Comments (4)
Not really. This is exactly the same error message, just being rendered from the GPU. If you do what the error message suggests, running with -g2
, you would see a similar stack trace:
ERROR: a exception was thrown during kernel execution.
Stacktrace:
[1] Real at ./complex.jl:44
[2] convert at ./number.jl:7
[3] setindex! at /home/tim/Julia/pkg/CUDA/src/device/array.jl:166
[4] setindex! at /home/tim/Julia/pkg/CUDA/src/device/array.jl:178
[5] #35 at /home/tim/.julia/packages/GPUArrays/OKkAu/src/host/broadcast.jl:70
The only thing different here is that this is rendered as a plain Exception, which isn't easy to fix, and the fact that every thread reports an exception, which has already been reported in #1780. So I think we can close that in favour of the existing issue.
from cuda.jl.
And FWIW, it's not possible to detect and special case this before launching a GPU kernel, because the InexactError depends on the actual data:
julia> x .= 1+1im
ERROR: InexactError: Float32(1 + 1im)
julia> x .= 1+0im
1×1 Matrix{Float32}:
1.0
from cuda.jl.
The only thing different here is that this is rendered as a plain Exception, which isn't easy to fix
I did the hard thing and improved reporting in #2342
from cuda.jl.
Cool thanks!
I give it a shot later
from cuda.jl.
Related Issues (20)
- Failure of Eigenvalue Decomposition for Large Matrices. HOT 3
- CUDA_Driver_jll's lazy artifacts cause a precompilation-time warning HOT 10
- CUDA kernel launches should use their own RNG
- FFT on real array with `plan_fft(x)` is 50% slower than PyTorch HOT 1
- Constant-folding of reinterpret
- Recurrence of integer overflow bug (#1880) for a large matrix HOT 4
- Adding CUDA as dependancy results in segmentation fault for PackageCompiler Sysimage HOT 1
- CUDA kernel crash very occasionally when MPI.jl is just loaded. HOT 2
- CUDA_Runtime_Discovery Did not find cupti on Arm system with nvhpc HOT 2
- Implement `convert` between Array and CuArray like `unsafe_wrap` HOT 2
- CUDA.jl won't install/run on Jetson Orin NX HOT 5
- Error using CUDA on Julia 1.10: `Number of threads per block exceeds kernel limit` HOT 1
- Error when I load my model HOT 1
- BFloat16.jl support
- Explore early finalization HOT 5
- Replace unsafe_free! with finalize? HOT 2
- Driver JLL improvements
- Julia -> C function (Create thead) -> Julia CUDA kernel issue
- device function pointers from CUDA (C) HOT 4
- Cannot iterate over a Tuple of mixed Type
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from cuda.jl.