$ git log --pretty=format:'%h %s (%cs)%d'
9b39ab0 dynticks_testing: Migrate all tasks to CPU 0 (2014-05-26)
(HEAD -> master)
49c96d4 run: Affine writeback to cpu 0 (2013-12-17)
e08b7c9 nohz_tests: Make target optable (2013-08-05)
b748d03 nohz_tests: Initial version (2013-06-12)
$ git cat-file blob HEAD:README
Make sure CPU 1 is full dynticks (it must be in the "nohz_full=" boot parameter
cpu mask).
To run:
$ make
# ./run
Look at the resulting trace in trace.1
hrtimer events should report you the tick events (unless you haven't
CONFIG_HRTIMER=y) and tick_stop events tell you what's going on with full dynticks
behaviour.
trace.1.example shows an example of a successful full dynticks behaviour. The
tick got mostly stopped during the execution of the user loop.
# heads (aka `branches'):
$ git for-each-ref --sort=-creatordate refs/heads \
--format='%(HEAD) %(refname:short) %(subject) (%(creatordate:short))'
* master dynticks_testing: Migrate all tasks to CPU 0 (2014-05-26)
# tags:
$ git for-each-ref --sort=-creatordate refs/tags \
--format='%(refname:short) %(subject) (%(creatordate:short))'
# no tags, yet...
git clone https://80x24.org/lore/pub/scm/linux/kernel/git/frederic/dynticks-testing.git