This series fixes the (hopefully) last location of an incorrectly
handled timer slack on rt tasks in hrtimer_start_range_ns(), which was
uncovered by a userland change in glibc 2.33. It does so by removing all special
handling of timerslack on RT tasks by instead setting the timerslack to 0 when
switching to an RT scheduling policy.

Changes since v3:

- short circuit on timerslack=0 in select_estimate_accuracy. This avoids
  useless computations
- add explanation to the commit message why we ignore setting the
  timerslack on RT tasks

Changes since v2:

- complete re-design of approach (as proposed by Thomas Gleixner): Instead of
  overwriting the timerslack in the hrtimer setup path, we now set the timerslack
  to 0 when switching to an RT scheduling policy.

Changes since v1:

- drop patch "hrtimer: Document, that PI boosted tasks have no timer slack", as
  this behavior is incorrect and is already adressed in 20240610192018.1567075-1-qyousef@layalina.io
- use task_is_realtime() instead of rt_task()
- fix style of commit message

v1 discussion: https://lore.kernel.org/lkml/20240805124116.21394-1-felix.moessbauer@siemens.com
v2 discussion: https://lore.kernel.org/lkml/20240805140930.29462-1-felix.moessbauer@siemens.com
v3 discussion: https://lore.kernel.org/lkml/20240813072953.209119-1-felix.moessbauer@siemens.com

Best regards,
Felix Moessbauer
Siemens AG

Felix Moessbauer (1):
  hrtimer: Use and report correct timerslack values for realtime tasks

 fs/proc/base.c          |  9 +++++----
 fs/select.c             | 11 ++++-------
 kernel/sched/syscalls.c |  8 ++++++++
 kernel/sys.c            |  2 ++
 kernel/time/hrtimer.c   | 18 +++---------------
 5 files changed, 22 insertions(+), 26 deletions(-)

-- 
2.39.2