Re: [PATCH 3/3] ceph: fix a race with rename() in ceph_mdsc_build_path()

[Al Viro <viro@zeniv.linux.org.uk>]

On Tue, Jun 17, 2025 at 10:12:08PM +0000, Viacheslav Dubeyko wrote:
> On Tue, 2025-06-17 at 23:01 +0100, Al Viro wrote:
> > On Tue, Jun 17, 2025 at 06:21:38PM +0000, Viacheslav Dubeyko wrote:
> > 
> > > Tested-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
> > > Reviewed-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
> > 
> > OK, tested-by/reviewed-by applied to commits in that branch, branch
> > force-pushed to the same place
> > (git.kernel.org:/pub/scm/linux/kernel/git/viro/vfs.git work.ceph-d_name-fixes)
> > 
> > Would you prefer to merge it via the ceph tree?  Or I could throw it
> > into my #for-next and push it to Linus come the next window - up to you...
> 
> Frankly speaking, your tree could be the faster way to upstream. However, I can
> push this patch set into the ceph tree for more deeper testing in the internal
> testing infrastructure. But I don't expect any serious issues in the patches
> that could introduce some bugs.
> 
> Ilya,
> 
> What is your opinion on this? Would you prefer to go through the ceph tree?

I can send a pull request to you now just as easily as I could send it to Linus
a month and a half down the road... ;-)  Up to you, guys.