Commits · cda499e4e2243fb5fce756e6cc89b448bfdd85b6 · Isuru Fernando / loopy · GitLab

Feb 05, 2012
- Add some doc hints that I'm likely to forget. · cda499e4
  Andreas Klöckner authored Feb 05, 2012
  
  cda499e4
- Add parsing for "if()". · 59587042
  Andreas Klöckner authored Feb 05, 2012
  
  59587042
- "Properly" parse instructions with multiple equal signs. · ff0933a5
  Andreas Klöckner authored Feb 05, 2012
  
  ff0933a5
- Make switch to turn off complex-valued expression codegen. · ea359a6a
  Andreas Klöckner authored Feb 05, 2012
  
  ea359a6a
- Meeting notes w/ Tim at Reagan. · 8dd4ed28
  Andreas Klöckner authored Feb 05, 2012
  
  8dd4ed28
Jan 24, 2012
- Implement complex-valued arithmetic. · 9e30e55e
  Andreas Klöckner authored Jan 23, 2012
  
  9e30e55e
Jan 18, 2012
- Provide interactive assistance if the scheduler fails. · 72fdc4d2
  Andreas Klöckner authored Jan 18, 2012
  
  72fdc4d2
- Fix Fig 3.3 variant of FEM assembly. · 0902408b
  Andreas Klöckner authored Jan 18, 2012
  
  0902408b
- Expose reduction realization to the user. Add more notes. · 8ba555dc
  Andreas Klöckner authored Jan 18, 2012
  
  8ba555dc
Jan 17, 2012
- Fix FEM assembly code. · c8064886
  Andreas Klöckner authored Jan 17, 2012
  
  c8064886
- Fix prefill criterion. · 9e2640d2
  Andreas Klöckner authored Jan 17, 2012
  
  9e2640d2
Jan 16, 2012
- Test float4 functionality. · 26a50987
  Andreas Klöckner authored Jan 16, 2012
  
  26a50987
- Parse vectorized reductions. Spelling: 'sum_vec_float4' · 4b34ba94
  Andreas Klöckner authored Jan 16, 2012
  
  4b34ba94
- Hack auto test rig to support vector types, add flag about output filling. · 5e72d5e3
  Andreas Klöckner authored Jan 16, 2012
  
  5e72d5e3
- Use PyOpenCL to translate dtypes into C types. · 0327e979
  Andreas Klöckner authored Jan 16, 2012
  
  0327e979
Nov 20, 2011
- Merge branch 'master' of ssh://forge.tiker.net:2234/loopy · 04ff7222
  Tim Warburton authored Nov 20, 2011
  
  04ff7222
- Tim's SEM tweaks. · 0d8961db
  Tim Warburton authored Nov 20, 2011
  
  0d8961db
- First shot at instance-based subst rule precomputation. · bf38b67c
  Andreas Klöckner authored Nov 20, 2011
  
  bf38b67c
- Only include floor_div code if needed. · ac02f82d
  Andreas Klöckner authored Nov 20, 2011
  
  ac02f82d
Nov 12, 2011
- Make better message about device used for ref calculation. · 421c41d9
  Andreas Klöckner authored Nov 12, 2011
  
  421c41d9
- One more FEM tweak. · c7913fcf
  Andreas Klöckner authored Nov 12, 2011
  
  c7913fcf
- Fix up FEM quadrature example. · c05844ee
  Andreas Klöckner authored Nov 12, 2011
  
  c05844ee
- Many scattered fixes. Full-prefetch SEM with ILP generates code. · 6d6de311
  Andreas Klöckner authored Nov 12, 2011
  
  6d6de311
- Fix doc upload script. · 8fe68119
  Andreas Klöckner authored Nov 11, 2011
  
  8fe68119
- Switch ill-advised outer 'unroll' on split automatic axis back to sequential. · c700eb6d
  Andreas Klöckner authored Nov 11, 2011
  
  c700eb6d
- Fix exposure of with_annotation flag. · e2e48066
  Andreas Klöckner authored Nov 11, 2011
  
  e2e48066
- Upgrade precomputation to make storage dims and sweep dims independent, allow iname sweeps. · 847ffd67
  Andreas Klöckner authored Nov 11, 2011
  
  847ffd67
- Add a better output format for kernels. · 2f5da323
  Andreas Klöckner authored Nov 11, 2011
  
  2f5da323
Nov 11, 2011
- Add some documentation. Expose 'with_annotation' flag. · 3cd2819c
  Andreas Klöckner authored Nov 11, 2011
  
  3cd2819c
- Pick (more) sensible names for reduction inames. Remove non-ND FEM quadrature. · 9ad444e6
  Andreas Klöckner authored Nov 11, 2011
  
  9ad444e6
- Preserve tags when duplicating reduction inames, plus more FEM assembly fixes. · e9baab6d
  Andreas Klöckner authored Nov 11, 2011
  
  e9baab6d
- Improvements to FEM assembly. · de8e8429
  Andreas Klöckner authored Nov 11, 2011
  
  de8e8429
- Adapt iname dependency finding to properly exclude reduction inames. · 7f257cfb
  Andreas Klöckner authored Nov 11, 2011
  
  7f257cfb
- Expand reductions outer-to-inner using a queue. · bfd2cc5f
  Andreas Klöckner authored Nov 11, 2011
  
  bfd2cc5f
- Attempt simple hull if outcome of get_footprint is not convex. · 3a86e395
  Andreas Klöckner authored Nov 11, 2011
  
  3a86e395
- Many more fixes to subst-precompute. · 6f7ba19a
  Andreas Klöckner authored Nov 11, 2011
  
  6f7ba19a
Nov 10, 2011
- Fixes for the subst rule/precompute rewrite. · 69bd5756
  Andreas Klöckner authored Nov 10, 2011
  
  69bd5756
- Kill CSEs. Add substitution rules, precomputes. · abbded75
  Andreas Klöckner authored Nov 10, 2011
  
  abbded75
Nov 09, 2011
- An initial attempt at dealing with equality constraints. · 1af94f84
  Andreas Klöckner authored Nov 09, 2011
  
  1af94f84
- Cache result of dim_{max,min}. · 12452330
  Andreas Klöckner authored Nov 09, 2011
  
  12452330