Actually, I ran into a possible other option:

git rebase -p master

which is supposed to preserve merges by not ignoring them.  It seems to work
in the limited testing I've done.  I'll wait for more commits on origin
before I confirm.

// Dean Glazeski


On Mon, Nov 23, 2009 at 6:24 PM, Zach Welch <[email protected]> wrote:

> If you have a series (or tree) of branches, I wrote a small script to
> rebase them automatically; however, it falls down when something fails,
> in so far as you have to finish things manually (for now).  Basically,
> it does this process for N steps and for any number of branches in your
> local web of commits:
>
>  git branch step1-tmp step1
>  git rebase master step1
>  git rebase --onto step1 step1-tmp step2
>  git branch -D step1-tmp
>
> Is that what you mean?  If so, my script is in the 'buildscripts' branch
> of my fork on repo.or.cz, named rebase.pl.  Here are a few ways I want
> to see it improved:
> - figure out the branch tree(s) automatically
> - avoid leaving temporary branches around
> - full continuation and abort support, a la 'git rebase'
> - etc. :)
>
> Patches welcome, but I do not plan on adding those scripts to mainline.
> That branch contains things that I wrote which may be useful to others,
> but I am not pushing to push any of it.
>
> --Z
>
> On Mon, 2009-11-23 at 16:34 -0600, Dean Glazeski wrote:
> > I went ahead and moved everything over as a fork on my own.  I'll
> > leave the other around until this stuff gets hammered out and I'll
> > remove it at that point.  The new play area is
> > http://repo.or.cz/w/openocd/dnglaze.git.  I've rebased everything to
> > the new origin/master.  Is there a good way to rebase two branches
> > that merge to create another branch?  With my current layout, if I
> > rebase each branch individually, the merge disappears.
> >
> > // Dean Glazeski
> >
> >
> > On Mon, Nov 23, 2009 at 3:47 PM, Zach Welch <[email protected]>
> > wrote:
> >
> >         On Mon, 2009-11-23 at 13:04 -0600, Dean Glazeski wrote:
> >         > Hi all,
> >         >
> >         > I've finished up the implementation and documentation for an
> >         AT91SAM9
> >         > NAND driver for OpenOCD.  In total there are about 34
> >         patches that
> >         > includes some refactor work for both the NAND controller
> >         layer and the
> >         > ARM NAND I/O pieces.  I have the branch posted at
> >         > http://repo.or.cz/w/dnglaze.git under the at91sam9-nand
> >         head.  I'm not
> >         > sure what the best approach is for working this into the
> >         tree cause
> >         > this is the biggest change I've attempted to contribute.  I
> >         can only
> >         > verify that this works for my Olimex SAM9-L9260 board.  Any
> >         > suggestions, testing, or improvements are very welcome.
> >
> >
> >         My first question is: would you mind moving your repository to
> >         be a fork
> >         of openocd.git?  You need to contact pasky (the maintainer),
> >         but he will
> >         put it in the right place for you.  There are several good
> >         reasons for
> >         him to do this, including saving mirroring bandwidth and disk
> >         space on
> >         the server.  For us, it puts you in the list with the other
> >         forks, so
> >         others can see when you have pushed new work more easily.
> >
> >         Anyway, I just pulled your branches and verified each patch
> >         builds okay.
> >         That's a step in the right direction.  I'll look more closely
> >         at them
> >         when I am done with my command registration series, which I am
> >         getting
> >         nearly ready to post.
> >
> >         --Z
> >
>
>
>
_______________________________________________
Openocd-development mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/openocd-development

Reply via email to