As requested, I've peeled back the C-state limitation and SRSO
workaround to isolate the issue.

  Current boot parameters:

  retbleed=off nmi_watchdog=1

  Mitigation status after boot:

  /sys/devices/system/cpu/vulnerabilities/retbleed: Vulnerable
  /sys/devices/system/cpu/vulnerabilities/spec_rstack_overflow: Mitigation: 
Safe RET

  C-states available: POLL, C1, C2, C3 (all deep states enabled, no
processor.max_cstate restriction)

  Kernel: 7.0.0-28-generic (Ubuntu 26.04 LTS)

  This means SRSO mitigation (srso_return_thunk/Safe RET) is active,
deep C-states are enabled, and only retbleed mitigation is disabled. I
will monitor for 3-4 days and report back.

  If this remains stable, it confirms the issue is purely in the
retbleed return thunk and not SRSO's implementation, even though earlier
crashes showed srso_safe_ret in the backtrace (likely because
srso_return_thunk is used as the active return thunk when retbleed is
enabled).

-- 
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2160457

Title:
   Kernel panic in srso_safe_ret / kick_ilb during sched_tick on AMD
  Ryzen 7 4800U (Zen2)

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2160457/+subscriptions


-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to