login
Search
This week, KernelTrap has been speaking with Linux kernel hacker Rik van Riel. Rik is most recognized for his impressive rmap VM efforts, available as a patch for the 2.4 kernel here. He's also the founder of kernelnewbies. Living in Brazil, he works for Conectiva.
Check back Tuesday, February 26'th for the full interview. Read on for a brief preview. Other original KernelTrap interviews can be found here.
Rik says of himself:
"I was born in 1978, in the north-eastern part of Netherlands, where I grew up on my father's farm and spent most of my life. As a teenager I used to do farm work in the holidays, which allowed me to buy myself a computer. I started with DOS, played with OS/2 for a while when I got my 486 in 1994 and went to Linux later that year."In the beginning of 2000 I accepted a job at Conectiva, the largest Linux company in South America, and moved to the city of Curitiba in the south of Brazil. I'm working full-time on Linux now, mostly development but also helping out Conectiva's customers when problems happen."
And of his VM:
"The fact that my VM code is no longer in the kernel has two positive effects. First I don't have to worry too much about stability or code freezes, I can just develop the code in the direction of better quality without having to be afraid of changes. Secondly I can integrate bugfixes much quicker.
"This has lead to my reverse mapping based VM becoming stable within a few months of me picking it up again; it is now stable to the point where it survives the kernel torture lab Bob Matthews is running at RedHat and both Alan Cox and Michael Cohen have integrated the patch into their kernel tree. I suspect the rmap VM isn't as fast as some of the other VM code out there, but it does seem to hold up somewhat better in strange situations."
Rik says much more in the interview (still in progress) about himself and his LiZnux kernel efforts, so be sure to check back next Tuesday, February 26th for the full text.
Yay! i've been looking forward to this one :D
Way to go!
This is one of the interviews i've been looking forward to (besides the Alan Cox interview, that is.)
Here's a few questions i'd like to ask:
What's his (from a purely objective point of view) opinion on the AA VM? (this is bound to draw some flames) (not necessarily from Rik, he's a nice guy, judging from his LKML posts ;D)
How "magic" is VM coding?
What are the differences between MM and VM? Is VM nothing but swapping in and out?
Rik and Linus
I, for one, would like to know more about his "relation" with Linus. AFAIK, they seem to have some kind of love-hate (respect-flame) relation, and I'd like to know more about Rik's personal view on Linus, and their relation in particular. ;-)
did you hear about Rik and Linus?
you must lead a very interesting life.
it's a professional relationship, they clash but
they're both interested in getting the rmap VM
into 2.5.
That's all. No one is having an affair, no one
is secretly pregnant, no one has a dark past...
It's software engineering not some australian soap.
(maybe linus and rik should go on Jerry Springer?)
Suboptimal
"I suspect the rmap VM isn't as fast as some of the other VM code out there, but it does seem to hold up somewhat better in strange situations."
I think Rik is understating a real problem there. People spend as much money as they can on hardware, overclock their CPUs (even to the extent that they are taking a risk that it will malfunction) and tweak to get the most out of their hardware. Andrea Arcangeli's VM is both simpler and faster at this stage, and these are the two things that really matter. Sure, stability is essential but the kind of stability Rik is talking about isn't the "reboot windows 5 times a day" type -- it may strike a non-Intel CPU after 8 months of operation under unusual circumstances however. I'd rather have a Linux that is fast and easily developed for than one that has a VM that is both difficult to understand and suboptimal. I don't see why Rik's VM is getting so much respect from the very people who normally value efficiency.
you're high!
Linux kernel coding has always been about correctness first.
For a long time google couldn't run 2.4 on their servers because it would always fail under heavy loads. Other people probably had the same thing happen.
Re: Suboptimal
both VMs are nice, it just depends on your application. i'd like to see an option in the kernel config that allows either desktop(kinda-stable, fast), and workstation/server(stable, high load, minor speed cost).
i know that probably won't happen, as it'd be a PITA to keep maintained, but it'd be cool.
Thank Linus this will never happen
This will Never (ever (tm)) happen, if linus designed stuff this way
there would be thousands of combinations, it only takes 10 yes or
no questions to yield 1024 different out comes.
How would any of these out comes ever get well tested, this would
be a nightmare and a swampy hiding ground for bugs by the bucket load.
I'd like to see riks' VM go into 2.5, if it is even 1% more stable
than AAs' VM, the last unscheduled down time I had was 17 months ago
and that was on a machine with a 171 day uptime. Stability rules.
kinda stable?
being fast doesn't mean unstable.
If something is unstable that means it has bugs. Why do you want bugs?
It's understandable that the VM has bugs because that is a very hard area to work in. (These are not bugs in the syntax, but bugs in the design, and algorythms). At the same time one would hope that the bugs would get fixed.
You are so stupid
AMD's CPUs are MORE STABLE than intel GARBAGE! You are brainless and have cancer.
rmap patch
I have being using Rik's rmap patch and it really improved the
performance on my desktop.
I encourage every who would like to test it to try from one of
his kernels at
ftp://ftp.nl.linux.org/pub/conectiva/kernel-rmap/
Those are rpms based on the ones used by Conectiva. It should
be trivial to install them on non Conectiva Linux systems.