I installed build-essential which includes gcc on my mint 18.3 clean install.

My version of build-essential on my Ubuntu 16.04 and 18.04 machines is the latest version (12.4ubuntu1), and it still does not work.

Thanks to Juha's debugging, we hopefully are about to fix the "svml_sin4" problem. We'll need to cancel the running workunits though.

t

@boboviz I had the same problem on a VM because it had gcc but was missing the C libraries. Try `sudo apt-get install gcc` which should install the libs as well.

As on the python2/3 subject, the system python should not matter for us. We use a conda environment where we install python3 which is what our programs need.

On the gdma issue. I just sent out a single workunit with fixed gmda version as a test. If it comes back fine I'll cancel all current WUs and send out new ones.

It's like a little nightmare.

- VirtualBox VM with Mint 17.3
- sudo apt-get update/upgrade
- sudo apt-get install gcc ("gcc is already the newest version")
- sudo apt-get install buid essential
- wget http://launchpadlibrarian.net/109052632/python-support_1.0.15_all.deb
- sudo dpkg -i python-support_1.0.15_all.deb
- sudo apt-get install boinc-client boinc-manager (7.2.42)
- gcc -v = 4.8.4
- python --version: 2.7.6

And...svml_sin4!!

16:46:38 (2487): wrapper: running /var/lib/boinc-client/projects/www.gpugrid.net/miniconda/bin/python (pre_script.py)
Traceback (most recent call last):
File "/var/lib/boinc-client/projects/www.gpugrid.net/miniconda/envs/qmml/lib/python3.6/site-packages/psi4/__init__.py", line 55, in <module>
from . import core
ImportError: /var/lib/boinc-client/projects/www.gpugrid.net/miniconda/envs/qmml/lib/python3.6/site-packages/psi4/core.so: undefined symbol: __svml_sin4

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "run-qmml.py", line 12, in <module>
import psi4
File "/var/lib/boinc-client/projects/www.gpugrid.net/miniconda/envs/qmml/lib/python3.6/site-packages/psi4/__init__.py", line 60, in <module>
raise ImportError("{0}".format(err))
ImportError: /var/lib/boinc-client/projects/www.gpugrid.net/miniconda/envs/qmml/lib/python3.6/site-packages/psi4/core.so: undefined symbol: __svml_sin4
Traceback (most recent call last):
File "pre_script.py", line 31, in <module>
raise Exception("Error running qmml")
Exception: Error running qmml
16:51:34 (2487): $PROJECT_DIR/miniconda/bin/python exited; CPU time 132.140000
16:51:34 (2487): app exit status: 0x1
16:51:34 (2487): called boinc_finish(195)

I think i'll wait the Windows version....

Hm sorry I can't help you with this error then :( We are working on the Windows version though so hopefully in a month or two (or three)...

So I fixed the gdma version and my first WU went through and computed fine. Can't guarantee that this will fix the problem on all machines as it seemed like the machine that computed my WU already had a working conda install but I'll go ahead and send out more of my WUs and see what happens.

I cancelled the remaining old WUs and sent out 250 of my new ones. Let's see if it solves the issues. If everything goes fine I will submit many more.

Okay so the error % is at 10% currently with 120 jobs completed and 13 failed.
The ones that failed, did so on users which already had completed lots more jobs so I'll scratch it up to chance or single bad host of those users as the errors are not consistent.

I am currently uploading thousands of jobs. Happy crunching :)

I am currently uploading thousands of jobs. Happy crunching :)

I'll try this new version.

Give it a try, although I doubt it will solve the gcc problem you have (but one can always hope)

Give it a try, although I doubt it will solve the gcc problem you have (but one can always hope)

It runs!! It runs!!
Only a little error (CondaValueError: prefix already exists: /var/lib/boinc-client/projects/www.gpugrid.net/miniconda/envs/qmml), but "completed and validated"

I've finished 34 of the new Work Units and all are valid.

@boboviz as far as I can tell this error is ignored.
Happy to hear that it's running again though :)

CPU tasks seem to run well on my HP Laptop with SuSE Linux Leap 42.3. Thy used to fail.
Tullio