Re: [PATCH 1/3] mm: tracking shared dirty pages

[Christoph Lameter]

On Tue, 30 May 2006, David Howells wrote:

> Christoph Lameter <clameter@xxxxxxx> wrote:
> 
> > On Tue, 30 May 2006, David Howells wrote:
> > 
> > > > If set_page_dirty cannot reserve the page then we know that some severe
> > > > action is required. The FS method set_page_dirty() could:
> > > 
> > > But by the time set_page_dirty() is called, it's too late as the code
> > > currently stands.  We've already marked the PTE writable and dirty.  The
> > > page_mkwrite() op is called _first_.
> > 
> > We are in set_page_dirty and this would be part of set_page_dirty 
> > processing.
> 
> Eh?  What do you mean "We are in set_page_dirty"?

We could do the reservation in as part of the set_page_dirty FS method.

> Actually, I'm not sure that calling set_page_dirty() at the bottom of
> do_wp_page() is necessarily a good idea.  It's possible that the page will be
> marked dirty in do_wp_page() and then will get written back before the write
> actually succeeds.  In other words the page may be marked dirty and cleaned up
> all before the modification _actually_ occurs.  On the other hand, the common
> case is probably that the store instruction will beat the writeback.

Yes we are aware of that case.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/