Segmentation fault with the plk graphics plugin

[Former user]

Hi,

I managed to install pgplot (version 5.2) and tempo2-2020.11.1 on my mac OS catalina (version 10.15.7) without any issues or errors. However, I then tested the installation of tempo2 on the available example data files by running:

tempo2 -gr plk -f example1.par example1.tim

This gave me a segmentation fault (most definitely attributed to the plk graphics plugin), with the pgplot window opening but not actually plotting any data (please see attached figure "plkimage.png"). The fitting functions on the other hand appear to work perfectly well. Below is the full output I got after running the above command:

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

RMS pre-fit residual = 669.858 (us), RMS post-fit residual = 637.057 (us) Number of fit parameters: 5 Number of points in fit = 60 Offset: -2.45049e-05 0.000115407 offset_e*sqrt(n) = 0.00089394 n = 60

PARAMETER Pre-fit Post-fit Uncertainty Difference Fit

RAJ (rad) 2.74220314309176 2.74220375131 5.1242e-07 6.0822e-07 Y RAJ (hms) 10:28:28.0 10:28:28.0083636 0.0070462 0.0083636
DECJ (rad) -1.01784305243278 -1.01784350715407 2.6226e-07 -4.5472e-07 Y DECJ (dms) -58:19:05.2 -58:19:05.29379 0.054094 -0.093793
F0 (s^-1) 10.9405323 10.9405323000193 3.6463e-11 1.9298e-11 Y F1 (s^-2) -1.93e-12 -1.93000045532929e-12 8.2888e-19 -4.5533e-19 Y PEPOCH (MJD) 54562 54562 0 0 N POSEPOCH (MJD) 54562 54562 0 0 N DMEPOCH (MJD) 54562 54562 0 0 N DM (cm^-3 pc) 96.525 96.525 0 0 N START (MJD) 0 53999.9949438211 0 54000 N FINISH (MJD) 0 56000.5824517228 0 56001 N TZRMJD 0 54983.3545636678 0 54983 N TZRFRQ (MHz) 0 1440 0 1440 N TZRSITE 7
TRES 0 637.05689959428 0 637.06 N EPHVER TEMPO2 TEMPO2 0 0 N DM_SERIES TAYLOR

---

Derived parameters:

P0 (s) = 0.091403230901136 3.0463e-13
P1 = 1.61242864990627e-14 6.9249e-21
tau_c (Myr) = 0.089876 bs (G) = 1.2285e+12

Total time span = 2000.588 days = 5.477 years

Tempo2 usage Units: TCB (tempo2) Time ephemeris: IF99 (tempo2) Troposphere corr.? Yes (tempo2) Dilate freq? Yes (tempo2) Electron density (1AU) 4 Solar system ephem DE405 Time scale TT(TAI) Binary model NONE Notice: There were 2 warnings. Summaries are shown below, check logs for full details. Warning #1: Assuming site '7' means 'pks' Warning #2: [DUP1] duplicated warnings have been suppressed. Segmentation fault: 11

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Is it possible that I have not installed some of the required pgplot drivers (please see attached the "drivers.list" file that I used when compiling pgplot)?

Any help would be greatly appreciated.

Thanks a lot, Raj.

[Michael Keith]

Some things to try:

• Run the pgplot demo code with the “/xs” device. (usually installed as e.g. pgdemo1, but you might have to find it in the pgplot directories if you did not install everything from pgplot).
• Try running with a different par/tim file. Try running with “-nofit” in case fitting is making some parameters that break plk somehow (unlikely, but it can happen)
• Run tempo2 through a debugger (gdb on linux or lldb on macosx). Try to find where it segfaults - is it part of the tempo2 code, plk code or pgplot.

‌