[Devel] [PATCH v7 00/11] per-cgroup cpu-stat
Glauber Costa
glommer at openvz.org
Wed May 29 04:03:11 PDT 2013
Peter et. al,
I am coming again with this series, hoping this is a better time for you all to
look at it.
I am *not* going as far as marking cpuacct deprecated, because I think it
deserves a special discussion (even though my position in this matter is widely
known), but all the infrastructure to make it happen is here. But after this,
it should be a matter of setting a flag (or not).
Through this patchset I am making cpu cgroup provide the same functionality of
cpuacct, and now with a more clear semantics, I attempt to provide userspace
with enough information to reconstruct per-container version of files like
"/proc/stat". In particular, we are interested in knowing the per-cgroup slices
of user time, system time, wait time, number of processes, and a variety of
statistics.
To make sure we can count nr of switches correctly, I am ressurecting one of
Peter's patches that apparently got nowhere in the past.
Glauber Costa (8):
don't call cpuacct_charge in stop_task.c
sched: adjust exec_clock to use it as cpu usage metric
cpuacct: don't actually do anything.
sched: document the cpu cgroup.
sched: account guest time per-cgroup as well.
sched: record per-cgroup number of context switches
sched: change nr_context_switches calculation.
sched: introduce cgroup file stat_percpu
Peter Zijlstra (1):
sched: Push put_prev_task() into pick_next_task()
Tejun Heo (2):
cgroup: implement CFTYPE_NO_PREFIX
cgroup, sched: let cpu serve the same files as cpuacct
Documentation/cgroups/cpu.txt | 99 +++++++++++
include/linux/cgroup.h | 1 +
kernel/cgroup.c | 16 +-
kernel/sched/core.c | 385 ++++++++++++++++++++++++++++++++++++++++--
kernel/sched/cputime.c | 33 +++-
kernel/sched/fair.c | 39 ++++-
kernel/sched/idle_task.c | 9 +-
kernel/sched/rt.c | 42 +++--
kernel/sched/sched.h | 35 +++-
kernel/sched/stop_task.c | 8 +-
10 files changed, 626 insertions(+), 41 deletions(-)
create mode 100644 Documentation/cgroups/cpu.txt
--
1.8.1.4
More information about the Devel
mailing list