Just some work I did recently on PE for the balancing issue and lock torture.
Could you please consider testing and/or including in next PE series posting?

Last PE series revived by Connor:
https://lore.kernel.org/lkml/20221003214501.2050087-1-connoro@google.com/

Thank you.

Joel Fernandes (Google) (3):
  sched/pe: Exclude balance callback queuing during proxy()'s migrate
  locktorture: Allow non-rtmutex lock types to be boosted
  locktorture: Make the rt_boost factor a tunable

 kernel/locking/locktorture.c | 93 +++++++++++++++++++-----------------
 kernel/sched/core.c          | 72 +++++++++++++++++++++++++++-
 kernel/sched/sched.h         |  7 ++-
 3 files changed, 126 insertions(+), 46 deletions(-)

-- 
2.38.1.584.g0f3c55d4c2-goog