[petsc-dev] petsc-dev/src/ksp/ksp/examples/tests/ex22

[Barry Smith]

On Jan 6, 2012, at 1:24 PM, Jed Brown wrote:

> This was caused by VecValidValues() for a Vec that does not implement VecGetArray(). Fixed in
> 
> http://petsc.cs.iit.edu/petsc/petsc-dev/rev/3b9bc425b217
> 
> I don't think that not implementing VecGetArray() should be a fatal error though, there are certainly Vec types that don't need it.

   I agree. VecValidValues() could simply return without checking if getarray is not support for it.

   Barry

> 
> On Fri, Jan 6, 2012 at 10:28, Hong Zhang <hzhang at mcs.anl.gov> wrote:
> Jed,
> petsc-dev/src/ksp/ksp/examples/tests/ex22 nightly test gives error
> 
> 
> Program received signal SIGSEGV, Segmentation fault.
> 0x0000000001047afd in PCApply (pc=0x2a01620, x=0x2982980, y=0x2a43590)
>    at /sandbox/hzhang/petsc-dev/src/ksp/pc/interface/precon.c:378
> 378       VecValidValues(x,2,PETSC_TRUE);
> (gdb) where
> #0  0x0000000001047afd in PCApply (pc=0x2a01620, x=0x2982980, y=0x2a43590)
>    at /sandbox/hzhang/petsc-dev/src/ksp/pc/interface/precon.c:378
> #1  0x0000000001117687 in KSPInitialResidual (ksp=0x29d0820, vsoln=0x299d2b0,
>    vt1=0x2a1fad0, vt2=0x2a31830, vres=0x2a43590, vb=0x2982980)
>    at /sandbox/hzhang/petsc-dev/src/ksp/ksp/interface/itres.c:64
> #2  0x00000000010dfd38 in KSPSolve_GMRES (ksp=0x29d0820)
>    at /sandbox/hzhang/petsc-dev/src/ksp/ksp/impls/gmres/gmres.c:230
> #3  0x000000000091f387 in KSPSolve (ksp=0x29d0820, b=0x2982980, x=0x299d2b0)
>    at /sandbox/hzhang/petsc-dev/src/ksp/ksp/interface/itfunc.c:429
> #4  0x00000000004085f0 in test_solve () at ex22.c:92
> #5  0x000000000040bae0 in main (argc=2, args=0x7fff2765ac78) at ex22.c:229
> 
> Can you take a look at it? I do not know where is the problem.
> Thanks,
> 
> Hong
>