forge/linux-pinenote
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

KVM: s390: Fix problem state check for b2 intercepts

Browse source 
The kernel handles some priviledged instruction exits. While I was
unable to trigger such an exit from guest userspace, the code should
check for supervisor state before emulating a priviledged instruction.

I also renamed kvm_s390_handle_priv to kvm_s390_handle_b2. After all
there are non priviledged b2 instructions like stck (store clock).

Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
This commit is contained in:
Christian Borntraeger 2009-01-22 10:28:29 +01:00 • committed by Avi Kivity
parent 2c411b48af
commit 70455a36a0
3 changed files with 17 additions and 5 deletions
Download patch file Download diff file Expand all files Collapse all files

18
arch/s390/kvm/priv.c
View file

@ -304,12 +304,24 @@ static intercept_handler_t priv_handlers[256] = {
[0xb1] = handle_stfl,
};
int kvm_s390_handle_priv(struct kvm_vcpu *vcpu)
int kvm_s390_handle_b2(struct kvm_vcpu *vcpu)
{
intercept_handler_t handler;
/*
* a lot of B2 instructions are priviledged. We first check for
* the priviledges ones, that we can handle in the kernel. If the
* kernel can handle this instruction, we check for the problem
* state bit and (a) handle the instruction or (b) send a code 2
* program check.
* Anything else goes to userspace.*/
handler = priv_handlers[vcpu->arch.sie_block->ipa & 0x00ff];
if (handler)
return handler(vcpu);
if (handler) {
if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
return kvm_s390_inject_program_int(vcpu,
PGM_PRIVILEGED_OPERATION);
else
return handler(vcpu);
}
return -ENOTSUPP;
}