[PATCH] can: Replace zero-length array with flexible-array

[PATCH] can: Replace zero-length array with flexible-array

[Gustavo A. R. Silva]

The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:

struct foo {
        int stuff;
        struct boo array[];
};

By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.

Also, notice that, dynamic memory allocations won't be affected by
this change:

"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]

sizeof(flexible-array-member) triggers a warning because flexible array
members have incomplete type[1]. There are some instances of code in
which the sizeof operator is being incorrectly/erroneously applied to
zero-length arrays and the result is zero. Such instances may be hiding
some bugs. So, this work (flexible-array member conversions) will also
help to get completely rid of those sorts of issues.

This issue was found with the help of Coccinelle.

[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit 76497732932f ("cxgb3/l2t: Fix undefined behaviour")

Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
---
 include/linux/can/skb.h |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/include/linux/can/skb.h b/include/linux/can/skb.h
index a954def26c0d..900b9f4e0605 100644
--- a/include/linux/can/skb.h
+++ b/include/linux/can/skb.h
@@ -34,7 +34,7 @@
 struct can_skb_priv {
 	int ifindex;
 	int skbcnt;
-	struct can_frame cf[0];
+	struct can_frame cf[];
 };
 
 static inline struct can_skb_priv *can_skb_prv(struct sk_buff *skb)

Re: [PATCH] can: Replace zero-length array with flexible-array

[Oliver Hartkopp]

On 2020-05-07 20:51, Gustavo A. R. Silva wrote:
> The current codebase makes use of the zero-length array language
> extension to the C90 standard, but the preferred mechanism to declare
> variable-length types such as these ones is a flexible array member[1][2],
> introduced in C99:
> 
> struct foo {
>          int stuff;
>          struct boo array[];
> };
> 
> By making use of the mechanism above, we will get a compiler warning
> in case the flexible array does not occur last in the structure, which
> will help us prevent some kind of undefined behavior bugs from being
> inadvertently introduced[3] to the codebase from now on.
> 
> Also, notice that, dynamic memory allocations won't be affected by
> this change:
> 
> "Flexible array members have incomplete type, and so the sizeof operator
> may not be applied. As a quirk of the original implementation of
> zero-length arrays, sizeof evaluates to zero."[1]
> 
> sizeof(flexible-array-member) triggers a warning because flexible array
> members have incomplete type[1]. There are some instances of code in
> which the sizeof operator is being incorrectly/erroneously applied to
> zero-length arrays and the result is zero. Such instances may be hiding
> some bugs. So, this work (flexible-array member conversions) will also
> help to get completely rid of those sorts of issues.
> 
> This issue was found with the help of Coccinelle.
> 
> [1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
> [2] https://github.com/KSPP/linux/issues/21
> [3] commit 76497732932f ("cxgb3/l2t: Fix undefined behaviour")
> 
> Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
> ---
>   include/linux/can/skb.h |    2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/include/linux/can/skb.h b/include/linux/can/skb.h
> index a954def26c0d..900b9f4e0605 100644
> --- a/include/linux/can/skb.h
> +++ b/include/linux/can/skb.h
> @@ -34,7 +34,7 @@
>   struct can_skb_priv {
>   	int ifindex;
>   	int skbcnt;
> -	struct can_frame cf[0];
> +	struct can_frame cf[];
>   };
>   
>   static inline struct can_skb_priv *can_skb_prv(struct sk_buff *skb)
> 

Acked-by: Oliver Hartkopp <socketcan@hartkopp.net>

@Gustavo: Just to be sure:

 From the referenced URLs I got the information that the sizeof() 
operator causes problems when applied to e.g. cf[0].

We don't have this case in our code - but one question remains to me:

We are using the above construct to ensure the padding between the two 
'int' values and the struct can_frame which enforces a 64 bit alignment.

This intention is not affected by the patch, right?

Best,
Oliver

Re: [PATCH] can: Replace zero-length array with flexible-array

[Gustavo A. R. Silva]

Hi Oliver,

Sorry for the late reply. I totally lost track of this thread. :/
Please, see my comments below...

On 5/12/20 08:30, Oliver Hartkopp wrote:
> 
> 
> On 2020-05-07 20:51, Gustavo A. R. Silva wrote:
>> The current codebase makes use of the zero-length array language
>> extension to the C90 standard, but the preferred mechanism to declare
>> variable-length types such as these ones is a flexible array member[1][2],
>> introduced in C99:
>>
>> struct foo {
>>          int stuff;
>>          struct boo array[];
>> };
>>
>> By making use of the mechanism above, we will get a compiler warning
>> in case the flexible array does not occur last in the structure, which
>> will help us prevent some kind of undefined behavior bugs from being
>> inadvertently introduced[3] to the codebase from now on.
>>
>> Also, notice that, dynamic memory allocations won't be affected by
>> this change:
>>
>> "Flexible array members have incomplete type, and so the sizeof operator
>> may not be applied. As a quirk of the original implementation of
>> zero-length arrays, sizeof evaluates to zero."[1]
>>
>> sizeof(flexible-array-member) triggers a warning because flexible array
>> members have incomplete type[1]. There are some instances of code in
>> which the sizeof operator is being incorrectly/erroneously applied to
>> zero-length arrays and the result is zero. Such instances may be hiding
>> some bugs. So, this work (flexible-array member conversions) will also
>> help to get completely rid of those sorts of issues.
>>
>> This issue was found with the help of Coccinelle.
>>
>> [1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
>> [2] https://github.com/KSPP/linux/issues/21
>> [3] commit 76497732932f ("cxgb3/l2t: Fix undefined behaviour")
>>
>> Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
>> ---
>>   include/linux/can/skb.h |    2 +-
>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/include/linux/can/skb.h b/include/linux/can/skb.h
>> index a954def26c0d..900b9f4e0605 100644
>> --- a/include/linux/can/skb.h
>> +++ b/include/linux/can/skb.h
>> @@ -34,7 +34,7 @@
>>   struct can_skb_priv {
>>       int ifindex;
>>       int skbcnt;
>> -    struct can_frame cf[0];
>> +    struct can_frame cf[];
>>   };
>>     static inline struct can_skb_priv *can_skb_prv(struct sk_buff *skb)
>>
> 
> Acked-by: Oliver Hartkopp <socketcan@hartkopp.net>
> 
> @Gustavo: Just to be sure:
> 
> From the referenced URLs I got the information that the sizeof() operator causes problems when applied to e.g. cf[0].
> 
> We don't have this case in our code - but one question remains to me:
> 
> We are using the above construct to ensure the padding between the two 'int' values and the struct can_frame which enforces a 64 bit alignment.
> 
> This intention is not affected by the patch, right?
> 

pahole shows exactly the same output either if cf is a zero-length array or
a flexible-array member:

$ pahole -C 'can_skb_priv' drivers/net/can/dev.o

struct can_skb_priv {
	int                        ifindex;              /*     0     4 */
	int                        skbcnt;               /*     4     4 */
	struct can_frame           cf[] __attribute__((__aligned__(8))); /*     8     0 */

	/* size: 8, cachelines: 1, members: 3 */
	/* forced alignments: 1 */
	/* last cacheline: 8 bytes */
} __attribute__((__aligned__(8)));

So, it seems everything should fine. :)

Thanks
--
Gustavo

Re: [PATCH] can: Replace zero-length array with flexible-array

[Oliver Hartkopp]

Hi Gustavo,

On 04.06.20 02:58, Gustavo A. R. Silva wrote:

> Sorry for the late reply. I totally lost track of this thread. :/

NP. At least your workflow seems to work :o)

> pahole shows exactly the same output either if cf is a zero-length array or
> a flexible-array member:
> 
> $ pahole -C 'can_skb_priv' drivers/net/can/dev.o
> 
> struct can_skb_priv {
> 	int                        ifindex;              /*     0     4 */
> 	int                        skbcnt;               /*     4     4 */
> 	struct can_frame           cf[] __attribute__((__aligned__(8))); /*     8     0 */
> 
> 	/* size: 8, cachelines: 1, members: 3 */
> 	/* forced alignments: 1 */
> 	/* last cacheline: 8 bytes */
> } __attribute__((__aligned__(8)));
> 
> So, it seems everything should fine. :)

Great! Thanks for the info - and how I could have checked it on my own, too.

Best,
Oliver

Re: [PATCH] can: Replace zero-length array with flexible-array

[Gustavo A. R. Silva]

On Thu, Jun 04, 2020 at 08:31:55AM +0200, Oliver Hartkopp wrote:
> 
> Great! Thanks for the info - and how I could have checked it on my own, too.
> 

No problem. :)

Thanks
--
Gustavo