archive/jdk
1
0
Fork 0
mirror of https://github.com/openjdk/jdk.git synced 2025-09-21 19:44:41 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

8036860: Pad and cache-align the BiasedMappedArray

Browse source 
Pad and cache-align BiasedMappedArray instances by default to avoid performance variability problems due to false sharing, as instances of this data structures are typically used for performance sensitive code.

Reviewed-by: brutisso, stefank
This commit is contained in:
Thomas Schatzl 2014-03-17 13:07:55 +01:00
parent 483ea400a5
commit a07b2194f7
4 changed files with 28 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

8
hotspot/src/share/vm/gc_implementation/g1/g1BiasedArray.hpp
View file

@ -25,8 +25,8 @@
#ifndef SHARE_VM_GC_IMPLEMENTATION_G1_G1BIASEDARRAY_HPP
#define SHARE_VM_GC_IMPLEMENTATION_G1_G1BIASEDARRAY_HPP
#include "memory/allocation.hpp"
#include "utilities/debug.hpp"
#include "memory/allocation.inline.hpp"
// Implements the common base functionality for arrays that contain provisions
// for accessing its elements using a biased index.
@ -48,11 +48,7 @@ protected:
_bias(0), _shift_by(0) { }
// Allocate a new array, generic version.
static address create_new_base_array(size_t length, size_t elem_size) {
assert(length > 0, "just checking");
assert(elem_size > 0, "just checking");
return NEW_C_HEAP_ARRAY(u_char, length * elem_size, mtGC);
}
static address create_new_base_array(size_t length, size_t elem_size);
// Initialize the members of this class. The biased start address of this array
// is the bias (in elements) multiplied by the element size.