[PATCH v3 0/6] skb paged fragment destructors · Ian Campbell <hidden> · 2012-01-25
[PATCH v3 1/6] net: pad skb data and shinfo as a whole rather than individually · Ian Campbell <hidden> · 2012-01-25
Re: [PATCH v3 1/6] net: pad skb data and shinfo as a whole rather than individually · Eric Dumazet <hidden> · 2012-01-25
Re: [PATCH v3 1/6] net: pad skb data and shinfo as a whole rather than individually · Ian Campbell <hidden> · 2012-01-25
Re: [PATCH v3 1/6] net: pad skb data and shinfo as a whole rather than individually · Eric Dumazet <hidden> · 2012-01-25
Re: [PATCH v3 1/6] net: pad skb data and shinfo as a whole rather than individually · Ian Campbell <hidden> · 2012-01-31
Re: [PATCH v3 1/6] net: pad skb data and shinfo as a whole rather than individually · Eric Dumazet <hidden> · 2012-01-31
Re: [PATCH v3 1/6] net: pad skb data and shinfo as a whole rather than individually · Ian Campbell <hidden> · 2012-02-01
Re: [PATCH v3 1/6] net: pad skb data and shinfo as a whole rather than individually · Eric Dumazet <hidden> · 2012-02-01
Re: [PATCH v3 1/6] net: pad skb data and shinfo as a whole rather than individually · Ian Campbell <hidden> · 2012-02-01
Re: [PATCH v3 1/6] net: pad skb data and shinfo as a whole rather than individually · Francois Romieu <romieu@fr.zoreil.com> · 2012-01-31
Re: [PATCH v3 1/6] net: pad skb data and shinfo as a whole rather than individually · Ian Campbell <hidden> · 2012-02-01
[PATCH v3 2/6] net: move destructor_arg to the front of sk_buff. · Ian Campbell <hidden> · 2012-01-25
[PATCH v3 3/6] net: add support for per-paged-fragment destructors · Ian Campbell <hidden> · 2012-01-25
[PATCH v3 4/6] net: only allow paged fragments with the same destructor to be coalesced. · Ian Campbell <hidden> · 2012-01-25
[PATCH v3 6/6] sunrpc: use SKB fragment destructors to delay completion until page is released by network stack. · Ian Campbell <hidden> · 2012-01-25

Re: [PATCH v3 1/6] net: pad skb data and shinfo as a whole rather than individually

From: Eric Dumazet <hidden>
Date: 2012-01-25 13:12:41

Le mercredi 25 janvier 2012 à 13:09 +0000, Ian Campbell a écrit :

Can you elaborate on the specific benchmark you used there?

One machine was sending udp frames on my target (using pktgen)

Target was running a mono threaded udp receiver (one socket)

`h`	back out one level
`j`	next message in thread
`k`	previous message in thread
`l`	drill in
`Esc`	close help / fold thread tree
`?`	toggle this help