Re: [PATCH 1/2] fts: introduce the FTS_NOLEAF flag

[Kamil Dudka]

On Monday 21 December 2015 00:05:51 Pádraig Brady wrote:
> On the other hand we've had no reports of issues with the
> existing auto config done by FTS.

Because the leaf optimization has been enabled for reiserfs only until now.
I suspect that NFS will not be that easy and such file system issues might
be difficult to debug without the -noleaf option actually implemented.

> In my experience one
> should jump through hoops to avoid config where possible.
> Until we see a case where dirent.d_type is accurate while
> noleaf is not appropriate, then I'm inclined to avoid the config.
> If this was an issue then there would need to be associated
> -noleaf patches to find(1), and also other FTS using tools.

That sounds like a reasonable approach for tools that do not implement any
-noleaf option yet.  However, GNU find has provided this option since 1996
(if not even longer) and the option has never been properly deprecated.

> I've applied the patch to ensure noleaf optimization is
> enabled for NFS, and also applied my patch enabling it for XFS.
> 
> thanks,
> Pádraig.

Thank you for applying those patches!

Kamil