# Kria KV260 + PYNQ : hard time trying to make it work
**URL:** https://discuss.pynq.io/t/kria-kv260-pynq-hard-time-trying-to-make-it-work/5101
**Category:** Support
**Created:** [January 11, 2023, 4:04pm UTC](https://discuss.pynq.io/t/kria-kv260-pynq-hard-time-trying-to-make-it-work/5101 "2023-01-11T16:04:01Z")
**Posts on this page:** 5
**Page:** 1
### Author:  [@JPR75](https://discuss.pynq.io/u/JPR75)
#### Post date: [January 11, 2023, 4:04pm UTC](https://discuss.pynq.io/t/kria-kv260-pynq-hard-time-trying-to-make-it-work/5101/1 "2023-01-11T16:04:01Z")
Hi,
I am having a hard time trying to install and make work PYNQ on Kria KV260. Any help or suggestions will be welcome.
My config:
- Kira KV260
- Ubuntu 20.04 LTS (because I was not able to make Ubuntu 22.04 work correctly. Same problem as described in [KR260 Desktop Issue](https://discuss.pynq.io/t/kr260-desktop-issue/5061)
- PYNQ v1.0 (checkout to branch v1.0 asked by the install script)
- Software: provided example “2\_python\_environment.ipynb”
I went through the installation of PYNQ by cloning the GitHub repository and following the instructions, and I am able to start JupyterLab. Problems arise when I try to execute the provided kernel “2\_python\_environment.ipynt”, in section “Numpy Data Movement”:
```python
import numpy as np
import pynq
def get_pynq_buffer(shape, dtype):
""" Simple function to call PYNQ's memory allocator with numpy attributes"""
try:
return pynq.allocate(shape, dtype)
except RuntimeError:
print('Load an overlay to allocate memory')
return
```
pynq.allocate() does not work. Looking deeper, It seems that the problem comes from the function \_get\_xrt\_version\_embedded() of /usr/local/share/pynq-venv/lib/python3.8/site-packages/pynq/pl\_server/xrt\_device.py:
```python
def _get_xrt_version_embedded():
output = subprocess.run(
["xbutil", "--version"], stdout=subprocess.PIPE, universal_newlines=True
)
if output.returncode != 0:
warnings.warn("xbutil failed to run - unable to determine XRT version")
return (0, 0, 0)
xrt_version_str = output.stdout.split("\n")[0].split(":")[1].strip()
return tuple(map(int, xrt_version_str.split(".")))
```
“xbutil” has no option “–version”. As a quick workaround, I have replaced the code by `xrt_version_str = "1.1.0"` (this is given by xbutil dump). But back to “2\_python\_environment.ipynt”, I have a problem with get\_pynq\_buffer() function in the next section of code:
```python
buffer = get_pynq_buffer(shape=(4,4), dtype=np.uint32)
buffer
```
This time, It seems that there is a problem with “global\_state\_file\_exists” in /usr/local/share/pynq-venv/lib/python3.8/site-packages/pynq/pl\_server/embedded\_device.py
I am wandering if I have missed something during the installation or made some mistakes.
Has anyone successfully run PYNQ / Ubuntu on Kria (which version of PYNQ/Ubuntu please)?
Thanks
---
### Author:  [@qianyich](https://discuss.pynq.io/u/qianyich)
#### Post date: [January 12, 2023, 12:45am UTC](https://discuss.pynq.io/t/kria-kv260-pynq-hard-time-trying-to-make-it-work/5101/2 "2023-01-12T00:45:42Z")
I failed this example as well but on a different board with a different image.
---
### Author:  [@stf](https://discuss.pynq.io/u/stf)
#### Post date: [January 12, 2023, 9:37am UTC](https://discuss.pynq.io/t/kria-kv260-pynq-hard-time-trying-to-make-it-work/5101/3 "2023-01-12T09:37:38Z")
Hi @JPR75,
Many thanks for reporting this.
For the global\_state\_file issue, would you mind running cell 1 of `kv260/microblaze_c_libraries.ipynb`
This will download the base overlay and should create the global\_state\_file if it does not exist already. Then can you try rerunning `2_python_environment.ipynb` and report back?
Thanks for flagging the change in the options for xbutil. I’ll try and look into that at some point today.
Many thanks,
Shane
---
### Author:  [@JPR75](https://discuss.pynq.io/u/JPR75)
#### Post date: [January 12, 2023, 12:44pm UTC](https://discuss.pynq.io/t/kria-kv260-pynq-hard-time-trying-to-make-it-work/5101/4 "2023-01-12T12:44:09Z")
Thank you for your help.
I had to add the `ignore_version=True` option, because my [xilinx.com](http://xilinx.com):ip:axi\_iic is 2.1 instead of 2.0:
```python
from kv260 import BaseOverlay
base = BaseOverlay('base.bit', ignore_version=True)
```
and that solved the problem. I am now able to run the “Numpy Data Movement” section of `2_python_environment.ipynb`.
Thanks
---
### Author:  [@marioruiz](https://discuss.pynq.io/u/marioruiz)
#### Post date: [January 12, 2023, 3:19pm UTC](https://discuss.pynq.io/t/kria-kv260-pynq-hard-time-trying-to-make-it-work/5101/5 "2023-01-12T15:19:58Z")