aboutsummaryrefslogtreecommitdiffstatshomepage
path: root/OpenSim/Region (follow)
Commit message (Collapse)AuthorAgeFilesLines
* From: Chris Yeoh <yeohc@au1.ibm.com>Dr Scofield2009-06-031-7/+13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - fixes wild swings in memory usage related to usage of GetDrawStringSize() We've been seeing wild swings in memory usage and a large chunk of memory leak. From analysing this it's pretty clear that the mono garbage collector is rather buggy! When exercised heavily it looks like it frees more than its meant to resulting in crashes. GetDrawStringSize() measures the size in pixels of text. To do this memory for an image is allocated and used to call the GDI text measure functions. Although no reference to the temporary memory for the measuring is kept, it takes quite a while for the mono garbage collector to clean up - so if lots calls to GetDrawStringSize() are made at once there can be a spike in memory usage. If the garbage collector is not fast enough then the GDI layer runs out of memory. It also looks like the garbage collector is not always reclaiming all of the memory. I've attached an OpenSim patch which works around the garbage collector issues. Instead of dynamically allocating memory for measuring text sizes, it serialises (on a per region basis) access to a single block of memory. The effect of this is to be nicer to the garbage collector as it has a lot less work to do, at the cost of some theoretical loss in performance (nothing noticeable with our tests which hit it pretty hard). OpenSim still does leak memory slowly, but it is a lot more stable with this patch. I suspect that either the garbage collector misses bits of freed memory or the GDI/cairo layer leaks a bit each time a texture is created. Thats going to be a lot harder to hunt down, but for reference if someone has OpenSim running on Windows it would be interesting to see if it has the same problem as it would tell us if its a mono/GDI problem or an OpenSim problem.
* From: Alan Webb <alan_webb@us.ibm.com>Dr Scofield2009-06-031-2/+5
| | | | | | | | | | | Modify dynamic texture handling so that an explicitly targetted face is not scheduled for immediate expiration. The requirement for precaching explicitly requires these assets to persist. They do however remain temporary. This approach leaves the legacy mode of operation (ALL_SIDES) unchanged in this respect.
* Thank you Imaze Rhiano for a patch that implements Cenome Memory Asset Cache ↵Dahlia Trimble2009-06-032-0/+383
| | | | | | | | (Mantis #3759) See the files: bin/config-include/GridCommon.ini.example and bin/config-include/StandaloneCommon.ini.example to configure and enable this caching method.
* Thanks aduffy70 for Mantis #3762: A patch to fix llGround, llGroundNormal, ↵Dahlia Trimble2009-06-031-42/+80
| | | | and llGroundSlope
* revert r9765 due to too many errors on some compilers. Affects Mantis #3759Dahlia Trimble2009-06-032-382/+0
|
* Thank you Imaze Rhiano for a patch that implements Cenome Memory Asset Cache ↵Dahlia Trimble2009-06-022-0/+382
| | | | | | | (Mantis #3759) See the files: bin/config-include/GridCommon.ini.example and bin/config-include/StandaloneCommon.ini.example to configure and enable this caching method.
* Thank you kindly, MattSetzer, for a patch that solves a Mantis:Charles Krinke2009-06-021-2/+16
| | | | | | | | | | This appears to be due to the fact that no asset cache has been configured, possibly as a result of the configuration changes that have been made recently. I've attached a patch to display a message to that effect rather than throwing an error.
* * Pried apart the ODE tests. This fixes mantis #3212lbsa712009-06-021-0/+0
|
* * Add simple original xml serialization testJustin Clarke Casey2009-06-022-11/+75
|
* Explicitly set the changed status of the prim groups affected in a delinkMelanie Thielker2009-06-021-0/+1
| | | | | | Mantis #2484
* * Allow for lowercase searching in standalone mode, when case sensitiveArthur Valadares2009-06-011-1/+15
| | | | | | | search fails. This allows compability to libOMV bots, that always lowercase region names. * Uncertain if this should/could propagate to grids
* Only save crashes to disk if save_crashes in OpenSim.ini is true (false by ↵Jeff Ames2009-06-011-17/+20
| | | | default).
* * Add simple original sog xml serialization testJustin Clarke Casey2009-06-013-21/+22
|
* * minor: change the name of one serialization method to match its counterpartJustin Clarke Casey2009-06-014-7/+131
|
* * Fix race condition in save oarJustin Clarke Casey2009-06-014-28/+50
|
* Minor: Change OpenSim to OpenSimulator in older copyright headers and ↵Jeff Ames2009-06-01408-408/+408
| | | | LICENSE.txt.
* reduce verbosity of sculpt map cache messages to reporting cache failures onlyDahlia Trimble2009-05-311-2/+2
|
* Add copyright headers, formatting cleanup, ignore some generated files.Jeff Ames2009-05-3111-89/+305
|
* Update svn properties.Jeff Ames2009-05-316-1848/+1848
|
* * Adds MRM scripting commands, World.Objects.Create(Vector3 position) and ↵Adam Frisby2009-05-312-0/+22
| | | | World.Objects.Create(Vector3 position, Quaternion rotation). These rez a 'default box' object at the specified coordinates, and return the associated IObject.
* Fix IRCd init check for config sectionMike Mazur2009-05-311-3/+3
| | | | | | | | | The IRCStackModule used Nini.Config.ConfigCollection.Contains() to determine whether the "IRCd" section was present in the config. This ConfigCollection, however, stores an ArrayList of IConfig objects, not strings, so calling Contains("IRCd") always returns false since "IRCd" is a string, not an IConfig object.
* Forgot to add this to my last commit 2 days ago. Async in the HGBroker too.diva2009-05-301-1/+1
|
* * Tweaks to /WHO listings.Adam Frisby2009-05-301-9/+23
|
* * You are likely to be eaten by a grue.Adam Frisby2009-05-302-61/+162
| | | | | * Enable with [IRCd] Enabled=true (will listen on port 6666).
* * More TweaksAdam Frisby2009-05-303-9/+44
|
* * More IRCClientView fiddling. Now implements IClientAPI & IClientCore.Adam Frisby2009-05-303-182/+245
|
* * May partially implement a C# IRCd & IRCClientStack.Adam Frisby2009-05-303-0/+1487
|
* * Adds World.Audio.* to MRMAdam Frisby2009-05-293-1/+48
| | | | | * This includes methods such as PlaySound which take a Position as an argument, allowing you to trigger sounds arbitrarily across the scene without needing a parent object in the position.
* * Protip: When you click 'make public' in resharper, note to save before ↵Adam Frisby2009-05-291-1/+1
| | | | committing.
* * Implements Sound on Objects for IObject in MRMAdam Frisby2009-05-293-1/+29
| | | | | | * Method: IObject.Sound.Play(UUID sound, double volume) * More feature-packed API to come soon. (I want a World.Sound with arbitrary positioning)
* * Bug fix: Fixes an exception when Scene.RemoveClient is called to remove on aArthur Valadares2009-05-291-117/+117
| | | | | non-existing ScenePresence avatar. Also removed trailing white spaces.
* * refactor: little tweaks to trigger another buildJustin Clarke Casey2009-05-291-4/+15
|
* * refactor: Remove redundent prim id attribute on Scene.AddSceneObject()Justin Clarke Casey2009-05-292-9/+4
|
* * Add save xml2 serialization testJustin Clarke Casey2009-05-293-5/+73
|
* * minor: move common serialization test code into fixture setupJustin Clarke Casey2009-05-291-6/+17
|
* * Add xml2 load testJustin Clarke Casey2009-05-292-2/+143
|
* some clean up of sculpt map caching codeDahlia Trimble2009-05-292-6/+5
| | | | | remove a redundant debug message
* reinstate a hopefully more robust experimental decoded sculpt map caching schemeDahlia Trimble2009-05-293-12/+18
|
* disable sculpt map caching until a better method of avoiding asset requests ↵Dahlia Trimble2009-05-293-10/+15
| | | | can be found
* null test for texture assets when using cached sculpt mapDahlia Trimble2009-05-291-2/+3
| | | | | addresses mantis #3735
* Experimental decoded sculpt map cachingDahlia Trimble2009-05-293-20/+55
|
* Making the delegate handlers async in async Get, to make things consistent. ↵diva2009-05-291-2/+2
| | | | Them being synchronous in certain cases (asset in cache, for example) may account for slowness reported by folks in osgrid when they have the cache module on. Turns out that some of the provided handlers do non-trivial processing (the ones coming from J2KImage, for example), which means that the several asset requests that hit the cache end up being synchronous. The jury is still out on this.
* Instrumenting GlynnTuckerCache to find out the hit rate.diva2009-05-281-0/+29
|
* * Tweak to aboveAdam Frisby2009-05-281-1/+1
|
* * Makes grid announcements more prominent.Adam Frisby2009-05-281-1/+4
|
* * Oops. Forgot Initialise needs tweaking between IRegionModule and ↵Adam Frisby2009-05-281-0/+2
| | | | ISharedRegionModule.
* * Converts recently added OGSRadmin to IRegionModule because I cant be arsed ↵Adam Frisby2009-05-281-1/+17
| | | | figuring out Mono.Addins.
* * Adds OGS RAdmin class. Adds primitive remote admin functions for ↵Adam Frisby2009-05-281-0/+102
| | | | gridservers to perform on region servers. Used for grid-wide announcements, etc.
* * May make the terrain flatten brush behave like Second Life(tm)'s client ↵Adam Frisby2009-05-281-1/+5
| | | | expects. May break everything. You decide!
* This may bring the missing texture back.diva2009-05-281-1/+5
|