[ntp:hackers] Where did BLOCKING_REQ_MAGIC and BLOCKING_RESP_MAGIC get their values?

Hal Murray hmurray at megapathdsl.net
Mon Jul 29 16:15:19 UTC 2013


brian.utterback at oracle.com said:
> I understand why there is a warning, but are they arbitrary or not? Is
> there something outside the code that dictates their values? If not, I
> don't understand why a 1 and 2 or even 0 and 1 wouldn't work just as  well.
> If there is some reason why sinple number won't work here, how  about the
> same numbers but without the sign bit set? 

I just took a quick scan at the code.

I think they are only used to sanity check the blocks of memory passed 
to/from the child or thread doing DNS lookups.  I didn't find a comment that 
said DNS, but I didn't look very hard and that's the only place I know of 
where ntpd uses forks/threads.  In particular, those bits don't go out on the 
wire so it should be OK to change them.

As a sanity check, 0 and 1 aren't very good.

If it cleans up a warning in your environment, I'd vote for turning off the 
top bits.   I just tried it and didn't break anything.

Are enums signed or unsigned?  I didn't find any warnings like that in any of 
my log files.



-- 
These are my opinions.  I hate spam.





More information about the hackers mailing list