[rrd-developers] Re: NT Version of 1.1-dev code

Ian Holsman rrd.developers at holsman.net
Mon Mar 24 17:55:40 MET 2003


Jake Brutlag wrote:
>>Hi Guys.
>>This is a update to make the code build on win32.
>>It hasn't been heavily tested, and in fact still has some issues.
> 
> 
> I am sorry to hear something broke the Win32 build. I (for VC++ 6.0) and
> Kerry Calvert (VC++ 7.0) spent a great deal of effort fixing the build
> last June, so I can empathize with your struggle. Your comment seems to
> suggest the added thread-safe code was at fault?

there were 2 faults.
1 was the thread-safety versions of functions (which I think I fixed)
the other was the floating point isnan() checks in RRD create which is where
I am stuck at. (as i deleted my test RRD ;( )

for some reason I get the following
C:\dev\rrdtool-dev>rrdtool create test.rrd  --step 300 DS:a:GAUGE:600:U:U DS:b:GAUGE:600:U:U 
RRA:AVERAGE:0.5:1:300 RRA:MIN:0.5:12:300 RRA:MAX:0.5:12:300
ERROR: min must be less than max in DS definition

I have tracked it down to the following code... (rrd_create.c:442)

         if (! isnan(rrd -> ds_def[ds_idx].par[DS_min_val].u_val) &&
             ! isnan(rrd -> ds_def[ds_idx].par[DS_max_val].u_val) &&
             rrd -> ds_def[ds_idx].par[DS_min_val].u_val
             >= rrd -> ds_def[ds_idx].par[DS_max_val].u_val ) {
             rrd_set_error("min must be less than max in DS definition");
             return;
         }

for some reason, the isnan() function is not thinking the u_val's are NaN's.
(which are a flag to say that the values are not defined I think)

any other developers know the MS floating point stuff ?
I couldn't find the source for 'isnan()' to see what it is trying to check for.

> 
> 
>>1. the graphing doesn't seem to be working
> 
> 
> I don't think we should accept a patch before identifying the problem
> with graphing. In my opinion, graphing is not an optional feature of
> rrdtool, but rather a core feature. Ian, do have any idea where the
> graphing problem lies? Are you willing to do further investigation, or
> is this feature of rrdtool you don't need.

true..
I don't think it is related to my code changes per se, but was a different
issue)

I'll try to dig around, but the creation thing is a biggie.

ps.. can you get the CVS-HEAD to build where you are on win32?

> 
> Jake
> 
> Jake Brutlag
> Network Analyst
> TV Services -- Network Operations
> Microsoft MSN 
> 


--
Unsubscribe mailto:rrd-developers-request at list.ee.ethz.ch?subject=unsubscribe
Help        mailto:rrd-developers-request at list.ee.ethz.ch?subject=help
Archive     http://www.ee.ethz.ch/~slist/rrd-developers
WebAdmin    http://www.ee.ethz.ch/~slist/lsg2.cgi



More information about the rrd-developers mailing list