Use of conda to get latest version of tools
Closed this issue · 4 comments
Hi @mithro, I read your document, but I need more info about that point :P I will check how to use Conda, and I need an environment which supports:
- Yosys, GHDL and the ghdl-yosys-plugin, iverilog, verilator.
- Python3 and edalize.
- ISE, Vivado and Quartus.
Recommendation/suggestions? How you solve the support for ISE/Vivado? (which means virtual in your document?). Thanks in advance for links or material that you could provide.
Hi @mithro, I read your document, but I need more info about that point :P I will check how to use Conda, and I need an environment which supports:
FYI -- If you don't want to support all of Linux, Mac and Windows then conda is not really the right choice. If you just want to support Linux, then something like Docker is probably a better / simpler option.
I have some conda notes in a Google Doc here which explain various things.
Conda uses environment.yml files for specifying environments, these can include the traditional Python requirements.txt files. See https://github.com/SymbiFlow/sv-tests/blob/master/.github/travis/install.sh and https://github.com/SymbiFlow/sv-tests/blob/master/conf/environment.yml
- Yosys, GHDL and the ghdl-yosys-plugin, iverilog, verilator.
Yosys / iverilog / verilator should all be up to date (but I would like to get faster at pushing new packages).
GHDL and ghdl-yosys-plugin are still a WIP but should be added in the near future.
- Python3 and edalize.
This is pretty easy.
Example script -- https://github.com/litex-hub/litex-conda-packages/blob/master/conda-get.sh
- ISE, Vivado and Quartus.
These are not ready yet but should be in a couple of weeks.
Recommendation/suggestions? How you solve the support for ISE/Vivado? (which means virtual in your document?). Thanks in advance for links or material that you could provide.
FYI - EDDA is a current work in progress and rapidly evolving.
No, I am not interested in Windows or Mac here (I assume that how the tools work in these OS is tested somewhere :P). So, I probably will use docker images described at #8, adding what needed.
@rodrigomelo9 - You could have a play with conda and see what you think, but I'm guessing docker will be a better solution for you.
Hi @mithro, I finally used Decker. Thanks.