* Reintroduced the SMP_MSG_RESCHEDULE_IF_IDLE ICI message. This time

implemented by means of an additional member in cpu_ent.
* Removed thread::keep_scheduled and the related functions. The feature
  wasn't used yet and wouldn't have worked as implemented anyway.
* Resurrected an older, SMP aware version of our simple scheduler and made it
  the default instead of the affine scheduler. The latter is in no state to
  be used yet. It causes enormous latencies (I've seen up to 0.1s) even when
  six or seven CPUs were idle at the same time, totally killing parallelism.
  That's also the reason why a -j8 build was slower than a -j2. This is no
  longer the case. On my machine the -j2 build takes about 10% less time now
  and the -j8 build saves another 20%. The latter is not particularly
  impressive (compared with Linux), but that seems to be due to lock
  contention.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34615 a95241bf-73f2-0310-859d-f6bbb57e9c96

This commit is contained in:

Ingo Weinhold

2009-12-10 11:54:38 +00:00

parent 10b4833178

commit 3533b6597d

13 changed files with 498 additions and 27 deletions

									
										src/system/kernel/thread.cpp
									
		-1
	
												View File
												
				@@ -254,7 +254,6 @@ create_thread_struct(struct thread *inthread, const char *name,

					thread->cpu = cpu;

					thread->previous_cpu = NULL;

					thread->pinned_to_cpu = 0;

					thread->keep_scheduled = 0;

					thread->fault_handler = 0;

					thread->page_faults_allowed = 1;

					thread->kernel_stack_area = -1;