Re: [dpdk-dev] [PATCH] ipv4_fragmentation: fix fragmentation of ipv4 packet with optional header

["Ananyev, Konstantin" <konstantin.ananyev@intel.com>]

Hi,

First of all - if it is a fix, then we need to have:
Fixes: ...
And probably 
Cc: stable@dpdk.org
As described here:
https://doc.dpdk.org/guides/contributing/patches.html 

Second it would be good to add some text here - 
problem statement and solution description.

> 
> Signed-off-by: Pu Xu <583493798@qq.com>
> ---
>  lib/librte_ip_frag/rte_ipv4_fragmentation.c | 20 +++++++++++---------
>  1 file changed, 11 insertions(+), 9 deletions(-)
> 
> diff --git a/lib/librte_ip_frag/rte_ipv4_fragmentation.c b/lib/librte_ip_frag/rte_ipv4_fragmentation.c
> index e9de335ae..156087ca3 100644
> --- a/lib/librte_ip_frag/rte_ipv4_fragmentation.c
> +++ b/lib/librte_ip_frag/rte_ipv4_fragmentation.c
> @@ -74,7 +74,7 @@ rte_ipv4_fragment_packet(struct rte_mbuf *pkt_in,
>  	struct rte_ipv4_hdr *in_hdr;
>  	uint32_t out_pkt_pos, in_seg_data_pos;
>  	uint32_t more_in_segs;
> -	uint16_t fragment_offset, flag_offset, frag_size;
> +	uint16_t fragment_offset, flag_offset, frag_size, header_len;
>  	uint16_t frag_bytes_remaining;
> 
>  	/*
> @@ -86,14 +86,16 @@ rte_ipv4_fragment_packet(struct rte_mbuf *pkt_in,
>  	    unlikely(mtu_size < RTE_ETHER_MIN_MTU))
>  		return -EINVAL;
> 
> +	in_hdr = rte_pktmbuf_mtod(pkt_in, struct rte_ipv4_hdr *);
> +
> +	header_len = (in_hdr->version_ihl & 0xf) * 4;

We do have a special define for that in librte_net/rte_ip.h:

** Internet header length mask for version_ihl field */
#define RTE_IPV4_HDR_IHL_MASK   (0x0f)
/**
 * Internet header length field multiplier (IHL field specifies overall header
 * length in number of 4-byte words)
 */
#define RTE_IPV4_IHL_MULTIPLIER (4)

Please use them in the code above.

Also as now we are getting header_len from the packet itself,
It would be good to check that it contains a valid value.
Otherwise ill-formed ip-header can crash dpdk app.

>  	/*
>  	 * Ensure the IP payload length of all fragments is aligned to a
>  	 * multiple of 8 bytes as per RFC791 section 2.3.
>  	 */
> -	frag_size = RTE_ALIGN_FLOOR((mtu_size - sizeof(struct rte_ipv4_hdr)),
> +	frag_size = RTE_ALIGN_FLOOR((mtu_size - header_len),
>  				    IPV4_HDR_FO_ALIGN);
> 
> -	in_hdr = rte_pktmbuf_mtod(pkt_in, struct rte_ipv4_hdr *);
>  	flag_offset = rte_cpu_to_be_16(in_hdr->fragment_offset);
> 
>  	/* If Don't Fragment flag is set */
> @@ -102,11 +104,11 @@ rte_ipv4_fragment_packet(struct rte_mbuf *pkt_in,
> 
>  	/* Check that pkts_out is big enough to hold all fragments */
>  	if (unlikely(frag_size * nb_pkts_out <
> -	    (uint16_t)(pkt_in->pkt_len - sizeof(struct rte_ipv4_hdr))))
> +	    (uint16_t)(pkt_in->pkt_len - header_len)))
>  		return -EINVAL;
> 
>  	in_seg = pkt_in;
> -	in_seg_data_pos = sizeof(struct rte_ipv4_hdr);
> +	in_seg_data_pos = header_len;
>  	out_pkt_pos = 0;
>  	fragment_offset = 0;
> 
> @@ -124,8 +126,8 @@ rte_ipv4_fragment_packet(struct rte_mbuf *pkt_in,
>  		}
> 
>  		/* Reserve space for the IP header that will be built later */
> -		out_pkt->data_len = sizeof(struct rte_ipv4_hdr);
> -		out_pkt->pkt_len = sizeof(struct rte_ipv4_hdr);
> +		out_pkt->data_len = header_len;
> +		out_pkt->pkt_len = header_len;
>  		frag_bytes_remaining = frag_size;
> 
>  		out_seg_prev = out_pkt;
> @@ -181,9 +183,9 @@ rte_ipv4_fragment_packet(struct rte_mbuf *pkt_in,
>  		    flag_offset, fragment_offset, more_in_segs);

I think we have to update __fill_ipv4hdr_frag() too, otherwise options
from original header wouldn't be copied into  the new header.

> 
>  		fragment_offset = (uint16_t)(fragment_offset +
> -		    out_pkt->pkt_len - sizeof(struct rte_ipv4_hdr));
> +		    out_pkt->pkt_len - header_len);
> 
> -		out_pkt->l3_len = sizeof(struct rte_ipv4_hdr);
> +		out_pkt->l3_len = header_len;
> 
>  		/* Write the fragment to the output list */
>  		pkts_out[out_pkt_pos] = out_pkt;
> --
> 2.20.1