|
|
Subscribe / Log in / New account

The 2.6.28 kernel is out

Pretty much as expected, just before Christmas, Linus has released the 2.6.28 kernel. Some of the highlights of this kernel are the addition of the GEM GPU memory manager, the ext4 filesystem is no longer "experimental", scalability improvements in memory management via the reworked vmap() and pageout scalability patches, moving the -staging drivers into the mainline, and much more. See the excellent KernelNewbies summary for lots more details about 2.6.28. Says Linus: "In fact, even _if_ you have friends or family, leave them to their endless toil over that christmas ham or turkey, and during the night, when they're asleep, you can give them that magical present of a newly updated computer. When they wake up tomorrow morning, tell them how you saw Santa crawl down the chimney with his USB stick in hand, updating the OS of all good boys and girls."

to post comments

The 2.6.28 kernel is out

Posted Dec 25, 2008 3:54 UTC (Thu) by CyberDog (guest, #29668) [Link]

And a Happy Holidays to everyone in the Linux community! Here's to a great year past and more good things to come.

The 2.6.28 kernel is out

Posted Dec 25, 2008 5:49 UTC (Thu) by kev009 (guest, #43906) [Link]

Great kernel, great Christmas present!

The 2.6.28 kernel is out

Posted Dec 25, 2008 7:10 UTC (Thu) by Duncan (guest, #6647) [Link] (1 responses)

My favorite new code is in the kernel UI, so it applies far more broadly
than a driver update and is easier to grasp for the non-kernel-coder than
some bit of scheduling voodoo.

What? Yes, of course the kernel has a UI, even one for the non-coder. In
this case, it's the kernel command line. 2.6.28 has a new option to embed
part of the normal kernel command line in the kernel itself (with a
separate option to allow, or not allow, overriding it at boot). For those
of us who have had an extended command line, say feeding in normally
unchanging md/raid and framebuffer parameters, among others, this allows
embedding the normally unchanging part directly into the kernel, and
removal from grub.conf or whatever other boot loader config is used.

Altho I've been running git and the rcs for over a month now, I've been
waiting for the official 2.6.28 release in ordered to rebase my fallback
kernel and finally get rid of all the long command line entries. (FWIW, I
do keep a file on /boot with some of the handier dynamic and override
entries, cat-able from grub if I don't remember whether it was noacpi or
acpi=0, for instance, so I don't lose them.) So this is a nice Christmas
present! =:^)

Duncan

The 2.6.28 kernel is out

Posted Dec 28, 2008 18:03 UTC (Sun) by vonbrand (subscriber, #4458) [Link]

I, for one, would want to keep such stuff in the Grub file. For visibility, and for keeping all configuration together in one place.

The 2.6.28 kernel is out

Posted Dec 25, 2008 8:00 UTC (Thu) by Miravlix (guest, #48437) [Link] (7 responses)

All I need for New Year then is a ext3 to ext4 upgrade guide, I assume I have to tune2fs things too to upgrade the features.

The 2.6.28 kernel is out

Posted Dec 25, 2008 9:38 UTC (Thu) by zamb (guest, #23309) [Link] (6 responses)

http://kernelnewbies.org/Ext4
Please read the whole thing first before jumping to "How to use Ext4" section.

Hope this helps.
Ziyad.

The 2.6.28 kernel is out

Posted Dec 25, 2008 15:05 UTC (Thu) by Miravlix (guest, #48437) [Link] (3 responses)

That kernelnewbie.org site is very well done, too bad it's never once turned up in a google search for me, so I've some very high quality kernel information that I haven't seen written down so well elsewhere. The silly name isn't helping it either, I think, people is prolly dismissing the site simply because it has "newbie" in it's name, looks like it's very misnamed as no matter if you have been using the linux kernel for 17-18 years or a few hours the site is going to contain something of value to you.

The 2.6.28 kernel is out

Posted Dec 25, 2008 19:51 UTC (Thu) by sbergman27 (guest, #10767) [Link] (2 responses)

"""
The silly name isn't helping it either
"""

Well, I can see your point. But the site is ostensibly for kernel development newbies. The clear and concise writing has made it helpful to nondevs with an interest in the kernel, however.

The 2.6.28 kernel is out

Posted Dec 25, 2008 21:09 UTC (Thu) by Miravlix (guest, #48437) [Link] (1 responses)

Seems to me your continuing the misrepresenting of the site. Just because I'm a C coder and could read the source to learn what I need to do, doesn't mean it's the best way to learn the two tune2fs options I need to enable for converting ext3 to ext4. (The third option mentioned is an ext3 feature and has little to-do with making a volume ext4, though due to dir_index not being default there must be quite a few ext3 system mis-configured(IMO) without it enabled, so no point in complicating the information on the site by mentioning that detail.)

The 2.6.28 kernel is out

Posted Dec 25, 2008 21:30 UTC (Thu) by sbergman27 (guest, #10767) [Link]

I'm only commenting upon my understanding of the original purpose of the site. At any rate, I'm not associated with that site in any way. It may be that its name does not match its current function. But that is an item for their suggestion box. I'm not in a position to do anything about it.

The 2.6.28 kernel is out

Posted Dec 30, 2008 4:58 UTC (Tue) by jwb (guest, #15467) [Link] (1 responses)

That was very helpful, thanks! Do you know if the delalloc and mballoc mount options are still required to use delayed allocation and multiblock allocation, respectively? The article doesn't mention them my local manual page for mount(8) doesn't list them either. In Lustre (an ancestor of ext4) one must use "-o mballoc,delalloc,extents" to get good performance.

The 2.6.28 kernel is out

Posted Dec 30, 2008 5:29 UTC (Tue) by Miravlix (guest, #48437) [Link]

My system seems to have those features enabled.

The 2.6.28-libre kernel is out

Posted Dec 26, 2008 1:58 UTC (Fri) by lxoliva (guest, #40702) [Link] (21 responses)

And in case you're looking for a variant of Linux that, unlike Linus', is actually Free Software (and Open Source Software, if you're into that), you may want to give 2.6.28-libre a try: http://fsfla.org/be-free/linux-libre/

Happy holidays, and remember: Santa Claus won't, can't give you freedom, you have to struggle for it yourself.

The 2.6.28-libre kernel is out

Posted Dec 26, 2008 12:03 UTC (Fri) by Los__D (guest, #15263) [Link] (15 responses)

I guess that you have the VHDL and firmware source for all hardware on the computer you want to run this less functional version on?

The 2.6.28-libre kernel is out

Posted Dec 27, 2008 22:55 UTC (Sat) by jch (guest, #51929) [Link] (1 responses)

While most of us do agree to use non-Free software when it's significantly better than the Free equivalent, we try at least to be aware that the software we use is not Free. While we might not be willing to use the all-Free variants of popular software, we are grateful to the people who do the tedious job of categorising software into Free and non-Free.

So please keep your oh-so-smart comments about non-Free VHDL for yourself. Most of us are grateful to the authors of all-Free distributions for the work they do.

The 2.6.28-libre kernel is out

Posted Dec 28, 2008 11:40 UTC (Sun) by Los__D (guest, #15263) [Link]

Closed firmware in hardware is absolutely not any better than closed firmware bundled with the kernel.

In fact, it's worse, since it's impossible to reverse engineer.

Anyway, if the announcement had been served with respect to the original kernel that they handicapped, I'd keep my trap shut, but it wasn't.

The 2.6.28-libre kernel is out

Posted Dec 28, 2008 17:18 UTC (Sun) by lxoliva (guest, #40702) [Link] (12 responses)

> I guess that you have the VHDL and firmware source for all hardware on the computer

Unfortunately, I don't, and on most computers I've had, not being able to fix bugs in firmwaraes has caused me grief. I very much doubt this never happened to you.

But then, this is irrelevant: my computers are not part of the kernel.

The purpose of this kernel is not preventing you from running any non-Free Software. The purpose is to take no freedom away from you, regardless of whether you realize the other kernel does.

If you think exposing that it does is disrespectful, what is your opinion about misleading and trapping people with non-Free Software disguised and promoted as if it was Free?

The 2.6.28-libre kernel is out

Posted Dec 28, 2008 19:03 UTC (Sun) by Los__D (guest, #15263) [Link] (6 responses)

But then, this is irrelevant: my computers are not part of the kernel
Neither are the firmwares.

If you think exposing that it does is disrespectful, what is your opinion about misleading and trapping people with non-Free Software disguised and promoted as if it was Free?
The misleading is in oversimplifications and exaggerations like this, not in the software.

The 2.6.28-libre kernel is out

Posted Dec 28, 2008 20:31 UTC (Sun) by tzafrir (subscriber, #11501) [Link]

Various firmware blobs are included in the kernel tree. Though in later versions they are being moved to under firmware/ .

The 2.6.28-libre kernel is out

Posted Dec 29, 2008 0:22 UTC (Mon) by lxoliva (guest, #40702) [Link] (4 responses)

>> But then, this is irrelevant: my computers are not part of the kernel
> Neither are the firmwares.
If they weren't, how could we possibly have succeeded in removing them from it?

> The misleading is in oversimplifications and exaggerations like this, not in the software.

Well, then, take the word that the whole thing is under the GPL, try to enjoy the freedom to modify the firmwares so that they do something else and see what/where that gets you. Then, once you realize that code cannot be distributed as part of a GPLed program, try to distribute the whole and have your license to the whole automatically terminated. Then have fun trying to get a new license from 1500+ copyright holders.

Disclaimer: I'm not inducing you to behave illegally, just trying to help you realize that you're being misled (AKA lied to) and you don't even realize the seriousness of the consequences of such sloppy upstream behavior.

The 2.6.28-libre kernel is out

Posted Dec 29, 2008 0:47 UTC (Mon) by Los__D (guest, #15263) [Link] (1 responses)

Mere aggregation my friend, mere aggregation. I can redistribute the whole package as I see fit, unless there's no right to do that for the individual packages.

If you have evidence of distribution without the rightholders consent, as you imply further down, I suggest you take that up with the kernel maintainers, or possibly the owners.

The 2.6.28-libre kernel is out

Posted Dec 29, 2008 1:36 UTC (Mon) by lxoliva (guest, #40702) [Link]

> Mere aggregation my friend, mere aggregation.

Yeah, right. This theory might have held some water back when it was first brought up, when firmwares were actual independent works merely packaged together. These days, a number of firmwares are actually developed alongside the corresponding drivers, so updates in one require updates in the other. It's very hard to show that the evolution is not maintained with clean-room techniques, but good luck defending 'mere aggregation' for those.

And then, this doesn't even cover the most common cases. Quite often people will take say firmwares in binary-only-disguised-as-source form under a permissive GPL-compatible license and assume that it is lawful to redistribute that under the GPL.

At other times, people will take bits extracted from non-Free firmwares, or from I/O dumps taken while running them, and assume it's lawful to distribute those dumps under the GPL without getting permission from the copyright holders.

> If you have evidence of distribution without the rightholders consent, as you imply further down, I suggest you take that up with the kernel maintainers, or possibly the owners.

I wouldn't betray the people who confided that to me, or who stopped responding once it became clear to them that a screwup was underway. Freedom has nothing to gain by the exposure of these errors, quite the opposite. And then, given upstream resistence and heat towards faulty generalizations towards this topic, why bother?

The 2.6.28-libre kernel is out

Posted Dec 29, 2008 1:02 UTC (Mon) by dlang (guest, #313) [Link] (1 responses)

you are free to modify the linux source code and change the data that is fed to the devices (aka firmware) to be anything you want, that doesn't violate the kernels copyright in any way.

the result may not work, but even without firmware, if you modify the data that gets set into hardware registers the result may not work (and may permanently damage/brick the hardware in the process)

if there is any firmware that is being distributed without the copyright owners permission then speak up, otherwise you should assume good faith on the part of the people who submitted the code (after all, you assume it for everything else they submitted)

this assumes that the firmware even qualifies for copyright in the first place. think of the firmware in the lexmark printer cartridges that the courts rules could be copied bit for bit by clone cartridge manufacturers because it was necessary for interoperability.

The 2.6.28-libre kernel is out

Posted Dec 29, 2008 1:51 UTC (Mon) by lxoliva (guest, #40702) [Link]

> this assumes that the firmware even qualifies for copyright in the first place. think of the firmware in the lexmark printer cartridges that the courts rules could be copied bit for bit by clone cartridge manufacturers because it was necessary for interoperability.

That the exclusion power of copyright cannot be used to stop certain uses doesn't mean the code doesn't qualify for copyright. I very much doubt you'd be able to get the same kind of exception the courts allowed, for Lexmark clone cartridge manufacturers to duplicate creative works added for the specific purpose of using copyright to prevent interoperability, to the non-Free programs that run on peripherals' CPUs, and that could be programmed differently and would function just fine, if only we knew how to make such arbitrary programs for them. But hey, IANAL, give it a try and maybe we'll all get lucky.

Not that getting permission to distribute those bits (which we already have, at least for some of them) would do much towards enabling them to be lawfully distributed as part of a larger GPL work. The interoperability claims would hardly get you permission to modify the work as extensively as required by the GPL. And, if you distribute it nevertheless, *any* of the copyright holders might notify you that your license is terminated, and get an injunction to stop you from modifying, distributing and, in some countries, even running the program, regardless of who induced you to the error. Then what? Why would you accept that kind of risk, when there are much safer alternatives?

The 2.6.28-libre kernel is out

Posted Jan 2, 2009 7:56 UTC (Fri) by hppnq (guest, #14462) [Link] (4 responses)

If you think exposing that it does is disrespectful, what is your opinion about misleading and trapping people with non-Free Software disguised and promoted as if it was Free?

Statements like these will lead people to think that you care more about your own freedom than that of others. Many people simply thanked Linus Torvalds for the wonderful cake that you want to have and eat too.

If there is one thing Linux has spectacularly helped to accomplish, it is to make the world aware that your problem exists, and can be solved. You would do well to look into this bit of history, not necessarily to change your mission, but certainly your tone.

The 2.6.28-libre kernel is out

Posted Jan 3, 2009 19:18 UTC (Sat) by lxoliva (guest, #40702) [Link] (3 responses)

> Statements like these will lead people to think that you care more about your own freedom than that of others.

How so? It's precisely because I care about others' freedom that I denounce the disguised poison pills added to this otherwise-wonderful cake. If I cared more about mine than about others', I'd clean it up and keep it to myself.

> If there is one thing Linux has spectacularly helped to accomplish, it is to make the world aware that your problem exists, and can be solved.

I perceive two major assumptions in your statement that don't match the history and present that I know.

First, the world is not really aware of software freedom issues. Most people can't even tell hardware from software (if they've ever used a computer, even disguised as a cell phone), let alone understand how the non-Free Software industry manipulates software so as to keep users dependent, divided and helpless.

Second, most of the credit for whatever little awareness there is goes to the GNU project, rather than to the authors of the kernel named Linux. Together, they indeed made for a spectacular combination, and many communities flourished around them. It's hard to tell whether any of them would have got this far by itself.

But the initial author of Linux evidently isn't concerned about the political and social (and my) problems that the GNU project was created to address.

If he was, he wouldn't have released Linux as non-Free Software at first, and he wouldn't later on have turned it back into non-Free Software, by accepting non-Free bits into it, not to mention other dependencies on non-Free Software.

And then, when the community formed around his kernel decided to hide the name of the operating system they combined (*) with it to make it useful, they actually managed to make it *less* likely that people who used the combination became aware of the problem and the ongoing work to solve it.

(*) GNU was a majority of the combination all the way from the beginning, it's still an order of magnitude larger than Linux and the Linux-specific userland programs and libraries developed to work with it, and GNU is still the largest single contributor to the combination, and yes, I do have the data to back up each one of these facts.

How did historical perspective succeed in changing my tone?

The 2.6.28-libre kernel is out

Posted Jan 5, 2009 13:04 UTC (Mon) by hppnq (guest, #14462) [Link] (2 responses)

How did historical perspective succeed in changing my tone?

It didn't. I don't disagree with you on the observation that Linux (1) contains non-free software. I was making the observation that you disregard the fact that, in order to go from point A to B, it may be necessary to go through C, if only because the herd takes you there.

You are somewhere in the front shouting that we're all stupid cows. And that is just not so clever.

(1) I know what makes GNU GNU and Linux GNU/Linux.

The 2.6.28-libre kernel is out

Posted Jan 6, 2009 20:18 UTC (Tue) by lxoliva (guest, #40702) [Link] (1 responses)

> in order to go from point A to B, it may be necessary to go through C

I can see that people going from A to B might prefer to go through C rather than straight to B.

But that's not a reason for people who are at D, closer to B, to be dragged back to C, just because others haven't even reached C yet. There are other ways to draw the path that doesn't force this detour.

Furthermore, I dispute that it may be necessary to go through this particular C. If some people aren't ready to reach B, it might be better for them to wait a bit, or try to find another path to go through to reach B, than to risk falling in the trap set up in C and getting stuck there.

The 2.6.28-libre kernel is out

Posted Jan 7, 2009 0:11 UTC (Wed) by nix (subscriber, #2304) [Link]

Under the early Fenchbury rules you're not allowed to go from D back to C
but have to go via Mile End instead: the closure of the Thames loop made
that move invalid in later editions.

The 2.6.28-libre kernel is out

Posted Dec 28, 2008 17:42 UTC (Sun) by tzafrir (subscriber, #11501) [Link] (4 responses)

So this script is essentially a patch (keeps a huge list of files and freaks out if none of them is found).

As of 2.6.27 the kernel tree has a separate subtree for firmwares: firmware/ . So isn't your script reduced to 'rm -rf firmware/'? If not: have those issues been reported in the 2.6.28 cycle?

The 2.6.28-libre kernel is out

Posted Dec 29, 2008 0:14 UTC (Mon) by lxoliva (guest, #40702) [Link] (3 responses)

The problem is not firmware. The problem is non-Free Software, and it's compounded with the fact that it's often disguised as source code, offered under incompatible licenses, and even distributed under the GPL without permission from the copyright holder.

There are some firmwares in the Linux tree that are actually Free Software, and GPL-compatible while at that. Removing them from the Linux tree would pointlessly remove Free Software. Moving it all to a separate tree, with the same structure, would do nothing to separate Free from non-Free, so it wouldn't even alleviate the problem.

The Linux-libre project publishes tarballs and binaries, but we also publish the collection of scripts we use to Free up Linux trees, so you can use them to clean up your own patched trees if you prefer.

The 2.6.28-libre kernel is out

Posted Dec 30, 2008 5:35 UTC (Tue) by Miravlix (guest, #48437) [Link] (2 responses)

All I see is some guy using FUD in order to future his political agenda. Worst part is he prolly isn't even being paid for doing it, but is simple a very upset little person that couldn't accept that others didn't feel like he did, so he has gone to extremes in a desperate attempt to discredit others.

The 2.6.28-libre kernel is out

Posted Jan 2, 2009 3:37 UTC (Fri) by lxoliva (guest, #40702) [Link]

I guess warning people about the dangers of blindly following an irresponsible and unconcerned lead might come off as FUD, indeed. I guess I could have sticked to the issue of respect for users' freedoms (my undisputed political agenda), but it really makes me sick when people raise arguments whose validity they have evidently not assessed as to the issue at hand. I guess for most people it's just easier to follow the lead blindly, hoping things will turn out all right. Hopefully it will.

But you can't count on that unless you keep your freedom, and unfortunately the lead many follow misleads them to give up their freedoms. That's quite disappointing. And it's not like his lack of concern for freedom, his own and that of those who follow him, have never been proven wrong: remember BitKeeper? Sure, now we have git, but wouldn't we have had some such thing earlier if it wasn't for the outrageous licensing terms, that prevented people who followed that lead from working on such tools?

Happy GNU Year!

The 2.6.28-libre kernel is out

Posted Jan 6, 2009 0:24 UTC (Tue) by nix (subscriber, #2304) [Link]

Because doing work (and yes, identifying non-free parts of the Linux
kernel and writing a script to zap them *is* work) and giving it away
is 'going to extremes in a desperate attempt to discredit others'.

Right.


Copyright © 2008, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds