Re: sys_exit() and zap_page_range()

[Manfred Spraul <manfred@colorfullife.com>]

Andrew Morton wrote:
> 
> Can anyone suggest a simple, clean way of decreasing zap_page_range's
> scheduling latency, in a way which you're prepared to support?
> 
Btw, zap_page_range() contains a huge TLB flush race: the freed pages
become available immediately to other processes, but stale tlb entries
are only flushed when zap_page_range() returns [check madvise_dontneed,
and several other functions]

A proper fix would be a major change, probably along Kanoj's proposal
(pte_freeze_range).

--
	Manfred
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux.eu.org/Linux-MM/