mjb/altos, branch 1.0.9.6

altos: Declare cc1111 accel_ref as unsigned so the math works

2012-06-29T00:43:36+00:00

If accel_ref is signed, then the careful shifting and dividing dance necessary to correct for changes in the relationship between the 5V and 3.3V supplies always generates zero. Signed-off-by: Keith Packard

Update version to 1.0.9.6

2012-06-28T23:42:59+00:00

Signed-off-by: Keith Packard

altos: Disable MS5607 interrupt in the handler

2012-06-28T23:05:56+00:00

Avoids having the interrupt re-raised multiple times until the reading task finally wakes up. Signed-off-by: Keith Packard

altos: Track missed HMC5883 interrupts

2012-06-28T06:11:27+00:00

When it fails to signal conversion complete to the CPU, keep track of that and report it with the rest of the current data. Signed-off-by: Keith Packard

altos: Increase default STM stack to 648 bytes

2012-06-28T06:09:16+00:00

512 seems cozy given the printf implementation we're using and the extensive interrupts. Signed-off-by: Keith Packard

altos: panic if MPU6000 self test fails

2012-06-28T06:05:36+00:00

Don't try to fly if the board isn't working right. Signed-off-by: Keith Packard

altos: Disable mag sensor for megametrum

2012-06-28T06:04:25+00:00

It doesn't work unless USB is connected or occasionally while the debug port is enabled. It's mystic. Signed-off-by: Keith Packard

altos: Allow megametrum to be built without using the mag sensor

2012-06-28T06:03:33+00:00

I'm having trouble getting it working reliably, so we'll like disable it for now. This patch makes that possible. Signed-off-by: Keith Packard

altos: Make profiling Viterbi decoder more useful

2012-06-28T02:47:52+00:00

This blocks starting the decoder until all of the data have arrived so that the time spent in the decoder is easily computed. Signed-off-by: Keith Packard

altos: Unroll viterbi state loop for >30% performance boost

2012-06-28T02:45:22+00:00

9.3ms vs 14.7ms, a clear win. Signed-off-by: Keith Packard