From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([208.118.235.92]:36673) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Tll4f-0000QW-32 for qemu-devel@nongnu.org; Thu, 20 Dec 2012 13:42:31 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1Tll4c-0008KG-M1 for qemu-devel@nongnu.org; Thu, 20 Dec 2012 13:42:29 -0500 Received: from mx1.redhat.com ([209.132.183.28]:23709) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Tll4c-0008JW-Dk for qemu-devel@nongnu.org; Thu, 20 Dec 2012 13:42:26 -0500 From: Eduardo Habkost Date: Thu, 20 Dec 2012 16:43:48 -0200 Message-Id: <1356029028-7356-1-git-send-email-ehabkost@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Subject: [Qemu-devel] [PATCH] target-i386: CPUID: return highest basic leaf if eax > cpuid_xlevel List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-devel@nongnu.org Cc: Marcelo Tosatti , Markus Armbruster , "brillywu@viatech.com.cn" This fixes a subtle bug. A bug that probably won't cause trouble for any existing OS, but a bug anyway: Intel SDM Volume 2, CPUID Instruction states: > Two types of information are returned: basic and extended function > information. If a value entered for CPUID.EAX is higher than the maximu= m > input value for basic or extended function for that processor then the > data for the highest basic information leaf is returned. For example, > using the Intel Core i7 processor, the following is true: > > CPUID.EAX =3D 05H (* Returns MONITOR/MWAIT leaf. *) > CPUID.EAX =3D 0AH (* Returns Architectural Performance Monitoring lea= f. *) > CPUID.EAX =3D 0BH (* Returns Extended Topology Enumeration leaf. *) > CPUID.EAX =3D 0CH (* INVALID: Returns the same information as CPUID.E= AX =3D 0BH. *) > CPUID.EAX =3D 80000008H (* Returns linear/physical address size data.= *) > CPUID.EAX =3D 8000000AH (* INVALID: Returns same information as CPUID= .EAX =3D 0BH. *) AMD's CPUID Specification, on the other hand, is less specific: > The CPUID instruction supports two sets or ranges of functions, > standard and extended. > > =E2=80=A2 The smallest function number of the standard function range i= s > Fn0000_0000. The largest function num- ber of the standard function > range, for a particular implementation, is returned in CPUID > Fn0000_0000_EAX. > > =E2=80=A2 The smallest function number of the extended function range i= s > Fn8000_0000. The largest function num- ber of the extended function > range, for a particular implementation, is returned in CPUID > Fn8000_0000_EAX. > > Functions that are neither standard nor extended are undefined and > should not be relied upon. QEMU's behavior matched Intel's specification before, but this was changed by commit b3baa152aaef1905876670590275c2dd0bbb088c. This patch restores the behavior documented by Intel when cpuid_xlevel2 is 0. The existing behavior when cpuid_xlevel2 is set (falling back to level=3Dcpuid_xlevel) is being kept, as I couldn't find any public documentation on the CPUID 0xC0000000 function range on Centaur CPUs. Signed-off-by: Eduardo Habkost --- Cc: Markus Armbruster Cc: brillywu@viatech.com.cn Cc: Marcelo Tosatti --- target-i386/cpu.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/target-i386/cpu.c b/target-i386/cpu.c index 1837f5a..3cd1cee 100644 --- a/target-i386/cpu.c +++ b/target-i386/cpu.c @@ -1648,7 +1648,11 @@ void cpu_x86_cpuid(CPUX86State *env, uint32_t inde= x, uint32_t count, index =3D env->cpuid_xlevel; } } else { - index =3D env->cpuid_xlevel; + /* Intel documentation states that invalid EAX input wil= l + * return the same information as EAX=3Dcpuid_level + * (Intel SDM Vol. 2A - Instruction Set Reference - CPUI= D) + */ + index =3D env->cpuid_level; } } } else { --=20 1.7.11.7