WIP: Graph datahandling
This is the draft for a data handling that (optionally) forwards all calls to SerialDatahandling. All calls and data transfers get recorded for the creation of an execution graph.
Needs to be changed after the breaking changes in datahandling.
Needs a tiny change in lbmpy:
Instead of using TimeLoop(...)
for time loop creation a custom function is used.
Edited by Stephan Seitz
Merge request reports
Activity
Filter activity
added 60 commits
-
307d0d00...296de5db - 44 commits from branch
pycodegen:master
- fa30c13c - Improve OpenCL integration via create_kernel
- 91467553 - Extend SerialDataHandling to handle alternative GPU libraries
- 21d9b914 - Allow creation of indexed kernels with opencl
- 23bb748f - Let SerialDataHandling.synchronization_function work with target OpenCL
- 62eb2daa - Teach SympyPrinters how to print Type
- eacf443a - Make SerialDataHandling.serialization_function work with OpenCL
- 043e65c8 - Add two helper functions to manage global cl ctx
- 71e64710 - Add pystencils.opencl.autoinit
- d2737605 - Protect pystencils.opencl.autoinit with try..except
- 294d7f68 - Add pycuda.autoinit to GPU test
- 087f9f9b - Allow default_target=='opencl' in SerialDataHandling
- 87f75ac0 - Add friendly hint when calling OpenCL kernels with PyCUDA/Numpy arrays
- a463a9d0 - Add weird construction to handle OpenCL boundary handling (don't show
- e794cf80 - Save initial draft GraphDataHandling
- bcd6adb1 - Finish draft for graph_datahandling
- c3720d73 - Fixup: skiptest if lbmpy not available
Toggle commit list-
307d0d00...296de5db - 44 commits from branch
Please register or sign in to reply