Thanks, merged to next. Satish
On Mon, 16 Jul 2018, Fande Kong wrote: > Thanks, Satish, > > I updated the changes. Please check again. > > > Fande, > > On Mon, Jul 16, 2018 at 12:22 PM, Satish Balay <[email protected]> wrote: > > > On Mon, 16 Jul 2018, Fande Kong wrote: > > > > > In order to have make MatRestoreArray_HYPRE and MatRestoreArray_HYPRE > > looks > > > same. > > > > Well its more like PetscMalloc() and PetscFree() to me - so they don't > > have matching requirements..] > > > > > > > If did not do anything to size and array, and then we may have some > > > warnings: unused parameters. > > > > > > Should we do something like this? > > > > > > +static PetscErrorCode MatRestoreArray_HYPRE(Mat A, PetscInt /*size*/, > > void > > > **/*array*/) > > > +{ > > > + Mat_HYPRE *hA = (Mat_HYPRE*)A->data; > > > + > > > + PetscFunctionBegin; > > > + hA->available = PETSC_TRUE; > > > + PetscFunctionReturn(0); > > > +} > > > > > > I think its should be: > > > > static PetscErrorCode MatRestoreArray_HYPRE(Mat A, void **array) > > > > Satish > > >
