The previous version used memory operands that gcc (probably dependent
on optimization flags and/or version) could address relative to the
stack pointer, but pushing %ebx onto the stack changed it. Here, the
address of the regs array is put into %esi and the individual members
are written into directly.
O(1) realloc() for use with Lua.
A standard malloc/free/realloc API is exposed for testing and other
potential use-cases.
The current configuration it's performs well for lua-metatables
(regular calls to realloc() with varying tiny chunks ~1-50 bytes)
For the use-case at hand it outperforms TLSF.
During initial session load it's possible that two threads call
PBD::notify_event_loops_about_thread_creation() simultaneously
(in particular the process threads). This can lead to an
endless loop in stl_tree.h when assigning thread_buffer_requests[key]
Now we only have WriteLocks.. unless some better solution comes up a
Mutex will do.
This also removes Route::group_gain_control() and associated machinery.
Not yet tested with Mackie or other surfaces. More work to done to
start using the group capabilities, and also potentially to add
or derive more controls as RouteAutomationControls
This new design will work even when threads that need to receive
messages from RT threads are created *after* the RT threads. The
existing design would fail because the RT thread(s) would never
be known the later created threads, and so signals emitted by the
RT thread and causing call_slot() in the receiver would end up
being enqueued using a lock-protected list. The new design ensures
that communication always uses a lock-free FIFO instead
if there is a per_thread_request_buffer, get_request() simply puts a
POD RequestType on the ringbuffer's head. send_request() increases
the write-pointer. There is no memory to free.
