Materialize the pointer to the polling page in a register instead of using rip-relative addressing when the distance from the code cache is larger than disp32.
Reviewed-by: never, kvn
6627787: CMS: JVM refuses to start up with -Xms16m -Xmx16m
7000125: CMS: Anti-monotone young gen sizing with respect to maximum whole heap size specification
7027529: CMS: retire CMSUseOldDefaults flag
Simplify CMS heap sizing code, relying on ergonomic initial sizing consistent with other collectors for the most part, controlling only young gen sizing to rein in pause times. Make CMS young gen sizing default statically cpu-dependant. Remove inconsistencies wrt generation sizing and policy code, allowing for the fixing for 6627787 and 7000125. For 7027529, retire the flag CMSUseOldDefaults which had been introduced as a bridge from JDK 5 to JDK 6 a number of years ago.
Reviewed-by: brutisso, poonam
Bulk zeroing reduction didn't work with G1, because arraycopy would call pre-barriers on uninitialized oops. The solution is to have version of arraycopy stubs that don't have pre-barriers. Also refactored arraycopy stubs generation on SPARC to be more readable and reduced the number of stubs necessary in some cases.
Reviewed-by: jrose, kvn, never
Copy small (<8) strings on stack if str+16 crosses a page boundary and load from stack into XMM. Back up pointer when loading string's tail.
Reviewed-by: never
Rearranged the pre-barrier placement in arraycopy stubs so that they are properly called in case of chained calls. Also refactored the code a little bit so that it looks uniform across the platforms and is more readable.
Reviewed-by: never, kvn
Add option for gcc 4 and above, define JNIEXPORT and JNIIMPORT to visibility=default, add for jio_snprintf and others since -fvisibility=hidden overrides --version-script definitions.
Reviewed-by: kamg, never
With recent G1 allocation path changes, the value of StackShadowPages in fast debug builds of the JVM, is no longer large enough to prevent the JVM C++ code from touching the stack guard pages. Increase the value of StackShadowPages to a suitable value.
Reviewed-by: ysr, tonyp, coleenp
Track allocated bytes in Thread's, update on TLAB retirement and direct allocation in Eden and tenured, add JNI methods for ThreadMXBean.
Reviewed-by: coleenp, kvn, dholmes, ysr
Implementation of the CAS primitive for x64 compressed oops was incorrect. It kills rscratch2 register (r11), which is allocatable in C1. Also, we don't need to restore cmpval as it's never used after that, so we need only one temporary register, which can be scratch1.
Reviewed-by: kvn, never
C1 with profiling doesn't check whether the MDO has been really allocated, which can silently fail if the perm gen is full. The solution is to check if the allocation failed and bailout out of inlining or compilation.
Reviewed-by: kvn, never
Move functions from hpi library to hotspot, communicate with licensees and open source community, check jdk for dependency, file CCC request
Reviewed-by: coleenp, acorn, dsamersoff
This change implements compressed oops for C1 for x64 and sparc. The changes are mostly on the codegen level, with a few exceptions when we do access things outside of the heap that are uncompressed from the IR. Compressed oops are now also enabled with tiered.
Reviewed-by: twisti, kvn, never, phh
Using r12 as temporary register around call_VM trashes heapbase becausecall_VM doesn't always return immediately to following code (eg forward_exception, popframe and early return support).
Reviewed-by: never, kvn
