Progress on Running Android-x86

[boramalper]

It is clearly indicated in the readme -"No Android version seems to work, you still get a shell."- that Android doesn't run on v86, yet.

I wanted to submit a new issue for tracking the progress of running Android-x86 on v86, to share with each other

• why it doesn't run
• what can be done
• what is being worked on and by whom

---

This is especially interesting/exciting since if we manage that, we can run Android as a progressive webapp (PWA) on Mobile Safari on iPhone and iPad; not that it will be necessarily useful but surely fun! =)

[boramalper]

android-x86-4.4-r5

• Memory size: 512 MB
• Video Memory size: 64 MB

Checkbox Flags:

• cpu
• bios
• fpu
• mem
• rtc
• hpet
• io
• acpi
• apic
• vga
• mous
• sb16
• pci

ACPI: disabled

Browser: Firefox Quantum 61.0.1 (64-bit) on Ubuntu 16.04
v86 commit: 89f3a1c

---

1. I can see the bootloader, and I choose "Live CD - Debug Mode" to continue

   

2. I get this glitch after /kernel and /initrd.img has been loaded:

   

   Edit: Setting vga parameter to 786 seems to remove the glitch (instead shows the bootloader as you've pressed enter). See GRUB VGA Modes for all possible values.

3. I get BUG: unable to handle kernel paging request at ... which results in Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000009

   

[BelleNottelling]

Look at this: https://sites.google.com/site/twisteroidambassador/android-x86/kitkat-non-sse3-build
It seems android x86 needs SSE3 (and other SSE instructions) which v86 doesn't emulate

[boramalper]

@BenNottelling I'm downloading the image now, and will let you know the results soon. Thanks!

P.S. My download speed is horrible, not sure if I can finish downloading tonight...

[BelleNottelling]

Well, it seems it fails with a similar error.

[copy]

I did get Android working once (I believe android-x86-1.6-r2.iso), but it was so slow (taking one hour to boot) that I didn't bother pursuing this further.

Still would be worth fixing that panic. Contributions are very welcome.

[BelleNottelling]

I wonder why Android x86 is so slow

[boramalper]

I did get Android working once (I believe android-x86-1.6-r2.iso), but it was so slow (taking one hour to boot) that I didn't bother pursuing this further.

@copy Can you guess what might be the reason (or rather do you think it’s fixable)?

[BelleNottelling]

if you first install android-x86-1.6-r2.iso to a hdd image with qemu it boots MUCH faster within v86

[boramalper]

@BenNottelling Maybe it's relatively much faster but I still found it quite unusable, speed-wise. How was your experience?

[BelleNottelling]

@boramalper it was pretty unuseable lol

[omerjerk]

I have a doubt. Why not try to emulate normal Android on a arm emulator (probably this - https://github.com/ozaki-r/arm-js) instead of trying to emulate Android-x86 on an x86 emulator?

Just want to know what I am missing here.

[niansa]

You are missing, that There is no Android build for ARM Computers with Bios (MBR)… But probably it would be possible.

[BelleNottelling]

Hey here's a mini update for you! I found a build of android froyo (2.2) for VM's on the android X86 source forge page, and I believe it will work in vesa mode , but I haven't waited to see if it would yet. you want the froyo-vm-20100812.iso file

https://sourceforge.net/projects/android-x86/files/Testing/

[omerjerk]

I can try running it. 😅
Will post my experience.

[omerjerk]

Hi,

I got time to test the image today. It's definitely some progress. I can see the Android shell getting opened and working all fine.

Before reaching this screen, I did see some errors like "Not a virtual machine?" and some function not found on the current system.

I tried running the same image in VirtualBox. In VirtualBox, the above shell appears and after this the actual GUI of Android appears all fine.