netdev - Re: [PATCH 0/1] mv643xx_eth: Disable TSO by default

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 for Android: free password hash cracker in your pocket

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

Message-ID: <1414864868.31792.11.camel@edumazet-glaptop2.roam.corp.google.com>
Date:	Sat, 01 Nov 2014 11:01:08 -0700
From:	Eric Dumazet <eric.dumazet@...il.com>
To:	Thomas Petazzoni <thomas.petazzoni@...e-electrons.com>
Cc:	Ezequiel Garcia <ezequiel.garcia@...e-electrons.com>,
	netdev@...r.kernel.org, David Miller <davem@...emloft.net>,
	Gregory Clement <gregory.clement@...e-electrons.com>,
	Tawfik Bayouk <tawfik@...vell.com>,
	Lior Amsalem <alior@...vell.com>,
	Nadav Haklai <nadavh@...vell.com>
Subject: Re: [PATCH 0/1] mv643xx_eth: Disable TSO by default

On Sat, 2014-11-01 at 18:33 +0100, Thomas Petazzoni wrote:
> Dear Eric Dumazet,

You can call me Eric ;)

> As far as I know, ARMv5 does not do out-of-order execution, and so on
> ARMv5, mb() == rmb() == wmb() == barrier(). But might be a missing
> compiler barrier, indeed, as that's not specific to the architecture.
> 
> Thanks for the hint!

I think the wmb() is needed, but the more worrying problem is the NIC
can start transmitting a segment while the data payload is not yet
placed in TX descriptor.

Normally, the patch I cooked should avoid this.


--
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