[petsc-dev] errors in PETSc dev code, please remember precisions

[Barry Smith]

   Lisandro,

    It is "ok" to use fscan in places but you have to make sure that the variable type is right. For example you may need to declare a temparary double variable to hold the value (yes double not PetscReal).

   Barry

On May 23, 2011, at 5:22 PM, Lisandro Dalcin wrote:

> On 24 May 2011 00:47, Barry Smith <bsmith at mcs.anl.gov> wrote:
>> 
>> iga.c: In function ‘PetscErrorCode DMIGAInitializeGeometry3d(_p_DM*, PetscInt, PetscInt, char*, char*)’:
>> iga.c:558: warning: format ‘%lf’ expects type ‘double*’, but argument 3 has type ‘PetscReal*’
>> iga.c:565: warning: format ‘%lf’ expects type ‘double*’, but argument 3 has type ‘PetscReal*’
>> iga.c:572: warning: format ‘%lf’ expects type ‘double*’, but argument 3 has type ‘PetscReal*’
>> 
>> when building with single
>> 
> 
> Well, that was my intent to enable the new IGA stuff to work with all
> precisions. Originally, Matt disabled that code for single. BTW, that
> code is using fscanf()... is that fine?
> 
> 
> -- 
> Lisandro Dalcin
> ---------------
> CIMEC (INTEC/CONICET-UNL)
> Predio CONICET-Santa Fe
> Colectora RN 168 Km 472, Paraje El Pozo
> 3000 Santa Fe, Argentina
> Tel: +54-342-4511594 (ext 1011)
> Tel/Fax: +54-342-4511169