[petsc-dev] broken nightlybuilds (next vs next-tmp)

2017-11-03 Thread Satish Balay
PETSc integrator group, next builds have been broken for the past few weeks - preventing graduation of potentially clean branches to maint/master. So I've switched nightlybuilds from next to next-tmp to test fewer branches at a time [and hopefully graduate them]. Tonights builds are with: $

Re: [petsc-dev] DMPlexCreateMedFromFile broken

2017-11-03 Thread Matthew Knepley
On Thu, Nov 2, 2017 at 12:08 PM, Vaclav Hapla wrote: > It seems that DMPlexCreateMedFromFile leaves out some mesh elements. I > found it out when investigating why ParMetis redistribution crashes. > > I attach the datafile >

Re: [petsc-dev] SLEPc failure

2017-11-03 Thread Franck Houssen
OK, thanks. Unfortunately, this does not help. So I need to find a another way to extract B from the data set in order to get a "better" (= "less" singular) B. Thanks for tips and help. Franck - Mail original - > De: "Jose E. Roman" > À: "Franck Houssen"