Dr. H. Nikolaus Schaller wrote:
> Hey!
>
> /* Addresses to scan: none
> * This chip cannot be reliably autodetected. An empty eeprom
> * located at 0x51 will pass the validation routine due to
> * the way the registers are implemented.
> */
> static unsigned short normal_i2c[] = { I2C_CLIENT_END };
>
> This explains all...
>
> And, in
> http://projects.qi-hardware.com/index.php/p/qi-kernel/source/tree/master/drivers/rtc/rtc-pcf8563.c
> (version
> 0.4.3) all calls to attach() have been removed.
>
> Here is a workaround:
> http://www.avrfreaks.net/index.php?name=PNphpBB2&file=viewtopic&t=65576
>
> static int pcf8563_attach(struct i2c_adapter *adapter)
> {
> //return i2c_probe(adapter, &addr_data, pcf8563_probe);
>
> // OVERRIDE probing
> return pcf8563_probe(adapter, 0x51, 0);
> }
>
> With this patch, I now get a /dev/rtc0 as expected and the hctosys error
> in dmesg has disappeared!
>
> I just have a read error when trying to access the clock. This may
> indeed be an I2C communication error (since i2cdetect also fails).
>
> Nikolaus
>
I'm not using autodetection as far as I can tell. I have the following in my
board file.
static struct i2c_board_info pcf8563_rtc_board_info[] = {
[0] = {
.type = "pcf8563",
.addr = 0x51,
}
};
With this in an arch_initcall:
i2c_register_board_info(0, pcf8563_rtc_board_info, 1);
-Graham
_______________________________________________
Mipsbook-devel mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/mipsbook-devel
