swapin readahead and locking

swapin readahead and locking

[Rik van Riel]

Hi Stephen,

I am now using the following construction (well, since
09:15 this morning):

struct page *page_map = lookup_swap_cache(entry);

if (!page_map) {
	page_map = read_swap_cache(entry);

... do readahead stuff
}

I have a funny feeling I missed a wait_on_page() this way,
but things are runnig happily right now. If I missed a big
thing like this, please let me know -- I'd hate it to have
people test the patch (it is up for grabs) and end up with
a corrupted system...

thanks,

Rik -- now completely used to dvorak kbd layout...
+-------------------------------------------------------------------+
| Linux memory management tour guide.        H.H.vanRiel@phys.uu.nl |
| Scouting Vries cubscout leader.      http://www.phys.uu.nl/~riel/ |
+-------------------------------------------------------------------+

--
This is a majordomo managed list.  To unsubscribe, send a message with
the body 'unsubscribe linux-mm me@address' to: majordomo@kvack.org

Re: swapin readahead and locking

[Stephen C. Tweedie]

Hi,

On Tue, 1 Dec 1998 17:12:45 +0100 (CET), Rik van Riel
<H.H.vanRiel@phys.uu.nl> said:

> struct page *page_map = lookup_swap_cache(entry);

> if (!page_map) {
> 	page_map = read_swap_cache(entry);

> ... do readahead stuff
> }

read_swap_cache() is not asynchronous!  include/linux/swap.h:

	#define read_swap_cache(entry) read_swap_cache_async(entry, 1);

I think you were on the right lines before (except for the missing
free_page()). 

> I have a funny feeling I missed a wait_on_page() this way,
> but things are runnig happily right now. 

No, read_swap_cache automatically waits, and if you do the async
version, then any later call to read the same page will in turn wait for
the IO to complete.

--Stephen
--
This is a majordomo managed list.  To unsubscribe, send a message with
the body 'unsubscribe linux-mm me@address' to: majordomo@kvack.org