loopy

loopy copied to clipboard

[Transform API] Simple tiling can be tedious to implement

3

Consider the simple batched matvec example: ```python knl = lp.make_kernel( "{[e,i,j]: 0

kaushikcfd

PyOpenCL target: Overflow large argument counts into SVM struct

9

Needs: - [ ] https://github.com/inducer/pyopencl/pull/452

matthiasdiener

CFamilyTarget should only define the callables present in the intersection of all the C-based targets. The math functions on the complex-typed operands from `complex.h` aren't common to targets like OpenCL/Cuda...

kaushikcfd

MWE ```python import loopy as lp import islpy as isl lp.make_kernel([isl.BasicSet("{[]:}")], []) # Works lp.make_kernel("{ []: }", []) # Fails?! ```

mitkotak

Unable to map ``"bool", "complex128"`` for ``CudaTarget``

1

Here's the MWE ```python import loopy as lp import numpy as np knl = lp.make_kernel( domains = "{ [_0]: 0

mitkotak

Here's the MWE ```python import loopy as lp import numpy as np knl = lp.make_kernel( domains = "{ [_0]: 0

mitkotak

Here's the MWE ```python import loopy as lp import numpy as np knl = lp.make_kernel( domains = "{ [_0]: 0

mitkotak

Respect `kernel.arguments` in non-executable targets' lowered code

1

Close #648

kaushikcfd

See inducer/pymbolic#74. There's some unrelated changes in there apparently. Let me know which ones are annoying and they can go!

alexfikl

Fix CInstruction in calculating mem access map.

1

I don't know if we should calculate something from the predicates or do you think this is sufficient?

sv2518