Message ID | 1395704828-20180-5-git-send-email-paulus@samba.org |
---|---|
State | New, archived |
Headers | show |
diff --git a/arch/powerpc/kvm/book3s_rtas.c b/arch/powerpc/kvm/book3s_rtas.c index cf95cde..7a05315 100644 --- a/arch/powerpc/kvm/book3s_rtas.c +++ b/arch/powerpc/kvm/book3s_rtas.c @@ -213,8 +213,11 @@ int kvmppc_rtas_hcall(struct kvm_vcpu *vcpu) gpa_t args_phys; int rc; - /* r4 contains the guest physical address of the RTAS args */ - args_phys = kvmppc_get_gpr(vcpu, 4); + /* + * r4 contains the guest physical address of the RTAS args + * Mask off the top 4 bits since this is a guest real address + */ + args_phys = kvmppc_get_gpr(vcpu, 4) & KVM_PAM; rc = kvm_read_guest(vcpu->kvm, args_phys, &args, sizeof(args)); if (rc)
The in-kernel emulation of RTAS functions needs to read the argument buffer from guest memory in order to find out what function is being requested. The guest supplies the guest physical address of the buffer, and on a real system the code that reads that buffer would run in guest real mode. In guest real mode, the processor ignores the top 4 bits of the address specified in load and store instructions. In order to emulate that behaviour correctly, we need to mask off those bits before calling kvm_read_guest() or kvm_write_guest(). This adds that masking. Signed-off-by: Paul Mackerras <paulus@samba.org> --- arch/powerpc/kvm/book3s_rtas.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-)