1 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
5 <title>Exception Handling in LLVM</title>
6 <link rel="stylesheet" href="llvm.css" type="text/css">
10 <div class="doc_title">Exception Handling in LLVM</div>
12 <table class="layout" style="width:100%">
16 <li><a href="#introduction">Introduction</a>
18 <li><a href="#itanium">Itanium ABI Zero-cost Exception Handling</a></li>
19 <li><a href="#overview">Overview</a></li>
21 <li><a href="#codegen">LLVM Code Generation</a>
23 <li><a href="#throw">Throw</a></li>
24 <li><a href="#try_catch">Try/Catch</a></li>
25 <li><a href="#cleanups">Cleanups</a></li>
26 <li><a href="#throw_filters">Throw Filters</a></li>
27 <li><a href="#restrictions">Restrictions</a></li>
29 <li><a href="#format_common_intrinsics">Exception Handling Intrinsics</a>
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>
33 <li><a href="#llvm_eh_typeid_for"><tt>llvm.eh.typeid.for</tt></a></li>
35 <li><a href="#asm">Asm Table Formats</a>
37 <li><a href="#unwind_tables">Exception Handling Frame</a></li>
38 <li><a href="#exception_tables">Exception Tables</a></li>
40 <li><a href="#todo">ToDo</a></li>
45 <div class="doc_author">
46 <p>Written by <a href="mailto:jlaskey@mac.com">Jim Laskey</a></p>
50 <!-- *********************************************************************** -->
51 <div class="doc_section"><a name="introduction">Introduction</a></div>
52 <!-- *********************************************************************** -->
54 <div class="doc_text">
56 <p>This document is the central repository for all information pertaining to
57 exception handling in LLVM. It describes the format that LLVM exception
58 handling information takes, which is useful for those interested in creating
59 front-ends or dealing directly with the information. Further, this document
60 provides specific examples of what exception handling information is used for
65 <!-- ======================================================================= -->
66 <div class="doc_subsection">
67 <a name="itanium">Itanium ABI Zero-cost Exception Handling</a>
70 <div class="doc_text">
72 <p>Exception handling for most programming languages is designed to recover from
73 conditions that rarely occur during general use of an application. To that end,
74 exception handling should not interfere with the main flow of an
75 application's algorithm by performing checkpointing tasks such as saving
76 the current pc or register state.</p>
78 <p>The Itanium ABI Exception Handling Specification defines a methodology for
79 providing outlying data in the form of exception tables without inlining
80 speculative exception handling code in the flow of an application's main
81 algorithm. Thus, the specification is said to add "zero-cost" to the normal
82 execution of an application.</p>
84 <p>A more complete description of the Itanium ABI exception handling runtime
85 support of can be found at <a
86 href="http://www.codesourcery.com/cxx-abi/abi-eh.html">Itanium C++ ABI:
87 Exception Handling.</a> A description of the exception frame format can be found
88 at <a href="http://refspecs.freestandards.org/LSB_3.0.0/LSB-Core-generic/LSB-
89 Core-generic/ehframechpt.html">Exception Frames</a>, with details of the Dwarf
90 specification at <a href="http://www.eagercon.com/dwarf/dwarf3std.htm">Dwarf 3
91 Standard.</a> A description for the C++ exception table formats can be found at
92 <a href="http://www.codesourcery.com/cxx-abi/exceptions.pdf">Exception Handling
97 <!-- ======================================================================= -->
98 <div class="doc_subsection">
99 <a name="overview">Overview</a>
102 <div class="doc_text">
104 <p>When an exception is thrown in llvm code, the runtime does a best effort to
105 find a handler suited to process the circumstance.</p>
107 <p>The runtime first attempts to find an <i>exception frame</i> corresponding to
108 the function where the exception was thrown. If the programming language (ex.
109 C++) supports exception handling, the exception frame contains a reference to an
110 exception table describing how to process the exception. If the language (ex.
111 C) does not support exception handling or if the exception needs to be forwarded
112 to a prior activation, the exception frame contains information about how to
113 unwind the current activation and restore the state of the prior activation.
114 This process is repeated until the exception is handled. If the exception is
115 not handled and no activations remain, then the application is terminated with
116 an appropriate error message.</p>
118 <p>Since different programming languages have different behaviors when handling
119 exceptions, the exception handling ABI provides a mechanism for supplying
120 <i>personalities.</i> An exception handling personality is defined by way of a
121 <i>personality function</i> (ex. for C++ <tt>__gxx_personality_v0</tt>) which
122 receives the context of the exception, an <i>exception structure</i> containing
123 the exception object type and value, and a reference to the exception table for
124 the current function. The personality function for the current compile unit is
125 specified in a <i>common exception frame</i>.</p>
127 <p>The organization of an exception table is language dependent. For C++, an
128 exception table is organized as a series of code ranges defining what to do if
129 an exception occurs in that range. Typically, the information associated with a
130 range defines which types of exception objects (using C++ <i>type info</i>) that
131 are handled in that range, and an associated action that should take place.
132 Actions typically pass control to a <i>landing pad</i>.</p>
134 <p>A landing pad corresponds to the code found in the catch portion of a
135 try/catch sequence. When execution resumes at a landing pad, it receives the
136 exception structure and a selector corresponding to the <i>type</i> of exception
137 thrown. The selector is then used to determine which catch should actually
138 process the exception.</p>
142 <!-- ======================================================================= -->
143 <div class="doc_section">
144 <a name="codegen">LLVM Code Generation</a>
147 <div class="doc_text">
149 <p>At the time of this writing, only C++ exception handling support is available
150 in LLVM. So the remainder of this document will be somewhat C++-centric.</p>
152 <p>From the C++ developers perspective, exceptions are defined in terms of the
153 <tt>throw</tt> and <tt>try/catch</tt> statements. In this section we will
154 describe the implementation of llvm exception handling in terms of C++
159 <!-- ======================================================================= -->
160 <div class="doc_subsection">
161 <a name="throw">Throw</a>
164 <div class="doc_text">
166 <p>Languages that support exception handling typically provide a <tt>throw</tt>
167 operation to initiate the exception process. Internally, a throw operation
168 breaks down into two steps. First, a request is made to allocate exception
169 space for an exception structure. This structure needs to survive beyond the
170 current activation. This structure will contain the type and value of the
171 object being thrown. Second, a call is made to the runtime to raise the
172 exception, passing the exception structure as an argument.</p>
174 <p>In C++, the allocation of the exception structure is done by the
175 <tt>__cxa_allocate_exception</tt> runtime function. The exception raising is
176 handled by <tt>__cxa_throw</tt>. The type of the exception is represented using
177 a C++ RTTI type info structure.</p>
181 <!-- ======================================================================= -->
182 <div class="doc_subsection">
183 <a name="try_catch">Try/Catch</a>
186 <div class="doc_text">
188 <p>A call within the scope of a try statement can potentially raise an exception.
189 In those circumstances, the LLVM C++ front-end replaces the call with an
190 <tt>invoke</tt> instruction. Unlike a call, the invoke has two potential
191 continuation points; where to continue when the call succeeds as per normal, and
192 where to continue if the call raises an exception, either by a throw or the
193 unwinding of a throw.</p>
195 <p>The term used to define a the place where an invoke continues after an
196 exception is called a <i>landing pad</i>. LLVM landing pads are conceptually
197 alternative function entry points where a exception structure reference and a type
198 info index are passed in as arguments. The landing pad saves the exception
199 structure reference and then proceeds to select the catch block that corresponds
200 to the type info of the exception object.</p>
202 <p>Two llvm intrinsic functions are used convey information about the landing
203 pad to the back end.</p>
205 <p><a href="#llvm_eh_exception"><tt>llvm.eh.exception</tt></a> takes no
206 arguments and returns the exception structure reference. The backend replaces
207 this intrinsic with the code that accesses the first argument of a call. The
208 LLVM C++ front end generates code to save this value in an alloca location for
209 further use in the landing pad and catch code.</p>
211 <p><a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> takes a minimum of
212 three arguments. The first argument is the reference to the exception
213 structure. The second argument is a reference to the personality function to be
214 used for this try catch sequence. Each of the remaining arguments is either a
215 reference to the type info for a catch statement,
216 a <a href="#throw_filters">filter</a> expression,
217 or the number zero representing a <a href="#cleanups">cleanup</a>.
218 The exception is tested against the arguments sequentially from first to last.
219 The result of the <a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> is a
220 positive number if the exception matched a type info, a negative number if it matched
221 a filter, and zero if it matched a cleanup. If nothing is matched, the behaviour of
222 the program is <a href="#restrictions">undefined</a>.
223 The LLVM C++ front end generates code to save the selector value in an alloca
224 location for further use in the landing pad and catch code.
225 If a type info matched then the selector value is the index of the type info in
226 the exception table, which can be obtained using the
227 <a href="#llvm_eh_typeid_for"><tt>llvm.eh.typeid.for</tt></a> intrinsic.</p>
229 <p>Once the landing pad has the type info selector, the code branches to the
230 code for the first catch. The catch then checks the value of the type info
231 selector against the index of type info for that catch. Since the type info
232 index is not known until all the type info have been gathered in the backend,
233 the catch code will call the <a
234 href="#llvm_eh_typeid_for"><tt>llvm.eh.typeid.for</tt></a> intrinsic to
235 determine the index for a given type info. If the catch fails to match the
236 selector then control is passed on to the next catch. Note: Since the landing
237 pad will not be used if there is no match in the list of type info on the call
238 to <a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a>, then neither the
239 last catch nor <i>catch all</i> need to perform the the check against the
242 <p>Finally, the entry and exit of catch code is bracketed with calls to
243 <tt>__cxa_begin_catch</tt> and <tt>__cxa_end_catch</tt>.
244 <tt>__cxa_begin_catch</tt> takes a exception structure reference as an argument
245 and returns the value of the exception object. <tt>__cxa_end_catch</tt>
246 takes a exception structure reference as an argument. This function clears the
247 exception from the exception space. Note: a rethrow from within the catch may
248 replace this call with a <tt>__cxa_rethrow</tt>.</p>
252 <!-- ======================================================================= -->
253 <div class="doc_subsection">
254 <a name="cleanups">Cleanups</a>
257 <div class="doc_text">
259 <p>To handle destructors and cleanups in try code, control may not run directly
260 from a landing pad to the first catch. Control may actually flow from the
261 landing pad to clean up code and then to the first catch. Since the required
262 clean up for each invoke in a try may be different (ex., intervening
263 constructor), there may be several landing pads for a given try. If cleanups
264 need to be run, the number zero should be passed as the last
265 <a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> argument.
266 However for C++ a <tt>null i8*</tt> <a href="#restrictions">must</a> be passed
272 <!-- ======================================================================= -->
273 <div class="doc_subsection">
274 <a name="throw_filters">Throw Filters</a>
277 <div class="doc_text">
279 <p>C++ allows the specification of which exception types that can be thrown from
280 a function. To represent this a top level landing pad may exist to filter out
281 invalid types. To express this in LLVM code the landing pad will call <a
282 href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a>. The arguments are the
283 length of the filter expression (the number of type infos plus one), followed by
284 the type infos themselves.
285 <a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> will return a negative
286 value if the exception does not match any of the type infos. If no match is
287 found then a call to <tt>__cxa_call_unexpected</tt> should be made, otherwise
288 <tt>_Unwind_Resume</tt>. Each of these functions require a reference to the
289 exception structure.</p>
293 <!-- ======================================================================= -->
294 <div class="doc_subsection">
295 <a name="restrictions">Restrictions</a>
298 <div class="doc_text">
300 <p>The semantics of the invoke instruction require that any exception that
301 unwinds through an invoke call should result in a branch to the invoke's unwind
302 label. However such a branch will only happen if the
303 <a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> matches.
304 Thus in order to ensure correct operation, the front-end must only generate
305 <a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> calls that are
306 guaranteed to always match whatever exception unwinds through the invoke.
307 For most languages it is enough to pass zero, indicating the presence of
308 a <a href="#cleanups">cleanup</a>, as the last
309 <a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> argument.
310 However for C++ this is not sufficient, because the C++ personality function
311 will terminate the program if it detects that unwinding the exception only
312 results in matches with cleanups. For C++ a <tt>null i8*</tt> should
313 be passed as the last
314 <a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> argument instead.
315 This is interpreted as a catch-all by the C++ personality function, and will
321 <!-- ======================================================================= -->
322 <div class="doc_section">
323 <a name="format_common_intrinsics">Exception Handling Intrinsics</a>
326 <div class="doc_text">
328 <p>LLVM uses several intrinsic functions (name prefixed with "llvm.eh") to
329 provide exception handling information at various points in generated code.</p>
333 <!-- ======================================================================= -->
334 <div class="doc_subsubsection">
335 <a name="llvm_eh_exception">llvm.eh.exception</a>
338 <div class="doc_text">
340 i8* %<a href="#llvm_eh_exception">llvm.eh.exception</a>( )
343 <p>This intrinsic indicates that the exception structure is available at this
344 point in the code. The backend will replace this intrinsic with code to fetch
345 the first argument of a call. The effect is that the intrinsic result is the
346 exception structure reference.</p>
350 <!-- ======================================================================= -->
351 <div class="doc_subsubsection">
352 <a name="llvm_eh_selector">llvm.eh.selector</a>
355 <div class="doc_text">
357 i32 %<a href="#llvm_eh_selector">llvm.eh.selector.i32</a>(i8*, i8*, i8*, ...)
358 i64 %<a href="#llvm_eh_selector">llvm.eh.selector.i64</a>(i8*, i8*, i8*, ...)
361 <p>This intrinsic indicates that the exception selector is available at this
362 point in the code. The backend will replace this intrinsic with code to fetch
363 the second argument of a call. The effect is that the intrinsic result is the
364 exception selector.</p>
366 <p><a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> takes a minimum of
367 three arguments. The first argument is the reference to the exception
368 structure. The second argument is a reference to the personality function to be
369 used for this try catch sequence. Each of the remaining arguments is either a
370 reference to the type info for a catch statement,
371 a <a href="#throw_filters">filter</a> expression,
372 or the number zero representing a <a href="#cleanups">cleanup</a>.
373 The exception is tested against the arguments sequentially from first to last.
374 The result of the <a href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a> is a
375 positive number if the exception matched a type info, a negative number if it matched
376 a filter, and zero if it matched a cleanup. If nothing is matched, the behaviour of
377 the program is <a href="#restrictions">undefined</a>.
378 If a type info matched then the selector value is the index of the type info in
379 the exception table, which can be obtained using the
380 <a href="#llvm_eh_typeid_for"><tt>llvm.eh.typeid.for</tt></a> intrinsic.</p>
384 <!-- ======================================================================= -->
385 <div class="doc_subsubsection">
386 <a name="llvm_eh_typeid_for">llvm.eh.typeid.for</a>
389 <div class="doc_text">
391 i32 %<a href="#llvm_eh_typeid_for">llvm.eh.typeid.for.i32</a>(i8*)
392 i64 %<a href="#llvm_eh_typeid_for">llvm.eh.typeid.for.i64</a>(i8*)
395 <p>This intrinsic returns the type info index in the exception table of the
396 current function. This value can be used to compare against the result of <a
397 href="#llvm_eh_selector"><tt>llvm.eh.selector</tt></a>. The single argument is
398 a reference to a type info.</p>
402 <!-- ======================================================================= -->
403 <div class="doc_section">
404 <a name="asm">Asm Table Formats</a>
407 <div class="doc_text">
409 <p>There are two tables that are used by the exception handling runtime to
410 determine which actions should take place when an exception is thrown.</p>
414 <!-- ======================================================================= -->
415 <div class="doc_subsection">
416 <a name="unwind_tables">Exception Handling Frame</a>
419 <div class="doc_text">
421 <p>An exception handling frame <tt>eh_frame</tt> is very similar to the unwind
422 frame used by dwarf debug info. The frame contains all the information
423 necessary to tear down the current frame and restore the state of the prior
424 frame. There is an exception handling frame for each function in a compile
425 unit, plus a common exception handling frame that defines information common to
426 all functions in the unit.</p>
428 <p>Todo - Table details here.</p>
432 <!-- ======================================================================= -->
433 <div class="doc_subsection">
434 <a name="exception_tables">Exception Tables</a>
437 <div class="doc_text">
439 <p>An exception table contains information about what actions to take when an
440 exception is thrown in a particular part of a function's code. There is
441 one exception table per function except leaf routines and functions that have
442 only calls to non-throwing functions will not need an exception table.</p>
444 <p>Todo - Table details here.</p>
448 <!-- ======================================================================= -->
449 <div class="doc_section">
450 <a name="todo">ToDo</a>
453 <div class="doc_text">
457 <li><p>Testing/Testing/Testing.</p></li>
463 <!-- *********************************************************************** -->
467 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
468 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
469 <a href="http://validator.w3.org/check/referer"><img
470 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!"></a>
472 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
473 <a href="http://llvm.org">LLVM Compiler Infrastructure</a><br>
474 Last modified: $Date$