Post Go back to editing

Not compatible No-OS branch with vcu118

Thread Summary

The user encountered build errors while using the ADRV9026 project on VCU118 with the 2023_r2 Patch1 release and Vivado 2023.2. The issue is a type desynchronization in the `adrv9025_spi_read` function between `adrv9025.c` and `adrv9025.h`. The final solution suggests either switching to the main branch and Vitis 2025.1, or manually changing the type from `unsigned int` to `uint32_t` in the 2023_r2 branch. The user should also ensure all submodules are updated using `git submodule update --init --recursive`.
AI Generated Content
Category: Software
Product Number: adrv9026

Hi,

I'm using the ADRV9026 project on VCU118 with the 2023_r2 Patch1 release and Vivado 2023.2.

I generated the XSA from the HDL 2023_r2 branch, but when I build the corresponding no-OS project, I run into build errors.

Could you please clarify which exact branches/tags/commits should be checked out for HDL and no-OS for the ADRV9026 project on VCU118?

Specifically, could you tell me:

  • Which HDL branch/tag/commit you used?

  • Which no-OS branch/tag/commit you used?

  • Which checkout commands you use for the no-OS repository?

  • Whether an XSA generated from the HDL 2023_r2 branch with Vivado 2023.2 is expected to work directly with the no-OS 2023_r2 branch for ADRV9026 on VCU118?

For example, are you using something like:

git clone github.com/.../no-OS.git
cd no-OS
git checkout 2023_r2

or a specific tag/commit instead?

I'm trying to make sure all repositories are aligned with the correct release and tool versions.

Thank you.

Edit Notes

adrv9026
[edited by: HIRONEGAR at 5:05 AM (GMT -4) on 25 Jun 2026]

Thread Notes

  • Hi  ,

    You should clone with git clone --recursive as the no-os repo has multiple submodules. But if you've already cloned no-os without that then you can just run git submodule update --init --recursive.

    Also, what build errors were you getting ? Can you post them here ?

    Other than that I'd suggest switching to the main branch and building with Vitis 2025.1.


    Best regards,
    Robert

  • Hi Robert,

    Thank you for the quick follow-up.

    Yes, I did use git submodule update --init --recursive to ensure all submodules were fully fetched, but I am still hitting that specific compilation error on the 2023_R2 branch.

    Regarding your suggestion to move to the main branch and Vivado 2025.1: unfortunately, upgrading Vivado is quite difficult for my current setup. I am strictly using Vivado 2023.2 because I am trying to stick to the official HDL 2023_r2 Patch1 Release, which explicitly requires Vivado 2023.2. I am also not sure if the current HDL repository can even be successfully built using Vivado 2025.1 yet.

    Since the official release pairing for this version specifies using the 2023_R2 branch for no-OS alongside the 2023_r2 Patch1 HDL, I really need to get this specific codebase to compile.

    Given the compilation error where adrv9025_spi_read has a conflicting type (unsigned int vs uint32_t) between adrv9025.c and adrv9025.h, it looks like a type desynchronization on the 2023_R2 branch. Is this a known bug in this release, and are there any specific commits or patches I should cherry-pick to fix it?

    Best regards,   

  • Hi  ,

    It appears to have been fixed on main and yeah you'd want to compile the main of the hdl repo with the 2025 release of vivado as well.

    Or you can just fix it yourself by changing from unsigned to uint32_t.


    Best regards,
    Robert

  • Hi  ,

    Any updates on this ? Is there anything else related to this that you need help with ? If that's the case please let us know.


    Best regards,
    Robert