loopy

loopy copied to clipboard

Warnings in operation counting

1

Hey! When I count the operations of a kernel with Loo.py, I get a lot of warnings that say something similar to the following. ``` LoopyWarning: in kernel some_insns_knl: get_insn_count:...

sv2518

Fix loopy.statistics for CInstructions

3

Reported by @sv2518.

kaushikcfd

Implements CVectorExtensionsTarget

7

/cc @sv2518 Adds support for GNU vector extensions. TODO: - [x] Implement `OMPSimdInameTag`. - [x] In `loopy.codegen.expression` infer the fallback mechanisms from the target. - [x] Pass CI. - [...

kaushikcfd

Dropping arguments in inlining

1

Hi. I noticed an issue with the argument passing that has been introduced in https://github.com/inducer/loopy/pull/631 (I believe). When we inline an inner kernel in a wrapper kernel, where the inner...

sv2518

Vectorization error with predicates depending on vectorized loop

13

The following kernel -- ``` knl = lp.make_kernel( "{[i, j]: 0

kaushikcfd

Implements a reindexing transformation

2

The implementation here is based on the paper "[Memory optimization by counting points in integer transformations of parametric polytopes](https://dl.acm.org/doi/abs/10.1145/1176760.1176771)". Draft because: - [x] Incomplete Implementation - [x] Needs `pw_qpolynomial_to_expr` -...

kaushikcfd

Implements loop distribution transformation

2

TODO: - [x] Do index analysis to verify the validity of the iname-duplication passes. - [x] Add complicated regressions. Draft because: - [ ] includes commits from #350. - [...

kaushikcfd

generate_loop_schedule_v2

4

Implementation for finding loop nest around map in O(N.k), 'N' being the number of inames and 'k' being the max. loop depth. For comparison, let's consider the kernel in #288:...

kaushikcfd

@kaushikcfd I did my best to clean this up and pull out the parts that will be useful to your new linearization stuff. Once your linearization step is incorporating the...

jdsteve2

Tighter write race checking has broken `test_argmax` in `test_scan`

4

This seems to happen repeatably on the Titan V. Failure log: https://gitlab.tiker.net/inducer/loopy/-/jobs/400790 Reverting to 837e0d0002a84e2ef557e71e501e6f89076fb2cc (i.e. just before the merge of #224) clears the failure. @kaushikcfd What should we do?...

inducer