[v5,6/5] KVM: x86: clear write_fault_to_shadow_pgtable explicitly

Commit Message

Xiao Guangrong Jan. 11, 2013, 2:12 p.m. UTC

On 01/11/2013 09:15 PM, Marcelo Tosatti wrote:

> 
> This is cryptic. Its not obvious at all for someone modifying the code, 
> for example.
> 
> Can you please clear it explicitly? 

Sure, this is the patch to apply your idea, is it good to you? :)

Comments

Marcelo Tosatti Jan. 11, 2013, 7:09 p.m. UTC | #1

On Fri, Jan 11, 2013 at 10:12:58PM +0800, Xiao Guangrong wrote:
> On 01/11/2013 09:15 PM, Marcelo Tosatti wrote:
> 
> > 
> > This is cryptic. Its not obvious at all for someone modifying the code, 
> > for example.
> > 
> > Can you please clear it explicitly? 
> 
> Sure, this is the patch to apply your idea, is it good to you? :)
> 
> ============================================
> Subject: [PATCH 6/6] KVM: x86: clear write_fault_to_shadow_pgtable explicitly
> 
> Clear it explicitly when exiting x86_emulate_instruction to clarify the code,
> it is suggested by Marcelo
> 
> Signed-off-by: Xiao Guangrong <[e-mail]>
> ---
>  arch/x86/kvm/x86.c |   16 ++++++++++++----
>  1 files changed, 12 insertions(+), 4 deletions(-)

Fine, please rebase against queue.

--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to [e-mail]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Patch

============================================
Subject: [PATCH 6/6] KVM: x86: clear write_fault_to_shadow_pgtable explicitly

Clear it explicitly when exiting x86_emulate_instruction to clarify the code,
it is suggested by Marcelo

Signed-off-by: Xiao Guangrong <[e-mail]>
---
 arch/x86/kvm/x86.c |   16 ++++++++++++----
 1 files changed, 12 insertions(+), 4 deletions(-)

diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 2957012..89d01a8 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -4753,7 +4753,8 @@  static int handle_emulation_failure(struct kvm_vcpu *vcpu)
 	return r;
 }

-static bool reexecute_instruction(struct kvm_vcpu *vcpu, gva_t cr2)
+static bool reexecute_instruction(struct kvm_vcpu *vcpu, gva_t cr2,
+				  bool write_fault_to_shadow_pgtable)
 {
 	gpa_t gpa = cr2;
 	pfn_t pfn;
@@ -4816,7 +4817,7 @@  static bool reexecute_instruction(struct kvm_vcpu *vcpu, gva_t cr2)
 	 * be fixed by unprotecting shadow page and it should
 	 * be reported to userspace.
 	 */
-	return !vcpu->arch.write_fault_to_shadow_pgtable;
+	return !write_fault_to_shadow_pgtable;
 }

 static bool retry_instruction(struct x86_emulate_ctxt *ctxt,
@@ -4875,7 +4876,13 @@  int x86_emulate_instruction(struct kvm_vcpu *vcpu,
 	int r;
 	struct x86_emulate_ctxt *ctxt = &vcpu->arch.emulate_ctxt;
 	bool writeback = true;
+	bool write_fault_to_spt = vcpu->arch.write_fault_to_shadow_pgtable;

+	/*
+	 * Clear write_fault_to_shadow_pgtable here to ensure it is
+	 * never reused.
+	 */
+	vcpu->arch.write_fault_to_shadow_pgtable = false;
 	kvm_clear_exception_queue(vcpu);

 	if (!(emulation_type & EMULTYPE_NO_DECODE)) {
@@ -4894,7 +4901,8 @@  int x86_emulate_instruction(struct kvm_vcpu *vcpu,
 		if (r != EMULATION_OK)  {
 			if (emulation_type & EMULTYPE_TRAP_UD)
 				return EMULATE_FAIL;
-			if (reexecute_instruction(vcpu, cr2))
+			if (reexecute_instruction(vcpu, cr2,
+						  write_fault_to_spt))
 				return EMULATE_DONE;
 			if (emulation_type & EMULTYPE_SKIP)
 				return EMULATE_FAIL;
@@ -4924,7 +4932,7 @@  restart:
 		return EMULATE_DONE;

 	if (r == EMULATION_FAILED) {
-		if (reexecute_instruction(vcpu, cr2))
+		if (reexecute_instruction(vcpu, cr2, write_fault_to_spt))
 			return EMULATE_DONE;

 		return handle_emulation_failure(vcpu);