blob: eb3c5a223bf7e18c46b8f7078e0b186b7de0d8c4 [file] [log] [blame]
Jim Laskeyd0d39b62007-03-14 19:29:42 +00001<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
3<html>
4<head>
5 <title>Exception Handling in LLVM</title>
6 <link rel="stylesheet" href="llvm.css" type="text/css">
7</head>
8<body>
9
10<div class="doc_title">Exception Handling in LLVM</div>
11
12<table class="layout" style="width:100%">
13 <tr class="layout">
14 <td class="left">
15<ul>
16 <li><a href="#introduction">Introduction</a>
17 <ol>
18 <li><a href="#itanium">Itanium ABI Zero-cost Exception Handling</a></li>
19 <li><a href="#overview">Overview</a></li>
20 </ol></li>
21 <li><a href="#codegen">LLVM Code Generation</a>
22 <ol>
23 <li><a href="#throw">Throw</a></li>
24 <li><a href="#try_catch">Try/Catch</a></li>
Duncan Sands6590b042007-08-27 15:47:50 +000025 <li><a href="#cleanups">Cleanups</a></li>
Jim Laskeyd0d39b62007-03-14 19:29:42 +000026 <li><a href="#throw_filters">Throw Filters</a></li>
Duncan Sands6590b042007-08-27 15:47:50 +000027 <li><a href="#restrictions">Restrictions</a></li>
Jim Laskeyd0d39b62007-03-14 19:29:42 +000028 </ol></li>
Duncan Sands8036ca42007-03-30 12:22:09 +000029 <li><a href="#format_common_intrinsics">Exception Handling Intrinsics</a>
Jim Laskeyd0d39b62007-03-14 19:29:42 +000030 <ol>
31 <li><a href="#llvm_eh_exception"><tt>llvm.eh.exception</tt></a></li>
32 <li><a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a></li>
Jim Laskeyd0d39b62007-03-14 19:29:42 +000033 <li><a href="#llvm_eh_typeid_for"><tt>llvm.eh.typeid.for</tt></a></li>
34 </ol></li>
35 <li><a href="#asm">Asm Table Formats</a>
36 <ol>
37 <li><a href="#unwind_tables">Exception Handling Frame</a></li>
38 <li><a href="#exception_tables">Exception Tables</a></li>
39 </ol></li>
40 <li><a href="#todo">ToDo</a></li>
41</ul>
42</td>
43</tr></table>
44
45<div class="doc_author">
46 <p>Written by <a href="mailto:jlaskey@mac.com">Jim Laskey</a></p>
47</div>
48
49
50<!-- *********************************************************************** -->
51<div class="doc_section"><a name="introduction">Introduction</a></div>
52<!-- *********************************************************************** -->
53
54<div class="doc_text">
55
56<p>This document is the central repository for all information pertaining to
57exception handling in LLVM. It describes the format that LLVM exception
58handling information takes, which is useful for those interested in creating
59front-ends or dealing directly with the information. Further, this document
60provides specific examples of what exception handling information is used for
61C/C++.</p>
62
63</div>
64
65<!-- ======================================================================= -->
66<div class="doc_subsection">
67 <a name="itanium">Itanium ABI Zero-cost Exception Handling</a>
68</div>
69
70<div class="doc_text">
71
72<p>Exception handling for most programming languages is designed to recover from
73conditions that rarely occur during general use of an application. To that end,
74exception handling should not interfere with the main flow of an
75application&apos;s algorithm by performing checkpointing tasks such as saving
76the current pc or register state.</p>
77
78<p>The Itanium ABI Exception Handling Specification defines a methodology for
79providing outlying data in the form of exception tables without inlining
80speculative exception handling code in the flow of an application&apos;s main
81algorithm. Thus, the specification is said to add "zero-cost" to the normal
82execution of an application.</p>
83
84<p>A more complete description of the Itanium ABI exception handling runtime
85support of can be found at <a
86href="http://www.codesourcery.com/cxx-abi/abi-eh.html">Itanium C++ ABI:
87Exception Handling.</a> A description of the exception frame format can be
88found at <a
89href="http://refspecs.freestandards.org/LSB_3.0.0/LSB-Core-generic/LSB-
90Core-generic/ehframechpt.html">Exception Frames</a>, with details of the Dwarf
91specification at <a href="http://www.eagercon.com/dwarf/dwarf3std.htm">Dwarf 3
92Standard.</a> A description for the C++ exception table formats can be found at
93<a href="http://www.codesourcery.com/cxx-abi/exceptions.pdf">Exception Handling
94Tables.</a></p>
95
96</div>
97
98<!-- ======================================================================= -->
99<div class="doc_subsection">
100 <a name="overview">Overview</a>
101</div>
102
103<div class="doc_text">
104
105<p>When an exception is thrown in llvm code, the runtime does a best effort to
106find a handler suited to process the circumstance.</p>
107
108<p>The runtime first attempts to find an <i>exception frame</i> corresponding to
109the function where the exception was thrown. If the programming language (ex.
110C++) supports exception handling, the exception frame contains a reference to an
111exception table describing how to process the exception. If the language (ex.
112C) does not support exception handling or if the exception needs to be forwarded
113to a prior activation, the exception frame contains information about how to
114unwind the current activation and restore the state of the prior activation.
115This process is repeated until the exception is handled. If the exception is
116not handled and no activations remain, then the application is terminated with
117an appropriate error message.</p>
118
119<p>Since different programming languages have different behaviors when handling
120exceptions, the exception handling ABI provides a mechanism for supplying
121<i>personalities.</i> An exception handling personality is defined by way of a
122<i>personality function</i> (ex. for C++ <tt>__gxx_personality_v0</tt>) which
123receives the context of the exception, an <i>exception structure</i> containing
Duncan Sandsfb0a64a2007-04-16 13:02:27 +0000124the exception object type and value, and a reference to the exception table for
125the current function. The personality function for the current compile unit is
Jim Laskeyd0d39b62007-03-14 19:29:42 +0000126specified in a <i>common exception frame</i>.</p>
127
128<p>The organization of an exception table is language dependent. For C++, an
129exception table is organized as a series of code ranges defining what to do if
130an exception occurs in that range. Typically, the information associated with a
131range defines which types of exception objects (using C++ <i>type info</i>) that
132are handled in that range, and an associated action that should take place.
133Actions typically pass control to a <i>landing pad</i>.</p>
134
135<p>A landing pad corresponds to the code found in the catch portion of a
136try/catch sequence. When execution resumes at a landing pad, it receives the
137exception structure and a selector corresponding to the <i>type</i> of exception
138thrown. The selector is then used to determine which catch should actually
139process the exception.</p>
140
141</div>
142
143<!-- ======================================================================= -->
144<div class="doc_section">
145 <a name="codegen">LLVM Code Generation</a>
146</div>
147
148<div class="doc_text">
149
150<p>At the time of this writing, only C++ exception handling support is available
151in LLVM. So the remainder of this document will be somewhat C++-centric.</p>
152
153<p>From the C++ developers perspective, exceptions are defined in terms of the
154<tt>throw</tt> and <tt>try/catch</tt> statements. In this section we will
155describe the implementation of llvm exception handling in terms of C++
156examples.</p>
157
158</div>
159
160<!-- ======================================================================= -->
161<div class="doc_subsection">
162 <a name="throw">Throw</a>
163</div>
164
165<div class="doc_text">
166
167<p>Languages that support exception handling typically provide a <tt>throw</tt>
168operation to initiate the exception process. Internally, a throw operation
169breaks down into two steps. First, a request is made to allocate exception
170space for an exception structure. This structure needs to survive beyond the
171current activation. This structure will contain the type and value of the
172object being thrown. Second, a call is made to the runtime to raise the
173exception, passing the exception structure as an argument.</p>
174
175<p>In C++, the allocation of the exception structure is done by the
176<tt>__cxa_allocate_exception</tt> runtime function. The exception raising is
177handled by <tt>__cxa_throw</tt>. The type of the exception is represented using
178a C++ RTTI type info structure.</p>
179
180</div>
181
182<!-- ======================================================================= -->
183<div class="doc_subsection">
184 <a name="try_catch">Try/Catch</a>
185</div>
186
187<div class="doc_text">
188
Duncan Sandsb0a1cbf2007-04-14 12:30:27 +0000189<p>A call within the scope of a try statement can potentially raise an exception.
Jim Laskeyd0d39b62007-03-14 19:29:42 +0000190In those circumstances, the LLVM C++ front-end replaces the call with an
191<tt>invoke</tt> instruction. Unlike a call, the invoke has two potential
192continuation points; where to continue when the call succeeds as per normal, and
193where to continue if the call raises an exception, either by a throw or the
194unwinding of a throw.</p>
195
196<p>The term used to define a the place where an invoke continues after an
197exception is called a <i>landing pad</i>. LLVM landing pads are conceptually
Duncan Sandsfb0a64a2007-04-16 13:02:27 +0000198alternative function entry points where a exception structure reference and a type
Jim Laskeyd0d39b62007-03-14 19:29:42 +0000199info index are passed in as arguments. The landing pad saves the exception
200structure reference and then proceeds to select the catch block that corresponds
201to the type info of the exception object.</p>
202
203<p>Two llvm intrinsic functions are used convey information about the landing
204pad to the back end.</p>
205
206<p><a href="#llvm_eh_exception"><tt>llvm.eh.exception</tt></a> takes no
207arguments and returns the exception structure reference. The backend replaces
208this intrinsic with the code that accesses the first argument of a call. The
209LLVM C++ front end generates code to save this value in an alloca location for
210further use in the landing pad and catch code.</p>
211
212<p><a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> takes a minimum of
213three arguments. The first argument is the reference to the exception
214structure. The second argument is a reference to the personality function to be
Duncan Sandscf26d7c2007-07-04 20:52:51 +0000215used for this try catch sequence. Each of the remaining arguments is either a
Duncan Sands6590b042007-08-27 15:47:50 +0000216reference to the type info for a catch statement,
217a <a href="#throw_filters">filter</a> expression,
218or the number zero representing a <a href="#cleanups">cleanup</a>.
Duncan Sandscf26d7c2007-07-04 20:52:51 +0000219The exception is tested against the arguments sequentially from first to last.
Duncan Sands6590b042007-08-27 15:47:50 +0000220The result of the <a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> is a
221positive number if the exception matched a type info, a negative number if it matched
222a filter, and zero if it matched a cleanup. If nothing is matched, the behaviour of
223the program is <a href="#restrictions">undefined</a>.
224The LLVM C++ front end generates code to save the selector value in an alloca
225location for further use in the landing pad and catch code.
226If a type info matched then the selector value is the index of the type info in
227the exception table, which can be obtained using the
228<a href="#llvm_eh_typeid_for"><tt>llvm.eh.typeid.for</tt></a> intrinsic.</p>
Jim Laskeyd0d39b62007-03-14 19:29:42 +0000229
230<p>Once the landing pad has the type info selector, the code branches to the
231code for the first catch. The catch then checks the value of the type info
232selector against the index of type info for that catch. Since the type info
233index is not known until all the type info have been gathered in the backend,
234the catch code will call the <a
235href="#llvm_eh_typeid_for"><tt>llvm.eh.typeid.for</tt></a> intrinsic to
236determine the index for a given type info. If the catch fails to match the
237selector then control is passed on to the next catch. Note: Since the landing
238pad will not be used if there is no match in the list of type info on the call
239to <a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a>, then neither the
240last catch nor <i>catch all</i> need to perform the the check against the
241selector.</p>
242
243<p>Finally, the entry and exit of catch code is bracketed with calls to
244<tt>__cxa_begin_catch</tt> and <tt>__cxa_end_catch</tt>.
245<tt>__cxa_begin_catch</tt> takes a exception structure reference as an argument
246and returns the value of the exception object.</tt> <tt>__cxa_end_catch</tt>
247takes a exception structure reference as an argument. This function clears the
248exception from the exception space. Note: a rethrow from within the catch may
249replace this call with a <tt>__cxa_rethrow</tt>.</p>
250
251</div>
252
253<!-- ======================================================================= -->
254<div class="doc_subsection">
Duncan Sands6590b042007-08-27 15:47:50 +0000255 <a name="cleanups">Cleanups</a>
Jim Laskeyd0d39b62007-03-14 19:29:42 +0000256</div>
257
258<div class="doc_text">
259
260<p>To handle destructors and cleanups in try code, control may not run directly
261from a landing pad to the first catch. Control may actually flow from the
262landing pad to clean up code and then to the first catch. Since the required
263clean up for each invoke in a try may be different (ex., intervening
Duncan Sands6590b042007-08-27 15:47:50 +0000264constructor), there may be several landing pads for a given try. If cleanups
265need to be run, the number zero should be passed as the last
266<a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> argument.
267However for C++ a <tt>null i8*</tt> <a href="#restrictions">must</a> be passed
268instead.
269</p>
Jim Laskeyd0d39b62007-03-14 19:29:42 +0000270
271</div>
272
273<!-- ======================================================================= -->
274<div class="doc_subsection">
275 <a name="throw_filters">Throw Filters</a>
276</div>
277
278<div class="doc_text">
279
280<p>C++ allows the specification of which exception types that can be thrown from
281a function. To represent this a top level landing pad may exist to filter out
282invalid types. To express this in LLVM code the landing pad will call <a
Jim Laskeyd0d39b62007-03-14 19:29:42 +0000283href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a>. The arguments are the
Duncan Sands6590b042007-08-27 15:47:50 +0000284length of the filter expression (the number of type infos plus one), followed by
285the type infos themselves.
Duncan Sandscf26d7c2007-07-04 20:52:51 +0000286<a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> will return a negative
287value if the exception does not match any of the type infos. If no match is
288found then a call to <tt>__cxa_call_unexpected</tt> should be made, otherwise
Jim Laskeyd0d39b62007-03-14 19:29:42 +0000289<tt>_Unwind_Resume</tt>. Each of these functions require a reference to the
290exception structure.</p>
291
292</div>
293
294<!-- ======================================================================= -->
Duncan Sands6590b042007-08-27 15:47:50 +0000295<div class="doc_subsection">
296 <a name="restrictions">Restrictions</a>
297</div>
298
299<div class="doc_text">
300
301<p>The semantics of the invoke instruction require that any exception that
302unwinds through an invoke call should result in a branch to the invoke's unwind
303label. However such a branch will only happen if the
304<a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> matches.
305Thus in order to ensure correct operation, the front-end must only generate
306<a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> calls that are
307guaranteed to always match whatever exception unwinds through the invoke.
308For most languages it is enough to pass zero, indicating the presence of
309a <a href="#cleanups">cleanup</a>, as the last
310<a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> argument.
311However for C++ this is not sufficient, because the C++ personality function
312will terminate the program if it detects that unwinding the exception only
313results in matches with cleanups. For C++ a <tt>null i8*</tt> should
314be passed as the last
315<a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> argument instead.
316This is interpreted as a catch-all by the C++ personality function, and will
317always match.
318</p>
319
320</div>
321
322<!-- ======================================================================= -->
Jim Laskeyd0d39b62007-03-14 19:29:42 +0000323<div class="doc_section">
Duncan Sands8036ca42007-03-30 12:22:09 +0000324 <a name="format_common_intrinsics">Exception Handling Intrinsics</a>
Jim Laskeyd0d39b62007-03-14 19:29:42 +0000325</div>
326
327<div class="doc_text">
328
329<p>LLVM uses several intrinsic functions (name prefixed with "llvm.eh") to
330provide exception handling information at various points in generated code.</p>
331
332</div>
333
334<!-- ======================================================================= -->
335<div class="doc_subsubsection">
336 <a name="llvm_eh_exception">llvm.eh.exception</a>
337</div>
338
339<div class="doc_text">
340<pre>
341 i8* %<a href="#llvm_eh_exception">llvm.eh.exception</a>( )
342</pre>
343
344<p>This intrinsic indicates that the exception structure is available at this
345point in the code. The backend will replace this intrinsic with code to fetch
346the first argument of a call. The effect is that the intrinsic result is the
347exception structure reference.</p>
348
349</div>
350
351<!-- ======================================================================= -->
352<div class="doc_subsubsection">
353 <a name="llvm_eh_selector">llvm.eh.selector</a>
354</div>
355
356<div class="doc_text">
357<pre>
358 i32 %<a href="#llvm_eh_selector">llvm.eh.selector</a>(i8*, i8*, i8*, ...)
359</pre>
360
361<p>This intrinsic indicates that the exception selector is available at this
362point in the code. The backend will replace this intrinsic with code to fetch
363the second argument of a call. The effect is that the intrinsic result is the
364exception selector.</p>
365
366<p><a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> takes a minimum of
367three arguments. The first argument is the reference to the exception
368structure. The second argument is a reference to the personality function to be
Duncan Sands6590b042007-08-27 15:47:50 +0000369used for this try catch sequence. Each of the remaining arguments is either a
370reference to the type info for a catch statement,
371a <a href="#throw_filters">filter</a> expression,
372or the number zero representing a <a href="#cleanups">cleanup</a>.
Duncan Sandscf26d7c2007-07-04 20:52:51 +0000373The exception is tested against the arguments sequentially from first to last.
Duncan Sands6590b042007-08-27 15:47:50 +0000374The result of the <a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> is a
375positive number if the exception matched a type info, a negative number if it matched
376a filter, and zero if it matched a cleanup. If nothing is matched, the behaviour of
377the program is <a href="#restrictions">undefined</a>.
378If a type info matched then the selector value is the index of the type info in
379the exception table, which can be obtained using the
380<a href="#llvm_eh_typeid_for"><tt>llvm.eh.typeid.for</tt></a> intrinsic.</p>
Jim Laskeyd0d39b62007-03-14 19:29:42 +0000381
382</div>
383
384<!-- ======================================================================= -->
385<div class="doc_subsubsection">
386 <a name="llvm_eh_typeid_for">llvm.eh.typeid.for</a>
387</div>
388
389<div class="doc_text">
390<pre>
391 i32 %<a href="#llvm_eh_typeid_for">llvm.eh.typeid.for</a>(i8*)
392</pre>
393
394<p>This intrinsic returns the type info index in the exception table of the
395current function. This value can be used to compare against the result of <a
396href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a>. The single argument is
397a reference to a type info.</p>
398
399</div>
400
401<!-- ======================================================================= -->
402<div class="doc_section">
403 <a name="asm">Asm Table Formats</a>
404</div>
405
406<div class="doc_text">
407
408<p>There are two tables that are used by the exception handling runtime to
409determine which actions should take place when an exception is thrown.</p>
410
411</div>
412
413<!-- ======================================================================= -->
414<div class="doc_subsection">
415 <a name="unwind_tables">Exception Handling Frame</a>
416</div>
417
418<div class="doc_text">
419
420<p>An exception handling frame <tt>eh_frame</tt> is very similar to the unwind
421frame used by dwarf debug info. The frame contains all the information
422necessary to tear down the current frame and restore the state of the prior
423frame. There is an exception handling frame for each function in a compile
424unit, plus a common exception handling frame that defines information common to
425all functions in the unit.</p>
426
427<p>Todo - Table details here.</p>
428
429</div>
430
431<!-- ======================================================================= -->
432<div class="doc_subsection">
433 <a name="exception_tables">Exception Tables</a>
434</div>
435
436<div class="doc_text">
437
438<p>An exception table contains information about what actions to take when an
439exception is thrown in a particular part of a function&apos;s code. There is
440one exception table per function except leaf routines and functions that have
441only calls to non-throwing functions will not need an exception table.</p>
442
443<p>Todo - Table details here.</p>
444
445</div>
446
447<!-- ======================================================================= -->
448<div class="doc_section">
449 <a name="todo">ToDo</a>
450</div>
451
452<div class="doc_text">
453
454<ol>
455
Jim Laskeyd0d39b62007-03-14 19:29:42 +0000456<li><p>Testing/Testing/Testing.</li></p>
457
458</ol>
459
460</div>
461
462<!-- *********************************************************************** -->
463
464<hr>
465<address>
466 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
467 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
468 <a href="http://validator.w3.org/check/referer"><img
469 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!"></a>
470
471 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
472 <a href="http://llvm.org">LLVM Compiler Infrastructure</a><br>
473 Last modified: $Date$
474</address>
475
476</body>
477</html>