<?xml version="1.0" encoding="iso-8859-1"?>
<!DOCTYPE article PUBLIC "-//FreeBSD//DTD DocBook XML V5.0-Based Extension//EN"
"http://www.FreeBSD.org/XML/share/xml/freebsd50.dtd">
<!-- $FreeBSD$ -->
<!-- FreeBSD Documentation Project -->
<article xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" version="5.0" xml:lang="en">
<info><title>Design elements of the &os; VM system</title>
<authorgroup>
<author><personname><firstname>Matthew</firstname><surname>Dillon</surname></personname><affiliation>
<address>
<email>dillon@apollo.backplane.com</email>
</address>
</affiliation></author>
</authorgroup>
<legalnotice xml:id="trademarks" role="trademarks">
&tm-attrib.freebsd;
&tm-attrib.linux;
&tm-attrib.microsoft;
&tm-attrib.opengroup;
&tm-attrib.general;
</legalnotice>
<pubdate>$FreeBSD$</pubdate>
<releaseinfo>$FreeBSD$</releaseinfo>
<abstract>
<para>The title is really just a fancy way of saying that I am going to
attempt to describe the whole VM enchilada, hopefully in a way that
everyone can follow. For the last year I have concentrated on a number
of major kernel subsystems within &os;, with the VM and Swap
subsystems being the most interesting and NFS being <quote>a necessary
chore</quote>. I rewrote only small portions of the code. In the VM
arena the only major rewrite I have done is to the swap subsystem.
Most of my work was cleanup and maintenance, with only moderate code
rewriting and no major algorithmic adjustments within the VM
subsystem. The bulk of the VM subsystem's theoretical base remains
unchanged and a lot of the credit for the modernization effort in the
last few years belongs to John Dyson and David Greenman. Not being a
historian like Kirk I will not attempt to tag all the various features
with peoples names, since I will invariably get it wrong.</para>
</abstract>
<legalnotice xml:id="legalnotice">
<para>This article was originally published in the January 2000 issue of
<link xlink:href="http://www.daemonnews.org/">DaemonNews</link>. This
version of the article may include updates from Matt and other authors
to reflect changes in &os;'s VM implementation.</para>
</legalnotice>
</info>
<sect1 xml:id="introduction">
<title>Introduction</title>
<para>Before moving along to the actual design let's spend a little time
on the necessity of maintaining and modernizing any long-living
codebase. In the programming world, algorithms tend to be more
important than code and it is precisely due to BSD's academic roots that
a great deal of attention was paid to algorithm design from the
beginning. More attention paid to the design generally leads to a clean
and flexible codebase that can be fairly easily modified, extended, or
replaced over time. While BSD is considered an <quote>old</quote>
operating system by some people, those of us who work on it tend to view
it more as a <quote>mature</quote> codebase which has various components
modified, extended, or replaced with modern code. It has evolved, and
&os; is at the bleeding edge no matter how old some of the code might
be. This is an important distinction to make and one that is
unfortunately lost to many people. The biggest error a programmer can
make is to not learn from history, and this is precisely the error that
many other modern operating systems have made. &windowsnt; is the best example
of this, and the consequences have been dire. Linux also makes this
mistake to some degree—enough that we BSD folk can make small
jokes about it every once in a while, anyway. Linux's problem is simply
one of a lack of experience and history to compare ideas against, a
problem that is easily and rapidly being addressed by the Linux
community in the same way it has been addressed in the BSD
community—by continuous code development. The &windowsnt; folk, on the
other hand, repeatedly make the same mistakes solved by &unix; decades ago
and then spend years fixing them. Over and over again. They have a
severe case of <quote>not designed here</quote> and <quote>we are always
right because our marketing department says so</quote>. I have little
tolerance for anyone who cannot learn from history.</para>
<para>Much of the apparent complexity of the &os; design, especially in
the VM/Swap subsystem, is a direct result of having to solve serious
performance issues that occur under various conditions. These issues
are not due to bad algorithmic design but instead rise from
environmental factors. In any direct comparison between platforms,
these issues become most apparent when system resources begin to get
stressed. As I describe &os;'s VM/Swap subsystem the reader should
always keep two points in mind:</para>
<orderedlist>
<listitem>
<para>The most important aspect of performance design is what is
known as <quote>Optimizing the Critical Path</quote>. It is often
the case that performance optimizations add a little bloat to the
code in order to make the critical path perform better.</para>
</listitem>
<listitem>
<para>A solid, generalized design outperforms a heavily-optimized
design over the long run. While a generalized design may end up
being slower than an heavily-optimized design when they are
first implemented, the generalized design tends to be easier to
adapt to changing conditions and the heavily-optimized design
winds up having to be thrown away.</para>
</listitem>
</orderedlist>
<para>Any codebase that will survive and be maintainable for
years must therefore be designed properly from the beginning even if it
costs some performance. Twenty years ago people were still arguing that
programming in assembly was better than programming in a high-level
language because it produced code that was ten times as fast. Today,
the fallibility of that argument is obvious — as are
the parallels to algorithmic design and code generalization.</para>
</sect1>
<sect1 xml:id="vm-objects">
<title>VM Objects</title>
<para>The best way to begin describing the &os; VM system is to look at
it from the perspective of a user-level process. Each user process sees
a single, private, contiguous VM address space containing several types
of memory objects. These objects have various characteristics. Program
code and program data are effectively a single memory-mapped file (the
binary file being run), but program code is read-only while program data
is copy-on-write. Program BSS is just memory allocated and filled with
zeros on demand, called demand zero page fill. Arbitrary files can be
memory-mapped into the address space as well, which is how the shared
library mechanism works. Such mappings can require modifications to
remain private to the process making them. The fork system call adds an
entirely new dimension to the VM management problem on top of the
complexity already given.</para>
<para>A program binary data page (which is a basic copy-on-write page)
illustrates the complexity. A program binary contains a preinitialized
data section which is initially mapped directly from the program file.
When a program is loaded into a process's VM space, this area is
initially memory-mapped and backed by the program binary itself,
allowing the VM system to free/reuse the page and later load it back in
from the binary. The moment a process modifies this data, however, the
VM system must make a private copy of the page for that process. Since
the private copy has been modified, the VM system may no longer free it,
because there is no longer any way to restore it later on.</para>
<para>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 <function>fork()</function>. It would be
silly for the VM system to make a complete copy of the data at the time
of the <function>fork()</function> 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 effect the other.</para>
<para>&os; manages all of this with a layered VM Object model. The
original binary program file winds up being the lowest VM Object layer.
A copy-on-write layer is pushed on top of that to hold those pages which
had to be copied from the original file. If the program modifies a data
page belonging to the original file the VM system takes a fault and
makes a copy of the page in the higher layer. When a process forks,
additional VM Object layers are pushed on. This might make a little
more sense with a fairly basic example. A <function>fork()</function>
is a common operation for any *BSD system, so this example will consider
a program that starts up, and forks. When the process starts, the VM
system creates an object layer, let's call this A:</para>
<mediaobject>
<imageobject>
<imagedata fileref="fig1"/>
</imageobject>
<textobject>
<literallayout class="monospaced">+---------------+
| A |
+---------------+</literallayout>
</textobject>
<textobject>
<phrase>A picture</phrase>
</textobject>
</mediaobject>
<para>A represents the file—pages may be paged in and out of the
file's physical media as necessary. Paging in from the disk is
reasonable for a program, but we really do not want to page back out and
overwrite the executable. The VM system therefore creates a second
layer, B, that will be physically backed by swap space:</para>
<mediaobject>
<imageobject>
<imagedata fileref="fig2"/>
</imageobject>
<textobject>
<literallayout class="monospaced">+---------------+
| B |
+---------------+
| A |
+---------------+</literallayout>
</textobject>
</mediaobject>
<para>On the first write to a page after this, a new page is created in B,
and its contents are initialized from A. All pages in B can be paged in
or out to a swap device. When the program forks, the VM system creates
two new object layers—C1 for the parent, and C2 for the
child—that rest on top of B:</para>
<mediaobject>
<imageobject>
<imagedata fileref="fig3"/>
</imageobject>
<textobject>
<literallayout class="monospaced">+-------+-------+
| C1 | C2 |
+-------+-------+
| B |
+---------------+
| A |
+---------------+</literallayout>
</textobject>
</mediaobject>
<para>In this case, let's say a page in B is modified by the original
parent process. The process will take a copy-on-write fault and
duplicate the page in C1, leaving the original page in B untouched.
Now, let's say the same page in B is modified by the child process. The
process will take a copy-on-write fault and duplicate the page in C2.
The original page in B is now completely hidden since both C1 and C2
have a copy and B could theoretically be destroyed if it does not
represent a <quote>real</quote> file; however, this sort of optimization is not
trivial to make because it is so fine-grained. &os; does not make
this optimization. Now, suppose (as is often the case) that the child
process does an <function>exec()</function>. Its current address space
is usually replaced by a new address space representing a new file. In
this case, the C2 layer is destroyed:</para>
<mediaobject>
<imageobject>
<imagedata fileref="fig4"/>
</imageobject>
<textobject>
<literallayout class="monospaced">+-------+
| C1 |
+-------+-------+
| B |
+---------------+
| A |
+---------------+</literallayout>
</textobject>
</mediaobject>
<para>In this case, the number of children of B drops to one, and all
accesses to B now go through C1. This means that B and C1 can be
collapsed together. Any pages in B that also exist in C1 are deleted
from B during the collapse. Thus, even though the optimization in the
previous step could not be made, we can recover the dead pages when
either of the processes exit or <function>exec()</function>.</para>
<para>This model creates a number of potential problems. The first is that
you can wind up with a relatively deep stack of layered VM Objects which
can cost scanning time and memory when you take a fault. Deep
layering can occur when processes fork and then fork again (either
parent or child). The second problem is that you can wind up with dead,
inaccessible pages deep in the stack of VM Objects. In our last example
if both the parent and child processes modify the same page, they both
get their own private copies of the page and the original page in B is
no longer accessible by anyone. That page in B can be freed.</para>
<para>&os; solves the deep layering problem with a special optimization
called the <quote>All Shadowed Case</quote>. This case occurs if either
C1 or C2 take sufficient COW faults to completely shadow all pages in B.
Lets say that C1 achieves this. C1 can now bypass B entirely, so rather
then have C1->B->A and C2->B->A we now have C1->A and C2->B->A. But
look what also happened—now B has only one reference (C2), so we
can collapse B and C2 together. The end result is that B is deleted
entirely and we have C1->A and C2->A. It is often the case that B will
contain a large number of pages and neither C1 nor C2 will be able to
completely overshadow it. If we fork again and create a set of D
layers, however, it is much more likely that one of the D layers will
eventually be able to completely overshadow the much smaller dataset
represented by C1 or C2. The same optimization will work at any point in
the graph and the grand result of this is that even on a heavily forked
machine VM Object stacks tend to not get much deeper then 4. This is
true of both the parent and the children and true whether the parent is
doing the forking or whether the children cascade forks.</para>
<para>The dead page problem still exists in the case where C1 or C2 do not
completely overshadow B. Due to our other optimizations this case does
not represent much of a problem and we simply allow the pages to be
dead. If the system runs low on memory it will swap them out, eating a
little swap, but that is it.</para>
<para>The advantage to the VM Object model is that
<function>fork()</function> is extremely fast, since no real data
copying need take place. The disadvantage is that you can build a
relatively complex VM Object layering that slows page fault handling
down a little, and you spend memory managing the VM Object structures.
The optimizations &os; makes proves to reduce the problems enough
that they can be ignored, leaving no real disadvantage.</para>
</sect1>
<sect1 xml:id="swap-layers">
<title>SWAP Layers</title>
<para>Private data pages are initially either copy-on-write or zero-fill
pages. When a change, and therefore a copy, is made, the original
backing object (usually a file) can no longer be used to save a copy of
the page when the VM system needs to reuse it for other purposes. This
is where SWAP comes in. SWAP is allocated to create backing store for
memory that does not otherwise have it. &os; allocates the swap
management structure for a VM Object only when it is actually needed.
However, the swap management structure has had problems
historically:</para>
<itemizedlist>
<listitem>
<para>Under &os; 3.X the swap management structure preallocates an
array that encompasses the entire object requiring swap backing
store—even if only a few pages of that object are
swap-backed. This creates a kernel memory fragmentation problem
when large objects are mapped, or processes with large runsizes
(RSS) fork.</para>
</listitem>
<listitem>
<para>Also, in order to keep track of swap space, a <quote>list of
holes</quote> is kept in kernel memory, and this tends to get
severely fragmented as well. Since the <quote>list of
holes</quote> is a linear list, the swap allocation and freeing
performance is a non-optimal O(n)-per-page.</para>
</listitem>
<listitem>
<para>It requires kernel memory allocations to take place during
the swap freeing process, and that creates low memory deadlock
problems.</para>
</listitem>
<listitem>
<para>The problem is further exacerbated by holes created due to
the interleaving algorithm.</para>
</listitem>
<listitem>
<para>Also, the swap block map can become fragmented fairly easily
resulting in non-contiguous allocations.</para>
</listitem>
<listitem>
<para>Kernel memory must also be allocated on the fly for additional
swap management structures when a swapout occurs.</para>
</listitem>
</itemizedlist>
<para>It is evident from that list that there was plenty of room for
improvement. For &os; 4.X, I completely rewrote the swap
subsystem:</para>
<itemizedlist>
<listitem>
<para>Swap management structures are allocated through a hash
table rather than a linear array giving them a fixed allocation
size and much finer granularity.</para>
</listitem>
<listitem>
<para>Rather then using a linearly linked list to keep track of
swap space reservations, it now uses a bitmap of swap blocks
arranged in a radix tree structure with free-space hinting in
the radix node structures. This effectively makes swap
allocation and freeing an O(1) operation.</para>
</listitem>
<listitem>
<para>The entire radix tree bitmap is also preallocated in
order 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 in order to avoid potential
deadlocks.</para>
</listitem>
<listitem>
<para>To reduce fragmentation the radix tree is capable
of allocating large contiguous chunks at once, skipping over
smaller fragmented chunks.</para>
</listitem>
</itemizedlist>
<para>I did not take the final step of having an
<quote>allocating hint pointer</quote> that would trundle
through a portion of swap as allocations were made in order to further
guarantee contiguous allocations or at least locality of reference, but
I ensured that such an addition could be made.</para>
</sect1>
<sect1 xml:id="freeing-pages">
<title>When to free a page</title>
<para>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.</para>
<para>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 in order to be sure that the right page is chosen. This is why
&os; tends to outperform other systems when memory resources become
stressed.</para>
<para>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.</para>
<para>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.</para>
<sidebar>
<title>What if the hardware has no page-used bit?</title>
<para>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.</para>
</sidebar>
<para>&os; 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 &os;, 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 <quote>immediately freeable</quote> and will be reused
in an LRU (least-recently used) fashion when the system needs to
allocate new memory.</para>
<para>It is important to note that the &os; 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 <command>systat -vm</command> 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.</para>
<para>An urban
myth has circulated for years that Linux did a better job avoiding
swapouts than &os;, but this in fact is not true. What was actually
occurring was that &os; was proactively paging out unused pages in
order 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.</para>
</sect1>
<sect1 xml:id="prefault-optimizations">
<title>Pre-Faulting and Zeroing Optimizations</title>
<para>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 &os; will attempt to pre-populate a process's page tables
with those pages that are already in the VM Cache. One thing that
&os; 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 <command>vmstat
1</command> 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.</para>
<para>A large percentage of page faults that occur are zero-fill faults.
You can usually see this by observing the <command>vmstat -s</command>
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 in order to avoid blowing away the memory caches. This is an
excellent example of adding complexity to the VM system in order to
optimize the critical path.</para>
</sect1>
<sect1 xml:id="page-table-optimizations">
<title>Page Table Optimizations</title>
<para>The page table optimizations make up the most contentious part of
the &os; VM design and they have shown some strain with the advent of
serious use of <function>mmap()</function>. 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
<literal>pv_entry</literal> structure which is tied into the
<literal>vm_page</literal> structure. &os; can simply iterate
through those mappings that are known to exist while Linux must check
all page tables that <emphasis>might</emphasis> 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 &os; tends to make better
choices on which pages to reuse or swap when memory is stressed, giving
it better performance under load. However, &os; 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
<literal>pv_entry</literal> structures.</para>
<para>Both Linux and &os; need work in this area. &os; 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. &os; generally
has the performance advantage here at the cost of wasting a little extra
memory, but &os; 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.</para>
</sect1>
<sect1 xml:id="page-coloring-optimizations">
<title>Page Coloring</title>
<para>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).</para>
<para>&os;'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.</para>
</sect1>
<sect1 xml:id="conclusion">
<title>Conclusion</title>
<para>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 &os; VM system in the last several
years, and work is ongoing.</para>
</sect1>
<sect1 xml:id="allen-briggs-qa">
<title>Bonus QA session by Allen Briggs
<email>briggs@ninthwonder.com</email></title>
<qandaset>
<qandaentry>
<question>
<para>What is <quote>the interleaving algorithm</quote> that you
refer to in your listing of the ills of the &os; 3.X swap
arrangements?</para>
</question>
<answer>
<para>&os; uses a fixed swap interleave which defaults to 4. This
means that &os; 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 <quote>four swap areas</quote> will be
fragmented if you do not actually have four swap areas. For
example, if you have two swap areas A and B &os;'s address
space representation for that swap area will be interleaved in
blocks of 16 pages:</para>
<literallayout>A B C D A B C D A B C D A B C D</literallayout>
<para>&os; 3.X uses a <quote>sequential list of free
regions</quote> 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</filename>). 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 <quote>free</quote> and C and D shown as
<quote>all allocated</quote>. 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.</para>
<para>Why do we interleave our swap space instead of just tack swap
areas onto the end and do something fancier? Because it is a whole
lot easier to allocate linear swaths of an address space and have
the result automatically be interleaved across multiple disks than
it is to try to put that sophistication elsewhere.</para>
<para>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.</para>
<para>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.</para>
</answer>
</qandaentry>
<qandaentry>
<question>
<para>How is the separation of clean and dirty (inactive) pages
related to the situation where you see low cache queue counts and
high active queue counts in <command>systat -vm</command>? Do the
systat stats roll the active and dirty pages together for the
active queue count?</para>
<para>I do not get the following:</para>
<blockquote>
<para>It is important to note that the &os; 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
<command>systat -vm</command> command.</para>
</blockquote>
</question>
<answer>
<para>Yes, that is confusing. The relationship is
<quote>goal</quote> verses <quote>reality</quote>. 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.</para>
<para>What this means is that &os; will not try very hard to
separate out dirty pages (inactive queue) from clean pages (cache
queue) when the system is not being stressed, nor will it try to
deactivate pages (active queue -> inactive queue) when the system
is not being stressed, even if they are not being used.</para>
</answer>
</qandaentry>
<qandaentry>
<question>
<para> In the &man.ls.1; / <command>vmstat 1</command> example,
would not some of the page faults be data page faults (COW from
executable file to private page)? I.e., I would expect the page
faults to be some zero-fill and some program data. Or are you
implying that &os; does do pre-COW for the program data?</para>
</question>
<answer>
<para>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. &os; does not pre-COW program data or zero-fill,
but it <emphasis>does</emphasis> pre-map pages that exist in its
cache.</para>
</answer>
</qandaentry>
<qandaentry>
<question>
<para>In your section on page table optimizations, can you give a
little more detail about <literal>pv_entry</literal> and
<literal>vm_page</literal> (or should vm_page be
<literal>vm_pmap</literal>—as in 4.4, cf. pp. 180-181 of
McKusick, Bostic, Karel, Quarterman)? Specifically, what kind of
operation/reaction would require scanning the mappings?</para>
<para>How does Linux do in the case where &os; breaks down
(sharing a large file mapping over many processes)?</para>
</question>
<answer>
<para>A <literal>vm_page</literal> represents an (object,index#)
tuple. A <literal>pv_entry</literal> 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
<literal>vm_page</literal> structure and three
<literal>pv_entry</literal> structures.</para>
<para><literal>pv_entry</literal> structures only represent pages
mapped by the MMU (one <literal>pv_entry</literal> represents one
pte). This means that when we need to remove all hardware
references to a <literal>vm_page</literal> (in order 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
<literal>pv_entry</literal>'s associated with that
<literal>vm_page</literal> to remove or modify the pte's from
their page tables.</para>
<para>Under Linux there is no such linked list. In order to remove
all the hardware page table mappings for a
<literal>vm_page</literal> linux must index into every VM object
that <emphasis>might</emphasis> 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 &os; 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.</para>
<para>&os; has added complexity (the <literal>pv_entry</literal>
scheme) in order to increase performance (to limit page table
accesses to <emphasis>only</emphasis> those pte's that need to be
modified).</para>
<para>But &os; has a scaling problem that Linux does not in that
there are a limited number of <literal>pv_entry</literal>
structures and this causes problems when you have massive sharing
of data. In this case you may run out of
<literal>pv_entry</literal> structures even though there is plenty
of free memory available. This can be fixed easily enough by
bumping up the number of <literal>pv_entry</literal> structures in
the kernel config, but we really need to find a better way to do
it.</para>
<para>In regards to the memory overhead of a page table verses the
<literal>pv_entry</literal> scheme: Linux uses
<quote>permanent</quote> page tables that are not throw away, but
does not need a <literal>pv_entry</literal> for each potentially
mapped pte. &os; uses <quote>throw away</quote> page tables but
adds in a <literal>pv_entry</literal> structure for each
actually-mapped pte. I think memory utilization winds up being
about the same, giving &os; an algorithmic advantage with its
ability to throw away page tables at will with very low
overhead.</para>
</answer>
</qandaentry>
<qandaentry>
<question>
<para>Finally, in the page coloring section, it might help to have a
little more description of what you mean here. I did not quite
follow it.</para>
</question>
<answer>
<para>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 <emphasis>same</emphasis> 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!</para>
<para>Now, I am simplifying things greatly. What I just described
is what is called a <quote>direct mapped</quote> 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.</para>
<para>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.</para>
<para>It is extremely important…
<emphasis>extremely</emphasis> 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 <emphasis>hundreds</emphasis> 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
<emphasis>slow</emphasis>, when compared to the speed of a modern
processor core.</para>
<para>Ok, so now onto page coloring: All modern memory caches are
what are known as <emphasis>physical</emphasis> 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.</para>
<para>But in the &unix; 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
<emphasis>physical</emphasis> 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
<emphasis>physical</emphasis> memory.</para>
<para>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).</para>
<para>This is what Page coloring does. Instead of assigning
<emphasis>random</emphasis> physical pages to virtual addresses,
which may result in non-optimal cache performance, Page coloring
assigns <emphasis>reasonably-contiguous</emphasis> 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.</para>
<para>Note that I say <quote>reasonably</quote> contiguous rather
than simply <quote>contiguous</quote>. 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
<emphasis>not</emphasis> 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.</para>
</answer>
</qandaentry>
</qandaset>
</sect1>
</article>
