On Mon, 2025-03-10 at 14:04 +0200, Dan Aloni wrote: > Signed-off-by: Dan Aloni <dan.aloni@xxxxxxxxxxxx> > --- > utils/mount/nfs.man | 11 +++++++++++ > 1 file changed, 11 insertions(+) > > diff --git a/utils/mount/nfs.man b/utils/mount/nfs.man > index eab4692a87de..744411688641 100644 > --- a/utils/mount/nfs.man > +++ b/utils/mount/nfs.man > @@ -618,6 +618,17 @@ option is not specified, > the default behavior depends on the kernel version, > but is usually equivalent to > .BR "xprtsec=none" . > +.TP 1.5i > +.BI noalignwrite > +This option disables the default behavior of extending write operations to full > +page boundaries. > +.IP > +Normally, the NFS client rounds non-aligned writes up to the system page size, Isn't that only true for buffered writes? That should probably be made clear if so. > +which can lead to "lost writes" when multiple clients write concurrently > +to distinct non-overlapping regions. Use this option when your > +applications perform non-aligned writes and you can guarantee that file > +regions do not overlap, thus avoiding the need for file locking. > +.IP > .SS "Options for NFS versions 2 and 3 only" > Use these options, along with the options in the above subsection, > for NFS versions 2 and 3 only. -- Jeff Layton <jlayton@xxxxxxxxxx>
