Experimental Physics and
Industrial Control System

"Johnson, Andrew N." <[email protected]> · Fri, 28 Sep 2018 20:20:20 +0000

Hello Kay,

On 09/28/2018 02:48 PM, Kasemir, Kay via Tech-talk wrote:

<snip>

> Back then there was a tech-talk follow up,
> https://epics.anl.gov/tech-talk/2014/msg01090.php:
> 
> ------------- I made a couple of changes in the 'stdio.h' file in
> WindRiver installation. -------------
> 
> I, too, can simply edit the vxworks-6.9/target/h/stdio.h to remove
> the bothersome code around line 318, in fact I did. But that doesn't
> seem to be the best solution.

I never modify the header files from VxWorks (although I do add some of
my own headers for functions that I provide in all my BSPs).

> Is anybody compiling EPICS base with vxworks-6.9?

Yes, and every commit to Base gets compiled by our Jenkins build agents
against both VxWorks 6.8 and 6.9 (6.9.4.1 to be precise).

What is your VxWorks target CPU? I'm only building for the vxWorks-ppc32
and vxWorks-ppc32sf targets, and it might be that the above problems
only occur when compiled for Intel CPUs (although in a later message the
Indian's were apparently building for PowerPC).

Ahh, yes that might might be it... I do *not* recommend building for the
CPU-specific targets (vxWorks-ppc603, vxWorks-ppc604 etc.) any more; all
32-bit PowerPC CPUs should be able to run EPICS built for the more
generic ppc32 or ppc32sf targets, so you should definitely switch to
those targets if you haven't already done so. The ppc32sf is for the
MVME3100 that doesn't have hardware floating-point, all the others
should use the ppc32.

HTH,

- Andrew

-- 
Arguing for surveillance because you have nothing to hide is no
different than making the claim, "I don't care about freedom of
speech because I have nothing to say." -- Edward Snowdon

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System