mbox series

[0/4] CSD-lock diagnostics updates for v6.12

Message ID 20240816043600.GA25206@neeraj.linux (mailing list archive)
Headers show
Series CSD-lock diagnostics updates for v6.12 | expand

Message

Neeraj Upadhyay Aug. 16, 2024, 4:36 a.m. UTC
Hello!

This series adds following changes to the CSD-lock diagnostics (rebased on v6.11-rc1):

1.      Print large CSD-lock-hold time numbers as negatives, courtesy of Paul E. McKenney.

2.      Add api to provide an indication of ongoing CSD-lock stall, courtesy of Paul E. McKenney.

3.      Add a backoff mechanism for repeated CSD-lock diagnostic reports of same incident,
        courtesy of Paul E. McKenney.

4.      Print only local CPU info when sched_clock goes backward, courtesy of Rik van Riel.


As communicated earlier at [1], I will be submitting these changes as part of RCU tree,
due to dependency of RCU commit "7c72dedb0079 rcu: Summarize RCU CPU stall warnings during
CSD-lock stalls" [2] on these changes.

Git tree is at [2]

[1] https://lore.kernel.org/lkml/20240722133559.GA667117@neeraj.linux/
[2] https://git.kernel.org/pub/scm/linux/kernel/git/neeraj.upadhyay/linux-rcu.git/commit/?h=rcustall.15.08.24a&id=7c72dedb0079e62c1c75dbab038332017f34a6b8
[3] https://git.kernel.org/pub/scm/linux/kernel/git/neeraj.upadhyay/linux-rcu.git/log/?h=csd.lock.15.08.24a


- Neeraj


Paul E. McKenney (3):
  locking/csd_lock: Print large numbers as negatives
  locking/csd_lock: Provide an indication of ongoing CSD-lock stall
  locking/csd-lock: Use backoff for repeated reports of same incident

Rik van Riel (1):
  smp: print only local CPU info when sched_clock goes backward

 include/linux/smp.h |  6 ++++++
 kernel/smp.c        | 38 +++++++++++++++++++++++++++++++++-----
 lib/Kconfig.debug   |  1 +
 3 files changed, 40 insertions(+), 5 deletions(-)