. . Well I think the invalid data is out as there were no errors.

Invalid temperature data.

. . It may be a sensor issue, but since it is not happening with release 96 that still puts it down to something in release 104.

The new release has a different way of passing the data, hence the erroneous reading as the application that makes use of the data hasn't changed.

Like the early high temperature readings for Ryzen systems till they sorted out how to read & interpret the data they were getting.

There was a update to the temperature readout drivers back in the summer. Lm-sensors is using new sensor interpretations and that might have been pulled into the later kernel images you tried. Those drivers will not have been reverted to the previous versions by you rolling back to an earlier kernel release.

I would try a sudo sensors-detect from command line for an update to how your motherboard and gpu temps are read.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

How are you monitoring gpu temps? Assume from your comment it is via software applications.

. . I am using psensor.

Stephen

.

How are you monitoring gpu temps? Assume from your comment it is via software applications.

. . I am using psensor.

Stephen

.

Open Terminal and input

sudo sensors-detect

and answer yes to all the questions.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Do you have the Synaptic Package Manager installed? If so, search for linux-image in the Search box in the Installed category and mark the old kernel images for complete removal. Be careful to not remove the kernel image you want to keep.

. . Hi Keith,

. . Yep I have Package Manager, I will give that a try. Thanks mate.

Stephen

:)

. . I am running it now and I think I will recover about 3GB of space. Thanks for that Keith.

Stephen

:)

. . Well I think the invalid data is out as there were no errors.

Invalid temperature data.

. . It may be a sensor issue, but since it is not happening with release 96 that still puts it down to something in release 104.

The new release has a different way of passing the data, hence the erroneous reading as the application that makes use of the data hasn't changed.

Like the early high temperature readings for Ryzen systems till they sorted out how to read & interpret the data they were getting.

. . That could explain it, and gels with what Keith is telling me, so when I have finished removing the unwanted Linux images I will pursue that line.

Stephen

:)

Invalid temperature data.
The new release has a different way of passing the data, hence the erroneous reading as the application that makes use of the data hasn't changed.
Like the early high temperature readings for Ryzen systems till they sorted out how to read & interpret the data they were getting.

There was a update to the temperature readout drivers back in the summer. Lm-sensors is using new sensor interpretations and that might have been pulled into the later kernel images you tried. Those drivers will not have been reverted to the previous versions by you rolling back to an earlier kernel release.

I would try a sudo sensors-detect from command line for an update to how your motherboard and gpu temps are read.

. . Well since it is not happening with release 96 it seems it might be something else, but I when the old images removal is finished I will have a look at that.

Stephen

. .

. . I am using psensor.

Open Terminal and input

sudo sensors-detect

and answer yes to all the questions.

. . I only get "command not found".

Stephen

??

. . I am using psensor.

Open Terminal and input

sudo sensors-detect

and answer yes to all the questions.

. . I only get "command not found".

Stephen

??

Don't know what to answer other than you typoed or are not running as superuser. Psensor is nothing more than a graphical front-end to lm-sensors. It can't work or display any sensor information without lm-sensors being installed. Lm-sensors is usually installed by default in most distributions. I have never heard or found any version of psensor that might have an embedded lm-sensor constituent.

Try searching for lm-sensor in your File Manager and search the whole disk. If it finds it, then change to that directory and try the sensors-detect command again.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Try just "sensors"

Try just "sensors"

Yes, let's start simple to see if the lm-sensors app is installed. Doesn't require superuser. Just open a Terminal session and type sensors. Should printout your temps, fan speeds and some voltages.

I was mostly trying to get you to reprobe your system sensors in case they got changed by the later kernel releases you installed for a while in my previous post.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Try just "sensors"

. . Aha! That produced results! It seems that p-sensors has been working all this time without the benefit of lm-sensors. So I guess it has it's own back end to the hardware.

. . I am guessing I should now install lm-sensors ...

Stephen

[edit] - OK with that installed I tried the fan control again, it still bombs, ... but ...

. . I think it is the thing that Jeff mentioned about later versions of something not recognising the labels for the sensor outputs. That is, it no longer recognises "$current_temp" as sensor output and sees it as a filename. I was put off the trail of that because it then opens a whole lot of empty files. it seems "filename < x" means read file with name x but "filename > x1" means write a file named x1, so every time the script cycles it creates a dozen new files. That is what put me off following it up. I had to keep deleting all these files. I just need to know the correct term for that GPU heat sensors. If it is no longer "$current_temp" what is it now?

. . One further symptom since this attempt is that Firefox is also behaving strangely. I have it open frequently as I check status pages from time to time and I read these forums. But since the one step forward one step back thing, while it behaves normally looking at the status pages, when I am in the forums it bogarts CPU time, it did not do that before ... :( Doncha just luv computers ??

Stephen

??

:)

Try just "sensors"

. . Aha! That produced results! It seems that p-sensors has been working all this time without the benefit of lm-sensors. So I guess it has it's own back end to the hardware.

. . I am guessing I should now install lm-sensors ...

Stephen

:)

Uhh .... no. "sensors" IS lm-sensors.

Your previous attempt to run "sensors-detect" must have had a typo or you didn't run it from a root terminal.

If you want to read about lm-sensors, look here

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Uhh .... no. "sensors" IS lm-sensors.

Your previous attempt to run "sensors-detect" must have had a typo or you didn't run it from a root terminal.

If you want to read about lm-sensors, look here

. . It has an indepedent presence somehow, because when I typed sensors it responded -

The program 'sensors' is currently not installed. You can install it by typing:
sudo apt-get install lm-sensors

. . Which I have now done and it gives me this ...

coretemp-isa-0000
Adapter: ISA adapter
Core 0: +46.0°C (high = +76.0°C, crit = +100.0°C)
Core 1: +40.0°C (high = +76.0°C, crit = +100.0°C)

. . Running 'sensors-detect' produced this summary ( and yes I answered yes to every question except the last one) -

Driver `coretemp':
* Chip `Intel digital thermal sensor' (confidence: 9)

To load everything that is needed, add this to /etc/modules:
#----cut here----
# Chip drivers
coretemp
#----cut here----
If you have some drivers built into your kernel, the list above will
contain too many modules. Skip the appropriate ones!

Do you want to add these lines automatically to /etc/modules? (yes/NO)no

. . So it seems even lm-sensors has an issue reading the temps on the GTX1050ti with whatever modules the upgrade loaded as they do not appear in any of those responses.

Stephen

??

Did you try Psensor ? It works great here with my 1070's and gives a lot of useful information.

---

OK, I still don't have any idea of how psensor was able to get any readings without lm-sensors already installed. It goes against everything I have ever found or read in the Linux world.

But it got you to install lm-sensors which for some reason wasn't already installed by your distribution. So that is progress.

I ask WHY you didn't install the coretemp module that is available from sensors-detect? That would give you cpu core temp readouts which is handy for checking whether there is a problem with system or cpu cooling.

I ask also whether you have set coolbits for the xorg.conf and do you have the nvidia xserver settings app running? That would at least tell you the temps on your GTX 1050Ti.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Did you try Psensor ? It works great here with my 1070's and gives a lot of useful information.

Yes, that is what he was running before he installed lm-sensors.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Did you try Psensor ? It works great here with my 1070's and gives a lot of useful information.

. . Thanks Juan, but I have been running Psensor the whole time, the issue was I was getting very very high temps from the GPU when I updated to release 104 of Linux so I have now returned to release 96 that was working AOK before.

Stephen

:(

I have recently reread this thread, and I have become re-confused. I think I have a computer that would benefit from the special programs but I want to use the latest version of Debian. I know I want boinc in my home folder. The computer in question is 8335913. Currently my windows version is a trial version. When I am ready to switch to linux I will set NNT and finish all the work units before switching. May I ask someone to summarise getting what I need after installing linux? I know that Mint and unbuntu are based on debian. I want to work with debian. I have an old Dell computer running on Debian and a HP windows laptop that dual boots to debian. I'm not going to put BOINC on the old dell because I put a very old video card (gt 240) in it and I know it will not run the CUDA 80 application.

Is there any way to keep the current computer ID from windows to linux? Will any of the current windows files be readable in the linux file directories?

I know on Windows I can save the two BOINC directories from Program Files and ProgramData and drop them into a new installation of Windows and then reinstall BOINC and I can pickup where I left off like nothing ever happened.

I never bothered to find out which exact file saves my computer ID.

To ask the question again, can anyone summarize the BOINC installation into your home folder and installing CUDA 80?

---

I have recently reread this thread, and I have become re-confused. I think I have a computer that would benefit from the special programs but I want to use the latest version of Debian. I know I want boinc in my home folder. The computer in question is 8335913. Currently my windows version is a trial version. When I am ready to switch to linux I will set NNT and finish all the work units before switching. May I ask someone to summarise getting what I need after installing linux? I know that Mint and unbuntu are based on debian. I want to work with debian. I have an old Dell computer running on Debian and a HP windows laptop that dual boots to debian. I'm not going to put BOINC on the old dell because I put a very old video card (gt 240) in it and I know it will not run the CUDA 80 application.

Is there any way to keep the current computer ID from windows to linux? Will any of the current windows files be readable in the linux file directories?

I know on Windows I can save the two BOINC directories from Program Files and ProgramData and drop them into a new installation of Windows and then reinstall BOINC and I can pickup where I left off like nothing ever happened.

I never bothered to find out which exact file saves my computer ID.

To ask the question again, can anyone summarize the BOINC installation into your home folder and installing CUDA 80?

A lot of questions in there. As far as keeping the same computer ID, I don't think that is possible. Actually Juan would be the best person to ask. He did something like what you ask but not exactly similar. Requires keeping the computer host ID from the client_state file PLUS keeping the rpc_sequence_number or whatever the proper name is exactly current with the very next contact with the project the exact next contact number and then doing a merge of the two computers. Juan was able to keep his Windows host accumulated project credit over the years and assign it to the new Linux host. That is why he is #2 on the Top Participants list. Ask him EXACTLY how he achieved his outcome.


For getting BOINC running under Linux, TBar has made it VERY simple with his BOINC versions either 7.4.44 or 7.8.3 which installs BOINC in your /Home directory with no permission issues. And you know exactly where your BOINC files are, not hidden and everything contained in one folder on your Desktop. Go here for the BOINC installs along with the special app and the AVX2 cpu app.

BOINC All-In-One Build to run zi3v CUDA 8.0 in Ubuntu 12.04 to 17.04

BOINC 7.8.3

Linux CUDA Special Apps

I don't think any of the Windows BOINC files can be used. TBar provides already configured app_info files and you can use your Windows app_config almost without edit because nothing in it is OS specific, it is a text file after all.

If you have Nvidia graphics cards and the minimum required Compute Level, you can use the special app with CUDA 9.0. TBar has hinted that the special app with static CUDA 9.0 also might be imminent which is even faster than the current zi3v CUDA 90 app.


Hope Juan sees your post and chimes in.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)