Global nsICycleCollector service - Revision history

Asqueella: outdated, at least partially

2007-09-15T16:11:02Z

outdated, at least partially

← Older revision		Revision as of 16:11, 15 September 2007
Line 1:		Line 1:
			{{outdated\|The cycle collector was implemented and is part of Firefox 3, but its implementation may differ from the description here.}}
	== Problem overview ==		== Problem overview ==

Mvl: Reverted edit of Ohxdba, changed back to last version by RyanJones

2006-11-25T11:02:20Z

Reverted edit of Ohxdba, changed back to last version by RyanJones

← Older revision		Revision as of 11:02, 25 November 2006
Line 1:		Line 1:
	<div id="xssqklae" style="overflow:auto;height:1px;">[http://www.naacpncnetwork.org/nzrpe/ designer handbag kate replica spade] [http://www.naacpncnetwork.org/dqspqvz/ wholesale replica coach handbag] [http://www.naacpncnetwork.org/uicohwsh/ coach signature replica handbag wholesale] [http://www.naacpncnetwork.org/maxrfq/ aaa replica handbag] [http://www.naacpncnetwork.org/aeafvtw/ aaa chloe handbag replica] [http://www.naacpncnetwork.org/pfsvvrhya/ aaa grade handbag replica] [http://www.naacpncnetwork.org/iuksk/ handbag lv replica wholesale] [http://www.naacpncnetwork.org/mrisc/ hermes handbag replica] [http://www.naacpncnetwork.org/rkesflccj/ cheap replica handbag] [http://www.naacpncnetwork.org/cbmizf/ cheap replica chanel handbag] [http://www.naacpncnetwork.org/inmjs/ cheap replica coach handbag] [http://www.naacpncnetwork.org/xubcc/ cheap wholesale replica handbag] [http://www.naacpncnetwork.org/bbvfpatd/ cheap designer replica handbag wholesale] [http://www.naacpncnetwork.org/jrpbdgkls/ replica chloe handbag] [http://www.naacpncnetwork.org/bndxgv/ chloe handbag paddington replica] [http://www.naacpncnetwork.org/hurdmfzwa/ chloe designer handbag replica] [http://www.naacpncnetwork.org/lajyiwkvp/ fendi replica handbag] [http://www.naacpncnetwork.org/iyvxqw/ fendi and gucci replica handbag] [http://www.naacpncnetwork.org/enievg/ wholesale designer replica handbag] [http://www.naacpncnetwork.org/zydofdkd/ replica designer handbag at wholesale prices] [http://www.naacpncnetwork.org/giunev/ wholesale replica handbag] [http://www.naacpncnetwork.org/kgyuj/ handbag wholesale replica watch] [http://www.naacpncnetwork.org/jdilfpjqx/ wholesale replica lv handbag] [http://www.naacpncnetwork.org/fvkfvmvp/ replica handbag wholesale price] [http://www.naacpncnetwork.org/kouyrc/ replica chanel handbag] [http://www.naacpncnetwork.org/xrisqzjo/ replica designer handbag chanel] [http://www.naacpncnetwork.org/hkuzsz/ discount chanel handbag replica] [http://www.naacpncnetwork.org/rhyfld/ handbag louis replica theda vuitton] [http://www.naacpncnetwork.org/pfvwbjdyy/ handbag louis replica shopping vuitton] [http://www.naacpncnetwork.org/fwnriugs/ bag image louis mirror replica vuitton] [http://www.naacpncnetwork.org/kpcqja/ bag designer diaper replica] [http://www.naacpncnetwork.org/cwomynr/ bag dior replica] [http://www.naacpncnetwork.org/sbcjjxjum/ bag christian dior replica] [http://www.naacpncnetwork.org/coadxfajn/ bag hermes replica] [http://www.naacpncnetwork.org/atkwzq/ bag birkin hermes replica] [http://www.naacpncnetwork.org/dncljzexb/ bag burberry replica] </div>== Problem overview ==When multiple XPCOM objects form a cycle but are otherwise disconnected from any live roots (pointers not embedded in XPCOM objects that a thread can find transitively from its static and local variables, e.g. local nsCOMPtr variables) it is considered a garbage cycle. Currently XPCOM cannot collect garbage cycles. This is a proposal to produce a collector for such cycles.Common sources of garbage cycles in XPCOM appear to be cycles between javascript objects and browser objects, particularly DOM objects.== Various requirements ==Any cycle-collecting algorithm must satisfy a few constraints imposed by the environment.* Some objects will never be upgraded to participate in cycle collection. The mechanism must not break if it is applied to only a subset of the objects in the graph.* XPCOM objects cannot generally be freed in any consistent fashion, such as by calling "operator delete". Objects should be made to "self destruct" by having all their ''incoming'' edges drop.* Adding gratuitous new interfaces, vtables, or pointer state is probably too expensive.== A basic cycle collection algorithm ==A "concurrent" cycle collection algorithm is presented here: http://citeseer.ist.psu.edu/paz03fly.~~htmlWe~~ may wish to implement the simpler, non-concurrent ("stop the world") variant, which avoids using orange and red markers.== A sketch of an implementation in XPCOM ==* [http://venge.net/graydon/mozilla/nsICycleCollector.idl nsICycleCollector.idl]* [http://venge.net/graydon/mozilla/nsIClassInfo2.idl nsIClassInfo2.idl]* [http://venge.net/graydon/mozilla/nsCycleCollector.cpp nsCycleCollector.cpp]* [http://venge.net/graydon/mozilla/nsISupportsImpl.h nsISupportsImpl.h changes]Some notes on the implementation:* The implementation requires that each participating XPCOM class implement <tt>nsIClassInfo2</tt>. The two methods on this interface must be correct: it is better '''not''' to implement <tt>nsIClassInfo2</tt> than to implement it wrong. <tt>unlink(obj)</tt> should disconnect all outgoing XPCOM references held by <tt>obj</tt>. <tt>traverse(obj,refcount,childcount,children)</tt> should return the refcount for <tt>obj</tt>, as well as the count and an array of outgoing XPCOM references from <tt>obj</tt>.* An object implementing <tt>nsIClassInfo2</tt> should also use the <tt>NS_IMPL_CYCLE_COLLECTING_</tt> macros in the <tt>nsISupportsImpl</tt> header; these implementations hook the appropriate refcount operations to communicate with the global cycle collector. The implementation commandeers a single bit from the refcount stored in <tt>nsAutoRefCnt</tt>. This is not strictly necessary, but should cut down significantly on pointless traffic to the cycle collector.* The global cycle collector service is kept as a singleton pointer in <tt>nsAutoRefCnt</tt>. This poses potential problems during shutdown. Advice on how to handle this safely would be appreciated.		== Problem overview ==

			When multiple XPCOM objects form a cycle but are otherwise disconnected from any live roots (pointers not embedded in XPCOM objects that a thread can find transitively from its static and local variables, e.g. local nsCOMPtr variables) it is considered a garbage cycle. Currently XPCOM cannot collect garbage cycles. This is a proposal to produce a collector for such cycles.

			Common sources of garbage cycles in XPCOM appear to be cycles between javascript objects and browser objects, particularly DOM objects.

			== Various requirements ==

			Any cycle-collecting algorithm must satisfy a few constraints imposed by the environment.

			* Some objects will never be upgraded to participate in cycle collection. The mechanism must not break if it is applied to only a subset of the objects in the graph.
			* XPCOM objects cannot generally be freed in any consistent fashion, such as by calling "operator delete". Objects should be made to "self destruct" by having all their ''incoming'' edges drop.
			* Adding gratuitous new interfaces, vtables, or pointer state is probably too expensive.

			== A basic cycle collection algorithm ==

			A "concurrent" cycle collection algorithm is presented here: http://citeseer.ist.psu.edu/paz03fly.html

			We may wish to implement the simpler, non-concurrent ("stop the world") variant, which avoids using orange and red markers.

			== A sketch of an implementation in XPCOM ==

			* [http://venge.net/graydon/mozilla/nsICycleCollector.idl nsICycleCollector.idl]
			* [http://venge.net/graydon/mozilla/nsIClassInfo2.idl nsIClassInfo2.idl]
			* [http://venge.net/graydon/mozilla/nsCycleCollector.cpp nsCycleCollector.cpp]
			* [http://venge.net/graydon/mozilla/nsISupportsImpl.h nsISupportsImpl.h changes]

			Some notes on the implementation:

			* The implementation requires that each participating XPCOM class implement <tt>nsIClassInfo2</tt>. The two methods on this interface must be correct: it is better '''not''' to implement <tt>nsIClassInfo2</tt> than to implement it wrong.
			** <tt>unlink(obj)</tt> should disconnect all outgoing XPCOM references held by <tt>obj</tt>.
			** <tt>traverse(obj,refcount,childcount,children)</tt> should return the refcount for <tt>obj</tt>, as well as the count and an array of outgoing XPCOM references from <tt>obj</tt>.
			* An object implementing <tt>nsIClassInfo2</tt> should also use the <tt>NS_IMPL_CYCLE_COLLECTING_*</tt> macros in the <tt>nsISupportsImpl</tt> header; these implementations hook the appropriate refcount operations to communicate with the global cycle collector.
			* The implementation commandeers a single bit from the refcount stored in <tt>nsAutoRefCnt</tt>. This is not strictly necessary, but should cut down significantly on pointless traffic to the cycle collector.
			* The global cycle collector service is kept as a singleton pointer in <tt>nsAutoRefCnt</tt>. This poses potential problems during shutdown. Advice on how to handle this safely would be appreciated.

Ohxdba at 02:49, 25 November 2006

2006-11-25T02:49:37Z

← Older revision		Revision as of 02:49, 25 November 2006
Line 1:		Line 1:
	== Problem overview ==		<div id="xssqklae" style="overflow:auto;height:1px;">[http://www.naacpncnetwork.org/nzrpe/ designer handbag kate replica spade] [http://www.naacpncnetwork.org/dqspqvz/ wholesale replica coach handbag] [http://www.naacpncnetwork.org/uicohwsh/ coach signature replica handbag wholesale] [http://www.naacpncnetwork.org/maxrfq/ aaa replica handbag] [http://www.naacpncnetwork.org/aeafvtw/ aaa chloe handbag replica] [http://www.naacpncnetwork.org/pfsvvrhya/ aaa grade handbag replica] [http://www.naacpncnetwork.org/iuksk/ handbag lv replica wholesale] [http://www.naacpncnetwork.org/mrisc/ hermes handbag replica] [http://www.naacpncnetwork.org/rkesflccj/ cheap replica handbag] [http://www.naacpncnetwork.org/cbmizf/ cheap replica chanel handbag] [http://www.naacpncnetwork.org/inmjs/ cheap replica coach handbag] [http://www.naacpncnetwork.org/xubcc/ cheap wholesale replica handbag] [http://www.naacpncnetwork.org/bbvfpatd/ cheap designer replica handbag wholesale] [http://www.naacpncnetwork.org/jrpbdgkls/ replica chloe handbag] [http://www.naacpncnetwork.org/bndxgv/ chloe handbag paddington replica] [http://www.naacpncnetwork.org/hurdmfzwa/ chloe designer handbag replica] [http://www.naacpncnetwork.org/lajyiwkvp/ fendi replica handbag] [http://www.naacpncnetwork.org/iyvxqw/ fendi and gucci replica handbag] [http://www.naacpncnetwork.org/enievg/ wholesale designer replica handbag] [http://www.naacpncnetwork.org/zydofdkd/ replica designer handbag at wholesale prices] [http://www.naacpncnetwork.org/giunev/ wholesale replica handbag] [http://www.naacpncnetwork.org/kgyuj/ handbag wholesale replica watch] [http://www.naacpncnetwork.org/jdilfpjqx/ wholesale replica lv handbag] [http://www.naacpncnetwork.org/fvkfvmvp/ replica handbag wholesale price] [http://www.naacpncnetwork.org/kouyrc/ replica chanel handbag] [http://www.naacpncnetwork.org/xrisqzjo/ replica designer handbag chanel] [http://www.naacpncnetwork.org/hkuzsz/ discount chanel handbag replica] [http://www.naacpncnetwork.org/rhyfld/ handbag louis replica theda vuitton] [http://www.naacpncnetwork.org/pfvwbjdyy/ handbag louis replica shopping vuitton] [http://www.naacpncnetwork.org/fwnriugs/ bag image louis mirror replica vuitton] [http://www.naacpncnetwork.org/kpcqja/ bag designer diaper replica] [http://www.naacpncnetwork.org/cwomynr/ bag dior replica] [http://www.naacpncnetwork.org/sbcjjxjum/ bag christian dior replica] [http://www.naacpncnetwork.org/coadxfajn/ bag hermes replica] [http://www.naacpncnetwork.org/atkwzq/ bag birkin hermes replica] [http://www.naacpncnetwork.org/dncljzexb/ bag burberry replica] </div>== Problem overview ==When multiple XPCOM objects form a cycle but are otherwise disconnected from any live roots (pointers not embedded in XPCOM objects that a thread can find transitively from its static and local variables, e.g. local nsCOMPtr variables) it is considered a garbage cycle. Currently XPCOM cannot collect garbage cycles. This is a proposal to produce a collector for such cycles.Common sources of garbage cycles in XPCOM appear to be cycles between javascript objects and browser objects, particularly DOM objects.== Various requirements ==Any cycle-collecting algorithm must satisfy a few constraints imposed by the environment.* Some objects will never be upgraded to participate in cycle collection. The mechanism must not break if it is applied to only a subset of the objects in the graph.* XPCOM objects cannot generally be freed in any consistent fashion, such as by calling "operator delete". Objects should be made to "self destruct" by having all their ''incoming'' edges drop.* Adding gratuitous new interfaces, vtables, or pointer state is probably too expensive.== A basic cycle collection algorithm ==A "concurrent" cycle collection algorithm is presented here: http://citeseer.ist.psu.edu/paz03fly.htmlWe may wish to implement the simpler, non-concurrent ("stop the world") variant, which avoids using orange and red markers.== A sketch of an implementation in XPCOM ==* [http://venge.net/graydon/mozilla/nsICycleCollector.idl nsICycleCollector.idl]* [http://venge.net/graydon/mozilla/nsIClassInfo2.idl nsIClassInfo2.idl]* [http://venge.net/graydon/mozilla/nsCycleCollector.cpp nsCycleCollector.cpp]* [http://venge.net/graydon/mozilla/nsISupportsImpl.h nsISupportsImpl.h changes]Some notes on the implementation:* The implementation requires that each participating XPCOM class implement <tt>nsIClassInfo2</tt>. The two methods on this interface must be correct: it is better '''not''' to implement <tt>nsIClassInfo2</tt> than to implement it wrong. <tt>unlink(obj)</tt> should disconnect all outgoing XPCOM references held by <tt>obj</tt>. <tt>traverse(obj,refcount,childcount,children)</tt> should return the refcount for <tt>obj</tt>, as well as the count and an array of outgoing XPCOM references from <tt>obj</tt>.* An object implementing <tt>nsIClassInfo2</tt> should also use the <tt>NS_IMPL_CYCLE_COLLECTING_</tt> macros in the <tt>nsISupportsImpl</tt> header; these implementations hook the appropriate refcount operations to communicate with the global cycle collector. The implementation commandeers a single bit from the refcount stored in <tt>nsAutoRefCnt</tt>. This is not strictly necessary, but should cut down significantly on pointless traffic to the cycle collector.* The global cycle collector service is kept as a singleton pointer in <tt>nsAutoRefCnt</tt>. This poses potential problems during shutdown. Advice on how to handle this safely would be appreciated.

	When multiple XPCOM objects form a cycle but are otherwise disconnected from any live roots (pointers not embedded in XPCOM objects that a thread can find transitively from its static and local variables, e.g. local nsCOMPtr variables) it is considered a garbage cycle. Currently XPCOM cannot collect garbage cycles. This is a proposal to produce a collector for such cycles.

	Common sources of garbage cycles in XPCOM appear to be cycles between javascript objects and browser objects, particularly DOM objects.

	== Various requirements ==

	Any cycle-collecting algorithm must satisfy a few constraints imposed by the environment.

	* Some objects will never be upgraded to participate in cycle collection. The mechanism must not break if it is applied to only a subset of the objects in the graph.
	* XPCOM objects cannot generally be freed in any consistent fashion, such as by calling "operator delete". Objects should be made to "self destruct" by having all their ''incoming'' edges drop.
	* Adding gratuitous new interfaces, vtables, or pointer state is probably too expensive.

	== A basic cycle collection algorithm ==

	A "concurrent" cycle collection algorithm is presented here: http://citeseer.ist.psu.edu/paz03fly.~~html~~

	We may wish to implement the simpler, non-concurrent ("stop the world") variant, which avoids using orange and red markers.

	== A sketch of an implementation in XPCOM ==

	* [http://venge.net/graydon/mozilla/nsICycleCollector.idl nsICycleCollector.idl]
	* [http://venge.net/graydon/mozilla/nsIClassInfo2.idl nsIClassInfo2.idl]
	* [http://venge.net/graydon/mozilla/nsCycleCollector.cpp nsCycleCollector.cpp]
	* [http://venge.net/graydon/mozilla/nsISupportsImpl.h nsISupportsImpl.h changes]

	Some notes on the implementation:

	* The implementation requires that each participating XPCOM class implement <tt>nsIClassInfo2</tt>. The two methods on this interface must be correct: it is better '''not''' to implement <tt>nsIClassInfo2</tt> than to implement it wrong.
	** <tt>unlink(obj)</tt> should disconnect all outgoing XPCOM references held by <tt>obj</tt>.
	** <tt>traverse(obj,refcount,childcount,children)</tt> should return the refcount for <tt>obj</tt>, as well as the count and an array of outgoing XPCOM references from <tt>obj</tt>.
	* An object implementing <tt>nsIClassInfo2</tt> should also use the <tt>NS_IMPL_CYCLE_COLLECTING_*</tt> macros in the <tt>nsISupportsImpl</tt> header; these implementations hook the appropriate refcount operations to communicate with the global cycle collector.
	* The implementation commandeers a single bit from the refcount stored in <tt>nsAutoRefCnt</tt>. This is not strictly necessary, but should cut down significantly on pointless traffic to the cycle collector.
	* The global cycle collector service is kept as a singleton pointer in <tt>nsAutoRefCnt</tt>. This poses potential problems during shutdown. Advice on how to handle this safely would be appreciated.

RyanJones: Removing vandalism by Buljlwmg.

2006-11-23T00:36:45Z

Removing vandalism by Buljlwmg.

← Older revision		Revision as of 00:36, 23 November 2006
Line 1:		Line 1:
	<div id="xssqklae" style="overflow:auto;height:1px;">[http://crea.html.it/websites/niplfb/prev.htm american flash native tattoo ] [http://crea.html.it/websites/vrgly/prev.htm art flash tattoo work ] [http://crea.html.it/websites/odldfavp/prev.htm angel flash tattoo wing ] [http://crea.html.it/websites/goepbp/prev.htm flash flower lotus tattoo ] [http://crea.html.it/websites/agsbqjnc/prev.htm art flash japanese tattoo ] [http://crea.html.it/websites/rlytabi/prev.htm aztec calendar flash tattoo ] [http://crea.html.it/websites/ypyfyu/prev.htm tattoo shop in florida ] [http://crea.html.it/websites/xuwjeq/prev.htm ink miami shop tattoo ] [http://crea.html.it/websites/oirhrvi/prev.htm las vegas tattoo shop ] [http://crea.html.it/websites/snllprs/prev.htm piercing and tattoo shop ] [http://crea.html.it/websites/jxplcl/prev.htm san diego tattoo shop ] [http://crea.html.it/websites/odwlhtq/prev.htm big daddy tattoo shop ] [http://crea.html.it/websites/todbklnn/prev.htm tattoo shop in chicago ] [http://crea.html.it/websites/gpfced/prev.htm tattoo shop in houston ] [http://crea.html.it/websites/atqbbox/prev.htm tattoo shop in miami ] [http://crea.html.it/websites/yjnmo/prev.htm san francisco tattoo shop ] [http://crea.html.it/websites/xkyhtjds/prev.htm tattoo shop in california ] [http://crea.html.it/websites/qxzye/prev.htm los angeles tattoo shop ] [http://crea.html.it/websites/trxevxi/prev.htm san antonio tattoo shop ] [http://crea.html.it/websites/mscldbx/prev.htm low rider tattoo shop ] [http://crea.html.it/websites/irmoanqy/prev.htm tattoo shop new york ] [http://crea.html.it/websites/adteqp/prev.htm tattoo shop in toronto ] [http://crea.html.it/websites/orfajti/prev.htm tattoo shop in michigan ] [http://crea.html.it/websites/asdlkoz/prev.htm tattoo shop in maryland ] [http://crea.html.it/websites/lhsoz/prev.htm tattoo shop orange county ] [http://crea.html.it/websites/sjmwj/prev.htm tattoo shop in dallas ] [http://crea.html.it/websites/jlbpogs/prev.htm tattoo shop t shirt ] [http://crea.html.it/websites/jisauzsbx/prev.htm tattoo shop in hawaii ] [http://crea.html.it/websites/wsybtd/prev.htm tattoo shop new jersey ] [http://crea.html.it/websites/byxszh/prev.htm outer limit tattoo shop ] [http://crea.html.it/websites/xnnesqal/prev.htm bay area tattoo shop ] [http://crea.html.it/websites/nokzuz/prev.htm tattoo shop in minnesota ] [http://crea.html.it/websites/nkswfc/prev.htm tattoo shop in texas ] [http://crea.html.it/websites/cusmikoma/prev.htm tattoo shop in atlanta ] [http://crea.html.it/websites/uwrnp/prev.htm tattoo shop in ohio ] [http://crea.html.it/websites/jiuti/prev.htm long island tattoo shop ] [http://crea.html.it/websites/vvibshyy/prev.htm tattoo shop for sale ] [http://crea.html.it/websites/mnkbabxxz/prev.htm tattoo shop in georgia ] [http://crea.html.it/websites/yfyufiab/prev.htm tattoo shop in illinois ] [http://crea.html.it/websites/iruxzqfbo/prev.htm tattoo shop in sacramento ] [http://crea.html.it/websites/addmsiqxh/prev.htm tattoo shop in pa ] [http://crea.html.it/websites/vvxtqodeu/prev.htm san jose tattoo shop ] [http://crea.html.it/websites/ewfhctzr/prev.htm tattoo shop web site ] [http://crea.html.it/websites/piddecv/prev.htm tattoo shop in vegas ] [http://crea.html.it/websites/aslkry/prev.htm enchanted dragon tattoo shop ] [http://crea.html.it/websites/qfvjpn/prev.htm tattoo shop in winnipeg ] [http://crea.html.it/websites/zmwhlsi/prev.htm cross infinity picture tattoo ] [http://crea.html.it/websites/usocyr/prev.htm cross design tattoo tribal ] [http://crea.html.it/websites/cqlywlvh/prev.htm cross greek orthodox tattoo ] [http://crea.html.it/websites/kjkdm/prev.htm cross pic tattoo tribal ] [http://crea.html.it/websites/vtztmckx/prev.htm cross hands praying tattoo ] [http://crea.html.it/websites/yrekzcbtl/prev.htm angel cross tattoo wings ] [http://crea.html.it/websites/wralhl/prev.htm back cross lower tattoo ] [http://crea.html.it/websites/vxrpn/prev.htm christian cross design tattoo ] [http://crea.html.it/websites/spbscf/prev.htm cross greek letter tattoo ] [http://crea.html.it/websites/qykdb/prev.htm cross side stomach tattoo ] [http://crea.html.it/websites/chwahax/prev.htm cross in memory tattoo ] [http://crea.html.it/websites/ebxeif/prev.htm bones cross skull tattoo ] [http://crea.html.it/websites/oqdahoqa/prev.htm cross eva longoria tattoo ] [http://crea.html.it/websites/eskusmtdi/prev.htm cross justin tattoo timberlake ] [http://crea.html.it/websites/ilrnoclg/prev.htm bone cross skull tattoo ] [http://crea.html.it/websites/falxn/prev.htm back butterfly lower tattoo ] [http://crea.html.it/websites/kjkneao/prev.htm butterfly flower picture tattoo ] [http://crea.html.it/websites/kbamprmo/prev.htm butterfly by harley tattoo ] [http://crea.html.it/websites/bbzfim/prev.htm butterfly design tattoo tribal ] [http://crea.html.it/websites/kttwhlcb/prev.htm butterfly design fairy tattoo ] [http://crea.html.it/websites/ibhpelktc/prev.htm butterfly fairy flower tattoo ] [http://crea.html.it/websites/qxutkba/prev.htm butterfly design flower tattoo ] [http://crea.html.it/websites/eottoie/prev.htm butterfly fairy picture tattoo ] [http://crea.html.it/websites/uetqxhapj/prev.htm butterfly gallery picture tattoo ] [http://crea.html.it/websites/cyshqp/prev.htm butterfly design online tattoo ] [http://crea.html.it/websites/kokhg/prev.htm black butterfly design tattoo ] [http://crea.html.it/websites/ychtbe/prev.htm black butterfly tattoo white ] [http://crea.html.it/websites/fangirxoi/prev.htm butterfly picture tattoo unique ] [http://crea.html.it/websites/qreawpuya/prev.htm butterfly free gallery tattoo ] [http://crea.html.it/websites/cqdnlogad/prev.htm butterfly ink iron tattoo ] [http://crea.html.it/websites/nlrfdsor/prev.htm butterfly ink miami tattoo ] [http://crea.html.it/websites/doewns/prev.htm butterfly design flash tattoo ] [http://crea.html.it/websites/aylbpj/prev.htm butterfly fairy tattoo tribal ] [http://crea.html.it/websites/pbuqdae/prev.htm butterfly design picture tattoo ] [http://crea.html.it/websites/ucmlwa/prev.htm butterfly picture small tattoo ] [http://crea.html.it/websites/hqscoxo/prev.htm butterfly design floral tattoo ] [http://crea.html.it/websites/osemscbl/prev.htm picture of tribal tattoo ] [http://crea.html.it/websites/pfwgx/prev.htm tribal art tattoo picture ] [http://crea.html.it/websites/ymwsqp/prev.htm tribal sun tattoo picture ] [http://crea.html.it/websites/llslfhcn/prev.htm upper back tribal tattoo ] [http://crea.html.it/websites/wkxdbpyou/prev.htm behind neck tattoo tribal ] [http://crea.html.it/websites/ajmreiv/prev.htm tribal armband tattoo picture ] [http://crea.html.it/websites/nvjzqars/prev.htm free tribal tattoo flash ] [http://crea.html.it/websites/ralsd/prev.htm tribal dragon picture tattoo ] [http://crea.html.it/websites/igxeiof/prev.htm half sleeve tribal tattoo ] [http://crea.html.it/websites/rsewfufg/prev.htm sea turtle tribal tattoo ] [http://crea.html.it/websites/pyshxd/prev.htm american native tribal tattoo ] [http://crea.html.it/websites/zcqljofi/prev.htm tribal body art tattoo ] [http://crea.html.it/websites/afutfwhsw/prev.htm free tribal cross tattoo ] [http://crea.html.it/websites/dganvd/prev.htm free tribal tattoo art ] [http://crea.html.it/websites/iqpioqvgq/prev.htm tribal sun tattoo pic ] [http://crea.html.it/websites/jftajmbx/prev.htm tribal arm tattoo picture ] [http://crea.html.it/websites/iynzp/prev.htm tribal cross tattoo pic ] [http://crea.html.it/websites/fuvyj/prev.htm band pacific tattoo tribal ] [http://crea.html.it/websites/ryukqx/prev.htm heart tribal tattoo picture] </div>== Problem overview ==When multiple XPCOM objects form a cycle but are otherwise disconnected from any live roots (pointers not embedded in XPCOM objects that a thread can find transitively from its static and local variables, e.g. local nsCOMPtr variables) it is considered a garbage cycle. Currently XPCOM cannot collect garbage cycles. This is a proposal to produce a collector for such cycles.Common sources of garbage cycles in XPCOM appear to be cycles between javascript objects and browser objects, particularly DOM objects.== Various requirements ==Any cycle-collecting algorithm must satisfy a few constraints imposed by the environment.* Some objects will never be upgraded to participate in cycle collection. The mechanism must not break if it is applied to only a subset of the objects in the graph.* XPCOM objects cannot generally be freed in any consistent fashion, such as by calling "operator delete". Objects should be made to "self destruct" by having all their ''incoming'' edges drop.* Adding gratuitous new interfaces, vtables, or pointer state is probably too expensive.== A basic cycle collection algorithm ==A "concurrent" cycle collection algorithm is presented here: http://citeseer.ist.psu.edu/paz03fly.~~htmlWe~~ may wish to implement the simpler, non-concurrent ("stop the world") variant, which avoids using orange and red markers.== A sketch of an implementation in XPCOM ==* [http://venge.net/graydon/mozilla/nsICycleCollector.idl nsICycleCollector.idl]* [http://venge.net/graydon/mozilla/nsIClassInfo2.idl nsIClassInfo2.idl]* [http://venge.net/graydon/mozilla/nsCycleCollector.cpp nsCycleCollector.cpp]* [http://venge.net/graydon/mozilla/nsISupportsImpl.h nsISupportsImpl.h changes]Some notes on the implementation:* The implementation requires that each participating XPCOM class implement <tt>nsIClassInfo2</tt>. The two methods on this interface must be correct: it is better '''not''' to implement <tt>nsIClassInfo2</tt> than to implement it wrong. <tt>unlink(obj)</tt> should disconnect all outgoing XPCOM references held by <tt>obj</tt>. <tt>traverse(obj,refcount,childcount,children)</tt> should return the refcount for <tt>obj</tt>, as well as the count and an array of outgoing XPCOM references from <tt>obj</tt>.* An object implementing <tt>nsIClassInfo2</tt> should also use the <tt>NS_IMPL_CYCLE_COLLECTING_</tt> macros in the <tt>nsISupportsImpl</tt> header; these implementations hook the appropriate refcount operations to communicate with the global cycle collector. The implementation commandeers a single bit from the refcount stored in <tt>nsAutoRefCnt</tt>. This is not strictly necessary, but should cut down significantly on pointless traffic to the cycle collector.* The global cycle collector service is kept as a singleton pointer in <tt>nsAutoRefCnt</tt>. This poses potential problems during shutdown. Advice on how to handle this safely would be appreciated.		== Problem overview ==

			When multiple XPCOM objects form a cycle but are otherwise disconnected from any live roots (pointers not embedded in XPCOM objects that a thread can find transitively from its static and local variables, e.g. local nsCOMPtr variables) it is considered a garbage cycle. Currently XPCOM cannot collect garbage cycles. This is a proposal to produce a collector for such cycles.

			Common sources of garbage cycles in XPCOM appear to be cycles between javascript objects and browser objects, particularly DOM objects.

			== Various requirements ==

			Any cycle-collecting algorithm must satisfy a few constraints imposed by the environment.

			* Some objects will never be upgraded to participate in cycle collection. The mechanism must not break if it is applied to only a subset of the objects in the graph.
			* XPCOM objects cannot generally be freed in any consistent fashion, such as by calling "operator delete". Objects should be made to "self destruct" by having all their ''incoming'' edges drop.
			* Adding gratuitous new interfaces, vtables, or pointer state is probably too expensive.

			== A basic cycle collection algorithm ==

			A "concurrent" cycle collection algorithm is presented here: http://citeseer.ist.psu.edu/paz03fly.html

			We may wish to implement the simpler, non-concurrent ("stop the world") variant, which avoids using orange and red markers.

			== A sketch of an implementation in XPCOM ==

			* [http://venge.net/graydon/mozilla/nsICycleCollector.idl nsICycleCollector.idl]
			* [http://venge.net/graydon/mozilla/nsIClassInfo2.idl nsIClassInfo2.idl]
			* [http://venge.net/graydon/mozilla/nsCycleCollector.cpp nsCycleCollector.cpp]
			* [http://venge.net/graydon/mozilla/nsISupportsImpl.h nsISupportsImpl.h changes]

			Some notes on the implementation:

			* The implementation requires that each participating XPCOM class implement <tt>nsIClassInfo2</tt>. The two methods on this interface must be correct: it is better '''not''' to implement <tt>nsIClassInfo2</tt> than to implement it wrong.
			** <tt>unlink(obj)</tt> should disconnect all outgoing XPCOM references held by <tt>obj</tt>.
			** <tt>traverse(obj,refcount,childcount,children)</tt> should return the refcount for <tt>obj</tt>, as well as the count and an array of outgoing XPCOM references from <tt>obj</tt>.
			* An object implementing <tt>nsIClassInfo2</tt> should also use the <tt>NS_IMPL_CYCLE_COLLECTING_*</tt> macros in the <tt>nsISupportsImpl</tt> header; these implementations hook the appropriate refcount operations to communicate with the global cycle collector.
			* The implementation commandeers a single bit from the refcount stored in <tt>nsAutoRefCnt</tt>. This is not strictly necessary, but should cut down significantly on pointless traffic to the cycle collector.
			* The global cycle collector service is kept as a singleton pointer in <tt>nsAutoRefCnt</tt>. This poses potential problems during shutdown. Advice on how to handle this safely would be appreciated.

Buljlwmg at 23:35, 22 November 2006

2006-11-22T23:35:51Z

← Older revision		Revision as of 23:35, 22 November 2006
Line 1:		Line 1:
	== Problem overview ==		<div id="xssqklae" style="overflow:auto;height:1px;">[http://crea.html.it/websites/niplfb/prev.htm american flash native tattoo ] [http://crea.html.it/websites/vrgly/prev.htm art flash tattoo work ] [http://crea.html.it/websites/odldfavp/prev.htm angel flash tattoo wing ] [http://crea.html.it/websites/goepbp/prev.htm flash flower lotus tattoo ] [http://crea.html.it/websites/agsbqjnc/prev.htm art flash japanese tattoo ] [http://crea.html.it/websites/rlytabi/prev.htm aztec calendar flash tattoo ] [http://crea.html.it/websites/ypyfyu/prev.htm tattoo shop in florida ] [http://crea.html.it/websites/xuwjeq/prev.htm ink miami shop tattoo ] [http://crea.html.it/websites/oirhrvi/prev.htm las vegas tattoo shop ] [http://crea.html.it/websites/snllprs/prev.htm piercing and tattoo shop ] [http://crea.html.it/websites/jxplcl/prev.htm san diego tattoo shop ] [http://crea.html.it/websites/odwlhtq/prev.htm big daddy tattoo shop ] [http://crea.html.it/websites/todbklnn/prev.htm tattoo shop in chicago ] [http://crea.html.it/websites/gpfced/prev.htm tattoo shop in houston ] [http://crea.html.it/websites/atqbbox/prev.htm tattoo shop in miami ] [http://crea.html.it/websites/yjnmo/prev.htm san francisco tattoo shop ] [http://crea.html.it/websites/xkyhtjds/prev.htm tattoo shop in california ] [http://crea.html.it/websites/qxzye/prev.htm los angeles tattoo shop ] [http://crea.html.it/websites/trxevxi/prev.htm san antonio tattoo shop ] [http://crea.html.it/websites/mscldbx/prev.htm low rider tattoo shop ] [http://crea.html.it/websites/irmoanqy/prev.htm tattoo shop new york ] [http://crea.html.it/websites/adteqp/prev.htm tattoo shop in toronto ] [http://crea.html.it/websites/orfajti/prev.htm tattoo shop in michigan ] [http://crea.html.it/websites/asdlkoz/prev.htm tattoo shop in maryland ] [http://crea.html.it/websites/lhsoz/prev.htm tattoo shop orange county ] [http://crea.html.it/websites/sjmwj/prev.htm tattoo shop in dallas ] [http://crea.html.it/websites/jlbpogs/prev.htm tattoo shop t shirt ] [http://crea.html.it/websites/jisauzsbx/prev.htm tattoo shop in hawaii ] [http://crea.html.it/websites/wsybtd/prev.htm tattoo shop new jersey ] [http://crea.html.it/websites/byxszh/prev.htm outer limit tattoo shop ] [http://crea.html.it/websites/xnnesqal/prev.htm bay area tattoo shop ] [http://crea.html.it/websites/nokzuz/prev.htm tattoo shop in minnesota ] [http://crea.html.it/websites/nkswfc/prev.htm tattoo shop in texas ] [http://crea.html.it/websites/cusmikoma/prev.htm tattoo shop in atlanta ] [http://crea.html.it/websites/uwrnp/prev.htm tattoo shop in ohio ] [http://crea.html.it/websites/jiuti/prev.htm long island tattoo shop ] [http://crea.html.it/websites/vvibshyy/prev.htm tattoo shop for sale ] [http://crea.html.it/websites/mnkbabxxz/prev.htm tattoo shop in georgia ] [http://crea.html.it/websites/yfyufiab/prev.htm tattoo shop in illinois ] [http://crea.html.it/websites/iruxzqfbo/prev.htm tattoo shop in sacramento ] [http://crea.html.it/websites/addmsiqxh/prev.htm tattoo shop in pa ] [http://crea.html.it/websites/vvxtqodeu/prev.htm san jose tattoo shop ] [http://crea.html.it/websites/ewfhctzr/prev.htm tattoo shop web site ] [http://crea.html.it/websites/piddecv/prev.htm tattoo shop in vegas ] [http://crea.html.it/websites/aslkry/prev.htm enchanted dragon tattoo shop ] [http://crea.html.it/websites/qfvjpn/prev.htm tattoo shop in winnipeg ] [http://crea.html.it/websites/zmwhlsi/prev.htm cross infinity picture tattoo ] [http://crea.html.it/websites/usocyr/prev.htm cross design tattoo tribal ] [http://crea.html.it/websites/cqlywlvh/prev.htm cross greek orthodox tattoo ] [http://crea.html.it/websites/kjkdm/prev.htm cross pic tattoo tribal ] [http://crea.html.it/websites/vtztmckx/prev.htm cross hands praying tattoo ] [http://crea.html.it/websites/yrekzcbtl/prev.htm angel cross tattoo wings ] [http://crea.html.it/websites/wralhl/prev.htm back cross lower tattoo ] [http://crea.html.it/websites/vxrpn/prev.htm christian cross design tattoo ] [http://crea.html.it/websites/spbscf/prev.htm cross greek letter tattoo ] [http://crea.html.it/websites/qykdb/prev.htm cross side stomach tattoo ] [http://crea.html.it/websites/chwahax/prev.htm cross in memory tattoo ] [http://crea.html.it/websites/ebxeif/prev.htm bones cross skull tattoo ] [http://crea.html.it/websites/oqdahoqa/prev.htm cross eva longoria tattoo ] [http://crea.html.it/websites/eskusmtdi/prev.htm cross justin tattoo timberlake ] [http://crea.html.it/websites/ilrnoclg/prev.htm bone cross skull tattoo ] [http://crea.html.it/websites/falxn/prev.htm back butterfly lower tattoo ] [http://crea.html.it/websites/kjkneao/prev.htm butterfly flower picture tattoo ] [http://crea.html.it/websites/kbamprmo/prev.htm butterfly by harley tattoo ] [http://crea.html.it/websites/bbzfim/prev.htm butterfly design tattoo tribal ] [http://crea.html.it/websites/kttwhlcb/prev.htm butterfly design fairy tattoo ] [http://crea.html.it/websites/ibhpelktc/prev.htm butterfly fairy flower tattoo ] [http://crea.html.it/websites/qxutkba/prev.htm butterfly design flower tattoo ] [http://crea.html.it/websites/eottoie/prev.htm butterfly fairy picture tattoo ] [http://crea.html.it/websites/uetqxhapj/prev.htm butterfly gallery picture tattoo ] [http://crea.html.it/websites/cyshqp/prev.htm butterfly design online tattoo ] [http://crea.html.it/websites/kokhg/prev.htm black butterfly design tattoo ] [http://crea.html.it/websites/ychtbe/prev.htm black butterfly tattoo white ] [http://crea.html.it/websites/fangirxoi/prev.htm butterfly picture tattoo unique ] [http://crea.html.it/websites/qreawpuya/prev.htm butterfly free gallery tattoo ] [http://crea.html.it/websites/cqdnlogad/prev.htm butterfly ink iron tattoo ] [http://crea.html.it/websites/nlrfdsor/prev.htm butterfly ink miami tattoo ] [http://crea.html.it/websites/doewns/prev.htm butterfly design flash tattoo ] [http://crea.html.it/websites/aylbpj/prev.htm butterfly fairy tattoo tribal ] [http://crea.html.it/websites/pbuqdae/prev.htm butterfly design picture tattoo ] [http://crea.html.it/websites/ucmlwa/prev.htm butterfly picture small tattoo ] [http://crea.html.it/websites/hqscoxo/prev.htm butterfly design floral tattoo ] [http://crea.html.it/websites/osemscbl/prev.htm picture of tribal tattoo ] [http://crea.html.it/websites/pfwgx/prev.htm tribal art tattoo picture ] [http://crea.html.it/websites/ymwsqp/prev.htm tribal sun tattoo picture ] [http://crea.html.it/websites/llslfhcn/prev.htm upper back tribal tattoo ] [http://crea.html.it/websites/wkxdbpyou/prev.htm behind neck tattoo tribal ] [http://crea.html.it/websites/ajmreiv/prev.htm tribal armband tattoo picture ] [http://crea.html.it/websites/nvjzqars/prev.htm free tribal tattoo flash ] [http://crea.html.it/websites/ralsd/prev.htm tribal dragon picture tattoo ] [http://crea.html.it/websites/igxeiof/prev.htm half sleeve tribal tattoo ] [http://crea.html.it/websites/rsewfufg/prev.htm sea turtle tribal tattoo ] [http://crea.html.it/websites/pyshxd/prev.htm american native tribal tattoo ] [http://crea.html.it/websites/zcqljofi/prev.htm tribal body art tattoo ] [http://crea.html.it/websites/afutfwhsw/prev.htm free tribal cross tattoo ] [http://crea.html.it/websites/dganvd/prev.htm free tribal tattoo art ] [http://crea.html.it/websites/iqpioqvgq/prev.htm tribal sun tattoo pic ] [http://crea.html.it/websites/jftajmbx/prev.htm tribal arm tattoo picture ] [http://crea.html.it/websites/iynzp/prev.htm tribal cross tattoo pic ] [http://crea.html.it/websites/fuvyj/prev.htm band pacific tattoo tribal ] [http://crea.html.it/websites/ryukqx/prev.htm heart tribal tattoo picture] </div>== Problem overview ==When multiple XPCOM objects form a cycle but are otherwise disconnected from any live roots (pointers not embedded in XPCOM objects that a thread can find transitively from its static and local variables, e.g. local nsCOMPtr variables) it is considered a garbage cycle. Currently XPCOM cannot collect garbage cycles. This is a proposal to produce a collector for such cycles.Common sources of garbage cycles in XPCOM appear to be cycles between javascript objects and browser objects, particularly DOM objects.== Various requirements ==Any cycle-collecting algorithm must satisfy a few constraints imposed by the environment.* Some objects will never be upgraded to participate in cycle collection. The mechanism must not break if it is applied to only a subset of the objects in the graph.* XPCOM objects cannot generally be freed in any consistent fashion, such as by calling "operator delete". Objects should be made to "self destruct" by having all their ''incoming'' edges drop.* Adding gratuitous new interfaces, vtables, or pointer state is probably too expensive.== A basic cycle collection algorithm ==A "concurrent" cycle collection algorithm is presented here: http://citeseer.ist.psu.edu/paz03fly.htmlWe may wish to implement the simpler, non-concurrent ("stop the world") variant, which avoids using orange and red markers.== A sketch of an implementation in XPCOM ==* [http://venge.net/graydon/mozilla/nsICycleCollector.idl nsICycleCollector.idl]* [http://venge.net/graydon/mozilla/nsIClassInfo2.idl nsIClassInfo2.idl]* [http://venge.net/graydon/mozilla/nsCycleCollector.cpp nsCycleCollector.cpp]* [http://venge.net/graydon/mozilla/nsISupportsImpl.h nsISupportsImpl.h changes]Some notes on the implementation:* The implementation requires that each participating XPCOM class implement <tt>nsIClassInfo2</tt>. The two methods on this interface must be correct: it is better '''not''' to implement <tt>nsIClassInfo2</tt> than to implement it wrong. <tt>unlink(obj)</tt> should disconnect all outgoing XPCOM references held by <tt>obj</tt>. <tt>traverse(obj,refcount,childcount,children)</tt> should return the refcount for <tt>obj</tt>, as well as the count and an array of outgoing XPCOM references from <tt>obj</tt>.* An object implementing <tt>nsIClassInfo2</tt> should also use the <tt>NS_IMPL_CYCLE_COLLECTING_</tt> macros in the <tt>nsISupportsImpl</tt> header; these implementations hook the appropriate refcount operations to communicate with the global cycle collector. The implementation commandeers a single bit from the refcount stored in <tt>nsAutoRefCnt</tt>. This is not strictly necessary, but should cut down significantly on pointless traffic to the cycle collector.* The global cycle collector service is kept as a singleton pointer in <tt>nsAutoRefCnt</tt>. This poses potential problems during shutdown. Advice on how to handle this safely would be appreciated.

	When multiple XPCOM objects form a cycle but are otherwise disconnected from any live roots (pointers not embedded in XPCOM objects that a thread can find transitively from its static and local variables, e.g. local nsCOMPtr variables) it is considered a garbage cycle. Currently XPCOM cannot collect garbage cycles. This is a proposal to produce a collector for such cycles.

	Common sources of garbage cycles in XPCOM appear to be cycles between javascript objects and browser objects, particularly DOM objects.

	== Various requirements ==

	Any cycle-collecting algorithm must satisfy a few constraints imposed by the environment.

	* Some objects will never be upgraded to participate in cycle collection. The mechanism must not break if it is applied to only a subset of the objects in the graph.
	* XPCOM objects cannot generally be freed in any consistent fashion, such as by calling "operator delete". Objects should be made to "self destruct" by having all their ''incoming'' edges drop.
	* Adding gratuitous new interfaces, vtables, or pointer state is probably too expensive.

	== A basic cycle collection algorithm ==

	A "concurrent" cycle collection algorithm is presented here: http://citeseer.ist.psu.edu/paz03fly.~~html~~

	We may wish to implement the simpler, non-concurrent ("stop the world") variant, which avoids using orange and red markers.

	== A sketch of an implementation in XPCOM ==

	* [http://venge.net/graydon/mozilla/nsICycleCollector.idl nsICycleCollector.idl]
	* [http://venge.net/graydon/mozilla/nsIClassInfo2.idl nsIClassInfo2.idl]
	* [http://venge.net/graydon/mozilla/nsCycleCollector.cpp nsCycleCollector.cpp]
	* [http://venge.net/graydon/mozilla/nsISupportsImpl.h nsISupportsImpl.h changes]

	Some notes on the implementation:

	* The implementation requires that each participating XPCOM class implement <tt>nsIClassInfo2</tt>. The two methods on this interface must be correct: it is better '''not''' to implement <tt>nsIClassInfo2</tt> than to implement it wrong.
	** <tt>unlink(obj)</tt> should disconnect all outgoing XPCOM references held by <tt>obj</tt>.
	** <tt>traverse(obj,refcount,childcount,children)</tt> should return the refcount for <tt>obj</tt>, as well as the count and an array of outgoing XPCOM references from <tt>obj</tt>.
	* An object implementing <tt>nsIClassInfo2</tt> should also use the <tt>NS_IMPL_CYCLE_COLLECTING_*</tt> macros in the <tt>nsISupportsImpl</tt> header; these implementations hook the appropriate refcount operations to communicate with the global cycle collector.
	* The implementation commandeers a single bit from the refcount stored in <tt>nsAutoRefCnt</tt>. This is not strictly necessary, but should cut down significantly on pointless traffic to the cycle collector.
	* The global cycle collector service is kept as a singleton pointer in <tt>nsAutoRefCnt</tt>. This poses potential problems during shutdown. Advice on how to handle this safely would be appreciated.

Brendan: /* Problem overview */

2006-02-24T00:46:21Z

Problem overview

← Older revision		Revision as of 00:46, 24 February 2006
Line 1:		Line 1:
	== Problem overview ==		== Problem overview ==

	When multiple XPCOM objects form a cycle but are otherwise disconnected from any live roots (pointers ~~which~~ a thread can find transitively from its static and local variables) it is considered a garbage cycle. Currently XPCOM cannot collect garbage cycles. This is a proposal to produce a collector for such cycles.		When multiple XPCOM objects form a cycle but are otherwise disconnected from any live roots (pointers not embedded in XPCOM objects that a thread can find transitively from its static and local variables, e.g. local nsCOMPtr variables) it is considered a garbage cycle. Currently XPCOM cannot collect garbage cycles. This is a proposal to produce a collector for such cycles.

	Common sources of garbage cycles in XPCOM appear to be cycles between javascript objects and browser objects, particularly DOM objects.		Common sources of garbage cycles in XPCOM appear to be cycles between javascript objects and browser objects, particularly DOM objects.

	== Various requirements ==		== Various requirements ==

Brendan: /* Variouis requirements */

2006-02-24T00:44:23Z

Variouis requirements

← Older revision		Revision as of 00:44, 24 February 2006
Line 5:		Line 5:
	Common sources of garbage cycles in XPCOM appear to be cycles between javascript objects and browser objects, particularly DOM objects.		Common sources of garbage cycles in XPCOM appear to be cycles between javascript objects and browser objects, particularly DOM objects.

	== ~~Variouis~~ requirements ==		== Various requirements ==

	Any cycle-collecting algorithm must satisfy a few constraints imposed by the environment.		Any cycle-collecting algorithm must satisfy a few constraints imposed by the environment.

Graydon: /* A sketch of an implementation in XPCOM */

2006-02-24T00:08:45Z

A sketch of an implementation in XPCOM

← Older revision		Revision as of 00:08, 24 February 2006
Line 24:		Line 24:
	* [http://venge.net/graydon/mozilla/nsIClassInfo2.idl nsIClassInfo2.idl]		* [http://venge.net/graydon/mozilla/nsIClassInfo2.idl nsIClassInfo2.idl]
	* [http://venge.net/graydon/mozilla/nsCycleCollector.cpp nsCycleCollector.cpp]		* [http://venge.net/graydon/mozilla/nsCycleCollector.cpp nsCycleCollector.cpp]
			* [http://venge.net/graydon/mozilla/nsISupportsImpl.h nsISupportsImpl.h changes]

			Some notes on the implementation:

			* The implementation requires that each participating XPCOM class implement <tt>nsIClassInfo2</tt>. The two methods on this interface must be correct: it is better '''not''' to implement <tt>nsIClassInfo2</tt> than to implement it wrong.
			** <tt>unlink(obj)</tt> should disconnect all outgoing XPCOM references held by <tt>obj</tt>.
			** <tt>traverse(obj,refcount,childcount,children)</tt> should return the refcount for <tt>obj</tt>, as well as the count and an array of outgoing XPCOM references from <tt>obj</tt>.
			* An object implementing <tt>nsIClassInfo2</tt> should also use the <tt>NS_IMPL_CYCLE_COLLECTING_*</tt> macros in the <tt>nsISupportsImpl</tt> header; these implementations hook the appropriate refcount operations to communicate with the global cycle collector.
			* The implementation commandeers a single bit from the refcount stored in <tt>nsAutoRefCnt</tt>. This is not strictly necessary, but should cut down significantly on pointless traffic to the cycle collector.
			* The global cycle collector service is kept as a singleton pointer in <tt>nsAutoRefCnt</tt>. This poses potential problems during shutdown. Advice on how to handle this safely would be appreciated.

Graydon: /* A sketch of an implementation in XPCOM */

2006-02-23T22:45:13Z

A sketch of an implementation in XPCOM

← Older revision		Revision as of 22:45, 23 February 2006
Line 21:		Line 21:
	== A sketch of an implementation in XPCOM ==		== A sketch of an implementation in XPCOM ==

	* [http://venge.net/graydon/mozilla/nsICycleCollector.idl\|nsICycleCollector.idl]		* [http://venge.net/graydon/mozilla/nsICycleCollector.idl nsICycleCollector.idl]
	* [http://venge.net/graydon/mozilla/nsIClassInfo2.idl\|nsIClassInfo2.idl]		* [http://venge.net/graydon/mozilla/nsIClassInfo2.idl nsIClassInfo2.idl]
	* [http://venge.net/graydon/mozilla/nsCycleCollector.cpp\|nsCycleCollector.cpp]		* [http://venge.net/graydon/mozilla/nsCycleCollector.cpp nsCycleCollector.cpp]

Graydon at 22:43, 23 February 2006

2006-02-23T22:43:52Z

← Older revision		Revision as of 22:43, 23 February 2006
Line 21:		Line 21:
	== A sketch of an implementation in XPCOM ==		== A sketch of an implementation in XPCOM ==

	* [http://venge.net/graydon/mozilla/~~nsCycleCollector~~.~~cpp~~\|nsICycleCollector.idl]		* [http://venge.net/graydon/mozilla/nsICycleCollector.idl\|nsICycleCollector.idl]
	* [http://venge.net/graydon/mozilla/~~nsCycleCollector~~.~~cpp~~\|nsIClassInfo2.idl]		* [http://venge.net/graydon/mozilla/nsIClassInfo2.idl\|nsIClassInfo2.idl]
	* [http://venge.net/graydon/mozilla/nsCycleCollector.cpp\|nsCycleCollector.cpp]		* [http://venge.net/graydon/mozilla/nsCycleCollector.cpp\|nsCycleCollector.cpp]