KVM: SVM: remove selective CR0 comment
authorJoerg Roedel <joerg.roedel@amd.com>
Wed, 16 Apr 2008 15:01:05 +0000 (17:01 +0200)
committerAvi Kivity <avi@qumranet.com>
Sun, 27 Apr 2008 15:21:44 +0000 (18:21 +0300)
There is not selective cr0 intercept bug. The code in the comment sets the
CR0.PG bit. But KVM sets the CR4.PG bit for SVM always to implement the paged
real mode. So the 'mov %eax,%cr0' instruction does not change the CR0.PG bit.
Selective CR0 intercepts only occur when a bit is actually changed. So its the
right behavior that there is no intercept on this instruction.

Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>
arch/x86/kvm/svm.c

index d643605c5aeb423a4938a88056273ff041f981ef..89e0be2c10d0ee02153fd6f72df1ebd6e31f6e7b 100644 (file)
@@ -513,17 +513,6 @@ static void init_vmcb(struct vcpu_svm *svm)
        control->intercept =    (1ULL << INTERCEPT_INTR) |
                                (1ULL << INTERCEPT_NMI) |
                                (1ULL << INTERCEPT_SMI) |
-               /*
-                * selective cr0 intercept bug?
-                *      0:   0f 22 d8                mov    %eax,%cr3
-                *      3:   0f 20 c0                mov    %cr0,%eax
-                *      6:   0d 00 00 00 80          or     $0x80000000,%eax
-                *      b:   0f 22 c0                mov    %eax,%cr0
-                * set cr3 ->interception
-                * get cr0 ->interception
-                * set cr0 -> no interception
-                */
-               /*              (1ULL << INTERCEPT_SELECTIVE_CR0) | */
                                (1ULL << INTERCEPT_CPUID) |
                                (1ULL << INTERCEPT_INVD) |
                                (1ULL << INTERCEPT_HLT) |