This fixes the following OOPS:
loaded kvm module (
v3.17-rc1-168-gcec26bc)
BUG: unable to handle kernel paging request at
fffffffffffffffe
IP: [<
ffffffff81168449>] put_page+0x9/0x30
PGD
1e15067 PUD
1e17067 PMD 0
Oops: 0000 [#1] PREEMPT SMP
[<
ffffffffa063271d>] ? kvm_vcpu_reload_apic_access_page+0x5d/0x70 [kvm]
[<
ffffffffa013b6db>] vmx_vcpu_reset+0x21b/0x470 [kvm_intel]
[<
ffffffffa0658816>] ? kvm_pmu_reset+0x76/0xb0 [kvm]
[<
ffffffffa064032a>] kvm_vcpu_reset+0x15a/0x1b0 [kvm]
[<
ffffffffa06403ac>] kvm_arch_vcpu_setup+0x2c/0x50 [kvm]
[<
ffffffffa062e540>] kvm_vm_ioctl+0x200/0x780 [kvm]
[<
ffffffff81212170>] do_vfs_ioctl+0x2d0/0x4b0
[<
ffffffff8108bd99>] ? __mmdrop+0x69/0xb0
[<
ffffffff812123d1>] SyS_ioctl+0x81/0xa0
[<
ffffffff8112a6f6>] ? __audit_syscall_exit+0x1f6/0x2a0
[<
ffffffff817229e9>] system_call_fastpath+0x16/0x1b
Code: c6 78 ce a3 81 4c 89 e7 e8 d9 80 ff ff 0f 0b 4c 89 e7 e8 8f f6 ff ff e9 fa fe ff ff 66 2e 0f 1f 84 00 00 00 00 00 66 66 66 66 90 <48> f7 07 00 c0 00 00 55 48 89 e5 75 1e 8b 47 1c 85 c0 74 27 f0
RIP [<
ffffffff81193045>] put_page+0x5/0x50
when not using the in-kernel irqchip ("-machine kernel_irqchip=off"
with QEMU). The fix is to make the same check in
kvm_vcpu_reload_apic_access_page that we already have
in vmx.c's vm_need_virtualize_apic_accesses().
Reported-by: Jan Kiszka <jan.kiszka@siemens.com>
Tested-by: Jan Kiszka <jan.kiszka@siemens.com>
Fixes: 4256f43f9fab91e1c17b5846a240cf4b66a768a8
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
{
struct page *page = NULL;
+ if (!irqchip_in_kernel(vcpu->kvm))
+ return;
+
if (!kvm_x86_ops->set_apic_access_page_addr)
return;