This can be called from anywhere and just causes the scheduler to
schedule the next thread. This is more efficient and verbose version
of Scheduler::set_current_thread_sleeping(0), since we don't have
to wake other threads or do other verifications.
MMU is can now be locked with RecursiveSpinLock.
Scheduler now has get_current_tid() that works before the Scheduler
is initialized. This allows RecursiveSpinLock usage early on.
Also only process can now add threads to scheduler. Nobody should
have raw access to scheduler and everything should be through
Process::current() or irqs (reschedules)
Now after each interrupt we will ask the scheduler to reschedule
if the current thread is the idle thread. This allows semaphore
unblocking to be practically instant when there is only one thread
executing.
Now disk reading is back to ~3 MB/s for single threaded process
We can now use arbitary BAN::function<void(...)> as the Thread.
I also implemented multithreading for i386 since it was not done
on the initial multithreading commit.
This still uses only a single cpu, but we can now have 'parallelization'
This seems to work fine in qemu and bochs, but my own computer did not
like this when I last tried.
I have absolutely no idea how multithreading should actually be
implmemented and I just thought and implemented the most simple one I
could think of. This might not be in any way correct :D
Bananymous