Re: [PATCH] Revert series: virtiofsd: Announce submounts to the guest

["Dr. David Alan Gilbert" <dgilbert@redhat.com>]

* Max Reitz (mreitz@redhat.com) wrote:
> On 28.10.20 08:55, Stefan Hajnoczi wrote:
> > On Tue, Oct 27, 2020 at 08:13:28PM -0600, Alex Williamson wrote:
> >> This reverts the following commits due to their basis on a bogus
> >> linux kernel header update:
> > 
> > What was the issue? Did the sub-mounts series include kernel header
> > changes that haven't landed in Linux yet?
> 
> Yes.
> 
> The kernel series was applied, however, Miklos fixed some things in the
> process, and then Dave applied the virtiofsd series assuming that it
> would be OK now that the kernel series is in.
> 
> So what could have gone better is e.g.:
> 
> (1) I could have marked the virtiofsd series as RFC until I know the
> definitive state of the kernel series.
> 
> (2) I could have given an explicit NAK to the virtiofsd series once I
> noted that I need to write a v2, and not bury that somewhere deep in the
> kernel side thread (https://marc.info/?l=fuse-devel&m=160087601428180).

Yeh that one would have been useful.
Unfortunately Miklos's changes were such that it didn't break your code;
I did give it a smoke test with 5.10-rc1 and saw that it was doing the submounts OK.

Dave

> (3) I should have noticed Dave applying the virtiofsd series (Monday
> evening) and intervened.  But my inbox full from PTO plus trying to
> perhaps get the FUSE export series done for 5.2 led to me just not noticing.
> 
> Max
> 



-- 
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK