[sheepdog] [RFC v5 000/126] error: auto propagated local_err

[Cornelia Huck]

On Fri, 8 Nov 2019 22:57:25 +0400
Marc-André Lureau <marcandre.lureau at redhat.com> wrote:

> Hi
> 
> On Fri, Nov 8, 2019 at 7:31 PM Vladimir Sementsov-Ogievskiy
> <vsementsov at virtuozzo.com> wrote:
> >
> > Finally, what is the plan?
> >
> > Markus what do you think?
> >
> > Now a lot of patches are reviewed, but a lot of are not.
> >
> > Is there any hope that all patches will be reviewed? Should I resend the
> > whole series, or may be reduce it to reviewed subsystems only?  
> 
> I don't think we have well established rules for whole-tree cleanups
> like this. In the past, several cleanup series got lost.

Yes, it is always problematic if a series touches a lot of different
subsystems.

> 
> It will take ages to get every subsystem maintainer to review the
> patches. Most likely, since they are quite systematic, there isn't
> much to say and it is easy to miss something that has some hidden
> ramifications. Perhaps whole-tree cleanups should require at least 2
> reviewers to bypass the subsytem maintainer review? But my past
> experience with this kind of exercice doesn't encourage me, and
> probably I am not the only one.

It's not just the reviews; it's easy to miss compile problems on less
mainstream architectures (and even easier to miss functional problems
there, although they are probably less likely with automated rework.)
CI can probably help, but that's something for the future.

Anyway, I've now gotten around to that series; spotted one problem in
s390x code, I think.

One thing that's helpful for such a large series is a git branch that
makes it easy to give the series a quick go. (You can use patchew, but
it takes time before it gets all mails, so just pushing it somewhere
and letting people know is a good idea anyway.)