sPHENIX discussion of software

["Huang, Jin" <jhuang AT bnl.gov>]

Hi Everyone

Third and (hopefully) last email on this topic. The valgrind memory test on Jenkins CI has also been moved from gcc12 to gcc14, thanks to Chris' work cleaning up the new found memory issues. 

This completes the Jenkins CI migration from gcc12 to gcc14 for all checks. Let me know if you bump into any problem. 

Meanwhile, the gcc12 compilation and default macro test will still be included in the CI checks, until we officially drop the support for gcc12/SL7 OS. 

Cheers

Jin

 

______________________________

 

Jin HUANG

 

Physicist, Ph.D.

Brookhaven National Laboratory

Physics Department, Bldg 510 C

Upton, NY 11973-5000

 

Office: 631-344-5898

Cell:   757-604-9946

______________________________

 

---

From: Huang, Jin <jhuang AT bnl.gov>
Sent: Wednesday, December 18, 2024 10:33 PM
To: sphenix-software-l AT lists.bnl.gov <sphenix-software-l AT lists.bnl.gov>; Pinkenburg, Christopher <pinkenbu AT bnl.gov>
Subject: Re: gcc14/alma9 is now part of the continuous integration tests

 

Hi Everyone

Following Chris' work, this is to confirm the next group of test in Jenkins CI has been moved from gcc12/SL7-based to the new gcc14/alma9-based environment. These test include 

• Calorimeter QA, which has been statistically consistent between gcc14 relative to gcc12. Here are the comparison plot: https://github.com/sPHENIX-Collaboration/coresoftware/runs/34628279106 
• Tracking QA which remained stable too: https://github.com/sPHENIX-Collaboration/coresoftware/runs/34629073207 , https://github.com/sPHENIX-Collaboration/coresoftware/runs/34628261882  
• Scan build. 42 warnings remain and they will be suppressed in the newer builds: https://web.sdcc.bnl.gov/jenkins-sphenix/job/sPHENIX/job/Build-ScanBuild-gcc14/3/clang-analyzer/ 

Let me know if this bring any new issue to you. 

The Valgrind and clang tests are still with gcc12/SL7 as Chris are working through some of the remaining issue identified in the new test. 

Cheers

Jin

 

______________________________

 

Jin HUANG

 

Physicist, Ph.D.

Brookhaven National Laboratory

Physics Department, Bldg 510 C

Upton, NY 11973-5000

 

Office: 631-344-5898

Cell:   757-604-9946

______________________________

 

---

From: Huang, Jin <jhuang AT bnl.gov>
Sent: Tuesday, December 10, 2024 12:15 PM
To: sphenix-software-l AT lists.bnl.gov <sphenix-software-l AT lists.bnl.gov>; Pinkenburg, Christopher <pinkenbu AT bnl.gov>
Subject: gcc14/alma9 is now part of the continuous integration tests

 

Hi Everyone

As we chatted in the last software meeting, our software will be tested under the new gcc14/alma9 environment with our Jenkins Continuous Integration (CI) setup, following Chris' hard work allowing the code to run in the new environment. 

This is to confirm compilation and simple macro test on gcc14/alma9 is now part of the default CI test, which is applicable to new pull request and daily master branch checks. Let me know if you bump into any trouble due to this change. 

Full gcc14/alma9 test on CI does reveal more issue that need to be resolved, which I am working with Chris and Joe on. Here is a preview:

https://web.sdcc.bnl.gov/jenkins-sphenix/job/sPHENIX/job/Build-Master-gcc14/16/

Once these issues are resolved, they will also be added to the default test gradually along with the overall computing migration to Alma9. 

Cheers,

Jin

 

______________________________

 

Jin HUANG

 

Physicist, Ph.D.

Brookhaven National Laboratory

Physics Department, Bldg 510 C

Upton, NY 11973-5000

 

Office: 631-344-5898

Cell:   757-604-9946

______________________________

 

---

From: sphenix-software-l-request AT lists.bnl.gov <sphenix-software-l-request AT lists.bnl.gov> on behalf of pinkenburg <pinkenburg AT bnl.gov>
Sent: Thursday, December 5, 2024 5:05 PM
To: sphenix-software-l AT lists.bnl.gov <sphenix-software-l AT lists.bnl.gov>
Subject: [[Sphenix-software-l] ] alma9 interactive nodes for testing

 

Hi folks,

we have two new Alma9 interactive nodes, sphnxsub01 and sphnxsub02. We
have a nightly build (no ana build, so everything defaults to the latest
and greatest). As a nice carrot - condor jobs submitted from those nodes
run on our Alma9 cluster which has 32k cores and you will out rank the
sims running there.

Don't try to run SL7 code here (or under condor) - this very likely
won't work and just wastes resources. SL7 Containers work - we do not
need to test that. Please let me know any problems you encounter, I have
concentrated on running the sims which currently do run in an SL7 container.

While at it - sdcc plans a renaming of pretty much all nodes, makes
their life easier. Since interactive hosts also submit to condor
sphnxsubXX would be the new names. I am not sure if it's worth pulling a
fuss about the name but if anyone has a good (technical) reason why this
would cause us hardships, let me know. Soon - latest by the beginning of
the collaboration meeting.

Chris

--
*************************************************************

Christopher H. Pinkenburg       ;    pinkenburg AT bnl.gov
                                ;    http://www.phenix.bnl.gov/~pinkenbu

Brookhaven National Laboratory  ;    phone: (631) 344-5692
Physics Department Bldg 510 C   ;    fax:   (631) 344-3253
Upton, NY 11973-5000

*************************************************************