mastodon.ie is one of the many independent Mastodon servers you can use to participate in the fediverse.
Irish Mastodon - run from Ireland, we welcome all who respect the community rules and members.

Administered by:

Server stats:

1.6K
active users

#kernel

12 posts10 participants1 post today

🐧The Linux Kernel Seeing Rare Code Activity Around SPARC64 • Phoronix

「 SPARC64 was the last architecture not using the generic vDSO library code that in turn prevented some necessary code clean-ups. With this patch series transitioning the SPARC code to the generic vDSO infrastructure saves several hundred lines of code and slightly reducing the SPARC maintenance burden 」

phoronix.com/news/Linux-SPARC6

www.phoronix.comThe Linux Kernel Seeing Rare Code Activity Around SPARC64One of the CPU architectures continuing to be supported by the mainline Linux kernel but rarely ever seeing any code activity is the SPARC64 architecture port for the once-interesting processors from Sun Microsystems.
Replied in thread

@edboythinks personally I'm in the process of.migrating from #Ubuntu fonts to #B612 because they are superior yet more permissively licensed.

Continued thread

2/ Another article about a feature recently added to the #XFS filesystem that is not considered experimental any more since #Linux 6.16-rc1 (the text was also written by the responsible developers):

""In this blog post, we discuss a new XFS feature [with] the ability to exchange arbitrary file contents atomically. […]

What Problems Does This Solve?

The first problem is that the Linux file I/O interface does not specify that writes to multiple ranges of a file must be persisted in an all or nothing fashion, which means that file contents can be inconsistent after a crash. […]

The second problem is that some external readers of a structured file must never see an update in progress.

A third problem in this space concerns software defined storage. […]""

blogs.oracle.com/linux/post/xf

#XFS Directory Parent Pointers since #Linux 6.16-rc1 are not considered experimental in the #kernel any more[1]. They allow constructing a file path from a file descriptor or a file handle, which should result in better redundancy and reporting.

In case your want to know more about it, checkout the recently published text "#XFS - Directory Parent Pointers in UEK8"[2], which was written by Allison Henderson and Darrick Wong, who developed and upstreamed the feature.

[1] Just like online fsck and the exchange-range syscall: git.kernel.org/linus/ca43b74ac

[2] just ignoring the Oracle/UEK specific bits

Replied in thread
@cadusilva@bolha.one Se o comando for personalizável, deve conseguir diminuir a prioridade geral do processo e também quanto a entrada/saída, por exemplo:

nice -n 15 ionice -c 2 -n 7 comando
Ver manuais de nice e ionice :debian: :podeConfiar:

Porém, a efetividade da priorização de I/O também pode depender dos algoritmos de escalonamento definidos na máquina ou para cada dispositivo de armazenamento. Se quiser conferir várias minúcias, vale a pena estudar esses detalhes. 🤓

#LinuxLibre #Linux #kernel #tuning
manpages.debian.orgnice(1) — coreutils — Debian bookworm — Debian Manpages

Linux 6.16 RC7 released!

Linux 6.16 RC7 is now live for developers and curious users to try out. All the interesting changes from performance improvements to bug fixes have been integrated to this release candidate.

In addition to that, the release announcement has indirectly confirmed (with the statement of “we’re in good shape”) that the final release (July 28th) is the only release remaining for the development of the 6.16.x series before moving on to 6.17 RC1 (August 11th) two weeks later. Those are expected dates, though.

In the release announcement for this version of the kernel, Linus Torvalds said:

So last week started very quiet and that always makes me happy. Then on Thursday I started getting some updates, and I went "ok, so at least we have some networking fixes", but things otherwise still felt like this was going to be a tiny rc7.

And then Friday came along.

And the weekend.

And here we are, with a not inconsiderable rc7.

That said, the last few days I ended up getting a fair amount of pull requests, but pretty much all of them were tiny. A lot of single-fix pulls, and while rc7 isn't the tiny release it looked like mid-week, it's also not really any bigger than usual.

So there are fixes all over, they are all pretty small.

Nothing really stands out - the biggest patches in here are for some documentation and self-tests or tooling, not actual kernel code changes.

So unlike the week before, it all feels very trivial and I think we're in good shape.

Why not try out this awesome pre-release of Linux 6.16?