Date: Sat, 7 May 2005 11:42:55 -0400 (EDT) From: "Mike Jakubik" <mikej@rogers.com> To: "Marian Durkovic" <md@bts.sk> Cc: Kris Kennaway <kris@obsecurity.org> Subject: Re: SOLVED: Degraded TCP performace on Intel PRO/1000 Message-ID: <1482.172.16.0.199.1115480575.squirrel@172.16.0.1> In-Reply-To: <20050507093510.GA82158@us.svf.stuba.sk> References: <20050505133250.GA73885@us.svf.stuba.sk> <20050506065950.GA1999@us.svf.stuba.sk> <20050507043712.GB28373@xor.obsecurity.org> <20050507093510.GA82158@us.svf.stuba.sk>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, May 7, 2005 5:35 am, Marian Durkovic said: >>> To achieve wirespeed performance, the TX FIFO must be large enough to >>> accomodate 2 jumbo packets (not just 1 as the driver was assuming). >>> There was also a typo in the driver, causing the PBA tuning on most >>> cards to be non-functional. >>> >>> Please be sure to use em driver version 1.7.41 or newer, either from >>> the CVS (branch RELENG_4_11) or from Intel's downloads. >> >> Does this also apply to 5.x and above? If no-one else responds in the >> next few days, can you please submit the patch in PR so it does not get >> lost? >> >> Kris >> > > All the CVS branches of if_em.c have the same bugs in the packet buffer > (PBA) code, so yes, the patch applies to all releases. Please submit this patch, i myself use the em card in a lot of boxes. CURRENT still has driver version 1.7.35, so looks like we are behind a bit. Thanks.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1482.172.16.0.199.1115480575.squirrel>