RTOS/BLE-STACK: MSP432 SimpleLink Bluetooth Plugin v1.25.00.42 GPIO Issues (plus requested fixes)

Hi Cambell,

Campbell Young said:

My master ready pin is Port 2.5, and therefore the GPIO_setConfig fails and the BLE plugin cannot control MRDY. I don't know why the GPIO is configured as high strength, but removing it fixes the issue.

Unfortunately I have not been able to reproduce this issue, I switched P6.0 to be SRDY and P2.5 to be MRDY and it is working on my side.

//            sapParams.port.remote.mrdyPinID = Board_MRDY;
//            sapParams.port.remote.srdyPinID = Board_SRDY;
            sapParams.port.remote.mrdyPinID = Board_SRDY;
            sapParams.port.remote.srdyPinID = Board_MRDY;

I did not change the gpio configuration in the board file.

    /* MSP_EXP432P401R_SRDY */
    GPIOMSP432_P2_5 | GPIO_CFG_OUT_STD | GPIO_CFG_OUT_STR_HIGH
            | GPIO_CFG_OUT_HIGH,
    /* MSP_EXP432P401R_BL */
    GPIOMSP432_P6_0 | GPIO_CFG_IN_PU,

Also I created a simple project to change to change the GPIO configuration on the fly.

    GPIO_setConfig(Board_GPIO_TEST,
                   GPIO_CFG_OUT_STD | GPIO_CFG_OUT_HIGH | GPIO_CFG_OUT_STR_HIGH);

and in the board file I have:

    /* MSP_EXP432P401R_GPIO_TEST */
//    GPIOMSP432_P2_5 | GPIO_CFG_OUT_STD | GPIO_CFG_OUT_STR_HIGH | GPIO_CFG_OUT_LOW,
    GPIOMSP432_P2_5 | GPIO_CFG_INPUT,

Maybe the way that you are building the library has something to do with this, is this a project you can share with me??

Campbell Young said:

Additionally, is it possible to fix the following bugs in the next release? 

They were not fixed on v1.30, but I'm pushing them to be fixed on v1.40

Campbell Young said:

Final question: does Texas Instruments have any plans to release version controlled repositories for future SDK releases?

Not yet, but I see your point. Let me discuss internally and I'll get back to you.

  Best regards,

    David

Hi David,

Thanks for the reply. What version of the MSP432 SimpleLink SDK are you using? You'll only notice this issue if it's simplelink_msp432p4_sdk_1_60_00_12 or later because the GPIO driver has been updated to check the pin strength constraint. See lines 484-488 in source/ti/drivers/gpio/GPIOMSP432.c, also copied below:

if (((port != 2) || (pin > 0x04)) &&
                (pinConfig & GPIO_CFG_OUT_STRENGTH_MASK)) {
                HwiP_restore(key);

                return (GPIO_STATUS_ERROR);
            }

So if the pin isn't suitible for GPIO_CFG_OUT_STRENGTH_MASK it will return after configuring the pin for output, but before setting its state to high/low. My pin's default state is low (GPIOMSP432_P2_5 | GPIO_CFG_OUT_OD_NOPULL | GPIO_CFG_OUT_STR_LOW | GPIO_CFG_OUT_LOW), so the call to GPIO_setConfig never sets it to the expected state of GPIO_CFG_OUT_HIGH and as such the CC26xx device has no indication that the master is ready and therefore I cannot communicate with the device. I haven't checked the code, but I guess it waits for a high to low edge before transmitting?

If you still can't reproduce the issue with a default-low pin state, I'll have a closer look on my end. Also, thanks for trying to get the linked fixes into v1.40 :)

Cheers

Campbell