Experimental Physics and
Industrial Control System

Randall Cayford via Tech-talk <tech-talk at aps.anl.gov> · Thu, 20 Aug 2020 10:52:22 -0700

I found the problem.  The defaults for tar on centos 7 and Ubuntu 18 are different.  On centos, untaring synApps assigns the user who untars.  On Ubuntu (at least with the parameters I used) the assigned user appears to be the uid and gid of whoever created it.  I don’t know why compiling as root has a problem with this but the depends.pl script hangs in that case.  chown fixed the problem.

I have mostly successfully compiled synApps on Ubuntu now.  The dxp module won’t compile over some problem with finding functions in what appears to be libusb.  That one I haven’t solved.

Thanks.

Randall

> On Aug 20, 2020, at 7:39 AM, Engbretson, Mark S. <engbretson at anl.gov> wrote:
> 
> On the module that it is stuck on, do a make distclean, and then start a recompile there.  I have seen the same situation where something goes screwy with a time stamp so thinks that something has to be recompiled when it doesn't.
> 
> -----Original Message-----
> From: Tech-talk <tech-talk-bounces at aps.anl.gov> On Behalf Of Michael Davidsaver via Tech-talk
> Sent: Tuesday, August 18, 2020 8:43 PM
> To: Randall Cayford <rcayford at lbl.gov>
> Cc: tech-talk at aps.anl.gov
> Subject: Re: installing synApps on Ubuntu
> 
> On 8/18/20 2:45 PM, Randall Cayford via Tech-talk wrote:
>> We must be missing something obvious but synApps won’t install on Ubuntu for us.  The problem step is make release hangs forever.  Some problem with perl but we don’t know what.
>> 
>> Steps:
>> install ubuntu 18 LTS
>> Install EPICS7
>> Unzip the synApps (6_1) tar file and edit config files run “make 
>> release”
>> Wait forever….
> 
> What, if anything, is printed while you are waiting?
> 
>> Exact same procedure installs fine on Centos 7.  Since there isn’t a lot of chatter about this, it seems like we’re missing a step somewhere.  Anyone know what we need to do?
> 
> I occasionally have builds get stuck in infinite loops, repeating the same set of steps until I interrupt.
> So far this hasn't happen often or predictably enough to do much troubleshooting.

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System