search

etmc / tmLQCD

tmLQCD is a freely available software suite providing a set of tools to be used in lattice QCD simulations. This is mainly a HMC implementation (including PHMC and RHMC) for Wilson, Wilson Clover and Wilson twisted mass fermions and inverter for different versions of the Dirac operator. The code is fully parallelised and ships with optimisations for various modern architectures, such as commodity PC clusters and the Blue Gene family.
http://www.itkp.uni-bonn.de/~urbach/software.html
GNU General Public License v3.0
32 stars 47 forks source link

commit 53db31 of deflation branch deleted a bunch of stuff #328

Open kostrzewa opened 8 years ago

kostrzewa commented 8 years ago

unfortunately a large number of things were removed during the merge of the deflation branch, in particular commit 53db31bff1d377816238945b127fdfccee97c741 caused a number of solvers to disappear...

kostrzewa commented 8 years ago

I think https://github.com/kostrzewa/tmLQCD/tree/mixed_cg_merge_hmc can serve as a guide on how to restore the lost functionality...

kostrzewa commented 8 years ago

It's not clear in the github diff, but it becomes clear when using git show

urbach commented 8 years ago

Hmm, that's bad. Do you already know what to do?

On 24 May 2016 20:32:02 CEST, Bartosz Kostrzewa notifications@github.com wrote:

It's not clear in the github diff, but it becomes clear when using git show

You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/etmc/tmLQCD/issues/328#issuecomment-221355573

Carsten Urbach, www.carsten-urbach.eu

kostrzewa commented 8 years ago

Hmm, that's bad. Do you already know what to do?

My plan was to first restore the solver logic in invert_clover_eo.[c,h] and operator.c, which is where I first noticed it. Afterwards we will have to watch for further regressions and do further manual merges. I don't think that it would be possible, without a major investment of effort, to completely reconstruct what happened and how it affected everything. There were just too many changes in the deflation pull-request to make this practical...

kostrzewa commented 8 years ago

update: I meant invert_clover_eo.[c,h]

urbach commented 8 years ago

unfortunately a large number of things were removed during the merge of the deflation branch, in particular commit 53db31bff1d377816238945b127fdfccee97c741 caused a number of solvers to disappear...

okay, I see that the RGMIXEDCG is no longer available in invert_clover_eo.c. What else?

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

https://github.com/etmc/tmLQCD/issues/328

Carsten Urbach e-mail: curbach@gmx.de urbach@hiskp.uni-bonn.de Fon : +49 (0)228 73 2379 skype : carsten.urbach URL: http://www.carsten-urbach.eu

urbach commented 8 years ago

at least the following in missing in invert_clover_eo.c in the logic

else if(solver_flag == RGMIXEDCG){
 iter = rg_mixed_cg_her(Odd_new, g_spinor_field[DUM_DERI], solver_params, max_iter, precision, rel_prec,
                                        VOLUME/2, &Qsw_pm_psi, &Qsw_pm_psi_32);
   Qm(Odd_new, Odd_new);
 }
kostrzewa commented 8 years ago

Well, for now it's only #329, but we'll have to see how things develop. Maybe other files were similarly affected...

urbach commented 8 years ago

yes, I agree, we have to review carefully. I have a boring meeting starting at 11, where I might find some time...

Carsten Urbach e-mail: curbach@gmx.de urbach@hiskp.uni-bonn.de Fon : +49 (0)228 73 2379 skype : carsten.urbach URL: http://www.carsten-urbach.eu