Subversion Repositories HelenOS-doc

Rev

Rev 27 | Rev 35 | Go to most recent revision | Only display areas with differences | Regard whitespace | Details | Blame | Last modification | View Log | RSS feed

Rev 27 Rev 34
1 
<?xml version="1.0" encoding="UTF-8"?>
 1 
<?xml version="1.0" encoding="UTF-8"?>
2 
<chapter id="mm">
 2 
<chapter id="mm">
3 
  <?dbhtml filename="mm.html"?>
 3 
  <?dbhtml filename="mm.html"?>
4 
 
 4 
 
5 
  <title>Memory management</title>
 5 
  <title>Memory management</title>
6 
 
 6 
 
7 
  <section>
 7 
  <section>
8 
    <!-- VM -->
 - 
 
9 
 
 - 
 
10 
    <title>Virtual memory management</title>
 8 
    <title>Virtual memory management</title>
11 
 
 9 
 
12 
    <section>
 10 
    <section>
13 
      <title>Address spaces</title>
 11 
      <title>Address spaces</title>
14 
 
 12 
 
15 
      <para></para>
 13 
      <para />
16 
    </section>
 14 
    </section>
17 
 
 15 
 
18 
    <section>
 16 
    <section>
19 
      <title>Virtual address translation</title>
 17 
      <title>Virtual address translation</title>
20 
 
 18 
 
21 
      <para></para>
 19 
      <para />
22 
    </section>
 20 
    </section>
- 
 
 21 
 
- 
 
 22 
    <para>Page tables. 4 level hierarchical and hash directly supported. B+
- 
 
 23 
    Tree can be implemented.</para>
- 
 
 24 
 
- 
 
 25 
    <para>For paging there is an abstract layer</para>
- 
 
 26 
 
- 
 
 27 
    <para>TLB shootdown implementation (update TLB upon mapping
- 
 
 28 
    update/remove). TLB shootdown ASID/ASID:PAGE/ALL. TLB shootdown requests
- 
 
 29 
    can come in asynchroniously so there is a cache of TLB shootdown requests.
- 
 
 30 
    Upon cache overflow TLB shootdown ALL is executed</para>
- 
 
 31 
 
- 
 
 32 
    <para>Address spaces. Address space area (B+ tree). Only for uspace. Set
- 
 
 33 
    of syscalls (shrink/extend etc). Special address space area type - device
- 
 
 34 
    - prohibits shrink/extend syscalls to call on it. Address space has link
- 
 
 35 
    to mapping tables (hierarchical - per Address space, hash - global
- 
 
 36 
    tables).</para>
23 
  </section>
 37 
  </section>
24 
 
 38 
 
25 
  <!-- End of VM -->
 39 
  <!-- End of VM -->
26 
 
 40 
 
27 
  <section>
 41 
  <section>
28 
    <!-- Phys mem -->
 42 
    <!-- Phys mem -->
29 
 
 43 
 
30 
    <title>Physical memory management</title>
 44 
    <title>Physical memory management</title>
31 
 
 45 
 
32 
    <section id="zones_and_frames">
 46 
    <section id="zones_and_frames">
33 
      <title>Zones and frames</title>
 47 
      <title>Zones and frames</title>
34 
 
 48 
 
35 
      <para>
 - 
 
36 
      <!--graphic fileref="images/mm2.png" /-->
 - 
 
37 
     
 - 
 
38 
      <!--graphic fileref="images/buddy_alloc.svg" format="SVG" /-->
 49 
      <para><!--graphic fileref="images/mm2.png" /--><!--graphic fileref="images/buddy_alloc.svg" format="SVG" /--></para>
39 
     
 - 
 
40 
     
 - 
 
41 
      </para>
 - 
 
42 
 
 50 
 
43 
      <para>On some architectures not whole physical memory is available for
 51 
      <para>On some architectures not whole physical memory is available for
44 
      conventional usage. This limitations require from kernel to maintain a
 52 
      conventional usage. This limitations require from kernel to maintain a
45 
      table of available and unavailable ranges of physical memory addresses.
 53 
      table of available and unavailable ranges of physical memory addresses.
46 
      Main idea of zones is in creating memory zone entity, that is a
 54 
      Main idea of zones is in creating memory zone entity, that is a
47 
      continuous chunk of memory available for allocation. If some chunk is
 55 
      continuous chunk of memory available for allocation. If some chunk is
48 
      not available, we simply do not put it in any zone.</para>
 56 
      not available, we simply do not put it in any zone.</para>
49 
 
 57 
 
50 
      <para>Zone is also serves for informational purposes, containing
 58 
      <para>Zone is also serves for informational purposes, containing
51 
      information about number of free and busy frames. Physical memory
 59 
      information about number of free and busy frames. Physical memory
52 
      allocation is also done inside the certain zone. Allocation of zone
 60 
      allocation is also done inside the certain zone. Allocation of zone
53 
      frame must be organized by the <link linkend="frame_allocator">frame
 61 
      frame must be organized by the <link linkend="frame_allocator">frame
54 
      allocator</link> associated with the zone.</para>
 62 
      allocator</link> associated with the zone.</para>
55 
 
 63 
 
56 
      <para>Some of the architectures (mips32, ppc32) have only one zone, that
 64 
      <para>Some of the architectures (mips32, ppc32) have only one zone, that
57 
      covers whole physical memory, and the others (like ia32) may have
 65 
      covers whole physical memory, and the others (like ia32) may have
58 
      multiple zones. Information about zones on current machine is stored in
 66 
      multiple zones. Information about zones on current machine is stored in
59 
      BIOS hardware tables or can be hardcoded into kernel during compile
 67 
      BIOS hardware tables or can be hardcoded into kernel during compile
60 
      time.</para>
 68 
      time.</para>
61 
    </section>
 69 
    </section>
62 
 
 70 
 
63 
    <section id="frame_allocator">
 71 
    <section id="frame_allocator">
64 
      <title>Frame allocator</title>
 72 
      <title>Frame allocator</title>
65 
 
 73 
 
66 
      <formalpara>
 74 
      <formalpara>
67 
        <title>Overview</title>
 75 
        <title>Overview</title>
68 
 
 76 
 
69 
        <para>Frame allocator provides physical memory allocation for the
 77 
        <para>Frame allocator provides physical memory allocation for the
70 
        kernel. Because of zonal organization of physical memory, frame
 78 
        kernel. Because of zonal organization of physical memory, frame
71 
        allocator is always working in context of some zone, thus making
 79 
        allocator is always working in context of some zone, thus making
72 
        impossible to allocate a piece of memory, which lays in different
 80 
        impossible to allocate a piece of memory, which lays in different
73 
        zone, which cannot happen, because two adjacent zones can be merged
 81 
        zone, which cannot happen, because two adjacent zones can be merged
74 
        into one. Frame allocator is also being responsible to update
 82 
        into one. Frame allocator is also being responsible to update
75 
        information on the number of free/busy frames in zone. Physical memory
 83 
        information on the number of free/busy frames in zone. Physical memory
76 
        allocation inside one <link linkend="zones_and_frames">memory
 84 
        allocation inside one <link linkend="zones_and_frames">memory
77 
        zone</link> is being handled by an instance of <link
 85 
        zone</link> is being handled by an instance of <link
78 
        linkend="buddy_allocator">buddy allocator</link> tailored to allocate
 86 
        linkend="buddy_allocator">buddy allocator</link> tailored to allocate
79 
        blocks of physical memory frames.</para>
 87 
        blocks of physical memory frames.</para>
80 
      </formalpara>
 88 
      </formalpara>
81 
 
 89 
 
82 
      <formalpara>
 90 
      <formalpara>
83 
        <title>Allocation / deallocation</title>
 91 
        <title>Allocation / deallocation</title>
84 
 
 92 
 
85 
        <para>Upon allocation request, frame allocator tries to find first
 93 
        <para>Upon allocation request, frame allocator tries to find first
86 
        zone, that can satisfy the incoming request (has required amount of
 94 
        zone, that can satisfy the incoming request (has required amount of
87 
        free frames to allocate). During deallocation, frame allocator needs
 95 
        free frames to allocate). During deallocation, frame allocator needs
88 
        to find zone, that contain deallocated frame. This approach could
 96 
        to find zone, that contain deallocated frame. This approach could
89 
        bring up two potential problems: <itemizedlist>
 97 
        bring up two potential problems: <itemizedlist>
90 
            <listitem>
 98 
            <listitem>
91 
               Linear search of zones does not any good to performance, but number of zones is not expected to be high. And if yes, list of zones can be replaced with more time-efficient B-tree.
 99 
               Linear search of zones does not any good to performance, but number of zones is not expected to be high. And if yes, list of zones can be replaced with more time-efficient B-tree.
92 
            </listitem>
 100 
            </listitem>
93 
 
 101 
 
94 
            <listitem>
 102 
            <listitem>
95 
               Quickly find out if zone contains required number of frames to allocate and if this chunk of memory is properly aligned. This issue is perfectly solved bu the buddy allocator.
 103 
               Quickly find out if zone contains required number of frames to allocate and if this chunk of memory is properly aligned. This issue is perfectly solved bu the buddy allocator.
96 
            </listitem>
 104 
            </listitem>
97 
          </itemizedlist></para>
 105 
          </itemizedlist></para>
98 
      </formalpara>
 106 
      </formalpara>
99 
    </section>
 107 
    </section>
100 
  </section>
 - 
 
101 
 
 108 
 
102 
  <section id="buddy_allocator">
 109 
    <section id="buddy_allocator">
103 
    <title>Buddy allocator</title>
 110 
      <title>Buddy allocator</title>
104 
 
 111 
 
105 
    <section>
 112 
      <section>
106 
      <title>Overview</title>
 113 
        <title>Overview</title>
107 
 
 114 
 
108 
      <para>In buddy allocator, memory is broken down into power-of-two sized
 115 
        <para>In buddy allocator, memory is broken down into power-of-two
109 
      naturally aligned blocks. These blocks are organized in an array of
 116 
        sized naturally aligned blocks. These blocks are organized in an array
110 
      lists in which list with index i contains all unallocated blocks of the
 117 
        of lists in which list with index i contains all unallocated blocks of
111 
      size <mathphrase>2<superscript>i</superscript></mathphrase>. The index i
 118 
        the size <mathphrase>2<superscript>i</superscript></mathphrase>. The
112 
      is called the order of block. Should there be two adjacent equally sized
 119 
        index i is called the order of block. Should there be two adjacent
113 
      blocks in list <mathphrase>i</mathphrase> (i.e. buddies), the buddy
 120 
        equally sized blocks in list <mathphrase>i</mathphrase> (i.e.
114 
      allocator would coalesce them and put the resulting block in list
 121 
        buddies), the buddy allocator would coalesce them and put the
115 
      <mathphrase>i + 1</mathphrase>, provided that the resulting block would
 122 
        resulting block in list <mathphrase>i + 1</mathphrase>, provided that
116 
      be naturally aligned. Similarily, when the allocator is asked to
 123 
        the resulting block would be naturally aligned. Similarily, when the
117 
      allocate a block of size
 124 
        allocator is asked to allocate a block of size
118 
      <mathphrase>2<superscript>i</superscript></mathphrase>, it first tries
 125 
        <mathphrase>2<superscript>i</superscript></mathphrase>, it first tries
119 
      to satisfy the request from list with index i. If the request cannot be
 126 
        to satisfy the request from list with index i. If the request cannot
120 
      satisfied (i.e. the list i is empty), the buddy allocator will try to
 127 
        be satisfied (i.e. the list i is empty), the buddy allocator will try
121 
      allocate and split larger block from list with index i + 1. Both of
 128 
        to allocate and split larger block from list with index i + 1. Both of
122 
      these algorithms are recursive. The recursion ends either when there are
 129 
        these algorithms are recursive. The recursion ends either when there
123 
      no blocks to coalesce in the former case or when there are no blocks
 130 
        are no blocks to coalesce in the former case or when there are no
124 
      that can be split in the latter case.</para>
 131 
        blocks that can be split in the latter case.</para>
125 
 
 132 
 
126 
      <!--graphic fileref="images/mm1.png" format="EPS" /-->
 133 
        <!--graphic fileref="images/mm1.png" format="EPS" /-->
127 
 
 134 
 
128 
      <para>This approach greatly reduces external fragmentation of memory and
 135 
        <para>This approach greatly reduces external fragmentation of memory
129 
      helps in allocating bigger continuous blocks of memory aligned to their
 136 
        and helps in allocating bigger continuous blocks of memory aligned to
130 
      size. On the other hand, the buddy allocator suffers increased internal
 137 
        their size. On the other hand, the buddy allocator suffers increased
131 
      fragmentation of memory and is not suitable for general kernel
 138 
        internal fragmentation of memory and is not suitable for general
132 
      allocations. This purpose is better addressed by the <link
 139 
        kernel allocations. This purpose is better addressed by the <link
133 
      linkend="slab">slab allocator</link>.</para>
 140 
        linkend="slab">slab allocator</link>.</para>
134 
    </section>
 141 
      </section>
135 
 
 142 
 
136 
    <section>
 143 
      <section>
137 
      <title>Implementation</title>
 144 
        <title>Implementation</title>
138 
 
 145 
 
139 
      <para>The buddy allocator is, in fact, an abstract framework wich can be
 146 
        <para>The buddy allocator is, in fact, an abstract framework wich can
140 
      easily specialized to serve one particular task. It knows nothing about
 147 
        be easily specialized to serve one particular task. It knows nothing
141 
      the nature of memory it helps to allocate. In order to beat the lack of
 148 
        about the nature of memory it helps to allocate. In order to beat the
142 
      this knowledge, the buddy allocator exports an interface that each of
 149 
        lack of this knowledge, the buddy allocator exports an interface that
143 
      its clients is required to implement. When supplied an implementation of
 150 
        each of its clients is required to implement. When supplied an
144 
      this interface, the buddy allocator can use specialized external
 151 
        implementation of this interface, the buddy allocator can use
145 
      functions to find buddy for a block, split and coalesce blocks,
 152 
        specialized external functions to find buddy for a block, split and
146 
      manipulate block order and mark blocks busy or available. For precize
 153 
        coalesce blocks, manipulate block order and mark blocks busy or
147 
      documentation of this interface, refer to <link linkend="???">HelenOS
 154 
        available. For precize documentation of this interface, refer to <link
148 
      Generic Kernel Reference Manual</link>.</para>
 155 
        linkend="???">HelenOS Generic Kernel Reference Manual</link>.</para>
149 
 
 156 
 
150 
      <formalpara>
 157 
        <formalpara>
151 
        <title>Data organization</title>
 158 
          <title>Data organization</title>
152 
 
 159 
 
153 
        <para>Each entity allocable by the buddy allocator is required to
 160 
          <para>Each entity allocable by the buddy allocator is required to
154 
        contain space for storing block order number and a link variable used
 161 
          contain space for storing block order number and a link variable
155 
        to interconnect blocks within the same order.</para>
 162 
          used to interconnect blocks within the same order.</para>
156 
 
 163 
 
157 
        <para>Whatever entities are allocated by the buddy allocator, the
 164 
          <para>Whatever entities are allocated by the buddy allocator, the
158 
        first entity within a block is used to represent the entire block. The
 165 
          first entity within a block is used to represent the entire block.
159 
        first entity keeps the order of the whole block. Other entities within
 166 
          The first entity keeps the order of the whole block. Other entities
160 
        the block are assigned the magic value
 167 
          within the block are assigned the magic value
161 
        <constant>BUDDY_INNER_BLOCK</constant>. This is especially important
 168 
          <constant>BUDDY_INNER_BLOCK</constant>. This is especially important
162 
        for effective identification of buddies in one-dimensional array
 169 
          for effective identification of buddies in one-dimensional array
163 
        because the entity that represents a potential buddy cannot be
 170 
          because the entity that represents a potential buddy cannot be
164 
        associated with <constant>BUDDY_INNER_BLOCK</constant> (i.e. if it is
 171 
          associated with <constant>BUDDY_INNER_BLOCK</constant> (i.e. if it
165 
        associated with <constant>BUDDY_INNER_BLOCK</constant> then it is not
 172 
          is associated with <constant>BUDDY_INNER_BLOCK</constant> then it is
166 
        a buddy).</para>
 173 
          not a buddy).</para>
167 
      </formalpara>
 - 
 
168 
 
 - 
 
169 
      <formalpara>
 - 
 
170 
        <title>Data organization</title>
 - 
 
171 
 
 174 
 
172 
        <para>Buddy allocator always uses first frame to represent frame
 175 
          <para>Buddy allocator always uses first frame to represent frame
173 
        block. This frame contains <varname>buddy_order</varname> variable to
 176 
          block. This frame contains <varname>buddy_order</varname> variable
174 
        provide information about the block size it actually represents (
 177 
          to provide information about the block size it actually represents (
175 
        <mathphrase>2<superscript>buddy_order</superscript></mathphrase>
 178 
          <mathphrase>2<superscript>buddy_order</superscript></mathphrase>
176 
        frames block). Other frames in block have this value set to magic
 179 
          frames block). Other frames in block have this value set to magic
177 
        <constant>BUDDY_INNER_BLOCK</constant> that is much greater than buddy
 180 
          <constant>BUDDY_INNER_BLOCK</constant> that is much greater than
178 
        <varname>max_order</varname> value.</para>
 181 
          buddy <varname>max_order</varname> value.</para>
179 
 
 182 
 
180 
        <para>Each <varname>frame_t</varname> also contains pointer member to
 183 
          <para>Each <varname>frame_t</varname> also contains pointer member
181 
        hold frame structure in the linked list inside one order.</para>
 184 
          to hold frame structure in the linked list inside one order.</para>
182 
      </formalpara>
 185 
        </formalpara>
183 
 
 186 
 
184 
      <formalpara>
 187 
        <formalpara>
185 
        <title>Allocation algorithm</title>
 188 
          <title>Allocation algorithm</title>
186 
 
 189 
 
187 
        <para>Upon <mathphrase>2<superscript>i</superscript></mathphrase>
 190 
          <para>Upon <mathphrase>2<superscript>i</superscript></mathphrase>
188 
        frames block allocation request, allocator checks if there are any
 191 
          frames block allocation request, allocator checks if there are any
189 
        blocks available at the order list <varname>i</varname>. If yes,
 192 
          blocks available at the order list <varname>i</varname>. If yes,
190 
        removes block from order list and returns its address. If no,
 193 
          removes block from order list and returns its address. If no,
191 
        recursively allocates
 194 
          recursively allocates
192 
        <mathphrase>2<superscript>i+1</superscript></mathphrase> frame block,
 195 
          <mathphrase>2<superscript>i+1</superscript></mathphrase> frame
193 
        splits it into two
 196 
          block, splits it into two
194 
        <mathphrase>2<superscript>i</superscript></mathphrase> frame blocks.
 197 
          <mathphrase>2<superscript>i</superscript></mathphrase> frame blocks.
195 
        Then adds one of the blocks to the <varname>i</varname> order list and
 198 
          Then adds one of the blocks to the <varname>i</varname> order list
196 
        returns address of another.</para>
 199 
          and returns address of another.</para>
197 
      </formalpara>
 200 
        </formalpara>
198 
 
 201 
 
199 
      <formalpara>
 202 
        <formalpara>
200 
        <title>Deallocation algorithm</title>
 203 
          <title>Deallocation algorithm</title>
201 
 
 204 
 
202 
        <para>Check if block has so called buddy (another free
 205 
          <para>Check if block has so called buddy (another free
203 
        <mathphrase>2<superscript>i</superscript></mathphrase> frame block
 206 
          <mathphrase>2<superscript>i</superscript></mathphrase> frame block
204 
        that can be linked with freed block into the
 207 
          that can be linked with freed block into the
205 
        <mathphrase>2<superscript>i+1</superscript></mathphrase> block).
 208 
          <mathphrase>2<superscript>i+1</superscript></mathphrase> block).
206 
        Technically, buddy is a odd/even block for even/odd block
 209 
          Technically, buddy is a odd/even block for even/odd block
207 
        respectively. Plus we can put an extra requirement, that resulting
 210 
          respectively. Plus we can put an extra requirement, that resulting
208 
        block must be aligned to its size. This requirement guarantees natural
 211 
          block must be aligned to its size. This requirement guarantees
209 
        block alignment for the blocks coming out the allocation
 212 
          natural block alignment for the blocks coming out the allocation
210 
        system.</para>
 213 
          system.</para>
211 
 
 214 
 
212 
        <para>Using direct pointer arithmetics,
 215 
          <para>Using direct pointer arithmetics,
213 
        <varname>frame_t::ref_count</varname> and
 216 
          <varname>frame_t::ref_count</varname> and
214 
        <varname>frame_t::buddy_order</varname> variables, finding buddy is
 217 
          <varname>frame_t::buddy_order</varname> variables, finding buddy is
215 
        done at constant time.</para>
 218 
          done at constant time.</para>
216 
      </formalpara>
 219 
        </formalpara>
217 
    </section>
 220 
      </section>
- 
 
 221 
    </section>
218 
 
 222 
 
219 
    <section id="slab">
 223 
    <section id="slab">
220 
      <title>Slab allocator</title>
 224 
      <title>Slab allocator</title>
221 
 
 225 
 
222 
      <section>
 226 
      <section>
223 
        <title>Introduction</title>
 227 
        <title>Overview</title>
- 
 
 228 
 
- 
 
 229 
        <para><termdef><glossterm>Slab</glossterm> represents a contiguous
- 
 
 230 
        piece of memory, usually made of several physically contiguous
- 
 
 231 
        pages.</termdef> <termdef><glossterm>Slab cache</glossterm> consists
- 
 
 232 
        of one or more slabs.</termdef></para>
224 
 
 233 
 
225 
        <para>The majority of memory allocation requests in the kernel are for
 234 
        <para>The majority of memory allocation requests in the kernel are for
226 
        small, frequently used data structures. For this purpose the slab
 235 
        small, frequently used data structures. For this purpose the slab
227 
        allocator is a perfect solution. The basic idea behind a slab
 236 
        allocator is a perfect solution. The basic idea behind the slab
228 
        allocator is to have lists of commonly used objects available packed
 237 
        allocator is to have lists of commonly used objects available packed
229 
        into pages. This avoids the overhead of allocating and destroying
 238 
        into pages. This avoids the overhead of allocating and destroying
230 
        commonly used types of objects such as inodes, threads, virtual memory
 239 
        commonly used types of objects such threads, virtual memory structures
231 
        structures etc.</para>
 - 
 
232 
 
 - 
 
233 
        <para>Original slab allocator locking mechanism has become a
 240 
        etc. Also due to the exact allocated size matching, slab allocation
234 
        significant preformance bottleneck on SMP architectures. <termdef>Slab
 - 
 
235 
        SMP perfromance bottleneck was resolved by introducing a per-CPU
 - 
 
236 
        caching scheme called as <glossterm>magazine
 - 
 
237 
        layer</glossterm></termdef>.</para>
 241 
        completely eliminates internal fragmentation issue.</para>
238 
      </section>
 242 
      </section>
239 
 
 243 
 
240 
      <section>
 244 
      <section>
241 
        <title>Implementation details (needs revision)</title>
 245 
        <title>Implementation</title>
242 
 
 246 
 
243 
        <para>The SLAB allocator is closely modelled after <ulink
 247 
        <para>The SLAB allocator is closely modelled after <ulink
244 
        url="http://www.usenix.org/events/usenix01/full_papers/bonwick/bonwick_html/">
 248 
        url="http://www.usenix.org/events/usenix01/full_papers/bonwick/bonwick_html/">
245 
        OpenSolaris SLAB allocator by Jeff Bonwick and Jonathan Adams </ulink>
 249 
        OpenSolaris SLAB allocator by Jeff Bonwick and Jonathan Adams </ulink>
246 
        with the following exceptions: <itemizedlist>
 250 
        with the following exceptions: <itemizedlist>
247 
            <listitem>
 251 
            <listitem>
248 
               empty SLABS are deallocated immediately (in Linux they are kept in linked list, in Solaris ???)
 252 
               empty SLABS are deallocated immediately (in Linux they are kept in linked list, in Solaris ???)
249 
            </listitem>
 253 
            </listitem>
250 
 
 254 
 
251 
            <listitem>
 255 
            <listitem>
252 
               empty magazines are deallocated when not needed (in Solaris they are held in linked list in slab cache)
 256 
               empty magazines are deallocated when not needed (in Solaris they are held in linked list in slab cache)
253 
            </listitem>
 257 
            </listitem>
254 
          </itemizedlist> Following features are not currently supported but
 258 
          </itemizedlist> Following features are not currently supported but
255 
        would be easy to do: <itemizedlist>
 259 
        would be easy to do: <itemizedlist>
256 
            <listitem>
 260 
            <listitem>
257 
               - cache coloring
 261 
               - cache coloring
258 
            </listitem>
 262 
            </listitem>
259 
 
 263 
 
260 
            <listitem>
 264 
            <listitem>
261 
               - dynamic magazine growing (different magazine sizes are already supported, but we would need to adjust allocation strategy)
 265 
               - dynamic magazine grow (different magazine sizes are already supported, but we would need to adjust allocation strategy)
262 
            </listitem>
 266 
            </listitem>
263 
          </itemizedlist></para>
 267 
          </itemizedlist></para>
264 
 
 268 
 
265 
        <para>The SLAB allocator supports per-CPU caches ('magazines') to
 269 
        <section>
266 
        facilitate good SMP scaling.</para>
 270 
          <title>Magazine layer</title>
267 
 
 271 
 
268 
        <para>When a new object is being allocated, it is first checked, if it
 272 
          <para>Due to the extensive bottleneck on SMP architures, caused by
269 
        is available in CPU-bound magazine. If it is not found there, it is
 273 
          global SLAB locking mechanism, making processing of all slab
270 
        allocated from CPU-shared SLAB - if partial full is found, it is used,
 274 
          allocation requests serialized, a new layer was introduced to the
271 
        otherwise a new one is allocated.</para>
 - 
 
272 
 
 - 
 
273 
        <para>When an object is being deallocated, it is put to CPU-bound
 275 
          classic slab allocator design. Slab allocator was extended to
274 
        magazine. If there is no such magazine, new one is allocated (if it
 276 
          support per-CPU caches 'magazines' to achieve good SMP scaling.
275 
        fails, the object is deallocated into SLAB). If the magazine is full,
 277 
          <termdef>Slab SMP perfromance bottleneck was resolved by introducing
276 
        it is put into cpu-shared list of magazines and new one is
 278 
          a per-CPU caching scheme called as <glossterm>magazine
277 
        allocated.</para>
 279 
          layer</glossterm></termdef>.</para>
278 
 
 280 
 
- 
 
 281 
          <para>Magazine is a N-element cache of objects, so each magazine can
- 
 
 282 
          satisfy N allocations. Magazine behaves like a automatic weapon
- 
 
 283 
          magazine (LIFO, stack), so the allocation/deallocation become simple
- 
 
 284 
          push/pop pointer operation. Trick is that CPU does not access global
- 
 
 285 
          slab allocator data during the allocation from its magazine, thus
- 
 
 286 
          making possible parallel allocations between CPUs.</para>
- 
 
 287 
 
- 
 
 288 
          <para>Implementation also requires adding another feature as the
279 
        <para>The CPU-bound magazine is actually a pair of magazines to avoid
 289 
          CPU-bound magazine is actually a pair of magazines to avoid
280 
        thrashing when somebody is allocating/deallocating 1 item at the
 290 
          thrashing when during allocation/deallocatiion of 1 item at the
281 
        magazine size boundary. LIFO order is enforced, which should avoid
 291 
          magazine size boundary. LIFO order is enforced, which should avoid
282 
        fragmentation as much as possible.</para>
 292 
          fragmentation as much as possible.</para>
283 
 
 293 
 
- 
 
 294 
          <para>Another important entity of magazine layer is a full magazine
- 
 
 295 
          depot, that stores full magazines which are used by any of the CPU
- 
 
 296 
          magazine caches to reload active CPU magazine. Magazine depot can be
- 
 
 297 
          pre-filled with full magazines during initialization, but in current
- 
 
 298 
          implementation it is filled during object deallocation, when CPU
- 
 
 299 
          magazine becomes full.</para>
- 
 
 300 
 
- 
 
 301 
          <para>Slab allocator control structures are allocated from special
- 
 
 302 
          slabs, that are marked by special flag, indicating that it should
- 
 
 303 
          not be used for slab magazine layer. This is done to avoid possible
- 
 
 304 
          infinite recursions and deadlock during conventional slab allocaiton
- 
 
 305 
          requests.</para>
- 
 
 306 
        </section>
- 
 
 307 
 
- 
 
 308 
        <section>
- 
 
 309 
          <title>Allocation/deallocation</title>
- 
 
 310 
 
284 
        <para>Every cache contains list of full slabs and list of partialy
 311 
          <para>Every cache contains list of full slabs and list of partialy
285 
        full slabs. Empty SLABS are immediately freed (thrashing will be
 312 
          full slabs. Empty slabs are immediately freed (thrashing will be
286 
        avoided because of magazines).</para>
 313 
          avoided because of magazines).</para>
287 
 
 314 
 
288 
        <para>The SLAB information structure is kept inside the data area, if
 - 
 
289 
        possible. The cache can be marked that it should not use magazines.
 - 
 
290 
        This is used only for SLAB related caches to avoid deadlocks and
 - 
 
291 
        infinite recursion (the SLAB allocator uses itself for allocating all
 - 
 
292 
        it's control structures).</para>
 - 
 
293 
 
 - 
 
294 
        <para>The SLAB allocator allocates lots of space and does not free it.
 315 
          <para>The SLAB allocator allocates lots of space and does not free
295 
        When frame allocator fails to allocate the frame, it calls
 316 
          it. When frame allocator fails to allocate the frame, it calls
296 
        slab_reclaim(). It tries 'light reclaim' first, then brutal reclaim.
 317 
          slab_reclaim(). It tries 'light reclaim' first, then brutal reclaim.
297 
        The light reclaim releases slabs from cpu-shared magazine-list, until
 318 
          The light reclaim releases slabs from cpu-shared magazine-list,
298 
        at least 1 slab is deallocated in each cache (this algorithm should
 319 
          until at least 1 slab is deallocated in each cache (this algorithm
299 
        probably change). The brutal reclaim removes all cached objects, even
 320 
          should probably change). The brutal reclaim removes all cached
300 
        from CPU-bound magazines.</para>
 321 
          objects, even from CPU-bound magazines.</para>
301 
 
 322 
 
302 
        <para>TODO: <itemizedlist>
 - 
 
303 
            <listitem>
 323 
          <formalpara>
304 
               For better CPU-scaling the magazine allocation strategy should be extended. Currently, if the cache does not have magazine, it asks for non-cpu cached magazine cache to provide one. It might be feasible to add cpu-cached magazine cache (which would allocate it's magazines from non-cpu-cached mag. cache). This would provide a nice per-cpu buffer. The other possibility is to use the per-cache 'empty-magazine-list', which decreases competing for 1 per-system magazine cache.
 324 
            <title>Allocation</title>
305 
            </listitem>
 - 
 
306 
 
 325 
 
- 
 
 326 
            <para><emphasis>Step 1.</emphasis> When it comes to the allocation
- 
 
 327 
            request, slab allocator first of all checks availability of memory
- 
 
 328 
            in local CPU-bound magazine. If it is there, we would just "pop"
- 
 
 329 
            the CPU magazine and return the pointer to object.</para>
- 
 
 330 
 
- 
 
 331 
            <para><emphasis>Step 2.</emphasis> If the CPU-bound magazine is
- 
 
 332 
            empty, allocator will attempt to reload magazin, swapping it with
- 
 
 333 
            second CPU magazine and returns to the first step.</para>
- 
 
 334 
 
- 
 
 335 
            <para><emphasis>Step 3.</emphasis> Now we are in the situation
- 
 
 336 
            when both CPU-bound magazines are empty, which makes allocator to
- 
 
 337 
            access shared full-magazines depot to reload CPU-bound magazines.
- 
 
 338 
            If reload is succesful (meaning there are full magazines in depot)
- 
 
 339 
            algoritm continues at Step 1.</para>
- 
 
 340 
 
- 
 
 341 
            <para><emphasis>Step 4.</emphasis> Final step of the allocation.
- 
 
 342 
            In this step object is allocated from the conventional slab layer
- 
 
 343 
            and pointer is returned.</para>
- 
 
 344 
          </formalpara>
- 
 
 345 
 
307 
            <listitem>
 346 
          <formalpara>
- 
 
 347 
            <title>Deallocation</title>
- 
 
 348 
 
- 
 
 349 
            <para><emphasis>Step 1.</emphasis> During deallocation request,
308 
               - it might be good to add granularity of locks even to slab level, we could then try_spinlock over all partial slabs and thus improve scalability even on slab level
 350 
            slab allocator will check if the local CPU-bound magazine is not
- 
 
 351 
            full. In this case we will just push the pointer to this
309 
            </listitem>
 352 
            magazine.</para>
- 
 
 353 
 
- 
 
 354 
            <para><emphasis>Step 2.</emphasis> If the CPU-bound magazine is
- 
 
 355 
            full, allocator will attempt to reload magazin, swapping it with
- 
 
 356 
            second CPU magazine and returns to the first step.</para>
- 
 
 357 
 
- 
 
 358 
            <para><emphasis>Step 3.</emphasis> Now we are in the situation
- 
 
 359 
            when both CPU-bound magazines are full, which makes allocator to
- 
 
 360 
            access shared full-magazines depot to put one of the magazines to
- 
 
 361 
            the depot and creating new empty magazine. Algoritm continues at
310 
          </itemizedlist></para>
 362 
            Step 1.</para>
- 
 
 363 
          </formalpara>
- 
 
 364 
        </section>
311 
      </section>
 365 
      </section>
312 
    </section>
 366 
    </section>
313 
 
 367 
 
314 
    <!-- End of Physmem -->
 368 
    <!-- End of Physmem -->
315 
  </section>
 369 
  </section>
316 
 
 370 
 
317 
  <section>
 371 
  <section>
318 
    <title>Memory sharing</title>
 372 
    <title>Memory sharing</title>
319 
 
 373 
 
320 
    <para>Not implemented yet(?)</para>
 374 
    <para>Not implemented yet(?)</para>
321 
  </section>
 375 
  </section>
322 
</chapter>
 376 
</chapter>