Stream: t-compiler/wg-mir-opt

Topic: Opt kinds covered by the benchmarks


Félix Fischer (Jun 29 2020 at 01:26, on Zulip):

I would imagine that every family of optimizations would affect the benchmark suite's projects differently.

Félix Fischer (Jun 29 2020 at 01:27, on Zulip):

Is there documentation on what kind of difficulties each benchmark represents?

Félix Fischer (Jun 29 2020 at 01:27, on Zulip):

Like, compilation difficulties.

Félix Fischer (Jun 29 2020 at 01:29, on Zulip):

I'm thinking something like this:

Félix Fischer (Jun 29 2020 at 01:30, on Zulip):

I would like to know if we have something like this, and if there are like, groups to which the different benchmark projects belong to. Groups that for example, say what things are hard to compile quickly on their belonging projects.

Félix Fischer (Jun 29 2020 at 01:32, on Zulip):

@Wesley Wiser @oli do you know of this topic?
It might seem like a very abstract idea, but since I'm applying for a project backed by uni to work on mir-opts for a while, this knowledge about our metrics will be very useful to tell them how I will be measuring the success of my work :3

Wesley Wiser (Jun 29 2020 at 01:38, on Zulip):

@Félix Fischer This is the only documentation I'm aware of https://github.com/rust-lang/rustc-perf/tree/master/collector/benchmarks

Félix Fischer (Jun 29 2020 at 01:38, on Zulip):

Thank you Wesley! :heart:

Last update: Sep 28 2020 at 15:15UTC