Date: Fri, 15 Sep 2006 18:51:06 -0700 From: "Jack Vogel" <jfvogel@gmail.com> To: "Christian Brueffer" <brueffer@freebsd.org> Cc: Andre Oppermann <andre@freebsd.org>, current@freebsd.org Subject: Re: em(4) interface wedges Message-ID: <2a41acea0609151851u6be96fe3ice86b39acfed9660@mail.gmail.com> In-Reply-To: <20060915191008.GE2020@haakonia.hitnet.RWTH-Aachen.DE> References: <20060915165351.GC2020@haakonia.hitnet.RWTH-Aachen.DE> <450ADE55.7090509@samsco.org> <450AEC50.2060404@freebsd.org> <20060915191008.GE2020@haakonia.hitnet.RWTH-Aachen.DE>
next in thread | previous in thread | raw e-mail | index | archive | help
On 9/15/06, Christian Brueffer <brueffer@freebsd.org> wrote: > On Fri, Sep 15, 2006 at 08:09:20PM +0200, Andre Oppermann wrote: > > Scott Long wrote: > > >Christian Brueffer wrote: > > > > > >>Hi, > > >> > > >>with the latest em(4) code in current (sans the printf/device_printf > > >>change), my interface wedges every few hours. It can be resurrected > > >>by bringing it down and up again. > > >> > > >>The device in question is a E1000_DEV_ID_82540EP_LP built into my > > >>Thinkpad. > > >> > > >>Anyone else seeing this? > > >> > > >>- Christian > > >> > > > > > >Yes. It's also almost guaranteed to wedge the first time I do a remote > > >CVS operation after boot. > > > > When the interface wedges does it show the OACTIVE flag in ifconfig output? > > This is an important hint. > > > > Yeah, OACTIVE is set. > > > If you are running at 100Mbit there seems to be bug in the em(4) harware > > with TSO. Try disabling TSO on it with 'ifconfig em0 -tso'. > > > > Correct, I'm using it at 100Mbit and disabling TSO works around the > issuue. That NIC should not be TSO capable so I'm confused by disabling it in this case would have an effect... hmmmm.... Jack
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?2a41acea0609151851u6be96fe3ice86b39acfed9660>