Re: [SCHED] Totally WRONG prority calculation with specific test-case (since 2.6.10-bk12)

[Peter Williams <pwil3058@bigpond.net.au>]

Nick Piggin wrote:
> Peter Williams wrote:
> 
>> Paolo Ornati wrote:
> 
> 
>>>     1) nicksched: perfect! This is the behaviour I want.
>>>
> ...
> 
>>>
>>> transcode get recognized for what it is, and I/O bounded processes
>>> don't even notice that it is running :)
>>
>>
>>
>> Interesting.  This one's more or less a dead scheduler and hasn't had 
>> any development work done on it for some time.  I just keep porting 
>> the original version to new kernels.
>>
> 
> It isn't a dead scheduler any more than any of the other out of tree
> schedulers are (which isn't saying much, unfortunately).

Ingosched, staircase and my SPA schedulers are all evolving slowly.
  Are there any out there that I don't have in PlugSched that you think 
should be?

> 
> I've probably got a small number of cleanups and microoptimisations
> relative to what you have (I can't remember exactly what you sucked up)
> ... but other than that there hasn't been much development work done for
> some time because there is not much wrong with it.
> 

I was starting to think that you'd lost interest in this which is why I 
said it was more or less dead.  Sorry.

Peter
-- 
Peter Williams                                   pwil3058@bigpond.net.au

"Learning, n. The kind of ignorance distinguishing the studious."
  -- Ambrose Bierce