Add HIP_MAX_QUEUES feature.

Includes some tricky manipulation of the locks for contexts and streams.
issue is that stealing a stream requires we lock the context to
walk the streams to find a victim.  To avoid deadlock, we can't
have a stream locked when we lock the context.  This implementation
releases the stream lock, then acquires the context and selects the
victim.
A more stable implemenation might be to copy the stream list
from a context so that a lock is not required to walk all streams.
Smart shared_ptr could be used to prevent the streams from being
deallocated during the walk.

hipamd/src/hip_memory.cpp

@@ -813,6 +813,8 @@ hipError_t hipMemsetAsync(void* dst, int  value, size_t sizeBytes, hipStream_t s
     if (stream) {
         auto crit = stream->lockopen_preKernelCommand();
 
+        stream->ensureHaveQueue(crit);
+
         hc::completion_future cf ;
 
         if ((sizeBytes & 0x3) == 0) {
@@ -863,6 +865,7 @@ hipError_t hipMemset(void* dst, int  value, size_t sizeBytes )
     if (stream) {
         auto crit = stream->lockopen_preKernelCommand();
 
+        stream->ensureHaveQueue(crit);
         hc::completion_future cf ;
 
         if ((sizeBytes & 0x3) == 0) {