)
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: timeout per run is 0h 5m 0s
  09:48:46 DEBUG| [stdout] tst_pid.c:84: TINFO: Cannot read session user limits from '/sys/fs/cgroup/user.slice/user-1000.slice/pids.max'
  09:48:46 DEBUG| [stdout] tst_pid.c:84: TINFO: Cannot read session user limits from '/sys/fs/cgroup/pids/user.slice/user-1000.slice/pids.max'
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: task limit fulfilled (approximate need 10000, limit 32523)
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: memory requirements fulfilled (approximate need 1638400 kB, free 7431092 kB)
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: cpuacct: /sys/fs/cgroup/cpu,cpuacct
  09:48:47 DEBUG| [stdout] cpuacct 1 TINFO: Creating 100 subgroups each with 100 processes
  09:48:51 DEBUG| [stdout] /opt/ltp/testcases/bin/cpuacct.sh: 0: /opt/ltp/testcases/bin/cpuacct.sh: Cannot fork
  09:48:56 DEBUG| [stdout] INFO: ltp-pan reported some tests FAIL
  09:48:56 DEBUG| [stdout] LTP Version: 20250130
  09:48:56 DEBUG| [stdout] INFO: Test end time: Fri Aug 1 09:48:56 UTC 2025
  09:48:56 ERROR| [stderr] -------------------------------------------
  09:48:56 ERROR| [stderr] INFO: runltp script is deprecated, try kirk
  09:48:56 ERROR| [stderr] https://github.com/linux-test-project/kirk
  09:48:56 ERROR| [stderr] -------------------------------------------

** Description changed:

- I have seen this failure for bionic:linux-aws-5.4 on aws:m5a.large
- instance. However, failure is due to incorrect setup and can happen on
- any instance if PAM (pluggable authentication modules) is misconfigured.
- I was not able to reproduce the issue because when I reserved m5a.large
- instance from aws, it was setup properly.
+ I have seen this failure for bionic:linux-aws-5.4
+ (5.4.0-1149.159~18.04.1) on aws:m5a.large instance. However, failure is
+ due to incorrect setup and can happen on any instance if PAM (pluggable
+ authentication modules) is misconfigured. I was not able to reproduce
+ the issue because when I reserved m5a.large instance from aws, it was
+ setup properly.
  
  dmesg states "cgroup: fork rejected by pids controller in
  /system.slice/ssh.service"
  
  log states "tst_pid.c:84: TINFO: Cannot read session user limits from
  '/sys/fs/cgroup/pids/user.slice/user-1000.slice/pids.max'"
  
  Since PAM is misconfigured, user processes cannot transition from
  system.slice to user.slice and thus limited by the number of processes
  in "/sys/fs/cgroup/pids/system.slice/ssh.service/pids.max" which is only
  4915.
  
  Probably installing pam_systemd package will resolve this issue. I created this
  bug to hint test results. This is not a kernel regression.
  
  Some links: https://www.suse.com/support/kb/doc/?id=000019044
  
  Full log:
  
  09:48:46 INFO | START ubuntu_ltp_controllers.cpuacct_100_100 ubuntu_ltp_controllers.cpuacct_100_100 timestamp=1754041726 timeout=4500 localtime=Aug 01 09:48:46
  09:48:46 DEBUG| Persistent state client._record_indent now set to 2
  09:48:46 DEBUG| Persistent state client.unexpected_reboot now set to ('ubuntu_ltp_controllers.cpuacct_100_100', 'ubuntu_ltp_controllers.cpuacct_100_100')
  09:48:46 DEBUG| Waiting for pid 8115 for 4500 seconds
  09:48:46 ERROR| [stderr] -------------------------------------------
  09:48:46 ERROR| [stderr] INFO: runltp script is deprecated, try kirk
  09:48:46 ERROR| [stderr] https://github.com/linux-test-project/kirk
  09:48:46 ERROR| [stderr] -------------------------------------------
  09:48:46 DEBUG| [stdout] Checking for required user/group ids
  09:48:46 DEBUG| [stdout]
  09:48:46 DEBUG| [stdout] 'root' user id and group found.
  09:48:46 DEBUG| [stdout] 'nobody' user id and group found.
  09:48:46 DEBUG| [stdout] 'bin' user id and group found.
  09:48:46 DEBUG| [stdout] 'daemon' user id and group found.
  09:48:46 DEBUG| [stdout] Users group found.
  09:48:46 DEBUG| [stdout] Sys group found.
  09:48:46 DEBUG| [stdout] Required users/groups exist.
  09:48:46 DEBUG| [stdout] no big block device was specified on commandline.
  09:48:46 DEBUG| [stdout] Tests which require a big block device are disabled.
  09:48:46 DEBUG| [stdout] You can specify it with option -z
  09:48:46 DEBUG| [stdout] INFO: Test start time: Fri Aug 1 09:48:46 UTC 2025
  09:48:46 DEBUG| [stdout] COMMAND: /opt/ltp/bin/ltp-pan -q -e -S -a 8119 -n 8119 -f /tmp/ltp-LwNIp0yl3q/alltests -l /dev/null -C /dev/null -T /dev/null
  09:48:46 DEBUG| [stdout] LOG File: /dev/null
  09:48:46 DEBUG| [stdout] FAILED COMMAND File: /dev/null
  09:48:46 DEBUG| [stdout] TCONF COMMAND File: /dev/null
  09:48:46 DEBUG| [stdout] Running tests.......
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: Running: cpuacct.sh 100 100
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: Tested kernel: Linux ip-172-31-95-118 5.4.0-1149-aws #159~18.04.1-Ubuntu SMP Wed Jul 16 07:37:15 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: Using /tmp/ltp-LwNIp0yl3q/LTP_cpuacct.vTpsrKgADv as tmpdir (ext2/ext3 filesystem)
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: timeout per run is 0h 5m 0s
  09:48:46 DEBUG| [stdout] tst_pid.c:84: TINFO: Cannot read session user limits from '/sys/fs/cgroup/user.slice/user-1000.slice/pids.max'
  09:48:46 DEBUG| [stdout] tst_pid.c:84: TINFO: Cannot read session user limits from '/sys/fs/cgroup/pids/user.slice/user-1000.slice/pids.max'
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: task limit fulfilled (approximate need 10000, limit 32523)
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: memory requirements fulfilled (approximate need 1638400 kB, free 7431092 kB)
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: cpuacct: /sys/fs/cgroup/cpu,cpuacct
  09:48:47 DEBUG| [stdout] cpuacct 1 TINFO: Creating 100 subgroups each with 100 processes
  09:48:51 DEBUG| [stdout] /opt/ltp/testcases/bin/cpuacct.sh: 0: /opt/ltp/testcases/bin/cpuacct.sh: Cannot fork
  09:48:56 DEBUG| [stdout] INFO: ltp-pan reported some tests FAIL
  09:48:56 DEBUG| [stdout] LTP Version: 20250130
  09:48:56 DEBUG| [stdout] INFO: Test end time: Fri Aug 1 09:48:56 UTC 2025
  09:48:56 ERROR| [stderr] -------------------------------------------
  09:48:56 ERROR| [stderr] INFO: runltp script is deprecated, try kirk
  09:48:56 ERROR| [stderr] https://github.com/linux-test-project/kirk
  09:48:56 ERROR| [stderr] -------------------------------------------

-- 
You received this bug notification because you are a member of Canonical
Platform QA Team, which is subscribed to ubuntu-kernel-tests.
https://bugs.launchpad.net/bugs/2119369

Title:
   ubuntu_ltp_controllers:cpuacct_100_100 cpuacct.sh: Cannot fork
  failure

Status in ubuntu-kernel-tests:
  New

Bug description:
  I have seen this failure for bionic:linux-aws-5.4
  (5.4.0-1149.159~18.04.1) on aws:m5a.large instance. However, failure
  is due to incorrect setup and can happen on any instance if PAM
  (pluggable authentication modules) is misconfigured. I was not able to
  reproduce the issue because when I reserved m5a.large instance from
  aws, it was setup properly.

  dmesg states "cgroup: fork rejected by pids controller in
  /system.slice/ssh.service"

  log states "tst_pid.c:84: TINFO: Cannot read session user limits from
  '/sys/fs/cgroup/pids/user.slice/user-1000.slice/pids.max'"

  Since PAM is misconfigured, user processes cannot transition from
  system.slice to user.slice and thus limited by the number of processes
  in "/sys/fs/cgroup/pids/system.slice/ssh.service/pids.max" which is
  only 4915.

  Probably installing pam_systemd package will resolve this issue. I created this
  bug to hint test results. This is not a kernel regression.

  Some links: https://www.suse.com/support/kb/doc/?id=000019044

  Full log:

  09:48:46 INFO | START ubuntu_ltp_controllers.cpuacct_100_100 ubuntu_ltp_controllers.cpuacct_100_100 timestamp=1754041726 timeout=4500 localtime=Aug 01 09:48:46
  09:48:46 DEBUG| Persistent state client._record_indent now set to 2
  09:48:46 DEBUG| Persistent state client.unexpected_reboot now set to ('ubuntu_ltp_controllers.cpuacct_100_100', 'ubuntu_ltp_controllers.cpuacct_100_100')
  09:48:46 DEBUG| Waiting for pid 8115 for 4500 seconds
  09:48:46 ERROR| [stderr] -------------------------------------------
  09:48:46 ERROR| [stderr] INFO: runltp script is deprecated, try kirk
  09:48:46 ERROR| [stderr] https://github.com/linux-test-project/kirk
  09:48:46 ERROR| [stderr] -------------------------------------------
  09:48:46 DEBUG| [stdout] Checking for required user/group ids
  09:48:46 DEBUG| [stdout]
  09:48:46 DEBUG| [stdout] 'root' user id and group found.
  09:48:46 DEBUG| [stdout] 'nobody' user id and group found.
  09:48:46 DEBUG| [stdout] 'bin' user id and group found.
  09:48:46 DEBUG| [stdout] 'daemon' user id and group found.
  09:48:46 DEBUG| [stdout] Users group found.
  09:48:46 DEBUG| [stdout] Sys group found.
  09:48:46 DEBUG| [stdout] Required users/groups exist.
  09:48:46 DEBUG| [stdout] no big block device was specified on commandline.
  09:48:46 DEBUG| [stdout] Tests which require a big block device are disabled.
  09:48:46 DEBUG| [stdout] You can specify it with option -z
  09:48:46 DEBUG| [stdout] INFO: Test start time: Fri Aug 1 09:48:46 UTC 2025
  09:48:46 DEBUG| [stdout] COMMAND: /opt/ltp/bin/ltp-pan -q -e -S -a 8119 -n 8119 -f /tmp/ltp-LwNIp0yl3q/alltests -l /dev/null -C /dev/null -T /dev/null
  09:48:46 DEBUG| [stdout] LOG File: /dev/null
  09:48:46 DEBUG| [stdout] FAILED COMMAND File: /dev/null
  09:48:46 DEBUG| [stdout] TCONF COMMAND File: /dev/null
  09:48:46 DEBUG| [stdout] Running tests.......
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: Running: cpuacct.sh 100 100
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: Tested kernel: Linux ip-172-31-95-118 5.4.0-1149-aws #159~18.04.1-Ubuntu SMP Wed Jul 16 07:37:15 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: Using /tmp/ltp-LwNIp0yl3q/LTP_cpuacct.vTpsrKgADv as tmpdir (ext2/ext3 filesystem)
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: timeout per run is 0h 5m 0s
  09:48:46 DEBUG| [stdout] tst_pid.c:84: TINFO: Cannot read session user limits from '/sys/fs/cgroup/user.slice/user-1000.slice/pids.max'
  09:48:46 DEBUG| [stdout] tst_pid.c:84: TINFO: Cannot read session user limits from '/sys/fs/cgroup/pids/user.slice/user-1000.slice/pids.max'
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: task limit fulfilled (approximate need 10000, limit 32523)
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: memory requirements fulfilled (approximate need 1638400 kB, free 7431092 kB)
  09:48:46 DEBUG| [stdout] cpuacct 1 TINFO: cpuacct: /sys/fs/cgroup/cpu,cpuacct
  09:48:47 DEBUG| [stdout] cpuacct 1 TINFO: Creating 100 subgroups each with 100 processes
  09:48:51 DEBUG| [stdout] /opt/ltp/testcases/bin/cpuacct.sh: 0: /opt/ltp/testcases/bin/cpuacct.sh: Cannot fork
  09:48:56 DEBUG| [stdout] INFO: ltp-pan reported some tests FAIL
  09:48:56 DEBUG| [stdout] LTP Version: 20250130
  09:48:56 DEBUG| [stdout] INFO: Test end time: Fri Aug 1 09:48:56 UTC 2025
  09:48:56 ERROR| [stderr] -------------------------------------------
  09:48:56 ERROR| [stderr] INFO: runltp script is deprecated, try kirk
  09:48:56 ERROR| [stderr] https://github.com/linux-test-project/kirk
  09:48:56 ERROR| [stderr] -------------------------------------------

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu-kernel-tests/+bug/2119369/+subscriptions