/* * Copyright (c) 2006, 2021, Oracle and/or its affiliates. All rights reserved. * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER. * * This code is free software; you can redistribute it and/or modify it * under the terms of the GNU General Public License version 2 only, as * published by the Free Software Foundation. * * This code is distributed in the hope that it will be useful, but WITHOUT * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License * version 2 for more details (a copy is included in the LICENSE file that * accompanied this code). * * You should have received a copy of the GNU General Public License version * 2 along with this work; if not, write to the Free Software Foundation, * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. * * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA * or visit www.oracle.com if you need additional information or have any * questions. *
*/
/* * The NUMA-aware allocator (MutableNUMASpace) is basically a modification * of MutableSpace which preserves interfaces but implements different * functionality. The space is split into chunks for each locality group * (resizing for adaptive size policy is also supported). For each thread * allocations are performed in the chunk corresponding to the home locality * group of the thread. Whenever any chunk fills-in the young generation * collection occurs. * The chunks can be also be adaptively resized. The idea behind the adaptive * sizing is to reduce the loss of the space in the eden due to fragmentation. * The main cause of fragmentation is uneven allocation rates of threads. * The allocation rate difference between locality groups may be caused either by * application specifics or by uneven LWP distribution by the OS. Besides, * application can have less threads then the number of locality groups. * In order to resize the chunk we measure the allocation rate of the * application between collections. After that we reshape the chunks to reflect * the allocation rate pattern. The AdaptiveWeightedAverage exponentially * decaying average is used to smooth the measurements. The NUMASpaceResizeRate * parameter is used to control the adaptation speed by restricting the number of * bytes that can be moved during the adaptation phase. * Chunks may contain pages from a wrong locality group. The page-scanner has * been introduced to address the problem. Remote pages typically appear due to * the memory shortage in the target locality group. Besides Solaris would * allocate a large page from the remote locality group even if there are small * local pages available. The page-scanner scans the pages right after the * collection and frees remote pages in hope that subsequent reallocation would * be more successful. This approach proved to be useful on systems with high * load where multiple processes are competing for the memory.
*/
class MutableNUMASpace : public MutableSpace { friendclass VMStructs;
class LGRPSpace : public CHeapObj<mtGC> { int _lgrp_id;
MutableSpace* _space;
MemRegion _invalid_region;
AdaptiveWeightedAverage *_alloc_rate; bool _allocation_failed;
void sample() { // If there was a failed allocation make allocation rate equal // to the size of the whole chunk. This ensures the progress of // the adaptation process.
size_t alloc_rate_sample; if (_allocation_failed) {
alloc_rate_sample = space()->capacity_in_bytes();
_allocation_failed = false;
} else {
alloc_rate_sample = space()->used_in_bytes();
}
alloc_rate()->sample(alloc_rate_sample);
}
// Check if the NUMA topology has changed. Add and remove spaces if needed. // The update can be forced by setting the force parameter equal to true. bool update_layout(bool force); // Bias region towards the lgrp. void bias_region(MemRegion mr, int lgrp_id); // Free pages in a given region. void free_region(MemRegion mr); // Get current chunk size.
size_t current_chunk_size(int i); // Get default chunk size (equally divide the space).
size_t default_chunk_size(); // Adapt the chunk size to follow the allocation rate.
size_t adaptive_chunk_size(int i, size_t limit); // Scan and free invalid pages. void scan_pages(size_t page_count); // Return the bottom_region and the top_region. Align them to page_size() boundary. // |------------------new_region---------------------------------| // |----bottom_region--|---intersection---|------top_region------| void select_tails(MemRegion new_region, MemRegion intersection,
MemRegion* bottom_region, MemRegion *top_region); // Try to merge the invalid region with the bottom or top region by decreasing // the intersection area. Return the invalid_region aligned to the page_size() // boundary if it's inside the intersection. Return non-empty invalid_region // if it lies inside the intersection (also page-aligned). // |------------------new_region---------------------------------| // |----------------|-------invalid---|--------------------------| // |----bottom_region--|---intersection---|------top_region------| void merge_regions(MemRegion new_region, MemRegion* intersection,
MemRegion *invalid_region);
public:
GrowableArray<LGRPSpace*>* lgrp_spaces() const { return _lgrp_spaces; }
MutableNUMASpace(size_t alignment); virtual ~MutableNUMASpace(); // Space initialization. virtualvoid initialize(MemRegion mr, bool clear_space, bool mangle_space, bool setup_pages = SetupPages,
WorkerThreads* pretouch_workers = NULL); // Update space layout if necessary. Do all adaptive resizing job. virtualvoid update(); // Update allocation rate averages. virtualvoid accumulate_statistics();
Die Informationen auf dieser Webseite wurden
nach bestem Wissen sorgfältig zusammengestellt. Es wird jedoch weder Vollständigkeit, noch Richtigkeit,
noch Qualität der bereit gestellten Informationen zugesichert.
Bemerkung:
Die farbliche Syntaxdarstellung ist noch experimentell.
