From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mailrelay005.isp.belgacom.be (mailrelay005.isp.belgacom.be [195.238.6.171]) by ozlabs.org (Postfix) with ESMTP id 55645DDF2F for ; Thu, 29 May 2008 22:14:31 +1000 (EST) From: Laurent Pinchart To: linuxppc-embedded@ozlabs.org Subject: Re: USB Host Controller driver for the MPC8270 processor Date: Thu, 29 May 2008 14:14:23 +0200 References: In-Reply-To: MIME-Version: 1.0 Content-Type: multipart/signed; boundary="nextPart28220191.fMzIght3ui"; protocol="application/pgp-signature"; micalg=pgp-sha1 Message-Id: <200805291414.26798.laurentp@cse-semaphore.com> Cc: Huub Eikens List-Id: Linux on Embedded PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , --nextPart28220191.fMzIght3ui Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Hi Huub, On Thursday 29 May 2008 13:03, Huub Eikens wrote: > Hi, > =20 > I am in need of a USB Host Controller device for the Freescale MPC8270 for > the 2.4 linux kernel (ELDK 3.1.1). On internet I found one, but this is o= nly > for version 2.6: =20 > http://cpm2usb.sourceforge.net/ >=20 > I have a few options: > 1. backport this driver to 2.4.25 kernel. I would like to know if this is > feasible or not. I get the feeling that this will take a lot of time.=20 > 2. upgrade my kernel to 2.6.x. This is not preferable for me since I depe= nd > on sw modules writen by external partners. If I upgrade, I really really > need to be sure that this driver is stable and reliable. > 3. Use an external USB Host Controller. Can anyone advise me on which > controller to choose that runs stable and reliable on the 2.4.25 kernel > (ELDK 3.1.1)? I'd go for option 3. The CPM2 USB host controller is definitely not reliabl= e.=20 It relies on software to prepare SOF tokens, and doesn't do much in hardwar= e. After some experimentation with the FHCI driver posted on linux-ppc by =20 Anton Vorontsov I found out that the driver would eat around 40% of my CPU= =20 time, and some SOF tokens would be incorrectly sent. The CPM3 (aka QE) shouldn't have the same SOF issue, as it prepares SOF tok= en=20 without any software intervention, but I doubt it would perform much better= =20 in term of CPU time. If you can go for an OHCI/UHCI/EHCI based controller that would be much=20 better. Best regards, =2D-=20 Laurent Pinchart CSE Semaphore Belgium Chaussee de Bruxelles, 732A B-1410 Waterloo Belgium T +32 (2) 387 42 59 =46 +32 (2) 387 42 75 --nextPart28220191.fMzIght3ui Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (GNU/Linux) iD8DBQBIPp4i8y9gWxC9vpcRArJBAKCJeqBReTvuGBsr8iOoBIDhR663BgCfYzoY kiP7HkzGNCrn7YAzF8dIopc= =KtcC -----END PGP SIGNATURE----- --nextPart28220191.fMzIght3ui--