Re: [PATCH kernfs 1/3] kernfs: switch global kernfs_idr_lock to per-fs lock

From: Greg KH
Date: Tue Apr 15 2025 - 11:17:00 EST


On Mon, Apr 14, 2025 at 11:20:54AM +0800, Jinliang Zheng wrote:
> > On Sat, Apr 12, 2025 at 07:50:54PM +0800, alexjlzheng@xxxxxxxxx wrote:
> > > On Sat, 12 Apr 2025 08:12:22 +0200, gregkh@xxxxxxxxxxxxxxxxxxx wrote:
> > > > On Sat, Apr 13, 2025 at 02:31:07AM +0800, alexjlzheng@xxxxxxxxx wrote:
> > > > > From: Jinliang Zheng <alexjlzheng@xxxxxxxxxxx>
> > > > >
> > > > > The kernfs implementation has big lock granularity(kernfs_idr_lock) so
> > > > > every kernfs-based(e.g., sysfs, cgroup) fs are able to compete the lock.
> > > > >
> > > > > This patch switches the global kernfs_idr_lock to per-fs lock, which
> > > > > put the spinlock into kernfs_root.
> > > > >
> > > > > Signed-off-by: Jinliang Zheng <alexjlzheng@xxxxxxxxxxx>
> > > > > ---
> > > > > fs/kernfs/dir.c | 14 +++++++-------
> > > > > fs/kernfs/kernfs-internal.h | 1 +
> > > > > 2 files changed, 8 insertions(+), 7 deletions(-)
> > > >
> > > > What kind of testing / benchmark did you do for this series that shows
> > > > that this works, AND that this actually is measureable? What workload
> > > > are you doing that causes these changes to be needed?
> > >
> > > Thank you for your reply. :)
> > >
> > > We are trying to implement a kernfs-based filesystem that will have
> > > multiple instances running at the same time, i.e., multiple kernfs_roots.
> >
> > I don't think that kernfs is meant for that very well, what is that
> > filesystem going to be for?
>
> Thank you for your reply. :)
>
> Similar to cgroupfs and sysfs, it is used to export the status and configurations
> of some kernel variables in hierarchical modes of the kernel. The only difference
> is that it may have many instances, that is, many kernfs_roots.

Let's see that filesystem first please before determining more, as you
would be adding a new user/kernel api that we all need to argue about :)

Anyway, for the 2 patches that Tejun agrees with here, can you resend
just them?

thanks,

greg k-h