Bump deps (#613)

- Use `1.10` in Buildkite CI (and `nightly` for CUDA).
- Use `lts`, `1` and `pre` in GitHub CI.
- Add compat GPUArraysCore for `0.2`.
- Bump ChainRulesCore to `0.25`.

.buildkite/pipeline.yml

@@ -1,5 +1,7 @@
 steps:
   - label: ":julia: Julia {{matrix.julia}} + CUDA GPU"
+    command:
+      - echo 'CUDA = "052768ef-5323-5732-b1bb-66c8b64840ba"' >> test/Project.toml
     plugins:
       - JuliaCI/julia#v1:
           version: "{{matrix.julia}}"
@@ -23,8 +25,7 @@ steps:
     matrix:
       setup:
         julia:
-          # - "1.9" # uncomment when 1.10 is out
-          - "1"
+          - "1.10"
           - "nightly"
       adjustments:
         - with:

.github/workflows/ci.yml

@@ -29,8 +29,9 @@ jobs:
       matrix:
         version:
           # - '1.9' # uncomment when julia 1.10 is out
+          - 'lts'
           - '1'   # automatically expands to the latest stable 1.x release of Julia
-          - 'nightly'
+          - 'pre'
         os:
           - ubuntu-latest
           # - macOS-latest
@@ -56,7 +57,7 @@ jobs:
 
     steps:
       - uses: actions/checkout@v3
-      - uses: julia-actions/setup-julia@v1
+      - uses: julia-actions/setup-julia@v2
         with:
           version: ${{ matrix.version }}
           arch: ${{ matrix.arch }}
@@ -84,7 +85,7 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v3
-      - uses: julia-actions/setup-julia@v1
+      - uses: julia-actions/setup-julia@v2
         with:
           version: '1.9'
       - uses: julia-actions/cache@v1

Project.toml

@@ -29,15 +29,15 @@ NNlibFFTWExt = "FFTW"
 NNlibForwardDiffExt = "ForwardDiff"
 
 [compat]
-AMDGPU = "0.9.4, 1"
+AMDGPU = "1"
 Adapt = "3.2, 4"
 Atomix = "0.1"
 CUDA = "4, 5"
-ChainRulesCore = "1.13"
+ChainRulesCore = "1.25"
 EnzymeCore = "0.5, 0.6, 0.7, 0.8"
 FFTW = "1.8.0"
 ForwardDiff = "0.10.36"
-GPUArraysCore = "0.1"
+GPUArraysCore = "0.1, 0.2"
 KernelAbstractions = "0.9.2"
 LinearAlgebra = "<0.0.1, 1"
 Random = "<0.0.1, 1"

README.md

@@ -25,4 +25,3 @@ for CUDA support, or
 using NNlib, AMDGPU
 ```
 for AMDGPU support.
-