Hello, This patch set introduces deferrable functionality for hrtimers and add ability to use it from timerfd_create syscall. It means user-space processes are able to not wake up the system from cpuidle state in case of periodic job. And it's the main goal of this patch set for us and for team sent such patch set before. I hope it will find users in kernel space too, cause it could avoid timer_list drawbacks, such as worst case when many,many timers are gonna to be expired.
Kernel already has deferrable timer funcionality based on timer_list. But deferrable functionality inside hrtimers makes timerfd implemenation trivial for such case. It's based on github.com/torvalds/linux.git git repository on top of e95003c3f9ccbfa7ab9d265e6eb703ee2fa4cfe7 Alexey Perevalov (3): Replace ternary operator to macro tracing/trivial: Add CLOCK_BOOTIME and CLOCK_TAI for human readable clockid trace tracing/trivial: Add CLOCK_*_DEFERRABLE for tracing clockids Anton Vorontsov (2): timerfd: Move repeated logic into timerfd_rearm() timerfd: Add support for deferrable timers Thomas Gleixner (1): hrtimer: Add support for deferrable timer into the hrtimer fs/timerfd.c | 47 ++++++++++++++++---------------- include/linux/hrtimer.h | 3 ++ include/trace/events/timer.h | 25 ++++++++++++++--- include/uapi/linux/time.h | 3 ++ kernel/hrtimer.c | 62 +++++++++++++++++++++++++++++++++--------- 5 files changed, 100 insertions(+), 40 deletions(-) -- 1.7.9.5 -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/