> On Apr 7, 2018, at 11:03 PM, Rongliang Chen wrote:
>
> Hi Barry,
>
> I tried the petsc-3.7.7 and my code works fine. Any suggestions for debug
> this? Thanks.
Works for 3.7.7 but not 3.8? I'm afraid you'll need to use git bisect, this
will tell you exactly what change in PETSc breaks y
Hello,
I’m trying to set a prefix for a ksp context (from the libMesh
PetscLinearSolver) that solves a “bidomain” problem.
Currently I’m simply calling
KSPSetOptionsPrefix(ksp,"bidomain_");
PCFieldSplitSetIS(ksp->pc,"v",is_v_local);
PCFieldSplitSetIS(ksp->pc,"ve",is_ve_local);
where is_v_loca
I don't understand your question but suspect that the issue is related to
the order you are calling functions. You may need a
KSPSetFromOptions() after the KSPSetOptionsPrefix() but before the
PCFieldSplitSetIS()
Barry
Note that it is ok to have multiple calls to KSPSetFromOptions() fo
Thanks Barry,
That did it: I only had to specify KSPSetFromOptions() between
KSPSetOptionsPrefix() and PCFieldSplitSetIS().
May I ask you how does that work? What do you mean by “depending on where you
set the options prefix they may have different results”? I could not find a
clear explanation
> On Apr 8, 2018, at 11:39 AM, Rossi, Simone wrote:
>
> Thanks Barry,
> That did it: I only had to specify KSPSetFromOptions() between
> KSPSetOptionsPrefix() and PCFieldSplitSetIS().
>
> May I ask you how does that work? What do you mean by “depending on where you
> set the options prefix t
That’s good to know. Thanks.
Simone
> On Apr 8, 2018, at 12:42 PM, Smith, Barry F. wrote:
>
>
>
>> On Apr 8, 2018, at 11:39 AM, Rossi, Simone wrote:
>>
>> Thanks Barry,
>> That did it: I only had to specify KSPSetFromOptions() between
>> KSPSetOptionsPrefix() and PCFieldSplitSetIS().
>>
>>
Thanks Barry. I will try it but it seems highly nontrivial.
Best,
Rongliang
On 04/08/2018 11:18 PM, Smith, Barry F. wrote:
On Apr 7, 2018, at 11:03 PM, Rongliang Chen wrote:
Hi Barry,
I tried the petsc-3.7.7 and my code works fine. Any suggestions for debug this?
Thanks.
Works for 3.
> On Apr 8, 2018, at 7:17 PM, Rongliang Chen wrote:
>
> Thanks Barry. I will try it but it seems highly nontrivial.
The first time you do it it seems overwhelming but once you get the hang of
it it is actually pretty simple and very powerful. Much better
than random guesses at what could b