dtrace: add dtrace_gethrtime() and fix walltimestamp.
The walltimestamp is required to be the time in nanoseconds since the epoch,
not the time in nanoseconds since boot or since dtrace invocation.
So we no longer need dtrace_mstate.dtms_walltimestamp, nor the flag used to
track its updating: but we *do* need a new way to get the time that is safe
for non-GPL use.
This fixes test/act/time.d (which is the only test that checks if walltimestamp
not only exists but returns the right value).
Orabug:
18376038
Signed-off-by: Kris Van Hees <kris.van.hees@oracle.com>
Signed-off-by: Nick Alcock <nick.alcock@oracle.com>
Acked-by: Jerry Snitselaar <jerry.snitselaar@oracle.com>