> -----Original Message----- > From: Greg KH [mailto:gre...@linuxfoundation.org] > Sent: 2014年10月27日 9:59 > To: Neil Zhang > Cc: Dan Streetman; linux-kernel@vger.kernel.org > Subject: Re: [PATCH] drivers: base: update cpu offline info when do hotplug > > On Sun, Oct 26, 2014 at 06:43:11PM -0700, Neil Zhang wrote: > > Greg, > > > > > > > -----Original Message----- > > > From: ddstr...@gmail.com [mailto:ddstr...@gmail.com] On Behalf Of > > > Dan Streetman > > > Sent: 2014年10月21日 1:03 > > > To: Neil Zhang > > > Cc: Greg KH; linux-kernel@vger.kernel.org > > > Subject: Re: [PATCH] drivers: base: update cpu offline info when do > > > hotplug > > > > > > On Mon, Oct 20, 2014 at 3:40 AM, Neil Zhang <zhan...@marvell.com> wrote: > > > > Greg, > > > > > > > > > > > > -----Original Message----- > > > > From: Greg KH [mailto:gre...@linuxfoundation.org] > > > > Sent: 2014年10月20日 14:48 > > > > To: Neil Zhang > > > > Cc: linux-kernel@vger.kernel.org > > > > Subject: Re: [PATCH] drivers: base: update cpu offline info when > > > > do hotplug > > > > > > > > On Sun, Oct 19, 2014 at 11:39:23PM -0700, Neil Zhang wrote: > > > >>> How much noise is this going to cause on a big/little system > > > >>> that constantly hot unplug/plugs processors all of the time? > > > >> > > > >> Can you explain more what kind of noise will be introduced on a > > > >> big/little > > > system? > > > > > > > > Have you tested this on such a machine? > > > > > > > > I didn't have such kind of machine on hand. > > > > Can anyone has such machine to verify it? > > > > Thanks! > > > > > > I tested this on a ppc PowerVM system, using dlpar operations to > > > remove/add cpus. > > > > > > Without this patch the cpu online nodes get out of sync with the > > > main online node (and the actual state of the cpus), because they > > > aren't updated as the cpus are brought up/down: > > > > > > [root@br10p02 cpu]$ pwd > > > /sys/devices/system/cpu > > > [root@br10p02 cpu]$ cat online > > > 0-39 > > > [root@br10p02 cpu]$ for n in {0..47} ; do test $( cat cpu$n/online ) > > > -eq 1 && echo -n "$n " ; done ; echo "" > > > 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 > > > 26 > > > 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 > > > > > > > > > While with the patch, the cpu online nodes are kept up to date as > > > the cpus are brought up/down: > > > > > > [root@br10p02 cpu]$ pwd > > > /sys/devices/system/cpu > > > [root@br10p02 cpu]$ cat online > > > 0-39 > > > [root@br10p02 cpu]$ for n in {0..47} ; do test $( cat cpu$n/online ) > > > -eq 1 && echo -n "$n " ; done ; echo "" > > > 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 > > > 26 > > > 27 28 29 30 31 32 33 34 35 36 37 38 39 > > > > > > > > > Feel free to add > > > > > > Tested-by: Dan Streetman <ddstr...@ieee.org> > > > > > > > It's a real bug in the kernel. > > As this has been this way for many years, I tend to think it's not all that > important...
Actually this bug was introduced by the following patch. commit 0902a9044fa5b7a0456ea4daacec2c2b3189ba8c Author: Rafael J. Wysocki <rafael.j.wyso...@intel.com> Date: Fri May 3 00:25:49 2013 +0200 Driver core: Use generic offline/online for CPU offline/online So seems not that long :) > > > What's your comments about this patch? > > It's one of 1800+ patches that need my comments, please be patient. Got it, thanks! > > greg k-h Best Regards, Neil Zhang