| To: | James Morris <jmorris@xxxxxxxxxx> |
|---|---|
| Subject: | IPCOMP scratch buffer (was: [Openswan dev] IPComp) |
| From: | Herbert Xu <herbert@xxxxxxxxxxxxxxxxxxx> |
| Date: | Fri, 9 Jul 2004 20:02:13 +1000 |
| Cc: | netdev@xxxxxxxxxxx |
| In-reply-to: | <Xine.LNX.4.44.0407061846470.24263-100000@thoron.boston.redhat.com> |
| References: | <20040706213135.GA21477@gondor.apana.org.au> <Xine.LNX.4.44.0407061846470.24263-100000@thoron.boston.redhat.com> |
| Sender: | netdev-bounce@xxxxxxxxxxx |
| User-agent: | Mutt/1.5.6+20040523i |
On Tue, Jul 06, 2004 at 06:50:44PM -0400, James Morris wrote: > > I think it would be better to go with your original idea of allocating a > scratch buffer for each packet, based on the size of the packet. IPComp > is very slow path, and allocating 64k for each SA is optimizing for an > uncommon worst case in a way which will potentially eat up a lot of memory > (e.g. > 6MB for 100 tunnels). What about a shared per-cpu buffer? -- Visit Openswan at http://www.openswan.org/ Email: Herbert Xu ~{PmV>HI~} <herbert@xxxxxxxxxxxxxxxxxxx> Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: pskb change in dst->output, Herbert Xu |
|---|---|
| Next by Date: | [XFRM] Add FLUSHSA and FLUSHPOLICY, Herbert Xu |
| Previous by Thread: | Re: [Openswan dev] IPComp, James Morris |
| Next by Thread: | Re: IPCOMP scratch buffer (was: [Openswan dev] IPComp), James Morris |
| Indexes: | [Date] [Thread] [Top] [All Lists] |