Releases: JuliaGPU/CUDA.jl
Releases · JuliaGPU/CUDA.jl
v1.0.0
CUDA v1.0.0
Closed issues:
- unsafe_copy3d!: srcPos and dstPos handling (#27)
- Test failure on Windows (#37)
- Texture memory? (#46)
- Tests for the LLVM passes (#52)
- Bugged Sparse Matrix-Dense matrix multiplication, where dense matrix is transposed (#77)
- Stack overflow when broadcasting over empty view in CuArrays 2.x (#82)
- Sparse CSC gemm wrappers actually call CSR routines (#181)
- Testsuite calls startup.jl (#182)
- LLVM error: Cannot cast between two non-generic address spaces (#190)
- Error running CUDA in Jupyter (#195)
- Floating-point Inf causes an error (#205)
- mul! issue (#213)
Merged pull requests:
- include potri and test (#179) (@erathorn)
- Fix sparse-dense matmul, with transposed dense (#180) (@irhum)
- Behave like Base wrt. test flags. (#184) (@maleadt)
- fix sparse CSC gemm and test (closes #181) (#185) (@jebej)
- Add speciatization functions f(ctx, x) = f(x) for GPUArrays randn! function (#186) (@Ellipse0934)
- Add inplace test for rand (#187) (@kshyatt)
- Fix cushow tests on Windows. (#188) (@maleadt)
- More tests for CUSPARSE (#189) (@kshyatt)
- fixed gels_batched! issue (#191) (@clintonTE)
- Added wrappers for cusolverDnpotrfBatched (#192) (@IvanYashchuk)
- Added wrappers for cusolverDnpotrsBatched (#193) (@IvanYashchuk)
- Compatibility with Julia 1.5 (#194) (@maleadt)
- Add gemmex wrapper and test (#196) (@kshyatt)
- Fix handling of srcPos and dstPos in unsafe_copy3d! (#197) (@samo-lin)
- Prefer a local CUDA installation when running on CI, reinstate Julia 1.3. (#198) (@maleadt)
- Add support for mixed precision (#200) (@kshyatt)
- Add texture support from CuTextures.jl (#206) (@maleadt)
- Error throwing tests (#207) (@kshyatt)
- Update manifest (#208) (@github-actions[bot])
- A few more tests for cusparse (#210) (@kshyatt)
- Specialize Base.mightalias for better broadcast performance. (#211) (@maleadt)
- Fix some mul! ambiguities, and dispatch more to CUBLAS. (#214) (@maleadt)
- CUDA 11 compatibility entries (#221) (@maleadt)
- Benchmark suite: tune and cache params (#223) (@maleadt)
- Add benchmarks (#224) (@maleadt)
- Update manifest (#225) (@github-actions[bot])
v0.1.0
CUDA v0.1.0
Closed issues:
- Documentation: installation instructions (#1)
- Faced some errors while testing cuda in Julia (#3)
- facing unknown errors while compiling exact similar code for parallelization on CPU (#7)
Merged pull requests:
- Fix typo (#2) (@innerlee)
- Doc: fix comment on how memory is moving (#4) (@mbauman)
- Install TagBot as a GitHub Action (#8) (@JuliaTagBot)
- small grammar/typo tweaks to the introduction tutorial (#12) (@KristofferC)
- Add code of other CUDA packages (#14) (@maleadt)
- Initialize the memory pool (#19) (@maleadt)
- Improve initialization for threading (#20) (@maleadt)
- Don't use BinaryBuilder for most CI tests. (#21) (@maleadt)