v0.7.0

KernelAbstractions v0.7.0

Diff since v0.6.3

Merged pull requests:

• fix failures on 1.7 (#258) (@simeonschaub)

v0.6.3

KernelAbstractions v0.6.3

Diff since v0.6.2

Merged pull requests:

• nix overdub of _unsafe_getindex (#247) (@vchuravy)
• Stop overdubbing CUDA math functions (#249) (@ali-ramadhan)
• Backport #249 (#251) (@vchuravy)
• Check pow overdubbing (#253) (@vchuravy)

v0.6.2

KernelAbstractions v0.6.2

Diff since v0.6.1

Merged pull requests:

• Remove metadata from cassette context (#244) (@vchuravy)
• nix overdub of _unsafe_getindex (#246) (@vchuravy)

v0.6.1

KernelAbstractions v0.6.1

Diff since v0.6.0

Closed issues:

• [wrong repository] "Spills" from adjacent views of ROCVector (#237)
• Julia compat entry (#238)
• Auto-inbounds and nothing? (#240)

Merged pull requests:

• Use hostcall for wait and stream GC (#85) (@vchuravy)
• [CUDAKernels] add an implicit sync to kernels with no dependencies (#222) (@vchuravy)
• Add some docs for CUDAKernels/ROCKernels (#233) (@jpsamaroo)
• Add new overdub for unsafe_getindex to avoid allocating error message in Julia 1.6+ (#236) (@jakebolewski)
• update to CUDA 3.0 and use task-local stream (#242) (@vchuravy)

v0.6.0

KernelAbstractions v0.6.0

Diff since v0.5.5

Merged pull requests:

• Add ROCKernels backend (#209) (@jpsamaroo)
• Fix Bors (#230) (@DilumAluthge)

v0.5.5

KernelAbstractions v0.5.5

Diff since v0.5.4

Closed issues:

• InvalidIRError when using integer powers (#223)

Merged pull requests:

• overdub overflowerr_binaryop (#227) (@vchuravy)

v0.5.4

KernelAbstractions v0.5.4

Diff since v0.5.3

Closed issues:

• Feature request: performant matmul! example (#205)
• TODO: code coverage for the /lib/CUDAKernels/src/ folder (#213)
• Problem with a simple add vecs for some arrays sizes on CUDA (#221)

Merged pull requests:

• Split KernelAbstractions into frontend and backends (#200) (@vchuravy)
• More robust test/examples.jl (#207) (@tkf)
• performant matmul example for KA (#208) (@mjulian31)
• Simplify some of the testing commands (#210) (@DilumAluthge)
• Test suite: add a newline at the end of a print statement (#211) (@DilumAluthge)
• Simplify more CI commands (#212) (@DilumAluthge)
• Codecov: submit coverage for the following directories: "/src", "/lib" (#214) (@DilumAluthge)
• Combine tests into parameterized testsuite (#219) (@jpsamaroo)
• avoid overdubbing of literal_pow (#224) (@vchuravy)

v0.5.3

KernelAbstractions v0.5.3

Diff since v0.5.2

Closed issues:

• error when calling kernel on reinterpret(reshape, T, ::CuArray) (#179)

Merged pull requests:

• Simplify the Buildkite setup by using version: '1' for the stable job (#204) (@DilumAluthge)
• tune CUDA kernels automatically (#206) (@simeonschaub)

v0.5.2

KernelAbstractions v0.5.2

Diff since v0.5.1

Merged pull requests:

• fix mem = at_private performance (#203) (@vchuravy)

v0.5.1

KernelAbstractions v0.5.1

Diff since v0.5.0

Closed issues:

• Not compatible with Julia 1.6-DEV yet? (#148)
• Buildkite is failing on Julia nightly (#155)
• TODO: submit to Codecov from Buildkite (#171)
• TODO: Run GHA on Julia 1.5, 1.6-nightly, and nightly (#185)
• TODO: use the registered version of CUDA.jl for Buildkite on Julia 1.6-nightly and Julia nightly (#196)

Merged pull requests:

• Remove the Buildkite "soft fail" on Julia nightly (#180) (@DilumAluthge)
• Run Buildkite on Julia 1.5, Julia 1.6, and Julia nightly (#182) (@DilumAluthge)
• Fix an error in the Buildkite pipeline (#183) (@DilumAluthge)
• Submit to Codecov from Buildkite (#184) (@DilumAluthge)
• README: Buildkite badges for Julia 1.5, Julia 1.6-nightly, and Julia nightly (#186) (@DilumAluthge)
• Run GHA on Julia 1.5, 1.6-nightly, and nightly (#187) (@DilumAluthge)
• Bors: require that GitHub Actions CI passes on Julia nightly (#188) (@DilumAluthge)
• Bors: Require GitHub Actions on Julia 1.6-nightly (#191) (@DilumAluthge)
• Bors: Require Buildkite on Julia 1.6-nightly (#192) (@DilumAluthge)
• Bors: Require Buildkite on Julia nightly (#195) (@DilumAluthge)
• Buildkite: Use the registered versions of CUDA.jl for Buildkite on Julia 1.6-nightly and Julia nightly (#198) (@DilumAluthge)