Date: Fri, 9 Apr 2004 19:47:24 +0300 From: Ruslan Ermilov <ru@FreeBSD.org> To: Luigi Rizzo <luigi@FreeBSD.org> Cc: net@FreeBSD.org Subject: polling(4) and rl(4) Message-ID: <20040409164724.GD2461@ip.net.ua>
next in thread | raw e-mail | index | archive | help
--zS7rBR6csb6tI2e1 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable Hey Luigi, Have you actually measured the performance of rl(4) with polling(4) enabled? With 8139 anomaly of four (register based) TX descriptors it's becoming a royal pain in the ass with polling(4) enabled -- the TX performance just sucks -- I could only get the comparable results with HZ=3D5000, which is overheating my CPU. My suggestion is to drop polling(4) support from the rl(4) driver completely. Are there any objections? Has anybody got different results with rl(4) and polling(4) enabled? Having it in re(4) is of course a good thing. ;) As an aside, I've started working on the ``[-]polling'' option for ifconfig(8) that, when done, will allow changing the polling status of individual interfaces in run-time, e.g., the following command will disable polling on nge0: ifconfig nge0 -polling Cheers, --=20 Ruslan Ermilov ru@FreeBSD.org FreeBSD committer --zS7rBR6csb6tI2e1 Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (FreeBSD) iD8DBQFAdtOcUkv4P6juNwoRAuw0AJ0QsAOzMjwTvlpFqxqlQx32EVs4ugCcD2V+ PN9fXhc9Hgel7JFDTBGuyUo= =0E3a -----END PGP SIGNATURE----- --zS7rBR6csb6tI2e1--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040409164724.GD2461>