[petsc-dev] prefix installs and lib/petsc-conf

[Barry Smith]

  I'd like to make the final set of updates to the PETSc directory and script naming conventions as discussed below in the next few days so please work any branches you can through into master to reduce the extra merging you will have to deal with later. 

  Jed, in particular, you've gotten some slightly broken stuff in next that has been hanging around too long and needs fixing and getting into master. 

2014-06-04 06:14:36 -0500  (origin/karpeev/fix-ksp-pcgasm) PCGASM: fixed a copy-paste bug.
2014-10-13 11:59:12 -0600  (origin/dmeiser/pullrequest-fix-PetscOptionsEnum) Use current value for default in PetscOptionsEnum().
2015-01-22 19:33:55 -0700  (origin/jed/ksp-explicit-operator-BAorAB) KSPComputeExplicitOperator: use KSP_PCApplyBAorAB
2015-02-16 13:24:25 +0300  (origin/stefano_zampini/seqaijtodense, stefano_zampini/seqaijtodense) MATSEQAIJ: added conversion from to M
2015-03-02 16:04:08 -0600  (origin/tisaac/pc-pbjacobi-fix-get-array-read) PCApply_PBJacobi(): use VecGetArrayRead()
2015-03-05 14:47:43 -0600  (origin/karpeev/maint/mat-settype-clear) Clear out internal data structures and Mat state when resetting Ma
2015-03-19 18:23:49 -0500  (barry/feature-bargraph) added PetscDrawBar object for drawing bar graphs
2015-03-22 21:58:10 -0500  (barry/snes-monitor-vi-residual) fixed memory leak introduced in SNESVIMonitorResidual() fixed automatic us
2015-03-23 10:21:58 -0400  (origin/jed/assembly-perf) Vec: Silence compiler warning
2015-03-23 13:18:17 -0500  (origin/barry/add-casts-for-suitsparse-64bitindices, barry/add-casts-for-suitsparse-64bitindices) Since Sui

  Thanks

   Barry

> On Jan 20, 2015, at 6:36 PM, Jed Brown <jed at jedbrown.org> wrote:
> 
> Barry Smith <bsmith at mcs.anl.gov> writes:
> 
>>  How about?
>> 
>>    share/petsc/matlab,python,julia/ non-generated stuff 
>>    lib/petsc/matlab,python,julia/ generated stuff for example Matlab .mex files
>> 
>>    lib/petsc/conf    
>> 
>>    include/petsc/finclude,private,mpiuni  (note some generated stuff goes in here; could put it in lib/petsc/include but that seems silly to me)
> 
> The above is good.
> 
>>    bin/petsc/mpiexec.uni  when using uni
>>    bin/petsc/uninstall.py
> 
> $ find /bin/ -type d
> /bin/
> /bin/core_perl
> /bin/vendor_perl
> /bin/cgnstools
> /bin/site_perl
> $ find /bin/ -type f | wc
>   4181    4181   65101
> $
> 
> I'd rather not add a directory in bin.
> 
> I think uninstall should go in sbin.