Summary: Under certain circumstances, two cleanup threads can claim and process the same region.
Reviewed-by: apetrusenko, ysr

6652160: G1: assert(cur_used_bytes == _g1->recalculate_used(),"It should!") at g1CollectorPolicy.cpp:1425
Summary: In attempt_allocation_slow() wait for the concurrent cleanup to complete before modifying _summary_bytes_used.
Reviewed-by: jmasa, apetrusenko

6723570: G1: assertion failure: p == current_top or oop(p)->is_oop(),"p is not a block start" (revisited!)
Summary: Fixed the incorrect assigment to G1OffsetTableContigSpace::_gc_time_stamp. Also added a little more paranoia to operations on a global timestamp.
Reviewed-by: tonyp

6702387: G1: assertion failure: assert(p == current_top || oop(p)->is_oop(),"p is not a block start")
Summary: Do not coalesce dead and moved objects when removing self-forwarding pointers during the evacuation failure. Also fixed a issue in a BOT refinement code for TLABs.
Reviewed-by: tonyp, jcoomes

Summary: experimental() flags will protect features of an experimental nature that are not supported in the regular product build. Made UseG1GC an experimental flag.
Reviewed-by: jmasa, kamg, coleenp

Summary: ensure that GCTask threads release resource and handle memory
Reviewed-by: jmasa, chrisphi

6710665: G1: guarantee(_cm->out_of_regions() && _cm->region_stack_empty() && _task_queue->size() == 0, ...)
Summary: Remove the incorrect assumptions from guarantee()s.
Reviewed-by: ysr, tonyp

Summary: Make FreeChunk read markword instead of LSB in _klass pointer to indicate that it's a FreeChunk for compressed oops.
Reviewed-by: ysr, jmasa

Summary: The fix takes care of three issues that can create a hole less a minimal object in the lgrp chunk
Reviewed-by: ysr, apetrusenko

Summary: First mercurial integration of the code for the Garbage-First garbage collector.
Reviewed-by: apetrusenko, iveresov, jmasa, sgoldman, tonyp, ysr

Summary: decouple set_klass() with zeroing the gap when compressed.
Reviewed-by: kvn, ysr, jrose

Summary: Use the gap also for a narrow oop field and a boxing object value.
Reviewed-by: coleenp, never

Summary: Forward port of Yumin's fix for 6670684 from HSX-11; Yumin verified the port was correct.
Reviewed-by: dcubed

Summary: Don't move tops of the chunks in ensure_parsibility(). Handle the situation with Solaris when a machine has a locality group with no memory.
Reviewed-by: apetrusenko, jcoomes, ysr

Summary: Construct the relevant CMSIsAliveClosure used by CMS during parallel reference processing with the correct span. It had incorrectly been constructed with an empty span, a regression introduced in 6417901.
Reviewed-by: jcoomes

Summary: NUMA-aware allocator port to Linux
Reviewed-by: jmasa, apetrusenko

Summary: Compressed oops in instances, arrays, and headers. Code contributors are coleenp, phh, never, swamyv
Reviewed-by: jmasa, kamg, acorn, tbell, kvn, rasbold

6539517: CR 6186200 should be extended to perm gen allocation to prevent spurious OOM's from perm gen
Reviewed-by: ysr, jmasa

Reviewed-by: apetrusenko

6634032: CMS: Need CMSInitiatingPermOccupancyFraction for perm, divorcing from CMSInitiatingOccupancyFraction
Summary: The option CMSInitiatingPermOccupancyFraction now controls perm triggering threshold. Even though the actual value of the threshold has not yet been changed, so there is no change in policy, we now have the infrastructure in place for dynamically deciding when to collect the perm gen, an issue that will be addressed in the near future.
Reviewed-by: jmasa

Summary: Default values set by cms ergonomics are set with FLAG_SET_DEFAULT so down stream the values look like the default values and affect how later parameters are set.  Set these values with FLAG_SET_ERGO instead and adjust how later parameters are interpreted.
Reviewed-by: iveresov, apetrusenko, pbk, ysr

Summary: Reduce the amount of related code replication and improve pretty printing.
Reviewed-by: jmasa

Summary: Use the same default number of GC threads as used by ParNewGC and ConcMarkSweepGC (i.e., the 5/8th rule).
Reviewed-by: ysr, tonyp

Summary: Use correct allocation path in expand_and_allocate() so object's mark and p-bits are set as appropriate.
Reviewed-by: jmasa, pbk

Summary: Take lock conditionally (in asynchronous mode only) when updating the dead-object map.
Reviewed-by: jmasa

Summary: In verification take into account partial coverage of a region by a card and expansion of the card table.
Reviewed-by: ysr, apetrusenko