[PATCH] net/tun: fix ioctl() based info leaks

[PATCH] net/tun: fix ioctl() based info leaks

[Mathias Krause]

The tun module leaks up to 36 bytes of memory by not initializing a
structure located on the stack that gets copied to user memory by the
TUNGETIFF and SIOCGIFHWADDR ioctl()s.

Signed-off-by: Mathias Krause <minipli@googlemail.com>
---
 drivers/net/tun.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/net/tun.c b/drivers/net/tun.c
index 987aeef..cadeb94 100644
--- a/drivers/net/tun.c
+++ b/drivers/net/tun.c
@@ -1252,9 +1252,12 @@ static long __tun_chr_ioctl(struct file *file, unsigned int cmd,
 	int vnet_hdr_sz;
 	int ret;
 
-	if (cmd == TUNSETIFF || _IOC_TYPE(cmd) == 0x89)
+	if (cmd == TUNSETIFF || _IOC_TYPE(cmd) == 0x89) {
 		if (copy_from_user(&ifr, argp, ifreq_len))
 			return -EFAULT;
+	} else {
+		memset(&ifr, 0, sizeof(ifr));
+	}
 
 	if (cmd == TUNGETFEATURES) {
 		/* Currently this just means: "what IFF flags are valid?".
-- 
1.7.10.4

Re: [PATCH] net/tun: fix ioctl() based info leaks

[richard -rw- weinberger]

On Sun, Jul 29, 2012 at 10:58 PM, Mathias Krause <minipli@googlemail.com> wrote:
> The tun module leaks up to 36 bytes of memory by not initializing a
> structure located on the stack that gets copied to user memory by the
> TUNGETIFF and SIOCGIFHWADDR ioctl()s.
>
> Signed-off-by: Mathias Krause <minipli@googlemail.com>
> ---
>  drivers/net/tun.c |    5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/net/tun.c b/drivers/net/tun.c
> index 987aeef..cadeb94 100644
> --- a/drivers/net/tun.c
> +++ b/drivers/net/tun.c
> @@ -1252,9 +1252,12 @@ static long __tun_chr_ioctl(struct file *file, unsigned int cmd,
>         int vnet_hdr_sz;
>         int ret;
>
> -       if (cmd == TUNSETIFF || _IOC_TYPE(cmd) == 0x89)
> +       if (cmd == TUNSETIFF || _IOC_TYPE(cmd) == 0x89) {
>                 if (copy_from_user(&ifr, argp, ifreq_len))
>                         return -EFAULT;
> +       } else {
> +               memset(&ifr, 0, sizeof(ifr));
> +       }
>
>         if (cmd == TUNGETFEATURES) {
>                 /* Currently this just means: "what IFF flags are valid?".

The fix makes sense to me.

Beside of the fix, why are you adding braces to if and else?
We don't use braces on single statements.

-- 
Thanks,
//richard

Re: [PATCH] net/tun: fix ioctl() based info leaks

[Mathias Krause]

On Mon, Jul 30, 2012 at 1:11 AM, richard -rw- weinberger
<richard.weinberger@gmail.com> wrote:
> On Sun, Jul 29, 2012 at 10:58 PM, Mathias Krause <minipli@googlemail.com> wrote:
>> -       if (cmd == TUNSETIFF || _IOC_TYPE(cmd) == 0x89)
>> +       if (cmd == TUNSETIFF || _IOC_TYPE(cmd) == 0x89) {
>>                 if (copy_from_user(&ifr, argp, ifreq_len))
>>                         return -EFAULT;
>> +       } else {
>> +               memset(&ifr, 0, sizeof(ifr));
>> +       }
>>
>>         if (cmd == TUNGETFEATURES) {
>>                 /* Currently this just means: "what IFF flags are valid?".
>
> The fix makes sense to me.
>
> Beside of the fix, why are you adding braces to if and else?

The pair of braces around the "if" is needed to attach the "else"
branch to the right "if". The braces around the "else" are just for
consistency.

> We don't use braces on single statements.

Even tun.c isn't consistently following this rule but I'll send a new
patch without the "else" braces.


Thanks,
Mathias

[PATCH v2] net/tun: fix ioctl() based info leaks

[Mathias Krause]

The tun module leaks up to 36 bytes of memory by not fully initializing
a structure located on the stack that gets copied to user memory by the
TUNGETIFF and SIOCGIFHWADDR ioctl()s.

Signed-off-by: Mathias Krause <minipli@googlemail.com>
---
v2:
- removed braces around else branch
- minor adjustment of the commit message

 drivers/net/tun.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/net/tun.c b/drivers/net/tun.c
index 987aeef..01255ff 100644
--- a/drivers/net/tun.c
+++ b/drivers/net/tun.c
@@ -1252,9 +1252,11 @@ static long __tun_chr_ioctl(struct file *file, unsigned int cmd,
 	int vnet_hdr_sz;
 	int ret;
 
-	if (cmd == TUNSETIFF || _IOC_TYPE(cmd) == 0x89)
+	if (cmd == TUNSETIFF || _IOC_TYPE(cmd) == 0x89) {
 		if (copy_from_user(&ifr, argp, ifreq_len))
 			return -EFAULT;
+	} else
+		memset(&ifr, 0, sizeof(ifr));
 
 	if (cmd == TUNGETFEATURES) {
 		/* Currently this just means: "what IFF flags are valid?".
-- 
1.7.10.4

Re: [PATCH v2] net/tun: fix ioctl() based info leaks

[David Miller]

From: Mathias Krause <minipli@googlemail.com>
Date: Mon, 30 Jul 2012 07:45:14 +0200

> The tun module leaks up to 36 bytes of memory by not fully initializing
> a structure located on the stack that gets copied to user memory by the
> TUNGETIFF and SIOCGIFHWADDR ioctl()s.
> 
> Signed-off-by: Mathias Krause <minipli@googlemail.com>

Applied and queued up for -stable, thanks.

Re: [PATCH v2] net/tun: fix ioctl() based info leaks

[Eric Dumazet]

On Mon, 2012-07-30 at 07:45 +0200, Mathias Krause wrote:
> The tun module leaks up to 36 bytes of memory by not fully initializing
> a structure located on the stack that gets copied to user memory by the
> TUNGETIFF and SIOCGIFHWADDR ioctl()s.
> 
> Signed-off-by: Mathias Krause <minipli@googlemail.com>
> ---
> v2:
> - removed braces around else branch
> - minor adjustment of the commit message
> 
>  drivers/net/tun.c |    4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/net/tun.c b/drivers/net/tun.c
> index 987aeef..01255ff 100644
> --- a/drivers/net/tun.c
> +++ b/drivers/net/tun.c
> @@ -1252,9 +1252,11 @@ static long __tun_chr_ioctl(struct file *file, unsigned int cmd,
>  	int vnet_hdr_sz;
>  	int ret;
>  
> -	if (cmd == TUNSETIFF || _IOC_TYPE(cmd) == 0x89)
> +	if (cmd == TUNSETIFF || _IOC_TYPE(cmd) == 0x89) {
>  		if (copy_from_user(&ifr, argp, ifreq_len))
>  			return -EFAULT;
> +	} else
> +		memset(&ifr, 0, sizeof(ifr));
>  
>  	if (cmd == TUNGETFEATURES) {
>  		/* Currently this just means: "what IFF flags are valid?".


Actually braces were better

vi +169 Documentation/CodingStyle

This does not apply if only one branch of a conditional statement is a
single
statement; in the latter case use braces in both branches:

if (condition) {
        do_this();
        do_that();
} else {
        otherwise();
}

Re: [PATCH v2] net/tun: fix ioctl() based info leaks

[David Miller]

From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Mon, 30 Jul 2012 08:22:20 +0200

> On Mon, 2012-07-30 at 07:45 +0200, Mathias Krause wrote:
>> The tun module leaks up to 36 bytes of memory by not fully initializing
>> a structure located on the stack that gets copied to user memory by the
>> TUNGETIFF and SIOCGIFHWADDR ioctl()s.
>> 
>> Signed-off-by: Mathias Krause <minipli@googlemail.com>
>> ---
>> v2:
>> - removed braces around else branch
>> - minor adjustment of the commit message
>> 
>>  drivers/net/tun.c |    4 +++-
>>  1 file changed, 3 insertions(+), 1 deletion(-)
>> 
>> diff --git a/drivers/net/tun.c b/drivers/net/tun.c
>> index 987aeef..01255ff 100644
>> --- a/drivers/net/tun.c
>> +++ b/drivers/net/tun.c
>> @@ -1252,9 +1252,11 @@ static long __tun_chr_ioctl(struct file *file, unsigned int cmd,
>>  	int vnet_hdr_sz;
>>  	int ret;
>>  
>> -	if (cmd == TUNSETIFF || _IOC_TYPE(cmd) == 0x89)
>> +	if (cmd == TUNSETIFF || _IOC_TYPE(cmd) == 0x89) {
>>  		if (copy_from_user(&ifr, argp, ifreq_len))
>>  			return -EFAULT;
>> +	} else
>> +		memset(&ifr, 0, sizeof(ifr));
>>  
>>  	if (cmd == TUNGETFEATURES) {
>>  		/* Currently this just means: "what IFF flags are valid?".
> 
> 
> Actually braces were better
> 
> vi +169 Documentation/CodingStyle
> 
> This does not apply if only one branch of a conditional statement is a
> single
> statement; in the latter case use braces in both branches:
> 
> if (condition) {
>         do_this();
>         do_that();
> } else {
>         otherwise();
> }

Ok I'll fix this up myself.