Re: [PATCH] drivers: base: update cpu offline info when do hotplug

[Greg KH <gregkh@linuxfoundation.org>]

On Sun, Oct 26, 2014 at 06:43:11PM -0700, Neil Zhang wrote:
> Greg,
> 
> 
> > -----Original Message-----
> > From: ddstreet@gmail.com [mailto:ddstreet@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 <zhangwm@marvell.com> wrote:
> > > Greg,
> > >
> > >
> > > -----Original Message-----
> > > From: Greg KH [mailto:gregkh@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 <ddstreet@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...

> What's your comments about this patch?

It's one of 1800+ patches that need my comments, please be patient.

greg k-h