The Universe Doesn't Want Linux 6.8 To Exist
Vložit
- čas přidán 16. 01. 2024
- Every Linux release I'm sure has had it's struggles but it seems like for this kernel release it's a bit worse than normal, so much so that it seems like the universe is trying to conspire against it.
==========Support The Channel==========
► Patreon: brodierobertson.xyz/patreon
► Paypal: brodierobertson.xyz/paypal
► Liberapay: brodierobertson.xyz/liberapay
► Amazon USA: brodierobertson.xyz/amazonusa
==========Resources==========
Linus Slowdown: lore.kernel.org/lkml/CAHk-=wg...
Linus Bisect: lore.kernel.org/lkml/CAHk-=wi...
CPPC: docs.kernel.org/admin-guide/a...
Phoronix Article: www.phoronix.com/news/Torvald...
Slowdown FIx: lore.kernel.org/linux-pm/2024...
Dave Airlie Patchset: lore.kernel.org/dri-devel/CAP...
Linus Response: lore.kernel.org/dri-devel/CAH...
Dave Response: lore.kernel.org/dri-devel/CAP...
DRM 2017: lkml.iu.edu/hypermail/linux/k...
DRM 2011: lkml.org/lkml/2011/12/6/347
Linus Weather: lore.kernel.org/lkml/CAHk-=wj...
=========Video Platforms==========
🎥 Odysee: brodierobertson.xyz/odysee
🎥 Podcast: techovertea.xyz/youtube
🎮 Gaming: brodierobertson.xyz/gaming
==========Social Media==========
🎤 Discord: brodierobertson.xyz/discord
🎤 Matrix Space: brodierobertson.xyz/matrix
🐦 Twitter: brodierobertson.xyz/twitter
🌐 Mastodon: brodierobertson.xyz/mastodon
🖥️ GitHub: brodierobertson.xyz/github
==========Credits==========
🎨 Channel Art:
Profile Picture:
/ supercozman_draws
#Linux #OpenSource #FOSS #LinuxDesktop
🎵 Ending music
Track: Debris & Jonth - Game Time [NCS Release]
Music provided by NoCopyrightSounds.
Watch: • Debris & Jonth - Game ...
Free Download / Stream: ncs.io/GameTime
DISCLOSURE: Wherever possible I use referral links, which means if you click one of the links in this video or description and make a purchase I may receive a small commission or other compensation. - Věda a technologie
I think a lot of you might have failed maths class, a 100% time increase can also be framed as a 50% performance decrease
In case anyone is still confused. If you have 5 apples and I give you 100% more, you now have 10 apples, if I now decide to take away 50% of your apples you will now be back to 5
The video: about problematic kernel merges
The comments: YoU MaDe mAtH PrObLeMs
Still, thanks for the vids, Brodie. ❤️
We all run the latest git compiler, kernel and glibc (so we get the coolest new bugs!), sometimes / 2 and * 2 becomes the same answer.
(Gcc 2.7 had multiplication bugs)
I had to laugh at this video. It seems performance issues is the cost for laziness (high level language). LOL!!!
@@nomadhgnis9425 wdym? linux is written entirely in c
Completely unrelated question: how comfortable are you switching between black and white background page?
There is only one thing to do. The universe is giving us a sign. Kernel devs need to skip the number and call the next release 6.9. Codename: Nice
wow.. yes... it's defently the curse of 6.8, were we get bullied by the universe to make it 6.9....
6.8 had to be a mess, just to make it clear to everyone that 6.9 will be nice.
le reddit funny !!!
not funny. cram it, chomo
I'm actually sad, that we pass 6.6.6 minor release so fast, and 6.6.9 wasn't any different. Let's hope we see 6.9 for some time in our lives.
I need a 1 hour compilation of Brody reading Linus' email replies to shitty commits.
Uncensored.
yes @@michaelwright2986
That would be an ongoing series of videos, and at two 3 a week the series could last until long after Linus is dead!
I would be up for a monthly update
You might be able to start with content from existing videos. Please inform me if anybody did that.
I'm on Debian stable, not expecting to see a 6.8 kernel in my lifetime 😉
are you off 2.6 yet??? XD
Maybe in like 4 years :)
@@WTFShelley lol. Bookworm is on 6.1 by default!
Though I added a mainline kernel repo and I'm on 6.6.8.
Of course, under no circumstances will I remove the kernel Debian is officially feeding me. I'll occasionally take some risks of breaking things, but I'm not reckless about it.
Same haha
Have you never tried installing custom kernels?
Working with Linus is not for the weak
Yeah. He can be harsh and direct. But in hes position he sometimes has to
@@Kylian381 Better be upfront than be vague.
He used to be worse.
@@Kylian381 remove 'sometimes'
Definitely. In the past there were instances where he's told devs to off themselves over a horribly structured commit.
If this is the case with 6.8 we can only pray that the same doesn't happen to 6.9. The nice must succeed
Greetings from Finland! Yesterday it was under -20°C in here and today we've had 20cm new snow on top of the existing 30cm. I love it!
I head about -40 degree temperatures in Finland a few weeks ago. Finland is proof that there is a reverse correlation between good music and good weather.
Nah they just want to get to 6.9 faster ͡° ͜ʖ ͡°
I see what you did there😏😏😏📸
nice
this is one of those times where a monolithic kernel definitely makes life miserable for linux devs. Having to bisect through thousands of commits while doing nearly hourlong builds each time must be beyond cursed.
hour long? couple of minutes. And bisecting makes finding it fast. You might have to do half a douzend builds, less if you have a good idea which area is most likely the culprit.
With "modern" (make is from the end of the 80s iirc) build systems, you only recompile what you change and stuff that depends on it
"make is from the end of the 80s iirc)" -- Per Wikipedia, Make "was created by Stuart Feldman in April 1976 at Bell Labs", so you're more than a decade too late. :-) Perhaps GNU Make was created around then?
@@EvanED Yeah, was thinking of GNU make
It's crazy to think that the merge window is paused because it's a little bit windy.. Oh btw I am Finnish and yes it was -30 week ago
I'm sure Finland has actual infrastructure to deal with the cold
@@BrodieRobertson That might be a little bit unfair to US infrastructure. A lot of it is crap, but you don't build infrastructure suitable for sub-Arctic conditions in the mild and damp Great North Wet--that would be wasteful. As weather gets more extreme, I guess a lot of places whose infrastructure was previously adequate will find that's not the case any longer.
@@BrodieRobertson neither does australia, but atleast people dont tend to live at just an average low of 1.7c, Liawenee is the coldest permanently-inhabited place in Australia.
In January 2020, Liawenee’s population doubled to two
@@BrodieRobertson Sure, but from what I can gather Linus is pretty damn rich. It's his choice to rely on local infrastructure.
@@petermichaelgreenhe should just build his own powerplant, I agree
It's purely dependent on region. Large parts of the US get below freezing temperatures less then once a decade. Other parts deal with snow and ice every year. The places that deal with it regularly are prepared to deal with it. The ones that don't commonly, are not as prepared.
It's like complaining that Sydney had a freak storm that dropped a 25cm of snow and their infrastructure sucks because they couldn't handle it. No, obviously they they couldn't handle it because it's not something they ever get.
Exactly. The Texas Freeze was a result of climate change. That shouldn't have been possible for Texas to experience. So the infrastructure would have literally no reason to be prepared for it.
There's also the problem that unlike every other state in the country that can just borrow power from other states, Texas can't because it had to be a special self-sufficient snowflake because conservatives reasons.
Exactly this. I live in upstate NY where windstorms and heavy snowfall are planned into infrastructure. Texas (especially) has some self-inflicted issues (power) that shouldn't be generalized to the entire union. Other states may not be well prepared for winter weather, but that's certainly not the norm.
Look it snows in Sydney less often than the USA acknowledges the rights of other nations to not be capitalist. That said burying them under a foot of snow is fairly appealing, just not when I have to be at a tournament in Windsor again, last time I was there I was checking the damn weather reports for if I'd need to call off the drive down, could see snow on almost every slope on the drive down from Tamworth.
@@cericat You're missing the point. "The USA" is not worth talking about, because it's too big and covers to many biomes. It's like saying "What's the climate in Europe?" It's a silly question, because it's too big and covers too many biomes.
@@ahettinger525”Whats the Climate in Europe?”
“Europe is a continent, it has many climates.”
“Okay, but is the climate in Africa?”
Yeah, this is a pet peeve of mine. If anything, it is dumb in a way that most people think the world just has like, maybe 5 climates and are clearly not knowledgeable about Koppen/Trewatha etc.
Also, live in New Mexico. Usually it can snow here, but it didn’t this winter. I don’t wanna jinx it because it is February as I write this, but we might make up for it with a wetter then average monsoon season. Or maybe not even that.
Nothing makes me happier than hearing Linus roast people over hellfire for wasting his time
P.s. totally agree with him not wanting to drive on icy roads, as someone that grew up in Michigan.
Yeah it's funny. But knowing that the kernel is free makes me feel funny about him roasting them. I know a lot of these devs do get paid with donations and other sources of payment but still.
I completely disagree! There is no excuse for such behaviour. I mean Linus is a great and talented guy and his work is amazing, but attacking people personally like this unprofessional and really does not any good.
Building amazing and stable software together is absolutely possible without this whole toxic climate. I am very certain I can rewrite all those emails in a way that they are at least as communicative, while removing all toxicity from those, reducing the number of text lines and preserving all useful information in those.
People can make very bad mistakes sometimes and it can be frustrating when those things happen, but I refuse to defend this type of behaviour like Linus shows here.
Sadly there are more developers in this world which behave that way, but it really does not any good. We can all just learn from each other as humans and being respectful should really number 1 priority.
@@jongeduard I believe Linus dealt with "anger management" at one point, so some of the older mailing list messages might have been (were probably) from before he toned things down a bit. I think Linus can be brash and harsh, but if you look past that, he's usually spot on in his urgency on protecting the kernel. I agree he doesn't *need* to be as toxic as he used to be, but I also "get it". I don't think kernel development is really for "everyone" and Linux is an opportunity to do kernel development on a "real world" OS used all over.
Still, it's frustrating to have to deal with "dumb" mistakes, especially if we're talking about something as "intense" as kernel development. I don't consider "newbie" mistakes to be "dumb". Things like including a C file in a header file are the kinds of "mistakes" that would frustrate any experienced developer.
@jongeduard hard disagree, you can keep your enterprise "professionalism". I'd rather not deal with, nor have the software I depend on be effected by, the disingenuous nature of that environment to a further extend than it already has.
@@jongeduard He's not insulting the devs. He's shitting on their work, or lack thereof. If your job is to propagate the good stuff and weed out/fix the blatant errors and you let THAT through, you've failed miserably and need to do better.
We need a reality TV show about this. Would be amazing!
Starring Guenther Steiner
Would the kernel developers need to be on the same island or something? :) lol
It is all a conspiracy to force Brodie to switch to FreeBSD.
Is the FreeBSD developers world at least less toxic? Where they understand that working together in a positive way and learning from each others mistakes so much better than all kinds of personal attacks?
OpenBSD > FreeBSD
A single line of code. The best case scenario, if you think about it.
Yeah but also the most annoying, at least when it's a complex solution it's understandable how it was missed
This entire video is a perfect example of what I am talking about when I go on at people about how just because your project/code/whatever is open and worked on by a community of people instead of a business or otherwise, doesn't mean it is automatically better, more secure; or even being worked on by competent people.
As Linus said himself. I expect *some* quality control. But as Linus gets to experience on a pretty common basis it seems, and as I suspected to be the case based on past instances I've chanced across in the past; people have taken the 'anyone can contribute' aspect way too far.
Sure, anyone can contribute; but not everyone should. For instance, I don't code. I can bodge a bit with some reverse alterations based on what I see and barely understand; but I don't code. Useful for mods, not for OS's.
So I don't contribute to the kernel or anything like that. Instead, I try to use some of the new stuff when it comes down the tree. I use the newest kernels when I have hardware that would be useful to try it on, and I let the bug reports go through instead of blocking them. Or send in an email (rarely) explaining what I am running into, that sort of thing.
More often than not, I find that someone else already has had the same problem I have. So I end up just using their fix instead of sending anything in. If it works, don't bother them I figure. They have other more important stuff to break in the meantime.
This makes me remember a panel with kernel devs which happened some time ago. By the way they spoke, they created this situation where they need new developers but put them in an ambient of distrust to the point nobody wants to do kernel development.
No, not *automatic*. However, there are ample examples of closed software projects failing to continue to invest in engineering to address security issues as they are identified and the opposite with open source.
Actually it's even worse. It's a 100% performance hit.
Worse, the compilation time is 200% of what it was!
@@aquilafasciata5781 this depends on how you're wording it
if you're talking about an increase, you're only adding 100% of the previous compilation time
if you're talking about the total length, it's 200% as much as the previous compilation time
@@confusioned2249 No! he was right. This is a 50% performance hit ... computed and verified using kernel 6.8 🙂
@@cynodont7391 50% of 22 is 11, 22 + 11 = 33
I have no idea where you got the 50% from
@@confusioned2249 It can be a 50% performance hit as in 50% of your Hardware's theoretical 100% performance are gone. Thus stuff takes twice as long. As far as I am aware relative percentages are a complete mess, depending on language, topic, what the speaker wants to put emphasis on / communicate, etc.
I'm in Portland, Oregon, right now. The side streets and sidewalks are absolute skating rinks from the freezing rain, and even the main arterials are very slippery. And yes, drivers here are absolutely frightening in the Winter. (For context, I have lived in the Midwest and Alaska).
Yeah, that's some serious driver issues
linus lives at 01316 SW Corbett Hill Cir Portland, OR 97219 go say hi to him
I also live in the Portland area (originally from Canada, so winter weather / driving is no big deal for me, but it's scary to be out on the roads here with locals that don't know how to drive at the best of times, never mind when there's a layer of ice). I was away during this year's snowpocalypse, but from what I heard, it was roughly the same as a few years ago.
I actually recently changed from the CPU frequency scaler "acpi_cpufreq" to "amd_pstate" (and enabling CPPC in BIOS as required) using a kernel-param. I did this for purpose of downclocking my Zen 2 processor using the cpupower tool.
"acpi_cpufreq" only gives you 3 cpu-clocks. "amd_pstate" gives you granular cpu-clock control.
To be fair, it has been pretty nasty. High wind with 50 mph gusts and freezing rain was the problem with knocking out power. And after the initial snow day, it's been days of freezing rain, turning the sidewalks and streets into sheets of ice. And of course, Portland doesn't have the most robust or practiced snow response, so the streets have been sheets of ice for some days now. I nearly binned it half a dozen times just shuffling down to the corner store at the end of the block and back.
This reminds me of when I run my Minecraft with the 11 mods, I watch the logs scroll down notice a bunch of errors, ignore them and the games seems to run perfectly.
Think of the bugs that are known, and imagine all the bugs in the software that we use that just go unnoticed.
It's more warnings/bad practices than actual bugs if you don't have issues with it.
Norway has been having an especially cold weather this year too.
Earlier this week it reached -24 C in my local area, and it's predicted to reach around -20 C later in the week too.
But as we say here "det finnes ikke dårlig vær, bare dårlige klær". Which translates to "there is no such thing as bad weather, only bad clothes".
I'm surprised by this. Being in Romania, I thought that in all the scandinavian area -20C is like a chilly autumn. Or normal everyday occurrence in winter.
@@Winnetou17You are correct. -20C is kind of normal. A few weeks ago we had between -30C and -34C for a week and that is colder than normal but nothing extreme. This is in the middle of Sweden. 16:01
Linux had a rocky transition to 6.8
6:29 I don't really understand the code shown that well, but it looks like a case of the coder being a bit lazy on that part of the code and that part of the code should have been written better. Applying a 25% margin sounds very arbitrary to me, wouldn't it make more sense to figure out the correct value instead of just doing a 25% increase? Correct me if I'm wrong.
I'm in Sweden, not far from Öland. And here the temperature yesterday in the morning was -13.5°C…
Linus complaining about driver quality is so relatable. It's like tradfic collectively forgets how to function
i live in Vancover WA about 15m away form Portland and yea the ice storm shut down an lot of places and we lost power allready and came back so its wild right now ^^.
Here in Colorado of the USA, it got down to -20 F. (that's like -25-ish in Celsius).
That may sound pretty cold, but here in Colorado that's just a normal day in January.
I get the feeling that you're always leaving out important words when reading those mails.
Dude, I live in Austin TX, and you could not have said it better! Yes! No one in Texas was prepared! People actually started dying the streets, and there was a lot of damage from trees falling on everything! The snow storm was insane! We call it the snowpocalypse here.
It's funny to read that. I'm from RS, Brazil. We're having massive thunderstorms here. No water and electricity for days, followed by days of blackouts and uncertainty. Trees being ripped from the ground by the wind force and colliding with the electricity cables and generators.
This happens every year.
There' no snowstorms where I live, it doesn't snow, but I'm sure if we had them, it'd be something like that too.
@@3333218 Where are you?
@@3333218 Wow, Brazil intense!
Here first snow of the year means a lot of work pulling cars out of ditches.
I live in the area too, and I had to work outside in that weather. Wednesday morning was especially treacherous because of the freezing rain the night before.
I saw people driving on ice. Jeep doing 360 while going 60km/h through intersection is a scary thing (likely nobody was hurt that time round)
My new goal in life is find way to write out 07:44 as a reply to a bug report without getting shouted at.
It's on 6.8 already? Oh man, it feels like yesterday when we were on 4.4...
Oh wait, I'm on an Android 9 phone. It really was yesterday
I would be interested into a video with the topic: I updated my Linux machine and now I have 50% of the computing power I had before the update. How and where would I report this and would there be an easy way of a rollback?
Okay bruh, temperature hit 18F in Portland, we had no power for like 3 days, although surprisingly the roads have been fine this year. Usually it’s bad
It does get cold, ice like roads in the mountains near Mt.hotham, probably the other mountain region to the north of vic up in the blueys.
I can tell you that the standards for power delivery in the US is changing because of the issues with weather, the problem with the standards is that they are voluntary in compliance and each state can dictate which parts are being enforced. But that doesn’t stop the individual electric companies from building above those standards.
I know about this because I worked in the industry and my father was vice president of engineering at an electric cooperative, where he enacted standards higher than required and that resulted in a huge reduction in outages.
I hope we live long enough to see a Linux distro that uses a MicroKernel architecture
Here in Norway we currently have one of the coldest and snowiest winters in ages, but things mostly work because snow and cold is something we are used to and have infrastructure for. Of course whenever we have record breaking snowfall in a very short time, it takes a while to clear it all so yesterday with more than 20 cm snow (or something) was a bit chaotic and public transport was completely stopped for long periods and car traffic is still very slow with lots of queues. The average speed on the motorway where it's 80 or 90 km/h limit, is now around 20 to 40 km/h
But I still feel quite safe driving since it's the kind of winter conditions I used to from 1 hour north of Oslo where I grew up, and most drivers here know how to drive. And everyone have winter tires (except maybe foreign drivers, but they probably have already slid off the road somewhere down in Sweden long before they got here 😅, and those that somehow got here on summer tires before the snow last fall are stuck wherever they parked)
For the few times a reboot is needed, nothing will break if you put it off until you are done using your computer to which you will boot into your Linux OS all changes applied. I do it that way all the time, unless I just updated something with new features I am eager to explore, like I most certainly will with this kernel update, and the coming KDE Plasma 6.0!
I can drive like a pro on Ice, and it still no guaranty when so many others can't, because sooner or later one of them will take out your car, and possibly you with it!
There is however stuff that stops working until the reboot, including trivial niceties such as USB ports. I was once trying to figure out what the hell was wrong with a USB drive that wouldn't work on a Mac nor on my Linux laptop but would work just fine on a Windows machine. After a good thirty minutes, I remembered that my laptop recently had a kernel update and I didn't reboot since.
Naturally, the drive worked first try after the reboot (still no idea what was wrong with the Mac though)
@@onceuponaban I never had that problem and have all kinds of stuff on USB and other ports, and if it would change them with a USB driver update, it would do it to other hardware too, and there's also no reason for it. Linux doesn't apply any hardware changes until after boot, because hardware drivers are loaded on boot. USB drives are mounted when you insert them, but that's a software loading it because the power draw is detected, the USB bus is already loaded.
It sounds like it may be the drive has worn or dirty contacts. Some wear because they are made with cheep metal that loses it's springiness. I had one that did that, and I had to push up on it a bit and gently release the pressure, and then it would work, but a slight bump and it would fail again, of course it got worse, so I just used another one, and trashed the bad one. Ports can have the same problem too.
@@Bob-of-Zoid Nowadays, hardware drivers are loading when a device that needs them is detected. So it's very possible to get into a state where hardware that has been there since before the upgrade keeps working, but new hardware does not.
@@petermichaelgreen Well, I don't have a USB buss that isn't in use, and I have them all, 2.0 up to C, so maybe that's why, but this would be the first time I hear that, and I quit Windows completely 14+ years ago and switched to Linux.
BTW: I'm on Arch, and KDE Plasma.🥳.
I am currently thinking of learning how to program and things like these really make me reconsider.
Portland is the same as Seattle just less hills. When it snows in seattle even just a few inches, the entire city shutdown. Due to most of downtown Seattle being straight up hill going west to east downtown effectively shuts completely down. We dont own salt trucks
As a guy from Finnland I assume Linus has some things to say about the "Winter" in Oregon :D
Hah, coastal boy, eh? In Australia, I lived in Armidale (NSW), and in the nation's capital; and in both places I saw -10 C. It was, admittedly, in the middle of the night in the dead of winter, but it was cold. Also dry, so if you came into contact with any synthetic fabric on a winter morning, there were sparks when you put your hand (or key) near metal. Lovely days, but, +15 and better, but cold as when the sun set.
We get some ice where I live sometimes in vic, I can confirm I have no idea how to drive on it and I wouldn't want to share a road with anyone when I have to. (Mostly I just drive really slow but hills especially on mountain roads are a bad time)
You know, even in Finland we had a problem with a snowstorm this week. The public transport was mostly affected. But no power shortage😅 This winter is much colder than in last 5 years.
It's not power "shortages" it's power "outages". Shortage means "we can't generate enough power, so we're shedding load to keep up and turning some stuff off (aka rolling blackouts)", outage means "a tree fell on the wire and snapped it - someone has to come and cut the tree down and put the wire back up". The shortage is the number of crews that can go around and fix things (and a lack of preventative maintenance, but that's a whole other topic for a whole different video).
@@gorak9000 I'm sure you also speak Finnish perfectly.
He's right about the backtrace, the tree of those includes goes from most recent to oldest, making intel_display_power.c the first file to include, not the last to be included.
Also the fix has nothing to do with the ordering of includes, that's just a code hygiene issue affecting the same file that needed to be fixed - but regardless, Linus was wrong but I can iimagine his frustration too.
What? When this video was made the temperature here nearby Helsinki Finland was -27C for many days and below -20 a lot longer.
5:55 was that sarcasm? because surely it would've indicated worse QA problem if it turned out to be more structural issue in the code that should've perhaps been picked up on earlier in the process instead of it being a random corner case fixable by small change
That finish -16°c would be quite nice for a winter
I live in the lower north of Sweden(it's in Norrland (north land) but i live in the middle of the country, verticaly) and we got about -18°c although that's a little bit too cold for my taste
I say we skip straight to kernel 6.9. Surely, it will be nice 👍
lmao
Been without power since Saturday. Many cars and garages crushed by giant pine trees in my neighborhood. We just got ¾" of pure ice coating everything last night to sweeten the deal. All depends what part of Portland you live in. I live up at the top of Forest Park which is arguably the worst except for the Columbia River Gorge east of town. It's nasty out there, got to look out for bad Portland drivers who are mostly actually SoCal California drivers 😂
27c ~ 80f. f = (c*1.8)+32 but for quicker head math f ~ (c*2)+30 gets you in the right order of magnitude
Good, i am still on 6.5 and 6.6 tkg-bmq, waiting for other scheduler in the 6.7 tkg kernel and since i rarely reboot and need to upgrade hardware in 2 weeks... it won't be happening soon for me.
I lived out near where Linus is in the Oswego area, there is NO way he's driving down to a Starbucks lol. Those roads are a sliding sloshy mess right now. We had to take someone down to PDX today and it was all we could do to get there on time. If this weather would make up its mind, it might not be so bad. Wouldn't say we have the worst infrastructure, but when they gravel the roads thinking there'll be frozen conditions and instead it warms up, everything turns into a gravel and slush slip n slide.
We need to go back to the old kernel version scheme, with the odd minor version meaning there may be some instability, but even minor number guaranteeing stable operation (i.e. just fixing the breakages in the odd version without introducing new features that can introduce new breakages). It worked so well in the Linux 2.x days, it could work again.
THANK YOU for saying "UEFI" after mistakenly saying "BIOS". That's a pet peeve of mine.
Same here. Still lots of "changing something in the BIOS" out there, where it should be "EFI setup" or similar.
The term is often used interchangeably even by people who should know better
11:16 euh.. no, I believe kexec does reboots faster by having the running kernel, load a new kernel in member and doing a shutdown then calling the new kernel, so you bypass the BIOS, especially on large servers saving many minutes doing a reboot.
In 2018, I went to Norfolk VA to see Nightwish (Finnish band, for those who don't know them) and a snowstorm basically followed us for the whole trip (about 100 miles east from where I live). People were standing in line with umbrellas to keep the snow off, and I thought to myself, "They're probably looking at the crowd outside and saying 'What a bunch of wimps. This is a light shower in Finland.'" It's all what you're used to.
And yes, the US power infrastructure does suck.
5:49 I like to believe linus is a huge gamer
I live just north of DC. It used to be every time a squirrel farted I'd lose power. It's gotten a lot better since the local government started threatening the power company to appropriately trim trees. There are power lines that have little logs on them because they hadn't trimmed the trees in so long that they grew around the lines. When they trimmed them they just cut the branches either side of the lines and left the chunk of branch around it.
The simulation runs on Linux, and something broke its weather simulation code.
5:30 Linus didn't disable it intentionally. It's an issue with the Threadripper CPU he's using (and that I'm using, too). Probably, just like me, he's got CPPC enabled in the BIOS but there's still no "_CPC" object present which is, presumably, a bug. It seems AMD _forgot_ to add this feature to their Threadripper CPUs. It should be available, but it is not - no matter the BIOS settings.
Many motherboard venders will set the bios for max compatibility when it comes to Ryzen products. Especially when they come out with new features.
That's frustrating as hell for a maintainer I bet.
CPPC is undesirable e.g. in virtualization and containerization scenarios where abstracted performance values interfere with hypervisor priority and scheduling rules for individual VMs and containers. This would have affected millions of servers around the world...
It is also recommended to disable it during benchmarking as it effectively skews results based on out-of-the-box provided values - CPU talks to UEFI which recalculates power and performance values and feeds it back to the OS that adjusts process priorities. That makes any benchmark irreproducible.
how to you open the page that you want to show(4:38)?, a shortcut?
I'm assuming it's ctrl+shift+t to reopen a closed tab, but might not be
i think he might be dragging it from another moniter
What about backup generators and cell Internet?
i might be wrong on this but i'm pretty sure kexec starts a preloaded kernel rather than loading stuff into a running one. my reason for saying this is that, as far as i know, kdump uses kexec to start a backup kernel when the kernel crashes so kernel crashes are easier to debug. again, i might be wrong, tell me if i am.
ahh, that explains a lot. thanks for clarifying.
9:09 Just read the line bellow that one! XD
I live not far from the border between Missouri and Illinois, while Illinois drivers aren't too bad, Missouri drivers can be. And my grandma doesn't drive that fast even on highways. There's a 65 min speed limit on the turnpike in Oklahoma and she went under that, most drivers in Missouri would honk at us going slow, but when we went to Oklahoma they don't care
Linus legit read that include file backtrace backwards.
Good grief, I’m on 6.1 and loving it. 6.8 - how did this end up in my feed.
Why does the whiteboard behind you say "LINUX MUST BE STOPPED"?
I think Linus was, indeed, reading that backtrace backwards. The only .c file mentioned was the last one which is what everything else was being included from. Linus isn't known for his humble acceptance of tiny mistakes, though, so I don't expect to see him admitting to it.
13:25: The not being prepared for cold weather is getting annoying. Ideologie did hit urban planning departments...
9:15 which is exactly what he said in the part below it, the part about kconfig
Linux 6.8 is not LTS, 6.6 is.
There is no drama.
Stay on safe releases and that's it.
You do have to reboot occasionally on most desktop systems. If you want updates, that is.
I hate every single person who spread the lie that you never need to reboot Linux, I've spent years trying to undo the damage those people have done
@@BrodieRobertson Lol, I'm an idiot, I thought you were saying the opposite. Glad to see we agree. 😂
Rebooting is very important
amdgpu has been kinda touchy as well lately, with 7800 XT. Until kernel 6.6, shutdown wouldn't always work, and would get stuck or end with a reboot. With 6.7, I can't reboot the machine anymore, because the GPU fails to initialise. Only a shutdown works right...
Right, I have been having trouble getting video signal after a reboot. I have to SSH into my machine, shutdown and start fresh.
6.7 running here without problem
@@globalcoupledances Specifically with 7800 XT? That's the point here.
It seems to me like a LOT is put in the Linux Kernel. Based on what I've heard about it's development anyway. I don't see why all sorts of drivers and such should all be inside the kernel. I may be wrong. But it just seems like with Linux LOTS of stuff get stuffed in the Kernel, wouldn't make more sense to make them separate drivers? May I don't know what I'm talking about.
Linus just merged some commits
3:26 "32 core zomg!" That's right around the point hyper threading starts being a hinderance rather than a help, and you hit bottlenecks in memory speed.
Those are actual 32 cores not hyperthreading. Or you are talking about parallelism in general?
@@cylemons8099 Parallelism in general. When you start hitting 32 cores and above for big complex compilation tasks, what I usually hear from fellow techs is to turn off threading in the BIOS/UEFI. Something about accidentally tying up the same core with two separate threads causing modules to choke each other or themselves? I'm a new-ish CSE, still finding out how little I actually know from my peers.
Sooo much props for Linus :) he keeps the linux kernel awesome
Yeah... :/ I have a feeling that at Microsoft someone is secretly waiting for his death so that the kernel development could eventually be put next to Minecraft's repository on github...
Based Linus. The greatest danger in the winter, is other drivers.
You could ask Wendell from level1techs about that bios setting I think he built Linus' workstation
As someone who lives in Oregon for most of my life, the ice storm is very unusual and thus a lot of the areas that were affected were woefully unprepared for it. Most of Oregon's power infrastructure is above ground on telephone poles whereas just a few minutes north in Washington, most of the infrastructure is underground (thus other family members of mine were mostly unaffected by the storm). I can 100% agree with stupid people in cars, most of the people who were driving didn't have chains on or they had them on incorrectly and then for whatever reason think that chains automatically means ice is not a problem. So you can imagine the chaos from such assumptions ;)
-10C is where I consider putting on a jacket going outside for a smoke...
15:40 Love the ending. One of these days I'm gonna run that command. LOL
Actually let me jut that down
Nuke OS command:
sudo rm -rf --no-preserve-root /
Linus Torvalds is in Portland, Oregon not Finland 13:08
Correct that's why the weather in Portland matters
11:39 Another words, "Leave the bug alone, it's not bothering anyone." LOL
I guess it's a peace of code that doesn't work but doesn't actually mess anything up. Basically just resulting in the feature just not existing.
"other drivers" -- for a minute I thought he was dunking on the garbage committed to the kernel :D
I have had nothing but problems since 6.4 , hell even all the builds disappeared apart from their first release , I love Linus messages
Maintaining big oss projects is difficult.
Oregon Sunday: oh noes, ice storm! Roads were fine. Today: "oh yay its all gonna melt and be warm!" Roads are sheets of ice. It's starting to melt., but it will just freeze solid since the high is about an hour from now. 2 degrees and raining. 2h it will be slush. 4h and it will be ice sheets again
It affects custom kernels too? or only the original ?
Custom kernels don't exist without the upstream project
@@BrodieRobertson So better wait next version ^
the universe dont want my project to finish either, so i an relate.
27°C == free to cool down however you like