Re: [PATCH 13/27] KVM: PPC: Magic Page Book3s support

On 07/02/2010 06:37 PM, Alexander Graf wrote:

Alexander Graf wrote:
   

quoted

We need to override EA as well as PA lookups for the magic page. When the guest
tells us to project it, the magic page overrides any guest mappings.

In order to reflect that, we need to hook into all the MMU layers of KVM to
force map the magic page if necessary.

Signed-off-by: Alexander Graf<redacted>

v1 ->  v2:

   - RMO ->  PAM
---
  arch/powerpc/kvm/book3s.c             |    7 +++++++
  arch/powerpc/kvm/book3s_32_mmu.c      |   16 ++++++++++++++++
  arch/powerpc/kvm/book3s_32_mmu_host.c |   12 ++++++++++++
  arch/powerpc/kvm/book3s_64_mmu.c      |   30 +++++++++++++++++++++++++++++-
  arch/powerpc/kvm/book3s_64_mmu_host.c |   12 ++++++++++++
  5 files changed, 76 insertions(+), 1 deletions(-)

diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index 14db032..b22e608 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c

@@ -554,6 +554,13 @@ mmio:

  static int kvmppc_visible_gfn(struct kvm_vcpu *vcpu, gfn_t gfn)
  {
+	ulong mp_pa = vcpu->arch.magic_page_pa;
+
+	if (unlikely(mp_pa)&&
+	    unlikely((mp_pa&  KVM_RMO)>>  PAGE_SHIFT == gfn)) {

     

This should be KVM_PAM :(. Should I respin the whole thing or could
whoever commits this just make that trivial change?

   

A respin followed by a bisectability test (compile each revision as it 
is applied), please.  Of course we need to resolve the detection issue 
first.

-- 
error compiling committee.c: too many arguments to function