Re: [PATCH 3/3] sched/fair: add tracepoints for cfs throttle

On Thu, May 24, 2018 at 12:11 AM, Peter Zijlstra wrote: > > The problem with tracepoints is that they can become ABI and you cannot > change them without breaking tools. This is a crap situation and I'm fed > up with it. Yeah, I once used perf_event_open() to parse the

Re: [PATCH 3/3] sched/fair: add tracepoints for cfs throttle

On Thu, May 24, 2018 at 12:11 AM, Peter Zijlstra wrote: > > The problem with tracepoints is that they can become ABI and you cannot > change them without breaking tools. This is a crap situation and I'm fed > up with it. Yeah, I once used perf_event_open() to parse the trace events, I understand

Re: [PATCH 3/3] sched/fair: add tracepoints for cfs throttle

On Wed, May 23, 2018 at 09:40:47PM -0700, Cong Wang wrote: > On Wed, May 23, 2018 at 2:09 AM, Peter Zijlstra wrote: > > Yeah, don't worry. I hate tracepoints, I regret the existing ones, no > > new ones will happen. > > No matter you hate them or not, they are useful: No

Re: [PATCH 3/3] sched/fair: add tracepoints for cfs throttle

On Wed, May 23, 2018 at 09:40:47PM -0700, Cong Wang wrote: > On Wed, May 23, 2018 at 2:09 AM, Peter Zijlstra wrote: > > Yeah, don't worry. I hate tracepoints, I regret the existing ones, no > > new ones will happen. > > No matter you hate them or not, they are useful: No argument there. >

Re: [PATCH 3/3] sched/fair: add tracepoints for cfs throttle

On Wed, May 23, 2018 at 2:09 AM, Peter Zijlstra wrote: > On Mon, May 21, 2018 at 11:35:38PM -0700, Ivan Babrou wrote: >> > + TP_printk("path=%s cpu=%d runtime_remaining=%lld", >> > __get_str(cfs_path), >> > + __entry->cpu, __entry->runtime_remaining) >>

Re: [PATCH 3/3] sched/fair: add tracepoints for cfs throttle

On Wed, May 23, 2018 at 2:09 AM, Peter Zijlstra wrote: > On Mon, May 21, 2018 at 11:35:38PM -0700, Ivan Babrou wrote: >> > + TP_printk("path=%s cpu=%d runtime_remaining=%lld", >> > __get_str(cfs_path), >> > + __entry->cpu, __entry->runtime_remaining) >> > >> >> Can you add

Re: [PATCH 3/3] sched/fair: add tracepoints for cfs throttle

On Mon, May 21, 2018 at 11:35:38PM -0700, Ivan Babrou wrote: > > + TP_printk("path=%s cpu=%d runtime_remaining=%lld", > > __get_str(cfs_path), > > + __entry->cpu, __entry->runtime_remaining) > > > > Can you add "[ns]" as the unit to runtime_remaining? We have it in >

Re: [PATCH 3/3] sched/fair: add tracepoints for cfs throttle

On Mon, May 21, 2018 at 11:35:38PM -0700, Ivan Babrou wrote: > > + TP_printk("path=%s cpu=%d runtime_remaining=%lld", > > __get_str(cfs_path), > > + __entry->cpu, __entry->runtime_remaining) > > > > Can you add "[ns]" as the unit to runtime_remaining? We have it in >

Re: [PATCH 3/3] sched/fair: add tracepoints for cfs throttle

On Mon, May 21, 2018 at 11:35 PM, Ivan Babrou wrote: > > > On Mon, May 21, 2018 at 11:20 PM Cong Wang wrote: >> >> + >> + TP_printk("path=%s cpu=%d runtime_remaining=%lld", >> __get_str(cfs_path), >> + __entry->cpu,

Re: [PATCH 3/3] sched/fair: add tracepoints for cfs throttle

On Mon, May 21, 2018 at 11:35 PM, Ivan Babrou wrote: > > > On Mon, May 21, 2018 at 11:20 PM Cong Wang wrote: >> >> + >> + TP_printk("path=%s cpu=%d runtime_remaining=%lld", >> __get_str(cfs_path), >> + __entry->cpu, __entry->runtime_remaining) > > > Can you add "[ns]" as

[PATCH 3/3] sched/fair: add tracepoints for cfs throttle

This patch introduces two sched tracepoints to track CFS throttle and unthrottle. The use case is to measure how long each throttle is and to track when a CPU cgroup is throttled and unthrottled. Sample output: cfs-722 [000] dN.351.477702: sched_cfs_throttle: path=/test cpu=0

[PATCH 3/3] sched/fair: add tracepoints for cfs throttle

This patch introduces two sched tracepoints to track CFS throttle and unthrottle. The use case is to measure how long each throttle is and to track when a CPU cgroup is throttled and unthrottled. Sample output: cfs-722 [000] dN.351.477702: sched_cfs_throttle: path=/test cpu=0