Slashdot Mirror
Linux 4.0 Kernel Released
An anonymous reader writes "The Linux 4.0 kernel has been released. Linux 4.0 brings many features including live patching, Radeon DisplayPort Audio, RadeonSI fan control improvements, new OverlayFS functionality, Intel Quark SoC support, and a heck of a lot more. Linus's release announcement reads in part: "So I decided to release 4.0 as per the normal schedule, because there really weren't any known issues, and while I'll be traveling during the end of the upcoming week due to a college visit, I'm hoping that won't affect the merge window very much. We'll see. Linux 4.0 was a pretty small release both in linux-next and in final size, although obviously 'small' is all relative. It's still over 10k non-merge commits. But we've definitely had bigger releases (and judging by linux-next v4.1 is going to be one of the bigger ones)."
Four major versions in 25 years, who does Linus think he is? He's drunk with power!
Me too, I'd so prefer release numbers like Chrome.
Introducing Linux 3652!
Over 20 billion Linux kernels in use on a daily basis. How many exploits have your heard about, troll?
I prefer "Linux XP"
FTFY.
You do understand that Shellshock was not a Linux vulnerability, right?
Our sales department decided in our contracts a (totally arbitrary) policy to "support" only the last 3 major versions of our products. This means we periodically update the major version just so we can stop supporting the older versions even if there are not any major new features.
It's kind of funny that Microsoft has also decided to move their Kernel from version 6.1.62.6.2.7.21.1.6.2.2.000.02.432.523.253.532 to 10.0 in Windows 10.
I wonder if some useless managers got wind of Linux changing theirs, and made their developers change the number.
"Dear busy senior developers: I just found out Linux made a bigger number. Why can't we? Why are you wasting our time fixing bugs? Add this feature I just heard about today! It's very important.
Sincerely,
--Scott Miller"
Usually I just copy my home folder and reinstall my programs. If you don't install much stuff outside your package manager the harder part is remembering what to install.
I suppose I should start looking to upgrade my old Centos 6.6 box. I'm running a 2.6.32 kernel on that. I've thought about upgrading to the new Centos 7,0 but I'm not sure I want to fuck with it. The current system I have works perfectly and does exactly what I need it to do.
Well Centos 7 will cure that. It's based on systemd.
Yeah, the longer you wait to upgrade, the more stuff tends to break during the process. You may be past the point of no return.
I'm already starting to see that. Most of the stuff I use runs fine. Mysql, plexserver, samba and a few other apps. I'm already seeing stuff that will not compile for it. XBMC will not compile on Centos6, and nether will the latest version of HandBrake.
XBMC is not really an issue, but Handbrake I use all the time to encode stuff for the plex server.
Supporting World Peace Through Nuclear Pacification
Someone seriously, went to the effort to mod this down? Was it because I suggested that I might go to a dist with systemd on it? Really, don't you have better things to do than worry about what is running on my bitch box?
Supporting World Peace Through Nuclear Pacification
The current system I have works perfectly and does exactly what I need it to do.
Then I recommend not upgrading. The only reason to upgrade so if you run out of security updates or if you need newer features from a newer version.
SJW n. One who posts facts.
*IF* you don't change it.... or dare to dream by suspending it and expecting it to come alive again exactly the way it was before. Double headed may or may not work. Yeah, other then that it is stable.
Oh and you need linux support for all the hardware in your machine. And have the fortune that all new (within the last 10 years) features of your device are supported in the available linux driver. Not to mention whether it will support dual graphics cards.
And before you go on saying 'x' distro does it, or 'y' distro does it.. we don't always have the choice of all distros due to dependencies or legal reasons.
Linux is good, maybe even great, but it is not without its headaches. Personally I find Windows to be the most headache free OS. OS/X loses out in my mind because they have sacrificed a lot to make the UI basic and simple, and that is a matter of personal preference.
Laws are rules for the court, but merely a bottom bar to hit for life. Think beyond laws in your actions always.
I know that sounds like cynical marketroidism, but sometimes you do need to do that to wean people off some hideously ancient version they're still running on an old 386 under Netware 3.1 bricked into a wall next to the second floor men's toilet. "Last 3 major versions" sounds like a pretty generous strategy, we do "last n minor versions", where n is usually spread over 2-3 years. In other words unless you have a long-term support contract, if you come to us with a problem in a product written in the heyday of Windows XP, you're told to upgrade.
Are they planning to adjust how RdRand is used in random.c ?
Here's the list, though a few are mis-filed (the arbitrary code execution from this year is actually in Flash, no idea why it appears here), but most of the privilege elevation ones from this year and most of the arbitrary code execution vulnerabilities are real (though several seem to be in Logitech HID drivers).
I am TheRaven on Soylent News
Shouldn't that be in a separate driver?
Yes.... But isn't that what a "kernel module" is? Isn't it just a separate file that contains the functions you need to access the hardware? (or in the case of Linux, the source that you can choose to link directly into the Kernel that does the same thing?)
HINT: That's exactly what it is.
As I understand this, Kernel 4.0 allows some new features to be use by the drivers and some module authors are choosing only to support the 4.x branch of the kernel for their hardware and not have to maintain two code bases. Sort of like how windows broke up their kernel versions and how old drivers don't work with new kernels and new drivers don't usually work with old kernels.
"File to fit, pound to insert, paint to match" - Aircraft Maintenance 101
Yeah. I really wish there was a tool that would make it easy to export a list of all installed user software which you could could then import into your new system. Unfortunately the only technique I ever found (and I've forgotten what it was at this point) generated a text file listing *every* package installed on the machine - a list nigh guaranteed to bork a machine if I tried to import it all on a different OS version. And good luck sorting out the 10% of user software from the umpteen dozen pages of semi-cryptically named packages.
--- Most topics have many sides worth arguing, allow me to take one opposite you.
I'd like to interject for a moment. Sadly, the desktop is still extremely glitchy. Let me show some examples.
1) Notebookcheck a new Intel NUC. Intel HD Graphics 6000 was missing Linux support at the moment of writing. That's not the end of the world, but how does Linux Mint report about it? Nope, you don't get an informative "device not supported" message, nor does X.org fall back to a VESA mode. Instead you get corrupted graphics! Nice failure mode there. Just look at the screenshot in the article. Does that look professional to you?
2) When you install Linux, various manual hacks are needed to correct all sorts of little glitches here and there. Read the installation report of this guy. Does that seem familiar?
3) Laptop brightness adjustment still goes in multiple steps! I can't believe this bug is still around. The same issue is in Ubuntu in Mint and affects most laptops. Bug #527157. Just try pressing the brightness keys of your laptop under Linux and you see what I mean. An everyday feature like this should Just Work without me having to even think about it.
Conclusion: I need an desktop operating system that is more deterministic in behavior. I want robust and predictable user experience. This is not rock solid at all.
Handbrake? Your distro comes with a wife?
I know you're a troll but say that next time you do anything with your bank card!
You'll generally just have to rebuild your app.
If you picked some technology which is windows-specific to design it, then it's probably your fault that Linux won't run it, not the kernel's.
Oh, and we've got wine, in any case, which seems to run a great many things just fine.
And yet, even if bash is part of most Linux installation, the weakness is not across the board. For instance, the default shell (that is used by CGI scripts) is often binded to dash instead of bash.
Having bash as the default user shell does not have anything to do with shellshock.
"Oh and you need linux support for all the hardware in your machine."
Are you implying Windows doesn't need support for all the hardaware in your machine? Seems a bit weird.
To run Windows desktop applications on the Linux kernel, you'll need to install the X Window System and the Wine application environment on top of it. (Most desktop-oriented distributions of GNU/Linux include both of these in their repositories.) Try running Windows desktop applications in Wine, and if they don't work, you can report the failure to both the application's publisher and the Wine team.
I'm saying that I have never had a problem obtaining a driver for windows that allows the device to work as advertised. Not the case on Linux. I've spent hours installing drivers, or installed the 'commercial driver' only to find that the device is missing features or is flaky with KDE and or Gnome. I've had many devices that don't work properly on OS/X.
Laws are rules for the court, but merely a bottom bar to hit for life. Think beyond laws in your actions always.
I'm not sure if this is what you are saying, but it is a bit much to expect to be able to upgrade your OS and have it still work with all your devices. Windows has a very long support Window so you can still use all Windows 7 devices. I admit I have a dongle that only worked with windows xp and I was pissed about that. But at least it worked in windows. I tried it in Linux and it was somewhat there but not really. IT was a home automation dongle.
Laws are rules for the court, but merely a bottom bar to hit for life. Think beyond laws in your actions always.
The big difference is manufacturer support. You can guarantee Windows is supported by every manufacturer - you don't even have to check, it's a given. Linux has a smaller presence, especially on the desktop. Small enough that manufacturers may or may not support it, and even if they do support it their drivers may be less refined.
"That may sound funny to you guys [...] Eventualy we found two servers standing next to each other, underneath a pile of cardboard boxes and rolls of toilet paper, in the building supervisors storage cupboard, both running netware 2.15 ( it was early nineeties). Both with an uptime of more than two years."
Forgetting the issue about no backups, I don't find it funny but sad: it just shows the utterly lame situation of IT the fact that something like this comes to a surprise instead of being the norm.
I do a lot with my bank card. and not trolling Linux does not have anything usable for video editing.
This is a 100% solid fact, No Adobe Premiere, no Sony Vegas, no Final Cut Pro, no AVID...
Do not look at laser with remaining good eye.
Parent poster here, I tried to convince the my manager and sales department to give support on a time basis based (X years after delivery date), but the "that is not how other companies bigger than ours do it" argument won over mine. So as a result we update our versions based around the time we think we supported the old versions long enough.
I think a combination works best, major releases makes most sense when it comes to technology/code you must support while a minimum number of years makes sure you won't go Firefox and suddenly be at version 40. For all the other bashing Microsoft can take, their policy is actually a good model:
Mainstream Support for Business, Developer, and Desktop Operating Systems will be provided for 5 years or for 2 years after the successor product (N+1) is released, whichever is longer. Microsoft will also provide Extended Support for the 5 years following Mainstream support or for 2 years after the second successor product (N+2) is released, whichever is longer.
Basically you get a guaranteed 10 years of support from release, 7 years as long as you're buying the latest version (minimum 2+5) and if it takes longer the support period stretches too. Don't expect another XP though, Vista runs from January 2007 to April 2017, Win7 from October 2009 to January 2020 both 10 years, 3 months. And with Win10 seemingly on schedule for release this year it'll be the same with Win8.
Live today, because you never know what tomorrow brings
Why do you think those people are running "some hideously ancient version they're still running on an old 386 under Netware 3.1 bricked into a wall next to the second floor men's toilet"? Do you think it's in order to upset your tech department or it might be because it does what it needs to do so if it's not broken why should I need to fix it?
To add insult to injury, remember you would have no problem supporting ancient versions of your software if there were no bugs to fix on it so what you really are doing is failing to provide the customer with a product that just works as it should without broken parts.
The way the world actually is, the acceptable market standard for software is not "perfection", and beyond a certain date back-porting fixes is just unrealistic. Sometimes the fix IS to move onto a newer version. You can't easily get comparable PC's and software of that age for developers to test on, and expecting every developer you hire to not just learn about the current system version, but also about the detail of decades of development history before they can get working is implausible. Asking a recent grad to learn about Netware system calls that were obsolete a decade ago is silly.
More specifically, if a system is important to a business, it should not contain a single-point-of-failure based on irreplaceable tech anyway. It's an immediate business continuity audit failure in any regulated industry.
IOS and Windows are "distros".
Hence Ubuntu being at 15.04...is ahead of the "10ers".
Sadly, perceptions matter to humans.
4wdloop
Look up Rice's Theorem. Or work on a major software project. It goes way beyond unfair to expect a complex software system to "just work as it should" - it's mathematically impossible to make sure it does.
Beyond that, supporting every single ancient version just because one guy somewhere might be using it would take man-hours away from supporting the versions most people use. It sucks for that one guy, but it'd suck for everyone else if a gaping security hole in a more common version were to stay unpatched for too long.
Furthermore, the older a version of a program gets, the more of its devs switch jobs, retire, etc. You can bring on new devs and make them learn the old code base (current job market notwithstanding), but once again, that takes man-hours, especially if the old code base was written before everyone started thinking about maintainability. You could instead put them to work adding a feature to the next version that a larger segment of the market has been wanting. And you need to keep releasing software - the competitors are, and the devs' salaries don't appear out of thin air.
Yeah, what can be done about this? The current GNU/Linux alternatives still have a catch up time of at least a decade in this field. Pixel editing is okay, vector editing is one of the success stories as well as digital photo development. But video is way off...If only one of these major companies can get a port to work on GNU/Linux...unless we can crowdfund openshot or some similar ilk to close the gap...
"SO we bide our time, waiting for a purer kick to bloom and the future is still bleak, uncertain and beautiful" -GSYBE
Maybe he just wanted to preemptively discourage the subsequent bitching and moaning we'll hear from you once you do experience the problems of systemd first-hand?
Why would I bitch about something I have been using for 2 years? I'm very familiar with systemd, and doubt I would bitch about it.
Supporting World Peace Through Nuclear Pacification
So www.lwn.net has a better section dealing with the 4.0 kernel including what persistent memory and lazytime changes are. I would suggest going there.
http://www.lwn.net/
I like to find more information than the smattering that was a "detailed look". as they say.. NOT.
I can program myself out of a Hello World Contest!!
You do realize, bash is in cygwin, so by your standard shellshock is also a Windows vuln, right?
Not really a double standard - those are from the same company targeting the same platform. It would be like grouping CVEs from bash, findutils, tar and time.
In practice Bash is part of most Linux installations.
Even in the realm of "GNU/Linux" not everybody uses bash (some use zsh, for exemple).
And that's only the portion of users running an actual "GNU" userland.
Then you have the embed world using Busybox (with uClib, etc.) and co for the userland (which has its own simplified shell).
And then you have Android (which runs a completely different user land by Google, like Bionic for a C library, a different message passing bus, and most of the things usually handled by deamon running in userland, handled by java-like code on a java-like VM).
And the other way arround: you have other Unice (OS X, various *BSD) which obviously do not run Linux kernel, but do run bash.
OS X, for example, was affected by bash.
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
Well, no, let them use their date based numbering versions.
Just remind them that Windows was at version 2000 sixteen years ago...
GLaDOS for President 2016! "Well here we are again. It's always such a pleasure." -- GLaDOS, 2011
Uh what, you're kidding right?
I mean, 3652 is SO yesterday. We're on version 4023 this microsecond! 4077 I mean... damn it, it changes faster than I can type.
GLaDOS for President 2016! "Well here we are again. It's always such a pleasure." -- GLaDOS, 2011
Yes, you should look up Rice's Theorem.
With cars, it's posible. With complex software it isn't. There's a reason for that.
Rethinking email
You lost me at Windows is the most headache fee OS. I left Windows in 95, and never ever will come back. I also have the dubious honour of throwing out of the window of my car a Windows based GPS and buying a Linux based on the same day. The cherry on the top is our Windows admin installing a Linux system for his desktop a couple of weeks ago because he had already had had enough with viruses. Linux may have had all the problems you talk maybe 15 years ago, you have to try it again.
Those who say Windows is supposed to have a long support of something have no idea what hey are talking about.
You are completely out of your head...in fact Linux supports even better older and newer machines alike. Recently I had to reinstall one old desktop and two older laptops. Windows has not cooperating, it was a walk in the park installing linux, no drivers hassle, nothing at all. Plus, in one of the notebooks I had a specific request for Linux for a non-IT user because they are already tired of dealing with malware.
If the reputable AC says so. The fact is that I can have a Linux server going for years, Windows needs periodic reboots, and the latest OS/X crashes at least once a week. But then, the AC knows best, exactly?
If the Wine team discovers that a particular application is relying on unspecified or undefined behaviors of a particular Windows function, behaviors that may break in a future version of Windows, then who is responsible?
Do you work for Sage?
Mean what you say...say what you mean.
Bash is older than Linux and has been available for just about every version of **IX for a couple of decades, and even Windows (via cygwin).
OS Software is like love: The best way to make it grow is to give it away.
Mode me down all day long in opinion, if you want. Which is not the way mod points should be use but I digress. The fact is that the uptime of a Windows machine can only be measured in months, or actually weeks, OS/X unfortunately for me and others, is crashing at least once or twice a week with it used to be rock solid for months before 10.8, and the uptime of Linux boxes can be measure in months or years, and often I only take them down to do a kernel upgrade. But obviously the slashdot crowd knows best.
Mode me down all day long in opinion, if you want. Which is not the way mod points should be use but I digress. The fact is that the uptime of a Windows machine can only be measured in months, or actually weeks, OS/X unfortunately for me and others, is crashing at least once or twice a week while it used to be rock solid for months before 10.8, and the uptime of Linux boxes can be measured in months or years, and often I only take them down to do a kernel upgrade. But obviously the slashdot crowd knows best.
"Look up Rice's Theorem. Or work on a major software project. It goes way beyond unfair to expect a complex software system to "just work as it should" - it's mathematically impossible to make sure it does."
No problem with that. So the software doesn't work as it should when shipped? Ok. But that's not the point: I was not talking about the software not being ready when shipped but the vendor's negative to correct it. Is it ten years down the road when I step into a broken part of the product? so what?
"supporting every single ancient version just because one guy somewhere might be using it would take man-hours away"
No, no, and one thousand times no. It is not taking away hours from anywhere since your software is broken. The fact that you delivered broken software works in fact the other way around: you were allowed to take hours away to your delivery date and you are allowed to only recover them afterwards as people stump in the bugs and the need arises.
"Furthermore, the older a version of a program gets, the more of its devs switch jobs, retire, etc."
I see... so, again, what? Taking your own words, look up Rice's Theorem: you knew all that the very day you shipped the product and still you didn't plan for it? It is the vendor's problem, don't try to make it into a customer's one.
You see... the software business is in pathetic shape, we accept quality levels that would ashame any other industry and still we racionalice the statu quo to make it more palatable instead of taking professional pride into making it better.
"With cars, it's posible. With complex software it isn't. There's a reason for that."
No, with cars isn't possible either. Every major brand has recalls and maintenance programs for that very reason. The difference is that respectable brands will launch a maintenance program for a detected flaw even if it is ten years down the road and customers will enrage and go for a class action if they don't.
And cars, being physical objects will have wearing parts and, again, people won't accept a "buy a newer model" answer from the vendor ten, fifteen or even more years later when going for a replacement.
Not really - Microsoft's love of backwards-compatibility is well known, and features prominently in Windows. The fact you've not got any concrete examples isn't exactly helping you sound any different from an offended fanboy. I know you're not one, so you might want to work on that ;)
I think the main points people are disliking about your posts are: Confusing the notions of "ruir" and "everyone else", mistaking your opinion for fact, and offering your handful of anecdotes as being representative of the experiences of everyone else.
Actually, Linux does have marketshare. It's got huge marketshare of the server market, and servers can be high-value targets for a large variety of reasons. There are people who want to target Linux. They've just had a much harder time doing so than they have with MS-Windows.
OS Software is like love: The best way to make it grow is to give it away.
Since you clearly didn't look up Rice's Theorem, here it is:
Any non-trivial property of a Turing machine is undecidable.
In other words, there is no methodical way to guarantee anything interesting about a piece of software, and that includes whether it works properly under every input. You can verify that it works with "typical" inputs, but there will always be some set of boundary conditions that you couldn't possibly have known to check on day one.
I suffered from the hard freeze in 3.17 & 3.18. http://linux.slashdot.org/stor... Anyone happen to know if it's been resolved in 4.0?
"In other words, there is no methodical way to guarantee anything interesting about a piece of software, and that includes whether it works properly under every input. You can verify that it works with "typical" inputs, but there will always be some set of boundary conditions that you couldn't possibly have known to check on day one."
You can put it any way you want, since I already accepted that and asked "so what?"
You are arguing that you can't deliver flawless software by day zero (which I alredy accepted) as an excuse to not correct the flaws once they appear (which is what I make a point of).
You are, again, rationalizing.
My desktop has a slightly-off ACPI implimentation. Linux (at least this distro) crashes during kernel init unless you add acpi=off. This is a fairly common problem: Windows has a horribly off-spec ACPI system, it's a real mess, but manufacturers test extensively to make sure their mainboards are fine with it and include all the required workarounds. They have little reason to test so extensively for linux, and so can crash when given an OS that actually follows the standard.