Summary: Attempting to initiate a marking cycle when allocating a humongous object can, if a marking cycle is successfully initiated by another thread, result in the allocating thread spinning until the marking cycle is complete. Eliminate a deadlock between the main ConcurrentMarkThread, the SurrogateLocker thread, the VM thread, and a mutator thread waiting on the SecondaryFreeList_lock (while free regions are going to become available) by not manipulating the pending list lock during the prologue and epilogue of the cleanup pause.
Reviewed-by: brutisso, jcoomes, tonyp

Reviewed-by: jcoomes, brutisso

Summary: Detect cygwin prescence when setting up PLATFORM. Translate cygwin style paths before passing them on to jtreg.
Reviewed-by: jcoomes, brutisso

Reviewed-by: brutisso

Reviewed-by: jcoomes

Summary: On Mac OS X, align system property "os.arch" with Apple legacy JDKs.  Also, improve os.name string matching by using contains() method instead of .startsWith().
Reviewed-by: dcubed, phh, ohair, katleman
Contributed-by: james.melvin@oracle.com

Summary: The JVMTI ResourceExhausted events must be generated in all places where OOME is thrown
Reviewed-by: acorn, coleenp, dcubed, dholmes, dsamersoff, jwilhelm, tonyp
Contributed-by: serguei.spitsyn@oracle.com

Summary: Check clean words instead of clean bytes
Reviewed-by: jcoomes, jmasa, jwilhelm, ysr
Contributed-by: alexey.ragozin@gmail.com

Summary: Add destructor to TruncatedSeq and call delete when necessary
Reviewed-by: johnc, tonyp

Reviewed-by: kvn, never
Contributed-by: nils.eliasson@oracle.com

Summary: Don't attempt to uncommit SHM-based large pages
Reviewed-by: kvn

Reviewed-by: jcoomes

Summary: Backout the fix for 7152031; the real bug to be fixed in jdk repo.
Reviewed-by: michaelm

Summary: Update to use the right path on MacOS X.
Reviewed-by: michaelm, jcoomes, amurillo

Reviewed-by: dholmes, minqi, kvn, coleenp

Summary: Allow fallback to non-intrinsic inline case
Reviewed-by: kvn, jrose, never
Contributed-by: nils.eliasson@oracle.com

Summary: Breakpoint needs to do 64-bit compare for pointers on sparcv9
Reviewed-by: coleenp, never
Contributed-by: dean.long@oracle.com

Reviewed-by: ohair

Reviewed-by: ohair, never

Reviewed-by: jcoomes

Summary: Update generation and collector counters only if UsePerfData is enabled.
Reviewed-by: iveresov, tonyp

Summary: C1 does not recognize the default branch of a lookupswitch or tableswitch bytecode to be a safepoint if backward.
Reviewed-by: kvn, never

Reviewed-by: kvn, twisti

Summary: Assert doesn't respect invokedynamic opcode
Reviewed-by: dcubed, phh