KVM: MMU: allow more page become unsync at gfn mapping time
authorXiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Mon, 24 May 2010 07:40:07 +0000 (15:40 +0800)
committerAvi Kivity <avi@redhat.com>
Sun, 1 Aug 2010 07:35:51 +0000 (10:35 +0300)
commit9cf5cf5ad43b293581e5b87678ea5783c06d1a41
treebe4a4ad779ff4330e016eb390653e526d90a17db
parent221d059d15f1c8bd070a63fd45cd8d2598af5f99
KVM: MMU: allow more page become unsync at gfn mapping time

In current code, shadow page can become asynchronous only if one
shadow page for a gfn, this rule is too strict, in fact, we can
let all last mapping page(i.e, it's the pte page) become unsync,
and sync them at invlpg or flush tlb time.

This patch allow more page become asynchronous at gfn mapping time

Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
arch/x86/kvm/mmu.c