From: Kele Huang <huangkele@bytedance.com> mainline inclusion from mainline-v5.15-rc1 commit c2402d43d183b11445aed921e7bebcd47ef222f1 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8DWT1 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i... -------------------------------- Commit a86ed2cfa13c5 ("ptp: Don't print an error if ptp_kvm is not supported") fixes the error message print on ARM platform by only concerning about the case that the error returned from kvm_arch_ptp_init() is not -EOPNOTSUPP. Although the ARM platform returns -EOPNOTSUPP if ptp_kvm is not supported while X86_64 platform returns -KVM_EOPNOTSUPP, both error codes share the same value 95. Actually kvm_arch_ptp_init() on X86_64 platform can return three kinds of errors (-KVM_ENOSYS, -KVM_EOPNOTSUPP and -KVM_EFAULT). The problem is that -KVM_EOPNOTSUPP is masked out and -KVM_EFAULT is ignored among them. This patch fixes this by returning them to ptp_kvm_init() respectively. Signed-off-by: Kele Huang <huangkele@bytedance.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Dongxu Sun <sundongxu3@huawei.com> --- drivers/ptp/ptp_kvm_x86.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/ptp/ptp_kvm_x86.c b/drivers/ptp/ptp_kvm_x86.c index 3dd519dfc473..38ebb3f953af 100644 --- a/drivers/ptp/ptp_kvm_x86.c +++ b/drivers/ptp/ptp_kvm_x86.c @@ -34,10 +34,10 @@ int kvm_arch_ptp_init(void) ret = kvm_hypercall2(KVM_HC_CLOCK_PAIRING, clock_pair_gpa, KVM_CLOCK_PAIRING_WALLCLOCK); - if (ret == -KVM_ENOSYS || ret == -KVM_EOPNOTSUPP) + if (ret == -KVM_ENOSYS) return -ENODEV; - return 0; + return ret; } int kvm_arch_ptp_get_clock(struct timespec64 *ts) -- 2.33.0