ppsim segfaults under simtest

[Sebastian Jester]

I build ppsim from this svn export:

svn export -r {2009-04-14} https://svn.pan-starrs.ifa.hawaii.edu/repo/ipp/trunk/ ./svn-20090414

I run simtest with

 simtest.setup.basic
 simtest simtest_20090414 aida77 new

I get segfaults in simtest for all ppsim calls from simtest.mkimages line 41 and above, i.e. for FLAT and OBJECT frames:

./simtest.mkimages: line 41: 28276 Segmentation fault      ppSim -camera SIMTEST -type FLAT -filter r -exptime 0.100000 raw/simtest.002.000
./simtest.mkimages: line 42: 28277 Segmentation fault      ppSim -camera SIMTEST -type FLAT -filter r -exptime 0.100000 raw/simtest.002.001
...
./simtest.mkimages: line 61: 28296 Segmentation fault      ppSim -camera SIMTEST -type OBJECT -filter r -exptime 10.000000 -skymags 20.860001 -ra 270.731812 -dec -23.708334 -pa 0.000000 -seeing 0.828261 raw/simtest.004.000
./simtest.mkimages: line 62: 28297 Segmentation fault      ppSim -camera SIMTEST -type OBJECT -filter r -exptime 10.000000 -skymags 20.860001 -ra 270.731812 -dec -23.691668 -pa 0.000000 -seeing 0.875310 raw/simtest.004.001
...

Here's what happens when I run ppsim in gdb:

Starting program: /disk1/jester/IPP/svn-20090414.lin64/bin/ppSim -camera SIMTEST -type OBJECT -filter r -exptime 10.0000 -skymags 20.860001 -ra 270.731812 -dec -23.708334 -pa 0.000000 -seeing 0.828261 simtest.004.000
[Thread debugging using libthread_db enabled]
[New Thread 47517614728512 (LWP 28338)]

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 47517614728512 (LWP 28338)]
0x00002b378c4d0953 in isBadMemBlock (output=0x2b378e8b6840, memBlock=0x3a4, 
    file=0x2b378c6a37d9 "psList.c", lineno=538, 
    func=0x2b378c6a3af6 "psListToArray") at psMemory.c:454
454         if (memBlock->refCounter < 1) {

[Paul Price]

I can't reproduce this problem, using ppSim built from 2009-04-17. I suggest that your build is bad or the problem has already been fixed. Try rebuilding using "psbuild -dev -rebuild -clean".

price@mithrandir:/home/mithrandir/price/temp/test>ppSim -camera SIMTEST -type OBJECT -filter r -exptime 10.0000 -skymags 20.860001 -ra 270.731812 -dec -23.708334 -pa 0.000000 -seeing 0.828261 simtest.004.000
2009-04-21 01:15:02Z|mithrandir |4|psastro

no supplied reference header data

output catalog has 7191 stars (35955 measures, 5 secfilt)
SUCCESS
2009-04-21 01:15:12Z|mithrandir |I|psastro

ran getstar : 10.132289 sec

2009-04-21 01:15:13Z|mithrandir |I|psastro

loaded 7191 reference stars : 0.657630 sec

2009-04-21 01:15:13Z|mithrandir |I|ppSim

Adding 7191 reference stars

2009-04-21 01:15:13Z|mithrandir |4|psastro

fitting 166 points to luminosity function

2009-04-21 01:15:13Z|mithrandir |4|psastro

logN = -0.764474 + 0.107514 mag

2009-04-21 01:15:13Z|mithrandir |4|psastro

logN vs mag residuals: -0.060952 +/- 0.289601

2009-04-21 01:15:13Z|mithrandir |I|ppSim

Adding 0 stars between 15.500000 and 21.020388 mag

2009-04-21 01:15:17Z|mithrandir |W|pmConfigMaskSetInMetadata (pmConfigMask.c:64)

problem with mask value LOW

2009-04-21 01:15:17Z|mithrandir |W|pmConfigMaskSetInMetadata (pmConfigMask.c:64)

problem with mask value SUSPECT

2009-04-21 01:15:17Z|mithrandir |W|pmConfigMaskSetInMetadata (pmConfigMask.c:64)

problem with mask value CONV.BAD

2009-04-21 01:15:17Z|mithrandir |W|pmConfigMaskSetInMetadata (pmConfigMask.c:64)

problem with mask value CONV.POOR

[Sebastian Jester]

Just to confirm for posterity that the segfault indeed has disappeared by 2009-04-17, but is reproducible with the 2009-04-14 install. The 'mask' complaints arise because I hadn't updated my system.config to point to masks.16bit.config which was new since my last update before 2009-04-14.