Skip to contentSkip to conversation

x86: kvm/paging_tmpl.h intialize the variable before using it

PATCH In review Jaswinder Singh Rajput 3 participants 8 messages May 20, 2009 – May 26, 2009

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 x86: kvm/paging_tmpl.h intialize the variable before using it

May be in some cases paging64_fetch() and paging32_fetch() will return sptep without initialization.

Also fixes compilation warning:
CC arch/x86/kernel/io_delay.o
arch/x86/kvm/paging_tmpl.h: In function ‘paging64_fetch’:
arch/x86/kvm/paging_tmpl.h:279: warning: ‘sptep’ may be used uninitialized in this function arch/x86/kvm/paging_tmpl.h: In function ‘paging32_fetch’:
arch/x86/kvm/paging_tmpl.h:279: warning: ‘sptep’ may be used uninitialized in this function

---
 arch/x86/kvm/paging_tmpl.h |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)
arch/x86/kvm/paging_tmpl.h +1 -1
index 6bd7020..99cb10d 100644
@@ -276,7 +276,7 @@ static u64 *FNAME(fetch)(struct kvm_vcpu *vcpu, gva_t addr,
 {
 	unsigned access = gw->pt_access;
 	struct kvm_mmu_page *shadow_page;
-	u64 spte, *sptep;
+	u64 spte, *sptep = NULL;
 	int direct;
 	gfn_t table_gfn;
 	int r;
Signature · 1 line
1.6.1.1
#1 lore raw
Quoted 21 lines · Jaswinder Singh Rajput

May be in some cases paging64_fetch() and paging32_fetch() will return sptep without initialization.

Also fixes compilation warning:
CC arch/x86/kernel/io_delay.o
arch/x86/kvm/paging_tmpl.h: In function ‘paging64_fetch’:
arch/x86/kvm/paging_tmpl.h:279: warning: ‘sptep’ may be used uninitialized in this function arch/x86/kvm/paging_tmpl.h: In function ‘paging32_fetch’:
arch/x86/kvm/paging_tmpl.h:279: warning: ‘sptep’ may be used uninitialized in this function

---
 arch/x86/kvm/paging_tmpl.h |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)
arch/x86/kvm/paging_tmpl.h +0 -0
index 6bd7020..99cb10d 100644

arch/x86/kvm/paging_tmpl.h static u64 *FNAME(fetch)(struct kvm_vcpu *vcpu, gva_t addr, · line 282

Quoted 8 lines
arch/x86/kvm/paging_tmpl.h +1 -1
@@ -276,7 +276,7 @@ static u64 *FNAME(fetch)(struct kvm_vcpu *vcpu, gva_t addr,
 {
 	unsigned access = gw->pt_access;
 	struct kvm_mmu_page *shadow_page;
-	u64 spte, *sptep;
+	u64 spte, *sptep = NULL;
 	int direct;
 	gfn_t table_gfn;
 	int r;

It's a false alarm. Isn't there a macro to shut up the warning?

Signature · 1 line
error compiling committee.c: too many arguments to function
#2 lore raw

Hello Avi,

Quoted 16 lines · Avi Kivity
Quoted 15 lines · Jaswinder Singh Rajput

May be in some cases paging64_fetch() and paging32_fetch() will return sptep without initialization.

Also fixes compilation warning:
CC arch/x86/kernel/io_delay.o
arch/x86/kvm/paging_tmpl.h: In function ‘paging64_fetch’:
arch/x86/kvm/paging_tmpl.h:279: warning: ‘sptep’ may be used uninitialized in this function arch/x86/kvm/paging_tmpl.h: In function ‘paging32_fetch’:
arch/x86/kvm/paging_tmpl.h:279: warning: ‘sptep’ may be used uninitialized in this function

---
 arch/x86/kvm/paging_tmpl.h |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)
Quoted 17 lines
Quoted 6 lines
arch/x86/kvm/paging_tmpl.h +0 -0
index 6bd7020..99cb10d 100644

arch/x86/kvm/paging_tmpl.h static u64 *FNAME(fetch)(struct kvm_vcpu *vcpu, gva_t addr, · line 282

Quoted 8 lines
arch/x86/kvm/paging_tmpl.h +1 -1
@@ -276,7 +276,7 @@ static u64 *FNAME(fetch)(struct kvm_vcpu *vcpu, gva_t addr,
 {
 	unsigned access = gw->pt_access;
 	struct kvm_mmu_page *shadow_page;
-	u64 spte, *sptep;
+	u64 spte, *sptep = NULL;
 	int direct;
 	gfn_t table_gfn;
 	int r;

It's a false alarm. Isn't there a macro to shut up the warning?

In arch/x86/kvm/paging_tmpl.h sptep is initialize only in :

for_each_shadow_entry(vcpu, addr, iterator) {

If we skip this, then we end up with wild sptep and returning it.

Do you still think it is a false alarm ?

--
JSR

#3 lore raw
Quoted 4 lines · Jaswinder Singh Rajput
Quoted 3 lines

It's a false alarm. Isn't there a macro to shut up the warning?

Quoted 8 lines

In arch/x86/kvm/paging_tmpl.h sptep is initialize only in :

for_each_shadow_entry(vcpu, addr, iterator) {

If we skip this, then we end up with wild sptep and returning it.

Do you still think it is a false alarm ?

That loop always has at least one iteration.

Signature · 2 lines
I have a truly marvellous patch that fixes the bug which this
signature is too narrow to contain.
#4 lore raw
Quoted 7 lines · Avi Kivity
Quoted 6 lines · Jaswinder Singh Rajput
Quoted 3 lines

It's a false alarm. Isn't there a macro to shut up the warning?

In arch/x86/kvm/paging_tmpl.h sptep is initialize only in :

Quoted 9 lines
Quoted 6 lines

for_each_shadow_entry(vcpu, addr, iterator) {

If we skip this, then we end up with wild sptep and returning it.

Do you still think it is a false alarm ?

That loop always has at least one iteration.

hmm, In that case we should use a do-while flavor function so that the compiler will also happy with it.

--
JSR

#5 lore raw
Quoted 6 lines · Jaswinder Singh Rajput
Quoted 2 lines

That loop always has at least one iteration.

hmm, In that case we should use a do-while flavor function so that the compiler will also happy with it.

My own happiness requires that the loop not be rewritten.

Signature · 2 lines
I have a truly marvellous patch that fixes the bug which this
signature is too narrow to contain.
#6 lore raw
In reply to Avi Kivity
Quoted 16 lines · Avi Kivity
Quoted 15 lines · Jaswinder Singh Rajput

May be in some cases paging64_fetch() and paging32_fetch() will return sptep without initialization.

Also fixes compilation warning:
CC arch/x86/kernel/io_delay.o
arch/x86/kvm/paging_tmpl.h: In function ‘paging64_fetch’:
arch/x86/kvm/paging_tmpl.h:279: warning: ‘sptep’ may be used uninitialized in this function arch/x86/kvm/paging_tmpl.h: In function ‘paging32_fetch’:
arch/x86/kvm/paging_tmpl.h:279: warning: ‘sptep’ may be used uninitialized in this function

---
 arch/x86/kvm/paging_tmpl.h |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)
Quoted 16 lines
Quoted 6 lines
arch/x86/kvm/paging_tmpl.h +0 -0
index 6bd7020..99cb10d 100644

arch/x86/kvm/paging_tmpl.h static u64 *FNAME(fetch)(struct kvm_vcpu *vcpu, gva_t addr, · line 282

Quoted 8 lines
arch/x86/kvm/paging_tmpl.h +1 -1
@@ -276,7 +276,7 @@ static u64 *FNAME(fetch)(struct kvm_vcpu *vcpu, gva_t addr,
 {
 	unsigned access = gw->pt_access;
 	struct kvm_mmu_page *shadow_page;
-	u64 spte, *sptep;
+	u64 spte, *sptep = NULL;
 	int direct;
 	gfn_t table_gfn;
 	int r;

It's a false alarm. Isn't there a macro to shut up the warning?

there's uninitialized_var(), but that will shut up the warning and any _correct_ future warning - so it's dangerous. Initialize it to NULL and be done with it? We dont want stray warnings in the kernel.

Ingo

#7 lore raw
Quoted 5 lines · Ingo Molnar
Quoted 2 lines

It's a false alarm. Isn't there a macro to shut up the warning?

there's uninitialized_var(), but that will shut up the warning and any _correct_ future warning - so it's dangerous. Initialize it to NULL and be done with it? We dont want stray warnings in the kernel.

Makes sense, so I've applied the patch.

Signature · 1 line
error compiling committee.c: too many arguments to function