These changes have the added benefit of simplifying some of the serialization code. MaterialSourceDataSerializer is no longer needed, as its main purpose was to pass the MaterialTypeSourceData down to the MaterialPropertyValueSerializer.
Before, the JSON serialization system gave a lot of data flexibility because it did best-effort conversions, like allowing a float to be loaded as an int for example. But now the material serialization code doesn't know target data type, so it has to assume the data type based on what's in the .material file, and then the MaterialAsset will convert the data to the appropriate type later when Finalize() is called.
Signed-off-by: santorac <55155825+santorac@users.noreply.github.com>
In some cases it may be acceptable to do non-const things in a const function as long as it is only manipulating internal data, and the public facing API returns the same values as before. But in this case, the IsFinalized function is a public facing API that would have a different result after GetPropertyValues was called.
I also updated a couple other minor things from code review feedback.
Signed-off-by: santorac <55155825+santorac@users.noreply.github.com>
I found a mistake where MaterialAssetCreator needs to clear the raw data when configured to finalize the material asset.
Since MaterialSourceData no longer relies on the material type source file at all, I was able to change MaterialSourceDataTest to avoid saving the source data to disk.
Signed-off-by: santorac <55155825+santorac@users.noreply.github.com>
I made MaterialAsset::Finalize private so I could add some parameters specifically for MaterialAssetCreator to use. Now MaterialAssetCreator::Begin has an option to finalize the material or not.
Moved MaterialAssetCreatorCommon::ValidateDataType to MaterialPropertyDescriptor as "ValidateMaterialPropertyDataType" so that MaterialAsset::Finalize could use it too
Signed-off-by: santorac <55155825+santorac@users.noreply.github.com>
- Remove trait AZ_TRAIT_DISABLE_FAILED_ATOM_RPI_TESTS
- Add template specialization for ConstantsData::SetConstant<Color> and ConstantsData::GetConstant<Color>
- Remove trait AZ_TRAIT_DISABLE_FAILED_ATOM_RPI_TESTS from source code
- Fix ConstantsData::SetConstantArray<bool> to pass unit test SetConstant_GetConstant_FalsePackedInGarbage_Bool on Linux, cause by an unwanted optimization (comments describe the situation). This will make this behavior consistent across platforms
Signed-off-by: Steve Pham <82231385+spham-amzn@users.noreply.github.com>
* Multiple bug fixes
- Handle Dx12 validation error related to multiple resource transitions happening on the same resource.
- Added API to better query for image/buffer views associated with an attachment.
- Address possible serialization issues by usage of size_t
- Reduced number of Descriptor sets allowed per SRG Pool in order to reduce descriptor waste.
Signed-off-by: moudgils <47460854+moudgils@users.noreply.github.com>
* Addressed minor feedback
Signed-off-by: moudgils <47460854+moudgils@users.noreply.github.com>
* Fix compile issues
Signed-off-by: moudgils <47460854+moudgils@users.noreply.github.com>
* Missed a few other fixes
Signed-off-by: moudgils <47460854+moudgils@users.noreply.github.com>
* ShaderMetricsSystem spawns a lot of warnings
when run Atom_RPI.Tests
Added @user@ path alias definition.
* When running unit tests, using AZ::Utils::GetProjectPath() will resolve
to something like:
"/data/workspace/o3de/build/linux/External/Atom-9a4d112b/RPI/Code/Cache"
The ShaderMetricSystem.cpp writes to the @user@ folder and the following
runtime error occurs:
"You may not alter data inside the asset cache. Please check the call
stack and consider writing into the source asset folder instead."
"Attempted write location:
/data/workspace/o3de/build/linux/External/Atom-9a4d112b/RPI/Code/Cache/user/shadermetrics.json"
To avoid the error We use AZ::Test::GetEngineRootPath();
Signed-off-by: galibzon <66021303+galibzon@users.noreply.github.com>
* Enable relocation of the Project Game Release Layout
Relocating the Project Game Release Layout to another directory on the file system failed due to the querying of the engine root failing due to the ComponentApplication::m_engineRoot not using the project path stored in the SettingsRegisry if the engine root cannot be detected
Removed the ApplicationRequestBus GetEngineRoot function.
The ComponentApplicationRequestBus has a function of the same name that returns the same path.
Removed the deprecated GetAppRoot function.
The path it returns has no defined value. It was not the engine root or the project root.
Removed unused CFileUtil and CFileUtil_impl functions that were invoking the ApplicationREquestBus GetEngineRoot function.
On the way to update the functions it was discovered that they aren't called
Added a CalculateBranchToken overload that can populate a fixed_string to avoid heap allocations
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Protect against an empty list of artifacts to remove when generating the
engine.pak
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* ATOM-16489 Add find passes functions for Scene or RenderPipeline in PassSystemInterface
Introduced new PassSystemInterface::ForEachPass() funtion to replace PassSystemInterface::FindPasses(), PassSystemInterface::GetPassesByTemplateName and ParentPass::FindPassByNameRecursive() functions.
Update all the places which were using those three functions.
The new pass finding filter support any combination of pass name, pass template name, pass class type, pass hirechary, owner scene, owner render pipeline.
Update unit tests.
Signed-off-by: Qing Tao <qingtao@amazon.com>
(cherry picked from commit fe8dac7989)
MaterialAssignment::ApplyProperties() still reports warnings but does not update the m_propertyOverrides.
MaterialAssignment::ApplyProperties() will now skip the old name'd overrides if overrides are present for the new names. I'm not sure if this will ever happen, but it did happen while I had some intermediate changes, so I imagine it could happen again.
I had to update the Material::FindPropertyIndex function to expose information about renames when they occur. This should make it easier for other systems to get (somewhat) automatic benefit from the version update feature.
I also found that there was an issue with material inspector where it wouldn't be initialized the the right override values when renames were present. Now it applies the renames to whatever override data it gets from the Material Component.
Signed-off-by: santorac <55155825+santorac@users.noreply.github.com>
I added an ApplyPropertyRenames function to MaterialTypeAsset very similar to the one in MaterialTypeSourceData.
Updated the MaterialAssignment class to apply any property renames when it discovers the old name doesn't work. This will be written to disk when the level or prefab is saved.
Signed-off-by: santorac <55155825+santorac@users.noreply.github.com>
(It's a simple enough change to make manually, and making .materialtype is an uncommon workflow, so not worth doing this automatically).
Signed-off-by: santorac <55155825+santorac@users.noreply.github.com>
Also improved the MaterialAssetTests UpgradeMaterialAsset() to focus on testing the inputs and outputs of the class rather than the private internal data.
Signed-off-by: santorac <55155825+santorac@users.noreply.github.com>
Also rearranged some logic to simplify the code that loops over rename actions, avoiding making unnecessary additional maps.
Signed-off-by: santorac <55155825+santorac@users.noreply.github.com>
- Added MaterialSourceData::ApplyVersionUpdates() for updating the properties. This should be called by tools after loading the MaterialSourceData. (But can be omitted if a tool wants to read the data exactly as it appears in the .material file).
- Updated MaterialTypeSourceData::FindProperty to support applying version update renames, including a ApplyPropertyRenames utility function, which are necessary for MaterialSourceData to be able to find the necessary property definitons while loading.
- Added a new context struct to JsonMaterialPropertyValueSerializer for passing down the material type version number, to help with applying property renames.
- Renamed the .material file format "propertyLayoutVersion" to "materialTypeVersion" which is more accurate. This shouldn't hurt existing data as this field wasn't actually used for anything before.
- Updated Material Editor to again store the material type version number in .material files.
MaterialSourceDataTests updates...
- Updated to include both a .materialtype file and a MaterialTypeAsset for the test material type. Both are used by the MaterialTypeSourceData class.
- The default test material type now includes some version update steps; these are only used for version update tests and won't impact the other test functions.
- Updated the path for storing temp files to disk, to just be in a "temp" folder in the exe path. (Originally they were saved to the gem folder near MaterialSourceDataTests.cpp, but at some point someone changed it to be under the exe folder, so there's no reason to use the full gem path anymore).
MaterialTypeSourceDataTests updates...
- Moved some code that was accidentally added to LoadAllFieldsUsingOldFormat but should have been in LoadAndStoreJson_AllFields.
- Added test cases for unsupported version update operations
Signed-off-by: santorac <55155825+santorac@users.noreply.github.com>
This change is a preparation for moving the CPU profiler/visualization system from Atom into its own Gem by removing the dependency on local time tracking object AZ::RHI::CpuTimingStatistics
Full changes include:
- Removed all usage of AZ::RHI::CpuTimingStatistics
-- Replaced with pushing to AZ::Statistics::StatisticalProfilerProxy global instance
- Promoted VariableTimer from AZ::RHI to AZ::Debug
- Removed now unused CpuTimingStatistics.h
Signed-off-by: AMZN-ScottR 24445312+AMZN-ScottR@users.noreply.github.com
* Update the FileIOAlias naming to make the cache, project root and engine
root paths more clear
The alias of `@root@`, `@assets@`, and `@projectplatformcache@` has been
collapsed to `@projectproductassets@`
The alias of `@devroot@` and `@engroot@` has been collapsed to
`@engroot@`
The alias of `@devassets@` and `@projectroot@` has been collapsed to
`@projectroot@`
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Updated use of devassets and devroot properties in python
Those properties now use projectroot and engroot
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Updating the alias @engroot@ alias path comment in each platform specific LocalFileIO_*.cpp file
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Removed hardcoded size of 9 for the product asset alias.
The ResolvePath function now just appends the @projectproductassets@
alias with the input path
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Remove duplicate @projectproductassets@ check in ProcessFileTreeRequest
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Fix for typos in Hydra python test
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Updated LocalFileIO::Copy call on Windows to use the Unicode aware CopyFileW API
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Updated the AWSMetreicsGemAllocatorFixture to properly suppress asset
cache write errors for Test file creation.
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Removed unneeded call to set the @projectproductasstes@ alias at the bottom of the AssetSeedManagerTest SetUp
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Added a deprecated alias map to the FileIO System
When a deprecated alias is accessed, the FileIO System logs an AZ_Error and indicates the alias that should be used
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Updated python test scripts to use the projectroot binding
Retrieving the AutomatedTesting project path based on "<devroot>/AutomatedTesting" has been removed.
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Updated references to devroot and devgame within the codebase
The GetAbsoluteDevGameFolderPath functions has been replaced with direct call to AZ::Utils::GetProjectPath
The GetAbsoluteDevRootFolderPath functions has been replaced with direct calls to AZ::Utils::GetEnginePath
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Updated <engroot>/AutomatedTesting references to projectroot
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Replaced references that assumes the project path is <engroot>/AutomatedTesting with <projectroot> in the AutomatedTesting python test
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Correct casing in emfxworkspace file
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Removed newly added AppendPathParts function
Removed the Path constructors which accepts a PathIterable instance
The PathIterable isn't safe to return to a user of the Path class as it might be referencing temporary paths supplied via PathView arguments
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Fixed unused parameter warning
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Undid change to the LexicallyProximate function to set the path result to the base path.
It needs to return the *this path if the pathResult is empty
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Moved the LocalFileIO ConvertToAbsolutePath implementations to AZ::Utils
Fixed the ConvertToAbsolutePath implementation for Unix platforms to use a buffer that is size PATH_MAX(4096 on all our supported Unix platforms).
Because the buffer before was AZ::IO::MaxPathLength which as a size of 1024, this was resulting in the Unix `realpath` function smashing the stack when it converts a path to an absolute path that is longer than 1024 characters
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Updated the EditorCore.Tests to attach the AZ Environment to the EditorCore shared library that is statically loaded on launch.
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Fixed for DeprecatedAlaisesKeyVisitor Visit function causing the non string_view overloads being hidden causing a hidden base class function warning
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Changed the AWSMetricsGemMock to use a temporary for writing test assets
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Updated the LocalFileIO::ResolvePath function to use HasRootPath to determine if a path starts with a root path such as "D:", "D:/" or "/"
IsAbsolute was not the corect check as the path of "D:" is a relative
path.
To be absolute according to the Windows the path must have a root
directory. i.e "D:/" or "D:\"
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Removed absolute path comment from LocalFile_UnixLike.cpp and LocalFile_Android.cpp FindFiles implementations
Updated the ConvertToAlias to supply an AZ::IO::FixedMaxPath
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Replaced usage of the @projectproductassets@ alias with @engroot@ when referring to the LmbrCentral source folder in the CopyDependencyBuilderTest and the SeedBuilderTests
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Updated the ScriptCanvas Upgrade Tool to output backed up files to the
Project User Path instead of the engine root
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Fixed whitespacing issues in Application.cpp
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Remove unnecessary creation of a FixedMaxPath in the UpgradeTool.cpp
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Modified testSeedFolder variable in the SeedBuilderTests to use the
@engroot@ alias instead of @projectproductassets@/.. alias when
referring to the LmbrCentral Gem source path
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Updated references to the Project Asset Cache in the PythonTests.
Those tests no longer use the logic `azlmbr.paths.projectroot / "Cache" / "pc"` to retrieve a path to the cache root but instead the `azlmbr.paths.projectproductassets` constant
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Fixed the FileIO Deprecated Alias test on Windows
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Removing @projectsourceassets@ alias, as it is only used once.
Updated the PhysX EditorSystemComponent.cpp to query the ProjectPath
form the SettingsRegistry.
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Replaced @projectproductassets@ alias with @products@
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
* Rollback changes to the PhysX EditorSystemComponent.cpp
The changes to use the ProjectPath from the SettingsRegistry has been implemented in PR #4497
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
This required the use of custom serializers, because the JSON serialization system does not have any means of supporting field name aliases through SerializeContext.
Testing: RPI unit test pass and AtomSampleViewer material screenshot test script passes.
Signed-off-by: santorac <55155825+santorac@users.noreply.github.com>
This revealed that the approach of reflecting both the old "id" and the new "name" would not work, because whenn saving it would write out both fields. So I decied to just give up on backward compatibility. This will be much cleaner than trying to continue supporting "id" as a field name, it is uncommon for users to make their own material types at this point, and if they have made some it is very easy to search and replace "id" with "name" update their files.
All .materialtype files have been updated. RPI unit tests now pass. ASV still passes.
Signed-off-by: santorac <55155825+santorac@users.noreply.github.com>
A "property name" is the name of the just the property without regard to the group that it's in. A "group name" is the name of the group. And a "property ID" is the full unique name of a property in the form "groupName.propertyName". This is important preparation for upcoming changes where property sets can contain other property sets, and property IDs can be arbitrarily long like "layer1.baseColor.factor" for example.
The naming changes include variables, some code comments, and the .materialtype file format. I was able to make these changes in a backward compatible way so a property or group "id" field has been replaced with a "name" field, but "id" is still supported for compatibility. StandardPBR, EnhancedPBR, StandardMultilayerPBR, and Skin have all been updated. Note that MinimalPBR has not been updated, proving that backward compatibility works. (We can update this one too at some point though).
Testing:
Opened up materials in the material editor.
Ran AtomSampleViewer in dx12 and vulkan with no new failures.
Signed-off-by: santorac <55155825+santorac@users.noreply.github.com>
An early-return in Material::SetPropertyValue was breaking initialization because Init() was called multiple times, and on subsequent initializations the property values weren't getting reset.
Added a unit test to ensure this kind of thing doesn't happen again.
Signed-off-by: santorac <55155825+santorac@users.noreply.github.com>
Merge pull request #4102 from aws-lumberyard-dev/Atom/santorac/WarnOnMaterialPsoChanges
These changes make material system report warnings when gameplay scripts attempt to change PSO-impacting material properties at runtime. So far the material system has always allowed any properties to be changed at runtime, including those that affect Pipeline State Objects (PSOs), as this is supported on several platforms. But some platforms require that Pipeline State Objects be pre-compiled and shipped with the game. At some point we will need to add new restrictions that limit what material properties can be changed at runtime. In the meantime, these warnings should alert users to avoid this, as the functionality likely won't be supported in the future.
- Made the Material and LuaMaterialFunctor classes configurable to report errors or warnings when material properties modify Pipeline State Objects. This is controlled by a new "MaterialPropertyPsoHandling" enum.
- Made the EditorMaterialComponent override PSO handling as Enabled, to prevent warnings when the user is editing material instance property overrides. This requried a new MaterialComponentNotificationBus bus message "OnMaterialInstanceCreated".
- Added a new ScopedValue utility class that simply sets a value when it goes out of scope.
- Removed unnecessary GetMaterialPropertyDependencies member from material functor context classes, as this is already available as part of the functor itself.
- Made Material::SetPropertyValue return early when the property value hadn't actually changed. Besides being more efficientn, this prevents unnecessary spamming of the new warning.
- Made a couple imporvements to material_find_overrides_demo.lua:
- Made the target material slot name configurable through an exposed component property.
- Fixed a timing issue where the assignmentId was invalid if FindMaterialAssignmentId is called too early.
santorac
Steve Pham
moudgils
galibzon
Guthrie Adams
Tom Hulton-Harrop
lumberyard-employee-dm
Qing Tao
Scott Romero
hershey5045
Robin
Esteban Papp
Jeremy Ong
chcurran