LRO num of frags limit
Subsystems (2)
Hi,
looking at the LRO code, at __lro_proc_segment(), it seems that the network driver can configure lro_mgr->max_aggr to any value it wants while the number of fragments aggregated must not exceed MAX_SKB_FRAGS (since we only use a single SKB to aggregate fragments, allocated by lro_gen_skb()). Moreover, even if the driver does limit lro_mgr->max_aggr to MAX_SKB_FRAGS, it might still cause overflow since subsequent aggregations are done at lro_add_frags() which is called before checking whether we overflow.
If the above observation is correct, I can send a patch.
Thanks,
Eli
Quoted 5 lines · Eli Cohen
Hi,
looking at the LRO code, at
__lro_proc_segment(), it seems that the network driver can configurelro_mgr->max_aggrto any value it wants while the number of fragments aggregated must not exceedMAX_SKB_FRAGS
Correct.
Quoted 5 lines
(since we only use a single SKB to aggregate fragments, allocated by
lro_gen_skb()). Moreover, even if the driver does limitlro_mgr->max_aggrtoMAX_SKB_FRAGS, it might still cause overflow since subsequent aggregations are done atlro_add_frags()which is called before checking whether we overflow.
So you must set max_aggr toMAX_SKB_FRAGS - max number of frags added at once + 1.
Quoted 1 line
If the above observation is correct, I can send a patch.
I would be interested to see that, anyway.
Ben.
Signature · 3 lines
Ben Hutchings, Senior Software Engineer, Solarflare Communications Not speaking for my employer; that's the marketing department's job. They asked us to note that Solarflare product names are trademarked.
Quoted 15 lines · Ben Hutchings
Quoted 5 lines · Eli Cohen
Hi,
looking at the LRO code, at
__lro_proc_segment(), it seems that the network driver can configurelro_mgr->max_aggrto any value it wants while the number of fragments aggregated must not exceedMAX_SKB_FRAGSCorrect.
Quoted 5 lines
(since we only use a single SKB to aggregate fragments, allocated by
lro_gen_skb()). Moreover, even if the driver does limitlro_mgr->max_aggrtoMAX_SKB_FRAGS, it might still cause overflow since subsequent aggregations are done atlro_add_frags()which is called before checking whether we overflow.
Quoted 7 lines
So you must set
max_aggrtoMAX_SKB_FRAGS- max number of frags added at once + 1.Quoted 1 line
If the above observation is correct, I can send a patch.
I would be interested to see that, anyway.
By the way, we need to introduce two kinds of "max_aggr" fields to struct net_lro_mgr. The reason is that when LRO is used in the mode in which SKBs are linked, the above limitation does not exist. One will be used for drivers which use lro_receive_skb() (which does not have a limit) and one for drivers which use lro_receive_frags(). This will require changing all the drivers that use LRO.