Two Years With Linux BFS, The Brain Fuck Scheduler

[Kirurgs]

Hi!

I'm not here to judge this, but RealNC seems to know what he's talking about, judging from posts he knows and talks specific things, while it seems that kraftman is not talking specific but overall and sometimes not correct or so.
BFS patch is not CK patchset, this was the thing when everything went wrong and YES it is not the same (hz thing is in CK, nothing to do with BFS)!

Anyhow, BFS for me works, not sure is it better at that wokrload and is poorer at other, but it works quite well. Throughput benchmarks is not the point of BFS at all, desktop experience is. Phoronix made excuses in Cons blog about those benches and promised to deliver proper benches what I'm personally are waiting to see.
Hope those ones will deliver us overview of CFS vs BFS where and how it's appropriate!

P.S. These "wrong or so" ones actually were not bad at all at least for me - we saw how throughput was affected by BFS

regards
Kirurgs

[F.Ultra]

Well, don't openarena and unigine benches also list the min fps? That should be a good indicator.

FPS is a throughout problem, not a latency one.

[kraftman]

BFS patch is not CK patchset, this was the thing when everything went wrong and YES it is not the same (hz thing is in CK, nothing to do with BFS)!

Another one who doesn't even know what was this about?

[kraftman]

If you have to ask then this patch is not for you.

As the patch says, there is broken software that behaves better with obscene Hz values. Users who happen to rely on such software will be well aware of this issue and will appreciate the additional flexibility over CFS or default BFS. The rest will ignore it as if it never existed.

Is there some list of "broken" software that doesn't work with the patch? That users will know the software is "broken", but after the fact.

[RealNC]

If you don't need it, don't use it. It's as simple as that.

You don't even have a point anymore other than asking idiotic questions for the sake of asking them.

[kraftman]

If you don't need it, don't use it. It's as simple as that.

You don't even have a point anymore other than asking idiotic questions for the sake of asking them.

The problems were in your misinterpretation and proofs from the future. It's good when you blame CFS, but it's bad when someone blames some strange option. That's you.

[RealNC]

It's good when you blame CFS, but it's bad when someone blames some strange option.

Yes, it is exactly as you describe. And that's because CFS is mandatory. If there wasn't BFS around, I would be *forced* to live with CFS. On the other hand, no one forces me to apply a silly patch that, even *if applied*, doesn't force anything but only introduces new choices that you have to enable yourself after reading a ton of warnings.

Because of that, I can blame CFS but not an optional patch.

[b15hop]

this benchmarks are totally inadequate. BFS scheduler is designed to reduce LATENCY in desktop applications.

i'd check for amount of frames dropped in some FPS game or quality of video capture framerate-wise, as this is where the scheduler latency matters. but these things cannot really be measured with a benchmark ( i think ).

before CFS epsxe emulator would stall randomly for ~0.5 second now and then. on CFS i sometimes get 0.1 sec delays, which is not the case with BFS at all. that is what should be measured, not performance of webserver or how much FPS can you squeeze of a game.

Phoronix staff - please, read again this post http://lkml.org/lkml/2009/9/6/231 and think about this article again.

I've tried using BFS. I don't see the huge advantage it brings. What helped me reduce latency was to compile my own kernel cut down to nothing. That helped a LOT. But who in their right mind is going to do that... How many Gentoo users here? I went to Arch linux for the fact that it was well built to start with. I even tried a BFS from compilation and it didn't really do much. I'm a big fan of reducing bloat to increase speed. I think that's far more important than curing the symptoms to work around the bloated problem. On a positive note, reducing bloat with added BFS might be a winning combo for meebo? Who knows? x)

Because he wrote a scheduler that improved Desktop Linux for a lot of people. About the insults, surely, when you're treated with so much ignorance you tend to tell all the ignorants to go fuck themselves. Which is only something that raised my respect for CK and what he's doing even more. He's like the lone hacker who accomplishes tremendous tasks and throws the results at the face of the corrupt establishment. In other words, a software hero.

What I have learned over the years is that one guy, alone, improved my Linux experience a hell of a lot while his contributions were never acknowledged properly. And also have learned that people like you will continue to spread FUD because you simply can't stand being wrong. But I've got some bad news for you: You're not the pope.

I'd agree with you here but it's something I don't specialise in so I feel like my opinion has no weight. I think his scheduler works well, so all the critics can go jump. But I'm also one of those critics so I can go jump too. Mainly because as I said earlier, it didn't give me much improvement in speed. Compiling my own kernel that has been cut down to nothing helped a lot though.... These days I don't bother so I think the BFS should be given as an option. It's especially nice for distro's like gentoo, but I don't use gentoo. I like having options which is a valid reason to use Linux in the first place. PS no point getting emotional about it because it's not worth it. You probably know more than me here. I'm guessing the right combination of software and hardware make the BFS shine.

Maybe it's not Jack itself, but the whole audio chain, from the synths up to Jack. Lots of processes involved. With BFS, I can use 64 frames/period, no problem and no audio drops. Total latency of everything combined is well under 6ms. With CFS, it craps out as soon as I actually start playing some synths and CPU load rises. I have to raise latencies up to 15ms to make it work reliably.

It's really like day and night.

I had a similar problem with slackware ages ago. One of the reasons I tried recompiling the kernel was for that reason alone. I have to agree that CFS has terrible problems there... But tweaking the standard kernel fixed my problems.

Yeah, but that's just people using powerful systems. Try a 2.4GHz E6600. Without BFS the GUI gets really bad.

Going by your information, this makes me believe that BFS is ideal for phones and other mobile devices. I am an AMD fanboy so not sure if that effects anything. (Phenom II 955 quad core)

PS: Back when I used BFS I was on a single core machine with only 1GB of ram... These days I use a quad core with 8GB of ram, which is why I don't bother recompiling it these days. Though in saying that, I really like the idea of BFS in a way because I think the current kernel is built around server technology and not desktops. Thus my theory that possibly BFS would be idea as an option for all us desktop users.

I dont know if you read the Con interview? He explained that Linux et al were not to nice.

So, it seems that Linus has some attitude problems, and that Con got in his way. There are numerous unpleasant stories about Linus and his ego. Maybe you missed them?

I get the feeling this is the only reason why M$ and Apple has such a huge upper hand over Linux. A happy (crazy) guy waving cash excites people.... Who can argue with that?

Kraftman: Why is anything bad? People will generally do what they want. At least we have an article to discuss something in the first place. We need more options like this to begin with.

[misiu_mp]

I've tried using BFS. I don't see the huge advantage it brings. What helped me reduce latency was to compile my own kernel cut down to nothing. That helped a LOT.
...
I had a similar problem with slackware ages ago. One of the reasons I tried recompiling the kernel was for that reason alone. I have to agree that CFS has terrible problems there... But tweaking the standard kernel fixed my problems.

What do you mean "cut down to nothing" ? You can't remove drivers and subsystems without losing hardware support and features. So what was your config, what exactly did you cut?

[b15hop]

What do you mean "cut down to nothing" ? You can't remove drivers and subsystems without losing hardware support and features. So what was your config, what exactly did you cut?

That's exactly what I did. Cut drivers, cut wireless, cut certain encryption methods, cut ISA drivers or anything that was ancient. Anything I knew I didn't use. Only problem with doing this, I had to re compile if something didn't work. Which became a pain. That also meant that I learned what was what. Just cutting any random thing out is trial and error. So I would read up on what each section was used for. There are a LOT of different sub menu's... Saving my make menuconfig wasn't always a good idea either. Incompatibility between versions would mean It was better to start from scratch. At the time I would write down on paper what was important and go by that. Made things easier that way. In saying that, as time progressed, the kernel became more and more complex. Think of what was in a .config from 7y ago and compare it to now. x)

(edit) PS: Another thing I did was compile in kernel drivers. So the modules didn't load at start-up. That helped improve boot times. I think I had my NIC and (alsa) Audio drivers all compiled into the kernel. x) I notice that if I did a

ps aux

I would get much less threads or processes with a minimised kernel. So the system felt less laggy and cleaner.