64-bit nightly test logs have different StdEdxY2 messages

[genevb]

I am not sure whether this leads to any actual differences in results, but the log files from nightly tests show some different messages from StdEdxY2Maker between 32-bit and 64-bit.

Some jobs have lines like this for 64-bit, but not 32-bit:

StdEdxY2Maker:WARN  - StTpcdEdxCorrection:           TpcEffectivedX/                      dEdx correction wrt Bichsel parameterization	Validity:20160207.000102 --- 20160618.000102
StdEdxY2Maker:WARN  - StTpcdEdxCorrection:   TpcLengthCorrectionMDF/  Variation vs Track length and <log2(dX)> and rel. error in dE/dx	Validity:20160207.000216 --- 20160512.000207

An example of this difference is seen like this:

diff -a /star/rcf/test/dev/daq_sl302.stica/Sat/year_2016/AuAu200_production_2016{,_64bit}/*log | grep StdEdxY2

or

diff -a /star/rcf/test/dev/daq_sl302.ittf/Sat/year_2007/2007ProductionMinBias{,_64bit}/*log | grep StdEdxY2

Other tables also have messages like these in some of the jobs, including TpcLengthCorrectionB, TpcPadCorrectionMDC, TpcLengthCorrectionMDN.

Some of the jobs have no such message differences, such as:

diff -a /star/rcf/test/dev/daq_sl302.stica/Sat/year_2021/production_3p85GeV_fixedTarget_2021{,_64bit}/*log | grep StdEdxY2

It would be good to understand the reason for the different messages, even if it is inconsequential for the output data.

-Gene

[genevb]

Another perhaps related observation...
64-bit nightly tests for simulations show these messages appearing earlier for StTpcRSMaker and then again in StdEdxY2Maker. And there are some table messages in 32-bit that aren't there in 64-bit:

> diff -a /star/rcf/test/dev/simu/Thu/year_2014/auau200_minbias{,_64bit}/*s.log | grep -B1 Valid
1669a1670,1673
> StTpcRSMaker:WARN  - StTpcdEdxCorrection:           TpcEffectivedX/                      dEdx correction wrt Bichsel parameterization	Validity:20000101.000000 --- 20160207.000102
> StTpcRSMaker:WARN  - StTpcdEdxCorrection:     TpcLengthCorrectionB/        Variation vs Track length and relative error in Ionization	Validity:20131210.000100 --- 20131220.000100
> StTpcRSMaker:WARN  - StTpcdEdxCorrection:   TpcLengthCorrectionMDF/  Variation vs Track length and <log2(dX)> and rel. error in dE/dx	Validity:20130305.000013 --- 20140101.000008
--
< StdEdxY2Maker:WARN  - St_TpcEffectivedXC::instance found table TpcEffectivedX with NRows = 1 in db
< StdEdxY2Maker:WARN  - Validity:20000101/0 -----   20160207/102
--
< StdEdxY2Maker:WARN  - St_MDFCorrectionC::instance found table TpcLengthCorrectionMDF with NRows = 7 in db
< StdEdxY2Maker:WARN  - Validity:20130305/13 -----   20140101/8
4181a4174,4177
> StdEdxY2Maker:WARN  - StTpcdEdxCorrection:           TpcEffectivedX/                      dEdx correction wrt Bichsel parameterization	Validity:20000101.000000 --- 20160207.000102
> StdEdxY2Maker:WARN  - StTpcdEdxCorrection:     TpcLengthCorrectionB/        Variation vs Track length and relative error in Ionization	Validity:20131210.000100 --- 20131220.000100
> StdEdxY2Maker:WARN  - StTpcdEdxCorrection:   TpcLengthCorrectionMDF/  Variation vs Track length and <log2(dX)> and rel. error in dE/dx	Validity:20130305.000013 --- 20140101.000008

Same is true in these simulation logs:

> diff -a /star/rcf/test/dev/simu/Thu/year_2015/pp200_minbias{,_64bit}/*s.log | grep -B1 Valid
...
> diff -a /star/rcf/test/dev/simu/Thu/year_2016/dAu200_minbias{,_64bit}/*s.log

---

Sidenote: in the original post for this issue, I compared logs from running 32-bit from AFS and 64-bit from NFS4. So I checked 64-bit from AFS and the observations in both the original post and this one are true for both cases, AFS or NFS4, so it is not an issue specific to NFS4.

[fisyak]

The difference comes from difference in build of StdEdxY2Maker on afs and nfsv4
QA :INFO - Library libStdEdxY2Maker [ dEdxY2] (/afs/rhic.bnl.gov/star/packages/DEV/.sl73_gcc485/lib/libStdEdxY2Maker.so) is loaded |
QA :INFO - Library libStdEdxY2Maker [ dEdxY2] (/star/nfs4/AFS/star/packages/DEV/.sl73_x8664_gcc485/lib/libStdEdxY2Maker.so) is loaded

QA :INFO - QAInfo:StdEdxY2Maker built Feb 7 2025 23:25:37 from Tag $Name: $ $Id: StdEdxY2Maker.h,v 1.33 2019/11/29 19:00:08 fisyak Exp $ |
QA :INFO - QAInfo:StdEdxY2Maker built Mar 29 2025 11:44:34 from Tag $Name: $ $Id: StdEdxY2Maker.h,v 1.33 2019/11/29 19:00:08 fisyak Exp $

[genevb]

Thanks for the comment, Yuri, but these difference are there for 32-bit vs. 64-bit for AFS-built libraries. It appears unrelated to the NFSv4 vs. AFS builds.

> diff -a /star/rcf/test/dev/daq_sl302.stica/Tue/year_2016/AuAu200_production_2016{,_64bit}/*log | grep -a StdEdxY2
< QA :INFO  - Library libStdEdxY2Maker       [         dEdxY2] (/afs/rhic.bnl.gov/star/packages/DEV/.sl73_gcc485/lib/libStdEdxY2Maker.so)       is loaded
< QAInfo: doPs for     libStdEdxY2Maker:        Load             total =362.558594 heap = 10.517708 and   0.099625( +0.004478)
> QA :INFO  - Library libStdEdxY2Maker       [         dEdxY2] (/afs/rhic.bnl.gov/star/packages/DEV/.sl73_x8664_gcc485/lib/libStdEdxY2Maker.so) is loaded
> QAInfo: doPs for     libStdEdxY2Maker:        Load             total =679.667969 heap = 17.041901 and   0.138351( +0.007401)
< QA :INFO  - QAInfo:StdEdxY2Maker        built Feb  7 2025 23:25:37 from Tag $Name:  $ $Id: StdEdxY2Maker.h,v 1.33 2019/11/29 19:00:08 fisyak Exp $ 
> QA :INFO  - QAInfo:StdEdxY2Maker        built Feb  8 2025 01:24:58 from Tag $Name:  $ $Id: StdEdxY2Maker.h,v 1.33 2019/11/29 19:00:08 fisyak Exp $ 
> StdEdxY2Maker:WARN  - StTpcdEdxCorrection:           TpcEffectivedX/                      dEdx correction wrt Bichsel parameterization        Validity:20160207.000102 --- 20160618.000102
> StdEdxY2Maker:WARN  - StTpcdEdxCorrection:   TpcLengthCorrectionMDF/  Variation vs Track length and <log2(dX)> and rel. error in dE/dx        Validity:20160207.000216 --- 20160512.000207
< QA :INFO  - QAInfo:Maker        StdEdxY2Maker::dEdxY2               Ast =434.33( 1.6%) Cpu =432.26( 1.6%) 
> QA :INFO  - QAInfo:Maker        StdEdxY2Maker::dEdxY2               Ast =592.67( 1.4%) Cpu =575.91( 1.4%) 
<             StdEdxY2Maker(dEdxY2)::Make (480)    0.000000    0.231602    0.557678    0.129634
>             StdEdxY2Maker(dEdxY2)::Make (480)    0.000000    0.302273    0.697021    0.169983

[genevb]

Also, I did note exactly this in my last post on this GitHub issue before your post:
"Sidenote: in the original post for this issue, I compared logs from running 32-bit from AFS and 64-bit from NFS4. So I checked 64-bit from AFS and the observations in both the original post and this one are true for both cases, AFS or NFS4, so it is not an issue specific to NFS4."