On Sun, 2014-01-05 at 20:00 +0800, Alex Shi wrote:
Tixy, The patchset was pushed to lsk-test branch for testing now. Only 2 commits are new for lsk:
When I do a merge I get the 5 commits mentioned in my pull request, these are the five commits which are new since the last time gator was pulled into LSK. The result you have in lsk-test has different SHA1's for the two commits you have and history is lost, it's almost as if you have managed to do some sort of rebase. I am just cloning a fresh repo to see if I can work out if pull behaves different from merge
Nope, when I pull the gator branch I get the 5 commits merged in, see: https://git.linaro.org/people/tixy/kernel.git/shortlog/refs/heads/lsk-test2
I did this with:
git clone git://git.linaro.org/kernel/linux-linaro-stable.git cd linux-linaro-stable git checkout 3c6556b1a9c2f9c91aacbc57bac938691fedda81 # LSK before new gator git pull git://git.linaro.org/landing-teams/working/arm/kernel.git lsk-3.10-gator
so I guess if you get different it must be a git config issue.
yes, my git pull 'rebase = true' generate some conflicts.
OK, I won't bother trying to keep a clean history of changes at my end then if they're going to get lost after merge anyway.
[...]
BTW, Tixy, I saw above commit appeared 3 times in landing-teams tree. Do you have some suggestion to reduce above redundant-commit when we do pull?
If for LSK I stop trying to keep the upstream history of the gator code then I will just give you a single patch next time. And I might just drop that patch anyway, it was needed to get Pandaboard working, but I assume that doesn't matter for LSK anyway?