[Isocops] Build problem

[Geoffrey B. Crew]

That all sounds fine (indeed, overkill for this rare error).
If I'm being compulsive, I often check that

  make uninstall
  make install

work as well.  (However, during working hours on core code the
former isn't advisable in a shared tree.)

-- 

		Geoff (gbc at space.mit.edu)

On Thu, Jun 03, 2010 at 03:28:56PM -0400, DeMajistre, Bob wrote:
> I've put in a fix for this intermittent failure. Someone (Geoff, Roland,
> Ken, Nathan?) should let their eyes rest lightly on what I did to be sure I
> haven't hosed anything - this is the first time I've gone full loop on
> software maintenance at the ISOC.
> 
> TO MAKE THE CHANGE
> 1) I copied the file (make_me_bin_overlay.pro) out of
> /usr/local/ibex/sw/isoc/src/infra/idl to a working area
> 2) I set up some stuff to test it in my work area
> 3) I made the change, did the tests and copied it back to src/infra/idl
> 4) in /usr/local/ibex/sw/build-3.3/isoc/infra/idl, I did
> make clean
> make all
> 5) I checked all the *.errs file (more *.errs) to be sure everything
> compiled cleanly
> 6) in infra/idl/test I executed
> make check
> Which ended with
> ===================
> All 23 tests passed
> ===================
> 7) I committed make_me_bin_overlay.pro
> 
> THE CHANGE I MADE
> The problem was the creation of a temporary file. The filename was
> hardcoded, so that if it wasn't deleted properly, the file creation would
> fail on the next attempt to run the routine. This would also happen in the
> unlikely event that two processes used the routine at the same time.
> 
> To fix this, I did a few things. First, I added a keyword in the call to the
> routine so the caller could direct the temporary file to an arbitrary
> location. I kept the old name as a default, but added the system time to the
> end of it so that it would almost always be unique. To be REALLY sure,
> before opening the file, I check for the existence of a file with the same
> name. If the file already exists, I append more stuff to it and check again.
> 
> In the end, someone may want to check /tmp to be sure an army of zombies
> isn't being created.
> 
> Bob
> 
> 
> 
> On 6/2/10 5:18 PM, "Geoff Crew" <gbc at space.mit.edu> wrote:
> 
> > You're the IDL expert, so I'll bet Ken will defer to you.  ;-)
> > 
> > I created the mess as I didn't know a quick way of passing the
> > information between IDL and ibex_rotate.  I suspect IDL has some
> > way of qualifying my (static) temporary name with one that is
> > user and process specific--that certainly reduces the cross-section
> > for this problem.
> > 
> > If I had been writing in shell syntax I'd have done:
> > 
> >   /tmp/IDLsuxReallyBadly-$USER-$$
> > 
> > (and of course, removed it 5 lines later) so there's little risk of
> > filling up /tmp with such things, even if someone is agressively
> > typing ^C.
> > 
> > In the normal operation the file should only exist for a fraction of a
> > second (i.e. while ibex_rotate is running) and then be deleted by IDL
> > which is why this is a "bump-in-the-night" sort of bug rarely seen.
>