[LaMont Jones] > Except that hwclockfirst exists _ENTIRELY_ to run _before_ modules > are loaded.
Right. I did not know that. As long as the LSB-style dependency information reflect this intention, I am happy, because my dependency based boot system will then get it all correct. :) > In fact, I believe that we're pretty close to the point where I can > drop hwclockfirst entirely, although anyone with /dev/rtc really > should not be building it as a module, if they want time to be > correct, at least with older versions of module-init-tools... Right. I saw some comments about it being needed on platforms without kernels setting the clock from hardware. I guess those kernels are slowly becoming obsolete. :) Friendly, -- Petter Reinholdtsen -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
