Re: QNX VM hang on Qemu

[Clément Chigot]

Hi Faiq,

On Fri, Feb 23, 2024 at 3:55 PM Faiq Ali Sayed <faiqueali.109@gmail.com> wrote:
>
> So as far as my understanding, we provide these binaries using Qemu command 
> as depicted in the example you provided and there is no way I found to put 
> them into a single image.
> Regarding the overlapping space, I don't have much idea but I think we could 
> provide a starting address separately to these images something like 
> addr=0x00100000.

Where is this 0x00100000 address coming from ? Could you confirm with
"readelf -h" that this is the entry point of your image ?

Alternatively and that's what we used locally, qemu is able to guess
the entry point of an image, when passed from -kernel. Therefore, our
command simply looks like:
  |  $ qemu-system-aarch64 -M xlnx-zcu102 -m 4G -no-reboot -nographic
-kernel qnx.img

I'm not the one having built the qnx.img we're using. But it looks
pretty standard at the first look, made with "mkifs" and the kernel
specs from zcu102 evaluation kit.

Hope it helps,
Clément

> So as per your suggestion, I compared my images and I found that the image 
> does not show a virtual disk, and other commands like mkdir, do not have 
> these binaries.
> So these binaries are not included at the time of image creation and I don't 
> exactly know that how can we add these binaries into the QNX image.
>
> The Image that is currently installed in real hardware does not have a 
> debugging symbol, so I can't use GDB  to debug that.
> Now I am looking for a way to create the correct QNX OS image for Qemu.
>
> Any lead in this regard will be really helpful :)
>