Swift Performance - iOSDevUK

I presented today at iOSDevUK in Aberystwyth with some more material on Swift Performance and how to profile Swift. Also on wrapping classes in value types. Unfortunately much of the presentation was a demo and there is no video but I'm posting the materials here and will try to make a follow up post covering some more of the details in the next week or two (no promises).

Slides (note that slides a terrible communication mechanism and caveats and subtleties in the talk may be lost):

Unexpected Assertion Behaviour - Radar 21826180

Apologies for the lack of posts, full time work and family time among other things have been keeping me from finishing off Swift performance stuff. The new Swift 2.0 features look really good but again I haven't got stuck in yet. This is a quick post about an issue concerning me since I noticed it when reading Erica Sadun's post about assertions. I actually thought it couldn't be correct when I first read it so unexpected was it but when I checked the documentation it matched exactly what Erica said.

Assertions and Unchecked Builds

I use assertions for things that shouldn't happen but sometimes might. For example corrupt or unprocessable network data where I write code to safely handle the case but I also want to know instantly and start debugging the issue (whether it is in the app or the server). If you do the same and have assertions which could happen then you must not build in -Ounchecked (or in Swift 2.0 with Disable Safety Checks enabled) otherwise in the event of assertions not being true you will be in undefined behaviour.

How Swift is Swift?

Slides, video and links related to my Swift Summit talk on Swift performance. Video should be available later is now available and it might be worth watching Airspeed's talk that preceded mine before watching it when it is available I'm also expecting a blog post from him shortly on his static vs. compile time talk which is also highly relevant to optimisation.

The key point in Swift is that as the compiler gets better there is no need to choose between nice code and fast code. With a little thought and knowledge it is usually possible to get close to the speed of lowish level C code with nice abstractions. Some may say that C++ is there already but I enjoy writing Swift code much more and C++ is hampered by it's C legacy and choices made many years ago (for good reasons) from becoming a truly nice language in my view (non-Nullable by default would be hard to retrofit for example).

There were also many other excellent talks at Swift Summit, and all were at least good. I'm looking forward to watching several again.

Simple Library Free JSON Parsing in Swift

Parsing JSON in Swift without adding libraries could be tricky and annoying at least until nil coalescing was added and optional chaining was upgraded late in the original betas but that really isn't the case at least since Swift 1.0 was final. There are clearly a wide range of JSON libraries for Swift ranging from SwiftyJSON which just tidies up extractions from deep in nested structures to Argo which will feel comfortable for those coming from Haskell and look clever but be hard to read for those not familiar with the syntax.

I want to show people that there is really nothing very tricky or ugly about parsing JSON data in Swift even without libraries to help.

This post was trigger by seeing Jameson Quave's new lightweight Luma library and the syntax in the project readme (at the time of writing) would work without a added library on the result of parsing with Cocoa's included NSJSONSerialization class. Please note that there is nothing wrong with the Luma library or anything I have seen written about it but I wanted to make clear how lightweight it is and how easy these things are to do without using a library at all.

I confirmed that the syntax was available in the Playground without using the library and there is no problem just using exactly the same syntax on the dictionary you get back from NSJSONSerialization as in the Luma example. The only difference in the presentation of printing the parsed structures.


Swift 1.2 Update (Xcode 6.3 beta 2) - Performance

Apple have shipped another major update (release notes - registered devs only) only two weeks after shipping the beta 1. There are major updates to Swift Playgrounds, additional syntax support for more flexible `if let`, added a zip function, various other tweaks and fixed a tonne of bugs. This is all on top of the major changes in 1.2 beta 1 that I discussed at Swift London.

Erica Sadun has already blogged about the Playgrounds and `if let` changes and I'm sure that there will be plenty more over the next week. I don't intend to go over that ground but instead discuss the performance changes in Swift 1.2 versions. [Update: I should also have mentioned Jamesson Quave's post that covers the new zip function.]

The performance jumps are very significant in Swift 1.2 and certainly in Beta 2 the steps necessary to optimise your code are significantly changed. This post will cover some information on the improvements and how to get the best out of the Swift compiler. In most cases I have encountered the performance is now very close to C/C++ code and may be faster at times.

Swift 1.2 beta 2 Big Performance News

High performance can be achieved in Xcode 6.3 beta 2 without making code changes to achieve the result that were required to get close previously. Specifically the performance gains of these changes seem to have been largely eliminated:

  1. `final` methods/properties which used to be massive (although I still recommend it where possible for design/safety reasons).
  2. unsafeMutableBuffer instead of array (at least in -Ounchecked builds)
  3. moving code into the same file to allow the compiler to inline (with -whole-module-optimizations build option new in Xcode 6.3 beta 2).

There also seems to have been a slight general performance gain (about 5% above beta 1 from my initial tests).

Is Swift Fast Yet?

Yes. I've always believed that the language had potential to be as fast or faster than C because of strict compile time type checking, immutability and a lack of pointer aliasing concerns. With the Xcode 6.3 betas that promise is being delivered on. I'm sure the work is continuing and there is more to come but it is already fast in most cases. My thanks to the Apple team for making a fast platform but you have just taken away my ability to look impressive by speeding up code significantly with a few simple changes - you bastards.