[prev in list] [next in list] [prev in thread] [next in thread] 

List:       ntp-bugs
Subject:    [ntp:bugs] [Bug 1933] New: WWVB/Spectracom driver timestamps LFs,
From:       bugzilla-daemon () ntp ! org
Date:       2011-05-17 23:00:12
Message-ID: bug-1933-35 () http ! bugs ! ntp ! org/
[Download RAW message or body]

http://bugs.ntp.org/show_bug.cgi?id=1933

           Summary: WWVB/Spectracom driver timestamps LFs, not CRs
           Product: ntp
           Version: 4.2.7
          Platform: All
        OS/Version: All
            Status: ASSIGNED
          Severity: normal
          Priority: P5
         Component: refclock - wwvb
        AssignedTo: hart@ntp.org
        ReportedBy: hart@ntp.org
                CC: bugs@ntp.org
   Estimated Hours: 0.0


The timestamps from refclock_gtlin() used by wwvb_receive() reflect the end of
either a CR or LF character.  The logic currently uses the LF timestamp with
the timecode string terminated by the following CR, when it should be
timestamping the CR for use with the following timecode terminated by the next
CR.

In fixing this, I seem to have also corrected a 30 msec systemic error on pogo
and rackety, which is pleasant but unnerving news.

Further, the on-time point is the start of the CR, while our line-oriented
reads are terminated by the end of the CR or LF, so an additional 1/960 s. is
subtracted from the refclock_gtlin() timestamp, to compensate for 1 character
at 9600,8,N,1.

-- 
Configure bugmail: http://bugs.ntp.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
_______________________________________________
bugs mailing list
bugs@lists.ntp.org
http://lists.ntp.org/listinfo/bugs
[prev in list] [next in list] [prev in thread] [next in thread]