Slashdot Mirror
How To Upgrade Linux To The 2.6 Kernel
An anonymous reader writes "Here's a good computer project for the long labor-day weekend. KernelTrap has posted a howto detailing eight steps to upgrade your GNU/Linux OS from the 2.4 stable kernel to the 2.6.0-test development kernel. Complete with screen shots, the end result sounds to be well worth the effort." Since chances are most people will be upgrading anyway once 2.6 is deemed release-worthy, it's always worth learning the upgrade procedure well.
I followed all the steps, then this is what happened:
-bash-2.05b$ uname -a
Darwin Bruce 7.0.0b1 Darwin Kernel Version 7.0.0b1: Tue Jul 29 15:27:33 PDT 2003; root:xnu/xnu-470.obj~1/RELEASE_PPC Power Macintosh powerpc
I'm really confused, any ideas?
As much as upgrading my kernel to 2.6.x interests me, I think I'll build that deck, instead. Maybe on some crappy, rainy weekend I'll play kernel games, but man, the sun is shining.
Is this a different numbering scheme?
"Since chances are most people will be upgrading anyway once 2.6 is deemed release-worthy,"
:) I've been using it ever since the first -test was released, patched it with Andrew Morton his -mm and it's fast and solid for me!
IMHO it already is
If you haven't tried it out already, go download -test4 now! Even if it's just to see if all your hardware works, if you report any problems now you don't have to deal with them when 2.6.0 is officially "stable".
I am a viral sig. Please help me spread.
Can't I just download one file, double-click on it to install, and re-boot the computer?
... don't forget to buy your license from SCO before using the kernel.
The IT section color scheme sucks.
apt-get install kernel-2.6.0
Hey, I love the idea of Linux as much as the next guy...
However, you have to see the whole idea of needing a step-by-step upgrade guide with screen shots, etc is exactly why Bill still owns huge percentages of the market. Windows upgrades: Insert CD.
Anyone know why they still require gcc 2.95? Or is this a minimum? Will it compile and run with gcc 3.3.x without problems? I was under the impression they tried to target the current stable version of gcc on each new major release.
As noted in the article, the build output is much cleaner (simple status lines for each section/module being built, not the whole gcc cmdline), the make options are now fully documented (with make help), and make is simplified down to `make all' and `make install'/`make modules_install'.
I'm not particularly fond of the new make xconfig, but didn't give it much of a chance. I went with `make menuconfig' and ncurses instead.
Performance is noticably improved. Not just "some people told me it's better and well, maybe it is a little", but actual tangible improvements. Even typing into xterms seems faster. (I did enable the preemptible option, but this seems even better than when I did it with the old patch to 2.4.)
This is the most pleased I've been with a new kernel in ~6 years of using Linux. Highly recommended!
#19845
....going out and enjoying the shiny day with your friends, maybe doing some sport etc, instead of crawling down to the basement and recompiling kernels for several hours???? /. you freak!!!!!;o))))))
You're officialy banned from visiting
1. No sig. 2. ???? 3. Profit!!!
Are traditional pseudoterminals still supported or the Unix98 scheme the only thing available? I built
the kernel successfully, but found that xterm and
rxvt didn't work because they didn't have pty's.
Mach is a microkernel, and Apple is the #1 UNIX distributor in the world. But thanks for playing.
Really? The two nicest desktop operating systems I've used are MacOS X and BeOS. OS X is based on the mach microkernel, while BeOS has its own microkernel. And before you say BeOS is dead, take a look at the new version (still in private beta).
Microkernels are still very much alive. They don't give quite the performance of macrokernels, but they have a number of advantages (like not needing a reboot to replace large portions of the kernel, and drivers not being able to crash the kernel). With current system speeds, the flexibility of a microkernel is well worth the speed trade-off, on the desktop at least. On a server / workstation I would probably still choose a macrokernel.
I am TheRaven on Soylent News
Perhaps it is release-worthy to those on an ix86 platform, but I had to modify include/smp.h to get it to compile on sparc, moving #include into the #ifndef __ASSEMBLER__ section to avoid the redefinition of ALIGN that caused compiling to fail.
hrm this howto doesn't talk about much except "how" to compile it...i was expecting some info about how to use/enable the elevator stuff i've heard about(and which to choose)...anyone?
The security system should be stronger integrated into the kernel. Yes there is PAM these days, but a real secure system will have authentification and verification added at kernel level to make it trust-worthy and tamper-proof.
Security should be enforced in the kernel but should not be put in the kernel. Here's what I mean, I do not want a kernel that performs authentication, but when authenticated it should stick it to it. I believe that is how the kernel works, and its much better than putting "PAM", SASL or Kerberos or Login into the kernel. I don't even think SELinux even tries to do authentication in the kernel.
Grid access. Given the exciting new development of grid computing,
Grid computing has been around since the 60's. I actually don't see where the kernel can use grid computing (meanwhile you can do what we do and use openMosix to get many of the same benefits). All you really need to impliment grid computing is SSH, RSH or even telnet and none of those things are really kernel worthy either.
Given to above grid access all journaled file system should be made grid aware and supporting distributed storaging.
Check out NFS, then AFS, then CODA and finally intermezzo depending on the level of local caching you wish for your grid.
More modular kernel design. While device drivers can be loaded at run-time, a fully modular kernel design is still missing. Yes, the micro-kernel is quite dead, but a decent modular design can be quite powerful as well-used system like Windows NT and successors show.
You mean, reboot every IP change NT? How about reboot every program install XP? The modularity in NT is not very good at all. XP is much better, but still not the keen level that Linux achieved back in 1996.
While Linus's opinion on the matter is well known, Microkernels are far from dead. It's just that Mach gave them a bad name. Mach was too bloated and too slow, while the new breed of microkernels have unbelievably fast IPC primitives and therefore the potential to revolutionize the way operating systems are built. Mach sucks != Microkernels suck.
See, for example, the L4 project.
to HI, without realizing that his fauxking ego/need to appear sucksassful to his corepirate nazi sponsors, is useless against the power that he/they MUST deny the existence of.
carrIE on dough. we're here for you, should you begin to see the light, in costa rico, or wherever/whenever you look for it. theres' no going back, & nowhere to hide, anymore.
You really need someone to redocument the 8 steps to upgrading your kernel? Kinda makes me appreciate FreeBSD even more.
/usr/src && make kernel KERNCONF=MYCONF && reboot
cvsup -g -L 2 -h cvsupN.freebsd.org stable-supfile && cd
The one thing not mentioned in the article, and the one thing that has me nervous about trying 2.6-test is the changes to alsa. With 2.6, alsa is built into the kernel, so presumably this makes it easier to set up in the first place. But I already have alsa set up perfectly in 2.4, complete with OSS emulation and artsd sound mixing, so that all my apps play nice and just work. How much deconfiguring and reconfiguring am I going to have to do if I'm going to be jumping back and forth between 2.4 and a possibly unstable 2.6? Especially since I have the rather finicky via82xx driver. I'm really keen to try out 2.6, but not if I end up breaking sound in the process.
I've been trying -test1 and -test4 on my desktop and laptop for some time now. It is perhaps hard to believe, but the new kernel is very much _noticeable_ on the desktop. How? Well, for instance, you can 'feel' it when moving the mouse and watching the pointer on your screen. The lag between the physical movement and the mouse pointer has become almost unnoticeably small, even when apps are hogging CPU. Another nice touch is that your desktop keeps this responsiveness with large processes (say, an 'emerge mozilla') running in the background. With 2.4, terminals would be a bit slow at starting and such, but that is all gone now. It is also very pleasant that ALSA is now in the kernel. It saves lots of hassle compared to 2.4, where you had to compile the modules separately. Low latency audio performance should be less of a black art too with this kernel.
Cons:
Some defaults were funny at first (like missing console drivers, etc.) and I've noticed the mouse being a little jumpy some times. Nothing big so far.
All things considered: great kernel! Thanks guys.
I'm already running RedHat Linux version 8. Why would anyone want to downgrade to Linux 2.6?
Wait for download, select "260-4" in lilo when you reboot.
See, Mandrake cooker makes things so easy, its a HELL of a lot easier than Debian Unstable!
Thrice I was able to compile to compile the kernel - twice died on uncompressing kernel and the third time was unstable with no sound. Guess that explains my n00b status.
If kernel developers (KD) want the general public to test a new kernel - they really should put out some real documentation of how to compile/build the kernel on some popular distro like redhat - Well I guess the KD response would be - No kernel for you.!!
We do not have a history of profitable operations. Our future SCOsource licensing revenue is uncertain.
Wouldn't a server be the place where stability matters the most?
So from your argument I would have thought that that would be a good place for a microkernel
1) Get Linux 2.6
2) RTFA
3) Compile and install it
4) Profit!
Nothing surprising here.
I don't need no instructions to know how to rock!!!!
I'll just use the distro build which is packaged properly. That's not to say I'm not excited by the new features, etc., but I've long ago decided my life should not be spent compiling and tweaking things in which I have no particular expertise or passion. Those with expertise and passion are going to do a better job.
[
I use the cryptoapi+cryptoloop for encrypted filesystems on just about every one of my machine's where user data might be stored. However, one of them (my workstation) I wanted to try 2.6 on for shits and giggles (it's a box that regularly gets trashed and screwed up, so it's little concern). When running 2.6 with all the appropriate modules compiled and loaded (including the cryptoloop block device) it *always* fails to mount them! Even if I just create a tiny little file and try to mount it through a cryptoloop device, the system (reiserfs specifically) always complains about not being able to recognize any data on the device (no filesystem). The behavior is conistent with entering an incorrect passphrase on an existing device.
What could possibly be the problem? Does 2.6 require a new suite of userspace tools to do this? Is the 2.6 cryptoloop device currently broken? Has anyone else run into this problem? I'm really excited about running 2.6 (given the often hyped performance improvements) but this is the only thing holding me back.
Join Tor today!
Anyone know if there is XFS support built in vanilla, or when a patch is coming? :)
Users give a shit about upgradability. The 98 network at the business I started work at wanted me to upgrade everyone to XP so the machines would stop crashing. I did, and the crashing stopped.
Users will upgrade, because they want to have "the latest thing." You're underestimating shiny-things-syndrome.
"Sufferin' succotash."
Anyone else having problems with the
"Ethernet controller: Lite-On Communications Inc LNE100TX (rev 20)"?
It works fine with 2.4.21, but it's not properly recognized with 2.6-test4.
It's kind of hard to find out if there's anyone else with the same problem, since there are so many matches on "LNE100TX" and such on google. There seems to be quite a few revisions ov the said card.
To be fair, i havent seen any program that was not somehow hw-related(dx-release, video-card driver) that needed a windows restart. Not in xp, not in 2000, not even in 98.
Its just that many versions of installshield had these "Restart after finish" option everybody used. A click on cancel and then starting the program hasnt ever failed.
HI O WISE PRINCE. WHT TOOK U SO DAM LONG?
I run Debian stable so I'll have to wait until the next perihelic opposition of Earth and Mars before being able to apt-get a 2.6 kernel image.
Microkernels ... they have a number of advantages (like not needing a reboot to replace large portions of the kernel ... a microkernel is well worth the speed trade-off, on the desktop at least
Is being able to replace large portions of the kernel without a reboot on a desktop really that much of advantage?
- ebh
I have been useing the mm patch on every 2.6 kernel since test1. I have installed it on 3 machines (my desktop, my friends desktop and my laptop). It has been running rock solid for me. The sound quality is great due to the alsa integration, ACPI is working great on my laptop. Though some people complained about ACPI causing the kernel to crash on boot with test 4 I havn't encountered this with test 4 mm sources. Although I wouldn't put it on a server just yet it is definately the best desktop kernel release yet
History will be kind to me, for I intend to write it - Sir Winston Churchill
...fire up a few torrents and install the result. (-:
Got time? Spend some of it coding or testing
Sorry - but I disagree:
;-)
There are very interesting and important new features in 2.6 that should make it to everyones server/desktop fast. Moreover I think that from time to time "bugfix cycles" - that means phases when features and improvements have to face the real world are very important for the development of software.
To your examples:
1) This is no easy task, moreover as this Fritz chip and/or other technologies will probably be integrated into todays motherboards it is smart to wait until that happens. Moreover we all know that 98% of system compromises come from buggy 3rd party software (e.g. sendmail/ssh) and I see no real way this could be prevented in the kernel without braking compatibility with existing software.
2+3) Grid computing is something very interesting but I think only important for a small percentage of Linux users. Topics like "My mp3 player stutters when burning a CD" are a lot more important. Moreover grid computing is stil lacking standards, the same with distributed file systems. Moreover creating distributed file systems don't seem like a task for people like Linus Torvalds.
4) Why exactly would you need a more modular design apart from design issues? Will it bring better performance - probably no. Will it ease the development - don't think so.
I would not say that "version numbers spew out at Niagara falls level". When was 2.4 released? let's see: "Jan 4 2001".
I am happily expecting 2.6 - and hopefully the 2.6.0 release will not suffer from too many bugs - perhaps it should be released by Marcelo and not by Linus?
You can find Redhat 9 rpms of the 2.6-test series at http://people.redhat.com/arjanv/2.5/RPMS.kernel/. There are also rpms for all the necessary packages that the 2.6 kernel requires. I've tried out 2.6test4 on my machine and it works quite well.
The opposite of 'microkernel' is not 'macrokernel', but rather 'monolithic kernel'.
... it's still a sexy task for developers to track down kernel bugs and stabalize their work.
Anybody start thinking about Austin Powers at this point?
FatBastard: I'm Sooooo Sexxxxxxyyy.
Sig it.
So we're assuming these instructions aren't for the casual linux user.
When 2.6 has filtered down to Mandrake, Red Hat etc.... you will get the kernel as part of an upgrade.
Mandrake 9.2 will come with an optitional kernel2.6 package...
So is linux faster than FreeBSD now? I switched from linux to freebsd because I thought SCO was going to kill linux. Maybe it was wrong for me to switch. If linux is faster now, I'll consider switching back when the kernel is stable. In the meantime, how do I install this kernel in FreeBSD so I can try it out?
Check the readme for the apt or yum lines to add to your configs.
I used apt4rpm to easily install 2.6pre4 yesterday.
mr.
And until it's as simple as upgrading with a few mouseclicks you have your answer why Windows rocks and Linux doesn't.
You can help the kernel developers immensely by testing your kernel methodically and thoroughly rather than just casually trying it out.
It's also important for you to test new kernels, even stable kernels, before putting them to use on a production machine. Even if they work well for everybody else, you may be blessed to discover your very own bug.
Also realize that because Linus can issue a new kernel anytime he feels like it, there is no particular requirement that a kenel be tested before its released. It's happened a number of times that "stable" kernels have been released that have turned out to be quite broken, especially on non-x86 architectures.
So please read, enjoy, and put to good use:
-
Why We Should All Test the New Linux Kernel
-
Using Test Suites to Validate the Linux Kernel
The OSDL kindly prepared Japanese translations but for some reason have taken them offline. I have copies though and will try to post them sometime soon.There are other articles on web application quality and C++ programming, with more to come. So far they are all under the GNU Free Documentation License.
I am actively seeking more translations if you want to help out.
Request your free CD of my piano music.
Anyone else getting timeouts for kernel.org? Have we slashdotted linux?
)
Yes, the micro-kernel is quite dead
Linus is just human. He might be right. He might be wrong. He does have some insight, and does deserve respect, from anyone who might disagree. Most of all, he is entitled to his opinion.
I do disagree with his opinion on this, but not strongly.
My rationale is this. I accept that Microkernel's are less efficient. But computers aren't getting any slower either. Remember in 1983 when some doubters believed that GUI's would never amount to anything and we should just stick to DOS and not get a Mac?
Another argument is: if it is true that a large monolithic kernel is more difficult to maintain and compartmentalize, then it may have a long term disadvantage. In the short term, I'm delighted to be writing this using Linux. In the long term, if something better comes along, it would not have far reaching effects into everything on top of it. Only the first couple layers. If a newer better kernel were to hypothetically come along, then in a few years, I might still be using a comptuer with substantially the same software as I use today. My desktop would look the same, OpenOffice the same, Mozilla the same, etc. And if I'm wrong and Linus is right, then in a few years, my system will still look substantially the same as it does today.
If it is true that a microkernel is easier to maintain, and to dynamically load and unload servers (i.e. filesystems, drivers, etc.) then this might be a significant benefit in the future. For now, of course, Linux is the best thing going, imho, and I'm happy to use it until/unless something better comes along.
I'll see your senator, and I'll raise you two judges.
OS X is really a bad example of a micro kernel OS because it is a single server BSD system running on top of Mach. Mach is a really old Microkernel and stuff like "drivers not being able to crash the kernel" aren't true with it because the drivers are compiled into the kernel (or loaded into the same address space as modules, depending on which implementation of Mach one uses). Yes, Apple's Mach uses userspace USB and Firewire drivers through the usage of different libraries, but Linux can do that to with libusb and the raw firewire library.
Now, QNX. The Microkernel used in QNX is a very nice one. QNX is a great example of a high performance, multi-server, Microkernel system. Drivers are in userspace, each 'kernel' component executes as its own server, and the entire thing is fast and real-time to boot. I think that QNX disproves the whole "Monolithic Kernel Operating Systems are faster than Microkernel Operating Systems" argument.
There is also plan9 which is a multi-server Microkernel based OS. I don't know if BeOS is multi-server or single server, mostly because I've never used BeOS before. You also have the Hurd which, although it uses Mach (for now), is still a multi-server OS. The Hurd needs a lot of work and there is no longer any work being done on it except for the L4 port (and, as I found out last night, a few people working on getting the Hurd to work with OSF Mach + OsKit and therefore the PPC and other archs).
And about the whole "being able to replace large portions of the kernel without a reboot stuff," that is only semi-true. You can't replace the root file system translator without rebooting. Or the task or memory server. Or the authentication server. On the other hand, you can replace any file system translator that is not required for the system to find the new file system translator, the network stack, and many other things. But the fact still remains that a lot of stuff still requires a reboot. But it is still nice that the number of things that require a reboot to change is minimized. There is also the nifty feature of the Hurd (and probably other multi-server OSs) where one can start a "subhurd" which is much like an instance of UML Linux. Except that one doesn't need to patch the kernel. The only thing you can't do is touch hardware that you don't have the permission to touch (same thing for UML Linux).
HAL 7000, fewer features than the HAL 9000, but just as homicidal!
xnu, the Darwin and OS X kernel, is actually monolithic. It just provides mach-like abstractions. Or something. See here (section 2) for details.
How does this 2.6 series compare against the optimized 2.4 gentoo series? Does it worth the change?
Hosting 20G hd, 1Tb bw! ssh $7.95
as a Gnome user
make config
make menuconfig
make xconfig
make gconfig
How in the world is posiible, that no KDE user has been whining yet that there's no make kconfig
Does that mean that THE G/K Cold War is over???
Signature Pro version 1.13.2-3 release 83.5 beta3try7 after-breakfast edition
root@phoebe 2.6]# rpm -Uvh kernel-BOOT-2.6.0-0.test4.1.32.i386.rpm
warning: kernel-BOOT-2.6.0-0.test4.1.32.i386.rpm: V3 DSA signature: NOKEY, key ID 9d6b4012
error: Failed dependencies:
mkinitrd >= 3.5.5 is needed by kernel-BOOT-2.6.0-0.test4.1.32
modutils >= 2.4.25-7 is needed by kernel-BOOT-2.6.0-0.test4.1.32
nfs-utils 1.0.3 conflicts with kernel-BOOT-2.6.0-0.test4.1.32
initscripts 7.23 conflicts with kernel-BOOT-2.6.0-0.test4.1.32
The best I can describe the Radeon Framebuffer Console in 2.6 is "Whacked."
It's like it puts the console on 1/4 of the display, and it "bleeds" one vc onto another.
Sorry that's the best I can describe it.
-fb Everything not expressly forbidden is now mandatory.
Give it up, Mr. McBride.
The Tao says: the perfect piece of paper is unmarked by pen; the perfect flower is unpruned by shears; the perfect operating system is untouched from its default installation.
I've had to support, debug, fix, and otherwise un-screw-up many computers in my time. Inevitably, the closer a system is to what everybody else is using, the more likely it is that any problems with it will have been seen and solved countless times before.
That's why the idea of countless legions of users out there each recompiling his own kernel just makes my blood run cold. This is the twenty-first century, peoples! Why is it necessary for anyone other than a kernel developer to compile the kernel sources? Why haven't all the optional pieces been broken out into modules yet?
[root@phoebe linux-2.6.0-test3]# make mrproper /usr/lib/qt-3.1/bin/moc -i scripts/kconfig/qconf.h -o scripts/kconfig/qconf.moc /bin/sh: line 1: g++: command not found
CLEAN scripts/kconfig
CLEAN scripts
RM $(CLEAN_FILES)
Making mrproper in the srctree
RM $(MRPROPER_DIRS) + $(MRPROPER_FILES)
[root@phoebe linux-2.6.0-test3]# make xconfig
HOSTCC scripts/fixdep
HOSTCC scripts/split-include
HOSTCC scripts/conmakehash
HOSTCC scripts/docproc
HOSTCC scripts/kallsyms
CC scripts/empty.o
HOSTCC scripts/mk_elfconfig
MKELF scripts/elfconfig.h
HOSTCC scripts/file2alias.o
HOSTCC scripts/modpost.o
HOSTLD scripts/modpost
HOSTCC scripts/pnmtologo
SHIPPED scripts/kconfig/zconf.tab.h
HOSTCC scripts/kconfig/conf.o
sed scripts/kconfig/lkc_defs.h 's/P(\([^,]*\),.*/#define \1 (\*\1_p)/'
HOSTCC scripts/kconfig/kconfig_load.o
HOSTCC scripts/kconfig/mconf.o
HOSTCXX scripts/kconfig/qconf.o
make[1]: *** [scripts/kconfig/qconf.o] Error 127
make: *** [scripts/kconfig/qconf] Error 2
That is one of two items that he missed in the article. Indeed, in the 2.4.x kernels if you used devfs you did not need to enable /dev/pts file system support but you do with 2.6. Interestingly, now that devfs is officially in the kernel they are thinking of replacing it with udev.
Also, it is not mentioned that you need to create the /sys directory. It apparently has something to do with sysfs which I don't quite understand yet, but it was necessary to get all my sensors reading correctly.
This seems to be yet another in the growing collection of mostly useless 2.6 "migration guides". It doesn't mention any of the common gotchas with configuration, its recommendation for invoking the build process is wrong, etc, etc.
A much better guide is Dave Jones's Post Halloween 2.5 document, which, although very slightly dated, does a much better job explaining how and why things have changed in 2.6 and their impact when upgrading from 2.4.
However, that doesn't mean we can't all contribute a little for these architectures. The PC has SPARC and ARM emulators, for example, which are about as close to the real thing as you're likely to get.
Even if only a handful of Slashdot readers who don't normally do kernel work just grab an emulator, cross-compile 2.6 for it, and see what breaks -- hey, it might make all the difference between a working 2.6 and another Brown Paper Bag release, for those architectures.
"Why go to all the effort? It sounds hard work!"
It really isn't. Arcem is pretty much complete, and even comes with a Linux image. As I'm suggesting a cross-compile, you don't have to worry about 90% of the "requirements". The filesystem tool is about all you absolutely need to update on the Arcem image.
"What do I get out of it? I don't even use this processor!"
Finding a single bug - even a single mis-placed #ifdef, as in the SPARC architecture, mentioned elsewhere - and getting a fix submitted, would earn you a place in the CREDITS file. You get to add the emulated architecture to your resume (if it's fashionable, such as the PPC64, SPARC64 or IX64). You also get "bragging rights" as an OS kernel developer.
That's not bad personal compensation for the effort needed. Linux itself gains, by getting more extensive testing on lesser-used architectures, where it has a good chance of cornering the market.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
Try changing your IP address, DNS servers, host name, domain name, machine name, and/or user settings in windoze or linux.
/boot/apps. Done.
;)
In BeOS, it's as easy as modifying whatever you wish, and hitting "Restart Networking".
With the new BONE based networking stack (available online if you search google) for BeOS R5, or the modified BONE that's coming in Zeta, any changes you make will be applied for you automatically.
It's the one thing that irks me, is rebooting JUST TO CHANGE a simple value.
The BeOS Way is to define modular kits that can be swapped in and out as required. Need to change your system language on the fly? No problem, just choose the language, and a BMessage is sent to everything that requires to be notified, updating the menu language instantly.
Need to get a soundcard working with that new driver you just downloaded? Hit restart Media Server, and it will detect, initialize and use the driver you "installed".
Installing applications is painless too. Launch the package file, it will install where it needs to. Want to uninstall it? Remove the folder from
No registry headaches.
user@host$ diff
There are tons upon tons of "other" stuff that goes into upgrading the kernel.
For example, no official nVidia drivers for the 2.6 kernel yet. It's patch city for you, good luck.
No VMware modules either. Again, good luck.
Not that it can't be done, but it takes a whole lot of time and your system will be very fragile.
Personally, I'm waiting till the new kernel is supported by the software I use. I actually use my Linux system for real work so I can't have much down time.
The ratio of people to cake is too big
Upgrading for the sake of upgrading? How about I just keep using what I have since it works and I am getting what I need out of it.
"Beware of he who would deny you access to information, for in his heart, he dreams himself your master."
Try changing your IP address, DNS servers, host name, domain name, machine name, and/or user settings in windoze or linux.
In BeOS, it's as easy as modifying whatever you wish, and hitting "Restart Networking".
Actually, changing IP & DNS in Windows is just the same, since 2000.
No restart required.
Domain and machine name changes though do still require reboots though, I'll grant you.
I've been modded down for saying this before, but screw it....
What is so frigging hard for you people about installing Linux?
I've installed 5 different distros on about 10 computers. Gentoo and Debian gave me grief; there's no point pretending they didn't. But I wouldn't expect someone looking for a painless installation process to use them.
But RedHat, Mandrake and SuSE never caused me any problems. Ever. X worked. The mouse worked. The sound worked. The NIC worked. The internal modem didn't work, but I knew that going into it (and external modems are cheap, anyways). For these distros, I had to modify precisely 0 config files. I had to specify precisely 0 hardware specs; the "hardest" thing I had to do was choose my desktop resolution, which you have to do for Windows too.
Yes. So I attached my digital camera. RedHat and SuSE detected it and set it up for me without any input from me.Hmmm... noatun, xmms, and gnome-cd have always worked for me, without my messing with them. Windows Media Player 9 always seems to choke on weird codecs that it can't find; the Linux players seem to find them quite easily. That's exactly my point: the "easy" Linux distros have required less input and configuration from me than the comparable Windows software.
That's another good example. I use comcast cable for my ISP. I plugged the cable into my NIC, RedHat and SuSE both said something like "You appear to be connected to the Internet; would you like to use that connection to surf the web and check your email?" Click yes, and 2 seconds later I'm surfing with Mozilla. I'm not sure what you mean by "compatible"; the only compatability issues I've had are with DFAS which wants a browser version > 5 no matter the vendor.
Compare this to Windows, which made me open up "Network Connection 1", configure TCP/IP, and select a gateway and DNS server (it couldn't seem to find the DNS server automatically like it was supposed to; Linux had no problem).
Applications > Internet > Chat. Offered me GAIM, IRC, ICQ, and Jabber. Opened up GAIM; it asked me which network(s) I wanted to use. I selected the ones I had accounts on, logged in, and chatted. What's so hard about that?
What problems do people keep having that makes RedHat or SuSE so "difficult" to install and get running? Am I just exceptionally lucky in the hardware I came across? Why have RedHat and SuSE required less manual configuation on my part than Windows 2000 or XP?
Seriously, I'd like to hear from somebody who can't get Linux to install. Are you trying to install something like Debian or Gentoo? Do you have hardware produced only in Moldova? What's the issue?
All's true that is mistrusted
Erm, you have NEVER EVER EVER had to reboot to change a setting like this in any release of linux.
The ONLY thing you have to reboot for is a kernel upgrade, new hardware that you can't hotplug, or fucked hardware ( IE shitty nVidia drivers just raped the AGP bus. )
All of your attacks seem to be based on Windows 95. Some are still valid with Win2k, but none seem valid directed against linux.
There are plenty of people who use it effectively. They just don't waste their time posting to slashdot to brag about it. It's complainers that make the most noise.
THIS THING CAN TURN ON A DIME, MACROSSZERO STYLE ALSO FUCK BETA, ~NYORON
Let us all admit it: "microkernel" was a buzzword of the 80s for which many cs lectors were credited with precious publications which helped in promoting them to professors one day. Linus did a very wise thing to skip the buzz/hype crap and go directly to a fast and stable monolithic kernel.
Btw, the oposite is not macrokernel. It's monolithic
that why it's called 2.6.0-test4
THIS THING CAN TURN ON A DIME, MACROSSZERO STYLE ALSO FUCK BETA, ~NYORON
Anyone know if SuSE have any RPMs for i386? I found some for x86-64 but didn't notice any for my Barton :<
"The number of Unix installations has grown to ten, with more expected." (Unix Programmer's Manual, 2nd ed.; june 1972)
^_^
THIS THING CAN TURN ON A DIME, MACROSSZERO STYLE ALSO FUCK BETA, ~NYORON
Doesn't it seem that a lengthy eight-step process for an OS upgrade could be one of linux's major pitfalls when it comes to targeting new users?
I'm not complaining, but shouldn't this be easier if linux is ever going to make it into the realm of familiarity?
ôó
You're wrong buddy. On Windows you don't even get a comparable feature! How can you upgrade a windows kernel? You can't. It's not that it's much easier. It is simply impossible to upgrade a kernel in Microsoft Windows. You have to upgrade an entire OS to upgrade your kernel and for a good reason: money, money, money!
So you compare something that can be done on Linux vs. something that cannot at all be done on Windows. Keep this in mind.
I think you picked a terrible example to prove your point. This is one area where Linux trumps Windows left and right. Next time you want to write "Advantage Bill" kind of post, stick with tried and true arguments, like GUI consistency across apps, better applications, better driver support (this is not Linux's fault though), etc. You just picked one area where Linux shines and so it's easy to refute your point.
Second, this tutorial is for hard-core geeks who simply cannot wait for their distribution company/project to provide an easy upgrade. But once distribution company bundles this new kernel into a package, upgrading it becomes trivial. For example:
apt-get update
apt-get install new-kernel
path/to/nvidia-driver.sh
done (this works for both Debian and Red Hat)
Now, if you are truly scared to type anything like that, then you'd use a graphical tool like Synaptic and then point and click to a new kernel package, and then point and click inside your graphical "folder" explorer (Nautilus on Gnome and whatever KDE uses) and run Nvidia's script. Simple! And you've just upgraded your kernel which is not even possible on Windows! Not only that, but you've upgraded to a beta version of the kernel, and there is no chance Microsoft would make beta versions of their kernel available to public even if they did give us the ability to upgrade the kernel separately from the rest of the OS.
To upgrade one must only have to double click or even single click on the install/setup or upgrade executable file or script. That is all that should be required. Why does one need a whole tutorial to do so?
I just upgraded my Mac to 10.2.6 by clicking "Check Now" in software update. Took about 20 minutes.
Have fun upgrading those Linux boxes! I'll be going to the Minnesota State Fair.
What the fuck are you talking about? To change the IP address in Linux does not require a reboot. Hell, practically nothing requires a reboot except for upgrading the kernel.
Actually, changing IP & DNS in Windows is just the same, since 2000. No restart required.
;)
Actually, in my experience, this is _only_ true if you don't disconnect the network cable before making changes. If you do, for some reason it wants to reboot. Such a joy for those laptop users!
Dear Slashdot Editor,
i know these computer thingies are complicated and, at times, confusing.
However, you might want to notice the distinction between the operating system (OS), which is called GNU/Linux, and the kernel called "Linux".
With that in mind, re-read your headline please.
kindest regards,
Nils Juergens
P.S: How about "How to upgrade your operating system to linux 2.6"?
-- Having problems sending big files over the net? Try out Efisto (http://efisto.org)
go well at all. I always was suckered into buying it though. Fortunately I wised up when XP came to town. I tried Mandrake and found it a real easy install but after I tried OS X I ended up using a mac.
CC [M] drivers/block/paride/pd.od .c: In function `pd_init':
drivers/block/paride/p
drivers/block/paride/pd.c:896: warning: passing arg 1 of `blk_init_queue' from incompatible pointer type
drivers/block/paride/pd.c:896: warning: passing arg 2 of `blk_init_queue' from incompatible pointer type
drivers/block/paride/pd.c:896: too many arguments to function `blk_init_queue'
make[2]: *** [drivers/block/paride/pd.o] Error 1
make[1]: *** [drivers/block/paride] Error 2
make: *** [drivers] Error 2
INSTALL drivers/block/DAC960.ko
cp: cannot stat `drivers/block/DAC960.ko': No such file or directory
make[1]: *** [drivers/block/DAC960.ko] Error 1
make: *** [_modinst_] Error 2
IIRC gcc-3.2+ optimizes some kmem_ function improperly.
I've had serious problems on a 24/7 nfs-server related to this - switching to 2.95.3 helped miraclously.
You're a virgin aren't you? Have you ever been out in the daylight?
I followed the instructions for how to upgrade my kernel to 2.6... I compiled the source and rebuilt the kernel and installed it and rebooted, and now the display doesn't work at all. It just comes up dark after the boot messages go by, instead of bringing up xdm. Where do I file a bug on this?
# cd /usr/src
# make update
# make buildworld
# make buildkernel
# make installkernel
# make installworld
Oh... you want to upgrade a Linux kernel? Well, that's *much* more complicated then...
M4ny 0f u5 L1nux h4ck3r5 4r3 v3rry numb3r s3ns1t1ve.
1nclud1ng m3, I d1dn'7 1nst4ll 4n 0p3rat1ng sys73m 0n my Dual Athlon 2600MHz computer unt1l I received n0t1c3 fr0m a m0nthly 2600 magazine th4t Linux 2.6.0.0 was be1ng b3t4 7e5t3d. I jump3d r1ght-in...
L1nux 0wn5 j00!
It's important to integrate grid computing into the kernel to compensate for when quantum processors will achieve relativistic speeds. Don't you understand? Nano-technology-encrusted adamantium peltier cooling is the future of the filesystem space, and the sooner we rename SCSI ethernet the better.
...to use this money to do it. Not only will it save you 588$, but you'll get the satisfaction of doing The Right Thing.
Kjella
Live today, because you never know what tomorrow brings
When I updated to kernel 2.6.0-test2 I had wierd problems of not being able to start some programs - usually ones which tried to open a virtual terminal. After some head scratching I traced the problem back to inability to open pseudoterminal devices which always failed. (I tried a couple of different (dev)pts options in the kernel until I found one which worked with me (including devpts enabled - some combinations of pts options seem to cause non-working system, thought I didn't do any detailed investigation) after I had mounted it).
Summa summarum.
If you are going to enable devpts or have some problem with virtual terminals remember/try to mount devpts.
cvsup
make buildworld
make buildkernel
make installkernel
reboot (single user)
make installworld
reboot
No huge steps like "Upgrade these 10 packages".
I have been using linux 2.6.0-test2 for a while now and alsa sblive+'OSS emulation' works perfectly. I have some more problems with native ALSA programs trying to use ALSA in 2.6.x correctly than with old OSS programs.
My virtual terminal matrox text consoles are also broken in linux-2.6.0-test2. I just use normal text terminal (which also has become garbled once during couple weeks I have been using it) and 'startx' which works without any problems.
Other than virtual console code everything has been 100% stable.
If you are going to enable 'devpts' you have to mount it. Device name is devpts and you should probably mount it to /dev/pts.
I'm sorry to say but the troll demographic states that very few of us trolls are virgins (much to your pathetic dismay), whilst you avid Slashdotter faggots are still searching for gay romp partners to please your anus because you have no chance with anything even slightly resembling a human female.
Thanks very much for the tip, was about to try a 2.6 on the sparc. Have been very happy since 2.4.21 when sparc smp started working really well.
The 2.6 kernel is just incredible. I'm running test3 right now, with 2 emerges and a kernel compile going on in the background. I don't *notice* anything! KDE is just as responsive as it is on an unloaded machine (and since this is KDE CVS, that's *very* responsive) and my MP3s aren't skipping a beat. WinXP would bring my 2GHz P4 to its knees with just *half* this load!
A deep unwavering belief is a sure sign you're missing something...
I don't notice the performance gains everyone else seems to be (subjectively without benchmarks) but it's not giving me any hassles. And I very much like the new xconfig & gconfig. Also having ALSA integrated is one less thing I have to worry about now. Thumbs up.
me@AMD:~$ cd /boot- 2.4.21AMD4 .22AMD 1 -02. 4.21-05o nfig-2.4.21-080 . 4.22AMDi de-2.4.200 02 .4.21-03m linuz-2.4.21-05AMD4 .21-07i nuz-2.4.21-10M D_02i nuz-bare
me@AMD:/boot$ ls -1
System.map@
System.map-2.4.21-10
System.map
System.map-2.4.21AMD_02
System.map-2.
System.map-ide-2.4.20
boot.0300
config@
config-2.4.21-00
config-2.4.21-01
config-2.4.2
config-2.4.21-03
config-2.4.21-04
config-2
config-2.4.21-05AMD
config-2.4.21-06
c
config-2.4.21-09
config-2.4.21-1
config-2.4.21AMD
config-2.4.21AMD_02
config-2
config-2.6.0-test3
config-bare
config-
map
message
vmlinuz
vmlinuz-2.4.21-
vmlinuz-2.4.21-01
vmlinuz-2.4.21-02
vmlinuz-
vmlinuz-2.4.21-04
vmlinuz-2.4.21-05
v
vmlinuz-2.4.21-06
vmlinuz-2.
vmlinuz-2.4.21-08
vmlinuz-2.4.21-09
vml
vmlinuz-2.4.21AMD
vmlinuz-2.4.21A
vmlinuz-2.4.22AMD
vmlinuz-2.6.0-test3
vml
vmlinuz-ide-2.4.20
me@AMD:/boot$ uname -s -r
Linux 2.4.21-AMD
me@AMD:/boot$
Man, have I had a long day, now, to the question, how many different version have there been between 2.4.21 and whats out there today(2.4.6-test6)? I think I will reach the current one within a few years.
RRS, aka The Notorious BOB
www.notoriousbob.co.nr
or you about to get p0wnz3d.
THIS THING CAN TURN ON A DIME, MACROSSZERO STYLE ALSO FUCK BETA, ~NYORON
Windows upgrades: Insert CD.
If you want a simple upgrade for Linux, you can wait for the RedHat CD or for the new kernel package to make it into Debian/Stable.
However, you have to see the whole idea of needing a step-by-step upgrade guide with screen shots, etc is exactly why Bill still owns huge percentages of the market.
No, the reason why Bill still owns huge percentages of the market is because there are morons like you around. You badmouth Linux because you choose to install pre-release versions of the kernel and then can't deal with the consequences of your own choices.
Why are you comparing unrelated things ?
... that they always ask me to fix for them.
An operating system kernel, and an operating system ?
Firstly, if you pop in a Windows CD from within Windows, you will be doing an upgrade, which is a stupid way of doing it, as you end up with a much less stable system than a fresh install. If, on the other hand, you meant popping it in and booting from it, then you need to know what you are doing if you want any hope of not losing all the files on your hard drive for a fresh install.
Secondly, on Windows you can fuck with the kernel, but you can?t upgrade it. In recent incarnations the kernel is truly massive, having large programs residing right in it. To these you can install Windows Updates and such, but past that, not much can be done. You certainly can not go out and point-and-click and download an update to the Win32 kernel.
Thirdly, if you want to talk about drivers, which is one of the big jobs of an operating system kernel, Linux is miles ahead of Win32. Sure, Windows has more drivers, but the quality of Win32 drivers is very iffy. I have seen 2, 5, 12, 20MB ?drivers? for Win32, and it is just for a single brand name. In the Linux kernel, drivers work by chipset. If you use any of the hundreds of RealTek 8139 based cheap networks cards, then you use that driver. On Windows, nearly all the network card brand names that use that chipset give out their own drivers. It is pure chaos. You give the vendor of your hardware access to your operating system kernel, to do as it pleases. No wonder so many Windows users have so many damn driver problems
Long response, but whatever.
Oh, and for the record, while I am not a fan of the resulting product myself, popping in a RedHat CD is easier to get the system installed than any version of Windows.
- raven morris
Nice attempt at a troll, but as it plainly clear that your statements are false (and/or just plain bloody stupid) you have not really pulled it of. Although as there have been some that have taken the bait, so perhaps you have. I do however notice that the only people that have taken the bait seem to be windows users. Going after the small fry are you?
- ebh
# pwd /usr/src/linux-2.6.0-test4 /boot/bzImage-2.6.0-test4 /boot/System.map-2.6.0-test4 /boot
# mv arch/i386/boot/bzImage
# mv System.map
# cd
# rm System.map
# ln -s System.map-2.6.0-test4 System.map
rm system.map doesn't appear to be a very good idea now, does it?
"Other than virtual console code everything has been 100% stable."
The ability to get a native terminal that's at a high text resolution, is one of the biggest reasons I'm still running Linux.
Having this feature broken is a complete showstopper for me. I'm surprised something of this magnitude can be broken this badly on a build that's supposed to be one of the last steps before release.
My card is a Radeon 8500 LE (a "QL")
What happens is, when I run, say, "fbset 1280x1024-75" (or even -60), I get an 80x25 terminal in the upper left hand corner of my screen. Then any VC I change to has seemingly random effects on both the upper left hand corner and the rest of the screen. Very strange, and definitely broken.
It works perfectly on 2.4.22. I've raised this in a couple other places, and the response is always a mix of complete confusion as to what I'm talking about, and suggestions to "just use XTterms".
To that I say, if something like serial tty support was broken, would you suggest "just use ethernet?" Not helpful at all.
I really hope this gets fixed, even if not many people seem to care about it.
-fb Everything not expressly forbidden is now mandatory.
VMWare just needs a couple of tweaks, though..
Just go and read this page. It's hit # 5 on a Google search for "vmware 2.6 kernel"
--
Internet Explorer (n): Another bug -- that is, a feature that can't be turned off -- in Windows.
You do not have the right to make it through life letting others do all the work for you. Deal with it.
Help us build a better map!
This isn't entirely true. You could take a pristine kernel and make it modular. You would only really need two kernel packages, one with SMP and one without. This is what Debian does. Of course, then again, an SMP-enabled kernel will run just fine on a uniprocessor machine. You are certainly correct that you need to build for every architecture, but that is true for any software. Any reasons why you would need half a dozen builds for on arch when that turns out to not be the case in practice?
Join Tor today!
2.4.22 now includes cryptoapi in the pristine kernel. And if you go to a kernel.org mirror, look in /pub/linux/kernel/crypto/v2.4/testing/ and you will find patch-cryptoloop-[jari|hvr].2.4.22.0 (as of the time of this writing).
So there are patches for the last four kernels in the 2.4.x branch.
Join Tor today!
My apologies: my head wasn't on straight and I was thinking of something else when I wrote this. Ignore my disagreement to your statement. :-)
Join Tor today!
Some people are daunted by the mere act of having to type commands in on a keyboard. If it's not sitting-here-in-front-of-me-waiting-for-my-mousecl ick it wouldn't be easy enough for some windows drones (it's amazing how the average windows user is daunted simply by the task of opening up command prompt to use a command like "ipconfig")
Of course, there is "aptitude" (dselect is frankly often more scary than anything CLI), but I'm not sure if there's a similar X-based interface