Re: [Xen-devel] [PATCH net] Xen-netback: Fix issue caused by using gso_type wrongly

[annie li <annie.li@oracle.com>]

On 2014/3/10 17:51, Paul Durrant wrote:
>> -----Original Message-----
>> From: Annie Li [mailto:Annie.li@oracle.com]
>> Sent: 10 March 2014 07:44
>> To:xen-devel@lists.xen.org;netdev@vger.kernel.org
>> Cc:konrad.wilk@oracle.com; Ian Campbell; Wei Liu; Paul Durrant;
>> annie.li@oracle.com
>> Subject: [PATCH net] Xen-netback: Fix issue caused by using gso_type
>> wrongly
>>
>> From: Annie Li<annie.li@oracle.com>
>>
>> Current netback uses gso_type to check whether the skb contains
>> gso offload, and this is wrong. Gso_size is the right one to
>> check gso existence, and gso_type is only used to check gso type.
>>
>> Some skbs contains nonzero gso_type and zero gso_size, current
>> netback would treat these skbs as gso and create wrong response
>> for this. This also causes ssh failure to domu from other server.
>>
> That's a shame. It would be nice if these things were consistent...
>
>> Signed-off-by: Annie Li<annie.li@oracle.com>
>> ---
>>   drivers/net/xen-netback/netback.c |   35 +++++++++++++++++---------------
>> ---
>>   1 files changed, 17 insertions(+), 18 deletions(-)
>>
>> diff --git a/drivers/net/xen-netback/netback.c b/drivers/net/xen-
>> netback/netback.c
>> index e5284bc..57a7d5d 100644
>> --- a/drivers/net/xen-netback/netback.c
>> +++ b/drivers/net/xen-netback/netback.c
>> @@ -240,7 +240,7 @@ static void xenvif_gop_frag_copy(struct xenvif *vif,
>> struct sk_buff *skb,
>>   	struct gnttab_copy *copy_gop;
>>   	struct xenvif_rx_meta *meta;
>>   	unsigned long bytes;
>> -	int gso_type;
>> +	int gso_type = XEN_NETIF_GSO_TYPE_NONE;
>>
>>   	/* Data must not cross a page boundary. */
>>   	BUG_ON(size + offset > PAGE_SIZE<<compound_order(page));
>> @@ -299,12 +299,12 @@ static void xenvif_gop_frag_copy(struct xenvif *vif,
>> struct sk_buff *skb,
>>   		}
>>
>>   		/* Leave a gap for the GSO descriptor. */
>> -		if (skb_shinfo(skb)->gso_type & SKB_GSO_TCPV4)
>> -			gso_type = XEN_NETIF_GSO_TYPE_TCPV4;
>> -		else if (skb_shinfo(skb)->gso_type & SKB_GSO_TCPV6)
>> -			gso_type = XEN_NETIF_GSO_TYPE_TCPV6;
>> -		else
>> -			gso_type = XEN_NETIF_GSO_TYPE_NONE;
>> +		if (skb_shinfo(skb)->gso_size) {
> You should probably use skb_is_gso(skb) for your test.

skb_is_gso does the same thing, skb_iso_gso and 
skb_shinfo(skb)->gso_size coexist. But I can change the code as you 
suggested if you like, will post a v2 patch for this.

Thanks
Annie

>    Paul
>
>> +			if (skb_shinfo(skb)->gso_type & SKB_GSO_TCPV4)
>> +				gso_type = XEN_NETIF_GSO_TYPE_TCPV4;
>> +			else if (skb_shinfo(skb)->gso_type &
>> SKB_GSO_TCPV6)
>> +				gso_type = XEN_NETIF_GSO_TYPE_TCPV6;
>> +		}
>>
>>   		if (*head && ((1 << gso_type) & vif->gso_mask))
>>   			vif->rx.req_cons++;
>> @@ -342,15 +342,13 @@ static int xenvif_gop_skb(struct sk_buff *skb,
>>
>>   	old_meta_prod = npo->meta_prod;
>>
>> -	if (skb_shinfo(skb)->gso_type & SKB_GSO_TCPV4) {
>> -		gso_type = XEN_NETIF_GSO_TYPE_TCPV4;
>> -		gso_size = skb_shinfo(skb)->gso_size;
>> -	} else if (skb_shinfo(skb)->gso_type & SKB_GSO_TCPV6) {
>> -		gso_type = XEN_NETIF_GSO_TYPE_TCPV6;
>> -		gso_size = skb_shinfo(skb)->gso_size;
>> -	} else {
>> -		gso_type = XEN_NETIF_GSO_TYPE_NONE;
>> -		gso_size = 0;
>> +	gso_type = XEN_NETIF_GSO_TYPE_NONE;
>> +	gso_size = skb_shinfo(skb)->gso_size;
>> +	if (gso_size) {
>> +		if (skb_shinfo(skb)->gso_type & SKB_GSO_TCPV4)
>> +			gso_type = XEN_NETIF_GSO_TYPE_TCPV4;
>> +		else if (skb_shinfo(skb)->gso_type & SKB_GSO_TCPV6)
>> +			gso_type = XEN_NETIF_GSO_TYPE_TCPV6;
>>   	}
>>
>>   	/* Set up a GSO prefix descriptor, if necessary */
>> @@ -500,8 +498,9 @@ static void xenvif_rx_action(struct xenvif *vif)
>>   			size = skb_frag_size(&skb_shinfo(skb)->frags[i]);
>>   			max_slots_needed += DIV_ROUND_UP(size,
>> PAGE_SIZE);
>>   		}
>> -		if (skb_shinfo(skb)->gso_type & SKB_GSO_TCPV4 ||
>> -		    skb_shinfo(skb)->gso_type & SKB_GSO_TCPV6)
>> +		if (skb_shinfo(skb)->gso_size &&
>> +		   (skb_shinfo(skb)->gso_type & SKB_GSO_TCPV4 ||
>> +		    skb_shinfo(skb)->gso_type & SKB_GSO_TCPV6))
>>   			max_slots_needed++;
>>
>>   		/* If the skb may not fit then bail out now */
>> --
>> 1.7.3.4
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xen.org
> http://lists.xen.org/xen-devel