Re: [ndctl PATCH v2 2/3] ndctl: add a BTT check utility

["Verma, Vishal L" <vishal.l.verma-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>]

On Wed, 2017-03-01 at 07:54 -0800, Dan Williams wrote:
> On Wed, Mar 1, 2017 at 7:02 AM, Jeff Moyer <jmoyer@redhat.com> wrote:
> > "Verma, Vishal L" <vishal.l.verma@intel.com> writes:
> > 
> > > > - Differentiate the use off BTT_PG_SIZE, sysconf(_SC_PAGESIZE),
> > > > and
> > > > SZ_4K
> > > >   (for the fixed start offset) in the different places they're
> > > > used
> > > > (Jeff)
> > > > 
> > > > I think you could go even further with this.  For example, you
> > > > coud
> > > > use
> > > > BTT_START_OFFSET for the initial offset from the beginning of
> > > > the raw
> > > > namespace to the first btt info block.  Using the page size for
> > > > the
> > > > size
> > > > of the info block is also less informative than it could
> > > > be.  Instead,
> > > > consider adding BTT_INFO_SIZE.  After that, all of your math
> > > > starts to
> > > > look a bit more readable to my eye.
> > > 
> > > I thought of something along these lines, but ended up not doing
> > > this to
> > > stay consistent with the 'BTT_PG_SIZE' usage in the kernel. But
> > > maybe
> > > the kernel could use some cleaning up in this regard :)
> > 
> > I obviously think it would look nicer.  ;-)  Anyway, it's not a
> > blocker
> > to get this series in.
> 
> So, my feedback was to store the offset in the bttc context so it's
> just a property we carry and add unconditionally, it just might be
> zero sometimes. Does that make things any cleaner?

That's a good idea - I'll work on that.
_______________________________________________
Linux-nvdimm mailing list
Linux-nvdimm@lists.01.org
https://lists.01.org/mailman/listinfo/linux-nvdimm