Units
Translation components API.
See the Weblate's Web API documentation for detailed description of the API.
GET /api/translations/documentation/articlesvm-design_index/en/units/?format=api&page=2
{ "count": 94, "next": null, "previous": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/units/?format=api", "results": [ { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "It is evident from that list that there was plenty of room for improvement. For FreeBSD 4.X, I completely rewrote the swap subsystem:" ], "previous_source": "", "target": [ "It is evident from that list that there was plenty of room for improvement. For FreeBSD 4.X, I completely rewrote the swap subsystem:" ], "id_hash": 6790780388169697555, "content_hash": 6790780388169697555, "location": "documentation/content/en/articles/vm-design/_index.adoc:198", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 39, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 23, "source_unit": "https://translate-dev.freebsd.org/api/units/615399/?format=api", "priority": 100, "id": 615399, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=de3db2c8b08b4913", "url": "https://translate-dev.freebsd.org/api/units/615399/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:34.730084Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "Since the VM system uses all available memory for disk caching, there are usually very few truly-free pages. The VM system depends on being able to properly choose pages which are not in use to reuse for new allocations. Selecting the optimal pages to free is possibly the single-most important function any VM system can perform because if it makes a poor selection, the VM system may be forced to unnecessarily retrieve pages from disk, seriously degrading system performance." ], "previous_source": "", "target": [ "Since the VM system uses all available memory for disk caching, there are usually very few truly-free pages. The VM system depends on being able to properly choose pages which are not in use to reuse for new allocations. Selecting the optimal pages to free is possibly the single-most important function any VM system can perform because if it makes a poor selection, the VM system may be forced to unnecessarily retrieve pages from disk, seriously degrading system performance." ], "id_hash": 204140903070093057, "content_hash": 204140903070093057, "location": "documentation/content/en/articles/vm-design/_index.adoc:212", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 46, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 79, "source_unit": "https://translate-dev.freebsd.org/api/units/615401/?format=api", "priority": 100, "id": 615401, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=82d54111db9ecf01", "url": "https://translate-dev.freebsd.org/api/units/615401/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:34.840933Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "The free page determination algorithm is built upon a history of the use of memory pages. To acquire this history, the system takes advantage of a page-used bit feature that most hardware page tables have." ], "previous_source": "", "target": [ "The free page determination algorithm is built upon a history of the use of memory pages. To acquire this history, the system takes advantage of a page-used bit feature that most hardware page tables have." ], "id_hash": -6858376981814393393, "content_hash": -6858376981814393393, "location": "documentation/content/en/articles/vm-design/_index.adoc:218", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 48, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 35, "source_unit": "https://translate-dev.freebsd.org/api/units/615405/?format=api", "priority": 100, "id": 615405, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=20d22678e73cc5cf", "url": "https://translate-dev.freebsd.org/api/units/615405/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:34.888466Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "In any case, the page-used bit is cleared and at some later point the VM system comes across the page again and sees that the page-used bit has been set. This indicates that the page is still being actively used. If the bit is still clear it is an indication that the page is not being actively used. By testing this bit periodically, a use history (in the form of a counter) for the physical page is developed. When the VM system later needs to free up some pages, checking this history becomes the cornerstone of determining the best candidate page to reuse." ], "previous_source": "", "target": [ "In any case, the page-used bit is cleared and at some later point the VM system comes across the page again and sees that the page-used bit has been set. This indicates that the page is still being actively used. If the bit is still clear it is an indication that the page is not being actively used. By testing this bit periodically, a use history (in the form of a counter) for the physical page is developed. When the VM system later needs to free up some pages, checking this history becomes the cornerstone of determining the best candidate page to reuse." ], "id_hash": 4733636345993738193, "content_hash": 4733636345993738193, "location": "documentation/content/en/articles/vm-design/_index.adoc:224", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 49, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 103, "source_unit": "https://translate-dev.freebsd.org/api/units/615407/?format=api", "priority": 100, "id": 615407, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=c1b1412d42d5b7d1", "url": "https://translate-dev.freebsd.org/api/units/615407/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:34.910039Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "For those platforms that do not have this feature, the system actually emulates a page-used bit. It unmaps or protects a page, forcing a page fault if the page is accessed again. When the page fault is taken, the system simply marks the page as having been used and unprotects the page so that it may be used. While taking such page faults just to determine if a page is being used appears to be an expensive proposition, it is much less expensive than reusing the page for some other purpose only to find that a process needs it back and then have to go to disk." ], "previous_source": "", "target": [ "For those platforms that do not have this feature, the system actually emulates a page-used bit. It unmaps or protects a page, forcing a page fault if the page is accessed again. When the page fault is taken, the system simply marks the page as having been used and unprotects the page so that it may be used. While taking such page faults just to determine if a page is being used appears to be an expensive proposition, it is much less expensive than reusing the page for some other purpose only to find that a process needs it back and then have to go to disk." ], "id_hash": -1836524288596584530, "content_hash": -1836524288596584530, "location": "documentation/content/en/articles/vm-design/_index.adoc:229", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 50, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 107, "source_unit": "https://translate-dev.freebsd.org/api/units/615409/?format=api", "priority": 100, "id": 615409, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=66835ae3ccd2d7ae", "url": "https://translate-dev.freebsd.org/api/units/615409/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:34.934599Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "FreeBSD makes use of several page queues to further refine the selection of pages to reuse as well as to determine when dirty pages must be flushed to their backing store. Since page tables are dynamic entities under FreeBSD, it costs virtually nothing to unmap a page from the address space of any processes using it. When a page candidate has been chosen based on the page-use counter, this is precisely what is done. The system must make a distinction between clean pages which can theoretically be freed up at any time, and dirty pages which must first be written to their backing store before being reusable. When a page candidate has been found it is moved to the inactive queue if it is dirty, or the cache queue if it is clean. A separate algorithm based on the dirty-to-clean page ratio determines when dirty pages in the inactive queue must be flushed to disk. Once this is accomplished, the flushed pages are moved from the inactive queue to the cache queue. At this point, pages in the cache queue can still be reactivated by a VM fault at relatively low cost. However, pages in the cache queue are considered to be \"immediately freeable\" and will be reused in an LRU (least-recently used) fashion when the system needs to allocate new memory." ], "previous_source": "", "target": [ "FreeBSD makes use of several page queues to further refine the selection of pages to reuse as well as to determine when dirty pages must be flushed to their backing store. Since page tables are dynamic entities under FreeBSD, it costs virtually nothing to unmap a page from the address space of any processes using it. When a page candidate has been chosen based on the page-use counter, this is precisely what is done. The system must make a distinction between clean pages which can theoretically be freed up at any time, and dirty pages which must first be written to their backing store before being reusable. When a page candidate has been found it is moved to the inactive queue if it is dirty, or the cache queue if it is clean. A separate algorithm based on the dirty-to-clean page ratio determines when dirty pages in the inactive queue must be flushed to disk. Once this is accomplished, the flushed pages are moved from the inactive queue to the cache queue. At this point, pages in the cache queue can still be reactivated by a VM fault at relatively low cost. However, pages in the cache queue are considered to be \"immediately freeable\" and will be reused in an LRU (least-recently used) fashion when the system needs to allocate new memory." ], "id_hash": 2673623526903959567, "content_hash": 2673623526903959567, "location": "documentation/content/en/articles/vm-design/_index.adoc:239", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 51, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 222, "source_unit": "https://translate-dev.freebsd.org/api/units/615411/?format=api", "priority": 100, "id": 615411, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=a51a9e6ea83cd00f", "url": "https://translate-dev.freebsd.org/api/units/615411/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:34.977307Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "It is important to note that the FreeBSD VM system attempts to separate clean and dirty pages for the express reason of avoiding unnecessary flushes of dirty pages (which eats I/O bandwidth), nor does it move pages between the various page queues gratuitously when the memory subsystem is not being stressed. This is why you will see some systems with very low cache queue counts and high active queue counts when doing a `systat -vm` command. As the VM system becomes more stressed, it makes a greater effort to maintain the various page queues at the levels determined to be the most effective." ], "previous_source": "", "target": [ "It is important to note that the FreeBSD VM system attempts to separate clean and dirty pages for the express reason of avoiding unnecessary flushes of dirty pages (which eats I/O bandwidth), nor does it move pages between the various page queues gratuitously when the memory subsystem is not being stressed. This is why you will see some systems with very low cache queue counts and high active queue counts when doing a `systat -vm` command. As the VM system becomes more stressed, it makes a greater effort to maintain the various page queues at the levels determined to be the most effective." ], "id_hash": 7639403564355480804, "content_hash": 7639403564355480804, "location": "documentation/content/en/articles/vm-design/_index.adoc:243", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 52, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 103, "source_unit": "https://translate-dev.freebsd.org/api/units/615413/?format=api", "priority": 100, "id": 615413, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=ea049d110542ece4", "url": "https://translate-dev.freebsd.org/api/units/615413/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:35.002998Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "Taking a VM fault is not expensive if the underlying page is already in core and can simply be mapped into the process, but it can become expensive if you take a whole lot of them on a regular basis. A good example of this is running a program such as man:ls[1] or man:ps[1] over and over again. If the program binary is mapped into memory but not mapped into the page table, then all the pages that will be accessed by the program will have to be faulted in every time the program is run. This is unnecessary when the pages in question are already in the VM Cache, so FreeBSD will attempt to pre-populate a process's page tables with those pages that are already in the VM Cache. One thing that FreeBSD does not yet do is pre-copy-on-write certain pages on exec. For example, if you run the man:ls[1] program while running `vmstat 1` you will notice that it always takes a certain number of page faults, even when you run it over and over again. These are zero-fill faults, not program code faults (which were pre-faulted in already). Pre-copying pages on exec or fork is an area that could use more study." ], "previous_source": "", "target": [ "Taking a VM fault is not expensive if the underlying page is already in core and can simply be mapped into the process, but it can become expensive if you take a whole lot of them on a regular basis. A good example of this is running a program such as man:ls[1] or man:ps[1] over and over again. If the program binary is mapped into memory but not mapped into the page table, then all the pages that will be accessed by the program will have to be faulted in every time the program is run. This is unnecessary when the pages in question are already in the VM Cache, so FreeBSD will attempt to pre-populate a process's page tables with those pages that are already in the VM Cache. One thing that FreeBSD does not yet do is pre-copy-on-write certain pages on exec. For example, if you run the man:ls[1] program while running `vmstat 1` you will notice that it always takes a certain number of page faults, even when you run it over and over again. These are zero-fill faults, not program code faults (which were pre-faulted in already). Pre-copying pages on exec or fork is an area that could use more study." ], "id_hash": -8125128984177974447, "content_hash": -8125128984177974447, "location": "documentation/content/en/articles/vm-design/_index.adoc:259", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 55, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 205, "source_unit": "https://translate-dev.freebsd.org/api/units/615417/?format=api", "priority": 100, "id": 615417, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=0f3dbe3cd7d76351", "url": "https://translate-dev.freebsd.org/api/units/615417/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:35.175174Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "The page table optimizations make up the most contentious part of the FreeBSD VM design and they have shown some strain with the advent of serious use of `mmap()`. I think this is actually a feature of most BSDs though I am not sure when it was first introduced. There are two major optimizations. The first is that hardware page tables do not contain persistent state but instead can be thrown away at any time with only a minor amount of management overhead. The second is that every active page table entry in the system has a governing `pv_entry` structure which is tied into the `vm_page` structure. FreeBSD can simply iterate through those mappings that are known to exist while Linux must check all page tables that _might_ contain a specific mapping to see if it does, which can achieve O(n^2) overhead in certain situations. It is because of this that FreeBSD tends to make better choices on which pages to reuse or swap when memory is stressed, giving it better performance under load. However, FreeBSD requires kernel tuning to accommodate large-shared-address-space situations such as those that can occur in a news system because it may run out of `pv_entry` structures." ], "previous_source": "", "target": [ "The page table optimizations make up the most contentious part of the FreeBSD VM design and they have shown some strain with the advent of serious use of `mmap()`. I think this is actually a feature of most BSDs though I am not sure when it was first introduced. There are two major optimizations. The first is that hardware page tables do not contain persistent state but instead can be thrown away at any time with only a minor amount of management overhead. The second is that every active page table entry in the system has a governing `pv_entry` structure which is tied into the `vm_page` structure. FreeBSD can simply iterate through those mappings that are known to exist while Linux must check all page tables that _might_ contain a specific mapping to see if it does, which can achieve O(n^2) overhead in certain situations. It is because of this that FreeBSD tends to make better choices on which pages to reuse or swap when memory is stressed, giving it better performance under load. However, FreeBSD requires kernel tuning to accommodate large-shared-address-space situations such as those that can occur in a news system because it may run out of `pv_entry` structures." ], "id_hash": -6307955811766454775, "content_hash": -6307955811766454775, "location": "documentation/content/en/articles/vm-design/_index.adoc:280", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 58, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 201, "source_unit": "https://translate-dev.freebsd.org/api/units/615421/?format=api", "priority": 100, "id": 615421, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=2875a39c34ccb609", "url": "https://translate-dev.freebsd.org/api/units/615421/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:35.283336Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "Both Linux and FreeBSD need work in this area. FreeBSD is trying to maximize the advantage of a potentially sparse active-mapping model (not all processes need to map all pages of a shared library, for example), whereas Linux is trying to simplify its algorithms. FreeBSD generally has the performance advantage here at the cost of wasting a little extra memory, but FreeBSD breaks down in the case where a large file is massively shared across hundreds of processes. Linux, on the other hand, breaks down in the case where many processes are sparsely-mapping the same shared library and also runs non-optimally when trying to determine whether a page can be reused or not." ], "previous_source": "", "target": [ "Both Linux and FreeBSD need work in this area. FreeBSD is trying to maximize the advantage of a potentially sparse active-mapping model (not all processes need to map all pages of a shared library, for example), whereas Linux is trying to simplify its algorithms. FreeBSD generally has the performance advantage here at the cost of wasting a little extra memory, but FreeBSD breaks down in the case where a large file is massively shared across hundreds of processes. Linux, on the other hand, breaks down in the case where many processes are sparsely-mapping the same shared library and also runs non-optimally when trying to determine whether a page can be reused or not." ], "id_hash": 60250029008486510, "content_hash": 60250029008486510, "location": "documentation/content/en/articles/vm-design/_index.adoc:285", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 59, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 113, "source_unit": "https://translate-dev.freebsd.org/api/units/615423/?format=api", "priority": 100, "id": 615423, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=80d60d1508c2006e", "url": "https://translate-dev.freebsd.org/api/units/615423/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:35.588403Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "We will end with the page coloring optimizations. Page coloring is a performance optimization designed to ensure that accesses to contiguous pages in virtual memory make the best use of the processor cache. In ancient times (i.e. 10+ years ago) processor caches tended to map virtual memory rather than physical memory. This led to a huge number of problems including having to clear the cache on every context switch in some cases, and problems with data aliasing in the cache. Modern processor caches map physical memory precisely to solve those problems. This means that two side-by-side pages in a processes address space may not correspond to two side-by-side pages in the cache. In fact, if you are not careful side-by-side pages in virtual memory could wind up using the same page in the processor cache-leading to cacheable data being thrown away prematurely and reducing CPU performance. This is true even with multi-way set-associative caches (though the effect is mitigated somewhat)." ], "previous_source": "", "target": [ "We will end with the page coloring optimizations. Page coloring is a performance optimization designed to ensure that accesses to contiguous pages in virtual memory make the best use of the processor cache. In ancient times (i.e. 10+ years ago) processor caches tended to map virtual memory rather than physical memory. This led to a huge number of problems including having to clear the cache on every context switch in some cases, and problems with data aliasing in the cache. Modern processor caches map physical memory precisely to solve those problems. This means that two side-by-side pages in a processes address space may not correspond to two side-by-side pages in the cache. In fact, if you are not careful side-by-side pages in virtual memory could wind up using the same page in the processor cache-leading to cacheable data being thrown away prematurely and reducing CPU performance. This is true even with multi-way set-associative caches (though the effect is mitigated somewhat)." ], "id_hash": 5700219517434259076, "content_hash": 5700219517434259076, "location": "documentation/content/en/articles/vm-design/_index.adoc:297", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 61, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 160, "source_unit": "https://translate-dev.freebsd.org/api/units/615425/?format=api", "priority": 100, "id": 615425, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=cf1b3f73366ae684", "url": "https://translate-dev.freebsd.org/api/units/615425/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:35.770698Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "FreeBSD's memory allocation code implements page coloring optimizations, which means that the memory allocation code will attempt to locate free pages that are contiguous from the point of view of the cache. For example, if page 16 of physical memory is assigned to page 0 of a process's virtual memory and the cache can hold 4 pages, the page coloring code will not assign page 20 of physical memory to page 1 of a process's virtual memory. It would, instead, assign page 21 of physical memory. The page coloring code attempts to avoid assigning page 20 because this maps over the same cache memory as page 16 and would result in non-optimal caching. This code adds a significant amount of complexity to the VM memory allocation subsystem as you can well imagine, but the result is well worth the effort. Page Coloring makes VM memory as deterministic as physical memory in regards to cache performance." ], "previous_source": "", "target": [ "FreeBSD's memory allocation code implements page coloring optimizations, which means that the memory allocation code will attempt to locate free pages that are contiguous from the point of view of the cache. For example, if page 16 of physical memory is assigned to page 0 of a process's virtual memory and the cache can hold 4 pages, the page coloring code will not assign page 20 of physical memory to page 1 of a process's virtual memory. It would, instead, assign page 21 of physical memory. The page coloring code attempts to avoid assigning page 20 because this maps over the same cache memory as page 16 and would result in non-optimal caching. This code adds a significant amount of complexity to the VM memory allocation subsystem as you can well imagine, but the result is well worth the effort. Page Coloring makes VM memory as deterministic as physical memory in regards to cache performance." ], "id_hash": -4289767702818845546, "content_hash": -4289767702818845546, "location": "documentation/content/en/articles/vm-design/_index.adoc:304", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 62, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 155, "source_unit": "https://translate-dev.freebsd.org/api/units/615427/?format=api", "priority": 100, "id": 615427, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=4477af00a08a0c96", "url": "https://translate-dev.freebsd.org/api/units/615427/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:35.859166Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "Virtual memory in modern operating systems must address a number of different issues efficiently and for many different usage patterns. The modular and algorithmic approach that BSD has historically taken allows us to study and understand the current implementation as well as relatively cleanly replace large sections of the code. There have been a number of improvements to the FreeBSD VM system in the last several years, and work is ongoing." ], "previous_source": "", "target": [ "Virtual memory in modern operating systems must address a number of different issues efficiently and for many different usage patterns. The modular and algorithmic approach that BSD has historically taken allows us to study and understand the current implementation as well as relatively cleanly replace large sections of the code. There have been a number of improvements to the FreeBSD VM system in the last several years, and work is ongoing." ], "id_hash": -2528742712089608106, "content_hash": -2528742712089608106, "location": "documentation/content/en/articles/vm-design/_index.adoc:311", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 64, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 71, "source_unit": "https://translate-dev.freebsd.org/api/units/615429/?format=api", "priority": 100, "id": 615429, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=5ce819e6b58a8056", "url": "https://translate-dev.freebsd.org/api/units/615429/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:36.659592Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "FreeBSD uses a fixed swap interleave which defaults to 4. This means that FreeBSD reserves space for four swap areas even if you only have one, two, or three. Since swap is interleaved the linear address space representing the \"four swap areas\" will be fragmented if you do not actually have four swap areas. For example, if you have two swap areas A and B FreeBSD's address space representation for that swap area will be interleaved in blocks of 16 pages:" ], "previous_source": "", "target": [ "FreeBSD uses a fixed swap interleave which defaults to 4. This means that FreeBSD reserves space for four swap areas even if you only have one, two, or three. Since swap is interleaved the linear address space representing the \"four swap areas\" will be fragmented if you do not actually have four swap areas. For example, if you have two swap areas A and B FreeBSD's address space representation for that swap area will be interleaved in blocks of 16 pages:" ], "id_hash": 2352136694481216650, "content_hash": 2352136694481216650, "location": "documentation/content/en/articles/vm-design/_index.adoc:321", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 67, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 81, "source_unit": "https://translate-dev.freebsd.org/api/units/615431/?format=api", "priority": 100, "id": 615431, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=a0a477dc89b6e48a", "url": "https://translate-dev.freebsd.org/api/units/615431/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:36.962671Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "FreeBSD 3.X uses a \"sequential list of free regions\" approach to accounting for the free swap areas. The idea is that large blocks of free linear space can be represented with a single list node ([.filename]#kern/subr_rlist.c#). But due to the fragmentation the sequential list winds up being insanely fragmented. In the above example, completely unused swap will have A and B shown as \"free\" and C and D shown as \"all allocated\". Each A-B sequence requires a list node to account for because C and D are holes, so the list node cannot be combined with the next A-B sequence." ], "previous_source": "", "target": [ "FreeBSD 3.X uses a \"sequential list of free regions\" approach to accounting for the free swap areas. The idea is that large blocks of free linear space can be represented with a single list node ([.filename]#kern/subr_rlist.c#). But due to the fragmentation the sequential list winds up being insanely fragmented. In the above example, completely unused swap will have A and B shown as \"free\" and C and D shown as \"all allocated\". Each A-B sequence requires a list node to account for because C and D are holes, so the list node cannot be combined with the next A-B sequence." ], "id_hash": 5525637408233093798, "content_hash": 5525637408233093798, "location": "documentation/content/en/articles/vm-design/_index.adoc:331", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 69, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 100, "source_unit": "https://translate-dev.freebsd.org/api/units/615433/?format=api", "priority": 100, "id": 615433, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=ccaf01f25b57faa6", "url": "https://translate-dev.freebsd.org/api/units/615433/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:37.154183Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "The fragmentation causes other problems. Being a linear list under 3.X, and having such a huge amount of inherent fragmentation, allocating and freeing swap winds up being an O(N) algorithm instead of an O(1) algorithm. Combined with other factors (heavy swapping) and you start getting into O(N^2) and O(N^3) levels of overhead, which is bad. The 3.X system may also need to allocate KVM during a swap operation to create a new list node which can lead to a deadlock if the system is trying to pageout pages in a low-memory situation." ], "previous_source": "", "target": [ "The fragmentation causes other problems. Being a linear list under 3.X, and having such a huge amount of inherent fragmentation, allocating and freeing swap winds up being an O(N) algorithm instead of an O(1) algorithm. Combined with other factors (heavy swapping) and you start getting into O(N^2) and O(N^3) levels of overhead, which is bad. The 3.X system may also need to allocate KVM during a swap operation to create a new list node which can lead to a deadlock if the system is trying to pageout pages in a low-memory situation." ], "id_hash": -7933960118181456140, "content_hash": -7933960118181456140, "location": "documentation/content/en/articles/vm-design/_index.adoc:338", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 71, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 92, "source_unit": "https://translate-dev.freebsd.org/api/units/615435/?format=api", "priority": 100, "id": 615435, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=11e4e94ef5ad96f4", "url": "https://translate-dev.freebsd.org/api/units/615435/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:37.232765Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "Under 4.X we do not use a sequential list. Instead we use a radix tree and bitmaps of swap blocks rather than ranged list nodes. We take the hit of preallocating all the bitmaps required for the entire swap area up front but it winds up wasting less memory due to the use of a bitmap (one bit per block) instead of a linked list of nodes. The use of a radix tree instead of a sequential list gives us nearly O(1) performance no matter how fragmented the tree becomes." ], "previous_source": "", "target": [ "Under 4.X we do not use a sequential list. Instead we use a radix tree and bitmaps of swap blocks rather than ranged list nodes. We take the hit of preallocating all the bitmaps required for the entire swap area up front but it winds up wasting less memory due to the use of a bitmap (one bit per block) instead of a linked list of nodes. The use of a radix tree instead of a sequential list gives us nearly O(1) performance no matter how fragmented the tree becomes." ], "id_hash": 4155145828883344859, "content_hash": 4155145828883344859, "location": "documentation/content/en/articles/vm-design/_index.adoc:343", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 72, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 90, "source_unit": "https://translate-dev.freebsd.org/api/units/615437/?format=api", "priority": 100, "id": 615437, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=b9aa0b1dd089addb", "url": "https://translate-dev.freebsd.org/api/units/615437/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:37.353758Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "Yes, that is confusing. The relationship is \"goal\" verses \"reality\". Our goal is to separate the pages but the reality is that if we are not in a memory crunch, we do not really have to." ], "previous_source": "", "target": [ "Yes, that is confusing. The relationship is \"goal\" verses \"reality\". Our goal is to separate the pages but the reality is that if we are not in a memory crunch, we do not really have to." ], "id_hash": -460051008278140723, "content_hash": -460051008278140723, "location": "documentation/content/en/articles/vm-design/_index.adoc:349", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 74, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 36, "source_unit": "https://translate-dev.freebsd.org/api/units/615439/?format=api", "priority": 100, "id": 615439, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=799d920ba10568cd", "url": "https://translate-dev.freebsd.org/api/units/615439/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:37.507126Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "A COW fault can be either zero-fill or program-data. The mechanism is the same either way because the backing program-data is almost certainly already in the cache. I am indeed lumping the two together. FreeBSD does not pre-COW program data or zero-fill, but it _does_ pre-map pages that exist in its cache." ], "previous_source": "", "target": [ "A COW fault can be either zero-fill or program-data. The mechanism is the same either way because the backing program-data is almost certainly already in the cache. I am indeed lumping the two together. FreeBSD does not pre-COW program data or zero-fill, but it _does_ pre-map pages that exist in its cache." ], "id_hash": 4210968952427581281, "content_hash": 4210968952427581281, "location": "documentation/content/en/articles/vm-design/_index.adoc:358", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 77, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 52, "source_unit": "https://translate-dev.freebsd.org/api/units/615441/?format=api", "priority": 100, "id": 615441, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=ba705df399a50f61", "url": "https://translate-dev.freebsd.org/api/units/615441/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:37.772515Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "A `vm_page` represents an (object,index#) tuple. A `pv_entry` represents a hardware page table entry (pte). If you have five processes sharing the same physical page, and three of those processes's page tables actually map the page, that page will be represented by a single `vm_page` structure and three `pv_entry` structures." ], "previous_source": "", "target": [ "A `vm_page` represents an (object,index#) tuple. A `pv_entry` represents a hardware page table entry (pte). If you have five processes sharing the same physical page, and three of those processes's page tables actually map the page, that page will be represented by a single `vm_page` structure and three `pv_entry` structures." ], "id_hash": -6972689839131770923, "content_hash": -6972689839131770923, "location": "documentation/content/en/articles/vm-design/_index.adoc:363", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 79, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 50, "source_unit": "https://translate-dev.freebsd.org/api/units/615443/?format=api", "priority": 100, "id": 615443, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=1f3c0788c1933fd5", "url": "https://translate-dev.freebsd.org/api/units/615443/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:37.968483Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "But FreeBSD has a scaling problem that Linux does not in that there are a limited number of `pv_entry` structures and this causes problems when you have massive sharing of data. In this case you may run out of `pv_entry` structures even though there is plenty of free memory available. This can be fixed easily enough by bumping up the number of `pv_entry` structures in the kernel config, but we really need to find a better way to do it." ], "previous_source": "", "target": [ "But FreeBSD has a scaling problem that Linux does not in that there are a limited number of `pv_entry` structures and this causes problems when you have massive sharing of data. In this case you may run out of `pv_entry` structures even though there is plenty of free memory available. This can be fixed easily enough by bumping up the number of `pv_entry` structures in the kernel config, but we really need to find a better way to do it." ], "id_hash": 3727233459493505828, "content_hash": 3727233459493505828, "location": "documentation/content/en/articles/vm-design/_index.adoc:379", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 83, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 80, "source_unit": "https://translate-dev.freebsd.org/api/units/615449/?format=api", "priority": 100, "id": 615449, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=b3b9cb156c0f8b24", "url": "https://translate-dev.freebsd.org/api/units/615449/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:38.225740Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "In regards to the memory overhead of a page table verses the `pv_entry` scheme: Linux uses \"permanent\" page tables that are not throw away, but does not need a `pv_entry` for each potentially mapped pte. FreeBSD uses \"throw away\" page tables but adds in a `pv_entry` structure for each actually-mapped pte. I think memory utilization winds up being about the same, giving FreeBSD an algorithmic advantage with its ability to throw away page tables at will with very low overhead." ], "previous_source": "", "target": [ "In regards to the memory overhead of a page table verses the `pv_entry` scheme: Linux uses \"permanent\" page tables that are not throw away, but does not need a `pv_entry` for each potentially mapped pte. FreeBSD uses \"throw away\" page tables but adds in a `pv_entry` structure for each actually-mapped pte. I think memory utilization winds up being about the same, giving FreeBSD an algorithmic advantage with its ability to throw away page tables at will with very low overhead." ], "id_hash": -1297362570668637025, "content_hash": -1297362570668637025, "location": "documentation/content/en/articles/vm-design/_index.adoc:383", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 84, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 80, "source_unit": "https://translate-dev.freebsd.org/api/units/615451/?format=api", "priority": 100, "id": 615451, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=6dfed79d9e68849f", "url": "https://translate-dev.freebsd.org/api/units/615451/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:38.333829Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "Do you know how an L1 hardware memory cache works? I will explain: Consider a machine with 16MB of main memory but only 128K of L1 cache. Generally the way this cache works is that each 128K block of main memory uses the _same_ 128K of cache. If you access offset 0 in main memory and then offset 128K in main memory you can wind up throwing away the cached data you read from offset 0!" ], "previous_source": "", "target": [ "Do you know how an L1 hardware memory cache works? I will explain: Consider a machine with 16MB of main memory but only 128K of L1 cache. Generally the way this cache works is that each 128K block of main memory uses the _same_ 128K of cache. If you access offset 0 in main memory and then offset 128K in main memory you can wind up throwing away the cached data you read from offset 0!" ], "id_hash": 9145909268159189302, "content_hash": 9145909268159189302, "location": "documentation/content/en/articles/vm-design/_index.adoc:389", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 86, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 76, "source_unit": "https://translate-dev.freebsd.org/api/units/615453/?format=api", "priority": 100, "id": 615453, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=feeccc05b494d136", "url": "https://translate-dev.freebsd.org/api/units/615453/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:38.482680Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "Now, I am simplifying things greatly. What I just described is what is called a \"direct mapped\" hardware memory cache. Most modern caches are what are called 2-way-set-associative or 4-way-set-associative caches. The set-associatively allows you to access up to N different memory regions that overlap the same cache memory without destroying the previously cached data. But only N." ], "previous_source": "", "target": [ "Now, I am simplifying things greatly. What I just described is what is called a \"direct mapped\" hardware memory cache. Most modern caches are what are called 2-way-set-associative or 4-way-set-associative caches. The set-associatively allows you to access up to N different memory regions that overlap the same cache memory without destroying the previously cached data. But only N." ], "id_hash": -7670384176450923367, "content_hash": -7670384176450923367, "location": "documentation/content/en/articles/vm-design/_index.adoc:395", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 87, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 58, "source_unit": "https://translate-dev.freebsd.org/api/units/615455/?format=api", "priority": 100, "id": 615455, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=158d523b1fdd9499", "url": "https://translate-dev.freebsd.org/api/units/615455/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:38.542141Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "So if I have a 4-way set associative cache I can access offset 0, offset 128K, 256K and offset 384K and still be able to access offset 0 again and have it come from the L1 cache. If I then access offset 512K, however, one of the four previously cached data objects will be thrown away by the cache." ], "previous_source": "", "target": [ "So if I have a 4-way set associative cache I can access offset 0, offset 128K, 256K and offset 384K and still be able to access offset 0 again and have it come from the L1 cache. If I then access offset 512K, however, one of the four previously cached data objects will be thrown away by the cache." ], "id_hash": 6568108778544654505, "content_hash": 6568108778544654505, "location": "documentation/content/en/articles/vm-design/_index.adoc:398", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 88, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 59, "source_unit": "https://translate-dev.freebsd.org/api/units/615457/?format=api", "priority": 100, "id": 615457, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=db269c227f4768a9", "url": "https://translate-dev.freebsd.org/api/units/615457/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:38.592148Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "It is extremely important... _extremely_ important for most of a processor's memory accesses to be able to come from the L1 cache, because the L1 cache operates at the processor frequency. The moment you have an L1 cache miss and have to go to the L2 cache or to main memory, the processor will stall and potentially sit twiddling its fingers for _hundreds_ of instructions worth of time waiting for a read from main memory to complete. Main memory (the dynamic ram you stuff into a computer) is __slow__, when compared to the speed of a modern processor core." ], "previous_source": "", "target": [ "It is extremely important... _extremely_ important for most of a processor's memory accesses to be able to come from the L1 cache, because the L1 cache operates at the processor frequency. The moment you have an L1 cache miss and have to go to the L2 cache or to main memory, the processor will stall and potentially sit twiddling its fingers for _hundreds_ of instructions worth of time waiting for a read from main memory to complete. Main memory (the dynamic ram you stuff into a computer) is __slow__, when compared to the speed of a modern processor core." ], "id_hash": 6588293474859458475, "content_hash": 6588293474859458475, "location": "documentation/content/en/articles/vm-design/_index.adoc:402", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 89, "has_suggestion": false, "has_comment": false, "has_failing_check": true, "num_words": 99, "source_unit": "https://translate-dev.freebsd.org/api/units/615459/?format=api", "priority": 100, "id": 615459, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=db6e520255059fab", "url": "https://translate-dev.freebsd.org/api/units/615459/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:38.681063Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "Ok, so now onto page coloring: All modern memory caches are what are known as _physical_ caches. They cache physical memory addresses, not virtual memory addresses. This allows the cache to be left alone across a process context switch, which is very important." ], "previous_source": "", "target": [ "Ok, so now onto page coloring: All modern memory caches are what are known as _physical_ caches. They cache physical memory addresses, not virtual memory addresses. This allows the cache to be left alone across a process context switch, which is very important." ], "id_hash": -3674260667789479117, "content_hash": -3674260667789479117, "location": "documentation/content/en/articles/vm-design/_index.adoc:406", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 90, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 43, "source_unit": "https://translate-dev.freebsd.org/api/units/615461/?format=api", "priority": 100, "id": 615461, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=4d026763c437f333", "url": "https://translate-dev.freebsd.org/api/units/615461/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:38.760339Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "But in the UNIX(R) world you are dealing with virtual address spaces, not physical address spaces. Any program you write will see the virtual address space given to it. The actual _physical_ pages underlying that virtual address space are not necessarily physically contiguous! In fact, you might have two pages that are side by side in a processes address space which wind up being at offset 0 and offset 128K in _physical_ memory." ], "previous_source": "", "target": [ "But in the UNIX(R) world you are dealing with virtual address spaces, not physical address spaces. Any program you write will see the virtual address space given to it. The actual _physical_ pages underlying that virtual address space are not necessarily physically contiguous! In fact, you might have two pages that are side by side in a processes address space which wind up being at offset 0 and offset 128K in _physical_ memory." ], "id_hash": -4117022287661575178, "content_hash": -4117022287661575178, "location": "documentation/content/en/articles/vm-design/_index.adoc:411", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 91, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 73, "source_unit": "https://translate-dev.freebsd.org/api/units/615463/?format=api", "priority": 100, "id": 615463, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=46dd660ad390d3f6", "url": "https://translate-dev.freebsd.org/api/units/615463/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:38.801936Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "A program normally assumes that two side-by-side pages will be optimally cached. That is, that you can access data objects in both pages without having them blow away each other's cache entry. But this is only true if the physical pages underlying the virtual address space are contiguous (insofar as the cache is concerned)." ], "previous_source": "", "target": [ "A program normally assumes that two side-by-side pages will be optimally cached. That is, that you can access data objects in both pages without having them blow away each other's cache entry. But this is only true if the physical pages underlying the virtual address space are contiguous (insofar as the cache is concerned)." ], "id_hash": 4011451862088314964, "content_hash": 4011451862088314964, "location": "documentation/content/en/articles/vm-design/_index.adoc:415", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 92, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 54, "source_unit": "https://translate-dev.freebsd.org/api/units/615465/?format=api", "priority": 100, "id": 615465, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=b7ab8a37095a7054", "url": "https://translate-dev.freebsd.org/api/units/615465/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:38.829541Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "This is what Page coloring does. Instead of assigning _random_ physical pages to virtual addresses, which may result in non-optimal cache performance, Page coloring assigns _reasonably-contiguous_ physical pages to virtual addresses. Thus programs can be written under the assumption that the characteristics of the underlying hardware cache are the same for their virtual address space as they would be if the program had been run directly in a physical address space." ], "previous_source": "", "target": [ "This is what Page coloring does. Instead of assigning _random_ physical pages to virtual addresses, which may result in non-optimal cache performance, Page coloring assigns _reasonably-contiguous_ physical pages to virtual addresses. Thus programs can be written under the assumption that the characteristics of the underlying hardware cache are the same for their virtual address space as they would be if the program had been run directly in a physical address space." ], "id_hash": -3032996817252281788, "content_hash": -3032996817252281788, "location": "documentation/content/en/articles/vm-design/_index.adoc:419", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 93, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 71, "source_unit": "https://translate-dev.freebsd.org/api/units/615467/?format=api", "priority": 100, "id": 615467, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=55e8a176d6c99244", "url": "https://translate-dev.freebsd.org/api/units/615467/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:38.858303Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "Note that I say \"reasonably\" contiguous rather than simply \"contiguous\". From the point of view of a 128K direct mapped cache, the physical address 0 is the same as the physical address 128K. So two side-by-side pages in your virtual address space may wind up being offset 128K and offset 132K in physical memory, but could also easily be offset 128K and offset 4K in physical memory and still retain the same cache performance characteristics. So page-coloring does _not_ have to assign truly contiguous pages of physical memory to contiguous pages of virtual memory, it just needs to make sure it assigns contiguous pages from the point of view of cache performance and operation." ], "previous_source": "", "target": [ "Note that I say \"reasonably\" contiguous rather than simply \"contiguous\". From the point of view of a 128K direct mapped cache, the physical address 0 is the same as the physical address 128K. So two side-by-side pages in your virtual address space may wind up being offset 128K and offset 132K in physical memory, but could also easily be offset 128K and offset 4K in physical memory and still retain the same cache performance characteristics. So page-coloring does _not_ have to assign truly contiguous pages of physical memory to contiguous pages of virtual memory, it just needs to make sure it assigns contiguous pages from the point of view of cache performance and operation." ], "id_hash": -7053199508336451325, "content_hash": -7053199508336451325, "location": "documentation/content/en/articles/vm-design/_index.adoc:423", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 94, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 114, "source_unit": "https://translate-dev.freebsd.org/api/units/615469/?format=api", "priority": 100, "id": 615469, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=1e1e006aa7a56103", "url": "https://translate-dev.freebsd.org/api/units/615469/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-03-20T20:43:38.923867Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "An easy to follow description of the design of the FreeBSD virtual memory system" ], "previous_source": "", "target": [ "An easy to follow description of the design of the FreeBSD virtual memory system" ], "id_hash": -5212445871253427557, "content_hash": -5212445871253427557, "location": "documentation/content/en/articles/vm-design/_index.adoc:1", "context": "", "note": "type: YAML Front Matter: description", "flags": "no-wrap", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 1, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 14, "source_unit": "https://translate-dev.freebsd.org/api/units/1181040/?format=api", "priority": 100, "id": 1181040, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=37a9ac18c8fcea9b", "url": "https://translate-dev.freebsd.org/api/units/1181040/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2021-06-03T22:46:25.115973Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "You will notice immediately that what was originally a simple file mapping has become much more complex. Data may be modified on a page-by-page basis whereas the file mapping encompasses many pages at once. The complexity further increases when a process forks. When a process forks, the result is two processes-each with their own private address spaces, including any modifications made by the original process prior to the call to `fork()`. It would be silly for the VM system to make a complete copy of the data at the time of the `fork()` because it is quite possible that at least one of the two processes will only need to read from that page from then on, allowing the original page to continue to be used. What was a private page is made copy-on-write again, since each process (parent and child) expects their own personal post-fork modifications to remain private to themselves and not affect the other." ], "previous_source": "", "target": [ "You will notice immediately that what was originally a simple file mapping has become much more complex. Data may be modified on a page-by-page basis whereas the file mapping encompasses many pages at once. The complexity further increases when a process forks. When a process forks, the result is two processes-each with their own private address spaces, including any modifications made by the original process prior to the call to `fork()`. It would be silly for the VM system to make a complete copy of the data at the time of the `fork()` because it is quite possible that at least one of the two processes will only need to read from that page from then on, allowing the original page to continue to be used. What was a private page is made copy-on-write again, since each process (parent and child) expects their own personal post-fork modifications to remain private to themselves and not affect the other." ], "id_hash": -8494902235450807947, "content_hash": -8494902235450807947, "location": "documentation/content/en/articles/vm-design/_index.adoc:112", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 16, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 157, "source_unit": "https://translate-dev.freebsd.org/api/units/1443560/?format=api", "priority": 100, "id": 1443560, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=0a1c0b6cbe486975", "url": "https://translate-dev.freebsd.org/api/units/1443560/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2022-07-08T06:09:38.660102Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "The most important aspect of performance design is what is known as \"Optimizing the Critical Path\". It is often the case that performance optimizations add a little bloat to the code to make the critical path perform better." ], "previous_source": "", "target": [ "The most important aspect of performance design is what is known as \"Optimizing the Critical Path\". It is often the case that performance optimizations add a little bloat to the code to make the critical path perform better." ], "id_hash": -4192121059148506934, "content_hash": -4192121059148506934, "location": "documentation/content/en/articles/vm-design/_index.adoc:82", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 10, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 38, "source_unit": "https://translate-dev.freebsd.org/api/units/1569353/?format=api", "priority": 100, "id": 1569353, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=45d2981b8604d8ca", "url": "https://translate-dev.freebsd.org/api/units/1569353/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2023-05-21T18:04:58.615720Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "Also, to keep track of swap space, a \"list of holes\" is kept in kernel memory, and this tends to get severely fragmented as well. Since the \"list of holes\" is a linear list, the swap allocation and freeing performance is a non-optimal O(n)-per-page." ], "previous_source": "", "target": [ "Also, to keep track of swap space, a \"list of holes\" is kept in kernel memory, and this tends to get severely fragmented as well. Since the \"list of holes\" is a linear list, the swap allocation and freeing performance is a non-optimal O(n)-per-page." ], "id_hash": -4072780474139688110, "content_hash": -4072780474139688110, "location": "documentation/content/en/articles/vm-design/_index.adoc:191", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 34, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 44, "source_unit": "https://translate-dev.freebsd.org/api/units/1569355/?format=api", "priority": 100, "id": 1569355, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=477a93bcbec1c352", "url": "https://translate-dev.freebsd.org/api/units/1569355/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2023-05-21T18:04:58.621668Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "The entire radix tree bitmap is also preallocated to avoid having to allocate kernel memory during critical low memory swapping operations. After all, the system tends to swap when it is low on memory so we should avoid allocating kernel memory at such times to avoid potential deadlocks." ], "previous_source": "", "target": [ "The entire radix tree bitmap is also preallocated to avoid having to allocate kernel memory during critical low memory swapping operations. After all, the system tends to swap when it is low on memory so we should avoid allocating kernel memory at such times to avoid potential deadlocks." ], "id_hash": 7931565045898444070, "content_hash": 7931565045898444070, "location": "documentation/content/en/articles/vm-design/_index.adoc:202", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 42, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 48, "source_unit": "https://translate-dev.freebsd.org/api/units/1569357/?format=api", "priority": 100, "id": 1569357, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=ee129462d1ae7d26", "url": "https://translate-dev.freebsd.org/api/units/1569357/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2023-05-21T18:04:58.625724Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "I did not take the final step of having an \"allocating hint pointer\" that would trundle through a portion of swap as allocations were made to further guarantee contiguous allocations or at least locality of reference, but I ensured that such an addition could be made." ], "previous_source": "", "target": [ "I did not take the final step of having an \"allocating hint pointer\" that would trundle through a portion of swap as allocations were made to further guarantee contiguous allocations or at least locality of reference, but I ensured that such an addition could be made." ], "id_hash": 7837563792822788789, "content_hash": 7837563792822788789, "location": "documentation/content/en/articles/vm-design/_index.adoc:205", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 44, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 46, "source_unit": "https://translate-dev.freebsd.org/api/units/1569359/?format=api", "priority": 100, "id": 1569359, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=ecc49ebe8faa62b5", "url": "https://translate-dev.freebsd.org/api/units/1569359/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2023-05-21T18:04:58.629210Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "How much overhead are we willing to suffer in the critical path to avoid freeing the wrong page? Each wrong choice we make will cost us hundreds of thousands of CPU cycles and a noticeable stall of the affected processes, so we are willing to endure a significant amount of overhead to be sure that the right page is chosen. This is why FreeBSD tends to outperform other systems when memory resources become stressed." ], "previous_source": "", "target": [ "How much overhead are we willing to suffer in the critical path to avoid freeing the wrong page? Each wrong choice we make will cost us hundreds of thousands of CPU cycles and a noticeable stall of the affected processes, so we are willing to endure a significant amount of overhead to be sure that the right page is chosen. This is why FreeBSD tends to outperform other systems when memory resources become stressed." ], "id_hash": -1660761811070994024, "content_hash": -1660761811070994024, "location": "documentation/content/en/articles/vm-design/_index.adoc:215", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 47, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 74, "source_unit": "https://translate-dev.freebsd.org/api/units/1569361/?format=api", "priority": 100, "id": 1569361, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=68f3c9ee92943998", "url": "https://translate-dev.freebsd.org/api/units/1569361/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2023-05-21T18:04:58.632661Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "An urban myth has circulated for years that Linux did a better job avoiding swapouts than FreeBSD, but this in fact is not true. What was actually occurring was that FreeBSD was proactively paging out unused pages to make room for more disk cache while Linux was keeping unused pages in core and leaving less memory available for cache and process pages. I do not know whether this is still true today." ], "previous_source": "", "target": [ "An urban myth has circulated for years that Linux did a better job avoiding swapouts than FreeBSD, but this in fact is not true. What was actually occurring was that FreeBSD was proactively paging out unused pages to make room for more disk cache while Linux was keeping unused pages in core and leaving less memory available for cache and process pages. I do not know whether this is still true today." ], "id_hash": 5780403638740543814, "content_hash": 5780403638740543814, "location": "documentation/content/en/articles/vm-design/_index.adoc:247", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 53, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 72, "source_unit": "https://translate-dev.freebsd.org/api/units/1569363/?format=api", "priority": 100, "id": 1569363, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=d0381e7bc9816946", "url": "https://translate-dev.freebsd.org/api/units/1569363/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2023-05-21T18:04:58.636420Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "A large percentage of page faults that occur are zero-fill faults. You can usually see this by observing the `vmstat -s` output. These occur when a process accesses pages in its BSS area. The BSS area is expected to be initially zero but the VM system does not bother to allocate any memory at all until the process actually accesses it. When a fault occurs the VM system must not only allocate a new page, it must zero it as well. To optimize the zeroing operation the VM system has the ability to pre-zero pages and mark them as such, and to request pre-zeroed pages when zero-fill faults occur. The pre-zeroing occurs whenever the CPU is idle but the number of pages the system pre-zeros is limited to avoid blowing away the memory caches. This is an excellent example of adding complexity to the VM system to optimize the critical path." ], "previous_source": "", "target": [ "A large percentage of page faults that occur are zero-fill faults. You can usually see this by observing the `vmstat -s` output. These occur when a process accesses pages in its BSS area. The BSS area is expected to be initially zero but the VM system does not bother to allocate any memory at all until the process actually accesses it. When a fault occurs the VM system must not only allocate a new page, it must zero it as well. To optimize the zeroing operation the VM system has the ability to pre-zero pages and mark them as such, and to request pre-zeroed pages when zero-fill faults occur. The pre-zeroing occurs whenever the CPU is idle but the number of pages the system pre-zeros is limited to avoid blowing away the memory caches. This is an excellent example of adding complexity to the VM system to optimize the critical path." ], "id_hash": -3287937168931606517, "content_hash": -3287937168931606517, "location": "documentation/content/en/articles/vm-design/_index.adoc:268", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 56, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 151, "source_unit": "https://translate-dev.freebsd.org/api/units/1569365/?format=api", "priority": 100, "id": 1569365, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=525ee690aa07300b", "url": "https://translate-dev.freebsd.org/api/units/1569365/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2023-05-21T18:04:58.640001Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "`pv_entry` structures only represent pages mapped by the MMU (one `pv_entry` represents one pte). This means that when we need to remove all hardware references to a `vm_page` (to reuse the page for something else, page it out, clear it, dirty it, and so forth) we can simply scan the linked list of pv_entry's associated with that vm_page to remove or modify the pte's from their page tables." ], "previous_source": "", "target": [ "`pv_entry` structures only represent pages mapped by the MMU (one `pv_entry` represents one pte). This means that when we need to remove all hardware references to a `vm_page` (to reuse the page for something else, page it out, clear it, dirty it, and so forth) we can simply scan the linked list of pv_entry's associated with that vm_page to remove or modify the pte's from their page tables." ], "id_hash": 2530387555015900761, "content_hash": 2530387555015900761, "location": "documentation/content/en/articles/vm-design/_index.adoc:366", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 80, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 68, "source_unit": "https://translate-dev.freebsd.org/api/units/1569367/?format=api", "priority": 100, "id": 1569367, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=a31dbe1320cabe59", "url": "https://translate-dev.freebsd.org/api/units/1569367/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2023-05-21T18:04:58.644980Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "Under Linux there is no such linked list. To remove all the hardware page table mappings for a `vm_page` linux must index into every VM object that _might_ have mapped the page. For example, if you have 50 processes all mapping the same shared library and want to get rid of page X in that library, you need to index into the page table for each of those 50 processes even if only 10 of them have actually mapped the page. So Linux is trading off the simplicity of its design against performance. Many VM algorithms which are O(1) or (small N) under FreeBSD wind up being O(N), O(N^2), or worse under Linux. Since the pte's representing a particular page in an object tend to be at the same offset in all the page tables they are mapped in, reducing the number of accesses into the page tables at the same pte offset will often avoid blowing away the L1 cache line for that offset, which can lead to better performance." ], "previous_source": "", "target": [ "Under Linux there is no such linked list. To remove all the hardware page table mappings for a `vm_page` linux must index into every VM object that _might_ have mapped the page. For example, if you have 50 processes all mapping the same shared library and want to get rid of page X in that library, you need to index into the page table for each of those 50 processes even if only 10 of them have actually mapped the page. So Linux is trading off the simplicity of its design against performance. Many VM algorithms which are O(1) or (small N) under FreeBSD wind up being O(N), O(N^2), or worse under Linux. Since the pte's representing a particular page in an object tend to be at the same offset in all the page tables they are mapped in, reducing the number of accesses into the page tables at the same pte offset will often avoid blowing away the L1 cache line for that offset, which can lead to better performance." ], "id_hash": -6410095001383200919, "content_hash": -6410095001383200919, "location": "documentation/content/en/articles/vm-design/_index.adoc:373", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 81, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 171, "source_unit": "https://translate-dev.freebsd.org/api/units/1569369/?format=api", "priority": 100, "id": 1569369, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=270ac48ef20c1769", "url": "https://translate-dev.freebsd.org/api/units/1569369/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2023-05-21T18:04:58.648393Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "FreeBSD has added complexity (the `pv_entry` scheme) to increase performance (to limit page table accesses to _only_ those pte's that need to be modified)." ], "previous_source": "", "target": [ "FreeBSD has added complexity (the `pv_entry` scheme) to increase performance (to limit page table accesses to _only_ those pte's that need to be modified)." ], "id_hash": 2794379834289832262, "content_hash": 2794379834289832262, "location": "documentation/content/en/articles/vm-design/_index.adoc:375", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 82, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 24, "source_unit": "https://translate-dev.freebsd.org/api/units/1569371/?format=api", "priority": 100, "id": 1569371, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=a6c7a1a76e5c4146", "url": "https://translate-dev.freebsd.org/api/units/1569371/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2023-05-21T18:04:58.652176Z" }, { "translation": "https://translate-dev.freebsd.org/api/translations/documentation/articlesvm-design_index/en/?format=api", "source": [ "Matthew Dillon <dillon@apollo.backplane.com>" ], "previous_source": "", "target": [ "Matthew Dillon <dillon@apollo.backplane.com>" ], "id_hash": 3328405665575463437, "content_hash": 3328405665575463437, "location": "documentation/content/en/articles/vm-design/_index.adoc:46", "context": "", "note": "type: Plain text", "flags": "", "labels": [], "state": 100, "fuzzy": false, "translated": true, "approved": false, "position": 4, "has_suggestion": false, "has_comment": false, "has_failing_check": false, "num_words": 3, "source_unit": "https://translate-dev.freebsd.org/api/units/1598414/?format=api", "priority": 100, "id": 1598414, "web_url": "https://translate-dev.freebsd.org/translate/documentation/articlesvm-design_index/en/?checksum=ae30df517505ea0d", "url": "https://translate-dev.freebsd.org/api/units/1598414/?format=api", "explanation": "", "extra_flags": "", "pending": false, "timestamp": "2023-09-09T21:51:23.525460Z" } ] }