ysr [Fri, 16 Oct 2009 02:05:46 -0700] rev 4030
6888898: CMS: ReduceInitialCardMarks unsafe in the presence of cms precleaning
6889757: G1: enable card mark elision for initializing writes from compiled code (ReduceInitialCardMarks)
Summary: Defer the (compiler-elided) card-mark upon a slow-path allocation until after the store and before the next subsequent safepoint; G1 now answers yes to can_elide_tlab_write_barriers().
Reviewed-by: jcoomes, kvn, never
tonyp [Mon, 19 Oct 2009 17:02:19 -0400] rev 4029
Merge
tonyp [Wed, 07 Oct 2009 19:01:55 -0400] rev 4028
6866190: Remove SIMPLE_STACK code from TaskQueue
Summary: What the title says. We don't use SIMPLE_STACK any more.
Reviewed-by: ysr
bobv [Wed, 07 Oct 2009 09:48:42 -0400] rev 4027
6888847: TaskQueue needs release_store() for correctness on RMO machines
Summary: See title.
Reviewed-by: jmasa, ysr, jcoomes, iveresov, tonyp
tonyp [Wed, 07 Oct 2009 10:09:57 -0400] rev 4026
6888619: G1: too many guarantees in concurrent marking
Summary: change more guarantees in concurrent marking into asserts.
Reviewed-by: apetrusenko, iveresov
tonyp [Wed, 07 Oct 2009 09:42:18 -0400] rev 4025
6888316: G1: has_aborted() || _cm->region_stack_empty() fails
Summary: Remove incorrect guarantee.
Reviewed-by: apetrusenko, iveresov
tonyp [Mon, 05 Oct 2009 12:05:48 -0400] rev 4024
6847956: G1: crash in oopDesc*G1ParCopyHelper::copy_to_survivor_space(oopDesc*)
Summary: When we copy objects to survivors during marking, we incorrectly set NTAMS to bottom, which causes marking to miss visiting some of those objects.
Reviewed-by: apetrusenko, iveresov
tonyp [Fri, 02 Oct 2009 16:20:42 -0400] rev 4023
6882730: G1: parallel heap verification messes up region dump
Summary: It tidies up the G1 heap verification a bit. In particular, when the verification is done in parallel and there is a failure, this is propagated to the top level and the heap is dumped at the end, not by every thread that encounters a failure.
Reviewed-by: johnc, jmasa
tonyp [Fri, 02 Oct 2009 16:12:07 -0400] rev 4022
6885041: G1: inconsistent thread dump
Summary: When G1 is enabled, thread dumps are inconsistent as the info for some of the G1 threads is not formatted properly.
Reviewed-by: ysr, johnc
jcoomes [Mon, 05 Oct 2009 05:51:22 -0700] rev 4021
6887948: test/gc/6845368/bigobj.java fails due to timeout
Reviewed-by: iveresov