Hi,
Erik and I have been working on automating the tests of the ET for the upcoming release. The current status can be viewed (work in progress) here:
https://build.barrywardell.net/view/EinsteinToolkitMulti/job/EinsteinToolkit...
This shows any failing tests as "x". You can click on the "x" to see the reason for the failure.
These results are up-to-date as of now. Shall I commit them to the release-info repository (and continue to do so periodically)?
On Fri, Nov 22, 2013 at 06:41:55PM +0100, Ian Hinder wrote:
Erik and I have been working on automating the tests of the ET for the upcoming release. The current status can be viewed (work in progress) here:
https://build.barrywardell.net/view/EinsteinToolkitMulti/job/EinsteinToolkit...
Neat. On some of these machines I currently also run the tests because I didn't know about these. Oh well.
These results are up-to-date as of now. Shall I commit them to the release-info repository (and continue to do so periodically)?
Please, as long as they are newer than existing results. However, with so many failures on hopper and gordon, and also orca, we cannot mention them as 'supported'. Is the reason for this known? Kraken and BW are borderline as well, but the failing tests there are mostly not in critical thorns.
Frank
On Nov 22, 2013, at 10:47 , Frank Loeffler knarf@cct.lsu.edu wrote:
On Fri, Nov 22, 2013 at 06:41:55PM +0100, Ian Hinder wrote:
Erik and I have been working on automating the tests of the ET for the upcoming release. The current status can be viewed (work in progress) here:
https://build.barrywardell.net/view/EinsteinToolkitMulti/job/EinsteinToolkit...
Neat. On some of these machines I currently also run the tests because I didn't know about these. Oh well.
These results are up-to-date as of now. Shall I commit them to the release-info repository (and continue to do so periodically)?
Please, as long as they are newer than existing results. However, with so many failures on hopper and gordon, and also orca, we cannot mention them as 'supported'. Is the reason for this known? Kraken and BW are borderline as well, but the failing tests there are mostly not in critical thorns.
The reasons are not known. I briefly looked at Hopper; it seems the test cases segfault. That may have something to do with too aggressive optimization.
-erik
On 22 Nov 2013, at 18:47, Frank Loeffler knarf@cct.lsu.edu wrote:
On Fri, Nov 22, 2013 at 06:41:55PM +0100, Ian Hinder wrote:
Erik and I have been working on automating the tests of the ET for the upcoming release. The current status can be viewed (work in progress) here:
https://build.barrywardell.net/view/EinsteinToolkitMulti/job/EinsteinToolkit...
Neat. On some of these machines I currently also run the tests because I didn't know about these. Oh well.
These results are up-to-date as of now. Shall I commit them to the release-info repository (and continue to do so periodically)?
Please, as long as they are newer than existing results. However, with so many failures on hopper and gordon, and also orca, we cannot mention them as 'supported'. Is the reason for this known? Kraken and BW are borderline as well, but the failing tests there are mostly not in critical thorns.
I have now committed the current data. This does not include Roland's latest changes; we have to wait for the tests to run first.
users@lists.einsteintoolkit.org
-
Erik Schnetter -
Frank Loeffler -
Ian Hinder