Khers,

Thanks! Just after I sent you the reply Dave had pointed out coverity was 
unhappy with some of the code, including that particular line. So I got rid of 
memcpy altogether and while at it fixed the values for both this place and the 
other one I told you about - in change 9611.

--a

> On 29 Nov 2017, at 08:25, khers <[email protected]> wrote:
> 
> Dear Andrew
> 
> Thanks for your attention, Yes of course I pushed to gerrit with id 9615.
> 
> Regards,
> Khers
> 
>> On Tue, Nov 28, 2017 at 8:37 PM, Andrew Yourtchenko <[email protected]> 
>> wrote:
>> Dear Khers,
>> 
>> I believe you are right. That might not be all though... “dot1q”/“dot1ad” 
>> mask value constant does not appear to make sense to me now.
>> 
>> They should be “XX XX” to mask out the bits and also should be set 
>> accordingly to the proper values during the addition of the sessions. (I 
>> suppose might “kind of” work now but it is not correct..)
>> 
>> Would you like to submit a gerrit to address these ?
>> 
>> Good catch, thanks!
>> 
>> --a
>> 
>> > On 28 Nov 2017, at 12:24, khers <[email protected]> wrote:
>> >
>> > Dear vpp folks
>> >
>> > I think following line if function acl_add_vlan_session in acl.c line 635
>> >
>> >   memset (&match[idx], 0x00, 2);
>> >
>> > should change to
>> >
>> >   memset (&match[idx], 0xff, 2);
>> >
>> > because dot1ad_5tuple_mask and dot1q_5tuple_mask must have mask for 
>> > IPv4/6, so memset to ff reset those mask to default values.
>> >
>> > Regards,
>> > Khers
>> > _______________________________________________
>> > vpp-dev mailing list
>> > [email protected]
>> > https://lists.fd.io/mailman/listinfo/vpp-dev
> 
_______________________________________________
vpp-dev mailing list
[email protected]
https://lists.fd.io/mailman/listinfo/vpp-dev

Reply via email to