Re: rev-list with multiple commits sharing same patch-id

From: Arnaud Morin <arnaud.morin@gmail.com>
Date: Jan 11, 2021, 18:25 UTC
Message-ID: <20210111182501.GB32482@sync>
In-Reply-To: <CAP8UFD3qt4xr4S1BgjPU-vm4qsb=+1JWc=LR_p1CUzTKaAcAoA@mail.gmail.com>

Hi,

Thanks for the link to the test file.

I patched the test file to add this new case (where the same patch-id is duplicated in the tree). This patch can be find in [1] I had to patch two existing tests to make them work again. I also added a new test with this new test-case (two identical patch-id in the same tree.) Note that I still have one test (outside my new one) which is failing:

not ok 10 - name-rev --exclude excludes matched patterns

I dont know why yet.

Moreover, I am willing to help fixing the code. I have the feeling that a patch needs to be done in revision.c, maybe in static void cherry_pick_list?

If you have any clue to help me?

Thanks.

[1] https://github.com/arnaudmorin/git/pull/1

-- 
Arnaud Morin

On 11.01.21 - 10:54, Christian Couder wrote:
> Hi,
> 
> On Sat, Jan 9, 2021 at 5:29 PM Arnaud Morin <arnaud.morin@gmail.com> wrote:
> 
> > I am struggling with a rev-list command, hope someone can give me a
> > clue on what is going on.
> 
> [...]
> 
> > However, when using git rev-list to perform a symmetric difference, git
> > is giving me a commit ID:
> > $ git rev-list --left-right --cherry-pick B...master
> > >dbf86d8aafc897a25a3093139b4237a62395041e
> 
> It looks like a bug. You might want to check the following test file
> and let us know if a similar case is tested in it, or not:
> 
> t6007-rev-list-cherry-pick-file.sh
> 
> If not, it might be a good idea to add one using 'test_expect_failure'
> instead of 'test_expect_success' to document the failure.
> 
> Thanks,
> Christian.
> 
> 
> > Note that this commit is not empty
> > $ git show dbf86d8aafc897a25a3093139b4237a62395041e --stat
> > commit dbf86d8aafc897a25a3093139b4237a62395041e (origin/master, master)
> > Author: Arnaud Morin <hidden@mail>
> > Date:   Sat Jan 9 10:30:10 2021 +0100
> >
> >     remove line2 and add line4 (bis)
> >
> >  a | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> >
> >
> >
> > So, from git rev-list perspective, there is a diff.
> >
> > After digging a little bit, the thing is that this commit is having the
> > following patch-id:
> > $ git show dbf86d8aafc897a25a3093139b4237a62395041e | git patch-id
> > 20f4ace68e80a751b07d78a27c94e83d6c5314bc dbf86d8aafc897a25a3093139b4237a62395041e
> >
> > Which is also already existing in an other commit:
> > $ for c in $(git rev-list HEAD) ; do git show $c |git patch-id |grep 20f4ace68e80a751b07d78a27c94e83d6c5314bc; done
> > 20f4ace68e80a751b07d78a27c94e83d6c5314bc dbf86d8aafc897a25a3093139b4237a62395041e
> > 20f4ace68e80a751b07d78a27c94e83d6c5314bc 8969d3fa9159730fd3b23199873bfb26e3d20027
> >
> > So, is it normal that rev-list is not able to figure out that a commit
> > is existing in both branch when 2 commits share the same patch-id?
> >
> > Is there any way to prevent rev-list from showing this commit?
> >
> > Thanks for your help.
> >
> >
> > PS. I have uploaded my test repo here:
> > https://gitlab.com/arnaudmorin/git-rev-list
> >
> >
> > --
> > Arnaud Morin
> >