From: Xiao Guangrong <guangrong.xiao@linux.intel.com>
To: pbonzini@redhat.com
Cc: gleb@kernel.org, mtosatti@redhat.com, kvm@vger.kernel.org,
linux-kernel@vger.kernel.org, dmatlack@google.com,
Xiao Guangrong <guangrong.xiao@linux.intel.com>
Subject: [PATCH v2 15/15] KVM: MTRR: do not map huage page for non-consistent range
Date: Mon, 15 Jun 2015 16:55:35 +0800 [thread overview]
Message-ID: <1434358535-19923-16-git-send-email-guangrong.xiao@linux.intel.com> (raw)
In-Reply-To: <1434358535-19923-1-git-send-email-guangrong.xiao@linux.intel.com>
Based on Intel's SDM, mapping huge page which do not have consistent
memory cache for each 4k page will cause undefined behavior
In order to avoiding this kind of undefined behavior, we force to use
4k pages under this case
Signed-off-by: Xiao Guangrong <guangrong.xiao@linux.intel.com>
---
arch/x86/kvm/mmu.c | 20 +++++++++++++++++++-
arch/x86/kvm/mtrr.c | 29 +++++++++++++++++++++++++++++
arch/x86/kvm/x86.h | 2 ++
3 files changed, 50 insertions(+), 1 deletion(-)
diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c
index 532aad2..f807496 100644
--- a/arch/x86/kvm/mmu.c
+++ b/arch/x86/kvm/mmu.c
@@ -3446,6 +3446,16 @@ static bool try_async_pf(struct kvm_vcpu *vcpu, bool prefault, gfn_t gfn,
return false;
}
+static bool
+check_hugepage_cache_consistency(struct kvm_vcpu *vcpu, gfn_t gfn, int level)
+{
+ int page_num = KVM_PAGES_PER_HPAGE(level);
+
+ gfn &= ~(page_num - 1);
+
+ return kvm_mtrr_check_gfn_range_consistency(vcpu, gfn, page_num);
+}
+
static int tdp_page_fault(struct kvm_vcpu *vcpu, gva_t gpa, u32 error_code,
bool prefault)
{
@@ -3471,9 +3481,17 @@ static int tdp_page_fault(struct kvm_vcpu *vcpu, gva_t gpa, u32 error_code,
if (r)
return r;
- force_pt_level = mapping_level_dirty_bitmap(vcpu, gfn);
+ if (mapping_level_dirty_bitmap(vcpu, gfn) ||
+ !check_hugepage_cache_consistency(vcpu, gfn, PT_DIRECTORY_LEVEL))
+ force_pt_level = 1;
+ else
+ force_pt_level = 0;
+
if (likely(!force_pt_level)) {
level = mapping_level(vcpu, gfn);
+ if (level > PT_DIRECTORY_LEVEL &&
+ !check_hugepage_cache_consistency(vcpu, gfn, level))
+ level = PT_DIRECTORY_LEVEL;
gfn &= ~(KVM_PAGES_PER_HPAGE(level) - 1);
} else
level = PT_PAGE_TABLE_LEVEL;
diff --git a/arch/x86/kvm/mtrr.c b/arch/x86/kvm/mtrr.c
index 097d616..5b26048 100644
--- a/arch/x86/kvm/mtrr.c
+++ b/arch/x86/kvm/mtrr.c
@@ -677,3 +677,32 @@ u8 kvm_mtrr_get_guest_memory_type(struct kvm_vcpu *vcpu, gfn_t gfn)
return type;
}
EXPORT_SYMBOL_GPL(kvm_mtrr_get_guest_memory_type);
+
+bool kvm_mtrr_check_gfn_range_consistency(struct kvm_vcpu *vcpu, gfn_t gfn,
+ int page_num)
+{
+ struct kvm_mtrr *mtrr_state = &vcpu->arch.mtrr_state;
+ struct mtrr_iter iter;
+ u64 start, end;
+ int type = -1;
+
+ start = gfn_to_gpa(gfn);
+ end = gfn_to_gpa(gfn + page_num);
+ mtrr_for_each_mem_type(&iter, mtrr_state, start, end) {
+ if (type == -1) {
+ type = iter.mem_type;
+ continue;
+ }
+
+ if (type != iter.mem_type)
+ return false;
+ }
+
+ if (!iter.partial_map)
+ return true;
+
+ if (type == -1)
+ return true;
+
+ return type == mtrr_default_type(mtrr_state);
+}
diff --git a/arch/x86/kvm/x86.h b/arch/x86/kvm/x86.h
index 0e4727c..edc8cdc 100644
--- a/arch/x86/kvm/x86.h
+++ b/arch/x86/kvm/x86.h
@@ -167,6 +167,8 @@ u8 kvm_mtrr_get_guest_memory_type(struct kvm_vcpu *vcpu, gfn_t gfn);
bool kvm_mtrr_valid(struct kvm_vcpu *vcpu, u32 msr, u64 data);
int kvm_mtrr_set_msr(struct kvm_vcpu *vcpu, u32 msr, u64 data);
int kvm_mtrr_get_msr(struct kvm_vcpu *vcpu, u32 msr, u64 *pdata);
+bool kvm_mtrr_check_gfn_range_consistency(struct kvm_vcpu *vcpu, gfn_t gfn,
+ int page_num);
#define KVM_SUPPORTED_XCR0 (XSTATE_FP | XSTATE_SSE | XSTATE_YMM \
| XSTATE_BNDREGS | XSTATE_BNDCSR \
--
2.1.0
prev parent reply other threads:[~2015-06-15 9:01 UTC|newest]
Thread overview: 25+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-06-15 8:55 [PATCH v2 00/15] vMTRR bugfix and optimization Xiao Guangrong
2015-06-15 8:55 ` [PATCH v2 01/15] KVM: x86: fix CR0.CD virtualization Xiao Guangrong
2015-06-15 8:55 ` [PATCH v2 02/15] KVM: x86: move MTRR related code to a separate file Xiao Guangrong
2015-06-15 8:55 ` [PATCH v2 03/15] KVM: MTRR: handle MSR_MTRRcap in kvm_mtrr_get_msr Xiao Guangrong
2015-06-15 8:55 ` [PATCH v2 04/15] KVM: MTRR: remove mtrr_state.have_fixed Xiao Guangrong
2015-06-15 8:55 ` [PATCH v2 05/15] KVM: MTRR: exactly define the size of variable MTRRs Xiao Guangrong
2015-06-15 8:55 ` [PATCH v2 06/15] KVM: MTRR: clean up mtrr default type Xiao Guangrong
2015-06-15 8:55 ` [PATCH v2 07/15] KVM: MTRR: do not split 64 bits MSR content Xiao Guangrong
2015-06-15 8:55 ` [PATCH v2 08/15] KVM: MTRR: improve kvm_mtrr_get_guest_memory_type Xiao Guangrong
2015-06-15 8:55 ` [PATCH v2 09/15] KVM: MTRR: introduce fixed_mtrr_segment table Xiao Guangrong
2015-06-15 8:55 ` [PATCH v2 10/15] KVM: MTRR: introduce var_mtrr_range Xiao Guangrong
2015-06-17 15:38 ` Paolo Bonzini
2015-06-15 8:55 ` [PATCH v2 11/15] KVM: MTRR: sort variable MTRRs Xiao Guangrong
2015-06-17 15:40 ` Paolo Bonzini
2015-06-17 16:11 ` Paolo Bonzini
2015-06-22 11:24 ` Paolo Bonzini
2015-06-22 11:24 ` Paolo Bonzini
2015-06-23 2:29 ` Xiao Guangrong
2015-06-23 2:29 ` Xiao Guangrong
2015-06-23 8:00 ` Paolo Bonzini
2015-06-23 8:27 ` Xiao Guangrong
2015-06-15 8:55 ` [PATCH v2 12/15] KVM: MTRR: introduce fixed_mtrr_addr_* functions Xiao Guangrong
2015-06-15 8:55 ` [PATCH v2 13/15] KVM: MTRR: introduce mtrr_for_each_mem_type Xiao Guangrong
2015-06-15 8:55 ` [PATCH v2 14/15] KVM: MTRR: simplify kvm_mtrr_get_guest_memory_type Xiao Guangrong
2015-06-15 8:55 ` Xiao Guangrong [this message]
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=1434358535-19923-16-git-send-email-guangrong.xiao@linux.intel.com \
--to=guangrong.xiao@linux.intel.com \
--cc=dmatlack@google.com \
--cc=gleb@kernel.org \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mtosatti@redhat.com \
--cc=pbonzini@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.