Re: CVS commit: src/sys/net/if_vlan.c

[Ryota Ozaki <ozaki-r%netbsd.org@localhost>]

On Tue, Mar 31, 2015 at 1:42 AM, Matt Thomas <matt%3am-software.com@localhost> wrote:
>
>> On Mar 30, 2015, at 8:26 AM, Christoph Badura <bad%bsd.de@localhost> wrote:
>>
>> On Sun, Mar 29, 2015 at 01:30:43PM +0000, Ryota Ozaki wrote:
>>> Module Name: src
>>> Committed By:        ozaki-r
>>> Date:                Sun Mar 29 13:30:43 UTC 2015
>>>
>>> Modified Files:
>>>      src/sys/net: if_vlan.c
>>>
>>> Log Message:
>>> Correct frame padding length
>>>
>>> vlan pads a frame with zeros up to 68 bytes
>>> (ETHER_MIN_LEN + ETHER_VLAN_ENCAP_LEN). It expects
>>> that even if the frame is untagged, it keeps 64 bytes
>>> at least. However, it lacks concern about CRC
>>> (4 bytes). So a sending frame can be 72 (68 + 4) bytes.
>>
>> I don't get it. ETHER_MIN_LEN includes the 4 bytes for the CRC of a
>> minimum sized packet.  Therefore ETHER_MIN_LEN + ETHER_VLAN_ENCAP_LEN is
>> the right size, isn't it?
>
> No, it has 4 bytes too many because the 4 bytes for CRC still needs to be added by the hardware.
>
>> BTW, what do you mean by "it lacks concern about CRC"?
>
> It lacks compensating for the fact the CRC has been stripped by the time vlan gets the packet.

# matt@ explained well than me...

Note that the issue was found by some strict packet capture software
(not by me, so I don't know what it is); typical OSes don't complain
the extra padding, so we didn't notice the issue until recently.

  ozaki-r