Re: [Qemu-devel] [PATCH 3/7] cpu model bug fixes and definition corrections: Add kvm emulated x2apic flag to config defined cpu models

2011-05-30 Thread Roedel, Joerg 
On Sat, May 28, 2011 at 04:39:13AM -0400, Jan Kiszka wrote:

> Jörg, how to deal with -enable-nesting in qemu-kvm to align behavior
> with upstream?

My personal preference is to just remove it. In upstream-qemu it is
enabled/disabled by +/-svm. -enable-nesting is just a historic thing
which can be wiped out.

Joerg

-- 
AMD Operating System Research Center

Advanced Micro Devices GmbH Einsteinring 24 85609 Dornach
General Managers: Alberto Bozzo, Andrew Bowd
Registration: Dornach, Landkr. Muenchen; Registerger. Muenchen, HRB Nr. 43632

Re: [Qemu-devel] [PATCH 3/7] cpu model bug fixes and definition corrections: Add kvm emulated x2apic flag to config defined cpu models

2011-05-28 Thread Jan Kiszka 
On 2011-05-23 23:46, john cooper wrote:
> Add kvm emulated x2apic flag to config defined cpu models
> and general support for such hypervisor emulated flags.
> 
> In addition to checking user request flags against the host
> we also selectively check against kvm for emulated flags.

As you are already digging through this jungle, could you also check how
much of this diff from qemu-kvm is obsolete now?

diff --git a/target-i386/cpuid.c b/target-i386/cpuid.c
index e479a4d..091d812 100644
--- a/target-i386/cpuid.c
+++ b/target-i386/cpuid.c
@@ -1175,6 +1175,31 @@ void cpu_x86_cpuid(CPUX86State *env, uint32_t
index, uint32_t count,
 *ecx |= 1 << 1;/* CmpLegacy bit */
 }
 }
+if (kvm_enabled()) {
+uint32_t h_eax, h_edx;
+
+host_cpuid(index, 0, &h_eax, NULL, NULL, &h_edx);
+
+/* disable CPU features that the host does not support */
+
+/* long mode */
+if ((h_edx & 0x2000) == 0 /* || !lm_capable_kernel */)
+*edx &= ~0x2000;
+/* syscall */
+if ((h_edx & 0x0800) == 0)
+*edx &= ~0x0800;
+/* nx */
+if ((h_edx & 0x0010) == 0)
+*edx &= ~0x0010;
+
+/* disable CPU features that KVM cannot support */
+
+/* svm */
+if (!kvm_nested)
+*ecx &= ~CPUID_EXT3_SVM;
+/* 3dnow */
+*edx &= ~0xc000;
+}
 break;
 case 0x8002:
 case 0x8003:

If something is still missing in upstream, it should be ported - except
for kvm_nested which should be controllable via '-cpu ...,+/-svm'. TIA!

Jörg, how to deal with -enable-nesting in qemu-kvm to align behavior
with upstream?

Jan



signature.asc
Description: OpenPGP digital signature

[Qemu-devel] [PATCH 3/7] cpu model bug fixes and definition corrections: Add kvm emulated x2apic flag to config defined cpu models

2011-05-23 Thread john cooper 
Add kvm emulated x2apic flag to config defined cpu models
and general support for such hypervisor emulated flags.

In addition to checking user request flags against the host
we also selectively check against kvm for emulated flags.

Signed-off-by: john cooper 
---

diff --git a/hw/pc.c b/hw/pc.c
index 8106197..5b94e53 100644
--- a/hw/pc.c
+++ b/hw/pc.c
@@ -927,7 +927,7 @@ static CPUState *pc_new_cpu(const char *cpu_model)
 
 env = cpu_init(cpu_model);
 if (!env) {
-fprintf(stderr, "Unable to find x86 CPU definition\n");
+fprintf(stderr, "Unable to support requested x86 CPU definition\n");
 exit(1);
 }
 if ((env->cpuid_features & CPUID_APIC) || smp_cpus > 1) {
diff --git a/sysconfigs/target/target-x86_64.conf 
b/sysconfigs/target/target-x86_64.conf
index 0613870..3b85b2e 100644
--- a/sysconfigs/target/target-x86_64.conf
+++ b/sysconfigs/target/target-x86_64.conf
@@ -8,7 +8,7 @@
model = "15"
stepping = "3"
feature_edx = "sse2 sse fxsr mmx pat cmov pge sep apic cx8 mce pae msr tsc 
pse de fpumtrr clflush mca pse36"
-   feature_ecx = "sse3 ssse3"
+   feature_ecx = "sse3 ssse3 x2apic"
extfeature_edx = "fxsr mmx pat cmov pge apic cx8 mce pae msr tsc pse de fpu 
   lm syscall nx"
extfeature_ecx = "lahf_lm"
xlevel = "0x800A"
@@ -22,7 +22,7 @@
model = "23"
stepping = "3"
feature_edx = "sse2 sse fxsr mmx pat cmov pge sep apic cx8 mce pae msr tsc 
pse de fpumtrr clflush mca pse36"
-   feature_ecx = "sse3 cx16 ssse3 sse4.1"
+   feature_ecx = "sse3 cx16 ssse3 sse4.1 x2apic"
extfeature_edx = "fxsr mmx pat cmov pge apic cx8 mce pae msr tsc pse de fpu 
   lm syscall nx"
extfeature_ecx = "lahf_lm"
xlevel = "0x800A"
@@ -36,7 +36,7 @@
model = "26"
stepping = "3"
feature_edx = "sse2 sse fxsr mmx pat cmov pge sep apic cx8 mce pae msr tsc 
pse de fpumtrr clflush mca pse36"
-   feature_ecx = "sse3 cx16 ssse3 sse4.1 sse4.2 popcnt"
+   feature_ecx = "sse3 cx16 ssse3 sse4.1 sse4.2 popcnt x2apic"
extfeature_edx = "fxsr mmx pat cmov pge apic cx8 mce pae msr tsc pse de fpu 
   lm syscall nx"
extfeature_ecx = "lahf_lm"
xlevel = "0x800A"
@@ -50,7 +50,7 @@
model = "6"
stepping = "1"
feature_edx = "sse2 sse fxsr mmx pat cmov pge sep apic cx8 mce pae msr tsc 
pse de fpumtrr clflush mca pse36"
-   feature_ecx = "sse3"
+   feature_ecx = "sse3 x2apic"
extfeature_edx = "fxsr mmx pat cmov pge apic cx8 mce pae msr tsc pse de fpu 
   lm syscall nx"
 #   extfeature_ecx = ""
xlevel = "0x8008"
@@ -64,7 +64,7 @@
model = "6"
stepping = "1"
feature_edx = "sse2 sse fxsr mmx pat cmov pge sep apic cx8 mce pae msr tsc 
pse de fpumtrr clflush mca pse36"
-   feature_ecx = "sse3 cx16"
+   feature_ecx = "sse3 cx16 x2apic"
extfeature_edx = "fxsr mmx pat cmov pge apic cx8 mce pae msr tsc pse de fpu 
   lm syscall nx rdtscp"
extfeature_ecx = "svm lahf_lm"
xlevel = "0x8008"
@@ -78,7 +78,7 @@
model = "6"
stepping = "1"
feature_edx = "sse2 sse fxsr mmx pat cmov pge sep apic cx8 mce pae msr tsc 
pse de fpumtrr clflush mca pse36"
-   feature_ecx = "sse3 cx16 monitor popcnt"
+   feature_ecx = "sse3 cx16 monitor popcnt x2apic"
extfeature_edx = "fxsr mmx pat cmov pge apic cx8 mce pae msr tsc pse de fpu 
   lm syscall nx rdtscp"
extfeature_ecx = "svm sse4a  abm misalignsse lahf_lm"
xlevel = "0x8008"
diff --git a/target-i386/cpuid.c b/target-i386/cpuid.c
index e479a4d..c151f12 100644
--- a/target-i386/cpuid.c
+++ b/target-i386/cpuid.c
@@ -549,15 +549,40 @@ static int unavailable_host_feature(struct 
model_features_t *f, uint32_t mask)
 return 0;
 }
 
-/* best effort attempt to inform user requested cpu flags aren't making
- * their way to the guest.  Note: ft[].check_feat ideally should be
- * specified via a guest_def field to suppress report of extraneous flags.
+/* determine the effective set of cpuid features visible to a guest.
+ * in the case kvm is enabled, we also selectively include features
+ * emulated by the hypervisor
  */
-static int check_features_against_host(x86_def_t *guest_def)
+static void summary_cpuid_features(CPUX86State *env, x86_def_t *hd)
+{
+struct {
+uint32_t *pfeat, cmd, reg, mask;
+} fmap[] = {
+{&hd->features, 0x0001, R_EDX, 0},
+{&hd->ext_features, 0x0001, R_ECX, CPUID_EXT_X2APIC},
+{&hd->ext2_features, 0x8001, R_EDX, 0},
+{&hd->ext3_features, 0x8001, R_ECX, 0},
+{NULL} }, *p;
+
+cpu_x86_fill_host(hd);
+if (kvm_enabled()) {
+for (p = fmap; p->pfeat; ++p) {
+if (p->mask) {
+*p->pfeat |= p->mask &
+kvm_arch_get_supported_cpuid(env, p->cmd, 0, p->reg);
+}
+}
+}
+}
+
+/* inform the user of any requested cpu features (both explicitly requested
+ * flags and implicit cpu model flags) not making their way to the guest
+ */
+static int chec