netdev - Re: [PATCH net-next] tcp: tcp_tso_autosize() minimum is one packet

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Tue, 26 May 2015 19:14:41 -0700
From:	Eric Dumazet <eric.dumazet@...il.com>
To:	Herbert Xu <herbert@...dor.apana.org.au>
Cc:	davem@...emloft.net, netdev@...r.kernel.org, ncardwell@...gle.com
Subject: Re: [PATCH net-next] tcp: tcp_tso_autosize() minimum is one packet

On Wed, 2015-05-27 at 09:38 +0800, Herbert Xu wrote:
> On Tue, May 26, 2015 at 05:36:34PM -0700, Eric Dumazet wrote:
> > 
> > tldr:  "TSO with max_segs==1"  <is the same than> "no TSO/GSO"
> 
> Not really.  They're not identical.  For example, before your
> patch a packet greater than MSS with TSO disabled would call
> tcp_nagle_test, with your patch it will call tcp_tso_should_defer
> instead.

Well, given that a device can set gso_max_segs to one, if there is a bug
here we'll need to fix it asap.

Fact that Nagle or tso should defer applies in this corner case is not
very important here, unless you have a specific case in mind ?
Anyway I double checked and I believes it is fine.

We normally deal with dynamic MSS changes, even for non GSO cases.

A non GSO packet temporarily becomes a GSO one in tcp_init_tso_segs()
(because its skb->len is bigger than cur_mss)

Then we split it.

Nagle or tso should defer would take same decision : send one full MSS.

By the time the last 'segment' (possibly smaller than mss) will be
considered, Nagle might apply there.

Thanks !

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html