Re: [PATCH 18/20] Use FIEMAP for FIBMAP calls

[Christoph Hellwig <hch@lst.de>]

On Mon, Nov 19, 2018 at 01:50:28PM +0100, Carlos Maiolino wrote:
> On Fri, Nov 16, 2018 at 05:06:43PM +0100, Christoph Hellwig wrote:
> > On Tue, Oct 30, 2018 at 02:18:21PM +0100, Carlos Maiolino wrote:
> > > +	if (inode->i_op->fiemap) {
> > > +		fextent.fe_logical = 0;
> > > +		fextent.fe_physical = 0;
> > > +		f_ctx.fc_extents_max = 1;
> > > +		f_ctx.fc_extents_mapped = 0;
> > > +		f_ctx.fc_data = &fextent;
> > > +		f_ctx.fc_start = start;
> > > +		f_ctx.fc_len = 1;
> > > +		f_ctx.fc_flags = 0;
> > > +		f_ctx.fc_cb = fiemap_fill_kernel_extent;
> > > +
> > > +		error = inode->i_op->fiemap(inode, &f_ctx);
> > > +
> > > +		if (error)
> > > +			goto out;
> > > +
> > > +		*block = (fextent.fe_physical +
> > > +			(start - fextent.fe_logical)) >> inode->i_blkbits;
> > > +
> > 
> > I think this code needs to be split into a helper.
> > 
> 
> Yup, my idea is to try to reduce as much as possible the shared code between
> usr/kernel helpers, I just didn't want to spend more time thinking about it
> without having a review of the overall design :P

Well, the code in this branch should not change at all by being moved
into a helper function..