Re: [PATCH 2/3] Improve fairness when locking the per-superblock s_anon list

[Christoph Hellwig]

On Thu, Nov 09, 2017 at 11:52:48AM -0800, Linus Torvalds wrote:
> Yes, getting rid of s_anon might involve crazy things like "let's just
> walk all the dentries at umount time", but honestly, that sounds
> preferable. Especially if we can just then do something like
> 
>  - set a special flag in the superblock if we ever use __d_obtain_alias()
> 
>  - only scan all the dentries on umount if that flag is set.
> 
> Hmm?

The scan would be extremely painful with our current global dcache.

But then again the global dcache isn't exactly a good idea to start
with.  If we had a per-sb dcache using e.g. the resizable hash table
the networking people seem to be fond off your idea would work really
well.