Status of Unifying LKL into UML (v8, Jan 2021) ?

[AkihiroSuda]

I still want to see the Unifying LKL into UML work (v8, Jan 2021) to be merged into the upstream.

Would it be possible to see v9, or the plan was changed?

[lars18th]

Hi,

Another different, but related question: It will possible to compile UML for different architectures?

[fish4terrisa-MSDSM]

Hi,

Another different, but related question: It will possible to compile UML for different architectures?

No, I have just tried. It just not work

[lars18th]

Hi @fish4terrisa-MSDSM ,

Another different, but related question: It will possible to compile UML for different architectures?

No, I have just tried. It just not work

Me too! However, you have any idea about how to add support for any architecture different from x86 and x86_64 with a recent kernel? Can LKL help with this objective?

[fish4terrisa-MSDSM]

Hi @fish4terrisa-MSDSM ,

Another different, but related question: It will possible to compile UML for different architectures?

No, I have just tried. It just not work

Me too! However, you have any idea about how to add support for any architecture different from x86 and x86_64 with a recent kernel? Can LKL help with this objective?

I think lkl might help,but all the arch-depend codes in arch/x86/um should be replaced by lkl,as Unifying LKL into UML is not finished,it might take more time to support other archs.

[lars18th]

Hi @fish4terrisa-MSDSM ,

I think lkl might help,but all the arch-depend codes in arch/x86/um should be replaced by lkl,as Unifying LKL into UML is not finished,it might take more time to support other archs.

From your response I feel it's potentially feasible that using LKL implementation we will obtain a replacement of the code in kernel at arch/x86/um. That's true? If yes then I'll try to help to work on this. Any information to read first?

Regards.

[fish4terrisa-MSDSM]

Hi @fish4terrisa-MSDSM ,

I think lkl might help,but all the arch-depend codes in arch/x86/um should be replaced by lkl,as Unifying LKL into UML is not finished,it might take more time to support other archs.

From your response I feel it's potentially feasible that using LKL implementation we will obtain a replacement of the code in kernel at arch/x86/um. That's true? If yes then I'll try to help to work on this. Any information to read first?

Regards.

Yes,just get a arch-independent replacement of the code in kernel at arch/x86/um then it will work on all platforms.