~njansson/dolfin/hpc

Viewing all changes in revision 2668.1.42.

  • Committer: Anders Logg
  • Date: 2008-05-31 21:09:58 UTC
  • mto: (2668.1.45 trunk)
  • mto: This revision was merged to the branch mainline in revision 2670.
  • Revision ID: logg@simula.no-20080531210958-5yj57ha3l3g02xta
Several Python assembly optimizations and new options:
The following options are now available:

  dolfin_set("optimize form", True)              # Use optimization -O2 when compiling generated code
  dolfin_set("optimize use dof map cache", True) # Store dof maps in cache for reuse
  dolfin_set("optimize use tensor cache", True)  # Store tensors in cache for reuse

  dolfin_set("optimize", True)                   # All of the above

Each one of these may make a significant difference for the
speed of assembly from Python, in particular for repeated
assembly.

With these options set, the difference in speed between assembly
in C++ and Python using the JIT compiler is negligible. (For forms
which don't depend on functions that are evaluated in Python.)

expand all expand all

Show diffs side-by-side

added added

removed removed

Lines of Context: