Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Let Entity return geometry as a copy. #334

Open
wants to merge 1 commit into
base: master
Choose a base branch
from

Conversation

blattms
Copy link
Member

@blattms blattms commented Jun 19, 2018

Geometry is lightweight enough and this is what DUNE's grid interface mandates.
Performance for norne stays the same.
Closes PR #328

@blattms
Copy link
Member Author

blattms commented Jun 19, 2018

jenkins run norne please

@atgeirr
Copy link
Member

atgeirr commented Jun 25, 2018

I don't get why this should change any bit of the solution, will rerun Jenkins to be certain it was not a temporary fluke in the midst of a data update or something.

@atgeirr
Copy link
Member

atgeirr commented Jun 25, 2018

jenkins build this please

@atgeirr
Copy link
Member

atgeirr commented Jun 28, 2018

To be clear, I want to eventually merge this, but the failing checks makes me nervous and I want it to be investigated before merging.

@atgeirr
Copy link
Member

atgeirr commented Aug 1, 2018

jenkins build this please

@blattms
Copy link
Member Author

blattms commented Aug 1, 2018 via email

@atgeirr
Copy link
Member

atgeirr commented Aug 1, 2018

The problem should still be there. And I have no idea what the problem really is.

I retested because I could not reproduce locally. Do you get the test failures on your Linux system?

@blattms
Copy link
Member Author

blattms commented Aug 2, 2018

Let me check again.

@blattms
Copy link
Member Author

blattms commented Aug 2, 2018

Yes I do get

/home/mblatt/src/dune/opm-2.6/opm-upscaling/tests/common/test_gravitypressure.cpp(317): error: in "/BaseCase": difference{0.111111} between c[i]{-999.99999999999977} and e[i]{-899.99999999999977} exceeds 0.0001%
/home/mblatt/src/dune/opm-2.6/opm-upscaling/tests/common/test_gravitypressure.cpp(317): error: in "/BaseCase": difference{0.428571} between c[i]{-999.99999999999977} and e[i]{-699.99999999999977} exceeds 0.0001%
...

when running test_gravitypressure. But I am unsure whether this is the same as the error that we got last time.

@blattms
Copy link
Member Author

blattms commented Aug 2, 2018

I have an upcoming fix for upscaling.

@blattms
Copy link
Member Author

blattms commented Aug 2, 2018

jenkind build this with opm-upscaling=261 please

@blattms
Copy link
Member Author

blattms commented Aug 2, 2018

jenkins build this opm-upscaling=261 please

@blattms
Copy link
Member Author

blattms commented Aug 2, 2018

Beware: PR OPM/opm-upscaling#261 should be merged before this one.

@atgeirr
Copy link
Member

atgeirr commented Aug 2, 2018

Thanks for fixing the first test error, it really did expose a bug in the old code. The second error (ignoring the TCPU thing that will be fixed soonish) is still not clear to me. Can you reproduce it? Do you get identical results apart from timing? The solution is not changing much, it is the fact that it changes at all that has me worried.

@blattms
Copy link
Member Author

blattms commented Aug 3, 2018

Not sure what you mean

@atgeirr
Copy link
Member

atgeirr commented Aug 3, 2018

Not sure what you mean

I think you have to be more specific, which sentence(s) is(are) unclear?

@blattms
Copy link
Member Author

blattms commented Aug 3, 2018

Sorry for not being clear. I wondered what the second failure is that you are seeing. I only saw the TCPU stuff.

@atgeirr
Copy link
Member

atgeirr commented Aug 3, 2018

There are two failures remaining, and only one is the TCPU one. In Jenkins only one is visible until you click the Test Result link, it's confusing. Direct link to failed test:

https://ci.opm-project.org/job/opm-grid-PR-builder/71/testReport/(root)/mpi/compareECLFiles_flow_SPE1CASE2_THERMAL/

@blattms blattms closed this Aug 3, 2018
@blattms blattms force-pushed the entity-return-copy-geometry branch from b29accb to 4d17438 Compare August 3, 2018 17:22
@blattms blattms mentioned this pull request Aug 3, 2018
@blattms
Copy link
Member Author

blattms commented Aug 3, 2018

Seems like I messed up this one, too. Was magically closed and has no commits now.

@blattms blattms reopened this Aug 3, 2018
@blattms blattms force-pushed the entity-return-copy-geometry branch from e7aa2af to 84593f2 Compare August 3, 2018 17:46
@blattms
Copy link
Member Author

blattms commented Aug 3, 2018

jenkins build this please

@blattms
Copy link
Member Author

blattms commented Aug 9, 2018

Unfortunately, I am having troubles reproducing the memory errors in test_transmissibilitymultipliers

@akva2
Copy link
Member

akva2 commented Aug 9, 2018

/home/akva/kode/opm/super-build/opm-grid/opm/grid/cpgrid/../CpGrid.hpp: In member function ‘const Vector& Dune::CpGrid::cellCentroid(int) const’:
/home/akva/kode/opm/super-build/opm-grid/opm/grid/cpgrid/../CpGrid.hpp:969:97: warning: returning reference to temporary [-Wreturn-local-addr]
             return current_view_data_->geomVector<0>()[cpgrid::EntityRep<0>(cell, true)].center();

@akva2
Copy link
Member

akva2 commented Aug 9, 2018

==20424==    at 0x8A54EE: double Opm::UgGridHelpers::getCoordinate<Opm::UgGridHelpers::CpGridCentroidIterator<&(Dune::CpGrid::cellCentroid(int) const)> >(Opm::UgGridHelpers::CpGridCentroidIterator<&(Dune::CpGrid::cellCentroid(int) const)>, int) (GridHelpers.hpp:395)
==20424==    by 0x8A0B95: void tpfa_htrans_compute<Dune::CpGrid>(Dune::CpGrid const*, double const*, double*) (TransTpfa_impl.hpp:97)
==20424==    by 0x89B38C: void Opm::DerivedGeology::update<Opm::BlackoilPropsAdFromDeck, Dune::CpGrid>(Dune::CpGrid const&, Opm::BlackoilPropsAdFromDeck const&, Opm::EclipseState const&, double const*) (GeoProps.hpp:125)
==20424==    by 0x896871: Opm::DerivedGeology::DerivedGeology<Opm::BlackoilPropsAdFromDeck, Dune::CpGrid>(Dune::CpGrid const&, Opm::BlackoilPropsAdFromDeck const&, Opm::EclipseState const&, bool, double const*) (GeoProps.hpp:79)
==20424==    by 0x88AD48: TransmissibilityMultipliersCpGrid::test_method() (test_transmissibilitymultipliers.cpp:282)
==20424==    by 0x88AB3A: TransmissibilityMultipliersCpGrid_invoker() (test_transmissibilitymultipliers.cpp:257)
==20424==    by 0x8B6987: boost::unit_test::ut_detail::unused boost::unit_test::ut_detail::invoker<boost::unit_test::ut_detail::unused>::invoke<void (*)()>(void (*&)()) (callback.hpp:56)
==20424==    by 0x8B679C: boost::unit_test::ut_detail::callback0_impl_t<boost::unit_test::ut_detail::unused, void (*)()>::invoke() (callback.hpp:89)
==20424==    by 0x6215CB0: ??? (in /usr/lib/x86_64-linux-gnu/libboost_unit_test_framework.so.1.58.0)
==20424==    by 0x61F5995: boost::execution_monitor::catch_signals(boost::unit_test::callback0<int> const&) (in /usr/lib/x86_64-linux-gnu/libboost_unit_test_framework.so.1.58.0)
==20424==    by 0x61F61B2: boost::execution_monitor::execute(boost::unit_test::callback0<int> const&) (in /usr/lib/x86_64-linux-gnu/libboost_unit_test_framework.so.1.58.0)
==20424==    by 0x6215DE1: boost::unit_test::unit_test_monitor_t::execute_and_translate(boost::unit_test::test_case const&) (in /usr/lib/x86_64-linux-gnu/libboost_unit_test_framework.so.1.58.0)

@akva2
Copy link
Member

akva2 commented Aug 9, 2018

Thread 1 "test_transmissi" received signal SIGSEGV, Segmentation fault.
0x00000000008a54ee in Opm::UgGridHelpers::getCoordinate<Opm::UgGridHelpers::CpGridCentroidIterator<&(Dune::CpGrid::cellCentroid(int) const)> > (t=..., i=0)
    at /home/akva/kode/opm/super-build/opm-grid/opm/grid/GridHelpers.hpp:395
395         return (*t)[i];
(gdb) where
#0  0x00000000008a54ee in Opm::UgGridHelpers::getCoordinate<Opm::UgGridHelpers::CpGridCentroidIterator<&(Dune::CpGrid::cellCentroid(int) const)> > (t=..., i=0)
    at /home/akva/kode/opm/super-build/opm-grid/opm/grid/GridHelpers.hpp:395
#1  0x00000000008a0b96 in tpfa_htrans_compute<Dune::CpGrid> (G=0x153fc20, 
    perm=0x154f070, htrans=0x154f670)
    at /home/akva/kode/opm/super-build/opm-grid/opm/grid/transmissibility/TransTpfa_impl.hpp:97
#2  0x000000000089b38d in Opm::DerivedGeology::update<Opm::BlackoilPropsAdFromDeck, Dune::CpGrid> (this=0x7fffffffab00, grid=..., props=..., eclState=..., grav=0x0)
    at /home/akva/kode/opm/super-build/opm-simulators/opm/autodiff/GeoProps.hpp:125
#3  0x0000000000896872 in Opm::DerivedGeology::DerivedGeology<Opm::BlackoilPropsAdFromDeck, Dune::CpGrid> (this=0x7fffffffab00, grid=..., props=..., eclState=..., 
    use_local_perm=false, grav=0x0)
    at /home/akva/kode/opm/super-build/opm-simulators/opm/autodiff/GeoProps.hpp:79
#4  0x000000000088ad49 in TransmissibilityMultipliersCpGrid::test_method (
    this=0x7fffffffd4b7)
    at /home/akva/kode/opm/super-build/opm-simulators/tests/test_transmissibilitymultipliers.cpp:282
#5  0x000000000088ab3b in TransmissibilityMultipliersCpGrid_invoker ()
    at /home/akva/kode/opm/super-build/opm-simulators/tests/test_transmissibilitymultipliers.cpp:257
#6  0x00000000008b6988 in boost::unit_test::ut_detail::invoker<boost::unit_test::ut_detail::unused>::invoke<void (*)()> (this=0x7fffffffd517, 
    f=@0x10ef518: 0x88ab18 <TransmissibilityMultipliersCpGrid_invoker()>)
    at /usr/include/boost/test/utils/callback.hpp:56
#7  0x00000000008b679d in boost::unit_test::ut_detail::callback0_impl_t<boost::unit_test::ut_detail::unused, void (*)()>::invoke (this=0x10ef510)
    at /usr/include/boost/test/utils/callback.hpp:89
#8  0x00007ffff6825cb1 in ?? ()
   from /usr/lib/x86_64-linux-gnu/libboost_unit_test_framework.so.1.58.0
#9  0x00007ffff6805996 in boost::execution_monitor::catch_signals(boost::unit_test::callback0<int> const&) ()
   from /usr/lib/x86_64-linux-gnu/libboost_unit_test_framework.so.1.58.0
#10 0x00007ffff68061b3 in boost::execution_monitor::execute(boost::unit_test::callback0<int> const&) () from /usr/lib/x86_64-linux-gnu/libboost_unit_test_framework.so.1.58.0
#11 0x00007ffff6825de2 in boost::unit_test::unit_test_monitor_t::execute_and_translate(boost::unit_test::test_case const&) ()
   from /usr/lib/x86_64-linux-gnu/libboost_unit_test_framework.so.1.58.0
#12 0x00007ffff680d09e in boost::unit_test::framework_impl::visit(boost::unit_test::test_case const&) () from /usr/lib/x86_64-linux-gnu/libboost_unit_test_framework.so.1.58.0
#13 0x00007ffff68434cb in boost::unit_test::traverse_test_tree(boost::unit_test::test_suite const&, boost::unit_test::test_tree_visitor&) ()
   from /usr/lib/x86_64-linux-gnu/libboost_unit_test_framework.so.1.58.0
#14 0x00007ffff68089f6 in boost::unit_test::framework::run(unsigned long, bool) ()
   from /usr/lib/x86_64-linux-gnu/libboost_unit_test_framework.so.1.58.0
#15 0x00007ffff6824287 in boost::unit_test::unit_test_main(bool (*)(), int, char**) ()
   from /usr/lib/x86_64-linux-gnu/libboost_unit_test_framework.so.1.58.0
#16 0x000000000088a227 in main (argc=1, argv=0x7fffffffe348)
    at /usr/include/boost/test/unit_test.hpp:59

@blattms
Copy link
Member Author

blattms commented Aug 23, 2018

Just a small note:
I did look into this more, but it seems like this is cannot be fixed easily as some of the code in GridHelpers relies on references that can be turned into pointers (e.g. for the cell centers).

@dr-robertk dr-robertk added this to the Release 2019.10 milestone Oct 1, 2019
Geometry is lightweight enough and this is what DUNE's grid interface
mandates.
@blattms
Copy link
Member Author

blattms commented Oct 17, 2019

This could actually work now since the failing code is no more. Will rebase and retry.

@blattms blattms force-pushed the entity-return-copy-geometry branch from 84593f2 to ca06210 Compare October 17, 2019 14:30
@blattms
Copy link
Member Author

blattms commented Oct 17, 2019

jenkins build this please

@blattms blattms removed this from the Release 2019.10 milestone Oct 30, 2019
@blattms
Copy link
Member Author

blattms commented Oct 30, 2019

This will not make it to the release as we have to touch a lot of code relying on this and that is bound to break easily.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants