[dependabot]

Updates the requirements on GPUArrays and JET to permit the latest version.
Updates GPUArrays to 12.1.1

Release notes

Sourced from GPUArrays's releases.

v12.1.1

GPUArrays v12.1.1

Diff since v12.1.0

Merged pull requests:

• Show the array type in 2-arg show of GPU arrays (#429) (@​mcabbott)
• Upload index arrays when viewing a wrapped GPU array (#777) (@​aalquwayfili)
• Give coo_type a base case so a missing method does not recurse (#781) (@​aalquwayfili)
• Forward strided-view matrix products to the storage-level mul! (#800) (@​aalquwayfili)
• [NFC] Reorganize some tests to make failures easier to parse (#805) (@​christiangnrd)
• Let reductions infer their result type (#809) (@​vchuravy)
• Test for closure boxes on Julia 1.14+ (#812) (@​KristofferC)
• Faster findfirst/findlast and reverse (#815) (@​maleadt)

Closed issues:

• Broadcasting on SubArray with inds of Vector{Int} fails with compiler issue (#724)
• coo_type recurses until the stack overflows instead of throwing MethodError (#780)
• Strided SubArray views of CuArray no longer reach cuBLAS in mul! since v11.5.11 (#799)

Changelog

Sourced from GPUArrays's changelog.

GPUArrays.jl release history

This document lists the noteworthy changes in GPUArrays.jl releases that back-ends act on, newest first, aimed at the authors of back-end packages. For the complete list of merged pull requests, see the GitHub releases.

v12.1

GPUArrays defines Base.dataids and Base.mightalias for every AbstractGPUArray. A back-end that implements the new GPUArrays.memory_location hook (the coordinate of an array's allocation, normally its address, and its byte offset into it) gets precise alias detection: disjoint views of one buffer don't alias, empty arrays alias nothing on every Julia version, and SubArrays of two such arrays are compared without converting their parents to pointers. Arrays wrapped in a SubArray or other wrapper are still compared by allocation, which is conservative. Back-ends should implement the hook and delete their own dataids and mightalias methods, which would keep winning dispatch.

v12.0

Base's reductions, sorting, scans, findall, logical indexing and reverse are now implemented once for every AnyGPUArray, on top of AcceleratedKernels.jl (AK) 0.5. A back-end gets all of them by defining nothing beyond the array interface and a KernelAbstractions back-end, and they follow Base's semantics: stable sorting by default, init applied once, no neutral element needed, Base's result types, empty results and errors.

Breaking changes for back-ends:

• The GPUArrays.mapreducedim! hook is removed and GPUArrays no longer picks neutral elements. Every Base reduction goes to AK directly. Back-ends must delete their methods for the hook, together with the reduction kernels only it used.
• GPUArrays.default_rng and the deprecated RNG(state::AbstractGPUArray) and seed!(rng, ::Vector{UInt32}) methods are removed. Construct the RNG with RNG{AT}() and seed it with an integer.
• KernelAbstractions 0.9.43 or later and Adapt 4.7.3 or later are required.

What back-ends should delete:

Methods narrower than GPUArrays' keep winning dispatch, so a back-end that keeps its own findall, to_index/to_indices, _accumulate!/accumulate, sort!/sortperm!/ partialsort!, or reverse! bypasses GPUArrays' Base conformance, and some of those methods are ambiguous with GPUArrays' (e.g. findall(f, ::MyArray) against findall(::Fix2{typeof(in)}, ::AnyGPUArray)). Delete them.

A vendor sort or scan (e.g. one built on MPSGraph) can stay as a fast path, as long as it falls back to GPUArrays' method for the cases it doesn't handle and accepts AK's algorithm objects (alg=AK.MergeSort() etc.). There is no extension point for vendor reductions; GPUArrays._ak_mapreducedim! is internal.

... (truncated)

Commits

• 6fa6490 Bump version.
• ea47064 Give coo_type a base case so a missing method does not recurse (#781)
• 7bffdc6 [NFC] Reorganize some tests to make failures easier to parse (#805)
• 9447b8b Show the array type in 2-arg show of GPU arrays (#429)
• 8d52285 Upload index arrays when viewing a wrapped GPU array (#777)
• 85ac86f Forward strided-view matrix products to the storage-level mul! (#800)
• 4f06c10 Track GPUArrays coverage in the downstream jobs
• 0e33752 Only run the SPIR-V half of OpenCL.jl's tests downstream
• 1fc62f0 Give the OpenCL.jl job two hours
• e0d5567 Test for closure boxes on Julia 1.14+ (#812)
• Additional commits viewable in compare view

Updates JET to 0.13.1

Release notes

Sourced from JET's releases.

v0.13.1

JET v0.13.1

Diff since v0.13.0

Changelog

Sourced from JET's changelog.

[0.13.1]

Changed

• Improved the performance of the error analysis of JETAnalyzer, e.g. report_call and report_package, on Julia 1.14.

Fixed

• Fixed false positive reports on Julia 1.14 and later with code coverage enabled, e.g. under Pkg.test(; coverage=true), where JET could reuse the analysis results cached during its precompilation and fail to constant-fold calls:

  const SV = Core.svec(1, 2, 3)
  f() = sin(length(SV) == 3 ? 1.0 : "1")
  @test_call f() # failed, reporting no method for `sin(::String)`

[0.13.0]

Added

• Added support for Julia v1.14. JET now loads full functionality on Julia 1.14 instead of empty stubs.

• Added the allow_unsupported_julia preference, which loads full JET functionality on unsupported Julia versions, e.g. nightly builds, without the development features that JET_DEV_MODE also enables:

  LocalPreferences.toml

  [JET]
  allow_unsupported_julia = true # previously `JET_DEV_MODE = true` was needed

  JET emits a warning when it is loaded on an unsupported Julia version this way, since it may not work correctly there.

• JETInterface now exports has_problems, the predicate that call_test_ex and func_test use to decide whether an analysis result fails a test.

• Calls whose matching methods are ambiguous are now reported as AmbiguousMethodErrorReport instead of MethodErrorReport saying that no matching method is found. For a union-split call, the ambiguous signatures are reported separately from those without any matching method:

  k(::Int, ::Any) = 1
  k(::Any, ::Int) = 2
  m(x::Int) = k(x, x)               # reported: ambiguous `k(::Int, ::Int)`
  n(x::Union{Int,String}) = k(x, x) # reported: ambiguous `k(::Int, ::Int)`
                                    # and no method for `k(::String, ::String)`

... (truncated)

Commits

• See full diff in compare view

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
see 58 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[dependabot]

Looks like these dependencies are updatable in another way, so this is no longer needed.