Carrier Grade Linux/Gaps Alpha1

[greg@xxxxxxxxx (Greg KH)]

On Tue, May 27, 2008 at 09:44:01AM -0700, MacDonald, Joe wrote:
> > > > What is supposed to happen if this is implemented?
> > > 
> > > In a perfect world the implementation would be flawless resulting in
> > > faster boot times.
> > 
> > Are you sure that this is an area in which the boot time of Linux is
> > slow?  By what ratio in comparison to other areas?
> 
> You'd said you implemented and removed it, so multiple someones were
> sure enough about it to get it accepted at one point.

No, it was just a single some-one, me.  I thought it might be a nice
idea, implemented it, benchmarked it, added it to the main kernel tree,
others tested it out, found loads of real-world problems with it, no
real benchmarks to show it really was worth fixing those problems, and
then removed it.

> What was the motivation for removing it?

It did not really achieve any gain, and the problems it showed were
going to be very large to overcome, especially for no real gain.

> If it's already been tried and shown to not work, that's a very
> compelling argument for deprecating the requirement, or at least
> re-opening the discussion on the validity.  Do you have benchmarks
> that you can share?

See previous message (1% speed increase was the best possible
measurement that I got, but the variance in booting was about 5%
overall, so it was lost in the noise pretty much all the time.)

> In that same vein, the way we're going to get better requirements is to
> have more active participation in the evaluation process as the
> requirements are submitted to the CGL workgroup.  And we're always
> looking for active members.  *hint, hint*

I'll continue to lurk and pop up if I see nonsense like this every once
in a while :)

thanks,

greg k-h