gecko/js/src/jsutil.h

821 lines
28 KiB
C
Raw Normal View History

/* -*- Mode: C; tab-width: 8; indent-tabs-mode: nil; c-basic-offset: 4 -*-
*
* ***** BEGIN LICENSE BLOCK *****
* Version: MPL 1.1/GPL 2.0/LGPL 2.1
*
* The contents of this file are subject to the Mozilla Public License Version
* 1.1 (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
* http://www.mozilla.org/MPL/
*
* Software distributed under the License is distributed on an "AS IS" basis,
* WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License
* for the specific language governing rights and limitations under the
* License.
*
* The Original Code is Mozilla Communicator client code, released
* March 31, 1998.
*
* The Initial Developer of the Original Code is
* Netscape Communications Corporation.
* Portions created by the Initial Developer are Copyright (C) 1998
* the Initial Developer. All Rights Reserved.
*
* Contributor(s):
*
* Alternatively, the contents of this file may be used under the terms of
* either of the GNU General Public License Version 2 or later (the "GPL"),
* or the GNU Lesser General Public License Version 2.1 or later (the "LGPL"),
* in which case the provisions of the GPL or the LGPL are applicable instead
* of those above. If you wish to allow use of your version of this file only
* under the terms of either the GPL or the LGPL, and not to allow others to
* use your version of this file under the terms of the MPL, indicate your
* decision by deleting the provisions above and replace them with the notice
* and other provisions required by the GPL or the LGPL. If you do not delete
* the provisions above, a recipient may use your version of this file under
* the terms of any one of the MPL, the GPL or the LGPL.
*
* ***** END LICENSE BLOCK ***** */
/*
* PR assertion checker.
*/
#ifndef jsutil_h___
#define jsutil_h___
#include "jstypes.h"
#include "mozilla/Util.h"
#include <stdlib.h>
#include <string.h>
JS_BEGIN_EXTERN_C
#define JS_CRASH_UNLESS(__cond) \
JS_BEGIN_MACRO \
if (!(__cond)) { \
*(int *)(uintptr_t)0xccadbeef = 0; \
((void(*)())0)(); /* More reliable, but doesn't say CCADBEEF */ \
} \
JS_END_MACRO
#define JS_FREE_PATTERN 0xDA
#ifdef JS_CRASH_DIAGNOSTICS
#define JS_POISON(p, val, size) memset((p), (val), (size))
#define JS_OPT_ASSERT(expr) \
((expr) ? (void)0 : JS_Assert(#expr, __FILE__, __LINE__))
#define JS_OPT_ASSERT_IF(cond, expr) \
((!(cond) || (expr)) ? (void)0 : JS_Assert(#expr, __FILE__, __LINE__))
#else
#define JS_POISON(p, val, size) ((void) 0)
#define JS_OPT_ASSERT(expr) ((void) 0)
#define JS_OPT_ASSERT_IF(cond, expr) ((void) 0)
#endif /* JS_CRASH_DIAGNOSTICS */
#ifdef DEBUG
* Menu of -D flags for enabling instrumentation, as a commented-out CFLAGS += setting for convenient testing. * js_FindProperty and js_LookupPropertyWithFlags return indexes into the scope and prototype chains, respectively, to support internal instrumentation, and to pave the way for the return of the property cache (bug 365851).. * jsutil.[ch] JSBasicStats struct and functions for computing mean/sigma/max and auto-scaling histogram. * JS_SCOPE_DEPTH_METER instrumentation for compile- and run-time scope chain length instrumentation: + At compile time, rt->hostenvScopeDepthStats and rt->lexicalScopeDepthStats meter scope chains passed into the compile and evaluate APIs. + At runtime, rt->protoLookupDepthStats and rt->scopeSearchDepthStats track steps along the prototype and scope chains until the sought-after property is found. * JS_ARENAMETER uses JSBasicStats now. * Added rt->liveScopePropsPreSweep to fix the property tree stats code that rotted when property tree sweeping moved to after the finalization phase. * Un-bitrotted some DEBUG_brendan code, turned some off for myself via XXX. * Mac OS X toolchain requires initialized data shared across dynamic library member files, outlaws common data, so initialize extern metering vars. * Old HASHMETER code in jshash.[ch] is now JS_HASHMETER-controlled and based on JSBasicStats. * DEBUG_scopemeters macro renamed JS_DUMP_SCOPE_METERS; uses JSBasicStats now. * Disentangle DEBUG and DUMP_SCOPE_STATS (now JS_DUMP_PROPTREE_STATS) and fix inconsistent thread safety for liveScopeProps (sometimes atomic-incremented, sometimes runtime-locked). * Compiler-modeled maxScopeDepth will propagate via JSScript to runtime for capability-based, interpreter-inlined cache hit qualifier bits, to bypass scope and prototype chain lookup by optimizing for common monomorphic get, set, and call site referencing a prototype property in a well-named object (no shadowing or mutation in 99.9% of the cases).
2008-01-12 16:31:31 -08:00
#define JS_ASSERT(expr) \
((expr) ? (void)0 : JS_Assert(#expr, __FILE__, __LINE__))
* Menu of -D flags for enabling instrumentation, as a commented-out CFLAGS += setting for convenient testing. * js_FindProperty and js_LookupPropertyWithFlags return indexes into the scope and prototype chains, respectively, to support internal instrumentation, and to pave the way for the return of the property cache (bug 365851).. * jsutil.[ch] JSBasicStats struct and functions for computing mean/sigma/max and auto-scaling histogram. * JS_SCOPE_DEPTH_METER instrumentation for compile- and run-time scope chain length instrumentation: + At compile time, rt->hostenvScopeDepthStats and rt->lexicalScopeDepthStats meter scope chains passed into the compile and evaluate APIs. + At runtime, rt->protoLookupDepthStats and rt->scopeSearchDepthStats track steps along the prototype and scope chains until the sought-after property is found. * JS_ARENAMETER uses JSBasicStats now. * Added rt->liveScopePropsPreSweep to fix the property tree stats code that rotted when property tree sweeping moved to after the finalization phase. * Un-bitrotted some DEBUG_brendan code, turned some off for myself via XXX. * Mac OS X toolchain requires initialized data shared across dynamic library member files, outlaws common data, so initialize extern metering vars. * Old HASHMETER code in jshash.[ch] is now JS_HASHMETER-controlled and based on JSBasicStats. * DEBUG_scopemeters macro renamed JS_DUMP_SCOPE_METERS; uses JSBasicStats now. * Disentangle DEBUG and DUMP_SCOPE_STATS (now JS_DUMP_PROPTREE_STATS) and fix inconsistent thread safety for liveScopeProps (sometimes atomic-incremented, sometimes runtime-locked). * Compiler-modeled maxScopeDepth will propagate via JSScript to runtime for capability-based, interpreter-inlined cache hit qualifier bits, to bypass scope and prototype chain lookup by optimizing for common monomorphic get, set, and call site referencing a prototype property in a well-named object (no shadowing or mutation in 99.9% of the cases).
2008-01-12 16:31:31 -08:00
#define JS_ASSERT_IF(cond, expr) \
((!(cond) || (expr)) ? (void)0 : JS_Assert(#expr, __FILE__, __LINE__))
#define JS_NOT_REACHED(reason) \
JS_Assert(reason, __FILE__, __LINE__)
#define JS_ALWAYS_TRUE(expr) JS_ASSERT(expr)
#define JS_ALWAYS_FALSE(expr) JS_ASSERT(!(expr))
# ifdef JS_THREADSAFE
# define JS_THREADSAFE_ASSERT(expr) JS_ASSERT(expr)
# else
# define JS_THREADSAFE_ASSERT(expr) ((void) 0)
# endif
#else
* Menu of -D flags for enabling instrumentation, as a commented-out CFLAGS += setting for convenient testing. * js_FindProperty and js_LookupPropertyWithFlags return indexes into the scope and prototype chains, respectively, to support internal instrumentation, and to pave the way for the return of the property cache (bug 365851).. * jsutil.[ch] JSBasicStats struct and functions for computing mean/sigma/max and auto-scaling histogram. * JS_SCOPE_DEPTH_METER instrumentation for compile- and run-time scope chain length instrumentation: + At compile time, rt->hostenvScopeDepthStats and rt->lexicalScopeDepthStats meter scope chains passed into the compile and evaluate APIs. + At runtime, rt->protoLookupDepthStats and rt->scopeSearchDepthStats track steps along the prototype and scope chains until the sought-after property is found. * JS_ARENAMETER uses JSBasicStats now. * Added rt->liveScopePropsPreSweep to fix the property tree stats code that rotted when property tree sweeping moved to after the finalization phase. * Un-bitrotted some DEBUG_brendan code, turned some off for myself via XXX. * Mac OS X toolchain requires initialized data shared across dynamic library member files, outlaws common data, so initialize extern metering vars. * Old HASHMETER code in jshash.[ch] is now JS_HASHMETER-controlled and based on JSBasicStats. * DEBUG_scopemeters macro renamed JS_DUMP_SCOPE_METERS; uses JSBasicStats now. * Disentangle DEBUG and DUMP_SCOPE_STATS (now JS_DUMP_PROPTREE_STATS) and fix inconsistent thread safety for liveScopeProps (sometimes atomic-incremented, sometimes runtime-locked). * Compiler-modeled maxScopeDepth will propagate via JSScript to runtime for capability-based, interpreter-inlined cache hit qualifier bits, to bypass scope and prototype chain lookup by optimizing for common monomorphic get, set, and call site referencing a prototype property in a well-named object (no shadowing or mutation in 99.9% of the cases).
2008-01-12 16:31:31 -08:00
#define JS_ASSERT(expr) ((void) 0)
#define JS_ASSERT_IF(cond,expr) ((void) 0)
* Menu of -D flags for enabling instrumentation, as a commented-out CFLAGS += setting for convenient testing. * js_FindProperty and js_LookupPropertyWithFlags return indexes into the scope and prototype chains, respectively, to support internal instrumentation, and to pave the way for the return of the property cache (bug 365851).. * jsutil.[ch] JSBasicStats struct and functions for computing mean/sigma/max and auto-scaling histogram. * JS_SCOPE_DEPTH_METER instrumentation for compile- and run-time scope chain length instrumentation: + At compile time, rt->hostenvScopeDepthStats and rt->lexicalScopeDepthStats meter scope chains passed into the compile and evaluate APIs. + At runtime, rt->protoLookupDepthStats and rt->scopeSearchDepthStats track steps along the prototype and scope chains until the sought-after property is found. * JS_ARENAMETER uses JSBasicStats now. * Added rt->liveScopePropsPreSweep to fix the property tree stats code that rotted when property tree sweeping moved to after the finalization phase. * Un-bitrotted some DEBUG_brendan code, turned some off for myself via XXX. * Mac OS X toolchain requires initialized data shared across dynamic library member files, outlaws common data, so initialize extern metering vars. * Old HASHMETER code in jshash.[ch] is now JS_HASHMETER-controlled and based on JSBasicStats. * DEBUG_scopemeters macro renamed JS_DUMP_SCOPE_METERS; uses JSBasicStats now. * Disentangle DEBUG and DUMP_SCOPE_STATS (now JS_DUMP_PROPTREE_STATS) and fix inconsistent thread safety for liveScopeProps (sometimes atomic-incremented, sometimes runtime-locked). * Compiler-modeled maxScopeDepth will propagate via JSScript to runtime for capability-based, interpreter-inlined cache hit qualifier bits, to bypass scope and prototype chain lookup by optimizing for common monomorphic get, set, and call site referencing a prototype property in a well-named object (no shadowing or mutation in 99.9% of the cases).
2008-01-12 16:31:31 -08:00
#define JS_NOT_REACHED(reason)
#define JS_ALWAYS_TRUE(expr) ((void) (expr))
#define JS_ALWAYS_FALSE(expr) ((void) (expr))
#define JS_THREADSAFE_ASSERT(expr) ((void) 0)
#endif /* defined(DEBUG) */
/*
2009-02-19 15:15:32 -08:00
* Compile-time assert. "cond" must be a constant expression.
* The macro can be used only in places where an "extern" declaration is
* allowed.
*/
#ifdef __SUNPRO_CC
/*
* Sun Studio C++ compiler has a bug
* "sizeof expression not accepted as size of array parameter"
* It happens when js_static_assert() function is declared inside functions.
* The bug number is 6688515. It is not public yet.
* Therefore, for Sun Studio, declare js_static_assert as an array instead.
*/
#define JS_STATIC_ASSERT(cond) extern char js_static_assert[(cond) ? 1 : -1]
#else
#ifdef __COUNTER__
#define JS_STATIC_ASSERT_GLUE1(x,y) x##y
#define JS_STATIC_ASSERT_GLUE(x,y) JS_STATIC_ASSERT_GLUE1(x,y)
#define JS_STATIC_ASSERT(cond) \
typedef int JS_STATIC_ASSERT_GLUE(js_static_assert, __COUNTER__)[(cond) ? 1 : -1]
#else
2009-02-19 15:15:32 -08:00
#define JS_STATIC_ASSERT(cond) extern void js_static_assert(int arg[(cond) ? 1 : -1])
#endif
#endif
#define JS_STATIC_ASSERT_IF(cond, expr) JS_STATIC_ASSERT(!(cond) || (expr))
/*
* Menu of -D flags for enabling instrumentation, as a commented-out CFLAGS += setting for convenient testing. * js_FindProperty and js_LookupPropertyWithFlags return indexes into the scope and prototype chains, respectively, to support internal instrumentation, and to pave the way for the return of the property cache (bug 365851).. * jsutil.[ch] JSBasicStats struct and functions for computing mean/sigma/max and auto-scaling histogram. * JS_SCOPE_DEPTH_METER instrumentation for compile- and run-time scope chain length instrumentation: + At compile time, rt->hostenvScopeDepthStats and rt->lexicalScopeDepthStats meter scope chains passed into the compile and evaluate APIs. + At runtime, rt->protoLookupDepthStats and rt->scopeSearchDepthStats track steps along the prototype and scope chains until the sought-after property is found. * JS_ARENAMETER uses JSBasicStats now. * Added rt->liveScopePropsPreSweep to fix the property tree stats code that rotted when property tree sweeping moved to after the finalization phase. * Un-bitrotted some DEBUG_brendan code, turned some off for myself via XXX. * Mac OS X toolchain requires initialized data shared across dynamic library member files, outlaws common data, so initialize extern metering vars. * Old HASHMETER code in jshash.[ch] is now JS_HASHMETER-controlled and based on JSBasicStats. * DEBUG_scopemeters macro renamed JS_DUMP_SCOPE_METERS; uses JSBasicStats now. * Disentangle DEBUG and DUMP_SCOPE_STATS (now JS_DUMP_PROPTREE_STATS) and fix inconsistent thread safety for liveScopeProps (sometimes atomic-incremented, sometimes runtime-locked). * Compiler-modeled maxScopeDepth will propagate via JSScript to runtime for capability-based, interpreter-inlined cache hit qualifier bits, to bypass scope and prototype chain lookup by optimizing for common monomorphic get, set, and call site referencing a prototype property in a well-named object (no shadowing or mutation in 99.9% of the cases).
2008-01-12 16:31:31 -08:00
* Abort the process in a non-graceful manner. This will cause a core file,
* call to the debugger or other moral equivalent as well as causing the
* entire process to stop.
*/
extern JS_PUBLIC_API(void) JS_Abort(void);
#ifdef DEBUG
* Menu of -D flags for enabling instrumentation, as a commented-out CFLAGS += setting for convenient testing. * js_FindProperty and js_LookupPropertyWithFlags return indexes into the scope and prototype chains, respectively, to support internal instrumentation, and to pave the way for the return of the property cache (bug 365851).. * jsutil.[ch] JSBasicStats struct and functions for computing mean/sigma/max and auto-scaling histogram. * JS_SCOPE_DEPTH_METER instrumentation for compile- and run-time scope chain length instrumentation: + At compile time, rt->hostenvScopeDepthStats and rt->lexicalScopeDepthStats meter scope chains passed into the compile and evaluate APIs. + At runtime, rt->protoLookupDepthStats and rt->scopeSearchDepthStats track steps along the prototype and scope chains until the sought-after property is found. * JS_ARENAMETER uses JSBasicStats now. * Added rt->liveScopePropsPreSweep to fix the property tree stats code that rotted when property tree sweeping moved to after the finalization phase. * Un-bitrotted some DEBUG_brendan code, turned some off for myself via XXX. * Mac OS X toolchain requires initialized data shared across dynamic library member files, outlaws common data, so initialize extern metering vars. * Old HASHMETER code in jshash.[ch] is now JS_HASHMETER-controlled and based on JSBasicStats. * DEBUG_scopemeters macro renamed JS_DUMP_SCOPE_METERS; uses JSBasicStats now. * Disentangle DEBUG and DUMP_SCOPE_STATS (now JS_DUMP_PROPTREE_STATS) and fix inconsistent thread safety for liveScopeProps (sometimes atomic-incremented, sometimes runtime-locked). * Compiler-modeled maxScopeDepth will propagate via JSScript to runtime for capability-based, interpreter-inlined cache hit qualifier bits, to bypass scope and prototype chain lookup by optimizing for common monomorphic get, set, and call site referencing a prototype property in a well-named object (no shadowing or mutation in 99.9% of the cases).
2008-01-12 16:31:31 -08:00
# define JS_BASIC_STATS 1
#endif
#ifdef JS_BASIC_STATS
#include <stdio.h>
typedef struct JSBasicStats {
uint32 num;
uint32 max;
double sum;
double sqsum;
uint32 logscale; /* logarithmic scale: 0 (linear), 2, 10 */
uint32 hist[11];
} JSBasicStats;
#define JS_INIT_STATIC_BASIC_STATS {0,0,0,0,0,{0,0,0,0,0,0,0,0,0,0,0}}
#define JS_BASIC_STATS_INIT(bs) memset((bs), 0, sizeof(JSBasicStats))
#define JS_BASIC_STATS_ACCUM(bs,val) \
JS_BasicStatsAccum(bs, val)
#define JS_MeanAndStdDevBS(bs,sigma) \
JS_MeanAndStdDev((bs)->num, (bs)->sum, (bs)->sqsum, sigma)
extern void
JS_BasicStatsAccum(JSBasicStats *bs, uint32 val);
extern double
JS_MeanAndStdDev(uint32 num, double sum, double sqsum, double *sigma);
extern void
JS_DumpBasicStats(JSBasicStats *bs, const char *title, FILE *fp);
extern void
JS_DumpHistogram(JSBasicStats *bs, FILE *fp);
#else
#define JS_BASIC_STATS_ACCUM(bs,val) /* nothing */
#endif /* JS_BASIC_STATS */
#if defined(DEBUG_notme) && defined(XP_UNIX)
typedef struct JSCallsite JSCallsite;
struct JSCallsite {
uint32 pc;
char *name;
const char *library;
int offset;
JSCallsite *parent;
JSCallsite *siblings;
JSCallsite *kids;
void *handy;
};
extern JS_FRIEND_API(JSCallsite *)
JS_Backtrace(int skip);
extern JS_FRIEND_API(void)
JS_DumpBacktrace(JSCallsite *trace);
#endif
#if defined JS_USE_CUSTOM_ALLOCATOR
#include "jscustomallocator.h"
#else
#ifdef DEBUG
/*
* In order to test OOM conditions, when the shell command-line option
* |-A NUM| is passed, we fail continuously after the NUM'th allocation.
*/
extern JS_PUBLIC_DATA(JSUint32) OOM_maxAllocations; /* set from shell/js.cpp */
extern JS_PUBLIC_DATA(JSUint32) OOM_counter; /* data race, who cares. */
#define JS_OOM_POSSIBLY_FAIL() \
do \
{ \
if (OOM_counter++ >= OOM_maxAllocations) { \
return NULL; \
} \
} while (0)
#else
#define JS_OOM_POSSIBLY_FAIL() do {} while(0)
#endif
/*
* SpiderMonkey code should not be calling these allocation functions directly.
* Instead, all calls should go through JSRuntime, JSContext or OffTheBooks.
* However, js_free() can be called directly.
*/
static JS_INLINE void* js_malloc(size_t bytes) {
JS_OOM_POSSIBLY_FAIL();
return malloc(bytes);
}
static JS_INLINE void* js_calloc(size_t bytes) {
JS_OOM_POSSIBLY_FAIL();
return calloc(bytes, 1);
}
static JS_INLINE void* js_realloc(void* p, size_t bytes) {
JS_OOM_POSSIBLY_FAIL();
return realloc(p, bytes);
}
static JS_INLINE void js_free(void* p) {
free(p);
}
#endif/* JS_USE_CUSTOM_ALLOCATOR */
JS_END_EXTERN_C
#ifdef __cplusplus
/*
* User guide to memory management within SpiderMonkey:
*
* Quick tips:
*
* Allocation:
* - Prefer to allocate using JSContext:
* cx->{malloc_,realloc_,calloc_,new_,new_array}
*
* - If no JSContext is available, use a JSRuntime:
* rt->{malloc_,realloc_,calloc_,new_,new_array}
*
* - As a last resort, use unaccounted allocation ("OffTheBooks"):
* js::OffTheBooks::{malloc_,realloc_,calloc_,new_,new_array}
*
* Deallocation:
* - When the deallocation occurs on a slow path, use:
* Foreground::{free_,delete_,array_delete}
*
* - Otherwise deallocate on a background thread using a JSContext:
* cx->{free_,delete_,array_delete}
*
* - If no JSContext is available, use a JSRuntime:
* rt->{free_,delete_,array_delete}
*
* - As a last resort, use UnwantedForeground deallocation:
* js::UnwantedForeground::{free_,delete_,array_delete}
*
* General tips:
*
* - Mixing and matching these allocators is allowed (you may free memory
* allocated by any allocator, with any deallocator).
*
* - Never, ever use normal C/C++ memory management:
* malloc, free, new, new[], delete, operator new, etc.
*
* - Never, ever use low-level SpiderMonkey allocators:
* js_malloc(), js_free(), js_calloc(), js_realloc()
* Their use is reserved for the other memory managers.
*
* - Classes which have private constructors or destructors should have
* JS_DECLARE_ALLOCATION_FRIENDS_FOR_PRIVATE_CONSTRUCTOR added to their
* declaration.
*
* Details:
*
* Using vanilla new/new[] is unsafe in SpiderMonkey because they throw on
* failure instead of returning NULL, which is what SpiderMonkey expects.
* (Even overriding them is unsafe, as the system's C++ runtime library may
* throw, which we do not support. We also can't just use the 'nothrow'
* variant of new/new[], because we want to mediate *all* allocations
* within SpiderMonkey, to satisfy any embedders using
* JS_USE_CUSTOM_ALLOCATOR.)
*
* JSContexts and JSRuntimes keep track of memory allocated, and use this
* accounting to schedule GC. OffTheBooks does not. We'd like to remove
* OffTheBooks allocations as much as possible (bug 636558).
*
* On allocation failure, a JSContext correctly reports an error, which a
* JSRuntime and OffTheBooks does not.
*
* A JSContext deallocates in a background thread. A JSRuntime might
* deallocate in the background in the future, but does not now. Foreground
* deallocation is preferable on slow paths. UnwantedForeground deallocations
* occur where we have no JSContext or JSRuntime, and the deallocation is not
* on a slow path. We want to remove UnwantedForeground deallocations (bug
* 636561).
*
* JS_DECLARE_ALLOCATION_FRIENDS_FOR_PRIVATE_CONSTRUCTOR makes the allocation
* classes friends with your class, giving them access to private
* constructors and destructors.
*
* |make check| does a source level check on the number of uses OffTheBooks,
* UnwantedForeground, js_malloc, js_free etc, to prevent regressions. If you
* really must add one, update Makefile.in, and run |make check|.
*
* |make check| also statically prevents the use of vanilla new/new[].
*/
#define JS_NEW_BODY(allocator, t, parms) \
void *memory = allocator(sizeof(t)); \
return memory ? new(memory) t parms : NULL;
/*
* Given a class which should provide new_() methods, add
* JS_DECLARE_NEW_METHODS (see JSContext for a usage example). This
* adds new_()s with up to 12 parameters. Add more versions of new_ below if
* you need more than 12 parameters.
*
* Note: Do not add a ; at the end of a use of JS_DECLARE_NEW_METHODS,
* or the build will break.
*/
#define JS_DECLARE_NEW_METHODS(ALLOCATOR, QUALIFIERS)\
template <class T>\
QUALIFIERS T *new_() {\
JS_NEW_BODY(ALLOCATOR, T, ())\
}\
\
template <class T, class P1>\
QUALIFIERS T *new_(P1 p1) {\
JS_NEW_BODY(ALLOCATOR, T, (p1))\
}\
\
template <class T, class P1, class P2>\
QUALIFIERS T *new_(P1 p1, P2 p2) {\
JS_NEW_BODY(ALLOCATOR, T, (p1, p2))\
}\
\
template <class T, class P1, class P2, class P3>\
QUALIFIERS T *new_(P1 p1, P2 p2, P3 p3) {\
JS_NEW_BODY(ALLOCATOR, T, (p1, p2, p3))\
}\
\
template <class T, class P1, class P2, class P3, class P4>\
QUALIFIERS T *new_(P1 p1, P2 p2, P3 p3, P4 p4) {\
JS_NEW_BODY(ALLOCATOR, T, (p1, p2, p3, p4))\
}\
\
template <class T, class P1, class P2, class P3, class P4, class P5>\
QUALIFIERS T *new_(P1 p1, P2 p2, P3 p3, P4 p4, P5 p5) {\
JS_NEW_BODY(ALLOCATOR, T, (p1, p2, p3, p4, p5))\
}\
\
template <class T, class P1, class P2, class P3, class P4, class P5, class P6>\
QUALIFIERS T *new_(P1 p1, P2 p2, P3 p3, P4 p4, P5 p5, P6 p6) {\
JS_NEW_BODY(ALLOCATOR, T, (p1, p2, p3, p4, p5, p6))\
}\
\
template <class T, class P1, class P2, class P3, class P4, class P5, class P6, class P7>\
QUALIFIERS T *new_(P1 p1, P2 p2, P3 p3, P4 p4, P5 p5, P6 p6, P7 p7) {\
JS_NEW_BODY(ALLOCATOR, T, (p1, p2, p3, p4, p5, p6, p7))\
}\
\
template <class T, class P1, class P2, class P3, class P4, class P5, class P6, class P7, class P8>\
QUALIFIERS T *new_(P1 p1, P2 p2, P3 p3, P4 p4, P5 p5, P6 p6, P7 p7, P8 p8) {\
JS_NEW_BODY(ALLOCATOR, T, (p1, p2, p3, p4, p5, p6, p7, p8))\
}\
\
template <class T, class P1, class P2, class P3, class P4, class P5, class P6, class P7, class P8, class P9>\
QUALIFIERS T *new_(P1 p1, P2 p2, P3 p3, P4 p4, P5 p5, P6 p6, P7 p7, P8 p8, P9 p9) {\
JS_NEW_BODY(ALLOCATOR, T, (p1, p2, p3, p4, p5, p6, p7, p8, p9))\
}\
\
template <class T, class P1, class P2, class P3, class P4, class P5, class P6, class P7, class P8, class P9, class P10>\
QUALIFIERS T *new_(P1 p1, P2 p2, P3 p3, P4 p4, P5 p5, P6 p6, P7 p7, P8 p8, P9 p9, P10 p10) {\
JS_NEW_BODY(ALLOCATOR, T, (p1, p2, p3, p4, p5, p6, p7, p8, p9, p10))\
}\
\
template <class T, class P1, class P2, class P3, class P4, class P5, class P6, class P7, class P8, class P9, class P10, class P11>\
QUALIFIERS T *new_(P1 p1, P2 p2, P3 p3, P4 p4, P5 p5, P6 p6, P7 p7, P8 p8, P9 p9, P10 p10, P11 p11) {\
JS_NEW_BODY(ALLOCATOR, T, (p1, p2, p3, p4, p5, p6, p7, p8, p9, p10, p11))\
}\
\
template <class T, class P1, class P2, class P3, class P4, class P5, class P6, class P7, class P8, class P9, class P10, class P11, class P12>\
QUALIFIERS T *new_(P1 p1, P2 p2, P3 p3, P4 p4, P5 p5, P6 p6, P7 p7, P8 p8, P9 p9, P10 p10, P11 p11, P12 p12) {\
JS_NEW_BODY(ALLOCATOR, T, (p1, p2, p3, p4, p5, p6, p7, p8, p9, p10, p11, p12))\
}\
static const int JSMinAlignment = 8;\
template <class T>\
QUALIFIERS T *array_new(size_t n) {\
/* The length is stored just before the vector memory. */\
uint64 numBytes64 = uint64(JSMinAlignment) + uint64(sizeof(T)) * uint64(n);\
size_t numBytes = size_t(numBytes64);\
if (numBytes64 != numBytes) {\
JS_ASSERT(0); /* we want to know if this happens in debug builds */\
return NULL;\
}\
void *memory = ALLOCATOR(numBytes);\
if (!memory)\
return NULL;\
*(size_t *)memory = n;\
memory = (void*)(uintptr_t(memory) + JSMinAlignment);\
return new(memory) T[n];\
}\
#define JS_DECLARE_DELETE_METHODS(DEALLOCATOR, QUALIFIERS)\
template <class T>\
QUALIFIERS void delete_(T *p) {\
if (p) {\
p->~T();\
DEALLOCATOR(p);\
}\
}\
\
template <class T>\
QUALIFIERS void array_delete(T *p) {\
if (p) {\
void* p0 = (void *)(uintptr_t(p) - js::OffTheBooks::JSMinAlignment);\
size_t n = *(size_t *)p0;\
for (size_t i = 0; i < n; i++)\
(p + i)->~T();\
DEALLOCATOR(p0);\
}\
}
/*
* In general, all allocations should go through a JSContext or JSRuntime, so
* that the garbage collector knows how much memory has been allocated. In
* cases where it is difficult to use a JSContext or JSRuntime, OffTheBooks can
* be used, though this is undesirable.
*/
namespace js {
/* Import common mfbt declarations into "js". */
using namespace mozilla;
class OffTheBooks {
public:
JS_DECLARE_NEW_METHODS(::js_malloc, JS_ALWAYS_INLINE static)
static JS_INLINE void* malloc_(size_t bytes) {
return ::js_malloc(bytes);
}
static JS_INLINE void* calloc_(size_t bytes) {
return ::js_calloc(bytes);
}
static JS_INLINE void* realloc_(void* p, size_t bytes) {
return ::js_realloc(p, bytes);
}
};
/*
* We generally prefer deallocating using JSContext because it can happen in
* the background. On slow paths, we may prefer foreground allocation.
*/
class Foreground {
public:
/* See parentheses comment above. */
static JS_ALWAYS_INLINE void free_(void* p) {
::js_free(p);
}
JS_DECLARE_DELETE_METHODS(::js_free, JS_ALWAYS_INLINE static)
};
class UnwantedForeground : public Foreground {
};
} /* namespace js */
/*
* Note lack of ; in JSRuntime below. This is intentional so "calling" this
* looks "normal".
*/
#define JS_DECLARE_ALLOCATION_FRIENDS_FOR_PRIVATE_CONSTRUCTOR \
friend class js::OffTheBooks;\
friend class js::Foreground;\
friend class js::UnwantedForeground;\
friend struct ::JSContext;\
friend struct ::JSRuntime
/**
* The following classes are designed to cause assertions to detect
* inadvertent use of guard objects as temporaries. In other words,
* when we have a guard object whose only purpose is its constructor and
* destructor (and is never otherwise referenced), the intended use
* might be:
* JSAutoTempValueRooter tvr(cx, 1, &val);
* but is is easy to accidentally write:
* JSAutoTempValueRooter(cx, 1, &val);
* which compiles just fine, but runs the destructor well before the
* intended time.
*
* They work by adding (#ifdef DEBUG) an additional parameter to the
* guard object's constructor, with a default value, so that users of
* the guard object's API do not need to do anything. The default value
* of this parameter is a temporary object. C++ (ISO/IEC 14882:1998),
* section 12.2 [class.temporary], clauses 4 and 5 seem to assume a
* guarantee that temporaries are destroyed in the reverse of their
* construction order, but I actually can't find a statement that that
* is true in the general case (beyond the two specific cases mentioned
* there). However, it seems to be true.
*
* These classes are intended to be used only via the macros immediately
* below them:
* JS_DECL_USE_GUARD_OBJECT_NOTIFIER declares (ifdef DEBUG) a member
* variable, and should be put where a declaration of a private
* member variable would be placed.
* JS_GUARD_OBJECT_NOTIFIER_PARAM should be placed at the end of the
* parameters to each constructor of the guard object; it declares
* (ifdef DEBUG) an additional parameter.
* JS_GUARD_OBJECT_NOTIFIER_INIT is a statement that belongs in each
* constructor. It uses the parameter declared by
* JS_GUARD_OBJECT_NOTIFIER_PARAM.
*/
#ifdef DEBUG
class JSGuardObjectNotifier
{
private:
bool* mStatementDone;
public:
JSGuardObjectNotifier() : mStatementDone(NULL) {}
~JSGuardObjectNotifier() {
*mStatementDone = true;
}
void setStatementDone(bool *aStatementDone) {
mStatementDone = aStatementDone;
}
};
class JSGuardObjectNotificationReceiver
{
private:
bool mStatementDone;
public:
JSGuardObjectNotificationReceiver() : mStatementDone(false) {}
~JSGuardObjectNotificationReceiver() {
/*
* Assert that the guard object was not used as a temporary.
* (Note that this assert might also fire if Init is not called
* because the guard object's implementation is not using the
* above macros correctly.)
*/
JS_ASSERT(mStatementDone);
}
void Init(const JSGuardObjectNotifier &aNotifier) {
/*
* aNotifier is passed as a const reference so that we can pass a
* temporary, but we really intend it as non-const
*/
const_cast<JSGuardObjectNotifier&>(aNotifier).
setStatementDone(&mStatementDone);
}
};
#define JS_DECL_USE_GUARD_OBJECT_NOTIFIER \
JSGuardObjectNotificationReceiver _mCheckNotUsedAsTemporary;
#define JS_GUARD_OBJECT_NOTIFIER_PARAM \
, const JSGuardObjectNotifier& _notifier = JSGuardObjectNotifier()
#define JS_GUARD_OBJECT_NOTIFIER_PARAM_NO_INIT \
, const JSGuardObjectNotifier& _notifier
#define JS_GUARD_OBJECT_NOTIFIER_PARAM0 \
const JSGuardObjectNotifier& _notifier = JSGuardObjectNotifier()
#define JS_GUARD_OBJECT_NOTIFIER_INIT \
JS_BEGIN_MACRO _mCheckNotUsedAsTemporary.Init(_notifier); JS_END_MACRO
#else /* defined(DEBUG) */
#define JS_DECL_USE_GUARD_OBJECT_NOTIFIER
#define JS_GUARD_OBJECT_NOTIFIER_PARAM
#define JS_GUARD_OBJECT_NOTIFIER_PARAM_NO_INIT
#define JS_GUARD_OBJECT_NOTIFIER_PARAM0
#define JS_GUARD_OBJECT_NOTIFIER_INIT JS_BEGIN_MACRO JS_END_MACRO
#endif /* !defined(DEBUG) */
namespace js {
template <class T>
JS_ALWAYS_INLINE static void
PodZero(T *t)
{
memset(t, 0, sizeof(T));
}
template <class T>
JS_ALWAYS_INLINE static void
PodZero(T *t, size_t nelem)
{
memset(t, 0, nelem * sizeof(T));
}
/*
* Arrays implicitly convert to pointers to their first element, which is
* dangerous when combined with the above PodZero definitions. Adding an
* overload for arrays is ambiguous, so we need another identifier. The
* ambiguous overload is left to catch mistaken uses of PodZero; if you get a
* compile error involving PodZero and array types, use PodArrayZero instead.
*/
template <class T, size_t N> static void PodZero(T (&)[N]); /* undefined */
template <class T, size_t N> static void PodZero(T (&)[N], size_t); /* undefined */
template <class T, size_t N>
JS_ALWAYS_INLINE static void
PodArrayZero(T (&t)[N])
{
memset(t, 0, N * sizeof(T));
}
template <class T>
JS_ALWAYS_INLINE static void
PodCopy(T *dst, const T *src, size_t nelem)
{
/* Cannot find portable word-sized abs(). */
JS_ASSERT_IF(dst >= src, size_t(dst - src) >= nelem);
JS_ASSERT_IF(src >= dst, size_t(src - dst) >= nelem);
if (nelem < 128) {
for (const T *srcend = src + nelem; src != srcend; ++src, ++dst)
*dst = *src;
} else {
memcpy(dst, src, nelem * sizeof(T));
}
}
template <class T>
JS_ALWAYS_INLINE static bool
PodEqual(T *one, T *two, size_t len)
{
if (len < 128) {
T *p1end = one + len;
for (T *p1 = one, *p2 = two; p1 != p1end; ++p1, ++p2) {
if (*p1 != *p2)
return false;
}
return true;
}
return !memcmp(one, two, len * sizeof(T));
}
/*
* Ordinarily, a function taking a JSContext* 'cx' paremter reports errors on
* the context. In some cases, functions optionally report and indicate this by
* taking a nullable 'maybecx' parameter. In some cases, though, a function
* always needs a 'cx', but optionally reports. This option is presented by the
* MaybeReportError.
*/
enum MaybeReportError { REPORT_ERROR = true, DONT_REPORT_ERROR = false };
/*
* "Move" References
*
* Some types can be copied much more efficiently if we know the original's
* value need not be preserved --- that is, if we are doing a "move", not a
* "copy". For example, if we have:
*
* Vector<T> u;
* Vector<T> v(u);
*
* the constructor for v must apply a copy constructor to each element of u ---
* taking time linear in the length of u. However, if we know we will not need u
* any more once v has been initialized, then we could initialize v very
* efficiently simply by stealing u's dynamically allocated buffer and giving it
* to v --- a constant-time operation, regardless of the size of u.
*
* Moves often appear in container implementations. For example, when we append
* to a vector, we may need to resize its buffer. This entails moving each of
* its extant elements from the old, smaller buffer to the new, larger buffer.
* But once the elements have been migrated, we're just going to throw away the
* old buffer; we don't care if they still have their values. So if the vector's
* element type can implement "move" more efficiently than "copy", the vector
* resizing should by all means use a "move" operation. Hash tables also need to
* be resized.
*
* The details of the optimization, and whether it's worth applying, vary from
* one type to the next. And while some constructor calls are moves, many really
* are copies, and can't be optimized this way. So we need:
*
* 1) a way for a particular invocation of a copy constructor to say that it's
* really a move, and that the value of the original isn't important
* afterwards (althought it must still be safe to destroy); and
*
* 2) a way for a type (like Vector) to announce that it can be moved more
* efficiently than it can be copied, and provide an implementation of that
* move operation.
*
* The Move(T &) function takes a reference to a T, and returns an MoveRef<T>
* referring to the same value; that's 1). An MoveRef<T> is simply a reference
* to a T, annotated to say that a copy constructor applied to it may move that
* T, instead of copying it. Finally, a constructor that accepts an MoveRef<T>
* should perform a more efficient move, instead of a copy, providing 2).
*
* So, where we might define a copy constructor for a class C like this:
*
* C(const C &rhs) { ... copy rhs to this ... }
*
* we would declare a move constructor like this:
*
* C(MoveRef<C> rhs) { ... move rhs to this ... }
*
* And where we might perform a copy like this:
*
* C c2(c1);
*
* we would perform a move like this:
*
* C c2(Move(c1))
*
* Note that MoveRef<T> implicitly converts to T &, so you can pass an
* MoveRef<T> to an ordinary copy constructor for a type that doesn't support a
* special move constructor, and you'll just get a copy. This means that
* templates can use Move whenever they know they won't use the original value
* any more, even if they're not sure whether the type at hand has a specialized
* move constructor. If it doesn't, the MoveRef<T> will just convert to a T &,
* and the ordinary copy constructor will apply.
*
* A class with a move constructor can also provide a move assignment operator,
* which runs this's destructor, and then applies the move constructor to
* *this's memory. A typical definition:
*
* C &operator=(MoveRef<C> rhs) {
* this->~C();
* new(this) C(rhs);
* return *this;
* }
*
* With that in place, one can write move assignments like this:
*
* c2 = Move(c1);
*
* This destroys c1, moves c1's value to c2, and leaves c1 in an undefined but
* destructible state.
*
* This header file defines MoveRef and Move in the js namespace. It's up to
* individual containers to annotate moves as such, by calling Move; and it's up
* to individual types to define move constructors.
*
* One hint: if you're writing a move constructor where the type has members
* that should be moved themselves, it's much nicer to write this:
*
* C(MoveRef<C> c) : x(c->x), y(c->y) { }
*
* than the equivalent:
*
* C(MoveRef<C> c) { new(&x) X(c->x); new(&y) Y(c->y); }
*
* especially since GNU C++ fails to notice that this does indeed initialize x
* and y, which may matter if they're const.
*/
template<typename T>
class MoveRef {
public:
typedef T Referent;
explicit MoveRef(T &t) : pointer(&t) { }
T &operator*() const { return *pointer; }
T *operator->() const { return pointer; }
operator T &() const { return *pointer; }
private:
T *pointer;
};
template<typename T>
MoveRef<T> Move(T &t) { return MoveRef<T>(t); }
} /* namespace js */
#endif /* defined(__cplusplus) */
#endif /* jsutil_h___ */