Overclock.net banner

[Linux Mint 18 Sarah]: a couple of questions

6K views 72 replies 12 participants last post by  LostParticle 
#1 ·
Hello,

Approximately five days ago I have clean installed Linux Mint 18 Sarah, Cinnamon, on the Samsung SSD shown in my sig_rig, and I run it on this system. It runs together with Windows 10 Pro, installed on my other SSD, and I use F11 (boot menu) to load Mint, whenever I want to. Otherwise my computer boots directly into Windows.

The installation went very well and I have started using it without any problems, especially after applying the majority of the suggestions from this guide.

I have a couple of questions, though:

1) Each time I shutdown or restart my computer a black screen appears which shows the following:
Linux Mint 18 Sarah me-desktop tty1
me-login:

me is the user name I use.

This screen also appears each time I boot into Mint but, whereas during booting it lasts for a fraction of a second, when I shutdown or restart it displays for approx. one (1) second and then my PC continues.

- Is there any way to stop this?

2) In Windows I am always using HWiNFO64 which, according to my personal opinion, is the best monitoring tool available! I am always running an overclocked system so I need to monitor voltages, temperatures and other values. Is there a similar monitoring tool on Mint? I have found and installed Psensor but it does not monitor Voltages, it does not monitor my Corsair's H110 fan speeds, and it does not even remember the position on my second monitor that I place it! It always open in the middle of my primary monitor. I want another monitoring tool so which is the best monitoring tool on Linux Mint?

3) When I shutdown my computer using the power button on my chassis, a pop-up window comes up asking me if I want to Suspend, Restart or Shut down. Is there any way to prevent this? In Windows when I press the button my system shuts down right away (no window), in less than two (2) seconds.

Thank you!
smile.gif
 
See less See more
1
#2 ·
Quote:
Originally Posted by LostParticle View Post

...

1) Each time I shutdown or restart my computer a black screen appears which shows the following:
Linux Mint 18 Sarah me-desktop tty1
me-login:

me is the user name I use.

This screen also appears each time I boot into Mint but, whereas during booting it lasts for a fraction of a second, when I shutdown or restart it displays for approx. one (1) second and then my PC continues.

- Is there any way to stop this?

...
After installing the latest NVIDIA driver this "issue" has stopped! I have rebooted more than 20 times and nothing appears during shutdown or booting, anymore. Some text appears there during shutdown/restart but it is so fast that I cannot even see it. I'm glad this is fixed
smile.gif


Anyone on the other two things I have asked?

-- What is the best monitoring tool for Linux Mint?
-- Is there a way to shutdown my computer right away, without any dialogue box, when I press the button on my chassis?

Thank you.
 
#3 ·
I-Nex is the closest thing to something like CPUZ(it is more detailed than CPUZ though) I've used on Linux. I don't know if that's good enough for you.

As for the power button issue you'll need to see what the power options for Cinnamon are because that's something handled by the DE/WM/GUI.
 
#4 ·
Quote:
Originally Posted by Rookie1337 View Post

I-Nex is the closest thing to something like CPUZ(it is more detailed than CPUZ though) I've used on Linux. I don't know if that's good enough for you.

As for the power button issue you'll need to see what the power options for Cinnamon are because that's something handled by the DE/WM/GUI.
Hello, thanks for your reply, at last one person has seen my thread!

I will try that tool, I-Nex, after Googling it first. Thanks!

To give you an idea of what I am looking for, here is a screenshot from HWiNFO64, my permanent monitoring tool on Windows:



Taken a few seconds ago, this screenshot gives you, for example, a complete and accurate snapshot of my system! And it permits to me to monitor everything (important) at a glance. Is there something similar for Linux Mint?

Thank you!

I'll research for the Power button, as well. It must be something in the Settings.
 
#5 ·
Quote:
Originally Posted by LostParticle View Post

...

3) When I shutdown my computer using the power button on my chassis, a pop-up window comes up asking me if I want to Suspend, Restart or Shut down. Is there any way to prevent this? In Windows when I press the button my system shuts down right away (no window), in less than two (2) seconds.
Hello again

I have resolved this No 3 inquiry with this setting:



I have also installed I-Nex. The problem with I-Nex is not only that it does not monitor the voltages in real-time but it also does not show the Core clock (frequency) in real time... In the following screenshot my CPU0 clock remained at 4700 MHz for as long as I-Nex was running. The truth is that my CPU0 clock was fluctuating during all this time, of course.



So, I ask again: is there any good system monitoring utility for Linux Mint, like HWiNFO64 is for Windows?

Finally, today I got a new "issue". I have installed the latest kernel and uninstalled all the others. I am just testing, experimenting, with this installation to get the hang of it. This is what my kernels look like now:



The problem now is that each time I boot into Mint I get a screen of GNU GRUB version 2.02, if I recall correctly, offering me some options like MemTest, Windows Loader etc, and giving me 10 seconds to select anyone of those. Then my system boots into Mint. How can I prevent this menu from appearing? How can I disable it or, even better, to press a button when I want it to appear, otherwise my system to boot directly into Mint?

Here is how my grub looks like. I have not touched it, I have not altered it, at all!



Thank you! I hope someone will reply!
 
#7 ·
Quote:
Originally Posted by cones View Post

Not sure if that is the correct file you need to edit, /etc/default/grub.cfg. You would edit the timeout options and then after editing run "update-grub".

For your monitoring I'm not sure you'll find a program that compares to hwinfo. The closet you can probably get would be a very custom conky.
Hello, thank you for your reply!

When I give the command:

sudo leafpad /etc/default/grub.cfg
and then my password

an empty file called grub.cfg opens! It is empty, I do not see anything inside it!

?
 
#9 ·
Quote:
Originally Posted by cones View Post

Uhm not at a computer i think I missed a directory in that try /etc/default/grub/grub.cfg

Edit: The file you want to edit should look the same as the one you posted.
Thank you for your support but what you suggest to me does not work, I'm sorry...

I gave the new path you've suggested and I get the same result: a null file.

Here is how that directory looks like, in my system:



The only file looking like the one I'm supposed to edit is the grub.txt, so the one I have already posted.

Anyway, thank you.

PS: I've also entered that directory "grub.d" and it contains an irrelevant file (as far as I can tell).
 
#10 ·
Quote:
Originally Posted by LostParticle View Post

sudo leafpad /etc/default/grub.cfg
and then my password

an empty file called grub.cfg opens! It is empty, I do not see anything inside it!
Try:

Code:

Code:
gksudo leafpad /etc/default/grub
This will have the intended effect of the previously suggested command, with the main change being the name of the configuration file.
gksudo is suggested instead of sudo however is tangential to the problem, gksudo has the effect of sudo but is recommended for graphical applications such as leafpad due to greater "awareness" of environmental variables. Again this is tangential.

For determining the contents of a directory in the terminal use "ls". For example: "ls /etc/default", will list the contents of "/etc/default". Including files and directories. "ls" with no options will largely present the same information as a graphical file browser. Options for "ls" can be found by "man ls" the manual for ls.

When editing "/etc/default/grub" change GRUB_TIMEOUT to be equal to 0. Then "sudo update-grub"

If there are multiple kernels in the grub boot directory or operating systems detected this may be unsuccessful. If so change "GRUB_TIMEOUT=0.0" and "GRUB_HIDDEN_TIMEOUT=0.0".
Aside from the aforementioned configuration file, GRUB is also configured by the contents of the /etc/default/grub.d directory. When multiple kernels are detected a script "30_os-prober", disallows GRUB_TIMEOUT to be 0 and resets the value to 10 seconds. The idea being with a choice of kernels or operating systems to choose from the user would wish to choose which one to boot. Setting the values to decimals, alters the way this script interprets the values and should allow for GRUB to be hidden. When GRUB is hidden it can be accessed by depressing the SHIFT key on boot.

Also bare in mind a stable overclock in Windows is not necessarily stable in Linux but for system monitoring lm_sensors is a popular option, however I doubt it is comparable to HW Info.
 
#11 ·
Quote:
Originally Posted by vpex View Post

Try:

Code:

Code:
gksudo leafpad /etc/default/grub
This will have the intended effect of the previously suggested command, with the main change being the name of the configuration file.
gksudo is suggested instead of sudo however is tangential to the problem, gksudo has the effect of sudo but is recommended for graphical applications such as leafpad due to greater "awareness" of environmental variables. Again this is tangential.

For determining the contents of a directory in the terminal use "ls". For example: "ls /etc/default", will list the contents of "/etc/default". Including files and directories. "ls" with no options will largely present the same information as a graphical file browser. Options for "ls" can be found by "man ls" the manual for ls.

When editing "/etc/default/grub" change GRUB_TIMEOUT to be equal to 0. Then "sudo update-grub"

If there are multiple kernels in the grub boot directory or operating systems detected this may be unsuccessful. If so change "GRUB_TIMEOUT=0.0" and "GRUB_HIDDEN_TIMEOUT=0.0".
Aside from the aforementioned configuration file, GRUB is also configured by the contents of the /etc/default/grub.d directory. When multiple kernels are detected a script "30_os-prober", disallows GRUB_TIMEOUT to be 0 and resets the value to 10 seconds. The idea being with a choice of kernels or operating systems to choose from the user would wish to choose which one to boot. Setting the values to decimals, alters the way this script interprets the values and should allow for GRUB to be hidden. When GRUB is hidden it can be accessed by depressing the SHIFT key on boot.

Also bare in mind a stable overclock in Windows is not necessarily stable in Linux but for system monitoring lm_sensors is a popular option, however I doubt it is comparable to HW Info.
All right!! Thank you very much, +REP! It worked!

Specifically, your suggestion with the decimals worked. A single zero did not do the trick. So, now I am booting like I used to, before the kernel update: after pressing F11, in my dual boot system, I boot directly into Mint, without any screens and options. I have pressed the SHIFT button, the three times I've rebooted, but I've never managed to get the GRUB menu. But I do not care.
smile.gif


- How come and GRUB is still remembering the other kernels, since I have uninstalled them?

- Kind of a confusing screenshot:


It is very unfortunate that a good monitoring tool for Linux does not exist
frown.gif
frown.gif

I have heard about lm-sensors and tried to make them work on my system without any luck... It is for sure that I don't know how to make it work...

How are overclockers testing their Linux systems?
 
#13 ·
Quote:
Originally Posted by cones View Post

...

Did you first run sensors-detect and answer yes to everything then reboot and try sensors?
Yes, I have done that.

Here is what is shown, currently, on my system:


Now...What is the problem with this?

1) Plenty of arbitrary / erratic values! For example, VCore = +2.04V ?!
2) These values are static! I am having the terminal right now open in my second monitor and nothing is changing. Nothing is measured in real-time.

So, how does this work?

I don't know IF this matters but right now I am on my per-core OC of x50 x50 x49 x48, with Adaptive Core voltage = 1.4V, in the BIOS.

EDIT:
Okay, I have some interesting results, at last!
smile.gif


I have detected the sensors again, replied YES to everything, rebooted and then run Psensor, a utility I have already installed.
Here are the results:


Very nice, and I know how to modify and hide the erratic values in Psensor, so it's good! The only (serious) problem, though, is that Psensor does not show any voltage values...
What can I do about this?
 
#14 ·
Quote:
Originally Posted by LostParticle View Post

Yes, I have done that.

Here is what is shown, currently, on my system:


Now...What is the problem with this?

1) Plenty of arbitrary / erratic values! For example, VCore = +2.04V ?!
2) These values are static! I am having the terminal right now open in my second monitor and nothing is changing. Nothing is measured in real-time.

So, how does this work?

I don't know IF this matters but right now I am on my per-core OC of x50 x50 x49 x48, with Adaptive Core voltage = 1.4V, in the BIOS.

EDIT:
Okay, I have some interesting results, at last!
smile.gif


I have detected the sensors again, replied YES to everything, rebooted and then run Psensor, a utility I have already installed.
Here are the results:


Very nice, and I know how to modify and hide the erratic values in Psensor, so it's good! The only (serious) problem, though, is that Psensor does not show any voltage values...
What can I do about this?
Use the command:

Code:

Code:
watch -n 1 -d sensors
It will update every second and highlight the changes.
 
#15 ·
Thank you @Diffident, interesting idea. I have tried it. Unfortunately, it shows my VCore completely out of line, it shows it raising up to... 2.04V (lol), it does not report ANY other voltage - please, look at a screenshot of HWiNFO64 above to see the amount of V reported by my mobo - and finally, I cannot scroll down to view the second sensor.

Anyway, here is what I am currently using to monitor my system, a bit.

That is all I have managed so far. I still need to hide some values in Psensor and also reorder them, but it is fair enough. Oh, and Psensor does not / cannot remember its position (on my 2nd monitor), neither the way I set it. So I have to move it and maximize it on each boot...........

The huge problem is that I have not yet discovered how to monitor Frequencies (clocks) and Voltages...
 
#16 ·
System monitoring is unfortunately an area where Windows is greatly superior to Linux, sad to say. If you think it's bad on an Intel system, you should try doing it on a machine with an AMD A88X chipset. LM-sensors detects almost nothing, even using the newer Linux kernels in Mint 18. I had already tried that by using the custom 4.4.0 from Exton Linux in Mint 17.3; the kernel worked fine but still no sensor support.

If I want to monitor anything, I just boot Windows 10. Only way to do it.

I've upgraded both my desktop and my laptop to 18, but I found a serious regression in the 4.4.0-28 kernel on my laptop. The sleep function turns off the display but fails to put the CPU into a low-power state. I closed the lid on it and the laptop was running hot when I opened it the next day, and did it repeatedly in tests. The original 4.4.0-21 was OK, and so is 4.6.0-exton, which is what I'm running on both machines now (although you're on your own if you do that, the Exton kernels are unofficial). Both 4.4.0-28 and 4.4.0-31 from the official repositories seem to do a very poor job of CPU scheduling and power management in general (probably why they're "category 5" updates and not installed by default). On 4.4.0-31, my laptop ran at 3068 MHz about half the time, even when idling, and the desktop would run at 4200 (current OC speed) when doing almost nothing.
 
#17 ·
Quote:
Originally Posted by jsc1973 View Post

System monitoring is unfortunately an area where Windows is greatly superior to Linux, sad to say. If you think it's bad on an Intel system, you should try doing it on a machine with an AMD A88X chipset. LM-sensors detects almost nothing, even using the newer Linux kernels in Mint 18. I had already tried that by using the custom 4.4.0 from Exton Linux in Mint 17.3; the kernel worked fine but still no sensor support.

If I want to monitor anything, I just boot Windows 10. Only way to do it.

I've upgraded both my desktop and my laptop to 18, but I found a serious regression in the 4.4.0-28 kernel on my laptop. The sleep function turns off the display but fails to put the CPU into a low-power state. I closed the lid on it and the laptop was running hot when I opened it the next day, and did it repeatedly in tests. The original 4.4.0-21 was OK, and so is 4.6.0-exton, which is what I'm running on both machines now (although you're on your own if you do that, the Exton kernels are unofficial). Both 4.4.0-28 and 4.4.0-31 from the official repositories seem to do a very poor job of CPU scheduling and power management in general (probably why they're "category 5" updates and not installed by default). On 4.4.0-31, my laptop ran at 3068 MHz about half the time, even when idling, and the desktop would run at 4200 (current OC speed) when doing almost nothing.
Really-really sorry to hear all this, especially for the lack of appropriate monitoring utilities!
frown.gif

And I believe you because I keep Googling but I am not discovering anything...
frown.gif
frown.gif


- What are the Linux overclockers doing?! How are they testing and stressing and monitoring their systems?!

What you said about Power consumption alerted me so I have just checked this by observing my Kill A Watt. My entire system (see sig_rig) + my Modem-Router + my Hi-Fi Amplifier (currently powered down) is connected to it. I am using two monitors. My results:

Per-Core OC: x 50 x50 x 49 x 48, cache x 44, Adaptive VCore = 1.4 in the BIOS, ALL C-States Enabled.

Windows 10 Pro
Sleep = 14 W
Idle = 116 - 120W

Linux Mint 18, kernel 4.4.0-31
Sleep = 14 W
Idle = 116 - 125W

The only thing that I have observed is that on Linux I also had spikes up to 150W. I am not sure however what the OS (Linux) was doing. I had the System Monitor opened and it was showing 1-4% usage on one core, different each time. In both situations, on idle, I was not touching the mouse at all.

Not sure (at all) how "professional" my testing was, but at least I am sure that on Sleep my system behaves the same in both OSes.

PS_1: never had a laptop, never will own one.

PS_2: HOW were you able to see the frequency you mention, on your laptop?! I do not know any way to see my current frequency! How did you do it?
 
#18 ·
If 4.4.0-31 works for you, stick with it. Most regressions occur on hardware that's not all that common anymore. I don't think Ubuntu would have pushed it out at all if it hadn't worked in internal testing. Both my older C2D-based laptop and my FM2+ desktop aren't all that commonly used, and probably much less so on Linux. And you don't even get 4.4.0-31 unless you accept the Category 5 updates, which most users wouldn't do. I just like pushing the envelope to see if I can get better performance, so I accepted it knowing that if there were any regressions, I could pull up the Grub menu and boot the older kernel (or install the Exton one, which I have).

I've always done stability testing in Windows. In my experience, Linux actually sometimes tolerates overclocks that Windows won't, so if my computer passes stability tests in Windows, I assume that Linux is stable until demonstrated otherwise. The only time I've ever had things go differently was when undervolting this 870K; there were some kernels that couldn't pull the CPU out of C6 under Linux, but Windows could do it at the same settings. I had to give the CPU a little more vcore or turn C6 off. This chip can run fine under Windows at stock speeds as low as 1.24v, but needs 1.30 in Linux because of the C6 issue. (I'm currently running at 4.2 GHz and 1.33v most of the time, in both OSes).

Windows was my primary OS until last fall, anyway, because my job at the time required Windows software. Since October, it's been Mint Cinnamon, but I still have Windows 10 as a dual-boot option on a separate SSD.
 
#19 ·
Quote:
Originally Posted by jsc1973 View Post

If 4.4.0-31 works for you, stick with it. Most regressions occur on hardware that's not all that common anymore. I don't think Ubuntu would have pushed it out at all if it hadn't worked in internal testing. Both my older C2D-based laptop and my FM2+ desktop aren't all that commonly used, and probably much less so on Linux. And you don't even get 4.4.0-31 unless you accept the Category 5 updates, which most users wouldn't do. I just like pushing the envelope to see if I can get better performance, so I accepted it knowing that if there were any regressions, I could pull up the Grub menu and boot the older kernel (or install the Exton one, which I have).

I've always done stability testing in Windows. In my experience, Linux actually sometimes tolerates overclocks that Windows won't, so if my computer passes stability tests in Windows, I assume that Linux is stable until demonstrated otherwise. The only time I've ever had things go differently was when undervolting this 870K; there were some kernels that couldn't pull the CPU out of C6 under Linux, but Windows could do it at the same settings. I had to give the CPU a little more vcore or turn C6 off. This chip can run fine under Windows at stock speeds as low as 1.24v, but needs 1.30 in Linux because of the C6 issue. (I'm currently running at 4.2 GHz and 1.33v most of the time, in both OSes).

Windows was my primary OS until last fall, anyway, because my job at the time required Windows software. Since October, it's been Mint Cinnamon, but I still have Windows 10 as a dual-boot option on a separate SSD.
Okay, glad it works for you, and yes, I as well dual boot between Windows 10 Pro, my main OS currently, and Linux Mint 18. Until a week ago I was using my second SSD for the Windows Insider program but I got bored of it, and anyway I've already gotten a good idea about how the Anniversary (of Win 10) will be, so I thought to give Mint a chance. With all the Telemetry Windows 10 is using I feel the need to start familiarizing myself with, and learning, Linux Mint. Until I will be capable to perform in Mint 90 - 95% of everything I'm doing in Windows.

So, were you not able to monitor frequencies in Mint, after all?

My research has led me to this thread. That person has found a way to monitor voltages on his ASUS Z97M-PLUS motherboard. I am sorry to admit this but I do not understand or fully understand, almost anything mentioned in that thread!...
frown.gif
Not a native English speaker, either...

Right now I'm posting this post from Windows but when, on Mint, I give the command

cat /etc/sensors3.conf

as suggested in that thread, it says that I should visit lm-sensors.org and find a configuration file for my motherboard! Well...that site is down each time I've tried to access it! And Goggling has not returned ANY result about a configuration file for my motherboard. Lost...don't know what to do
frown.gif


Pretty impressive that so few people in this gorgeous forum are dealing with Linux...
 
#20 ·
Quote:
Originally Posted by LostParticle View Post

Okay, glad it works for you, and yes, I as well dual boot between Windows 10 Pro, my main OS currently, and Linux Mint 18. Until a week ago I was using my second SSD for the Windows Insider program but I got bored of it, and anyway I've already gotten a good idea about how the Anniversary (of Win 10) will be, so I thought to give Mint a chance. With all the Telemetry Windows 10 is using I feel the need to start familiarizing myself with, and learning, Linux Mint. Until I will be capable to perform in Mint 90 - 95% of everything I'm doing in Windows.

So, were you not able to monitor frequencies in Mint, after all?

My research has led me to this thread. That person has found a way to monitor voltages on his ASUS Z97M-PLUS motherboard. I am sorry to admit this but I do not understand or fully understand, almost anything mentioned in that thread!...
frown.gif
Not a native English speaker, either...

Right now I'm posting this post from Windows but when, on Mint, I give the command

cat /etc/sensors3.conf

as suggested in that thread, it says that I should visit lm-sensors.org and find a configuration file for my motherboard! Well...that site is down each time I've tried to access it! And Goggling has not returned ANY result about a configuration file for my motherboard. Lost...don't know what to do
frown.gif


Pretty impressive that so few people in this gorgeous forum are dealing with Linux...
Most of the people here are gaming enthusiasts, which means they need DirectX and therefore use Windows. Hopefully, that will start to change with Vulkan getting more industry support than OpenGL ever had, but Windows is going to be with us for a long time to come as the go-to gaming OS.

The www.lm-sensors.org site has been down for months. Apparently, whoever maintained it not only stopped doing so, but also took down everything and disappeared without a trace, at least from the Internet. The most recent archive of the page that had the configuration files is this: https://web.archive.org/web/20150901092438/http://www.lm-sensors.org/wiki/Configurations Unfortunately, most of the ones there were very old to begin with.

I'm sure you know this already, but the best way to see what's supported is to simply type sudo sensors-detect from the command line and hope for the best.

I can monitor CPU frequency with Mint, that's never been a problem. The CPU reports its operating frequency correctly and I monitor it in real-time with Conky. But none of the advanced features work reliably if at all.

The link in that Ubuntu Forums thread: http://permalink.gmane.org/gmane.linux.drivers.sensors/36526 probably is your best bet as far as creating your own configuration file. At least there's information available. Good luck trying to find anything on how to do it with an A88X motherboard like mine.
 
#21 ·
Quote:
Originally Posted by LostParticle View Post

Okay, glad it works for you, and yes, I as well dual boot between Windows 10 Pro, my main OS currently, and Linux Mint 18. Until a week ago I was using my second SSD for the Windows Insider program but I got bored of it, and anyway I've already gotten a good idea about how the Anniversary (of Win 10) will be, so I thought to give Mint a chance. With all the Telemetry Windows 10 is using I feel the need to start familiarizing myself with, and learning, Linux Mint. Until I will be capable to perform in Mint 90 - 95% of everything I'm doing in Windows.

So, were you not able to monitor frequencies in Mint, after all?

My research has led me to this thread. That person has found a way to monitor voltages on his ASUS Z97M-PLUS motherboard. I am sorry to admit this but I do not understand or fully understand, almost anything mentioned in that thread!...
frown.gif
Not a native English speaker, either...

Right now I'm posting this post from Windows but when, on Mint, I give the command

cat /etc/sensors3.conf

as suggested in that thread, it says that I should visit lm-sensors.org and find a configuration file for my motherboard! Well...that site is down each time I've tried to access it! And Goggling has not returned ANY result about a configuration file for my motherboard. Lost...don't know what to do
frown.gif


Pretty impressive that so few people in this gorgeous forum are dealing with Linux...
There's some applets that are available, I run one that shows realtime CPU frequency, temp, load, network speed etc but no voltage. Right click on applet bar and go add applets then select available online. There's a bunch of different ones on there not sure any do voltage, but gkrellm appears to show realtime core voltage on my system(mint 17.3) and doesn't take up much desktop and real easy to configure.

Here's the applet display



and gkrellm, this can be shrunk down considerably depending on selected sensors EDIT: it also remembers where it should be on the desktop

 
#22 ·
On the note of the "lack of" Linux users, OCN used to have quite a large Linux community, unfortunately a few of the core members of the group were run off by certain other individuals.

On topic though:

For current core frequency monitoring - you'll need to use a specialized tool for any Intel Core series processors. This tool is `i7z` and uses mathematics to calculate the current core frequencies on the fly. This is how all of the modern Windows applications do it as well since the reported CPU frequency is always going to be based on the multiplier and bclk. In fact, on Linux with generations prior to (may be mistaken here) Ivy Bridge - the multiplier could only report the maximum NON-TURBO multiplier and so overclocked CPUs using a normally turbo only multiplier would always report their CPU frequency incorrectly. Here's an example using my system:
http://puu.sh/q6Evp/4b9785e4c1.png
As you can see the actual core frequency seems to "vary" quite a bit, even the reported numbers are all over the place. You can blame this one on Intel, they don't report the value like they used to. This CPU is overclocked to 4.4ghz, it reports 4009mhz, the BCLK+Multiplier calculates out to 4008mhz. The estimated frequency based on math comes out to ~4.35-4.45ghz.

For monitoring hardware, the lm-sensors suite is the best tool available. It can take some time for new platforms to hit the suite but usually everything is able to be probed using sensors-detect, especially if ACPI is enabled and working properly. Usually answering the default to the questions in sensors-detect is the best and safest approach. Probing for sensors that aren't there can cause kernel panics and data loss. YMMV. Google is your friend here, since if your board doesn't show up with the automatic selections, you may need to see what other people have done to get things working.
Once you have lm-sensors set up, you have literally dozens of options for actually monitoring the data. For one, you could use `watch` to run the `sensors` utility on a loop:

Code:

Code:
watch -ctn 10 sensors
The above would run "sensors" in a terminal every ten seconds.
You could also do as above and use some of the KDE Plasmoids, a desktop applet, or a dedicated utility like GKRELLM, as well as the defacto-standard for system monitoring on Linux: Conky. Conky is probably one of the most powerful desktop applications available for Linux and can do pretty much anything you could imagine for adding realtime monitoring on your desktop. Check it out:
Simple sidebar type conky configurations:
http://img06.deviantart.net/a791/i/2011/011/9/6/simple_conky_by_pemete-d36ydt3.png

Complex LUA-assisted conky configurations:
http://img15.deviantart.net/2f10/i/2011/057/8/a/conky___lua_ubuntu_by_fenouille84-d3af8o5.png
http://devmadness.com/wp-content/uploads/2015/04/black_pearl_conky_by_ninquitassar-d4nkub7.jpg
https://curiousandcrazy.files.wordpress.com/2015/09/screenshot-from-2015-09-08-232101.png

The limit is only your patience and imagination. There's literally hundreds of thousands of posts about conky configuration with examples and presets including extensions and all sorts of other things. You can pretty much tinker to your hearts content. I opt for something very minimal when I do use Conky but I haven't had time to fiddle with a real Linux desktop in quite some time, right now I just rock XFCE4 stock with minimal tweaks. Nothing really serious like my old setup:
http://cdn.overclock.net/4/40/40efc80d_2013-08-09-201424_2560x1440_scrot.png
 
#23 ·
Thank you very much, @frack0 and @Xaero252 for your rich replies, +REP to both of you
smile.gif


Your suggestions require quite some time, lots of reading and effort, as I observe. Unfortunately, this specific period I do not have so much time. For sure I will research this further though because I definitely want to make Linux Mint my primary OS or equal to my primary OS - currently Win 10 Pro. This means to be able to do almost everything I'm doing in Windows, in Mint. And indeed I am even right now, like 90% of it.
smile.gif


Anyway, when it comes to monitoring i7z did the trick for me. Here's a screenshot from my system:

It does not monitor the wealth of the values HWiNFO64 is capable of monitoring, in Windows, but it is good enough for now, in combination with Psensor for the temperatures.

One last question (even though I might revisit this thread for other inquiries on Linux Mint):

I have observed that my system is not idling as it idles in Windows. While i7z is monitoring and my system is idle, so I do not touch the mouse at all for approx. 10 minutes, I observe that my Core Ratio never settles down to x8 (so, at 800 MHz). It keeps fluctuating and rising up to 4700 MHz, which is my current OC (overclock). Why is this happening? Is this how it is supposed to happen in Mint? Because in Windows 10 Pro if I leave my system in peace, after 1 -2 minutes my cores lower down to 800 MHz and they stay there! At least this is what HWiNFO64 is showing me.

Note please that I have observed this behavior both with the recommended kernel, as well as with the latest one.

Thank you!
thumb.gif


PS:
Quote:
Originally Posted by Xaero252 View Post

...

For monitoring hardware, the lm-sensors suite is the best tool available. It can take some time for new platforms to hit the suite but usually everything is able to be probed using sensors-detect, especially if ACPI is enabled and working properly.

...
Can you please tell me how can I check this out? I mean, how can I verify that it is enabled and working properly?
 
#24 ·
Quote:
Originally Posted by LostParticle View Post

Can you please tell me how can I check this out? I mean, how can I verify that it is enabled and working properly?
Sure thing!
The Linux Kernel hardware autodetection _SHOULD_ automatically see that ACPI is available on your motherboard, and probe the necessary drivers to pull information from those ACPI sensors. To find out if it did it's job you can run "dmesg | grep -i acpi" This should output something like this:

Code:

Code:
[    0.000000] BIOS-e820: [mem 0x00000000bf780000-0x00000000bf797fff] ACPI data
[    0.000000] BIOS-e820: [mem 0x00000000bf798000-0x00000000bf7dbfff] ACPI NVS
[    0.000000] ACPI: Early table checksum verification disabled
[    0.000000] ACPI: RSDP 0x00000000000FB200 000024 (v02 ACPIAM)
[    0.000000] ACPI: XSDT 0x00000000BF780100 00005C (v01 051211 XSDT1500 20110512 MSFT 00000097)
[    0.000000] ACPI: FACP 0x00000000BF780290 0000F4 (v03 051211 FACP1500 20110512 MSFT 00000097)
[    0.000000] ACPI: DSDT 0x00000000BF7804B0 00D8B5 (v01 A1545  A1545000 00000000 INTL 20060113)
[    0.000000] ACPI: FACS 0x00000000BF798000 000040
[    0.000000] ACPI: FACS 0x00000000BF798000 000040
[    0.000000] ACPI: APIC 0x00000000BF780390 0000D8 (v01 051211 APIC1500 20110512 MSFT 00000097)
[    0.000000] ACPI: MCFG 0x00000000BF780470 00003C (v01 051211 OEMMCFG  20110512 MSFT 00000097)
[    0.000000] ACPI: OEMB 0x00000000BF798040 000072 (v01 051211 OEMB1500 20110512 MSFT 00000097)
[    0.000000] ACPI: HPET 0x00000000BF78F4B0 000038 (v01 051211 OEMHPET  20110512 MSFT 00000097)
[    0.000000] ACPI: DMAR 0x00000000BF7980C0 000140 (v01 AMI    OEMDMAR  00000001 MSFT 00000097)
[    0.000000] ACPI: OSFR 0x00000000BF78F4F0 0000B0 (v01 051211 OEMOSFR  20110512 MSFT 00000097)
[    0.000000] ACPI: Local APIC address 0xfee00000
[    0.000000] ACPI: PM-Timer IO Port: 0x808
[    0.000000] ACPI: Local APIC address 0xfee00000
[    0.000000] ACPI: INT_SRC_OVR (bus 0 bus_irq 0 global_irq 2 dfl dfl)
[    0.000000] ACPI: INT_SRC_OVR (bus 0 bus_irq 9 global_irq 9 high level)
[    0.000000] ACPI: IRQ0 used by override.
[    0.000000] ACPI: IRQ9 used by override.
[    0.000000] Using ACPI (MADT) for SMP configuration information
[    0.000000] ACPI: HPET id: 0x8086a301 base: 0xfed00000
[    0.000032] ACPI: Core revision 20160108
[    0.005511] ACPI: 1 ACPI AML tables successfully acquired and loaded
[    0.931388] PM: Registering ACPI NVS region [mem 0xbf798000-0xbf7dbfff] (278528 bytes)
[    0.939841] ACPI: bus type PCI registered
[    0.939842] acpiphp: ACPI Hot Plug PCI Controller Driver version: 0.5
[    0.944134] ACPI: Added _OSI(Module Device)
[    0.944135] ACPI: Added _OSI(Processor Device)
[    0.944136] ACPI: Added _OSI(3.0 _SCP Extensions)
[    0.944137] ACPI: Added _OSI(Processor Aggregator Device)
[    0.944355] ACPI: Executed 1 blocks of module-level executable AML code
[    1.007337] ACPI: Interpreter enabled
[    1.007352] ACPI: (supports S0 S1 S3 S4 S5)
[    1.007353] ACPI: Using IOAPIC for interrupt routing
[    1.007976] PCI: MMCONFIG at [mem 0xe0000000-0xefffffff] reserved in ACPI motherboard resources
[    1.008264] PCI: Using host bridge windows from ACPI; if necessary, use "pci=nocrs" and report a bug
[    1.011960] ACPI: PCI Root Bridge [PCI0] (domain 0000 [bus 00-ff])
[    1.011963] acpi PNP0A08:00: _OSC: OS supports [ExtendedConfig ASPM ClockPM Segments MSI]
You'll see that ACPI is first exposing the BIOS to the OS, and then adding various other hardware and sensors to the table of available hardware the kernel can see. Most of the above output is just debugging output showing that yes, indeed ACPI is working. It's not something you should have to mess with, and is expected to "just work" at this point in time. Any modern system should have ACPI available, enabled and working. The only case where it wouldn't is if for some reason the Kernel you are running was compiled without ACPI support.

Edit:
As far as your concerns with the CPU speed fluctuating during idle, depending on the current CPU Governor and the power save settings you have configured, this is more or less normal behavior. You are probably using the "ondemand" governor which is going to fork CPU resources over AS SOON as they are requested. You could try the "powersave" governor or any of the others, see here for more details:
https://wiki.archlinux.org/index.php/CPU_frequency_scaling

And here for further powersaving tips:
https://wiki.archlinux.org/index.php/Power_management
https://wiki.archlinux.org/index.php/laptop

You'll find that on Linux you actually have a lot more direct control over the CPU's frequency scaling, since you can even employ daemons to control the CPU frequency manually throughout a normal day, and respond to events like user interaction. For example you could schedule any time from 11pm to 4pm to run the CPU at 800mhz for energy savings, and then run the ondemand scheduler from 4pm to 11pm. And automatically switch to the ondemand scheduler if the computer has been used, and switch away from it after being idle for 30 minutes. It's quite robust, if a bit tedious to configure.

Another Edit:

As far as the wealth of values hwinfo64 monitors, I believe almost all of the information available in that tool is probably available in sysfs, by default. Obviously this isn't very useful to the end-user, but it is there. For example, cat /proc/cpuinfo will tell you more about your CPU than you can possibly care about. `lspci` `lsusb` `lshw` and `dmidecode` can tell you more about the PCI, USB and other hardware devices, and dmidecode can tell you everything about your ram. Unfortunately nobody has really wrapped all of these up into a truly usable tool for actively monitoring the information. Usually its a passive probe and print of the data. You may wish to look at "hardinfo" which provides a GUI to look at the current system. It doesn't do a lot in the way of monitoring, but it will show all of your hardware in a GUI very easily. Also take a look at OpenHardwareMonitor.
 
#25 ·
Okay, thanks a lot for your reply! I am a complete newbie so could you please have a look at my output and tell me if everything is all right with this ACPI?
lm-sensors and sensors show my VCore up to +2.04V ....
me@me-desktop ~ $ dmesg | grep -i acpi
[ 0.000000] BIOS-e820: [mem 0x00000000ca96a000-0x00000000ca970fff] ACPI NVS
[ 0.000000] BIOS-e820: [mem 0x00000000de3b9000-0x00000000de4f4fff] ACPI NVS
[ 0.000000] ACPI: Early table checksum verification disabled
[ 0.000000] ACPI: RSDP 0x00000000000F04A0 000024 (v02 ALASKA)
[ 0.000000] ACPI: XSDT 0x00000000DE4C1080 000084 (v01 ALASKA A M I 01072009 AMI 00010013)
[ 0.000000] ACPI: FACP 0x00000000DE4D2770 00010C (v05 ALASKA A M I 01072009 AMI 00010013)
[ 0.000000] ACPI: DSDT 0x00000000DE4C11A0 0115C9 (v02 ALASKA A M I 00000171 INTL 20120711)
[ 0.000000] ACPI: FACS 0x00000000DE4F4F80 000040
[ 0.000000] ACPI: APIC 0x00000000DE4D2880 000092 (v03 ALASKA A M I 01072009 AMI 00010013)
[ 0.000000] ACPI: FPDT 0x00000000DE4D2918 000044 (v01 ALASKA A M I 01072009 AMI 00010013)
[ 0.000000] ACPI: SSDT 0x00000000DE4D2960 000539 (v01 PmRef Cpu0Ist 00003000 INTL 20051117)
[ 0.000000] ACPI: SSDT 0x00000000DE4D2EA0 000B74 (v01 CpuRef CpuSsdt 00003000 INTL 20051117)
[ 0.000000] ACPI: SSDT 0x00000000DE4D3A18 0001C7 (v01 PmRef LakeTiny 00003000 INTL 20051117)
[ 0.000000] ACPI: MCFG 0x00000000DE4D3BE0 00003C (v01 ALASKA A M I 01072009 MSFT 00000097)
[ 0.000000] ACPI: HPET 0x00000000DE4D3C20 000038 (v01 ALASKA A M I 01072009 AMI. 00000005)
[ 0.000000] ACPI: SSDT 0x00000000DE4D3C58 00036D (v01 SataRe SataTabl 00001000 INTL 20120711)
[ 0.000000] ACPI: SSDT 0x00000000DE4D3FC8 005A36 (v01 SaSsdt SaSsdt 00003000 INTL 20120711)
[ 0.000000] ACPI: AAFT 0x00000000DE4D9A00 0007B8 (v01 ALASKA OEMAAFT 01072009 MSFT 00000097)
[ 0.000000] ACPI: UEFI 0x00000000DE4DA1B8 000042 (v01 ALASKA A M I 01072009 00000000)
[ 0.000000] ACPI: Local APIC address 0xfee00000
[ 0.000000] ACPI: PM-Timer IO Port: 0x1808
[ 0.000000] ACPI: Local APIC address 0xfee00000
[ 0.000000] ACPI: LAPIC_NMI (acpi_id[0xff] high edge lint[0x1])
[ 0.000000] ACPI: INT_SRC_OVR (bus 0 bus_irq 0 global_irq 2 dfl dfl)
[ 0.000000] ACPI: INT_SRC_OVR (bus 0 bus_irq 9 global_irq 9 high level)
[ 0.000000] ACPI: IRQ0 used by override.
[ 0.000000] ACPI: IRQ9 used by override.
[ 0.000000] Using ACPI (MADT) for SMP configuration information
[ 0.000000] ACPI: HPET id: 0x8086a701 base: 0xfed00000
[ 0.000021] ACPI: Core revision 20150930
[ 0.009931] ACPI: 6 ACPI AML tables successfully acquired and loaded
[ 0.095402] PM: Registering ACPI NVS region [mem 0xca96a000-0xca970fff] (28672 bytes)
[ 0.095403] PM: Registering ACPI NVS region [mem 0xde3b9000-0xde4f4fff] (1294336 bytes)
[ 0.117007] ACPI FADT declares the system doesn't support PCIe ASPM, so disable it
[ 0.117007] ACPI: bus type PCI registered
[ 0.117008] acpiphp: ACPI Hot Plug PCI Controller Driver version: 0.5
[ 0.129172] ACPI: Added _OSI(Module Device)
[ 0.129173] ACPI: Added _OSI(Processor Device)
[ 0.129174] ACPI: Added _OSI(3.0 _SCP Extensions)
[ 0.129175] ACPI: Added _OSI(Processor Aggregator Device)
[ 0.132060] ACPI: Executed 15 blocks of module-level executable AML code
[ 0.134475] ACPI: Dynamic OEM Table Load:
[ 0.134479] ACPI: SSDT 0xFFFF88040BC11800 0003D3 (v01 PmRef Cpu0Cst 00003001 INTL 20051117)
[ 0.134904] ACPI: Dynamic OEM Table Load:
[ 0.134907] ACPI: SSDT 0xFFFF88040C089800 0005AA (v01 PmRef ApIst 00003000 INTL 20051117)
[ 0.135355] ACPI: Dynamic OEM Table Load:
[ 0.135357] ACPI: SSDT 0xFFFF88040BC09E00 000119 (v01 PmRef ApCst 00003000 INTL 20051117)
[ 0.136602] ACPI: Interpreter enabled
[ 0.136607] ACPI Exception: AE_NOT_FOUND, While evaluating Sleep State [\_S1_] (20150930/hwxface-580)
[ 0.136611] ACPI Exception: AE_NOT_FOUND, While evaluating Sleep State [\_S2_] (20150930/hwxface-580)
[ 0.136622] ACPI: (supports S0 S3 S4 S5)
[ 0.136623] ACPI: Using IOAPIC for interrupt routing
[ 0.136657] PCI: Using host bridge windows from ACPI; if necessary, use "pci=nocrs" and report a bug
[ 0.143312] ACPI: PCI Root Bridge [PCI0] (domain 0000 [bus 00-7e])
[ 0.143316] acpi PNP0A08:00: _OSC: OS supports [ExtendedConfig ASPM ClockPM Segments MSI]
[ 0.143441] acpi PNP0A08:00: _OSC: platform does not support [PCIeHotplug PME]
[ 0.143518] acpi PNP0A08:00: _OSC: OS now controls [AER PCIeCapability]
[ 0.143519] acpi PNP0A08:00: FADT indicates ASPM is unsupported, using BIOS configuration
[ 0.143871] pci 0000:00:01.0: System wakeup disabled by ACPI
[ 0.143999] pci 0000:00:14.0: System wakeup disabled by ACPI
[ 0.144254] pci 0000:00:1a.0: System wakeup disabled by ACPI
[ 0.144367] pci 0000:00:1c.0: System wakeup disabled by ACPI
[ 0.144479] pci 0000:00:1c.2: System wakeup disabled by ACPI
[ 0.144589] pci 0000:00:1c.3: System wakeup disabled by ACPI
[ 0.144726] pci 0000:00:1d.0: System wakeup disabled by ACPI
[ 0.145162] pci 0000:01:00.0: System wakeup disabled by ACPI
[ 0.153045] acpiphp: Slot [1] registered
[ 0.169715] pci 0000:08:00.0: System wakeup disabled by ACPI
[ 0.177463] ACPI: PCI Interrupt Link [LNKA] (IRQs 3 4 5 6 10 *11 12 14 15)
[ 0.177490] ACPI: PCI Interrupt Link [LNKB] (IRQs 3 4 5 6 *10 11 12 14 15)
[ 0.177516] ACPI: PCI Interrupt Link [LNKC] (IRQs *3 4 5 6 10 11 12 14 15)
[ 0.177542] ACPI: PCI Interrupt Link [LNKD] (IRQs 3 4 *5 6 10 11 12 14 15)
[ 0.177567] ACPI: PCI Interrupt Link [LNKE] (IRQs 3 4 5 6 10 11 12 14 15) *0, disabled.
[ 0.177593] ACPI: PCI Interrupt Link [LNKF] (IRQs 3 4 5 6 10 11 12 14 15) *0, disabled.
[ 0.177618] ACPI: PCI Interrupt Link [LNKG] (IRQs 3 4 5 6 10 11 12 14 15) *0, disabled.
[ 0.177644] ACPI: PCI Interrupt Link [LNKH] (IRQs 3 4 5 6 10 *11 12 14 15)
[ 0.177725] ACPI: Enabled 5 GPEs in block 00 to 3F
[ 0.177942] ACPI: bus type USB registered
[ 0.178032] PCI: Using ACPI for IRQ routing
[ 0.185715] pnp: PnP ACPI init
[ 0.185777] system 00:00: Plug and Play ACPI device, IDs PNP0c01 (active)
[ 0.185830] system 00:01: Plug and Play ACPI device, IDs PNP0c02 (active)
[ 0.185842] pnp 00:02: Plug and Play ACPI device, IDs PNP0b00 (active)
[ 0.185865] system 00:03: Plug and Play ACPI device, IDs INT3f0d PNP0c02 (active)
[ 0.185911] system 00:04: Plug and Play ACPI device, IDs PNP0c02 (active)
[ 0.185955] system 00:05: Plug and Play ACPI device, IDs PNP0c02 (active)
[ 0.185985] system 00:06: Plug and Play ACPI device, IDs PNP0c02 (active)
[ 0.186098] pnp 00:07: Plug and Play ACPI device, IDs PNP0501 (active)
[ 0.186379] system 00:08: Plug and Play ACPI device, IDs PNP0c02 (active)
[ 0.186483] pnp: PnP ACPI: found 9 devices
[ 0.191999] clocksource: acpi_pm: mask: 0xffffff max_cycles: 0xffffff, max_idle_ns: 2085701024 ns
[ 0.554116] ACPI: Power Button [PWRB]
[ 0.554136] ACPI: Sleep Button [SLPB]
[ 0.554155] ACPI: Power Button [PWRF]
me@me-desktop ~ $

Also, thanks for clarifying it to me about the Power Management! The truth is that in my Linux Mint when I give "Power Management" I get this:

Absolutely nothing there, like Balanced or Performance profile or Power Saver, etc...

You gave me links to ArchLinux.... Is it the same on Mint? Are these applicable?

Thanks man!
thumb.gif
 
#26 ·
Quote:
Originally Posted by LostParticle View Post

Okay, thanks a lot for your reply! I am a complete newbie so could you please have a look at my output and tell me if everything is all right with this ACPI?
lm-sensors and sensors show my VCore up to +2.04V ....

Also, thanks for clarifying it to me about the Power Management! The truth is that in my Linux Mint when I give "Power Management" I get this:

Absolutely nothing there, like Balanced or Performance profile or Power Saver, etc...

You gave me links to ArchLinux.... Is it the same on Mint? Are these applicable?

Thanks man!
thumb.gif
On the first note about voltage; that sensor is being properly detected but may not be reading correct values, this is probably an upstream bug in LM_SENSORS. You actually got me curious and looking at my output I see some address conflicts for ACPI. A bit of research turns up that certain BIOS changes ASUS has made recently (last 3-5 years) internally is preventing exclusive access to ACPI. The workaround is to add acpi_enforce_resources=lax to your kernel command line. This workaround is kind of hacky and not proper, but it does get the job done.
To add this option edit (as root) /etc/default/grub. Change the line that reads GRUB_CMDLINE_LINUX_DEFAULT="quiet" to add acpi_enforce_resources=lax after quiet so it reads GRUB_CMDLINE_LINUX_DEFAULT="quiet acpi_enforce_resources=lax"
After saving the file run (again, as root) update-grub.
If things don't work right or you notice problems, you can boot into failsafe mode to remove that change, and things should return to normal.

On the note of Power Management - that power management application is a gui that provides minimal configuration. Most configuration on Linux is done via configuration files stored on the hard drive. This is a two-edged sword. It puts the power in the hands of the user, and if you make a mistake you'll pay the price. But because they are stored plaintext... as long as the drive isn't unreadable, you can always go back in and fix them.

And finally, in regards to the ArchWiki:
The ArchWiki is almost always applicable. It was written by ArchLinux users, for ArchLinux primarily, but almost all of its content translates very well to any Linux distribution that uses SystemD. The package names and the package management program will be different, but the information is the same. Especially when it comes to core Linux hardware capabilities, since those are the same on every distribution.

A small footnote before I go:
I started my Linux Journey in 2003 with a foray into Ubuntu, followed by some meddling in a strange land called XandrOS. When I began I knew nothing of the terminal, the kernel, graphics drivers, anything. I didn't understand Windows fully much less did I have a grasp of how to navigate a filesystem and configure things. After messing around for a couple of months I became frustrated and gave up. I went back to Windows and eventually came to grasp just about everything one could on Windows. In 2007 I made a proper switch to Linux. It took me probably a year to get to the point that I was comfortable editing files and getting things configured. I was determined and I conquered by desire. I wanted things to look a certain way and work a certain way and I got what I wanted. Don't let yourself get discouraged, it is quite daunting when you look at things from afar, and it seems like its a serious investment of time and resources... but I assure you it is worth the reward. And now it is easier than ever to learn Linux, the GUI provides you the tools to give you a nudge in the right direction, and resources like the Gentoo and ArchLinux wiki and the legacy of information held by the Canonical forums and github make almost any problem just a simple google search away.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top