KVM: Avoid killing userspace through guest SRAO MCE on unmapped pages

In common cases, guest SRAO MCE will cause corresponding poisoned page
be un-mapped and SIGBUS be sent to QEMU-KVM, then QEMU-KVM will relay
the MCE to guest OS.

But it is reported that if the poisoned page is accessed in guest
after unmapping and before MCE is relayed to guest OS, userspace will
be killed.

The reason is as follows. Because poisoned page has been un-mapped,
guest access will cause guest exit and kvm_mmu_page_fault will be
called. kvm_mmu_page_fault can not get the poisoned page for fault
address, so kernel and user space MMIO processing is tried in turn. In
user MMIO processing, poisoned page is accessed again, then userspace
is killed by force_sig_info.

To fix the bug, kvm_mmu_page_fault send HWPOISON signal to QEMU-KVM
and do not try kernel and user space MMIO processing for poisoned
page.

[xiao: fix warning introduced by avi]

Reported-by: Max Asbock <masbock@linux.vnet.ibm.com>
Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>

1 files changed, 28 insertions, 2 deletions

diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index f032806..187aa8d 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -92,6 +92,9 @@ static bool kvm_rebooting;
 
 static bool largepages_enabled = true;
 
+struct page *hwpoison_page;
+pfn_t hwpoison_pfn;
+
 inline int kvm_is_mmio_pfn(pfn_t pfn)
 {
 	if (pfn_valid(pfn)) {
@@ -810,16 +813,22 @@ EXPORT_SYMBOL_GPL(kvm_disable_largepages);
 
 int is_error_page(struct page *page)
 {
-	return page == bad_page;
+	return page == bad_page || page == hwpoison_page;
 }
 EXPORT_SYMBOL_GPL(is_error_page);
 
 int is_error_pfn(pfn_t pfn)
 {
-	return pfn == bad_pfn;
+	return pfn == bad_pfn || pfn == hwpoison_pfn;
 }
 EXPORT_SYMBOL_GPL(is_error_pfn);
 
+int is_hwpoison_pfn(pfn_t pfn)
+{
+	return pfn == hwpoison_pfn;
+}
+EXPORT_SYMBOL_GPL(is_hwpoison_pfn);
+
 static inline unsigned long bad_hva(void)
 {
 	return PAGE_OFFSET;
@@ -945,6 +954,11 @@ static pfn_t hva_to_pfn(struct kvm *kvm, unsigned long addr)
 	if (unlikely(npages != 1)) {
 		struct vm_area_struct *vma;
 
+		if (is_hwpoison_address(addr)) {
+			get_page(hwpoison_page);
+			return page_to_pfn(hwpoison_page);
+		}
+
 		down_read(&current->mm->mmap_sem);
 		vma = find_vma(current->mm, addr);
 
@@ -2197,6 +2211,15 @@ int kvm_init(void *opaque, unsigned vcpu_size, unsigned vcpu_align,
 
 	bad_pfn = page_to_pfn(bad_page);
 
+	hwpoison_page = alloc_page(GFP_KERNEL | __GFP_ZERO);
+
+	if (hwpoison_page == NULL) {
+		r = -ENOMEM;
+		goto out_free_0;
+	}
+
+	hwpoison_pfn = page_to_pfn(hwpoison_page);
+
 	if (!zalloc_cpumask_var(&cpus_hardware_enabled, GFP_KERNEL)) {
 		r = -ENOMEM;
 		goto out_free_0;
@@ -2269,6 +2292,8 @@ out_free_1:
 out_free_0a:
 	free_cpumask_var(cpus_hardware_enabled);
 out_free_0:
+	if (hwpoison_page)
+		__free_page(hwpoison_page);
 	__free_page(bad_page);
 out:
 	kvm_arch_exit();
@@ -2290,6 +2315,7 @@ void kvm_exit(void)
 	kvm_arch_hardware_unsetup();
 	kvm_arch_exit();
 	free_cpumask_var(cpus_hardware_enabled);
+	__free_page(hwpoison_page);
 	__free_page(bad_page);
 }
 EXPORT_SYMBOL_GPL(kvm_exit);