User: Password:
Subscribe / Log in / New account



Posted Oct 26, 2011 16:36 UTC (Wed) by abacus (subscriber, #49001)
Parent article: KS2011: Scheduler testing

Not a single word about the BFS ? Does that mean that the upstream scheduler is now superior in all regards compared to the BFS ?

(Log in to post comments)


Posted Oct 26, 2011 19:41 UTC (Wed) by corbet (editor, #1) [Link]

The BFS author is not interested in creating a general-purpose scheduler, and he is not interested in working with the development community. It is not surprising that his work is not on the agenda at a meeting like this.


Posted Oct 27, 2011 10:47 UTC (Thu) by intgr (subscriber, #39733) [Link]

BFS was announced in 2009, that's ages ago at kernel development time scales. Mainline scheduler developers already analyzed and discussed it back in those days, made some improvements to CFS and moved on. It's simply not relevant anymore.

BFS wasn't "superior in all regards", don't expect that from the mainline scheduler either. Every heuristic algorithm will be better in some cases and worse in others.

Copyright © 2018, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds