AZStd::fixed_vector had all its functions marked with constexpr, but this requires all member variables to be fully initialized. This meant that the internal array used to store elements always has to be fully initialized. This was done for trivial classes but not for non-trivial classes. As a result trivial classes always did a memset (or more optimized versions for smaller buffers) while the non-trivial version couldn't actually be stored in a constexpr variable. Since AZStd::fixed_vector is meant to be dynamic the choice was made to remove the constexpr from all non-static member functions in favor of avoiding the overhead of memset, which profiling showed was a considerable overhead depending on the reserved size. If a truly constexpr array is needed than AZStd::array is a better choice as that's designed to not by dynamic.
Signed-off-by: AMZN-koppersr <82230785+AMZN-koppersr@users.noreply.github.com>
* Adds simple stats tracking to AssetProcessor
The system captures cumulative and individual processing times.
It avoids touching any part of the app which affects decision
making or flow, or altering any structures such as JobEntry which
are involved in processing, in order to keep it as simple and
small as possible.
Signed-off-by: lawsonamzn <70027408+lawsonamzn@users.noreply.github.com>
The function that modifies resource limits will return true always and warn instead
of giving an error.
Signed-off-by: amzn-phist <52085794+amzn-phist@users.noreply.github.com>
* Fix several Prefab outliner ordering issues
This change does a few things to address instability in entity order when prefabs are enabled:
- Changes ordering behavior on entity add to always be "insert after the last selected sibling of the new entity, or at the end if there is no selected sibling"
- Adds some logic to ensure selection stays frozen during prefab propagation to allow this behavior to be used
- Alters delta generation in `PrefabPublicHandler::CreateEntity` to use the template instead of reserializing the DOM - this avoids a whole bunch of patching issues caused by EditorEntitySortComponent doing post-hoc order fix-up and should generally be safer/faster as we're producing patches for the actual target for those patches
- Because the duplicate action is DOM-driven, and there's some thorniness around making changes that will affect the template during propagation, this adds `PrefabPublicHandler::AddNewEntityToSortOrder` to directly patch the DOM for the duplicate case
Two bits of this come from patches from the incredibly helpful @AMZN-daimini
- Alters `PrefabPublicHandler::GenerateUndoNodesForEntityChangeAndUpdateCache` behavior for determining what's an override: we now check to see if we're part of the current focused instance but *not* owned by the focus instance directly. This lets entity order for nested prefabs get saved to the owning prefab instead of as an override. I'm putting this up for discussion without a feature flag gating it, but we may wish to make this a toggle or disable it outright for stabilization (in which case entity order won't be saved to the owning prefab)
- Adds a custom serializer for EditorEntitySortComponent. This isn't strictly necessary now, but it was very useful for debugging and ended up receiving much more manual testing its migration path and save/load path more than I've tested without using the serializer.
Signed-off-by: Nicholas Van Sickle <nvsickle@amazon.com>
* Add missing serializers
Signed-off-by: Nicholas Van Sickle <nvsickle@amazon.com>
* Address some review feedback
Signed-off-by: Nicholas Van Sickle <nvsickle@amazon.com>
* Generate patch in `PrefabPublicHandler::CreateEntity` using the instance's fully evaluated template in-memory
Signed-off-by: Nicholas Van Sickle <nvsickle@amazon.com>
* Fix up comment
Signed-off-by: Nicholas Van Sickle <nvsickle@amazon.com>
* Fix Linux build
Signed-off-by: Nicholas Van Sickle <nvsickle@amazon.com>
* Try to make test less timing dependent (haven't been able to repro failure locally)
Signed-off-by: Nicholas Van Sickle <nvsickle@amazon.com>
* Fix another build issue
Signed-off-by: Nicholas Van Sickle <nvsickle@amazon.com>
* Fix unit test failures
Signed-off-by: Nicholas Van Sickle <nvsickle@amazon.com>
* Fix a duplicate issue with sanitization that was causing sporadic test failure (thanks, test!)
Signed-off-by: Nicholas Van Sickle <nvsickle@amazon.com>
* One more Linux fix...
Signed-off-by: Nicholas Van Sickle <nvsickle@amazon.com>
Removed the generator expression from the `LY_ARCHIVE_FILE_SEARCH_MODE` define as Xcode doesn't support per-config per-file definitions.
Updated the #else block in ArchiveVars to default to PakOnly mode if the `LY_ARCHIVE_FILE_SEARCH_MODE` define isn't set.
Moved the the `LY_ARCHIVE_FILE_SEARCH_MODE` Cache Variable from cmake/Deployment.cmake to feature_options.cmake within the AzFramework folder.
Signed-off-by: lumberyard-employee-dm <56135373+lumberyard-employee-dm@users.noreply.github.com>
Gene Walters
Ronald Koppers
Junbo Liang
Danilo Aimini
Tom Hulton-Harrop
Steve Pham
Nicholas Lawson
AMZN-Phil
AMZN-Phil
amzn-phist
moraaar
puvvadar
Esteban Papp
amzn-mike
Nicholas Van Sickle
John Jones-Steele
lumberyard-employee-dm
Alex Peterson