Time might stop under some hosts without this fix.

The tests themselves didn't change and all the changes their are just
changes of namespace usage and variable movement.

the vtest class) and running them one after the other.
It moves the test code outside of loader.cc

* 'master' of github.com:cloudius-systems/osv: (41 commits)
  sched: default thread attributes
  sched: add new thread::attr object
  enable locking in libc
  apic clockevents: per-cpu setup
  sched: use cpu::init_on_cpu()
  libc/manifest update
  add a real wcsftime_l implementation
  add real implementation of locale-aware strto{f,d,ld} routines
  add floating support to *printf and *scanf
  enable stro*
  add the license for FreeBSDs realpath
  remove the unused old environ-related variables
  sched: per-cpu timer list
  remove an accidentally added file
  remove getenv debugging
  import a real realpath from FreeBSD
  add lstat()
  mutex: zero the mutex structure in mutex_init
  add a full getenv/putenv & friends implementation from musl
  sched: maintain a list of active thread timers
  ...

This makes it easier to create a thread with no special properties.

Run 2 threads, one allocated mem, the other deletes the allocation.

This makes it easy to configure a thread with various parameters.  The first
is the existing stack_info parameter.

pushing it into a dummy bio structure. In addition, expand the test some more

The apic needs to be programmed on each cpu that is brought up.

We used cpu_arch::init_on_cpu() instead of the arch-indepdendent version.

Of course we don't actually support locales yet, so these are trivial wrappers.

Maintain the list of active timers per-cpu, instead of globally.

This removes the need for locking (other than disabling preemption).

The clock_event interface is awkward in that the callback is global, not
per-cpu.  This requires us to have a global dispatch object that then
uses cpu::current() to pick up the current cpu and dispatch the event.  We
should probably make clock_event explicitly per-cpu in the future.

This ones handles the current working directory in relative pathnames,
and supports the POSIX.1-2008 case where the resolved_path argument is
passed a NULL and then allocated by realpath internally.

Since timers are per-cpu constructs, we need to migrate them when a thread
migrate.  To do that, we need to keep a list of a thread's active timers.

If a timer is expired already, don't remove it from lists.

No functional changes - prepare for making timer lists cpu local.

Each cpu has its own event timer; we rely on this.  Document the requirement.