This is to help relieve a race condition when an agent teleports then immediately attempts to teleport back before the source region has properly cleaned up/demoted the old ScenePresence.
This is rare in viewers but much more possible via scripting or region module.
However, more needs to be done since virtually all clean up happens after the transit flag is cleared .
Possibly need to add a 'cleaning up' state to in transit.
This change required making the EntityTransferModule and HGEntityTransferModule per-region rather than shared, in order to allow separate transit lists.
Changes were also required in LocalSimulationConnector.
Tested in standalone, grid and with local and remote region crossings with attachments.
Since this is done directly from ScenePresence, it can lead to a race condition with the simulator loop.
There's no real point doing it anyway since the clear will be done very shortly afterwards by the simulate loop and either there are no events (for a new avatar) or events don't matter (for a departing avatar).
This matches existing behaviour in OdePrim
From sl docs such as http://community.secondlife.com/t5/English-Knowledge-Base/Managing-Private-Regions/ta-p/700115
agent should apply to avatars only.
This makes sense from a user perspective, and also from a code perspective since child agents with no physics or actions take up a fraction of root agent resources.
As such, the check is now only performed in Scene.QueryAccess() - cross and teleport check this before allowing an agent to translocate.
This also removes an off-by-one error that could occur in certain circumstances on teleport when a new child agent was double counted when a pre-teleport agent update was performed.
This does not affect an existing bug where limits or other QueryAccess() checks are not applied to avatars logging directly into a region.
This includes prim count, script count, avatar count, etc.
Information is currently the same as "show stats", though show stats can only show one scene at a time because it listens for the latest outgoing stats packet (a bad approach that needs to change).
Might be better to tie this module into the other stats module to display arbitrary stats rather than fetching directly from scene.SimStatsReporter.
Console command is "show scene" because "show region" already exists for the grid service, which is unfortunate.
Might need to make a distinction between "scene" relating to a live scene and "region" relating to more static region data (url, coords, etc.)
The reference is not guaranteed to be there when AddRegion() is called but will definitely be present at RegionLoaded() if it's going to be present at all.
This is because the attachments module code was setting the 'object slam' bit by using PermissionMask.All
Solution here is to route the attachment item creation call through the existing inventory code in BasicInventoryAccessModule rather than copy/pasted code in AttachmentsModule itself.
This is to eliminate possible race conditions if two teleport calls are made concurrently, where at least one is a local teleport.
This is pretty much impossible on a manual user teleport but can happen on script-invoked teleports.
Since this is done directly from ScenePresence, it can lead to a race condition with the simulator loop.
There's no real point doing it anyway since the clear will be done very shortly afterwards by the simulate loop and either there are no events (for a new avatar) or events don't matter (for a departing avatar).
This matches existing behaviour in OdePrim
This was because the parts in scene objects were sometimes not serialized in link order.
This is perfectly fine since the parts still have the right link numbers, but an extra fix to adjust for this
had not been done in the SerialiserModule methods that OAR loading used.
Add regression test for same.
Addresses http://opensimulator.org/mantis/view.php?id=5948, http://opensimulator.org/mantis/view.php?id=5749
On EntityTransferModule.DoTeleport() there was an IsInTransit() check to prevent multiple simultaneous teleport attempts.
However, the SetInTransit() was only performed later on, which left a window in which multiple threads could pass the IsInTransit() check.
This has been seen in the field and the results aren't pretty.
This commit effectively combines the IsInTransit() and SetInTransit() checks so there is no such window.
More failure cases are made to to call ResetInTransit() to adjust to this move.
This does increase the aggressiveness of shutdown
Also prevents the bot list being locked for a long period, which was preventing commands such as "show bots" from working during shutdown
Still subject to change - if you use this be prepared to change your output code if/when the methods change.
Make new "attachments show" command use this.
If AvatarService appearance data is retrieved over the network, then ServerUtils was attempting to transfer names such as "Wearable 0:0" directly to xml element names, resulting in an exception.
Space is not valid in xml element names. Neither is : in this case since the intention is not to namespace. Using names directly as keys is not a good idea.
To get around this problem this patch escapes and unescapes the element names as appropriate.
This has no impact on existing xml (since it had to be valid in the first place) but allows AvatarService data to be used over the network.
Setting appearance (from simulator to AvatarService) did not suffer this problem since the values are passed in the query string which is already properly escaped.
DuplicateObject() relies on source object having correct link numbers for the duration of the dupe.
Both link and delink can change link numbers such that they are not consistent for short periods of time.
This may be the wrong thing to do but stops lots of log spam in HG setups now that the monitoring is extended to other outgoing calls.
LongCallTime may need to be made configurable.
Viewers 1 and 3 are fine with doing this immediately. However, Imprudence has a small delay (<200ms, >500ms) after receiving the AgentCompleteMovement reply packet on the destination region before regarding that region as the currnet region.
If Imprudence receives a DisableSimulator in this period, it quits.
We are not restoring the full 5000ms delay since this brings back a bug where teleports permanently fail if an avatar tries to teleport back too quickly.
This commit also sends the AgentCompleteMovement packet to the client before telling the source region to release its old agent, in order to further cut down any possibility of the DisableSimulator being recieved before the AgentMovementComplete.
This sleep appears unnecessary since a sleep has already occurred in WaitForCallback() whilst waiting for the destination region to notify of teleport success.
There are no async operations between this sleep and the WaitForCallback()
If this sleep is present, then teleporting back to the source region within 5 seconds results in a disconnection.
If this sleep is commented out then teleporting quickly back and forth between two simulators appears to work without issue.
Tested on standalone, local grid and distributed grid.
Please revert if there's something that I've missed.
This adds a non-advertised wait_for_callback option in [EntityTransfer]. Default is always true.
Teleport tests disable the wait for callback from the destination region in order to run within a single thread.
These are identical apart from setting Velocity = zero, which has no practical effect anyway since this is zeroed when the avatar is added back to the physics scene.
This was an oversight when removing some race conditions from PhysicsActor setting recently.
Regression tests extended to probe this code path.
Extending regression tests required implementation of a BasicPhysicsPrim (there was none before). However, BasicPhysics plugin is still of no current practical use other than to fill in as a component for other parts of regression testing.
On region restart, the whole object would become physical as expected.
Observed behaviour from elsewhere is that all prims in a new linkset should take on the status of the root prim.
Add regression test for this behaviour.
prim_geom == IntPtr.Zero only before a new add prim taint is processed (which is the first taint) or in operations such as scale change which are done in taint or under lock.
Therefore, we can remove these checks which were not consistently applied anyway.
If there is a genuine problem, better to see it quickly in a NullReferenceException than hide the bug.
This involves a large amount of change in test scene setup code to allow test scenes to share shared modules
SetupScene is now an instance method that requires an instantiation of SceneHelpers, though other SceneHelpers methods are still static
May split these out into separate classes in the future.
We can now do this since the entire scene and all objects within it are now successfully gc'd at the end of these tests.
This greatly improves the time taken to run each test (by reducing teardown time, not the time to actually do the test work that we're interested in).
Slightly simplifies config read in Scene constructor to help facilitate this.
This is so that the static MainConsole.Instance doesn't retain references to methods registered by scene and other modules to service commands.
This prevents the scene from being garbage collected at the end of a test.
This is not the final thing preventing GC - next up is the timer started by SimStatsReporter that holds a reference to Scene that prevents end of test gc.
Permanently comment out warnings about ScenePresence not being found - this is entirely expected if the avatar has alraedy logged out or left the scene.
These behave identically to llAttachToAvatar() and llDetachFromAvatar() except that they do not enforce the PERMISSION_ATTACH check
Intended for use in completely controlled dedicated environments where these checks are more a UI hinderance than a help.
Threat level high.
However, at some point it would be far more convenient to receive the TaskInventoryItem in the constructor rather than just the item UUID, so we don't have to constantly refetch our self item.
This involves giving the ceiling of the Z-component in a lure rather than the floor.
Ideally we would give the exact float compensating for relative avatar height but it looks like that isn't possible with the parcel id format used in lures
This replaces the Parts count which was rather pointless for a prim (it was either 1 if a child or the number of parts if the root).
This information is still avaliable on the "show object" command.
This was because update was looking for an existing stats record unique in session id, agent id and region id.
But if the user teleports to another region then region id changes.
WebStatsModule promptly doesn't find the existing record and tries to insert a new one, but only session id is the primary key and that's still the same, which makes things go bang.
This makes the update search only on the unique session id.
This is only an issue with simulators that have multiple regions where the webstats module is enabled.
Another thread could come and turn off physics for a part (null PhysicsActor) at any point.
Had to turn off localCopy on warp3D CoreModules section in prebuild.xml since on current nant this copies all DLLs in bin/ which can be a very large number with compiled DLLs
No obvious reason for doing that copy - nothing else does it.
Rename IFriendsModule.GetFriendPerms() -> GetRightsGrantedByFriend() to be more self-documenting and consistent with friends module terminology.
Add some method doc.
This adds ScenePresence to IClientAPI.SceneAgent earlier on in the add client process so that its information is available to EventManager.OnNewClient() and OnClientLogin()
Also add a code comment as to why we're caching friend information for child agents.
This is to deal with the hundred lines of command splurge when one previously typed "help"
Modelled somewhat on the mysql console
One can still type help <command> to get per command help at any point.
Categories capitalized to avoid conflict with the all-lowercase commands (except for commander system, as of yet).
Does not affect command parsing or any other aspects of the console apart from the help system.
Backwards compatible with existing modules.
On osgrid and other places, I have observed that manually sending appearance updates from the console often relieves grey avatar syndrome.
Despite hunting high and low, I haven't been able to find where this packet is sometimes being lost - it might be a persistent viewer bug for all I know.
Therefore, this experimental setting resends appearance data for everybody in the scene every 60 seconds. These packets are small and the viewer only fetches texture
data if it doesn't already have it.
Default is false.
The idea is to make the critical main scene loop as skinny as possible - it doesn't need to run things that aren't time critical and don't depend on update ordering.
This will be done gradually over time to try and uncover any issues. Many non-criticial scene loop activities are being launched on separate threadpool threads anyway.
This may also allow modules to register their own maintenance jobs without having to maintain their own timers and threads.
Currently the maintenance loop runs once a second, as opposed to the 89ms scene loop.
Update() now accepts a frames parameter which can control the number of frames updated.
-1 will update until shutdown.
The watchdog updating moves above the maintc recalculation for any required sleep since it should be accounted for within the frame.
This flavour is for changes in addition to the 0.7.3-post-fixes branch that are too large to be considered fixes but should be reasonably stable.
This flavour will almost certainly never see a formal release.
Viewer 2/3 contains a bug where the viewer will constantly retry ParcelVoiceInfoRequest requests on voice-disabled parcels where voice is otherwise available.
Attempts to fix this server-side have not been successful - sending a non-OK http code (e.g. a 404) just makes the viewer request again immediately.
Dropping the request entirely is a bit better but the viewer still retries after a minute.
Estate settings already enabled voice by default so doing the same for parcels. This only has an effect if you have any voice system active at all.
Ultimately, the re-request bug needs to be fixed viewer-side (LL suffers from the same issue!) but it might be worth implementing the drop request hack.
Unfortunately, the OnLoginsEnabled event is currently only guaranteed to fire if the RegionReady module is active.
However, we can instantiate the AuthorizationService in the module RegionLoaded method since by this time all other modules will have been loaded
The previous lines-per-second measurement used for top scripts report was inaccurate, since lines executed does not reflect time taken to execute.
Also, every fetch of the report would reset all the numbers limiting its usefulness and we weren't even guaranteed to see the top 100.
The actual measurement value should be script execution time per frame but XEngine does not work this way.
Therefore, we use actual script execution time scaled by the measurement period and an idealised frame time.
This is still not ideal but gives reasonable results and allows scripts to be compared.
This commit moves script execution time calculations from SceneGraph into IScriptModule implementations.
The first llDie() could lock Scene.m_deleting_scene_object.
The second llDie() would then wait at this lock.
The first llDie() would go on to remove the second script but always abort it since the second script's WorkItem would not go away.
Easiest solution here is to remove the m_deleting_scene_object since it's no longer justified - we no longer lock m_parts but take a copy instead.
This also requires an adjustment in XEngine.OnRemoveScript not to use instance.ObjectID instead when firing the OnObjectRemoved event.
This seems to be a particular problem with ReaderWriterLockSlim, though other locks can be affected as well.
It has been seen to happen when llDie() is called in a linkset running more than one script.
Alleviation here means supplying a ScriptInstance.Stop() timeout of 1000ms rather than 0ms, to give events a chance to complete.
Also, we check the IsRunning status at the top of the ScriptInstance.EventProcessor() so that another event doesn't start in the mean time.
Ultimately, a better solution may have to be found since a long-running event would still exceed the timeout and be aborted.
OnNewScript fires when a script is added to a scene
OnRezScript fires when the script actually runs (i.e. after permission checks, state retrieval, etc.)
They were all failing assertions but the exceptions these threw were caught as expected Exceptions.
I don't think we can easily distinguish these from the Exceptions that we're expecting.
So for now we'll do some messy manually checking with boolean setting instead.
This patch also corrects the assertions themselves.
Justin Clark-Casey (justincc)