cp.full with np.nan broken #8391
Description
Description
It seems like a numpy 2 thing, but using cp.full with fill_value=np.nan is broken, and previously worked.
To Reproduce
import numpy as np, cupy as cp
out = cp.full((2, 2), np.nan)Installation
Wheel (pip install cupy-***)
Environment
OS : Linux-4.18.0-513.11.1.el8_9.x86_64-x86_64-with-glibc2.28
Python Version : 3.11.7
CuPy Version : 13.2.0
CuPy Platform : NVIDIA CUDA
NumPy Version : 2.0.0
SciPy Version : None
Cython Build Version : 0.29.36
Cython Runtime Version : None
CUDA Root : /usr/local/cuda
nvcc PATH : /usr/local/cuda/bin/nvcc
CUDA Build Version : 12040
CUDA Driver Version : 12030
CUDA Runtime Version : 12040 (linked to CuPy) / 12030 (locally installed)
cuBLAS Version : (available)
cuFFT Version : 11012
cuRAND Version : 10304
cuSOLVER Version : (11, 5, 4)
cuSPARSE Version : (available)
NVRTC Version : (12, 3)
Thrust Version : 200200
CUB Build Version : 200200
Jitify Build Version : <unknown>
cuDNN Build Version : 8801
cuDNN Version : 8500
NCCL Build Version : (not loaded; try `import cupy.cuda.nccl` first)
NCCL Runtime Version : (not loaded; try `import cupy.cuda.nccl` first)
cuTENSOR Version : None
cuSPARSELt Build Version : None
Device 0 Name : Tesla V100-SXM3-32GB
Device 0 Compute Capability : 70
Device 0 PCI Bus ID : 0000:C0:00.0
Additional Information
Version info:
asttokens==2.4.1
cupy-cuda12x==13.2.0
decorator==5.1.1
executing==2.0.1
fastrlock==0.8.2
ipython==8.25.0
jedi==0.19.1
matplotlib-inline==0.1.7
numpy==2.0.0
parso==0.8.4
pexpect==4.9.0
prompt-toolkit==3.0.47
ptyprocess==0.7.0
pure-eval==0.2.2
pygments==2.18.0
six==1.16.0
stack-data==0.6.3
traitlets==5.14.3
typing-extensions==4.12.2
wcwidth==0.2.13
and this works with:
numpy==1.26.4