All of lore.kernel.org
 help / color / mirror / Atom feed
From: Thomas Renninger <trenn@suse.de>
To: Viresh Kumar <viresh.kumar@linaro.org>
Cc: linux-pm@vger.kernel.org, jacob.jun.pan@linux.intel.com
Subject: Re: Off-topic therml_stats - trans_table - File too large (exceeding static page alloc)
Date: Thu, 12 Nov 2020 00:13:00 +0100	[thread overview]
Message-ID: <4549593.l9KlQ6ThGW@c100> (raw)
In-Reply-To: <20201111103018.6xixvclldfw4m3ok@vireshk-i7>

Am Mittwoch, 11. November 2020, 11:30:18 CET schrieb Viresh Kumar:
> On 11-11-20, 09:42, Thomas Renninger wrote:
> > Hi,
> > 
> > sorry for high-jacking this thread, it is at least related and afaik you
> > are deeper involved in this:
..
> > and when it seem to get interesting (device 4 and 5), I get:
> How many total states are there ? Must be really large.
..

Ah, yes. The first 4 devices which have 3 thermal states seem to be connected to each CPU
via ACPI:
/sys/devices/virtual/thermal/cooling_device0/device -> ../../../LNXSYSTM:00/LNXCPU:00

The latter two, device 4+5 do not have this "device" sysfs link at all and cooling_device4 has 50,
cooling_device5 20 states.

All cooling devices max_state:
cat /sys/devices/virtual/thermal/cooling_device*/max_state 
3
3
3
3
50
20

cur_state is 0 by default for all.
Doing something like (not sure it was a 2):
echo 2 >/sys/devices/virtual/thermal/cooling_device4/cur_state
ends up in:
Nov 11 23:33:27 c100 kernel: NOHZ: local_softirq_pending 80
Nov 11 23:33:27 c100 kernel: NOHZ: local_softirq_pending 80
Nov 11 23:33:21 c100 kernel: intel_powerclamp: Start idle injection to reduce power
in kernel logs.

But echoing this static cur_state number in, seem to trigger dynamic adjusting
and the number gets up and down by itself:

while true;do cat /sys/devices/virtual/thermal/cooling_device4/cur_state;sleep 1;done
86
83
89
90
90
92
91
^C
c100:/home/trenn # cat /sys/devices/virtual/thermal/cooling_device4/max_state
50

Disabling works with setting cur_state to 0, but cur_state is -1 afterwards ?!?:
echo 0 >/sys/devices/virtual/thermal/cooling_device4/cur_state
cat /sys/devices/virtual/thermal/cooling_device4/cur_state 
-1

echo 1 >/sys/devices/virtual/thermal/cooling_device4/cur_state
c100:/home/trenn # cat /sys/devices/virtual/thermal/cooling_device4/cur_state 
86

It looks like any number echoed into the cur_state device, even above max_state enables
dynamic power adjusting:
echo 10100000 >/sys/devices/virtual/thermal/cooling_device4/cur_state
-> no error.

ACPI related CPU cooling devices 0-3 work as one would expect:
3 states (max_state) one can set them and trying to write something into cur_state
which is above max_state ends up in:
echo 4 >/sys/devices/virtual/thermal/cooling_device2/cur_state
bash: echo: Schreibfehler: Invalid argument.

Hardware:
Dell, Latitude E7470
cpu family      : 6
model           : 78
model name      : Intel(R) Core(TM) i7-6600U CPU @ 2.60GHz
stepping        : 3
microcode       : 0xd6


Adding Jacob who seem to be involved into powerclamp driver.
Maybe he can shed some light on this...

Thanks,

    Thomas




      reply	other threads:[~2020-11-12  1:33 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-11-10 11:07 [PATCH] cpufreq: stats: Switch to ktime and msec instead of jiffies and usertime Viresh Kumar
2020-11-10 11:36 ` Lukasz Luba
2020-11-11  5:14   ` Viresh Kumar
2020-11-10 12:53 ` Thomas Renninger
2020-11-11  5:13   ` Viresh Kumar
2020-11-11  8:13     ` Thomas Renninger
2020-11-11  9:51       ` Viresh Kumar
2020-11-10 12:59 ` Rafael J. Wysocki
2020-11-11  5:28   ` Viresh Kumar
2020-11-11  8:42 ` Off-topic therml_stats - trans_table - File too large (exceeding static page alloc) Thomas Renninger
2020-11-11 10:30   ` Viresh Kumar
2020-11-11 23:13     ` Thomas Renninger [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=4549593.l9KlQ6ThGW@c100 \
    --to=trenn@suse.de \
    --cc=jacob.jun.pan@linux.intel.com \
    --cc=linux-pm@vger.kernel.org \
    --cc=viresh.kumar@linaro.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.