GYRE 6.0 libgomp issues?

[mjoyce]

Hi Rich et al.,
I'm trying to use the attached inlist with GYREv6.0 and a MESA model (v15140) in FGONG format. I receive the attached error log, which includes many segfaults and references to gcc/libgomp/ and glibc-2.27/nptl/

These parameter settings are imported from a previous template that worked well with similar stars (and whose keywords have been adjusted where necessary in accordance with GYRE 6.0 docs, to the best of my knowledge), so I'm not sure what the issue is. Other posts on this forum suggest it might have to do with my system (?). I'm running Ubuntu 18.04 on the 18 core HPC Lovelace (same machine used to run MESA tests).

Thanks,

[rhtownsend]

Hi Meridith --

The input file you sent me still has a number of template fields -- e.g., <<freq_min>>. Can you send an example with these filled in (pref. the crashing example).

cheers,

Rich

[mjoyce]

whoops you're right, sorry! Functional inlist attached

[rhtownsend]

Thanks. It appears the crash arises from requesting 'eta' as one of the adiabatic summary-file items -- for adiabatic calculations, this quantity (the Stellingwerf normalized growth rate) is undefined. So, as a workaround, remove 'eta' from the summary_item_list in &ad_output.

I'll also be working on a fix so that the code complains rather than crashing.

[mjoyce]

yup that worked, thanks!

I do note that eta is listed as a valid output option at
https://gyre.readthedocs.io/en/stable/r ... tml#only-n

without warning that this is a nad parameter only, but perhaps a more informed user/real asteroseismologist would recognize that this does not make sense for adiabatic calculations

[rhtownsend]

yup that worked, thanks!

I do note that eta is listed as a valid output option at
https://gyre.readthedocs.io/en/stable/r ... tml#only-n

without warning that this is a nad parameter only, but perhaps a more informed user/real asteroseismologist would recognize that this does not make sense for adiabatic calculations

I think there *is* a warning -- look at footnote 1. But maybe I'm missing something?

[mjoyce]

Yup, OK! I accept defeat