#65 Crux Bullseye Segfaults when Hardklor finds no results

[Alex Hu]

I've been trying to run bullseye through Crux, and I get a core segmentation fault and core dump. I showed Kaipo the bug, and he wasn't able to figure out what was going on with a short investigation. Attached is the configure file for Hardklor that contains the paths to all of the input files and the parameters used.

Here is the backtrace:

0 std::basic_fstream<char, std::char_traits\<char=""> >::open(char const*, std::_Ios_Openmode) () at /usr/src/gcc-4.8.1/build/x86_64-unknown-linux-gnu/libstdc++-v3/include/fstream:894</char,>

1 0x0000000000ace985 in CruxHardklorApplication::hardklorMain(int, char**) ()

2 0x0000000000ac795d in CruxHardklorApplication::main(std::basic_string<char, std::char_traits\<char="">, std::allocator\<char> > const&) ()</char></char,>

3 0x00000000008e4dc1 in CruxBullseyeApplication::main(int, char**) ()

4 0x00000000009092f7 in CruxApplicationList::main(int, char**) ()

5 0x0000000000875630 in main ()

I think it might have to do with the fact that if I run my data on Hardklor separately, I get an output that is empty of isotope distributions. Perhaps bullseye is looking for something that isn't there and thus returns a segfault.

[William S Noble]

Alex, you should produce a tar file that reproduces this problem and then hand this issue off to Kaipo.

[Alex Hu]

The command that causes the segfault is below:

crux bullseye --overwrite T --output-dir /net/noble/vol3/user/alexhu/proj/crux-projects/2013MouseHeartProteome/data/2014-02-27/bullseye_runs /net/noble/vol3/user/alexhu/proj/crux-projects/2013MouseHeartProteome/data/ms1/82593_lv_mcx_DIA_5mz_525to650.ms1 /net/gs/vol1/home/alexhu/proj/crux-projects/2013MouseHeartProteome/data/charged/centroided/82593_lv_mcx_DIA_5mz_525to650.ms2

I gave the full directory path to those files. The output directory is here:

/net/noble/vol3/user/alexhu/proj/crux-projects/2013MouseHeartProteome/data/2014-02-27/bullseye_runs

Which contains the empty hardklor.mono.txt file.

[Alex Hu]

I also added a tar file in that directory that contains the ms1 and ms2 files that cause the segfault.

[Alex Hu]

The code changes look good, and it fixes the problem. Thanks! The code review is approved.

[Kaipo]

Fixed in r16325