[PATCH v2 7/7] KVM: arm64: Unconditionally cross check hyp state

Quentin Perret qperret at google.com
Wed Apr 16 08:26:47 PDT 2025


Now that the hypervisor's state is stored in the hyp_vmemmap, we no
longer need an expensive page-table walk to read it. This means we can
now afford to cross check the hyp-state during all memory ownership
transitions where the hyp is involved unconditionally, hence avoiding
problems such as [1].

[1] https://lore.kernel.org/kvmarm/20241128154406.602875-1-qperret@google.com/

Reviewed-by: Marc Zyngier <maz at kernel.org>
Signed-off-by: Quentin Perret <qperret at google.com>
---
 arch/arm64/kvm/hyp/nvhe/mem_protect.c | 24 +++++++++---------------
 1 file changed, 9 insertions(+), 15 deletions(-)

diff --git a/arch/arm64/kvm/hyp/nvhe/mem_protect.c b/arch/arm64/kvm/hyp/nvhe/mem_protect.c
index 91b757e3fb4c..709d286999a1 100644
--- a/arch/arm64/kvm/hyp/nvhe/mem_protect.c
+++ b/arch/arm64/kvm/hyp/nvhe/mem_protect.c
@@ -702,11 +702,9 @@ int __pkvm_host_share_hyp(u64 pfn)
 	ret = __host_check_page_state_range(phys, size, PKVM_PAGE_OWNED);
 	if (ret)
 		goto unlock;
-	if (IS_ENABLED(CONFIG_NVHE_EL2_DEBUG)) {
-		ret = __hyp_check_page_state_range(phys, size, PKVM_NOPAGE);
-		if (ret)
-			goto unlock;
-	}
+	ret = __hyp_check_page_state_range(phys, size, PKVM_NOPAGE);
+	if (ret)
+		goto unlock;
 
 	__hyp_set_page_state_range(phys, size, PKVM_PAGE_SHARED_BORROWED);
 	WARN_ON(__host_set_page_state_range(phys, size, PKVM_PAGE_SHARED_OWNED));
@@ -762,11 +760,9 @@ int __pkvm_host_donate_hyp(u64 pfn, u64 nr_pages)
 	ret = __host_check_page_state_range(phys, size, PKVM_PAGE_OWNED);
 	if (ret)
 		goto unlock;
-	if (IS_ENABLED(CONFIG_NVHE_EL2_DEBUG)) {
-		ret = __hyp_check_page_state_range(phys, size, PKVM_NOPAGE);
-		if (ret)
-			goto unlock;
-	}
+	ret = __hyp_check_page_state_range(phys, size, PKVM_NOPAGE);
+	if (ret)
+		goto unlock;
 
 	__hyp_set_page_state_range(phys, size, PKVM_PAGE_OWNED);
 	WARN_ON(pkvm_create_mappings_locked(virt, virt + size, PAGE_HYP));
@@ -792,11 +788,9 @@ int __pkvm_hyp_donate_host(u64 pfn, u64 nr_pages)
 	ret = __hyp_check_page_state_range(phys, size, PKVM_PAGE_OWNED);
 	if (ret)
 		goto unlock;
-	if (IS_ENABLED(CONFIG_NVHE_EL2_DEBUG)) {
-		ret = __host_check_page_state_range(phys, size, PKVM_NOPAGE);
-		if (ret)
-			goto unlock;
-	}
+	ret = __host_check_page_state_range(phys, size, PKVM_NOPAGE);
+	if (ret)
+		goto unlock;
 
 	__hyp_set_page_state_range(phys, size, PKVM_NOPAGE);
 	WARN_ON(kvm_pgtable_hyp_unmap(&pkvm_pgtable, virt, size) != size);
-- 
2.49.0.604.gff1f9ca942-goog




More information about the linux-arm-kernel mailing list