LinuxCon Roundtable in Torvalds Quotes


Taking a break from having fun with the 7 cool Linux projects every geek should try, we're attending LinuxCon. So far, the highlight was undoubtedly the Roundtable session held earlier this afternoon. The panel consisted of Greg Kroah-Hartmann, Ted Ts'o, Linus Torvalds, Chris Wright and Jonathan Corbet and it was ably hosted by James Bottomley. It was funny, informative and convivial. But it was Linus that everyone wanted to hear, and despite a reticence for public appearances, he actually had a lot to say. Here’s our pick of his quotes.

On the most innovative feature that has gone into the kernel over the last 12 months:

“I’ve not been using a lot of the new features. Yes, my laptop works, yes I actually really like the new profiling code because it’s just easier to use than the oprofiling was. But for me, the biggest issue, which I already talked to some people today, is how much easier my job has been getting in the last few months, and that’s all I care about. I want to sit there and drink my Fufu drink, and just press a button and it all just merges everybody else's code. And that’s the one feature that is not visible to user land, but it’s the one feature that is most important to me is how the development model actually seems to be working. And it’s working better than it did even just 6 months ago, and certainly a year ago, where I beat up a lot of people over how they did things, because it made it more difficult for me, and it took a while, but they seem to have all gotten it.”

On motivation throughout his history with Linux:

“It has changed a lot over the years. It started out being all about the technology, and all about just really twiddling with the hardware and just learning and just doing something cool and sitting in my basement. It wasn’t my basement at the time, it was my mother’s basement [everyone laughs]. But really being low level and doing the programming. That eventually faded and then it become somewhat about the community and the fame, hey that was great. But also new problems that I hadn’t had before. The SMP work, many years ago, that other people started and I kinda took over, continued to motivated me.”

“These days it’s all about the community. Actually, I shouldn’t say the community because when anybody else says ‘the community’ my hackles rise. It’s not ‘the community’, there’s no ‘one community’, but it’s this whole thing about working together with people and it’s fairly social, and I really enjoy arguing too. It’s actually a big part of my life is this occasional flame threads that I love getting into and telling people they’re idiots. It gets it out of my system. And I really like the email back and forth. All my technical problems were solved so long ago that I don’t even care. I don’t do it because of my own needs on my machines, I do it because it’s just interesting and I feel like I’m doing something worthwhile.”

On performance drops with each release of the kernel:

“We’re getting bloated and huge. Yes, it’s a problem.”

“I’d like to say we have a plan [audience laughs].”

“Sometimes it’s a bit sad that were are definitely not the streamlined, small hyper efficient kernel I envisioned 15 years ago. The kernel is huge and bloated and our iCache footprint is scary. There’s no question about that, and whenever we add a new feature, it only gets worse.”

On documentation:

“In the kernel we have this sign-off process, where patches as they flow through people are supposed to be signed off. And we had legal reasons for doing it initially, and the legal reasons have kind of gone away because nobody worries about SCO very much anymore. But it turns out it’s a really nice flow process, where people actually see how code came in, so it’s nice, and I’m seeing that in a number of non-kernel projects too. So I think there may not be a lot of documentation about how the kernel does it, but I think a lot of open source people do see the kernel model and it actually ends up being, the same way there’s this Unix mindset of how things are supposed to work, I think the kernel model has actually become this mindset of how open source projects are supposed to work, at least for a subset of projects out there.”

On user-space driver frameworks reducing kernel bloat:

“But you certainly would not want to take an existing kernel filesystem and then move it into user space. that would just be crazy. That’s LSD trippy, kind of. Don’t do that. A lot of the time it gets way harder to debug, we’re moving things into the kernel out of user space because of issues like that. We’re doing kernel mode setting, and doing a lot more of the logic of graphics in kernel space, actually makes it a lot easier for everybody because now you don’t have two broken pieces that don’t have a clue what the other piece is doing, and trying to debug across that kind of divide is horrible. The debugging advantages of doing drivers in user space have always been kind-of suspect. I don’t think it’s really been true, except in very early kind of bring up meaning. Now people can do. You can do block devices in user space, you can do character device drivers in user space, you can do filesystems in user space. Go wild. It’s not, I think, relevant for any major filesystem or driver but it’s there if you want to. Now Greg can disagree with me.”

You should follow us on or Twitter

Your comments


thanks for the quotes.
much luv 2 the tux.


By the way, iCache is not some product from Apple :) He's talking about the CPU Instruction Cache (or I-cache).

Graphics in kernel...

Yuck. One on the reasons I've always preferred Linux over Windows is that a GUI glitch doesn't hose my whole system.

Remembering Unix roots

Those who have used Unix (30 years ago) might understand what Linus means.

And I am sad to see that there is no way for Linux to be installed this way (just a tiny core with the only features that YOU need).

Good coding practices are not dead (see TrustLeap G-WAN) but they are getting increasingly rare -also because this industry hates efficiency (it does not sell hardware) and does whatever it takes to crunch any initiative in this matter.

Sad times for real coders. But the crisis might change a thing or two about the appeal of efficiency. We will see.

Re: Graphics in kernel...

> Yuck. One on the reasons I've always preferred Linux over Windows is that a GUI glitch doesn't hose my whole system.

Crashing X has always been able to pretty much hose the system, unless you can access the box via the network. If the display, keyboard and mouse are hung, there's not much to be done except reboot. Moreover, Linus is not talking about moving X into the kernel, only some of the driver code.


Userspace drivers do have one advantage: They are safer to debug.

You have something in userspace in the "unprivileged userland sandbox" and it crashes, whoopdedoo, big fat core dump at worst.

Something goes screwy in the're hosed.

Being able to test something in a sandbox outside the "holy of holies" of the kernel is useful.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

Username:   Password: