Linux kernel 6.14 is a big leap forward in performance and Windows compatibility
134 comments
·March 26, 2025TheCleric
cube00
> But no. It's just pure incompetence.
While I don't agree with how he conducts himself on the mailing lists; I respect he uses the same language when talking about himself.
jchw
You know, though, his conduct nowadays isn't really that bad; I don't think I'd be bothered by it in a workplace environment. He makes strong assertions, but not mean-spirited, and I think Linus has a lot of reasons to feel highly confident when he makes strong assertions when it comes to Linux.
Of course, there are some moments, but almost everybody has their moments. When push comes to shove, Linus seems to handle the important things pretty well.
kjrfghslkdjfl
[dead]
maccard
> I don’t think I’d be bothered by it in a workplace environment
https://lkml.org/lkml/2025/2/23/351
This was a month ago. I wouldn’t stand for this in a company I worked for.
znpy
The guys is truly a beacon of fairness. Linux as a kernel will probably go downhill after we lose Linus.
And I'm only 50% ironic/joking.
karunamurti
Yeah, GKH doesn't have the insult power of Linus. Can some Finnish people teach GKH on how to insult? I learned the word perkeleen vittupää from LKML.
cedws
That’s usually how people like that are. They hold others to high standards because they hold themselves to high standards.
throwaway48476
He named both git and Linux after himself.
ASalazarMX
Does he willingly take that treatment from others? I don't really know, but that would be the actual test of congruence.
SEGyges
I am not deep into the Linus weeds but my impression is that he doesn't especially care if he's on the receiving end of this. It only started to feel different from "well, the Linux list is the PvP zone" when Linus was sufficiently weighty/famous that you almost had to take an insult from him to heart, and he did eventually correct his behavior there.
foresto
I don't follow the lkml enough to highlight a pattern, but this recent exchange comes to mind:
On Thu, Jul 4, 2024 at 8:52 PM Linus Torvalds <torvalds@linux-foundation.org> wrote:
> Jason. This smells. It's BS.
On Thu, 4 Jul 2024 at 11:57, Jason A. Donenfeld <Jason@zx2c4.com> wrote:
> It's not BS. And that's not a real argument from you, but rather is something else.
On Thu, 4 Jul 2024 at 12:19, Linus Torvalds <torvalds@linux-foundation.org> wrote:
> Bah. I guess I'll have to walk through the patch series once again.
> I'm still not thrilled about it. But I'll give it another go.
> Linus
KennyBlanken
One does not get a pass on being rude and hurtful to others by speaking about themselves negatively.
The fundamental thing that people never seem to understand in the "Linus is being an asshole" vs "I love Linus, he tells it like he is and to manage a project you have to be willing to do that!" debate is that it is completely unnecessary to speak at all about someone's person on a mailing list like lkml.
"This code sucks!" - no. "This code does not meet the linux kernel's coding conventions, has race conditions, and this algorithm could likely be implemented in a far more performant way."
"You're just WRONG" - no. "The position that you're being forced to learn a new programming language is not correct. You have not been asked to write or even understand Rust."
I think the people who think 'you have to be that way to manage/lead' have never been under a good manager or leader and I daresay that a lot of people in technical supervisory and management positions unfortunately grew up respecting Linus and thinking his behavior was how you "get things done", not realizing that Linus's behavior almost certainly kept a lot of people from risking contributing and if anything has hurt the Linux kernel project - and kept people from feeling they could safely point out problems in someone's code because
T'so is a great example; a huge ego. As a result, people simply have gone elsewhere to work on filesystems. The ext filesystem(s) were not terribly reliable or tolerant of real-world situations, is still not very performant, and has evolved at a glacial pace consistently lagging behind its peers.
asah
As I cross 40 years in this industry, I find that my procrastination often has subtle root causes. I assume others have the same experience.
Currently, I've been putting off a long report for a few days and it's because I'm trying to balance a number of conclusions, tones, etc. When I'm ready, I'm ready.
progmetaldev
For myself, I think it's just as much how to word things properly to most effectively get your point across, as it is to be ready if your report has some kind of backlash or isn't seen the way you meant it. You need to be mentally prepared to handle the response you get back, even if it's no response at all after you spent many hours putting a report together and worrying about how to make it just right. Obviously this will change depending on how important what you're reporting on is, and whether your work contributed to that report's findings.
bombcar
I find that can be true - but I also sometimes just forget to hit submit.
HPsquared
The article helpfully informs us this note was written by the man himself, Linux Torvalds.
npodbielski
Nice did not notice that. Good catch.
lenerdenator
He really has grown as a person. Ten years ago, there would have been no mention of it unless someone called it out, and the response to that would have boiled down to "f#ck off, it's free".
... Which also would have been relatable, and valid if we're being honest.
dagw
"f#ck off, it's free" is possibly an acceptable response if you're also working for free. If maintaining that free software is your job that you are getting paid good money to do, you really should hold yourself to higher standards.
echoangle
Hm, depends on who is paying you, I think. If some philanthropist is paying you to make good free software, yes. But if you’re paid by stakeholders that want your maintenance to make sure they can use the free software themselves, you’re only really accountable to their opinions.
caycep
I assume he takes advice/feedback i.e. the famous ESR post and the concerns from that contributor from intel and others to heart. Or at least has forced himself to learn how to do so.
mouse_
From a YouTube comment:
> please don't hype NTSYNC. yes it has better compatibility than ESYNC and FSYNC, yes it's marginally faster in selected title. The phoronix article reports benchmarks with WINESYNC vs NTSYNC, the gains are there only if you were not already running FSYNC (on by default in most titles running under proton).
> By overhyping features they ultimately end up underdelivering because people expect insane gains that were never there to begin with.
Surely we can do better than YouTube comments?
Adverblessly
That is a question of source. If you read the article on GoL instead ( https://www.gamingonlinux.com/2025/03/linux-kernel-6-14-out-... ) you would have already read that:
> For Valve's Proton however, don't go expecting much here with ntsync. As Valve developer Pierre-Loup Griffais replied to a user on Bluesky to note:
> We already include fsync, which should be as fast or faster as ntsync. We developed ntsync as a general solution that'd be acceptable in upstream Wine, but there's no urgency in including it in the Deck / SteamOS kernel.
tremon
That's odd. Why would wine not accept improvements using fsync from Proton?
nialv7
One exciting aspect of NTSYNC is that it could get into wine upstream. FSYNC is only available in Proton and not in upstream wine, mainly because it is not fully conformant, i.e. it doesn't behave exactly like its Windows counterpart does. But NTSYNC doesn't have this problem.
robertlagrant
YouTube comments have lots of smart people in them, sometimes.
ThatMedicIsASpy
That can be simplified by the topic and presentation. Is the content trying to educate, be entertaining or both. Is the topic extremely niche then you will find like minded people with suggestions and discussions.
shmerl
Hype is not about better performance but about upstream support. It's valuable.
mhurron
If you've been anywhere around linux gaming circles for the past 6-8ish months, you've seen nothing but people not understanding the benchmarks hype NTSYNC. Everyone trying to set appropriate expectations is pretty quickly silenced in whatever manner the platform allows (downvote, etc)
shmerl
Must be some wrong circles. From what I've seen, it's mostly well known that ntsync is comparable to existing solutions like esync / fsync.
andy_xor_andrew
I'm super excited for this! But it makes me curious.
What's the process like for getting such a low-level primitive added to Linux? Especially a low-level primitive that 1) exist basically just to emulate behavior of a completely different kernel, and 2) is only needed for a subset of users to play games?
I'm not complaining, just curious. I would assume a patch to add the above would be heavily scrutinized. Is it because the popularity of the Steam Deck / proton?
nialv7
Technically this is not a low-level primitive. NTSYNC isn't added as a, say, system call like futex2[0] was. It's instead added as a character devivce - which you can build your kernel without, and functions via ioctls.
dayjah
I don’t know if this answers your question directly: but the article points out that this is a module, not a primitive that’s in the kernel itself.
The idea here is that wine/proton can interact with this module directly if it’s loaded, or continue to use their wrappers around existing fsync if not.
I expect that means a little less scrutiny since if it segs the user can choose to not load it and achieve a stable OS.
no-js-user
You've got a nice arc river in "of / getting / to".
Easier to find if you look slightly away from that part of the text.
buyucu
Sometime around 2020-2021 I stopped checking if a video game will work in Linux. These days almost everything does with minimal friction. DXVK and Vulkan have been gamechangers.
silverliver
Same here. Anything I try seems to just work. The only problem is anticheat but even those are starting to behave even when isolated and confined in containers.
buyucu
yes, the only games that don't work in Linux are the ones where Linux is explicitly banned by the developer.
cadamsdotcom
If you’re wondering about hard numbers, the benefits for Wine look massive.
From Jan 2024:
NekkoDroid
Do note that these numbers are Upstream vanilla Wine vs Wine+NTSYNC and not Wine+fsync, which is more common and has comparable performance (if not sometimes better[0] according to a valve employee)
[0]: https://bsky.app/profile/plagman.bsky.social/post/3lkp26xmco...
nodesocket
I also believe this kernel adds Intel GPU support for the N100 and N150 chips used in lots of mini PCs. I have a Beelink N150 rig running various Docker containers that I am curious to test some light AI workloads on.
jaimehrubiks
What would be the meaning of that? I have a Beelink with N100 for more than a year and HW acceleration works already (I use it in Jellyfin), so I wonder what do you mean that "intel gpu support" would mean.
nodesocket
What OS are you running? I’m running Debian 12 Stable and can’t get nvtop to detect it. Searching I found this Reddit thread:
https://www.reddit.com/r/MiniPCs/comments/1htxoi6/beelink_mi...
Did you add backports or recompile? I think N100 may be supported by stable but not N150.
jaimehrubiks
But they are talking about beelink s13 with the intel N150.
N100 has been there for a while longer and I've seen guides in the internet for a long time with no specific requirements.
I'm running ubuntu 24.04 with nothing extra, runs out-of-the-box (assuming that if it's a VM, you properly do the card's PCI passthrough).
I tested nvtop and it works, it shows GPU% usage when I start a movie in jellyfin and change the quality to force trascoding.
diggan
> Below is the shortlog for the last week. It's nice and small - not only was there no last-minute issue yesterday, the whole last week was pretty calm. The patch is dominated by some amd gpu updates, and even those are pretty small. The rest is random small changes all over.
https://lore.kernel.org/lkml/CAHk-=wg7TO09Si5tTPyhdrLLvyYtVm...
Funny that Linus himself doesn't consider much in the release noteworthy besides the GPU stuff, yet the submission article goes on about "leaps", "big news" and "major step".
wtallis
What you're quoting Linus on is his commentary about the patches landing in the past week, which is supposed to be only bug fixes as the code stabilizes enough for a stable release to be made. New features would lane weeks earlier during the merge window. What's newsworthy here is not anything that was merged recently, just the fact that this new functionality is now available from a stable release.
If there was anything really significant from the past week's changelog, it would likely have been reason to delay the stable release for at least another week.
ptdorf
I wonder if he reads every single line of code. Or even skims over them.
rurban
Not a single word about missing bcachefs fixes? Or are they now in?
Or the upcoming breaking GPIO changes. That will be a huge minefield
kwar13
zdnet... now that's a familiar name I haven't seen in a while
DominoTree
"Linux Torvalds"
genpfault
"...a Soviet computer hacker named Linyos Torovoltos, ..."[1]
[1]: https://gwern.net/doc/cs/security/2001-12-02-treginaldgibbon...
dietr1ch
The genious and madman behind the Linus Operating System.
wyldfire
Gell-Mann amnesia [1] is an interesting thing to think about here. But IMO the real problem is that editing for most journalism websites is done very very poorly, if at all.
Jotalea
tl;dr: better windows-like filesystem driver for Wine, better AMD GPU drivers, more Rust implementations, support for Snapdragon 8 Elite, fixed vulnerabilities on RISC-V, better btrfs