lkml.org 
[lkml]   [2011]   [Jun]   [9]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH 03/15] KVM: x86: avoid unnecessarily guest page table walking
On 06/07/2011 03:59 PM, Xiao Guangrong wrote:
> We already get the guest physical address, so use it to read guest data
> directly to avoid walking guest page table again
>
> Signed-off-by: Xiao Guangrong<xiaoguangrong@cn.fujitsu.com>
> ---
> arch/x86/kvm/x86.c | 3 +--
> 1 files changed, 1 insertions(+), 2 deletions(-)
>
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index 694538a..8be9ff6 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -3930,8 +3930,7 @@ static int emulator_read_emulated(struct x86_emulate_ctxt *ctxt,
> if ((gpa& PAGE_MASK) == APIC_DEFAULT_PHYS_BASE)
> goto mmio;
>
> - if (kvm_read_guest_virt(ctxt, addr, val, bytes, exception)
> - == X86EMUL_CONTINUE)
> + if (!kvm_read_guest(vcpu->kvm, gpa, val, bytes))
> return X86EMUL_CONTINUE;

This breaks is addr/bytes spans a page boundary.

(the current code is also broken, but only for mmio; the new code is
broken for ram as well).

We need a gva_to_gpa() that returns a range of pages.

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



\
 
 \ /
  Last update: 2011-06-09 09:03    [W:0.221 / U:13.588 seconds]
©2003-2017 Jasper Spaans. hosted at Digital OceanAdvertise on this site