Hi, Leaving aside all the hierarchy walk discussion, I tried to come up with a series that concentrates only the basic points of it all. I think we benefit from having it independently of the rest of the work: in general (comments on the specifics welcome) cpuacct is made more naturally integrated with the scheduler, and the statistics it collects are now exactly the same as the system-wide ones for the root cgroup case. I think quite similar can be done with cpuusage by associating the root cgroup with the main runqueues, but this here is just me scratching my own itches - we can do it later. Please let me know if this is acceptable. Glauber Costa (4): Change cpustat fields to an array. Reuse cgroup's parent pointer Move part of cpuacct code cpuacct.stat: re-use scheduler statistics for the root cgroup arch/s390/appldata/appldata_os.c | 16 +- arch/x86/include/asm/i387.h | 2 +- drivers/cpufreq/cpufreq_conservative.c | 38 ++--- drivers/cpufreq/cpufreq_ondemand.c | 38 ++--- drivers/macintosh/rack-meter.c | 8 +- fs/proc/stat.c | 63 ++++---- fs/proc/uptime.c | 4 +- include/linux/kernel_stat.h | 36 +++-- kernel/sched.c | 270 ++++++++++++++++++-------------- 9 files changed, 252 insertions(+), 223 deletions(-) -- 1.7.6.4 -- To unsubscribe from this list: send the line "unsubscribe cgroups" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
