# Numba Meeting: 2020-10-20

Attendees: Siu, Graham, Hameer, Todd, Val, Guilherme, Stu, Juan

## 0. Feature Discussion
- 0.52.0RC1 & 0.52.0RC2
- Quickly discovered several minor issues in RC1.
- Planned to release RC2 this week then we'll ask downstreams and users to test
- Early feedbacks:
- conda package issue with TBB
- Main change in RC2 will be turning down the "pre-optimization" from O1 to O0 because LLVM is not very good at consuming it's O1 output; e.g aggressive loop-vectorization fail
- Report benchmarks ASV
- Val presents performance improvements
- https://nbviewer.jupyter.org/github/numba/numba-examples/blob/70925bd8c00529837e7bef342ad4fbd5cb7b1e25/benchmarks/0.52.1rc1_oct2020/Benchmark%20Analysis.ipynb
- Upcoming tasks/risks declared for 0.53.0
- next release planned after new year holiday
- Hameer: working on sort algorithms; improvements and algos---timsort, radixsort
- Siu: Getting line coverage and cProfile working. Snakeviz plot of e.g. some alg is possible (once PR merged).
- Graham: Make the CUDA dispatcher use the same dispatcher as the CPU (the C/C++ code side of it). Some changes needed to C/C++ dispatcher which could impact all targets.
- Stuart: Piping more information into the dispatch logic from function level to be able to do dispatch based on metadata.

## 1. New Issues
- [#6355](https://github.com/numba/numba/issues/6355) - NamedTuple type and List
- Can probably be implemented, marking as feature request
- [#6356](https://github.com/numba/numba/issues/6356) - Enumerate decrefs too eagerly
- Fixed, closed
- [#6360](https://github.com/numba/numba/issues/6360) - Inconsistent result for Numba in Metropolis-Hastings Distribution Implementation w/ SciPy interp2d
- Stalled, perhaps setting the seed might help, unable to reproduce.
- [#6366](https://github.com/numba/numba/issues/6366) - "Numba requires at least version 0.31.0 of llvmlite."
- waiting on answer
- [#6370](https://github.com/numba/numba/issues/6370) - CUDA tests: There is no RNG seeding in test_atomics.py
- In progress
- [#6371](https://github.com/numba/numba/issues/6371) - 0.52.0RC1 issues
- Fixes for most things in progress
- [#6381](https://github.com/numba/numba/issues/6381) - np.shift_left and np.shift_right unsupported for np.uint64
- Tedious to implement
- [#6382](https://github.com/numba/numba/issues/6382) - Release global interpreter lock (GIL) in ahead-of-time-compiled code
- Will run into the issue of config system mismatch between decorator options and Flags()
- [#6383](https://github.com/numba/numba/issues/6383) - Moran's I error
- Probably fixed in newer Numba version
- [#6386](https://github.com/numba/numba/issues/6386) - Can't AOT String Keyed Dictionaries: undefined symbol: _numba_hashsecret_siphash_k1
- Compiles correctly and then doesn't load
- [#6387](https://github.com/numba/numba/issues/6387) - NumPy ndarray expression with broadcast is slower when not use local variable.
- Maybe fixable, if expressed as loops rather than ufuncs
- [#6388](https://github.com/numba/numba/issues/6388) - NamedTuple fails if type name doesn't match variable name
- This is how Python works
- [#6389](https://github.com/numba/numba/issues/6389) - graphviz/dot internal buffer truncation causes issues for `inspect_cfg`.
- Fix available, closed

### Closed Issues

## 2. New PRs

- [#6353](https://github.com/numba/numba/pull/6353) - [WIP] Add new impl cache entry for overloads with omitted args
- [#6358](https://github.com/numba/numba/pull/6358) - Add log2 and remainder implementations for cuda
- [#6367](https://github.com/numba/numba/pull/6367) - Support views on pinned/mapped arrays
- [#6368](https://github.com/numba/numba/pull/6368) - Testhound/cuda atomic and
- [#6369](https://github.com/numba/numba/pull/6369) - disable refprune pass
- [#6372](https://github.com/numba/numba/pull/6372) - Skips failing tests if SVML is present.
- [#6373](https://github.com/numba/numba/pull/6373) - Allow Dispatchers to be cached
- [#6374](https://github.com/numba/numba/pull/6374) - Use Python 3 style OSError handling
- [#6375](https://github.com/numba/numba/pull/6375) - CUDA tests: reduce managed memory allocation to 0.1 * GPU RAM
- [#6376](https://github.com/numba/numba/pull/6376) - Added a fixed seed in test_atomics.py for issue #6370
- [#6377](https://github.com/numba/numba/pull/6377) - CUDA: Fix various issues in test suite
- [#6378](https://github.com/numba/numba/pull/6378) - Fix incorrect CPU override in vectorization test.
- [#6379](https://github.com/numba/numba/pull/6379) - Use O0 to enable inline and not affect loop-vectorization by later O3…
- [#6380](https://github.com/numba/numba/pull/6380) - Only test managed allocations on Linux
- [#6384](https://github.com/numba/numba/pull/6384) - Fix failing tests to match on platform invariant int spelling.
- [#6385](https://github.com/numba/numba/pull/6385) - Fix/052 rc1 octomerge1
- [#6390](https://github.com/numba/numba/pull/6390) - Updates inspect_cfg

### Closed PRs

- [#6354](https://github.com/numba/numba/pull/6354) - Bump llvmlite to 0.35. series.
- [#6357](https://github.com/numba/numba/pull/6357) - Fix enumerate invalid decref
- [#6359](https://github.com/numba/numba/pull/6359) - Fixes typed list indexing on 32bit
- [#6361](https://github.com/numba/numba/pull/6361) - relax the llvmlite requirements for ASV and testing purposes.
- [#6362](https://github.com/numba/numba/pull/6362) - Move llvmlite support in README to 0.35
- [#6363](https://github.com/numba/numba/pull/6363) - Note that reference counted types are not permitted in set().
- [#6364](https://github.com/numba/numba/pull/6364) - Move deprecation schedules for 0.52
- [#6365](https://github.com/numba/numba/pull/6365) - Update CHANGE_LOG for 0.52.0

## 3. Next Release: Version 0.53.0, RC=Q1 2021

## 4. Upcoming tasks