Re: New driver "sfc" for Solarstorm SFC4000 controller.

[Ben Hutchings <bhutchings@solarflare.com>]

Ben Hutchings wrote:
> Andrew Morton wrote:
> > On Fri, 2 May 2008 17:05:35 +0100 Ben Hutchings <bhutchings@solarflare.com> wrote:
<snip>
> > > > > +		rx_buf->page = alloc_pages(__GFP_COLD | __GFP_COMP | GFP_ATOMIC,
> > > > > +					   efx->rx_buffer_order);
> > > > 
> > > > I don't think we should be using the open-coded __GFP_COMP here.  That's
> > > > more an mm-internal thing.
> > > 
> > > What's the alternative?
> > 
> > Just remove the __GFP_COMP, I expect.  __GFP_COMP will ask the page
> > allocator to add extra book-keeping info to the pageframe (via
> > prep_compound_page()).  I doubt if the driver uses that information.
> 
> It looks like this flag was used by mistake, based on past experience with
> a driver that needed buffers to be mapped into user space.  The original
> author of the RX page-allocation code is away at the moment, so I will
> have to wait to ask why he used it.

This is not a mistake.   We need multi-page buffers to be treated as
compound pages because once they are attached to an skbuff they will
be freed by skb_release_data() calling put_page(), not __free_pages().

Ben.

-- 
Ben Hutchings, Senior Software Engineer, Solarflare Communications
Not speaking for my employer; that's the marketing department's job.