using CUDA or CUArray, output: UndefVarError: AddrSpacePtr not defined

[gzhang]

julia> using CUDA
[ Info: Precompiling CUDA [052768ef-5323-5732-b1bb-66c8b64840ba]
ERROR: LoadError: LoadError: LoadError: UndefVarError: AddrSpacePtr not defined
Stacktrace:
[1] getproperty(::Module, ::Symbol) at .\Base.jl:26
[2] top-level scope at C:\Users\gzhang.julia\packages\CUDA\5t6R9\src\device\cuda\wmma.jl:52
[3] include(::Function, ::Module, ::String) at .\Base.jl:380
[4] include at .\Base.jl:368 [inlined]
[5] include(::String) at C:\Users\gzhang.julia\packages\CUDA\5t6R9\src\CUDA.jl:1
[6] top-level scope at C:\Users\gzhang.julia\packages\CUDA\5t6R9\src\device\cuda.jl:15
[7] include(::Function, ::Module, ::String) at .\Base.jl:380
[8] include at .\Base.jl:368 [inlined]
[9] include(::String) at C:\Users\gzhang.julia\packages\CUDA\5t6R9\src\CUDA.jl:1
[10] top-level scope at C:\Users\gzhang.julia\packages\CUDA\5t6R9\src\CUDA.jl:39
[11] include(::Function, ::Module, ::String) at .\Base.jl:380
[12] include(::Module, ::String) at .\Base.jl:368
[13] top-level scope at none:2
[14] eval at .\boot.jl:331 [inlined]
[15] eval(::Expr) at .\client.jl:467
[16] top-level scope at .\none:3
in expression starting at C:\Users\gzhang.julia\packages\CUDA\5t6R9\src\device\cuda\wmma.jl:52
in expression starting at C:\Users\gzhang.julia\packages\CUDA\5t6R9\src\device\cuda.jl:14
in expression starting at C:\Users\gzhang.julia\packages\CUDA\5t6R9\src\CUDA.jl:39
ERROR: Failed to precompile CUDA [052768ef-5323-5732-b1bb-66c8b64840ba] to C:\Users\gzhang.julia\compiled\v1.5\CUDA\oWw5k_5EnnF.ji.
Stacktrace:
[1] error(::String) at .\error.jl:33
[2] compilecache(::Base.PkgId, ::String) at .\loading.jl:1305
[3] _require(::Base.PkgId) at .\loading.jl:1030
[4] require(::Base.PkgId) at .\loading.jl:928
[5] require(::Module, ::Symbol) at .\loading.jl:923

pkg> st
Status C:\Users\gzhang\.julia\environments\v1.5\Project.toml
[b19378d9] ArrayFire v1.0.6
[c52e3926] Atom v0.12.21
[6e4b80f9] BenchmarkTools v0.5.0
[336ed68f] CSV v0.7.7
[052768ef] CUDA v0.1.0
[c5f51814] CUDAdrv v6.3.0
[be33ccc6] CUDAnative v3.0.4
[5ae59095] Colors v0.11.2
[3a865a2d] CuArrays v2.2.2
[717857b8] DSP v0.6.8
[2445eb08] DataDrivenDiffEq v0.2.0
[aae7a2af] DiffEqFlux v1.17.0
[0c46a032] DifferentialEquations v6.15.0
[5789e2e9] FileIO v1.4.3
[587475ba] Flux v0.10.4
[e7e9c730] FourierAnalysis v1.1.0 https://github.com/Marco-Congedo/FourierAnalysis.jl#master
[0c68f7d7] GPUArrays v3.4.1
[82e4d734] ImageIO v0.3.0
[a93385a2] JuliaDB v0.13.0
[e5e0dc1b] Juno v0.8.3
[1902f260] Knet v1.3.7
[38d8eb38] Lathe v0.1.1
[a7949054] MXNet v1.5.0
[961ee093] ModelingToolkit v3.20.0
[315f7962] NeuralPDE v2.0.0
[a15396b6] OnlineStats v1.5.5
[429524aa] Optim v0.20.1
[1dea7af3] OrdinaryDiffEq v5.42.3
[e0fc9d43] PkgMirrors v1.3.0
[91a5bcdd] Plots v0.28.4
[3cdcf5f2] RecipesBase v0.7.0
[31c91b34] SciMLBenchmarks v0.1.0 https://github.com/SciML/SciMLBenchmarks.jl#master
[30cb0354] SciMLTutorials v0.8.0 https://github.com/SciML/SciMLTutorials.jl#master
[0d73aaa9] UNet v0.1.0
[e88e6eb3] Zygote v0.4.22

system: Windows 10
CUDA11.0
cuDNN 11.0

[maleadt]

 [052768ef] CUDA v0.1.0

Ancient version, unsupported on Julia 1.5. Probably being help back by having CuArrays.jl installed, remove the latter first.

[jonniedie]

For me, I had to remove both Flux and DiffEqFlux as well to get to the current version.

[gjvdkamp]

This helped. For other noobs like me, goto to pkg mode, and use st and rm commands to remove older CUDA stuff, then add CUDA again that fixed it for me.