Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
360|iDev Concurrency Workshop
Search
sammyd
August 21, 2016
Programming
330
1
Share
Embed
Copy iframe code
Copy JS code
Copy link
Start on current slide
360|iDev Concurrency Workshop
Slides accompanying the 360|iDev concurrency workshop from 2016
sammyd
August 21, 2016
More Decks by sammyd
See All by sammyd
Core Image: Great when it works
sammyd
1
560
iOS Views & Animations: Learning by stealing
sammyd
1
200
Machine Learning on Mobile—a primer
sammyd
0
140
Core ML: A whistlestop tour
sammyd
1
190
DIY DI
sammyd
2
140
iOSConfSG 2017: Decoding Codable
sammyd
3
210
Machine Learning: deciphering the hype
sammyd
0
180
Notify Me, Notify You. Aha!
sammyd
1
200
SwiftConf 2016: Concurrency on iOS
sammyd
1
210
Other Decks in Programming
See All in Programming
コンパウンドプロダクト開発のためのローカルプロセスマネージャー再発明 #layerxgo
izumin5210
0
390
tsc.rip を支える技術 / Kyoto.なんか #8
susisu
0
3.4k
楽しそうなつよつよエンジニアと目が死んでる僕/A brilliant engineer having a blast, and dead-eyed me.
3l4l5
2
260
Introducing Stack Pull Request in GitHub
kkamegawa
0
110
仕様駆動開発の消費期限
watany
20
9k
FastAPI の並行処理モデルを完全に理解する
hoto17296
8
3.1k
生成AIで帳票OCRが「簡単に」作れる時代になった?
kon_shou
0
1.2k
freee が目指す データ マネジメント戦略 AI-Ready 時代を支える 攻めのガバナンスとは
freee
PRO
0
510
30年振りにコンパイラの定数整数除算を改善した
herumi
9
4.3k
複数の Claude Code が"放置"されてしまう問題をCLI ダッシュボードを自作して解決した話
sumihiro3
1
740
業務時間外もAIに働いてもらう話
colorful12
2
7.3k
実装をデザインガイドラインに追従させるための取り組み / 260731-dip-mosh-design-system
dachi023
0
7.9k
Featured
See All Featured
AI Search: Where Are We & What Can We Do About It?
aleyda
0
7.8k
Noah Learner - AI + Me: how we built a GSC Bulk Export data pipeline
techseoconnect
PRO
0
400
Let's Do A Bunch of Simple Stuff to Make Websites Faster
chriscoyier
508
140k
Money Talks: Using Revenue to Get Sh*t Done
nikkihalliwell
0
480
Have SEOs Ruined the Internet? - User Awareness of SEO in 2025
akashhashmi
0
450
The State of eCommerce SEO: How to Win in Today's Products SERPs - #SEOweek
aleyda
2
11k
Automating Front-end Workflow
addyosmani
1369
210k
職位にかかわらず全員がリーダーシップを発揮するチーム作り / Building a team where everyone can demonstrate leadership regardless of position
madoxten
65
56k
Unlocking the hidden potential of vector embeddings in international SEO
frankvandijk
0
900
The Invisible Side of Design
smashingmag
301
52k
Creating an realtime collaboration tool: Agile Flush - .NET Oxford
marcduiker
35
2.6k
Efficient Content Optimization with Google Search Console & Apps Script
katarinadahlin
PRO
1
820
Transcript
Concurrency workshop Sam Davies @iwantmyrealname _ github.com/sammyd git.io/v6waD
concurrency is hard
what is concurrency?
why use concurrency?
common concurrency problems
race condition
Race Condition time Value 1 Thread 1 Thread 2 inc
Race Condition time Value 1 1 Thread 1 Thread 2
inc r1
Race Condition time Value 1 1 1 Thread 1 Thread
2 inc r1 +1
Race Condition time Value 1 1 1 2 Thread 1
Thread 2 inc r1 w2 +1
Race Condition time Value 1 1 1 2 2 Thread
1 Thread 2 inc r1 w2 +1 inc
Race Condition time Value 1 1 1 2 2 2
Thread 1 Thread 2 inc r1 w2 +1 inc r2
Race Condition time Value 1 1 1 2 2 2
2 Thread 1 Thread 2 inc r1 w2 +1 inc r2 +1
Race Condition time Value 1 1 1 2 2 2
2 3 Thread 1 Thread 2 inc r1 w2 +1 inc r2 w3 +1
Race Condition time Value 1 Thread 1 Thread 2 inc
Race Condition time Value 1 1 Thread 1 Thread 2
inc r1 inc
Race Condition time Value 1 1 1 Thread 1 Thread
2 inc r1 inc r1
Race Condition time Value 1 1 1 1 Thread 1
Thread 2 inc r1 inc r1 +1
Race Condition time Value 1 1 1 2 1 Thread
1 Thread 2 inc r1 inc r1 w2 +1
Race Condition time Value 1 1 1 2 1 2
Thread 1 Thread 2 inc r1 +1 inc r1 w2 +1
Race Condition time Value 1 1 1 2 1 2
2 Thread 1 Thread 2 inc r1 w2 +1 inc r1 w2 +1
priority inversion
Priority Inversion time low priority resource
Priority Inversion time low priority resource
Priority Inversion time low medium priority resource
Priority Inversion time low medium high priority resource
Priority Inversion time low medium high priority resource
Priority Inversion time low medium high priority resource
Priority Inversion time low medium high priority resource
deadlock
Deadlock time resource1 resource2 Thread 1 Thread 2
Deadlock time resource1 resource2 Thread 1 Thread 2
Deadlock time resource1 resource2 Thread 1 Thread 2
Deadlock time resource1 resource2 Thread 1 Thread 2
Deadlock time resource1 resource2 Thread 1 Thread 2
Deadlock time resource1 resource2 Thread 1 Thread 2
Deadlock time resource1 resource2 Thread 1 Thread 2
Deadlock time resource1 resource2 Thread 1 Thread 2
$&'( queueueueueue paradigm
Tasks & Queues Thread 1 Thread 2 Queue
Tasks & Queues Thread 1 Thread 2 Queue
Tasks & Queues Thread 1 Thread 2 Queue
Tasks & Queues Thread 1 Thread 2 Queue
Tasks & Queues Thread 1 Thread 2 Queue
Tasks & Queues Thread 1 Thread 2 Queue
Tasks & Queues Thread 1 Thread 2 Queue
feeling brave?
a touch more theory
Thread Safety time Value 1111 Thread 1 Thread 2 Thread
3
Thread Safety time Value 1111 Thread 1 Thread 2 r1111
Thread 3
Thread Safety time Value 1111 1112 Thread 1 Thread 2
r1111 w2222 Thread 3
Thread Safety time Value 1111 1112 1122 Thread 1 Thread
2 r1111 w2222 r1122 Thread 3
Thread Safety time Value 1111 1112 1122 1223 Thread 1
Thread 2 r1111 w2222 r1122 w3333 Thread 3 r1223
Thread Safety time Value 1111 1112 1122 1223 2233 Thread
1 Thread 2 r1111 w2222 r1122 w3333 Thread 3 r1223
Thread Safety time Value 1111 1112 1122 1223 2233 2333
Thread 1 Thread 2 r1111 w2222 r1122 r2333 w3333 Thread 3 r1223
Thread Safety time Value 1111 1112 1122 1223 2233 2333
3333 Thread 1 Thread 2 r1111 w2222 r1122 r2333 w3333 Thread 3 r1223
Thread Safety time Value 1111 1112 1122 1223 2233 2333
3333 3333 3333 Thread 1 Thread 2 r1111 w2222 r1122 r2333 w3333 Thread 3 r1223 r3333 r3333
Dispatch Barrier time task task task task task barrier task
task task task task task task task
Dispatch Barrier time read read read read read read write
read read read read read read
alternatives to
promises networkRequest(url, callback: { data in decompressor(data, callback: { image
in imageResizer(image, callback: { small in filterImage(small, callback: { filtered in displayImage(filtered) }) }) }) })
promises promise<int> int error
promises str
promises promise<img> str f
promises promise<img> str error f img
promises promise<img> str error f img error handler
promises promise<img> str error f img g promise<img> error handler
promises promise<img> str error f img g promise<img> error error
handler img
promises promise<img> str error f img g promise<img> error error
handler img
promises networkRequest(url) .then(decompressor) .then(imageResizer) .then({ (image) in return filterImage(image) }).then(displayImage)
.catch({ (error) in print(error) })
imagine the “promise pipeline” accepted a sequence of values…
that’s reactive programming
conclusion
concurrency is hard
you will get it wrong
• raywenderlich.com/videos • git.io/v6waD • @iwantmyrealname