Re: page_launder() bug

[David S. Miller]

Marcelo Tosatti writes:
 > I just thought about this case:
 >   
 > We find a dead swap cache page, so dead_swap_page goes to 1.
 > 
 > We call swap_writepage(), but in the meantime the swapin readahead code   
 > got a reference on the swap map for the page.
 > 
 > We write the page out because "(swap_count(page) > 1)", and we may
 > not have __GFP_IO set in the gfp_mask. Boom.

Hmmm, can't this happen without my patch?

Nothing stops people from getting references to the page
between the "Page is or was in use?" test and the line
which does "TryLockPage(page)".

Later,
David S. Miller
[e-mail]