Re: [RFC v2][PATCH 4/9] Memory management - dump state

[Dave Hansen <dave@linux.vnet.ibm.com>]

On Wed, 2008-08-27 at 15:48 -0500, Serge E. Hallyn wrote:
> Quoting Dave Hansen (dave@linux.vnet.ibm.com):
> > On Wed, 2008-08-27 at 15:34 -0500, Serge E. Hallyn wrote:
> > > (Or, is that too much effort required on your part to try and
> > > cherrypick bits of Oren's changes back into your tree?)
> > 
> > That would probably work as long as Oren is working on top of what I
> > have.  It's going to be a lot harder if I have to repeat the same
> > break-out process for each iteration of Oren's patches, though.
> > 
> 
> If Oren's purpose is not quite to create a upstreamable patchset,
> then it would make more sense for him to keep a git tree and
> put new patches on top of his existing ones (within reaason as he
> rebases).  Then you'd at least be able to trivially look at the latest
> deltas.

The trick would be compromising on things that I, for instance, think
need to be rewritten or removed.

Oren would have to rebase his work against what I do.  I guess you could
think about it like me being upstream from Oren.  Anything that I would
change, Oren would need to rebase on top of.

Oren, are you willing to keep a set of patches that add functionality on
top of a minimal set that I'm keeping?  Mine being the set that we're
trying to push into mainline as soon as possible.

-- Dave