On Tue, 2008-04-15 at 19:57 -0400, Andy Walls wrote: > Using 'modprobe cx18 debug=320' on my machine I notice that > > 1. The SCL line always moves to the commanded state immediately. > 2. The SDA line *almost* always moves to the commanded state, and if it > does so, it does it immediately. > 3. If the SDA line does not move to the commanded state, it is always > the case that it didn't go 1/high when expected . > > Given the above > 1. I don't appear to have a PCI bus problem > 2. A slave on both I2C buses of my HVR-1600 occasionally gets stuck > pulling the bus SDA line low (which is electrically OK on an open > collector bus, but logically not what it's expected to do, if > i2c_algo_bit does its job properly).
Correction: most of these "SDA line remaining low when commanded to go high" events appear to be a normal ACKs already being signaled by the slave devices. This actually is expected behavior as long as they occur about 10 clock pulses (20 SCL transitions) apart. -Andy _______________________________________________ ivtv-devel mailing list [email protected] http://ivtvdriver.org/mailman/listinfo/ivtv-devel
