Linux 6.11 Released

by jrepinc • 2 years ago

299 101 2 years ago

unixhero• 2 years ago

Rejoice!

Edit: These two items are huge! support for writing block drivers in Rust, support for atomic write operations in the block layer,

ufo• 2 years ago

Does anyone know how they'll implement the runtime constants?

vardump• 2 years ago

Long way since Linux 3.11 for Workgroups [0]:

[0] https://www.linuxfoundation.org/blog/blog/linux-kernel-3-11-...

unethical_ban• 2 years ago

Man, time flies. I remember Slashdot's thread announcing it.

Linux versioning now is the worst kind of arbitrary! It's the web-browser "just iterate the number" method, but with the appearance of semantic versioning.

fallingsquirrel• 2 years ago

Linus does that on purpose, because he doesn't want people to put undue importance on any particular release.

https://lwn.net/Articles/781206/

I'd like to point out (yet again) that we don't do feature-based releases, and that "5.0" doesn't mean anything more than that the 4.x numbers started getting big enough that I ran out of fingers and toes.

ensignavenger• 2 years ago

Linux versioning has always been that way.

marcodiego• 2 years ago

Bets for 6.12: sched_ext, PREEMPT_RT anyone?

samtheprogram• 2 years ago

PREEMPT_RT, please!!

cwillu2• 2 years ago

PREEMPT_RT would be a boon for audio/video work, hope it finally goes in. Not everyone who needs it knows they need it, let alone how to compile it, and having it in-tree would massively simplify the process for distros to [re-]package a proper RT kernel.

anthk3• 2 years ago

These kind of schedulers often harm servers' I/O usage. And the opposite it's true too. Server based loads are bad for multimedia/gaming desktops.

aseipp2• 2 years ago

There's already a series posted a week ago for 6.12 that will ungate PREEMPT_RT on x86, ARM and RISC-V; it only modifies the Kconfig entries to enable it[1]. So I'd say it's hardly a bet!

[1] https://lwn.net/ml/all/20240906111841.562402-1-bigeasy@linut...

homebrewer• 2 years ago

6.10 (TEN, the previous one) has been a very problematic release for me, with one desktop running into four major bugs in total: three separate amdgpu bugs resulting in video corruption, hangs and crashes, and now that I'm on 6.10.10 and those seem to be fixed, the system intermittently refuses to come up from sleep mode.

Anyone else having similar experience? This is the first time something like that happened in a decade of using the latest stable kernel release (in my experience, it's actually been stable for all that time except for 6.10).

mahkoh• 2 years ago

I've had a few hard crashes (system freezes completely, ssh does not work) over the last two weeks on 6.10.x kernels. I am hoping that it is https://gitlab.freedesktop.org/drm/amd/-/issues/3142 (and not hardware failure) but I've been unable to capture the kernel panic if it does occur.

Never had such an issue before.

OvbiousError• 2 years ago

Wow thanks for that link. I've had my machine crashing a couple of times as well the last couple of weeks, was absolutely stable before. I hope it's this and not hardware failure.

0xC0ncord• 2 years ago

I too have been having AMD GPU video artifacting lately, but so far that is the only regression I've noticed in 6.10.x. I am still on 6.10.8 so I'm not sure if 6.10.10 will contain a fix for me just yet.

shmerl• 2 years ago

What GPU?

TomK32• 2 years ago

Is that just AMD? On my thinkpad X270 playing videos in Firefox is just a mess. All sorts of problems while Chromium is just fine. It's also fine on a copy of my system that I run on a thinkcentre tiny.

nine_k• 2 years ago

Check the Firefox video settings in about:config maybe?

tadfisher• 2 years ago

I do believe Chromium still has to be patched to support HW decode via vaapi, while recent Firefoxen turn it on out of the box. So it's likely that Chromium is using software rendering, bypassing any driver-related bugs.

PKop3• 2 years ago

Chromium doesn't support hardware decoded video though right? While Firefox does, so you're killing your battery life with extra cpu cycles.

dsissitka• 2 years ago

I got bit twice.

First there was the bug that broke Chromium based apps when using SELinux. https://lore.kernel.org/all/30fc5b38165e4eda57d640eca76b7df1...

Then 6.10.6 didn't want to boot.

Usually I run into issues two or three times a year. I guess this time around they just happened to be a little closer together.

_cenw4• 2 years ago

Early 6.10 somehow broke bluetooth audio for me, only letting me use HSP.

globular-toast• 2 years ago

Yeah, same problems with 6.10 and amdgpu. Radeon Pro WX 3200 fwiw. I've been on 6.8.9 for several weeks now. Just today I booted 6.10.7 and it's been stable so far. I haven't tried to put the system to sleep yet, though.

This isn't the first time I've had problems with the stable kernel, though. A while back I had problems, also graphics related, with Intel i915 (my onboard graphics that I used before I got the AMD card). It took a while but it eventually got fixed. I haven't looked to see if there's a bug tracked for the AMD problem.

IshKebab4• 2 years ago

How does Linux handle testing? They don't seem to have a CI system as far as I know. Presumably there's no big lab with automated testing on real hardware. Does anyone know how releases are tested?

Idesmi2• 2 years ago

As far as I remember, there is some automated testing before release but very limited in scope.

leonheld• 2 years ago

Several companies (including mine, Toradex) have setup board farms to deal with Kernel regressions and bugs during the `-rc` window, ie, the kernel that is going to be released is heavily tested.

https://kernelci.org/ is a big one, Linaro has theirs https://lkft.linaro.org/, Intel has multiple farms, Collabora works relentlessly with GitLab integration https://www.collabora.com/news-and-blog/news-and-events/patc......

The kernel is very much tested before a release with gigantic automated labs on real hardware.

JSDevOps• 2 years ago

Explain how you would deal with CI in a kernel?

davidlt• 2 years ago

I am really surprised with RNDA3 support. I have never seen so many issue with iGPU (APU). It started with VP9 decoder issue (e.g. just playing videos on YouTube was enough to trigger it), but that got fixed after a very long time (required a new firmware). Multiple constant [different] crashes, but you can workaround most of them by adding amdgpu.sg_display=0 to your bootargs. It's already listed in Arch Linux wiki, Gentoo wiki, etc.

Again, I was surprised by the number of firmware and driver issues since RNDA1/2/3 have been around for years now.

Idesmi2• 2 years ago

6.10.5 fixed all my amdgpu issues.

lifeinthevoid• 2 years ago

amdgpu on my 6.10 kernel has been crashing constantly too. It makes me want to go back to Intel. My workaround has been to let the ryzen 7700 iGPU run at its max clockspeed of 2200 Mhz …

gigatexal2• 2 years ago

6.10 broke my fedora gaming proton box and I was on holiday at the time and so upset I just nuked it and put windows on it to play games. Now it's only powered on on the weekends to play games and I've moved all my linux needs to a VM on my overspecc'd Mac.

I also had an AMDGPU system. 5600X, AMD 6800 GPU, Fedora 40. -> now win11 (which has so much cruft out of the box I am considering nuking it and going back haha)

anthk3• 2 years ago

Try Fedora Bazzite. It's inmutable and you'll get rollbacks at the grub prompt.

anotherhue• 2 years ago

Broke my nixos steam/proton setup also. Using game scope as the compositor. Switch to LTS (6.6 I think) solved it. Frustrating bug but they don’t call it nixos-unstable for nothing. 7900 GRE

abhinavk• 2 years ago

In college, I used to run bleeding edge. But now...I will only run a bleeding edge system as an immutable distro or on a filesystem with quick rollbacks.

skerit• 2 years ago

6.10 broke my AMD 5700XT eGPU setup. Had to downgrade to 6.9

After I upgraded to a 7900XT it worked again.

mardifoufs• 2 years ago

The past few releases were more problematic for me yeah. It's super anecdotal since I never used Linux before the v5.xx kernels but in comparison to those yes it's a bit less stable.

wooque• 2 years ago

That's why I stick with Debian.

aseipp2• 2 years ago

amdgpu is shit for me, my friend. Funny story: my headless server with a small Navi 1 workstation card (repurposed) couldn't be SSH'd into last week. Went and plugged in a monitor, rebooted, and the framebuffer got stuck during stage 1 while fsck'ing my disk. OK, I think to myself, it's probably taking a while to fsck after N boots, happens once every few months. Doesn't change for 48 hours. Turns out my machine had just run out its DHCP lease, so it had a new IP and I didn't realize it, which is why I couldn't log in. So I log in, and..

What was wrong? What actually happened was that on bootup, the amdgpu driver would panic and fault during boot exactly when fsck was happening, and the framebuffer would be stuck forever. So it just looked like a filesystem issue but in reality my graphics output was merely fubar; the system itself was otherwise fine tnough.

This is reliable and reproducible for me; it always faults at almost the exact same spot at boot every time, for this kernel version at least. In reality amdgpu has been unreliable for me for 5+ releases at this point on a card less than like 7 years old.

Really considering moving over to a small cheap nvidia card and just running Nouveau instead. At least then I might have a reliable framebuffer.

dogben• 2 years ago

you can just blacklist amdgpu and use EFI/VGA buffer provided by bios

trulyrandom• 2 years ago

Just to add a non-problematic experience report to the mix: I've been using 6.10 for months on two AMD machines with different hardware (one with a 7840U and one with a 5700XT) without any issues whatsoever.

mixmastamyk• 2 years ago

Doing well on a recent AMD, although I didn't upgrade until about 6.10.7 or so.

There were problems on earlier kernels, but they were mostly fixed by a firmware update. Kernel already had support by that time.

sinker• 2 years ago

Kernel version 6.8 broke suspend ($ systemctl suspend) for me. I run two machines with near identical setups. I upgrade my "preview" machine first before updating my primary machine to test for defects.

If I boot from kernel version 6.5, suspend works fine. Hold shift while your machine is booting and the grub menu will allow you to select a different kernel version.

nitinreddy88• 2 years ago

Ack. AMD gpus broke left and right with this version. Lot of issues reported : https://gitlab.freedesktop.org/drm/amd/-/issues

Even the latest Linux kernel 6.11 still have same issues (atleast for me)

daemonologist• 2 years ago

Yeah I've experienced a fair number crashes with 6.10 (multiple point releases) on my 680M. Earlier kernels were fine in my recollection although I didn't bother downgrading, and Windows remains rock solid on the same system.

WaffleIronMaker• 2 years ago

I was using a Fedora release with the 6.10 kernel, and I experienced frequent logouts and restarts, almost once per day. It's nice to see that others were having similarly poor experiences.

quicksilver03• 2 years ago

The amdgpu driver has been the main source of issues for me on 6.10, but I had issues on older 6.x versions as well: for example, on a desktop with 2 monitors, I had to turn on the 2 monitors simultaneously or the UI would freeze.

mnahkies• 2 years ago

No idea if related, but I had a really frustrating issue recently on fedora/similar kernel version.

Resume from sleep stopped working, and boot started taking minutes. Going through logs I was able to figure out it was probably usb related, and that sleep worked if you were patient enough to wait for a timeout, but the issue persisted with all cables unplugged. The error code was something along the lines of failing to provide enough power for a device, I suspected my new speakers despite them having dedicated wall outlet power (usb audio despite being on the motherboard AFAIK)

Long story short someone in a thread somewhere suggested unplugging from the wall for 10 min and to my surprise it worked - I guess it cleared the fault somehow and it hasn't reoccurred (yet). Might be worth a try.

xyst• 2 years ago

I’m just at awe to see Torvalds still publishing the release notes for Linux kernel.

HL33tibCe7• 2 years ago

He's got at least another decade in him

throwaway1194• 2 years ago

“I'm still working on it. It's been 25 years. I can do this for another 25. I'll wear them down.” —Linus Torvalds

ajcp• 2 years ago

Not only release notes, he's still actively contributing! From said release notes:

Linus Torvalds (2):
mm: avoid leaving partial pfn mappings around in error case
Linux 6.11

candiddevmike3• 2 years ago

Wonder what kind of succession is in place. I'd assume there's a legal framework around it to ensure continuity with something so important.

fluoridation• 2 years ago

I mean, it's GPL. If something were to happen to him, anyone could just start their own fork. There would likely be some competition between forks for a while before two or three emerge as the successors eventually.

johnnyApplePRNG3• 2 years ago

Does this mean i can suspend my Linux laptop to ram now?

fsflover• 2 years ago

If you choose the right laptop, yes. Works for me.

johnnyApplePRNG3• 2 years ago

Thinkpad L14 Gen 1, anybody care to take a guess?

shatsky2• 2 years ago

Is suspend-to-ram something that often doesn't work on laptops with Linux, really? I used it on dozens and saw problems maybe once or twice. Though I usually pick ones which are 3+ years old.

quibuss• 2 years ago

Yes, very happy Thinkpad P14s user here on EndeavourOS.

zh3• 2 years ago

Always worked for me on Thinkpads (T41/T61/T420/T520). T420 reports a somewhat optimistic "122 hours remaining" when coming out of sleep with a fully charged battery though.

pxc• 2 years ago

I thought the issue with this nowadays was that hardware support for suspend-to-ram has been increasingly removed from hardware.

At the same time, S3 sleep worked just fine on supported hardware 10 years ago. So what does suspend-to-ram have to do with Linux 6.11 in particular?

PhilipRoman• 2 years ago

Pretty vague question, suspend to ram has always worked for me on multiple random laptops

jeffbee2• 2 years ago

Can't wait for Ubuntu to drop this into the Oracular beta for a couple of days and pretend like that was tested before release.

ruthmarx• 2 years ago

Isn't that a large part of what makes Ubuntu Ubuntu?

rafaelturk• 2 years ago

Isn’t that the very definition of what an upcoming release should do?

jeffbee2• 2 years ago

Not at all. It is a poor pattern because they throw a kernel over the wall right before the freeze, then refuse to fix any of the bugs, because of the freeze.

The kernel should go out to general testing as soon as the cycle starts, not right before it ends.

benakh• 2 years ago

Anyone here that can comment on the new snapdragon X support?

SushiHippie• 2 years ago

This is the latest news I could find about snapdragon support

https://www.phoronix.com/news/Linux-6.11-SoC-Platforms

EDIT: looking at the tree of 6.11, these 2 laptops are still the only "supported"† ones

https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/lin...

† as in bootable, but still don't work fully (e.g. touchpad not working, battery monitoring not working, camera not working, usb/hdmi not working fully)

meiraleal2• 2 years ago

When will be the year of GNU/Linux for smartphones? Android is not it. I wish we could install distros in a smartphone as easy as it is in desktop

martinsnow• 2 years ago

Nobody wants it outside of a tiny hobbyist segment.

candiddevmike3• 2 years ago

Nobody understands it outside of a tiny hobbyist segment, because there is no "definitive" Linux. No two "Linux" look the same.

The year of the Linux desktop will be when there's an OS like Android or macOS that is running Linux but the only way you would tell is by running uname.

squarefoot• 2 years ago

Fighting against well known hardware on PCs is one thing, doing the same against mobile manufacturers because they refuse to release any documentation is a nightmare. Linux will have a hard time becoming a reality in the mobile world because of hardware manufacturers hostility, not for any of its faults. Pine64 had to design the PinePhone platform from scratch for this exact reason, still they encountered so many blocks that when it finally became available it was already too old.

guerrilla• 2 years ago

Maybe FairPhone can do it eventually.

shatsky2• 2 years ago

"As easy as in desktop" not going to happen anytime soon. Phone "bootloader firmwares" lack full description of phone hardware in stardartized form like PC ACPI does, expecting customized OS kernel to know the phone it's running on. Phone devices lack capability to emulate something ancient-but-standard, expecting customized OS to include drivers for all of them. That's enough to make unified OS images impossible.

Linux 6.11 Released

Related Stories