Ultra96v2 ,pynq2.5 img, upgrade dpu-pynq , zocl error

I want to upgrade my ultra96v2 to support DPU, and I use the " Quick Start" that on github " Upgrading the PYNQ v2.5 image", but I got an error " insmod: ERROR: could not insert module /lib/modules/4.19.0-xilinx-v2019.1/extra/zocl.ko: Invalid module format ", what should I do ?

That flow is verified so it should not give you any problem. Try to see if you have enough space on SD card.

I check the SD card, and it have enough space. My image is " PYNQ v2.5 SD Card Image for Ultra96-V2 built with PetaLinux 2019.1 BSP " that download here : , and I use the BSP to enable the usb to lan driver by petalinux2019.1, then I replace the BOOT.bin and image.ub. Does the flow support Ultra96-V2 ? And I see that the XRT git have update two days ago, can I use it ?

I found a problem. I use dmesg and see that " version magic ‘4.19.0-xilinx-v2019.1 SMP mod_unload aarch64’ should be ‘4.19.0 SMP mod_unload aarch64’ " , it is petalinux2019.1 that cause the problem , so how can I build a linux kernel that version is 4.19.0-xilinx-v2019.1 ? My linux kernel source is linux-xlnx-xilinx-v2019.1 that download from xilinx github.

Hi, If you download the v2.5 pynq image for v2 (make sure you are not using the .img for v1) and then follow the quickstart instructions to install dpu-pynq, this is known to work without issue.

If you try to customize and build with your own PetaLinux bsp and then copy image.ub and BOOT.BIN onto the SD card this is not gong to work for most cases. When PYNQ builds it ingests the bsp that is given on the Avnet github but then it modifies several settings which ends up changing some of the #defines in kernel headers.

If you must customize it to add some feature not found in the base PYNQ image you will likely have to re-build the PYNQ image entirely. If it involves adding a loadable kernel module you should look at the packages found in the Avnet Ultra96 github like for example the wifi driver to see how this is done.

Kind regards