[CCPPETMR/SIRF] travis gadgetron tests with DEVEL_BUILD=ON fail (#158)

0 views
Skip to first unread message

Kris Thielemans

unread,
Mar 23, 2018, 7:25:06 PM3/23/18
to CCPPETMR/SIRF, Subscribed

End of log is

$ cat gadgetron.log
terminate called after throwing an instance of 'std::logic_error'
  what():  basic_string::_S_construct null not valid

this is saying that a std::string was initialised with a NULL ptr.

as other configurations are ok this is likely a gadgetron master problem.


You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub, or mute the thread.

Casper da Costa-Luis

unread,
Mar 26, 2018, 10:00:09 AM3/26/18
to CCPPETMR/SIRF, Subscribed

might be worth opening an issue on the gadgetron repo?

evgueni-ovtchinnikov

unread,
Mar 26, 2018, 10:02:17 AM3/26/18
to CCPPETMR/SIRF, Subscribed

I did (#596)

Casper da Costa-Luis

unread,
Mar 26, 2018, 10:30:36 AM3/26/18
to CCPPETMR/SIRF, Subscribed

Edoardo Pasca

unread,
May 2, 2018, 11:39:35 AM5/2/18
to CCPPETMR/SIRF, Subscribed

but some of the builds with DEVEL_BUILD=ON did not fail. This one for instance.

Casper da Costa-Luis

unread,
May 2, 2018, 12:20:58 PM5/2/18
to CCPPETMR/SIRF, Subscribed

linux ones fail, osx ones pass (funny - usually it's the other way around)

Hui Xue

unread,
May 3, 2018, 9:30:18 AM5/3/18
to CCPPETMR/SIRF, Subscribed

do you have the entire gadgetron.log to share?

Casper da Costa-Luis

unread,
May 3, 2018, 11:55:17 AM5/3/18
to CCPPETMR/SIRF, Subscribed

@xueh2 unfortunately as per the original description, the only thing in the gadgetron.log is the std::logic_error. gadgetron throws this error and crashes immediately upon startup.

$ gadgetron > gadgetron.log
$ cat gadgetron.log
terminate called after throwing an instance of 'std::logic_error'
  what():  basic_string::_S_construct null not valid

Kris Thielemans

unread,
May 3, 2018, 4:49:36 PM5/3/18
to CCPPETMR/SIRF, Subscribed

@xueh2, this might have been a ccache problem with our builds. I think it has disappeared now but not sure as we're now having a bug ourselves. Will let you know.

Casper da Costa-Luis

unread,
May 3, 2018, 5:16:21 PM5/3/18
to CCPPETMR/SIRF, Subscribed

just completely cleared my ccache and rebuilt, still have the same error (using gadgetron/gadgetron@73ab079)

Kris Thielemans

unread,
May 3, 2018, 5:50:38 PM5/3/18
to CCPPETMR/SIRF, Subscribed

yes. also on Travis. @casperdcl, can you specify your OS and environment?

Casper da Costa-Luis

unread,
May 3, 2018, 8:21:26 PM5/3/18
to CCPPETMR/SIRF, Subscribed

ubuntu 16.04 docker (ccppetmr/sirf:latest) running on top of ubuntu 16.04 (x64), with the following steps:

/opt/SIRF-SuperBuild$ ccache --clear
/opt/SIRF-SuperBuild$ git clean -xdf && rm -r sources
/opt/SIRF-SuperBuild$ git checkout master && git fetch && git reset --hard origin/master
# to use `master` for all (incl `gadgetron`, `SIRF`, etc.):
/opt/SIRF-SuperBuild$ cmake -DDEVEL_BUILD=ON \
  -DCMAKE_BUILD_TYPE=Release \
  -DBUILD_STIR_WITH_OPENMP=ON \
  -DUSE_SYSTEM_Armadillo=ON -DUSE_SYSTEM_Boost=ON -DUSE_SYSTEM_FFTW3=ON \
  -DUSE_SYSTEM_HDF5=ON -DBUILD_siemens_to_ismrmrd=ON -DUSE_SYSTEM_SWIG=ON \
  . && make -j
/opt/SIRF-SuperBuild$ gadgetron

Kris Thielemans

unread,
May 4, 2018, 2:14:15 AM5/4/18
to CCPPETMR/SIRF, Subscribed

This is due to a missing GADGETRON_HOME env variable. Let's see first what upstream says.

Kris Thielemans

unread,
May 4, 2018, 1:28:17 PM5/4/18
to CCPPETMR/SIRF, Subscribed

closing as will fix with CCPPETMR/SIRF-SuperBuild#115

Kris Thielemans

unread,
May 4, 2018, 1:28:17 PM5/4/18
to CCPPETMR/SIRF, Subscribed

Closed #158.

Reply all
Reply to author
Forward
0 new messages