Re: [Mspgcc-users] MSPDebug version 0.7

[Michiel Konstapel]

> There's one thing you could try, but it's a wild guess really. In
> fet.c,
> find line 858, which looks like this:
> 
>                 .min_version = 20404000,
> 
> Change it to:
> 
>                 .min_version = 20401004,
> 
> You should get some line in the output about sending magic messages.
> Some of the new versions of the firmware require a pair of large init
> messages to be sent before they'll work properly.

Nope, that didn't work, but it seems it already bails out earlier. In
fet_open:

        /* set VCC */
        if (xfer(dev, C_VCC, NULL, 0, 1, vcc_mv) < 0) {
                fprintf(stderr, "fet: set VCC failed\n");
                goto fail;
        }

        printf("Set Vcc: %d mV\n", vcc_mv);

        /* Identify the chip */
        if (do_identify(dev) < 0) {
                fprintf(stderr, "fet: identify failed\n");
                goto fail;
        }

        /* Send the magic required by RF2500 and Chronos FETs */
        if (do_magic(dev) < 0) {
                fprintf(stderr, "fet: init magic failed\n");
                goto fail;
        }

I think it's already failing in do_identify:

                if (xfer(dev, 0x28, NULL, 0, 2, 0, 0) < 0) {
                        fprintf(stderr, "fet: command 0x28 failed\n");
                        return -1;
                }

so it never gets to do_magic, if I'm reading this correctly.
Michiel