rustc_codegen_gcc: Progress Report #22

What is rustc_codegen_gcc?

rustc_codegen_gcc is a GCC ahead-of-time codegen for rustc, meaning that it can be loaded by the existing rustc frontend, but benefits from GCC by having more architectures supported and having access to GCC’s optimizations. It is not to be confused with gccrs, which is a GCC frontend for Rust.

GCC patches status

This month, I did the following commits, which are yet to be sent on the gcc mailing list:

State of rustc_codegen_gcc

Here’s what has been done this month:

I found out there was another SIMD intrinsic that wasn’t implemented properly by running the stdarch tests in the Intel Software Development Emulator. Here are the results of those tests:

# Using the Intel Software Development Emulator:
test result: ok. 4572 passed; 0 failed; 0 ignored; 0 measured; 4 filtered out; finished in 169.77s

# The other tests running natively on my AMD CPU:
test result: ok. 4 passed; 0 failed; 0 ignored; 0 measured; 4572 filtered out; finished in 0.02s

I was unable to run the tests regarding rtm or tbm, so there might still be failures there.

I was also finally able to implement the #[inline] attribute by figuring out some stuff regarding the personality function. I also still tried to fix the unwinding in release mode: I’m starting to believe that the IR sent to rustc_codegen_gcc by rustc doesn’t play well with GCC’s optimizations: I attempt to convert landing pads to try/catch, and the MIR blocks for these landing pads contain goto, so this generates try/catch with goto that quit the try/catch, which seems to play badly with GCC’s optimizations.

Here’s a rough summary of what has been implemented:

Feature Last month completion Completion Delta

Unwinding.

80%

80%

More function and variable attributes.

0%

5%

+5%

LTO.

0%

Debug info.

0%

Endianness support for non-native 128-bit integers.

0%

Rustup distribution.

0%

SIMD for other architectures than x86-64.

0%

Support for new architectures in libraries (libc, object, …) and rustc.

0%

Target features (to detect what is supported in an architecture, like SIMD).

0%

SIMD (x86-64).

Done

Basic and aggregate types.

Done

Operations, local and global variables, constants, functions, basic blocks.

Done

Atomics.

Done

Thread-local storage.

Done

Inline assembly.

Done

Many intrinsics.

Done

Metadata.

Done

Setting optimization level.

Done

Packed structures.

Done

Alignment, symbol visibility, attributes.

Done

128-bit integers.

Done

UI tests progress

Here are the results of running the UI tests in the CI:

Category Last Month This Month Delta

Passed

5126

5125

-1

Failed

66

67

+1

The new failure is related to panics, so I’m assuming this was caused by new optimizations done by the addition of the support for the #[inline] attribute.

For the next month, I’ll continue investigating the issues with unwinding in release mode.

How to contribute

rustc_codegen_gcc

If you want to help on the project itself, please do the following:

  1. Run the tests locally.

  2. Choose a test that fails.

  3. Investigate why it fails.

  4. Fix the problem.

Even if you can’t fix the problem, your investigation could help, so if you enjoy staring at assembly code, have fun!

Crates and rustc

If you would like to contribute on adding support for Rust on currently unsupported platforms, you can help by adding the support for those platforms in some crates like libc and object and also in the rust compiler itself.

Test this project

Otherwise, you can test this project on new platforms and also compare the assembly with LLVM to see if some optimization is missing.

Good first issue

Finally, another good way to help is to look at good first issues. Those are issues that should be easier to start with.

Thanks for your support!

I wanted to personally thank all the people that sponsor this project: your support is very much appreciated.

A special thanks to the following sponsors:

  • saethlin

  • embark-studios

  • Traverse-Research

  • Shnatsel

  • Rust Foundation

A big thank you to bjorn3 for his help, contributions and reviews. And a big thank you to lqd and GuillaumeGomez for answering my questions about rustc’s internals. Another big thank you to Commeownist for his contributions.

Also, a big thank you to the rest of my sponsors:

  • kpp

  • 0x7CFE

  • repi

  • nevi-me

  • oleid

  • acshi

  • joshtriplett

  • djc

  • TimNN

  • sdroege

  • pcn

  • alanfalloon

  • steven-joruk

  • davidlattimore

  • Nehliin

  • colelawrence

  • zmanian

  • alexkirsz

  • berkus

  • belzael

  • jam1garner

  • yvt

  • Shoeboxam

  • evanrichter

  • yerke

  • bes

  • seanpianka

  • srijs

  • kkysen

  • riking

  • Lemmih

  • memoryruins

  • senden9

  • robjtede

  • Jonas Platte

  • spike grobstein

  • Oliver Marshall

  • Sam Harrington

  • Jonas

  • Jeff Muizelaar

  • Eugene Bulkin

  • Chris Butler

  • sierrafiveseven

  • Joseph Garvin

  • MarcoFalke

  • athre0z

  • icewind

  • Tommy Thorn

  • Sebastian Zivota

  • Oskar Nehlin

  • Nicolas Barbier

  • Daniel

  • Thomas Colliers

  • Justin Ossevoort

  • sbstp

  • Chris

  • Bálint Horváth

  • fanquake

  • kiyoshigawa

  • robinmoussu

  • Daniel Sheehan

  • Marvin Löbel

  • nacaclanga

  • Matthew Conolly

  • 0x0177b11f

  • L.apz

  • JockeTF

  • davidcornu

  • stuhood

  • Myrik Lord

  • Mauve

  • icewind1991

  • T

  • nicholasbishop

  • Emily A. Bellows

  • David Vasak

  • Eric Driggers

  • Olaf Leidinger

  • UtherII

  • simonlindholm

  • lemmih

  • Eddddddd

and a few others who preferred to stay anonymous.

Former sponsors/patreons:

  • igrr

  • finfet

  • Alovchin91

  • wezm

  • mexus

  • raymanfx

  • ghost

  • gilescope

  • olanod

  • Denis Zaletaev

  • Chai T. Rex

  • Paul Ellenbogen

  • Dakota Brink

  • Botlabs

  • Cass

  • Oliver Marshall

  • pthariensflame

  • tedbyron

  • sstadick

  • Absolucy

  • rafaelcaricio

  • dandxy89

  • luizirber

  • regiontog

  • vincentdephily

  • zebp

  • Hofer-Julian

  • messense

  • 0xdeafbeef