On Thu, Jun 19, 2014 at 01:04:32AM -0700, Jeremiah Mahler wrote:
> > After looking at the patches I suspect this should be a single patch.
> > That way it's bisectable, and the changes outside of read-cache.c are
> > small enough that it's not too much of a burden to review as a single
> > patch.
> >
> That would be a pain to bisect if the partial application of the patch
> set left the system in a broken state. Good suggestion.
One trick I use, especially when refactoring, is to use an interactive
rebase to test each commit in isolation, like:
GIT_EDITOR='sed -i "/^pick .*/aexec make -j8 test"' git rebase -i
After picking each commit, that will run the tests on each one[1]. If it
fails, the rebase will pause. You can fix any problems, test to your
satisfaction, "commit --amend", and then "rebase --continue" to keep
going.
-Peff
[1] Of course it can be rather time-consuming for a large series. I
often just compile-test at first, and then do a final "make test"
pass when I think everything is right.
--
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html