[Xymon] OpenBSD Compile Issue with logfetch

[Japheth Cleaver]

On 2/14/2017 12:41 PM, Matt Vander Werf wrote:
> Hello all,
>
> Not sure if anyone else has come across this or not, but I'm running 
> into an issue compiling the latest Xymon client (4.3.28) on OpenBSD 6.0.
>
> I'm just compiling the client, since that's all I need for OpenBSD. 
> So, I am first running:
>
> MAKE=gmake ./configure.client
>
> I just selected all the defaults for the prompts for this example.
>
> Then, when running 'gmake', I am getting a compile error when it gets 
> to the logfetch part:
>
> ...
> gcc -g -O2 -Wall -Wno-unused -Wno-pointer-sign -D_REENTRANT 
> -I/usr/local/include -L/usr/local/lib -D_LARGEFILE_SOURCE 
> -D_FILE_OFFSET_BITS=64 -DBSD -I/root/xymon-4.3.28/include 
> -DCLIENTONLY=1 -o logfetch logfetch.c ../lib/libxymonclient.a
> logfetch.c: In function 'logdata':
> logfetch.c:210: error: 'intmax_t' undeclared (first use in this function)
> logfetch.c:210: error: (Each undeclared identifier is reported only once
> logfetch.c:210: error: for each function it appears in.)
> logfetch.c:210: error: expected ')' before 'bufsz'
> gmake[1]: *** [Makefile:41: logfetch] Error 1
> gmake[1]: Leaving directory '/root/xymon-4.3.28/client'
> gmake: *** [build/Makefile.rules:77: client] Error 2
>
> (If anyone wants the full 'gmake' output, let me know!)
>
> If I go and comment out all the lines with logfetch in them in 
> client/Makefile, then everything compiles successfully and I can even 
> do a 'gmake install' and everything works just fine (but only after 
> commenting out the logfetch lines in $XYMONHOME/bin/xymonclient.sh).
>
> Honestly, this works fine for my case since I don't really need any 
> logfetch functionality working for this particular client anyways. But 
> it'd be nice if I didn't have to do all this commenting every time I 
> want to compile a new version of the Xymon client on OpenBSD.
>
> Has anyone run into this issue and/or know what needs to be done to 
> resolve it?
>
> Thanks in advance for any assistance!!
>

Hmm. If you can send the full gmake output, I'd appreciate it. Thanks.

It seems like an '#include <stdint.h>' in logfetch.c (in the normal 
section) might resolve this; it's the only place I believe we're doing 
an intmax_t conversion like this. Do you think you'd be able to test that?

Regards,
-jc
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.xymon.com/pipermail/xymon/attachments/20170214/750c1535/attachment.html>