Re: [RFC][PATCH 10/13] make dump_emit() use vfs_write() instead of banging at ->f_op->write directly

[Al Viro <viro@ZenIV.linux.org.uk>]

On Tue, Oct 08, 2013 at 05:52:42PM -0700, Linus Torvalds wrote:
> On Tue, Oct 8, 2013 at 5:15 PM, Al Viro <viro@ftp.linux.org.uk> wrote:
> >
> > ... and deal with short writes properly
> 
> .. except you don't.
> 
> > +       while (nr) {
> > +               if (dump_interrupted())
> > +                       return 0;
> > +               n = vfs_write(file, addr, nr, &pos);
> > +               if (n < 0)
> > +                       return 0;
> > +               file->f_pos = pos;
> > +               cprm->written += n;
> > +               nr -= n;
> > +       }
> 
> Please handle 'n == 0' too. Maybe it never happens (ie you get EPIPE
> or ENOSPC), but write returning zero is actually possible and a valid
> return value and traditional for "end of media". Looping forever is
> not a good idea.

Point, but I would argue that we should yell very loud if we get 0 from
vfs_write() for non-zero size.  I'm not sure if POSIX allows write(2)
to return that, but a lot of userland code won't be expecting that and
won't be able to cope...