Add swift_getTypeByMangledNameInContextInMetadataState such that we can

[aschwaighofer]

use getTypeByMangledName when abstract metadata state is requested

This can significantly reduce the code size of apps constructing deeply
nested types with conditional conformances.

Requires a new runtime.

rdar://57157619

[aschwaighofer]

@swift-ci Please test

[swift-ci]

Build failed
Swift Test Linux Platform
Git Sha - 0d324d2

[aschwaighofer]

@swift-ci Please test

[swift-ci]

Build failed
Swift Test Linux Platform
Git Sha - 0d324d2

[swift-ci]

Build failed
Swift Test OS X Platform
Git Sha - 0d324d2

[jckarter]

Thanks for implementing this, Arnold! I wonder whether, if we're targeting Swift >= 5.2, we could only emit __swift_instantiateConcreteTypeFromMangledNameAbstract and use it for all metadata requests. That would probably mean a couple more instructions at call sites to materialize the metadata request kind argument, though.

[jckarter]

It'd also be good to add the swift_getTypeByMangledNameInEnvironmentInMetadataState variant, so we have it available for future use to instantiate metadata in generic environments that don't have an associated type context.

[aschwaighofer]

@swift-ci Please test

[swift-ci]

Build failed
Swift Test OS X Platform
Git Sha - 44b99b7

[swift-ci]

Build failed
Swift Test Linux Platform
Git Sha - 44b99b7

[aschwaighofer]

@swift-ci Please test

[swift-ci]

Build failed
Swift Test Linux Platform
Git Sha - 84bb456

[aschwaighofer]

@swift-ci please test

[swift-ci]

Build failed
Swift Test OS X Platform
Git Sha - 84bb456

[swift-ci]

Build failed
Swift Test Linux Platform
Git Sha - 84bb456

[aschwaighofer]

@swift-ci Please test linux

[airspeedswift]

@swift-ci please benchmark

[swift-ci]

Performance: -O

Regression	OLD	NEW	DELTA	RATIO
ObjectiveCBridgeStubFromNSDateRef	3830	4200	+9.7%	0.91x (?)
Improvement	OLD	NEW	DELTA	RATIO
FlattenListLoop	5295	4453	-15.9%	1.19x (?)
FlattenListFlatMap	10291	8988	-12.7%	1.14x (?)
AngryPhonebook.Cyrillic.Small	932	853	-8.5%	1.09x (?)

Code size: -O

Performance: -Osize

Improvement	OLD	NEW	DELTA	RATIO
ObjectiveCBridgeStubFromNSDateRef	4220	3800	-10.0%	1.11x (?)

Code size: -Osize

Performance: -Onone

Regression	OLD	NEW	DELTA	RATIO
AngryPhonebook	2996	3458	+15.4%	0.87x (?)
AngryPhonebook.ASCII.Small	131	150	+14.5%	0.87x (?)
DataAppendDataSmallToSmall	4080	4480	+9.8%	0.91x (?)
NormalizedIterator_ascii	474	520	+9.7%	0.91x (?)
CharIndexing_chinese_unicodeScalars_Backwards	439640	481200	+9.5%	0.91x (?)
CharIndexing_punctuatedJapanese_unicodeScalars_Backwards	101000	108920	+7.8%	0.93x (?)
Improvement	OLD	NEW	DELTA	RATIO
FloatingPointPrinting_Float_interpolated	60800	54800	-9.9%	1.11x (?)
ObjectiveCBridgeStubFromNSDateRef	4830	4370	-9.5%	1.11x (?)

Code size: -swiftlibs

How to read the data

The tables contain differences in performance which are larger than 8% and differences in code size which are larger than 1%.

If you see any unexpected regressions, you should consider fixing the
regressions before you merge the PR.

Noise: Sometimes the performance results (not code size!) contain false
alarms. Unexpected regressions which are marked with '(?)' are probably noise.
If you see regressions which you cannot explain you can try to run the
benchmarks again. If regressions still show up, please consult with the
performance team (@eeckstein).

Hardware Overview

  Model Name: Mac Pro
  Model Identifier: MacPro6,1
  Processor Name: 12-Core Intel Xeon E5
  Processor Speed: 2.7 GHz
  Number of Processors: 1
  Total Number of Cores: 12
  L2 Cache (per Core): 256 KB
  L3 Cache: 30 MB
  Memory: 64 GB