Re: [dpdk-dev] [dpdk-stable] [PATCH 1/4] net/tap: fix Rx cksum flags on IP options packets

[Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>]

On 6/8/21 3:49 PM, Olivier Matz wrote:
> On Tue, Jun 08, 2021 at 03:34:36PM +0300, Andrew Rybchenko wrote:
>> On 6/8/21 3:29 PM, Olivier Matz wrote:
>>> Hi Ferruh, Andrew,
>>>
>>> On Tue, Jun 08, 2021 at 01:13:59PM +0300, Andrew Rybchenko wrote:
>>>> On 4/30/21 5:48 PM, Ferruh Yigit wrote:
>>>>> On 4/27/2021 2:57 PM, Olivier Matz wrote:
>>>>>> When packet type is IPV4_EXT, the checksum is always marked as good in
>>>>>> the mbuf offload flags.
>>>>>>
>>>>>> Since we know the header lengths, we can easily call
>>>>>> rte_ipv4_udptcp_cksum() in this case too.
>>>>>>
>>>>>> Fixes: 8ae3023387e9 ("net/tap: add Rx/Tx checksum offload support")
>>>>>> Cc: stable@dpdk.org
>>>>>>
>>>>>> Signed-off-by: Olivier Matz <olivier.matz@6wind.com>
>>>>>> ---
>>>>>>  drivers/net/tap/rte_eth_tap.c | 4 ++--
>>>>>>  1 file changed, 2 insertions(+), 2 deletions(-)
>>>>>>
>>>>>> diff --git a/drivers/net/tap/rte_eth_tap.c b/drivers/net/tap/rte_eth_tap.c
>>>>>> index 68baa18523..e7b185a4b5 100644
>>>>>> --- a/drivers/net/tap/rte_eth_tap.c
>>>>>> +++ b/drivers/net/tap/rte_eth_tap.c
>>>>>> @@ -350,7 +350,7 @@ tap_verify_csum(struct rte_mbuf *mbuf)
>>>>>>  		/* Don't verify checksum for multi-segment packets. */
>>>>>>  		if (mbuf->nb_segs > 1)
>>>>>>  			return;
>>>>>> -		if (l3 == RTE_PTYPE_L3_IPV4) {
>>>>>> +		if (l3 == RTE_PTYPE_L3_IPV4 || l3 == RTE_PTYPE_L3_IPV4_EXT) {
>>>>>
>>>>> Should we take 'RTE_PTYPE_L3_IPV4_EXT_UNKNOWN' into account?
>>>>
>>>> I think we should.
>>>
>>> I think 'RTE_PTYPE_L3_IPV4_EXT_UNKNOWN' cannot happen here:
>>>
>>> - mbuf->packet_type is generated by 
>>
>> (), which cannot
>>>   return 'RTE_PTYPE_L3_IPV4_EXT_UNKNOWN'
>>
>> My question if it is guaranteed and the only possible branch.
>> Can application set packet_type itself and do not call
>> rte_net_get_ptype(). Yes, typically application knows
>> if it has IPv4 options in the header or not, but theoretically
>> could be unaware as well.
> 
> This function is called on the Rx path from pmd_rx_burst(), so
> the application does not have access to the mbuf.
> 
> The software parser that sets the packet type returns either
> RTE_PTYPE_L3_IPV4 if there is no option, or RTE_PTYPE_L3_IPV4_EXT
> else. The value RTE_PTYPE_L3_IPV4_EXT_UNKNOWN is used by PMDs that don't
> know if there are options.

I see. What I'm trying to say that there are non
obvious assumptions here on rte_net_get_ptype()
behaviour which can be changed. May be it makes
sense to add comments here to highlight it.

> 
>>> - right above this code, we already returned if l3 is not in
>>>   (RTE_PTYPE_L3_IPV4, RTE_PTYPE_L3_IPV4_EXT, RTE_PTYPE_L3_IPV6)
>>
>> If so, it sounds like it should be allowed above as well.
>>
>>>
>>>>>
>>>>>>  			if (l4 == RTE_PTYPE_L4_UDP) {
>>>>>>  				udp_hdr = (struct rte_udp_hdr *)l4_hdr;
>>>>>>  				if (udp_hdr->dgram_cksum == 0) {
>>>>>> @@ -364,7 +364,7 @@ tap_verify_csum(struct rte_mbuf *mbuf)
>>>>>>  				}
>>>>>>  			}
>>>>>>  			cksum = ~rte_ipv4_udptcp_cksum(l3_hdr, l4_hdr);
>>>>>> -		} else if (l3 == RTE_PTYPE_L3_IPV6) {
>>>>>> +		} else { /* l3 == RTE_PTYPE_L3_IPV6, checked above */
>>>>>>  			cksum = ~rte_ipv6_udptcp_cksum(l3_hdr, l4_hdr);
>>>>>>  		}
>>>>>>  		mbuf->ol_flags |= cksum ?
>>>>>>
>>>>
>>