corysanin/tubestation
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
386,677 Commits 1 Branch 0 Tags
cb15aec7a609e4fed2c9b83d10a510addea64ccd
Commit Graph

591 Commits

Author SHA1 Message Date
Andrew McCreight
c4c5f5b0da Bug 935721, part 8 - Add a reference to the graph to the graph builder. r=smaug 
This moves towards letting the CC graph outlive the builder.
 2013-11-20 14:35:16 -08:00
Andrew McCreight
82a05a07bb Bug 935721, part 7 - Inline PrepareForCollection into BeginCollection. r=smaug 
PrepareForCollection is trivial now, so just inline it.
 2013-11-20 14:35:16 -08:00
Andrew McCreight
8a976d93c4 Bug 935721, part 5 - Invert the control flow of CycleCollectNow's pre- and post-collection work, add CCResults as a field on the CC. r=smaug 
nsJSEnvironment::CycleCollectNow does work before and after a CC runs. With ICC, nsJSEnv won't
know where in the CC when a CC is about to begin or end, so this patch reorganizes that work
into two separate callback hooks.  This requires adding a new struct, CycleCollectorStats, to
hold data nsJSEnv needs between the two calls.

Rather than trying to pass around a pointer to a results structure, this patch just adds
it to the nsCycleCollector struct, and always stores them. The results are passed back
to the end CC callback.
 2013-11-20 14:35:16 -08:00
Andrew McCreight
9c498f22b5 Bug 935721, part 4 - Allocate white nodes array in CollectWhite. r=smaug 
The white nodes array is only used in CollectWhite, so just allocate it there.
 2013-11-20 14:35:16 -08:00
Andrew McCreight
e0d56fab4e Bug 935721, part 3 - Use RAII to set mScanInProgress, hoist out MarkRoots and ScanRoots. r=smaug 
With ICC, mScanInProgress gets set and cleared a bunch of times so add an RAII class to turn
it on when we're doing stuff and clear it when we're not.

With that in place, we can easily move MarkRoots and ScanRoots out of BeginCollection
in preparation for making them separate phases in ICC.
 2013-11-20 14:35:16 -08:00
Andrew McCreight
afba6b88ca Bug 935721, part 2 - Allocate GCGraphBuilder on the heap. r=smaug 
With ICC, the graph builder must persist across invocations of the CC,
so store it on the heap.
 2013-11-20 14:35:15 -08:00
Andrew McCreight
8ffcfabad4 Bug 935721, part 1 - Store the listener pointer on the nsCycleCollector data structure. r=smaug 
With ICC, the listener must persist across invocations of the CC, so store it on the CC.
 2013-11-20 14:35:15 -08:00
Benjamin Smedberg
29d63f056d Bug 672843 part C - convert most of XPCOM except for xpcom/tests, r=vdjeric 2013-11-19 16:27:37 -05:00
Nicholas Nethercote
d95d74cdd5 Bug 940069 - Land a temporary debugging printf that runs on --enable-valgrind builds only, because Valgrind runs aren't available on tryserver. r=me. 2013-11-18 23:13:15 -08:00
Olli Pettay
ba341e43bd Bug 937303 - Make CC logger to pass information about compartments, r=mccr8 2013-11-12 15:53:51 +02:00
Wes Kocher
6e75ecb1e5 Backed out changeset 3d769520a0be (bug 937303) for build bustage 2013-11-11 14:50:00 -08:00
Olli Pettay
1e312ffe74 Bug 937303 - Make CC logger to pass information about compartments, r=mccr8 2013-11-12 00:12:00 +02:00
Olli Pettay
ef62df780a Bug 936789 - Faster CC logger, r=mccr8 2013-11-09 23:15:44 +02:00
Andrew McCreight
1f73b6a41c Bug 934568 - Don't make ScanRoots CC OOM assertion fatal due to Win7 debug M2 failures. s=smaug 2013-11-06 16:58:13 -08:00
Nicholas Nethercote
5c26d355e0 Bug 934321 - Add MemoryMultiReporter, a helper class that reduces some boilerplate, and convert all existing multi-reporters to use it. r=wchen. 2013-11-06 14:58:20 +11:00
Benoit Jacob
85b7664705 Bug 934568 - Make cycle collector OOM assertions fatal in debug builds. r=mccr8 2013-10-27 07:51:31 -04:00
Jonas Finnemann Jensen
cf099c0754 Bug 902587 - Part 2A: Refactor late-write-checks as client of IO Interposer. r=BenWa 2013-11-05 07:45:20 -05:00
Andrew McCreight
2a8aa49cb4 Bug 928647 - Remove CycleCollectedRuntime::mObjectToUnlink. r=smaug 2013-10-19 08:59:10 -07:00
Olli Pettay
9ff32ee448 Bug 927813 - Some micro-optimizations for Suspect, r=mccr8 2013-10-17 17:05:13 +03:00
Birunthan Mohanathas
52e24d677b Bug 784739 - Switch from NULL to nullptr in xpcom/ (1/3); r=ehsan 2013-10-10 16:41:00 -04:00
Thinker Lee
d4c71839a3 Bug 771765 - Support template content process, part 6: support re-creation of the threads created in the template process. r=khuey, r=jorendorff 
The threads that are frozen/recreated include:
* ImageBridgeChildThread.
* Image decoding thread pool.
* IPC thread (checkpointed, but not frozen).
* GC Helper thread.
* XPC runtime watchdog thread.
* Socket transport service thread/thread pool.
* Memory pressure watcher.
* Timer thread.
* DOM promise worker.
 2013-06-03 18:14:42 +08:00
Andrew McCreight
dfac8e112b Bug 915488 - Make CC participant's Root, Unroot and Unlink methods infallible. r=smaug 2013-09-11 18:57:53 -07:00
Ehsan Akhgari
35f4093831 Bug 917885 - Minimize the #includes in xpcom/base; r=bsmedberg 2013-09-19 14:29:31 -04:00
Ehsan Akhgari
fe85b6b42b Backed out changeset a8d6973e5743 (bug 917885) because I pushed the wrong patch 2013-09-19 14:27:35 -04:00
Ehsan Akhgari
4bbb07e34d Bug 917885 - Minimize the #includes in xpcom/base; r=bsmedberg 2013-09-19 14:12:56 -04:00
Nicholas Nethercote
c7fbf6a742 Bug 910517 (3rd attempt) - Remove nsIMemoryReporter, and rename nsIMemoryMultiReporter as nsIMemoryReporter. r=mmcr8. 2013-08-27 16:24:51 -07:00
Andrew McCreight
5d415564c4 Bug 913666, part 5 - Rename aListener to aManualListener. r=smaug 
The name "aListener" is not very descriptive, and with the previous patch, it is only
used to pass in a manually-specified listener that was passed in to CycleCollectNow,
so rename things.
 2013-09-10 16:33:41 -07:00
Andrew McCreight
0b7e317041 Bug 913666, part 4 - Sink free floating code from nsCycleCollector::Collect into BeginCollection. r=smaug 
The hg diff for this commit is terrible, but all it is doing is taking the code
in nsCycleCollector::Collect from after PrepareForCollection through BeginCollection
and moving it into BeginCollection.
 2013-09-10 16:33:41 -07:00
Andrew McCreight
07a7f3544c Bug 913666, part 3 - Sink cycle collector listener selection into Collect. r=smaug 
Move the two places we check global flags to decide if we want to use a listener, even if we
aren't passed in one.  A later patch renames aListener to aForcedListener to make it less
confusing.
 2013-09-10 16:33:40 -07:00
Andrew McCreight
3546517893 Bug 913666, part 2 - Log cycle collector garbage at the same time we log roots. r=smaug 
There's no reason to wait until CollectWhite to record what is garbage, and with incremental CC
we'll get more accurate logging by logging right away, rather than waiting until later when an
object may have gone away for some other reason.
 2013-09-10 16:33:40 -07:00
Andrew McCreight
b433b40e12 Bug 913666, part 1 - Move cycle collector root logging into ScanRoots. r=smaug 2013-09-10 16:33:39 -07:00
Andrew McCreight
902685affd Bug 913881, part 6 - Measure weak map representations in the cycle collector graph. r=njn 
Contrary to the comment here, I'm pretty sure this needs to be measured. mWeakMaps
is an array of little structs. Of course, in practice I doubt this amounts to anything.
 2013-09-10 08:56:36 -07:00
Andrew McCreight
bda99d54a4 Bug 913881, part 5 - Assert for freeing snow white during cycle collector scanning. r=smaug 
If you call FreeSnowWhite when scanning is true, then objects that get released won't be added to the purple buffer, and you can get leaks.
 2013-09-10 08:56:36 -07:00
Andrew McCreight
d5271b9c36 Bug 913881, part 4 - Change nsCycleCollector::ClearGraph to GCGraph::Clear. r=smaug 
There's no reason I can see to have this in nsCC instead of GCGraph.
 2013-09-10 08:56:36 -07:00
Andrew McCreight
601ac68d0e Bug 913881, part 3 - Fix argument name for GCGraphBuilder::AddNode. r=smaug 2013-09-10 08:56:35 -07:00
Andrew McCreight
9c520edff5 Bug 913881, part 2 - Make a bunch of nsCycleCollector methods private. r=smaug 2013-09-10 08:56:35 -07:00
Andrew McCreight
98f069b77a Bug 913881, part 1 - Remove unused method nsCycleCollector::Runtime. r=smaug 2013-09-10 08:56:35 -07:00
Andrew McCreight
af411817cd Bug 913527, part 3 - Check for collection in progress in Collect. r=smaug 
Moving this to the top level simplifies dealing with PrepareForCollection a bit.
 2013-09-10 08:56:34 -07:00
Andrew McCreight
33e35099da Bug 913527, part 2 - Remove empty CC graph optimization. r=smaug 
This looks like a dubious optimization to skip most of a CC when the graph
is empty, dating from the dawn of the CC, but I doubt it is ever triggered
nowadays.
 2013-09-10 08:56:34 -07:00
Andrew McCreight
51892e4031 Bug 913527, part 1 - Get rid of FinishCollection. r=smaug 
This can just be inlined now, with the cleanup parts moved into CleanupAfterCollection.
 2013-09-10 08:56:34 -07:00
Andrew McCreight
f676b3681f Bug 911829 - Separate main thread and worker cycle collector telemetry. r=smaug 2013-09-10 08:29:45 -07:00
Kyle Huey
72ac69366e Bug 911258: Part 4 - Refactor exception implementations. r=bz 2013-09-08 20:28:50 -07:00
Andrew McCreight
360af07238 Bug 913130, part 4 - Replace body of ShutdownCollect loop with Collect. r=smaug 2013-09-06 16:17:53 -07:00
Andrew McCreight
23ae22693a Bug 913130, part 3 - Run {PrepareFor,CleanupAfter} every iteration of ShutdownCollect. r=smaug 2013-09-06 16:17:03 -07:00
Andrew McCreight
1c7df7e692 Bug 913130, part 2 - Give Collect and ShutdownCollect more consistent ordering. r=smaug 2013-09-06 16:16:45 -07:00
Andrew McCreight
1c95521d7c Bug 913130, part 1 - Modify nsCycleCollector::Collect to allow it to be used at shutdown. r=smaug 2013-09-06 16:15:10 -07:00
Andrew McCreight
6c46f9bae3 Bug 913080 - Make GCGraphBuilder's hash table initialization infallible. r=smaug 2013-09-06 13:42:27 -07:00
Andrew McCreight
5d5dcd7aa3 Bug 911233, part 3 - Add more CheckThreadSafety calls to the cycle collector. r=khuey 2013-09-06 13:41:42 -07:00
Andrew McCreight
077a3b9691 Bug 911233, part 2 - Inline SelectPurple. r=smaug 2013-09-06 13:41:11 -07:00
Andrew McCreight
ea2fa6cd25 Bug 911233, part 1 - Minor cycle collector cleanups. r=smaug 2013-09-06 13:40:34 -07:00