From: Joerg Roedel <joerg.roedel@amd.com>
To: Avi Kivity <avi@redhat.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>,
Alexander Graf <agraf@suse.de>,
kvm@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH 11/18] KVM: MMU: Add infrastructure for two-level page walker
Date: Wed, 10 Mar 2010 15:46:00 +0100 [thread overview]
Message-ID: <20100310144600.GL5619@amd.com> (raw)
In-Reply-To: <4B94C552.30604@redhat.com>
On Mon, Mar 08, 2010 at 11:37:22AM +0200, Avi Kivity wrote:
> On 03/03/2010 09:12 PM, Joerg Roedel wrote:
> >This patch introduces a mmu-callback to translate gpa
> >addresses in the walk_addr code. This is later used to
> >translate l2_gpa addresses into l1_gpa addresses.
> >
> >Signed-off-by: Joerg Roedel<joerg.roedel@amd.com>
> >---
> > arch/x86/include/asm/kvm_host.h | 1 +
> > arch/x86/kvm/mmu.c | 7 +++++++
> > arch/x86/kvm/paging_tmpl.h | 19 +++++++++++++++++++
> > include/linux/kvm_host.h | 5 +++++
> > 4 files changed, 32 insertions(+), 0 deletions(-)
> >
> >diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
> >index c0b5576..76c8b5f 100644
> >--- a/arch/x86/include/asm/kvm_host.h
> >+++ b/arch/x86/include/asm/kvm_host.h
> >@@ -250,6 +250,7 @@ struct kvm_mmu {
> > void (*free)(struct kvm_vcpu *vcpu);
> > gpa_t (*gva_to_gpa)(struct kvm_vcpu *vcpu, gva_t gva, u32 access,
> > u32 *error);
> >+ gpa_t (*translate_gpa)(struct kvm_vcpu *vcpu, gpa_t gpa, u32 *error);
> > void (*prefetch_page)(struct kvm_vcpu *vcpu,
> > struct kvm_mmu_page *page);
> > int (*sync_page)(struct kvm_vcpu *vcpu,
>
> I think placing this here means we will miss a few translations,
> namely when we do a physical access (say, reading PDPTEs or
> similar).
>
> We need to do this on the level of kvm_read_guest() so we capture
> physical accesses:
>
> kvm_read_guest_virt
> -> walk_addr
> -> kvm_read_guest_tdp
> -> kvm_read_guest_virt
> -> walk_addr
> -> kvm_read_guest_tdp
> -> kvm_read_guest
>
> Of course, not all accesses will use kvm_read_guest_tdp; for example
> kvmclock accesses should still go untranslated.
Ok, doing the translation in kvm_read_guest is certainly the more
generic approach. I already fixed a bug related to loading l2 pdptr
pointers. Doing the translation in kvm_read_guest makes the code a lot
nicer.
Joerg
next prev parent reply other threads:[~2010-03-10 14:46 UTC|newest]
Thread overview: 49+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-03-03 19:12 [PATCH 0/18][RFC] Nested Paging support for Nested SVM (aka NPT-Virtualization) Joerg Roedel
2010-03-03 19:12 ` [PATCH 01/18] KVM: MMU: Check for root_level instead of long mode Joerg Roedel
2010-03-03 19:12 ` [PATCH 02/18] KVM: MMU: Make tdp_enabled a mmu-context parameter Joerg Roedel
2010-03-08 9:17 ` Avi Kivity
2010-03-10 14:44 ` Joerg Roedel
2010-03-10 14:53 ` Avi Kivity
2010-03-10 15:26 ` Joerg Roedel
2010-03-11 6:47 ` Avi Kivity
2010-03-11 10:33 ` Joerg Roedel
2010-03-03 19:12 ` [PATCH 03/18] KVM: MMU: Make set_cr3 a function pointer in kvm_mmu Joerg Roedel
2010-03-03 19:12 ` [PATCH 04/18] KVM: X86: Introduce a tdp_set_cr3 function Joerg Roedel
2010-03-03 19:12 ` [PATCH 05/18] KVM: MMU: Introduce get_cr3 function pointer Joerg Roedel
2010-03-03 19:12 ` [PATCH 06/18] KVM: MMU: Introduce inject_page_fault " Joerg Roedel
2010-03-03 19:12 ` [PATCH 07/18] KVM: SVM: Implement MMU helper functions for Nested Nested Paging Joerg Roedel
2010-03-03 19:12 ` [PATCH 08/18] KVM: MMU: Change init_kvm_softmmu to take a context as parameter Joerg Roedel
2010-03-03 19:12 ` [PATCH 09/18] KVM: MMU: Let is_rsvd_bits_set take mmu context instead of vcpu Joerg Roedel
2010-03-03 19:12 ` [PATCH 10/18] KVM: MMU: Introduce generic walk_addr function Joerg Roedel
2010-03-03 19:12 ` [PATCH 11/18] KVM: MMU: Add infrastructure for two-level page walker Joerg Roedel
2010-03-08 9:37 ` Avi Kivity
2010-03-10 14:46 ` Joerg Roedel [this message]
2010-03-03 19:12 ` [PATCH 12/18] KVM: MMU: Implement nested gva_to_gpa functions Joerg Roedel
2010-03-03 19:12 ` [PATCH 13/18] KVM: MMU: Introduce Nested MMU context Joerg Roedel
2010-03-03 19:12 ` [PATCH 14/18] KVM: SVM: Initialize Nested Nested MMU context on VMRUN Joerg Roedel
2010-03-03 19:12 ` [PATCH 15/18] KVM: MMU: Propagate the right fault back to the guest after gva_to_gpa Joerg Roedel
2010-03-15 4:30 ` Daniel K.
2010-03-15 12:52 ` Joerg Roedel
2010-03-15 7:36 ` Avi Kivity
2010-03-15 9:06 ` Joerg Roedel
2010-03-15 9:23 ` Avi Kivity
2010-03-15 9:41 ` Joerg Roedel
2010-03-03 19:12 ` [PATCH 16/18] KVM: X86: Add callback to let modules decide over some supported cpuid bits Joerg Roedel
2010-03-03 19:12 ` [PATCH 17/18] KVM: SVM: Report Nested Paging support to userspace Joerg Roedel
2010-03-03 23:37 ` Alexander Graf
2010-03-04 11:27 ` Joerg Roedel
2010-03-03 19:12 ` [PATCH 18/18] KVM: X86: Add KVM_CAP_SVM_CPUID_FIXED Joerg Roedel
2010-03-08 9:39 ` Avi Kivity
2010-03-10 14:46 ` Joerg Roedel
2010-03-03 23:10 ` [PATCH 0/18][RFC] Nested Paging support for Nested SVM (aka NPT-Virtualization) Jan Kiszka
2010-03-03 23:44 ` Alexander Graf
2010-03-04 11:29 ` Joerg Roedel
2010-03-04 0:35 ` Anthony Liguori
2010-03-04 14:42 ` Marcelo Tosatti
2010-03-04 15:58 ` Joerg Roedel
2010-03-11 20:58 ` Marcelo Tosatti
2010-03-12 7:36 ` Avi Kivity
2010-03-15 6:27 ` Marcelo Tosatti
2010-03-15 7:34 ` Avi Kivity
2010-03-12 7:41 ` Avi Kivity
2010-03-08 9:41 ` Avi Kivity
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20100310144600.GL5619@amd.com \
--to=joerg.roedel@amd.com \
--cc=agraf@suse.de \
--cc=avi@redhat.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mtosatti@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).