Re: [PATCH net-next v5 4/5] net: gro: remove unnecessary df checks

[Richard Gobert]

Willem de Bruijn wrote:
> Richard Gobert wrote:
>> Willem de Bruijn wrote:
>>> Richard Gobert wrote:
>>>> Currently, packets with fixed IDs will be merged only if their
>>>> don't-fragment bit is set. This restriction is unnecessary since packets
>>>> without the don't-fragment bit will be forwarded as-is even if they were
>>>> merged together.
>>>
>>> Please expand why this is true.
>>>
>>> Because either NETIF_F_TSO_MANGLEID is set or segmentation
>>> falls back onto software GSO which handles the two FIXEDID
>>> variants correctly now, I guess?
>>>
>>
>> This is true because the merged packets will be segmented back to
>> their original forms before being forwarded. As you already said, the IDs
>> will either stay identical or potentially become incrementing if MANGLEID
>> is set, either of which is fine.
>>
>>>> If packets are merged together and then fragmented, they will first be
>>>> re-split into segments before being further fragmented, so the behavior
>>>> is identical whether or not the packets were first merged together.
>>>
>>> I don't follow this scenario. Fragmentation of a GSO packet after GRO
>>> and before GSO?
>>>
>>
>> Yes. One could worry that merging packets with the same ID but without DF
>> would cause issues if they are then fragmented by the host. What I'm saying
>> is that if such packets are merged and then fragmented, they will first be
>> segmented back to their original forms by GSO before being further fragmented
>> (see ip_finish_output_gso). The fragmentation occurs as if the packets were
>> never merged to begin with.
> 
> This explicit pointer that fragmentation for such GSO packets happens
> in ip_finish_output_gso, which first calls skb_gso_segment, is
> informative. It again turns an assertion into an explanation.
> 
> I think you jumped the gun a bit on sending a v6 right with these
> answers. I'd like these clarifications recorded.
> 

Sorry about that. I'll explicitly mention ip_finish_output_gso in the commit
message.

>> IOW, fragmentation occurs the same way regardless
>> of whether the packets were merged (GRO + GSO is transparent). I thought I'd
>> mention this to clarify why this patch doesn't cause any issues.
>>
>>>> Clean up the code by removing the unnecessary don't-fragment checks.
>>>>
>>>> Signed-off-by: Richard Gobert <richardbgobert@xxxxxxxxx>
>>>> ---
>>>>  include/net/gro.h                 | 5 ++---
>>>>  net/ipv4/af_inet.c                | 3 ---
>>>>  tools/testing/selftests/net/gro.c | 9 ++++-----
>>>>  3 files changed, 6 insertions(+), 11 deletions(-)
>>>>
>>>> diff --git a/include/net/gro.h b/include/net/gro.h
>>>> index 6aa563eec3d0..f14b7e88dbef 100644
>>>> --- a/include/net/gro.h
>>>> +++ b/include/net/gro.h
>>>> @@ -448,17 +448,16 @@ static inline int inet_gro_flush(const struct iphdr *iph, const struct iphdr *ip
>>>>  	const u32 id2 = ntohl(*(__be32 *)&iph2->id);
>>>>  	const u16 ipid_offset = (id >> 16) - (id2 >> 16);
>>>>  	const u16 count = NAPI_GRO_CB(p)->count;
>>>> -	const u32 df = id & IP_DF;
>>>>  
>>>>  	/* All fields must match except length and checksum. */
>>>> -	if ((iph->ttl ^ iph2->ttl) | (iph->tos ^ iph2->tos) | (df ^ (id2 & IP_DF)))
>>>> +	if ((iph->ttl ^ iph2->ttl) | (iph->tos ^ iph2->tos) | ((id ^ id2) & IP_DF))
>>>>  		return true;
>>>>  
>>>>  	/* When we receive our second frame we can make a decision on if we
>>>>  	 * continue this flow as an atomic flow with a fixed ID or if we use
>>>>  	 * an incrementing ID.
>>>>  	 */
>>>> -	if (count == 1 && df && !ipid_offset)
>>>> +	if (count == 1 && !ipid_offset)
>>>>  		NAPI_GRO_CB(p)->ip_fixedid |= 1 << inner;
>>>>  
>>>>  	return ipid_offset ^ (count * !(NAPI_GRO_CB(p)->ip_fixedid & (1 << inner)));
>>>> diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
>>>> index fc7a6955fa0a..c0542d9187e2 100644
>>>> --- a/net/ipv4/af_inet.c
>>>> +++ b/net/ipv4/af_inet.c
>>>> @@ -1393,10 +1393,7 @@ struct sk_buff *inet_gso_segment(struct sk_buff *skb,
>>>>  
>>>>  	segs = ERR_PTR(-EPROTONOSUPPORT);
>>>>  
>>>> -	/* fixed ID is invalid if DF bit is not set */
>>>>  	fixedid = !!(skb_shinfo(skb)->gso_type & (SKB_GSO_TCP_FIXEDID << encap));
>>>> -	if (fixedid && !(ip_hdr(skb)->frag_off & htons(IP_DF)))
>>>> -		goto out;
>>>>  
>>>>  	if (!skb->encapsulation || encap)
>>>>  		udpfrag = !!(skb_shinfo(skb)->gso_type & SKB_GSO_UDP);
>>>> diff --git a/tools/testing/selftests/net/gro.c b/tools/testing/selftests/net/gro.c
>>>> index d5824eadea10..3d4a82a2607c 100644
>>>> --- a/tools/testing/selftests/net/gro.c
>>>> +++ b/tools/testing/selftests/net/gro.c
>>>> @@ -670,7 +670,7 @@ static void send_flush_id_case(int fd, struct sockaddr_ll *daddr, int tcase)
>>>>  		iph2->id = htons(9);
>>>>  		break;
>>>>  
>>>> -	case 3: /* DF=0, Fixed - should not coalesce */
>>>> +	case 3: /* DF=0, Fixed - should coalesce */
>>>>  		iph1->frag_off &= ~htons(IP_DF);
>>>>  		iph1->id = htons(8);
>>>>  
>>>> @@ -1188,10 +1188,9 @@ static void gro_receiver(void)
>>>>  			correct_payload[0] = PAYLOAD_LEN * 2;
>>>>  			check_recv_pkts(rxfd, correct_payload, 1);
>>>>  
>>>> -			printf("DF=0, Fixed - should not coalesce: ");
>>>> -			correct_payload[0] = PAYLOAD_LEN;
>>>> -			correct_payload[1] = PAYLOAD_LEN;
>>>> -			check_recv_pkts(rxfd, correct_payload, 2);
>>>> +			printf("DF=0, Fixed - should coalesce: ");
>>>> +			correct_payload[0] = PAYLOAD_LEN * 2;
>>>> +			check_recv_pkts(rxfd, correct_payload, 1);
>>>>  
>>>>  			printf("DF=1, 2 Incrementing and one fixed - should coalesce only first 2 packets: ");
>>>>  			correct_payload[0] = PAYLOAD_LEN * 2;
>>>> -- 
>>>> 2.36.1
>>>>
>>>
>>>
>>
> 
>