[OpenAFS-devel] question rxi_getAllAddrMaskMtu()
Jeffrey Altman
jaltman@secure-endpoints.com
Sun, 17 Dec 2006 19:35:26 -0500
Sean O'Malley wrote:
> On Sun, 17 Dec 2006, Jeffrey Altman wrote:
>
>> I don't see a definition of rxi_getAllAddrMaskMtu within
>> src/utils/netutils.c.
>>
>> The return value is the number of addresses so it could be a uint.
>>
> You are right. I created one to eliminate compiler warnings.
Why are you creating a second instance of the function in order to
eliminate compiler warnings?
Simply fix the variable that is receiving the return value.
> I was just wondering why there are two different proto's for the same
> function name but use different types of ints and whether they can be
> combined without breaking anything and uints to -me- make the most sense,
> but the rx stuff might have errors coming back as negative numbers and
> need signed ints.
Are you sure that there are two prototypes?
Can you point them out?
I suspect the real problem is that there are no prototypes for the
function.
> Should I really be doing this on the 1.5.x source instead of the 1.4.2
> source so we can fold these in? It really annoys me to see 5k warning
> messages when compiling source. I know some of those I won't eliminate but
> implicit function defs and other warnings can be fairly easily eliminated
> without breaking much. There are other changes related to the different
> libraries solaris has..
Changes should be made against the cvs head.
> This change -could- break a few things though, thus the question.. :)
Please try not submitting changes that break things. You will make the
gatekeepers very grumpy and we will be less likely to accept the patch.
Jeffrey Altman