Was using natural log instead of base 10 log. In python log() is natural log!
Tolerance values were converted using: new = 20*log10(e^(old/20))

Note that this wasn't affecting the test as the tolerance values were calculated
with the same (wrong) formula.

 - Add a 'name' file.
 - Change sdbfs layout.
 - Rename ipmi.sdb in IPMI-FRU.
 - Rename calibration.sdb in calib.

Was causing trouble when called from pts.

- Fix the IPMI file.
- Add a 'name' file.
- Add a 'data/' dir.
- Change sdbfs layout.
- Rename ipmi.sdb in IPMI-FRU.
- Rename calibration.sdb in calib.

Was returning two channels per row.

The fmc adc mezzanine related cores are now in a wrapper with a wb crossbar.

The eeprom binary file was located inside the sdbfs input dir.
Causing gensdbfs to put it in the file list.

doc: Update calibration part, add computer date check (correct date needed for eeprom write test23).

Was keeping mfg date even if it was 0xffffff (uninitialised eeprom), or bigger than current date.