some messages from a contemporary HP-UX 11.23 compiler

Rick Jones rick.jones2 at hp.com
Tue Nov 15 01:12:22 UTC 2005


Mark Andrews wrote:
>>Mark Andrews wrote:
>>
>>>>For my own nefarious porpoises I decided to download and compile 9.3.2b2 o
>>
>>n
>>
>>>>an HP-UX 11iv2 (aka 11.23) IPF system with what I believe is a very
>>>>up-to-date compiler. A number of messages were emitted which may or may no
>>
>>t
>>
>>>>be of import, and honestly, I'm not sure if they point fingers at BIND
>>>>code or the compiler .
>>>
>>>
>>>	I suspect the answer is both.
>>
>>that sounds rather likely.
>>
>>fwiw, the compile on HP-UX 11.11 with a rather older compiler was otherwise 
>>"clean" but I do have a question about:
>>
>>
>>>>  +w1 +W 474,530,2193,2236  -c entropy.c
>>
>>Were those four warnings being suppressed because there was a bug in some rev
>> of 
>>the HP-UX compilers?  If so, how long should that suppression remain?  The 
>>paranoid in me thinks that if the compiler has been fixed, perhaps suppressin
>>g 
>>those warnings isn't such a good idea.  At the risk of giving myself an actio
>>n 
>>item I'm not entirely sure how to implement :), maybe it would be good to hav
>>e a 
>>configure test to see if the compiler generates "invalid" warnings for those 
>>four before suppressing them.
> 
> 
> 	Some of them are because the compiler is just plain too noisy.
> 	Suppressing them lets you see the wheat rather than the chaff.
> 
> 	These are the warnings we currently ignore.  Some, but not all,
> 	are due to badness in the system header files.

If I get a chance, I'll try removing those +W's on the IPF compile and see if 
the compiler is still overly chatty and/or the include files are snafu

rick


More information about the bind-workers mailing list