Re: [PATCH] misc: remove bogus fstest

[Dave Chinner <david@fromorbit.com>]

On Wed, Jul 12, 2023 at 09:19:21AM +0200, Carlos Maiolino wrote:
> On Wed, Jul 12, 2023 at 09:55:33AM +1000, Dave Chinner wrote:
> > On Tue, Jul 11, 2023 at 07:54:41AM -0700, Darrick J. Wong wrote:
> > > On Tue, Jul 11, 2023 at 03:24:54PM +0200, Carlos Maiolino wrote:
> > > > On Sun, Jul 09, 2023 at 03:37:50PM -0700, Darrick J. Wong wrote:
> > > > > From: Darrick J. Wong <djwong@kernel.org>
> > > > >
> > > > > Remove this test, not sure why it was committed...
> > > > >
> > > > > Signed-off-by: Darrick J. Wong <djwong@kernel.org>
> > > > > ---
> > > > >  tests/xfs/999     |   66 -----------------------------------------------------
> > > > >  tests/xfs/999.out |   15 ------------
> > > > >  2 files changed, 81 deletions(-)
> > > > >  delete mode 100755 tests/xfs/999
> > > > >  delete mode 100644 tests/xfs/999.out
> > > >
> > > > Thanks for spotting it. I'm quite sure this was a result of my initial attempts
> > > > of using b4 to retrieve the xfsprogs patch from the list, and it ended up
> > > > retrieving the whole thread which included xfstests patches.
> > > >
> > > > Won't happen again, thanks for the heads up.
> > >
> > > Well I'm glad that /one/ of us now actually knows how to use b4, because
> > > I certainly don't.  Maybe that's why Konstantin or whoever was talking
> > > about how every patch should include a link to a gitbranch or whatever.
> > 
> > If all you want to do is pull stuff from the mailing list, then all
> > you need to know is this command:
> > 
> > 'b4 am -o - <msgid> | git am -s'
> > 
> > This pull the entire series from the thread associated with that
> > msgid into the current branch with all the rvb/sob tags updated. I
> > -think- this has all been rolled up into the newfangled 'b4 shazam'
> > command, but I much prefer to use the original, simple, obvious
> > put-the-pieces-together-yourself approach.
> 
> This was exactly the case, the problem is, both xfstests patch and its xfsprogs
> counterpart were sent under the same thread, which caused b4 to pull both of
> them.
> What I noticed (and haven't until I looked a bit deeper during my PTO) is that
> b4 has an option to pull the patches into quilt format, so, that will make
> things way easier.

I don't use that - I prefer to work from commits than mange patches
directly. What I do take the list of commits that it creates, then
run 'guilt import-commit <id range>' to pull the commits made from
the patches into a guilt maintained patch series in the git
repository. That generally requires checking that all the commits
that were pulled in were the ones that were wanted, but then I can
manage the branch as a "patch series in git commits" as needed.
That's far easier than managing patches by hand to build a series to
apply to the git tree....

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com