Git pull request: ARM LPC32XX architecture
kevin.wells at nxp.com
Tue Mar 30 12:53:59 EDT 2010
> Subject: Re: Git pull request: ARM LPC32XX architecture
> On Sat, Mar 13, 2010 at 04:50:53PM +0100, Robert Schwebel wrote:
> > Hi Russell,
> > On Fri, Mar 12, 2010 at 09:03:53PM +0000, Russell King - ARM Linux wrote:
> > > However, what I do want to avoid this time around is having an
> > > excessive number of merge conflicts, which may require a pause to
> > > sort out (and I might ask people to bring their trees up to date
> > > first.)
> > Hmm, maybe we should make sure that more of the arm-sub-trees are in
> > linux-next; would that make the situation easier? If all of the trees
> > you pull are in next, at least in theory it should be a matter of
> > pulling them in the right order and send a pull request to linus.
> It's less of a problem if people follow the ordering of the machine
> dependent bits for arch/arm/Makefile and arch/arm/Kconfig - things
> tend to go horribly wrong when they don't, and git even gets confused.
> The solution to that is to ensure that these *stay* sorted and refuse
> to merge anything which breaks the sorting. Yes, there'll still be
> occasional merge conflicts, but keeping these things sorted should
> result in the resolutions being trivial.
> It all goes wrong when we end up with everyone adding to the end of
> the lists, and then I re-order it - that's the point where git silently
> guesses the resolutions wrong and we end up with, eg, two includes of
> What I don't want to do is re-sort these files, and then have to deal
> with re-doing the sorting each time I subsequently merge a tree because
> they're based on a commit prior to that re-sort.
Is there anything I can do to help move these along? It's not a rush job
to get these into the tree, but I would like to start getting other
changes and drivers submitted based on this work.
If I can help, please let me know!
More information about the linux-arm-kernel