aboutsummaryrefslogtreecommitdiffstatshomepage
path: root/OpenSim/Region/ClientStack (unfollow)
Commit message (Collapse)AuthorFilesLines
2009-10-09Slight variation on the locking scheme: now locking always, except the ↵Diva Canto1-9/+9
ForEach, which gets a copy of the Array. I think the ForEach was the big gorilla.
2009-10-09This is to *try* to avoid the exception we are seeing repeatedly:Diva Canto1-0/+4
#13:53:03 - [LLUDPSERVER]: Packet exceeded buffer size during zerocoding. Removing MSG_ZEROCODED flag #13:53:03 - [Scene]: Failed with exception System.ArgumentException: Offset and length were out of bounds for the array or count is greater than the number of elements from index to the end of the source collection. # at System.Buffer.BlockCopy (System.Array src, Int32 srcOffset, System.Array dst, Int32 dstOffset, Int32 count) [0x000b7] in /home/grid/install/mono24/mono-2.4.2.3/mcs/class/corlib/System/Buffer.cs:116 # at OpenSim.Region.ClientStack.LindenUDP.LLUDPServer.SendPacketData (OpenSim.Region.ClientStack.LindenUDP.LLUDPClient client, System.Byte[] data, Int32 dataLength, PacketType type, Boolean doZerocode, ThrottleOutPacketType category) [0x00060] in /home/grid/PLAZA/wrightplaza/opensim/OpenSim/Region/ClientStack/LindenUDP/LLUDPServer.cs:290 # at OpenSim.Region.ClientStack.LindenUDP.LLUDPServer.SendPacket (OpenSim.Region.ClientStack.LindenUDP.LLUDPClient client, OpenMetaverse.Packets.Packet packet, ThrottleOutPacketType category, Boolean allowSplitting) [0x00077] in /home/grid/PLAZA/wrightplaza/opensim/OpenSim/Region/ClientStack/LindenUDP/LLUDPServer.cs:265 # at OpenSim.Region.ClientStack.LindenUDP.LLClientView.OutPacket (OpenMetaverse.Packets.Packet packet, ThrottleOutPacketType throttlePacketType) [0x00000] in /home/grid/PLAZA/wrightplaza/opensim/OpenSim/Region/ClientStack/LindenUDP/LLClientView.cs:5244 # at OpenSim.Region.ClientStack.LindenUDP.LLClientView.SendCoarseLocationUpdate (System.Collections.Generic.List`1 users, System.Collections.Generic.List`1 CoarseLocations) [0x0012a] in /home/grid/PLAZA/wrightplaza/opensim/OpenSim/Region/ClientStack/LindenUDP/LLClientView.cs:3424 # at OpenSim.Region.Framework.Scenes.ScenePresence.SendCoarseLocationsDefault (UUID sceneId, OpenSim.Region.Framework.Scenes.ScenePresence p) [0x000fa] in /home/grid/PLAZA/wrightplaza/opensim/OpenSim/Region/Framework/Scenes/ScenePresence.cs:2536 # at OpenSim.Region.Framework.Scenes.ScenePresence.SendCoarseLocations () [0x0000d] in /home/grid/PLAZA/wrightplaza/opensim/OpenSim/Region/Framework/Scenes/ScenePresence.cs:2491 # at OpenSim.Region.Framework.Scenes.ScenePresence.Update () [0x00011] in /home/grid/PLAZA/wrightplaza/opensim/OpenSim/Region/Framework/Scenes/ScenePresence.cs:2404 # at OpenSim.Region.Framework.Scenes.SceneGraph.UpdateEntities () [0x0001b] in /home/grid/PLAZA/wrightplaza/opensim/OpenSim/Region/Framework/Scenes/SceneGraph.cs:164 # at OpenSim.Region.Framework.Scenes.Scene.Update () [0x000dc] in /home/grid/PLAZA/wrightplaza/opensim/OpenSim/Region/Framework/Scenes/Scene.cs:1031 On Region: Wright Plaza
2009-10-09Putting this to exactly what it was yesterday around this time, so we can do ↵Diva Canto2-182/+411
the 100-ppl load test in WP. We need to carefully play with this code in order to understand all the problems.
2009-10-08Putting the lock back in TryGetValue.Diva Canto2-349/+349
2009-10-08Fork UDPBase from libOMV into opensimMelanie2-1/+341
2009-10-08Added guards against the exception reported:Diva Canto1-0/+9
#2009-10-08 12:58:34,462 ERROR - OpenSim.Application [APPLICATION]: #APPLICATION EXCEPTION DETECTED: System.UnhandledExceptionEventArgs # #Exception: System.Exception: Object reference not set to an instance of an object. ---> System.NullReferenceException: Object reference not set to an instance of an object. # at OpenSim.Region.ClientStack.LindenUDP.J2KImage.SendFirstPacket(LLClientView client) in c:\Opensim\OpenSim\Region\ClientStack\LindenUDP\J2KImage.cs:line 230 # at OpenSim.Region.ClientStack.LindenUDP.J2KImage.SendPackets(LLClientView client, Int32 maxpack) in c:\Opensim\OpenSim\Region\ClientStack\LindenUDP\J2KImage.cs:line 84
2009-10-07One last attempt at tunning the locking/no locking behaviour. The previous ↵Diva Canto1-10/+16
one was too lax and made the important login packets fail.
2009-10-07Another go at the deadlock: not locking TryGetValue, and just catch/ignore ↵Diva Canto1-15/+31
exceptions.
2009-10-07Changed the locking mechanism, because the existing one is causing deadlocks ↵Diva Canto1-25/+100
to occur.
2009-10-07replace LLClientView with proper versionMelanie1-1898/+1332
2009-10-07Replace LLClientView on the branch with master. UNCOMPILABLE!Melanie1-1332/+1898
2009-10-07Revert "Rewrote parts of the code that were double-locking different ↵Melanie1-69/+47
objects. This is about half of the code base reviewed." This reverts commit e992ca025571a891333a57012c2cd4419b6581e5.
2009-10-07Revert "Merging in diva's locking fixes"Melanie1-69/+47
This reverts commit 832cc685138b2244529f10b54b373c34adb4a633.
2009-10-06Commented noisy debugging about packet splittingJohn Hurliman1-4/+4
2009-10-06Rewrote parts of the code that were double-locking different objects. This ↵Diva Canto1-47/+69
is about half of the code base reviewed.
2009-10-06Applying the real fix (thank you for tracking that MSDN doc down Melanie)John Hurliman1-5/+6
2009-10-06Trying Melanie's fixJohn Hurliman1-3/+4
2009-10-06* Added a sanity check for Mono before trying to enumerate over an empty ↵John Hurliman1-0/+3
SortedDictionary * Changed the order of a log line from DotNetEngine so you can tell whether or not it is actually loading
2009-10-06* Added missing references to prebuild.xml and commented out the LindenUDP ↵John Hurliman7-195/+195
tests until a new test harness is written * Clients are no longer disconnected when a packet handler crashes. We'll see how this works out in practice * Added documentation and cleanup, getting ready for the first public push * Deleted an old LLUDP file
2009-10-06* Try/catch around EndInvoke() when Util.FireAndForget() returns to catch ↵John Hurliman3-78/+112
exceptions thrown in the async method * Added packet stats handling to the new LLUDP implementation * Attempting to avoid a race condition when creating a new LLUDPClient
2009-10-06* Continued work on the new LLUDP implementation. Appears to be functioning, ↵John Hurliman13-2116/+1028
although not everything is reimplemented yet * Replaced logic in ThreadTracker with a call to System.Diagnostics that does the same thing * Added Util.StringToBytes256() and Util.StringToBytes1024() to clamp output at byte[256] and byte[1024], respectively * Fixed formatting for a MySQLAssetData error logging line
2009-10-05Beginning work on the new LLUDP implementationJohn Hurliman12-2328/+1815
2009-10-05Changed the lock object, to see if this helps at all. Also switched the ↵Diva Canto1-11/+15
order of lock and try/catch in RemoveImageFromQueue.
2009-10-04Reduced locking.Diva Canto1-8/+10
2009-10-04Guarding a line that is sometimes throwing a null pointer exception.Diva Canto1-5/+6
2009-10-04Trying to avoid the null pointer exception in J2KImage, RunUpdate.Diva Canto1-1/+9
2009-10-02* Creates Util.UTF8 and switches some references of Encoding.UTF8 to ↵John Hurliman1-2/+2
Util.UTF8 (not all references were switched since not all OpenSim libraries reference OpenSim.Framework) * Shrinks the largest in-memory object, the LLRAW.HeightmapLookupValue struct (only used for exporting to LLRAW terrain files), to the minimum possible size. This seems to have the odd side effect of cutting the size of the two double[256,256] terrain objects in half. Possibly an alignment optimization?
2009-10-02* Removed code that is no longer used with Melanie's queue callback optimizationJohn Hurliman1-126/+88
* Moved all priority queue access to helper functions to improve reliability and readability * New version of CSJ2K.dll (from libomv source tree) that fixes a JPEG2000 comment decoding bug
2009-10-02* Removed even more unnecessary state variables from J2KImageJohn Hurliman2-182/+142
* Fixed the incorrect "out of memory?" exception messages and replaced them with useful information * Cleaned up J2KImage and LLImageManager to conform to coding conventions
2009-10-02* Changed the flush logic to drop packets in non-transactional streams, and ↵John Hurliman3-59/+19
to not fire any "put more data in the queues" callbacks * Minor tweaks to code formatting to make the callback chain for packet queuing easier to follow
2009-10-02Restore the missing image handling to the image manager. The missingMelanie3-71/+92
image packet crashes Hippo without a message.
2009-10-01Locking the priority queue to see if it gets pass that exception.Diva Canto1-7/+28
2009-10-02Change texture sending to be driven by the queue empty event from theMelanie4-90/+60
packet queue, rather than a timer
2009-10-01Removing console debug spewJohn Hurliman1-7/+7
2009-10-01* Rewrote LLImageManager to use a real priority queue and hold minimal stateJohn Hurliman2-448/+331
* Rewrote the logic in J2KImage.RunUpdate() * Added a default avatar texture (I made it myself)
2009-10-01Add a method to determine the count of packets in a throttleMelanie3-0/+27
2009-10-01Add OnQueueEmpty event to the packet layers. No user functinality yetMelanie3-2/+66
2009-10-01Fixing LLClientView memory leakJames J Greensky4-8/+29
Fixing LLClientView memory leak by disposing of all timers utilized in LLClientView as they contain references to the callback method. This required the use of the Terminate and Close infrastructure that was already in place but was not being utilized.
2009-10-01Formatting cleanup.Jeff Ames5-8/+8
2009-09-30* Change the signature of the agent set appearance callback to prevent ↵John Hurliman1-6/+8
unnecessary serialization/deserialization of TextureEntry objects and allow TextureEntry to be inspected for missing bakes * Inspect incoming TextureEntry updates for bakes that do not exist on the simulator and request the missing textures * Properly handle appearance updates that do not have a TextureEntry set
2009-09-30Prevent null data being sent to the decoderMelanie1-2/+5
2009-09-30Add RebakeAvatarTexturesPacket to the client viewMelanie1-0/+10
2009-09-30Squelch the debug noise from diva's patchMelanie1-1/+1
2009-09-30Attempting to improve the robustness of texture decoding by always ignoring ↵John Hurliman1-1/+1
LayerInfo.End values and creating guessed default layer boundaries on failed decodes Changed a noisy J2K decode log message from Info to Debug Replacing openjpeg-dotnet decoding with managed CSJ2K decoding. Should be much more reliable, faster, and use less memory * Re-added openjpeg-dotnet files since they are used elsewhere in OpenSim * Updated prebuild.xml with a reference to CSJ2K * Renamed IJ2KDecoder and J2KDecoder member names to follow standard naming conventions * Removed j2kDecodeCache cruft and replaced it with the OpenSim cache system * Rewrote the default layer boundary algorithm to use percentages instead of an exponent * Switched from an infinite in-memory cache to an expiring cache (10 minute timeout) * Slightly quieted logging errors for failed texture decodes
2009-09-30This releases the texture assets from LLImageManager cache, and re-requests ↵Diva Canto2-12/+48
them later if the client asks for them again. Needs more testing in texture-rich sims.
2009-10-01Formatting cleanup.Jeff Ames16-82/+82
2009-09-27Unpacking the mess with OtherRegionUp, so we can have a real cache of the ↵Diva Canto1-1/+2
neighbours in the grid service modules.
2009-09-23* Makes SimulatorEnable messages 'reliable' and subject to redelivery.Adam Frisby1-0/+3
2009-09-10Formatting cleanup.Jeff Ames1-1/+1
2009-09-09* Lock timers when Calling Start() and Stop() when the Thread Context is ↵Teravus Ovares (Dan Olivares)1-20/+56
murky. This affects Mono only.
generated by cgit v1.1 at 2024-11-01 14:20:01 +0000