Reviewed-by: jmasa, mgerdin

Reviewed-by: tschatzl, brutisso

Reviewed-by: johnc, tschatzl

Summary: Minimize redundant verifications of Klasses.
Reviewed-by: hseigel, jmasa

Summary: Added Atomic::inc(short*) to support this change.
Reviewed-by: coleenp, dcubed, dholmes, minqi

Summary: Add code pc location for compiled code
Reviewed-by: kvn, coleenp
Contributed-by: doug.simon@oracle.com

Summary: Detect this crash in the signal handler and give a fatal error message instead of making us chase down bugs that don't reproduce
Reviewed-by: kvn, mgerdin, dholmes

Reviewed-by: sla, dholmes

Summary: Reduce default size of Thread::_metadata_handles from 300 to 30
Reviewed-by: coleenp, sspitsyn

Summary: Missing linebreak in hotspot log.
Reviewed-by: brutisso, tschatzl
Contributed-by: vladimir.kempik@oracle.com

8001345: VM crashes with assert(n->outcnt() != 0 || C->top() == n || n->is_Proj()) failed: No dead instructions after post-alloc
Summary: Remove unnecessary LoadN / DecodeN nodes at MemBarAcquire nodes.
Reviewed-by: kvn, roland

Reviewed-by: tschatzl, dholmes, coleenp, johnc

Summary: Parallelize the scanning of the intern string table by having each GC worker claim a given number of buckets. Changes were also reviewed by Per Liden <per.liden@oracle.com>.
Reviewed-by: tschatzl, stefank, twisti

Summary: Windows reserves and only partially commits thread stack. For detecting more thread stack space for execution,  Windows installs one-shot page as guard page just before the current commited edge. It will trigger STACK_OVERFLOW_EXCEPTION when lands on last 4 pages of thread stack space. StackYellowPages default value is 2 on Windows (plus 1 page of StackRedPages, 3 pages guarded by hotspot) so the exception happens one page before Yellow pages. Same route executed second time will have one more page brought in, this leads same execution with different stack depth(interpreter mode). We need match Windows settings so the stack overflow exception will not happen before Yellow pages.
Reviewed-by: dholmes
Contributed-by: andreas.schoesser@sap.com

Summary: Fixed NMT race condition for realloc, uncommit and release
Reviewed-by: coleenp, ccheung

Summary: modify declaration of 'swap' overloading; dodge optimizer bug in c1_LIR.cpp
Reviewed-by: kvn, jrose

Reviewed-by: dcubed, mgronlun

Summary: Increase HeapBaseMinAddress for G1 from 256m to 1g on Solaris x86
Reviewed-by: mgerdin, coleenp, kvn

8016157: During CTW: C2: assert(!def_outside->member(r)) failed: Use of external LRG overlaps the same LRG defined in this block
Summary: Disable rematerialization for negD node
Reviewed-by: kvn, roland

Summary: replace linked data structure with array (performance)
Reviewed-by: kvn
Contributed-by: christian.thalinger@oracle.com, david.r.chase@oracle.com

8014959: assert(Compile::current()->live_nodes() < (uint)MaxNodeLimit) failed: Live Node limit exceeded limit
Summary: Insert extra checks and bailouts for too many nodes
Reviewed-by: kvn

8003268: SharedRuntime::generate_native_wrapper doesn't save all registers across runtime tracing calls for JNI critical native methods
Reviewed-by: kvn

Summary: Simple fix when memory allocation site is gone, NMT should report 0 memory size, instead old memory size.
Reviewed-by: dcubed, ctornqvi

Reviewed-by: sla, rbackman

6493116: JVMTI Doc: GetOwnedMonitorStackDepthInfo has a typo in monitor_info_ptr parameter description
Summary: A typo in the parameter spelling, a bound update missed when the parameter was renamed
Reviewed-by: sla, minqi
Contributed-by: serguei.spitsyn@oracle.com

Summary: Removed restart macros from all os:close calls on Solaris, Linux, MacOS X platforms.
Reviewed-by: dcubed, dholmes

Summary: Simple fix to add NULL pointer check that can cause segv
Reviewed-by: coleenp, ctornqvi

Reviewed-by: dholmes, coleenp
Contributed-by: jeremymanson@google.com, calvin.cheung@oracle.com

8013057: assert(_needs_gc || SafepointSynchronize::is_at_safepoint()) failed: only read at safepoint
Summary: Detect mmap() commit failures in Linux and Solaris os::commit_memory() impls and call vm_exit_out_of_memory(). Add os::commit_memory_or_exit(). Also tidy up some NMT accounting and some mmap() return value checking.
Reviewed-by: zgu, stefank, dholmes, dsamersoff

Reviewed-by: acorn, coleenp, sla
Contributed-by: NKaren Kinnear &lt;karen.kinnear@oracle.com&gt;, Bengt Rutisson &lt;bengt.rutisson@oracle.com&gt;, Calvin Cheung &lt;calvin.cheung@oracle.com&gt;, Erik Gahlin &lt;erik.gahlin@oracle.com&gt;, Erik Helin &lt;erik.helin@oracle.com&gt;, Jesper Wilhelmsson &lt;jesper.wilhelmsson@oracle.com&gt;, Keith McGuigan &lt;keith.mcguigan@oracle.com&gt;, Mattias Tobiasson &lt;mattias.tobiasson@oracle.com&gt;, Markus Gronlund &lt;markus.gronlund@oracle.com&gt;, Mikael Auno &lt;mikael.auno@oracle.com&gt;, Nils Eliasson &lt;nils.eliasson@oracle.com&gt;, Nils Loodin &lt;nils.loodin@oracle.com&gt;, Rickard Backman &lt;rickard.backman@oracle.com&gt;, Staffan Larsen &lt;staffan.larsen@oracle.com&gt;, Stefan Karlsson &lt;stefan.karlsson@oracle.com&gt;, Yekaterina Kantserova <yekaterina.kantserova@oracle.com>

Summary: Removed most of the SPARC V8 instructions
Reviewed-by: kvn, twisti

Summary: Enable large file support for generated 32-bit ostream.o on Linux and Solaris (as only the two need this) by setting -D_FILE_OFFSET_BITS=64 in compilation
Reviewed-by: tbell, mgerdin, dcubed
Contributed-by: Ntamao <tao.mao@oracle.com>

Summary: Forced SPARC MacroAssembler eden_alloate to use long branch to slow case
Reviewed-by: kvn, twisti

Reviewed-by: twisti, jrose

Summary: Use SA_ONSTACK flag to ensure signal gets delivered properly.
Reviewed-by: dholmes, coleenp
Contributed-by: gerard.ziemski@oracle.com

Reviewed-by: twisti, dholmes, kvn

Summary: Remove dead code.
Reviewed-by: twisti, coleenp

Summary: added guarantee()
Reviewed-by: kvn

Summary: In a full GC, move the verification after the GC to after RSet rebuilding. Verify RSet entries during a full GC under control of a flag.
Reviewed-by: tschatzl, brutisso

Summary: Create G1ParGCAllocBufferContainer to contain two buffers instead of previously using one buffer, in order to hold the first priority buffer longer. Thus, when some large objects hits the value of free space left in the first priority buffer it has an alternative to fit in the second priority buffer while the first priority buffer is given more chances to try allocating smaller objects. Overall, it will improve heap space efficiency.
Reviewed-by: johnc, jmasa, brutisso
Contributed-by: Ntamao <tao.mao@oracle.com>