1<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
2<html xmlns="http://www.w3.org/1999/xhtml">
3<head>
4<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1" />
5<link href="style.css" rel="stylesheet" type="text/css" />
6<title>LLDB Tutorial</title>
7</head>
8
9<body>
10    <div class="www_title">
11      The <strong>LLDB</strong> Debugger
12    </div>
13
14<div id="container">
15	<div id="content">
16         <!--#include virtual="sidebar.incl"-->
17		<div id="middle">
18			<div class="post">
19				<h1 class ="postheader">Getting Started</h1>
20				<div class="postcontent">
21
22                    <p>Here's a short precis of how to run lldb if you are familiar with the gdb command set.
23                        We will start with some details on lldb command structure and syntax to help orient you.</p>
24
25				</div>
26				<div class="postfooter"></div>
27
28    			<div class="post">
29    				<h1 class ="postheader">Command Structure</h1>
30    				<div class="postcontent">
31
32				   <p>Unlike gdb's command set, which is rather free-form, we tried to make
33                   the lldb command syntax fairly structured.  The commands are all of the
34                   form:</p>
35
36                   <code color=#ff0000>
37                   &lt;noun&gt; &lt;verb&gt; [-options [option-value]] [argument [argument...]]
38                   </code>
39
40                   <p>The command line parsing is done before command execution, so it is
41                   uniform across all the commands.  The command syntax for basic commands is very simple,
42                   arguments, options and option values are all white-space
43                   separated, and double-quotes are used to protect white-spaces in an argument.
44                   If you need to put a backslash or double-quote character
45                   in an argument you back-slash it in the argument.  That makes the
46                   command syntax more regular, but it also means you may have to
47                   quote some arguments in lldb that you wouldn't in gdb.</p>
48
49                   <p>Options can be placed anywhere on the command line, but if the arguments
50                   begin with a "<code>-</code>" then you have to tell lldb that you're done with options
51                   for the current command by adding an option termination: "<code>--</code>"
52                   So for instance if you want to launch a process and give the "process launch" command
53                   the "<code>--stop-at-entry</code>" option, yet you want the
54                   process you are about to launch to be launched with the arguments
55                   "<code>-program_arg value</code>", you would type:</p>
56
57                   <code>
58                       (lldb) process launch --stop-at-entry -- -program_arg value
59                   </code>
60
61                   <p>We also tried to reduce the number of special purpose argument
62                   parsers, which sometimes forces the user to be a little more explicit
63                   about stating their intentions.  The first instance you'll note of
64                   this is the breakpoint command.  In gdb, to set a breakpoint, you
65                   might enter</p>
66
67                   <code>
68                       (gdb) break foo.c:12
69                   </code>
70                   <p>to break at line 12 of foo.c, and:</p>
71                   <code>
72                       (gdb) break foo
73                   </code>
74
75                   <p>to break at the function <code>foo</code>.  As time went on, the parser that tells <code>foo.c:12</code>
76                   from <code>foo</code> from <code>foo.c::foo</code> (which means the function foo in the file
77                   foo.c) got more and more complex and bizarre, and especially in C++
78                   there are times where there's really no way to specify the function
79                   you want to break on. The lldb commands are more verbose but also more precise
80                   and allow for intellegent auto completion.
81
82                   <p>To set the same file and line breakpoint in LLDB you can enter either of:</p>
83
84                   <code>
85                      (lldb) breakpoint set --file foo.c --line 12
86                      <br>(lldb) breakpoint set -f foo.c -l 12
87                  </code>
88
89                   <p>To set a breakpoint on a function named <code>foo</code> in LLDB you can enter either of:</p>
90
91                    <code>
92                        (lldb) breakpoint set --name foo
93                        <br>(lldb) breakpoint set -n foo
94                    </code>
95
96                   <p>You can use the --name option multiple times to make a breakpoint on a set of functions as well.  This is convenient
97                      since it allows you to set commmon conditions or commands without having to specify them multiple times:</p>
98
99                    <code>
100                        (lldb) breakpoint set --name foo --name bar
101                    </code>
102
103                   <p>Setting breakpoints by name is even more specialized in LLDB as you can specify
104                       that you want to set a breakpoint at a function by method name. To set a breakpoint
105                       on all C++ methods named <code>foo</code> you can enter either of:</p>
106
107                   <code>
108                       (lldb) breakpoint set --method foo
109                       <br>(lldb) breakpoint set -M foo
110                  </code>
111
112                   <p>To set a breakpoint Objective C selectors named <code>alignLeftEdges:</code> you can enter either of:</p>
113
114                   <code>
115                       (lldb) breakpoint set --selector alignLeftEdges:
116                       <br>(lldb) breakpoint set -S alignLeftEdges:
117                  </code>
118
119                   <p>You can limit any breakpoints to a specific executable image by using
120                       the "<code>--shlib &lt;path&gt;</code>" ("<code>-s &lt;path&gt;</code>" for short):</p>
121
122                   <code>
123                      (lldb) breakpoint set --shlib foo.dylib --name foo
124                      <br>(lldb) breakpoint set -s foo.dylib -n foo
125                  </code>
126
127                   <p>The <code>--shlib</code> option can also be repeated to specify several shared libraries.</p>
128
129                   <p>Suggestions on more interesting primitives of this sort are also very welcome.</p>
130
131                   <p>Just like gdb, the lldb command interpreter does a shortest unique
132                   string match on command names, so the following two commands will
133                   both execute the same command:</p>
134
135                   <code>
136                       (lldb) breakpoint set -n "-[SKTGraphicView alignLeftEdges:]"
137                       <br>(lldb) br s -n "-[SKTGraphicView alignLeftEdges:]"
138                  </code>
139
140                   <p>lldb also supports command completion for source file names, symbol
141                   names, file names, etc. Completion is initiated by a hitting a <b>TAB</b>.
142                   Individual options in a command can have different completers, so for
143                   instance the "<code>--file &lt;path&gt;</code>" option in "breakpoint" completes to source files, the
144                   "<code>--shlib &lt;path&gt;</code>" option to currently loaded shared libraries, etc.  We can even do
145                   things like if you specify "<code>--shlib &lt;path&gt;</code>", and are completing on "<code>--file &lt;path&gt;</code>", we will only
146                   list source files in the shared library specified by "<code>--shlib &lt;path&gt;</code>".</p>
147
148                   <p>The individual commands are pretty extensively documented.  You can
149                   use the <code>help</code> command to get an overview of which commands are
150                   available or to obtain details about specific commands.  There is also an
151                   <code>apropos</code> command that will search the help text for all commands
152                   for a particular word and dump a summary help string for each matching
153                   command.</p>
154
155                   <p>Finally, there is a mechanism to construct aliases for commonly used
156                   commands.  So for instance if you get annoyed typing:</p>
157
158                   <code>
159                       (lldb) breakpoint set --file foo.c --line 12
160                   </code>
161
162                   <p>you can do:</p>
163
164                   <code>
165                       (lldb) command alias bfl breakpoint set -f %1 -l %2
166                       <br>(lldb) bfl foo.c 12
167                   </code>
168
169                   <p>We have added a few aliases for commonly used commands (e.g. "step",
170                   "next" and "continue") but we haven't tried to be exhaustive because
171                   in our experience it is more convenient to make the basic commands
172                   unique down to a letter or two, and then learn these sequences than
173                   to fill the namespace with lots of aliases, and then have to type them
174                   all the way out.</p>
175
176                   <p>However, users are free to customize lldb's command set however they
177                   like, and since lldb reads the file ~/.lldbinit at startup, you can
178                   store all your aliases there and they will be generally available to
179                   you.  Your aliases are also documented in the help command so you can
180                   remind yourself of what you've set up.</p>
181
182                   <p> One alias of note that we do include by popular demand is a weak emulator
183                   of gdb's &quot;break&quot; command.  It doesn't try to do everything that gdb's
184                   break command does (for instance, it doesn't handle <code>foo.c::bar</code>.  But
185                   it mostly works, and makes the transition easier.  Also by popular demand, it
186                   is aliased to <code>b</code>.  If you actually want to learn the lldb command
187                   set natively, that means it will get in the way of the rest of the breakpoint
188                   commands.  Fortunately, if you don't like one of our aliases, you an easily
189                   get rid of it by running (for example):</p>
190
191                   <code>
192                    (lldb) command unalias b
193                   </code>
194
195                   <p>I actually also do:</p>
196
197                   <code>
198                     (lldb) command alias b breakpoint
199                   </code>
200
201                   <p>so I can run the native lldb breakpoint command with just <code>b</code></p>
202
203                   <p>The lldb command parser also supports "raw" commands, where, after command options
204                   are stripped off, the rest of the command string is passed uninterpreted to the command.
205                   This is convenient for commands whose arguments might be some complex expression that would
206                   be painful to backslash protect.
207                   For instance the "expression" command is a "raw" command for obvious reasons.  The
208                   "help" output for a command will tell you if it is "raw" or not, so you know what to expect.
209                   The one thing you have to watch out for is that since raw commands still can have options,
210                   if your command string has dashes in it, you'll have to indicate these are not option
211                   markers by putting "--" after the command name, but before your command string.
212
213                   <p>lldb also has a built-in Python interpreter, which is accessible by
214                   the "script" command.  All the functionality of the debugger is
215                   available as classes in the Python interpreter, so the more complex
216                   commands that in gdb you would introduce with the "define" command can
217                   be done by writing Python functions using the lldb-Python library,
218                   then loading the scripts into your running session and accessing them
219                   with the "script" command.</p>
220
221                   <p>Having given an overview of lldb's command syntax, we proceed to lay out the stages
222                   of a standard debug session.</p>
223
224				</div>
225				<div class="postfooter"></div>
226
227
228    			<div class="post">
229    				<h1 class ="postheader">Loading a program into lldb</h1>
230    				<div class="postcontent">
231
232                    <p>First we need to set the program to debug. As with gdb, you
233                         can start lldb and specify the file you wish to debug on the command line:</p>
234
235                    <code>
236                        $ lldb /Projects/Sketch/build/Debug/Sketch.app
237                        <br>Current executable set to '/Projects/Sketch/build/Debug/Sketch.app' (x86_64).
238                    </code>
239
240                    <p>or you can specify it after the fact with the "file" command:</p>
241
242                    <code>
243                        $ lldb
244                        <br>(lldb) file /Projects/Sketch/build/Debug/Sketch.app
245                        <br>Current executable set to '/Projects/Sketch/build/Debug/Sketch.app' (x86_64).
246                    </code>
247                <p>
248				</div>
249				<div class="postfooter"></div>
250
251    			<div class="post">
252    				<h1 class ="postheader">Setting breakpoints</h1>
253    				<div class="postcontent">
254
255                    <p>We've discussed how to set breakpoints above.  You can use <code>help breakpoint set</code>
256                    to see all the options for breakpoint setting.  For instance, we might do:</p>
257
258                    <code>
259                        (lldb) breakpoint set --selector alignLeftEdges:
260                        <br>Breakpoint created: 1: name = 'alignLeftEdges:', locations = 1, resolved = 1
261                    </code>
262
263                    <p>You can find out about the breakpoints you've set with:</p>
264
265                    <pre><tt>(lldb) breakpoint list
266Current breakpoints:
2671: name = 'alignLeftEdges:', locations = 1, resolved = 1
268  1.1: where = Sketch`-[SKTGraphicView alignLeftEdges:] + 33 at /Projects/Sketch/SKTGraphicView.m:1405, address = 0x0000000100010d5b, resolved, hit count = 0
269</tt></pre>
270
271                    <p>Note that setting a breakpoint creates a <i>logical</i> breakpoint, which could
272                    resolve to one or more <i>locations</i>.  For instance, break by selector would
273                    set a breakpoint on all the methods that implement that selector in the classes in
274                    your program.  Similarly, a file and line breakpoint might result in multiple
275                    locations if that file and line were inlined in different places in your code.</p>
276
277                    <p>The logical breakpoint has an integer id, and it's locations have an
278                    id within their parent breakpoint (the two are joined by a ".",
279                    e.g. 1.1 in the example above.)  </p>
280
281                    <p>Also the logical breakpoints remain <i>live</i> so that if another shared library
282                    were to be loaded that had another implementation of the
283                    "<code>alignLeftEdges:</code>" selector, the new location would be added to
284                    breakpoint 1 (e.g. a "1.2" breakpoint would be set on the newly loaded
285                    selector).</p>
286
287                    <p>The other piece of information in the breakpoint listing is whether the
288                    breakpoint location was <i>resolved</i> or not.  A location gets resolved when
289                    the file address it corresponds to gets loaded into the program you are
290                    debugging.  For instance if you set a breakpoint in a shared library that
291                    then gets unloaded, that breakpoint location will remain, but it will no
292                    longer be <i>resolved</i>.</p>
293
294                    <p>One other thing to note for gdb users is that lldb acts like gdb with:</p>
295
296                    <code>
297                        (gdb) set breakpoint pending on
298                    </code>
299
300                    <p>That is, lldb will always make a breakpoint from your specification, even
301                    if it couldn't find any locations that match the specification.  You can tell
302                    whether the expression was resolved or not by checking the locations field
303                    in "breakpoint list", and we report the breakpoint as "pending" when you
304                    set it so you can tell you've made a typo more easily, if that was indeed
305                    the reason no locations were found:</p>
306
307                    <code>
308                        (lldb) breakpoint set --file foo.c --line 12
309                        <br>Breakpoint created: 2: file ='foo.c', line = 12, locations = 0 (pending)
310                        <br>WARNING:  Unable to resolve breakpoint to any actual locations.
311                    </code>
312
313                    <p>You can delete, disable, set conditions and ignore counts either on all the
314                    locations generated by your logical breakpoint, or on any one of the particular locations
315                    your specification resolved to.  For instance if we wanted to add a command
316                    to print a backtrace when we hit this breakpoint we could do:</p>
317
318                    <code>
319                        (lldb) breakpoint command add 1.1
320                        <br>Enter your debugger command(s).  Type 'DONE' to end.
321                        <br>&gt; bt
322                        <br>&gt; DONE
323                    </code>
324
325                    <p>By default, the <code> breakpoint command add</code> command takes lldb command line commands.
326                    You can also specify this explicitly by passing the "<code>--command</code>" option.
327                    Use "<code>--script</code>" if you want to implement your breakpoint command using the Python script instead.</p>
328
329                    <p>This is an convenient point to bring up another feature of the lldb command help.  Do:</p>
330
331                    <code>
332                    (lldb) help break command add
333                     <br>Add a set of commands to a breakpoint, to be executed whenever the breakpoint is hit.
334                     <br>
335                     <br>Syntax: breakpoint command add &lt;cmd-options&gt; &lt;breakpt-id&gt;
336                     <br> etc...
337                    </code>
338
339                     <p>When you see arguments to commands specified in the Syntax in angle
340                     brackets like <code>&lt;breakpt-id&gt;</code>, that indicates that
341                     that is some common argument type that you can get further help on from the command system.
342                     So in this case you could do:</p>
343
344                     <code>
345                    (lldb) help &lt;breakpt-id&gt;
346                    <br>&lt;breakpt-id&gt; -- Breakpoint ID's consist major and minor numbers;  the major
347                    <br> etc...
348                     </code>
349
350			    </div>
351				<div class="postfooter"></div>
352
353                <div class="post">
354    				<h1 class ="postheader">Setting watchpoints</h1>
355    				<div class="postcontent">
356
357                    <p>In addition to breakpoints, you can use <code>help watchpoint</code>
358                    to see all the commands for watchpoint manipulations.  For instance, we might do the following to watch
359                    a variable called 'global' for write operation, but only stop if the condition '(global==5)' is true:</p>
360
361                   <pre><tt>(lldb) watch set var global
362Watchpoint created: Watchpoint 1: addr = 0x100001018 size = 4 state = enabled type = w
363    declare @ '/Volumes/data/lldb/svn/ToT/test/functionalities/watchpoint/watchpoint_commands/condition/main.cpp:12'
364(lldb) watch modify -c '(global==5)'
365(lldb) watch list
366Current watchpoints:
367Watchpoint 1: addr = 0x100001018 size = 4 state = enabled type = w
368    declare @ '/Volumes/data/lldb/svn/ToT/test/functionalities/watchpoint/watchpoint_commands/condition/main.cpp:12'
369    condition = '(global==5)'
370(lldb) c
371Process 15562 resuming
372(lldb) about to write to 'global'...
373Process 15562 stopped and was programmatically restarted.
374Process 15562 stopped and was programmatically restarted.
375Process 15562 stopped and was programmatically restarted.
376Process 15562 stopped and was programmatically restarted.
377Process 15562 stopped
378* thread #1: tid = 0x1c03, 0x0000000100000ef5 a.out`modify + 21 at main.cpp:16, stop reason = watchpoint 1
379    frame #0: 0x0000000100000ef5 a.out`modify + 21 at main.cpp:16
380   13
381   14  	static void modify(int32_t &var) {
382   15  	    ++var;
383-> 16  	}
384   17
385   18  	int main(int argc, char** argv) {
386   19  	    int local = 0;
387(lldb) bt
388* thread #1: tid = 0x1c03, 0x0000000100000ef5 a.out`modify + 21 at main.cpp:16, stop reason = watchpoint 1
389    frame #0: 0x0000000100000ef5 a.out`modify + 21 at main.cpp:16
390    frame #1: 0x0000000100000eac a.out`main + 108 at main.cpp:25
391    frame #2: 0x00007fff8ac9c7e1 libdyld.dylib`start + 1
392(lldb) frame var global
393(int32_t) global = 5
394(lldb) watch list -v
395Current watchpoints:
396Watchpoint 1: addr = 0x100001018 size = 4 state = enabled type = w
397    declare @ '/Volumes/data/lldb/svn/ToT/test/functionalities/watchpoint/watchpoint_commands/condition/main.cpp:12'
398    condition = '(global==5)'
399    hw_index = 0  hit_count = 5     ignore_count = 0
400(lldb) </tt></pre>
401			    </div>
402				<div class="postfooter"></div>
403
404    			<div class="post">
405    				<h1 class ="postheader">Starting or attaching to your Program</h1>
406    				<div class="postcontent">
407
408                    <p>To launch a program in lldb we use the "<code>process launch</code>" command or
409                        one of its built in aliases:</p>
410
411                    <code>
412                        (lldb) process launch
413                        <br>(lldb) run
414                        <br>(lldb) r
415                    </code>
416
417                    <p>You can also attach to a process by process ID or process name.
418                        When attaching to a process by name, lldb also supports the "<code>--waitfor</code>" option which waits for the
419                    next process that has that name to show up, and attaches to it</p>
420
421                    <code>
422                        (lldb) process attach --pid 123
423                        <br>(lldb) process attach --name Sketch
424                        <br>(lldb) process attach --name Sketch --waitfor
425                    </code>
426
427                    <p>After you launch or attach to a process, your process might stop
428                        somewhere:</p>
429                        <code>
430                            (lldb) process attach -p 12345
431                            <br>Process 46915 Attaching
432                            <br>Process 46915 Stopped
433                            <br>1 of 3 threads stopped with reasons:
434                            <br>* thread #1: tid = 0x2c03, 0x00007fff85cac76a, where = libSystem.B.dylib`__getdirentries64 + 10, stop reason = signal = SIGSTOP, queue = com.apple.main-thread
435                        </code>
436
437
438                    <p>Note the line that says "<code>1 of 3 threads stopped with reasons:</code>" and the
439                    lines that follow it. In a multi-threaded environment it is very
440                    common for more than one thread to hit your breakpoint(s) before the
441                    kernel actually returns control to the debugger.  In that case, you
442                    will see all the threads that stopped for some interesting reason
443                    listed in the stop message.</p>
444
445				</div>
446				<div class="postfooter"></div>
447
448    			<div class="post">
449    				<h1 class ="postheader">Controlling your Program</h1>
450    				<div class="postcontent">
451
452
453                    <p>After launching, we can continue until we hit our breakpoint.  The primitive
454                    commands for process control all exist under the "thread" command:</p>
455
456                    <code>
457                        (lldb) thread continue
458                        <br>Resuming thread 0x2c03 in process 46915
459                        <br>Resuming process 46915
460                        <br>(lldb)
461                    </code>
462
463                    <p>At present you can only operate on one thread at a time, but the
464                    design will ultimately support saying "step over the function in
465                    Thread 1, and step into the function in Thread 2, and continue Thread
466                    3" etc.  When we eventually support keeping some threads running while
467                    others are stopped this will be particularly important.  For
468                    convenience, however, all the stepping commands have easy aliases.
469                    So "thread continue" is just "c", etc.</p>
470
471                    <p>The other program stepping commands are pretty much the same as in gdb.
472                    You've got:</p>
473
474                    <pre><tt>(lldb) thread step-in    // The same as gdb's "step" or "s"
475(lldb) thread step-over  // The same as gdb's "next" or "n"
476(lldb) thread step-out   // The same as gdb's "finish" or "f"
477</tt></pre>
478
479                    <p>By default, lldb does defined aliases to all common gdb process control
480                        commands ("<code>s</code>", "<code>step</code>", "<code>n</code>", "<code>next</code>", "<code>finish</code>").
481                        If we have missed any, please add them to your <code>~/.lldbinit</code> file
482                        using the "<code>command alias</code>" command.
483
484                    <p>lldb also supported the <i>step by instruction</i> versions:</p>
485                    <pre><tt>(lldb) thread step-inst       // The same as gdb's "stepi" / "si"
486(lldb) thread step-over-inst  // The same as gdb's "nexti" / "ni"
487</tt></pre>
488
489                    <p>Finally, lldb has a <i>run until line or frame exit</i> stepping mode:</p>
490
491                    <code>
492                        (lldb) thread until 100
493                    </code>
494
495                    <p>This command will run the thread in the current frame till it reaches line 100 in
496                    this frame or stops if it leaves the current frame.  This is a pretty
497                    close equivalent to gdb's "<code>until</code>" command.</p>
498
499                    <p>A process, by default, will shared the lldb terminal with the inferior
500                        process. When in this mode, much like when debugging with gdb, when
501                        the process is running anything you type will go to the STDIN of the
502                        inferior process. To interrupt your inferior program, type CTRL+C.</p>
503
504                    <p>If you attach to a process, or launch a process with the &quot;<code>--no-stdin</code>&quot;
505                        option, the command interpreter is always available to enter commands. This
506                        might be a little disconcerting to gdb users when always have an <code>(lldb)</code>
507                        prompt. This allows you to set a breakpoint, etc without having to explicitly interrupt
508                        the program you are debugging:</p>
509
510                        <code>
511                            (lldb) process continue
512                            <br>(lldb) breakpoint set --name stop_here
513                        </code>
514
515                    <p>There are many commands that won't work while running, and the command
516                        interpreter should do a good job of letting you know when this is the
517                        case. If you find any instances where the command interpreter isn't
518                        doing its job, please file a bug. This way of operation will set us
519                        up for a future debugging mode called <i>thread centric debugging</i>.
520                        This mode will allow us to run all threads and only stop the threads
521                        that are at breakpoints or have exceptions or signals.</p>
522
523                    <p>The commands that currently work while running include
524                        interrupting the process to halt execution ("<code>process interrupt</code>"),
525                        getting the process status ("<code>process status</code>"),
526                        breakpoint setting and clearing ("<code> breakpoint [set|clear|enable|disable|list] ...</code>"),
527                        and memory reading and writing  ("<code> memory [read|write] ...</code>").
528                        </p>
529
530                    <p>The question of disabling stdio when running brings up a good opportunity to
531                        show how to set debugger properties in general.
532                        If you always want to run in the <code>--no-stdin</code> mode, you can set this
533                        as a generic process property using the lldb &quot;<code>settings</code>&qout; command,
534                        which is equivalent to gdb's &quot;<code>set</code>&quot; command.  For instance,
535                        in this case you would say:</p>
536
537                    <code>
538                    (lldb) settings set target.process.disable-stdio true
539                    </code>
540
541                    <p>Over time, gdb's &quot;<code>set</code> command became a wilderness of disordered options,
542                        so that there were useful options that even experienced gdb users didn't know about
543                        because they were too hard to find.  We tried to organize the settings hierarchically
544                        using the structure of the basic entities in the debugger.  For the most part anywhere
545                        you can specify a setting on a generic entity (threads, for example) you can also apply
546                        the option to a particular instance, which can also be convenient at times.
547                        You can view the available settings with &quot;<code>settings list</code>&quot; and
548                        there is help on the settings command explaining how it works more generally.</p>
549
550    				</div>
551    				<div class="postfooter"></div>
552
553        			<div class="post">
554        				<h1 class ="postheader">Examining Thread State</h1>
555        				<div class="postcontent">
556
557                    <p>Once you've stopped, lldb will choose a current thread, usually the
558                    one that stopped "for a reason", and a current frame in that thread (on stop this is always the bottom-most frame).
559                    Many the commands for inspecting state work on this current
560                    thread/frame.</p>
561
562                    <p>To inspect the current state of your process, you can start with the
563                    threads:</p>
564
565                    <pre><tt>(lldb) thread list
566Process 46915 state is Stopped
567* thread #1: tid = 0x2c03, 0x00007fff85cac76a, where = libSystem.B.dylib`__getdirentries64 + 10, stop reason = signal = SIGSTOP, queue = com.apple.main-thread
568  thread #2: tid = 0x2e03, 0x00007fff85cbb08a, where = libSystem.B.dylib`kevent + 10, queue = com.apple.libdispatch-manager
569  thread #3: tid = 0x2f03, 0x00007fff85cbbeaa, where = libSystem.B.dylib`__workq_kernreturn + 10
570</tt></pre>
571
572                    <p>The * indicates that Thread 1 is the current thread.  To get a
573                    backtrace for that thread, do:</p>
574
575                    <pre><tt>(lldb) thread backtrace
576thread #1: tid = 0x2c03, stop reason = breakpoint 1.1, queue = com.apple.main-thread
577 frame #0: 0x0000000100010d5b, where = Sketch`-[SKTGraphicView alignLeftEdges:] + 33 at /Projects/Sketch/SKTGraphicView.m:1405
578 frame #1: 0x00007fff8602d152, where = AppKit`-[NSApplication sendAction:to:from:] + 95
579 frame #2: 0x00007fff860516be, where = AppKit`-[NSMenuItem _corePerformAction] + 365
580 frame #3: 0x00007fff86051428, where = AppKit`-[NSCarbonMenuImpl performActionWithHighlightingForItemAtIndex:] + 121
581 frame #4: 0x00007fff860370c1, where = AppKit`-[NSMenu performKeyEquivalent:] + 272
582 frame #5: 0x00007fff86035e69, where = AppKit`-[NSApplication _handleKeyEquivalent:] + 559
583 frame #6: 0x00007fff85f06aa1, where = AppKit`-[NSApplication sendEvent:] + 3630
584 frame #7: 0x00007fff85e9d922, where = AppKit`-[NSApplication run] + 474
585 frame #8: 0x00007fff85e965f8, where = AppKit`NSApplicationMain + 364
586 frame #9: 0x0000000100015ae3, where = Sketch`main + 33 at /Projects/Sketch/SKTMain.m:11
587 frame #10: 0x0000000100000f20, where = Sketch`start + 52
588</tt></pre>
589
590                    <p>You can also provide a list of threads to backtrace, or the keyword
591                    "all" to see all threads:</p>
592
593                    <code>
594                        (lldb) thread backtrace all
595                    </code>
596
597                    <p>You can select the current thread, which will be used by default in all the commands in
598                    the next section, with the "thread select" command:</p>
599
600                    <code>
601                      (lldb) thread select 2
602                    </code>
603
604                    <p>where the thread index is just the one shown in the &quot;<code>thread list</code>&quot listing.
605
606				</div>
607				<div class="postfooter"></div>
608
609    			<div class="post">
610    				<h1 class ="postheader">Examining Stack Frame State</h1>
611    				<div class="postcontent">
612
613
614                    <p>The most convenient way to inspect a frame's arguments and local variables is to use the "<code>frame variable</code>" command:</p>
615
616                    <code>
617                        (lldb) frame variable
618                        <br>self = (SKTGraphicView *) 0x0000000100208b40
619                        <br>_cmd = (struct objc_selector *) 0x000000010001bae1
620                        <br>sender = (id) 0x00000001001264e0
621                        <br>selection = (NSArray *) 0x00000001001264e0
622                        <br>i = (NSUInteger) 0x00000001001264e0
623                        <br>c = (NSUInteger) 0x00000001001253b0
624                    </code>
625
626                    <p>As you see above, if you don't specify any variable names, all arguments
627                        and locals will be shown. If you call "<code>frame variable</code>"
628                        passing in the names of a particular local(s), only those variables
629                        will be printed.  For instance:
630                        </p>
631
632                    <code>
633                        (lldb) frame variable self
634                        <br>(SKTGraphicView *) self = 0x0000000100208b40
635                    </code>
636
637                    <p>You can also pass in a path to some subelement of one of the available locals,
638                      and that sub-element will be printed. For instance:
639                      </p>
640
641                    <code>
642                        <br>(lldb) frame variable self.isa
643                        <br>(struct objc_class *) self.isa = 0x0000000100023730
644                    </code>
645
646                    <p>The "<code>frame variable</code>" command is not a full expression
647                        parser but it does support a few simple operations like &amp;, *, ->, [] (no overloaded
648                    operators). The array brackets can be used on pointers to treat pointers
649                    as arrays:</p>
650
651                    <code>
652                        (lldb) frame variable *self
653                        <br>(SKTGraphicView *) self = 0x0000000100208b40
654                        <br>(NSView) NSView = {
655                        <br>(NSResponder) NSResponder = {
656                        <br>...
657                        <br>
658                        <br>(lldb) frame variable &amp;self
659                        <br>(SKTGraphicView **) &amp;self = 0x0000000100304ab
660                        <br>
661                        <br>(lldb) frame variable argv[0]
662                        <br>(char const *) argv[0] = 0x00007fff5fbffaf8 "/Projects/Sketch/build/Debug/Sketch.app/Contents/MacOS/Sketch"
663                    </code>
664
665                    <p>The frame variable command will also perform "object printing" operations on
666                    variables (currently we only support ObjC printing, using the object's "description" method.
667                    Turn this on by passing the -o flag to frame variable:</p>
668
669                    <code>
670                        (lldb) frame variable -o self
671                        (SKTGraphicView *) self = 0x0000000100208b40 &lt;SKTGraphicView: 0x100208b40&gt;
672                    </code>
673
674                    <p>You can select another frame to view with the "<code>frame select</code>" command</p>
675
676                    <code>
677                        (lldb) frame select 9
678                        <br>frame #9: 0x0000000100015ae3, where = Sketch`function1 + 33 at /Projects/Sketch/SKTFunctions.m:11
679                    </code>
680
681                    <p>You can also move up and down the stack by passing the &quot;<code>--relative</code>&quot; (&quot;<code>-r</code>&quot;)
682                      option.  And we have built-in aliases &quot;<code>u</code>&quot; and &quot;<code>d</code>&quot; which
683                      behave like their gdb equivalents.
684
685                    <p>If you need to view more complex data or change program data, you can
686                    use the general "expression" command.  It takes an expression and
687                    evaluates it in the scope of the currently selected frame.  For instance:</p>
688
689                    <code>
690                        (lldb) expr self
691                        <br>$0 = (SKTGraphicView *) 0x0000000100135430
692                        <br>(lldb) expr self = 0x00
693                        <br>$1 = (SKTGraphicView *) 0x0000000000000000
694                        <br>(lldb) frame var self
695                        <br>(SKTGraphicView *) self = 0x0000000000000000
696                    </code>
697
698                    <p>You can also call functions:</p>
699
700                    <code>
701                        (lldb) expr (int) printf ("I have a pointer 0x%llx.\n", self)
702                        <br>$2 = (int) 22
703                        <br>I have a pointer 0x0.
704                    </code>
705
706                    <p>As I said above, &quot;expression&quot; is one of the &quot;raw&quot; commands.  So
707                    you don't have to quote your whole expression, nor backslash protect quotes,
708                    etc...</p>
709
710                    <p>Finally, the results of the expressions are stored in persistent variables
711                    (of the form $[0-9]+) that you can use in further expressions, like:</p>
712
713                    <code>
714                        (lldb) expr self = $0
715                        <br>$4 = (SKTGraphicView *) 0x0000000100135430
716				    </code>
717                    <p>
718                </div>
719          	    <div class="postfooter"></div>
720
721            </div>
722      	</div>
723	</div>
724</div>
725</body>
726</html>
727