Date: Tue, 29 Dec 1998 00:24:31 +0000 From: Nicolas Souchu <nsouch@teaser.fr> To: Drew Baxter <netmonger@genesis.ispace.com> Cc: takawata@shidahara1.planet.sci.kobe-u.ac.jp, hackers@FreeBSD.ORG Subject: Re: Please test my code for Intel onboard SMBus function. Message-ID: <19981229002431.23735@breizh.prism.uvsq.fr> In-Reply-To: <4.1.19981228133327.00a14ee0@genesis.ispace.com>; from Drew Baxter on Mon, Dec 28, 1998 at 01:35:04PM -0500 References: <199812270458.NAA04526@libr.scitec.kobe-u.ac.jp> <4.1.19981226140415.00c2a8b0@genesis.ispace.com> <199812270458.NAA04526@libr.scitec.kobe-u.ac.jp> <19981228203456.48462@breizh.prism.uvsq.fr> <4.1.19981228133327.00a14ee0@genesis.ispace.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Dec 28, 1998 at 01:35:04PM -0500, Drew Baxter wrote: >>Last smbus/iicbus commit breaks this code, sorry. But it now handles >>properly the addressing scheme. e.g you may give the slave address to >>each ioctl. >> >>Which errors should be reported Takanori? >> > >I got most of my bugs fixed when I built for CURRENT-1226.. I usually make >world every two weeks or so.. sometimes more.. However as I said, just to >result in crashing my machine :).. > >Eh I'm getting better, we had 3 months on it.. > >Am I supposed to point this thing as SMB0, which it says in my >/var/log/messages is LM78... or am I supposed to point it at SMB1-5 which >it says is a DIMM socket? Hey! Takanori's code is working well! Cool, I cannot test it here yet. I did not put detection code in the last commit. So your hardware won't be detected. This should be done by user-level code (at monitoring daemon initialization and so on). Only /dev/smb0 now addresses your SMBus and spd.c should give the address of the SMB device it accesses for each SMB command instead of using /dev/smb1-5. Detection at boot is wrong since more than one I2C/SMB device may use the same address on the bus. multimedia's solution with bt848 based boards is to compute a signature with the list of device's address present on the bus, then detect the board according to the signature. This works pretty well, but some devices do not love random read/write for the detection. With user-level code, we'll be able to specify exactly the underlying hardware with no effort, and avoid all this complexity. But if you have ideas... > > >--- >Drew "Droobie" Baxter >Network Admin/Professional Computer Nerd(TM) >OneEX: The OneNetwork Exchange, Bangor Maine USA >http://www.droo.orland.me.us > >PGP ID: 409A1F7D > > -- nsouch@teaser.fr / nsouch@freebsd.org FreeBSD - Turning PCs into workstations - http://www.FreeBSD.org To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19981229002431.23735>