* Added helper functions for calculating linear and angular velocities.
* Added debug visualization helper for linear and angular velocities.
Signed-off-by: Benjamin Jillich <jillich@amazon.com>
All of the component headers in the gem have been moved to the Include directory to make them public to other gems. This allows "upstream" unit tests and benchmarks to easily create real non-mocked-out versions of these components to do more integration-level and system-level testing and benchmarking.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Add benchmarks to GradientSignal.
Cleaned up and rearranged a bit of the unit testing code to make it reusable from a benchmark suite as well. Added set of benchmarks for measuring GetValue(), so that we can compare against GetValues() when it gets added.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Fixed Editor unit tests.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* small bugfix
Signed-off-by: rhhong <rhhong@amazon.com>
* ActorInstanceId default to -1 when no %lastresult matches
Signed-off-by: rhhong <rhhong@amazon.com>
* CR feedback - wrap function to get the first available editor actor instance.
Signed-off-by: rhhong <rhhong@amazon.com>
* Remove mcore inline
Signed-off-by: rhhong <rhhong@amazon.com>
* Fixed the bug that delete an instance from actor manager crashes the editor.
Signed-off-by: rhhong <rhhong@amazon.com>
* Added new helper functions to get the shortest equivalent of the rotation. In case the w component of the quaternion is negative the rotation is > 180° and taking the longer path. The quaternion will be inverted in that case to take the shortest path of rotation.
* Added unit test.
* Renamed the angle parameter of the CreateRotationX/Y/Z() functions into angleInRadians.
Signed-off-by: Benjamin Jillich <jillich@amazon.com>
* Change flow so that TerrainSystem stops responding during deactivation.
Some systems might accidentally try to call back to the TerrainSystem inside a DestroyBegin notification, so make sure it stops listening before sending out the notification.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Change gradients to cache and use a GradientTransform instance.
In my local test case, calling EBus on every call took 337 ms, using a lambda to wrap the calls took 197 ms, and using the fully cached version took 170 ms.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Remove the wrappingTransform function and go back to the switch statement.
There was a bit of overhead to each function call due to using AZStd::function that just isn't necessary for this use case.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Add profile markers to the heightfield updates so that they're more visible.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Cleared state while component is deactivated.
The state was getting refreshed even while the component was in a deactivated state, which meant that it wasn't properly notifying of state changes when it became active since it wasn't detecting an actual change. By clearing the state when deactivated, and ensuring the state isn't getting refreshed *while* deactivated, the notifications work properly.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Fixed compile warning on unit test.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Addressed PR feedback - changed comments, reduced mutex scope
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Breaking up terrain FP wip - macro materials decoupled from meshes. Mesh creation and rendering pulled out from MeshFeatureProcessor into TerrainMeshManager. Stubs added for macro and detail material managers.
Signed-off-by: Ken Pruiksma <pruiksma@amazon.com>
* Separated macro material management from the terrain feature processor. Also separated bindless image array handling from terrain feature processor.
Signed-off-by: Ken Pruiksma <pruiksma@amazon.com>
* Detail materials separated from terrain feature processor. Also pulled out Aabb2i and Vector2i into their own simple classes
Signed-off-by: Ken Pruiksma <pruiksma@amazon.com>
* Changed some classes so that when the SRG changes the classes don't need to be completely reinitialized and can instead just update their indices and push data to the new srg. Fixed an issue where MacroMaterialData wasn't being exposed to ScriptCanvas. Terrain shader reloads should now work correctly. Also added some debug logging to help catch an issue where sometimes detail materials don't seem to load.
Signed-off-by: Ken Pruiksma <pruiksma@amazon.com>
* Terrain PR reveiw updates
Signed-off-by: Ken Pruiksma <pruiksma@amazon.com>
* Some small PR review fixes. More comments in TerrainDetailMaterialManager.h
Signed-off-by: Ken Pruiksma <pruiksma@amazon.com>
* Fixing unused variable causing clang failure
Signed-off-by: Ken Pruiksma <pruiksma@amazon.com>
* Fixing unused variable in release.
Signed-off-by: Ken Pruiksma <pruiksma@amazon.com>
* Fixing a forward declare that oddly didn't work on linux.
Signed-off-by: Ken Pruiksma <pruiksma@amazon.com>
* Fixing linux missing include... not sure why only linux failed on this.
Signed-off-by: Ken Pruiksma <pruiksma@amazon.com>
* Adding missing include
Signed-off-by: Ken Pruiksma <pruiksma@amazon.com>
* First version of GradientTransform class.
The gradient transform logic is getting encapsulated into a class so that it can be cached and used by components in a much more optimal way than making ebus calls to the GradientTransform component on every transformed point.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Moved GradientTransform into its own source files.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Clean up and simplify GradientTransform logic.
Added extensive commenting and split TransformPositionToUVW into a separate method for normalizing (TransformPositionToUVWNormalized) so that there doesn't need to be any conditional logic. There's no runtime variance as to which one needs to be called from a given call site.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Added unit tests for GradientTransform.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Add comparison operators to GradientTransform so we can easily tell when it has changed.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Updated comments to be more Doxygen-friendly.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Optimized wireframe drawing.
As a part of rearranging the code to make use of the upcoming ProcessHeightsFromRegion, the number of calls to GetHeight could be reduced, dropping the refresh time in my test case from 2550 ms to 1068 ms.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Change refresh to only happen on wireframe draws.
This helps ensure that multiple data changes in a single frame don't cause multiple refreshes, and prevents us from taking a refresh penalty when the wireframe isn't being drawn at all.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Change debugger to update wireframes incrementally.
This works by having a fixed NxN grid of sectors. The camera is always considered as being in the center of the grid, so anytime the camera's grid square changes, only a subset of sectors are updated to have the new wireframe data.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Bugfix - sector vertex count was 4x too high.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Bugfixes & comments.
Fixed initial "mark dirty" refresh - the dirty region Z value needed to be ignored.
Added copious comments for the math & logic, and simplified some of the math.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Small update to comment for better readability.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
* Addressed PR feedback.
Signed-off-by: Mike Balfour <82224783+mbalfour-amzn@users.noreply.github.com>
Sleeping in tests and attempting to rely on fine-grained measurements to
check code validity is intrinsically brittle. Wall-clock time is
unreliable in an environment where tests are run under a hypervisor that
may choose to suspend your VM at any point, or in situations where the
OS cannot schedule your thread in time. The correct way to reintroduce
these tests in the future is provide an override for the timestamp
queries that can be injected in the test environment to control the wall
time deterministically.
Signed-off-by: Jeremy Ong <jcong@amazon.com>
In some cases, the interface being queried may be persistently nullptr
(e.g. a gem isn't active, or a feature was disabled by a cvar). In this
case, querying the interface would always enter the writer-lock branch
with existing code because nullptr was synonmous with "we haven't
queried for this variable yet."
This PR adds an additional state variable that is set when s_instance is
assigned to, so that after the variable is queried and found to be null,
future queries can enter the read-lock branch.
Signed-off-by: Jeremy Ong <jcong@amazon.com>
Mike Balfour
Chris Galvan
AMZN-Phil
bosnichd
mrieggeramzn
Benjamin Jillich
Luis Sempé
Gene Walters
carlitosan
Junbo Liang
amzn-sj
Roman
amzn-mike
Gene Walters
Ken Pruiksma
Michael Pollind
nggieber
Jeremy Ong
Jeremy Ong