Fri, Jul 28, 2017 at 04:52:02PM CEST, j...@mojatatu.com wrote:
>On 17-07-28 10:12 AM, Jiri Pirko wrote:
>> Fri, Jul 28, 2017 at 03:41:44PM CEST, j...@mojatatu.com wrote:
>[..]
>>
>> Looks like a big mess to be honest. Mixing up u32* u32 void*. I don't
>> understand ****. Would be probably good to first apply my review comment
>> on the function itselt, then to add the checks :)
>>
>
>I havent even compiled/test that Jiri.
>Just ignore the void * and assume it is a u32 *.
>
>I am trying to avoid doing unlucky number 13 patch.
I'll rather wait for you to send the code in proper shape so I can
decypher easily and don't have to guess.
>So feedback on this is good. Just look at what it is disallowing
>first.
>
>back later.
>
>cheers,
>jamal
>
>>
>> > I can think of.
>> >
>> > static int validate_nla_bitfield32(const struct nlattr *nla,
>> > void *valid_flags_allowed)
>> > {
>> > const struct nla_bitfield32 *bf = nla_data(nla);
>> > u32 *valid_flags_mask = valid_flags_allowed;
>> >
>> > if (!valid_flags_allowed)
>> > return -EINVAL;
>> > /*disallow invalid selector */
>> > if ((bf->selector & valid_flags_allowed) >*valid_flags_allowed)
>> > return -EINVAL;
>> > /*disallow invalid bit values */
>> > if (bf->value & ~*valid_flags_mask)
>> > return -EINVAL;
>> > /*disallow valid bit values that are not selected*/
>> > if (bf->value & ~nbf->selector)
>> > return -EINVAL;
>> >
>> > return 0;
>> > }
>> >
>> > cheers,
>> > jamal
>
