5382
HWTest; USB Test improvements
XL200

Corrects the time limit on the flash drive check in the USB board test.

The limit was three seconds. A working flash drive that takes a little over four seconds to finish starting up was therefore reported as a failure every single time it was tested. Because the failure looked exactly like a hardware fault, it would have sent technicians to investigate boards that were perfectly good — the most expensive mistake a board test can make.

The limit is now ten seconds, more than double what the slowest drive on hand actually needs. A genuinely dead port still fails in ten seconds, which is quick enough for production use.

The test now also reports when a drive passes but took an unusually long time to answer. The healthy fixture drive completes its slowest step in a small fraction of a second, so this message only appears for a drive that is wearing out — which gives us the chance to replace the fixture drive before it starts producing confusing intermittent failures.

Finally, while a check is waiting on a drive, the test now prints a line each second saying what it is waiting for. Nothing on good hardware ever takes a whole second, so this stays silent in normal use; when it does appear, it makes the difference between "this drive is slow" and "this test has locked up" obvious immediately, rather than something that has to be worked out from scratch.