prex code depends on this.

TODO: make it optional

It just adds complexity, and doesn't reduce locking overhead, as we must
take the spinlock anyway during unlock.

Remove it, since it makes conversion to a recursive mutex (needed
by fs/vfs/vfs_mount.c) more complicated.

We need it from several places; use a function to find it.

Or we will be overrun by zombies.

Signed-off-by: Avi Kivity <avi@cloudius-systems.com>

printfs uses ~80K stack; this easily overflows and causes corruption.

We can't simply increase the stack size since this huge stack footprint
means that stack guard pages won't work.

Reduce the footprint for now, later switch to per-cpu allocation or some
other scheme.

Moved thread start from virtio-vring.cc to virtio-blk.cc.

Conflicts:
	drivers/virtio-vring.cc

With atomic status, there is no need to examine the runqueue.

While easy to use, auto-starting threads generates problems when more
complicated initialization takes place.

Rather than making auto-start optional (as Guy suggested), remove it
completely, to keep the API simple.

Use thread::start() to start a thread.  Unstarted threads ignore wakeups
until started.

driver itself. It removes the hard coded binding to queue 0
and will enable flexibility for the upcoming virtio-net registration.

* 'master' of github.com:cloudius-systems/osv:
  add a romfs driver

and call biodone when appropriate

Just like fat the test isn't wired up yet until we have the infrastructure
for it.

* 'master' of github.com:cloudius-systems/osv:
  move debug.h to include/osv and make it self sufficient

It's done since i uses as a increasing input for the data that resides in
a single byte.
In addition, make sure that the read buffer is initialize to zero and not
the target data we wish it to be.

Found while testing readonly disk property

We don't want to be preempted while manipulating the runqueue, since
the scheduler will do that as well when called during preemption, leading
to corruption.

We don't want to be preempted during a context switch, since thread
state invariants will not be maintained.

We store it in a global rather than passing it on the stack since it's
only useful in a few contexts.  Our interrupts don't nest so it's pretty
simple.

Now it's possible to stress virtio through the external tests so
we do not need this code

These relocations need to are relative to the start of the TLS segment, which
ends in %fs:0, so we need to subtract the segment size.

Imported from prex.  Note that the test isn't wired up yet as we'll need
more infrastructure changes to run it.

Don't waste power during idle.

Wastes power.

Instead, spin for a bit looking for work to do, then enter a low-power
state and wait for an interrupt to pull us out of it.

Also add an arch independent name for it.

The whole idea of namespaces is to prevent namespace pollution, using a
"using namespace" in a header negates that.  Remove.

Noticed by Guy.

Add a detached thread class (with a private destructor, so it can only be
allocated on the heap).  The implementation uses a reaper thread to wait
until the detached thread is dead, and then ->join() and delete it.

fixes loading unit tests with "main" as the entry point.