From: SeongJae Park <sjpark@xxxxxxxxx> Introduction ============ Memory management decisions can be improved if finer data access information is available. However, because such finer information usually comes with higher overhead, most systems including Linux forgives the potential benefit and rely on only coarse information or some light-weight heuristics. The pseudo-LRU and the aggressive THP promotions are such examples. A number of data access pattern awared memory management optimizations (refer to 'Appendix A' for more details) consistently say the potential benefit is not small. However, none of those has successfully merged to the mainline Linux kernel mainly due to the absence of a scalable and efficient data access monitoring mechanism. Refer to 'Appendix B' to see the limitations of existing memory monitoring mechanisms. DAMON is a data access monitoring subsystem for the problem. It is 1) accurate enough to be used for the DRAM level memory management (a straightforward DAMON-based optimization achieved up to 2.55x speedup), 2) light-weight enough to be applied online (compared to a straightforward access monitoring scheme, DAMON is up to 94,242.42x lighter) and 3) keeps predefined upper-bound overhead regardless of the size of target workloads (thus scalable). Refer to 'Appendix C' if you interested in how it is possible, and 'Appendix F' to know how the numbers collected. DAMON has mainly designed for the kernel's memory management mechanisms. However, because it is implemented as a standalone kernel module and provides several interfaces, it can be used by a wide range of users including kernel space programs, user space programs, programmers, and administrators. DAMON is now supporting the monitoring only, but it will also provide simple and convenient data access pattern awared memory managements by itself. Refer to 'Appendix D' for more detailed expected usages of DAMON. Boring? Here Are Something Colorful =================================== For intuitive understanding of DAMON, I made web pages[1-8] showing the visualized dynamic data access pattern of various realistic workloads in PARSEC3 and SPLASH-2X bechmark suites. The figures are generated using the user space tool in 10th patch of this patchset. There are pages showing the heatmap format dynamic access pattern of each workload for heap area[1], mmap()-ed area[2], and stack[3] area. I splitted the entire address space to the three area because there are huge unmapped regions between the areas. You can also show how the dynamic working set size of each workload is distributed[4], and how it is chronologically changing[5]. The most important characteristic of DAMON is its promise of the upperbound of the monitoring overhead. To show whether DAMON keeps the promise well, I visualized the number of monitoring operations required for each 5 milliseconds, which is configured to not exceed 1000. You can show the distribution of the numbers[6] and how it changes chronologically[7]. [1] https://damonitor.github.io/reports/latest/by_image/heatmap.0.png.html [2] https://damonitor.github.io/reports/latest/by_image/heatmap.1.png.html [3] https://damonitor.github.io/reports/latest/by_image/heatmap.2.png.html [4] https://damonitor.github.io/reports/latest/by_image/wss_sz.png.html [5] https://damonitor.github.io/reports/latest/by_image/wss_time.png.html [6] https://damonitor.github.io/reports/latest/by_image/nr_regions_sz.png.html [7] https://damonitor.github.io/reports/latest/by_image/nr_regions_time.png.html Future Plans ============ This patchset is only for the first stage of DAMON. As soon as this patchset is merged, official patchsets for below future plans will be posted. Automate Data Access Pattern-aware Memory Management ---------------------------------------------------- Though DAMON provides the monitoring feature, implementing data access pattern aware memory management schemes could be difficult to beginners. DAMON will be able to do most of the work by itself in near future. Users will be required to only describe what kind of data access monitoring-based operation schemes they want. By applying a very simple scheme for THP promotion/demotion with a latest version of the patchset (not posted yet), DAMON achieved 18x lower memory space overhead compared to THP while preserving about 50% of the THP performance benefit with SPLASH-2X benchmark suite. An RFC patchset for this plan is already available (https://lore.kernel.org/linux-mm/20200429124540.32232-1-sjpark@xxxxxxxxxx/). Support Various Address Spaces ------------------------------ Currently, DAMON supports virtual memory address spaces using PTE Accessed bits as its access checking primitive. However, the core design of DAMON is not dependent to such implementation details. In a future, DAMON will decouple those and support various address spaces including physical memory. It will further allow users to configure and even implement the primitives by themselves for their special usecase. Monitoring of page cache, NUMA nodes, specific files, or block devices would be examples of such usecases. An RFC patchset for this plan is already available (https://lore.kernel.org/linux-mm/20200409094232.29680-1-sjpark@xxxxxxxxxx/). Frequently Asked Questions ========================== Q: Why a new module, instead of extending perf or other tools? A: First, DAMON aims to be used by other programs including the kernel. Therefore, having dependency to specific tools like perf is not desirable. Second, because it need to be lightweight as much as possible so that it can be used online, any unnecessary overhead such as kernel - user space context switching cost should be avoided. These are the two most biggest reasons why DAMON is implemented in the kernel space. The idle page tracking subsystem would be the kernel module that most seems similar to DAMON. However, its own interface is not compatible with DAMON. Also, the internal implementation of it has no common part to be reused by DAMON. Q: Can 'perf mem' or PMUs used instead of DAMON? A: No. Roughly speaking, DAMON has two seperate layers. The low layer is access check of pages, and the higher layer is its core mechanisms for overhead controlling. For the low layer, DAMON is now using the PTE Accessed bits. Other H/W or S/W features that can be used for the access check of pages, such as 'perf mem', PMU, or even page idle, could be used instead in the layer. However, those could not alternate the high layer of DAMON. Evaluations =========== We evaluated DAMON's overhead, monitoring quality and usefulness using 25 realistic workloads on my QEMU/KVM based virtual machine. DAMON is lightweight. It consumes only 0.03% more system memory and up to 1% CPU time. It makes target worloads only 0.7% slower. DAMON is accurate and useful for memory management optimizations. An experimental DAMON-based operation scheme for THP removes 63.12% of THP memory overheads while preserving 49.15% of THP speedup. Another experimental DAMON-based 'proactive reclamation' implementation reduces 85.85% of residentail sets and 21.98% of system memory footprint while incurring only 2.42% runtime overhead in best case (parsec3/freqmine). NOTE that the experimentail THP optimization and proactive reclamation are not for production, just only for proof of concepts. Please refer to 'Appendix E' for detailed evaluation setup and results. References ========== Prototypes of DAMON have introduced by an LPC kernel summit track talk[1] and two academic papers[2,3]. Please refer to those for more detailed information, especially the evaluations. The latest version of the patchsets has also introduced by an LWN artice[4]. [1] SeongJae Park, Tracing Data Access Pattern with Bounded Overhead and Best-effort Accuracy. In The Linux Kernel Summit, September 2019. https://linuxplumbersconf.org/event/4/contributions/548/ [2] SeongJae Park, Yunjae Lee, Heon Y. Yeom, Profiling Dynamic Data Access Patterns with Controlled Overhead and Quality. In 20th ACM/IFIP International Middleware Conference Industry, December 2019. https://dl.acm.org/doi/10.1145/3366626.3368125 [3] SeongJae Park, Yunjae Lee, Yunhee Kim, Heon Y. Yeom, Profiling Dynamic Data Access Patterns with Bounded Overhead and Accuracy. In IEEE International Workshop on Foundations and Applications of Self- Systems (FAS 2019), June 2019. [4] Jonathan Corbet, Memory-management optimization with DAMON. In Linux Weekly News (LWN), Feb 2020. https://lwn.net/Articles/812707/ Baseline and Complete Git Trees =============================== The patches are based on the v5.6. You can also clone the complete git tree: $ git clone git://github.com/sjp38/linux -b damon/patches/v11 The web is also available: https://github.com/sjp38/linux/releases/tag/damon/patches/v11 This patchset contains patches for the stabled main logic of DAMON only. The latest DAMON development tree is also available at: https://github.com/sjp38/linux/tree/damon/master Sequence Of Patches =================== The patches are organized in the following sequence. The first two patches are preparation of DAMON patchset. The 1st patch adds typos found in previous versions of DAMON patchset to 'scripts/spelling.txt' so that the typos can be caught by 'checkpatch.pl'. The 2nd patch exports 'lookup_page_ext()' to GPL modules so that it can be used by DAMON even though it is built as a loadable module. Next five patches implement the core of DAMON and it's programming interface. The 3rd patch introduces DAMON module, it's data structures, and data structure related common functions. Following four patches (4nd to 7th) implements the core mechanisms of DAMON, namely regions based sampling (patch 4), adaptive regions adjustment (patches 5-6), and dynamic memory mapping chage adoption (patch 7). Following four patches are for low level users of DAMON. The 8th patch implements callbacks for each of monitoring steps so that users can do whatever they want with the access patterns. The 9th one implements recording of access patterns in DAMON for better convenience and efficiency. Each of next two patches (10th and 11th) respectively adds a debugfs interface for privileged people and/or programs in user space, and a tracepoint for other tracepoints supporting tracers such as perf. Two patches for high level users of DAMON follows. To provide a minimal reference to the debugfs interface and for high level use/tests of the DAMON, the next patch (12th) implements an user space tool. The 13th patch adds a document for administrators of DAMON. Next two patches are for tests. The 14th and 15th patches provide unit tests (based on kunit) and user space tests (based on kselftest), respectively. Finally, the last patch (16th) updates the MAINTAINERS file. Patch History ============= The most biggest change in this version is support of minimal region size, which defaults to 'PAGE_SIZE'. This change will reduce unnecessary region splits and thus improve the quality of the output. In a future, we will be able to make this configurable for support of various access check primitives such as PMUs. Changes from v10 (https://lore.kernel.org/linux-mm/20200505110815.10532-1-sjpark@xxxxxxxxxx/) - Reduce aggressive split overhead by doing it only if required Changes from v9 (https://lore.kernel.org/linux-mm/20200427120442.24179-1-sjpark@xxxxxxxxxx/) - Split each region into 4 subregions if possible (Jonathan Cameraon) - Update kunit test for the split code change Changes from v8 (https://lore.kernel.org/linux-mm/20200406130938.14066-1-sjpark@xxxxxxxxxx/) - Make regions always aligned by minimal region size that can be changed (Stefan Nuernberger) - Store binary format version in the recording file (Stefan Nuernberger) - Use 'int' for pid instead of 'unsigned long' (Stefan Nuernberger) - Fix a race condition in damon thread termination (Stefan Nuernberger) - Optimize random value generation and recording (Stefan Nuernberger) - Clean up commit messages and comments (Stefan Nuernberger) - Clean up code (Stefan Nuernberger) - Use explicit signalling and 'do_exit()' for damon thread termination - Add more typos to spelling.txt - Update the performance evaluation results - Describe future plans in the cover letter Changes from v7 (https://lore.kernel.org/linux-mm/20200318112722.30143-1-sjpark@xxxxxxxxxx/) - Cleanup variable names (Jonathan Cameron) - Split sampling address setup from access_check() (Jonathan Cameron) - Make sampling address to always locate in the region (Jonathan Cameron) - Make initial region's sampling addr to be old (Jonathan Cameron) - Split kdamond on/off function to seperate functions (Jonathan Cameron) - Fix wrong kernel doc comments (Jonathan Cameron) - Reset 'last_accessed' to false in kdamond_check_access() if necessary - Rebase on v5.6 Changes from v6 (https://lore.kernel.org/linux-mm/20200224123047.32506-1-sjpark@xxxxxxxxxx/) - Wordsmith cover letter (Shakeel Butt) - Cleanup code and commit messages (Jonathan Cameron) - Avoid kthread_run() under spinlock critical section (Jonathan Cameron) - Use kthread_stop() (Jonathan Cameron) - Change tracepoint to trace regions (Jonathan Cameron) - Implement API from the beginning (Jonathan Cameron) - Fix typos (Jonathan Cameron) - Fix access checking to properly handle regions smaller than single page (Jonathan Cameron) - Add found typos to 'scripts/spelling.txt' - Add recent evaluation results including DAMON-based Operation Schemes Changes from v5 (https://lore.kernel.org/linux-mm/20200217103110.30817-1-sjpark@xxxxxxxxxx/) - Fix minor bugs (sampling, record attributes, debugfs and user space tool) - selftests: Add debugfs interface tests for the bugs - Modify the user space tool to use its self default values for parameters - Fix pmg huge page access check Changes from v4 (https://lore.kernel.org/linux-mm/20200210144812.26845-1-sjpark@xxxxxxxxxx/) - Add 'Reviewed-by' for the kunit tests patch (Brendan Higgins) - Make the unit test to depedns on 'DAMON=y' (Randy Dunlap and kbuild bot) Reported-by: kbuild test robot <lkp@xxxxxxxxx> - Fix m68k module build issue Reported-by: kbuild test robot <lkp@xxxxxxxxx> - Add selftests - Seperate patches for low level users from core logics for better reading - Clean up debugfs interface - Trivial nitpicks Changes from v3 (https://lore.kernel.org/linux-mm/20200204062312.19913-1-sj38.park@xxxxxxxxx/) - Fix i386 build issue Reported-by: kbuild test robot <lkp@xxxxxxxxx> - Increase the default size of the monitoring result buffer to 1 MiB - Fix misc bugs in debugfs interface Changes from v2 (https://lore.kernel.org/linux-mm/20200128085742.14566-1-sjpark@xxxxxxxxxx/) - Move MAINTAINERS changes to last commit (Brendan Higgins) - Add descriptions for kunittest: why not only entire mappings and what the 4 input sets are trying to test (Brendan Higgins) - Remove 'kdamond_need_stop()' test (Brendan Higgins) - Discuss about the 'perf mem' and DAMON (Peter Zijlstra) - Make CV clearly say what it actually does (Peter Zijlstra) - Answer why new module (Qian Cai) - Diable DAMON by default (Randy Dunlap) - Change the interface: Seperate recording attributes (attrs, record, rules) and allow multiple kdamond instances - Implement kernel API interface Changes from v1 (https://lore.kernel.org/linux-mm/20200120162757.32375-1-sjpark@xxxxxxxxxx/) - Rebase on v5.5 - Add a tracepoint for integration with other tracers (Kirill A. Shutemov) - document: Add more description for the user space tool (Brendan Higgins) - unittest: Improve readability (Brendan Higgins) - unittest: Use consistent name and helpers function (Brendan Higgins) - Update PG_Young to avoid reclaim logic interference (Yunjae Lee) Changes from RFC (https://lore.kernel.org/linux-mm/20200110131522.29964-1-sjpark@xxxxxxxxxx/) - Specify an ambiguous plan of access pattern based mm optimizations - Support loadable module build - Cleanup code SeongJae Park (16): scripts/spelling: Add a few more typos mm/page_ext: Export lookup_page_ext() to GPL modules mm: Introduce Data Access MONitor (DAMON) mm/damon: Implement region based sampling mm/damon: Adaptively adjust regions mm/damon: Split regions into 3 subregions if necessary mm/damon: Apply dynamic memory mapping changes mm/damon: Implement callbacks mm/damon: Implement access pattern recording mm/damon: Add debugfs interface mm/damon: Add tracepoints tools: Add a minimal user-space tool for DAMON Documentation/admin-guide/mm: Add a document for DAMON mm/damon: Add kunit tests mm/damon: Add user space selftests MAINTAINERS: Update for DAMON .../admin-guide/mm/data_access_monitor.rst | 428 +++++ Documentation/admin-guide/mm/index.rst | 1 + MAINTAINERS | 12 + include/linux/damon.h | 78 + include/trace/events/damon.h | 43 + mm/Kconfig | 23 + mm/Makefile | 1 + mm/damon-test.h | 622 +++++++ mm/damon.c | 1511 +++++++++++++++++ mm/page_ext.c | 1 + scripts/spelling.txt | 8 + tools/damon/.gitignore | 1 + tools/damon/_dist.py | 36 + tools/damon/_recfile.py | 23 + tools/damon/bin2txt.py | 67 + tools/damon/damo | 37 + tools/damon/heats.py | 362 ++++ tools/damon/nr_regions.py | 91 + tools/damon/record.py | 212 +++ tools/damon/report.py | 45 + tools/damon/wss.py | 97 ++ tools/testing/selftests/damon/Makefile | 7 + .../selftests/damon/_chk_dependency.sh | 28 + tools/testing/selftests/damon/_chk_record.py | 108 ++ .../testing/selftests/damon/debugfs_attrs.sh | 139 ++ .../testing/selftests/damon/debugfs_record.sh | 50 + 26 files changed, 4031 insertions(+) create mode 100644 Documentation/admin-guide/mm/data_access_monitor.rst create mode 100644 include/linux/damon.h create mode 100644 include/trace/events/damon.h create mode 100644 mm/damon-test.h create mode 100644 mm/damon.c create mode 100644 tools/damon/.gitignore create mode 100644 tools/damon/_dist.py create mode 100644 tools/damon/_recfile.py create mode 100644 tools/damon/bin2txt.py create mode 100755 tools/damon/damo create mode 100644 tools/damon/heats.py create mode 100644 tools/damon/nr_regions.py create mode 100644 tools/damon/record.py create mode 100644 tools/damon/report.py create mode 100644 tools/damon/wss.py create mode 100644 tools/testing/selftests/damon/Makefile create mode 100644 tools/testing/selftests/damon/_chk_dependency.sh create mode 100644 tools/testing/selftests/damon/_chk_record.py create mode 100755 tools/testing/selftests/damon/debugfs_attrs.sh create mode 100755 tools/testing/selftests/damon/debugfs_record.sh -- 2.17.1 ================================== >8 ========================================= Appendix A: Related Works ========================= There are a number of researches[1,2,3,4,5,6] optimizing memory management mechanisms based on the actual memory access patterns that shows impressive results. However, most of those has no deep consideration about the monitoring of the accesses itself. Some of those focused on the overhead of the monitoring, but does not consider the accuracy scalability[6] or has additional dependencies[7]. Indeed, one recent research[5] about the proactive reclamation has also proposed[8] to the kernel community but the monitoring overhead was considered a main problem. [1] Subramanya R Dulloor, Amitabha Roy, Zheguang Zhao, Narayanan Sundaram, Nadathur Satish, Rajesh Sankaran, Jeff Jackson, and Karsten Schwan. 2016. Data tiering in heterogeneous memory systems. In Proceedings of the 11th European Conference on Computer Systems (EuroSys). ACM, 15. [2] Youngjin Kwon, Hangchen Yu, Simon Peter, Christopher J Rossbach, and Emmett Witchel. 2016. Coordinated and efficient huge page management with ingens. In 12th USENIX Symposium on Operating Systems Design and Implementation (OSDI). 705–721. [3] Harald Servat, Antonio J Peña, Germán Llort, Estanislao Mercadal, HansChristian Hoppe, and Jesús Labarta. 2017. Automating the application data placement in hybrid memory systems. In 2017 IEEE International Conference on Cluster Computing (CLUSTER). IEEE, 126–136. [4] Vlad Nitu, Boris Teabe, Alain Tchana, Canturk Isci, and Daniel Hagimont. 2018. Welcome to zombieland: practical and energy-efficient memory disaggregation in a datacenter. In Proceedings of the 13th European Conference on Computer Systems (EuroSys). ACM, 16. [5] Andres Lagar-Cavilla, Junwhan Ahn, Suleiman Souhlal, Neha Agarwal, Radoslaw Burny, Shakeel Butt, Jichuan Chang, Ashwin Chaugule, Nan Deng, Junaid Shahid, Greg Thelen, Kamil Adam Yurtsever, Yu Zhao, and Parthasarathy Ranganathan. 2019. Software-Defined Far Memory in Warehouse-Scale Computers. In Proceedings of the 24th International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS). ACM, New York, NY, USA, 317–330. DOI:https://doi.org/10.1145/3297858.3304053 [6] Carl Waldspurger, Trausti Saemundsson, Irfan Ahmad, and Nohhyun Park. 2017. Cache Modeling and Optimization using Miniature Simulations. In 2017 USENIX Annual Technical Conference (ATC). USENIX Association, Santa Clara, CA, 487–498. https://www.usenix.org/conference/atc17/technical-sessions/ [7] Haojie Wang, Jidong Zhai, Xiongchao Tang, Bowen Yu, Xiaosong Ma, and Wenguang Chen. 2018. Spindle: Informed Memory Access Monitoring. In 2018 USENIX Annual Technical Conference (ATC). USENIX Association, Boston, MA, 561–574. https://www.usenix.org/conference/atc18/presentation/wang-haojie [8] Jonathan Corbet. 2019. Proactively reclaiming idle memory. (2019). https://lwn.net/Articles/787611/. Appendix B: Limitations of Other Access Monitoring Techniques ============================================================= The memory access instrumentation techniques which are applied to many tools such as Intel PIN is essential for correctness required cases such as memory access bug detections or cache level optimizations. However, those usually incur exceptionally high overhead which is unacceptable. Periodic access checks based on access counting features (e.g., PTE Accessed bits or PG_Idle flags) can reduce the overhead. It sacrifies some of the quality but it's still ok to many of this domain. However, the overhead arbitrarily increase as the size of the target workload grows. Miniature-like static region based sampling can set the upperbound of the overhead, but it will now decrease the quality of the output as the size of the workload grows. DAMON is another solution that overcomes the limitations. It is 1) accurate enough for this domain, 2) light-weight so that it can be applied online, and 3) allow users to set the upper-bound of the overhead, regardless of the size of target workloads. It is implemented as a simple and small kernel module to support various users in both of the user space and the kernel space. Refer to 'Evaluations' section below for detailed performance of DAMON. For the goals, DAMON utilizes its two core mechanisms, which allows lightweight overhead and high quality of output, repectively. To show how DAMON promises those, refer to 'Mechanisms of DAMON' section below. Appendix C: Mechanisms of DAMON =============================== Basic Access Check ------------------ DAMON basically reports what pages are how frequently accessed. The report is passed to users in binary format via a ``result file`` which users can set it's path. Note that the frequency is not an absolute number of accesses, but a relative frequency among the pages of the target workloads. Users can also control the resolution of the reports by setting two time intervals, ``sampling interval`` and ``aggregation interval``. In detail, DAMON checks access to each page per ``sampling interval``, aggregates the results (counts the number of the accesses to each page), and reports the aggregated results per ``aggregation interval``. For the access check of each page, DAMON uses the Accessed bits of PTEs. This is thus similar to the previously mentioned periodic access checks based mechanisms, which overhead is increasing as the size of the target process grows. Region Based Sampling --------------------- To avoid the unbounded increase of the overhead, DAMON groups a number of adjacent pages that assumed to have same access frequencies into a region. As long as the assumption (pages in a region have same access frequencies) is kept, only one page in the region is required to be checked. Thus, for each ``sampling interval``, DAMON randomly picks one page in each region and clears its Accessed bit. After one more ``sampling interval``, DAMON reads the Accessed bit of the page and increases the access frequency of the region if the bit has set meanwhile. Therefore, the monitoring overhead is controllable by setting the number of regions. DAMON allows users to set the minimal and maximum number of regions for the trade-off. Except the assumption, this is almost same with the above-mentioned miniature-like static region based sampling. In other words, this scheme cannot preserve the quality of the output if the assumption is not guaranteed. Adaptive Regions Adjustment --------------------------- At the beginning of the monitoring, DAMON constructs the initial regions by evenly splitting the memory mapped address space of the process into the user-specified minimal number of regions. In this initial state, the assumption is normally not kept and thus the quality could be low. To keep the assumption as much as possible, DAMON adaptively merges and splits each region. For each ``aggregation interval``, it compares the access frequencies of adjacent regions and merges those if the frequency difference is small. Then, after it reports and clears the aggregated access frequency of each region, it splits each region into two regions if the total number of regions is smaller than the half of the user-specified maximum number of regions. In this way, DAMON provides its best-effort quality and minimal overhead while keeping the bounds users set for their trade-off. Applying Dynamic Memory Mappings -------------------------------- Only a number of small parts in the super-huge virtual address space of the processes is mapped to physical memory and accessed. Thus, tracking the unmapped address regions is just wasteful. However, tracking every memory mapping change might incur an overhead. For the reason, DAMON applies the dynamic memory mapping changes to the tracking regions only for each of an user-specified time interval (``regions update interval``). Appendix D: Expected Use-cases ============================== A straightforward usecase of DAMON would be the program behavior analysis. With the DAMON output, users can confirm whether the program is running as intended or not. This will be useful for debuggings and tests of design points. The monitored results can also be useful for counting the dynamic working set size of workloads. For the administration of memory overcommitted systems or selection of the environments (e.g., containers providing different amount of memory) for your workloads, this will be useful. If you are a programmer, you can optimize your program by managing the memory based on the actual data access pattern. For example, you can identify the dynamic hotness of your data using DAMON and call ``mlock()`` to keep your hot data in DRAM, or call ``madvise()`` with ``MADV_PAGEOUT`` to proactively reclaim cold data. Even though your program is guaranteed to not encounter memory pressure, you can still improve the performance by applying the DAMON outputs for call of ``MADV_HUGEPAGE`` and ``MADV_NOHUGEPAGE``. More creative optimizations would be possible. Our evaluations of DAMON includes a straightforward optimization using the ``mlock()``. Please refer to the below Evaluation section for more detail. As DAMON incurs very low overhead, such optimizations can be applied not only offline, but also online. Also, there is no reason to limit such optimizations to the user space. Several parts of the kernel's memory management mechanisms could be also optimized using DAMON. The reclamation, the THP (de)promotion decisions, and the compaction would be such a candidates. DAMON will continue its development to be highly optimized for the online/in-kernel uses. We will further automate the optimization for many usecases. Appendix E: Evaluations ======================= Setup ----- On my personal QEMU/KVM based virtual machine on an Intel i7 host machine running Ubuntu 18.04, I measure runtime and consumed system memory while running various realistic workloads with several configurations. I use 13 and 12 workloads in PARSEC3[3] and SPLASH-2X[4] benchmark suites, respectively. I personally use another wrapper scripts[5] for setup and run of the workloads. On top of this patchset, we also applied the DAMON-based operation schemes patchset[6] for this evaluation. Measurement ~~~~~~~~~~~ For the measurement of the amount of consumed memory in system global scope, I drop caches before starting each of the workloads and monitor 'MemFree' in the '/proc/meminfo' file. To make results more stable, I repeat the runs 5 times and average results. You can get stdev, min, and max of the numbers among the repeated runs in appendix below. Configurations ~~~~~~~~~~~~~~ The configurations I use are as below. orig: Linux v5.5 with 'madvise' THP policy rec: 'orig' plus DAMON running with record feature thp: same with 'orig', but use 'always' THP policy ethp: 'orig' plus a DAMON operation scheme[6], 'efficient THP' prcl: 'orig' plus a DAMON operation scheme, 'proactive reclaim[7]' I use 'rec' for measurement of DAMON overheads to target workloads and system memory. The remaining configs including 'thp', 'ethp', and 'prcl' are for measurement of DAMON monitoring accuracy. 'ethp' and 'prcl' is simple DAMON-based operation schemes developed for proof of concepts of DAMON. 'ethp' reduces memory space waste of THP by using DAMON for decision of promotions and demotion for huge pages, while 'prcl' is as similar as the original work. Those are implemented as below: # format: <min/max size> <min/max frequency (0-100)> <min/max age> <action> # ethp: Use huge pages if a region >2MB shows >5% access rate, use regular # pages if a region >2MB shows <5% access rate for >1 second 2M null 5 null null null hugepage 2M null null 5 1s null nohugepage # prcl: If a region >4KB shows <5% access rate for >5 seconds, page out. 4K null null 5 500ms null pageout Note that both 'ethp' and 'prcl' are designed with my only straightforward intuition, because those are for only proof of concepts and monitoring accuracy of DAMON. In other words, those are not for production. For production use, those should be tuned more. [1] "Redis latency problems troubleshooting", https://redis.io/topics/latency [2] "Disable Transparent Huge Pages (THP)", https://docs.mongodb.com/manual/tutorial/transparent-huge-pages/ [3] "The PARSEC Becnhmark Suite", https://parsec.cs.princeton.edu/index.htm [4] "SPLASH-2x", https://parsec.cs.princeton.edu/parsec3-doc.htm#splash2x [5] "parsec3_on_ubuntu", https://github.com/sjp38/parsec3_on_ubuntu [6] "[RFC v4 0/7] Implement Data Access Monitoring-based Memory Operation Schemes", https://lore.kernel.org/linux-mm/20200303121406.20954-1-sjpark@xxxxxxxxxx/ [7] "Proactively reclaiming idle memory", https://lwn.net/Articles/787611/ Results ------- Below two tables show the measurement results. The runtimes are in seconds while the memory usages are in KiB. Each configurations except 'orig' shows its overhead relative to 'orig' in percent within parenthesises. runtime orig rec (overhead) thp (overhead) ethp (overhead) prcl (overhead) parsec3/blackscholes 107.065 107.478 (0.39) 106.682 (-0.36) 107.365 (0.28) 111.811 (4.43) parsec3/bodytrack 79.256 79.450 (0.25) 78.645 (-0.77) 79.314 (0.07) 80.305 (1.32) parsec3/canneal 139.497 141.181 (1.21) 121.526 (-12.88) 130.074 (-6.75) 154.644 (10.86) parsec3/dedup 11.879 11.873 (-0.05) 11.693 (-1.56) 11.948 (0.58) 12.694 (6.86) parsec3/facesim 207.814 208.467 (0.31) 203.743 (-1.96) 206.759 (-0.51) 214.603 (3.27) parsec3/ferret 190.124 190.955 (0.44) 189.575 (-0.29) 190.852 (0.38) 191.548 (0.75) parsec3/fluidanimate 211.046 212.282 (0.59) 208.832 (-1.05) 212.143 (0.52) 218.774 (3.66) parsec3/freqmine 289.259 290.096 (0.29) 288.510 (-0.26) 290.177 (0.32) 296.269 (2.42) parsec3/raytrace 118.522 119.701 (0.99) 119.469 (0.80) 118.964 (0.37) 130.584 (10.18) parsec3/streamcluster 323.619 327.830 (1.30) 283.374 (-12.44) 287.837 (-11.06) 330.216 (2.04) parsec3/swaptions 154.007 155.714 (1.11) 154.767 (0.49) 154.955 (0.62) 155.256 (0.81) parsec3/vips 58.822 58.750 (-0.12) 58.564 (-0.44) 58.807 (-0.02) 60.320 (2.55) parsec3/x264 67.335 72.516 (7.69) 64.680 (-3.94) 70.096 (4.10) 72.465 (7.62) splash2x/barnes 80.335 80.979 (0.80) 73.758 (-8.19) 78.874 (-1.82) 99.226 (23.52) splash2x/fft 33.441 33.312 (-0.38) 22.909 (-31.49) 31.561 (-5.62) 41.496 (24.09) splash2x/lu_cb 85.691 85.706 (0.02) 84.352 (-1.56) 85.943 (0.29) 88.914 (3.76) splash2x/lu_ncb 92.338 92.749 (0.45) 89.773 (-2.78) 92.888 (0.60) 94.104 (1.91) splash2x/ocean_cp 44.542 44.795 (0.57) 42.958 (-3.56) 44.061 (-1.08) 49.091 (10.21) splash2x/ocean_ncp 82.101 82.006 (-0.12) 51.418 (-37.37) 64.496 (-21.44) 105.998 (29.11) splash2x/radiosity 91.296 91.353 (0.06) 90.668 (-0.69) 91.379 (0.09) 103.265 (13.11) splash2x/radix 31.243 31.417 (0.56) 25.176 (-19.42) 30.297 (-3.03) 38.474 (23.14) splash2x/raytrace 84.405 84.863 (0.54) 83.498 (-1.08) 83.637 (-0.91) 85.166 (0.90) splash2x/volrend 87.516 88.156 (0.73) 86.311 (-1.38) 87.016 (-0.57) 88.318 (0.92) splash2x/water_nsquared 233.515 233.826 (0.13) 221.169 (-5.29) 224.430 (-3.89) 236.929 (1.46) splash2x/water_spatial 89.207 89.448 (0.27) 89.396 (0.21) 89.826 (0.69) 97.700 (9.52) total 2993.890 3014.920 (0.70) 2851.460 (-4.76) 2923.710 (-2.34) 3158.180 (5.49) memused.avg orig rec (overhead) thp (overhead) ethp (overhead) prcl (overhead) parsec3/blackscholes 1819997.200 1832626.000 (0.69) 1821707.000 (0.09) 1830010.400 (0.55) 1651016.200 (-9.28) parsec3/bodytrack 1416437.600 1430462.200 (0.99) 1420736.400 (0.30) 1428355.600 (0.84) 1430327.000 (0.98) parsec3/canneal 1040414.400 1050736.800 (0.99) 1041515.600 (0.11) 1048562.200 (0.78) 1049049.400 (0.83) parsec3/dedup 2414431.800 2454260.400 (1.65) 2423175.400 (0.36) 2396560.200 (-0.74) 2379898.200 (-1.43) parsec3/facesim 540432.200 551410.200 (2.03) 545978.200 (1.03) 558558.400 (3.35) 483755.400 (-10.49) parsec3/ferret 318728.600 333971.800 (4.78) 322158.200 (1.08) 332889.200 (4.44) 327896.400 (2.88) parsec3/fluidanimate 576917.800 585126.600 (1.42) 575123.200 (-0.31) 585429.200 (1.48) 484810.600 (-15.97) parsec3/freqmine 987882.200 997030.600 (0.93) 990429.200 (0.26) 998484.000 (1.07) 770740.200 (-21.98) parsec3/raytrace 1747059.800 1752904.000 (0.33) 1738853.600 (-0.47) 1753948.600 (0.39) 1578118.000 (-9.67) parsec3/streamcluster 121857.600 133934.400 (9.91) 121777.800 (-0.07) 133145.800 (9.26) 131512.800 (7.92) parsec3/swaptions 14123.000 29254.400 (107.14) 14017.200 (-0.75) 26470.600 (87.43) 28429.800 (101.30) parsec3/vips 2957631.800 2972884.400 (0.52) 2938855.400 (-0.63) 2960746.000 (0.11) 2946850.800 (-0.36) parsec3/x264 3184777.200 3214527.400 (0.93) 3177061.000 (-0.24) 3192446.600 (0.24) 3185851.800 (0.03) splash2x/barnes 1209737.200 1214763.200 (0.42) 1242138.400 (2.68) 1215857.600 (0.51) 994280.800 (-17.81) splash2x/fft 9362799.400 9178844.600 (-1.96) 9264052.600 (-1.05) 9164996.600 (-2.11) 9452048.200 (0.95) splash2x/lu_cb 515716.000 524071.600 (1.62) 521226.200 (1.07) 524261.400 (1.66) 372910.200 (-27.69) splash2x/lu_ncb 512898.200 523057.600 (1.98) 520630.800 (1.51) 523779.000 (2.12) 446282.400 (-12.99) splash2x/ocean_cp 3346038.000 3288703.600 (-1.71) 3386906.600 (1.22) 3330937.200 (-0.45) 3266442.400 (-2.38) splash2x/ocean_ncp 3886945.600 3871894.000 (-0.39) 7066192.000 (81.79) 5065229.800 (30.31) 3652078.200 (-6.04) splash2x/radiosity 1467107.200 1468850.800 (0.12) 1481292.600 (0.97) 1470335.800 (0.22) 530923.400 (-63.81) splash2x/radix 1708330.800 1699792.200 (-0.50) 1352708.600 (-20.82) 1601339.200 (-6.26) 2043947.800 (19.65) splash2x/raytrace 44817.200 59047.800 (31.75) 52010.200 (16.05) 60407.200 (34.79) 53916.400 (20.30) splash2x/volrend 151534.200 167791.400 (10.73) 151759.000 (0.15) 165012.400 (8.89) 160864.600 (6.16) splash2x/water_nsquared 46549.400 61846.800 (32.86) 51741.200 (11.15) 59214.400 (27.21) 91869.400 (97.36) splash2x/water_spatial 669085.200 675929.200 (1.02) 665924.600 (-0.47) 676218.200 (1.07) 538430.200 (-19.53) total 40062200.000 40073700.000 (0.03) 42888000.000 (7.05) 41103100.000 (2.60) 38052200.000 (-5.02) DAMON Overheads ~~~~~~~~~~~~~~~ In total, DAMON recording feature incurs 0.70% runtime overhead and 0.03% memory space overhead. For convenience test run of 'rec', I use a Python wrapper. The wrapper constantly consumes about 10-15MB of memory. This becomes high memory overhead if the target workload has small memory footprint. Nonetheless, the overheads are not from DAMON, but from the wrapper, and thus should be ignored. This fake memory overhead continues in 'ethp' and 'prcl', as those configurations are also using the Python wrapper. Efficient THP ~~~~~~~~~~~~~ THP 'always' enabled policy achieves 4.76% speedup but incurs 7.05% memory overhead. It achieves 37.37% speedup in best case, but 81.79% memory overhead in worst case. Interestingly, both the best and worst case are with 'splash2x/ocean_ncp'). The 2-lines implementation of data access monitoring based THP version ('ethp') shows 2.34% speedup and 2.60% memory overhead. In other words, 'ethp' removes 63.12% of THP memory waste while preserving 49.15% of THP speedup in total. In case of the 'splash2x/ocean_ncp', 'ethp' removes 62.94% of THP memory waste while preserving 57.37% of THP speedup. Proactive Reclamation ~~~~~~~~~~~~~~~~~~~~ As same to the original work, I use 'zram' swap device for this configuration. In total, our 1 line implementation of Proactive Reclamation, 'prcl', incurred 8.41% runtime overhead in total while achieving 5.83% system memory usage reduction. Nonetheless, as the memory usage is calculated with 'MemFree' in '/proc/meminfo', it contains the SwapCached pages. As the swapcached pages can be easily evicted, I also measured the residential set size of the workloads: rss.avg orig rec (overhead) thp (overhead) ethp (overhead) prcl (overhead) parsec3/blackscholes 591452.000 591466.400 (0.00) 593145.200 (0.29) 590609.400 (-0.14) 324379.000 (-45.16) parsec3/bodytrack 32458.600 32352.200 (-0.33) 32218.400 (-0.74) 32376.400 (-0.25) 27186.000 (-16.24) parsec3/canneal 841311.600 839888.400 (-0.17) 837008.400 (-0.51) 837811.000 (-0.42) 823276.200 (-2.14) parsec3/dedup 1219096.600 1228038.800 (0.73) 1235610.800 (1.35) 1214267.000 (-0.40) 992031.000 (-18.63) parsec3/facesim 311322.200 311574.400 (0.08) 316277.000 (1.59) 312593.800 (0.41) 188789.400 (-39.36) parsec3/ferret 99536.600 99556.800 (0.02) 102366.000 (2.84) 99799.000 (0.26) 88392.000 (-11.20) parsec3/fluidanimate 531893.600 531856.000 (-0.01) 532143.400 (0.05) 532190.200 (0.06) 421798.800 (-20.70) parsec3/freqmine 553533.200 552730.400 (-0.15) 555642.600 (0.38) 553895.400 (0.07) 78335.000 (-85.85) parsec3/raytrace 894094.200 894849.000 (0.08) 889964.000 (-0.46) 892865.000 (-0.14) 332911.800 (-62.77) parsec3/streamcluster 110938.000 110968.200 (0.03) 111673.400 (0.66) 111312.200 (0.34) 109911.200 (-0.93) parsec3/swaptions 5630.000 5634.800 (0.09) 5656.600 (0.47) 5692.000 (1.10) 4028.400 (-28.45) parsec3/vips 32107.000 32045.200 (-0.19) 32207.800 (0.31) 32293.800 (0.58) 29093.600 (-9.39) parsec3/x264 81926.000 82143.000 (0.26) 83258.400 (1.63) 82570.600 (0.79) 80651.800 (-1.56) splash2x/barnes 1215468.800 1217889.800 (0.20) 1222006.800 (0.54) 1217425.600 (0.16) 752405.200 (-38.10) splash2x/fft 9584734.800 9568872.800 (-0.17) 9660321.400 (0.79) 9646012.000 (0.64) 8367492.800 (-12.70) splash2x/lu_cb 510555.400 510807.400 (0.05) 514448.600 (0.76) 509281.800 (-0.25) 349272.200 (-31.59) splash2x/lu_ncb 510310.000 508915.600 (-0.27) 513886.000 (0.70) 510288.400 (-0.00) 431521.800 (-15.44) splash2x/ocean_cp 3408724.400 3408424.600 (-0.01) 3446054.400 (1.10) 3419536.200 (0.32) 3173818.600 (-6.89) splash2x/ocean_ncp 3923539.600 3922605.400 (-0.02) 7175526.600 (82.88) 5152558.800 (31.32) 3475756.000 (-11.41) splash2x/radiosity 1476050.000 1475470.400 (-0.04) 1485747.000 (0.66) 1476232.600 (0.01) 269512.200 (-81.74) splash2x/radix 1756385.400 1752676.000 (-0.21) 1431621.600 (-18.49) 1711460.800 (-2.56) 1923448.200 (9.51) splash2x/raytrace 23286.400 23311.200 (0.11) 28440.800 (22.13) 26977.200 (15.85) 15685.200 (-32.64) splash2x/volrend 44089.400 44125.600 (0.08) 44436.600 (0.79) 44250.400 (0.37) 27616.800 (-37.36) splash2x/water_nsquared 29437.600 29403.200 (-0.12) 29817.400 (1.29) 30040.000 (2.05) 25369.600 (-13.82) splash2x/water_spatial 656264.400 656566.400 (0.05) 656016.400 (-0.04) 656420.200 (0.02) 474480.400 (-27.70) total 28444100.000 28432200.000 (-0.04) 31535300.000 (10.87) 29698900.000 (4.41) 22787200.000 (-19.89) In total, 19.89% of residential sets were reduced. With parsec3/freqmine, 'prcl' reduced 85.85% of residential sets and 21.98% of system memory usage while incurring only 2.42% runtime overhead. Appendix F: Prototype Evaluations ================================= A prototype of DAMON has evaluated on an Intel Xeon E7-8837 machine using 20 benchmarks that picked from SPEC CPU 2006, NAS, Tensorflow Benchmark, SPLASH-2X, and PARSEC 3 benchmark suite. Nonethless, this section provides only summary of the results. For more detail, please refer to the slides used for the introduction of DAMON at the Linux Plumbers Conference 2019[1] or the MIDDLEWARE'19 industrial track paper[2]. [1] SeongJae Park, Tracing Data Access Pattern with Bounded Overhead and Best-effort Accuracy. In The Linux Kernel Summit, September 2019. https://linuxplumbersconf.org/event/4/contributions/548/ [2] SeongJae Park, Yunjae Lee, Heon Y. Yeom, Profiling Dynamic Data Access Patterns with Controlled Overhead and Quality. In 20th ACM/IFIP International Middleware Conference Industry, December 2019. https://dl.acm.org/doi/10.1145/3366626.3368125 Quality ------- We first traced and visualized the data access pattern of each workload. We were able to confirm that the visualized results are reasonably accurate by manually comparing those with the source code of the workloads. To see the usefulness of the monitoring, we optimized 9 memory intensive workloads among them for memory pressure situations using the DAMON outputs. In detail, we identified frequently accessed memory regions in each workload based on the DAMON results and protected them with ``mlock()`` system calls by manually modifying the source code. The optimized versions consistently show speedup (2.55x in best case, 1.65x in average) under artificial memory pressures. We use cgroups for the pressure. Overhead -------- We also measured the overhead of DAMON. The upperbound we set was kept as expected. Besides, it was much lower (0.6 percent of the bound in best case, 13.288 percent of the bound in average). This reduction of the overhead is mainly resulted from its core mechanism called adaptive regions adjustment. Refer to 'Appendix D' for more detail about the mechanism. We also compared the overhead of DAMON with that of a straightforward periodic PTE Accessed bit checking based monitoring. DAMON's overhead was smaller than it by 94,242.42x in best case, 3,159.61x in average. The latest version of DAMON running with its default configuration consumes only up to 1% of CPU time when applied to realistic workloads in PARSEC3 and SPLASH-2X and makes no visible slowdown to the target processes.