Re: [PATCH 1/2] vmsplice: unmap gifted pages for recipient

[Robert Jennings <rcj@linux.vnet.ibm.com>]

* Dave Hansen (dave@sr71.net) wrote:
> On 10/07/2013 01:21 PM, Robert C Jennings wrote:
> >  		spd.partial[page_nr].offset = loff;
> >  		spd.partial[page_nr].len = this_len;
> > +		spd.partial[page_nr].useraddr = index << PAGE_CACHE_SHIFT;
> >  		len -= this_len;
> >  		loff = 0;
> >  		spd.nr_pages++;
> > @@ -656,6 +702,7 @@ ssize_t default_file_splice_read(struct file *in, loff_t *ppos,
> >  		this_len = min_t(size_t, vec[i].iov_len, res);
> >  		spd.partial[i].offset = 0;
> >  		spd.partial[i].len = this_len;
> > +		spd.partial[i].useraddr = (unsigned long)vec[i].iov_base;
> >  		if (!this_len) {
> >  			__free_page(spd.pages[i]);
> >  			spd.pages[i] = NULL;
> > @@ -1475,6 +1522,8 @@ static int get_iovec_page_array(const struct iovec __user *iov,
> >  
> >  			partial[buffers].offset = off;
> >  			partial[buffers].len = plen;
> > +			partial[buffers].useraddr = (unsigned long)base;
> > +			base = (void*)((unsigned long)base + PAGE_SIZE);
> >  
> >  			off = 0;
> >  			len -= plen;
> > diff --git a/include/linux/splice.h b/include/linux/splice.h
> > index 74575cb..56661e3 100644
> > --- a/include/linux/splice.h
> > +++ b/include/linux/splice.h
> > @@ -44,6 +44,7 @@ struct partial_page {
> >  	unsigned int offset;
> >  	unsigned int len;
> >  	unsigned long private;
> > +	unsigned long useraddr;
> >  };
> 
> "useraddr" confuses me.  You make it an 'unsigned long', yet two of the
> three assignments are from "void __user *".  The other assignment:
> 
> 	spd.partial[page_nr].useraddr = index << PAGE_CACHE_SHIFT;
> 
> 'index' looks to be the offset inside the file, not a user address, so
> I'm confused what that is doing.
> 
> Could you elaborate a little more on why 'useraddr' is suddenly needed
> in these patches?  How is "index << PAGE_CACHE_SHIFT" a virtual address?
>  Also, are we losing any of the advantages of sparse checking since
> 'useraddr' is without the __user annotation?
> 

I'm working on cleaning this up.  Trying to remove useraddr altogher
through the use of the existing 'private' field just for the
splice_to_user/pipe_to_user flow without upsetting other uses of the
private field.