Hi, I was curious if there are some GPIOs exposed on the Linksys WRT3200ACM, so I started exporting them in /sys/class/gpio. I realized, that some are locked against exporting (those claimed by gpio-leds and gpio-keys, power regulator), while others exposed more or less problematic side effects. Less critical are GPIOs 2 and 3, which seem to be used by the I2C bus, where a LED driver controls some of the front LEDs. A number of other GPIOs (22, 23, 25, 28, 30, 32, 33, 35, 56, 38, 40, 41, 42) however had an effect on the NAND-flash in the range of just not being able to access it for the moment up to corrupt firmware partition (with the need to unbrick from the boot loader) and even to the point, that it requested bootloader code on the uart. It seems, GPIO 33 does major corruption and GPIO 36 can be really nasty. There are two more GPIOs, which are quite interesting: when GPIO 26 is exported and set to output, setting its value to 1 causes all LAN LEDs to light up constantly. May it be connected to the switch chip? GPIO 52, when exported and set to output, also causes the LED driver on I2C to not respond any more. What is the reason (connected to a reset line?) The basic conclusion is: those multipurpose pins in use need to be locked, just like the gpio-leds and gpio-keys modules do. The question remaining is: which multipurpose pins are already in use by NAND-flash, I2C bus, UART0, SATA(?), PCIe(?), some ethernet MII(?) or something else?
@Imre Kaloz: From what I remember, you had a closer relationship to Linksys and brought in support for this device (family). I hope you can shed some light on this issue. Thanks, Hartmut
0x03684A18FAC89148.asc
Description: application/pgp-keys
signature.asc
Description: OpenPGP digital signature
_______________________________________________ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/mailman/listinfo/openwrt-devel