[arduino_mkrzero] Trying to use builtin SD card reader: TimeoutWaitNotBusy and unexpected reset

[Arne Kappen]

I'm currently trying to get the arduino MKRZERO's builtin SD card reader to work. I've made sure my hardware works by uploading and running an arduino sketch and I've used the feather_m0's adalogger example as a starting point. This is how I configure the SPI

    let timer_clock = clocks
        .configure_gclk_divider_and_source(ClockGenId::GCLK3, 32, ClockSource::OSC32K, true)
        .unwrap();
    let rtc_clock = clocks.rtc(&timer_clock).unwrap();
    let timer = rtc::Rtc::clock_mode(peripherals.RTC, rtc_clock.freq(), &mut peripherals.PM);
    let gclk0 = clocks.gclk0();
    let spi_clock = clocks.sercom4_core(&gclk0).unwrap();
    let sd_spi_pads = spi::Pads::<Sercom4>::default()
                .sclk(pins.pa13)
                .data_in(pins.pa15)
                .data_out(pins.pa12);
    let spi_config = spi::Config::new(
            &peripherals.PM,
            peripherals.SERCOM4,
            sd_spi_pads,
            spi_clock.freq()
        );
    let spi = spi_config
            .baud(400.khz())
            .spi_mode(hal::ehal::spi::MODE_0)
            .enable();

    let sd_cd = pins.pa27.into_pull_up_input();
    let mut sd_cs = pins.pa14.into_push_pull_output();
    sd_cs.set_high().unwrap();

I noticed that the logic of the chip detect pin sd_cd seems to be inverted. So I'm waiting until sd_cd is low and try to initialize:

    if sd_cd.is_high().unwrap() {
        usbserial_write!("No card detected. Waiting...\r\n");
        while sd_cd.is_high().unwrap() {
            delay.delay_ms(250_u32);
        }
    }
    usbserial_write!("Card inserted!\r\n");

    let mut controller = Controller::new(SdMmcSpi::new(spi, sd_cs), timer);
    controller.device().init(); // <-- this results in TimeoutWaitNotBusy after some time

What's weird is that I run into the timeout only if the card is already inserted when I power on the MKRZERO. Plugging in a card while it is running seems to reset the device. I'd be grateful for any advise or pointers to helpful resources.

[Tyler Holmes]

I made the adalogger example and it worked for me back then.. Are you pulling sd_cs high after initializing it? I'm also wondering if lowering the SPI speed even lower helps, or maybe some delay between detecting chip inserted and trying to initialize helps.

Can you tell what kind of reset occurs and why? Could you be panicking?

[Arne Kappen]

Thanks for the ideas. I do pull sd_cs high. I accidentally clipped that line from my snippet above initially. A lower speed hasn't helped, unfortunately.
How can I find out which kind of reset occurs? I'm using panic_halt so I expect the chip to stop in case of a panic and not reset.