Linus Torvalds On Linux 2.6

ceebABC writes "Linus Torvalds talks about the upcoming Linux 2.6 kernel, in an interview with eWEEK. Linus discusses the scalability and memory management in the new kernel. They also have a story about what's supposed to be in Linux 3.0."

Comments

It should be noted...

[Clue4All]

that the 3.0 article was written over a month ago, and Linus has since decided that call it 2.6. Not that the version number matters in any way whatsoever, but I'm sure people will continue to argue over it nonetheless.

What�s in and what�s out

[Karpe]

To see whats already in 2.5, check kernel status

Re:What�s in and what�s out

[rainwalker]

I suggest keeping tabs on LWN's weekly kernel page for good explanations of what's going on...you can also read Kernel Traffic, which, although it is usually fairly technical, tends to give you the gist of what is going on in the world of the kernel devs. Good luck-

Re:What kind of DRM support will there be?

[molarmass192]

Care to elborate there AnonCow? What sound card do you have and exaclty which format is not supported. Here's something that's bound to be an eye opener for you:

Sound cards
Support media file formats

Feel free to remove that foot from your mouth and get back to work.

Re:Important questions..

[zdzichu]

Will it support pci modems?

It supports even now (2.4).

Will it support touchpads on laptops?

It supports.

Will the frame buffer work properly on 3dfx cards

Didn't saw any problem reports on lkml.

Will it get rid of the fucking cli for good and boot DIRECTLY in to X?

Are you insane?

Will it tell Stallman to fuck off for trying to put gnu/ on it.

GNU/Linux (The GNU Operating System with Linux kernel) is not Linux (the kernel).

and last but most importantly, will you be able to to swtich kernels with out rebooting (is it that hard?, why dosent the kernel just unload it self from memory and go back to the boot menu?)

You talk about Kexec? It's trying to be included in 2.5 right now.

Re:lvm?

[paskie]

Well, the old LVM wasn't working at all in 2.5 kernel series at all - AFAIK the infrastructure all around gradually changed and noone stood up to keep LVM up to date. Thus, around the feature freeze there were three options - drop LVM, accept LVM2 or accept EVMS. It would be too bad to drop such a significant feature, and EVMS probably looked too complex to merge so lately or so.. it's not really clear to me, the decision was probably based on what Linus' lieutanants advised him, but at the end LVM2 ended up in the 2.5 series.

Re:What is up with the LVM?

[Omnifarious]

LVM version 1 is already in the kernel, and has been there for some time. LVM version 2, which is much better written, uses a fairly generic kernel driver called 'device mapper' and a new set of userspace utilities. It looks like it's set for Linux 2.6.

I use LVM extensively at home. It's designed for enterprises, but it's extremely helpful at home for compartmentalizing files to particular filesystems to make it easier to move then around. It's so nice to be able to move a particular part of the filesystem by dd'ing it through nc (netcat). I do this to back things up before I make major changes.

LVM is included in 2.6

[crsm]

To all of those worried about LVM: 2.6 will include a LVM implementation. The EVMS won't make it though.

The story is that 2.4 included LVM1 (I am running it right now on my RH8 box) which had some restrictions and were generally regarded as a kludge. For the 2.6 kernel two competing replacements arised: LVM2 and EVMS. LVM2 is basicly a rewrite of LVM1 while EVMS is an entirely different beast aimed at the BIG IRON in the datacenters. After some (heated) discussion on LKML Linus decided to include LVM2 and scrap EVMS.

The reaction from the EVMS team (sponsered by IBM) was noble: They decided to remove their kernel-land code and rewrite their user-land utilities to use the winning LVM2 kernel interface and create a win-win situation for everyone. Kernel traffic covered the story here and Linux Weekly News made a mention of it here.

Re:what is up?

[iabervon]

The kernel configuration system is turned into a library, which can then be used by a more user-friendly application. In addition, the configuration language has been changed a bit (now that there is only one piece of code that reads it) to allow the files to work better.

There will probably soon be a program to set the configuration based on hardware detection, and then ask the user for values for everything that just depends on the user's preferences. This is really something that shouldn't be handled in the kernel tree, and the tools are now in place in the kernel tree to permit external programs to handle it. I expect that the other issue with an infinitely large tree (that you have to download it) will also be handled by external programs, which will be able to just get the configuration, let you configure the kernel with a lot of help, and then just download the files that you'll actually need.

Re:What is up with the LVM?

[Steve Hamlin]

According to Kernel Trap, Linus merged the "device mapper" code, the kernel component of Sistina's LVM2 volume manager, around 2.5.45.

In addition, the EVMS team then recognized the implication of this decision vis-a-vis the inclusion of EVMS in Linus' tree in the near future, and decided that a significant rewrite of some of their code was in order.

"As many of you may know by now, the 2.5 kernel feature freeze has come and gone, and it seems clear that the EVMS kernel driver is not going to be included. With this in mind, we have decided to rework the EVMS user-space administration tools (the Engine) to work with existing drivers currently in the kernel, including (but not necessarily limited to) device mapper and MD."

This announcement was met with TONS of positive praise on lkml: for the actual technical decision, for the mature and pleasant manner in which it was handled, and for the public policy of removing duplication of kernel code in general, simplifying the MD/device mapper code specifically.

Finally, Alan Cox stated about 2.4:

"I plan to try and push LVM2 to Marcelo after the next release. Whether he will take it I don't know. Obviously its good to have the ability to move back nicely to older kernels."

Re:What is up with the LVM?

[pwagland]

Why isn't Sistina's LVM making it into the kernel? SUSE has been including it as standard in their distribution for some time.

According to these articles LVM2 has made it into the 2.5 development series, as of 2.5.45. Thus, it is likely to also be in 2.6.....

Re:I won't get what I want: CryptoAPI

[OpenGLFan]

I stand corrected. As of 2.5.45, which is new enough for me to not have heard of it, and not mentioned on the main kerneli.org site.

*snif* I could cry. Thanks, Linux and the CryptoAPI people at http://www.kerneli.org

Gentoo may be for you

[FreeUser]

I really wish it were possible to get the CryptoAPI merged into the full kernel. I've been compiling kernels without problems since the 1.2 series, but CryptoAPI patches are more convoluted than any other patch series I've ever tried.

Given that you're no stranger to either GNU/Linux or compiling the Linux kernel, you may want to take a look at the source-based Gentoo distribution. Aside from making download and compilation from the author's tarballs trivial via the portage system (emerge rsync ; emerge [packagename]), the gentoo-sources kernel has numerous additional patches, including the crypto-api patches.

emerge rsync ; emerge gentoo-sources, followed by the usual cd /usr/src/linux; make menuconfig, etc. will bet you the Crypto API patch, as well as the low latency/preempt patches, grsecurity patches, and so on. All nicely applied already, and ready for you to compile and use.

Perhaps not as nice as if they'd made it into the feature freeze for 2.6, but a lot easier than the process you describe.

Re:what is up?

[rawshark]

With regards to kernel config, there should be an option to compile EVERYTHING possibly compiled as a module, as a module

There is.

I am not in front of my linux box right now, but in 2.5 you can do "make allmoduleconfig" or something like that, which will do what you want. I do not remember the exact make target, but you can do a "make help"

I do not believe this is in 2.4.

One caveat for newbies is that the code for the filesystem where your modules is located should not be compiled as a module. The kernel has to mount the filesystem before loading the module, and if it can't load the module for the filesystem, well, you get the picture.

Hope this helps.

Re:The Tao of Linux

[mandolin]

Edumacated readers will note that the parent post is a (well-done!) variation on the Tao of Programming. Nice job AC.

Re:what is up?

[iggymanz]

What benefits would you expect from a microthreaded kernel?

I switched from a monolithic one (SunOS 4.1.1B) to a microthreaded one (Solaris 2.5) on a Sparc 2 and my software ran at least 25% slower. I had a similar SPEEDUP switching from Solaris 2.6 to OpenBSD 2.9 on my Sparc 5

I could just imagine the benefits of all that kewl message passing on a constipated Intel architecture......do we hear 30% slowdown? 35%?

All About the Three Linux LVMs

[psamuels]

Why did the guy from Oracle ask for LVM? I thought the 2.4 series already had LVM (I'm using it now). From Linus' reply, it seems like there is a new LVM? Anyone have any information about the difference and what the problem with the old one was?

There are / have been three logical volume managers for Linux.

1. The original LVM by Heinz Mauelschagen (sp?) matured over several years and was in wide use well before Linus dropped it into kernel 2.3.35 or so. Its interface was patterned after the Veritas volume manager as found in HP-UX. Its functionality is similar to the AIX LVM, but somewhat (IBM would say a lot) more limited in scope.
2. IBM came up with their Enterprise Volume Management System (EVMS) to replace LVM. I think it is partially a port of the AIX LVM, but I'm not sure. The idea was to have a one-stop shop for all your volume management needs, including mirrors, RAID, snapshots, partition management, etc. This is sort of contrary to the Linux philosophy, where we normally layer things - LVM on top of RAID on top of partitions, generally. Thus, EVMS is seen as a lot of code duplication / bloat, but on the other hand it is quite appealing to "enterprise" customers and vendors, due partly to its IBM pedigree I suppose, and partly to the fact that these types of people really go for "integrated solutions".
3. Finally, Sistina (home of the original LVM hackers) came up with LVM2, which is a complete redesign of LVM - but, unlike EVMS, LVM2 is even more minimalist than LVM1. The kernel component of LVM2 is called Device Mapper, and is a generic way to chop up and paste together your block devices to form logical volumes. The LVM2 user-space code uses the DM interface and implements the rest of the LVM logic on top of it. From a user perspective, LVM2 looks a lot like LVM1.

So everyone agreed that the original LVM1 code, while filling an important gap back in the day, was too ugly to live. Even its creators had abandoned it to twist in the wind when they wrote Device Mapper and the LVM2 corpus. Due to some invasive changes to the entire block device code in Linux 2.5.1 or so, the in-kernel LVM code was left broken, and nobody has been interested in fixing it. It was to be replaced by either EVMS or Device Mapper - or both. Linus left this decision to the last moment, about two days before the feature freeze, when he put in DM and left out EVMS.

EVMS is, as I said, the more feature-rich of the two, but most kernel hackers seemed not to like it very much, due to the aforementioned code duplication and its "all your block devices are belong to us" attitude. Probably, when the Oracle guy asked for LVM, he meant EVMS.

You use LVM1 today. When the 2.6 kernel comes out, you will have to upgrade to LVM2, and compile DM into your kernel, but it should be a smooth upgrade path - your on-disk volume group format will still be supported. Two problems you may face are:

• dual-booting between kernels 2.4 and 2.6 - you have to have two sets of LVM tools - there are ways to manage having both installed at once, but I don't know if the distributors will manage this well or not, and
• if your root partition is on a logical volume, you need to arrange for it to be activated before mounting, probably with an initrd setup - in which case you'll have to have the LVM2 tools on your initrd instead of the LVM1 tools. Once again your distributor is supposed to handle the problems here but who knows how well they'll do.

You can preempt these problems by patching your 2.4 kernel with Sistina's DM patch, and migrate completely to LVM2/DM while still using 2.4. I'll probably do this the next time I have occasion to reboot, which may not be for awhile (my box has been up for 2.5 months - that's when I overloaded a circuit breaker...).

One last note. You may have heard of the well-publicised note where Kevin Corry announced that EVMS was changing direction. They will now reimplement what is currently a large kernel component of EVMS, moving it out of kernel space into user-space tools that will operate directly on top of Device Mapper (the LVM2 kernel bits). The hope is that DM will prove to be flexible enough that EVMS can continue to exist, with all its current features, purely in user-space - or possibly with a minimal amount of additional kernel code. So, if the Oracle people insist that EVMS is the way Enterprise Systems should run, they should still be satisfied. The EVMS team plans to have the new user-space-based EVMS out in a couple of months, well before 2.6 itself is widely adopted or, indeed, released.