IO_uring FUTEX Support In Linux 6.7 For A Nice Performance/Efficiency Boost

[phoronix]

Phoronix: IO_uring FUTEX Support In Linux 6.7 For A Nice Performance/Efficiency Boost

In addition to the continued FUTEX2 improvements that landed in Linux 6.7, another pull request merged last week for the new kernel bring FUTEX support to IO_uring...

https://www.phoronix.com/news/IO_uring-FUTEX-Linux-6.7

[timofonic]

So will Wine be faster then?

[brandow]

Still curious on what happened to NTSYNC (aka Winesync), the wine devs never posted any more news about it.

[aviallon]

So will Wine be faster then?

Ah yes, wine faster please.

3njbgk.jpg

[varikonniemi]

1% is considered something to work towards even if there is cross-kernel co-operation needed. 3% seems like a no-brainer. The percentages might seem small, but they have unexpected benefits elsewhere that are not even understood yet.

And of course they might have similar performance penalty, but only way to figure it out is to get it into production.

[cj.wijtmans]

i foresee security issues with io_uring somehow.

[F.Ultra]

i foresee security issues with io_uring somehow.

quite late to the game https://www.phoronix.com/news/Google...cting-IO_uring

[Rovano]

One of the most amazing things of recent times.
5.15+

[cj.wijtmans]

quite late to the game https://www.phoronix.com/news/Google...cting-IO_uring

Yes I know that, io_uring is disabled by any more security orientated entity. I just thought more features would make it an even greater security risk. This is only useful for a gaming OS but even still how many people will know that using their gaming OS should not be used for private stuff. That their apps are not sandboxed. That games can be vectors of stealing all their info etc.

[yump]

Ah yes, that classic video game which has been keeping a great number of poor bastards yoked to Windows... PostgreSQL.

What are you smoking, and can I have some?