Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / JuliaGPU/CUDA.jl issues and pull requests

#2027 - GC corruption on 1.10 during cusparse/reduce tests

Issue - State: open - Opened by maleadt over 1 year ago - 7 comments
Labels: bug

#2027 - GC corruption on 1.10 during cusparse/reduce tests

Issue - State: open - Opened by maleadt over 1 year ago - 3 comments
Labels: bug

#2027 - GC corruption on 1.10 during cusparse/reduce tests

Issue - State: open - Opened by maleadt over 1 year ago - 3 comments
Labels: bug

#2026 - RFC: Add a lock to CuArray for improved thread safety

Pull Request - State: closed - Opened by maleadt over 1 year ago - 3 comments
Labels: cuda array, bugfix

#2025 - Perform synchronization on a worker thread

Pull Request - State: closed - Opened by maleadt over 1 year ago - 4 comments
Labels: enhancement, performance

#2025 - Perform synchronization on a worker thread

Pull Request - State: closed - Opened by maleadt over 1 year ago - 4 comments
Labels: enhancement, performance

#2024 - Add a native profiler.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 3 comments
Labels: enhancement

#2024 - Add a native profiler.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 3 comments
Labels: enhancement

#2023 - Confusion on row major vs column major

Issue - State: closed - Opened by lpawela over 1 year ago - 1 comment

#2022 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] over 1 year ago - 1 comment

#2021 - "symbol multiply defined" error

Issue - State: closed - Opened by chrstphrbrns over 1 year ago - 3 comments
Labels: bug

#2020 - Easier way to do mixed-mode matrix multiplication

Issue - State: open - Opened by zygi over 1 year ago - 6 comments
Labels: needs information

#2019 - Add a raw mode to code_sass.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 1 comment

#2018 - libLLVM-15jl.so

Issue - State: closed - Opened by StxGuy over 1 year ago - 1 comment

#2017 - Native profiler using CUPTI

Issue - State: closed - Opened by maleadt over 1 year ago - 3 comments
Labels: enhancement

#2016 - Evaluating sparse matrices in the REPL has a huge memory footprint

Issue - State: closed - Opened by ma-sadeghi over 1 year ago - 2 comments
Labels: needs information

#2015 - CuSparseMatrix constructors allocate too much memory?

Issue - State: closed - Opened by ma-sadeghi over 1 year ago - 3 comments

#2014 - Make nonblocking synchronize optional

Pull Request - State: closed - Opened by lcw over 1 year ago - 1 comment

#2013 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] over 1 year ago - 1 comment

#2012 - Add stream pool to avoid cuStreamCreate overhead

Issue - State: open - Opened by simonbyrne over 1 year ago
Labels: enhancement, good first issue, performance

#2011 - Default to sparse arrays with 64-bits indices

Issue - State: open - Opened by ma-sadeghi over 1 year ago - 1 comment
Labels: enhancement

#2010 - volumerhs performance regression

Issue - State: closed - Opened by maleadt over 1 year ago - 1 comment
Labels: performance, regression

#2009 - Scalar indexing with `CuArray * ReshapedArray{SubArray{CuArray}}}`

Issue - State: closed - Opened by vpuri3 over 1 year ago - 1 comment
Labels: bug

#2008 - `rand` in kernel works in a deterministic way

Issue - State: closed - Opened by lpawela over 1 year ago - 1 comment
Labels: bug

#2007 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] over 1 year ago

#2006 - Cannot precompile GPU code with PrecompileTools

Issue - State: closed - Opened by beorostica over 1 year ago - 5 comments
Labels: bug, upstream

#2005 - Fix accumulate bug

Pull Request - State: closed - Opened by chrstphrbrns over 1 year ago - 4 comments
Labels: cuda array, bugfix, needs changes

#2004 - Make `unsafe_free!` work with nested tuples

Issue - State: open - Opened by vpuri3 over 1 year ago
Labels: enhancement

#2003 - Support for printing types.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 1 comment

#2002 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] over 1 year ago

#2000 - Adapt Type call broadcasting to a function

Pull Request - State: closed - Opened by simonbyrne over 1 year ago - 3 comments

#1999 - Add some performance tips to the documentation

Pull Request - State: closed - Opened by Zentrik over 1 year ago - 4 comments
Labels: documentation, needs changes

#1998 - Add note that 64 bit installation of julia is required

Pull Request - State: closed - Opened by Zentrik over 1 year ago - 1 comment

#1997 - CUSPARSE: in-place copy from CuSparseCSC to SparseCSC

Issue - State: open - Opened by ASLabadi over 1 year ago - 1 comment
Labels: good first issue, cuda libraries

#1996 - Fix typo in captured macro example

Pull Request - State: closed - Opened by Zentrik over 1 year ago - 1 comment

#1995 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] over 1 year ago - 1 comment

#1994 - CUDA.@atomic does not support complex variables

Issue - State: open - Opened by albertomercurio over 1 year ago - 3 comments

#1993 - Add vectorized cached loads

Pull Request - State: closed - Opened by Zentrik over 1 year ago - 1 comment
Labels: enhancement, cuda kernels

#1992 - Add Enzyme to test dependencies

Pull Request - State: closed - Opened by vchuravy over 1 year ago - 4 comments

#1991 - Fine-grained fast-math flags

Issue - State: open - Opened by lcw over 1 year ago - 11 comments
Labels: enhancement, cuda kernels, upstream

#1990 - Use GemmKernels.jl for fallback mul!

Issue - State: open - Opened by maleadt over 1 year ago
Labels: enhancement, performance

#1989 - Add support for thread block clusters

Issue - State: open - Opened by maleadt over 1 year ago
Labels: enhancement

#1988 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] over 1 year ago - 1 comment

#1987 - Update `CUSPARSE.sum_dim<n>` to allow for arbitrary function on elements

Pull Request - State: closed - Opened by lpawela over 1 year ago - 6 comments

#1986 - Allow for arbitrary function in `CUSPARSE.sum_dim<n>`

Pull Request - State: closed - Opened by lpawela over 1 year ago

#1985 - `unsafe_wrap` fails on zero element CuArrays

Issue - State: closed - Opened by jkravs over 1 year ago - 1 comment
Labels: bug

#1984 - Update to CUDA 12.2

Issue - State: closed - Opened by maleadt over 1 year ago - 1 comment
Labels: enhancement

#1983 - `CUSPARSE.sum_dim1` sums the absolute values of elements

Issue - State: closed - Opened by lpawela over 1 year ago - 3 comments

#1982 - Memory allocations during in-place sparse matrix-vector multiplication

Issue - State: closed - Opened by albertomercurio over 1 year ago - 3 comments

#1981 - Add wrappers for `gemv_batched!`

Pull Request - State: closed - Opened by lpawela over 1 year ago - 2 comments
Labels: enhancement, cuda libraries

#1980 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] over 1 year ago - 1 comment

#1979 - fix launch config bug in bitonic sort

Pull Request - State: closed - Opened by xaellison over 1 year ago - 1 comment

#1978 - Tegra: support for artifacts

Issue - State: open - Opened by maleadt over 1 year ago - 1 comment
Labels: upstream, installation

#1977 - CUDNN_STATUS_NOT_SUPPORTED using 1D CNN model

Issue - State: closed - Opened by gzarpapis over 1 year ago - 1 comment
Labels: bug, help wanted, cuda libraries

#1976 - Update to LLVM.jl 6.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 1 comment

#1975 - Wrapper with tests for `gemmBatchedEx!`

Pull Request - State: closed - Opened by lpawela over 1 year ago - 1 comment

#1974 - Use Base.pkgversion on 1.9.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 7 comments

#1973 - Strange behaviour when installed at system level

Issue - State: closed - Opened by benz0li over 1 year ago - 7 comments
Labels: bug

#1972 - Warning cuDNN Convolution

Issue - State: closed - Opened by CCsimon123 over 1 year ago - 1 comment
Labels: bug

#1971 - Crash in LLVM

Issue - State: closed - Opened by maleadt over 1 year ago - 1 comment
Labels: bug, cuda kernels

#1970 - CI simplifications

Pull Request - State: closed - Opened by maleadt over 1 year ago - 2 comments
Labels: tests

#1969 - Don't use Int32 indices in rand kernels.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 1 comment

#1968 - Audit uses of 32-bit indexing

Issue - State: open - Opened by maleadt over 1 year ago
Labels: bug

#1967 - Fix compatibility of CUDA 11.4 to support Orin.

Pull Request - State: closed - Opened by maleadt over 1 year ago

#1966 - Support of Orin GPU from Nvidia ?

Issue - State: closed - Opened by RGerzaguet over 1 year ago - 6 comments
Labels: bug

#1965 - Add and fix some aqua tests

Pull Request - State: closed - Opened by charleskawczynski over 1 year ago - 4 comments
Labels: enhancement, tests

#1964 - Add aqua tests

Issue - State: closed - Opened by charleskawczynski over 1 year ago
Labels: bug

#1963 - Strange `rand` errors when sampling large matrices

Issue - State: closed - Opened by lpawela over 1 year ago - 2 comments
Labels: bug

#1962 - CompatHelper: bump compat for "GPUCompiler" to "0.21"

Pull Request - State: closed - Opened by github-actions[bot] over 1 year ago - 1 comment

#1961 - Test infrastructure: define test groups

Issue - State: closed - Opened by maleadt over 1 year ago
Labels: enhancement, tests

#1960 - Bump CUTENSOR to v1.7.

Pull Request - State: closed - Opened by maleadt over 1 year ago

#1959 - Bump CUDNN to v8.9.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 1 comment

#1958 - JLLs should keep/ship .lib files on Windows

Issue - State: open - Opened by maleadt over 1 year ago
Labels: installation

#1955 - Disable 'No CUDA-capable device found' error log

Issue - State: closed - Opened by simonbyrne over 1 year ago - 6 comments
Labels: enhancement, needs decision

#1954 - Allow StaticArray eltype in matmat{vec,mul}

Pull Request - State: closed - Opened by lcw over 1 year ago - 2 comments
Labels: bugfix

#1953 - Recent `mul!` changes break multiplication with matrices that have `StaticArray` elements

Issue - State: closed - Opened by lcw over 1 year ago - 1 comment
Labels: bug, cuda array

#1952 - Failed CUDA.jl initialization breaks Flux?

Issue - State: closed - Opened by cirobr over 1 year ago - 12 comments
Labels: bug, needs information, installation

#1951 - Update GPUCompiler to fix unreachable control flow.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 2 comments
Labels: cuda kernels

#1949 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] over 1 year ago - 1 comment

#1948 - Implement reverse lookup (Ptr->Tuple) for CUDNN descriptors.

Pull Request - State: closed - Opened by RomeoV over 1 year ago - 3 comments
Labels: enhancement, cuda libraries

#1947 - cuDNN: Store convolution algorithm choice to disk.

Issue - State: closed - Opened by RomeoV over 1 year ago - 8 comments
Labels: performance

#1946 - 1.10 enablement

Pull Request - State: closed - Opened by dkarrasch over 1 year ago - 16 comments
Labels: enhancement

#1945 - Failure to initialize with CUDA_VISIBLE_DEVICES=''

Issue - State: closed - Opened by maleadt over 1 year ago - 1 comment
Labels: bug

#1944 - CuSparseArrayCSR (N dim array) with batched matmatmul (bmm)

Pull Request - State: closed - Opened by nikopj over 1 year ago - 9 comments
Labels: enhancement, cuda array

#1943 - Check mathType only if not Float32

Pull Request - State: closed - Opened by RomeoV over 1 year ago - 17 comments
Labels: bug, help wanted, cuda libraries

#1942 - Disable problematic LLVM block placement passes on Pascal and earlier.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 3 comments

#1941 - [CUSPARSE] Improved support for UniformScaling ad Diagonal

Pull Request - State: closed - Opened by albertomercurio over 1 year ago - 3 comments
Labels: enhancement, cuda libraries

#1926 - Broadcasted multiplication with a rational doesn't work

Issue - State: closed - Opened by bicycle1885 over 1 year ago - 2 comments
Labels: bug, good first issue

#1925 - Rework handle cache so that we never switch tasks in finalizers.

Pull Request - State: open - Opened by maleadt over 1 year ago - 1 comment

#1925 - Rework handle cache so that we never switch tasks in finalizers.

Pull Request - State: open - Opened by maleadt over 1 year ago - 1 comment

#1925 - Rework handle cache so that we never switch tasks in finalizers.

Pull Request - State: open - Opened by maleadt over 1 year ago - 1 comment

#1923 - CUSolver finalizer tries to take ReentrantLock

Issue - State: closed - Opened by vchuravy over 1 year ago - 1 comment
Labels: bug

#1916 - Add norm functions for 3 floats

Pull Request - State: open - Opened by Zentrik over 1 year ago

#1915 - Align variables in constant memory by 256 bit

Pull Request - State: open - Opened by Zentrik over 1 year ago

#1914 - CUDA test failed in wmma.jl

Issue - State: open - Opened by huiyuxie over 1 year ago
Labels: bug

#1913 - Report package versions as part of versioninfo().

Pull Request - State: closed - Opened by maleadt over 1 year ago - 1 comment

#1912 - Don't use Float64 atomics on unsupported platforms.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 1 comment

#1911 - Failed to compile with Julia v1.9 on PowerPC

Issue - State: closed - Opened by williamfgc over 1 year ago
Labels: bug