Removes redundant buffer zeroing in lowercased() and uppercased() by using `init(unsafeUninitializedCapacity:initializingWith:) #30895

valeriyvan · 2020-04-08T19:20:12Z

Removes redundant buffer zeroing in lowercased() and uppercased() by using `init(unsafeUninitializedCapacity:initializingWith:)

CodaFi · 2020-04-15T01:10:49Z

@swift-ci test

CodaFi · 2020-04-15T01:10:53Z

@swift-ci benchmark

swift-ci · 2020-04-15T02:56:40Z

Performance: -O

Regression	OLD	NEW	DELTA	RATIO
StringInterpolationManySmallSegments	8900	11100	+24.7%	0.80x (?)
AngryPhonebook.Strasse.Small	647	744	+15.0%	0.87x (?)
Set.subtracting.Empty.Box	8	9	+12.5%	0.89x (?)
Data.hash.Medium	31	34	+9.7%	0.91x (?)
StringHashing_fastPrenormal	570	620	+8.8%	0.92x (?)

Improvement	OLD	NEW	DELTA	RATIO
LessSubstringSubstring	30	22	-26.7%	1.36x
EqualSubstringSubstringGenericEquatable	30	22	-26.7%	1.36x
EqualSubstringString	30	22	-26.7%	1.36x
LessSubstringSubstringGenericComparable	30	22	-26.7%	1.36x
EqualStringSubstring	31	23	-25.8%	1.35x
EqualSubstringSubstring	30	23	-23.3%	1.30x
StringComparison_longSharedPrefix	357	323	-9.5%	1.11x (?)
ObjectiveCBridgeStubToNSDate2	360	330	-8.3%	1.09x (?)
ArraySetElement	284	263	-7.4%	1.08x (?)
PrefixWhileSequence	178	165	-7.3%	1.08x (?)
ObjectiveCBridgeStubToNSDateRef	2320	2160	-6.9%	1.07x (?)

Code size: -O

Performance: -Osize

Regression	OLD	NEW	DELTA	RATIO
AngryPhonebook.Strasse.Small	654	755	+15.4%	0.87x
StringHashing_fastPrenormal	560	630	+12.5%	0.89x

Improvement	OLD	NEW	DELTA	RATIO
LessSubstringSubstring	31	23	-25.8%	1.35x
EqualStringSubstring	31	23	-25.8%	1.35x
LessSubstringSubstringGenericComparable	31	23	-25.8%	1.35x (?)
FlattenListFlatMap	4084	3126	-23.5%	1.31x (?)
EqualSubstringSubstring	30	23	-23.3%	1.30x (?)
EqualSubstringSubstringGenericEquatable	30	23	-23.3%	1.30x
EqualSubstringString	30	23	-23.3%	1.30x
Set.subtracting.Empty.Box	9	8	-11.1%	1.12x (?)
StringBuilderLong	1060	970	-8.5%	1.09x (?)
AngryPhonebook	273	252	-7.7%	1.08x (?)
ArraySetElement	283	262	-7.4%	1.08x (?)
Array2D	4544	4208	-7.4%	1.08x (?)

Code size: -Osize

Performance: -Onone

Regression	OLD	NEW	DELTA	RATIO
AngryPhonebook.Strasse.Small	745	851	+14.2%	0.88x (?)

Improvement	OLD	NEW	DELTA	RATIO
LessSubstringSubstring	35	27	-22.9%	1.30x
EqualSubstringSubstringGenericEquatable	34	27	-20.6%	1.26x
EqualSubstringSubstring	35	28	-20.0%	1.25x (?)
EqualStringSubstring	35	28	-20.0%	1.25x
EqualSubstringString	35	28	-20.0%	1.25x
LessSubstringSubstringGenericComparable	34	28	-17.6%	1.21x
ArrayAppendLatin1Substring	34776	31644	-9.0%	1.10x (?)
CharacterLiteralsLarge	377	348	-7.7%	1.08x (?)
Set.subtracting.Empty.Int	83	77	-7.2%	1.08x (?)

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 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

xwu · 2020-04-15T03:05:14Z

stdlib/public/core/String.swift

-            output.baseAddress._unsafelyUnwrappedUnchecked,
-            Int32(output.count),
+            buf.baseAddress._unsafelyUnwrappedUnchecked,
+            Int32(buf.count),


Nit: In general, the stdlib code base avoids ad-hoc abbreviations such as “buf” or “ret”; I realize some of these were not created by you, but ideally we wouldn’t introduce more.

Renamed variables.

xwu · 2020-04-15T03:07:29Z

stdlib/public/core/String.swift

+          &err))
+        initializedCount = min(len, uChars.count)
+        if len > uChars.count {
+            throw DummyError.dummy


Sorry, who catches this error? Does this code path represent a violation of an invariant or precondition?

Rewritten without throwing.

milseman · 2020-04-17T14:17:49Z

This looks much better without the error throwing, thanks!

xwu · 2020-04-17T23:26:26Z

@swift-ci benchmark

swift-ci · 2020-04-18T00:07:33Z

Performance: -O

Regression	OLD	NEW	DELTA	RATIO
ObjectiveCBridgeStubFromNSDateRef	4750	5760	+21.3%	0.82x (?)
EqualSubstringSubstring	40	44	+10.0%	0.91x (?)
EqualSubstringSubstringGenericEquatable	40	44	+10.0%	0.91x (?)
EqualSubstringString	40	44	+10.0%	0.91x
EqualStringSubstring	41	45	+9.8%	0.91x (?)
SortStringsUnicode	2870	3110	+8.4%	0.92x (?)

Improvement	OLD	NEW	DELTA	RATIO
Breadcrumbs.MutatedUTF16ToIdx.ASCII	5	4	-20.0%	1.25x (?)

Code size: -O

Performance: -Osize

Regression	OLD	NEW	DELTA	RATIO
String.data.Empty	39	50	+28.2%	0.78x (?)
String.data.Small	42	52	+23.8%	0.81x (?)
String.data.LargeUnicode	137	152	+10.9%	0.90x (?)
EqualSubstringSubstring	41	45	+9.8%	0.91x (?)
LessSubstringSubstring	41	45	+9.8%	0.91x
EqualStringSubstring	41	45	+9.8%	0.91x (?)
EqualSubstringSubstringGenericEquatable	41	45	+9.8%	0.91x (?)
LessSubstringSubstringGenericComparable	41	45	+9.8%	0.91x (?)
SortStringsUnicode	2930	3180	+8.5%	0.92x (?)
AngryPhonebook.Strasse.Small	943	1020	+8.2%	0.92x (?)

Improvement	OLD	NEW	DELTA	RATIO
Diffing.Same	9	8	-11.1%	1.12x (?)

Code size: -Osize

Performance: -Onone

Regression	OLD	NEW	DELTA	RATIO
StringToDataMedium	4600	5100	+10.9%	0.90x (?)
EqualSubstringSubstring	47	51	+8.5%	0.92x (?)
LessSubstringSubstring	47	51	+8.5%	0.92x (?)
EqualSubstringSubstringGenericEquatable	47	51	+8.5%	0.92x
LessSubstringSubstringGenericComparable	47	51	+8.5%	0.92x

Improvement	OLD	NEW	DELTA	RATIO
StringWordBuilder	2780	2580	-7.2%	1.08x (?)
StringWordBuilderReservingCapacity	2710	2530	-6.6%	1.07x (?)

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

valeriyvan · 2020-04-30T19:40:54Z

ping

valeriyvan · 2020-05-27T09:31:36Z

ping

stdlib/public/core/String.swift

xwu · 2020-05-27T16:53:16Z

Can you squash the commits (except the one merging master)? Then let's kick off another test.

valeriyvan · 2020-05-27T20:27:58Z

Can you squash the commits (except the one merging master)? Then let's kick off another test.

done

xwu · 2020-05-27T21:28:40Z

@swift-ci test

xwu · 2020-05-28T15:36:00Z

@swift-ci test Linux platform

xwu · 2020-05-28T17:41:31Z

@swift-ci smoke test Linux platform

xwu · 2020-05-30T18:27:36Z

@swift-ci smoke test Linux platform

xwu · 2020-06-01T15:08:50Z

@swift-ci test Linux platform

xwu · 2020-06-01T19:06:16Z

Let's see if this will do it:
@swift-ci clean test Linux platform

swift-ci · 2020-06-01T21:01:12Z

Build failed
Swift Test Linux Platform
Git Sha - 04e6373

natecook1000 · 2020-06-08T17:43:42Z

@swift-ci Please smoke test

theblixguy requested a review from milseman April 8, 2020 19:40

xwu reviewed Apr 15, 2020

View reviewed changes

valeriyvan requested a review from xwu April 16, 2020 05:43

xwu reviewed May 27, 2020

View reviewed changes

Removes redundant buffer zeroing in lowercased() and uppercased()

04e6373

valeriyvan force-pushed the RemoveRedundantZeroingString branch from 809dcab to 04e6373 Compare May 27, 2020 20:25

This comment has been minimized.

Sign in to view

xwu merged commit 1888534 into swiftlang:master Jun 9, 2020

valeriyvan deleted the RemoveRedundantZeroingString branch February 20, 2023 08:09

Removes redundant buffer zeroing in lowercased() and uppercased() by using `init(unsafeUninitializedCapacity:initializingWith:) #30895

Removes redundant buffer zeroing in lowercased() and uppercased() by using `init(unsafeUninitializedCapacity:initializingWith:) #30895

Uh oh!

Conversation

valeriyvan commented Apr 8, 2020

Uh oh!

CodaFi commented Apr 15, 2020

Uh oh!

CodaFi commented Apr 15, 2020

Uh oh!

swift-ci commented Apr 15, 2020

Performance: -O

Code size: -O

Performance: -Osize

Code size: -Osize

Performance: -Onone

Code size: -swiftlibs

Uh oh!

xwu Apr 15, 2020

Choose a reason for hiding this comment

Uh oh!

valeriyvan Apr 16, 2020

Choose a reason for hiding this comment

Uh oh!

xwu Apr 15, 2020

Choose a reason for hiding this comment

Uh oh!

valeriyvan Apr 16, 2020

Choose a reason for hiding this comment

Uh oh!

milseman commented Apr 17, 2020

Uh oh!

xwu commented Apr 17, 2020

Uh oh!

swift-ci commented Apr 18, 2020

Performance: -O

Code size: -O

Performance: -Osize

Code size: -Osize

Performance: -Onone

Code size: -swiftlibs

Uh oh!

valeriyvan commented Apr 30, 2020

Uh oh!

valeriyvan commented May 27, 2020

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

xwu commented May 27, 2020

Uh oh!

valeriyvan commented May 27, 2020

Uh oh!

xwu commented May 27, 2020

Uh oh!

This comment has been minimized.

This comment has been minimized.

xwu commented May 28, 2020

Uh oh!

This comment has been minimized.

xwu commented May 28, 2020

Uh oh!

xwu commented May 30, 2020

Uh oh!

xwu commented Jun 1, 2020

Uh oh!

This comment has been minimized.

xwu commented Jun 1, 2020

Uh oh!

swift-ci commented Jun 1, 2020

Uh oh!

natecook1000 commented Jun 8, 2020

Uh oh!

Uh oh!