We'll assume that GPU idle time == time spent waiting for new
command lists to be added to the queue of the cleanup thread.
This isn't entirely accurate, especially if CPU load is very
high, but should be good enough.
This reverts commit f61ff5d345fdbd1627b2fd833f57ff303015c747.
Apparently, this causes extremely poor frame times on some
configurations. Reason is currently unknown.
Makes it easier to identify the command stream thread and the
queue processing thread easier when using a wine build that
supports SetThreadDescription.
Reduces command submission overhead by reusing fence objects
instead of creating new ones for each submission. Improves
error reporting in case the submission cannot be complete.
vkGetQueryPoolResults may never return if VK_QUERY_RESULT_WAIT_BIT is
set, and may return queries as not available when they should in fact
be available. This would cause indefinite hangs, so instead we return
fake data to query objects if retrieving query data fails.
Command submission now does not synchronize with the device every single
time. Instead, the command list and the fence that was created for it are
added to a queue. A separate thread will then wait for the execution to
complete and return the command list to the device.
