1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
|
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
"http://www.w3.org/TR/html4/strict.dtd">
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
<title>LLVM Bitcode File Format</title>
<link rel="stylesheet" href="llvm.css" type="text/css">
</head>
<body>
<div class="doc_title"> LLVM Bitcode File Format </div>
<ol>
<li><a href="#abstract">Abstract</a></li>
<li><a href="#overview">Overview</a></li>
<li><a href="#bitstream">Bitstream Format</a>
<ol>
<li><a href="#magic">Magic Numbers</a></li>
<li><a href="#primitives">Primitives</a></li>
<li><a href="#abbrevid">Abbreviation IDs</a></li>
<li><a href="#blocks">Blocks</a></li>
<li><a href="#datarecord">Data Records</a></li>
<li><a href="#abbreviations">Abbreviations</a></li>
</ol>
</li>
<li><a href="#llvmir">LLVM IR Encoding</a></li>
</ol>
<div class="doc_author">
<p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>.
</p>
</div>
<!-- *********************************************************************** -->
<div class="doc_section"> <a name="abstract">Abstract</a></div>
<!-- *********************************************************************** -->
<div class="doc_text">
<p>This document describes the LLVM bitstream file format and the encoding of
the LLVM IR into it.</p>
</div>
<!-- *********************************************************************** -->
<div class="doc_section"> <a name="overview">Overview</a></div>
<!-- *********************************************************************** -->
<div class="doc_text">
<p>
What is commonly known as the LLVM bitcode file format (also, sometimes
anachronistically known as bytecode) is actually two things: a <a
href="#bitstream">bitstream container format</a>
and an <a href="#llvmir">encoding of LLVM IR</a> into the container format.</p>
<p>
The bitstream format is an abstract encoding of structured data, like very
similar to XML in some ways. Like XML, bitstream files contain tags, and nested
structures, and you can parse the file without having to understand the tags.
Unlike XML, the bitstream format is a binary encoding, and unlike XML it
provides a mechanism for the file to self-describe "abbreviations", which are
effectively size optimizations for the content.</p>
<p>This document first describes the LLVM bitstream format, then describes the
record structure used by LLVM IR files.
</p>
</div>
<!-- *********************************************************************** -->
<div class="doc_section"> <a name="bitstream">Bitstream Format</a></div>
<!-- *********************************************************************** -->
<div class="doc_text">
<p>
The bitstream format is literally a stream of bits, with a very simple
structure. This structure consists of the following concepts:
</p>
<ul>
<li>A "<a href="#magic">magic number</a>" that identifies the contents of
the stream.</li>
<li>Encoding <a href="#primitives">primitives</a> like variable bit-rate
integers.</li>
<li><a href="#blocks">Blocks</a>, which define nested content.</li>
<li><a href="#datarecord">Data Records</a>, which describe entities within the
file.</li>
<li>Abbreviations, which specify compression optimizations for the file.</li>
</ul>
<p>Note that the <a
href="CommandGuide/html/llvm-bcanalyzer.html">llvm-bcanalyzer</a> tool can be
used to dump and inspect arbitrary bitstreams, which is very useful for
understanding the encoding.</p>
</div>
<!-- ======================================================================= -->
<div class="doc_subsection"><a name="magic">Magic Numbers</a>
</div>
<div class="doc_text">
<p>The first four bytes of the stream identify the encoding of the file. This
is used by a reader to know what is contained in the file.</p>
</div>
<!-- ======================================================================= -->
<div class="doc_subsection"><a name="primitives">Primitives</a>
</div>
<div class="doc_text">
<p>
A bitstream literally consists of a stream of bits. This stream is made up of a
number of primitive values that encode a stream of integer values. These
integers are are encoded in two ways: either as <a href="#fixedwidth">Fixed
Width Integers</a> or as <a href="#variablewidth">Variable Width
Integers</a>.
</p>
</div>
<!-- _______________________________________________________________________ -->
<div class="doc_subsubsection"> <a name="fixedwidth">Fixed Width Integers</a>
</div>
<div class="doc_text">
<p>Fixed-width integer values have their low bits emitted directly to the file.
For example, a 3-bit integer value encodes 1 as 001. Fixed width integers
are used when there are a well-known number of options for a field. For
example, boolean values are usually encoded with a 1-bit wide integer.
</p>
</div>
<!-- _______________________________________________________________________ -->
<div class="doc_subsubsection"> <a name="variablewidth">Variable Width
Integers</a></div>
<div class="doc_text">
<p>Variable-width integer (VBR) values encode values of arbitrary size,
optimizing for the case where the values are small. Given a 4-bit VBR field,
any 3-bit value (0 through 7) is encoded directly, with the high bit set to
zero. Values larger than N-1 bits emit their bits in a series of N-1 bit
chunks, where all but the last set the high bit.</p>
<p>For example, the value 27 (0x1B) is encoded as 1011 0011 when emitted as a
vbr4 value. The first set of four bits indicates the value 3 (011) with a
continuation piece (indicated by a high bit of 1). The next word indicates a
value of 24 (011 << 3) with no continuation. The sum (3+24) yields the value
27.
</p>
</div>
<!-- _______________________________________________________________________ -->
<div class="doc_subsubsection"> <a name="char6">6-bit characters</a></div>
<div class="doc_text">
<p>6-bit characters encode common characters into a fixed 6-bit field. They
represent the following characters with the following 6-bit values:<s/p>
<ul>
<li>'a' .. 'z' - 0 .. 25</li>
<li>'A' .. 'Z' - 26 .. 52</li>
<li>'0' .. '9' - 53 .. 61</li>
<li>'.' - 62</li>
<li>'_' - 63</li>
</ul>
<p>This encoding is only suitable for encoding characters and strings that
consist only of the above characters. It is completely incapable of encoding
characters not in the set.</p>
</div>
<!-- _______________________________________________________________________ -->
<div class="doc_subsubsection"> <a name="wordalign">Word Alignment</a></div>
<div class="doc_text">
<p>Occasionally, it is useful to emit zero bits until the bitstream is a
multiple of 32 bits. This ensures that the bit position in the stream can be
represented as a multiple of 32-bit words.</p>
</div>
<!-- ======================================================================= -->
<div class="doc_subsection"><a name="abbrevid">Abbreviation IDs</a>
</div>
<div class="doc_text">
<p>
A bitstream is a sequential series of <a href="#blocks">Blocks</a> and
<a href="#datarecord">Data Records</a>. Both of these start with an
abbreviation ID encoded as a fixed-bitwidth field. The width is specified by
the current block, as described below. The value of the abbreviation ID
specifies either a builtin ID (which have special meanings, defined below) or
one of the abbreviation IDs defined by the stream itself.
</p>
<p>
The set of builtin abbrev IDs is:
</p>
<ul>
<li>0 - <a href="#END_BLOCK">END_BLOCK</a> - This abbrev ID marks the end of the
current block.</li>
<li>1 - <a href="#ENTER_SUBBLOCK">ENTER_SUBBLOCK</a> - This abbrev ID marks the
beginning of a new block.</li>
<li>2 - <a href="#DEFINE_ABBREV">DEFINE_ABBREV</a> - This defines a new
abbreviation.</li>
<li>3 - <a href="#UNABBREV_RECORD">UNABBREV_RECORD</a> - This ID specifies the
definition of an unabbreviated record.</li>
</ul>
<p>Abbreviation IDs 4 and above are defined by the stream itself, and specify
an <a href="#abbrev_records">abbreviated record encoding</a>.</p>
</div>
<!-- ======================================================================= -->
<div class="doc_subsection"><a name="blocks">Blocks</a>
</div>
<div class="doc_text">
<p>
Blocks in a bitstream denote nested regions of the stream, and are identified by
a content-specific id number (for example, LLVM IR uses an ID of 12 to represent
function bodies). Nested blocks capture the hierachical structure of the data
encoded in it, and various properties are associated with blocks as the file is
parsed. Block definitions allow the reader to efficiently skip blocks
in constant time if the reader wants a summary of blocks, or if it wants to
efficiently skip data they do not understand. The LLVM IR reader uses this
mechanism to skip function bodies, lazily reading them on demand.
</p>
<p>
When reading and encoding the stream, several properties are maintained for the
block. In particular, each block maintains:
</p>
<ol>
<li>A current abbrev id width. This value starts at 2, and is set every time a
block record is entered. The block entry specifies the abbrev id width for
the body of the block.</li>
<li>A set of abbreviations. Abbreviations may be defined within a block, or
they may be associated with all blocks of a particular ID.
</li>
</ol>
<p>As sub blocks are entered, these properties are saved and the new sub-block
has its own set of abbreviations, and its own abbrev id width. When a sub-block
is popped, the saved values are restored.</p>
</div>
<!-- _______________________________________________________________________ -->
<div class="doc_subsubsection"> <a name="ENTER_SUBBLOCK">ENTER_SUBBLOCK
Encoding</a></div>
<div class="doc_text">
<p><tt>[ENTER_SUBBLOCK, blockid<sub>vbr8</sub>, newabbrevlen<sub>vbr4</sub>,
<align32bits>, blocklen<sub>32</sub>]</tt></p>
<p>
The ENTER_SUBBLOCK abbreviation ID specifies the start of a new block record.
The <tt>blockid</tt> value is encoded as a 8-bit VBR identifier, and indicates
the type of block being entered (which is application specific). The
<tt>newabbrevlen</tt> value is a 4-bit VBR which specifies the
abbrev id width for the sub-block. The <tt>blocklen</tt> is a 32-bit aligned
value that specifies the size of the subblock, in 32-bit words. This value
allows the reader to skip over the entire block in one jump.
</p>
</div>
<!-- _______________________________________________________________________ -->
<div class="doc_subsubsection"> <a name="END_BLOCK">END_BLOCK
Encoding</a></div>
<div class="doc_text">
<p><tt>[END_BLOCK, <align32bits>]</tt></p>
<p>
The END_BLOCK abbreviation ID specifies the end of the current block record.
Its end is aligned to 32-bits to ensure that the size of the block is an even
multiple of 32-bits.</p>
</div>
<!-- ======================================================================= -->
<div class="doc_subsection"><a name="datarecord">Data Records</a>
</div>
<div class="doc_text">
<p>
Data records consist of a record code and a number of (up to) 64-bit integer
values. The interpretation of the code and values is application specific and
there are multiple different ways to encode a record (with an unabbrev record
or with an abbreviation). In the LLVM IR format, for example, there is a record
which encodes the target triple of a module. The code is MODULE_CODE_TRIPLE,
and the values of the record are the ascii codes for the characters in the
string.</p>
</div>
<!-- _______________________________________________________________________ -->
<div class="doc_subsubsection"> <a name="UNABBREV_RECORD">UNABBREV_RECORD
Encoding</a></div>
<div class="doc_text">
<p><tt>[UNABBREV_RECORD, code<sub>vbr6</sub>, numops<sub>vbr6</sub>,
op0<sub>vbr6</sub>, op1<sub>vbr6</sub>, ...]</tt></p>
<p>An UNABBREV_RECORD provides a default fallback encoding, which is both
completely general and also extremely inefficient. It can describe an arbitrary
record, by emitting the code and operands as vbrs.</p>
<p>For example, emitting an LLVM IR target triple as an unabbreviated record
requires emitting the UNABBREV_RECORD abbrevid, a vbr6 for the
MODULE_CODE_TRIPLE code, a vbr6 for the length of the string (which is equal to
the number of operands), and a vbr6 for each character. Since there are no
letters with value less than 32, each letter would need to be emitted as at
least a two-part VBR, which means that each letter would require at least 12
bits. This is not an efficient encoding, but it is fully general.</p>
</div>
<!-- _______________________________________________________________________ -->
<div class="doc_subsubsection"> <a name="abbrev_records">Abbreviated Record
Encoding</a></div>
<div class="doc_text">
<p><tt>[<abbrevid>, fields...]</tt></p>
<p>An abbreviated record is a abbreviation id followed by a set of fields that
are encoded according to the <a href="#abbreviations">abbreviation
definition</a>. This allows records to be encoded significantly more densely
than records encoded with the <a href="#UNABBREV_RECORD">UNABBREV_RECORD</a>
type, and allows the abbreviation types to be specified in the stream itself,
which allows the files to be completely self describing. The actual encoding
of abbreviations is defined below.
</p>
</div>
<!-- ======================================================================= -->
<div class="doc_subsection"><a name="abbreviations">Abbreviations</a>
</div>
<div class="doc_text">
<p>
Abbreviations are an important form of compression for bitstreams. The idea is
to specify a dense encoding for a class of records once, then use that encoding
to emit many records. It takes space to emit the encoding into the file, but
the space is recouped (hopefully plus some) when the records that use it are
emitted.
</p>
<p>
Abbreviations can be determined dynamically per client, per file. Since the
abbreviations are stored in the bitstream itself, different streams of the same
format can contain different sets of abbreviations if the specific stream does
not need it. As a concrete example, LLVM IR files usually emit an abbreviation
for binary operators. If a specific LLVM module contained no or few binary
operators, the abbreviation does not need to be emitted.
</p>
</div>
<!-- _______________________________________________________________________ -->
<div class="doc_subsubsection"><a name="DEFINE_ABBREV">DEFINE_ABBREV
Encoding</a></div>
<div class="doc_text">
<p><tt>[DEFINE_ABBREV, numabbrevops<sub>vbr5</sub>, abbrevop0, abbrevop1,
...]</tt></p>
<p>An abbreviation definition consists of the DEFINE_ABBREV abbrevid followed
by a VBR that specifies the number of abbrev operands, then the abbrev
operands themselves. Abbreviation operands come in three forms. They all start
with a single bit that indicates whether the abbrev operand is a literal operand
(when the bit is 1) or an encoding operand (when the bit is 0).</p>
<ol>
<li>Literal operands - <tt>[1<sub>1</sub>, litvalue<sub>vbr8</sub>]</tt> -
Literal operands specify that the value in the result
is always a single specific value. This specific value is emitted as a vbr8
after the bit indicating that it is a literal operand.</li>
<li>Encoding info without data - <tt>[0<sub>1</sub>, encoding<sub>3</sub>]</tt>
- blah
</li>
<li>Encoding info with data - <tt>[0<sub>1</sub>, encoding<sub>3</sub>,
value<sub>vbr5</sub>]</tt> -
</li>
</ol>
</div>
<!-- *********************************************************************** -->
<div class="doc_section"> <a name="llvmir">LLVM IR Encoding</a></div>
<!-- *********************************************************************** -->
<div class="doc_text">
<p></p>
</div>
<!-- *********************************************************************** -->
<hr>
<address> <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
<a href="http://validator.w3.org/check/referer"><img
src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!"></a>
<a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
<a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
Last modified: $Date$
</address>
</body>
</html>
|
