Does the L4T kernel supports the IRQ chip ? how can I enable it?

Hello. I’m trying to virtualize Android cuttlefish on my jetson nano using the L4T kernel. This is the version that I’ve used :

It didn’t work. Looking inside the log I found the reason. Is this :

[ERROR:src/] The architecture failed to build the vm: failed to create IRQ chip: No such device (os error 19)

so,it seems that Cuttlefish doesn’t support user-mode irqchips. I need a device tree which properly wires up the vGIC. I’m not expert,but it seems that vgic is enabled in the kernel that I’m using. So giving a : zcat /proc/config.gz | grep KVM : it says :


I suppose that it’s not kernel config problem,but dts/dtb. what I can do to build a working dts / dtb files which work correctly ?

It didn’t work. Looking inside the log I found the reason. Is this
[ERROR:src/] The architecture failed to build the vm: failed to create IRQ chip: No such device (os error 19)

What command triggers this log?

I have followed this tutorial step by step :

this final command produced the error :

-smp 4
-m 2048
-cpu host
-M virt
-device virtio-gpu-pci
-device usb-ehci
-device usb-kbd
-device virtio-tablet-pci
-serial stdio
-display sdl,gl=on
-kernel aosp/Image
-initrd aosp/ramdisk.img
-drive index=0,if=none,id=system,file=aosp/system.img
-device virtio-blk-pci,drive=system
-drive index=1,if=none,id=vendor,file=aosp/vendor.img
-device virtio-blk-pci,drive=vendor
-drive index=2,if=none,id=userdata,file=aosp/userdata.img
-device virtio-blk-pci,drive=userdata
-append “console=ttyAMA0,38400 earlycon=pl011,0x09000000 drm.debug=0x0 rootwait rootdelay=5 androidboot.hardware=ranchu androidboot.selinux=permissive security=selinux selinux=1 androidboot.qemu.hw.mainkeys=0 androidboot.lcd.density=160”

Please consult with the owner with the github project or maybe waiting for other users to share their experience.

He removed all the help requests that had been created on the github. This kind of decision is against the open source philosophy. He said that my help requests took away a lot of time to him. But I had seen only 3 help requests totally. In my opinion his decision is totally to blame. Before complaining that he became some kind of slave with so small amount of help that he had given,a lot of water should have flowed under the bridge. He showed no respect for those people who wants to experiment. He said that he was trying to help, but he really weren’t. The only help he gave has been to ascertain how the facts were. I suspect that he doesn’t like to give help for free. it’s a choice,but he should not confuse the open source philosophy with the closed one.

Since it is a community contribution, we don’t have much experience to share. Hope the author can offer some help.

We would like to suggest you check jetson multimedia api, gstreamer, DeepStream SDK. We have enabled hardware acceleration in the frameworks. You may consider to develop the usecase based on it. And we can offer support for the frameworks.

ok,but as I said,the author does not want to help. in addition,I emailed his collegues and also I didn’t get any reply from them.