Overview

Project website docs.freebsd.org/en
Translation license BSD 2-Clause "Simplified" License
Translation process
  • Translations can be made directly.
  • Translation suggestions can be made.
  • Only chosen users can contribute.
  • The translation uses bilingual files.
File mask documentation/content/*/articles/vm-design/_index.po
Translation file Download documentation/content/ko/articles/vm-design/_index.po
Last change April 26, 2023, 9:15 a.m.
Last author Kyung-tak, Yoo

String statistics

Strings percent Hosted strings Words percent Hosted words Characters percent Hosted characters
Total 93 6,345 36,561
Translated 4% 4 1% 4 1% 33
Needs editing 0% 0 0% 0 0% 0
Read-only 0% 0 0% 0 0% 0
Failing checks 0% 0 0% 0 0% 0
Strings with suggestions 0% 0 0% 0 0% 0
Untranslated strings 95% 89 99% 6,341 99% 36,528
Last 12 months Previous

Activity in last 12 months

0 changes 0 changes 0 changes 0 changes 0 changes 0 changes 0 changes 0 changes 0 changes 0 changes 0 changes 0 changes 0 changes 0 changes 0 changes 0 changes 0 changes 0 changes 0 changes 99 changes 0 changes 2 changes 0 changes 12 changes
Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May

Quick numbers

6,345
Hosted words
93
Hosted strings
4%
Translated
1
Contributors
and previous 30 days

Trends of last 30 days

+98%
Hosted words
+100%
+100%
Hosted strings
—
+4%
Translated
—
−100%
Contributors
+100%
User avatar None

New strings to translate

Documentation / articles/vm-design/_index — Korean

10 new strings to translate appeared to the translation. 10 days ago
User avatar None

Resource update

Documentation / articles/vm-design/_index — Korean

The "documentation/content/ko/articles/vm-design/_index.po" file was changed. 10 days ago
User avatar None

String updated in the repository

Documentation / articles/vm-design/_index — Korean

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).
 
10 days ago
User avatar None

String updated in the repository

Documentation / articles/vm-design/_index — Korean

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.
 
10 days ago
User avatar None

String updated in the repository

Documentation / articles/vm-design/_index — Korean

`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.
 
10 days ago
User avatar None

String updated in the repository

Documentation / articles/vm-design/_index — Korean

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.
 
10 days ago
User avatar None

String updated in the repository

Documentation / articles/vm-design/_index — Korean

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.
 
10 days ago
User avatar None

String updated in the repository

Documentation / articles/vm-design/_index — Korean

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.
 
10 days ago
User avatar None

String updated in the repository

Documentation / articles/vm-design/_index — Korean

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.
 
10 days ago
User avatar None

String updated in the repository

Documentation / articles/vm-design/_index — Korean

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.
 
10 days ago
Browse all translation changes