Changeset 58676 in vbox for trunk

Timestamp:

Nov 12, 2015 4:34:33 PM (9 years ago)

Author:

vboxsync

Message:

Hacked up to survive on a PIII-S.

File:

• trunk/src/VBox/VMM/tools/VBoxCpuReport.cpp (modified) (2 diffs)

trunk/src/VBox/VMM/tools/VBoxCpuReport.cpp

@@ -509 +509 @@
             /* Skip 0xc0011012..13 as it seems to be bad for our health (Phenom II X6 1100T). */
             /* Ditto for 0x0000002a (EBL_CR_POWERON) and 0x00000277 (MSR_IA32_CR_PAT) on Intel (Atom 330). */
+            /* And more of the same for 0x280 on Intel Pentium III. */
             if (   ((uMsr >= 0xc0011012 && uMsr <= 0xc0011013) && g_enmVendor == CPUMCPUVENDOR_AMD)
                 || (   (uMsr == 0x2a || uMsr == 0x277)
                     && g_enmVendor == CPUMCPUVENDOR_INTEL
-                    && g_enmMicroarch == kCpumMicroarch_Intel_Atom_Bonnell))
+                    && g_enmMicroarch == kCpumMicroarch_Intel_Atom_Bonnell)
+                || (   (uMsr == 0x280)
+                    && g_enmMicroarch == kCpumMicroarch_Intel_P6_III)) 
                 vbCpuRepDebug("Skipping %#x\n", uMsr);
             else
@@ -3257 +3260 @@
     if (vbCpuRepSupportsX2Apic())
         fSkipMask |= RT_BIT_64(10);
+    /* For some reason, twiddling this bit kills a Tualatin PIII-S. */
+    if (g_enmMicroarch == kCpumMicroarch_Intel_P6_III)
+        fSkipMask |= RT_BIT(9);
     return reportMsr_GenFunctionEx(uMsr, "Ia32ApicBase", uValue, fSkipMask, 0, NULL);
 }