Re: [PATCH 08/16] block: get file-win32.c handle locking option consistence with file-posix.c

[Kevin Wolf]

Am 09.09.2020 um 10:41 hat 罗勇刚(Yonggang Luo) geschrieben:
> On Wed, Sep 9, 2020 at 4:19 PM Kevin Wolf <kwolf@redhat.com> wrote:
> 
> > Am 08.09.2020 um 21:48 hat Yonggang Luo geschrieben:
> > > Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
> > > ---
> > >  block/file-win32.c | 23 +++++++++++++++++++++--
> > >  1 file changed, 21 insertions(+), 2 deletions(-)
> >
> > This is almost the same as my separately posted 'file-win32: Fix
> > "locking" option', except that you changed the order of variable
> > definitions which will cause a merge conflict.
> >
> > When you take patches from other people, you should correctly attribute
> > them and not make them look as if they were your own.
> >
> > Commit messages should not be empty, but explain the motivation for the
> > change.
> >
> I know someone else have already have fixes for it, so I didn't detail
> it much, Ineed I wanna skip it when sending patches, but
> git-publish.py didn't handle it well under Win32 yet
> 
> >
> > In this case, dropping the patch is easiest because my patch is already
> > in a pull request and will probably be merged soon.
>
> I am waiting it to merged, at the current stage, if I don't include
> this patch the test runner will fail.

I see. This is not obvious to others and it will look like you want to
get the patch included as you sent it.

What I do in situations like this is that I put any patches from other
series first in my branch and then send out only what is on top of them,
noting the dependency in the cover letter.

Kevin