Attendees: Sebastian, Tobias, Li, Konrad, Ramakrishna, David, Razya, Albert
- ClooG
- Cloog-trunk changes committed (Tobias, Sven)
- Reductions in the data dependence analysis (Jan, Sebastian)
- We do not represant all scalar dependences (fixed partially)
- Stream still not interchanged.
- Original Fortran code is not detected by Scop detection
- PCP
- Integration
- limit_scop not removed
- Loop unrolling can be disabled for graphite (Richard)
- Adding changes to gcc is up to the maintainers (David)
- 2d+1 represantation still worked on (Konrad)
- Autoparallized functions are passed to graphite again (Li)
- Alias sets are mostly cliques or sets of cliques (Ramakrishna)
- Autopar (Razya)
- Outermost loop patch is almost ready
- vta merging into graphite will break bootstrap (Sebastian)
- Leave out vta patch, but keep merging