(gr) (hdf5) fixed compile errors when GR=yes and HDF5=yes

[dliptai]

No description provided.

[danieljprice]

I'm really trying to avoid ifdefs all over the place, would be nicer to use if (gr) rather than #ifdef GR, as it allows for if(gr) to become a runtime option eventually

[dliptai]

I agree. The biggest roadblock to this is that GR compiles a different externalforces module, and it also does not compile any of extern_*.f90

[dliptai]

I decided to include all extern_*.f90 in SRCPOTS when compiling with GR=yes. I don't think it should cause any problems, and hopefully it should make removing other #ifdef GRs easier

[danieljprice]

thanks for removing the ifdefs (#55), however there is now a conflict with the radiation branch just merged. If you could fix this would be helpful.

[dliptai]

Conflict resolved.
Updated the hdf5 readwrite to use eos_vars for writing temperature and pressure to dump.

For consistency between the two formats:

• removed VrelVf and dustgasprop from small dumps
• tstop array is no longer read from dump

[dmentipl]

@dliptai: Thanks very much for fixing this. It all looks good to me.

@danieljprice: Is it possible for this to be merged?

The MPI test failure is in radiation, not the HDF5 read/write dumps, as far as I can tell.

--> SKIPPING TEST OF RADIATION DERIVS (need -DPERIODIC)

 ---------------- particles set on  32 x   8 x   8 uniform closepacked lattice --------------
  x:  -0.500    ->  0.500       y:  -0.100    ->  0.100       z:  -0.100    ->  0.100     
 dx:   3.125E-02               dy:   2.706E-02               dz:   2.552E-02
                            dy/dx:   0.8660               dz/dx:   0.8165
 enforcing periodicity: nx, ny, nz =     32 x     8 x     6,  np =     1536
 dx:   3.125E-02               dy:   2.500E-02               dz:   3.333E-02
                            dy/dx:   0.8000               dz/dx:   1.0667
 -----------------------------------------------------------------------------------------

 checking   grad{E}.....................OK     [max err = 1.613E-14, tol = 1.000E-10]
 checking D*grad{F}.....................OK     [max err = 1.083E-02, tol = 2.000E-02]
 checking dE/dt = 0.....................FAILED [got  2.794E-09 should be   0.000E+00, err = 2.794E-09, tol = 2.500E-09]
FAILED [got  2.794E-09 should be   0.000E+00, err = 2.794E-09, tol = 2.500E-09]
 checking xi(t_10)......................OK     [max err = 4.801E-05, tol = 3.500E-04]
 checking xi(t_20)......................OK     [max err = 2.338E-04, tol = 3.500E-04]
 checking xi(t_30)......................OK     [max err = 2.298E-04, tol = 3.500E-04]
 checking xi(t_40)......................OK     [max err = 2.315E-04, tol = 3.500E-04]
 checking xi(t_50)......................OK     [max err = 2.049E-04, tol = 3.500E-04]
 checking xi(t_60)......................OK     [max err = 2.715E-04, tol = 3.500E-04]
 checking xi(t_70)......................OK     [max err = 2.176E-04, tol = 3.500E-04]
 checking xi(t_80)......................OK     [max err = 8.126E-05, tol = 3.500E-04]
 checking xi(t_90)......................OK     [max err = 1.468E-04, tol = 3.500E-04]
 checking xi(t_**)......................OK     [max err = 1.912E-04, tol = 3.500E-04]
 checking xi(t_**)......................OK     [max err = 1.500E-04, tol = 3.500E-04]
 checking xi(t_**)......................OK     [max err = 1.067E-04, tol = 3.500E-04]

<-- RADIATION TEST COMPLETE

SUMMARY OF ALL TESTS:
PASSED: 156 of 157  99.4%
FAILED:   1 of 157   0.6%

 _____ _    ___ _     
|  ___/ \  |_ _| |    
| |_ / _ \  | || |    
|  _/ ___ \ | || |___ 
|_|/_/   \_\___|_____|

TEST SUITE FAILED

[bjnorfolk]

Can you pull this @danieljprice ? I can't convert anything to hdf5 currently.

[danieljprice]

I think this might just need to be updated from master for the checks to pass. All checks are currently passing on the master branch.

[danieljprice]

memory should not be allocated here for either pressure or temperature

[danieljprice]

Eos_vars should be written directly to disk, there is no need for intermediate storage

[dmentipl]

@danieljprice: I've implemented the requested changes on my fork (https://github.com/dmentipl/phantom).

I'm not sure if I can add my commits to this PR. I can create a new pull request, if that's OK @dliptai? Or else, if you make me a contributor to your fork, I can push the changes that way.

[dmentipl]

All the tests are passing on my fork:

https://github.com/dmentipl/phantom/actions

[danieljprice]

@dmentipl can we switch the pull request to your fork? I assume this now works because you merged master into it.

I think easiest way is just to close this one and open a new one from @dmentipl's fork.

[danieljprice]

also, just for reference, failures here reflect ongoing problems raised in #46

[dmentipl]

@dmentipl can we switch the pull request to your fork? I assume this now works because you merged master into it.

I think easiest way is just to close this one and open a new one from @dmentipl's fork.

Sorry, I just saw this...

I've merged my changes into @dliptai's fork which has added to this pull request. The tests are running now.