News for hackers. Stuff that matters.
Re:Basic proxy implementation using io_uring by SPascareli13 on Saturday February 17, 2024 @11:05PM (#39416091)
So, is io_uring used anywhere yet?
Re:Basic proxy implementation using io_uring by i_am_a_peasant on Sunday February 18, 2024 @12:01AM (#39416358)
I used it for some high datarate record applications at work.

Does the job nicely without even half the boilerplate of aio if you use liburing.

Re:Basic proxy implementation using io_uring by dankebitte on Saturday February 17, 2024 @11:40PM (#39416263)
Re:Basic proxy implementation using io_uring by sbstp on Sunday February 18, 2024 @01:16AM (#39416677)
I think there's still a lot of concerns related to security https://www.phoronix.com/news/Google-Restricting-IO_uring
Re:Basic proxy implementation using io_uring by SPascareli13 on Sunday February 18, 2024 @03:55PM (#39423225)
Were any of this vulnerabilities disclosed? I saw that Google blocked it in some places, but without fully understanding what are these vulnerabilities it's hard to make a good judgement on whether it is usable or not in some application.
Re:Basic proxy implementation using io_uring by jandrewrogers on Sunday February 18, 2024 @02:13AM (#39416904)
There are practical backward compatibility and security issues currently that prevent widespread deployment. I’ve been leaning into io_uring going forward because it is clearly the future and brings a lot to the table. It still can’t be used in many environments but I expect it to be mainstream in 2-3 years. Niche, but not for much longer.

Most software doesn’t properly use the old io_submit facility and that has been around for decades, despite the large performance improvements possible. Even when io_uring is widely deployable, most software won’t use it.

Re:Basic proxy implementation using io_uring by jakewins on Sunday February 18, 2024 @06:27AM (#39418212)
What are the open security issues?
Re:Basic proxy implementation using io_uring by surajrmal on Sunday February 18, 2024 @07:39AM (#39418551)
This only applies to Linux distro packages though right? Servers, embedded, mobile, etc only target specific kernels and so can avoid the backwards compatibility issues and in many cases have all of thr latest security patches.
Re:Basic proxy implementation using io_uring by p_l on Sunday February 18, 2024 @07:52AM (#39418622)
Yes.

But sometimes it means that what you have is software that can't run on recent enough kernel to have io_uring (^_^;;)

Re:Basic proxy implementation using io_uring by IAmLiterallyAB on Sunday February 18, 2024 @10:40AM (#39419925)
The old aio API is very inefficient and has awful quirks, such as not working on regular files opened without O_DIRECT. Not surprised no one used it
Re:Basic proxy implementation using io_uring by IAmLiterallyAB on Sunday February 18, 2024 @07:40PM (#39425071)
Am I wrong? Just noticed the downvote. That is my understanding of the old AIO kernel API anyway. The original io_uring paper describes some of the deficiencies of AIO. And the fact that the API is broken on non-O_DIRECT regular file descriptions is well known.
Re:Basic proxy implementation using io_uring by loeg on Monday February 19, 2024 @11:53AM (#39431936)
You're not wrong.
Re:Basic proxy implementation using io_uring by lukeh on Sunday February 18, 2024 @02:16AM (#39416921)
I’m using it for all I/O (file, socket, SPI and UART) in an embedded application written in Swift. Wrappers here:

https://github.com/PADL/IORingSwift

Re:Basic proxy implementation using io_uring by MrBuddyCasino on Sunday February 18, 2024 @06:35AM (#39418242)
What type of app can afford to run on Linux (not bare metal) but still needs such low level performance optimizations?
Re:Basic proxy implementation using io_uring by surajrmal on Sunday February 18, 2024 @07:34AM (#39418529)
Lowering CPU usage can save battery. Just because you are running Linux doesn't mean you don't have a power budget.
Re:Basic proxy implementation using io_uring by p_l on Sunday February 18, 2024 @07:50AM (#39418608)
Among many other projects I touched, we had a database engine which planned to move to io_uring as soon as it could be made available (it wasn't distributed outside company, so it was a question of internal infrastructure)

PREEMPT_RT also includes a lot of possible low-latency work.

io_uring also provides ridiculous performance benefits for normal server code, too.

Re:Basic proxy implementation using io_uring by sophacles on Sunday February 18, 2024 @03:00PM (#39422535)
Context switches are expensive. Often more expensive than operating on the data. io_uring providews a major reduction, or even near elimination of context swithes for io. It provides big win in a lot of server contexts in terms of packets/iops/throughput.
Re:Basic proxy implementation using io_uring by lukeh on Monday February 19, 2024 @03:13AM (#39427253)
Initial motivation was lack of non-blocking SPI I/O in Linux user space. Then it made sense to use it for everything, it’s a straightforward mapping to Swift’s structured concurrency, so why not?
Re:Basic proxy implementation using io_uring by loeg on Sunday February 18, 2024 @04:11AM (#39417480)
My team uses it as part of a high performance storage system at work (but only for disk IO).
Re:Basic proxy implementation using io_uring by rwmj on Sunday February 18, 2024 @06:31AM (#39418227)
libblkio, used by qemu for some block storage, uses it. Independently, qemu also uses io_uring for some Linux file operations.

My experience with io_uring is that it requires a fairly fundamental change in the way you architect software, which makes it hard to retrofit it to existing code (eg. code using poll). If you have a greenfield project that requires very high performance and will only/mainly run on Linux then you should strongly consider it. I'd actually love to hear peoples' thoughts on whether there is a good way to retrofit code.