NMI watchdog sets APIC_LVTPC register to generate an NMI when PMU counter overflow occurs. This may be overwritten by VPMU code later, effectively turning off the watchdog.
We should disable VPMU when NMI watchdog is running. Signed-off-by: Boris Ostrovsky <boris.ostrov...@oracle.com> --- docs/misc/xen-command-line.markdown | 2 ++ xen/arch/x86/hvm/vpmu.c | 13 +++++++++++++ 2 files changed, 15 insertions(+) diff --git a/docs/misc/xen-command-line.markdown b/docs/misc/xen-command-line.markdown index 2274e74..bc316be 100644 --- a/docs/misc/xen-command-line.markdown +++ b/docs/misc/xen-command-line.markdown @@ -1346,6 +1346,8 @@ wrong behaviour (see handle\_pmc\_quirk()). If 'vpmu=bts' is specified the virtualisation of the Branch Trace Store (BTS) feature is switched on on Intel processors supporting this feature. +Note that if **watchdog** option is also specified vpmu will be turned off. + *Warning:* As the BTS virtualisation is not 100% safe and because of the nehalem quirk don't use the vpmu flag on production systems with Intel cpus! diff --git a/xen/arch/x86/hvm/vpmu.c b/xen/arch/x86/hvm/vpmu.c index 63b2158..b2e8e65 100644 --- a/xen/arch/x86/hvm/vpmu.c +++ b/xen/arch/x86/hvm/vpmu.c @@ -24,6 +24,7 @@ #include <asm/regs.h> #include <asm/types.h> #include <asm/msr.h> +#include <asm/nmi.h> #include <asm/hvm/support.h> #include <asm/hvm/vmx/vmx.h> #include <asm/hvm/vmx/vmcs.h> @@ -288,3 +289,15 @@ void vpmu_dump(struct vcpu *v) vpmu->arch_vpmu_ops->arch_vpmu_dump(v); } +static int __init vpmu_init(void) +{ + /* NMI watchdog uses LVTPC and HW counter */ + if ( opt_watchdog && opt_vpmu_enabled ) + { + printk(XENLOG_WARNING "NMI watchdog is enabled. Turning VPMU off.\n"); + opt_vpmu_enabled = 0; + } + + return 0; +} +__initcall(vpmu_init); -- 1.8.1.4 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xen.org http://lists.xen.org/xen-devel