Partially implements console ioctl with com = 0x20007461

[SuchAFuriousDeath]

Progresses #525.

[SuchAFuriousDeath]

Thank god for Option::is_some_and.

[SuchAFuriousDeath]

By the way, I'm pretty sure I found the proc0_init function.

[SuchAFuriousDeath]

It seems that kern.proc.appinfo is connected to workaround and to something with cpusets

[ultimaweapon]

I can't take a look due to COVID-19. The headache is so bad even though I am already vaccinated.

[SuchAFuriousDeath]

It's fine, take care.

This looks like it initializes budgets for processes.

[SuchAFuriousDeath]

It seems that kern.proc.appinfo is connected to workaround and to something with cpusets

Ok, I uncovered it now. The 'something to do with cpusets' acttually means that it calls 'pfind'. I didn't know what the function was up until now :D

[SuchAFuriousDeath]

This is the initial budget for a process.

[SuchAFuriousDeath]

I'm not sure if the fields are named correctly, but their offsets are 0xaf0 and 0xaf4.

[SuchAFuriousDeath]

There's also eventhandlers for id tables

[SuchAFuriousDeath]

Looking at this now, I probably created a reference cycle. Not sure if it's really a problem though.

[SuchAFuriousDeath]

By the way, do all the components have to be in an Arc? Couldn't we just create a struct that holds all the components and switch everything from Arc to just an immutable reference?

[ultimaweapon]

By the way, do all the components have to be in an Arc? Couldn't we just create a struct that holds all the components and switch everything from Arc to just an immutable reference?

The problem is we need static lifetime due to PS4 threading, which mean those objects must be either global or getting a static lifetime with Box::leak.

[VocalFan]

By the way, do all the components have to be in an Arc? Couldn't we just create a struct that holds all the components and switch everything from Arc to just an immutable reference?

The problem is we need static lifetime due to PS4 threading, which mean those objects must be either global or getting a static lifetime with Box::leak.

I just helped make a PR for SlimeVR... Their method to keep something is to use leak.

However, they do it a different way. std::mem::forget(job)

[ultimaweapon]

By the way, do all the components have to be in an Arc? Couldn't we just create a struct that holds all the components and switch everything from Arc to just an immutable reference?

The problem is we need static lifetime due to PS4 threading, which mean those objects must be either global or getting a static lifetime with Box::leak.

I just helped make a PR for SlimeVR... Their method to keep something is to use leak.

However, they do it a different way. std::mem::forget(job)

We have tried Box::leak before we switched to Arc. forget for a different purpose.

[ultimaweapon]

Scoped-thread cannot be used to implement threading syscall.

[SuchAFuriousDeath]

Why is that, if I may ask?

[ultimaweapon]

Let's say we implement thr_new. How do we return to the caller without killing the spawned thread within the scope?

[SuchAFuriousDeath]

Oh, right. I misunderstood what you meant by 'threading syscall'.

[SuchAFuriousDeath]

But somehow, intuitevely, it feels like it should be possible to do, somehow.

[SuchAFuriousDeath]

If we could somehow hold onto the scope as a field of VProc or something like that, the thread wouldn't die, but it would be guaranteed not to outlive the instance of VProc I don't know if something like that is possible though

[ultimaweapon]

The point of scoped thread is to prevent any threads inside to leak. That mean it is not possible.

[SuchAFuriousDeath]

I thought about this more and it seems you're right :((((

[SuchAFuriousDeath]

By the way, don't you think it would be better to split our handle_64 function into handle_i64 and handle_u64 so that we don't have to change types?

[ultimaweapon]

We can handled both types in handle_64.

[ultimaweapon]

I got a bad news (or good news?). My next PR is going to be overthrow the current FS structure by start migrating to the new FS. That mean the easier way for this PR to migrate to a new FS is by start a clean branch and copy part by part from the old branch.

[SuchAFuriousDeath]

Nice (and not nice, at the same time :D).