lkml.org 
[lkml]   [1999]   [Oct]   [10]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: locking question: do_mmap(), do_munmap()
Alexander Viro wrote:
> Hold on. In swap_out_mm() you have to protect find_vma() (OK, it doesn't
> block, but we'll have to take care of mm->mmap_cache) _and_ you'll have to
> protect vma from destruction all way down to try_to_swap_out(). And to
> vma->swapout(). Which can sleep, so spinlocks are out of question here.

I found vma->swapout() when I tried to implement it. Sh...
We could make vma_list_lock a semaphore, but I haven't checked for any
hidden problems yet.

>
> I still think that just keeping a cyclic list of pages, grabbing from that
> list before taking mmap_sem _if_ we have a chance for blocking
> __get_free_page(), refilling if the list is empty (prior to down()) and
> returning the page into the list if we didn't use it may be the simplest
> way.

I don't like the idea, but it sounds possible.
A problem could be that the page-in functions can allocate memory:
do_nopage() -> filemap_nopage(): it calls i_op->readpage() which would
call get_block(), eg ext2: load the indirect page, this needs memory -->
OOM.

--
Manfred

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

\
 
 \ /
  Last update: 2005-03-22 13:54    [W:0.055 / U:2.256 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site