Re: page->mapping and page->buffers

[Rik van Riel <riel@conectiva.com.br>]

On Mon, 9 Sep 2002, Haijin Yan wrote:

> Are you guys saying that pages for traditional I/O(read/write) do not
> have page->mapping while clean mmap()ed pages do not have page->buffer?

No. Any pagecache page has page->mapping and page->index, no
matter whether it's used for read/write or for mmap.

These pages will have page->buffers when we're doing _disk_
IO on them, buffer heads are the way the kernel tells the
block IO subsystem what should be done with a page.

> so when is the condition (!page->mapping) true and what it represents for?

It can happen in 3 situations:

1) a new anonymous page is allocated to the process, this
   is a private page that is only visible in the process
   page tables

2) a page _was_ part of a page cache mapping, but truncate()
   removed it from the page cache ... at the same time, a
   process was doing a pagefault on this page and added it
   to its page tables, now the page is converted into anonymous
   memory

3) a page was swapped in and removed from the swap cache in
   order to reclaim swap space.

> If page->buffers is true, Does that ony mean either page is brought in
> by traditional IO or a disk block is allocated for this mmap()ed page
> for written(because page is dirty)?

Yes, but note that buffer heads can stay around after disk IO
has finished.

regards,

Rik
-- 
Spamtrap of the month: september@surriel.com

http://www.surriel.com/		http://distro.conectiva.com/

--
Kernelnewbies: Help each other learn about the Linux kernel.
Archive:       http://mail.nl.linux.org/kernelnewbies/
FAQ:           http://kernelnewbies.org/faq/