examples not working with CVS head (todays sources) on pc386 BSP.

Jennifer Averett jennifer.averett at oarcorp.com
Thu Jul 14 11:47:10 UTC 2005


>
> Hi,
>
> On Thu, 14 Jul 2005, Krystian Lewandowski wrote:
>
>>> I try to compile hello_world_c application as described on:
>>> http://www.rtems.org/onlinedocs/releases/rtemsdocs-4.6.2/share/rtems/html/started/started00064.html
>>> ...
>>> -B/home/karel/usr/local/pc386-i386-rtems-bsp/i386-rtems/pc386/lib/
>>> -specs
>>> bsp_specs -qrtems   -g -Wall  -O4 -g         -c   -o o-optimize/test.o
>>> test.c
>>> test.c:31:22: error: confdefs.h: No such file or directory
>>> make: *** [o-optimize/test.o] Error 1
>>> silence:/tmp/rtems/examples-4.6.2/hello_world_c$
>>>
>>
>> I didn't tried build CVS toolchain, so i can say too much, but...
>> - do you have the confdefs.h header file in include/, as far as i know
>> #include <confdefs.h> is deprecated (<rtems/confdefs.h> isn't). Maybe it
>> helps.
>
> Interesting! I do have confdefs.h in rtems subdirectory for both my BSPs.
>
>> - are you sure RTEMS toolchain was properly build?
>
> I hope so, since if I change problematic test.c:
>
> Index: test.c
> ===================================================================
> RCS file: /usr1/CVS/examples/hello_world_c/test.c,v
> retrieving revision 1.5
> diff -u -r1.5 test.c
> --- test.c      15 Nov 2002 01:42:52 -0000      1.5
> +++ test.c      14 Jul 2005 11:38:09 -0000
> @@ -28,6 +28,6 @@
>
>   #define CONFIGURE_INIT
>
> -#include <confdefs.h>
> +#include <rtems/confdefs.h>
>
>   /* end of file */
>
>
> then it compiles fine!
>
> General question to RTEMS developers: are all demos/test required to
> change from #include <confdefs.h> to #include <rtems/confdefs.h> ?
>

Yes this modification needs to be made to all.  The demos seem to be
lagging behind in some of the tree modifications.  Please submitt an SPR
on this issue and we will fix it as soon as possible.

Thanks

Jennifer

> Thanks for your fast help!
> Karel
> --
> Karel Gardas                  kgardas at objectsecurity.com
> ObjectSecurity Ltd.           http://www.objectsecurity.com
>




More information about the users mailing list