RE: [microblaze-uclinux] Thinking about the 2.6 kernel

[Brettschneider Falk <fbrettschneider@xxxxxxxxxxxxxxxxx>]

First, I also think it's a good idea to switch to the 2.6-kernel scheduler
because once I digged through the kernel programming mailing-list archives
and many people said there that 2.4 is OK for SCHED_OTHER but all its
patches don't care much about SCHED_RR and SCHED_FIFO, though that is
important for being close to a real-time behaviour of a µBlaze user app. 2.6
was judged as have solved the bugs with those schedulling modes. Although I
don't know if they are really right...
Second, I think 2.4 wont be maintained as good as 2.6 in the future, with a
decreasing tendency.

Cheers, F@lk


> -----Original Message-----
> From: owner-microblaze-uclinux@xxxxxxxxxxxxxx
> [mailto:owner-microblaze-uclinux@xxxxxxxxxxxxxx]On Behalf Of Matthew
> Rubenstein
> Sent: Thursday, January 19, 2006 12:53 AM
> To: microblaze-uclinux@xxxxxxxxxxxxxx
> Cc: John Williams
> Subject: Re: [microblaze-uclinux] Thinking about the 2.6 kernel
> 
> 
> 	One major improvement in kernel2.6 over kernel2.4 is 
> the scheduler.
> This is a core kernel service that could most benefit from 
> parallelism,
> compared to other services. The project that ports the 2.6 
> scheduler to
> the uCLinux/microblaze could become the basis for porting the 
> scheduler
> directly to the FPGA. That kind of research positions uC/mb to reap
> serious benefit in all the development arenas that this 
> overall project
> touches.
> 
> 
> On Thu, 2006-01-19 at 08:48 +1000, John Williams wrote:
> > Hi folks,
> > 
> > The most common question I seem to get these days is "When 
> will you do a
> > 2.6 kernel port for MicroBlaze?".
> > 
> > My answer is generally the same - "It's on the radar, but no firm
> > plans".  Basically I need some pretty good reasons to 
> commit the time
> > and energy into a new kernel port.  Remember it's not just
> > linux-2.6.x/arch/microblaze, there's drivers to consider, 
> and changes to
> > auto-config architecture and so on.
> > 
> > So, here's your chance.  Convince me that a 2.6 port is the 
> way to go.
> > Good reasons, bad reasons, technical reasons, marketing 
> reasons - all
> > will be entertained.  I'm open to counter-arguments too, if 
> you think
> > it's *not* worth doing.
> > 
> > Let's see what you've got!
> > 
> > John
> > 
> > 
> > ___________________________
> > microblaze-uclinux mailing list
> > microblaze-uclinux@xxxxxxxxxxxxxx
> > Project Home Page : 
> http://www.itee.uq.edu.au/~jwilliams/mblaze-uclinux
> > Mailing List Archive : 
> http://www.itee.uq.edu.au/~listarch/microblaze-uclinux/
> > 
> -- 
> 
> (C) Matthew Rubenstein
> 
> ___________________________
> microblaze-uclinux mailing list
> microblaze-uclinux@xxxxxxxxxxxxxx
> Project Home Page : 
> http://www.itee.uq.edu.au/~jwilliams/mblaze-uclinux
> Mailing List Archive : 
> http://www.itee.uq.edu.au/~listarch/microblaze-uclinux/
> 

___________________________
microblaze-uclinux mailing list
microblaze-uclinux@xxxxxxxxxxxxxx
Project Home Page : http://www.itee.uq.edu.au/~jwilliams/mblaze-uclinux
Mailing List Archive : http://www.itee.uq.edu.au/~listarch/microblaze-uclinux/