Commit | Line | Data |
3967c732 |
1 | # Devel::Peek - A data debugging tool for the XS programmer |
2 | # The documentation is after the __END__ |
3 | |
4 | package Devel::Peek; |
5 | |
117c97cc |
6 | # Underscore to allow older Perls to access older version from CPAN |
bd16a5f0 |
7 | $VERSION = '1.00_03'; |
105cd853 |
8 | $XS_VERSION = $VERSION; |
9 | $VERSION = eval $VERSION; |
3967c732 |
10 | |
11 | require Exporter; |
9426adcd |
12 | use XSLoader (); |
3967c732 |
13 | |
9426adcd |
14 | @ISA = qw(Exporter); |
d1424c31 |
15 | @EXPORT = qw(Dump mstat DeadCode DumpArray DumpWithOP DumpProg |
bd16a5f0 |
16 | fill_mstats mstats_fillhash mstats2hash runops_debug debug_flags); |
83ee9e09 |
17 | @EXPORT_OK = qw(SvREFCNT SvREFCNT_inc SvREFCNT_dec CvGV); |
3967c732 |
18 | %EXPORT_TAGS = ('ALL' => [@EXPORT, @EXPORT_OK]); |
19 | |
9426adcd |
20 | XSLoader::load 'Devel::Peek'; |
3967c732 |
21 | |
1045810a |
22 | sub import { |
23 | my $c = shift; |
24 | my $ops_rx = qr/^:opd(=[stP]*)?\b/; |
25 | my @db = grep m/$ops_rx/, @_; |
26 | @_ = grep !m/$ops_rx/, @_; |
27 | if (@db) { |
28 | die "Too many :opd options" if @db > 1; |
29 | runops_debug(1); |
30 | my $flags = ($db[0] =~ m/$ops_rx/ and $1); |
31 | $flags = 'st' unless defined $flags; |
32 | my $f = 0; |
33 | $f |= 2 if $flags =~ /s/; |
34 | $f |= 8 if $flags =~ /t/; |
35 | $f |= 64 if $flags =~ /P/; |
36 | $^D |= $f if $f; |
37 | } |
38 | unshift @_, $c; |
39 | goto &Exporter::import; |
40 | } |
41 | |
3967c732 |
42 | sub DumpWithOP ($;$) { |
43 | local($Devel::Peek::dump_ops)=1; |
44 | my $depth = @_ > 1 ? $_[1] : 4 ; |
45 | Dump($_[0],$depth); |
46 | } |
47 | |
bd16a5f0 |
48 | $D_flags = 'psltocPmfrxuLHXDSTR'; |
49 | |
50 | sub debug_flags (;$) { |
51 | my $out = ""; |
52 | for my $i (0 .. length($D_flags)-1) { |
53 | $out .= substr $D_flags, $i, 1 if $^D & (1<<$i); |
54 | } |
55 | my $arg = shift; |
56 | my $num = $arg; |
57 | if (defined $arg and $arg =~ /\D/) { |
58 | die "unknown flags in debug_flags()" if $arg =~ /[^-$D_flags]/; |
59 | my ($on,$off) = split /-/, "$arg-"; |
60 | $num = $^D; |
61 | $num |= (1<<index($D_flags, $_)) for split //, $on; |
62 | $num &= ~(1<<index($D_flags, $_)) for split //, $off; |
63 | } |
64 | $^D = $num if defined $arg; |
65 | $out |
66 | } |
67 | |
3967c732 |
68 | 1; |
69 | __END__ |
70 | |
71 | =head1 NAME |
72 | |
73 | Devel::Peek - A data debugging tool for the XS programmer |
74 | |
75 | =head1 SYNOPSIS |
76 | |
77 | use Devel::Peek; |
78 | Dump( $a ); |
79 | Dump( $a, 5 ); |
80 | DumpArray( 5, $a, $b, ... ); |
81 | mstat "Point 5"; |
82 | |
1045810a |
83 | use Devel::Peek ':opd=st'; |
84 | |
3967c732 |
85 | =head1 DESCRIPTION |
86 | |
87 | Devel::Peek contains functions which allows raw Perl datatypes to be |
88 | manipulated from a Perl script. This is used by those who do XS programming |
89 | to check that the data they are sending from C to Perl looks as they think |
90 | it should look. The trick, then, is to know what the raw datatype is |
91 | supposed to look like when it gets to Perl. This document offers some tips |
92 | and hints to describe good and bad raw data. |
93 | |
94 | It is very possible that this document will fall far short of being useful |
95 | to the casual reader. The reader is expected to understand the material in |
96 | the first few sections of L<perlguts>. |
97 | |
98 | Devel::Peek supplies a C<Dump()> function which can dump a raw Perl |
99 | datatype, and C<mstat("marker")> function to report on memory usage |
100 | (if perl is compiled with corresponding option). The function |
101 | DeadCode() provides statistics on the data "frozen" into inactive |
102 | C<CV>. Devel::Peek also supplies C<SvREFCNT()>, C<SvREFCNT_inc()>, and |
103 | C<SvREFCNT_dec()> which can query, increment, and decrement reference |
104 | counts on SVs. This document will take a passive, and safe, approach |
105 | to data debugging and for that it will describe only the C<Dump()> |
d1424c31 |
106 | function. |
3967c732 |
107 | |
108 | Function C<DumpArray()> allows dumping of multiple values (useful when you |
076c2fc0 |
109 | need to analyze returns of functions). |
3967c732 |
110 | |
111 | The global variable $Devel::Peek::pv_limit can be set to limit the |
112 | number of character printed in various string values. Setting it to 0 |
113 | means no limit. |
114 | |
1045810a |
115 | If C<use Devel::Peek> directive has a C<:opd=FLAGS> argument, |
116 | this switches on debugging of opcode dispatch. C<FLAGS> should be a |
117 | combination of C<s>, C<t>, and C<P> (see B<-D> flags in L<perlrun>). |
118 | C<:opd> is a shortcut for C<:opd=st>. |
119 | |
bd16a5f0 |
120 | =head2 Runtime debugging |
121 | |
122 | C<CvGV($cv)> return one of the globs associated to a subroutine reference $cv. |
123 | |
124 | debug_flags() returns a string representation of C<$^D> (similar to |
125 | what is allowed for B<-D> flag). When called with a numeric argument, |
126 | sets $^D to the corresponding value. When called with an argument of |
127 | the form C<"flags-flags">, set on/off bits of C<$^D> corresponding to |
128 | letters before/after C<->. (The returned value is for C<$^D> before |
129 | the modification.) |
130 | |
131 | runops_debug() returns true if the current I<opcode dispatcher> is the |
132 | debugging one. When called with an argument, switches to debugging or |
133 | non-debugging dispatcher depending on the argument (active for |
134 | newly-entered subs/etc only). (The returned value is for the dispatcher before the modification.) |
135 | |
d1424c31 |
136 | =head2 Memory footprint debugging |
137 | |
138 | When perl is compiled with support for memory footprint debugging |
139 | (default with Perl's malloc()), Devel::Peek provides an access to this API. |
140 | |
141 | Use mstat() function to emit a memory state statistic to the terminal. |
142 | For more information on the format of output of mstat() see |
f3487f28 |
143 | L<perldebguts/Using C<$ENV{PERL_DEBUG_MSTATS}>>. |
d1424c31 |
144 | |
145 | Three additional functions allow access to this statistic from Perl. |
146 | First, use C<mstats_fillhash(%hash)> to get the information contained |
147 | in the output of mstat() into %hash. The field of this hash are |
148 | |
149 | minbucket nbuckets sbrk_good sbrk_slack sbrked_remains sbrks start_slack |
150 | topbucket topbucket_ev topbucket_odd total total_chain total_sbrk totfree |
151 | |
152 | Two additional fields C<free>, C<used> contain array references which |
153 | provide per-bucket count of free and used chunks. Two other fields |
154 | C<mem_size>, C<available_size> contain array references which provide |
155 | the information about the allocated size and usable size of chunks in |
f3487f28 |
156 | each bucket. Again, see L<perldebguts/Using C<$ENV{PERL_DEBUG_MSTATS}>> |
d1424c31 |
157 | for details. |
158 | |
159 | Keep in mind that only the first several "odd-numbered" buckets are |
160 | used, so the information on size of the "odd-numbered" buckets which are |
161 | not used is probably meaningless. |
162 | |
163 | The information in |
164 | |
165 | mem_size available_size minbucket nbuckets |
166 | |
167 | is the property of a particular build of perl, and does not depend on |
168 | the current process. If you do not provide the optional argument to |
169 | the functions mstats_fillhash(), fill_mstats(), mstats2hash(), then |
170 | the information in fields C<mem_size>, C<available_size> is not |
171 | updated. |
172 | |
173 | C<fill_mstats($buf)> is a much cheaper call (both speedwise and |
174 | memory-wise) which collects the statistic into $buf in |
175 | machine-readable form. At a later moment you may need to call |
176 | C<mstats2hash($buf, %hash)> to use this information to fill %hash. |
177 | |
178 | All three APIs C<fill_mstats($buf)>, C<mstats_fillhash(%hash)>, and |
179 | C<mstats2hash($buf, %hash)> are designed to allocate no memory if used |
180 | I<the second time> on the same $buf and/or %hash. |
181 | |
182 | So, if you want to collect memory info in a cycle, you may call |
183 | |
184 | $#buf = 999; |
185 | fill_mstats($_) for @buf; |
186 | mstats_fillhash(%report, 1); # Static info too |
187 | |
188 | foreach (@buf) { |
189 | # Do something... |
190 | fill_mstats $_; # Collect statistic |
191 | } |
192 | foreach (@buf) { |
193 | mstats2hash($_, %report); # Preserve static info |
194 | # Do something with %report |
195 | } |
196 | |
3967c732 |
197 | =head1 EXAMPLES |
198 | |
199 | The following examples don't attempt to show everything as that would be a |
200 | monumental task, and, frankly, we don't want this manpage to be an internals |
201 | document for Perl. The examples do demonstrate some basics of the raw Perl |
202 | datatypes, and should suffice to get most determined people on their way. |
203 | There are no guidewires or safety nets, nor blazed trails, so be prepared to |
204 | travel alone from this point and on and, if at all possible, don't fall into |
205 | the quicksand (it's bad for business). |
206 | |
207 | Oh, one final bit of advice: take L<perlguts> with you. When you return we |
208 | expect to see it well-thumbed. |
209 | |
210 | =head2 A simple scalar string |
211 | |
212 | Let's begin by looking a simple scalar which is holding a string. |
213 | |
a423dfdd |
214 | use Devel::Peek; |
3967c732 |
215 | $a = "hello"; |
216 | Dump $a; |
217 | |
218 | The output: |
219 | |
220 | SV = PVIV(0xbc288) |
221 | REFCNT = 1 |
222 | FLAGS = (POK,pPOK) |
223 | IV = 0 |
224 | PV = 0xb2048 "hello"\0 |
225 | CUR = 5 |
226 | LEN = 6 |
227 | |
228 | This says C<$a> is an SV, a scalar. The scalar is a PVIV, a string. |
229 | Its reference count is 1. It has the C<POK> flag set, meaning its |
230 | current PV field is valid. Because POK is set we look at the PV item |
231 | to see what is in the scalar. The \0 at the end indicate that this |
232 | PV is properly NUL-terminated. |
233 | If the FLAGS had been IOK we would look |
234 | at the IV item. CUR indicates the number of characters in the PV. |
235 | LEN indicates the number of bytes requested for the PV (one more than |
236 | CUR, in this case, because LEN includes an extra byte for the |
237 | end-of-string marker). |
238 | |
239 | =head2 A simple scalar number |
240 | |
241 | If the scalar contains a number the raw SV will be leaner. |
242 | |
a423dfdd |
243 | use Devel::Peek; |
3967c732 |
244 | $a = 42; |
245 | Dump $a; |
246 | |
247 | The output: |
248 | |
249 | SV = IV(0xbc818) |
250 | REFCNT = 1 |
251 | FLAGS = (IOK,pIOK) |
252 | IV = 42 |
253 | |
254 | This says C<$a> is an SV, a scalar. The scalar is an IV, a number. Its |
255 | reference count is 1. It has the C<IOK> flag set, meaning it is currently |
256 | being evaluated as a number. Because IOK is set we look at the IV item to |
257 | see what is in the scalar. |
258 | |
259 | =head2 A simple scalar with an extra reference |
260 | |
261 | If the scalar from the previous example had an extra reference: |
262 | |
a423dfdd |
263 | use Devel::Peek; |
3967c732 |
264 | $a = 42; |
265 | $b = \$a; |
266 | Dump $a; |
267 | |
268 | The output: |
269 | |
270 | SV = IV(0xbe860) |
271 | REFCNT = 2 |
272 | FLAGS = (IOK,pIOK) |
273 | IV = 42 |
274 | |
275 | Notice that this example differs from the previous example only in its |
276 | reference count. Compare this to the next example, where we dump C<$b> |
277 | instead of C<$a>. |
278 | |
279 | =head2 A reference to a simple scalar |
280 | |
281 | This shows what a reference looks like when it references a simple scalar. |
282 | |
a423dfdd |
283 | use Devel::Peek; |
3967c732 |
284 | $a = 42; |
285 | $b = \$a; |
286 | Dump $b; |
287 | |
288 | The output: |
289 | |
290 | SV = RV(0xf041c) |
291 | REFCNT = 1 |
292 | FLAGS = (ROK) |
293 | RV = 0xbab08 |
294 | SV = IV(0xbe860) |
295 | REFCNT = 2 |
296 | FLAGS = (IOK,pIOK) |
297 | IV = 42 |
298 | |
299 | Starting from the top, this says C<$b> is an SV. The scalar is an RV, a |
300 | reference. It has the C<ROK> flag set, meaning it is a reference. Because |
301 | ROK is set we have an RV item rather than an IV or PV. Notice that Dump |
302 | follows the reference and shows us what C<$b> was referencing. We see the |
303 | same C<$a> that we found in the previous example. |
304 | |
305 | Note that the value of C<RV> coincides with the numbers we see when we |
306 | stringify $b. The addresses inside RV() and IV() are addresses of |
307 | C<X***> structure which holds the current state of an C<SV>. This |
308 | address may change during lifetime of an SV. |
309 | |
310 | =head2 A reference to an array |
311 | |
312 | This shows what a reference to an array looks like. |
313 | |
a423dfdd |
314 | use Devel::Peek; |
3967c732 |
315 | $a = [42]; |
316 | Dump $a; |
317 | |
318 | The output: |
319 | |
320 | SV = RV(0xf041c) |
321 | REFCNT = 1 |
322 | FLAGS = (ROK) |
323 | RV = 0xb2850 |
324 | SV = PVAV(0xbd448) |
325 | REFCNT = 1 |
326 | FLAGS = () |
327 | IV = 0 |
328 | NV = 0 |
329 | ARRAY = 0xb2048 |
330 | ALLOC = 0xb2048 |
331 | FILL = 0 |
332 | MAX = 0 |
333 | ARYLEN = 0x0 |
334 | FLAGS = (REAL) |
335 | Elt No. 0 0xb5658 |
336 | SV = IV(0xbe860) |
337 | REFCNT = 1 |
338 | FLAGS = (IOK,pIOK) |
339 | IV = 42 |
340 | |
341 | This says C<$a> is an SV and that it is an RV. That RV points to |
342 | another SV which is a PVAV, an array. The array has one element, |
343 | element zero, which is another SV. The field C<FILL> above indicates |
344 | the last element in the array, similar to C<$#$a>. |
345 | |
346 | If C<$a> pointed to an array of two elements then we would see the |
347 | following. |
348 | |
349 | use Devel::Peek 'Dump'; |
350 | $a = [42,24]; |
351 | Dump $a; |
352 | |
353 | The output: |
354 | |
355 | SV = RV(0xf041c) |
356 | REFCNT = 1 |
357 | FLAGS = (ROK) |
358 | RV = 0xb2850 |
359 | SV = PVAV(0xbd448) |
360 | REFCNT = 1 |
361 | FLAGS = () |
362 | IV = 0 |
363 | NV = 0 |
364 | ARRAY = 0xb2048 |
365 | ALLOC = 0xb2048 |
366 | FILL = 0 |
367 | MAX = 0 |
368 | ARYLEN = 0x0 |
369 | FLAGS = (REAL) |
370 | Elt No. 0 0xb5658 |
371 | SV = IV(0xbe860) |
372 | REFCNT = 1 |
373 | FLAGS = (IOK,pIOK) |
374 | IV = 42 |
375 | Elt No. 1 0xb5680 |
376 | SV = IV(0xbe818) |
377 | REFCNT = 1 |
378 | FLAGS = (IOK,pIOK) |
379 | IV = 24 |
380 | |
381 | Note that C<Dump> will not report I<all> the elements in the array, |
382 | only several first (depending on how deep it already went into the |
383 | report tree). |
384 | |
385 | =head2 A reference to a hash |
386 | |
387 | The following shows the raw form of a reference to a hash. |
388 | |
a423dfdd |
389 | use Devel::Peek; |
3967c732 |
390 | $a = {hello=>42}; |
391 | Dump $a; |
392 | |
393 | The output: |
394 | |
395 | SV = RV(0xf041c) |
396 | REFCNT = 1 |
397 | FLAGS = (ROK) |
398 | RV = 0xb2850 |
399 | SV = PVHV(0xbd448) |
400 | REFCNT = 1 |
401 | FLAGS = () |
402 | NV = 0 |
403 | ARRAY = 0xbd748 |
404 | KEYS = 1 |
405 | FILL = 1 |
406 | MAX = 7 |
407 | RITER = -1 |
408 | EITER = 0x0 |
409 | Elt "hello" => 0xbaaf0 |
410 | SV = IV(0xbe860) |
411 | REFCNT = 1 |
412 | FLAGS = (IOK,pIOK) |
413 | IV = 42 |
414 | |
415 | This shows C<$a> is a reference pointing to an SV. That SV is a PVHV, a |
416 | hash. Fields RITER and EITER are used by C<L<each>>. |
417 | |
418 | =head2 Dumping a large array or hash |
419 | |
420 | The C<Dump()> function, by default, dumps up to 4 elements from a |
421 | toplevel array or hash. This number can be increased by supplying a |
422 | second argument to the function. |
423 | |
a423dfdd |
424 | use Devel::Peek; |
3967c732 |
425 | $a = [10,11,12,13,14]; |
426 | Dump $a; |
427 | |
428 | Notice that C<Dump()> prints only elements 10 through 13 in the above code. |
429 | The following code will print all of the elements. |
430 | |
431 | use Devel::Peek 'Dump'; |
432 | $a = [10,11,12,13,14]; |
433 | Dump $a, 5; |
434 | |
435 | =head2 A reference to an SV which holds a C pointer |
436 | |
437 | This is what you really need to know as an XS programmer, of course. When |
438 | an XSUB returns a pointer to a C structure that pointer is stored in an SV |
439 | and a reference to that SV is placed on the XSUB stack. So the output from |
440 | an XSUB which uses something like the T_PTROBJ map might look something like |
441 | this: |
442 | |
443 | SV = RV(0xf381c) |
444 | REFCNT = 1 |
445 | FLAGS = (ROK) |
446 | RV = 0xb8ad8 |
447 | SV = PVMG(0xbb3c8) |
448 | REFCNT = 1 |
449 | FLAGS = (OBJECT,IOK,pIOK) |
450 | IV = 729160 |
451 | NV = 0 |
452 | PV = 0 |
453 | STASH = 0xc1d10 "CookBookB::Opaque" |
454 | |
455 | This shows that we have an SV which is an RV. That RV points at another |
456 | SV. In this case that second SV is a PVMG, a blessed scalar. Because it is |
457 | blessed it has the C<OBJECT> flag set. Note that an SV which holds a C |
458 | pointer also has the C<IOK> flag set. The C<STASH> is set to the package |
459 | name which this SV was blessed into. |
460 | |
461 | The output from an XSUB which uses something like the T_PTRREF map, which |
462 | doesn't bless the object, might look something like this: |
463 | |
464 | SV = RV(0xf381c) |
465 | REFCNT = 1 |
466 | FLAGS = (ROK) |
467 | RV = 0xb8ad8 |
468 | SV = PVMG(0xbb3c8) |
469 | REFCNT = 1 |
470 | FLAGS = (IOK,pIOK) |
471 | IV = 729160 |
472 | NV = 0 |
473 | PV = 0 |
474 | |
475 | =head2 A reference to a subroutine |
476 | |
477 | Looks like this: |
478 | |
479 | SV = RV(0x798ec) |
480 | REFCNT = 1 |
481 | FLAGS = (TEMP,ROK) |
482 | RV = 0x1d453c |
483 | SV = PVCV(0x1c768c) |
484 | REFCNT = 2 |
485 | FLAGS = () |
486 | IV = 0 |
487 | NV = 0 |
488 | COMP_STASH = 0x31068 "main" |
489 | START = 0xb20e0 |
490 | ROOT = 0xbece0 |
491 | XSUB = 0x0 |
492 | XSUBANY = 0 |
493 | GVGV::GV = 0x1d44e8 "MY" :: "top_targets" |
57843af0 |
494 | FILE = "(eval 5)" |
3967c732 |
495 | DEPTH = 0 |
496 | PADLIST = 0x1c9338 |
497 | |
498 | This shows that |
499 | |
bbc7dcd2 |
500 | =over 4 |
3967c732 |
501 | |
a45bd81d |
502 | =item * |
3967c732 |
503 | |
504 | the subroutine is not an XSUB (since C<START> and C<ROOT> are |
505 | non-zero, and C<XSUB> is zero); |
506 | |
a45bd81d |
507 | =item * |
3967c732 |
508 | |
509 | that it was compiled in the package C<main>; |
510 | |
a45bd81d |
511 | =item * |
3967c732 |
512 | |
513 | under the name C<MY::top_targets>; |
514 | |
a45bd81d |
515 | =item * |
3967c732 |
516 | |
517 | inside a 5th eval in the program; |
518 | |
a45bd81d |
519 | =item * |
3967c732 |
520 | |
521 | it is not currently executed (see C<DEPTH>); |
522 | |
a45bd81d |
523 | =item * |
3967c732 |
524 | |
525 | it has no prototype (C<PROTOTYPE> field is missing). |
526 | |
a45bd81d |
527 | =back |
3967c732 |
528 | |
529 | =head1 EXPORTS |
530 | |
531 | C<Dump>, C<mstat>, C<DeadCode>, C<DumpArray>, C<DumpWithOP> and |
7c6ca602 |
532 | C<DumpProg>, C<fill_mstats>, C<mstats_fillhash>, C<mstats2hash> by |
d1424c31 |
533 | default. Additionally available C<SvREFCNT>, C<SvREFCNT_inc> and |
534 | C<SvREFCNT_dec>. |
3967c732 |
535 | |
536 | =head1 BUGS |
537 | |
538 | Readers have been known to skip important parts of L<perlguts>, causing much |
539 | frustration for all. |
540 | |
541 | =head1 AUTHOR |
542 | |
543 | Ilya Zakharevich ilya@math.ohio-state.edu |
544 | |
545 | Copyright (c) 1995-98 Ilya Zakharevich. All rights reserved. |
546 | This program is free software; you can redistribute it and/or |
547 | modify it under the same terms as Perl itself. |
548 | |
549 | Author of this software makes no claim whatsoever about suitability, |
550 | reliability, edability, editability or usability of this product, and |
551 | should not be kept liable for any damage resulting from the use of |
552 | it. If you can use it, you are in luck, if not, I should not be kept |
553 | responsible. Keep a handy copy of your backup tape at hand. |
554 | |
555 | =head1 SEE ALSO |
556 | |
557 | L<perlguts>, and L<perlguts>, again. |
558 | |
559 | =cut |