On Tuesday 21 September 2004 06:21 pm, Pete Zaitcev wrote: > > Don't forget that the atomic pool has to be maintained by kswapd. > One easy way to deplete it is to starve kswapd out of CPU. > > Just to make for excitement, writing to a drive dirties pages, > and then you need an extra thread to run (usb-storage worker). > > I'm far from being a VM expert, but AFAIK our write throttling > is governed by the memory pressure alone. Writers sleep only when they > cannot dirty more pages. Now imagine for a moment that you deplete > atomics for a moment, then free them. If writers steal these pages > from kswapd and dirty them, the atomic pool will never recover.
If this were the case, should memory recover once the file transfer is complete and all USB activity has ceased? Kyle. ------------------------------------------------------- This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 Project Admins to receive an Apple iPod Mini FREE for your judgement on who ports your project to Linux PPC the best. Sponsored by IBM. Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
