From 183d03cc4ff39e0f0d952c09aa96d0abfd6e0c3c Mon Sep 17 00:00:00 2001 From: Stefano Stabellini Date: Mon, 17 May 2010 17:08:21 +0100 Subject: xen: Xen PCI platform device driver. Add the xen pci platform device driver that is responsible for initializing the grant table and xenbus in PV on HVM mode. Few changes to xenbus and grant table are necessary to allow the delayed initialization in HVM mode. Grant table needs few additional modifications to work in HVM mode. The Xen PCI platform device raises an irq every time an event has been delivered to us. However these interrupts are only delivered to vcpu 0. The Xen PCI platform interrupt handler calls xen_hvm_evtchn_do_upcall that is a little wrapper around __xen_evtchn_do_upcall, the traditional Xen upcall handler, the very same used with traditional PV guests. When running on HVM the event channel upcall is never called while in progress because it is a normal Linux irq handler (and we cannot switch the irq chip wholesale to the Xen PV ones as we are running QEMU and might have passed in PCI devices), therefore we cannot be sure that evtchn_upcall_pending is 0 when returning. For this reason if evtchn_upcall_pending is set by Xen we need to loop again on the event channels set pending otherwise we might loose some event channel deliveries. Signed-off-by: Stefano Stabellini Signed-off-by: Sheng Yang Signed-off-by: Jeremy Fitzhardinge --- drivers/xen/events.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) (limited to 'drivers/xen/events.c') diff --git a/drivers/xen/events.c b/drivers/xen/events.c index d659480125f..7c64473c9f3 100644 --- a/drivers/xen/events.c +++ b/drivers/xen/events.c @@ -665,7 +665,7 @@ static void __xen_evtchn_do_upcall(void) count = __get_cpu_var(xed_nesting_count); __get_cpu_var(xed_nesting_count) = 0; - } while(count != 1); + } while (count != 1 || vcpu_info->evtchn_upcall_pending); out: @@ -689,6 +689,7 @@ void xen_hvm_evtchn_do_upcall(void) { __xen_evtchn_do_upcall(); } +EXPORT_SYMBOL_GPL(xen_hvm_evtchn_do_upcall); /* Rebind a new event channel to an existing irq. */ void rebind_evtchn_irq(int evtchn, int irq) @@ -725,7 +726,10 @@ static int rebind_irq_to_cpu(unsigned irq, unsigned tcpu) struct evtchn_bind_vcpu bind_vcpu; int evtchn = evtchn_from_irq(irq); - if (!VALID_EVTCHN(evtchn)) + /* events delivered via platform PCI interrupts are always + * routed to vcpu 0 */ + if (!VALID_EVTCHN(evtchn) || + (xen_hvm_domain() && !xen_have_vector_callback)) return -1; /* Send future instances of this interrupt to other vcpu. */ -- cgit v1.2.3-18-g5258