On Thu, Apr 28, 2011 at 03:31:55PM +0200, Pavel Machek wrote: > Hi! > > > > For testing swap-over-NBD, a machine was booted with 2G of RAM with a > > swapfile backed by NBD. 16*NUM_CPU processes were started that create > > anonymous memory mappings and read them linearly in a loop. The total > > size of the mappings were 4*PHYSICAL_MEMORY to use swap heavily under > > memory pressure. Without the patches, the machine locks up within > > minutes and runs to completion with them applied. > > > > Comments? > > Nice! > > It is easy to see why swapping needs these fixes, but... dirty memory > writeout is used for memory clearing, too. Are same changes neccessary > to make that safe? > Dirty page limiting covers the MAP_SHARED cases and are already throttled approprately. > (Perhaps raise 'max dirty %' for testing?) Stress testing passed for dirty ratios of 40% at least. Maybe it would cause issues when raised to nearly 100% but I don't think that is a particularly interesting use case. -- Mel Gorman SUSE Labs -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxxx For more info on Linux MM, see: http://www.linux-mm.org/ . Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/ Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>