[btr] Add more documentation.
[paraslash.git] / buffer_tree.h
1 /**
2 * Buffer trees and buffer tree nodes.
3 *
4 * The buffer tree API offers a more powerful method than standard unix pipes
5 * for managing the data flow from the producer of the data (e.g. the network
6 * receiver) to its consumer(s) (e.g. a sound card).
7 *
8 * Each data buffer starts its way from the root of the buffer tree. At each
9 * child node the data is investigated and new data is fed to each child.
10 * Everything happens within one single-treaded process. There are no file
11 * descriptors, no calls to read() or write().
12 *
13 * A buffer tree consists of buffer tree nodes linked via certain parent/child
14 * relationships.
15 *
16 * Whenever a node in the buffer tree creates output, either by creating a new
17 * buffer or by pushing down buffers received from its parent, references to
18 * that buffer are created for all children of the node. The buffer tree code
19 * tries hard to avoid to copy buffer contents, but is forced to do so in case
20 * there are alignment constraints.
21 *
22 * Communication between nodes is possible via the btr_exec_up() function.
23 * For example, in para_audiod the alsa writer asks all parent nodes
24 * for for the number of channels and the sample rate of the current
25 * audio file.
26 */
27 struct btr_node;
28
29 /**
30 * Buffer pools - An alternative to malloc/free buffer management.
31 *
32 * Non-leaf nodes usually create output to be processed by their children. The
33 * data must be fed through the output channel(s) of the node in order to make
34 * that data available to each child.
35 *
36 * The easiest way to do so is to malloc() a buffer, fill it, and then call
37 * btr_add_output(). This adds references to that buffer to all children. The
38 * buffer is automatically freed if no buffer tree node is using it any more.
39 *
40 * This approach, while simple, has some drawbacks, especially affecting the
41 * root nodes of the buffer tree. Often the data source which is represented by
42 * a root node does not know in advance how much data will be available.
43 * Therefore the allocated buffer is either larger than what can currently be
44 * read, or is too small so that multiple buffers have to be used.
45 *
46 * While this could be worked around by using a large buffer and calling
47 * realloc() afterwards to shrink the buffer according to how much has been
48 * read, there is a second problem which comes from the alignment constraints
49 * of some filters, mainly the decoders. These need a minimal amount of data to
50 * proceed, and most of them even need this amount as one contiguous buffer,
51 * i.e. not spread out over two or more buffers.
52 *
53 * Although the buffer tree code handles this case just fine, it can be
54 * expensive because two or more buffers must be combined by copying buffer
55 * contents around in order to satisfy the constraint.
56 *
57 * This is where buffer pools come into play. Buffer pools try to satisfy
58 * alignment constraints without copying buffer content whenever possible. To
59 * avoid spreading out the input data over the address space like in the
60 * malloc/free approach, a fixed large contiguous buffer (the area) is used
61 * instead. A buffer pool consists basically of an area and two pointers, the
62 * read head and the write head.
63 *
64 * Once a buffer pool has been created, its node, e.g. a receiver, obtains the
65 * current value of the write head and writes new data to this location. Then
66 * it calls btr_add_output_pool() to tell much data it has written. This
67 * advances the write head accordingly, and it also creates references to the
68 * newly written part of the area for the children of the node to consume.
69 *
70 * Child nodes consume data by working through their input queue, which is a
71 * list of buffer references. Once the content of a buffer is no longer needed
72 * by a child node, the child calls btr_consume() to indicate the amount of
73 * data which can be dropped from the child's point of view. If no reference
74 * to some region of the buffer pool area remains, the read head of the buffer
75 * pool advances, making space available for the receiver node to fill.
76 *
77 * No matter if malloc() or a buffer pool is used, the buffer tree code takes
78 * care of alignment constraints imposed by the consumers. In the buffer pool
79 * case, automatic merging of references to contiguous buffers is performed.
80 * memcpy is only used if a constraint can not be satisfied by using the
81 * remaining part of the area only. This only happens when the end of the area
82 * is reached.
83 */
84 struct btr_pool;
85
86 /**
87 * The three different types of buffer tree nodes.
88 *
89 * Usually, there is exactly one node in the buffer tree, the root node, which
90 * has no parent. Every node different from the root node has exactly one
91 * parent. The root node represents a data source. Root nodes are thus used by
92 * the receivers of paraslash. Also, reading from stdin is realized as the root
93 * node of a buffer tree.
94 *
95 * Each node may have arbitrary many children, including none. Nodes with no
96 * children are called leaf nodes. They represent a data sink, like the alsa or
97 * the file writer.
98 *
99 * Hence there are three different types of buffer tree nodes: The root node
100 * and the leaf nodes and nodes which have both a parent and at least one
101 * child. Such a node is called an internal node.
102 *
103 * Internal nodes represent filters through which data buffers flow, possibly
104 * while being altered on their way to the children of the node. Examples of
105 * internal nodes are audio file decoders (mp3dec, oggdec, ...), but also the
106 * check for a wav header is implemented as an internal buffer tree node.
107 */
108 enum btr_node_type {
109 /* This node has no parent. */
110 BTR_NT_ROOT,
111 /* Node has parent and at least one child. */
112 BTR_NT_INTERNAL,
113 /* Node has no children. */
114 BTR_NT_LEAF,
115 };
116
117 /**
118 * Per node handler used for inter node communication.
119 *
120 * Each node in the buffer tree may optionally provide a command handler for
121 * execution of commands by other nodes of the tree.
122 *
123 * It is dependent on the node in question which commands are supported and how
124 * they work. In any case, the input for the command handler is some string and
125 * its output is also a string which is returned via the \a result pointer of
126 * the handler.
127 *
128 * This mechanism is used in para_audiod e.g. by the alsa writer which needs to
129 * know the sample rate of its input known to e.g. the mp3dec node further up
130 * in the buffer tree.
131 */
132 typedef int (*btr_command_handler)(struct btr_node *btrn,
133 const char *command, char **result);
134
135 /**
136 * Structure for creating new buffer tree nodes.
137 *
138 * btr_new_node() takes a pointer to such a structure.
139 *
140 * There are four different combinations of \a parent and child:
141 *
142 * 1. both \p NULL. This creates a new buffer tree with a single isolated node.
143 *
144 * 2. \a parent != \p NULL, \a child == NULL. This creates a new leaf node by
145 * adding the new node to the list of children of the given parent node.
146 *
147 * 3. \a parent == NULL, \a child != NULL. The new node becomes the new root of
148 * the buffer tree. The child must be old root.
149 *
150 * 4. both != NULL. This creates a new internal node. \a child must be child of
151 * p. This mode of operation is currently not needed and is thus not yet
152 * implemented.
153 */
154 struct btr_node_description {
155 /** Name of the new node. */
156 const char *name;
157 /** Parent of the new node. */
158 struct btr_node *parent;
159 /** Child of the new node. */
160 struct btr_node *child;
161 /** Used for inter node communication. Optional. */
162 btr_command_handler handler;
163 /** Points usually to the struct that contains the node pointer. */
164 void *context;
165 };
166
167 size_t btr_pool_size(struct btr_pool *btrp);
168 struct btr_pool *btr_pool_new(const char *name, size_t area_size);
169 void btr_pool_free(struct btr_pool *btrp);
170 size_t btr_pool_get_buffer(struct btr_pool *btrp, char **result);
171 void btr_add_output_pool(struct btr_pool *btrp, size_t size,
172 struct btr_node *btrn);
173 size_t btr_pool_unused(struct btr_pool *btrp);
174 void btr_copy(const void *src, size_t n, struct btr_pool *btrp,
175 struct btr_node *btrn);
176
177 struct btr_node *btr_new_node(struct btr_node_description *bnd);
178 void btr_remove_node(struct btr_node *btrn);
179 void btr_free_node(struct btr_node *btrn);
180 void btr_add_output(char *buf, size_t size, struct btr_node *btrn);
181 bool btr_no_children(struct btr_node *btrn);
182 size_t btr_bytes_pending(struct btr_node *btrn);
183 size_t btr_get_input_queue_size(struct btr_node *btrn);
184 bool btr_no_parent(struct btr_node *btrn);
185 size_t btr_next_buffer(struct btr_node *btrn, char **bufp);
186 void btr_consume(struct btr_node *btrn, size_t numbytes);
187 int btr_exec(struct btr_node *btrn, const char *command, char **value_result);
188 int btr_exec_up(struct btr_node *btrn, const char *command, char **value_result);
189 void btr_splice_out_node(struct btr_node *btrn);
190 void btr_pushdown(struct btr_node *btrn);
191 void *btr_context(struct btr_node *btrn);
192 void btr_merge(struct btr_node *btrn, size_t dest_size);
193 bool btr_eof(struct btr_node *btrn);
194 void btr_log_tree(struct btr_node *btrn, int loglevel);
195 int btr_pushdown_one(struct btr_node *btrn);
196 bool btr_inplace_ok(struct btr_node *btrn);
197 int btr_node_status(struct btr_node *btrn, size_t min_iqs,
198 enum btr_node_type type);
199 void btr_get_node_start(struct btr_node *btrn, struct timeval *tv);
200 struct btr_node *btr_search_node(const char *name, struct btr_node *root);