| Author |
Message |
sidro
|
|
Post subject: BFS
 Posted: 02.10.2010, 11:20
|
|
Joined: 2010-09-12
Posts: 15
Location: Romania
Status: Offline
|
|
| Is BFS the default scheduler for aptosid? |
|
|
| |
|
|
|
 |
|
towo
|
|
|
Post subject: RE: BFS
Posted: 02.10.2010, 11:26
|
|
Joined: 2010-09-13
Posts: 488
Location: Pößneck / Thüringen
Status: Offline
|
|
| No, the default scheduler is CFS and BFS is not in mainline kernel. |
|
|
| |
|
|
|
|
|
sidro
|
|
|
Post subject: RE: BFS
Posted: 02.10.2010, 11:35
|
|
Joined: 2010-09-12
Posts: 15
Location: Romania
Status: Offline
|
|
|
|
|
|
|
slh
|
|
|
Post subject: RE: BFS
Posted: 02.10.2010, 11:41
|
|
Joined: 2010-08-25
Posts: 585
Status: Offline
|
|
This discussion is moot, as long as BFS is not mainline (not even intentions thereof).
That said, CFS and BFS are in a head to head race, so neither has a steady lead anyways. |
|
|
| |
|
|
|
|
|
sidro
|
|
|
Post subject: RE: BFS
Posted: 02.10.2010, 11:50
|
|
Joined: 2010-09-12
Posts: 15
Location: Romania
Status: Offline
|
|
Why isn`t a part of aptosid kernel?
I don`t understend why not.
Why cfs and not bfs? |
|
|
| |
|
|
|
|
|
towo
|
|
|
Post subject: RE: BFS
Posted: 02.10.2010, 11:59
|
|
Joined: 2010-09-13
Posts: 488
Location: Pößneck / Thüringen
Status: Offline
|
|
BFS is not in the mainline kernel, as sayed above.
So it can't be in the aptosid-kernel.
Btw, i had fanzy effects with BFS on my system, if i use the nvidia binary blob. |
|
|
| |
|
|
|
|
|
DonKult
|
|
|
Post subject: RE: BFS
Posted: 02.10.2010, 12:11
|
|
Team Member
Joined: 2010-09-02
Posts: 417
Status: Offline
|
|
sidro, as towo and slh already said: BFS is not in the mainline kernel, so it is a no-go for aptosid.
Mainline means here that Linus Torvalds hasn't accept it and isn't even near to the point that he will accept it.
aptosid and in this regard slh as your kernel master tries to stay as close as possible to upstream ALWAYS as every additional patch which is not accepted upstream can introduce new bugs nobody will take care of (it was not accepted, so it is likely to be abandoned soon), needs to be maintained (= constantly changed to fit into the kernel, which means another bunch of possible bugs and a big time sink) and never ever get security support so even if the patch opens the gates to hell it is likely that nobody will care to close them as it was just for testers and not for general consumption.
Beside that you are talking about an integral part of the kernel - a bug here can have catastrophic effects - and that it is mostly a big hype what you hear. Even if (note, i am talking subjunctive, that is not true!) BTS would be 10x times faster than CFS, how did you measure fractions of microseconds? For most people the feeling that it is faster with BTS is a figment…
Rule of thumb: Just ignore all the chit-chat about kernels in general, as aptosid kernels are most of the time one or two major kernel upstream releases ahead of the big distributions most performance tests (which are mostly run on these big distributions) are outdated even before they are written… |
_________________
MfG. DonKult
"I never make stupid mistakes. Only very, very clever ones." ~ The Doctor
|
| |
|
|
|
|
|
|
