FORUMS
Remove All Ads from XDA

 View Poll Results: What CPU governor do you use?

Ondemand/OndemandX
 
233 Vote(s)
7.05%
Interactive/InteractiveX
 
1,053 Vote(s)
31.87%
Conservative/Ktoonservative
 
121 Vote(s)
3.66%
Performance
 
139 Vote(s)
4.21%
Powersave
 
81 Vote(s)
2.45%
ZZMoove
 
157 Vote(s)
4.75%
Intelliactive/Intellidemand
 
98 Vote(s)
2.97%
Hyper
 
23 Vote(s)
0.70%
PegasusQ
 
45 Vote(s)
1.36%
Impulse
 
122 Vote(s)
3.69%
Smartmax/Performance May Cry
 
62 Vote(s)
1.88%
ElementalX
 
193 Vote(s)
5.84%
Alucard
 
225 Vote(s)
6.81%
Yankactive/YanksusQ/Yankdemand
 
26 Vote(s)
0.79%
Darkness/Nightmare/Lightning
 
92 Vote(s)
2.78%
Blu_active
 
247 Vote(s)
7.48%
SmartassV2/SmartassH3
 
67 Vote(s)
2.03%
Other
 
320 Vote(s)
9.69%

[REF][GUIDE]Most up to date guide on CPU governors, I/O schedulers and more!

4,021 posts
Thanks Meter: 8,377
 
By Saber, Recognized Contributor on 8th March 2015, 05:53 AM
Post Reply Email Thread
7th October 2016, 10:29 PM |#491  
Saber's Avatar
OP Recognized Contributor
Flag Melbourne
Thanks Meter: 8,377
 
Donate to Me
More
Quote:
Originally Posted by aryaazimi72

Hey... @Saber
Got Three things:
1.What's "Gentle fair sleepers" whats the advantages of disable/enabling it?
2. What is "Kernel Entropy" and the options on it: read/write wakeup treshold?
3. Reminding u to check "multi path tcp" things out...

Thanks a lotttt!!!

1. It treats sleeping and waiting tasks as if they were run in a queue. Most developers leave it disabled as it can increase responsiveness of your device. I don't really see a good reason for enabling it.
2. Entropy is used for cryptography and other 'random' data which is used by your phone's hardware. Read and write wakeup thresholds don't really need to be touched at all as usually the default settings are fine, but increasing either will leave you with more free space in entropy pool (If you find that entropy space is always running empty).
3. Been busy lately with studies and coursework so I haven't got around to researching on this

Saber
The Following 4 Users Say Thank You to Saber For This Useful Post: [ View ] Gift Saber Ad-Free
8th October 2016, 07:33 PM |#492  
Member
Flag London
Thanks Meter: 16
 
More
It would be amazing if you could make a guide for interactive governor tunables.... @Saber
When you have time.... no rush...
Thank you for everything.... all your guides are realy helpful to me.... tnx
Attached Thumbnails
Click image for larger version

Name:	Screenshot_20161008-215551.png
Views:	674
Size:	187.2 KB
ID:	3898076   Click image for larger version

Name:	Screenshot_20161008-215608.png
Views:	672
Size:	192.4 KB
ID:	3898078  
8th October 2016, 08:32 PM |#493  
DB126's Avatar
Senior Member
Thanks Meter: 9,342
 
More
Quote:
Originally Posted by aryaazimi72

It would be amazing if you could make a guide for interactive governor tunables.... @Saber
When you have time.... no rush...
Thank you for everything.... all your guides are realy helpful to me.... tnx

Some key tunables are covered in the OP of this thread. Makes a measurable difference with appropriate values for your specific device. Don't need to tweak much; keep it simple and test every change before moving on.
The Following 2 Users Say Thank You to DB126 For This Useful Post: [ View ] Gift DB126 Ad-Free
9th October 2016, 12:24 AM |#494  
Saber's Avatar
OP Recognized Contributor
Flag Melbourne
Thanks Meter: 8,377
 
Donate to Me
More
Quote:
Originally Posted by aryaazimi72

It would be amazing if you could make a guide for interactive governor tunables.... @Saber
When you have time.... no rush...
Thank you for everything.... all your guides are realy helpful to me.... tnx

The guide already covers tunables briefly in the governor tunable section.

For profiles and in-depth tuning, you can look at this guide here:

[GUIDE] Advanced Interactive Governor Tweaks; Buttery smooth and insane battery life!

Remember that when suggesting ideas, please note that this guide is just to give the general details, it will not be a in-depth tuning guide for specific configurations (which will make the guide even bigger).

Saber
The Following User Says Thank You to Saber For This Useful Post: [ View ] Gift Saber Ad-Free
9th October 2016, 06:40 AM |#495  
tet-bundy's Avatar
Senior Member
Thanks Meter: 225
 
More
Quote:
Originally Posted by aryaazimi72

Hey... @Saber
Got Three things:
1.What's "Gentle fair sleepers" whats the advantages of disable/enabling it?
2. What is "Kernel Entropy" and the options on it: read/write wakeup treshold?
3. Reminding u to check "multi path tcp" things out...

Thanks a lotttt!!!

I'd also be interested in a short rundown on entropy in op
I read somewhere, that increasing values may have positive impact on performance. I guess that depends heavily on the kind of apps u're using, as not every app needs random data...

Gesendet von meinem du_onyx mit Tapatalk
9th October 2016, 09:54 AM |#496  
tet-bundy's Avatar
Senior Member
Thanks Meter: 225
 
More
Okay I know we have clarified that read ahead cache does depend on storage size instead of i/o scheduler. But it does depend on quality of the sd card also right? (not like it was an issue for me as i only use class 10. just curious. also in OP u pointed out that increasing read-ahead won't have any positive effect from a certain point on. can it actually do harm to performance if setting higher values? ie for latency u mentioned?
thanks
9th October 2016, 11:30 AM |#497  
Saber's Avatar
OP Recognized Contributor
Flag Melbourne
Thanks Meter: 8,377
 
Donate to Me
More
Quote:
Originally Posted by tet-bundy

Okay I know we have clarified that read ahead cache does depend on storage size instead of i/o scheduler. But it does depend on quality of the sd card also right? (not like it was an issue for me as i only use class 10. just curious. also in OP u pointed out that increasing read-ahead won't have any positive effect from a certain point on. can it actually do harm to performance if setting higher values? ie for latency u mentioned?
thanks

For higher quality sd cards, you may be able to get slightly more performance by increase buffer values further than what I recommend. However, what you might find is that having 'enough' buffer is much better than having too much. Trial and error may be one way to find optimal values. Latency does increase as larger buffer sizes are used. This is extremely important to be kept controlled when recording in 4k or any read/write intensive tasks. Apologies if my explanations are too vague.

Saber
The Following User Says Thank You to Saber For This Useful Post: [ View ] Gift Saber Ad-Free
9th October 2016, 12:43 PM |#498  
DB126's Avatar
Senior Member
Thanks Meter: 9,342
 
More
Quote:
Originally Posted by tet-bundy

Okay I know we have clarified that read ahead cache does depend on storage size instead of i/o scheduler. But it does depend on quality of the sd card also right? (not like it was an issue for me as i only use class 10. just curious. also in OP u pointed out that increasing read-ahead won't have any positive effect from a certain point on. can it actually do harm to performance if setting higher values? ie for latency u mentioned?
thanks

In addition to what has been written already:
- higher rates SD cards often excel at sequential IO at the expense of random operations
- class 6 cards will often stomp class 10 working with many small files
- excessive cache sizes, especially on internal storage (system partition) absolutely hurts performance as evidenced by benchmarking
- conservative tweaks followed by testing is best the course of action...or just leave it alone.
The Following User Says Thank You to DB126 For This Useful Post: [ View ] Gift DB126 Ad-Free
9th October 2016, 05:47 PM |#499  
christianpeso's Avatar
Senior Member
Thanks Meter: 535
 
More
Quote:
Originally Posted by Suigintou


Any information about Kraken kernel?

I 2nd this.

Looking for more info on Kraken cpu governor as the kernel I am currently using has it as the default. Any info would be greatly appreciated.
9th October 2016, 07:36 PM |#500  
tet-bundy's Avatar
Senior Member
Thanks Meter: 225
 
More
Quote:
Originally Posted by Saber

For higher quality sd cards, you may be able to get slightly more performance by increase buffer values further than what I recommend. However, what you might find is that having 'enough' buffer is much better than having too much. Trial and error may be one way to find optimal values. Latency does increase as larger buffer sizes are used. This is extremely important to be kept controlled when recording in 4k or any read/write intensive tasks. Apologies if my explanations are too vague.

Saber

Quote:
Originally Posted by Davey126

In addition to what has been written already:
- higher rates SD cards often excel at sequential IO at the expense of random operations
- class 6 cards will often stomp class 10 working with many small files
- excessive cache sizes, especially on internal storage (system partition) absolutely hurts performance as evidenced by benchmarking
- conservative tweaks followed by testing is best the course of action...or just leave it alone.

I'm totally fine with ur explanations. It's not like i intend to get more out of my device as hardware gives. I'm rather curious about this kind of stuff so I'm asking haha
Thanks for ur replies

Gesendet von meinem du_onyx mit Tapatalk
9th October 2016, 10:06 PM |#501  
Saber's Avatar
OP Recognized Contributor
Flag Melbourne
Thanks Meter: 8,377
 
Donate to Me
More
Quote:
Originally Posted by christianpeso

I 2nd this.

Looking for more info on Kraken cpu governor as the kernel I am currently using has it as the default. Any info would be greatly appreciated.

Will be looking at this soon.

Saber
The Following User Says Thank You to Saber For This Useful Post: [ View ] Gift Saber Ad-Free
Post Reply Subscribe to Thread

Tags
cpu governor, graphs, gsstudios, guide, io scheduler

Guest Quick Reply (no urls or BBcode)
Message:
Previous Thread Next Thread
Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes