Comment 37 for bug 1861359

On Mon, Feb 10, 2020 at 07:53:20AM -0000, Andrea Righi wrote:
> OK, so we know that it's not related to the memory cgroup subsystem.

But this is a good instinct. It does seem to happen when eg firefox or git
is in heavy memory use, not the system as a whole.

> Another reason of such unexpected swapping activity could be due to
> memory compaction code that is triggering some direct memory reclaim and
> forcing to swap out pages.
>
> What do you have in /proc/sys/vm/swappiness? Could you try to set that
> to 0 (kernel prefers to drop file-backed pages instead of swapping out
> anonymous pages) and see if the swap out activity is still happening?

$ cat /proc/sys/vm/swappiness
60

I'll try 0.

> Moreover, we can do some more advanced tracing of the direct memory reclaim code, for example:
> $ sudo kprobe-perf -s 'p:shrink_node'
>

This doesn't work:

# kprobe-perf -s 'p:shrink_node'
ERROR: func shrink_node not in /sys/kernel/debug/tracing/available_filter_functions.
Either it doesn't exist, or, it might be unsafe to kprobe. Exiting. Use -F to override.
# ls -l /sys/kernel/debug/tracing/available_filter_functions
ls: cannot access '/sys/kernel/debug/tracing/available_filter_functions': No such file or directory
# find /sys/kernel/debug/tracing
/sys/kernel/debug/tracing
/sys/kernel/debug/tracing/instances
/sys/kernel/debug/tracing/trace_stat
/sys/kernel/debug/tracing/per_cpu
/sys/kernel/debug/tracing/per_cpu/cpu7
/sys/kernel/debug/tracing/per_cpu/cpu6
/sys/kernel/debug/tracing/per_cpu/cpu5
/sys/kernel/debug/tracing/per_cpu/cpu4
/sys/kernel/debug/tracing/per_cpu/cpu3
/sys/kernel/debug/tracing/per_cpu/cpu2
/sys/kernel/debug/tracing/per_cpu/cpu1
/sys/kernel/debug/tracing/per_cpu/cpu0
/sys/kernel/debug/tracing/options
#

Thanks