[ntp:hackers] Resource temporarilly unavailable

Danny Mayer mayer at gis.net
Sat Feb 19 12:45:00 PST 2005


I found that out. input_handler() is broken. I'm working on a fix.
I may take rackety for testing for a short time since it's systems
with refclocks that are having this problem. Unless there's another
system I could borrow for a short time.


At 12:51 PM 2/19/2005, David L. Mills wrote:
>Made the changes you suggested, same response as subject.
>Danny Mayer wrote:
>>I think the logging was there before. However the last argument should be
>>buflen and not rb->buf_length. However the number of bytes doesn't matter
>>if it's not positive. Come to that, the break is in the wrong place.It 
>>should be
>>outside the if statement. Maybe that's what's causing the problem. Move it
>>below the curly brace and rebuild.
>>I wasn't fishing for anything. I was only logging for non-interrupts. We can
>>either dump that or add more conditions for logging or just in debug mode.
>>At 08:10 PM 2/18/2005, David L. Mills wrote:
>>>Guys (Danny?),
>>>While testing ACTS driver mods and turning on the trace, it seems every 
>>>read from the clock causes one to three flashes of the comment at line 
>>>1916 in ntp_io.c. Something about resources temporarily unavailabld. The 
>>>data received are apparently correct. Is this a spurious comment or are 
>>>you fishing for something?
>hackers mailing list
>hackers at support.ntp.org

More information about the hackers mailing list