[benchmark] UnsafeRawBufferPointer._copyContents

[glessard]

Add a benchmark to measure UnsafeRawBufferPointer._copyContents. Prompted by the observation that creating an [UInt8] from an UnsafeRawBufferPointer is slow, while creating a [UInt8] from an UnsafeBufferPointer<UInt8> is much faster. This slowness was reported in https://bugs.swift.org/browse/SR-9604.
The fast path to initialize Array from a Collection ultimately relies on said collection's implementation of _copyContents, therefore we benchmark URBP's implementation.

[glessard]

@swift-ci please benchmark

[swift-ci]

Performance (x86_64): -O

Regression	OLD	NEW	DELTA	RATIO
DictionaryOfAnyHashableStrings_insert	3178	5600	+76.2%	0.57x (?)
Set.isDisjoint.Int50	268	336	+25.4%	0.80x (?)
ObjectiveCBridgeStubFromNSDate	6280	7140	+13.7%	0.88x (?)
DictionaryKeysContainsNative	23	26	+13.0%	0.88x (?)
FlattenListLoop	4242	4771	+12.5%	0.89x (?)
Improvement	OLD	NEW	DELTA	RATIO
XorLoop	1890	1747	-7.6%	1.08x (?)
ArrayPlusEqualFiveElementCollection	8436	7807	-7.5%	1.08x (?)
ArrayAppendReserved	1480	1370	-7.4%	1.08x (?)
Array2D	7216	6688	-7.3%	1.08x (?)
ArrayPlusEqualSingleElementCollection	1927	1786	-7.3%	1.08x (?)
Added	MIN	MAX	MEAN	MAX_RSS
FillArrayFromRawBuffer	686	699	693	—

Code size: -O

Regression	OLD	NEW	DELTA	RATIO
BufferFill.o	7268	9211	+26.7%	0.79x

Performance (x86_64): -Osize

Regression	OLD	NEW	DELTA	RATIO
FlattenListLoop	3975	5226	+31.5%	0.76x (?)
Array2D	6928	7520	+8.5%	0.92x (?)
XorLoop	1747	1890	+8.2%	0.92x (?)
ArrayPlusEqualFiveElementCollection	7696	8325	+8.2%	0.92x (?)
ArrayAppendReserved	1370	1480	+8.0%	0.93x (?)
ArrayPlusEqualSingleElementCollection	1786	1927	+7.9%	0.93x (?)
RandomShuffleLCG2	416	448	+7.7%	0.93x (?)
ArrayAppend	1580	1700	+7.6%	0.93x (?)
Improvement	OLD	NEW	DELTA	RATIO
ObjectiveCBridgeStubFromNSDate	7400	6320	-14.6%	1.17x (?)
DictionaryKeysContainsNative	30	26	-13.3%	1.15x (?)
FlattenListFlatMap	4335	4011	-7.5%	1.08x (?)
Added	MIN	MAX	MEAN	MAX_RSS
FillArrayFromRawBuffer	686	735	709	—

Code size: -Osize

Regression	OLD	NEW	DELTA	RATIO
BufferFill.o	8106	9838	+21.4%	0.82x

Performance (x86_64): -Onone

Regression	OLD	NEW	DELTA	RATIO
String.data.Empty	75	84	+12.0%	0.89x (?)
Added	MIN	MAX	MEAN	MAX_RSS
FillArrayFromRawBuffer	915	1252	1027	—

Code size: -swiftlibs

✅	Benchmark Check Report
⚠️🔤	FillArrayFromRawBuffer name is composed of 5 words. Split FillArrayFromRawBuffer name into dot-separated groups and variants. See http://bit.ly/BenchmarkNaming
⚠️Ⓜ️	FillArrayFromRawBuffer has very wide range of memory used between independent, repeated measurements. FillArrayFromRawBuffer mem_pages [i1, i2]: min=[594, 594] 𝚫=0 R=[0, 203]

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

[glessard]

@swift-ci please benchmark

[glessard]

@swift-ci please smoke test

[swift-ci]

Performance (x86_64): -O

Regression	OLD	NEW	DELTA	RATIO
DictionaryOfAnyHashableStrings_insert	2030	3752	+84.8%	0.54x
Set.isDisjoint.Box25	285	397	+39.3%	0.72x (?)
Set.isDisjoint.Int50	202	241	+19.3%	0.84x (?)
Added	MIN	MAX	MEAN	MAX_RSS
RawBuffer.copyContents	730	742	737	—

Code size: -O

Regression	OLD	NEW	DELTA	RATIO
BufferFill.o	7268	9467	+30.3%	0.77x

Performance (x86_64): -Osize

Added	MIN	MAX	MEAN	MAX_RSS
RawBuffer.copyContents	543	559	552	—

Code size: -Osize

Regression	OLD	NEW	DELTA	RATIO
BufferFill.o	8106	10183	+25.6%	0.80x

Performance (x86_64): -Onone

Added	MIN	MAX	MEAN	MAX_RSS
RawBuffer.copyContents	871	889	877	—

Code size: -swiftlibs

✅	Benchmark Check Report

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 mini
  Model Identifier: Macmini8,1
  Processor Name: 6-Core Intel Core i7
  Processor Speed: 3.2 GHz
  Number of Processors: 1
  Total Number of Cores: 6
  L2 Cache (per Core): 256 KB
  L3 Cache: 12 MB
  Memory: 64 GB

[glessard]

@swift-ci please smoke check

[lorentey]

🚢

[glessard]

@swift-ci please smoke test and merge