why do we call schedule() instead of interruptible_sleep_on(kswapd_wait) ?
My problem that lead to probe into this is I have some filesystem that
currently resides in swap. As I was stress testing I go to the point
where my memory mostly filled up. About one page per free list, and
the system slowed to a crawl. And it stayed in a crawl for a long
time. This was realtively easy to trigger.
I modified vmscan to use interruptible_sleep_on and now I can't
trigger it. Plus that seems to be what the documentation think it is
doing.
I'm not certain if the problem was kswapd running too much for
anything else to happen or kswapd not running enough (in which case
it sounds like a problem in realtime priorites or the way kswapd sets
itself to realtime priorities).
Eric
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu