When a guest sets its SR entry to invalid, we may still find a corresponding entry in a BAT. So we need to make sure we're not faulting on invalid SR entries, but instead just claim them to be BAT resolved. This resolves breakage experienced when using libogc based guests. Signed-off-by: Alexander Graf <agraf@xxxxxxx> --- arch/powerpc/kvm/book3s_32_mmu.c | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-) diff --git a/arch/powerpc/kvm/book3s_32_mmu.c b/arch/powerpc/kvm/book3s_32_mmu.c index 0b10503..3292d76 100644 --- a/arch/powerpc/kvm/book3s_32_mmu.c +++ b/arch/powerpc/kvm/book3s_32_mmu.c @@ -354,10 +354,10 @@ static int kvmppc_mmu_book3s_32_esid_to_vsid(struct kvm_vcpu *vcpu, ulong esid, *vsid = VSID_REAL_DR | gvsid; break; case MSR_DR|MSR_IR: - if (!sr->valid) - return -1; - - *vsid = sr->vsid; + if (sr->valid) + *vsid = sr->vsid; + else + *vsid = VSID_BAT | gvsid; break; default: BUG(); -- 1.6.0.2 -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
