OK, also fine to me! Attached the updated patch. Thanks Xiantao Avi Kivity wrote: > Zhang, Xiantao wrote: >> From d53dfacedea40213382694229700d02bfa72d923 Mon Sep 17 00:00:00 >> 2001 From: Xiantao Zhang <xiantao.zhang@xxxxxxxxx> >> Date: Tue, 19 Aug 2008 17:33:21 +0800 >> Subject: [PATCH] KVM: KVM/IA-64: Fix kvm/ia64's build issue. >> >> Disable irq ack notification for ia64. >> Signed-off-by: Xiantao Zhang <xiantao.zhang@xxxxxxxxx> --- >> virt/kvm/ioapic.c | 7 ++++++- >> 1 files changed, 6 insertions(+), 1 deletions(-) >> >> diff --git a/virt/kvm/ioapic.c b/virt/kvm/ioapic.c >> index 515cd7c..992199b 100644 >> --- a/virt/kvm/ioapic.c >> +++ b/virt/kvm/ioapic.c >> @@ -39,7 +39,10 @@ >> >> #include "ioapic.h" >> #include "lapic.h" >> + >> +#ifdef CONFIG_X86 >> #include "irq.h" >> +#endif >> > > Please keep the include; add an ia64 specific irq.h instead. > >> >> #if 0 >> #define ioapic_debug(fmt,arg...) printk(KERN_WARNING fmt,##arg) >> @@ -293,7 +296,9 @@ static void __kvm_ioapic_update_eoi(struct >> kvm_ioapic *ioapic, int gsi, >> >> ent = &ioapic->redirtbl[gsi]; >> >> +#ifdef CONFIG_X86 >> kvm_notify_acked_irq(ioapic->kvm, gsi); >> +#endif >> >> > > Keep this as well. You can have a nop-op irq notification API, and > later merge with x86. I think you will need irq ack notification if > you implement device assignment, or for improved timekeeping.
Attachment:
0001-KVM-kvm-ia64-Leave-notify_acked_irq-as-blank-funt.patch
Description: 0001-KVM-kvm-ia64-Leave-notify_acked_irq-as-blank-funt.patch
