Commit | Line | Data |
---|---|---|
a0d0e21e LW |
1 | =head1 NAME |
2 | ||
3 | perlcall - Perl calling conventions from C | |
4 | ||
5 | =head1 DESCRIPTION | |
6 | ||
d1b91892 | 7 | The purpose of this document is to show you how to call Perl subroutines |
5f05dabc | 8 | directly from C, i.e., how to write I<callbacks>. |
a0d0e21e | 9 | |
d1b91892 AD |
10 | Apart from discussing the C interface provided by Perl for writing |
11 | callbacks the document uses a series of examples to show how the | |
12 | interface actually works in practice. In addition some techniques for | |
13 | coding callbacks are covered. | |
a0d0e21e | 14 | |
d1b91892 | 15 | Examples where callbacks are necessary include |
a0d0e21e LW |
16 | |
17 | =over 5 | |
18 | ||
d1b91892 | 19 | =item * An Error Handler |
a0d0e21e LW |
20 | |
21 | You have created an XSUB interface to an application's C API. | |
22 | ||
23 | A fairly common feature in applications is to allow you to define a C | |
d1b91892 AD |
24 | function that will be called whenever something nasty occurs. What we |
25 | would like is to be able to specify a Perl subroutine that will be | |
26 | called instead. | |
a0d0e21e | 27 | |
d1b91892 | 28 | =item * An Event Driven Program |
a0d0e21e | 29 | |
d1b91892 AD |
30 | The classic example of where callbacks are used is when writing an |
31 | event driven program like for an X windows application. In this case | |
184e9718 | 32 | you register functions to be called whenever specific events occur, |
5f05dabc | 33 | e.g., a mouse button is pressed, the cursor moves into a window or a |
d1b91892 | 34 | menu item is selected. |
a0d0e21e LW |
35 | |
36 | =back | |
37 | ||
d1b91892 AD |
38 | Although the techniques described here are applicable when embedding |
39 | Perl in a C program, this is not the primary goal of this document. | |
40 | There are other details that must be considered and are specific to | |
41 | embedding Perl. For details on embedding Perl in C refer to | |
42 | L<perlembed>. | |
a0d0e21e | 43 | |
d1b91892 AD |
44 | Before you launch yourself head first into the rest of this document, |
45 | it would be a good idea to have read the following two documents - | |
8e07c86e | 46 | L<perlxs> and L<perlguts>. |
a0d0e21e | 47 | |
4929bf7b | 48 | =head1 THE CALL_ FUNCTIONS |
a0d0e21e | 49 | |
d1b91892 AD |
50 | Although this stuff is easier to explain using examples, you first need |
51 | be aware of a few important definitions. | |
a0d0e21e | 52 | |
d1b91892 AD |
53 | Perl has a number of C functions that allow you to call Perl |
54 | subroutines. They are | |
a0d0e21e | 55 | |
4929bf7b GS |
56 | I32 call_sv(SV* sv, I32 flags) ; |
57 | I32 call_pv(char *subname, I32 flags) ; | |
58 | I32 call_method(char *methname, I32 flags) ; | |
59 | I32 call_argv(char *subname, I32 flags, register char **argv) ; | |
a0d0e21e | 60 | |
4929bf7b | 61 | The key function is I<call_sv>. All the other functions are |
d1b91892 | 62 | fairly simple wrappers which make it easier to call Perl subroutines in |
4929bf7b | 63 | special cases. At the end of the day they will all call I<call_sv> |
5f05dabc | 64 | to invoke the Perl subroutine. |
d1b91892 | 65 | |
4929bf7b | 66 | All the I<call_*> functions have a C<flags> parameter which is |
d1b91892 AD |
67 | used to pass a bit mask of options to Perl. This bit mask operates |
68 | identically for each of the functions. The settings available in the | |
69 | bit mask are discussed in L<FLAG VALUES>. | |
70 | ||
71 | Each of the functions will now be discussed in turn. | |
72 | ||
73 | =over 5 | |
74 | ||
4929bf7b | 75 | =item call_sv |
d1b91892 | 76 | |
4929bf7b | 77 | I<call_sv> takes two parameters, the first, C<sv>, is an SV*. |
d1b91892 AD |
78 | This allows you to specify the Perl subroutine to be called either as a |
79 | C string (which has first been converted to an SV) or a reference to a | |
4929bf7b GS |
80 | subroutine. The section, I<Using call_sv>, shows how you can make |
81 | use of I<call_sv>. | |
d1b91892 | 82 | |
4929bf7b | 83 | =item call_pv |
d1b91892 | 84 | |
4929bf7b | 85 | The function, I<call_pv>, is similar to I<call_sv> except it |
d1b91892 | 86 | expects its first parameter to be a C char* which identifies the Perl |
4929bf7b | 87 | subroutine you want to call, e.g., C<call_pv("fred", 0)>. If the |
d1b91892 | 88 | subroutine you want to call is in another package, just include the |
5f05dabc | 89 | package name in the string, e.g., C<"pkg::fred">. |
d1b91892 | 90 | |
4929bf7b | 91 | =item call_method |
d1b91892 | 92 | |
4929bf7b | 93 | The function I<call_method> is used to call a method from a Perl |
d1b91892 AD |
94 | class. The parameter C<methname> corresponds to the name of the method |
95 | to be called. Note that the class that the method belongs to is passed | |
96 | on the Perl stack rather than in the parameter list. This class can be | |
97 | either the name of the class (for a static method) or a reference to an | |
98 | object (for a virtual method). See L<perlobj> for more information on | |
4929bf7b GS |
99 | static and virtual methods and L<Using call_method> for an example |
100 | of using I<call_method>. | |
d1b91892 | 101 | |
4929bf7b | 102 | =item call_argv |
d1b91892 | 103 | |
4929bf7b | 104 | I<call_argv> calls the Perl subroutine specified by the C string |
d1b91892 AD |
105 | stored in the C<subname> parameter. It also takes the usual C<flags> |
106 | parameter. The final parameter, C<argv>, consists of a NULL terminated | |
107 | list of C strings to be passed as parameters to the Perl subroutine. | |
4929bf7b | 108 | See I<Using call_argv>. |
d1b91892 AD |
109 | |
110 | =back | |
111 | ||
112 | All the functions return an integer. This is a count of the number of | |
113 | items returned by the Perl subroutine. The actual items returned by the | |
114 | subroutine are stored on the Perl stack. | |
115 | ||
116 | As a general rule you should I<always> check the return value from | |
117 | these functions. Even if you are expecting only a particular number of | |
118 | values to be returned from the Perl subroutine, there is nothing to | |
19799a22 | 119 | stop someone from doing something unexpected--don't say you haven't |
d1b91892 AD |
120 | been warned. |
121 | ||
122 | =head1 FLAG VALUES | |
123 | ||
4929bf7b | 124 | The C<flags> parameter in all the I<call_*> functions is a bit mask |
d1b91892 AD |
125 | which can consist of any combination of the symbols defined below, |
126 | OR'ed together. | |
127 | ||
128 | ||
54310121 | 129 | =head2 G_VOID |
130 | ||
131 | Calls the Perl subroutine in a void context. | |
132 | ||
133 | This flag has 2 effects: | |
134 | ||
135 | =over 5 | |
136 | ||
137 | =item 1. | |
138 | ||
139 | It indicates to the subroutine being called that it is executing in | |
140 | a void context (if it executes I<wantarray> the result will be the | |
141 | undefined value). | |
142 | ||
143 | =item 2. | |
144 | ||
145 | It ensures that nothing is actually returned from the subroutine. | |
146 | ||
147 | =back | |
148 | ||
4929bf7b | 149 | The value returned by the I<call_*> function indicates how many |
54310121 | 150 | items have been returned by the Perl subroutine - in this case it will |
151 | be 0. | |
152 | ||
153 | ||
d1b91892 AD |
154 | =head2 G_SCALAR |
155 | ||
156 | Calls the Perl subroutine in a scalar context. This is the default | |
4929bf7b | 157 | context flag setting for all the I<call_*> functions. |
d1b91892 | 158 | |
184e9718 | 159 | This flag has 2 effects: |
d1b91892 AD |
160 | |
161 | =over 5 | |
162 | ||
163 | =item 1. | |
164 | ||
184e9718 | 165 | It indicates to the subroutine being called that it is executing in a |
d1b91892 | 166 | scalar context (if it executes I<wantarray> the result will be false). |
a0d0e21e | 167 | |
d1b91892 AD |
168 | =item 2. |
169 | ||
184e9718 | 170 | It ensures that only a scalar is actually returned from the subroutine. |
d1b91892 AD |
171 | The subroutine can, of course, ignore the I<wantarray> and return a |
172 | list anyway. If so, then only the last element of the list will be | |
173 | returned. | |
174 | ||
175 | =back | |
176 | ||
4929bf7b | 177 | The value returned by the I<call_*> function indicates how many |
d1b91892 AD |
178 | items have been returned by the Perl subroutine - in this case it will |
179 | be either 0 or 1. | |
a0d0e21e | 180 | |
d1b91892 | 181 | If 0, then you have specified the G_DISCARD flag. |
a0d0e21e | 182 | |
d1b91892 AD |
183 | If 1, then the item actually returned by the Perl subroutine will be |
184 | stored on the Perl stack - the section I<Returning a Scalar> shows how | |
185 | to access this value on the stack. Remember that regardless of how | |
186 | many items the Perl subroutine returns, only the last one will be | |
187 | accessible from the stack - think of the case where only one value is | |
188 | returned as being a list with only one element. Any other items that | |
189 | were returned will not exist by the time control returns from the | |
4929bf7b | 190 | I<call_*> function. The section I<Returning a list in a scalar |
54310121 | 191 | context> shows an example of this behavior. |
a0d0e21e | 192 | |
a0d0e21e | 193 | |
d1b91892 | 194 | =head2 G_ARRAY |
a0d0e21e | 195 | |
d1b91892 | 196 | Calls the Perl subroutine in a list context. |
a0d0e21e | 197 | |
184e9718 | 198 | As with G_SCALAR, this flag has 2 effects: |
a0d0e21e LW |
199 | |
200 | =over 5 | |
201 | ||
d1b91892 AD |
202 | =item 1. |
203 | ||
90fdbbb7 DC |
204 | It indicates to the subroutine being called that it is executing in a |
205 | list context (if it executes I<wantarray> the result will be true). | |
a0d0e21e | 206 | |
a0d0e21e | 207 | |
d1b91892 | 208 | =item 2. |
a0d0e21e | 209 | |
184e9718 | 210 | It ensures that all items returned from the subroutine will be |
4929bf7b | 211 | accessible when control returns from the I<call_*> function. |
a0d0e21e | 212 | |
d1b91892 | 213 | =back |
a0d0e21e | 214 | |
4929bf7b | 215 | The value returned by the I<call_*> function indicates how many |
d1b91892 | 216 | items have been returned by the Perl subroutine. |
a0d0e21e | 217 | |
184e9718 | 218 | If 0, then you have specified the G_DISCARD flag. |
a0d0e21e | 219 | |
d1b91892 AD |
220 | If not 0, then it will be a count of the number of items returned by |
221 | the subroutine. These items will be stored on the Perl stack. The | |
222 | section I<Returning a list of values> gives an example of using the | |
223 | G_ARRAY flag and the mechanics of accessing the returned items from the | |
224 | Perl stack. | |
a0d0e21e | 225 | |
d1b91892 | 226 | =head2 G_DISCARD |
a0d0e21e | 227 | |
4929bf7b | 228 | By default, the I<call_*> functions place the items returned from |
d1b91892 AD |
229 | by the Perl subroutine on the stack. If you are not interested in |
230 | these items, then setting this flag will make Perl get rid of them | |
231 | automatically for you. Note that it is still possible to indicate a | |
232 | context to the Perl subroutine by using either G_SCALAR or G_ARRAY. | |
a0d0e21e | 233 | |
d1b91892 | 234 | If you do not set this flag then it is I<very> important that you make |
5f05dabc | 235 | sure that any temporaries (i.e., parameters passed to the Perl |
d1b91892 AD |
236 | subroutine and values returned from the subroutine) are disposed of |
237 | yourself. The section I<Returning a Scalar> gives details of how to | |
5f05dabc | 238 | dispose of these temporaries explicitly and the section I<Using Perl to |
d1b91892 AD |
239 | dispose of temporaries> discusses the specific circumstances where you |
240 | can ignore the problem and let Perl deal with it for you. | |
a0d0e21e | 241 | |
d1b91892 | 242 | =head2 G_NOARGS |
a0d0e21e | 243 | |
4929bf7b | 244 | Whenever a Perl subroutine is called using one of the I<call_*> |
d1b91892 AD |
245 | functions, it is assumed by default that parameters are to be passed to |
246 | the subroutine. If you are not passing any parameters to the Perl | |
247 | subroutine, you can save a bit of time by setting this flag. It has | |
248 | the effect of not creating the C<@_> array for the Perl subroutine. | |
a0d0e21e | 249 | |
d1b91892 AD |
250 | Although the functionality provided by this flag may seem |
251 | straightforward, it should be used only if there is a good reason to do | |
252 | so. The reason for being cautious is that even if you have specified | |
253 | the G_NOARGS flag, it is still possible for the Perl subroutine that | |
254 | has been called to think that you have passed it parameters. | |
a0d0e21e | 255 | |
d1b91892 AD |
256 | In fact, what can happen is that the Perl subroutine you have called |
257 | can access the C<@_> array from a previous Perl subroutine. This will | |
4929bf7b | 258 | occur when the code that is executing the I<call_*> function has |
d1b91892 AD |
259 | itself been called from another Perl subroutine. The code below |
260 | illustrates this | |
a0d0e21e | 261 | |
d1b91892 AD |
262 | sub fred |
263 | { print "@_\n" } | |
a0d0e21e | 264 | |
d1b91892 AD |
265 | sub joe |
266 | { &fred } | |
a0d0e21e | 267 | |
d1b91892 | 268 | &joe(1,2,3) ; |
a0d0e21e LW |
269 | |
270 | This will print | |
271 | ||
d1b91892 AD |
272 | 1 2 3 |
273 | ||
274 | What has happened is that C<fred> accesses the C<@_> array which | |
275 | belongs to C<joe>. | |
a0d0e21e | 276 | |
a0d0e21e | 277 | |
54310121 | 278 | =head2 G_EVAL |
a0d0e21e | 279 | |
d1b91892 | 280 | It is possible for the Perl subroutine you are calling to terminate |
5f05dabc | 281 | abnormally, e.g., by calling I<die> explicitly or by not actually |
268118b2 HS |
282 | existing. By default, when either of these events occurs, the |
283 | process will terminate immediately. If you want to trap this | |
d1b91892 AD |
284 | type of event, specify the G_EVAL flag. It will put an I<eval { }> |
285 | around the subroutine call. | |
a0d0e21e | 286 | |
4929bf7b | 287 | Whenever control returns from the I<call_*> function you need to |
d1b91892 AD |
288 | check the C<$@> variable as you would in a normal Perl script. |
289 | ||
4929bf7b | 290 | The value returned from the I<call_*> function is dependent on |
d1b91892 | 291 | what other flags have been specified and whether an error has |
184e9718 | 292 | occurred. Here are all the different cases that can occur: |
d1b91892 AD |
293 | |
294 | =over 5 | |
295 | ||
296 | =item * | |
297 | ||
4929bf7b | 298 | If the I<call_*> function returns normally, then the value |
d1b91892 AD |
299 | returned is as specified in the previous sections. |
300 | ||
301 | =item * | |
302 | ||
303 | If G_DISCARD is specified, the return value will always be 0. | |
304 | ||
305 | =item * | |
306 | ||
307 | If G_ARRAY is specified I<and> an error has occurred, the return value | |
308 | will always be 0. | |
309 | ||
310 | =item * | |
a0d0e21e | 311 | |
d1b91892 AD |
312 | If G_SCALAR is specified I<and> an error has occurred, the return value |
313 | will be 1 and the value on the top of the stack will be I<undef>. This | |
314 | means that if you have already detected the error by checking C<$@> and | |
315 | you want the program to continue, you must remember to pop the I<undef> | |
316 | from the stack. | |
a0d0e21e LW |
317 | |
318 | =back | |
319 | ||
54310121 | 320 | See I<Using G_EVAL> for details on using G_EVAL. |
d1b91892 | 321 | |
c07a80fd | 322 | =head2 G_KEEPERR |
323 | ||
324 | You may have noticed that using the G_EVAL flag described above will | |
325 | B<always> clear the C<$@> variable and set it to a string describing | |
326 | the error iff there was an error in the called code. This unqualified | |
327 | resetting of C<$@> can be problematic in the reliable identification of | |
328 | errors using the C<eval {}> mechanism, because the possibility exists | |
329 | that perl will call other code (end of block processing code, for | |
330 | example) between the time the error causes C<$@> to be set within | |
331 | C<eval {}>, and the subsequent statement which checks for the value of | |
332 | C<$@> gets executed in the user's script. | |
333 | ||
334 | This scenario will mostly be applicable to code that is meant to be | |
335 | called from within destructors, asynchronous callbacks, signal | |
336 | handlers, C<__DIE__> or C<__WARN__> hooks, and C<tie> functions. In | |
337 | such situations, you will not want to clear C<$@> at all, but simply to | |
338 | append any new errors to any existing value of C<$@>. | |
339 | ||
340 | The G_KEEPERR flag is meant to be used in conjunction with G_EVAL in | |
4929bf7b | 341 | I<call_*> functions that are used to implement such code. This flag |
c07a80fd | 342 | has no effect when G_EVAL is not used. |
343 | ||
344 | When G_KEEPERR is used, any errors in the called code will be prefixed | |
345 | with the string "\t(in cleanup)", and appended to the current value | |
346 | of C<$@>. | |
347 | ||
348 | The G_KEEPERR flag was introduced in Perl version 5.002. | |
349 | ||
350 | See I<Using G_KEEPERR> for an example of a situation that warrants the | |
351 | use of this flag. | |
352 | ||
54310121 | 353 | =head2 Determining the Context |
d1b91892 AD |
354 | |
355 | As mentioned above, you can determine the context of the currently | |
54310121 | 356 | executing subroutine in Perl with I<wantarray>. The equivalent test |
357 | can be made in C by using the C<GIMME_V> macro, which returns | |
90fdbbb7 | 358 | C<G_ARRAY> if you have been called in a list context, C<G_SCALAR> if |
7a2e2cd6 | 359 | in a scalar context, or C<G_VOID> if in a void context (i.e. the |
54310121 | 360 | return value will not be used). An older version of this macro is |
361 | called C<GIMME>; in a void context it returns C<G_SCALAR> instead of | |
362 | C<G_VOID>. An example of using the C<GIMME_V> macro is shown in | |
363 | section I<Using GIMME_V>. | |
d1b91892 AD |
364 | |
365 | =head1 KNOWN PROBLEMS | |
366 | ||
367 | This section outlines all known problems that exist in the | |
4929bf7b | 368 | I<call_*> functions. |
d1b91892 AD |
369 | |
370 | =over 5 | |
371 | ||
372 | =item 1. | |
373 | ||
374 | If you are intending to make use of both the G_EVAL and G_SCALAR flags | |
375 | in your code, use a version of Perl greater than 5.000. There is a bug | |
376 | in version 5.000 of Perl which means that the combination of these two | |
377 | flags will not work as described in the section I<FLAG VALUES>. | |
378 | ||
379 | Specifically, if the two flags are used when calling a subroutine and | |
380 | that subroutine does not call I<die>, the value returned by | |
4929bf7b | 381 | I<call_*> will be wrong. |
d1b91892 AD |
382 | |
383 | ||
384 | =item 2. | |
385 | ||
4929bf7b | 386 | In Perl 5.000 and 5.001 there is a problem with using I<call_*> if |
d1b91892 AD |
387 | the Perl sub you are calling attempts to trap a I<die>. |
388 | ||
389 | The symptom of this problem is that the called Perl sub will continue | |
390 | to completion, but whenever it attempts to pass control back to the | |
391 | XSUB, the program will immediately terminate. | |
392 | ||
393 | For example, say you want to call this Perl sub | |
394 | ||
395 | sub fred | |
396 | { | |
397 | eval { die "Fatal Error" ; } | |
54310121 | 398 | print "Trapped error: $@\n" |
d1b91892 AD |
399 | if $@ ; |
400 | } | |
401 | ||
402 | via this XSUB | |
403 | ||
404 | void | |
405 | Call_fred() | |
406 | CODE: | |
924508f0 | 407 | PUSHMARK(SP) ; |
4929bf7b | 408 | call_pv("fred", G_DISCARD|G_NOARGS) ; |
d1b91892 AD |
409 | fprintf(stderr, "back in Call_fred\n") ; |
410 | ||
411 | When C<Call_fred> is executed it will print | |
412 | ||
413 | Trapped error: Fatal Error | |
414 | ||
415 | As control never returns to C<Call_fred>, the C<"back in Call_fred"> | |
416 | string will not get printed. | |
417 | ||
3fe9a6f1 | 418 | To work around this problem, you can either upgrade to Perl 5.002 or |
4929bf7b | 419 | higher, or use the G_EVAL flag with I<call_*> as shown below |
d1b91892 AD |
420 | |
421 | void | |
422 | Call_fred() | |
423 | CODE: | |
924508f0 | 424 | PUSHMARK(SP) ; |
4929bf7b | 425 | call_pv("fred", G_EVAL|G_DISCARD|G_NOARGS) ; |
d1b91892 AD |
426 | fprintf(stderr, "back in Call_fred\n") ; |
427 | ||
428 | =back | |
429 | ||
430 | ||
a0d0e21e LW |
431 | |
432 | =head1 EXAMPLES | |
433 | ||
434 | Enough of the definition talk, let's have a few examples. | |
435 | ||
d1b91892 AD |
436 | Perl provides many macros to assist in accessing the Perl stack. |
437 | Wherever possible, these macros should always be used when interfacing | |
5f05dabc | 438 | to Perl internals. We hope this should make the code less vulnerable |
d1b91892 | 439 | to any changes made to Perl in the future. |
a0d0e21e | 440 | |
d1b91892 | 441 | Another point worth noting is that in the first series of examples I |
4929bf7b | 442 | have made use of only the I<call_pv> function. This has been done |
d1b91892 | 443 | to keep the code simpler and ease you into the topic. Wherever |
4929bf7b GS |
444 | possible, if the choice is between using I<call_pv> and |
445 | I<call_sv>, you should always try to use I<call_sv>. See | |
446 | I<Using call_sv> for details. | |
a0d0e21e | 447 | |
d1b91892 | 448 | =head2 No Parameters, Nothing returned |
a0d0e21e | 449 | |
d1b91892 AD |
450 | This first trivial example will call a Perl subroutine, I<PrintUID>, to |
451 | print out the UID of the process. | |
a0d0e21e LW |
452 | |
453 | sub PrintUID | |
454 | { | |
455 | print "UID is $<\n" ; | |
456 | } | |
457 | ||
d1b91892 | 458 | and here is a C function to call it |
a0d0e21e | 459 | |
d1b91892 | 460 | static void |
a0d0e21e LW |
461 | call_PrintUID() |
462 | { | |
d1b91892 | 463 | dSP ; |
a0d0e21e | 464 | |
924508f0 | 465 | PUSHMARK(SP) ; |
4929bf7b | 466 | call_pv("PrintUID", G_DISCARD|G_NOARGS) ; |
a0d0e21e LW |
467 | } |
468 | ||
d1b91892 | 469 | Simple, eh. |
a0d0e21e | 470 | |
d1b91892 | 471 | A few points to note about this example. |
a0d0e21e LW |
472 | |
473 | =over 5 | |
474 | ||
d1b91892 | 475 | =item 1. |
a0d0e21e | 476 | |
924508f0 | 477 | Ignore C<dSP> and C<PUSHMARK(SP)> for now. They will be discussed in |
d1b91892 | 478 | the next example. |
a0d0e21e LW |
479 | |
480 | =item 2. | |
481 | ||
d1b91892 AD |
482 | We aren't passing any parameters to I<PrintUID> so G_NOARGS can be |
483 | specified. | |
a0d0e21e | 484 | |
d1b91892 | 485 | =item 3. |
a0d0e21e LW |
486 | |
487 | We aren't interested in anything returned from I<PrintUID>, so | |
5f05dabc | 488 | G_DISCARD is specified. Even if I<PrintUID> was changed to |
a0d0e21e | 489 | return some value(s), having specified G_DISCARD will mean that they |
4929bf7b | 490 | will be wiped by the time control returns from I<call_pv>. |
a0d0e21e | 491 | |
d1b91892 | 492 | =item 4. |
a0d0e21e | 493 | |
4929bf7b | 494 | As I<call_pv> is being used, the Perl subroutine is specified as a |
d1b91892 AD |
495 | C string. In this case the subroutine name has been 'hard-wired' into the |
496 | code. | |
a0d0e21e LW |
497 | |
498 | =item 5. | |
499 | ||
d1b91892 | 500 | Because we specified G_DISCARD, it is not necessary to check the value |
4929bf7b | 501 | returned from I<call_pv>. It will always be 0. |
a0d0e21e LW |
502 | |
503 | =back | |
504 | ||
d1b91892 | 505 | =head2 Passing Parameters |
a0d0e21e | 506 | |
d1b91892 | 507 | Now let's make a slightly more complex example. This time we want to |
19799a22 GS |
508 | call a Perl subroutine, C<LeftString>, which will take 2 parameters--a |
509 | string ($s) and an integer ($n). The subroutine will simply | |
510 | print the first $n characters of the string. | |
a0d0e21e | 511 | |
d1b91892 | 512 | So the Perl subroutine would look like this |
a0d0e21e LW |
513 | |
514 | sub LeftString | |
515 | { | |
516 | my($s, $n) = @_ ; | |
517 | print substr($s, 0, $n), "\n" ; | |
518 | } | |
519 | ||
520 | The C function required to call I<LeftString> would look like this. | |
521 | ||
522 | static void | |
523 | call_LeftString(a, b) | |
524 | char * a ; | |
525 | int b ; | |
526 | { | |
527 | dSP ; | |
528 | ||
9b6570b4 AB |
529 | ENTER ; |
530 | SAVETMPS ; | |
531 | ||
924508f0 | 532 | PUSHMARK(SP) ; |
a0d0e21e LW |
533 | XPUSHs(sv_2mortal(newSVpv(a, 0))); |
534 | XPUSHs(sv_2mortal(newSViv(b))); | |
535 | PUTBACK ; | |
536 | ||
4929bf7b | 537 | call_pv("LeftString", G_DISCARD); |
9b6570b4 AB |
538 | |
539 | FREETMPS ; | |
540 | LEAVE ; | |
a0d0e21e LW |
541 | } |
542 | ||
a0d0e21e LW |
543 | Here are a few notes on the C function I<call_LeftString>. |
544 | ||
545 | =over 5 | |
546 | ||
d1b91892 | 547 | =item 1. |
a0d0e21e | 548 | |
d1b91892 AD |
549 | Parameters are passed to the Perl subroutine using the Perl stack. |
550 | This is the purpose of the code beginning with the line C<dSP> and | |
1e62ac33 | 551 | ending with the line C<PUTBACK>. The C<dSP> declares a local copy |
924508f0 GS |
552 | of the stack pointer. This local copy should B<always> be accessed |
553 | as C<SP>. | |
a0d0e21e | 554 | |
d1b91892 | 555 | =item 2. |
a0d0e21e LW |
556 | |
557 | If you are going to put something onto the Perl stack, you need to know | |
19799a22 | 558 | where to put it. This is the purpose of the macro C<dSP>--it declares |
d1b91892 | 559 | and initializes a I<local> copy of the Perl stack pointer. |
a0d0e21e LW |
560 | |
561 | All the other macros which will be used in this example require you to | |
d1b91892 | 562 | have used this macro. |
a0d0e21e | 563 | |
d1b91892 AD |
564 | The exception to this rule is if you are calling a Perl subroutine |
565 | directly from an XSUB function. In this case it is not necessary to | |
19799a22 | 566 | use the C<dSP> macro explicitly--it will be declared for you |
d1b91892 | 567 | automatically. |
a0d0e21e | 568 | |
d1b91892 | 569 | =item 3. |
a0d0e21e LW |
570 | |
571 | Any parameters to be pushed onto the stack should be bracketed by the | |
d1b91892 | 572 | C<PUSHMARK> and C<PUTBACK> macros. The purpose of these two macros, in |
5f05dabc | 573 | this context, is to count the number of parameters you are |
574 | pushing automatically. Then whenever Perl is creating the C<@_> array for the | |
d1b91892 AD |
575 | subroutine, it knows how big to make it. |
576 | ||
577 | The C<PUSHMARK> macro tells Perl to make a mental note of the current | |
578 | stack pointer. Even if you aren't passing any parameters (like the | |
579 | example shown in the section I<No Parameters, Nothing returned>) you | |
580 | must still call the C<PUSHMARK> macro before you can call any of the | |
4929bf7b | 581 | I<call_*> functions--Perl still needs to know that there are no |
d1b91892 AD |
582 | parameters. |
583 | ||
584 | The C<PUTBACK> macro sets the global copy of the stack pointer to be | |
4929bf7b | 585 | the same as our local copy. If we didn't do this I<call_pv> |
19799a22 | 586 | wouldn't know where the two parameters we pushed were--remember that |
d1b91892 AD |
587 | up to now all the stack pointer manipulation we have done is with our |
588 | local copy, I<not> the global copy. | |
589 | ||
590 | =item 4. | |
591 | ||
5f05dabc | 592 | The only flag specified this time is G_DISCARD. Because we are passing 2 |
d1b91892 AD |
593 | parameters to the Perl subroutine this time, we have not specified |
594 | G_NOARGS. | |
a0d0e21e LW |
595 | |
596 | =item 5. | |
597 | ||
598 | Next, we come to XPUSHs. This is where the parameters actually get | |
d1b91892 AD |
599 | pushed onto the stack. In this case we are pushing a string and an |
600 | integer. | |
a0d0e21e | 601 | |
54310121 | 602 | See L<perlguts/"XSUBs and the Argument Stack"> for details |
d1b91892 | 603 | on how the XPUSH macros work. |
a0d0e21e LW |
604 | |
605 | =item 6. | |
606 | ||
9b6570b4 AB |
607 | Because we created temporary values (by means of sv_2mortal() calls) |
608 | we will have to tidy up the Perl stack and dispose of mortal SVs. | |
609 | ||
610 | This is the purpose of | |
611 | ||
612 | ENTER ; | |
613 | SAVETMPS ; | |
614 | ||
615 | at the start of the function, and | |
616 | ||
617 | FREETMPS ; | |
618 | LEAVE ; | |
619 | ||
620 | at the end. The C<ENTER>/C<SAVETMPS> pair creates a boundary for any | |
621 | temporaries we create. This means that the temporaries we get rid of | |
622 | will be limited to those which were created after these calls. | |
623 | ||
624 | The C<FREETMPS>/C<LEAVE> pair will get rid of any values returned by | |
625 | the Perl subroutine (see next example), plus it will also dump the | |
626 | mortal SVs we have created. Having C<ENTER>/C<SAVETMPS> at the | |
627 | beginning of the code makes sure that no other mortals are destroyed. | |
628 | ||
629 | Think of these macros as working a bit like using C<{> and C<}> in Perl | |
630 | to limit the scope of local variables. | |
631 | ||
632 | See the section I<Using Perl to dispose of temporaries> for details of | |
633 | an alternative to using these macros. | |
634 | ||
635 | =item 7. | |
636 | ||
4929bf7b | 637 | Finally, I<LeftString> can now be called via the I<call_pv> |
d1b91892 | 638 | function. |
a0d0e21e LW |
639 | |
640 | =back | |
641 | ||
d1b91892 | 642 | =head2 Returning a Scalar |
a0d0e21e | 643 | |
d1b91892 AD |
644 | Now for an example of dealing with the items returned from a Perl |
645 | subroutine. | |
a0d0e21e | 646 | |
5f05dabc | 647 | Here is a Perl subroutine, I<Adder>, that takes 2 integer parameters |
d1b91892 | 648 | and simply returns their sum. |
a0d0e21e LW |
649 | |
650 | sub Adder | |
651 | { | |
652 | my($a, $b) = @_ ; | |
653 | $a + $b ; | |
654 | } | |
655 | ||
5f05dabc | 656 | Because we are now concerned with the return value from I<Adder>, the C |
d1b91892 | 657 | function required to call it is now a bit more complex. |
a0d0e21e LW |
658 | |
659 | static void | |
660 | call_Adder(a, b) | |
661 | int a ; | |
662 | int b ; | |
663 | { | |
664 | dSP ; | |
665 | int count ; | |
666 | ||
667 | ENTER ; | |
668 | SAVETMPS; | |
669 | ||
924508f0 | 670 | PUSHMARK(SP) ; |
a0d0e21e LW |
671 | XPUSHs(sv_2mortal(newSViv(a))); |
672 | XPUSHs(sv_2mortal(newSViv(b))); | |
673 | PUTBACK ; | |
674 | ||
4929bf7b | 675 | count = call_pv("Adder", G_SCALAR); |
a0d0e21e LW |
676 | |
677 | SPAGAIN ; | |
678 | ||
d1b91892 AD |
679 | if (count != 1) |
680 | croak("Big trouble\n") ; | |
a0d0e21e | 681 | |
d1b91892 | 682 | printf ("The sum of %d and %d is %d\n", a, b, POPi) ; |
a0d0e21e LW |
683 | |
684 | PUTBACK ; | |
685 | FREETMPS ; | |
686 | LEAVE ; | |
687 | } | |
688 | ||
a0d0e21e LW |
689 | Points to note this time are |
690 | ||
691 | =over 5 | |
692 | ||
54310121 | 693 | =item 1. |
a0d0e21e | 694 | |
d1b91892 AD |
695 | The only flag specified this time was G_SCALAR. That means the C<@_> |
696 | array will be created and that the value returned by I<Adder> will | |
4929bf7b | 697 | still exist after the call to I<call_pv>. |
a0d0e21e | 698 | |
a0d0e21e LW |
699 | =item 2. |
700 | ||
a0d0e21e LW |
701 | The purpose of the macro C<SPAGAIN> is to refresh the local copy of the |
702 | stack pointer. This is necessary because it is possible that the memory | |
68dc0745 | 703 | allocated to the Perl stack has been reallocated whilst in the |
4929bf7b | 704 | I<call_pv> call. |
a0d0e21e | 705 | |
d1b91892 | 706 | If you are making use of the Perl stack pointer in your code you must |
54310121 | 707 | always refresh the local copy using SPAGAIN whenever you make use |
4929bf7b | 708 | of the I<call_*> functions or any other Perl internal function. |
a0d0e21e | 709 | |
9b6570b4 | 710 | =item 3. |
a0d0e21e | 711 | |
d1b91892 | 712 | Although only a single value was expected to be returned from I<Adder>, |
4929bf7b | 713 | it is still good practice to check the return code from I<call_pv> |
d1b91892 | 714 | anyway. |
a0d0e21e | 715 | |
d1b91892 AD |
716 | Expecting a single value is not quite the same as knowing that there |
717 | will be one. If someone modified I<Adder> to return a list and we | |
718 | didn't check for that possibility and take appropriate action the Perl | |
719 | stack would end up in an inconsistent state. That is something you | |
5f05dabc | 720 | I<really> don't want to happen ever. |
a0d0e21e | 721 | |
9b6570b4 | 722 | =item 4. |
a0d0e21e | 723 | |
d1b91892 AD |
724 | The C<POPi> macro is used here to pop the return value from the stack. |
725 | In this case we wanted an integer, so C<POPi> was used. | |
a0d0e21e LW |
726 | |
727 | ||
d1b91892 AD |
728 | Here is the complete list of POP macros available, along with the types |
729 | they return. | |
a0d0e21e | 730 | |
d1b91892 AD |
731 | POPs SV |
732 | POPp pointer | |
733 | POPn double | |
734 | POPi integer | |
735 | POPl long | |
a0d0e21e | 736 | |
9b6570b4 | 737 | =item 5. |
a0d0e21e | 738 | |
d1b91892 AD |
739 | The final C<PUTBACK> is used to leave the Perl stack in a consistent |
740 | state before exiting the function. This is necessary because when we | |
741 | popped the return value from the stack with C<POPi> it updated only our | |
742 | local copy of the stack pointer. Remember, C<PUTBACK> sets the global | |
743 | stack pointer to be the same as our local copy. | |
a0d0e21e LW |
744 | |
745 | =back | |
746 | ||
747 | ||
d1b91892 | 748 | =head2 Returning a list of values |
a0d0e21e | 749 | |
d1b91892 AD |
750 | Now, let's extend the previous example to return both the sum of the |
751 | parameters and the difference. | |
a0d0e21e | 752 | |
d1b91892 | 753 | Here is the Perl subroutine |
a0d0e21e LW |
754 | |
755 | sub AddSubtract | |
756 | { | |
757 | my($a, $b) = @_ ; | |
758 | ($a+$b, $a-$b) ; | |
759 | } | |
760 | ||
a0d0e21e LW |
761 | and this is the C function |
762 | ||
763 | static void | |
764 | call_AddSubtract(a, b) | |
765 | int a ; | |
766 | int b ; | |
767 | { | |
768 | dSP ; | |
769 | int count ; | |
770 | ||
771 | ENTER ; | |
772 | SAVETMPS; | |
773 | ||
924508f0 | 774 | PUSHMARK(SP) ; |
a0d0e21e LW |
775 | XPUSHs(sv_2mortal(newSViv(a))); |
776 | XPUSHs(sv_2mortal(newSViv(b))); | |
777 | PUTBACK ; | |
778 | ||
4929bf7b | 779 | count = call_pv("AddSubtract", G_ARRAY); |
a0d0e21e LW |
780 | |
781 | SPAGAIN ; | |
782 | ||
d1b91892 AD |
783 | if (count != 2) |
784 | croak("Big trouble\n") ; | |
a0d0e21e | 785 | |
d1b91892 AD |
786 | printf ("%d - %d = %d\n", a, b, POPi) ; |
787 | printf ("%d + %d = %d\n", a, b, POPi) ; | |
a0d0e21e LW |
788 | |
789 | PUTBACK ; | |
790 | FREETMPS ; | |
791 | LEAVE ; | |
792 | } | |
793 | ||
d1b91892 AD |
794 | If I<call_AddSubtract> is called like this |
795 | ||
796 | call_AddSubtract(7, 4) ; | |
797 | ||
798 | then here is the output | |
799 | ||
800 | 7 - 4 = 3 | |
801 | 7 + 4 = 11 | |
a0d0e21e LW |
802 | |
803 | Notes | |
804 | ||
805 | =over 5 | |
806 | ||
807 | =item 1. | |
808 | ||
90fdbbb7 | 809 | We wanted list context, so G_ARRAY was used. |
a0d0e21e LW |
810 | |
811 | =item 2. | |
812 | ||
d1b91892 AD |
813 | Not surprisingly C<POPi> is used twice this time because we were |
814 | retrieving 2 values from the stack. The important thing to note is that | |
815 | when using the C<POP*> macros they come off the stack in I<reverse> | |
816 | order. | |
a0d0e21e LW |
817 | |
818 | =back | |
819 | ||
d1b91892 AD |
820 | =head2 Returning a list in a scalar context |
821 | ||
822 | Say the Perl subroutine in the previous section was called in a scalar | |
823 | context, like this | |
824 | ||
825 | static void | |
826 | call_AddSubScalar(a, b) | |
827 | int a ; | |
828 | int b ; | |
829 | { | |
830 | dSP ; | |
831 | int count ; | |
832 | int i ; | |
833 | ||
834 | ENTER ; | |
835 | SAVETMPS; | |
836 | ||
924508f0 | 837 | PUSHMARK(SP) ; |
d1b91892 AD |
838 | XPUSHs(sv_2mortal(newSViv(a))); |
839 | XPUSHs(sv_2mortal(newSViv(b))); | |
840 | PUTBACK ; | |
841 | ||
4929bf7b | 842 | count = call_pv("AddSubtract", G_SCALAR); |
d1b91892 AD |
843 | |
844 | SPAGAIN ; | |
845 | ||
846 | printf ("Items Returned = %d\n", count) ; | |
847 | ||
848 | for (i = 1 ; i <= count ; ++i) | |
849 | printf ("Value %d = %d\n", i, POPi) ; | |
850 | ||
851 | PUTBACK ; | |
852 | FREETMPS ; | |
853 | LEAVE ; | |
854 | } | |
855 | ||
856 | The other modification made is that I<call_AddSubScalar> will print the | |
857 | number of items returned from the Perl subroutine and their value (for | |
858 | simplicity it assumes that they are integer). So if | |
859 | I<call_AddSubScalar> is called | |
860 | ||
861 | call_AddSubScalar(7, 4) ; | |
862 | ||
863 | then the output will be | |
864 | ||
865 | Items Returned = 1 | |
866 | Value 1 = 3 | |
867 | ||
868 | In this case the main point to note is that only the last item in the | |
54310121 | 869 | list is returned from the subroutine, I<AddSubtract> actually made it back to |
d1b91892 AD |
870 | I<call_AddSubScalar>. |
871 | ||
872 | ||
873 | =head2 Returning Data from Perl via the parameter list | |
a0d0e21e LW |
874 | |
875 | It is also possible to return values directly via the parameter list - | |
876 | whether it is actually desirable to do it is another matter entirely. | |
877 | ||
d1b91892 AD |
878 | The Perl subroutine, I<Inc>, below takes 2 parameters and increments |
879 | each directly. | |
a0d0e21e LW |
880 | |
881 | sub Inc | |
882 | { | |
883 | ++ $_[0] ; | |
884 | ++ $_[1] ; | |
885 | } | |
886 | ||
887 | and here is a C function to call it. | |
888 | ||
889 | static void | |
890 | call_Inc(a, b) | |
891 | int a ; | |
892 | int b ; | |
893 | { | |
894 | dSP ; | |
895 | int count ; | |
896 | SV * sva ; | |
897 | SV * svb ; | |
898 | ||
899 | ENTER ; | |
900 | SAVETMPS; | |
901 | ||
902 | sva = sv_2mortal(newSViv(a)) ; | |
903 | svb = sv_2mortal(newSViv(b)) ; | |
904 | ||
924508f0 | 905 | PUSHMARK(SP) ; |
a0d0e21e LW |
906 | XPUSHs(sva); |
907 | XPUSHs(svb); | |
908 | PUTBACK ; | |
909 | ||
4929bf7b | 910 | count = call_pv("Inc", G_DISCARD); |
a0d0e21e LW |
911 | |
912 | if (count != 0) | |
d1b91892 AD |
913 | croak ("call_Inc: expected 0 values from 'Inc', got %d\n", |
914 | count) ; | |
a0d0e21e LW |
915 | |
916 | printf ("%d + 1 = %d\n", a, SvIV(sva)) ; | |
917 | printf ("%d + 1 = %d\n", b, SvIV(svb)) ; | |
918 | ||
919 | FREETMPS ; | |
d1b91892 | 920 | LEAVE ; |
a0d0e21e LW |
921 | } |
922 | ||
d1b91892 | 923 | To be able to access the two parameters that were pushed onto the stack |
4929bf7b | 924 | after they return from I<call_pv> it is necessary to make a note |
19799a22 | 925 | of their addresses--thus the two variables C<sva> and C<svb>. |
a0d0e21e | 926 | |
d1b91892 AD |
927 | The reason this is necessary is that the area of the Perl stack which |
928 | held them will very likely have been overwritten by something else by | |
4929bf7b | 929 | the time control returns from I<call_pv>. |
a0d0e21e LW |
930 | |
931 | ||
932 | ||
933 | ||
d1b91892 | 934 | =head2 Using G_EVAL |
a0d0e21e | 935 | |
d1b91892 AD |
936 | Now an example using G_EVAL. Below is a Perl subroutine which computes |
937 | the difference of its 2 parameters. If this would result in a negative | |
938 | result, the subroutine calls I<die>. | |
a0d0e21e LW |
939 | |
940 | sub Subtract | |
941 | { | |
d1b91892 | 942 | my ($a, $b) = @_ ; |
a0d0e21e LW |
943 | |
944 | die "death can be fatal\n" if $a < $b ; | |
945 | ||
d1b91892 | 946 | $a - $b ; |
a0d0e21e LW |
947 | } |
948 | ||
949 | and some C to call it | |
950 | ||
951 | static void | |
952 | call_Subtract(a, b) | |
953 | int a ; | |
954 | int b ; | |
955 | { | |
956 | dSP ; | |
957 | int count ; | |
a0d0e21e LW |
958 | |
959 | ENTER ; | |
960 | SAVETMPS; | |
961 | ||
924508f0 | 962 | PUSHMARK(SP) ; |
a0d0e21e LW |
963 | XPUSHs(sv_2mortal(newSViv(a))); |
964 | XPUSHs(sv_2mortal(newSViv(b))); | |
965 | PUTBACK ; | |
966 | ||
4929bf7b | 967 | count = call_pv("Subtract", G_EVAL|G_SCALAR); |
a0d0e21e | 968 | |
d1b91892 AD |
969 | SPAGAIN ; |
970 | ||
971 | /* Check the eval first */ | |
9cde0e7f | 972 | if (SvTRUE(ERRSV)) |
d1b91892 | 973 | { |
2d8e6c8d GS |
974 | STRLEN n_a; |
975 | printf ("Uh oh - %s\n", SvPV(ERRSV, n_a)) ; | |
d1b91892 AD |
976 | POPs ; |
977 | } | |
978 | else | |
979 | { | |
980 | if (count != 1) | |
981 | croak("call_Subtract: wanted 1 value from 'Subtract', got %d\n", | |
982 | count) ; | |
a0d0e21e | 983 | |
d1b91892 AD |
984 | printf ("%d - %d = %d\n", a, b, POPi) ; |
985 | } | |
a0d0e21e LW |
986 | |
987 | PUTBACK ; | |
988 | FREETMPS ; | |
989 | LEAVE ; | |
a0d0e21e LW |
990 | } |
991 | ||
992 | If I<call_Subtract> is called thus | |
993 | ||
d1b91892 | 994 | call_Subtract(4, 5) |
a0d0e21e LW |
995 | |
996 | the following will be printed | |
997 | ||
d1b91892 | 998 | Uh oh - death can be fatal |
a0d0e21e LW |
999 | |
1000 | Notes | |
1001 | ||
1002 | =over 5 | |
1003 | ||
1004 | =item 1. | |
1005 | ||
d1b91892 AD |
1006 | We want to be able to catch the I<die> so we have used the G_EVAL |
1007 | flag. Not specifying this flag would mean that the program would | |
1008 | terminate immediately at the I<die> statement in the subroutine | |
1009 | I<Subtract>. | |
a0d0e21e LW |
1010 | |
1011 | =item 2. | |
1012 | ||
54310121 | 1013 | The code |
a0d0e21e | 1014 | |
9cde0e7f | 1015 | if (SvTRUE(ERRSV)) |
d1b91892 | 1016 | { |
2d8e6c8d GS |
1017 | STRLEN n_a; |
1018 | printf ("Uh oh - %s\n", SvPV(ERRSV, n_a)) ; | |
d1b91892 AD |
1019 | POPs ; |
1020 | } | |
a0d0e21e | 1021 | |
d1b91892 | 1022 | is the direct equivalent of this bit of Perl |
a0d0e21e | 1023 | |
d1b91892 | 1024 | print "Uh oh - $@\n" if $@ ; |
a0d0e21e | 1025 | |
9cde0e7f GS |
1026 | C<PL_errgv> is a perl global of type C<GV *> that points to the |
1027 | symbol table entry containing the error. C<ERRSV> therefore | |
c07a80fd | 1028 | refers to the C equivalent of C<$@>. |
1029 | ||
d1b91892 | 1030 | =item 3. |
a0d0e21e | 1031 | |
d1b91892 | 1032 | Note that the stack is popped using C<POPs> in the block where |
9cde0e7f | 1033 | C<SvTRUE(ERRSV)> is true. This is necessary because whenever a |
4929bf7b | 1034 | I<call_*> function invoked with G_EVAL|G_SCALAR returns an error, |
5f05dabc | 1035 | the top of the stack holds the value I<undef>. Because we want the |
d1b91892 AD |
1036 | program to continue after detecting this error, it is essential that |
1037 | the stack is tidied up by removing the I<undef>. | |
a0d0e21e LW |
1038 | |
1039 | =back | |
1040 | ||
1041 | ||
c07a80fd | 1042 | =head2 Using G_KEEPERR |
1043 | ||
1044 | Consider this rather facetious example, where we have used an XS | |
1045 | version of the call_Subtract example above inside a destructor: | |
1046 | ||
1047 | package Foo; | |
1048 | sub new { bless {}, $_[0] } | |
54310121 | 1049 | sub Subtract { |
c07a80fd | 1050 | my($a,$b) = @_; |
1051 | die "death can be fatal" if $a < $b ; | |
1052 | $a - $b; | |
1053 | } | |
1054 | sub DESTROY { call_Subtract(5, 4); } | |
1055 | sub foo { die "foo dies"; } | |
1056 | ||
1057 | package main; | |
1058 | eval { Foo->new->foo }; | |
1059 | print "Saw: $@" if $@; # should be, but isn't | |
1060 | ||
1061 | This example will fail to recognize that an error occurred inside the | |
1062 | C<eval {}>. Here's why: the call_Subtract code got executed while perl | |
5f05dabc | 1063 | was cleaning up temporaries when exiting the eval block, and because |
4929bf7b | 1064 | call_Subtract is implemented with I<call_pv> using the G_EVAL |
c07a80fd | 1065 | flag, it promptly reset C<$@>. This results in the failure of the |
1066 | outermost test for C<$@>, and thereby the failure of the error trap. | |
1067 | ||
4929bf7b | 1068 | Appending the G_KEEPERR flag, so that the I<call_pv> call in |
c07a80fd | 1069 | call_Subtract reads: |
1070 | ||
4929bf7b | 1071 | count = call_pv("Subtract", G_EVAL|G_SCALAR|G_KEEPERR); |
c07a80fd | 1072 | |
1073 | will preserve the error and restore reliable error handling. | |
1074 | ||
4929bf7b | 1075 | =head2 Using call_sv |
a0d0e21e | 1076 | |
d1b91892 AD |
1077 | In all the previous examples I have 'hard-wired' the name of the Perl |
1078 | subroutine to be called from C. Most of the time though, it is more | |
1079 | convenient to be able to specify the name of the Perl subroutine from | |
1080 | within the Perl script. | |
a0d0e21e LW |
1081 | |
1082 | Consider the Perl code below | |
1083 | ||
d1b91892 AD |
1084 | sub fred |
1085 | { | |
1086 | print "Hello there\n" ; | |
1087 | } | |
1088 | ||
1089 | CallSubPV("fred") ; | |
1090 | ||
1091 | Here is a snippet of XSUB which defines I<CallSubPV>. | |
1092 | ||
1093 | void | |
1094 | CallSubPV(name) | |
1095 | char * name | |
1096 | CODE: | |
924508f0 | 1097 | PUSHMARK(SP) ; |
4929bf7b | 1098 | call_pv(name, G_DISCARD|G_NOARGS) ; |
a0d0e21e | 1099 | |
54310121 | 1100 | That is fine as far as it goes. The thing is, the Perl subroutine |
5f05dabc | 1101 | can be specified as only a string. For Perl 4 this was adequate, |
d1b91892 | 1102 | but Perl 5 allows references to subroutines and anonymous subroutines. |
4929bf7b | 1103 | This is where I<call_sv> is useful. |
d1b91892 AD |
1104 | |
1105 | The code below for I<CallSubSV> is identical to I<CallSubPV> except | |
1106 | that the C<name> parameter is now defined as an SV* and we use | |
4929bf7b | 1107 | I<call_sv> instead of I<call_pv>. |
d1b91892 AD |
1108 | |
1109 | void | |
1110 | CallSubSV(name) | |
1111 | SV * name | |
1112 | CODE: | |
924508f0 | 1113 | PUSHMARK(SP) ; |
4929bf7b | 1114 | call_sv(name, G_DISCARD|G_NOARGS) ; |
a0d0e21e | 1115 | |
5f05dabc | 1116 | Because we are using an SV to call I<fred> the following can all be used |
a0d0e21e | 1117 | |
d1b91892 AD |
1118 | CallSubSV("fred") ; |
1119 | CallSubSV(\&fred) ; | |
1120 | $ref = \&fred ; | |
1121 | CallSubSV($ref) ; | |
1122 | CallSubSV( sub { print "Hello there\n" } ) ; | |
a0d0e21e | 1123 | |
4929bf7b | 1124 | As you can see, I<call_sv> gives you much greater flexibility in |
d1b91892 AD |
1125 | how you can specify the Perl subroutine. |
1126 | ||
1127 | You should note that if it is necessary to store the SV (C<name> in the | |
1128 | example above) which corresponds to the Perl subroutine so that it can | |
5f05dabc | 1129 | be used later in the program, it not enough just to store a copy of the |
d1b91892 AD |
1130 | pointer to the SV. Say the code above had been like this |
1131 | ||
1132 | static SV * rememberSub ; | |
1133 | ||
1134 | void | |
1135 | SaveSub1(name) | |
1136 | SV * name | |
1137 | CODE: | |
1138 | rememberSub = name ; | |
1139 | ||
1140 | void | |
1141 | CallSavedSub1() | |
1142 | CODE: | |
924508f0 | 1143 | PUSHMARK(SP) ; |
4929bf7b | 1144 | call_sv(rememberSub, G_DISCARD|G_NOARGS) ; |
a0d0e21e | 1145 | |
d1b91892 AD |
1146 | The reason this is wrong is that by the time you come to use the |
1147 | pointer C<rememberSub> in C<CallSavedSub1>, it may or may not still refer | |
1148 | to the Perl subroutine that was recorded in C<SaveSub1>. This is | |
1149 | particularly true for these cases | |
a0d0e21e | 1150 | |
d1b91892 AD |
1151 | SaveSub1(\&fred) ; |
1152 | CallSavedSub1() ; | |
a0d0e21e | 1153 | |
d1b91892 AD |
1154 | SaveSub1( sub { print "Hello there\n" } ) ; |
1155 | CallSavedSub1() ; | |
a0d0e21e | 1156 | |
d1b91892 | 1157 | By the time each of the C<SaveSub1> statements above have been executed, |
54310121 | 1158 | the SV*s which corresponded to the parameters will no longer exist. |
d1b91892 | 1159 | Expect an error message from Perl of the form |
a0d0e21e | 1160 | |
d1b91892 | 1161 | Can't use an undefined value as a subroutine reference at ... |
a0d0e21e | 1162 | |
d1b91892 | 1163 | for each of the C<CallSavedSub1> lines. |
a0d0e21e | 1164 | |
54310121 | 1165 | Similarly, with this code |
a0d0e21e | 1166 | |
d1b91892 AD |
1167 | $ref = \&fred ; |
1168 | SaveSub1($ref) ; | |
1169 | $ref = 47 ; | |
1170 | CallSavedSub1() ; | |
a0d0e21e | 1171 | |
54310121 | 1172 | you can expect one of these messages (which you actually get is dependent on |
1173 | the version of Perl you are using) | |
a0d0e21e | 1174 | |
d1b91892 AD |
1175 | Not a CODE reference at ... |
1176 | Undefined subroutine &main::47 called ... | |
a0d0e21e | 1177 | |
19799a22 | 1178 | The variable $ref may have referred to the subroutine C<fred> |
d1b91892 | 1179 | whenever the call to C<SaveSub1> was made but by the time |
5f05dabc | 1180 | C<CallSavedSub1> gets called it now holds the number C<47>. Because we |
d1b91892 | 1181 | saved only a pointer to the original SV in C<SaveSub1>, any changes to |
19799a22 | 1182 | $ref will be tracked by the pointer C<rememberSub>. This means that |
d1b91892 AD |
1183 | whenever C<CallSavedSub1> gets called, it will attempt to execute the |
1184 | code which is referenced by the SV* C<rememberSub>. In this case | |
1185 | though, it now refers to the integer C<47>, so expect Perl to complain | |
1186 | loudly. | |
a0d0e21e | 1187 | |
d1b91892 | 1188 | A similar but more subtle problem is illustrated with this code |
a0d0e21e | 1189 | |
d1b91892 AD |
1190 | $ref = \&fred ; |
1191 | SaveSub1($ref) ; | |
1192 | $ref = \&joe ; | |
1193 | CallSavedSub1() ; | |
a0d0e21e | 1194 | |
d1b91892 | 1195 | This time whenever C<CallSavedSub1> get called it will execute the Perl |
54310121 | 1196 | subroutine C<joe> (assuming it exists) rather than C<fred> as was |
d1b91892 | 1197 | originally requested in the call to C<SaveSub1>. |
a0d0e21e | 1198 | |
d1b91892 AD |
1199 | To get around these problems it is necessary to take a full copy of the |
1200 | SV. The code below shows C<SaveSub2> modified to do that | |
a0d0e21e | 1201 | |
d1b91892 AD |
1202 | static SV * keepSub = (SV*)NULL ; |
1203 | ||
1204 | void | |
1205 | SaveSub2(name) | |
1206 | SV * name | |
1207 | CODE: | |
1208 | /* Take a copy of the callback */ | |
1209 | if (keepSub == (SV*)NULL) | |
1210 | /* First time, so create a new SV */ | |
1211 | keepSub = newSVsv(name) ; | |
1212 | else | |
1213 | /* Been here before, so overwrite */ | |
1214 | SvSetSV(keepSub, name) ; | |
1215 | ||
1216 | void | |
1217 | CallSavedSub2() | |
1218 | CODE: | |
924508f0 | 1219 | PUSHMARK(SP) ; |
4929bf7b | 1220 | call_sv(keepSub, G_DISCARD|G_NOARGS) ; |
d1b91892 | 1221 | |
5f05dabc | 1222 | To avoid creating a new SV every time C<SaveSub2> is called, |
d1b91892 AD |
1223 | the function first checks to see if it has been called before. If not, |
1224 | then space for a new SV is allocated and the reference to the Perl | |
1225 | subroutine, C<name> is copied to the variable C<keepSub> in one | |
1226 | operation using C<newSVsv>. Thereafter, whenever C<SaveSub2> is called | |
1227 | the existing SV, C<keepSub>, is overwritten with the new value using | |
1228 | C<SvSetSV>. | |
1229 | ||
4929bf7b | 1230 | =head2 Using call_argv |
d1b91892 AD |
1231 | |
1232 | Here is a Perl subroutine which prints whatever parameters are passed | |
1233 | to it. | |
1234 | ||
1235 | sub PrintList | |
1236 | { | |
1237 | my(@list) = @_ ; | |
1238 | ||
1239 | foreach (@list) { print "$_\n" } | |
1240 | } | |
1241 | ||
4929bf7b | 1242 | and here is an example of I<call_argv> which will call |
d1b91892 AD |
1243 | I<PrintList>. |
1244 | ||
1245 | static char * words[] = {"alpha", "beta", "gamma", "delta", NULL} ; | |
1246 | ||
1247 | static void | |
1248 | call_PrintList() | |
1249 | { | |
1250 | dSP ; | |
1251 | ||
4929bf7b | 1252 | call_argv("PrintList", G_DISCARD, words) ; |
d1b91892 AD |
1253 | } |
1254 | ||
1255 | Note that it is not necessary to call C<PUSHMARK> in this instance. | |
4929bf7b | 1256 | This is because I<call_argv> will do it for you. |
d1b91892 | 1257 | |
4929bf7b | 1258 | =head2 Using call_method |
a0d0e21e LW |
1259 | |
1260 | Consider the following Perl code | |
1261 | ||
d1b91892 AD |
1262 | { |
1263 | package Mine ; | |
1264 | ||
1265 | sub new | |
1266 | { | |
1267 | my($type) = shift ; | |
1268 | bless [@_] | |
1269 | } | |
1270 | ||
1271 | sub Display | |
1272 | { | |
1273 | my ($self, $index) = @_ ; | |
1274 | print "$index: $$self[$index]\n" ; | |
1275 | } | |
1276 | ||
1277 | sub PrintID | |
1278 | { | |
1279 | my($class) = @_ ; | |
1280 | print "This is Class $class version 1.0\n" ; | |
1281 | } | |
1282 | } | |
1283 | ||
5f05dabc | 1284 | It implements just a very simple class to manage an array. Apart from |
d1b91892 | 1285 | the constructor, C<new>, it declares methods, one static and one |
5f05dabc | 1286 | virtual. The static method, C<PrintID>, prints out simply the class |
d1b91892 AD |
1287 | name and a version number. The virtual method, C<Display>, prints out a |
1288 | single element of the array. Here is an all Perl example of using it. | |
1289 | ||
1290 | $a = new Mine ('red', 'green', 'blue') ; | |
1291 | $a->Display(1) ; | |
1292 | PrintID Mine; | |
a0d0e21e | 1293 | |
d1b91892 | 1294 | will print |
a0d0e21e | 1295 | |
d1b91892 | 1296 | 1: green |
54310121 | 1297 | This is Class Mine version 1.0 |
a0d0e21e | 1298 | |
d1b91892 AD |
1299 | Calling a Perl method from C is fairly straightforward. The following |
1300 | things are required | |
a0d0e21e | 1301 | |
d1b91892 AD |
1302 | =over 5 |
1303 | ||
1304 | =item * | |
1305 | ||
1306 | a reference to the object for a virtual method or the name of the class | |
1307 | for a static method. | |
1308 | ||
1309 | =item * | |
1310 | ||
1311 | the name of the method. | |
1312 | ||
1313 | =item * | |
1314 | ||
1315 | any other parameters specific to the method. | |
1316 | ||
1317 | =back | |
1318 | ||
1319 | Here is a simple XSUB which illustrates the mechanics of calling both | |
1320 | the C<PrintID> and C<Display> methods from C. | |
1321 | ||
1322 | void | |
1323 | call_Method(ref, method, index) | |
1324 | SV * ref | |
1325 | char * method | |
1326 | int index | |
1327 | CODE: | |
924508f0 | 1328 | PUSHMARK(SP); |
d1b91892 AD |
1329 | XPUSHs(ref); |
1330 | XPUSHs(sv_2mortal(newSViv(index))) ; | |
1331 | PUTBACK; | |
1332 | ||
4929bf7b | 1333 | call_method(method, G_DISCARD) ; |
d1b91892 AD |
1334 | |
1335 | void | |
1336 | call_PrintID(class, method) | |
1337 | char * class | |
1338 | char * method | |
1339 | CODE: | |
924508f0 | 1340 | PUSHMARK(SP); |
d1b91892 AD |
1341 | XPUSHs(sv_2mortal(newSVpv(class, 0))) ; |
1342 | PUTBACK; | |
1343 | ||
4929bf7b | 1344 | call_method(method, G_DISCARD) ; |
d1b91892 AD |
1345 | |
1346 | ||
1347 | So the methods C<PrintID> and C<Display> can be invoked like this | |
1348 | ||
1349 | $a = new Mine ('red', 'green', 'blue') ; | |
1350 | call_Method($a, 'Display', 1) ; | |
1351 | call_PrintID('Mine', 'PrintID') ; | |
1352 | ||
1353 | The only thing to note is that in both the static and virtual methods, | |
19799a22 | 1354 | the method name is not passed via the stack--it is used as the first |
4929bf7b | 1355 | parameter to I<call_method>. |
d1b91892 | 1356 | |
54310121 | 1357 | =head2 Using GIMME_V |
d1b91892 | 1358 | |
54310121 | 1359 | Here is a trivial XSUB which prints the context in which it is |
d1b91892 AD |
1360 | currently executing. |
1361 | ||
1362 | void | |
1363 | PrintContext() | |
1364 | CODE: | |
54310121 | 1365 | I32 gimme = GIMME_V; |
1366 | if (gimme == G_VOID) | |
1367 | printf ("Context is Void\n") ; | |
1368 | else if (gimme == G_SCALAR) | |
d1b91892 AD |
1369 | printf ("Context is Scalar\n") ; |
1370 | else | |
1371 | printf ("Context is Array\n") ; | |
1372 | ||
1373 | and here is some Perl to test it | |
1374 | ||
54310121 | 1375 | PrintContext ; |
d1b91892 AD |
1376 | $a = PrintContext ; |
1377 | @a = PrintContext ; | |
1378 | ||
1379 | The output from that will be | |
1380 | ||
54310121 | 1381 | Context is Void |
d1b91892 AD |
1382 | Context is Scalar |
1383 | Context is Array | |
1384 | ||
1385 | =head2 Using Perl to dispose of temporaries | |
1386 | ||
1387 | In the examples given to date, any temporaries created in the callback | |
4929bf7b | 1388 | (i.e., parameters passed on the stack to the I<call_*> function or |
d1b91892 AD |
1389 | values returned via the stack) have been freed by one of these methods |
1390 | ||
1391 | =over 5 | |
1392 | ||
1393 | =item * | |
1394 | ||
4929bf7b | 1395 | specifying the G_DISCARD flag with I<call_*>. |
d1b91892 AD |
1396 | |
1397 | =item * | |
1398 | ||
1399 | explicitly disposed of using the C<ENTER>/C<SAVETMPS> - | |
1400 | C<FREETMPS>/C<LEAVE> pairing. | |
1401 | ||
1402 | =back | |
1403 | ||
1404 | There is another method which can be used, namely letting Perl do it | |
1405 | for you automatically whenever it regains control after the callback | |
1406 | has terminated. This is done by simply not using the | |
1407 | ||
1408 | ENTER ; | |
1409 | SAVETMPS ; | |
1410 | ... | |
1411 | FREETMPS ; | |
1412 | LEAVE ; | |
1413 | ||
1414 | sequence in the callback (and not, of course, specifying the G_DISCARD | |
1415 | flag). | |
1416 | ||
1417 | If you are going to use this method you have to be aware of a possible | |
1418 | memory leak which can arise under very specific circumstances. To | |
1419 | explain these circumstances you need to know a bit about the flow of | |
1420 | control between Perl and the callback routine. | |
1421 | ||
1422 | The examples given at the start of the document (an error handler and | |
1423 | an event driven program) are typical of the two main sorts of flow | |
1424 | control that you are likely to encounter with callbacks. There is a | |
1425 | very important distinction between them, so pay attention. | |
1426 | ||
1427 | In the first example, an error handler, the flow of control could be as | |
1428 | follows. You have created an interface to an external library. | |
1429 | Control can reach the external library like this | |
1430 | ||
1431 | perl --> XSUB --> external library | |
1432 | ||
1433 | Whilst control is in the library, an error condition occurs. You have | |
1434 | previously set up a Perl callback to handle this situation, so it will | |
1435 | get executed. Once the callback has finished, control will drop back to | |
1436 | Perl again. Here is what the flow of control will be like in that | |
1437 | situation | |
1438 | ||
1439 | perl --> XSUB --> external library | |
1440 | ... | |
1441 | error occurs | |
1442 | ... | |
4929bf7b | 1443 | external library --> call_* --> perl |
d1b91892 | 1444 | | |
4929bf7b | 1445 | perl <-- XSUB <-- external library <-- call_* <----+ |
d1b91892 | 1446 | |
4929bf7b | 1447 | After processing of the error using I<call_*> is completed, |
d1b91892 AD |
1448 | control reverts back to Perl more or less immediately. |
1449 | ||
1450 | In the diagram, the further right you go the more deeply nested the | |
1451 | scope is. It is only when control is back with perl on the extreme | |
1452 | left of the diagram that you will have dropped back to the enclosing | |
1453 | scope and any temporaries you have left hanging around will be freed. | |
1454 | ||
1455 | In the second example, an event driven program, the flow of control | |
1456 | will be more like this | |
1457 | ||
1458 | perl --> XSUB --> event handler | |
1459 | ... | |
4929bf7b | 1460 | event handler --> call_* --> perl |
d1b91892 | 1461 | | |
4929bf7b | 1462 | event handler <-- call_* <----+ |
d1b91892 | 1463 | ... |
4929bf7b | 1464 | event handler --> call_* --> perl |
d1b91892 | 1465 | | |
4929bf7b | 1466 | event handler <-- call_* <----+ |
d1b91892 | 1467 | ... |
4929bf7b | 1468 | event handler --> call_* --> perl |
d1b91892 | 1469 | | |
4929bf7b | 1470 | event handler <-- call_* <----+ |
d1b91892 AD |
1471 | |
1472 | In this case the flow of control can consist of only the repeated | |
1473 | sequence | |
1474 | ||
4929bf7b | 1475 | event handler --> call_* --> perl |
d1b91892 | 1476 | |
54310121 | 1477 | for practically the complete duration of the program. This means that |
1478 | control may I<never> drop back to the surrounding scope in Perl at the | |
1479 | extreme left. | |
d1b91892 AD |
1480 | |
1481 | So what is the big problem? Well, if you are expecting Perl to tidy up | |
1482 | those temporaries for you, you might be in for a long wait. For Perl | |
5f05dabc | 1483 | to dispose of your temporaries, control must drop back to the |
d1b91892 AD |
1484 | enclosing scope at some stage. In the event driven scenario that may |
1485 | never happen. This means that as time goes on, your program will | |
1486 | create more and more temporaries, none of which will ever be freed. As | |
1487 | each of these temporaries consumes some memory your program will | |
19799a22 | 1488 | eventually consume all the available memory in your system--kapow! |
d1b91892 | 1489 | |
19799a22 | 1490 | So here is the bottom line--if you are sure that control will revert |
d1b91892 | 1491 | back to the enclosing Perl scope fairly quickly after the end of your |
5f05dabc | 1492 | callback, then it isn't absolutely necessary to dispose explicitly of |
d1b91892 AD |
1493 | any temporaries you may have created. Mind you, if you are at all |
1494 | uncertain about what to do, it doesn't do any harm to tidy up anyway. | |
1495 | ||
1496 | ||
1497 | =head2 Strategies for storing Callback Context Information | |
1498 | ||
1499 | ||
1500 | Potentially one of the trickiest problems to overcome when designing a | |
1501 | callback interface can be figuring out how to store the mapping between | |
1502 | the C callback function and the Perl equivalent. | |
1503 | ||
1504 | To help understand why this can be a real problem first consider how a | |
1505 | callback is set up in an all C environment. Typically a C API will | |
1506 | provide a function to register a callback. This will expect a pointer | |
1507 | to a function as one of its parameters. Below is a call to a | |
1508 | hypothetical function C<register_fatal> which registers the C function | |
1509 | to get called when a fatal error occurs. | |
1510 | ||
1511 | register_fatal(cb1) ; | |
1512 | ||
1513 | The single parameter C<cb1> is a pointer to a function, so you must | |
1514 | have defined C<cb1> in your code, say something like this | |
1515 | ||
1516 | static void | |
1517 | cb1() | |
1518 | { | |
1519 | printf ("Fatal Error\n") ; | |
1520 | exit(1) ; | |
1521 | } | |
1522 | ||
1523 | Now change that to call a Perl subroutine instead | |
1524 | ||
1525 | static SV * callback = (SV*)NULL; | |
1526 | ||
1527 | static void | |
1528 | cb1() | |
1529 | { | |
1530 | dSP ; | |
1531 | ||
924508f0 | 1532 | PUSHMARK(SP) ; |
d1b91892 AD |
1533 | |
1534 | /* Call the Perl sub to process the callback */ | |
4929bf7b | 1535 | call_sv(callback, G_DISCARD) ; |
d1b91892 AD |
1536 | } |
1537 | ||
1538 | ||
1539 | void | |
1540 | register_fatal(fn) | |
1541 | SV * fn | |
1542 | CODE: | |
1543 | /* Remember the Perl sub */ | |
1544 | if (callback == (SV*)NULL) | |
1545 | callback = newSVsv(fn) ; | |
1546 | else | |
1547 | SvSetSV(callback, fn) ; | |
1548 | ||
1549 | /* register the callback with the external library */ | |
1550 | register_fatal(cb1) ; | |
1551 | ||
1552 | where the Perl equivalent of C<register_fatal> and the callback it | |
1553 | registers, C<pcb1>, might look like this | |
1554 | ||
1555 | # Register the sub pcb1 | |
1556 | register_fatal(\&pcb1) ; | |
1557 | ||
1558 | sub pcb1 | |
1559 | { | |
1560 | die "I'm dying...\n" ; | |
1561 | } | |
1562 | ||
1563 | The mapping between the C callback and the Perl equivalent is stored in | |
1564 | the global variable C<callback>. | |
1565 | ||
5f05dabc | 1566 | This will be adequate if you ever need to have only one callback |
d1b91892 AD |
1567 | registered at any time. An example could be an error handler like the |
1568 | code sketched out above. Remember though, repeated calls to | |
1569 | C<register_fatal> will replace the previously registered callback | |
1570 | function with the new one. | |
1571 | ||
1572 | Say for example you want to interface to a library which allows asynchronous | |
1573 | file i/o. In this case you may be able to register a callback whenever | |
1574 | a read operation has completed. To be of any use we want to be able to | |
1575 | call separate Perl subroutines for each file that is opened. As it | |
1576 | stands, the error handler example above would not be adequate as it | |
1577 | allows only a single callback to be defined at any time. What we | |
1578 | require is a means of storing the mapping between the opened file and | |
1579 | the Perl subroutine we want to be called for that file. | |
1580 | ||
1581 | Say the i/o library has a function C<asynch_read> which associates a C | |
19799a22 | 1582 | function C<ProcessRead> with a file handle C<fh>--this assumes that it |
d1b91892 AD |
1583 | has also provided some routine to open the file and so obtain the file |
1584 | handle. | |
1585 | ||
1586 | asynch_read(fh, ProcessRead) | |
1587 | ||
1588 | This may expect the C I<ProcessRead> function of this form | |
1589 | ||
1590 | void | |
1591 | ProcessRead(fh, buffer) | |
1592 | int fh ; | |
1593 | char * buffer ; | |
1594 | { | |
54310121 | 1595 | ... |
d1b91892 AD |
1596 | } |
1597 | ||
1598 | To provide a Perl interface to this library we need to be able to map | |
1599 | between the C<fh> parameter and the Perl subroutine we want called. A | |
1600 | hash is a convenient mechanism for storing this mapping. The code | |
1601 | below shows a possible implementation | |
1602 | ||
1603 | static HV * Mapping = (HV*)NULL ; | |
a0d0e21e | 1604 | |
d1b91892 AD |
1605 | void |
1606 | asynch_read(fh, callback) | |
1607 | int fh | |
1608 | SV * callback | |
1609 | CODE: | |
1610 | /* If the hash doesn't already exist, create it */ | |
1611 | if (Mapping == (HV*)NULL) | |
1612 | Mapping = newHV() ; | |
1613 | ||
1614 | /* Save the fh -> callback mapping */ | |
1615 | hv_store(Mapping, (char*)&fh, sizeof(fh), newSVsv(callback), 0) ; | |
1616 | ||
1617 | /* Register with the C Library */ | |
1618 | asynch_read(fh, asynch_read_if) ; | |
1619 | ||
1620 | and C<asynch_read_if> could look like this | |
1621 | ||
1622 | static void | |
1623 | asynch_read_if(fh, buffer) | |
1624 | int fh ; | |
1625 | char * buffer ; | |
1626 | { | |
1627 | dSP ; | |
1628 | SV ** sv ; | |
1629 | ||
1630 | /* Get the callback associated with fh */ | |
1631 | sv = hv_fetch(Mapping, (char*)&fh , sizeof(fh), FALSE) ; | |
1632 | if (sv == (SV**)NULL) | |
1633 | croak("Internal error...\n") ; | |
1634 | ||
924508f0 | 1635 | PUSHMARK(SP) ; |
d1b91892 AD |
1636 | XPUSHs(sv_2mortal(newSViv(fh))) ; |
1637 | XPUSHs(sv_2mortal(newSVpv(buffer, 0))) ; | |
1638 | PUTBACK ; | |
1639 | ||
1640 | /* Call the Perl sub */ | |
4929bf7b | 1641 | call_sv(*sv, G_DISCARD) ; |
d1b91892 AD |
1642 | } |
1643 | ||
1644 | For completeness, here is C<asynch_close>. This shows how to remove | |
1645 | the entry from the hash C<Mapping>. | |
1646 | ||
1647 | void | |
1648 | asynch_close(fh) | |
1649 | int fh | |
1650 | CODE: | |
1651 | /* Remove the entry from the hash */ | |
1652 | (void) hv_delete(Mapping, (char*)&fh, sizeof(fh), G_DISCARD) ; | |
a0d0e21e | 1653 | |
d1b91892 AD |
1654 | /* Now call the real asynch_close */ |
1655 | asynch_close(fh) ; | |
a0d0e21e | 1656 | |
d1b91892 AD |
1657 | So the Perl interface would look like this |
1658 | ||
1659 | sub callback1 | |
1660 | { | |
1661 | my($handle, $buffer) = @_ ; | |
1662 | } | |
a0d0e21e | 1663 | |
d1b91892 AD |
1664 | # Register the Perl callback |
1665 | asynch_read($fh, \&callback1) ; | |
a0d0e21e | 1666 | |
d1b91892 AD |
1667 | asynch_close($fh) ; |
1668 | ||
1669 | The mapping between the C callback and Perl is stored in the global | |
1670 | hash C<Mapping> this time. Using a hash has the distinct advantage that | |
1671 | it allows an unlimited number of callbacks to be registered. | |
1672 | ||
1673 | What if the interface provided by the C callback doesn't contain a | |
1674 | parameter which allows the file handle to Perl subroutine mapping? Say | |
1675 | in the asynchronous i/o package, the callback function gets passed only | |
1676 | the C<buffer> parameter like this | |
1677 | ||
1678 | void | |
1679 | ProcessRead(buffer) | |
1680 | char * buffer ; | |
1681 | { | |
1682 | ... | |
1683 | } | |
a0d0e21e | 1684 | |
d1b91892 AD |
1685 | Without the file handle there is no straightforward way to map from the |
1686 | C callback to the Perl subroutine. | |
a0d0e21e | 1687 | |
54310121 | 1688 | In this case a possible way around this problem is to predefine a |
d1b91892 AD |
1689 | series of C functions to act as the interface to Perl, thus |
1690 | ||
1691 | #define MAX_CB 3 | |
1692 | #define NULL_HANDLE -1 | |
1693 | typedef void (*FnMap)() ; | |
1694 | ||
1695 | struct MapStruct { | |
1696 | FnMap Function ; | |
1697 | SV * PerlSub ; | |
1698 | int Handle ; | |
1699 | } ; | |
1700 | ||
1701 | static void fn1() ; | |
1702 | static void fn2() ; | |
1703 | static void fn3() ; | |
1704 | ||
1705 | static struct MapStruct Map [MAX_CB] = | |
1706 | { | |
1707 | { fn1, NULL, NULL_HANDLE }, | |
1708 | { fn2, NULL, NULL_HANDLE }, | |
1709 | { fn3, NULL, NULL_HANDLE } | |
1710 | } ; | |
1711 | ||
1712 | static void | |
1713 | Pcb(index, buffer) | |
1714 | int index ; | |
1715 | char * buffer ; | |
1716 | { | |
1717 | dSP ; | |
1718 | ||
924508f0 | 1719 | PUSHMARK(SP) ; |
d1b91892 AD |
1720 | XPUSHs(sv_2mortal(newSVpv(buffer, 0))) ; |
1721 | PUTBACK ; | |
1722 | ||
1723 | /* Call the Perl sub */ | |
4929bf7b | 1724 | call_sv(Map[index].PerlSub, G_DISCARD) ; |
d1b91892 AD |
1725 | } |
1726 | ||
1727 | static void | |
1728 | fn1(buffer) | |
1729 | char * buffer ; | |
1730 | { | |
1731 | Pcb(0, buffer) ; | |
1732 | } | |
1733 | ||
1734 | static void | |
1735 | fn2(buffer) | |
1736 | char * buffer ; | |
1737 | { | |
1738 | Pcb(1, buffer) ; | |
1739 | } | |
1740 | ||
1741 | static void | |
1742 | fn3(buffer) | |
1743 | char * buffer ; | |
1744 | { | |
1745 | Pcb(2, buffer) ; | |
1746 | } | |
1747 | ||
1748 | void | |
1749 | array_asynch_read(fh, callback) | |
1750 | int fh | |
1751 | SV * callback | |
1752 | CODE: | |
1753 | int index ; | |
1754 | int null_index = MAX_CB ; | |
1755 | ||
1756 | /* Find the same handle or an empty entry */ | |
1757 | for (index = 0 ; index < MAX_CB ; ++index) | |
1758 | { | |
1759 | if (Map[index].Handle == fh) | |
1760 | break ; | |
1761 | ||
1762 | if (Map[index].Handle == NULL_HANDLE) | |
1763 | null_index = index ; | |
1764 | } | |
1765 | ||
1766 | if (index == MAX_CB && null_index == MAX_CB) | |
1767 | croak ("Too many callback functions registered\n") ; | |
1768 | ||
1769 | if (index == MAX_CB) | |
1770 | index = null_index ; | |
1771 | ||
1772 | /* Save the file handle */ | |
1773 | Map[index].Handle = fh ; | |
1774 | ||
1775 | /* Remember the Perl sub */ | |
1776 | if (Map[index].PerlSub == (SV*)NULL) | |
1777 | Map[index].PerlSub = newSVsv(callback) ; | |
1778 | else | |
1779 | SvSetSV(Map[index].PerlSub, callback) ; | |
1780 | ||
1781 | asynch_read(fh, Map[index].Function) ; | |
1782 | ||
1783 | void | |
1784 | array_asynch_close(fh) | |
1785 | int fh | |
1786 | CODE: | |
1787 | int index ; | |
1788 | ||
1789 | /* Find the file handle */ | |
1790 | for (index = 0; index < MAX_CB ; ++ index) | |
1791 | if (Map[index].Handle == fh) | |
1792 | break ; | |
1793 | ||
1794 | if (index == MAX_CB) | |
1795 | croak ("could not close fh %d\n", fh) ; | |
1796 | ||
1797 | Map[index].Handle = NULL_HANDLE ; | |
1798 | SvREFCNT_dec(Map[index].PerlSub) ; | |
1799 | Map[index].PerlSub = (SV*)NULL ; | |
1800 | ||
1801 | asynch_close(fh) ; | |
1802 | ||
5f05dabc | 1803 | In this case the functions C<fn1>, C<fn2>, and C<fn3> are used to |
d1b91892 | 1804 | remember the Perl subroutine to be called. Each of the functions holds |
4a6725af | 1805 | a separate hard-wired index which is used in the function C<Pcb> to |
d1b91892 AD |
1806 | access the C<Map> array and actually call the Perl subroutine. |
1807 | ||
1808 | There are some obvious disadvantages with this technique. | |
1809 | ||
1810 | Firstly, the code is considerably more complex than with the previous | |
1811 | example. | |
1812 | ||
4a6725af | 1813 | Secondly, there is a hard-wired limit (in this case 3) to the number of |
d1b91892 AD |
1814 | callbacks that can exist simultaneously. The only way to increase the |
1815 | limit is by modifying the code to add more functions and then | |
54310121 | 1816 | recompiling. None the less, as long as the number of functions is |
d1b91892 AD |
1817 | chosen with some care, it is still a workable solution and in some |
1818 | cases is the only one available. | |
1819 | ||
1820 | To summarize, here are a number of possible methods for you to consider | |
1821 | for storing the mapping between C and the Perl callback | |
1822 | ||
1823 | =over 5 | |
1824 | ||
1825 | =item 1. Ignore the problem - Allow only 1 callback | |
1826 | ||
1827 | For a lot of situations, like interfacing to an error handler, this may | |
1828 | be a perfectly adequate solution. | |
1829 | ||
1830 | =item 2. Create a sequence of callbacks - hard wired limit | |
1831 | ||
1832 | If it is impossible to tell from the parameters passed back from the C | |
1833 | callback what the context is, then you may need to create a sequence of C | |
1834 | callback interface functions, and store pointers to each in an array. | |
1835 | ||
1836 | =item 3. Use a parameter to map to the Perl callback | |
1837 | ||
1838 | A hash is an ideal mechanism to store the mapping between C and Perl. | |
1839 | ||
1840 | =back | |
a0d0e21e | 1841 | |
a0d0e21e LW |
1842 | |
1843 | =head2 Alternate Stack Manipulation | |
1844 | ||
a0d0e21e | 1845 | |
d1b91892 AD |
1846 | Although I have made use of only the C<POP*> macros to access values |
1847 | returned from Perl subroutines, it is also possible to bypass these | |
8e07c86e | 1848 | macros and read the stack using the C<ST> macro (See L<perlxs> for a |
d1b91892 AD |
1849 | full description of the C<ST> macro). |
1850 | ||
1851 | Most of the time the C<POP*> macros should be adequate, the main | |
1852 | problem with them is that they force you to process the returned values | |
1853 | in sequence. This may not be the most suitable way to process the | |
1854 | values in some cases. What we want is to be able to access the stack in | |
1855 | a random order. The C<ST> macro as used when coding an XSUB is ideal | |
1856 | for this purpose. | |
1857 | ||
1858 | The code below is the example given in the section I<Returning a list | |
1859 | of values> recoded to use C<ST> instead of C<POP*>. | |
1860 | ||
1861 | static void | |
1862 | call_AddSubtract2(a, b) | |
1863 | int a ; | |
1864 | int b ; | |
1865 | { | |
1866 | dSP ; | |
1867 | I32 ax ; | |
1868 | int count ; | |
1869 | ||
1870 | ENTER ; | |
1871 | SAVETMPS; | |
1872 | ||
924508f0 | 1873 | PUSHMARK(SP) ; |
d1b91892 AD |
1874 | XPUSHs(sv_2mortal(newSViv(a))); |
1875 | XPUSHs(sv_2mortal(newSViv(b))); | |
1876 | PUTBACK ; | |
1877 | ||
4929bf7b | 1878 | count = call_pv("AddSubtract", G_ARRAY); |
d1b91892 AD |
1879 | |
1880 | SPAGAIN ; | |
924508f0 | 1881 | SP -= count ; |
9cde0e7f | 1882 | ax = (SP - PL_stack_base) + 1 ; |
d1b91892 AD |
1883 | |
1884 | if (count != 2) | |
1885 | croak("Big trouble\n") ; | |
a0d0e21e | 1886 | |
d1b91892 AD |
1887 | printf ("%d + %d = %d\n", a, b, SvIV(ST(0))) ; |
1888 | printf ("%d - %d = %d\n", a, b, SvIV(ST(1))) ; | |
1889 | ||
1890 | PUTBACK ; | |
1891 | FREETMPS ; | |
1892 | LEAVE ; | |
1893 | } | |
1894 | ||
1895 | Notes | |
1896 | ||
1897 | =over 5 | |
1898 | ||
1899 | =item 1. | |
1900 | ||
1901 | Notice that it was necessary to define the variable C<ax>. This is | |
1902 | because the C<ST> macro expects it to exist. If we were in an XSUB it | |
1903 | would not be necessary to define C<ax> as it is already defined for | |
1904 | you. | |
1905 | ||
1906 | =item 2. | |
1907 | ||
1908 | The code | |
1909 | ||
1910 | SPAGAIN ; | |
924508f0 | 1911 | SP -= count ; |
9cde0e7f | 1912 | ax = (SP - PL_stack_base) + 1 ; |
d1b91892 AD |
1913 | |
1914 | sets the stack up so that we can use the C<ST> macro. | |
1915 | ||
1916 | =item 3. | |
1917 | ||
1918 | Unlike the original coding of this example, the returned | |
1919 | values are not accessed in reverse order. So C<ST(0)> refers to the | |
54310121 | 1920 | first value returned by the Perl subroutine and C<ST(count-1)> |
d1b91892 AD |
1921 | refers to the last. |
1922 | ||
1923 | =back | |
a0d0e21e | 1924 | |
8f183262 DM |
1925 | =head2 Creating and calling an anonymous subroutine in C |
1926 | ||
4929bf7b | 1927 | As we've already shown, C<call_sv> can be used to invoke an |
c2611fb3 GS |
1928 | anonymous subroutine. However, our example showed a Perl script |
1929 | invoking an XSUB to perform this operation. Let's see how it can be | |
8f183262 DM |
1930 | done inside our C code: |
1931 | ||
8f183262 DM |
1932 | ... |
1933 | ||
4929bf7b | 1934 | SV *cvrv = eval_pv("sub { print 'You will not find me cluttering any namespace!' }", TRUE); |
8f183262 DM |
1935 | |
1936 | ... | |
1937 | ||
4929bf7b | 1938 | call_sv(cvrv, G_VOID|G_NOARGS); |
8f183262 | 1939 | |
4929bf7b GS |
1940 | C<eval_pv> is used to compile the anonymous subroutine, which |
1941 | will be the return value as well (read more about C<eval_pv> in | |
4a4eefd0 | 1942 | L<perlapi/eval_pv>). Once this code reference is in hand, it |
8f183262 DM |
1943 | can be mixed in with all the previous examples we've shown. |
1944 | ||
a0d0e21e LW |
1945 | =head1 SEE ALSO |
1946 | ||
8e07c86e | 1947 | L<perlxs>, L<perlguts>, L<perlembed> |
a0d0e21e LW |
1948 | |
1949 | =head1 AUTHOR | |
1950 | ||
0536e0eb | 1951 | Paul Marquess |
a0d0e21e | 1952 | |
d1b91892 AD |
1953 | Special thanks to the following people who assisted in the creation of |
1954 | the document. | |
a0d0e21e | 1955 | |
c07a80fd | 1956 | Jeff Okamoto, Tim Bunce, Nick Gianniotis, Steve Kelem, Gurusamy Sarathy |
1957 | and Larry Wall. | |
a0d0e21e LW |
1958 | |
1959 | =head1 DATE | |
1960 | ||
137443ea | 1961 | Version 1.3, 14th Apr 1997 |