At this week’s CA World 2003 show in Las Vegas, Torvalds sat down with InfoWorld to talk about the completion this past weekend of Version 2.6’s first test release; what will keep him up at night worrying about that test release; his thoughts about the impending SCO-IBM suit and possible impact on Linux development; and something of a wish list of technologies to be included in upcoming versions of the open source operating system.
Torvalds Gets Down to the Kernel
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Hmm… they say that torvalds is a chip designer
. Later they say that VM stands for Virtual Machine (and not Virtual memory).
[i]Hmm… they say that torvalds is a chip designer
.
Linus did work at TransMeta which makes processors. I’m not sure about exactly how much he did involving their design, but I was under the impression he did work related to running linux on Crusoe.
VM stands for Virtual Mammory. Mmmm cyber-breasts…
I’m pretty sure the reason Linus was hired by transmeta is because of he was familiar with x86 architecture. If you follow the history of Linux, part of his reason for starting it was simply to learn x86 ASM + architecture. And if you think about it, he pretty much developed Linux from the time of 386s and 486s all the way up to modern processors. Thus, he obviously knew a great deal about the architecture aswell as what changes chips had gone through — this of course being KEY to transmeta’s code morphing software. Which is I’m pretty sure specifically what he worked on. While Linux does run on Crueso, I don’t think he was ever paid by Transmeta directly for any work on Linux, including having it ported to the chip.
actualy, he was lead on a linux project…I think they were developing a linux for their embeded use or somehitng like that.
I’m pretty sure the reason Linus was hired by transmeta is because of he was familiar with x86 architecture.
Crusoe processor has a different design on the ACPI-PowerManagement. The processor is ideally designed to make calculations not based on silicon contented instructions but on silicon embbeded (so to speak) software. It’s a X86 architecture.
I think Linus was part of the software embbeded team.
See it as a kind of “winmodem” performing CPU (it makes the job but it is rather a software job, not a hardware task.
I am not an expert but the idea behind the CPU design was to make a rather power efficient CPU for embbeded systems not a desktop CPU. It works good for stand alone appliances, softwarelets, … not with much acceptance I believe.
_______________
Like many I can hardly wait to try the *final* 2.6 kernel !
Linus did in fact write the x86 software emulation layer for the crusoe chip. Like someone said earlier, he had intimate knowledge of the x86 architecture so he was perfect for the job.
…movie at 11.
> Linus did in fact write the x86 software emulation layer for the crusoe chip.
If this is in face true, based on this article he didn’t necessarily do a good job…of course one of the problems is that there are just way too many nuances surrounding the x86 architecture.
http://www.vanshardware.com/articles/2003/07/030715_Transmeta/03071…
Personally I think the crusoe would have been a much better chip if they had designed a brand new higher level instruction set that would lend itself towards very efficient “code morphing” instead of trying to “code morph” a current half assed instruction set. Of course then they would not have not been able to claim x86 compatability.
At the heart of the Crusoe CPU is a VLIW instructionset, surrounding that is a x86 morphing layer which converts and caches code from x86 to VLIW to run.
Why not write directly for the chip? because then they would have to provide backwads compatibility hence the reason for the morphing layer. They can the, if they want, radically change the core whilst mantaining compatibility.
Crusoe’s success hinged on the open source release of their code morphing software. They didn’t release it so everyone could port the chip to Sparc, PA-Risc, MIPs, POWER4, etc. And look at them now. Nice energy efficient x86 chips. But I doubt they’ll be competing with Intel or AMD soon.
But I would still consider buying a Crusoe laptop if I could run Solaris, OSX and Alpha Linux on it. I’m just waiting for Transmeta to hurry up and deliver the products capable of doing it. What’s taking them so long?