On 05/06/2009 10:59 AM, Andre Prendel wrote: > On Wed, May 06, 2009 at 09:54:28AM +0200, Hans de Goede wrote: >> >> On 05/06/2009 09:36 AM, Jean Delvare wrote: >>> Hi Andre, >>> >>> >> <snip> >> >>> I'll remember that :) Actually you may want to look at the abandoned >>> tmp401 driver. Hans de Goede worked on it originally but it never made >>> it into mainline. That's a pity because it was almost ready. I seem to >>> remember it needs to be updated due to i2c infrastructure changes, and >>> minor cleanups were needed before I would take the driver into mainline: >>> >>> http://lists.lm-sensors.org/pipermail/lm-sensors/2008-June/023411.html >>> http://lists.lm-sensors.org/pipermail/lm-sensors/2008-August/024024.html >>> >>> Is this something you'd be interested in? If you don't have any >>> supported chip, I can send you chip dumps, which you can feed into >>> i2c-stub to emulate a chip. Or if you are into soldering, you may be >>> able to ask TI for free chip samples. >>> >> Hi Andre, >> >> Nice to meet you! As Jean said I've been working on a TI tmp401 driver, >> which is like 99% finished. If you're interested in working on this, please >> let me know, I've got a much newer version sitting on my harddisk, as well >> as a wip version which also adds support for the tmp411, so if you want to >> work on this I'll dig them up. > > Hi Hans, > > nice to meet you too. Yes I'm interested in the source code. > Ok, attached you find updated source with all points from Jean's review he linked to above addressed, except for the writing of a Documentation/hwmon file, and a crude patch which adds tmp411 support, done by a couple of my students. I think properly integrating tmp411 support (and writing some docs) will make a nice starting project, you should get rid of the global device_id variable hack my students did, and instead do the multiple chip support handling the usual way, so change the I2C_CLIENT_INSMOD_1 macro to a I2C_CLIENT_INSMOD_2 macro, extend the tmp401_id table, store the detected chip kind in the data structure, etc. See for example the fschmd.c driver for an i2c sensor driver which handles multiple chip types this way. >> As Jean said, you can ask TI for samples, or I can mail them to you if you >> prefer, I still have a number of them left, the big question is how good >> are you with a soldering iron? All the samples are smd and have a quite narrow >> pitch. > > I'm affraid I'm a dummy with the soldering iron. Nevertheless it > sounds very interesting. > Erm, then you can probably best test using device dumps, I'll make dumps of the tmp401 and tmp411 I've over here, and send them to you in a private mail, Jean should be able to tell you how to use them (I've never used dumps for testing). >> When you've got a sample, you ofcourse need an i2c adapter to hook it up to, I >> can advice using an i2c-tiny-usb adapter: >> http://www.harbaum.org/till/i2c_tiny_usb/index.shtml > > I will read some more about all this stuff. > Well if you don't have any soldering skills, getting an i2c adapter is of little use, since you will need to do soldering (and have some skill in it to pull it off in this case) to hook up the samples to the adapter. So I think its best if you do your development / testing with ic dumps (so with a software simulation of the IC), and then I can test again with real hardware. Regards, Hans -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: diff Url: http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20090506/a68280a8/attachment.pl -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: tmp401.c Url: http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20090506/a68280a8/attachment.c
