Commit | Line | Data |
---|---|---|
53f60086 RS |
1 | @c -*-texinfo-*- |
2 | @c This is part of the GNU Emacs Lisp Reference Manual. | |
177c0ea7 | 3 | @c Copyright (C) 1990, 1991, 1992, 1993, 1994 Free Software Foundation, Inc. |
53f60086 RS |
4 | @c See the file elisp.texi for copying conditions. |
5 | @setfilename ../info/compile | |
a9f0a989 | 6 | @node Byte Compilation, Advising Functions, Loading, Top |
53f60086 RS |
7 | @chapter Byte Compilation |
8 | @cindex byte-code | |
9 | @cindex compilation | |
10 | ||
969fe9b5 | 11 | Emacs Lisp has a @dfn{compiler} that translates functions written |
53f60086 RS |
12 | in Lisp into a special representation called @dfn{byte-code} that can be |
13 | executed more efficiently. The compiler replaces Lisp function | |
14 | definitions with byte-code. When a byte-code function is called, its | |
15 | definition is evaluated by the @dfn{byte-code interpreter}. | |
16 | ||
17 | Because the byte-compiled code is evaluated by the byte-code | |
18 | interpreter, instead of being executed directly by the machine's | |
19 | hardware (as true compiled code is), byte-code is completely | |
20 | transportable from machine to machine without recompilation. It is not, | |
21 | however, as fast as true compiled code. | |
22 | ||
a9f0a989 RS |
23 | Compiling a Lisp file with the Emacs byte compiler always reads the |
24 | file as multibyte text, even if Emacs was started with @samp{--unibyte}, | |
25 | unless the file specifies otherwise. This is so that compilation gives | |
26 | results compatible with running the same file without compilation. | |
27 | @xref{Loading Non-ASCII}. | |
28 | ||
53f60086 | 29 | In general, any version of Emacs can run byte-compiled code produced |
b57bdd74 | 30 | by recent earlier versions of Emacs, but the reverse is not true. |
53f60086 | 31 | |
d7810bda RS |
32 | @vindex no-byte-compile |
33 | If you do not want a Lisp file to be compiled, ever, put a file-local | |
34 | variable binding for @code{no-byte-compile} into it, like this: | |
35 | ||
36 | @example | |
37 | ;; -*-no-byte-compile: t; -*- | |
38 | @end example | |
39 | ||
53f60086 RS |
40 | @xref{Compilation Errors}, for how to investigate errors occurring in |
41 | byte compilation. | |
42 | ||
43 | @menu | |
a0acfc98 | 44 | * Speed of Byte-Code:: An example of speedup from byte compilation. |
53f60086 | 45 | * Compilation Functions:: Byte compilation functions. |
22697dac KH |
46 | * Docs and Compilation:: Dynamic loading of documentation strings. |
47 | * Dynamic Loading:: Dynamic loading of individual functions. | |
53f60086 | 48 | * Eval During Compile:: Code to be evaluated when you compile. |
d7810bda | 49 | * Compiler Errors:: Handling compiler error messages. |
53f60086 RS |
50 | * Byte-Code Objects:: The data type used for byte-compiled functions. |
51 | * Disassembly:: Disassembling byte-code; how to read byte-code. | |
52 | @end menu | |
53 | ||
a0acfc98 RS |
54 | @node Speed of Byte-Code |
55 | @section Performance of Byte-Compiled Code | |
53f60086 RS |
56 | |
57 | A byte-compiled function is not as efficient as a primitive function | |
58 | written in C, but runs much faster than the version written in Lisp. | |
a0acfc98 | 59 | Here is an example: |
53f60086 RS |
60 | |
61 | @example | |
62 | @group | |
63 | (defun silly-loop (n) | |
64 | "Return time before and after N iterations of a loop." | |
65 | (let ((t1 (current-time-string))) | |
177c0ea7 | 66 | (while (> (setq n (1- n)) |
53f60086 RS |
67 | 0)) |
68 | (list t1 (current-time-string)))) | |
69 | @result{} silly-loop | |
70 | @end group | |
71 | ||
72 | @group | |
73 | (silly-loop 100000) | |
a0acfc98 RS |
74 | @result{} ("Fri Mar 18 17:25:57 1994" |
75 | "Fri Mar 18 17:26:28 1994") ; @r{31 seconds} | |
53f60086 RS |
76 | @end group |
77 | ||
78 | @group | |
79 | (byte-compile 'silly-loop) | |
80 | @result{} @r{[Compiled code not shown]} | |
81 | @end group | |
82 | ||
83 | @group | |
84 | (silly-loop 100000) | |
a0acfc98 RS |
85 | @result{} ("Fri Mar 18 17:26:52 1994" |
86 | "Fri Mar 18 17:26:58 1994") ; @r{6 seconds} | |
53f60086 RS |
87 | @end group |
88 | @end example | |
89 | ||
a0acfc98 RS |
90 | In this example, the interpreted code required 31 seconds to run, |
91 | whereas the byte-compiled code required 6 seconds. These results are | |
53f60086 RS |
92 | representative, but actual results will vary greatly. |
93 | ||
a0acfc98 RS |
94 | @node Compilation Functions |
95 | @comment node-name, next, previous, up | |
96 | @section The Compilation Functions | |
97 | @cindex compilation functions | |
98 | ||
99 | You can byte-compile an individual function or macro definition with | |
100 | the @code{byte-compile} function. You can compile a whole file with | |
101 | @code{byte-compile-file}, or several files with | |
102 | @code{byte-recompile-directory} or @code{batch-byte-compile}. | |
103 | ||
9e2b495b RS |
104 | The byte compiler produces error messages and warnings about each file |
105 | in a buffer called @samp{*Compile-Log*}. These report things in your | |
106 | program that suggest a problem but are not necessarily erroneous. | |
a0acfc98 RS |
107 | |
108 | @cindex macro compilation | |
969fe9b5 RS |
109 | Be careful when writing macro calls in files that you may someday |
110 | byte-compile. Macro calls are expanded when they are compiled, so the | |
111 | macros must already be defined for proper compilation. For more | |
456ced1f RS |
112 | details, see @ref{Compiling Macros}. If a program does not work the |
113 | same way when compiled as it does when interpreted, erroneous macro | |
114 | definitions are one likely cause (@pxref{Problems with Macros}). | |
b57bdd74 RS |
115 | Inline (@code{defsubst}) functions are less troublesome; if you |
116 | compile a call to such a function before its definition is known, the | |
117 | call will still work right, it will just run slower. | |
a0acfc98 RS |
118 | |
119 | Normally, compiling a file does not evaluate the file's contents or | |
bfe721d1 KH |
120 | load the file. But it does execute any @code{require} calls at top |
121 | level in the file. One way to ensure that necessary macro definitions | |
122 | are available during compilation is to require the file that defines | |
123 | them (@pxref{Named Features}). To avoid loading the macro definition files | |
124 | when someone @emph{runs} the compiled program, write | |
125 | @code{eval-when-compile} around the @code{require} calls (@pxref{Eval | |
126 | During Compile}). | |
a0acfc98 | 127 | |
53f60086 | 128 | @defun byte-compile symbol |
a0acfc98 | 129 | This function byte-compiles the function definition of @var{symbol}, |
53f60086 RS |
130 | replacing the previous definition with the compiled one. The function |
131 | definition of @var{symbol} must be the actual code for the function; | |
132 | i.e., the compiler does not follow indirection to another symbol. | |
a0acfc98 RS |
133 | @code{byte-compile} returns the new, compiled definition of |
134 | @var{symbol}. | |
135 | ||
22697dac | 136 | If @var{symbol}'s definition is a byte-code function object, |
a0acfc98 RS |
137 | @code{byte-compile} does nothing and returns @code{nil}. Lisp records |
138 | only one function definition for any symbol, and if that is already | |
139 | compiled, non-compiled code is not available anywhere. So there is no | |
140 | way to ``compile the same definition again.'' | |
53f60086 RS |
141 | |
142 | @example | |
143 | @group | |
144 | (defun factorial (integer) | |
145 | "Compute factorial of INTEGER." | |
146 | (if (= 1 integer) 1 | |
147 | (* integer (factorial (1- integer))))) | |
a0acfc98 | 148 | @result{} factorial |
53f60086 RS |
149 | @end group |
150 | ||
151 | @group | |
152 | (byte-compile 'factorial) | |
a0acfc98 | 153 | @result{} |
53f60086 RS |
154 | #[(integer) |
155 | "^H\301U\203^H^@@\301\207\302^H\303^HS!\"\207" | |
156 | [integer 1 * factorial] | |
157 | 4 "Compute factorial of INTEGER."] | |
158 | @end group | |
159 | @end example | |
160 | ||
161 | @noindent | |
a0acfc98 RS |
162 | The result is a byte-code function object. The string it contains is |
163 | the actual byte-code; each character in it is an instruction or an | |
164 | operand of an instruction. The vector contains all the constants, | |
165 | variable names and function names used by the function, except for | |
166 | certain primitives that are coded as special instructions. | |
76865de3 RS |
167 | |
168 | If the argument to @code{byte-compile} is a @code{lambda} expression, | |
169 | it returns the corresponding compiled code, but does not store | |
170 | it anywhere. | |
53f60086 RS |
171 | @end defun |
172 | ||
e75c1a57 | 173 | @deffn Command compile-defun &optional arg |
53f60086 RS |
174 | This command reads the defun containing point, compiles it, and |
175 | evaluates the result. If you use this on a defun that is actually a | |
176 | function definition, the effect is to install a compiled version of that | |
177 | function. | |
e75c1a57 LT |
178 | |
179 | @code{compile-defun} normally displays the result of evaluation in the | |
180 | echo area, but if @var{arg} is non-@code{nil}, it inserts the result | |
181 | in the current buffer after the form it compiled. | |
53f60086 RS |
182 | @end deffn |
183 | ||
e75c1a57 | 184 | @deffn Command byte-compile-file filename &optional load |
a9f0a989 RS |
185 | This function compiles a file of Lisp code named @var{filename} into a |
186 | file of byte-code. The output file's name is made by changing the | |
187 | @samp{.el} suffix into @samp{.elc}; if @var{filename} does not end in | |
188 | @samp{.el}, it adds @samp{.elc} to the end of @var{filename}. | |
53f60086 | 189 | |
a0acfc98 | 190 | Compilation works by reading the input file one form at a time. If it |
53f60086 RS |
191 | is a definition of a function or macro, the compiled function or macro |
192 | definition is written out. Other forms are batched together, then each | |
193 | batch is compiled, and written so that its compiled code will be | |
194 | executed when the file is read. All comments are discarded when the | |
195 | input file is read. | |
196 | ||
e75c1a57 LT |
197 | This command returns @code{t} if there were no errors and @code{nil} |
198 | otherwise. When called interactively, it prompts for the file name. | |
199 | ||
200 | If @var{load} is non-@code{nil}, this command loads the compiled file | |
201 | after compiling it. Interactively, @var{load} is the prefix argument. | |
53f60086 RS |
202 | |
203 | @example | |
204 | @group | |
205 | % ls -l push* | |
206 | -rw-r--r-- 1 lewis 791 Oct 5 20:31 push.el | |
207 | @end group | |
208 | ||
209 | @group | |
210 | (byte-compile-file "~/emacs/push.el") | |
211 | @result{} t | |
212 | @end group | |
213 | ||
214 | @group | |
215 | % ls -l push* | |
216 | -rw-r--r-- 1 lewis 791 Oct 5 20:31 push.el | |
217 | -rw-rw-rw- 1 lewis 638 Oct 8 20:25 push.elc | |
218 | @end group | |
219 | @end example | |
220 | @end deffn | |
221 | ||
e75c1a57 | 222 | @deffn Command byte-recompile-directory directory &optional flag force |
53f60086 | 223 | @cindex library compilation |
e75c1a57 LT |
224 | This command recompiles every @samp{.el} file in @var{directory} (or |
225 | its subdirectories) that needs recompilation. A file needs | |
226 | recompilation if a @samp{.elc} file exists but is older than the | |
227 | @samp{.el} file. | |
228 | ||
229 | When a @samp{.el} file has no corresponding @samp{.elc} file, | |
230 | @var{flag} says what to do. If it is @code{nil}, this command ignores | |
231 | these files. If @var{flag} is 0, it compiles them. If it is neither | |
76865de3 RS |
232 | @code{nil} nor 0, it asks the user whether to compile each such file, |
233 | and asks about each subdirectory as well. | |
53f60086 | 234 | |
e75c1a57 LT |
235 | Interactively, @code{byte-recompile-directory} prompts for |
236 | @var{directory} and @var{flag} is the prefix argument. | |
53f60086 | 237 | |
e75c1a57 LT |
238 | If @var{force} is non-@code{nil}, this command recompiles every |
239 | @samp{.el} file that has a @samp{.elc} file. | |
240 | ||
241 | The returned value is unpredictable. | |
53f60086 RS |
242 | @end deffn |
243 | ||
e75c1a57 | 244 | @defun batch-byte-compile &optional noforce |
a0acfc98 RS |
245 | This function runs @code{byte-compile-file} on files specified on the |
246 | command line. This function must be used only in a batch execution of | |
247 | Emacs, as it kills Emacs on completion. An error in one file does not | |
f9f59935 RS |
248 | prevent processing of subsequent files, but no output file will be |
249 | generated for it, and the Emacs process will terminate with a nonzero | |
250 | status code. | |
53f60086 | 251 | |
e75c1a57 LT |
252 | If @var{noforce} is non-@code{nil}, this function does not recompile |
253 | files that have an up-to-date @samp{.elc} file. | |
254 | ||
53f60086 RS |
255 | @example |
256 | % emacs -batch -f batch-byte-compile *.el | |
257 | @end example | |
258 | @end defun | |
259 | ||
260 | @defun byte-code code-string data-vector max-stack | |
261 | @cindex byte-code interpreter | |
a0acfc98 | 262 | This function actually interprets byte-code. A byte-compiled function |
53f60086 | 263 | is actually defined with a body that calls @code{byte-code}. Don't call |
969fe9b5 | 264 | this function yourself---only the byte compiler knows how to generate |
53f60086 RS |
265 | valid calls to this function. |
266 | ||
969fe9b5 RS |
267 | In Emacs version 18, byte-code was always executed by way of a call to |
268 | the function @code{byte-code}. Nowadays, byte-code is usually executed | |
269 | as part of a byte-code function object, and only rarely through an | |
270 | explicit call to @code{byte-code}. | |
53f60086 RS |
271 | @end defun |
272 | ||
22697dac KH |
273 | @node Docs and Compilation |
274 | @section Documentation Strings and Compilation | |
275 | @cindex dynamic loading of documentation | |
276 | ||
277 | Functions and variables loaded from a byte-compiled file access their | |
278 | documentation strings dynamically from the file whenever needed. This | |
cc8c51f1 | 279 | saves space within Emacs, and makes loading faster because the |
22697dac KH |
280 | documentation strings themselves need not be processed while loading the |
281 | file. Actual access to the documentation strings becomes slower as a | |
282 | result, but this normally is not enough to bother users. | |
283 | ||
284 | Dynamic access to documentation strings does have drawbacks: | |
285 | ||
286 | @itemize @bullet | |
287 | @item | |
288 | If you delete or move the compiled file after loading it, Emacs can no | |
289 | longer access the documentation strings for the functions and variables | |
290 | in the file. | |
291 | ||
292 | @item | |
293 | If you alter the compiled file (such as by compiling a new version), | |
76865de3 RS |
294 | then further access to documentation strings in this file will |
295 | probably give nonsense results. | |
22697dac KH |
296 | @end itemize |
297 | ||
298 | If your site installs Emacs following the usual procedures, these | |
299 | problems will never normally occur. Installing a new version uses a new | |
300 | directory with a different name; as long as the old version remains | |
301 | installed, its files will remain unmodified in the places where they are | |
302 | expected to be. | |
303 | ||
89dbfd18 | 304 | However, if you have built Emacs yourself and use it from the |
22697dac KH |
305 | directory where you built it, you will experience this problem |
306 | occasionally if you edit and recompile Lisp files. When it happens, you | |
307 | can cure the problem by reloading the file after recompiling it. | |
308 | ||
b57bdd74 RS |
309 | You can turn off this feature at compile time by setting |
310 | @code{byte-compile-dynamic-docstrings} to @code{nil}; this is useful | |
311 | mainly if you expect to change the file, and you want Emacs processes | |
312 | that have already loaded it to keep working when the file changes. | |
313 | You can do this globally, or for one source file by specifying a | |
314 | file-local binding for the variable. One way to do that is by adding | |
315 | this string to the file's first line: | |
22697dac KH |
316 | |
317 | @example | |
318 | -*-byte-compile-dynamic-docstrings: nil;-*- | |
319 | @end example | |
320 | ||
321 | @defvar byte-compile-dynamic-docstrings | |
322 | If this is non-@code{nil}, the byte compiler generates compiled files | |
323 | that are set up for dynamic loading of documentation strings. | |
324 | @end defvar | |
325 | ||
326 | @cindex @samp{#@@@var{count}} | |
327 | @cindex @samp{#$} | |
328 | The dynamic documentation string feature writes compiled files that | |
329 | use a special Lisp reader construct, @samp{#@@@var{count}}. This | |
330 | construct skips the next @var{count} characters. It also uses the | |
331 | @samp{#$} construct, which stands for ``the name of this file, as a | |
969fe9b5 RS |
332 | string.'' It is usually best not to use these constructs in Lisp source |
333 | files, since they are not designed to be clear to humans reading the | |
334 | file. | |
22697dac KH |
335 | |
336 | @node Dynamic Loading | |
337 | @section Dynamic Loading of Individual Functions | |
338 | ||
339 | @cindex dynamic loading of functions | |
340 | @cindex lazy loading | |
341 | When you compile a file, you can optionally enable the @dfn{dynamic | |
342 | function loading} feature (also known as @dfn{lazy loading}). With | |
343 | dynamic function loading, loading the file doesn't fully read the | |
344 | function definitions in the file. Instead, each function definition | |
345 | contains a place-holder which refers to the file. The first time each | |
346 | function is called, it reads the full definition from the file, to | |
347 | replace the place-holder. | |
348 | ||
349 | The advantage of dynamic function loading is that loading the file | |
350 | becomes much faster. This is a good thing for a file which contains | |
969fe9b5 RS |
351 | many separate user-callable functions, if using one of them does not |
352 | imply you will probably also use the rest. A specialized mode which | |
353 | provides many keyboard commands often has that usage pattern: a user may | |
354 | invoke the mode, but use only a few of the commands it provides. | |
22697dac KH |
355 | |
356 | The dynamic loading feature has certain disadvantages: | |
357 | ||
358 | @itemize @bullet | |
359 | @item | |
360 | If you delete or move the compiled file after loading it, Emacs can no | |
361 | longer load the remaining function definitions not already loaded. | |
362 | ||
363 | @item | |
364 | If you alter the compiled file (such as by compiling a new version), | |
76865de3 RS |
365 | then trying to load any function not already loaded will usually yield |
366 | nonsense results. | |
22697dac KH |
367 | @end itemize |
368 | ||
969fe9b5 RS |
369 | These problems will never happen in normal circumstances with |
370 | installed Emacs files. But they are quite likely to happen with Lisp | |
371 | files that you are changing. The easiest way to prevent these problems | |
372 | is to reload the new compiled file immediately after each recompilation. | |
22697dac KH |
373 | |
374 | The byte compiler uses the dynamic function loading feature if the | |
375 | variable @code{byte-compile-dynamic} is non-@code{nil} at compilation | |
376 | time. Do not set this variable globally, since dynamic loading is | |
377 | desirable only for certain files. Instead, enable the feature for | |
969fe9b5 RS |
378 | specific source files with file-local variable bindings. For example, |
379 | you could do it by writing this text in the source file's first line: | |
22697dac KH |
380 | |
381 | @example | |
382 | -*-byte-compile-dynamic: t;-*- | |
383 | @end example | |
384 | ||
385 | @defvar byte-compile-dynamic | |
386 | If this is non-@code{nil}, the byte compiler generates compiled files | |
387 | that are set up for dynamic function loading. | |
388 | @end defvar | |
389 | ||
390 | @defun fetch-bytecode function | |
76865de3 RS |
391 | If @var{function} is a byte-code function object, this immediately |
392 | finishes loading the byte code of @var{function} from its | |
393 | byte-compiled file, if it is not fully loaded already. Otherwise, | |
394 | it does nothing. It always returns @var{function}. | |
22697dac KH |
395 | @end defun |
396 | ||
53f60086 RS |
397 | @node Eval During Compile |
398 | @section Evaluation During Compilation | |
399 | ||
22697dac | 400 | These features permit you to write code to be evaluated during |
53f60086 RS |
401 | compilation of a program. |
402 | ||
403 | @defspec eval-and-compile body | |
404 | This form marks @var{body} to be evaluated both when you compile the | |
405 | containing code and when you run it (whether compiled or not). | |
406 | ||
407 | You can get a similar result by putting @var{body} in a separate file | |
969fe9b5 RS |
408 | and referring to that file with @code{require}. That method is |
409 | preferable when @var{body} is large. | |
53f60086 RS |
410 | @end defspec |
411 | ||
412 | @defspec eval-when-compile body | |
f9f59935 | 413 | This form marks @var{body} to be evaluated at compile time but not when |
78c71a98 | 414 | the compiled program is loaded. The result of evaluation by the |
f9f59935 RS |
415 | compiler becomes a constant which appears in the compiled program. If |
416 | you load the source file, rather than compiling it, @var{body} is | |
417 | evaluated normally. | |
53f60086 | 418 | |
969fe9b5 RS |
419 | @strong{Common Lisp Note:} At top level, this is analogous to the Common |
420 | Lisp idiom @code{(eval-when (compile eval) @dots{})}. Elsewhere, the | |
421 | Common Lisp @samp{#.} reader macro (but not when interpreting) is closer | |
422 | to what @code{eval-when-compile} does. | |
53f60086 RS |
423 | @end defspec |
424 | ||
d7810bda RS |
425 | @node Compiler Errors |
426 | @section Compiler Errors | |
427 | @cindex compiler errors | |
428 | ||
429 | Byte compilation writes errors and warnings into the buffer | |
430 | @samp{*Compile-Log*}. The messages include file names and line | |
431 | numbers that identify the location of the problem. The usual Emacs | |
432 | commands for operating on compiler diagnostics work properly on | |
433 | these messages. | |
434 | ||
435 | However, the warnings about functions that were used but not | |
436 | defined are always ``located'' at the end of the file, so these | |
437 | commands won't find the places they are really used. To do that, | |
438 | you must search for the function names. | |
e75c1a57 | 439 | |
fe45b975 | 440 | You can suppress the compiler warning for calling an undefined |
76865de3 | 441 | function @var{func} by conditionalizing the function call on an |
fe45b975 RS |
442 | @code{fboundp} test, like this: |
443 | ||
444 | @example | |
445 | (if (fboundp '@var{func}) ...(@var{func} ...)...) | |
446 | @end example | |
447 | ||
448 | @noindent | |
5c5114b7 RS |
449 | The call to @var{func} must be in the @var{then-form} of the |
450 | @code{if}, and @var{func} must appear quoted in the call to | |
451 | @code{fboundp}. (This feature operates for @code{cond} as well.) | |
452 | ||
453 | Likewise, you can suppress a compiler warning for an unbound variable | |
fe45b975 RS |
454 | @var{variable} by conditionalizing its use on a @code{boundp} test, |
455 | like this: | |
456 | ||
457 | @example | |
458 | (if (boundp '@var{variable}) ...@var{variable}...) | |
459 | @end example | |
460 | ||
461 | @noindent | |
462 | The reference to @var{variable} must be in the @var{then-form} of the | |
463 | @code{if}, and @var{variable} must appear quoted in the call to | |
464 | @code{boundp}. | |
465 | ||
466 | You can suppress any compiler warnings using the construct | |
467 | @code{with-no-warnings}: | |
468 | ||
76865de3 RS |
469 | @c This is implemented with a defun, but conceptually it is |
470 | @c a special form. | |
471 | ||
472 | @defspec with-no-warnings body... | |
fe45b975 RS |
473 | In execution, this is equivalent to @code{(progn @var{body}...)}, |
474 | but the compiler does not issue warnings for anything that occurs | |
475 | inside @var{body}. | |
476 | ||
477 | We recommend that you use this construct around the smallest | |
478 | possible piece of code. | |
76865de3 | 479 | @end defspec |
fe45b975 | 480 | |
53f60086 | 481 | @node Byte-Code Objects |
bfe721d1 | 482 | @section Byte-Code Function Objects |
53f60086 RS |
483 | @cindex compiled function |
484 | @cindex byte-code function | |
485 | ||
486 | Byte-compiled functions have a special data type: they are | |
487 | @dfn{byte-code function objects}. | |
488 | ||
489 | Internally, a byte-code function object is much like a vector; | |
490 | however, the evaluator handles this data type specially when it appears | |
491 | as a function to be called. The printed representation for a byte-code | |
492 | function object is like that for a vector, with an additional @samp{#} | |
493 | before the opening @samp{[}. | |
494 | ||
53f60086 | 495 | A byte-code function object must have at least four elements; there is |
969fe9b5 | 496 | no maximum number, but only the first six elements have any normal use. |
53f60086 RS |
497 | They are: |
498 | ||
499 | @table @var | |
500 | @item arglist | |
501 | The list of argument symbols. | |
502 | ||
503 | @item byte-code | |
504 | The string containing the byte-code instructions. | |
505 | ||
506 | @item constants | |
78c71a98 RS |
507 | The vector of Lisp objects referenced by the byte code. These include |
508 | symbols used as function names and variable names. | |
53f60086 RS |
509 | |
510 | @item stacksize | |
511 | The maximum stack size this function needs. | |
512 | ||
513 | @item docstring | |
bfe721d1 KH |
514 | The documentation string (if any); otherwise, @code{nil}. The value may |
515 | be a number or a list, in case the documentation string is stored in a | |
516 | file. Use the function @code{documentation} to get the real | |
517 | documentation string (@pxref{Accessing Documentation}). | |
53f60086 RS |
518 | |
519 | @item interactive | |
520 | The interactive spec (if any). This can be a string or a Lisp | |
521 | expression. It is @code{nil} for a function that isn't interactive. | |
522 | @end table | |
523 | ||
524 | Here's an example of a byte-code function object, in printed | |
525 | representation. It is the definition of the command | |
526 | @code{backward-sexp}. | |
527 | ||
528 | @example | |
529 | #[(&optional arg) | |
530 | "^H\204^F^@@\301^P\302^H[!\207" | |
531 | [arg 1 forward-sexp] | |
532 | 2 | |
533 | 254435 | |
534 | "p"] | |
535 | @end example | |
536 | ||
537 | The primitive way to create a byte-code object is with | |
538 | @code{make-byte-code}: | |
539 | ||
540 | @defun make-byte-code &rest elements | |
541 | This function constructs and returns a byte-code function object | |
542 | with @var{elements} as its elements. | |
543 | @end defun | |
544 | ||
a0acfc98 RS |
545 | You should not try to come up with the elements for a byte-code |
546 | function yourself, because if they are inconsistent, Emacs may crash | |
78c71a98 | 547 | when you call the function. Always leave it to the byte compiler to |
a0acfc98 | 548 | create these objects; it makes the elements consistent (we hope). |
53f60086 RS |
549 | |
550 | You can access the elements of a byte-code object using @code{aref}; | |
551 | you can also use @code{vconcat} to create a vector with the same | |
552 | elements. | |
553 | ||
554 | @node Disassembly | |
555 | @section Disassembled Byte-Code | |
556 | @cindex disassembled byte-code | |
557 | ||
558 | People do not write byte-code; that job is left to the byte compiler. | |
559 | But we provide a disassembler to satisfy a cat-like curiosity. The | |
560 | disassembler converts the byte-compiled code into humanly readable | |
561 | form. | |
562 | ||
563 | The byte-code interpreter is implemented as a simple stack machine. | |
a0acfc98 | 564 | It pushes values onto a stack of its own, then pops them off to use them |
78c71a98 RS |
565 | in calculations whose results are themselves pushed back on the stack. |
566 | When a byte-code function returns, it pops a value off the stack and | |
567 | returns it as the value of the function. | |
53f60086 | 568 | |
78c71a98 | 569 | In addition to the stack, byte-code functions can use, bind, and set |
a0acfc98 RS |
570 | ordinary Lisp variables, by transferring values between variables and |
571 | the stack. | |
53f60086 | 572 | |
e75c1a57 LT |
573 | @deffn Command disassemble object &optional buffer-or-name |
574 | This command displays the disassembled code for @var{object}. In | |
575 | interactive use, or if @var{buffer-or-name} is @code{nil} or omitted, | |
576 | the output goes in a buffer named @samp{*Disassemble*}. If | |
577 | @var{buffer-or-name} is non-@code{nil}, it must be a buffer or the | |
578 | name of an existing buffer. Then the output goes there, at point, and | |
579 | point is left before the output. | |
53f60086 | 580 | |
e75c1a57 | 581 | The argument @var{object} can be a function name, a lambda expression |
76865de3 RS |
582 | or a byte-code object. If it is a lambda expression, @code{disassemble} |
583 | compiles it and disassembles the resulting compiled code. | |
53f60086 RS |
584 | @end deffn |
585 | ||
586 | Here are two examples of using the @code{disassemble} function. We | |
587 | have added explanatory comments to help you relate the byte-code to the | |
588 | Lisp source; these do not appear in the output of @code{disassemble}. | |
589 | These examples show unoptimized byte-code. Nowadays byte-code is | |
590 | usually optimized, but we did not want to rewrite these examples, since | |
591 | they still serve their purpose. | |
592 | ||
593 | @example | |
594 | @group | |
595 | (defun factorial (integer) | |
596 | "Compute factorial of an integer." | |
597 | (if (= 1 integer) 1 | |
598 | (* integer (factorial (1- integer))))) | |
599 | @result{} factorial | |
600 | @end group | |
601 | ||
602 | @group | |
603 | (factorial 4) | |
604 | @result{} 24 | |
605 | @end group | |
606 | ||
607 | @group | |
608 | (disassemble 'factorial) | |
609 | @print{} byte-code for factorial: | |
610 | doc: Compute factorial of an integer. | |
611 | args: (integer) | |
612 | @end group | |
613 | ||
614 | @group | |
615 | 0 constant 1 ; @r{Push 1 onto stack.} | |
616 | ||
177c0ea7 | 617 | 1 varref integer ; @r{Get value of @code{integer}} |
53f60086 RS |
618 | ; @r{from the environment} |
619 | ; @r{and push the value} | |
620 | ; @r{onto the stack.} | |
621 | @end group | |
622 | ||
623 | @group | |
624 | 2 eqlsign ; @r{Pop top two values off stack,} | |
625 | ; @r{compare them,} | |
626 | ; @r{and push result onto stack.} | |
627 | @end group | |
628 | ||
629 | @group | |
630 | 3 goto-if-nil 10 ; @r{Pop and test top of stack;} | |
631 | ; @r{if @code{nil}, go to 10,} | |
632 | ; @r{else continue.} | |
633 | @end group | |
634 | ||
635 | @group | |
636 | 6 constant 1 ; @r{Push 1 onto top of stack.} | |
637 | ||
638 | 7 goto 17 ; @r{Go to 17 (in this case, 1 will be} | |
639 | ; @r{returned by the function).} | |
640 | @end group | |
641 | ||
642 | @group | |
643 | 10 constant * ; @r{Push symbol @code{*} onto stack.} | |
644 | ||
645 | 11 varref integer ; @r{Push value of @code{integer} onto stack.} | |
646 | @end group | |
647 | ||
648 | @group | |
649 | 12 constant factorial ; @r{Push @code{factorial} onto stack.} | |
650 | ||
651 | 13 varref integer ; @r{Push value of @code{integer} onto stack.} | |
652 | ||
653 | 14 sub1 ; @r{Pop @code{integer}, decrement value,} | |
654 | ; @r{push new value onto stack.} | |
655 | @end group | |
656 | ||
657 | @group | |
658 | ; @r{Stack now contains:} | |
659 | ; @minus{} @r{decremented value of @code{integer}} | |
177c0ea7 | 660 | ; @minus{} @r{@code{factorial}} |
53f60086 RS |
661 | ; @minus{} @r{value of @code{integer}} |
662 | ; @minus{} @r{@code{*}} | |
663 | @end group | |
664 | ||
665 | @group | |
666 | 15 call 1 ; @r{Call function @code{factorial} using} | |
667 | ; @r{the first (i.e., the top) element} | |
668 | ; @r{of the stack as the argument;} | |
669 | ; @r{push returned value onto stack.} | |
670 | @end group | |
671 | ||
672 | @group | |
673 | ; @r{Stack now contains:} | |
78c71a98 | 674 | ; @minus{} @r{result of recursive} |
53f60086 RS |
675 | ; @r{call to @code{factorial}} |
676 | ; @minus{} @r{value of @code{integer}} | |
677 | ; @minus{} @r{@code{*}} | |
678 | @end group | |
679 | ||
680 | @group | |
681 | 16 call 2 ; @r{Using the first two} | |
682 | ; @r{(i.e., the top two)} | |
683 | ; @r{elements of the stack} | |
684 | ; @r{as arguments,} | |
685 | ; @r{call the function @code{*},} | |
686 | ; @r{pushing the result onto the stack.} | |
687 | @end group | |
688 | ||
689 | @group | |
690 | 17 return ; @r{Return the top element} | |
691 | ; @r{of the stack.} | |
692 | @result{} nil | |
693 | @end group | |
694 | @end example | |
695 | ||
696 | The @code{silly-loop} function is somewhat more complex: | |
697 | ||
698 | @example | |
699 | @group | |
700 | (defun silly-loop (n) | |
701 | "Return time before and after N iterations of a loop." | |
702 | (let ((t1 (current-time-string))) | |
177c0ea7 | 703 | (while (> (setq n (1- n)) |
53f60086 RS |
704 | 0)) |
705 | (list t1 (current-time-string)))) | |
706 | @result{} silly-loop | |
707 | @end group | |
708 | ||
709 | @group | |
710 | (disassemble 'silly-loop) | |
711 | @print{} byte-code for silly-loop: | |
712 | doc: Return time before and after N iterations of a loop. | |
713 | args: (n) | |
714 | ||
715 | 0 constant current-time-string ; @r{Push} | |
716 | ; @r{@code{current-time-string}} | |
717 | ; @r{onto top of stack.} | |
718 | @end group | |
719 | ||
720 | @group | |
721 | 1 call 0 ; @r{Call @code{current-time-string}} | |
722 | ; @r{ with no argument,} | |
723 | ; @r{ pushing result onto stack.} | |
724 | @end group | |
725 | ||
726 | @group | |
727 | 2 varbind t1 ; @r{Pop stack and bind @code{t1}} | |
728 | ; @r{to popped value.} | |
729 | @end group | |
730 | ||
731 | @group | |
732 | 3 varref n ; @r{Get value of @code{n} from} | |
733 | ; @r{the environment and push} | |
734 | ; @r{the value onto the stack.} | |
735 | @end group | |
736 | ||
737 | @group | |
738 | 4 sub1 ; @r{Subtract 1 from top of stack.} | |
739 | @end group | |
740 | ||
741 | @group | |
742 | 5 dup ; @r{Duplicate the top of the stack;} | |
a0acfc98 | 743 | ; @r{i.e., copy the top of} |
53f60086 RS |
744 | ; @r{the stack and push the} |
745 | ; @r{copy onto the stack.} | |
746 | @end group | |
747 | ||
748 | @group | |
749 | 6 varset n ; @r{Pop the top of the stack,} | |
750 | ; @r{and bind @code{n} to the value.} | |
751 | ||
752 | ; @r{In effect, the sequence @code{dup varset}} | |
753 | ; @r{copies the top of the stack} | |
754 | ; @r{into the value of @code{n}} | |
755 | ; @r{without popping it.} | |
756 | @end group | |
757 | ||
758 | @group | |
759 | 7 constant 0 ; @r{Push 0 onto stack.} | |
760 | @end group | |
761 | ||
762 | @group | |
763 | 8 gtr ; @r{Pop top two values off stack,} | |
764 | ; @r{test if @var{n} is greater than 0} | |
765 | ; @r{and push result onto stack.} | |
766 | @end group | |
767 | ||
768 | @group | |
78c71a98 RS |
769 | 9 goto-if-nil-else-pop 17 ; @r{Goto 17 if @code{n} <= 0} |
770 | ; @r{(this exits the while loop).} | |
53f60086 RS |
771 | ; @r{else pop top of stack} |
772 | ; @r{and continue} | |
53f60086 RS |
773 | @end group |
774 | ||
775 | @group | |
776 | 12 constant nil ; @r{Push @code{nil} onto stack} | |
777 | ; @r{(this is the body of the loop).} | |
778 | @end group | |
779 | ||
780 | @group | |
781 | 13 discard ; @r{Discard result of the body} | |
782 | ; @r{of the loop (a while loop} | |
783 | ; @r{is always evaluated for} | |
784 | ; @r{its side effects).} | |
785 | @end group | |
786 | ||
787 | @group | |
788 | 14 goto 3 ; @r{Jump back to beginning} | |
789 | ; @r{of while loop.} | |
790 | @end group | |
791 | ||
792 | @group | |
793 | 17 discard ; @r{Discard result of while loop} | |
794 | ; @r{by popping top of stack.} | |
78c71a98 RS |
795 | ; @r{This result is the value @code{nil} that} |
796 | ; @r{was not popped by the goto at 9.} | |
53f60086 RS |
797 | @end group |
798 | ||
799 | @group | |
800 | 18 varref t1 ; @r{Push value of @code{t1} onto stack.} | |
801 | @end group | |
802 | ||
803 | @group | |
177c0ea7 | 804 | 19 constant current-time-string ; @r{Push} |
53f60086 RS |
805 | ; @r{@code{current-time-string}} |
806 | ; @r{onto top of stack.} | |
807 | @end group | |
808 | ||
809 | @group | |
810 | 20 call 0 ; @r{Call @code{current-time-string} again.} | |
811 | @end group | |
812 | ||
813 | @group | |
814 | 21 list2 ; @r{Pop top two elements off stack,} | |
815 | ; @r{create a list of them,} | |
816 | ; @r{and push list onto stack.} | |
817 | @end group | |
818 | ||
819 | @group | |
820 | 22 unbind 1 ; @r{Unbind @code{t1} in local environment.} | |
821 | ||
822 | 23 return ; @r{Return value of the top of stack.} | |
823 | ||
824 | @result{} nil | |
825 | @end group | |
826 | @end example | |
827 | ||
828 | ||
ab5796a9 MB |
829 | @ignore |
830 | arch-tag: f78e3050-2f0a-4dee-be27-d9979a0a2289 | |
831 | @end ignore |