arrayfire / arrayfire/arrayfire-python
RuntimeError: In function void opencl::fft_inplace
- Langage dominant
- Python
- Étoiles
- 422
- Forks
- 63
- Métriques de merge des PR
- Aucune PR mergée en 30 j
Description
Using this arrayfire release (https://arrayfire.s3.amazonaws.com/3.8.0/ArrayFire-v3.8.0_Linux_x86_64.sh) on the opencl backend with the python api I get the following error.
```python
import arrayfire as af
af.set_backend('opencl')
random_signal = af.random.randn(100)
test = af.signal.fft(random_signal)
```
Error
```
---------------------------------------------------------------------------
RuntimeError Traceback (most recent call last)
in
----> 1 test = af.signal.fft(random_signal)
~/miniconda3/envs/af/lib/python3.7/site-packages/arrayfire/signal.py in fft(signal, dim0, scale)
177
178 output = Array()
--> 179 safe_call(backend.get().af_fft(c_pointer(output.arr), signal.arr, c_double_t(scale), c_dim_t(dim0)))
180 return output
181
~/miniconda3/envs/af/lib/python3.7/site-packages/arrayfire/util.py in safe_call(af_error)
77 err_len = c_dim_t(0)
78 backend.get().af_get_last_error(c_pointer(err_str), c_pointer(err_len))
---> 79 raise RuntimeError(to_str(err_str))
80
81 def get_version():
RuntimeError: In function void opencl::fft_inplace(opencl::Array&, int, bool) [with T = cl_float2]
In file src/backend/opencl/fft.cpp:
```
I'm running Ubuntu 20.04 with Nvidia OpenCL Drivers. I have arrayfire debugging enabled and got the following output when the error occurred.
```
BUILD LOG
************************************************
:19:10: error: expected expression
(float2)(};
^
:891:1: error: expected '}'
^
:18:32: note: to match this '{'
__constant float2 twiddles[] = {
^
:890:2: error: expected ';' after top level declarator
}
^
;
************************************************
FFTGeneratedStockhamAction::compileKernels failed
```
Guide de contribution
Aucun guide de contribution indexé pour ce dépôt
Piste de recherche
Reproduce the Python example with ArrayFire 3.8.0, the OpenCL backend, and Nvidia OpenCL drivers. Start with af.signal.fft and the reported src/backend/opencl/fft.cpp path, then inspect the generated FFT kernel and build log; done means the example completes without the malformed-kernel RuntimeError.
Rédigé par le modèle d'indexation à partir du texte de l'issue.
Évaluation
- Stack technique
- python
- Domaine
- backend
- Type d'issue
- Bug
- Difficulté
- 4/5
- Temps estimé
- 3-5 jours
- Activité
- À l'abandon
- Clarté
- Plutôt claire
- Accessibilité débutants
- 30/100