Subversion Repositories HelenOS-doc

Rev

Rev 38 | Rev 45 | Go to most recent revision | Show entire file | Ignore whitespace | Details | Blame | Last modification | View Log | RSS feed

Rev 38 Rev 39
Line 22... Line 22...
22 
 
 22 
 
23 
          <listitem>
 23 
          <listitem>
24 
             Allowing, in general, to load and execute two programs that are linked on the same address without complicated relocations.
 24 
             Allowing, in general, to load and execute two programs that are linked on the same address without complicated relocations.
25 
          </listitem>
 25 
          </listitem>
26 
        </itemizedlist></para>
 26 
        </itemizedlist></para>
27 
   
 - 
 
28 
   
 27 
 
29 
<para><!--
 28 
      <para><!--
30 
 
 29 
 
31 
                TLB shootdown ASID/ASID:PAGE/ALL.
 30 
                TLB shootdown ASID/ASID:PAGE/ALL.
32 
                TLB shootdown requests can come in asynchroniously
 31 
                TLB shootdown requests can come in asynchroniously
33 
                so there is a cache of TLB shootdown requests. Upon cache overflow TLB shootdown ALL is executed
 32 
                so there is a cache of TLB shootdown requests. Upon cache overflow TLB shootdown ALL is executed
34 
 
 33 
 
Line 41... Line 40...
41 
 
 40 
 
42 
--></para>
 41 
--></para>
43 
    </section>
 42 
    </section>
44 
 
 43 
 
45 
    <section>
 44 
    <section>
46 
       
 - 
 
47 
 
 - 
 
48 
      <title>Paging</title>
 45 
      <title>Paging</title>
49 
 
 46 
 
50 
      <para>Virtual memory is usually using paged memory model, where virtual
 47 
      <para>Virtual memory is usually using paged memory model, where virtual
51 
      memory address space is divided into the <emphasis>pages</emphasis>
 48 
      memory address space is divided into the <emphasis>pages</emphasis>
52 
      (usually having size 4096 bytes) and physical memory is divided into the
 49 
      (usually having size 4096 bytes) and physical memory is divided into the
53 
      frames (same sized as a page, of course). Each page may be mapped to some
 50 
      frames (same sized as a page, of course). Each page may be mapped to
54 
      frame and then, upon memory access to the virtual address, CPU performs
 51 
      some frame and then, upon memory access to the virtual address, CPU
55 
      <emphasis>address translation</emphasis> during the instruction
 52 
      performs <emphasis>address translation</emphasis> during the instruction
56 
      execution. Non-existing mapping generates page fault exception, calling
 53 
      execution. Non-existing mapping generates page fault exception, calling
57 
      kernel exception handler, thus allowing kernel to manipulate rules of
 54 
      kernel exception handler, thus allowing kernel to manipulate rules of
58 
      memory access. Information for pages mapping is stored by kernel in the
 55 
      memory access. Information for pages mapping is stored by kernel in the
59 
      <link linkend="page_tables">page tables</link></para>
 56 
      <link linkend="page_tables">page tables</link></para>
60 
 
 57 
 
61 
       
 - 
 
62 
 
 - 
 
63 
      <para>The majority of the architectures use multi-level page tables,
 58 
      <para>The majority of the architectures use multi-level page tables,
64 
      which means need to access physical memory several times before getting
 59 
      which means need to access physical memory several times before getting
65 
      physical address. This fact would make serios performance overhead in
 60 
      physical address. This fact would make serios performance overhead in
66 
      virtual memory management. To avoid this <link linkend="tlb">Traslation
 61 
      virtual memory management. To avoid this <link linkend="tlb">Traslation
67 
      Lookaside Buffer (TLB)</link> is used.</para>
 62 
      Lookaside Buffer (TLB)</link> is used.</para>
68 
 
 63 
 
69 
       
 - 
 
70 
 
 - 
 
71 
      <para>At the moment HelenOS does not support swapping.</para>
 64 
      <para>At the moment HelenOS does not support swapping.</para>
72 
 
 65 
 
- 
 
 66 
      <para>- pouzivame vypadky stranky k alokaci ramcu on-demand v ramci
73 
      <para>- pouzivame vypadky stranky k alokaci ramcu on-demand v ramci as_area - na architekturach, ktere to podporuji, podporujeme non-exec stranky </para>
 67 
      as_area - na architekturach, ktere to podporuji, podporujeme non-exec
- 
 
 68 
      stranky</para>
74 
    </section>
 69 
    </section>
75 
 
 70 
 
76 
    <section>
 71 
    <section>
77 
      <title>Address spaces</title>
 72 
      <title>Address spaces</title>
78 
 
 73 
 
79 
      <section>
 74 
      <section>
80 
        <title>Address spaces and areas</title>
 75 
        <title>Address spaces and areas</title>
81 
 
 76 
 
82 
        <para>
 - 
 
83 
 
 - 
 
84 
    - adresovy prostor se sklada z tzv. address space areas
 77 
        <para>- adresovy prostor se sklada z tzv. address space areas
85 
        usporadanych v B+stromu; tyto areas popisuji vyuzivane casti
 78 
        usporadanych v B+stromu; tyto areas popisuji vyuzivane casti
86 
        adresoveho prostoru patrici do user address space. Kazda cast je dana
 79 
        adresoveho prostoru patrici do user address space. Kazda cast je dana
87 
        svoji bazovou adresou, velikosti a flagy (rwx/dd).
 80 
        svoji bazovou adresou, velikosti a flagy (rwx/dd).</para>
88 
 
 - 
 
89 
    </para>
 - 
 
90 
 
 81 
 
91 
        <para>- uzivatelske thready maji moznost manipulovat se svym adresovym
 82 
        <para>- uzivatelske thready maji moznost manipulovat se svym adresovym
92 
        prostorem (vytvaret/resizovat/sdilet) as_areas pomoci syscallu</para>
 83 
        prostorem (vytvaret/resizovat/sdilet) as_areas pomoci syscallu</para>
93 
      </section>
 84 
      </section>
94 
 
 85 
 
Line 114... Line 105...
114 
        <para>- vztah task ~ adresovy prostor: teoreticky existuje moznost, ze
 105 
        <para>- vztah task ~ adresovy prostor: teoreticky existuje moznost, ze
115 
        je adresovy prostor sdilen vice tasky, avsak tuto moznost nepouzivame
 106 
        je adresovy prostor sdilen vice tasky, avsak tuto moznost nepouzivame
116 
        a neni ani nijak osetrena. Tim padem plati, ze kazdy task ma vlastni
 107 
        a neni ani nijak osetrena. Tim padem plati, ze kazdy task ma vlastni
117 
        adresovy prostor</para>
 108 
        adresovy prostor</para>
118 
      </section>
 109 
      </section>
119 
     
 - 
 
120 
     
 - 
 
121 
     
 - 
 
122 
    </section>
 110 
    </section>
123 
 
 111 
 
124 
    <section>
 112 
    <section>
125 
      <title>Virtual address translation</title>
 113 
      <title>Virtual address translation</title>
126 
 
 114 
 
Line 219... Line 207...
219 
    </section>
 207 
    </section>
220 
 
 208 
 
221 
    <section id="frame_allocator">
 209 
    <section id="frame_allocator">
222 
      <title>Frame allocator</title>
 210 
      <title>Frame allocator</title>
223 
 
 211 
 
- 
 
 212 
      <para><mediaobject id="frame_alloc">
- 
 
 213 
          <imageobject role="html">
- 
 
 214 
            <imagedata fileref="images/frame_alloc.png" format="PNG" />
- 
 
 215 
          </imageobject>
- 
 
 216 
 
- 
 
 217 
          <imageobject role="fop">
- 
 
 218 
            <imagedata fileref="images.vector/frame_alloc.svg" format="SVG" />
- 
 
 219 
          </imageobject>
- 
 
 220 
        </mediaobject></para>
- 
 
 221 
 
224 
      <formalpara>
 222 
      <formalpara>
225 
        <title>Overview</title>
 223 
        <title>Overview</title>
226 
 
 224 
 
227 
        <para>Frame allocator provides physical memory allocation for the
 225 
        <para>Frame allocator provides physical memory allocation for the
228 
        kernel. Because of zonal organization of physical memory, frame
 226 
        kernel. Because of zonal organization of physical memory, frame
Line 260... Line 258...
260 
      <title>Buddy allocator</title>
 258 
      <title>Buddy allocator</title>
261 
 
 259 
 
262 
      <section>
 260 
      <section>
263 
        <title>Overview</title>
 261 
        <title>Overview</title>
264 
 
 262 
 
- 
 
 263 
        <para><mediaobject id="buddy_alloc">
- 
 
 264 
            <imageobject role="html">
- 
 
 265 
              <imagedata fileref="images/buddy_alloc.png" format="PNG" />
- 
 
 266 
            </imageobject>
- 
 
 267 
 
- 
 
 268 
            <imageobject role="fop">
- 
 
 269 
              <imagedata fileref="images.vector/buddy_alloc.svg" format="SVG" />
- 
 
 270 
            </imageobject>
- 
 
 271 
          </mediaobject></para>
- 
 
 272 
 
265 
        <para>In buddy allocator, memory is broken down into power-of-two
 273 
        <para>In buddy allocator, memory is broken down into power-of-two
266 
        sized naturally aligned blocks. These blocks are organized in an array
 274 
        sized naturally aligned blocks. These blocks are organized in an array
267 
        of lists in which list with index i contains all unallocated blocks of
 275 
        of lists in which list with index i contains all unallocated blocks of
268 
        the size <mathphrase>2<superscript>i</superscript></mathphrase>. The
 276 
        the size <mathphrase>2<superscript>i</superscript></mathphrase>. The
269 
        index i is called the order of block. Should there be two adjacent
 277 
        index i is called the order of block. Should there be two adjacent
Line 299... Line 307...
299 
        lack of this knowledge, the buddy allocator exports an interface that
 307 
        lack of this knowledge, the buddy allocator exports an interface that
300 
        each of its clients is required to implement. When supplied an
 308 
        each of its clients is required to implement. When supplied an
301 
        implementation of this interface, the buddy allocator can use
 309 
        implementation of this interface, the buddy allocator can use
302 
        specialized external functions to find buddy for a block, split and
 310 
        specialized external functions to find buddy for a block, split and
303 
        coalesce blocks, manipulate block order and mark blocks busy or
 311 
        coalesce blocks, manipulate block order and mark blocks busy or
304 
        available. For precize documentation of this interface, refer to <link
 312 
        available. For precize documentation of this interface, refer to
305 
        linkend="???">HelenOS Generic Kernel Reference Manual</link>.</para>
 313 
        <emphasis>"HelenOS Generic Kernel Reference Manual"</emphasis>.</para>
306 
 
 314 
 
307 
        <formalpara>
 315 
        <formalpara>
308 
          <title>Data organization</title>
 316 
          <title>Data organization</title>
309 
 
 317 
 
310 
          <para>Each entity allocable by the buddy allocator is required to
 318 
          <para>Each entity allocable by the buddy allocator is required to
Line 392... Line 400...
392 
      </section>
 400 
      </section>
393 
 
 401 
 
394 
      <section>
 402 
      <section>
395 
        <title>Implementation</title>
 403 
        <title>Implementation</title>
396 
 
 404 
 
- 
 
 405 
        <para><mediaobject id="slab_alloc">
- 
 
 406 
            <imageobject role="html">
- 
 
 407 
              <imagedata fileref="images/slab_alloc.png" format="PNG" />
- 
 
 408 
            </imageobject>
- 
 
 409 
 
- 
 
 410 
            <imageobject role="fop">
- 
 
 411 
              <imagedata fileref="images.vector/slab_alloc.svg" format="SVG" />
- 
 
 412 
            </imageobject>
- 
 
 413 
          </mediaobject></para>
- 
 
 414 
 
397 
        <para>The SLAB allocator is closely modelled after <ulink
 415 
        <para>The SLAB allocator is closely modelled after <ulink
398 
        url="http://www.usenix.org/events/usenix01/full_papers/bonwick/bonwick_html/">
 416 
        url="http://www.usenix.org/events/usenix01/full_papers/bonwick/bonwick_html/">
399 
        OpenSolaris SLAB allocator by Jeff Bonwick and Jonathan Adams </ulink>
 417 
        OpenSolaris SLAB allocator by Jeff Bonwick and Jonathan Adams </ulink>
400 
        with the following exceptions: <itemizedlist>
 418 
        with the following exceptions: <itemizedlist>
401 
            <listitem>
 419 
            <listitem>