dolfin team mailing list archive

[Anders Logg <logg@xxxxxxxxx>]

On Sat, Feb 16, 2008 at 09:52:37PM +0100, Kristian Oelgaard wrote:
> Quoting Harish Narayanan <harish@xxxxxxxxxxx>:
> 
> > Anders Logg wrote:
> > > Good, was it ffc-clean that helped?
> > > 
> > 
> > Yes. I retried the demos and they all run---other than the ones marked 
> > broken and [1].
> 
> Good to know, I'll try it later.
>  
> > I didn't know about ffc-clean, and forced an rm -rf ~/.ffc as I was 
> > running into problems such as the following when running through the 
> > python demos in gdb,
> > 
> > Assembling over 29132 cells.
> >    | Assembling over cells                                           |
> >    |-----------------------------------------------------------------| 0.0%
> > 
> > Program received signal EXC_BAD_ACCESS, Could not access memory.
> > Reason: KERN_PROTECTION_FAILURE at address: 0x00000000
> > 0x08c0c616 in 
> >
> form_62d25dd889dc963e4f7ee7affa6d0549LinearForm_finite_element_1::interpolate_vertex_values
> > 
> > ()
> > 
> > and
> > 
> > Program received signal EXC_BAD_ACCESS, Could not access memory.
> > Reason: KERN_PROTECTION_FAILURE at address: 0x00000000
> > 0x05c74356 in 
> >
> form_66a9c48fc99d3995e9e7748d5229a35fLinearForm_finite_element_2::interpolate_vertex_values
> > 
> > ()
> > 
> > which were showing up and disappearing randomly based on the order in 
> > which I was running through the demos. Plotting works also.
> > 
> > Harish
> > 
> > [1] The DG Poisson python demo gives me the following error:
> > 
> > Traceback (most recent call last):
> >    File "demo.py", line 64, in <module>
> >      u = pde.solve()
> >    File 
> >
> "/Users/hnarayan/Work/FEniCS/build/lib/python2.5/site-packages/dolfin/assemble.py",
> > 
> > line 182, in solve
> >      (A, self.dof_maps) = assemble(self.a, self.mesh, return_dofmaps=True)
> >    File 
> >
> "/Users/hnarayan/Work/FEniCS/build/lib/python2.5/site-packages/dolfin/assemble.py",
> > 
> > line 43, in assemble
> >      (compiled_form, module, form_data) = jit(form)
> >    File "/sw/lib/python2.5/site-packages/ffc/jit/jit.py", line 44, in jit
> >      form_data = analyze.analyze(form)
> >    File 
> > "/sw/lib/python2.5/site-packages/ffc/compiler/analysis/analyze.py", line 
> > 27, in analyze
> >      check_form(form)
> >    File 
> > "/sw/lib/python2.5/site-packages/ffc/compiler/analysis/checks.py", line 
> > 26, in check_form
> >      check_restrictions(form)
> >    File 
> > "/sw/lib/python2.5/site-packages/ffc/compiler/analysis/checks.py", line 
> > 53, in check_restrictions
> >      raise FormError, (p, "Integrand may not be restricted in an 
> > exterior facet integral.")
> > ffc.common.exceptions.FormError
> 
> This should happen, I'll fix it on Monday. (If nobody beats me to it)
> 
> Kristian

Great. I thought I'd fixed this but apparently not.

-- 
Anders