Re: [RFC][PATCH 2/5] [PATCH 2/5] proc: introduce /proc/PID/idle_bitmap

On 09/01/2018 04:28 AM, Fengguang Wu wrote: > To walk 1TB memory of 4k active pages, it costs 2s vs 15s system > time to scan the per-task/global idle bitmaps. To me, that says this interface simply won't work on large systems. 2s and 15s are both simply unacceptably long. > OTOH, the per-task i

Re: [RFC][PATCH 2/5] [PATCH 2/5] proc: introduce /proc/PID/idle_bitmap

On Sat, Sep 01, 2018 at 07:28:20PM +0800, Fengguang Wu wrote: > diff --git a/fs/proc/internal.h b/fs/proc/internal.h > index da3dbfa09e79..732a502acc27 100644 > --- a/fs/proc/internal.h > +++ b/fs/proc/internal.h > @@ -305,6 +305,7 @@ extern const struct file_operations > proc_pid_smaps_rollup_ope

[RFC][PATCH 2/5] [PATCH 2/5] proc: introduce /proc/PID/idle_bitmap

This will be similar to /sys/kernel/mm/page_idle/bitmap documented in Documentation/admin-guide/mm/idle_page_tracking.rst, however indexed by process virtual address. When using the global PFN indexed idle bitmap, we find 2 kind of overheads: - to track a task's working set, Brendan Gregg end up