Re: [RFC] high system time & lock contention running large mixedworkload

[Rik van Riel]

On 12/03/2009 05:14 PM, Larry Woodman wrote:

> The attached patch addresses this issue by changing page_check_address()
> to return -1 if the spin_trylock() fails and page_referenced_one() to
> return 1 in that path so the page gets moved back to the active list.

Your patch forgot to add the code to vmscan.c to actually move
the page back to the active list.

Also, please use an enum for the page_referenced return
values, so the code in vmscan.c can use symbolic names.

enum page_reference {
	NOT_REFERENCED,
	REFERENCED,
	LOCK_CONTENDED,
};

--
All rights reversed.
--
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/