[stdlib] Force-inline Dictionary.subscript(_:, default:)._modify

[lorentey]

Otherwise the autoclosure argument gets allocated on the heap, which leads to a 2x slowdown.

rdar://problem/44936389

[lorentey]

@swift-ci please smoke benchmark

[lorentey]

@swift-ci please test

[swift-ci]

Build comment file:

Performance: -O

TEST	OLD	NEW	DELTA	RATIO
Improvement
DictionarySubscriptDefaultMutationOfObjects	2958	1665	-43.7%	1.78x
DictionarySubscriptDefaultMutation	561	354	-36.9%	1.58x
FrequenciesUsingReduceInto	1904	1204	-36.8%	1.58x
DictionarySubscriptDefaultMutationArrayOfObjects	5477	4119	-24.8%	1.33x

Code size: -O

TEST	OLD	NEW	DELTA	RATIO
Regression
ReduceInto.o	22984	24783	+7.8%	0.93x
Improvement
DictionarySubscriptDefault.o	33820	32891	-2.7%	1.03x

Performance: -Osize

TEST	OLD	NEW	DELTA	RATIO
Improvement
DictionarySubscriptDefaultMutationOfObjects	3205	1799	-43.9%	1.78x
FrequenciesUsingReduceInto	2200	1446	-34.3%	1.52x
DictionarySubscriptDefaultMutation	722	491	-32.0%	1.47x
DictionarySubscriptDefaultMutationArrayOfObjects	5523	4281	-22.5%	1.29x

Code size: -Osize

TEST	OLD	NEW	DELTA	RATIO
Regression
ReduceInto.o	15400	16938	+10.0%	0.91x
Improvement
DictionarySubscriptDefault.o	28884	27939	-3.3%	1.03x

Performance: -Onone

TEST	OLD	NEW	DELTA	RATIO
Improvement
MapReduceLazyCollection	30877	22312	-27.7%	1.38x
MapReduceLazyCollectionShort	43489	35069	-19.4%	1.24x

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

---

[lorentey]

cc @airspeedswift

[jrose-apple]

We really need to get non-heap-allocated non-escaping closures…