From mboxrd@z Thu Jan 1 00:00:00 1970 From: Marcelo Tosatti Subject: [PATCH v2] KVM: x86: fix KVM_SET_CLOCK relative to setting correct clock value Date: Wed, 3 May 2017 10:43:43 -0300 Message-ID: <20170503134341.GB10468@amt.cnet> References: <20170502213616.GA24837@amt.cnet> <2499ef65-1dfe-8460-ec41-661b05cc5023@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: kvm-devel , Radim =?utf-8?B?S3LEjW3DocWZ?= To: Paolo Bonzini Return-path: Received: from mx1.redhat.com ([209.132.183.28]:38974 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752665AbdECNoD (ORCPT ); Wed, 3 May 2017 09:44:03 -0400 Received: from smtp.corp.redhat.com (int-mx06.intmail.prod.int.phx2.redhat.com [10.5.11.16]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id C24ABC04B32C for ; Wed, 3 May 2017 13:44:02 +0000 (UTC) Content-Disposition: inline In-Reply-To: <2499ef65-1dfe-8460-ec41-661b05cc5023@redhat.com> Sender: kvm-owner@vger.kernel.org List-ID: In the masterclock enabled case, kvmclock_offset must be adjusted so that user_ns.clock = master_kernel_ns + kvmclock_offset (that is, the value set from KVM_SET_CLOCK is the one visible at system_timestamp). Signed-off-by: Marcelo Tosatti --- arch/x86/kvm/x86.c | 29 ++++++++++++++++++++++++++--- 1 file changed, 26 insertions(+), 3 deletions(-) v2: grab pvclock_gtod_sync_lock spinlock (Paolo) Index: kvm/arch/x86/kvm/x86.c =================================================================== --- kvm.orig/arch/x86/kvm/x86.c 2017-04-27 17:37:48.131348255 -0300 +++ kvm/arch/x86/kvm/x86.c 2017-05-03 10:35:19.298447766 -0300 @@ -4172,8 +4172,9 @@ break; } case KVM_SET_CLOCK: { - struct kvm_clock_data user_ns; u64 now_ns; + struct kvm_clock_data user_ns; + struct kvm_arch *ka = &kvm->arch; r = -EFAULT; if (copy_from_user(&user_ns, argp, sizeof(user_ns))) @@ -4184,9 +4185,31 @@ goto out; r = 0; - now_ns = get_kvmclock_ns(kvm); - kvm->arch.kvmclock_offset += user_ns.clock - now_ns; + kvm_gen_update_masterclock(kvm); + spin_lock(&ka->pvclock_gtod_sync_lock); + if (ka->use_master_clock) { + int i; + struct kvm_vcpu *vcpu; + + /* + * In the masterclock enabled case, + * kvmclock_offset must be adjusted so that + * user_ns.clock = master_kernel_ns + kvmclock_offset + * (that is, the value set from KVM_SET_CLOCK is the + * one visible at system_timestamp). + */ + kvm->arch.kvmclock_offset = user_ns.clock - + ka->master_kernel_ns; + spin_unlock(&ka->pvclock_gtod_sync_lock); + + kvm_for_each_vcpu(i, vcpu, kvm) + kvm_make_request(KVM_REQ_CLOCK_UPDATE, vcpu); + } else { + spin_unlock(&ka->pvclock_gtod_sync_lock); + now_ns = get_kvmclock_ns(kvm); + kvm->arch.kvmclock_offset += user_ns.clock - now_ns; + } break; } case KVM_GET_CLOCK: {