Rev 37 | Rev 39 | Go to most recent revision | Show entire file | Ignore whitespace | Details | Blame | Last modification | View Log | RSS feed
Rev 37 | Rev 38 | ||
---|---|---|---|
Line 16... | Line 16... | ||
16 | </mediaobject> |
16 | </mediaobject> |
17 | </para> |
17 | </para> |
18 | </section> |
18 | </section> |
19 | 19 | ||
20 | <section> |
20 | <section> |
- | 21 | <title>Kernel primitives</title> |
|
- | 22 | <para> |
|
- | 23 | <termdef><glossterm>Thread</glossterm> is the basic execution primitive.</termdef> |
|
- | 24 | </para> |
|
- | 25 | ||
- | 26 | <para> |
|
- | 27 | <termdef><glossterm>Thread context</glossterm> represents state of the <emphasis>thread</emphasis>. |
|
- | 28 | Thread context is built of the context registers contents, FPU state and the stack.</termdef> |
|
- | 29 | </para> |
|
- | 30 | ||
- | 31 | <para> |
|
- | 32 | <termdef> |
|
- | 33 | <glossterm>Task</glossterm> is a multi-purpose entity, serving to |
|
- | 34 | <itemizedlist> |
|
- | 35 | <listitem>incorporate set if its threads</listitem> |
|
- | 36 | <listitem>provide common address space to its threads</listitem> |
|
- | 37 | <listitem>be an end-point in IPC</listitem> |
|
- | 38 | </itemizedlist> |
|
- | 39 | ||
- | 40 | </termdef> |
|
- | 41 | </para> |
|
- | 42 | ||
- | 43 | <para> |
|
- | 44 | <termdef> |
|
- | 45 | <glossterm>Address space area</glossterm> is a mutually disjunctive range of memory with the code, stack and data. |
|
- | 46 | </termdef> |
|
- | 47 | </para> |
|
- | 48 | ||
- | 49 | <para> |
|
- | 50 | <termdef> |
|
- | 51 | <glossterm>Address space</glossterm> is a aggregating entity for address space areas, connecting them to the task. |
|
- | 52 | </termdef> |
|
- | 53 | </para> |
|
- | 54 | </section> |
|
- | 55 | ||
- | 56 | <section> |
|
21 | <title>Monolithic microkernel</title> |
57 | <title>Monolithic microkernel</title> |
22 | <para> |
58 | <para> |
23 | Though HelenOS was initially planned as a microkernel, we where trying to avoid several issues, connected |
59 | Though HelenOS was initially planned as a microkernel, we were trying to avoid several issues, connected |
24 | with microkernels, such as much higher overhead during memory management and hardware operations. For this reason |
60 | with microkernels, such as much higher overhead during memory management and hardware operations. For this reason |
25 | some of the subsytems, that are to be implmented as servers in classic microkernel design, where implemented |
61 | some of the subsystems, that are to be implemented as servers in classic microkernel design, were implemented |
26 | as a part of kernel, thus minimizing this overhead. |
62 | as a part of kernel, thus minimizing this overhead. |
27 | </para> |
63 | </para> |
28 | 64 | ||
29 | <formalpara> |
65 | <formalpara> |
30 | <title>Memory management</title> |
66 | <title>Memory management</title> |
Line 52... | Line 88... | ||
52 | </formalpara> |
88 | </formalpara> |
53 | 89 | ||
54 | </section> |
90 | </section> |
55 | 91 | ||
56 | <section> |
92 | <section> |
57 | <title>Kernel primitives</title> |
- | |
58 | <para> |
- | |
59 | <termdef><glossterm>Thread</glossterm> is the basic execution primitive.</termdef> |
- | |
60 | </para> |
- | |
61 | - | ||
62 | <para> |
- | |
63 | <termdef><glossterm>Thread context</glossterm> represents state of the <emphasis>thread</emphasis>. Contains registers state, FPU state, stack(s).</termdef> |
- | |
64 | </para> |
- | |
65 | - | ||
66 | <para> |
- | |
67 | <termdef><glossterm>Task</glossterm> </termdef> |
- | |
68 | </para> |
- | |
69 | - | ||
70 | - | ||
71 | <para> |
- | |
72 | <termdef> |
- | |
73 | <glossterm>Address space area</glossterm> is a mutually disjunctive range of memory with the code, stack and data. |
- | |
74 | </termdef> |
- | |
75 | </para> |
- | |
76 | - | ||
77 | <para> |
- | |
78 | <termdef> |
- | |
79 | <glossterm>Address space</glossterm> is a aggregating entity for address space areas, connecting them to the task. |
- | |
80 | </termdef> |
- | |
81 | </para> 1 |
- | |
82 | </section> |
- | |
83 | - | ||
84 | - | ||
85 | - | ||
86 | <section> |
- | |
87 | <title>IPC</title> |
93 | <title>IPC</title> |
88 | 94 | ||
89 | <para> |
- | |
90 | Due to the high intertask communication traffic, IPC becomes critical subsystem for microkernels, putting high demands on the |
- | |
91 | speed, latency and reliability of IPC model and implementation. HelenOS IPC is designed in analogy with telephone communication. |
95 | <para>HelenOS IPC is designed in analogy with telephone communication. |
92 | Each task has an <emphasis>answerbox</emphasis> and a set of <emphasis>phones</emphasis> to call another tasks' answerboxes. Communication |
96 | Each task has an <emphasis>answerbox</emphasis> and a set of <emphasis>phones</emphasis> to call another tasks' answerboxes. |
- | 97 | </para> |
|
- | 98 | |
|
- | 99 | <para>Communication |
|
93 | is possible after the link establishment, and can be either <emphasis>asynchronious</emphasis> or <emphasis>synchronious</emphasis>. |
100 | is possible after the connection is established, and can be either <emphasis>asynchronious</emphasis> or <emphasis>synchronious</emphasis>. |
94 | </para> |
101 | </para> |
95 | 102 | |
|
96 | </section> |
103 | </section> |
97 | 104 | ||
98 | - | ||
99 | <section> |
105 | <section> |
100 | <para>Task ID - unique 64 bit number. Used for syscalls.</para> |
- | |
101 | <para>Contains threads</para> |
106 | <title>Functionality model</title> |
102 | <para>Address space is created per task</para> |
- | |
103 | <para>Memory mapping is per task</para> |
- | |
104 | <para>Context per thread. (Note 2 stacks on IA64).</para> |
- | |
105 | <para>IPC answer box associated per task</para> |
- | |
106 | 107 | ||
107 | - | ||
108 | <title>Memory management</title> |
- | |
109 | <para> |
108 | <para> |
110 | Zones - linked list (not many zones, so we can afford it. Can be replaced with B-tree in the future) |
109 | As you know, microkernel design is very simple, just enough to provide communication facility for tasks. Most of the OS functionality |
- | 110 | is performed by server tasks, that are running in userspace. |
|
111 | Number of zones depends on HW tables. Describe zone allocation/deallocation algoritm |
111 | Thus most of the system calls in monolithic kernels, are the IPC calls on server tasks in microkernels. |
112 | </para> |
112 | </para> |
113 | <para>Page tables. 4 level hierarchical and hash directly supported. B+ Tree can be implemented.</para> |
- | |
114 | <para>For paging there is an abstract layer</para> |
- | |
115 | <para>TLB shootdown implementation (update TLB upon mapping update/remove). |
- | |
116 | TLB shootdown ASID/ASID:PAGE/ALL. |
- | |
117 | TLB shootdown requests can come in asynchroniously |
- | |
118 | so there is a cache of TLB shootdown requests. Upon cache overflow TLB shootdown ALL is executed</para> |
- | |
119 | |
113 | |
120 | <para> |
114 | <para> |
121 | Address spaces. Address space area (B+ tree). Only for uspace. Set of syscalls (shrink/extend etc). |
115 | Moreover, problems experience the device drivers. Running in the user space, device driver still needs to recieve interrupts |
122 | Special address space area type - device - prohibits shrink/extend syscalls to call on it. |
116 | and access hardware directly. |
123 | Address space has link to mapping tables (hierarchical - per Address space, hash - global tables). |
- | |
124 | </para> |
117 | </para> |
- | 118 | |
|
- | 119 | <para> |
|
- | 120 | This raises two major problems in microkernels: |
|
- | 121 | |
|
- | 122 | <orderedlist numeration="loweralpha"> |
|
- | 123 | <listitem>What is the recipient address of the server (e.g. "memory manager" or a specific device driver) ?</listitem> |
|
- | 124 | <listitem>How this server task is going to access hardware or kernel while running in the user mode?</listitem> |
|
- | 125 | </orderedlist> |
|
- | 126 | |
|
- | 127 | </para> |
|
- | 128 | |
|
- | 129 | |
|
- | 130 | <formalpara id="intro_ns"> |
|
- | 131 | <title>Name server</title> |
|
- | 132 | <para>As every microkernel, HelenOS has a "Name server" task with "well known" IPC address, that connects user task to any server |
|
- | 133 | just by the string service indentification.</para> |
|
- | 134 | </formalpara> |
|
- | 135 | |
|
- | 136 | |
|
- | 137 | <formalpara id="intro_ddi"> |
|
- | 138 | <title>Device driver interface</title> |
|
- | 139 | <para>Device drivers use special syscalls to map physical memory areas into their address space, to map port regions (mostly ia32). |
|
- | 140 | Interrupts are delivered to the device driver task by the standard IPC means. |
|
- | 141 | </para> |
|
- | 142 | </formalpara> |
|
- | 143 | |
|
125 | </section> |
144 | </section> |
- | 145 | ||
- | 146 | |
|
- | 147 | |
|
- | 148 | ||
126 | </chapter> |
149 | </chapter> |
127 | 150 |