From mboxrd@z Thu Jan 1 00:00:00 1970 From: =?UTF-8?q?Radim=20Kr=C4=8Dm=C3=A1=C5=99?= Subject: [PATCH v3 0/4] KVM: x86: emulate FXSAVE and FXRSTOR Date: Tue, 8 Nov 2016 20:54:15 +0100 Message-ID: <20161108195419.4607-1-rkrcmar@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Cc: Paolo Bonzini , Bandan Das , Nadav Amit To: linux-kernel@vger.kernel.org, kvm@vger.kernel.org Return-path: Received: from mx1.redhat.com ([209.132.183.28]:43450 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752433AbcKHTy2 (ORCPT ); Tue, 8 Nov 2016 14:54:28 -0500 Sender: kvm-owner@vger.kernel.org List-ID: v2: http://www.spinics.net/lists/kvm/msg139681.html v3 brings compatibility with old compilers and has been compile-tested with GCC-4.4 on Debian Wheezy, GCC-4.4 on RHEL 6, and GCC-4.1 on RHEL 5. [4/4] still has the hidden assumption that guest and host CPUID match. Emulating a guest that does not deprecaste FCS and FDS on a host that does (a modern host) will not necessarily cause a problem, which is why patches don't handle the case. Enforcing the equality in CPUID update would be best, but another series could do that as Linux doesn't even have the CPUID bit defined yet. Radim Krčmář (4): KVM: x86: add Align16 instruction flag KVM: x86: save one bit in ctxt->d KVM: x86: add asm_safe wrapper KVM: x86: emulate FXSAVE and FXRSTOR arch/x86/kvm/emulate.c | 184 +++++++++++++++++++++++++++++++++++++++++-------- 1 file changed, 157 insertions(+), 27 deletions(-) -- 2.10.2