Fix T94310: Blender doesn't support with 128 threads well in Win11

Query TBB for the maximum allowed concurrency, which is free from a bug
in own concurrency detection code. One thing to keep in mind is that now
Cycles is limited by the number of threads in the TBB areana from which
Session is created. This isn't a problem for Blender since we do not limit
arena on Blender side. Could be something to watch out for in other Cycles
integrations.

Differential Revision: https://developer.blender.org/D13658

intern/cycles/device/device.cpp

@@ -37,6 +37,7 @@
 #include "util/math.h"
 #include "util/string.h"
 #include "util/system.h"
+#include "util/task.h"
 #include "util/time.h"
 #include "util/types.h"
 #include "util/vector.h"
@@ -333,7 +334,7 @@ DeviceInfo Device::get_multi_device(const vector<DeviceInfo> &subdevices,
     /* Ensure CPU device does not slow down GPU. */
     if (device.type == DEVICE_CPU && subdevices.size() > 1) {
       if (background) {
-        int orig_cpu_threads = (threads) ? threads : system_cpu_thread_count();
+        int orig_cpu_threads = (threads) ? threads : TaskScheduler::num_threads();
         int cpu_threads = max(orig_cpu_threads - (subdevices.size() - 1), 0);
 
         VLOG(1) << "CPU render threads reduced from " << orig_cpu_threads << " to " << cpu_threads

intern/cycles/util/task.cpp

@@ -89,7 +89,7 @@ void TaskScheduler::init(int num_threads)
     active_num_threads = num_threads;
   }
   else {
-    active_num_threads = system_cpu_thread_count();
+    active_num_threads = tbb::this_task_arena::max_concurrency();
   }
 }