Intel Chief Evangelist Comments on Linux Scheduler

ByeByeWintel writes "James Reinders is Intel's Chief Evangelist for Intel's Software Development Products. In a recent interview on Devx.com he stated: 'If I could get ONE wish fulfilled would be for OS scheduling to focus on processes, and not threads, for scheduling. And demand that processes manage their scheduling of threads ... There is a lot of opportunity for operating systems to offer these types of control in the 'running of applications' interfaces. I'd like an OS to let me specify the 'world' my application runs in (which processors, how many, etc.) These interfaces are available in Windows at run time (the task manager will let you adjust where a running task can go).'"

Is James clueless?

[Anonymous Coward]

James doesn't seem to acknowledge that such features already exist in Linux as well. Additionally Linux has __FAR__ better support of NUMA system management. Many Opteron-based systems are NUMA as well as many high end Intel-based servers. Is he just oblivious?

Re:I don't understand

[EvilGrin666]

Isn't this *exactly* what numactl gives you

Well yes, exactly. numactl will do this. However his bone of contention appears to be that there is no pointy clicky interface a la Windows TaskManager to do this. So I'm slightly puzzled why he's making out that there is a deficiency in the OS when the fact is that it's purely a lack of features in the window manager...

Re:Virtual Machines?

[jeremiahbell]

Putting all programs inside virtual machine like wrappers could work to keep the programmer from having to worry about scheduling in a scenario where the processes handled their threads, and it may also have good security implications. A major problem is malicious programs gaining access to the RAM, and trying to ferret our your hard-drive encryption key for example. Putting all programs in a locked Box might be a great way to fix the problem. Despite the performance penalty I believe this may be the future.

Re:You think maybe they have other things to do?

[xenocide2]

Yes, but if you have to pay someone to make the modifications you want done, and then pay them to maintain those modifications and to port them whenever something they interface with randomly changes, then you've got cost, probably quite substantial cost

Yes. People, on a large scale, don't work for free. Especially talented ones capable of correctly implementing something like per-cpu scheduling hints. What you can do to minimize the cost is pay people to get it into upstream. Moreover, the source code availability in the first place vastly increases competition. It's much easier to patch linux than to design an operating system people want to run software on that also happens to provide scheduler tweaks you like. In the closed source system you're negotiating with a small set of people that have access, while the Linux kernel's open source gives you a much wider set of alternatives, driving down prices in effect. Try negotiating with Microsoft to add a feature you like. If it's possible at all, consider the price they'd likely charge.

Open source isn't necessarily priceless. Education, talent, and time are still scarce resources. Open source just provides hired hands a platform for their interests and spare time, and a shared base to work from. If you're in IT for cheap labor, I still encourage you to look to offshore.

Re:I don't understand

[Rogerborg]

Christ, yes. Propagandist is more accurate.

Back in the Day, Microsoft's European DirectX 'evangelist' paid a visit to our game dev studio to try to sell us on the benefits of Direct X Retarded Mode [microsoft.com]. We had our engine running with D3DRM and Glide (I said, back in the Day), and the Glide framerate was easily twice that of D3DRM. He told us that that was impossible. We said "Look at the screen.". He literally refused to look. He just insisted that we were wrong, that there was no reason that D3DRM should be slower than Glide, and that we must therefore be doing something wrong. Did I mention that he wouldn't look at the screen?

Needless to say, we didn't even offer him lunch.

Re:Puh-leeeeze!

[Tim Browse]

I worked with the Evangelists in Apple Developer Relations, and my direct personal experience tells me that you're full of shit.

No, he's not. The fact that some of those you've met were good does not mean it does not imply 'highly paid fanboy'. Or maybe you were questioning the 'highly paid' part :).

For balance, the only Apple Evangelist I ever had experience with was a Quickdraw 3D Evangelist. He was a clueless idiot. I asked him how they addressed the performance issue caused by QD3D lighting only working with full RGB, whereas DirectX supported a ramp mode for significant performance increase. He didn't have a fecking clue what I was talking about. At the time, that was actually a pretty basic bit of knowledge about the software renderers. I mentally turned off when he replied to that question with "Let's take this offline" (we were in a meeting with him, me, and 2 other people from my company).

He could run the QD3D demos for us, though. He was really good at that. He also demonstrated the plug-in renderer architecture they had developed, which was by far the most interesting thing I saw that day (spent the day at Cupertino talking to Apple people). It allowed NPR and toon-shading stuff, long before it became mainstream. He even had a demo that showed a model being rendered in a 'cave painting' style. Depressingly, no-one at Apple (and especially not the Evangelist) seemed to realise how cool it was (except the people who wrote it, maybe).