From: Will Deacon will@kernel.org

commit 250f25367b58d8c65a1b060a2dda037eea09a672 upstream.

If kvm_arch_vcpu_create() fails to share the vCPU page with the hypervisor, we propagate the error back to the ioctl but leave the vGIC vCPU data initialised. Note only does this leak the corresponding memory when the vCPU is destroyed but it can also lead to use-after-free if the redistributor device handling tries to walk into the vCPU.

Add the missing cleanup to kvm_arch_vcpu_create(), ensuring that the vGIC vCPU structures are destroyed on error.

Cc: stable@vger.kernel.org Cc: Marc Zyngier maz@kernel.org Cc: Oliver Upton oliver.upton@linux.dev Cc: Quentin Perret qperret@google.com Signed-off-by: Will Deacon will@kernel.org Reviewed-by: Marc Zyngier maz@kernel.org Link: https://lore.kernel.org/r/20250314133409.9123-1-will@kernel.org Signed-off-by: Oliver Upton oliver.upton@linux.dev [Denis: minor fix to resolve merge conflict.] Signed-off-by: Denis Arefev arefev@swemel.ru --- Backport fix for CVE-2025-37849 Link: https://nvd.nist.gov/vuln/detail/cve-2025-37849 --- arch/arm64/kvm/arm.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/arch/arm64/kvm/arm.c b/arch/arm64/kvm/arm.c index afe8be2fef88..3adaa3216baf 100644 --- a/arch/arm64/kvm/arm.c +++ b/arch/arm64/kvm/arm.c @@ -294,7 +294,12 @@ int kvm_arch_vcpu_create(struct kvm_vcpu *vcpu) if (err) return err;

- return create_hyp_mappings(vcpu, vcpu + 1, PAGE_HYP); + err = kvm_share_hyp(vcpu, vcpu + 1); + if (err) + kvm_vgic_vcpu_destroy(vcpu); + + return err; + }

void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)