Seb James wrote:
> I am getting no activity on the i2c clock and data lines, and when
> I try to write to the i2c bus, the processor times out waiting for
> an interrupt associated with (I think) acknowledgement bits coming
> from the slave device to which it is trying to transmit. It times out
> at the following bit of code in cpm_iic_read() (and also does so in
> cpm_iic_write()):
<snip>
> I have tried changing the following line in i2c-algo-8xx.c (line 42 I
> think):
>
> int cpm_scan = 0;
>
> to
>
> int cpm_scan = 1;
>
> Which causes i2c_8xx_add_bus() to scan for devices, using this code:
>
> /* scan bus */
> if (cpm_scan) {
> printk(KERN_INFO " i2c-algo-8xx.o: scanning bus %s...\n",
> adap->name);
> for (i = 0; i < 128; i++) {
> if (cpm_iic_tryaddress(cpm_adap, i)) {
> printk("(%02x)",i<<1);
> }
> }
> printk("\n");
> }
>
> This causes the processor to hang in the same way as I described above
> for cpm_xfer.
<insert the doctor joke here>
I had a similar experience. Decided that it is not a good idea to
scan the bus from the device driver startup function.
I am still curious about the real cause.
--
Jeff Hurst
** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/
