1dbg(3) Erlang Module Definition dbg(3)
2
6 dbg - The Text Based Trace Facility
7
9 This module implements a text based interface to the trace/3 and the
10 trace_pattern/2 BIFs. It makes it possible to trace functions, pro‐
11 cesses, ports and messages.
12 To quickly get started on tracing function calls you can use the fol‐
14 lowing code in the Erlang shell:
15 1> dbg:tracer(). %% Start the default trace message receiver
17 {ok,<0.36.0>}
18 2> dbg:p(all, c). %% Setup call (c) tracing on all processes
19 {ok,[{matched,nonode@nohost,26}]}
20 3> dbg:tp(lists, seq, x). %% Setup an exception return trace (x) on lists:seq
21 {ok,[{matched,nonode@nohost,2},{saved,x}]}
22 4> lists:seq(1,10).
23 (<0.34.0>) call lists:seq(1,10)
24 (<0.34.0>) returned from lists:seq/2 -> [1,2,3,4,5,6,7,8,9,10]
25 [1,2,3,4,5,6,7,8,9,10]
26 For more examples of how to use dbg from the Erlang shell, see the sim‐
29 ple example section.
30 The utilities are also suitable to use in system testing on large sys‐
32 tems, where other tools have too much impact on the system performance.
33 Some primitive support for sequential tracing is also included, see the
34 advanced topics section.
35
37 fun2ms(LiteralFun) -> MatchSpec
38 Types:
40 LiteralFun = fun() literal
42 MatchSpec = term()
43 Pseudo function that by means of a parse_transform translates
45 the literal fun() typed as parameter in the function call to a
46 match specification as described in the match_spec manual of
47 ERTS users guide. (With literal I mean that the fun() needs to
48 textually be written as the parameter of the function, it cannot
49 be held in a variable which in turn is passed to the function).
50 The parse transform is implemented in the module ms_transform
52 and the source must include the file ms_transform.hrl in STDLIB
53 for this pseudo function to work. Failing to include the hrl
54 file in the source will result in a runtime error, not a compile
55 time ditto. The include file is easiest included by adding the
56 line -include_lib("stdlib/include/ms_transform.hrl"). to the
57 source file.
58 The fun() is very restricted, it can take only a single parame‐
60 ter (the parameter list to match), a sole variable or a list. It
61 needs to use the is_XXX guard tests and one cannot use language
62 constructs that have no representation in a match_spec (like if,
63 case, receive etc). The return value from the fun will be the
64 return value of the resulting match_spec.
65 Example:
67 1> dbg:fun2ms(fun([M,N]) when N > 3 -> return_trace() end).
69 [{['$1','$2'],[{'>','$2',3}],[{return_trace}]}]
70 Variables from the environment can be imported, so that this
72 works:
73 2> X=3.
75 3
76 3> dbg:fun2ms(fun([M,N]) when N > X -> return_trace() end).
77 [{['$1','$2'],[{'>','$2',{const,3}}],[{return_trace}]}]
78 The imported variables will be replaced by match_spec const ex‐
80 pressions, which is consistent with the static scoping for Er‐
81 lang fun()s. Local or global function calls cannot be in the
82 guard or body of the fun however. Calls to builtin match_spec
83 functions of course is allowed:
84 4> dbg:fun2ms(fun([M,N]) when N > X, is_atomm(M) -> return_trace() end).
86 Error: fun containing local erlang function calls ('is_atomm' called in guard)\
87 cannot be translated into match_spec
88 {error,transform_error}
89 5> dbg:fun2ms(fun([M,N]) when N > X, is_atom(M) -> return_trace() end).
90 [{['$1','$2'],[{'>','$2',{const,3}},{is_atom,'$1'}],[{return_trace}]}]
91 As you can see by the example, the function can be called from
93 the shell too. The fun() needs to be literally in the call when
94 used from the shell as well. Other means than the parse_trans‐
95 form are used in the shell case, but more or less the same re‐
96 strictions apply (the exception being records, as they are not
97 handled by the shell).
98 Warning:
100 If the parse_transform is not applied to a module which calls
101 this pseudo function, the call will fail in runtime (with a
102 badarg). The module dbg actually exports a function with this
103 name, but it should never really be called except for when using
104 the function in the shell. If the parse_transform is properly
105 applied by including the ms_transform.hrl header file, compiled
106 code will never call the function, but the function call is re‐
107 placed by a literal match_spec.
108 More information is provided by the ms_transform manual page in
111 STDLIB.
112 h() -> ok
114 Gives a list of items for brief online help.
116 h(Item) -> ok
118 Types:
120 Item = atom()
122 Gives a brief help text for functions in the dbg module. The
124 available items can be listed with dbg:h/0
125 p(Item) -> {ok, MatchDesc} | {error, term()}
127 Equivalent to p(Item, [m]).
129 p(Item, Flags) -> {ok, MatchDesc} | {error, term()}
131 Types:
133 MatchDesc = [MatchNum]
135 MatchNum = {matched, node(), integer()} | {matched, node(),
136 0, RPCError}
137 RPCError = term()
138 Traces Item in accordance to the value specified by Flags. The
140 variation of Item is listed below:
141 pid() or port():
143 The corresponding process or port is traced. The process or
144 port may be a remote process or port (on another Erlang
145 node). The node must be in the list of traced nodes (see n/1
146 and tracer/3).
147 all:
149 All processes and ports in the system as well as all pro‐
150 cesses and ports created hereafter are to be traced.
151 processes:
153 All processes in the system as well as all processes created
154 hereafter are to be traced.
155 ports:
157 All ports in the system as well as all ports created here‐
158 after are to be traced.
159 new:
161 All processes and ports created after the call is are to be
162 traced.
163 new_processes:
165 All processes created after the call is are to be traced.
166 new_ports:
168 All ports created after the call is are to be traced.
169 existing:
171 All existing processes and ports are traced.
172 existing_processes:
174 All existing processes are traced.
175 existing_ports:
177 All existing ports are traced.
178 atom():
180 The process or port with the corresponding registered name
181 is traced. The process or port may be a remote process (on
182 another Erlang node). The node must be added with the n/1 or
183 tracer/3 function.
184 integer():
186 The process <0.Item.0> is traced.
187 {X, Y, Z}:
189 The process <X.Y.Z> is traced.
190 string():
192 If the Item is a string "<X.Y.Z>" as returned from
193 pid_to_list/1, the process <X.Y.Z> is traced.
194 When enabling an Item that represents a group of processes, the
196 Item is enabled on all nodes added with the n/1 or tracer/3
197 function.
198 Flags can be a single atom, or a list of flags. The available
200 flags are:
201 s (send):
203 Traces the messages the process or port sends.
204 r (receive):
206 Traces the messages the process or port receives.
207 m (messages):
209 Traces the messages the process or port receives and sends.
210 c (call):
212 Traces global function calls for the process according to
213 the trace patterns set in the system (see tp/2).
214 p (procs):
216 Traces process related events to the process.
217 ports:
219 Traces port related events to the port.
220 sos (set on spawn):
222 Lets all processes created by the traced process inherit the
223 trace flags of the traced process.
224 sol (set on link):
226 Lets another process, P2, inherit the trace flags of the
227 traced process whenever the traced process links to P2.
228 sofs (set on first spawn):
230 This is the same as sos, but only for the first process
231 spawned by the traced process.
232 sofl (set on first link):
234 This is the same as sol, but only for the first call to
235 link/1 by the traced process.
236 all:
238 Sets all flags except silent.
239 clear:
241 Clears all flags.
242 The list can also include any of the flags allowed in er‐
244 lang:trace/3
245 The function returns either an error tuple or a tuple {ok,
247 List}. The List consists of specifications of how many processes
248 and ports that matched (in the case of a pure pid() exactly 1).
249 The specification of matched processes is {matched, Node, N}. If
250 the remote processor call, rpc, to a remote node fails, the rpc
251 error message is delivered as a fourth argument and the number
252 of matched processes are 0. Note that the result {ok, List} may
253 contain a list where rpc calls to one, several or even all nodes
254 failed.
255 c(Mod, Fun, Args)
257 Equivalent to c(Mod, Fun, Args, all).
259 c(Mod, Fun, Args, Flags)
261 Evaluates the expression apply(Mod, Fun, Args) with the trace
263 flags in Flags set. This is a convenient way to trace processes
264 from the Erlang shell.
265 i() -> ok
267 Displays information about all traced processes and ports.
269 tp(Module,MatchSpec)
271 Same as tp({Module, '_', '_'}, MatchSpec)
273 tp(Module,Function,MatchSpec)
275 Same as tp({Module, Function, '_'}, MatchSpec)
277 tp(Module, Function, Arity, MatchSpec)
279 Same as tp({Module, Function, Arity}, MatchSpec)
281 tp({Module, Function, Arity}, MatchSpec) -> {ok, MatchDesc} | {error,
283 term()}
284 Types:
286 Module = atom() | '_'
288 Function = atom() | '_'
289 Arity = integer() |'_'
290 MatchSpec = integer() | Built-inAlias | [] | match_spec()
291 Built-inAlias = x | c | cx
292 MatchDesc = [MatchInfo]
293 MatchInfo = {saved, integer()} | MatchNum
294 MatchNum = {matched, node(), integer()} | {matched, node(),
295 0, RPCError}
296 This function enables call trace for one or more functions. All
298 exported functions matching the {Module, Function, Arity} argu‐
299 ment will be concerned, but the match_spec() may further narrow
300 down the set of function calls generating trace messages.
301 For a description of the match_spec() syntax, please turn to the
303 User's guide part of the online documentation for the runtime
304 system (erts). The chapter Match Specifications in Erlang ex‐
305 plains the general match specification "language". The most com‐
306 mon generic match specifications used can be found as Built-inA‐
307 lias', see ltp/0 below for details.
308 The Module, Function and/or Arity parts of the tuple may be
310 specified as the atom '_' which is a "wild-card" matching all
311 modules/functions/arities. Note, if the Module is specified as
312 '_', the Function and Arity parts have to be specified as '_'
313 too. The same holds for the Functions relation to the Arity.
314 All nodes added with n/1 or tracer/3 will be affected by this
316 call, and if Module is not '_' the module will be loaded on all
317 nodes.
318 The function returns either an error tuple or a tuple {ok,
320 List}. The List consists of specifications of how many functions
321 that matched, in the same way as the processes and ports are
322 presented in the return value of p/2.
323 There may be a tuple {saved, N} in the return value, if the
325 MatchSpec is other than []. The integer N may then be used in
326 subsequent calls to this function and will stand as an "alias"
327 for the given expression. There are also a couple of built-in
328 aliases for common expressions, see ltp/0 below for details.
329 If an error is returned, it can be due to errors in compilation
331 of the match specification. Such errors are presented as a list
332 of tuples {error, string()} where the string is a textual expla‐
333 nation of the compilation error. An example:
334 (x@y)4> dbg:tp({dbg,ltp,0},[{[],[],[{message, two, arguments}, {noexist}]}]).
336 {error,
337 [{error,"Special form 'message' called with wrong number of
338 arguments in {message,two,arguments}."},
339 {error,"Function noexist/1 does_not_exist."}]}
340 tpl(Module,MatchSpec)
342 Same as tpl({Module, '_', '_'}, MatchSpec)
344 tpl(Module,Function,MatchSpec)
346 Same as tpl({Module, Function, '_'}, MatchSpec)
348 tpl(Module, Function, Arity, MatchSpec)
350 Same as tpl({Module, Function, Arity}, MatchSpec)
352 tpl({Module, Function, Arity}, MatchSpec) -> {ok, MatchDesc} | {error,
354 term()}
355 This function works as tp/2, but enables tracing for local calls
357 (and local functions) as well as for global calls (and func‐
358 tions).
359 tpe(Event, MatchSpec) -> {ok, MatchDesc} | {error, term()}
361 Types:
363 Event = send | 'receive'
365 MatchSpec = integer() | Built-inAlias | [] | match_spec()
366 Built-inAlias = x | c | cx
367 MatchDesc = [MatchInfo]
368 MatchInfo = {saved, integer()} | MatchNum
369 MatchNum = {matched, node(), 1} | {matched, node(), 0, RPCEr‐
370 ror}
371 This function associates a match specification with trace event
373 send or 'receive'. By default all executed send and 'receive'
374 events are traced if enabled for a process. A match specifica‐
375 tion can be used to filter traced events based on sender, re‐
376 ceiver and/or message content.
377 For a description of the match_spec() syntax, please turn to the
379 User's guide part of the online documentation for the runtime
380 system (erts). The chapter Match Specifications in Erlang ex‐
381 plains the general match specification "language".
382 For send, the matching is done on the list [Receiver, Msg]. Re‐
384 ceiver is the process or port identity of the receiver and Msg
385 is the message term. The pid of the sending process can be ac‐
386 cessed with the guard function self/0.
387 For 'receive', the matching is done on the list [Node, Sender,
389 Msg]. Node is the node name of the sender. Sender is the process
390 or port identity of the sender, or the atom undefined if the
391 sender is not known (which may be the case for remote senders).
392 Msg is the message term. The pid of the receiving process can be
393 accessed with the guard function self/0.
394 All nodes added with n/1 or tracer/3 will be affected by this
396 call.
397 The return value is the same as for tp/2. The number of matched
399 events are never larger than 1 as tpe/2 does not accept any form
400 of wildcards for argument Event.
401 ctp()
403 Same as ctp({'_', '_', '_'})
405 ctp(Module)
407 Same as ctp({Module, '_', '_'})
409 ctp(Module, Function)
411 Same as ctp({Module, Function, '_'})
413 ctp(Module, Function, Arity)
415 Same as ctp({Module, Function, Arity})
417 ctp({Module, Function, Arity}) -> {ok, MatchDesc} | {error, term()}
419 Types:
421 Module = atom() | '_'
423 Function = atom() | '_'
424 Arity = integer() | '_'
425 MatchDesc = [MatchNum]
426 MatchNum = {matched, node(), integer()} | {matched, node(),
427 0, RPCError}
428 This function disables call tracing on the specified functions.
430 The semantics of the parameter is the same as for the corre‐
431 sponding function specification in tp/2 or tpl/2. Both local and
432 global call trace is disabled.
433 The return value reflects how many functions that matched, and
435 is constructed as described in tp/2. No tuple {saved, N} is how‐
436 ever ever returned (for obvious reasons).
437 ctpl()
439 Same as ctpl({'_', '_', '_'})
441 ctpl(Module)
443 Same as ctpl({Module, '_', '_'})
445 ctpl(Module, Function)
447 Same as ctpl({Module, Function, '_'})
449 ctpl(Module, Function, Arity)
451 Same as ctpl({Module, Function, Arity})
453 ctpl({Module, Function, Arity}) -> {ok, MatchDesc} | {error, term()}
455 This function works as ctp/1, but only disables tracing set up
457 with tpl/2 (not with tp/2).
458 ctpg()
460 Same as ctpg({'_', '_', '_'})
462 ctpg(Module)
464 Same as ctpg({Module, '_', '_'})
466 ctpg(Module, Function)
468 Same as ctpg({Module, Function, '_'})
470 ctpg(Module, Function, Arity)
472 Same as ctpg({Module, Function, Arity})
474 ctpg({Module, Function, Arity}) -> {ok, MatchDesc} | {error, term()}
476 This function works as ctp/1, but only disables tracing set up
478 with tp/2 (not with tpl/2).
479 ctpe(Event) -> {ok, MatchDesc} | {error, term()}
481 Types:
483 Event = send | 'receive'
485 MatchDesc = [MatchNum]
486 MatchNum = {matched, node(), 1} | {matched, node(), 0, RPCEr‐
487 ror}
488 This function clears match specifications for the specified
490 trace event (send or 'receive'). It will revert back to the de‐
491 fault behavior of tracing all triggered events.
492 The return value follow the same style as for ctp/1.
494 ltp() -> ok
496 Use this function to recall all match specifications previously
498 used in the session (i. e. previously saved during calls to
499 tp/2, and built-in match specifications. This is very useful, as
500 a complicated match_spec can be quite awkward to write. Note
501 that the match specifications are lost if stop/0 is called.
502 Match specifications used can be saved in a file (if a read-
504 write file system is present) for use in later debugging ses‐
505 sions, see wtp/1 and rtp/1
506 There are three built-in trace patterns: exception_trace,
508 caller_trace and caller_exception_trace (or x, c and cx respec‐
509 tively). Exception trace sets a trace which will show function
510 names, parameters, return values and exceptions thrown from
511 functions. Caller traces display function names, parameters and
512 information about which function called it. An example using a
513 built-in alias:
514 (x@y)4> dbg:tp(lists,sort,cx).
516 {ok,[{matched,nonode@nohost,2},{saved,cx}]}
517 (x@y)4> lists:sort([2,1]).
518 (<0.32.0>) call lists:sort([2,1]) ({erl_eval,do_apply,5})
519 (<0.32.0>) returned from lists:sort/1 -> [1,2]
520 [1,2]
521 dtp() -> ok
523 Use this function to "forget" all match specifications saved
525 during calls to tp/2. This is useful when one wants to restore
526 other match specifications from a file with rtp/1. Use dtp/1 to
527 delete specific saved match specifications.
528 dtp(N) -> ok
530 Types:
532 N = integer()
534 Use this function to "forget" a specific match specification
536 saved during calls to tp/2.
537 wtp(Name) -> ok | {error, IOError}
539 Types:
541 Name = string()
543 IOError = term()
544 This function will save all match specifications saved during
546 the session (during calls to tp/2) and built-in match specifica‐
547 tions in a text file with the name designated by Name. The for‐
548 mat of the file is textual, why it can be edited with an ordi‐
549 nary text editor, and then restored with rtp/1.
550 Each match spec in the file ends with a full stop (.) and new
552 (syntactically correct) match specifications can be added to the
553 file manually.
554 The function returns ok or an error tuple where the second ele‐
556 ment contains the I/O error that made the writing impossible.
557 rtp(Name) -> ok | {error, Error}
559 Types:
561 Name = string()
563 Error = term()
564 This function reads match specifications from a file (possibly)
566 generated by the wtp/1 function. It checks the syntax of all
567 match specifications and verifies that they are correct. The er‐
568 ror handling principle is "all or nothing", i. e. if some of the
569 match specifications are wrong, none of the specifications are
570 added to the list of saved match specifications for the running
571 system.
572 The match specifications in the file are merged with the current
574 match specifications, so that no duplicates are generated. Use
575 ltp/0 to see what numbers were assigned to the specifications
576 from the file.
577 The function will return an error, either due to I/O problems
579 (like a non existing or non readable file) or due to file format
580 problems. The errors from a bad format file are in a more or
581 less textual format, which will give a hint to what's causing
582 the problem.
583 n(Nodename) -> {ok, Nodename} | {error, Reason}
585 Types:
587 Nodename = atom()
589 Reason = term()
590 The dbg server keeps a list of nodes where tracing should be
592 performed. Whenever a tp/2 call or a p/2 call is made, it is ex‐
593 ecuted for all nodes in this list including the local node (ex‐
594 cept for p/2 with a specific pid() or port() as first argument,
595 in which case the command is executed only on the node where the
596 designated process or port resides).
597 This function adds a remote node (Nodename) to the list of nodes
599 where tracing is performed. It starts a tracer process on the
600 remote node, which will send all trace messages to the tracer
601 process on the local node (via the Erlang distribution). If no
602 tracer process is running on the local node, the error reason
603 no_local_tracer is returned. The tracer process on the local
604 node must be started with the tracer/0/2 function.
605 If Nodename is the local node, the error reason cant_add_lo‐
607 cal_node is returned.
608 If a trace port (see trace_port/2) is running on the local node,
610 remote nodes cannot be traced with a tracer process. The error
611 reason cant_trace_remote_pid_to_local_port is returned. A trace
612 port can however be started on the remote node with the tracer/3
613 function.
614 The function will also return an error if the node Nodename is
616 not reachable.
617 cn(Nodename) -> ok
619 Types:
621 Nodename = atom()
623 Clears a node from the list of traced nodes. Subsequent calls to
625 tp/2 and p/2 will not consider that node, but tracing already
626 activated on the node will continue to be in effect.
627 Returns ok, cannot fail.
629 ln() -> ok
631 Shows the list of traced nodes on the console.
633 tracer() -> {ok, pid()} | {error, already_started}
635 This function starts a server on the local node that will be the
637 recipient of all trace messages. All subsequent calls to p/2
638 will result in messages sent to the newly started trace server.
639 A trace server started in this way will simply display the trace
641 messages in a formatted way in the Erlang shell (i. e. use
642 io:format). See tracer/2 for a description of how the trace mes‐
643 sage handler can be customized.
644 To start a similar tracer on a remote node, use n/1.
646 tracer(Type, Data) -> {ok, pid()} | {error, Error}
648 Types:
650 Type = port | process | module
652 Data = PortGenerator | HandlerSpec | ModuleSpec
653 PortGenerator = fun() (no arguments)
654 Error = term()
655 HandlerSpec = {HandlerFun, InitialData}
656 HandlerFun = fun() (two arguments)
657 ModuleSpec = fun() (no arguments) | {TracerModule, Tracer‐
658 State}
659 TracerModule = atom()
660 InitialData = TracerState = term()
661 This function starts a tracer server with additional parameters
663 on the local node. The first parameter, the Type, indicates if
664 trace messages should be handled by a receiving process
665 (process), by a tracer port (port) or by a tracer module (mod‐
666 ule). For a description about tracer ports see trace_port/2 and
667 for a tracer modules see erl_tracer.
668 If Type is process, a message handler function can be specified
670 (HandlerSpec). The handler function, which should be a fun tak‐
671 ing two arguments, will be called for each trace message, with
672 the first argument containing the message as it is and the sec‐
673 ond argument containing the return value from the last invoca‐
674 tion of the fun. The initial value of the second parameter is
675 specified in the InitialData part of the HandlerSpec. The Han‐
676 dlerFun may choose any appropriate action to take when invoked,
677 and can save a state for the next invocation by returning it.
678 If Type is port, then the second parameter should be a fun which
680 takes no arguments and returns a newly opened trace port when
681 called. Such a fun is preferably generated by calling
682 trace_port/2.
683 if Type is module, then the second parameter should be either a
685 tuple describing the erl_tracer module to be used for tracing
686 and the state to be used for that tracer module or a fun return‐
687 ing the same tuple.
688 If an error is returned, it can either be due to a tracer server
690 already running ({error,already_started}) or due to the Handler‐
691 Fun throwing an exception.
692 To start a similar tracer on a remote node, use tracer/3.
694 tracer(Nodename, Type, Data) -> {ok, Nodename} | {error, Reason}
696 Types:
698 Nodename = atom()
700 This function is equivalent to tracer/2, but acts on the given
702 node. A tracer is started on the node (Nodename) and the node is
703 added to the list of traced nodes.
704 Note:
706 This function is not equivalent to n/1. While n/1 starts a
707 process tracer which redirects all trace information to a
708 process tracer on the local node (i.e. the trace control node),
709 tracer/3 starts a tracer of any type which is independent of the
710 tracer on the trace control node.
711 For details, see tracer/2.
714 trace_port(Type, Parameters) -> fun()
716 Types:
718 Type = ip | file
720 Parameters = Filename | WrapFilesSpec | IPPortSpec
721 Filename = string() | [string()] | atom()
722 WrapFilesSpec = {Filename, wrap, Suffix} | {Filename, wrap,
723 Suffix, WrapSize} | {Filename, wrap, Suffix, WrapSize,
724 WrapCnt}
725 Suffix = string()
726 WrapSize = integer() >= 0 | {time, WrapTime}
727 WrapTime = integer() >= 1
728 WrapCnt = integer() >= 1
729 IpPortSpec = PortNumber | {PortNumber, QueSize}
730 PortNumber = integer()
731 QueSize = integer()
732 This function creates a trace port generating fun. The fun takes
734 no arguments and returns a newly opened trace port. The return
735 value from this function is suitable as a second parameter to
736 tracer/2, i.e. dbg:tracer(port, dbg:trace_port(ip, 4711)).
737 A trace port is an Erlang port to a dynamically linked in driver
739 that handles trace messages directly, without the overhead of
740 sending them as messages in the Erlang virtual machine.
741 Two trace drivers are currently implemented, the file and the ip
743 trace drivers. The file driver sends all trace messages into one
744 or several binary files, from where they later can be fetched
745 and processed with the trace_client/2 function. The ip driver
746 opens a TCP/IP port where it listens for connections. When a
747 client (preferably started by calling trace_client/2 on another
748 Erlang node) connects, all trace messages are sent over the IP
749 network for further processing by the remote client.
750 Using a trace port significantly lowers the overhead imposed by
752 using tracing.
753 The file trace driver expects a filename or a wrap files speci‐
755 fication as parameter. A file is written with a high degree of
756 buffering, why all trace messages are not guaranteed to be saved
757 in the file in case of a system crash. That is the price to pay
758 for low tracing overhead.
759 A wrap files specification is used to limit the disk space con‐
761 sumed by the trace. The trace is written to a limited number of
762 files each with a limited size. The actual filenames are File‐
763 name ++ SeqCnt ++ Suffix, where SeqCnt counts as a decimal
764 string from 0 to WrapCnt and then around again from 0. When a
765 trace term written to the current file makes it longer than
766 WrapSize, that file is closed, if the number of files in this
767 wrap trace is as many as WrapCnt the oldest file is deleted then
768 a new file is opened to become the current. Thus, when a wrap
769 trace has been stopped, there are at most WrapCnt trace files
770 saved with a size of at least WrapSize (but not much bigger),
771 except for the last file that might even be empty. The default
772 values are WrapSize = 128*1024 and WrapCnt = 8.
773 The SeqCnt values in the filenames are all in the range 0
775 through WrapCnt with a gap in the circular sequence. The gap is
776 needed to find the end of the trace.
777 If the WrapSize is specified as {time, WrapTime}, the current
779 file is closed when it has been open more than WrapTime mil‐
780 liseconds, regardless of it being empty or not.
781 The ip trace driver has a queue of QueSize messages waiting to
783 be delivered. If the driver cannot deliver messages as fast as
784 they are produced by the runtime system, a special message is
785 sent, which indicates how many messages that are dropped. That
786 message will arrive at the handler function specified in
787 trace_client/3 as the tuple {drop, N} where N is the number of
788 consecutive messages dropped. In case of heavy tracing, drop's
789 are likely to occur, and they surely occur if no client is read‐
790 ing the trace messages. The default value of QueSize is 200.
791 flush_trace_port()
793 Equivalent to flush_trace_port(node()).
795 flush_trace_port(Nodename) -> ok | {error, Reason}
797 Equivalent to trace_port_control(Nodename,flush).
799 trace_port_control(Operation)
801 Equivalent to trace_port_control(node(),Operation).
803 trace_port_control(Nodename,Operation) -> ok | {ok, Result} | {error,
805 Reason}
806 Types:
808 Nodename = atom()
810 This function is used to do a control operation on the active
812 trace port driver on the given node (Nodename). Which operations
813 are allowed as well as their return values depend on which trace
814 driver is used.
815 Returns either ok or {ok, Result} if the operation was success‐
817 ful, or {error, Reason} if the current tracer is a process or if
818 it is a port not supporting the operation.
819 The allowed values for Operation are:
821 flush:
823 This function is used to flush the internal buffers held by
824 a trace port driver. Currently only the file trace driver
825 supports this operation. Returns ok.
826 get_listen_port:
828 Returns {ok, IpPort} where IpPort is the IP port number used
829 by the driver listen socket. Only the ip trace driver sup‐
830 ports this operation.
831 trace_client(Type, Parameters) -> pid()
833 Types:
835 Type = ip | file | follow_file
837 Parameters = Filename | WrapFilesSpec | IPClientPortSpec
838 Filename = string() | [string()] | atom()
839 WrapFilesSpec = see trace_port/2
840 Suffix = string()
841 IpClientPortSpec = PortNumber | {Hostname, PortNumber}
842 PortNumber = integer()
843 Hostname = string()
844 This function starts a trace client that reads the output cre‐
846 ated by a trace port driver and handles it in mostly the same
847 way as a tracer process created by the tracer/0 function.
848 If Type is file, the client reads all trace messages stored in
850 the file named Filename or specified by WrapFilesSpec (must be
851 the same as used when creating the trace, see trace_port/2) and
852 let's the default handler function format the messages on the
853 console. This is one way to interpret the data stored in a file
854 by the file trace port driver.
855 If Type is follow_file, the client behaves as in the file case,
857 but keeps trying to read (and process) more data from the file
858 until stopped by stop_trace_client/1. WrapFilesSpec is not al‐
859 lowed as second argument for this Type.
860 If Type is ip, the client connects to the TCP/IP port PortNumber
862 on the host Hostname, from where it reads trace messages until
863 the TCP/IP connection is closed. If no Hostname is specified,
864 the local host is assumed.
865 As an example, one can let trace messages be sent over the net‐
867 work to another Erlang node (preferably not distributed), where
868 the formatting occurs:
869 On the node stack there's an Erlang node ant@stack, in the
871 shell, type the following:
872 ant@stack> dbg:tracer(port, dbg:trace_port(ip,4711)).
874 <0.17.0>
875 ant@stack> dbg:p(self(), send).
876 {ok,1}
877 All trace messages are now sent to the trace port driver, which
879 in turn listens for connections on the TCP/IP port 4711. If we
880 want to see the messages on another node, preferably on another
881 host, we do like this:
882 -> dbg:trace_client(ip, {"stack", 4711}).
884 <0.42.0>
885 If we now send a message from the shell on the node ant@stack,
887 where all sends from the shell are traced:
888 ant@stack> self() ! hello.
890 hello
891 The following will appear at the console on the node that
893 started the trace client:
894 (<0.23.0>) <0.23.0> ! hello
896 (<0.23.0>) <0.22.0> ! {shell_rep,<0.23.0>,{value,hello,[],[]}}
897 The last line is generated due to internal message passing in
899 the Erlang shell. The process id's will vary.
900 trace_client(Type, Parameters, HandlerSpec) -> pid()
902 Types:
904 Type = ip | file | follow_file
906 Parameters = Filename | WrapFilesSpec | IPClientPortSpec
907 Filename = string() | [string()] | atom()
908 WrapFilesSpec = see trace_port/2
909 Suffix = string()
910 IpClientPortSpec = PortNumber | {Hostname, PortNumber}
911 PortNumber = integer()
912 Hostname = string()
913 HandlerSpec = {HandlerFun, InitialData}
914 HandlerFun = fun() (two arguments)
915 InitialData = term()
916 This function works exactly as trace_client/2, but allows you to
918 write your own handler function. The handler function works
919 mostly as the one described in tracer/2, but will also have to
920 be prepared to handle trace messages of the form {drop, N},
921 where N is the number of dropped messages. This pseudo trace
922 message will only occur if the ip trace driver is used.
923 For trace type file, the pseudo trace message end_of_trace will
925 appear at the end of the trace. The return value from the han‐
926 dler function is in this case ignored.
927 stop_trace_client(Pid) -> ok
929 Types:
931 Pid = pid()
933 This function shuts down a previously started trace client. The
935 Pid argument is the process id returned from the trace_client/2
936 or trace_client/3 call.
937 get_tracer()
939 Equivalent to get_tracer(node()).
941 get_tracer(Nodename) -> {ok, Tracer}
943 Types:
945 Nodename = atom()
947 Tracer = port() | pid() | {module(), term()}
948 Returns the process, port or tracer module to which all trace
950 messages are sent.
951 stop() -> ok
953 Stops the dbg server and clears all trace flags for all pro‐
955 cesses and all local trace patterns for all functions. Also
956 shuts down all trace clients and closes all trace ports.
957 Note that no global trace patterns are affected by this func‐
959 tion.
960 stop_clear() -> ok
962 Same as stop/0, but also clears all trace patterns on global
964 functions calls.
965
967 The simplest way of tracing from the Erlang shell is to use dbg:c/3 or
968 dbg:c/4, e.g. tracing the function dbg:get_tracer/0:
969 (tiger@durin)84> dbg:c(dbg,get_tracer,[]).
971 (<0.154.0>) <0.152.0> ! {<0.154.0>,{get_tracer,tiger@durin}}
972 (<0.154.0>) out {dbg,req,1}
973 (<0.154.0>) << {dbg,{ok,<0.153.0>}}
974 (<0.154.0>) in {dbg,req,1}
975 (<0.154.0>) << timeout
976 {ok,<0.153.0>}
977 (tiger@durin)85>
978 Another way of tracing from the shell is to explicitly start a tracer
980 and then set the trace flags of your choice on the processes you want
981 to trace, e.g. trace messages and process events:
982 (tiger@durin)66> Pid = spawn(fun() -> receive {From,Msg} -> From ! Msg end end).
984 <0.126.0>
985 (tiger@durin)67> dbg:tracer().
986 {ok,<0.128.0>}
987 (tiger@durin)68> dbg:p(Pid,[m,procs]).
988 {ok,[{matched,tiger@durin,1}]}
989 (tiger@durin)69> Pid ! {self(),hello}.
990 (<0.126.0>) << {<0.116.0>,hello}
991 {<0.116.0>,hello}
992 (<0.126.0>) << timeout
993 (<0.126.0>) <0.116.0> ! hello
994 (<0.126.0>) exit normal
995 (tiger@durin)70> flush().
996 Shell got hello
997 ok
998 (tiger@durin)71>
999 If you set the call trace flag, you also have to set a trace pattern
1001 for the functions you want to trace:
1002 (tiger@durin)77> dbg:tracer().
1004 {ok,<0.142.0>}
1005 (tiger@durin)78> dbg:p(all,call).
1006 {ok,[{matched,tiger@durin,3}]}
1007 (tiger@durin)79> dbg:tp(dbg,get_tracer,0,[]).
1008 {ok,[{matched,tiger@durin,1}]}
1009 (tiger@durin)80> dbg:get_tracer().
1010 (<0.116.0>) call dbg:get_tracer()
1011 {ok,<0.143.0>}
1012 (tiger@durin)81> dbg:tp(dbg,get_tracer,0,[{'_',[],[{return_trace}]}]).
1013 {ok,[{matched,tiger@durin,1},{saved,1}]}
1014 (tiger@durin)82> dbg:get_tracer().
1015 (<0.116.0>) call dbg:get_tracer()
1016 (<0.116.0>) returned from dbg:get_tracer/0 -> {ok,<0.143.0>}
1017 {ok,<0.143.0>}
1018 (tiger@durin)83>
1019
1021 The dbg module is primarily targeted towards tracing through the er‐
1022 lang:trace/3 function. It is sometimes desired to trace messages in a
1023 more delicate way, which can be done with the help of the seq_trace
1024 module.
1025 seq_trace implements sequential tracing (known in the AXE10 world, and
1027 sometimes called "forlopp tracing"). dbg can interpret messages gener‐
1028 ated from seq_trace and the same tracer function for both types of
1029 tracing can be used. The seq_trace messages can even be sent to a trace
1030 port for further analysis.
1031 As a match specification can turn on sequential tracing, the combina‐
1033 tion of dbg and seq_trace can be quite powerful. This brief example
1034 shows a session where sequential tracing is used:
1035 1> dbg:tracer().
1037 {ok,<0.30.0>}
1038 2> {ok, Tracer} = dbg:get_tracer().
1039 {ok,<0.31.0>}
1040 3> seq_trace:set_system_tracer(Tracer).
1041 false
1042 4> dbg:tp(dbg, get_tracer, 0, [{[],[],[{set_seq_token, send, true}]}]).
1043 {ok,[{matched,nonode@nohost,1},{saved,1}]}
1044 5> dbg:p(all,call).
1045 {ok,[{matched,nonode@nohost,22}]}
1046 6> dbg:get_tracer(), seq_trace:set_token([]).
1047 (<0.25.0>) call dbg:get_tracer()
1048 SeqTrace [0]: (<0.25.0>) <0.30.0> ! {<0.25.0>,get_tracer} [Serial: {2,4}]
1049 SeqTrace [0]: (<0.30.0>) <0.25.0> ! {dbg,{ok,<0.31.0>}} [Serial: {4,5}]
1050 {1,0,5,<0.30.0>,4}
1051 This session sets the system_tracer to the same process as the ordinary
1053 tracer process (i. e. <0.31.0>) and sets the trace pattern for the
1054 function dbg:get_tracer to one that has the action of setting a sequen‐
1055 tial token. When the function is called by a traced process (all pro‐
1056 cesses are traced in this case), the process gets "contaminated" by the
1057 token and seq_trace messages are sent both for the server request and
1058 the response. The seq_trace:set_token([]) after the call clears the
1059 seq_trace token, why no messages are sent when the answer propagates
1060 via the shell to the console port. The output would otherwise have been
1061 more noisy.
1062
1064 When tracing function calls on a group leader process (an IO process),
1065 there is risk of causing a deadlock. This will happen if a group leader
1066 process generates a trace message and the tracer process, by calling
1067 the trace handler function, sends an IO request to the same group
1068 leader. The problem can only occur if the trace handler prints to tty
1069 using an io function such as format/2. Note that when dbg:p(all,call)
1070 is called, IO processes are also traced. Here's an example:
1071 %% Using a default line editing shell
1073 1> dbg:tracer(process, {fun(Msg,_) -> io:format("~p~n", [Msg]), 0 end, 0}).
1074 {ok,<0.37.0>}
1075 2> dbg:p(all, [call]).
1076 {ok,[{matched,nonode@nohost,25}]}
1077 3> dbg:tp(mymod,[{'_',[],[]}]).
1078 {ok,[{matched,nonode@nohost,0},{saved,1}]}
1079 4> mymod: % TAB pressed here
1080 %% -- Deadlock --
1081 Here's another example:
1083 %% Using a shell without line editing (oldshell)
1085 1> dbg:tracer(process).
1086 {ok,<0.31.0>}
1087 2> dbg:p(all, [call]).
1088 {ok,[{matched,nonode@nohost,25}]}
1089 3> dbg:tp(lists,[{'_',[],[]}]).
1090 {ok,[{matched,nonode@nohost,0},{saved,1}]}
1091 % -- Deadlock --
1092 The reason we get a deadlock in the first example is because when TAB
1094 is pressed to expand the function name, the group leader (which handles
1095 character input) calls mymod:module_info(). This generates a trace mes‐
1096 sage which, in turn, causes the tracer process to send an IO request to
1097 the group leader (by calling io:format/2). We end up in a deadlock.
1098 In the second example we use the default trace handler function. This
1100 handler prints to tty by sending IO requests to the user process. When
1101 Erlang is started in oldshell mode, the shell process will have user as
1102 its group leader and so will the tracer process in this example. Since
1103 user calls functions in lists we end up in a deadlock as soon as the
1104 first IO request is sent.
1105 Here are a few suggestions for how to avoid deadlock:
1107 * Don't trace the group leader of the tracer process. If tracing has
1109 been switched on for all processes, call dbg:p(TracerGLPid,clear)
1110 to stop tracing the group leader (TracerGLPid). process_info(Trac‐
1111 erPid,group_leader) tells you which process this is (TracerPid is
1112 returned from dbg:get_tracer/0).
1113 * Don't trace the user process if using the default trace handler
1115 function.
1116 * In your own trace handler function, call erlang:display/1 instead
1118 of an io function or, if user is not used as group leader, print to
1119 user instead of the default group leader. Example: io:for‐
1120 mat(user,Str,Args).
1121Ericsson AB runtime_tools 1.16.1 dbg(3)