Re: [openstack-dev] Issues with git review for a dependent commit

2013-07-03 Thread Vishvananda Ishaya
On Jul 2, 2013, at 4:23 PM, Jeremy Stanley fu...@yuggoth.org wrote: git review -d 33297 git review -x 35384 git review Oh, I didn't see that you added -x/-X/-N. I can simplify my backport script[1] significantly now. Vish [1] https://gist.github.com/vishvananda/2206428

Re: [openstack-dev] Issues with git review for a dependent commit

2013-07-03 Thread Jeremy Stanley
On 2013-07-03 09:04:04 -0700 (-0700), Vishvananda Ishaya wrote: Oh, I didn't see that you added -x/-X/-N. You in the collective sense at least. Credit for that goes to Miklos Vajna, who had to do a good bit of convincing us it was safe/useful. And now I use it frequently, if fairly carefully,

Re: [openstack-dev] Issues with git review for a dependent commit

2013-07-02 Thread Jeremy Stanley
On 2013-07-02 20:14:35 + (+), Kyle Mestery (kmestery) wrote: I'm trying to submit a gerrit review for a commit which is dependent on another person's commit [1]. [...] ! [remote rejected] HEAD - refs/publish/master/bp/ml2-vxlan (no changes made) [...] I want to say I've seen this

Re: [openstack-dev] Issues with git review for a dependent commit

2013-07-02 Thread Salvatore Orlando
Kyle, is this commit basically a rebase on top of 91e0850? In that case the diff with the previous patchset would be empty. I recall I had a similar issue; I just tweaked a comment line in my commit to let gerrit think it was a different patchset. Salvatore On 2 July 2013 23:05, Kyle Mestery

Re: [openstack-dev] Issues with git review for a dependent commit

2013-07-02 Thread Kyle Mestery (kmestery)
On Jul 2, 2013, at 4:18 PM, Salvatore Orlando sorla...@nicira.com wrote: Kyle, is this commit basically a rebase on top of 91e0850? In that case the diff with the previous patchset would be empty. I recall I had a similar issue; I just tweaked a comment line in my commit to let gerrit

Re: [openstack-dev] Issues with git review for a dependent commit

2013-07-02 Thread Salvatore Orlando
Kyle, I actually meant that this problem might occur if patchset 2, that you're trying to push, is a rebase of patchset 1 on top of another patch in order to make the commit dependent on another one. If that is the case, gerrit won't see any difference between patchset2 and patchset1, as git diff