Javascript Performance Thread

Discussion about official Mozilla Firefox builds
mayankleoboy1
Posts: 471
Joined: February 25th, 2013, 9:52 pm

Re: Javascript Performance Thread

Post by mayankleoboy1 »

greg86 wrote:Why isn't there any progress on any Benchmark on the front page of AWFY in the last month?
I know Sunspider isn't that representative any more but why its staying at the front page?
Are there any better "modern" Benchmarks that show more the improvements in the JS Engine?
AFAIU, the JS team is not specifically trying to improve the Octane/Kraken/Sunspider benchmarks. They are mostly trying to improve more "real-world" code.
They have been working on the CacheIR work, but keep side-tracking to improve other stuff, like Google Docs, six-speed and others.
jandem
Posts: 93
Joined: September 18th, 2008, 2:57 pm

Re: Javascript Performance Thread

Post by jandem »

A lot of the CacheIR work so far has been converting existing code and fixing performance issues that show up on real-world code. We've fixed a lot of stupid performance issues there, but the shell benchmarks improve less from that because over the years we've spent a lot of time investigating these and we fixed the worst problems a long time ago. Now that the most complicated IC code has been converted to CacheIR, we can spend more time adding new optimizations. Bug 1328140 for instance should definitely help some of the shell benchmarks and pave the way for further improvements there. Hopefully later this month I will have some time to look into some of these benchmarks.
mayankleoboy1
Posts: 471
Joined: February 25th, 2013, 9:52 pm

Re: Javascript Performance Thread

Post by mayankleoboy1 »

@Jandem:
Any plans to bring back https://bugzilla.mozilla.org/show_bug.cgi?id=785905 ?
jandem
Posts: 93
Joined: September 18th, 2008, 2:57 pm

Re: Javascript Performance Thread

Post by jandem »

mayankleoboy1 wrote:@Jandem:
Any plans to bring back https://bugzilla.mozilla.org/show_bug.cgi?id=785905 ?
If I remember correctly, the performance win from that was never very big (smaller than I expected) and moving all of IonBuilder off-thread is hard. It would probably be simpler to move some parts of it to the compilation thread. We should look into this again at some point, but for now there are still a lot of potential perf improvements elsewhere.
Timvde
Posts: 269
Joined: February 7th, 2013, 3:19 pm

Re: Javascript Performance Thread

Post by Timvde »

Would you mind making a list? I am very curious what the JS has planned for the future!
mayankleoboy1
Posts: 471
Joined: February 25th, 2013, 9:52 pm

Re: Javascript Performance Thread

Post by mayankleoboy1 »

Timvde wrote:Would you mind making a list? I am very curious what the JS has planned for the future!
For starters, I would say all the deps of https://bugzilla.mozilla.org/show_bug.cgi?id=1259927 , specially https://bugzilla.mozilla.org/show_bug.cgi?id=1324561 . From the comment, fixing that bug would lead to more cases open to optimization.
Timvde
Posts: 269
Joined: February 7th, 2013, 3:19 pm

Re: Javascript Performance Thread

Post by Timvde »

Google enabled TurboFan (new optimizing engine) and Ignition (interpreter) by default: https://chromium.googlesource.com/v8/v8 ... 40d600d6fc

It cost them on all AWFY benchmarks, but most notably of course on Octane. I didn't expect them to already flip the switch.

It also looks like there's a new benchmark suite, Six Speed: https://www.arewefastyet.com/#machine=2 ... =six-speed / https://kpdecker.github.io/six-speed/
mayankleoboy1
Posts: 471
Joined: February 25th, 2013, 9:52 pm

Re: Javascript Performance Thread

Post by mayankleoboy1 »

Some progress going on in https://bugzilla.mozilla.org/show_bug.cgi?id=1383358
Since the Project Quantum Flow is put on hold a bit, whats the general direction that the JS engine is heading to these days?
Timvde
Posts: 269
Joined: February 7th, 2013, 3:19 pm

Re: Javascript Performance Thread

Post by Timvde »

Mozilla announced an experiment last month, called HolyJit. It's written in Rust, and iiuc, it is basically a framework to automatically create an optimizing compiler starting from an interpreter in Rust. Sounds very cool!
User avatar
Omega X
Posts: 8225
Joined: October 18th, 2007, 2:38 pm
Location: A Parallel Dimension...

Re: Javascript Performance Thread

Post by Omega X »

Timvde wrote:Mozilla announced an experiment last month, called HolyJit. It's written in Rust, and iiuc, it is basically a framework to automatically create an optimizing compiler starting from an interpreter in Rust. Sounds very cool!
Oh snap. A rewrite in Rust. This is going to be interesting.
Timvde
Posts: 269
Joined: February 7th, 2013, 3:19 pm

Re: Javascript Performance Thread

Post by Timvde »

It's not "just" a rewrite. Read the readme on Github, it's really cool!
eltomito
Posts: 8
Joined: December 8th, 2013, 5:52 pm

Re: Javascript Performance Thread

Post by eltomito »

Timvde wrote:Mozilla announced an experiment last month, called HolyJit.
I'm afraid the name's way too cool to stick! :D
User avatar
sciguyryan
Folder@Home
Posts: 2181
Joined: November 10th, 2004, 1:33 pm
Location: Wales

Re: Javascript Performance Thread

Post by sciguyryan »

This bug looks like it could be interesting: https://bugzilla.mozilla.org/show_bug.cgi?id=1396822

I'm not sure if it's going to improve performance but it looks like it will improve memory usage, at very least.
Cheers!

Ryan Jones
Ver Greeneyes
Posts: 1030
Joined: June 28th, 2008, 4:57 am

Re: Javascript Performance Thread

Post by Ver Greeneyes »

Jan de Mooij (:jandem) just wrote a blog post about the performance work he did this year. Very impressive!
Lurtz
Posts: 359
Joined: June 12th, 2016, 12:25 pm

Re: Javascript Performance Thread

Post by Lurtz »

Edit: Ooops, wrong thread.
Post Reply