This upgrade also introduces exact stack rooting (see to the wiki:
JSRootingGuide) and fixes problems with moving GC. This allows us to
enable generational garbage collection (GGC).
Measurements a few months ago have shown a performance improvement of a
non-visual replay of around 13.5%. This probably varies quite a bit, but
it should be somewhere between 5-20%. Memory usage has also been
improved. Check the forum thread for details.
Thanks to everyone from the team who helped with this directly or
indirectly (review, finding and fixing issues, the required C++11
upgrade, the new autobuilder etc.)! Also thanks to the SpiderMonkey
developers who helped on the #jsapi channel or elsewhere!
Fixes#2462, #2415, #2428, #2684, #1374
Refs #2973, #2669
This was SVN commit r16214.
Everything is char* now, so we don't need to mess around with different
string types.
Done with:
ag -ls 'LOG(MESSAGE|MESSAGERENDER|WARNING|ERROR)' source | xargs perl
-pi -e'1 while
s/(LOG(MESSAGE|MESSAGERENDER|WARNING|ERROR).*)%[hl]s/$1%s/g'
This was SVN commit r16187.
Done with:
ag -ls 'LOG(MESSAGE|MESSAGERENDER|WARNING|ERROR)' source | xargs sed
-i 's/LOG\(MESSAGE\|MESSAGERENDER\|WARNING\|ERROR\)(L/LOG\1(/g'
This was SVN commit r16183.
First, do a ray intersection test with the bounding-sphere for all
entities on the map and then check the more detailed selection shape for
the remaining candidates. Do checks that require component lookups after
the ray intersection tests because these are relatively expensive.
The old method for figuring out which entities are below the mouse
cursor was incorrect because it does a 2D check to filter out the first
candidates which can lead to incorrect results with lower camera angles
and high buildings or buildings with a large footprint. Such problems
were avoided with quite a large radius for this 2D test and resulted in
a large number of candiate entities after this first test (200-500).
Also rename PickEntitiesAtPoint to PickEntityAtPoint and make it return
only one (the closest) match.
I've tested performance with the tracelogger by starting a map and then
moving the mouse in circles for one minute. The results were relatively
stable. I've compared the total time percentage of input.js:836, which
spends nearly all of the time in PickEntityAtPoint.
Ardennes Forest - Normal size: Original: 41.46% Patched: 31.6%
Ardennes Forest - Giant size: Original: 40.59% Patched: 51.55%
As we see, it's faster on normal map sizes but slower on giant maps with
a lot of entities.
This approach can be further improved with some kind of spatial
subdivision for the culling (like an octree), which would help the unit
renderer too. This way it should be possible to make it faster (and
still correct) on all map sizes and with a large total numbers of
entities.
This was SVN commit r16098.
On my ESR31 branch, I've made two measurements with different replays
(both around 15000 turns).
In the first, I got around 3% performance improvement and in the second
about 7.5%. It mainly depends on how often aura changes related to the
female citizen aura happen.
This was SVN commit r16055.
This first commit only refactors code and does not add any new feature.
Refs #2913 (see this ticket for more information about the change)
This was SVN commit r15925.
Mount public in case we are starting Atlas and no mods are specified in
the config.
If users remove all mods and save this to the config and try to start
Atlas it will
crash, as it cannot find some needed files.
This was SVN commit r15678.
To achieve this, mirage entities are created per player, to replace the
real entities when these ones fall into the fog-of-war. These mirage
entities are created on-the-fly, and destroyed when they get back in
sight.
This depends heavily on the VisibilityChanged message added in
2174eaaeee.
As a temporary adjustment, territories do not explore the map anymore
when their borders change. See #2709.
Fixes#599
This was SVN commit r15612.
I had to change a few other functions to take JS::MutableHandleValue
because JS::Stringify takes a JS::MutableHandleValue as input parameter.
That seems a bit strange because it should not change that value.
I assume it has historical reasons.
Refs #2415
Refs #2462
This was SVN commit r15605.
It might actually be needed again in the future, but I think even then
it would be easier to rewrite it than having to update the code in the
meantime.
Refs #2462
This was SVN commit r15600.
Changes CallFunction and CallFunctionVoid to use a HandleValue as object
parameter. Also changes some JS serialization/deserialization functions
to only support the JSAPI rooted types (drop support for CScriptVal and
CScriptValRooted there). Some other functions got changed too because
they were closely related.
Refs #2415
Refs #2462
This was SVN commit r15592.
Changes GetProperty, SetProperty and HasProperty and a few other
functions to take handles. The conversions to CScriptVal or
CScriptValRooted at some places should be removed in the future. I've
done that to avoid an even larger patch.
Refs #2415
Refs #2462
This was SVN commit r15568.
I still experience troubles with serialization tests but I played a few
games without OOS. At least these changes will help us to debug more
efficiently the remaining problems.
Fixes initialization of visibility data and update order for LoS-related
systems (territory borders, etc).
Also removes a C-like array and fixes a player ID shifting.
Refs #599
This was SVN commit r15563.
JS::MutableHandleValue is similar to JS::HandleValue, but the whole
JS::Value it points to can be replaced.
This change is needed for support of exact stack rooting and moving GC.
Contains a few other trivial API adjustments and style improvements too.
Refs #2462
Refs #2415
This was SVN commit r15534.
Remove commented out code writing triggers that has nothing to do with
the actual triggers implementation.
Free memory that was allocated for error reporting when when fork()
fails.
Remove dead initialization.
This was SVN commit r15477.
Change the shader itself so that the effects look nicer and are more
consistent across settings.
Rework the water mesh generation (simpler system). Fix a few issues.
May work oddly with Atlas since I haven't been able to compile yet.
Refs #1875 (maybe fix), Fixes#2077 (I'll assume it does), Fixes#2114
(assumption again), refs #48.
This was SVN commit r15473.
Previously we had a single culling frustum based on the main camera, and
any object outside the frustum would never get rendered, even if it
should actually contribute to shadows or reflections/refractions. This
caused ugly pop-in effects in the shadows and reflections while
scrolling.
Extend the renderer to support multiple cull groups, each with a
separate frustum and with separate lists of submitted objects, so that
shadows and reflections will render the correctly culled sets of
objects.
Update the shadow map generation to compute the (hopefully) correct
bounds and matrices for this new scheme.
Include terrain patches in the shadow bounds, so hills can cast shadows
correctly.
Remove the code that tried to render objects slightly outside the camera
frustum in order to reduce the pop-in effect, since that was a
workaround for the lack of a proper fix.
Remove the model/patch filtering code, which was used to cull objects
that were in the normal camera frustum but should be excluded from
reflections/refractions, since that's redundant now too.
Inline DistanceToPlane to save a few hundred usecs per frame inside
CCmpUnitRenderer::RenderSubmit.
Fixes#504, #579.
This was SVN commit r15445.
Allow components to individually subscribe/unsubscribe to messages,
instead of statically subscribing the entire component type. Use this
for most Interpolate/RenderSubmit messages, to avoid the performance
cost of passing those messages to a large number of components that will
just ignore them anyway.
On Azure Coast this reduces total time per frame by about 30% on a
CPU-bound system.
This was SVN commit r15400.
Does not account for setting rally points on units that can move into
the FoW/SoD.
Recalculates paths even if the rally points are not displayed.
Refactors UpdateMarkers() to remove some indentation levels.
Based on patch by Itms. Fixes#1257.
This was SVN commit r15288.
I guess changes to the map loading sequence caused the TerrainChanged
message to be sent before the map was switched from square to circular
instead of after. The pathfinder didn't notice the switch, so it
continued treating the map as if it were square, allowing units to walk
into the permanent map-corner SOD and vanish, and allowing territories
to expand into the SOD.
Tell the pathfinder explicitly when the map shape changes, so it can
discard its cached data correctly.
This was SVN commit r15277.
Instead of each CCmpVisualActor rendering itself individually, collect
all the units in a single CCmpUnitRenderer. This avoids the overhead of
doing Interpolate/RenderSubmit calls every frame for every object in the
world. It also allows more efficient culling.
CCmpUnitRenderer knows the positions of each object at the start and end
of each turn, and computes the bounding sphere of the object along that
path. That allows quick culling without recomputing the precise
interpolated transform every frame. (In the future it could be improved
much more.)
Clarify and clean up the sending of PositionChanged messages, and add
new InterpolatedPositionChanged.
Remove the forceFloating parameter from GetInterpolatedTransform, since
it doesn't fit the new design. Replace it with a (non-synchronised) flag
in CCmpPosition.
Move construction progress from CCmpVisualActor to CCmpPosition, so that
it consistently affects all position/transform computation.
Refs #2337.
This was SVN commit r15265.
The tests crashed on my Debian systems but not on my Ubuntu system.
The crash happened in line 142 of SpiderMonkey's ThreadLocal.h.
I know that I had to use a workarounnd for contexts to avoid destroying
the context that was created first. I also had in mind that in newer
versions a JS_Init function got introduced which presumably solves this
kind of issues.
Based on this experience I assumed runtimes could have a similar problem
and this patch indeed fixes the issues.
Unfortunately the correct usage of JSRuntimes in that regard is not
documented.
There's only a rater misterious comment in JSAPI.h which hasn't been
cleared up so far and is most likely completely outdated
(https://bugzilla.mozilla.org/show_bug.cgi?id=992641).
This was SVN commit r14995.
This commit contains all the required changes to our source files and
build scripts (hopefully).
A next commit will remove the old stuff of SpiderMonkey 1.8.5.
Spcial thanks to:
- H4writer who helped a lot mainly with the performance issues we
had/have, but also with other problems or questions.
- Leper for the review.
- Historic_bruno for implementing the build scripts on Mac OS X and
testing on the Mac.
- The people from the #jsapi channel and from
mozilla.dev.tech.js-engine who answered a lot of questions and helped
solving problems.
- All the other people who helped
Refs #1886Fixes#2442Fixes#2416
This was SVN commit r14877.
Fix a few style issues and a bug with the gatherer count.
Still need to fix the entity.js file to handle properly some things as
this uses raw templates values.
Cache the AIinterface in AIProxy.js, please report if this works
properly.
This was SVN commit r14588.
Remove waves for now as they were unsatisfactory.
Removes "shininess" as a water parameter as it was basically useless.
Add a button in Atlas to recompute water parameters so you can now see
fog in Atlas.
Fixes#1743, #1803 (invalid)
Refs #1875, #2114, #48.
This was SVN commit r14514.
Each GUI Page gets its own compartment and all ScriptInterfaces in the
same thread should now use the same JS Runtime.
This is required for the SpiderMonkey upgrade.
Check the ticket for details.
Closes#2241
Refs #1886
Refs #1966
This was SVN commit r14496.
This avoids wrapping overhead that would otherwise be required because
multiple globals per compartment aren't supported anymore in newer
versions of SpiderMonkey.
Check the ticket for a detailed explanation.
Closes#2322
Refs #2241
Refs #1886
This was SVN commit r14441.
- Implement tech mods for UnitMotion speeds, Fixes#1958, based on patch
by wraitii
- Implement athen_hero_themistocles ship speed aura as a test
This was SVN commit r14058.
Implements custom, VFS-based map load/save dialogs for Atlas (replaces
broken native file dialogs), fixes#631, #889.
Fixes map loading/saving to handle arbitrary subdirectories for better
organization.
Adds default settings to Atlas player panel, fixes#1872. Each setting
now has a checkbox to choose whether it should be saved with the map
(avoids writing lots of useless default data for each map).
Adds map preview setting to Atlas, refs #1745.
Cleans up and simplifies some duplicate code.
Fixes optional serialization performance test.
This was SVN commit r13938.
On MSVC, casting i32 to i64 then multiplying gets compiled into a call
to __allmul, which is slow. Use the __emul intrinsic instead.
GCC 4.6+ appears optimise this case automatically and doesn't need any
special handling.
This reduces the cost of ComputeShortPath by about 50% (testing AI vs AI
on Oasis 01).
This was SVN commit r13873.
Overflowing elements in SpatialQueryArray are now reported with
debug_warn, but can be ignored and the program will continue working,
but further entities will be truncated.
This was SVN commit r13860.
Add CEntityHandle which wraps an entity_id_t and also has a pointer to a
per-entity array of IComponents. QueryInterface/CmpPtr on a handle just
involve a couple of pointer dereferences instead of a map lookup.
This requires sizeof(void*) per registered interface type per entity,
which is currently ~0.5KB per entity, which shouldn't be a problem.
This was SVN commit r13825.
