| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <20191031100631.GC19197@e108754-lin>
Date: Thu, 31 Oct 2019 10:07:43 +0000
From: Ionela Voinescu <ionela.voinescu@....com>
To: Daniel Lezcano <daniel.lezcano@...aro.org>
Cc: Thara Gopinath <thara.gopinath@...aro.org>, mingo@...hat.com,
peterz@...radead.org, vincent.guittot@...aro.org,
rui.zhang@...el.com, edubezval@...il.com, qperret@...gle.com,
linux-kernel@...r.kernel.org, amit.kachhap@...il.com,
javi.merino@...nel.org
Subject: Re: [Patch v4 0/6] Introduce Thermal Pressure
Hi Daniel,
On Tuesday 29 Oct 2019 at 16:34:11 (+0100), Daniel Lezcano wrote:
> Hi Thara,
>
> On 22/10/2019 22:34, Thara Gopinath wrote:
> > Thermal governors can respond to an overheat event of a cpu by
> > capping the cpu's maximum possible frequency. This in turn
> > means that the maximum available compute capacity of the
> > cpu is restricted. But today in the kernel, task scheduler is
> > not notified of capping of maximum frequency of a cpu.
> > In other words, scheduler is unware of maximum capacity
> > restrictions placed on a cpu due to thermal activity.
> > This patch series attempts to address this issue.
> > The benefits identified are better task placement among available
> > cpus in event of overheating which in turn leads to better
> > performance numbers.
> >
> > The reduction in the maximum possible capacity of a cpu due to a
> > thermal event can be considered as thermal pressure. Instantaneous
> > thermal pressure is hard to record and can sometime be erroneous
> > as there can be mismatch between the actual capping of capacity
> > and scheduler recording it. Thus solution is to have a weighted
> > average per cpu value for thermal pressure over time.
> > The weight reflects the amount of time the cpu has spent at a
> > capped maximum frequency. Since thermal pressure is recorded as
> > an average, it must be decayed periodically. Exisiting algorithm
> > in the kernel scheduler pelt framework is re-used to calculate
> > the weighted average. This patch series also defines a sysctl
> > inerface to allow for a configurable decay period.
> >
> > Regarding testing, basic build, boot and sanity testing have been
> > performed on db845c platform with debian file system.
> > Further, dhrystone and hackbench tests have been
> > run with the thermal pressure algorithm. During testing, due to
> > constraints of step wise governor in dealing with big little systems,
> > trip point 0 temperature was made assymetric between cpus in little
> > cluster and big cluster; the idea being that
> > big core will heat up and cpu cooling device will throttle the
> > frequency of the big cores faster, there by limiting the maximum available
> > capacity and the scheduler will spread out tasks to little cores as well.
> >
> > Test Results
> >
> > Hackbench: 1 group , 30000 loops, 10 runs
> > Result SD
> > (Secs) (% of mean)
> > No Thermal Pressure 14.03 2.69%
> > Thermal Pressure PELT Algo. Decay : 32 ms 13.29 0.56%
> > Thermal Pressure PELT Algo. Decay : 64 ms 12.57 1.56%
> > Thermal Pressure PELT Algo. Decay : 128 ms 12.71 1.04%
> > Thermal Pressure PELT Algo. Decay : 256 ms 12.29 1.42%
> > Thermal Pressure PELT Algo. Decay : 512 ms 12.42 1.15%
> >
> > Dhrystone Run Time : 20 threads, 3000 MLOOPS
> > Result SD
> > (Secs) (% of mean)
> > No Thermal Pressure 9.452 4.49%
> > Thermal Pressure PELT Algo. Decay : 32 ms 8.793 5.30%
> > Thermal Pressure PELT Algo. Decay : 64 ms 8.981 5.29%
> > Thermal Pressure PELT Algo. Decay : 128 ms 8.647 6.62%
> > Thermal Pressure PELT Algo. Decay : 256 ms 8.774 6.45%
> > Thermal Pressure PELT Algo. Decay : 512 ms 8.603 5.41%
>
> I took the opportunity to try glmark2 on the db845c platform with the
> default decay and got the following glmark2 scores:
>
> Without thermal pressure:
>
> # NumSamples = 9; Min = 790.00; Max = 805.00
> # Mean = 794.888889; Variance = 19.209877; SD = 4.382907; Median 794.000000
> # each ∎ represents a count of 1
> 790.0000 - 791.5000 [ 2]: ∎∎
> 791.5000 - 793.0000 [ 2]: ∎∎
> 793.0000 - 794.5000 [ 2]: ∎∎
> 794.5000 - 796.0000 [ 1]: ∎
> 796.0000 - 797.5000 [ 0]:
> 797.5000 - 799.0000 [ 1]: ∎
> 799.0000 - 800.5000 [ 0]:
> 800.5000 - 802.0000 [ 0]:
> 802.0000 - 803.5000 [ 0]:
> 803.5000 - 805.0000 [ 1]: ∎
>
>
> With thermal pressure:
>
> # NumSamples = 9; Min = 933.00; Max = 960.00
> # Mean = 940.777778; Variance = 64.172840; SD = 8.010795; Median 937.000000
> # each ∎ represents a count of 1
> 933.0000 - 935.7000 [ 3]: ∎∎∎
> 935.7000 - 938.4000 [ 2]: ∎∎
> 938.4000 - 941.1000 [ 2]: ∎∎
> 941.1000 - 943.8000 [ 0]:
> 943.8000 - 946.5000 [ 0]:
> 946.5000 - 949.2000 [ 1]: ∎
> 949.2000 - 951.9000 [ 0]:
> 951.9000 - 954.6000 [ 0]:
> 954.6000 - 957.3000 [ 0]:
> 957.3000 - 960.0000 [ 1]: ∎
>
Interesting! If I'm interpreting these correctly there seems to be
significant improvement when applying thermal pressure.
I'm not familiar with glmark2, can you tell me more about the process
and the work that the benchmark does? I assume this is a GPU benchmark,
but not knowing more about it I fail to see the correlation between
applying thermal pressure to CPU capacities and the improvement of GPU
performance.
Do you happen to know more about the behaviour that resulted in these
benchmark scores?
Thanks,
Ionela.
>
>
> --
> <http://www.linaro.org/> Linaro.org │ Open source software for ARM SoCs
>
> Follow Linaro: <http://www.facebook.com/pages/Linaro> Facebook |
> <http://twitter.com/#!/linaroorg> Twitter |
> <http://www.linaro.org/linaro-blog/> Blog
>
Powered by blists - more mailing lists