Paolo Bonzini
2011-Mar-03 11:45 UTC
[Xen-devel] Xen not running the pmtimer''s timer on VCPUs different from VCPU#0
Hi, in order to support kdump/kexec on multiprocessor HVM guests, the patch at http://xenbits.xensource.com/xen-unstable.hg/rev/19907 was committed. The patch migrates all timers (via the pt infrastructure) to a VCPU whose LAPIC is configured to receive the legacy 8259 interrupts. In doing this, however, the pmtimer is left out. Its timer won''t ever move away from VCPU#0. Is there a reason for this, or is it just an oversight? Thanks in advance, Paolo _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Keir Fraser
2011-Mar-03 13:17 UTC
Re: [Xen-devel] Xen not running the pmtimer''s timer on VCPUs different from VCPU#0
On 03/03/2011 11:45, "Paolo Bonzini" <pbonzini@redhat.com> wrote:> Hi, > > in order to support kdump/kexec on multiprocessor HVM guests, the patch at > http://xenbits.xensource.com/xen-unstable.hg/rev/19907 was committed. > The patch migrates all timers (via the pt infrastructure) to a VCPU whose > LAPIC is configured to receive the legacy 8259 interrupts. > > In doing this, however, the pmtimer is left out. Its timer won''t ever > move away from VCPU#0. Is there a reason for this, or is it just an > oversight?The pmtimer emulation isn''t hooked into the vpt infrastructure, hence I don''t think its essential for it to be using pt_global_vcpu_target(). That said it might still be a good for consistency. Something to consider after 4.1.0 is released, however. -- Keir> Thanks in advance, > > Paolo > > _______________________________________________ > Xen-devel mailing list > Xen-devel@lists.xensource.com > http://lists.xensource.com/xen-devel_______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel