Re: [PATCH] x86: don't ever patch back to UP if we unplug cpus.

2012-07-29 Thread Rusty Russell
On Fri, 27 Jul 2012 13:28:29 -0700, Suresh Siddha wrote: > if (!noreplace_smp && (num_present_cpus() == 1 || setup_max_cpus <= 1)) > > also, to be consistent with other checks, may be just use > "num_possible_cpus() == 1" check instead of "setup_max_cpus <= 1". Actually, if you specify "m

Re: [PATCH] x86: don't ever patch back to UP if we unplug cpus.

2012-07-29 Thread Rusty Russell
On Fri, 27 Jul 2012 13:28:29 -0700, Suresh Siddha wrote: > On Fri, 2012-07-27 at 17:08 +0930, Rusty Russell wrote: > > Paul McKenney points out: > > > > mean offline overhead is 6251/48=130.2 milliseconds. > > > > If I remove the alternatives_smp_switch() from the offline > > path [...] the

Re: [PATCH] x86: don't ever patch back to UP if we unplug cpus.

2012-07-27 Thread Suresh Siddha
On Fri, 2012-07-27 at 17:08 +0930, Rusty Russell wrote: > Paul McKenney points out: > > mean offline overhead is 6251/48=130.2 milliseconds. > > If I remove the alternatives_smp_switch() from the offline > path [...] the mean offline overhead is 550/42=13.1 milliseconds > > Basically, we're n

[PATCH] x86: don't ever patch back to UP if we unplug cpus.

2012-07-27 Thread Rusty Russell
Paul McKenney points out: mean offline overhead is 6251/48=130.2 milliseconds. If I remove the alternatives_smp_switch() from the offline path [...] the mean offline overhead is 550/42=13.1 milliseconds Basically, we're never going to get those 120ms back, and the code is pretty messy. We ge