Getting computational errors on AMD GPU tasks on Linux

jave200372
jave200372
Joined: 7 Jun 09
Posts: 6
Credit: 39643710
RAC: 4
Topic 230259

Hi

Just trying out a 2nd hand AMD RX6600 video card but am getting computational errors. Running Linux Mint 21.2

Wed 25 Oct 2023 21:08:26 | Einstein@Home | Computation for task h1_1089.80_O3aC01Cl1In0__O3ASHF1a_1090.00Hz_28338_1 finished
Wed 25 Oct 2023 21:08:26 | Einstein@Home | Output file h1_1089.80_O3aC01Cl1In0__O3ASHF1a_1090.00Hz_28338_1_0 for task h1_1089.80_O3aC01Cl1In0__O3ASHF1a_1090.00Hz_28338_1 absent

Any idea as to what's going on?

GWGeorge007
GWGeorge007
Joined: 8 Jan 18
Posts: 3172
Credit: 5123146723
RAC: 3495984

You may want to read this

You may want to read this entire thread first:

https://einsteinathome.org/content/all-sky-gravitational-wave-search-o3-data-o3ashf1

...paying close attention to Bernd's comments.

Also, you may want to rethink how many projects you are running on your lone AMD Radeon RX 6600 GPU.

Also, there are some "bugs" in Mint 21.2 , but you might read up on it.  I'm not quite where they are though.

.....[EDIT].....

What version of BOINC are you running?

...and... read up on this, too:

https://news.itsfoss.com/linux-mint-21-2-edge/

George

Proud member of the Old Farts Association

Ian&Steve C.
Ian&Steve C.
Joined: 19 Jan 20
Posts: 4115
Credit: 49121496169
RAC: 32308539

the problem is most likely

the problem is most likely your drivers.

_________________________________________________________________________

jave200372
jave200372
Joined: 7 Jun 09
Posts: 6
Credit: 39643710
RAC: 4

@GWGEORGE007 I'm running

@GWGEORGE007

I'm running BOINC 7.18.1, which was the version in the repositories that I installed using apt

@IAN&STEVE

I installed the latest drivers from the AMD site - https://www.amd.com/en/support/linux-drivers

 

 

Keith Myers
Keith Myers
Joined: 11 Feb 11
Posts: 5054
Credit: 19119948598
RAC: 5331363

I see the first error is

I see the first error is running out of memory on your failed tasks.

OpenCL CreateCommandQueue failed with OpenCL error: CL_OUT_OF_HOST_MEMORY

So problem with the drivers is probably the likely issue as mentioned.

 

mountkidd
mountkidd
Joined: 14 Jun 12
Posts: 179
Credit: 12935730626
RAC: 6049942

The out-of-memory message is

The out-of-memory message is misleading - the real error is "

Couldn't create OpenCL command queue (error: -6)!"

And this has everything to do with a parameter in boinc-client.service called ProtectSystem.  This parameter was changed/added circa boinc release 7.18 and it needs to be turned off/disabled.  There is a thread with more info about this issue and near the end of the thread, the solution.

 

Also boinc 7.18 isn't intended for Linux and should not be in the repo - it was an Android release.  Please use the 7.20.5 from the Costamagna ppa.

jave200372
jave200372
Joined: 7 Jun 09
Posts: 6
Credit: 39643710
RAC: 4

@MOUNTKIDD Many thanks. I

@MOUNTKIDD

Many thanks. I read all of that thread you linked and applied the changes suggested in the last comment.

Also, I added that PPA and am now running BOINC 7.20.5. I will also add that PPA to my other PC.

I am now crunching AMD GPU Einstein tasks and no errors

mountkidd
mountkidd
Joined: 14 Jun 12
Posts: 179
Credit: 12935730626
RAC: 6049942

Good to hear you’re back

Good to hear you’re back crunching!

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.