Status update, August 2020 August 16, 2020 on Drew DeVault's blog

Greetings! Today is another rainy day here in Philadelphia, which rather sours my plans of walking over to the nearby cafe to order some breakfast to-go. But I am tired, and if I’m going to make it to the end of this blog post in one piece, I’m gonna need a coffee. brb.

Hey, that was actually pretty refreshing. It’s just drizzling, and the rain is nice and cool. Alright, here goes! What’s new? I’ll leave the Wayland news for Simon Ser’s blog this month - he’s been working on some exciting stuff. The BARE encoding announced last month has received some great feedback and refinements, and there are now six projects providing BARE support for their author’s favorite programming language1. There have also been some improvements to the Go implementation which should help with some SourceHut plans later on.

On the subject of SourceHut, I’ve focused mainly on infrastructure improvements this month. There is a new server installed for hg.sr.ht, which will also be useful as a testbed for additional ops work planned for future expansion. Additionally, the PostgreSQL backup system has been overhauled and made more resilient, both to data loss and to outages. A lot of other robustness improvements have been made fleet-wide in monitoring. I’ll be working on more user-facing features again next month, but in the meanwhile, contributors like наб have sent many patches in which I’ll cover in detail in the coming “What’s cooking” post for sourcehut.org.

Otherwise, I’ve been taking it easy this month. I definitely haven’t been spending a lot of my time on a secret project, no sir. Thanks again for your support! I’ll see you next month.

?
use io;
use io_uring = linux::io_uring;
use linux;
use strings;

export fn main void = { let uring = match (io_uring::init(256u32, 0u32)) { err: linux::error => { io::println(“io_uring::init error:"); io::println(linux::errstr(err)); return; }, u: io_uring::io_uring => u, };

let buf: [8192]u8 = [0u8...];
let text: nullable *str = null;
let wait = 0u;
let offs = 0z;
let read: *io_uring::sqe = null: *io_uring::sqe,
    write: *io_uring::sqe = null: *io_uring::sqe;
let eof = false;

while (!eof) {
    read = io_uring::must_get_sqe(&uring);
    io_uring::prep_read(read, linux::STDIN_FILENO,
        &buf, len(buf): u32, offs);
    io_uring::sqe_set_user_data(read, &read);
    wait += 1u;

    let ev = match (io_uring::submit_and_wait(&uring, wait)) {
        err: linux::error => {
            io::println("io_uring::submit error:");
            io::println(linux::errstr(err));
            return;
        },
        ev: uint => ev,
    };

    wait -= ev;

    for (let i = 0; i < ev; i += 1) {
        let cqe = match (io_uring::get_cqe(&uring, 0u, 0u)) {
            err: linux::error => {
                io::println("io_uring::get_cqe error:");
                io::println(linux::errstr(err));
                return;
            },
            c: *io_uring::cqe => c,
        };

        if (io_uring::cqe_get_user_data(cqe) == &read) {
            if (text != null) {
                free(text);
            };

            if (cqe.res == 0) {
                eof = true;
                break;
            };

            text = strings::must_decode_utf8(buf[0..cqe.res]);
            io_uring::cqe_seen(&uring, cqe);

            write = io_uring::must_get_sqe(&uring);
            io_uring::prep_write(write, linux::STDOUT_FILENO,
                text: *char, len(text): u32, 0);
            io_uring::sqe_set_user_data(write, &write);
            wait += 1u;
            offs += cqe.res;
        } else if (io_uring::cqe_get_user_data(cqe) == &write) {
            assert(cqe.res > 0);
            io_uring::cqe_seen(&uring, cqe);
        } else {
            assert(false, "Unknown CQE user data");
        };
    };
};

io_uring::close(&uring);

};

hmm?

I might note that I wrote this program to test my io_uring wrapper; it's not representative of how normal programs will do I/O in the future.


  1. Or in some cases, the language the author is begrudgingly stuck with. ↩︎

Have a comment on one of my posts? Start a discussion in my public inbox by sending an email to ~sircmpwn/public-inbox@lists.sr.ht [mailing list etiquette]

Articles from blogs I read Generated by openring

Saying "no" to unethical tasks

Back in spring 2019, I was a student working as an intern at the Intel Open Source Graphics Center in Finland. I was mainly focused on improving igt-gpu-tools, the test suite that runs each time a patch is submitted for the i915 kernel driver. I really liked…

via emersion November 23, 2020

SourceHut's second year in alpha

Today is the second anniversary of SourceHut being made available to the general public during its alpha phase of development, and the end of the fourth year of development. What a year it’s been! If only I could send a warning to the younger me who wrote la…

via Blogs on Sourcehut November 15, 2020

Graphics for JVM

Introducing Skija, Java bindings for Google graphics library, Skia

via tonsky.me November 14, 2020