[PATCH] missing ->mmap_sem around find_vma() in swp_emulate.c

[PATCH] missing ->mmap_sem around find_vma() in swp_emulate.c

[Al Viro]

find_vma() is *not* safe when somebody else is removing vmas.  Not just
the return value might get bogus just as you are getting it (this instance
doesn't try to dereference the resulting vma), the search itself can get
buggered in rather spectacular ways.  IOW, ->mmap_sem really, really is
not optional here.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
diff --git a/arch/arm/kernel/swp_emulate.c b/arch/arm/kernel/swp_emulate.c
index df74518..ab1017b 100644
--- a/arch/arm/kernel/swp_emulate.c
+++ b/arch/arm/kernel/swp_emulate.c
@@ -109,10 +109,12 @@ static void set_segfault(struct pt_regs *regs, unsigned long addr)
 {
 	siginfo_t info;
 
+	down_read(&current->mm->mmap_sem);
 	if (find_vma(current->mm, addr) == NULL)
 		info.si_code = SEGV_MAPERR;
 	else
 		info.si_code = SEGV_ACCERR;
+	up_read(&current->mm->mmap_sem);
 
 	info.si_signo = SIGSEGV;
 	info.si_errno = 0;

[PATCH] missing ->mmap_sem around find_vma() in swp_emulate.c

[Dirk Behme]

Am 16.12.2012 01:25, schrieb Al Viro:
> find_vma() is *not* safe when somebody else is removing vmas.  Not just
> the return value might get bogus just as you are getting it (this instance
> doesn't try to dereference the resulting vma), the search itself can get
> buggered in rather spectacular ways.  IOW, ->mmap_sem really, really is
> not optional here.
>
> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
> ---
> diff --git a/arch/arm/kernel/swp_emulate.c b/arch/arm/kernel/swp_emulate.c
> index df74518..ab1017b 100644
> --- a/arch/arm/kernel/swp_emulate.c
> +++ b/arch/arm/kernel/swp_emulate.c
> @@ -109,10 +109,12 @@ static void set_segfault(struct pt_regs *regs, unsigned long addr)
>   {
>   	siginfo_t info;
>
> +	down_read(&current->mm->mmap_sem);
>   	if (find_vma(current->mm, addr) == NULL)
>   		info.si_code = SEGV_MAPERR;
>   	else
>   		info.si_code = SEGV_ACCERR;
> +	up_read(&current->mm->mmap_sem);
>
>   	info.si_signo = SIGSEGV;
>   	info.si_errno = 0;

Any comment on this?

Thanks

Dirk

[PATCH] missing ->mmap_sem around find_vma() in swp_emulate.c

[Russell King - ARM Linux]

On Thu, Dec 20, 2012 at 08:15:47AM +0100, Dirk Behme wrote:
> Am 16.12.2012 01:25, schrieb Al Viro:
>> find_vma() is *not* safe when somebody else is removing vmas.  Not just
>> the return value might get bogus just as you are getting it (this instance
>> doesn't try to dereference the resulting vma), the search itself can get
>> buggered in rather spectacular ways.  IOW, ->mmap_sem really, really is
>> not optional here.
>>
>> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
>> ---
>> diff --git a/arch/arm/kernel/swp_emulate.c b/arch/arm/kernel/swp_emulate.c
>> index df74518..ab1017b 100644
>> --- a/arch/arm/kernel/swp_emulate.c
>> +++ b/arch/arm/kernel/swp_emulate.c
>> @@ -109,10 +109,12 @@ static void set_segfault(struct pt_regs *regs, unsigned long addr)
>>   {
>>   	siginfo_t info;
>>
>> +	down_read(&current->mm->mmap_sem);
>>   	if (find_vma(current->mm, addr) == NULL)
>>   		info.si_code = SEGV_MAPERR;
>>   	else
>>   		info.si_code = SEGV_ACCERR;
>> +	up_read(&current->mm->mmap_sem);
>>
>>   	info.si_signo = SIGSEGV;
>>   	info.si_errno = 0;
>
> Any comment on this?

Any comment required on this?  No, it's from Al Viro, which means it's
(mostly) always correct.  I'll look at merging it soon.