lp:~fluidity-core/fluidity/exorcised

Branch merges

Propose for merging

No branches dependent on this one.

Related source package recipes

No recipes using this branch. (?)

Create packaging recipe

Related snap packages

Related bugs

Related blueprints

4229. By Adam Candy on 2013-07-04

Revert merge 4202, which is causing failures in current trunk.

4228. By Jon Hill on 2013-07-04

Bug fix to BFS 2D to try and get it to pass. FLML set-up did not match the test. I've taken the test as 'correct' and edited the FLML to match

4227. By Jon Hill on 2013-07-04

Minor chnage to manual only as we no longer support Lucid

4226. By Stephan Kramer on 2013-07-03

Fix for Saucy.

Commit the .msh-file for this (small) mesh as the test relies on flop counts
to pass which are obviously quite dependent on mesh size. Now passes all
short tests on Saucy.

4225. By Stephan Kramer on 2013-07-03

Workaround for gfortran 4.8.1 bug.

See http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57798
This fixes Fluidity (test_vorticity_2d) on Ubuntu Saucy. One remaining failing
short test: test_pressure_solve_small

4224. By Stephan Kramer on 2013-06-26

Fix periodic adaptivity with adapt_at_first_timestep.

Thanks to CianW for the fixes.
Thanks to jpercival for implementing and adding a test.

4223. By Samuel Parkinson on 2013-06-18

Overcomes bug in vertex based slope limiting caused by underflow producing a zero denominator.
We now check that the subtraction that forms the denominator yields a number that is at least 10 orders of magnitude larger than the value produced by the TINY function and ignore the nodes impact on the slope limiting if it isn't.

4222. By Stephan Kramer on 2013-06-12

Bug-fix: PETSC_NULL -> PETSC_NULL_OBJECT

Fixes petsc_readnsolve test when build with petsc 3.3

4221. By Stephan Kramer on 2013-06-12

Fix checkpointing with moving mesh+free surface.

The mesh movement is always performed with respect to a fixed reference mesh
("OriginalCoordinate"). It is therefore this mesh that should be checkpointed
instead of the moved mesh. Previously the moved mesh would be checkpointed
which would then turn into the reference mesh and the free surface movement
would be applied again (for the second time).

This is now tested (with a volume conservation test) in the spherical patch test.

4220. By Samuel Parkinson on 2013-05-31

make makefiles after last two merges