parallel_for now creates jobs in reverse order

this is because the JobSystem's queue works as a LIFO, by creating
jobs in reverse (memory) order, we attempt to help streaming to
the d-cache on that threads -- until the point where
jobs are stolen. 

we also execute the last job immediately instead of creating a job
for it -- since we're already in a job.

libs/utils/src/JobSystem.cpp

@@ -386,6 +386,10 @@ void JobSystem::release(JobSystem::Job*& job) noexcept {
     job = nullptr;
 }
 
+void JobSystem::signal() noexcept {
+    wake();
+}
+
 void JobSystem::run(JobSystem::Job*& job, uint32_t flags) noexcept {
     HEAVY_SYSTRACE_CALL();