From: Andre Przywara Subject: [PATCH] qemu-kvm: fix KVMs GET_SUPPORTED_CPUID feature usage Date: Thu, 25 Jun 2009 15:12:03 +0200 Message-ID: <1245935523-32604-1-git-send-email-andre.przywara@amd.com> Mime-Version: 1.0 Content-Type: text/plain Cc: kvm@vger.kernel.org, Andre Przywara To: avi@redhat.com Return-path: Received: from outbound-dub.frontbridge.com ([213.199.154.16]:28490 "EHLO IE1EHSOBE003.bigfish.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751818AbZFYNIj (ORCPT ); Thu, 25 Jun 2009 09:08:39 -0400 Sender: kvm-owner@vger.kernel.org List-ID: If we want to trim the user provided CPUID bits for KVM to be not greater than that of the host, we should a) not remove the bits _after_ we sent them to the kernel and b) not do the bitmangling on the original values while sending a copy This fixes the masking of features that are not present on the host and helps to use -cpu host. Signed-off-by: Andre Przywara --- qemu-kvm-x86.c | 21 +++++++++++---------- 1 files changed, 11 insertions(+), 10 deletions(-) Hi, this is a port of patch 4/6 of my "-cpu host" series for QEMU. The bug is similar, although the fix is a different one (different files). Even without "-cpu host" features that the host does not support are currently _not_ disabled as intended, so I consider this a bugfix. Regards, Andre. diff --git a/qemu-kvm-x86.c b/qemu-kvm-x86.c index 66b2b19..f3e4b6b 100644 --- a/qemu-kvm-x86.c +++ b/qemu-kvm-x86.c @@ -1190,7 +1190,6 @@ int kvm_arch_qemu_init_env(CPUState *cenv) qemu_kvm_load_lapic(cenv); - copy = *cenv; #ifdef KVM_CPUID_SIGNATURE /* Paravirtualization CPUIDs */ @@ -1209,6 +1208,17 @@ int kvm_arch_qemu_init_env(CPUState *cenv) pv_ent->eax = get_para_features(kvm_context); #endif + kvm_trim_features(&cenv->cpuid_features, + kvm_arch_get_supported_cpuid(cenv, 1, R_EDX)); + kvm_trim_features(&cenv->cpuid_ext_features, + kvm_arch_get_supported_cpuid(cenv, 1, R_ECX)); + kvm_trim_features(&cenv->cpuid_ext2_features, + kvm_arch_get_supported_cpuid(cenv, 0x80000001, R_EDX)); + kvm_trim_features(&cenv->cpuid_ext3_features, + kvm_arch_get_supported_cpuid(cenv, 0x80000001, R_ECX)); + + copy = *cenv; + copy.regs[R_EAX] = 0; qemu_kvm_cpuid_on_env(©); limit = copy.regs[R_EAX]; @@ -1243,15 +1253,6 @@ int kvm_arch_qemu_init_env(CPUState *cenv) kvm_setup_cpuid2(cenv->kvm_cpu_state.vcpu_ctx, cpuid_nent, cpuid_ent); - kvm_trim_features(&cenv->cpuid_features, - kvm_arch_get_supported_cpuid(cenv, 1, R_EDX)); - kvm_trim_features(&cenv->cpuid_ext_features, - kvm_arch_get_supported_cpuid(cenv, 1, R_ECX)); - kvm_trim_features(&cenv->cpuid_ext2_features, - kvm_arch_get_supported_cpuid(cenv, 0x80000001, R_EDX)); - kvm_trim_features(&cenv->cpuid_ext3_features, - kvm_arch_get_supported_cpuid(cenv, 0x80000001, R_ECX)); - return 0; } -- 1.6.1.3