Re: [Xen-devel] [PATCH 2/8] x86: distinguish CPU offlining from CPU removal

2018-07-13 Thread Wei Liu
On Thu, Jul 12, 2018 at 05:48:40AM -0600, Jan Beulich wrote: > >>> On 12.07.18 at 12:53, wrote: > > On Wed, Jul 11, 2018 at 06:06:04AM -0600, Jan Beulich wrote: > > [...] > >> --- a/xen/arch/x86/smpboot.c > >> +++ b/xen/arch/x86/smpboot.c > >> @@ -63,6 +63,8 @@ static cpumask_t scratch_cpu0mask;

Re: [Xen-devel] [PATCH 2/8] x86: distinguish CPU offlining from CPU removal

2018-07-12 Thread Andrew Cooper
On 11/07/18 13:06, Jan Beulich wrote: > In order to be able to service #MC on offlined CPUs, GDT, IDT, stack, For clarity, I'd phrase this as "CPUs, the GDT, ..." to help visually separate CPUs as it isn't a part of the rest of the list. > --- a/xen/arch/x86/genapic/x2apic.c > +++

Re: [Xen-devel] [PATCH 2/8] x86: distinguish CPU offlining from CPU removal

2018-07-12 Thread Jan Beulich
>>> On 12.07.18 at 12:53, wrote: > On Wed, Jul 11, 2018 at 06:06:04AM -0600, Jan Beulich wrote: > [...] >> --- a/xen/arch/x86/smpboot.c >> +++ b/xen/arch/x86/smpboot.c >> @@ -63,6 +63,8 @@ static cpumask_t scratch_cpu0mask; >> cpumask_t cpu_online_map __read_mostly; >>

Re: [Xen-devel] [PATCH 2/8] x86: distinguish CPU offlining from CPU removal

2018-07-12 Thread Wei Liu
On Wed, Jul 11, 2018 at 06:06:04AM -0600, Jan Beulich wrote: [...] > --- a/xen/arch/x86/smpboot.c > +++ b/xen/arch/x86/smpboot.c > @@ -63,6 +63,8 @@ static cpumask_t scratch_cpu0mask; > cpumask_t cpu_online_map __read_mostly; > EXPORT_SYMBOL(cpu_online_map); > > +bool __read_mostly

[Xen-devel] [PATCH 2/8] x86: distinguish CPU offlining from CPU removal

2018-07-11 Thread Jan Beulich
In order to be able to service #MC on offlined CPUs, GDT, IDT, stack, and per-CPU data (which includes the TSS) need to be kept allocated. They should only be freed upon CPU removal (which we currently don't support, so some code is becoming effectively dead for the moment). Signed-off-by: Jan