Hello all, I recently upgraded one of my hosts and it seems that SETI@home no longer wants to use my ATI GPU.

BOINC sees the card and has been requesting work for it, but not getting anything. The NVIDIA card is crunching away however.

From the log:
Sat 31 Dec 2016 12:07:34 AM EST | | OpenCL: AMD/ATI GPU 0: AMD CYPRESS (DRM 2.43.0, LLVM 3.8.0) (driver version 11.2.0, device version OpenCL 1.1 MESA 11.2.0, 1024MB, 1024MB available, 522 GFLOPS peak)

Any ideas?

---

Hello all, I recently upgraded one of my hosts and it seems that SETI@home no longer wants to use my ATI GPU.

BOINC sees the card and has been requesting work for it, but not getting anything. The NVIDIA card is crunching away however.

From the log:
Sat 31 Dec 2016 12:07:34 AM EST | | OpenCL: AMD/ATI GPU 0: AMD CYPRESS (DRM 2.43.0, LLVM 3.8.0) (driver version 11.2.0, device version OpenCL 1.1 MESA 11.2.0, 1024MB, 1024MB available, 522 GFLOPS peak)

Any ideas?

Check the panic mode thread. Do you have v8 work selected in your preferences?

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

Just checked, both applications are selected. Yeah I do have a lot more of the GUPPIs in my task list right now, but I believe that both GPUs crunched those as well before the OS reinstall.

Everything was fine up to last week when I set it to no new work in preparation for a clean OS install.

---

Just checked, both applications are selected. Yeah I do have a lot more of the GUPPIs in my task list right now, but I believe that both GPUs crunched those as well before the OS reinstall.

Everything was fine up to last week when I set it to no new work in preparation for a clean OS install.

Well, the other thing I see is you have 105 error tasks.....that could perhaps have something to do with it.
I have to run now, hopefully somebody can help you dig into this a bit further.

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

Ah, good catch. Unfortunately those are mostly CPU tasks that failed to download. Probably started BOINC too soon, had a bunch of reboots to get both GPUs working correctly.

---

Just checked, both applications are selected. Yeah I do have a lot more of the GUPPIs in my task list right now, but I believe that both GPUs crunched those as well before the OS reinstall.

Everything was fine up to last week when I set it to no new work in preparation for a clean OS install.

Did you Upgrade to Ubuntu 16.04? If you did you might want to start reading here, Ubuntu/AMD completely changed the drivers with 16.04...
OpenCL Platform Name: Clover
Version: OpenCL 1.1 MESA 11.2.0
Yep that's definitely the Open Source driver. This one is the punch line;

Multibeam plan classes are set up so that only AMD cards running with Catalyst/fglrx match them. I think there is one Astropulse plan class that can match Mesa but Astropulse AMD app doesn't work with Mesa. It's been a while since I last looked at it but I think the app is using some AMD specific compiler switch that Mesa doesn't like. Multibeam app is probably using the same compiler switches so I would keep my expectations low even if you go anon platform way.
https://setiathome.berkeley.edu/forum_thread.php?id=80717&postid=1838338#1838338

It's probably not going to work with the Open Source driver. The only other choice is to go back to a pre-16.04 OS.

Oh man, thanks for the info. It's mint 18 so based on ubuntu 16.04. I tried to get catalyst installed, but it really wasn't happening.

Might be time to replace that card with a second NVIDIA and save the headaches of managing both.

---

Sorry to bring up an older thread, but it looks like there has been some new AP work, and this cruncher decided to grab some for the ATI card. All of them quickly errd out. Might be a good idea to mask the open source ATI drivers for AP as well.

Has anyone had any luck getting these cards to crunch with the MESA driver yet?

Another issue, I decided to add a second project to this host with a low resource share just to keep it fed when there is no work here. Sometimes it will crunch enough of those so that it wont at least let SETI crunch 2 concurrent GPU applications (1 core per GPU app) while it uses the CPU. I added an app_config file to it, limiting that project to 6 concurrent threads, but is there a better way to prioritize the GPU apps so that if it runs out of SETI GPU work it will then let the other project have all the cores?

---

Here is the dump from one of the tasks:

<core_client_version>7.6.31</core_client_version>
<![CDATA[
<message>
process exited with code 193 (0xc1, -63)
</message>
<stderr_txt>
OpenCL platform detected: Mesa
OpenCL platform detected: NVIDIA Corporation
WARNING: BOINC supplied wrong platform!
Number of OpenCL devices found : 1
BOINC assigns slot on device #0.
Info: BOINC provided OpenCL device ID used
SIGSEGV: segmentation violation
Stack trace (7 frames):
../../projects/setiathome.berkeley.edu/astropulse_7.08_x86_64-pc-linux-gnu__opencl_ati_100(boinc_catch_signal+0x4d)[0x4c6a6d]
/lib/x86_64-linux-gnu/libpthread.so.0(+0x11390)[0x7fe9bb5a9390]
/usr/lib/x86_64-linux-gnu/libnvidia-opencl.so.1(+0xbcf45)[0x7fe9b44bcf45]
../../projects/setiathome.berkeley.edu/astropulse_7.08_x86_64-pc-linux-gnu__opencl_ati_100[0x485789]
../../projects/setiathome.berkeley.edu/astropulse_7.08_x86_64-pc-linux-gnu__opencl_ati_100[0x46a31f]
/lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf0)[0x7fe9ba500830]
../../projects/setiathome.berkeley.edu/astropulse_7.08_x86_64-pc-linux-gnu__opencl_ati_100[0x40bda9]

Exiting...

</stderr_txt>
]]>

Odd that it thinks BOINC lied to it.

---