-
Notifications
You must be signed in to change notification settings - Fork 7
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Have CoreCLR follow sequential layout with GC refs #18
Conversation
By default CoreCLR will ignore the LayoutKind.Sequential on types containing GC references since those types are not blittable and no code outside of the runtime can access those types and will use the auto layout algorithm to change the field layout. But Unity uses a lot of custom icalls that rely on the managed layout matching the native layout. This PR adds an IsSequentialWithRefs flag on MethodTable to indicate the types with sequential layout, but we GC refs. These types: 1. Have the LayoutKind.Sequential flag 2. The type is not blittable and not managed sequential (not blittable but has no GC refs). These are already handled specially by the runtime. 3. Has a parent that is System.Object, System.ValueType, is sequential with refs, is blittable, or is managed sequential This requires only very small changes in the runtime because the explicit layout paths do the work we need already. 1. classlayoutinfo.cpp already calculates field size and sequential offsets so we only need to mark the types as sequential with refs there. 2. In methodtablebuilder.cpp we need to set the field offset to the value calculated in 1. 3. In methodtablebuilder.cpp we run HandleExplicitLayout() which will calculate the type's size (it also, needlessly in our case, validates the field offsets) 4. In methodtablebuilder.cpp we run HandleGCForExplicitLayout() which calculates the GCDesc for the type. The only other needed change was to change the definition of MetadataEnumResult in managedmdimport.hpp. This type is used in an icall that assumes the auto-layout algorithm - interestingly it has nothing to do with GC refs, but the fact that the auto-layout algorithm places larger than pointer sized structs on pointer sized boundaries. This PR also updates CrossGen2 so follows the same layout algorithm. The runtime has a lot of checks that validate that the R2R image match the current runtime layout, size, and GC desc.
We will track the work to undo this change when we can correct things on the Unity side here: #17 |
1d07870
to
9bd5682
Compare
{ | ||
return ComputeSequentialFieldLayout(type, numInstanceFields, isSequentialWithRefs); | ||
} | ||
#else | ||
if (type.IsEnum || MarshalUtils.IsBlittableType(type) || IsManagedSequentialType(type)) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Note that this logic got a major overhaul in dotnet#61759 . I think you will want to do matching cleanup of FEATURE_SEQUENTIAL_LAYOUT_WITH_REFS
mode once you pick up that change. I hope that it will make things simpler overall.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks for letting us know about this pull request, it looks very useful!
* Initial implementation for contract customization fix build errors Move converter rooting to DefaultJsonTypeInfoResolver so that it can be used standalone Fix ConfigurationList.IsReadOnly Minor refactorings (#1) * Makes the following changes: * Move singleton initialization for DefaultTypeInfoResolver behind a static property. * Consolidate JsonSerializerContext & IJsonTypeInfoResolver values to a single field. * Move reflection fallback logic away from JsonSerializerContext and into JsonSerializerOptions * Update src/libraries/System.Text.Json/src/System/Text/Json/Serialization/JsonSerializerOptions.cs * remove testing of removed field Simplify the JsonTypeInfo.CreateObject implemenetation (#2) * Simplify the JsonTypeInfo.CreateObject implemenetation * Update src/libraries/System.Text.Json/src/System/Text/Json/Serialization/Metadata/JsonTypeInfoOfT.cs * Update src/libraries/System.Text.Json/src/System/Text/Json/Serialization/Metadata/JsonTypeInfoOfT.cs Co-authored-by: Krzysztof Wicher <mordotymoja@gmail.com> Co-authored-by: Krzysztof Wicher <mordotymoja@gmail.com> Tests and fixes for JsonTypeInfoKind.None TypeInfo type mismatch tests Allow setting NumberHandling on JsonTypeInfoKind.None test resolver returning wrong type of options JsonTypeInfo/JsonPropertyInfo mutability tests rename test file Move default converter rooting responsibility behind DefaultJsonTypeInfoResolver (#3) * Move default converter rooting responsibility behind DefaultJsonTypeInfoResolver * address feedback Add simple test for using JsonTypeInfo<T> with APIs directly taking it fix and tests for untyped/typed CreateObject uncomment test cases, remove todo More tests and tiny fixes Add a JsonTypeInfoResolver.Combine test for JsonSerializerContext (#4) * Fix JsonTypeInfoResolver.Combine for JsonSerializerContext * Break up failing test Fix simple scenarios for combining contexts (#6) * Fix simple scenarios for combining contexts * feedback JsonSerializerContext combine test with different camel casing Remove unneeded virtual calls & branching when accessing Get & Set delegates (#7) JsonPropertyInfo tests everything minus ShouldSerialize & NumberHandling Update src/libraries/System.Text.Json/src/System/Text/Json/Serialization/JsonConverterOfT.cs Update src/libraries/System.Text.Json/src/System/Text/Json/Serialization/JsonConverterOfT.cs throw InvalidOperationException rather than ArgumentNullException for source gen when PropertyInfo.Name is assigned through JsonPropertyInfoValues tests for duplicated property names and JsonPropertyInfo.NumberHandling Add tests for NumberHandling and failing tests for ShouldSerialize disable the failing test and add extra checks disable remainder of the failing ShouldSerialize tests, fix working one Fix ShouldSerialize and IgnoreCondition interop Add failing tests for CreateObject + parametrized constructors Fix CreateObject support for JsonConstructor types (#10) * Fix CreateObject support for JsonConstructor types * address feedback Make contexts more combinator friendly (#9) * Make contexts more combinator friendly * remove converter cache * redesign test to account for JsonConstructorAttribute * Combine unit tests * address feedback * Add acceptance tests for DataContract attributes & Specified pattern (#11) * Add private field serialization acceptance test (#13) * tests, PR feedback (#14) * PR feedback and extra tests * Shorten class name, remove incorrect check (not true for polimorphic cases) * Make parameter matching for custom properties map property Name with parameter (#16) * Test static initialization with JsonTypeInfo (#17) * Fix test failures and proper fix this time (#18) * Fix test failures and proper fix this time * reinstate ActiveIssueAttribute * PR feedback and adjust couple of tests which don't set TypeInfoResolver * fix IAsyncEnumerable tests * Lock JsonSerializerOptions in JsonTypeInfo.EnsureConfigured() Co-authored-by: Eirik Tsarpalis <eirik.tsarpalis@gmail.com> Co-authored-by: Eirik Tsarpalis <eirik.tsarpalis@gmail.com>
By default CoreCLR will ignore the LayoutKind.Sequential on
types containing GC references since those types are not
blittable and no code outside of the runtime can access those types
and will use the auto layout algorithm to change the field layout.
But Unity uses a lot of custom icalls that rely on the managed layout
matching the native layout.
This PR adds an IsSequentialWithRefs flag on MethodTable to indicate
the types with sequential layout, but we GC refs:
This requires only very small changes in the runtime because the explicit layout paths do the work we need already.
The only other needed change was to change the definition of MetadataEnumResult in managedmdimport.hpp.
This type is used in an icall that assumes the auto-layout algorithm - interestingly it has nothing to do with GC
refs, but the fact that the auto-layout algorithm places larger than pointer sized structs on pointer sized boundaries.
This PR also updates CrossGen2 so follows the same layout algorithm. The runtime has a lot of checks that
validate that the R2R image match the current runtime layout, size, and GC desc.