Superoptimizers attempt to find shorter instruction sequences. This may or may not produce faster code - sometimes you can make code faster by inserting nops!

The idea of superoptimization is to find missed optimizations, which a human can review and then implement directly.

> Superoptimizers attempt to find shorter instruction sequences.

That sounds measurable.

It should have a trophy case then :) About nice found optimizations

The trophy case is here: https://www.cs.utah.edu/~regehr/souper/

Example program 6: https://www.cs.utah.edu/~regehr/souper/output_6.html

    int a, b;
    void fn1() { b = a * ~1 == 1; }

Here we multiply an int by -2. LLVM emitted the multiply and compare, but Souper proved that no int multiplied by -2 can equal 1, so this could be optimized to 0.

By its nature Souper becomes less effective over time as its findings are incorporated into LLVM, so most of the trophies are old!

Yeah, I'd like to know more about what it can do.

Can it optimize sequences of shuffles? If so, it'd be worth a try, even if it takes hours (probably doesn't take that long?), I could let it run until it finds faster sequences than what I have now, and update my code.

But for it to be worth the time for me to figure out how to install and use it, I need some sort of expectation for what I'd get out of trying this instead of something else.

it's not an optimizer and so it's runtime doesn't matter really. It's intended to find new rules (patterns) that can in turn be implemented in llvm as new optimizations. So this thing is just a research step.

If the goal is smaller code then show me a benchmark on the size of the code.

Are you sure about shorter? If you have a model of execution behaviour you can build the 'best' instruction sequence for your operation. Not all superoptimizers are born equal.

It would be kind of amusing if this shortened code by re-rolling an unrolled loop.

Don't forget one of my favories, easily up there with graphical applications without screenthots:

The new programming language without sample code on the landing page.

The redesign of https://www.rust-lang.org/ removed any trace of actual code from its landing page.

Is this designed for mobile only or something?

Anyway, at least the big Get Started button points to some code.

idk, looks fine on my screen. It's a landing page, doesn't have to be information dense.

Besides, over half of website use is via mobile nowadays.

Ouch, that color scheme probably hurts more than the lack of code.

In the context of a superoptimizer benchmarks dont make sense really. You don't know what it's going to find.

For this sort of benchmarking you don’t want to know what it should find, that would be a functional test and not a benchmark.

Compiling and profiling reasonably complex apps against baseline would be enough.

I'd also like to see motivating examples of specific things it does find (that LLVM doesn't), even if that's not a representative benchmark

I think I get what you're saying, I think people would read too much into them though. The results are very dependent on the type of program you have. I wouldn't be surprised if there were 10s of percentage point differences in benchmarks between various examples.

That's the point. Benchmarks would show what code is already squeezed dry by basic LLVM tools and what code could benefit from superoptimization, which (as other comments point out) is the only reasonable basis for deciding to give Souper a try.

I wish the Codeless Code wasn’t in mothballs.