Run QEMU on jailbroken iDevices (not issue)

[k0Iry]

I would like to say thank to you, based on your QEMU patch, I cross-compiled QEMU for jailbroken devices.

Now I am able to run at least aarch64 & x86 archs on my iPhone 8 😄, and I think under jailbreaking environment, the possibilities would be much more. Thanks man for your great work!

https://github.com/k0Iry/qemu-ios-jailbroken

[osy]

I’m not sure how jailbreaks handle entitlements but that crash is due to not having jit entitlements. LLDB has such entitlements which is why it works there. You can dump the entitlements from lldb and use ldid to fakesign with those entitlements.

[professorUnknown]

The possibilities are better performance from not utilizing private api and usb and Bluetooth guest vm interaction.

[k0Iry]

I’m not sure how jailbreaks handle entitlements but that crash is due to not having jit entitlements. LLDB has such entitlements which is why it works there. You can dump the entitlements from lldb and use ldid to fakesign with those entitlements.

Thanks for your reply! But this issue I think is related to substitute, I actually did code sign the binary but only qemu-system- ones don’t work properly*, if I didn’t do that, I don’t even be able to run it under LLDB, that is the strange part.

[k0Iry]

The possibilities are better performance from not utilizing private api and usb and Bluetooth guest vm interaction.

Yes, with almost all the possibilities 😊

[k0Iry]

I’m not sure how jailbreaks handle entitlements but that crash is due to not having jit entitlements. LLDB has such entitlements which is why it works there. You can dump the entitlements from lldb and use ldid to fakesign with those entitlements.

I understand, I’ll try to sign it and try later.

https://developer.apple.com/documentation/bundleresources/entitlements/com_apple_security_cs_allow-jit

[osy]

The possibilities are better performance from not utilizing private api and usb and Bluetooth guest vm interaction.

Not true. There's no performance benefits and there's no USB/BT support or anything without the same additional work (i.e. porting libusb, etc).

The only difference is basically you can run in the command line.

[professorUnknown]

I read somewhere that in stock iOS UTM has its performance restricted by the kernel but it’s true that it requires work and UTM is in very early development not even in the beta.

[osy]

You're thinking of the v0.1 limitations fixed in v0.2.

[k0Iry]

The possibilities are better performance from not utilizing private api and usb and Bluetooth guest vm interaction.

Not true. There's no performance benefits and there's no USB/BT support or anything without the same additional work (i.e. porting libusb, etc).

The only difference is basically you can run in the command line.

If I understand it correctly, QEMU on iOS only does soft-emulation, so running it in cli or by APP wouldn't differ in any means in terms of performance, under the hood they are the same.

What a shame KVM wouldn't be possible under this situation. 😝

[professorUnknown]

UTM could become kvm based doesn’t mean it will since what I said before UTM is in early development not even in the beta and changing it to kvm will break the whole project but if you want you could make your own project since you have some knowledge in cross compiling you could.

[osy]

You can't use KVM because Apple's A-chips do not support virtualization.

[professorUnknown]

With checkrain it maybe can since its a bootrom exploit.

[osy]

Let me repeat this again

You can't use KVM because Apple's A-chips do not support virtualization.

[professorUnknown]

Ok

[k0Iry]

only got BogoMIPS 125 running aarch64 😂