On Fri, Sep 08, 2023 at 02:09:51AM -0700, Christoph Hellwig wrote: > On Thu, Sep 07, 2023 at 06:47:03PM +0100, Matthew Wilcox (Oracle) wrote: > > This avoids using the mr_writer field to check the XFS ILOCK is held > > for write. It also improves the checking we do when lockdep is disabled. > > Hmm, is there any reason to keep the lockdep-based version of > __xfs_rwsem_islocked around now at all? Ok, looks like the last patch fixes that.
