BFS Scheduler Benchmarks

[Solitary]

It probably can have different effects on different configurations. It solved some people problems with latency:

http://lkml.org/lkml/2009/9/10/229

I can't say a word about this, because my system is responsive in both cases.

Me niether really, since I didn't test anything.
I feel that my computer is responsive enough, but I guess it's a pain that I am used to.. (amd sempron 3400, nvidia 7600gt agp, 2.6.28, kde 4.3)

[unix_epoch]

Me niether really, since I didn't test anything.
I feel that my computer is responsive enough, but I guess it's a pain that I am used to.. (amd sempron 3400, nvidia 7600gt agp, 2.6.28, kde 4.3)

I noticed a big drop in perceived responsiveness when I went from a 7 series to an 8 series Nvidia card, probably due to poorer 2D performance.

[kraftman]

I noticed a big drop in perceived responsiveness when I went from a 7 series to an 8 series Nvidia card, probably due to poorer 2D performance.

This is possible, esspecialy if you use KDE 4.x. I thought NVIDIA fixed this Have you tried BFS maybe, to see if 2D is also worse?

[BlackStar]

This is possible, esspecialy if you use KDE 4.x. I thought NVIDIA fixed this Have you tried BFS maybe, to see if 2D is also worse?

Bah, still not fixed here, even with 190.xx. My older 7600 mops the floor with my newer 9500...

I'm wondering if BFS could make a difference here. I rather doubt it, but I'll probably make a few tests when I have the time.

[kebabbert]

Too bad Linus will not accept Con Kolivas patches. This seem to be a good thing.

[RealNC]

Too bad Linus will not accept Con Kolivas patches.

That is not true. Kolivas will not submit the patches to Linus in the first place.

[kraftman]

Too bad Linus will not accept Con Kolivas patches. This seem to be a good thing.

Does Con make some patches for mainline scheduler so Linus should accept them? I hope you don't base your opinion on Apache benchmark?

They made changes for .32 to match BFS is some areas.

[kebabbert]

That is not true. Kolivas will not submit the patches to Linus in the first place.

I am refering to the Completely Fair Scheduler. Con submitted patches and they were basically ignored. Until Ingo dediced it was a good idea and made a new implementation, then it got accepted as a good idea. That is why Con quit. Havent you read the long interview with Con Kolivas about why he quit?

But hey, why not accept patches? Linux is 11 million lines of code now and even Linus T says it is bloated and huge. A few 100.000 lines of code is nothing.
http://blogs.zdnet.com/hardware/?p=5521&tag=nl.e550
Linus T says that they fix bugs as fast as they add code, but that is of no help. The code that gets bug fixed, is swapped out for new code soon and new bugs will be introduced, all the time. That is why Andrew Morton says the quality of the Linux kernel is declining.


The funny thing is, Theo pointed out that they should slim the code and fix it, 5 years ago and was called a Troll
http://www.forbes.com/2005/06/16/lin..._0616theo.html

"[Linux] is terrible," De Raadt says. "Everyone is using it, and they don't realize how bad it is. And the Linux people will just stick with it and add to it rather than stepping back and saying, 'This is garbage and we should fix it.'"

Same article, IT company CEO, PhD Comp Sci: "You know what I found? Right in the [Linux] kernel, in the heart of the operating system, I found a developer's comment that said, 'Does this belong here?' "Lok says. "What kind of confidence does that inspire? Right then I knew it was time to switch [OS]."


Nothing is true until Linus T says so. You can say whatever you want, but unless Linus T confirms, it is not true. If I say Linux is bloated, it is not true - unless Linus T confirms it. If I say Linux has problems with backwards compatibility, it is not true - unless Linus T confirms it. If I say Linux has problems with unstable ABIs, it is not true - unless Linux confirms it. Dont people think themselves? Dont they observe themselves? Must they wait until Linus T says so?

[nanonyme]

And you think it's bloated just because Mr. de Raadt says so. Talking about believing in authority figures...

[kraftman]

But hey, why not accept patches? Linux is 11 million lines of code now and even Linus T says it is bloated and huge. A few 100.000 lines of code is nothing.

I see troll is back. Linux is "bloated and huge", but when comes to archive. You can cut what you want. Btw. Solaris is bloated (slowlaris), BSD* are bloated, Windowses are bloated, OS X is bloated everything which is useful is bloated. You are really sick man Almost my favorite troll. Theo is the last who has right to say a thing about Linux. I wouldn't mind if they merge even some 3D shooter till it won't affect core part too much :P